Thermal Desktop 2b75t

and key commands (as are all input fields). 4.3.1.2

Numbering Tab

The Numbering tab of the Thin Shell Data dialog box, shown in Figure 4-9, specifies the submodels and node IDs for the Top/Out and the Bottom/In sides of a surface. The “top” side of a flat surface is in the +Z direction of the local coordinate system, determined by traversing the vertices in a counter-clockwise direction using the right-hand rule. The “out” side refers to the exterior side of curved Thermal Desktop custom surfaces. The “bottom” or “in” side is the opposing side.

Figure 4-9

Thin Shell Data Dialog Box Numbering Tab

Submodel names may be defined ahead of time using the Thermal > SINDA Submodels menu choice so they can be available on the Thin Shell Data dialog box (Section 4.2). Only defined submodels will appear in the drop-down list. The SINDA/FLUINT submodel name “MAIN” is automatically defined. Submodel names entered as text that are not in the Submodel Manager will automatically be added to the Submodel Manager after the confirms the operation by selecting OK on a pop-up message dialog box. The submodel name and the ID may be edited on Numbering tab or on the Node dialog box (see "Nodes" on page 4-62). The nodes on a surface may all be in the same submodel or may be in different submodels. If they are all in the same submodel, but not sequential, then the submodel name will be in the Submodel field, and the node IDs will be in the list.

4-12

Thermal Models

If nodes are in more than one submodel, than the submodel name of the first node will be in the Submodel field, only the IDs will be displayed for nodes in that submodel, and the entire submodel.id will be input for the nodes that are in different submodels. If more than one surface is selected for the Edit operation, node ID input is restricted to the submodel and whether or not same side node numbering should be used. Node IDs for an arbitrary group of surfaces may be modified using the Resequence IDs menu choice (see “Resequence IDs” on page 7-2). The Use same ID’s on both sides check box allows choosing between a true 2D thermal network for the surface (checked) or a 3D thermal network which includes two nodes through the thickness of the surface (unchecked). This second method, with the check box unchecked is used for simplified modeling of a sandwich or honeycomb . The materials and material thicknesses are specified on the Cond/Cap tab (Section 4.3.1.4). This option is only available for finite difference surfaces. Further description of the thermal network is provided in Section 9.1.1 and Section 9.2.1. Note: Unchecking Use same ID’s on both sides is typically used when a low conductivity material separates two higher conductivity materials. This would be typical of a foam-core sandwich or perhaps a honeycomb sandwich. If the gradient through the thickness is due to a low conductivity material on one or both sides of a high conductivity material, then the Insulation tab (Section 4.3.1.6) should be used instead. If the plate is sufficiently thick or the conductivity through the thickness sufficiently low to produce a significant gradient, a better option may be to use a finite difference solid (Section 4.4) or solid finite elements (Section 4.5). Node IDs may be numbered sequentially on a surface by selecting Use Start ID and entering the desired initial ID number. Alternatively, a list of node IDs may be entered by selecting Use List and entering the IDs in the list field. The number of IDs entered must match the number of nodes on the surface given by the subdivision page. Node IDs are ordered by traversing the first direction listed on the subdivision page before incrementing along the second direction. For example, a disk has the angular direction as the first direction, and the radial direction as the second direction. Nodes are input starting with the smallest radius value and traversing along the angular direction. When the first radius “row” is complete, values are entered for the first to last angular locations for the second radius “row”. Node ID assignments may be verified using the Model Checking commands. 4.3.1.3

Radiation Tab

The Radiation tab of the Thin Shell Data dialog box, shown in Figure 4-10, allows the selected surface(s) to be assigned to one or more analysis groups (see “Radiation Analysis Groups” on page 4-1) and also allows the optical properties of the surface to be defined. A RadCAD analysis group is used to compute view factors, radks, or heating rates. Surfaces

Thermal Models

4-13

may be placed in any number of RadCAD analysis groups, with different active sides. Analysis group names must be defined ahead of time with the Radiation Analysis Groups menu choice before they can be used in the Thin Shell Data dialog box.

Figure 4-10

Thin Shell Data Dialog Box Radiation Tab

The may edit the active side of an analysis group by either double clicking on the name in the list or by selecting one or more names in the list and then selecting Edit. Upon editing, an Edit Active Side dialog box will appear and can select the active side for the edited set (see "Analysis Group Active Sides" on page 4-5). This page also assigns optical property references to each side of the surface. Names may be typed directly into the fields, or they may be set from the pulldown list of defined properties. The pulldown list contains the list of currently defined property aliases, followed by the list of properties defined in the current property database file (Section 3.1). The property name “DEFAULT” is always defined, even if no property database file is associated with the current drawing. If only dialog box factors are being computed, the property “DEFAULT” may be used without having to perform the unnecessary step of creating a property database file. The property “DEFAULT” is defined internally to be opaque, diffuse, and black. The property “NORMAL” is also always defined and will cause all rays that are shot from the surface to be emitted normal to the surface. This is useful for using the surface to simulate a heating source. The surface will emit with an emissivity of unity, but will also appear 100% specularly transparent to the rest of the model, so that the surface only behaves as a heat source.

4-14

Thermal Models

The Top Side Overrides and Bottom Side Overrides buttons allow the to input a different optical property for each node. In the Optics Override window, type the node number in smn.# format (e.g. MAIN.1) followed by a comma. On the right side of the window, select the desired optical property from the drop-down list and click on Insert. This capability is similar to the TRASYS MODPR capability. 4.3.1.4

Conductance/Capacitance Tab

Output of capacitance and conductance data to SINDA/FLUINT may be specified by the Thin Shell Data dialog box Cond/Cap tab as shown in Figure 4-11.

Figure 4-11

Thin Shell Data Dialog Box Conductance-Capacitance Tab

Generate Cond/Cap or Not Generated. The Generate Cond/Cap (or Not Generated)

button opens an expression editor to allow the to define a symbolic expression to determine whether the nodes and conductors of the surface are included in the SINDA/ FLUINT solution. If the expression is left black or equals one, the surface will be part of the thermal solution. If expression is set to zero, then nodes and conductors, as well as any defined insulation data, will not be generated; the button reads Not Generated when the expression equals zero. Cond Submodel. This field allows the to specify which submodel the intra-surface conductors are placed in. Note that the Model Browser uses this field in the List By Surfaces/Solids and also Conductance modes.

Thermal Models

4-15

Gen Nodes. The may specify to have the nodes generated based on the material property or simply as arithmetic nodes. If based on material properties, the nodes will be diffusion nodes only if enough information is provided (density, specific heat, and thickness). Material. The drop-down field allows selecting a material from the thermophysical properties database (Section 3.2) or from alias names (Section 3.2.4) created in the current DWG file. When Use same ID’s on both sides check box is unchecked on the Numbering tab (Section 4.3.1.2), the material must be specified for both sides (Top/Out and Bottom/In) and the must also specify the material and distance between the nodes (Separation). See Section 9.2.1 "Double-Sided Surfaces" for information on the double-sided surface thermal network. The material use for the Separation can have an effective emissivity specified (Section 3.2.3.1). Surfaces that reference properties with anisotropic inputs will be modeled orthotropically. That is, the principle directions of conduction will line up with the principle coordinates of the surface. The principle directions are provided in Table 3-1. Finite element surfaces that reference anisotropic materials must have the name of a material orienter specified in the field Material Orientation name. See Section 3.2.6for more information about material orienters.

Multipliers The multiplier fields allow adjusting the thermophysical properties for the surface. The Density multiplier can be used to adjust the mass of a surface. The fields for U or X Cond, V or Y Cond, and W or Z Cond adjust the conductivity in the principle directions for primitives (Table 3-1) or the directions defined by the material orienter (Section 3.2.6). 4.3.1.5

Conductance Tab Note: When thermal connections must be modeled between surfaces or solids, three choices are available: global (discussed here), ors (see “ors” on page 4-74), and merge nodes (see “Merge Coincident Nodes” on page 4-108). For finite conductances, global is generally outdated: please consider using ors instead. For very large conductances or perfect (infinite conductance), use merge nodes, if possible.

The Thin Shell Data dialog box tab is shown in Figure 4-12. A surface may have associated with the top and bottom side and also along the edges. Simply check the box for the associated conductance and input the value in the associated field. Thermal Desktop will calculate which surface/edge is closest to the checked edge or face, and output the connection between the appropriate nodes. For a summary of how these calculations are made, please see “Area Calculations” on page 9-3 and "Edge Calculations" on page 9-6. Each conductance can be input as either absolute or as a function of the area or length of the edge. For the area calculations, the can select to have the test points generated at the exterior of the face (using the thickness of the surface) or at the mid plane of the 4-16

Thermal Models

Figure 4-12

Thin Shell Data Dialog Box Conductance () Tab

surface (where the thickness value is ignored). The can check these calculations by using the Thermal > Model Checks > Show Markers command. (See "Display /or Markers" on page 8-6). Caution to s: Thermal Desktop does not calculate the conductance from the edge of a surface to a center node. Therefore when adding , the value must include that conductance, or alternately the should use edge nodes. Please see see “Circuit Board Conduction Example” on page 20-67 for guidance. 4.3.1.6

Insulation Tab

Insulation may be placed on the top/out side or bottom/in side of a surface with the Insulation tab shown in Figure 4-13. By checking the Put on top/out side and/or the Put on bottom/in side boxes, the places insulation on that side of the surface. The P button allows the to specify insulation or no insulation based on a symbolic expression (Section 2.10.9). Restrictions and assumptions of insulation applied through the Insulation tab are: •

Insulation conductors are based on constant area. Area changes due to changing radius are not ed.

•

The thickness of the insulation is not used in radiation calculations

•

The optical properties of the insulation are applied on the Radiation tab as the optical properties of the surface. If insulation is turned off by programming or overrides (both described below), then the optical properties should be changed, as appropriate.

•

Insulation is assumed to be 1D with conductors only in the thickness direction. The

Thermal Models

4-17

exception to this is the use of Material Stacks (Section 3.2.5). •

The insulation will generate linear and radiation conductors within the insulation based on the material properties: • if the effective emissivity of the material is positive (non-zero), then a radiation conductor will be created with the value of eff*Area • if the conductivity of the material and the Thickness of the insulation are both positive, then a linear conductor will be created with a value of k*Area/Thickness • if both of the above conditions are met, then both a linear and radiation conductor will be created.

Figure 4-13

Thin Shell Data Dialog Box Insulation Tab

A schematic of the SINDA model is shown in Figure 4-14. The actual insulation SINDA nodes are not represented graphically in Thermal Desktop. The optical properties used for radiation calculations of the insulating material must be set on the radiation page (See "Radiation Tab" on page 4-13). A special postprocessing check box has been added so that the may select to display the calculated insulation node temperatures on the model (See "SINDA/FLUINT Dataset" on page 17-13).

4-18

Thermal Models

SINDA.ID from Node Numbering Page

Figure 4-14

G

Insulation Node Not in graphics, but created for SINDA Submodel.ID from Insulation Page

Conductor across the two is determined by the material defined on the insulation page. This can be a radiation conductor for MLI, a linear conductor, or both. Insulation Nodalization

Top/Out (Bottom/In) Side Material/Thickness Single Material. If the insulation can be assumed to be a single material, that material can be selected from the Material drop-down list of materials in the material property database (Section 3.2) or material aliases (Section 3.2.4) defined in the model. Along with the selection of the material, the specifies the total Thickness of the insulation and the Number of Nodes representing the insulation. The thickness can be zero for radiation-only insulation (multi-layer insulation or MLI). The Number of Nodes can be any positive number, but will be set to 1, automatically, if Use new submodel is selected, as described below. Anisotropic materials are not allowed for this option. Multiple Materials (Stack). To define an insulation made up of multiple materials, the specifies the material Stack from the drop-down list. Material Stacks can be created in the Material Stack Manager (Section 3.2.5), accessed through the Stack Manager button at the top of the form or Thermal > Thermophysical Properties > Material Stack Manager.

Top/Out (Bottom/In) Side Node Numbering/Creation Offset Node ID’s by. By choosing this option, the nodes representing the insulation will have ID numbers offset from the underlying surface node ID by the specified amount. For example, if a surface with node MAIN.1 has insulation with Number of Nodes set to 3 and Offset Node ID’s by equal to 1000, the insulation will be represented by nodes MAIN.1001, MAIN.2001, and MAIN.3001, from the surface, outward. Important: If the Number of Nodes on top and bottom of a surface are both greater than one, the should choose the Node ID Offset carefully to prevent duplication of IDs. Use new submodel. By choosing this option, the nodes representing the insulation will use the same ID number as the underlying surface, but will be placed in the specified submodel. This option limits the Number of Nodes to one. This option cannot be used with Recession (Section 3.2.7).

Thermal Models

4-19

Calc Type. The Calc Type drop-down gives the options: Based on material property and Arithmetic. By choosing arithmetic, the insulation nodes will be assumed massless (this option is not recommended for Recession, Section 3.2.7). By choosing Based on material property, the insulation nodes will be generated as diffusion nodes, if possible. Init temp. This field allows specifying the initial temperature of the insulation nodes for transient solutions. Overrides. The override button at the bottom of each insulation set allows the to specify which nodes have insulation. If no overrides are input, then the insulation is placed on the entire surface. The override list consists of a single column of node numbers in submodel.ID format. Note: If overriding surface insulation, be sure to include overrides on the Radiation tab, also, if the surface optical properties will be different. 4.3.1.7

Surface Tab

If a custom conic surface is being edited, surface parameters may be edited directly using the Thin Shell Data dialog box Surface tab (Figure 4-15). Labels for the input fields are adjusted according to the type of surface being edited and checking for validity of the parameters will be done when OK is selected or a different tab selected. Surface parameters may also be modified using interactive grips.

Figure 4-15

4-20

Thin Shell Data Dialog Box Surface Tab

Thermal Models

The Comment field is a multi-line field that is extremely useful for model documentation. See "Comment Field" on page 2-41 for more information. The first line of the comment field provides object identification in three locations: the Model Browser (see "Model Browser" on page 2-8), the tool tip that appears when the cursor is positioned over the object, and as a comment line in the SINDA/FLUINT conductance/capacitance file. 4.3.1.8

Trans/Rot Tab

The may specify additional translations and rotations for the surface using the TRANS/ROT tab. The tab is shown in Figure 4-16. These rotations and translations are

Figure 4-16

Thin Shell Data Dialog Box Translation/Rotation Tab

relative to the initial origin and local coordinate system of the surface. The has the option to specify the order of the rotations by using the drop-down menus for each axis. 4.3.2

Box

A box is a group of five or six surfaces created together using the Thermal > Surfaces/ Solids > Box menu choice. Initially, the Create Box form (Figure 4-17) opens and the selects the nodalization and the number of sides for the box (Note that a 5-sided box will not have a ZMIN face). The actual object(s) created will depend on the nodalization selected. 2x2 Edge Nodes will generate five or six individual rectangles with Subdivisions set to two, edge nodes selected and the nodes merged. Either 1 Node per Face or 1 Node for the box will generate a single mesh: the subdivisions cannot be changed and the sides cannot be individually edited. Selecting the mesh will provide grip points at each vertex that can be individually moved. The 1 Node per Face option will generate conductors between faces

Thermal Models

4-21

Figure 4-17

Create Box Form

when a SINDA/FLUINT file is created; the 1 Node for the box option only has one node and therefore has only one temperature and no conductors; the 2x2 Edge Nodes option has merged nodes at the vertices so no conductors are necessary. After selecting the nodalization, the is queried for the origin, the point for the +X axis and the X-size, the point to set the XY plane and the Y-size, and the point to set the Zsize or a length for the Z-size. When the dimensions have been specified, the edit form appropriate for the nodalization will be opened. 4.3.3

Cone

A cone is created using the Thermal > Surfaces/Solids > Cone menu choice. The following prompts will appear at the command line: >: Pick or enter point for base of cone <0,0,0>: Click a point on the screen for the

base of cone This point defines the base and partially defines the centerline. >: Pick or enter point for top of cone <@0,0,1>: Click a point on the screen for the

top of cone The vector from the base point to the top of the cone defines the centerline and the +Z axis. The local +X axis is computed so as to lie in the current Coordinate System (UCS) XY plane, and the +Y axis follows from the right-hand rule. The UCS is repositioned at the base point with the axes aligned with surface’s local coordinate system for the remainder of the inputs. Points entered at the command line prompt may be proceeded with an asterisk (*) to use the World Coordinate System, rather than the local coordinate system. The UCS is reset to the previous UCS after the surface has been created. >: Enter base radius or pick/enter point <0.0>: Enter a value or point, or pick a

point on the screen 4-22

Thermal Models

The base radius may be entered directly or computed from a point entry. A return without entering a value creates a cone with a base radius of zero. The base radius may be less than, equal to, or greater than the top radius. >: Enter top radius or pick/enter point <1.0>: Enter a value or point, or pick a point

on the screen The top radius may be entered directly or computed from a point entry. A return without entering a value creates a cone with a top radius of one. The top radius may be less than, equal to, or greater than the base radius (only one of the radii may be zero, however). >: Enter start angle or pick/enter point <0.0>: Enter a value or point, or pick a

point on the screen The starting angle for the cone is referenced counterclockwise from the local +X axis. A return without entering a value starts the cone at the local +X axis. If a point is input, the line connecting the origin to the point is projected into the local XY plane to determine the angle from the local X axis. >: Enter end angle or pick/enter point <360.0>: Enter a value or point, or pick a

point on the screen The end angle for the cone is referenced counterclockwise from the local +X axis. A return without entering a value ends the cone at the local +X axis. If a point is input, the line connecting the origin to the point is projected into the local XY plane to determine the angle from the local X axis. The Thin Shell Data dialog box (Figure 4-7) will appear after this prompt to allow specification of thin shell data. The grip points for the cone are shown in Figure 4-18. The move origin grip point moves the origin of the cone to the newly selected point. The cone is translated with the origin. The stretch top grip point stretches the cone along its centerline. The new height of the cone is set by projecting the line from the grip point to the picked (or entered) point along the centerline of the cone. The grip point three quarters of the way from the base to the top of the cone is the aim Z axis grip point. Selecting this grip and then entering or picking a point will orient the cone so that the +Z axis is aimed at the point by rotating about the base point in the plane formed by the current +Z axis and the vector from the base point to the pick point. The height is left unchanged. The set start angle and set end angle grip points are located at one quarter and three quarters of the height at the start and end edges of the cone. An angle is set by projecting the line from the origin to the picked point onto the local XY plane. The set base radius grip is located on the base of the cone, half way between the start and end angles. Selecting the base radius grip and picking a point will modify the cone’s base radius by projecting the line from the base point to the picked point onto the local XY plane. The distance of the projected line is used for the radius. Thermal Models

4-23

Set Top Radius Stretch Top Aim Z Axis Set End Angle Set Base Radius Set Start Angle Move Origin

Figure 4-18

Thermal Desktop Cone Grip Points

The set top radius grip is located on the top of the cone, half way between the start and end angles. Selecting the top radius grip and picking a point will modify the cone’s top radius by projecting the line from the base point to the picked point onto the local XY plane. The distance of the projected line is used for the radius. The top radius may be less than, equal to, or greater than the base radius, except that both cannot be zero. The orientation of the nodal breakdown is always counterclockwise from the start edge, then proceeding from the base to the top. If the top or base radius is zero, the radius grip and the stretch grips will coincide. In this case, the grip will always modify the radius. To modify the height of a cone that has a zero radius top or base, drag the grip slightly to set a non-zero radius, modify the height, and then set the radius back to zero. The cursor will automatically snap to displayed grips, making the input operation easy. 4.3.4

Cylinder

A cylinder is created using the Thermal > Surfaces/Solids > Cylinder menu choice. The following prompts will appear at the command line: >: Pick or enter point for base of cylinder <0,0,0>: Enter or pick a point on the

screen This point defines the base and partially defines the centerline. >: Pick or enter point for top of cylinder <0,0,0>: Enter or pick a point on the

screen

4-24

Thermal Models

The vector from the origin point to the top of the cylinder defines the centerline and the +Z axis. The local +X axis is computed so as to lie in the current UCS XY plane, and the +Y axis follows from the right-hand rule. The local +X axis is computed so as to lie in the current Coordinate System (UCS) XY plane, and the +Y axis follows from the right-hand rule. The UCS is repositioned at the base point with the axes aligned with surface’s local coordinate system for the remainder of the inputs. Points entered at the command line prompt may be proceeded with an asterisk to use the World Coordinate System, rather than the local coordinate system. The UCS is reset to the previous UCS after the surface has been created. >: Enter radius or pick/enter point <1.0>: Enter a value or point, or pick a point on

the screen The radius may be entered directly or computed from a point entry. A return without entering a value creates a cylinder with a radius of one. >: Enter start angle or pick/enter point <0.0>: Enter a value or point, or pick a

point on the screen The starting angle for the cylinder is referenced counterclockwise from the local +X axis. A return without entering a value starts the cylinder at the local +X axis. If a point is input, the line connecting the origin to the point is projected into the local XY plane to determine the angle from the local X axis. >: Enter end angle or pick/enter point <360.0>: Enter a value or point, or pick a

point on the screen The end angle for the cylinder is referenced counterclockwise from the local +X axis. A return without entering a value ends the cylinder at the local +X axis. If a point is input, the line connecting the origin to the point is projected into the local XY plane to determine the angle from the local X axis. The Thin Shell Data dialog box (Figure 4-7) will appear after this prompt. The grip points for the cylinder are shown in Figure 4-19. The move origin grip point moves the origin of the cylinder to the newly selected point. The cylinder is translated with the origin. The stretch top grip point stretches the cylinder along its centerline. The new height of the cylinder is set by projecting the line from the grip point to the picked (or entered) point along the centerline of the cylinder. The grip point three quarters of the way from the origin to the top of the cylinder along the centerline is the aim Z axis grip point. Selecting this grip and then entering or picking a point will orient the cylinder so that the +Z axis is aimed at the point by rotating about the base point in the plane formed by the current +Z axis and the vector from the base point to the pick point. The height is left unchanged.

Thermal Models

4-25

Stretch Top Aim Z Axis Set End Angle

Set Radius

Set Start Angle Move Origin

Figure 4-19

Thermal Desktop Cylinder Grip Points

The start angle grip is located along the starting edge, and up one quarter of the height along the Z axis. The end angle grip is located at the ending edge, up three quarters of the height along the Z axis. An angle is set by projecting the line from the origin to the picked point onto the local XY plane. The radius grip is located on the base of the cylinder, half way between the start and end angles of the cylinder. Selecting the radius grip and picking a point will modify the cylinder’s radius by projecting the line from the origin to the picked point onto the local XY plane. The distance of the projected line is used for the radius. Want "Hands-On" Information? Gain experience working with cylinders by completing Section 21.6 "Orbital Maneuvers" on page 21-87. 4.3.5

Disk

A disk is created using the Thermal > Surfaces/Solids > Disk menu choice. The following prompts will appear at the command line: >: Pick or enter point for center of disk <0,0,0>t: Enter or pick a point on the

screen This point defines the center of the disk. >: Pick or enter point for +Z axis of disk <@0,0,1>: Enter or pick a point on the

screen The vector from the origin point to this point defines the +Z axis. The local +X axis is computed so as to lie in the current UCS XY plane, and the +Y axis follows 4-26

Thermal Models

from the right-hand rule. The UCS is repositioned at the base point with the axes aligned with surface’s local coordinate system for the remainder of the inputs. Points entered at the command line prompt may be proceeded with an asterisk to use the World Coordinate System, rather than the local coordinate system. The UCS is reset to the previous UCS after the surface has been created. >: Enter maximum radius or pick/enter point <1.0>: Enter a value or point, or pick

a point on the screen The radius may be entered directly or computed from a point entry. A return without entering a value creates a disk with an outer radius of one. >: Enter minimum radius or pick/enter point <0.0>: Enter a value or point, or pick

a point on the screen The radius may be entered directly or computed from a point entry. A return without entering a value creates a disk with no hole in the center. >: Enter start angle or pick/enter point <0.0>: Enter a value or point, or pick a

point on the screen The starting angle for the disk is referenced counterclockwise from the local +X axis. A return without entering a value starts the disk at the local +X axis. If a point is input, the line connecting the origin to the point is projected into the local XY plane to determine the angle from the local X axis. >: Enter end angle or pick/enter point <360.0>: Enter a value or point, or pick a

point on the screen The end angle for the disk is referenced counterclockwise from the local +X axis. A return without entering a value ends the disk at the local +X axis. If a point is input, the line connecting the origin to the point is projected into the local XY plane to determine the angle from the local X axis. The Thin Shell Data dialog box (Figure 4-7) will appear after this prompt. The grip points for the disc are shown in Figure 4-20. The move origin grip point moves the disc origin to the newly specified point. The aim Z axis grip point is located along the local +Z axis of the disk, at a distance of one quarter of the outer radius. Selecting this grip and then entering or picking a point will orient the disk so that the +Z axis is aimed at the point by rotating about the local origin in the plane formed by the current +Z axis and the vector from the origin point to the pick point. The set start angle and set end angle grip points are located at one quarter and three quarters of the way along the radial edges so that they do not coincide with themselves or other grips. An angle is set by projecting the line from the origin to the picked point onto the local XY plane. The set min radius and set max radius grips are located on the inner and outer circumferences of the disk, half way between the start and end angles. Selecting a radius grip and picking a point will modify the radius by projecting the line from the origin to the picked point onto the local XY plane. The distance of the projected line is used for the radius.

Thermal Models

4-27

Set Max Radius

Set Min Radius

Aim Z Axis

Move Origin Set Start Angle

Set End Angle

Figure 4-20

Thermal Desktop Disk Grip Points

If the minimum radius is set to zero, it will coincide with the move origin grip. In this case, the move origin grip becomes inactive and the grip at the origin is the set min radius grip. Want "Hands-On" Information? Gain experience working with disks by completing Section 20.6 "Beer Can Example" on page 20-89, and Section 21.6 "Orbital Maneuvers" on page 21-87. 4.3.6

Ellipse

An ellipse is created using the Thermal > Surfaces/Solids > Ellipse menu choice. The will be prompted to select the following points. >: Pick or enter point for center of ellipse <0,0,0>: >: Pick or enter point for +Z axis of ellipse <@0,0,1>: >: Enter semi major axis length or pick/enter point <2.0>: >: Enter semi major axis length or pick/enter point <1.0>: >: Enter semi major inner axis length or pick/enter point <0.0>: >: Enter start angle or pick/enter point <0.0>: >: Enter end angle or pick/enter point <360.0>:

The ellipse will have 7 grip points for editing. These are shown in Figure 4-21. As with the other conics, the start angle will always be 25% of the length on the start line, while the end angle will be 75% of the length on the end line.

4-28

Thermal Models

Set Semi Major Axis

Set Start Angle

Aim Z Axis Set Semi Minor Axis

Set Semi Major Inner Axis

Move Origin

Set End Angle

Figure 4-21

4.3.7

Thermal Desktop Ellipse Grip Points

Ellipsoid

An ellipsoid is created using the Thermal > Surfaces/Solids > Ellipsoid menu choice. The will be prompted to select the following points. >: Pick or enter point for center of Ellipsoid <0,0,0>: >: Pick or enter point to define +Z axis <@0,0,1>: >: Enter X-radius or pick/enter point <2.0>: >: Enter Y-radius or pick/enter point <1.0>: >: Enter Z-radius or pick/enter point <1.0>: >: Enter start angle or pick/enter point <0.0>: >: Enter end angle or pick/enter point <360.0>:

Thermal Models

4-29

The ellipse will have 7 grip points for editing. These are shown in Figure 4-22. As with the other conics, the start angle will always be 25% of the length on the start line, while the end angle will be 75% of the length on the end line. Aim Z Axis

Set Hmax

Set Z Radius

Set End Angle Set X Radius

Set Y Radius Move Origin

Set Start Angle

Set Hmin

Figure 4-22

4.3.8

Thermal Desktop Ellipsoid Grip Points

Elliptic Cone

An ellipse is created using the Thermal > Surfaces/Solids > Elliptic Cone menu choice. The will be prompted to select the following points. >: Pick or enter point for base of cone <0,0,0>: >: Pick or enter point for top of cone <@0,0,1>: >: Enter base semi-major or pick/enter point <0.0>: >: Enter top semi-major or pick/enter point <1.0>: >: Enter base semi-minor or pick/enter point <1.0>: >: Enter start angle or pick/enter point <0.0>: >: Enter end angle or pick/enter point <360.0>:

4-30

Thermal Models

The elliptic cone will have 9 grip points for editing. These are shown in Figure 4-23. As with the other conics, the start angle will always be 25% of the length on the start line, while the end angle will be 75% of the length on the end line.

Figure 4-23

4.3.9

Thermal Desktop Elliptic Cone Grip Points

Elliptic Cylinder

An ellipse is created using the Thermal > Surfaces/Solids > Elliptic Cylinder menu choice. The will be prompted to select the following points. >: Pick or enter point for base of cylinder <0,0,0>: >: Pick or enter point for top of cylinder <@0,0,1>: >: Enter semi major axis length or pick/enter point <2.0>: >: Enter semi minor axis length or pick/enter point <1.0>: >: Enter start angle or pick/enter point <0.0>: >: Enter end angle or pick/enter point <360.0>:

Thermal Models

4-31

The elliptic cylinder will have 7 grip points for editing. These are shown in Figure 424. As with the other conics, the start angle will always be 25% of the length on the start line, while the end angle will be 75% of the length on the end line.

Figure 4-24

4.3.10

Thermal Desktop Elliptic Cylinder Grip Points

Finite Element (FE) Shells

FE shells are elements whose shapes are defined by the vertex nodes and internal conduction networks are calculated using FE conduction calculations. Thermal Desktop recognizes and uses the following FE shell types: • Linear, 3-node, triangular elements • Linear, 4-node quad elements • Curved triangular elements • Curved quad elements Figure 4-25 shows an example of a linear, triangular finite element shell. The curved finite elements do not generate more nodes than the linear elements and do not gerenate more internal conductors. Finite elements can be extremely useful when the shape of the geometry does not fit one of the primitive surface shapes (e.g. - a plate with a hole) and provide an accurate conductance network in most situations. The should follow these guidelines when choosing to use finite elements: 1. Linear finite elements require more elements, and therefore more nodes, to define a curved shape. 2. When trying to minimize the number of finite elements, the element shapes should

4-32

Thermal Models

1

2

Figure 4-25

3

Linear Triangular Element

be as close to ideal as possible: square for quadrilateral and equilateral triangle for triangular. Element can be checked using Check Elements (Section 8.12). 3. For very small models, use as many finite elements as you would finite difference nodes. You will have more nodes with the finite element approach. If minimizing the number of nodes is important, than use finite difference objects. FE shells are created using one of the following methods: • Creating Linear FE Shells Manually - Section 4.3.10.1 • Surface Coat Free Solid FEM Faces - Section 4.16.2 • Convert AutoCAD Surface to Nodes/Elements - Section 4.16.4 • Convert Finite Difference to Finite Elements - Section 7.11 • Modeling with TD Mesher - Section 14 • Finite Element Model - Section 18.2.3 • Link to TD Direct - Section 18.6 Only linear elements can be created within Thermal Desktop. Second-order elements must be created using one of the last two methods above. 4.3.10.1

Creating Linear FE Shells Manually Note: The may wish to investigate TD Mesher (Section 14) or TD Direct (Section 18.6) as alternative methods for creating finite element shells.

Prerequisite: Nodes (Section 4.6) for the element vertices must already be created before creating a FE shell • Icon: • Command: RcLinearElement Thermal Models

4-33

• Menu: Thermal > FD/FEM Network > Element • Ribbon: Thermal > Create Network > Element • Toolbar: Network Objects When the command is issued, the receives the following prompt: >: Select nodes for linear element:

At each prompt, the selects a node representing a vertex of the element. Either three or four nodes can be selected. When all nodes have been selected, the hits <Enter> to complete the element. The order of selection of the elements will determine which side is the “top” side for radiation calculations. If the elements in Figure 4-25 are selected in a counter clockwise direction (1,2,3), the top side follows the right hand rule and will be the displayed side shown in the current view. If the nodes were selected in the reverse order (1,3,2), the top side would be facing down from the current view. The should always graphically the active sides using the display active sides functionality (Section 8.1). Finite element shapes are defined by their nodes. In order to move an element, its nodes must be selected and moved. 4.3.10.2

Thin Shell Data for FE Shells

• Radiation - Section 4.3.1.3 • Cond/Cap - Section 4.3.1.4 • Insulation - Section 4.3.1.6 • Surface - The Surface tab for Finite Element shells only has a comment field 4.3.11

From AutoCAD Surface

A finite difference surface may be created from the following types of AutoCAD surfaces: 3D faces, polymeshes, and polyface meshes. The resulting Thermal Desktop surface will be represented by one node in-plane (not ing for different node IDs on each side or insulation, which are both allowed). These AutoCAD surfaces may be created directly from input or by using AutoCAD techniques such as revolving a 2D polyline, constructing a ruled surface, or generating an edge defined patch. Extruded 2D polylines are not currently ed, however, the same results can be obtained by copying the 2D polyline and then constructing a ruled surface between the two curves. Please review AutoCAD documentation for techniques used to construct 3D surface geometry. A Thermal Desktop surface created from an AutoCAD surface initially contains one nodal region with the same or separate node IDs on each side. Mesh surfaces may be converted to one nodal region per mesh facet by using the Modeling Tools > Toggle FD Mesh Nodalization command described in Section 7.8 (performing the command a second time will change the mesh back to one node per side). An example is displayed in Figure 4-34

Thermal Models

4-26.

Default Figure 4-26

After Toggle Mesh Command

Thermal Desktop Surface from AutoCAD rulesurf Command

Note: Once a surface is created by this method, the only option for changing the number of nodes is by using the Toggle FD Mesh Nodalization. The only other option is to recreate the original AutoCAD surface with different subdivisions. If changing the number of nodes may be important, the may wish to investigate TDMesh or TD Direct for complex surfaces. Facets in AutoCAD generated meshes are defined to have a +Z normal given by the ordering of the vertices of the faces. The +Z normal is given by the right hand rule when traversing the vertices in a counter-clockwise manner. For flat surfaces, the +Z direction of the local coordinate system is called the “Top” side. The other side is called the “Bottom” side. Note that some AutoCAD surface generation techniques produce a mesh that have an outside and an inside. For example, a curve can be revolved into a closed surface. Depending on the orientation of the curve and the direction it was revolved, the inside of the surface may be either the active Top/Out or active Bottom/In (because of the ordering of the vertices for the faces). Thermal Desktop custom entities will always have the visual outside of a surface as the Top/Out side. It is a good idea to always check active sides with the Model Checking commands (see “Model Checks” on page 8-1). The connectivity of a mesh can also be reversed with Thermal > Modeling Tools > Reverse Connectivity of Planar Elements/Meshes (Section 7.9). More than one AutoCAD mesh may be converted at the same time. If multiple AutoCAD surfaces are selected, each surface will be automatically assigned default thin shell data parameters. The surfaces can then be modified using subsequent Thermal > Edit commands. Node IDs for an arbitrary group of Thermal Desktop surfaces may be resequenced using the Thermal > Modeling Tools > Resequence IDs menu choice (see “Resequence IDs” on page 7-2). The lateral conductance is calculated between nodes using finite difference approximations. As meshes become more skewed (less orthogonal), the approximations become less accurate. Please note that you can also convert AutoCAD surfaces to nodes and finite elements for more accurate lateral conductance calculations. For information on this, please see “Convert AutoCAD Surface to Nodes/Elements” on page 4-110. Thermal Models

4-35

Want "Hands-On" Information? Learn how to create a mesh and convert it to either Thermal Desktop polygons or finite elements in Section 20.4 "Simple Meshing Methods" on page 20-57. 4.3.12

Offset Paraboloid

An offset paraboloid is the surface that is created when a solid elliptical (or circular) cylinder is intersected with a paraboloid. An offset paraboloid is created using the Thermal > Surfaces/Solids > Offset Paraboloid menu choice. The will be prompted to select the following points in order to create the offset paraboloid. >: Pick or enter point for center of paraboloid <0,0,0>: >: Pick or enter point to define +Z axis and height of paraboloid <@0,0,1>: >: Enter top radius or pick/enter point <.5*Height>: >: Enter Ellipse Origin along X or pick/enter point <.5*Radius>: >: Enter semi major axis length or pick/enter point : >: Enter semi minor axis length or pick/enter point <SemiMajor>:

4.3.13

Ogive

An ogive is created using the Thermal > Surfaces/Solids > Ogive menu choice. The will be prompted to select the following points in order to create the ogive. >: Pick or enter point for center of Ogive <0,0,0>: >: Pick or enter point to define +Z axis <@0,0,1>: >: Enter small radius or pick/enter point <1.0>: >: Select point to set large radius <2.*Small Radius>: >: Enter start angle or pick/enter point <0.0>: >: Enter end angle or pick/enter point <360.0>:

4.3.14

Parabolic Trough

A parabolic trough is created using the Thermal > Surfaces/Solids > Parabolic trough menu choice. The following prompts will appear at the command line: >: Pick or enter point for origin of paraboloid <0,0,0>: Enter or pick a point on the

screen This point defines the vertex of the parabola and partially defines the baseline of the trough. >: Pick or enter point for focal point and Z axis <@0,0,.1>: Enter or pick a point on

the screen The vector from the vertex point to this point defines the +Z axis. The local +X axis is computed so as to lie in the current UCS XZ plane, and the +Y axis follows

4-36

Thermal Models

from the right-hand rule. The UCS is repositioned at the base point with the axes aligned with surface’s local coordinate system for the remainder of the inputs. Points entered at the command line prompt may be proceeded with an asterisk to use the World Coordinate System, rather than the local coordinate system. The UCS is reset to the previous UCS after the surface has been created. >: Enter XMIN or pick/enter point <-1.0>: Enter a value or point, or pick a point on

the screen The X-direction of the trough defines the width of the parabola. The XMIN value defines the minimum X value of the trough width. >: Enter XMAX or pick/enter point <1.0>: Enter a value or point, or pick a point on

the screen The XMAX value defines the maximum X value of the trough width. If the XMIN and XMAX of the trough are opposite signs, the trough will include the parabola’s vertex. >: Enter YMIN or pick/enter point <0.0>: Enter a value or point, or pick a point on

the screen The Y-direction of the trough defines the trough’s length. The YMIN value defines the minimum Y value of the trough length. >: Enter YMAX or pick/enter point <1.0>: Enter a value or point, or pick a point on

the screen The YMAX value defines the maximum Y value of the trough length. The grip points for the parabolic trough are shown in Figure 4-28. The move origin grip moves the origin of the trough without changing the orientation.

Figure 4-27

Thermal Desktop Paraboloid Grip Points

Thermal Models

4-37

The aim Z axis grip point is located along the local +Z axis of the parabolic trough. Selecting this grip and then entering or picking a point will orient the paraboloid so that the +Z axis is aimed at the point by rotating about the base point in the plane formed by the current +Z axis and the vector from the base point to the pick point. The XMIN grip is located on the base of the trough at the minimum X location in the XY plane. Selecting the XMIN grip and picking a point will modify the width (and depth) of the trough in -X direction. The distance of the projected line is used for the width. The depth is based on the parabolic shape for the given focal length and the width. The XMAX grip is located on the base of the trough at the maximum X location in the XY plane. Selecting the XMAX grip and picking a point will modify the width (and depth) of the trough in +X direction. The distance of the projected line is used for the width. The depth is based on the parabolic shape for the given focal length and the width. The YMIN grip is located on the base of the trough, between the Move Origin and XMAX grips. Selecting the YMIN grip and picking a point will modify the location of the base of the extruded length of the parabola. The distance the base is moved will be the difference between the original point and second point in the Y direction. The YMAX grip is located at the top (or length) of the trough. Selecting the YMAX grip and picking a point will modify the extruded length of the parabola. The distance the top is moved will be the difference between the original point and second point in the Y direction. 4.3.15

Paraboloid

A paraboloid is created using the Thermal > Surfaces/Solids > Paraboloid menu choice. The following prompts will appear at the command line: >: Pick or enter point for base of paraboloid <0,0,0>: Enter or pick a point on the

screen This point defines the base and partially defines the centerline. >: Pick or enter point for top of paraboloid <@0,0,1>: Enter or pick a point on the

screen The vector from the base point to this point defines the +Z axis. The local +X axis is computed so as to lie in the current UCS XY plane, and the +Y axis follows from the right-hand rule. The UCS is repositioned at the base point with the axes aligned with surface’s local coordinate system for the remainder of the inputs. Points entered at the command line prompt may be proceeded with an asterisk to use the World Coordinate System, rather than the local coordinate system. The UCS is reset to the previous UCS after the surface has been created. >: Enter base radius or pick/enter point <0.0>: Enter a value or point, or pick a

point on the screen The base radius may be entered directly or computed from a point entry. A return without entering a value sets the radius at the base of the paraboloid to zero.

4-38

Thermal Models

>: Enter top radius or pick/enter point : Enter a value or point, or

pick a point on the screen The top radius may be entered directly or computed from a point entry. A return without entering a value creates a paraboloid with a top radius equal to the base radius plus one half of the height. The top radius must be greater than the base radius. >: Enter start angle or pick/enter point <0.0>: Enter a value or point, or pick a

point on the screen The starting angle for the paraboloid is referenced counterclockwise from the local +X axis. A return without entering a value starts the sphere at the local +X axis. If a point is input, the line connecting the origin to the point is projected into the local XY plane to determine the angle from the local X axis. >: Enter end angle or pick/enter point <360.0>: Enter a value or point, or pick a

point on the screen The end angle for the paraboloid is referenced counterclockwise from the local +X axis. A return without entering a value ends the paraboloid at the local +X axis. If a point is input, the line connecting the origin to the point is projected into the local XY plane to determine the angle from the local X axis. The Thin Shell Data dialog box (Figure 4-7) will appear after this prompt. The grip points for the paraboloid are shown in Figure 4-28. The stretch base grip moves the base along the local Z axis. The stretch top grip locates the height of the paraboloid along the Z axis. The base and height are computed by projecting the line from the base point to the selected point onto the Z axis.

Figure 4-28

Thermal Desktop Paraboloid Grip Points

The aim Z axis grip point is located along the local +Z axis of the paraboloid, at a distance of three quarters of the height. Selecting this grip and then entering or picking a point will orient the paraboloid so that the +Z axis is aimed at the point by rotating about the base point in the plane formed by the current +Z axis and the vector from the base point to the pick point. Thermal Models

4-39

The start angle and end angle grip points are located on the surface of the paraboloid at one quarter and three quarters of the way along the height. An angle is set by projecting the line from the origin to the picked point onto the local XY plane. The base radius grip is located on the base of the paraboloid, half way between the start and end angles. Selecting the base radius grip and picking a point will modify the base radius by projecting the line from the base point to the picked point onto the local XY plane. The distance of the projected line is used for the radius. The top radius grip is located on the top of the paraboloid, half way between the start and end angles. Selecting the top radius grip and picking a point will modify the top radius by projecting the line from the base point to the picked point onto the local XY plane. The distance of the projected line is used for the radius. The top radius must be greater than the base radius, and the base radius may be zero. If the base radius is zero, the radius grip and the stretch grips will coincide. In this case, the grip will always modify the radius. To modify the base point position of paraboloid that has a zero radius base, drag the grip slightly to set a non-zero radius, modify the base, and then set the radius back to zero. The cursor will automatically snap to displayed grips, making the input operation easy. 4.3.16

Polygon

The Thermal > Surfaces/Solids > Polygon menu choice is the command to create an n-sided polygon. If three points are selected, then a triangle is created. If four points are input, then a quadrilateral is created. If five or more points are input, then an n-sided polygon is created. The created polygon is a single node. The may make each triangle in the polygon an individual node by using the Thermal > Modeling Tools > Toggle FD Mesh Nodalization command (see “From AutoCAD Surface” on page 4-34). Polygons may also be created from AutoCAD 3D faces or polymeshes. The may edit the points in the polygon by selecting the polygon, and then by moving the grip points. 4.3.17

Rectangle

A rectangle is created using the Thermal > Surfaces/Solids > Rectangle menu choice. The following prompts will appear at the command line: >: Origin point <0,0,0>: Enter or pick a point on the screen

This point defines the origin of the rectangle, which also defines the location of one corner. >: Point for +X axis and X-size <@1,0,0>: Enter or pick a point on the screen

The vector from the origin point to this point defines the +X axis. The length in the X direction is also set to coincide with the point. >: Point to set XY plane and Y-size <@0,1,0>: Enter or pick a point on the screen

4-40

Thermal Models

The local +Z axis is computed by taking the cross product of the local +X axis with the vector from the origin to the point. The +Y axis follows from the right-hand rule. After the +Y axis is computed, the Y-size is set by projecting the line from the origin to the point onto the +Y axis. The grip points for the rectangle are shown in Figure 4-29. The move origin grip relocates the origin to an arbitrary point. The stretch X size and stretch Y size grips modify the size of the rectangle without changing the orientation or location of the origin. A side is resized by projecting the line from the origin to the point onto the corresponding +X or +Y axis.

Figure 4-29

Thermal Desktop Rectangle Grip Points

The aim Z axis grip point is located along the local +Z axis of the disk, at a distance of one quarter of the largest side. Selecting this grip and then entering or picking a point will orient the rectangle so that the +Z axis is aimed at the point by rotating about the local origin in the plane formed by the current +Z axis and the vector from the origin point to the pick point. The aim X about Z and aim Y about Z grip points are located at one quarter of the way along the edges of the rectangle. An axis may be aimed at a point, restricted to rotating about the Z using these grips. The aim Y about X and aim X about Y grips can also be used to aim an axis constrained to rotate about the other axes. Want "Hands-On" Information? Gain experience working with rectangles by completing Section 20.5 "Circuit Board Conduction Example" on page 20-67, Section 21.1 "Radks for Parallel Plates" on page 21-3, Section 21.5 "Simple Satellite" on page 21-71, and Section 21.6 "Orbital Maneuvers" on page 21-87.

Thermal Models

4-41

4.3.18

Scarfed Cone

An scarfed cone is created using the Thermal > Surfaces/Solids > Scarfed Cone menu choice. The will be prompted to select the following points in order to create the scarfed cone. >: Pick or enter point for base of cone <0,0,0>: >: Pick or enter point for top of cone <@0,0,1>: >: Enter base radius or pick/enter point <0.0>: >: Enter top radius or pick/enter point <1.0>: >: Enter start angle or pick/enter point <0.0>: >: Enter end angle or pick/enter point <360.0>: >: Enter scarf angle or pick/enter point <0.0>:

4.3.19

Scarfed Cylinder

An scarfed cylinder is created using the Thermal > Surfaces/Solids > Scarfed Cylinder menu choice. The will be prompted to select the following points in order to create the scarfed cylinder. >: Pick or enter point for base of cylinder <0,0,0>: >: Pick or enter point for top of cylinder <@0,0,1>: >: Enter radius or pick/enter point <1.0>: >: Enter start angle or pick/enter point <0.0>: >: Enter end angle or pick/enter point <360.0>: >: Enter scarf angle or pick/enter point <0.0>:

4.3.20

Sphere

A sphere is created using the Thermal > Surfaces/Solids > Sphere menu choice. The following prompts will appear at the command line: >: Pick or enter point for center of sphere <0,0,0>: Enter or pick a point on the

screen This point defines the origin of the sphere. >: Pick or enter point to define +Z axis <@0,0,1>: Enter or pick a point on the

screen The vector from the origin point to this point defines the +Z axis. The local +X axis is computed so as to lie in the current UCS XY plane, and the +Y axis follows from the right-hand rule. The UCS is repositioned at the origin point with the axes aligned with surface’s local coordinate system for the remainder of the inputs. Points entered at the command line prompt may be proceeded with an asterisk to

4-42

Thermal Models

use the World Coordinate System, rather than the local coordinate system. The UCS is reset to the previous UCS after the surface has been created. >: Enter radius or pick/enter point <1.0>: Enter a value or point, or pick a point on

the screen The radius may be entered directly or computed from a point entry. A return without entering a value creates a sphere with a radius of one. If a point is input, the radius is computed using the length from the origin to the picked point. >: Enter start angle or pick/enter point <0.0>: Enter a value or point, or pick a

point on the screen The starting angle for the sphere is referenced counterclockwise from the local +X axis. A return without entering a value starts the sphere at the local +X axis. If a point is input, the line connecting the origin to the point is projected into the local XY plane to determine the angle from the local X axis. >: Enter end angle or pick/enter point <360.0>: Enter a value or point, or pick a

point on the screen The end angle for the sphere is referenced counterclockwise from the local +X axis. A return without entering a value ends the sphere at the local +X axis. If a point is input, the line connecting the origin to the point is projected into the local XY plane to determine the angle from the local X axis. The Thin Shell Data dialog box (Figure 4-7) will appear after this prompt. Top and bottom heights may be set using grips after the sphere is created. The grip points for the sphere are shown in Figure 4-30. The move origin grip relocates the origin to an arbitrary point. The top height and bottom height grips locate the dimensions of the sphere along the Z axis. The heights are computed by projecting the line from the origin to the selected point onto the Z axis. If a height is computed that is greater than the radius, the height is set equal to the radius. The aim Z axis grip point is located along the local +Z axis of the sphere, at a distance of 1.1 times the radius. The aim Z axis grip is located slightly outside the sphere to avoid the possibility of coinciding with the top height grip. Selecting this grip and then entering or picking a point will orient the sphere so that the +Z axis is aimed at the point by rotating about the local origin in the plane formed by the current +Z axis and the vector from the origin point to the pick point. The start angle and end angle grip points are located on the surface of the sphere at one quarter and three quarters of the way along the height. An angle is set by projecting the line from the origin to the picked point onto the local XY plane. The radius grip is located on the surface of the sphere half way between the start and end angles and half way between the top and bottom heights. Selecting the radius grip and picking a point will modify the radius using the distance from the origin to the picked point. The top and bottom heights are scaled along with the radius.

Thermal Models

4-43

Figure 4-30

4.3.21

Thermal Desktop Sphere Grip Points

Torus

An torus is created using the Thermal > Surfaces/Solids > Torus menu choice. The will be prompted to select the following points in order to create the torus. >: Pick or enter point for center of Torus <0,0,0>: >: Pick or enter point to define +Z axis <@0,0,1>: >: Select point to set large radius <2.0>: >: Enter small radius or pick/enter point <.25*large radius>: >: Enter start angle or pick/enter point <0.0>: >: Enter end angle or pick/enter point <360.0>:

4.4

Finite Difference (FD) Solids

Thermal Desktop s 3 kinds of finite difference solids. They are bricks, cylinders, and spheres. Arbitrary shaped solids may be built and analyzed in Thermal Desktop using finite elements (see “Finite Element (FE) Solid Elements” on page 4-58).

4-44

Thermal Models

4.4.1

FD Solid Data

Editing of FD Solids is very similar to editing surfaces. The FD Solid Edit dialog box uses a similar “tabbed” format, the tabs being: • Subdivision - Section 4.4.1.1 • Numbering - Section 4.4.1.2 • Cond/Cap - Section 4.4.1.3 • Radiation - Section 4.4.1.4 • - Section 4.4.1.5 • Advection - Section 4.4.1.6 • Insulation - Section 4.4.1.7 • Parameters - Section 4.4.1.8 • Translate/Rotate - Section 4.4.1.9 4.4.1.1

Subdivision Tab

The FD Solid Edit dialog box Subdivision tab allows the to set the subdivision for the three primary directions of the solid. The Subdivision tab is shown in Figure 4-31.

Figure 4-31

FD Solid Edit Dialog Box Subdivision Tab

Thermal Models

4-45

4.4.1.2

Numbering Tab

The FD Solid Edit dialog box Numbering tab allows the to set the submodel name and node IDs for the nodes that are attached to the solids. The Numbering tab is shown in Figure 4-32.

Figure 4-32

4.4.1.3

FD Solid Edit Dialog Box Numbering Tab

Cond/Cap Tab

The FD Solid Edit dialog box Cond/Cap tab allows the s to Select to Generate Cond/Cap as well as set the Material for the solid and the submodel that the conductors will be placed in. The may also specify to generate the nodes based on the material property or as arithmetic. The Cond/Cap tab is shown in Figure 4-33.

Figure 4-33

4-46

FD Solid Edit Dialog Box Cond/Cap Tab

Thermal Models

Scaling factors for density and conductivity may also be input. The density multiplier is really a capacitance multiplier in the solution. 4.4.1.4

Radiation Tab

The FD Solid Edit dialog box Radiation tab allows the to set the optical properties of each face as well as which faces will be active in each analysis group. The Radiation tab is shown in Figure 4-34.

Figure 4-34

FD Solid Edit Dialog Box Radiation Tab

To set the faces that are active in each group, simply double click on the Analysis Group name and a new dialog box will appear with check boxes for each face name. The radio buttons allow outside faces, inside faces, both, or none of the faces to be active in an analysis group. Faces that are checked will be included in the analysis group using the radio button selection; unchecked faces will not be in the analysis group. Please understand that if the is editing a solid cylinder that has minimum radius set to zero, that even though the can select the RMIN face, nothing will be generated for radiation since the resulting inner surface area would be zero. This is likewise true for GMIN and GMAX faces when a solid cylinder or solid sphere goes around the Z axis 360 degrees. (See "Analysis Group Active Sides" on page 4-5) The is not only able to set the analysis groups and optical properties for the outside faces of the solid, but also the inside faces. This allows modeling of transparent solids and internal reflecting faces. To assign optical properties to the inside faces, the selects the Inside Optical Props button and assigns the appropriate properties to the appropriate faces. Caution: If outside faces have transparent properties, be sure inside faces have properties assigned or radiation will be fully absorbed if the optical properties are left as DEFAULT.

Thermal Models

4-47

4.4.1.5

Tab Note: When thermal connection must be modeled between surfaces or solids, three choices are available: global (discussed below), ors (see “ors” on page 4-74), and merge nodes (see “Merge Coincident Nodes” on page 4-108). Each choice has advantages and disadvantages, so the is advised to review all three options to see which best fits the modeling needs.

The FD Solid Edit dialog box tab allows the to create conductance from the faces of the surface to other objects in the model. This works just like on surfaces, but there is not a thickness parameter since the FD Solid has a thickness defined (see “Area Calculations” on page 9-3). The tab is shown in Figure 4-35.

Figure 4-35

FD Solid Edit Dialog Box Tab

Each conductance can be input as either Absolute or as a function of the area or length of the edge. For the area calculations, the can select to have the test points generated at the exterior of the face (using the thickness of the surface) or at the mid plane of the surface (where the thickness value is ignored). Caution to s: Thermal Desktop does not calculate the conductance from the edge of a surface to a center node. Therefore when adding , the value must include that conductance, or alternately the should use edge nodes. Please see see “Circuit Board Conduction Example” on page 20-67 for guidance.

4-48

Thermal Models

4.4.1.6

Advection Tab

The Advection tab allows the to define mass flow in any of the major directions. This can be used to model relative motion of the solid such as an extrusion process or rotation of a wheel. See "Material Flow (Advection) Options" on page 9-8 for more discussion regarding this option.

Figure 4-36

4.4.1.7

FD Solid Edit Dialog Box Advection Tab

Insulation Tab

The FD Solid Edit dialog box Insulation tab allows the to add insulation to the face of any of the FD Solids. This feature works just like the insulation features that are associated with the planar surfaces. (see “Insulation Tab” on page 4-17). The Insulation tab

Thermal Models

4-49

is shown in Figure 4-37.

Figure 4-37

4.4.1.8

FD Solid Edit Dialog Box Insulation Tab

Parameters Tab

The FD Solid Edit dialog box Parameters tab allows the to change the parameters of the solid. These fields can all parameterized by double clicking in the field and then inputting a symbol. The Parameters tab is shown in Figure 4-38.

Figure 4-38

4-50

FD Solid Edit Dialog Box Parameters Tab

Thermal Models

The Comment field is a multi-line field that is extremely useful for model documentation. Text may be entered into the field or double-clicking will open up a text edit window. The first line of the comment field provides object identification in three locations: the Model Browser, the tool tip that appears when the cursor is positioned over the object, and as a comment line in the SINDA/FLUINT conductance/capacitance file. 4.4.1.9

Translate/Rotate Tab

The may specify additional translations and rotations for the surface using the TRANS/ROT tab. The tab is shown in Figure 4-16. These rotations and translations are

Figure 4-39

FD Solid Edit Dialog Box Translation/Rotation Tab

relative to the initial origin and local coordinate system of the surface. The has the option to specify the order of the rotations by using the drop-down menus for each axis. 4.4.2

Solid Brick

The finite difference solid brick is an orthogonal brick that can be broken down in all 3 directions (X, Y, and Z). When creating a brick, the will be prompted for 4 points. They are: >: Origin point <0,0,0>: >: Point for +X axis and X-size <@1,0,0>: >: Point to set XY plane and Y-size <@0,1,0>: >: Point to set Z-size or input length <1>:

The FD Brick will have 9 grip points for editing. These are shown in Figure 4-40. The FD Solid Brick has 6 rectangular faces. The can specify radiation, , and insulation for each face individually. Each face can have different optical properties, coefficients, or materials.

Thermal Models

4-51

Figure 4-40

4.4.3

Thermal Desktop FD Brick Grip Points

Solid Cone

The finite difference solid cone is cone that has breakdowns in 3 directions, radial, angular, and height. When creating a solid cone, the will be prompted for 6 points, They are: >: Pick or enter point for base of cone <0,0,0>: >: Pick or enter point for top of cone <@0,0,1>: >: Enter base max radius or pick/enter point <1.0>: >: Enter top max radius or pick/enter point <0.0>: >: Enter start angle or pick/enter point <0.0>: >: Enter end angle or pick/enter point <360.0>:

The can also enter different min radius for the base and top of the cone by editing the solid, or by using grip point manipulation. 4.4.4

Solid Cylinder

The finite difference solid cylinder is a cylinder that has a radial thickness which can be broken into nodal regions. When creating a solid cylinder, the will be prompted for 6 points. They are: >: Pick or enter point for base of cylinder <0,0,0>:

4-52

Thermal Models

>: Pick or enter point for top of cylinder <@0,0,1>: >: Enter max radius or pick/enter point <1.0>: >: Enter min radius or pick/enter point <0.0>: >: Enter start angle or pick/enter point <0.0>: >: Enter end angle or pick/enter point <360.0>:

The FD Solid Cylinder will have 7 grip points for editing. These are shown in Figure 4-41. The FD Solid Cylinder can have a varying number of faces based on the minimum radius being zero or not, and also if the Start and End angles form an entire circle. The maximum number of faces are 6, while the minimum number of faces is 3. Each face can have radiation, , and insulation, as well as different optical properties for each face.

Figure 4-41

4.4.5

Thermal Desktop FD Solid Cylinder Grip Points

Solid Ellipsoid

The finite difference solid ellipsoid is an ellipsoid that can be filled or hollow with a varying wall thickness and can be broken into nodal regions. If two radii are equal, then a spheroid is created and if all three radii are equal a sphere is created. 4.4.5.1

Parameters

The solid ellipsoid has 12 parameters for defining its shape. These parameters, their definitions and any limits are given below. • Outer X-Radius - the Nowadays radius of the ellipsoid’s outer wall along the ellipsoid’s X axis. • Outer Y-Radius - the equatorial radius of the ellipsoid’s outer wall along the ellipsoid’s Y axis.

Thermal Models

4-53

Figure 4-42

Thermal Desktop Finite Difference Solid Ellipsoid

• Outer Z-Radius - the polar radius of the ellipsoid’s outer wall along the ellipsoid’s Z axis. • Outer Min Height - the height (in the Z direction) of the lower (-Z) side of the ellipsoid’s outer wall. The Outer Min Height must be less than the Outer Max Height and its magnitude must be less than or equal to the Outer Z-Radius. When the magnitude is less than the Outer Z-Radius, the ellipsoid’s outer wall is truncated on the -Z side. • Outer Max Height - the height (in the Z direction) of the upper (+Z) side of the ellipsoid’s outer wall. The Outer Max Height must be greater than the Outer Min Height and its magnitude must be less than or equal to the Outer Z-Radius. When the magnitude is less than the Outer Z-Radius, the ellipsoid’s outer wall is truncated on the +Z side. • Start Angle - the angle measured about the +Z axis from the X axis defining the initial face of the ellipsoid. • End Angle - the angle measured about the +Z axis from the X axis defining the final face of the ellipsoid. The End angle must be greater than the Start Angle. • Inner X-Radius - the equatorial radius of the ellipsoid’s inner wall along the ellipsoid’s X axis. The Inner X-Radius must be less than the Outer X-Radius. Inner Radii greater than zero create a hollow ellipsoid.

4-54

Thermal Models

• Inner Y-Radius - the equatorial radius of the ellipsoid’s inner wall along the ellipsoid’s Y axis. The Inner Y-Radius must be less than the Outer Y-Radius. Inner Radii greater than zero create a hollow ellipsoid. • Inner Z-Radius - the polar radius of the ellipsoid’s inner wall along the ellipsoid’s Z axis. The Inner Z-Radius must be less than the Outer Z-Radius. Inner Radii greater than zero create a hollow ellipsoid. • Inner Min Height - the height (in the Z direction) of the lower (-Z) side of the ellipsoid’s inner wall. The Inner Min Height must be less than the Inner Max Height and its magnitude must be less than or equal to the Inner Z-Radius. When the magnitude is less than the Inner Z-Radius, the ellipsoid’s inner wall is truncated on the -Z side. • Inner Max Height - the height (in the Z direction) of the upper (+Z) side of the ellipsoid’s inner wall. The Inner Max Height must be greater than the Inner Min Height and its magnitude must be less than or equal to the Inner Z-Radius. When the magnitude is less than the Inner Z-Radius, the ellipsoid’s inner wall is truncated on the +Z side. 4.4.5.2

Creation

After selecting Thermal > Surfaces/Solids > Solid Ellipsoid, the will be prompted for 7 points or 2 points and 5 values. They are: >: Pick or enter point for center of Ellipsoid <0,0,0>: >: Pick or enter point to define +Z axis <@0,0,1>: >: Enter X-radius or pick/enter point <2.0>: >: Enter Y-radius or pick/enter point <.5*X>: >: Enter Z-radius or pick/enter point <Same as Y>: >: Enter start angle or pick/enter point <0.0>: >: Enter end angle or pick/enter point <360.0>:

Picking a point in the graphics area for the radii values will calculate the distance from the origin to the selected point in the direction of the radius. 4.4.5.3

Grip Points

The solid ellipsoid has 9 grip points for editing. These are shown in Figure 4-43. 4.4.6

Solid Sphere

The finite difference solid sphere is a sphere that has an inner and outer radius. When creating a solid sphere, the will be prompted for 6 points. They are: >: Pick or enter point for origin of sphere <0,0,0>:

Thermal Models

4-55

Figure 4-43

Thermal Desktop FD Solid Ellipsoid Grip Points

>: Pick or enter point to define Z axis of sphere <@0,0,1>: >: Enter max radius or pick/enter point <1.0>: >: Enter min radius or pick/enter point <0.0>: >: Enter start angle or pick/enter point <0.0>: >: Enter end angle or pick/enter point <360.0>:

The FD Solid Sphere will have 6 grip points for editing. These are shown in Figure 4-44. The FD Solid Sphere will have a varying number of faces based on the parameters of the surface. The maximum number of faces are 6, and the minimum is 1. Each face can have radiation, , and insulation. The can also specify different properties for each face. 4.4.7

Applying Boundary Conditions to Faces of Finite Difference Solids

A may desire to place a heat load, pressure load, or heater to one or faces of a finite difference solid brick, cylinder, or sphere. This can be accomplished in one of two ways. When the command is issued to create a heat load on surface, the is prompted to select the surfaces. If the selects a finite difference solid, the location that the selects on that solid will determine which face the heat load is applied to. Figure 4-45 shows the different locations for selecting faces of Finite Difference Solid Brick. If the selects where the cursor is shown, the XMIN face will be selected (as displayed in the tool tip). If the selects a line in the middle of the solid, then ALL of the faces will be selected. If the fails to select the desired face with the cursor, the can always directly edit the selected faces by selecting the object (heat load, pressure load, or heater) in the list of surfaces

4-56

Thermal Models

Figure 4-44

Thermal Desktop FD Solid Sphere Grip Points

on the proper Edit dialog box, and then selecting Edit. When that is done, the Select Faces dialog box (Figure 4-45) appears showing the face names.

Figure 4-45

Applying a Boundary Condition to a Finite Difference Solid Face

Please note that in the Solid Cylinder and Solid Sphere, if the minimum radius is zero, there is no face for that situation. If a selects that face for a boundary condition, the program would know that the face does not exist, and that selection would be ignored by the program. The faces of the solid objects are:

Thermal Models

4-57

Brick XMIN, XMAX. The faces perpendicular to the X axis of the brick as defined when the brick was created. YMIN, YMAX. The faces perpendicular to the Y axis of the brick as defined when the brick was created. ZMIN, ZMAX. The faces perpendicular to the Z axis of the brick as defined when the brick was created.

Cone/Cylinder/Ellpsoid GMIN, GMAX. The faces at the starting and ending angles around the Z axis, as defined when the object was created. If the shape revolves 360 degrees around the Z axis, then GMIN and GMAX are not used. RMIN, RMAX. The faces at the minimum and maximum radii. If the minimum radius is zero, then RMIN is not used. HMIN, HMAX. The faces at the minimum and maximum height in the Z direction, as defined when the shape was created. If the maximum radius of a cone is zero at the minimum or maximum height of the shape, then the HMIN or HMAX, respectively, is not used. If the minimum or maximum height of an ellipsoid is equal to the Z semi-major axis, then the HMIN or HMAX, respectively, is not used.

Sphere GMIN, GMAX. The faces at the starting and ending angles around the Z axis, as defined when the object was created. If the shape revolves 360 degrees around the Z axis, then GMIN and GMAX are not used. RMIN, RMAX. The faces at the minimum and maximum radii. If the minimum radius is zero, then RMIN is not used. BMIN, BMAX. The faces of the minimum and maximum beta angle (measured from the -Z axis to the +Z axis. If the minimum beta angle is zero or the maximum beta angle is 180, then the BMIN or BMAX, respectively, is not used.

4.5

Finite Element (FE) Solid Elements

Thermal Desktop s eight types of solid elements, four linear and four curved elements. All linear element types can be created by one of several methods within Thermal Desktop. All element types can be imported from TD Direct or a third-party mesher. The linear element types are: • four-node tetrahedron • five-node pyramid

4-58

Thermal Models

• six-node wedge • eight-node brick The curved elements have curved edges and faces, but do not have more nodes than their linear counterparts. Additionally, the number of internal conductors is not increased. • tetrahedron • pyramid • prism • brick Examples of each type of linear solid element are shown in Figure 4-46.

Figure 4-46

Linear Solid Element Node definition

Finite elements can be extremely useful when the shape of the geometry does not fit one of the primitive surface shapes (e.g. - a plate with a hole) and provide an accurate conductance network in most situations. The should follow these guidelines when choosing to use finite elements: 1. Linear finite elements require more elements, and therefore more nodes, to define a curved shape. 2. When trying to minimize the number of finite elements, the element shapes should be as close to ideal as possible: square for quadrilateral and equilateral triangle for triangular. Element can be checked using Check Elements (Section 8.12). 3. For very small models, use as many finite elements as you would finite difference nodes. You will have more nodes with the finite element approach. If minimizing the number of nodes is important, than use finite difference objects. FE solids may be created using one of the following methods: • Create FE Tetrahedron (Tet) Solids Manually - Section 4.5.1 • Create FE Pyramid, Prism, or Brick Solids Manually - Section 4.5.2 • Extrude/Revolve Planar Elements Into Solids - Section 4.16.5 • Map Solid Mesh Between Conics - Section 4.16.6

Thermal Models

4-59

• Convert Finite Difference to Finite Elements - Section 7.11 • Modeling with TD Mesher - Section 14 • Import Finite Element Model - Section 18.2.3 • Link to TD Direct - Section 18.6 The list above can be considered to be in ascending order of preference. Manually create elements should be reserved to special requirements, such as filling in holes. 4.5.1

Create FE Tetrahedron (Tet) Solids Manually

Prerequisite: Nodes (Section 4.6) for the element vertices must have already been created before creating a FE tet • Icon: • Command: rcTet • Menu: Thermal > FD/FEM Network > Tet Element • Ribbon: Thermal > Create Network > Tet Element • Toolbar: Network Objects When the command is issued, the receives the following prompt: >: Select nodes for linear element:

At each prompt, the selects a node representing a vertex of the element. Four nodes can be selected. When all nodes have been selected, the hits <Enter> to complete the element. 4.5.2

Create FE Pyramid, Prism, or Brick Solids Manually

Prerequisite: Nodes (Section 4.6) for the element vertices must have already been created before creating a FE solid. • Icon: • Command: RcLinearElement • Menu: Thermal > FD/FEM Network > Element • Ribbon: Thermal > Create Network > Element • Toolbar: Network Objects When the command is issued, the receives the following prompt: >: Select nodes for linear element:

At each prompt, the selects a node representing a vertex of the element. Either five, six or eight nodes can be selected. When all nodes have been selected, the hits <Enter> to complete the element. 4-60

Thermal Models

4.5.3

FE Solid Data

When a solid FE element is edited (Section 2.3), the Solid Element Attributes dialog is displayed (Figure 4-47).

Figure 4-47

Solid Element Attributes Dialog Box

Comment. See Section 2.10.4 Material. Select the material or material alias name from the drop-down menu. Material Orienter. Type in the name of the material orienter (Section 3.2.6) if the material is anisotropic, a laminate, or an aggregate. Cond Submodel. Select of type the name of the submodel (Section 4.2) for the conductors. The node submodels are defined by editing the individual nodes.

Multiplication Factors Density. Enter a value or expression by which to multiply the material density for the node capacitances. X Conductivity. Enter a value or expression by which to multiply the material conductivity in the X direction of the material orienter. Y Conductivity. Enter a value or expression by which to multiply the material conductivity in the Y direction of the material orienter. Z Conductivity. Enter a value or expression by which to multiply the material conductivity in the Z direction of the material orienter.

Thermal Models

4-61

To apply radiation and conductance to a solid face, a planar element must be placed on that face. The command FD/FEM Network > Surface Coat Free Solid Faces will automatically place the planar elements on the solid (see Section 4.16.2). A surfacecoated element solely for radiation purposes should have zero thickness so the capacitance of the node will be calculated only from the solid elements. An alternative is to disable Generate Nodes and Conductors for the planar elements (Section 4.3.1.4). Another common way to create solid finite elements is to extrude planar surfaces or elements into solids (see “Extrude/Revolve Planar Elements Into Solids” on page 4-110) or to map mesh between two surfaces (see “Map Solid Mesh Between Conics” on page 4-111). Finite element shapes are defined by their nodes. In order to move an element, you must select and move its nodes (node visibility must be turned on). The may specify the material of the solid by selecting the solids and then the Thermal > Edit command. This command will bring up the Solid Element Attributes dialog box shown in . Input for a comment, thermophysical property material, and a material orienter (used for anisotropic properties) are available. The may also specify the conductor submodel that the network description for each element will be placed into. Finally, the may input some multiplication factors for the density and the conductivity in all three directions. Want "Hands-On" Information? Refer to tutorial exercise "Beer Can Example" on page 20-89 for experience working with solid elements.

4.6

Nodes

Thermal nodes are the most basic portion of the thermal network. Nodes store energy, represented by temperature. Some types of nodes can also gain and release energy as a function of time. Nodes are either defined by the (-defined Nodes) or defined by surfaces and elements (TD/RC Nodes). Node visibility can be controlled by the method of definition in the Preferences Graphics Visibility tab. All nodes must be connected to the thermal network for SINDA/FLUINT to run. If a node becomes detached or otherwise has no connection to the other nodes in the model, the node is considered a STRAY NODE. Stray nodes will not appear as problems until SINDA/ FLUINT starts. If SINDA/FLUINT fails due to stray nodes, an AutoCAD group named STRAYNODES will be automatically created to assist with finding those stray nodes. Want "Hands-On" Information? The following tutorial exercises offer the opportunities to work directly with nodes: "Beer Can Example" on page 20-89; "Mapping Temperatures From a Coarse Thermal Model to a Detailed NASTRAN Model" on page 20-157; "or Example" on page 20-171; "Manifolded Coldplate" on page 22-37; and "FEM

4-62

Thermal Models

Walled Pipe" on page 22-99. 4.6.1

Node Types

Thermal Desktop allows the to define four types of nodes: arithmetic, diffusion, boundary, and clone. The node types are defined in Table 4-1. Table 4-1 Node Types

Node Type

Description

Diffusion

Diffusion nodes have a finite capacitance and, therefore, store and release energy.

Arithmetic

Arithmetic nodes have zero capacitance and respond instantaneously to any change in energy balance. The only required input is the Initial Temp.

Boundary

A boundary has infinite capacitance and has a defined temperature. The boundary node temperature is either constant or time-varying temperature: a constant temperature is defined by the Initial Temp field and a time-varying temperature is defined in a Tabular Input accessed by the Edit button beside the Time Varying check box. The tabular data for a timevarying temperature is treated by SINDA/ FLUINT as a cyclical condition with the period being the last time in the tabular data. Therefore, if the does not wish the temperatures to repeat cyclically during the solution, then the last time in the tabular data must be greater than or equal to the solution end time.

Thermal Models

Graphical Image (wireframe and shaded)

4-63

Table 4-1 Node Types

Node Type Clone

4.6.2

Description

Graphical Image (wireframe and shaded)

A clone node can be considered a graphical representation of an already existing node. A clone node has no properties, but can be used to connect conductors or as a postprocessing visual aid. The submodel name and ID number must be set to match the submodel and ID of the node to be cloned. Create Nodes • Icon: • Command: RcNode • Menu: Thermal > FD/FEM Network > Node • Ribbon: Thermal > Create Network > Create Node • Toolbar: Network Objects

When this command is issued, the receives the following prompt: >Enter location of node:

After entering or selecting a location for the node, the node will be displayed if defined nodes are visible. A node defined by the above command is always considered a -defined node. 4.6.3

Edit Nodes

No matter how a node is defined, -defined or TD/RC, the node can be edited by selecting the node and selecting Thermal > Edit. When the edit command is issued for a node the Node dialog is opened (Figure 4-48). Enabled/Disabled. Selecting the Enabled/Disabled button opens an expression editor. The expression should equal 0 or 1. If the expression is equal to zero, the node will be diabled. See Section 4.6.4 for more information about the behavior of disabled nodes. Submodel. The Submodel field in the upper left defines the submodel for the node (Section 4.2). The drop down menu allows selection of an existing submodel or the may type in a submodel name to create a new submodel. Submodel names entered as text that are not in the Submodel Manager will automatically be added to the Submodel Manager after the confirms the operation by selecting OK on a pop-up message dialog box. ID. The ID field contains the node number.

4-64

Thermal Models

Figure 4-48

Node Dialog Box

Comment. The Comment field is used for model documentation (see Section 2.10.4 on page 2-41). The first line of the comment will appear in the Model Browser, the node’s tool tip, and in the NODE DATA block. Initial Temp. The Initial Temp field defines the initial temperature of diffusion or arithmetic nodes or defines a fixed temperature for boundary nodes.

Thermal Models

4-65

Type/(calculated by elements) This frame is available for all -defined nodes. If the node is defined by an element, surface, or solid, the frame will be grayed out and the title changed to (calculated by elements). Diffusion. When selected, the node will be a diffusion node. Diffusion nodes have a finite capacitance and, therefore, store and release energy over time. The capacitance is defined by the following options: • Thermal Mass: With Use Material unchecked, the capacitance is entered into this field. • Volume or Mass: When Use Material is checked, the Thermal Mass option changes to a drop-down list with the choices of Volume or Mass. When Volume is selected, the enters the volume of the node and the specific heat and density are obtained from the selected material. When Mass is selected, the enters the mass of the node and the specific heat is obtained from the selected material. • Use Material:When this box is checked, the must select a material property name from the drop-down list. The material property provides the specific heat and density for the node. Arithmetic. When selected, the node will be an arithmetic node. Arithmetic nodes have zero capacitance and respond instantaneously to any change in energy balance. Boundary. When selected, the node will be a boundary node. A boundary node has infinite capacitance and has a defined temperature. The boundary node temperature is either constant or time-varying temperature. A constant temperature is defined by the Initial Temp field. • Time Varying: When checked, the inputs the time-varying temperature in the Tabular Input form accessed by the Edit button. The tabular data for a timevarying temperature is treated by SINDA/FLUINT as a cyclical condition with the period being the last time in the tabular data. Therefore, if the does not wish the temperatures to repeat cyclically during the solution, then the last time in the tabular data must be greater than or equal to the solution end time. • Steady State: When Time varying is checked, the can select the behavior of the node during stead state calculations. The options are: 4. Current SINDA Value - The node temperature will not be altered from its state in the solution. 5. Time Average - The node temperature will be set to the time-averaged temperature of the time varying array. 6. Max Value - The node temperature will be set to the maximum temperature of the time varying array. 7. Min Value - The node temperature will be set to the minimum temperature of the

4-66

Thermal Models

time varying array. 8. Initial Temp - The node temperature will be set to the Initial Temp value. Clone. When selected, the node will be a clone node. A clone node can be considered a graphical representation of an already existing node. A clone node has no properties, but can be used to connect conductors or as a postprocessing visual aid. The submodel name and ID number must be set to match the submodel and ID of the node to be cloned. Override calculations by elements/surfaces. If the node is defined by an elements, surface, or solid, the node can be converted to a -defined node by checking this box. When checked, the must select the desired type and associated options. Put in sub-network. When checked, the node will be included in a sub-network (see “Super Network” on page 9-16). 4.6.4

Disabled Nodes

When nodes are disabled under the Enabled/Disabled feature of the Node edit dialog, the node is no longer considered part of the thermal network. Below are the behaviors of disabled nodes: 1. Nodes can be disabled or deactivated from the Node dialog using the Enabled/Disabled button 2. A disabled node is meant to not be part of the calculations 3. Disabled nodes are designed for thin-shell primitives and polygons. Nodes cannot be disabled for any finite element or on the face of a finite difference solid. A disabled node on the inside of a finite difference solid is essentially modeled as a void 4. When a node is disabled, its label will be drawn as *INACTIVE.ID, where ID is the ID of the node. 5. Disabled node surfaces are not drawn in solid shaded graphics. They are drawn in wireframe. In solid shading mode, color contouring is disabled for surfaces that have disabled nodes. 6. In RadCAD, a disabled node is internally modeled as a completely transparent surface meaning it does not participate in radiation calculations. 7. For conductance and capacitance calculations, a disabled node is filtered at cond/ cap generation time, thus a conductor or heat load to that node would not be output to the SINDA deck. 8. For double-sided surfaces, if one side is disabled, the other side is automatically disabled. This means you cannot have one side disabled and the other side not dis-

Thermal Models

4-67

abled.

4.7

Conductors

A conductor, or set of conductors, may be created by using one of three menu choices under the Thermal > FD/FEM Network menu: Node-to-Node Conductor, Node-toNodes Conductor, and Node-to-Surface Conductor. When the Node-to-Node Conductor command or Node-to-Nodes Conductor command is issued, the is prompted to select a single node (the primary node) followed by either another single node or a group of nodes (the target node or nodes), respectively. Once the conductor is created the number of nodes is not limited, however, only nodes can be the target of the conductor. When the Node-to-Surface Conductor command is issued, the is prompted to select a single node (the primary node) and any number of thin shells (finite difference or finite element surfaces), finite difference solids, or pipes (the conductor targets). Domain tag sets can also be used for the conductor target. Finite element solids cannot be selected directly, but can be surface coated with planar elements for the connection (Section 4.16.2 "Surface Coat Free Solid FEM Faces"). After completing the command, conductor lines will be generated from the primary node to each of the target nodes. For Node-to-Surface conductors, the target nodes are the nodes associated with the target objects and arrows will indicate the side or sides of an object to which the conductor is connected. The conductor definition may be edited by selecting any one of the conductor lines followed by the Thermal > Edit command. The Conductor dialog box is show in Figure 4-49. At the top of the form is the Enable/Disable button (see Section 2.10.8 "Enable/Disable"), the Comment field (see Section 2.10.4 "Comment Field") and the Submodel selection for the generated conductors (Section 4.2). By default the conductor numbers are automatically generated (Auto-number ID selected). As an option, the may specify the number of the first conductor (ID number radio button selected). When ID number is selected, the enters the initial conductor ID in the field and the Add Code button becomes enabled and the is able to add Network Element Logic (see Section 2.10.10 "Network Element Logic"). If Network Element Logic has already been entered, the Add Code button will read Edit Code. The conductor Type can be one of three general options: Generic, Natural Convection, or Function of Temperature Difference. The Natural Convection options are not allowed for Node-to-Node(s) conductors unless the target nodes listed in the To list of the Conductor form are associated with a geometric object. The options for the conductor types are discussed in Section 4.7.1 "Generic Conductor Type", Section 4.7.2 "Natural Convection Conductor Type", and Section 4.7.3 "Function of Temperature Difference Conductor Type".

4-68

Thermal Models

Figure 4-49

Conductor Dialog Box

The Insulation Nodes checkbox, when checked, specifies that the generated conductors will connect to the insulation nodes of surfaces or solids in the To list. The insulation nodes will only be used if they are on the side to which the conductor is applied. For Node-toNode(s) conductors, all target nodes must be associated with a geometric object if the Insulation Nodes checkbox is checked. At the bottom of the dialog box, the From Node and the To objects are listed. The may reselect any of the objects connected by the conductors and may add or delete any of the To objects, as long as one remains. Specific faces of the To objects can be included in the conductors by selecting the objects in the To list and using one of the following options: • selecting the edit button (an asterisk [*] and a pencil) on the right side of the Conductor form • double-clicking an individual object • right-clicking an object and selecting edit from the contextual menu Selecting two sides of a thin-shell object will double the area when the Per Area checkbox is checked.

Thermal Models

4-69

If the From node or all nodes or surfaces in the To list are deleted, then the conductor will also be deleted. To avoid this, use domain tag sets described in Section 2.5. 4.7.1

Generic Conductor Type

Generic conductors may be defined to be one way by checking the One way conductor checkbox. One-way conductors are typically used for abstract model duplication or simple (no pressure drop) flow calculations. See “one-way conductors” in the SINDA/FLUINT manual. Generic conductors may be a function of time (Vs.Time checkbox) or a function of the temperature difference between the two nodes of each generated conductor (Vs. Temp Diff checkbox). If either of these two options is selected, the must input an array to define the function by clicking the Array... button (see Section 2.10.1 "Tabular Input"). The temperature difference option is often used to simulate natural convection correlations. If the first value in this array is greater than or equal to zero, then the program will take the absolute value of the temperature difference before the array lookup. If the first value is negative, then the difference is calculated as the From Node (primary node) minus the To node (target node). For Generic conductor types (Type drop-down list), the conductor parameter is typed into the field just below the Type drop-down list. The name, use and units of the value are determined from the Use material, Radiation conductor, and Per Area checkboxes. With Use material, Radiation conductor and Per Area all unchecked, the Value field contains a conductance. This value is divided amongst the individual conductors weighted relative to the area of the To nodes (target nodes). For node-to-node(s) conductors, the value is evenly divided among the target nodes if any of the target nodes are free nodes (nodes not associated with geometric objects). The sum of all conductor conductances created by this conductor form will sum to the Value independent of the size of the surfaces. If the Per Area checkbox is checked and Use material and Radiation conductor are unchecked, then the Value is multiplied by the area of each node associated with objects in the To list. Therefore a change in the size of the To objects would result in a change in the total conductance. The Per Area option cannot be used with free nodes (nodes not associated with geometric objects) as the target nodes. If Use material is checked and Per Area is unchecked (Radiation conductor and Use Material are mutually exclusive), the input value is Area/length, where area is the crosssectional area of the heat transfer path and length is the distance along the heat transfer path. The conductance is calculated internally by multiplying the conductivity of the specified material by the value in the Area/length field. The resulting total conductance is divided amongst the individual conductors weighted relative to the area of the To nodes. For nodeto-node(s) conductors, the value is evenly divided among the target nodes if any of the target nodes are free nodes (nodes not associated with geometric objects).

4-70

Thermal Models

If Use material and Per Area are both checked, the input field is Length. The conductance is calculated internally by dividing the specified material’s conductivity by the Length value and then multiplying that value by the area of each node for each individual conductor. Again, the Per Area option cannot be used with free nodes (nodes not associated with geometric objects) as the target nodes. If only Radiation conductor is checked, then the input value is Radk. A radk is i*Ai*Bij, where i is the emissivity of the surfaces, Ai is the area of the surfaces, and Bij is the grey-body exchange factor from the surfaces to the node. If the node can be assumed to be a large black body (neglegible reflections) then Fij can be substituted for Bij. The Radk value is divided amongst the individual conductors weighted relative to the area of the target nodes. For node-to-node(s) conductors, the value is evenly divided among the target nodes if any of the target nodes are free nodes (nodes not associated with geometric objects). The sum of all conductor conductances created by this conductor form will sum to the Radk value independent of the size of the surfaces. The condutors are generated in the SINDA/FLUINT input deck as radiation conductors for which the heat rate is *radk*(Ti^4+Tj^4), where is the Stefan-Boltzmann constant†. If Radiation conductor and Per Area are both checked, then the input value is Bij for node-to-surface conductors and Bij*emissivity for node-to-node(s) conductors. This Bij is the exchange factor from the surface(s) to the node. The Bij is multiplied by the area of each target node and the emissivity of each of the target nodes to calculate the radk for each target node to the primary node. These conductors will be generated as radiation conductors as described in the previous paragraph. Note: For Node-to-node conductors, changing the subdivision of an object with which a target node is associated will cause the target node to become disassociated with the object keeping the conductor, but disconnecting the conductor from the object. Want "Hands-On" Information? Gain experience working with nodeto-surface conductors by completing Section 20.6 "Beer Can Example", and Section 22.5 "FEM Walled Pipe". 4.7.2

Natural Convection Conductor Type

Thermal Desktop provides access to natural convection correlations for a thermal-only model using air, without having to develop a fluid model in FloCAD. The can select the built-in natural convection correlations when creating either a node-to-node conductor or a node-to-surface conductor. The correlations used by Thermal Desktop are subroutines available in SINDA/FLUINT and are described in detail in the SINDA/FLUINT ’s Manual.

† The Stefan-Boltzmann constant is automatically included in the calculation in the system-of-units defined by the (Section 2.7.1)

Thermal Models

4-71

Note: The Natural Convection conductor type cannot be used with free nodes (nodes not associated with geometric objects) as the target nodes. The is cautioned that these correlations are approximations and many assumptions are made during their application, such as applying a constant heat transfer coefficient across the entire surface. Additionally the fluid is assumed to be free flowing in space around the surface, which is often not true. For example if you have a vertical circuit board mounted on a horizontal surface, the air flow at the bottom of the vertical board most likely has a near-zero velocity. To help overcome some of these assumptions, a multiplication factor is provided to give the the ability to scale the heat transfer coefficient. For more accuracy the fluid can be modeled using FloCAD. The can access the convection correlations through the Type drop-down list on the Conductor dialog box as shown in Figure 4-50.

Figure 4-50

Conductor Dialog Box, Natural Convection Routines

If the selects a natural convection correlation, the dialog box will update to display input parameters necessary for the type of natural convection selected. Two options, Horizontal flat plate upside and Horizontal flat plate downside, produce logic calling either NCHSU (subroutine for hot side up or cold side down) or NCHSD (subroutine for hot side down or cold side up) based on the temperature difference between the plate and the air. These two options are preferred over Horizontal flat plate upward heated or downside cooled and Horizontal flat plate downward heated or upside cooled since those options are fixed and will not change when conditions change.

4-72

Thermal Models

The characteristic dimension for the natural convection correlations must be provided by the since configurations are too variable for the parameter to be calculated automatically (for example, multiple surfaces may be combined together or the heat transfer area may only be a small fraction of a larger surface). The characteristic dimension for horizontal plates is the Area/Perimeter and for vertical plates is the height of the surface. When creating convection on a rectangular fin, the must provide inputs for height, width, and gap. These parameters are graphically defined in Figure 4-51.

Figure 4-51

Rectangular Fin Geometry Inputs

The subroutines implemented for these natural convection correlations are detailed in the SINDA/FLUINT manual. To assist the , Table 4-2 matches Thermal Desktop’s natural convection options with the SINDA/FLUINT subroutines as reference to further documentation in the SINDA/FLUINT ’s Manual.

4.7.3

Function of Temperature Difference Conductor Type

The Function of Temperature Difference conductor type (selected from the Type drop-down menu) allows the to specify a Multiplication Factor and an Exponent. The conductance of individual conductors is calculated internally as: area*(Multiplication Factor)*(deltaT)**Exponent

where: area is the area of the target nodes; and deltaT is the absolute value of the temperature difference between the primary and target nodes. This option is available for node-tonode(s) and node-to-surface conductors.

Thermal Models

4-73

Table 4-2 Convection Correlation Options for Conductors

Natural Convection Option

SINDA Routine Called

Vertical flat plate, isothermal

NCVFPT

Vertical flat plate, isoflux

NCVFPF

Vertical cylinder, isothermal

NCVCT

Vertical cylinder, isoflux

NCVCF

Horizontal flat plate

NCHSU

heated plate facing up (Tplate>Tair) cooled plate facing down (Tplate
NCHSD

cooled plate facing up (Tplate Tair) Horizontal flat plate, upside

NCHSD or NCHSU

Routine selected through logic based on actual temperatures Horizontal flat plate, downside

NCHSD or NCHSU

Routine selected through logic based on actual temperatures Horizontal cylinder, isothermal

NCHCT

Rectangular fins, isothermal

NCFINS

Sphere with gas inside

NCSPHERE

Note: The Function of Temperature Difference conductor type cannot be used with free nodes (nodes not associated with geometric objects) as the target nodes.

4.8

ors Note: When a thermal connection must be made between surfaces or solids, the should consider: ors (explicit connection via a conductance); or merge nodes (intimate connection, see “Merge Coincident Nodes” on page 4-108). For finite conductances, consider using ors. For very large conductances or perfect (infinite conductance), consider using merge nodes, if possible.

4-74

Thermal Models

ors are a -defined thermal conductance between two or more high-level objects, such as 2D surfaces‡, 3D solids (FE solids must first be surface coated), and FloCAD pipes (Section 5.4). ors are defined as “from” one or more objects “to” one or more objects. As explained later in this section, the designation of “from” and “to” has a significant effect on how the connection is calculated. ors have a number of distinct advantages over the outdated global option on the thin shell and solid edit forms (Section 4.3.1.5). These advantages are: • A or allows for a gap to exist between the From object(s) and the To object(s). A maximum allowable gap is set for each or on the or dialog box using the Tolerance field. • The or will only check for with objects in the To list. This is contrasted with the global (surface/solid edit form) method, which will check for with all objects in the model and will connect with any object with which the defining surface or edge is adjacent. The or allows the to explicitly define which objects are intended to be connected, even if they are not the closest objects. ors provides the with control over the connection in case of any ambiguity in the model. Not checking every object in the model for will save time in large models. • A or allows the to select between radiation or conduction (linear) type connections. Conductors can be specified as a function of time or temperature difference. Linear conductors can use a material property, including pressure dependent materials, to define the interface; only the x-direction conductivity is used for anisotropic materials. The FD/FEM Network > or command will prompt the to select objects between which conductors are to be created. Selected objects can be surfaces, solids (FE solids must first be surface coated) or FloCAD pipes. (Note that fluid-only pipes will not generate any conductors, but are still allowed in the list.) The first selects a group of objects from which ors are to be generated. Next the selects the objects to which the might be made. An object cannot have a or to itself. When selection is complete, the Contractor dialog box appears as shown in Figure 4-52. The or dialog box allows the to: • specify the thermal submodel to contain the generated conductors • specify whether the edges or the faces of the object are used to find connections • define the conduction coefficient • choose the number of integration intervals • set the tolerance (or allowable gap) for connections • choose the connection testing algorithm ‡Thermal Desktop polygon surfaces (see Section 4.3.16) can be used for face ors but not for edge ors. If the wishes to use an edge for a surface generated as a polygon, then the polygon should be converted to finite elements (see "Convert Finite Difference to Finite Elements" on page 7-9).

Thermal Models

4-75

Figure 4-52

or Dialog Box

• define scaling factors • choose the specific edges or faces of objects between which connections will be made, and • subtract, or change visibility of objects associated with the or Some typical uses for ors and suggested settings are provided in Appendix A "or Table". The Comment field (Section 2.10.4) and the Enable/Disable button (Section 2.10.8) are at the top of the form. The conductor submodel and the choice of edge or face connections are selected by using the Conductor Submodel and or From drop-down lists, respectively. From Edges uses the active edges of 2-D objects (or the length of a pipe) in the From list to calculate the interface with objects on the To list. From Faces uses the active faces of 2-D or 3-D objects in the From list to calculate the interface with objects on

4-76

Thermal Models

the To list. The choice of edge or face connections determines the units (areal versus lineal) of the conduction coefficient. This election also determines the options available when objects in the From list are edited when using the edit icon below that list. The conduction coefficient section of the or dialog box, along with the geometry of the From objects and the integrated area, form the basis for the conductors to be generated for SINDA/FLUINT. In this section, the can provide a fixed conduction coefficient or define the value as a function of time or temperature. The can also make selections to define the or to be based on a conductive material, material independent, or even based on radiation. For linear (non-radiation) conductors, the value written to SINDA/FLUINT is the equivalent of kA/L or hA, where k is the conductivity, A is the area of the interface, L is the thickness of the interface and h is the heat transfer coefficient. The result of the radiation or is a Radk, or Emis(i)*A(i)*Bij, where 'i' represents the emitting surface and 'j' represents the receiving surface. For the or, ‘i’ is the ‘from’ surface and ‘j’ is the ‘to’ surface. By default, the area is calculated by the or calculations. For grey-body radiation, the Bij is the amount of energy emitted from surface i and absorbed by surface j by all possible paths, including reflections. If the geometric view factor (Fij) is used instead of Bij, then Emis(i) should be an effective emissivity between the two surfaces instead of the emissivity of the 'from' surface. The Input Value Type determines how the Conduction Coefficient is distributed on the ing areas. • Per Area Or Length multiplies the Conduction Coefficient by the ed area of each node on the FROM surfaces. The total value of conductors will be (Conduction Coefficient)*(total ed area of FROM surfaces). • Absolute - (If not 100% connected, total cond output is less than input value) defines the total conductance of all FROM surfaces. If the FROM surfaces are not completely in with TO surfaces, then the total value of conductors will be less that the given Conduction Coefficient. The total value of conductors is distributed among the nodes of the FROM surfaces proportional to their ed areas. The total value of conductors will be (Conduction Coefficient)*(total ed area of FROM surfaces)/(total area of FROM surfaces). • Absolute - (Total conductance output equals input value) defines the total conductance of the or. The conductance coefficient is distributed among the nodes of the FROM surfaces proportional to their ed areas. Selecting Use Material will allow the to choose a material in the material properties database with the drop-down list; the conductivity of the material, even temperature- or pressure-dependent, is used in the calculation of the resulting conductances (for ors using pressure-dependent material properties, the pressure load must be applied to the FROM surfaces, see Section 4.12). Selecting Radiation will define the conductors as radiation conductors.

Thermal Models

4-77

Selecting Use Insulation Nodes will create the ors from insulation nodes on the From surfaces and also use any insulation nodes on the To surfaces, if a side with insulation is closest to the From surface. Selecting One Way will create one-way conductors** from the From surface to the To surface. The Integration Intervals option specifies the number of test points on nodes within the From surfaces for calculating the interface area or length of the or. See Section 9.3 "Area Calculations" on page 9-3 and Section 9.4 "Edge Calculations" on page 9-6 for more information regarding how the integration intervals are used. The Tolerance value defines the distance a To object may be from a test point to allow a connection. If the From surface or edge is larger than the To surfaces, then a reasonable tolerance must be set. In addition, adiabatic sections of the From surface can be defined by using small tolerances. When connecting from small objects to large objects and no adiabatic sections exist, then the tolerance may remain at a very large value. By default, the Apply Surface Thickness to Test Points check box is checked and the thickness of 2-D objects is used for the test points. Deselecting that check box (i.e., no check mark) causes Thermal Desktop to ignore the surface thickness and to use the midplane of the From surface when testing for proximity. Again, see Sections 9.3 and 9.4 for more information. The thickness of To surface(s) is always used for the Point Algorithm described in the next paragraph; the thickness of the To surface(s) is never used for the Ray Trace Algorithm. In the Testing Algorithm section of the or dialog box, the can choose which algorithm is to be used to calculate the interface length or area. The Point Algorithm, available for both edge and face ors, checks in all directions for the To object within the tolerance and closest to a test point on a From object. The point algorithm is very accurate, but as the From and To lists increase in size, this algorithm can be slow. Therefore, the Max Check Objects field is available to speed up the process for the point algorithm. When the number of objects in the To list exceed this value, a special procedure is used to reduce the number of objects that will be checked (for a more detailed explanation of the Max Check Objects parameter, see page 9-6). The Point Algorithm will work for coplanar surfaces. The Ray Trace Algorithm, available only for face (and not edge) ors, shoots rays normal (perpendicular) to the From surface and deems the closest node to be whatever node that ray first encounters. When the surfaces are parallel, this algorithm works very well and is extremely fast. Note that the tolerance for the ray trace algorithm is the length of the rays. With this algorithm, s should make sure the proper Top or Bottom side is correctly specified. Note: Thermal Desktop checks for coplanar surfaces only when performing radiation calculations that are run through the Case Set Manager. Using the Ray Trace Algorithm with a model that ** One-way conductors transfer energy in one direction (See Section 2.14 in the SINDA/FLUINT ’s Manual for more information regarding one-way conductors). They are normally used to represent advection (material or fluid flow), and are not the same as thermal diodes.

4-78

Thermal Models

has overlapping (coplanar) surfaces will result in erroneous results. If radiation calculations are not going to be run, or if the just wants to check for overlapping surfaces at any time, the command Thermal > Model Checks > Check overlapping surfaces can be used. Regardless of the algorithm chosen, it is highly recommended that the check the connections using the Show or Markers command (see “Display /or Markers” on page 8-6). The Coefficient Scaling field allows for the specification of a function along the edges of each object (designated as the U and V direction for each edge or surface) in the From list. The tabular input form accepts the scaling factor as a function of the normalized edge length (a value between 0 and 1 with 0 being the start of the edge and 1 being the end of the edge). This allows the same scaling-factor function to be used for different-sized objects. For an edge conductor, the conduction coefficient is multiplied by the function that corresponds with that edge. Scaling across a face is computed by multiplying the two functions, i.e. G(U,V) = (Conduction Coefficient) * fu(U) * fv(V), where fu(U) is the function along one edge, and fv(V) is the function for the other edge. The same functions apply to all selected faces of a solid object. The scaling factor is ignored unless the check box to use scaling has been checked. The From and To lists on the form display which items are associated with the or. Objects can be added to and deleted from each list using the icons below the From and To lists on the form. There are also buttons on the form to turn visibility on/off, node numbers on/off, and to display only selected or display all entities. Choosing items on the From list and selecting the edit icon (or double clicking on items) allows the to select individual faces or edges for inclusion in the or. A or is graphically displayed by using small arrows that are sized relative to the size of a node. The From surfaces are shown with green arrows. For face , the green arrows will be drawn from the face in the direction of . For edge , the green arrows will be drawn along the same edge as the , and in both directions. The To surfaces are marked with a gold arrow that is drawn on both sides of the object for a planar surface, or that is drawn on all the faces of a finite difference solid. The graphical or entity will not be shown unless ors are displayed (see “Graphics Size” on page 229). Finally, note that or calculations can be time consuming for large numbers of surfaces. Thermal Desktop or calculations have an automatic restart capability to avoid unnecessary recalculation. Please see "Conductance Capacitance Parameters" on page 9-17 for more information and control of this feature. To illustrate some of the usage of the or, an example is shown in Figure 4-53. This example shows three identical square surfaces (viewed at an angle), with one of the surfaces located above the common edge of the lower two surfaces. The or has been specified to use the lower edge of the upper surface in the From list, and the lower surfaces in the To list. With a tolerance larger than the distance between the surfaces, half of the upper surface will be connected to each of the lower surfaces. In SINDA/FLUINT,

Thermal Models

4-79

Figure 4-53

or Example

two conductors will be created connecting the node of the upper surface to the nodes of the lower surfaces. For this image, the or markers are turned on to show the connections made in yellow (see “Display /or Markers” on page 8-6). Now, consider what happens if the lists were to be reversed, so that the From list contained the two lower surfaces, and the To list contained the upper surface. If the tolerance is left at the default large value, the results will be very different. This instance is shown in Figure 4-54. In this case, each lower surface would connect to the upper surface. The conductors would have twice the edge as in the original example because they are computed based on the From surfaces, which have twice the length.

Figure 4-54

or Example: Reversal of From and To Objects

If the tolerance of the or is reduced to be equal to the gap between the rectangles, Thermal Desktop will yield the same length as in the first example (Figure 4-53). This happens because half of each of the lower surfaces (the From object) is used for the or, which is the same total length as that used in the first example. Figure 4-55 shows the connected points in yellow, with the test points that do not meet the tolerance distance test depicted in red.

4-80

Thermal Models

Figure 4-55

or Example With Reduced Tolerance

The conductor values generated by the or are a function only of the input value on the or dialog box, and do not take into conductivity within the From or To objects. To illustrate this point using the above example, note that the conductance from the node center to the edge of the surface is ignored. If that conductance is of concern, consider using edge nodes to relocate the nodes to the edge of the surface. Want "Hands-On" Information? Create and edit ors in Section 20.5 "Circuit Board Conduction Example", Section 20.9 "or Example", and in three FloCAD tutorials, "Heat Pipe Model" on page 2223, "Manifolded Coldplate" on page 22-37, and "Drawn Shape Heat Pipe" on page 22-85.

4.9

Heat Loads

The FD/FEM Network > Heat Load On Surfaces/Nodes/Solids commands allows the to apply a constant or time varying heat load to specific entities. After selecting the entities to receive the heat load, the Heat Load Edit Form dialog box shown in Figure 456 will appear. At the top of the form is the Enable/Disable button (see “"Enable/Disable" on page 246), the Comment field (see "Comment Field" on page 2-41) and the submodel selection for the generated heat load. The type of heat can also be selected and may be a constant value, time dependent, temperature dependent, or time and temperature dependent. When using Heat Load on Nodes, only nodes can be selected. If the node is associated with a surface or solid, the heat load can be defined as an absolute load or a flux. For a flux on a heat load on nodes, the surface area of the node is used based on the selection of top, bottom, or both. When using Heat Load on Surfaces, surfaces or solids may be selected. If a solid is selected, the heat load will be applied to the chosen faces of the solid, as described

Thermal Models

4-81

Figure 4-56

Heat Load Edit Form Dialog Box

later. Finite element solids must be surface coated (see Section 4.16.2). Again, the heat load can be specified as absolute or a flux. When using Heat Load on Solids, only solids or solid elements may be selected. The heat load may be specified as absolute or volumetric. When volumetric or flux heat loads are used, the value provided for the heat load is multiplied by the volume or area, respectively, of each node. When an absolute heat load is used, the value provided is the total for all objects and the heat load applied to each node is evenly divided among the nodes if the heat load was applied to nodes, proportional to the surface area of each node if the heat load was applied to surfaces, and proportional to the volume of each node if the heat load was applied to solids. When the checkbox Put heat load into Insulation nodes is checked, then the heat load will be applied to the outer nodes of any defined insulation (Section 4.3.1.6). The icons on the right can be used to manipulate the objects that the heat load is applied to. Selecting Add will prompt the to select either surfaces or nodes to add to the list. Selecting Delete will remove the currently selected objects from the list. The Edit function can be used for surfaces and also FD Solids to determine the face on which to place the heat load. (see “Heat Loads” on page 4-81) If the heat load is time dependent, the must input an array of time dependency with an optional multiplier of the array data values (see Figure 4-57). A time varying heat load is assumed to be periodic. If a periodic heat load is not desired, the should input values that represent the entire time of a transient run. Also, if a time varying heat load is input and a steady state solution is being solved, then the time average heat load will be used for the steady state calculations. 4-82

Thermal Models

Figure 4-57

Time Dependent Heat Load Edit Form Dialog Box

If the heat load is temperature dependent (see Figure 4-58), the must input an array of temperature dependency using the Array button or may specify an existing SINDA array. For both options, an optional multiplier of the array data values can be specified. For the SINDA array option, units for the SINDA array may be chosen so appropriate multipliers can be applied. If the heat load is time and temperature dependent (see Figure 4-59), the must input a bivariate array of time and temperature dependency or may specify an existing SINDA bivariate array. As with the temperature dependent heat load, a multiplier may be specified for the array data values and units for the SINDA array may be specified. Want "Hands-On" Information? Section 20.8 "Mapping Temperatures From a Coarse Thermal Model to a Detailed NASTRAN Model" offers the a chance to apply a heat load.

4.10

Heaters

The Thermal > FD/FEM Network > Heater command allows the to apply a temperature dependent heater to nodes or surfaces. A heater is a thermostatically controlled heat load. The temperature used for the control (“sensed” as with a thermocouple) can be a different location that wher the heat load is applied. The is first prompted to select the

Thermal Models

4-83

Figure 4-58

Temperature Dependent Heat Load Edit Form Dialog Box

Figure 4-59

Time and Temperature Dependent Heat Load Edit Form Dialog Box

4-84

Thermal Models

objects receiving heat, followed by the sensor objects. If no sensor objects are selected, the heated objects will be used as the sensing objects. Objects that may be used for heaters or sensors are nodes, finite difference surfaces and solids, planar elements, and FloCAD pipes. Measures (Section 13) can be used as sensors. Once the objects are selected, the will see the Heater Edit Form dialog box shown in Figure 4-60. At the top of the form are the Enable/Disable button (see "Enable/Disable"

Figure 4-60

Heater Edit Form Dialog Box

on page 2-46), the Name field (see "Comment Field" on page 2-41) and the Logic Submodel selection pull-down for selecting to which submodel the heater logic will be added. Input Values. On the left side of the dialog box the inputs the available Heater Power and the set-point temperatures (On Temp and Off Temp) to cycle the heater on and off. The heater power may be input as a flux or power. The heater power may be negative, which

Thermal Models

4-85

allows the heater to perform as a cooler with thermostat. For heaters and coolers, the On Temp must always be lower than the Off Temp; Thermal Desktop will write the appropriate logic for heaters or coolers. Note: Heater are, by nature, transient objects. The parameters entered for the Input Values are used for transient solutions. For steady state conditions the use of the Input Values will be based on the options selected in the Steady State section on the right side of the form. Note: A heater without hysteresis (no difference between the setpoint temperatures) should not be modeled and, in the physical world, is difficult to produce and control. Note: If using a symbol for the heater power and writing the expression to SINDA (Section 2.10.7), do not allow the power to change from positive to negative or vice versa. Doing so will invalidate the logic written by Thermal Desktop. The Proportional Off/On button opens an expression editor that allows the to define an expression to determine if the heater is proportional (expression = 1, button read Proportional On) or bang-bang (expression = 0, button reads Proportional Off). A proportional heater changes the power based on the temperature: for a heater (not a cooler), if T>= OnTemp, then power is 100%; if T<=Off Temp, then power is 0%; if OffTemp < T < OnTemp, then power is proportional to the range between OffTemp and OnTemp. A bangbang heater is either at 0% power or 100% power. The Transient Scaling option allows the to input a scaling factor of the power versus time after clicking on the Edit button (see "Tabular Input" on page 2-37). A value of 1 is full power, while a value of zero effectively turns off the heater. Sense Method allows the to choose how the temperature of sensor surfaces is calculated: area-weighted average, maximum, minimum or logic input through the Edit button on the Heater Sense Method form. For logic sense method, the final result of

the logic must be placed into the T__TD (double underscore between ‘T’ and ‘TD’). Everything typed into the Program T__TD to the sense temperature form must follow formatted FORTRAN rules. Steady State. For Steady State calculations, the can select to operate with the heated surface Set to Mid-Point Temperature with an optional Offset Temp, with the heater at a Set Power, or with Proportional heater power based on the sensor temperature. The Set to Mid-Point Temperature changes the heated surface to a boundary condition and calculates the required heat to maintain the temperature. The Set Power treats the heater as a heat load. These two options do not control the heater using the sensor surface. The steady-state proportional option allows the choice of Sense Method. The power is set proportional to the temperature within the hysteresis band. The Damp Factor controls the changing of the power level from iteration to iteration preventing oscillations in the results. The Program button allows the to change the steady state behavior based on an expression.

4-86

Thermal Models

Important: If the heater is set to the mid-point temperature, the temperature of the heated surface is maintained regardless of the heater power or sensor temperature. Therefore a set power, proportional power or a transient solution should be used to heater sizing. The may decide whether or not to place the heater on the outer-most insulation nodes if the selected surfaces have insulation by selecting the Use Insulation nodes if possible check box. The sets of buttons at the bottom of the dialog box allow the to manipulate the lists of heated objects (Apply Heat To field) and sensor objects (Sense Temperature From field). Like Heat Loads, the can specify which face(s) of surfaces and FD Solids the heat is applied to (see “Heat Loads” on page 4-81). Heaters are graphically represented with red arrows pointing toward each heated node with an “H” at the end opposite the node. If sensor objects are selected, red arrows will be shown pointed away from each sensor node with an “S” at the end opposite the node.

Heater Performance During Solutions Thermal Desktop automatically generates s for SINDA/FLUINT to record heater parameters and allow review of heater performance. The s will be: • CH* - the number of times a heater turns on • HT* - a flag: 0 for heater off; 1 for heater on • OT* - the total on time of the heater • PO* - the power of the heater • TP* - the total energy used by the heater (time-integrated power) • XP* - steady state proportional fraction (for steady state proportional option) where the * indicates a wildcard that represents an append string unique to the heater in the Thermal Desktop model. The may specify this append string in the append string field (below the Logic Submodel of the Heater Edit Form) to customize the names. The Append string is limited to 30 alphanumeric characters. Thermal Desktop will not allow duplicate strings for multiple heaters. These s may be postprocessed using EZXY or the Excel Plotter template installed with Thermal Desktop if s are included in the results files (see Section 15.2.3 "Case Set - Output Tab"). A summary of the heater performance is provided at the end of the solution in the *.out file. When heaters are included in the thermal model, logic may be added to the following subroutines by Thermal Desktop depending on the selections made on the Heater Edit Form: • TDHTR - This subroutine holds the nodes of a heater (the objects to which heat is applied) at the specified temperature, the midpoint temperature with offset, during a steady state analysis. Logic is added to this subroutine only if the Set to Midpoint Temperature option is selected for Steady State. By default, a call statement is included in the OPERATIONS block for a steady state solution. Thermal Models

4-87

• TDREL - This subroutine releases the nodes of a heater (the objects to which heat is applied) after a steady state analysis. Logic is added to this subroutine only if the Set to Midpoint Temperature option is selected for Steady State. By default, a call statement is included in the OPERATIONS block after a steady state solution. • TDHTOT - This subroutine write a heater summary to the ASCII output file for each heater in the thermal model. By default, a call statement is included in the OPERATIONS block after a transient solution. • TDHTRST - This subroutine resets the heater s to zero (with the exception of PO*). The should call this subroutine if heater statistical data must be reset, as between parametric runs.

4.11

Thermoelectric Coolers (TECs)

The Thermal > FD/FEM Network > Thermoelectric Cooler command allows the to model a thermoelectric cooler (TEC) or other Peltier device. Note that a restriction on the TEC option is that the cold side surface and the hot side surface must be identical in subdivision, and must line up from each face. For detailed information on the TEC subroutine and its input parameters, the should consult the SINDA/FLUINT manual. Multi-stage coolers can also be modeled with this functionality by stacking TEC devices. A thermoelectric device uses the Peltier effect to heat or cool a component. The device is two sided and characterized by a hot side and cold side. The TEC command will prompt the to select the cold and hot sides of the cooler during creation. After selecting the cold and hot sides, the TEC dialog box shown in Figure 4-61 will be displayed. The first section of the form allows the to provide a comment (Section 2.10.4), the submodel (Section 4.2) for the conductors and generated logic, and the append string (Section 2.10.5). The must select the Input Mode of the cooler (Current, Voltage, Power) and the value of the input mode. After that, the should input the Area/Thickness of an individual couple in the device. This parameter is usually available on technical data sheets be can be provided by the manufacturer. If modeling a Bismuth Telluride device, select the Bismuth Telluride option (if not already selected) and input the number of couples contained in the cooler. If the couples are made of another material, select the Defined radio button and input the Seebeck Coefficients (Seebeck Coeff field), Effective Resistivity (Eff. Resistivity field), and Thermal Conductivity (Conductivity field) of the device. By default the program will generate conductors representing back conduction through the couples. This can be disabled by unchecking the Generate Conductors for Couples check box. Please note that this option does not for any filler material that surrounds the couples. If the wants to for the filler, then create either a or between the cold and hot side, or fill the gap between the cold and hot sides with a FD Solid Brick.

4-88

Thermal Models

Figure 4-61

TEC (Thermoelectric Cooler) Dialog Box

When using the brick option, conductance can be obtained by subdividing the hot, cold, and filler the same, use edge nodes, and merge coincident nodes between the brick and the hot and cold sides. Since the brick will for the entire volume between the hot and cold sides, the conductivity of the filler material should be adjusted based on volume occupied by the couples or the conductivity multiplier on the Cond/Cab tab of the brick edit form should be used. The Use Deep Solution Method option allows the to switch to a more involved solution. The should only turn on this method if the model is having problems converging. In some situations, turning this method on, may actually cause convergence problems. Please refer to the SINDA manual for a more in-depth description.

Thermal Models

4-89

The right side of the TEC dialog box allows the to add temperature control to either the hot or cold side for transient simulations. This control simulates on/off thermostatic control by default. A proportional option is available using the check box. For steady state simulations, the can elect to apply a constant input or midpoint control. For constant input, the defined percentage will be applied to the parameter defined as the Input Mode (i.e. Current, Voltage, or Power). When midpoint control is selected, the side of the device being controlled will be maintained at the midpoint of the on/off set-points. An offset is available to allow the to bias the device with a temperature offset from the midpoint control. Important: Use of the midpoint control method can result in a nonsensical solution. The graphical representation of the cooler shows a blue arrow for the cold side and a red arrow for the hot side.l 4.11.1

TEC Definitions

The following s are generated for the SINDA/FLUINT solution and the Append String is added to their names: • CC_ - TEC Cycle • CO_ - Optimum coefficient of performance at IO_ and VO_ • DT_ - Maximum temperature lift at zero cold load • IM_ - Current at maximum net cold production • IO_ - Optimum current (maximum CO_ for cold production) • KK_ - Calculated conductivity • OO_ - On/Off status • P_ - Power • QC_ - Maximum cold production at IM_ and VM_ • QO_ - Net cold production at IO_ and VO_ • R_ - Calculated electrical resistivity • S_ - Calculated Seebeck coefficient • VM_ - Voltage at maximum net cold production • VO_ - Optimum voltage (maximum CO_ for cold production) • V_ - Voltage

4-90

Thermal Models

4.12

Pressure Loads

The Thermal > FD/FEM Network > Pressure Load command will allow the to place a pressure load on surfaces. This pressure load will be used if the objects have pressure dependent conductivities associated with them (Section 3.2.3.1). The may specify the pressure in three different ways. The pressure may be constant, time varying, or linked to a file. The file for the last option would contain the x, y, z coordinates and the SINDA array that corresponds to that location (the file must be spacedelimited). Thermal Desktop will read that file and find the closest point on that file for the object being generated and link the pressure to the SINDA array. The is responsible for including the arrays in the SINDA model via the Case Set Manager (Section 15). For Thermal Desktop surfaces and planar elements, if the pressure-dependent material is used for the surface, the top insulation, the top face of a double-sided surface, or the core of a double-sided surface, the pressure must be applied to the top or outside of the surface. If the pressure-dependent material is used for the bottom insulation or the bottom face of a double-sided surface, then the pressure must be applied to the bottom of the surface. If a Thermal Desktop solid, or insulation on a Thermal Desktop solid, is defined by a pressuredependent property then the pressure can be applied to any face of the solid. Only one pressure can be specified for any object. The exception being one pressure can be applied to the top of a surface and another pressure applied to the bottom.

4.13

Boundary Condition Mapper

The Boundary Condition Mapper is used to apply boundary conditions to Thermal Desktop surfaces. These boundary conditions are typically calculated from CFD (but not limited to) and are imposed on the Thermal Desktop model via a mesh comprised of planar triangular and quadrilateral elements. The applied mesh must overlap the Thermal Desktop surfaces in 3d space to accurately map the boundary conditions. There are currently three types of boundary conditions that can be applied via the Boundary Condition Mapper. There are: • Heat Fluxes as a function of temperature • Temperatures • Convection/Temperature pairs The Boundary Condition Mapper will always map values to the outer surface of insulation if it exists on the face to which the Boundary Condition Mapper is applied. The Boundary Condition Mapper cannot be in a dynamic submodel. Be sure to place it in a static submodel if Dynamic SINDA is used (Section 15.2.5).

Thermal Models

4-91

4.13.1

Boundary Condition Mapper Interface

Invoking the Thermal > FD/FEM Network > Boundary Condition Mapper command will bring up an Open dialog box. From here the will select the file that contains the boundary condition data (see Section 4.13.2): unit definitions; the reference temperatures; the external mesh points and connectivity; and appropriate data at each time point. The mesh stored in the boundary condition data file will then be displayed as a graphical object aligned with the WCS. This graphical object can be manipulated using AutoCAD commands (move, rotate, align, etc.) to align it with the thermal model. Once the two models are aligned, the must select the mapper graphical object and use the Thermal-> Edit command. This opens the Boundary Condition Mapper dialog box (Figure 4-62).

Figure 4-62

Boundary Condition Mapper Dialog Box

From this dialog box, the must choose the Thermal Desktop objects to which the heat fluxes will be mapped. To select or add objects in the Objects to map from using points field, the clicks on the icon with the '+' and then selects objects using a selection method (click, selection box, groups, etc.). From the dialog box, the may define the submodel

4-92

Thermal Models

for the heat fluxes, change the related input file, set mapping parameters, set graphical object display preferences, test the mapping, and explode the external mesh to create finite elements and nodes. Boundary Condition Mappers can be viewed in the Model Browser (Section 2.4 "Model Browser") by selecting List > Mesh Displayers from the Model Browser Menus. The Enable/Disable button is documented in Section 2.10.8 on page 2-46. 4.13.2

Boundary Condition Mapper File Format

The file format is an ASCII format that can be created in a text editor. All commands must start in the first column of each line. An option exists to read in the data in a binary format. BCM DATA Input The first line of the file must be the DATA input definition. The DATA input tells the program the type of data that is being mapped. The DATA input must be one of the following 4 options. DATA: TEMPERATURE DEPENDENT HEAT FLUX DATA: TEMPERATURE DATA: COND_PER_AREA_TEMP_PAIR The above input assumes that the data is being provided at each node location. If the data is being provided at each element, then the input must be preceded by words ELEMENT BASED. For example: DATA: ELEMENT BASED TEMPERATURE DEPENDENT HEAT FLUX DATA: ELEMENT BASED TEMPERATURE DATA: ELEMENT BASED COND_PER_AREA_TEMP_PAIR For both node-based and element-based data, both node (grid) points and elements (connectivity) must be provided. The difference is the location of the data in the input file. BCM UNITS Input The UNITS input section defines units for the data values. This section shows the units allowed for each type of value. For composite units, the order shown is required. When power is part of the data value, the can choose to use power units or energy and time units. The UNITS LENGTH input is used to set the length units for the input of the NODE locations. The available case independent input units are: centimeters, millimeter, meters, inch, feet. UNITS LENGTH meters The UNITS TEMPERATURE input is used to set the temperature units for the inputs of all the available DATA types. The available case independent input units are: R, F, C, K.

Thermal Models

4-93

UNITS TEMERATURE R The UNITS TIME input is used to set the time units that are input for the data. The available case independent inputs are: SEC, MIN, HOU. UNITS TIME sec The UNITS DATA†† input is used to set the FLUX units for the TEMPERATURE DEPENDENT HEAT FLUX data inputs. The case independent units are: • Energy: BTU, J, KJ • Time: SEC, MIN, HOUR • Power: W, KW • Area: MM, METER, CENTIMETER, FEET, INCH UNITS DATA BTU/min/centimeter^2 or UNITS DATA W/mm^2 The UNITS COND_PER_AREA input is used to set the units for the COND_PER_AREA_TEMP_PAIR inputs. The case independent units are: • Energy: BTU, J, KJ • Time: S, MIN, H • Power: W, KW • Area: MM, METER, CENTIMETER, FEET, INCH • Temperature: C, K, F, R UNITS COND_PER_AREA BTU/min/feet^2/R BCM TEMPERATURES Input The TEMPERATURES input is used to set the temperatures for the TEMPERATURE DEPENDENT HEAT FLUX input. This input consists of the TEMPERATURES followed by the number of temperatures. The temperatures values follow this line with one temperature on each lines. An example is: TEMPERATURES 2 300. 1000.

††The keyword DATA is maintained to be consistent with earlier versions of the Boundary Condition Mapper.

4-94

Thermal Models

BCM GEOMETRY Definition The GEOMETRY definition is defined through the definition of nodes and elements. The NODE input is used to define the nodes and must precede the element inputs. A sample node input is: NODE 100 .1 .2 .3 Which defines node number 100 at coordinate x=.1, y=.2, z=.3. There are 2 types of elements that can be defined, they are simply TRI and QUAD. Samples are: TRI 998

100 101 102

QUAD 999 200 201 202 203 The TRI example defines element ID 998 as a connection between nodes 100, 101, and 102. The QUAD example defines element ID 999 as a connection between nodes 200, 201, 202, and 203. BCM TIME and Data Value Input The TIME input is simply TIME followed by the time value. The data values for the input time follow with one data value per line. If the data input is ELEMENT BASED, the input values are for the elements and are in the order that the elements are input. If the data is node based, the inputs are in the order that the nodes are input. Please look at the sample files for the input orders. TEMPERATURE DEPENDENT HEAT FLUX BCM Sample Input Note: Please note that all information including and after the ‘!’ is for description and should not be in the actual file. DATA: TEMPERATURE DEPENDENT HEAT FLUX UNITS LENGTH meters UNITS TEMPERATURE R UNITS TIME SECONDS UNITS DATA W/cm2 TEMPERATURES 2 300.000000 1000.000000 NODE 1 0. 0. 0. NODE 2 0. 1. 0. NODE 3 1. 0. 0. NODE 4 1. 1. 0. Thermal Models

4-95

NODE 5 2. 0. 0. NODE 6 2. 1. 0. TRI 1 1 2 3 TRI 2 3 2 4 TRI 3 3 4 5 TRI 4 5 4 6 TIME 87.000000 1.01

! Flux for node 1 at T = 300

1.02


2.01


2.02 3.01 3.02 4.01 4.02 5.01 5.02 6.01 6.02 TIME 90.000000 11.01 11.02 12.01 12.02 13.01 13.02 14.01 14.02 15.01 15.02 16.01 16.02

4-96

Thermal Models

TEMPERATURE BCM Sample Input DATA: ELEMENT BASED TEMPERATURE UNITS LENGTH meters UNITS TIME SECONDS UNITS TEMPERATURE C NODE 1 0. 0. 0. NODE 2 1. 0. 0. NODE 3 2. 0. 0. NODE 4 3. 0. 0. NODE 5 0. 1. 0. NODE 6 1. 1. 0. NODE 7 2. 1. 0. NODE 8 3. 1. 0. NODE 9 0. 2. 0. NODE 10 1. 2. 0. NODE 11 2. 2. 0. NODE 12 3. 2. 0. QUAD 1 1 2 6 5 QUAD 2 2 3 7 6 QUAD 3 3 4 8 7 QUAD 4 5 6 10 9 QUAD 5 6 7 11 10 QUAD 6 7 8 12 11 TIME 0.000000 1.

! Temp Element 1

2.

! Temp Element 2

3.

! Temp Element 3

11.

! Temp Element 4

12.

! Temp Element 5

13.

! Temp Element 6

TIME 1.0

Thermal Models

4-97

101. 102. 103. 111. 112. 113. COND_PER_AREA_TEMP_PAIR BCM Sample Input DATA: ELEMENT BASED COND_PER_AREA_TEMP_PAIR UNITS LENGTH meters UNITS TIME SECONDS UNITS TEMPERATURE C UNITS COND_PER_AREA_TEMP_PAIR W/m^2/K NODE 1 0. 0. 0. NODE 2 1. 0. 0. NODE 3 2. 0. 0. NODE 4 3. 0. 0. NODE 5 0. 1. 0. NODE 6 1. 1. 0. NODE 7 2. 1. 0. NODE 8 3. 1. 0. NODE 9 0. 2. 0. NODE 10 1. 2. 0. NODE 11 2. 2. 0. NODE 12 3. 2. 0. QUAD 1 1 2 6 5 QUAD 2 2 3 7 6 QUAD 3 3 4 8 7 QUAD 4 5 6 10 9 QUAD 5 6 7 11 10 QUAD 6 7 8 12 11 TIME 0.000000

4-98

Thermal Models

1.

! Cond Per Area Value Node 1

2.


3.


11. 12. 13. 501.

! Temperature Value Node 1

502.


503.


511. 512. 513. TIME 1.0 101. 102. 103. 111. 112. 113. 601. 602. 603. 611. 612. 613. 4.13.3

Miscellaneous Input Commands

BINARY There may be instances where the wants to put all the time dependent data in a binary file. This can be done by simply putting the word BINARY or inputting BINARY extenstion.bin. Here are a couple of examples: BINARY BINARY mydata.bin

Thermal Models

4-99

If the boundary condition mapper input file name is fred.inp, the first case with no extension would look for the binary file name to be 'fredCfdResults.bin' and for the second case, the binary file name would be fredmydata.bin. POSTLOGIC/PRELOGIC The POSTLOGIC and PRELOGIC commands can be used to add logic before and after any BCM output is written. Note that you can have more than 1 of these commands and they are output cumulatively. A sample would be: POSTLOGIC

WRITE( NOUT, * ) 'BCM POST 1'

POSTLOGIC

WRITE( NOUT, * ) 'BCM POST 2'

ADD_MULT stest The ADD_MULT command allows the to multiply the calculated resultant data by an input string that is added to the SINDA output. An example would be: ADD_MULT

stest + 4.

Which would result in the output for a QVALUE to be: MAIN.Q1000 = MAIN.Q1000 + T__TD * (stest + 4.)

4.14

Articulators

Articulators may be used to group sets of surfaces into an assembly or to define motion of a set of surfaces. There are two types of Articulators: • assemblies - 6-degree-of-freedom control for translation and rotation • trackers - 3-degree-of-freedom control for rotation Articulators may be viewed in the Model Browser (Section 2.4 "Model Browser") by choosing List > Assemblies/Trackers from the Model Browser menu. In the TrackerAssembly Tree of the Model Browser, drag-and-drop methods may be used to add or remove objects from trackers and assemblies. Objects, including other assemblies and trackers, can be attached to and detached from an articulator using the Thermal > Articulators > Attach Geometry and Thermal > Articulators > Detach Geometry commands. When executed, the will by prompted to select the articulator and then to select the objects to be attached/detached to that articulator. While articulators can be nested (infinitely deep), surfaces, assemblies and trackers can only be attached to one articulator. Therefore, if a surface will be controlled by one of two articulators, then the one articulator should be nested inside the other, the surface attached to the nested articulator and only one articulator should be updated. For assemblies this means only changing the transformations of one of the assemblies and for trackers this means locking one of the trackers (see Section 4.14.1 and Section 4.14.2, respectively).

4-100

Thermal Models

Note: Finite elements are defined by the location of the nodes defining the element, so the nodes must be in the assembly in order for the element to move with the assembly. If a finite element is selected during the Attach Geometry command, the nodes will automatically be put in the assembly. Note: When using TDMesh, the Mesh Controller and part should be attached to the articulator. TDMesh, the Mesh Controller and part are described in the Advanced Modeling Techniques s Guide accessed by Windows Start > Programs > Thermal Desktop > s Manual - Meshing. The Thermal > Articulators > Highlight Geometry command will highlight the geometry associated with the selected articulator. Finally, the Articulators > Detach All command will remove all surfaces associated with an articulator. The may rotate the articulator by selecting it and manipulating the grip points as shown in Figure 4-63. The may also use the 3DROTATE command to manipulate the axes. The may move the entire articulator by selecting the origin. Please note that the may have trouble rotating if the UCS is in the same plane as the selected rotation point and the axis of rotation is also in that plane. If the UCS is aligned with the X and Y axes shown in Figure 4-63, the would not be able to rotate about the Y axis by selecting on the midpoint of the X axis. The would have to select the end point on the Z axis to rotate about Y. This problem occurs because the amount of rotation is calculated by projecting the selected point into the UCS plane. about Y about X about Z

about X

about Z

about Y

move to point Figure 4-63

Assembly Edit Grip Points

Want "Hands-On" Information? Complete Section 21.3 "Importing a TRASYS Model and Using Articulators" to gain some experience working with Thermal Desktop articulators. 4.14.1

Assemblies

An assembly is an articulator that can be freely translated and rotated. A collection of any Thermal Desktop or FloCAD objects including, but not limited to, nodes, surfaces, pipes, articulators or other assemblies can be attached to an assembly and translated or rotated with the assembly. The assembly coordinate system is displayed in the graphics area on the screen. When the assembly is modified, via a rotate or a move, the location of the objects attached to that assembly will also be modified. Thermal Models

4-101

An assembly is created using the Thermal > Articulators > Create Assembly command. The command will prompt the to input a location for the assembly. Once the location has been selected, the Edit Assembly dialog box shown in Figure 4-64 will be displayed. The Edit Assembly dialog box Assembly tab allows input of the assembly name and the size of the axis placed in the graphics area. The capability to make the assembly inactive is also on the dialog box. An inactive assembly does not modify the associated geometry when the assembly is repositioned. In the case of nested assemblies, an inactive assembly will break the chain of dependence. An assembly may be edited by selecting it in the graphics area and then selecting Thermal > Edit. The Edit Assembly dialog box Trans/Rot tab is used to move the location of the objects attached to the assembly along the X, Y or Z axes of the assembly. Both translation and rotation inputs can be defined as symbols to parameterize a model. Please see “Symbol Manager” on page 11-1 for more information. Want "Hands-On" Information? Create an assembly in tutorial exercise "Dynamic SINDA Example" on page 20-201.

Figure 4-64

4.14.2

Edit Assembly Dialog Box

Trackers

A tracker is an articulator that has a defined axis of rotation that can be used to make surfaces or a collection of surfaces track the planet, sun, or a star, while the model is orbiting a planet with sun or planet orientation. A good example of the use of a tracker is for a solar tracking the sun. A tracker is created using the Thermal > Articulators > Create Tracker command. The command will prompt the to input the location of the tracker. Once the location has been determined, the Single Axis Tracker dialog box shown in Figure 4-65 will be displayed. The tracker may be planet, sun, or star tracking. If it is star tracking,

4-102

Thermal Models

inputs must also be made for the right ascension and declination of the star. The tracker may also be locked at a certain input angle. A tracker can be functional for the entire orbit, in the sun or the shade, or between specified orbit angles. Trackers are only active during heat rate calculations and articulating radk calculations. For fixed radk calculations, radks are reset to their original orientations. to calculate radks for a locked tracker, the should create an orbit with a single position and run an articulating radk calculation. The tracker can also be disabled. Disabling a tracker breaks the dependence of any attached geometry, tracker or assembly from that tracker. The tracker may be disabled in one of four ways: by selecting the Disabled radio button; by setting the Working mode Program to a value of 3; by toggling the global activation of articulators as described in Section 4.14.3 "Toggle Global Activation"; or by deactivating the tracker as described in Section 10.1.7 "Disabling Specific Trackers". Important: Disabling trackers should only be used to make corrections to trackers and only after the tracker has been reset. The best method for this is to Reset Trackers and the Toggle Global Activation. Once the corrections have been made, then Toggle Global Activation again. Range limits may be placed on the tracker in the dialog box of degrees. The Track Mode and Working Mode methods can be programmed with an expression by selecting the appropriate “Program” button (see Section 2.10.9 on page 2-47). The graphical size of the tracker may also be input. Trackers may be nested to simulate multiple axes of control.

Figure 4-65

Single Axis Tracker Dialog Box

Thermal Models

4-103

The graphical representation of a tracker is shown in Figure 4-66. The orange arrow is the tracking point and it will rotate about the green circle. The arrow color specifies the type of tracker. The arrow colors are as follows: • Yellow: Sun • Blue: Nadir (Planet) • White: Star • Orange: if the tracker is locked.

Figure 4-66

Tracker Representation

The colors of the tracker will be dimmed if the tracker is disabled. The coordinate system is placed at the center of the tracker such that the X and Y axes will align with the current UCS. The graphical display of a tracker may be reset using the Thermal > Articulators > Reset Tracker command. This command is useful after calculations have been made or after the model has been displayed on the orbit. Trackers should be reset before any changes are made such as detaching geometry. The may need to modify the orientation of the tracker to locate it relative to the geometry which will be attached to it. The tracker should be oriented such that the arrow is normal to the tracking surface and the tracker circle is in the plane of rotation. This can be accomplished by selecting the tracker and manipulating the grip points as shown in Figure 4-63. If the tracker has an assembly or another tracker attached to it, and the coordinates lie at the same origin, the must be careful to know which coordinates they are actually manipulating. It is recommended that the size parameter is used so that the one coordinate system is larger than the other. The may easily select the tracker coordinate system by selecting on the green limits circle, while an assembly may be selected by picking any point near the end of the each axis. A tracker may be edited by selecting it in the graphics area and then selecting the Edit Thermal Desktop Object icon or the Thermal > Edit command. Want "Hands-On" Information? Create a tracker in tutorial exercise "Importing a TRASYS Model and Using Articulators" on page 21-35.

4-104

Thermal Models

4.14.3

Toggle Global Activation

The Thermal > Articulators > Toggle Global Activation command will globally turn off or on the use of articulators. This command is especially useful if the wants to move an articulator, but not move the objects that are attached to it. This functionality should be used when the decides to scale or move a model that has articulators. The global activation setting is NOT saved on the DWG file. Want "Hands-On" Information? Use the Toggle Global Activation command in Section 21.3 "Importing a TRASYS Model and Using Articulators".

4.15

Grip Manipulators

A grip manipulator is a type of articulator that allows positioning of grip points for Thermal Desktop primitives. Grip manipulators are created manually or within TD Direct. If a grip manipulator is created through TD Direct, the will connect the grip points to the grip manipulator and then the grip manipulator will be updated by TD Direct. If a grip manipulator is created manually, the will make all modifications to the manipulator. The connected grip points are always placed at the center of the grip manipulator. Therefore the grip manipulator can be moved or translations defined to reposition grip points. A grip manipulator can control multiple grip points, but only one key point grip per surface can be controlled by any one grip manipulator. Grip manipulators will not change object parameters that are defined by symbols or symbol expressions. Grip manipulators are represented by a cube and, optionally, a coordinate system. The grip manipulator’s coordinate system can be translated and rotated by selecting the grip manipulator and moving the grip points. A grip manipulator’s grip points cannot be connected to another grip manipulator. 4.15.1

Create Manipulator • Command: RcGripManipulator • Ribbon: Thermal > Grips > Create Manipulator • Icon:

When this command is issued, the receives the following prompt: >Enter origin of grip manipulator

When a point is entered, the Edit Grip Manipulator dialog is opened.(Figure 4-67)

Thermal Models

4-105

Figure 4-67

Edit Grip Manipulator dialog

Assembly Tab Name. A name given to the grip manipulator for identification Size. The dimension of the cube that represents the grip manipulator. Graphically Display Name. When checked, the name will be displayed in the graphics area with a leader to the grip manipulator. Display displacement vector and base coordinate system. When checked, a coordinate system will be displayed at the origin of the grip manipulator. If translations are applied on the Translation tab, then a line connecting the base coordinate system and the center of the grip manipulator will also be displayed. If unchecked, only the grip manipulator cube will be displayed. Active. When checked, the connected grip points will be adjusted based the position of the grip manipulator. When unchecked, the grip manipulator can moved without adjusting the connected grip points.

Translation Tab The translation fields can be defined using symbols and symbol expressions. Translation X. The translation of the grip manipulator from the grip manipulator origin in the X direction of the grip manipulator’s coordinate system.

4-106

Thermal Models

Translation Y. The translation of the grip manipulator from the grip manipulator origin in the Y direction of the grip manipulator’s coordinate system. Translation Z. The translation of the grip manipulator from the grip manipulator origin in the Z direction of the grip manipulator’s coordinate system.

Grip Selection Tab The grip selection tab lists the connected objects (under Name) and the selected grip points (under Grip) in the table. See Section 4.15.2 for connecting grip points to the manipulator. Change Grip. Selecting a grip in the table and then selecting the Change Grip button opens the Change Grip dialog that allows changing the grip point type (Select Mode) and the grip point (Select Grip). Remove. Selecting a grip in the table and then selecting the Remove button removes the selected grip from the table. 4.15.2

Connect • Command: RcConnectGrips • Ribbon: Thermal > Grips > Connect • Icon:

When this command is issued, the receives the following prompts: >: Select an entity: >: Select a grip manipulator:

An entity is any Thermal Desktop surface or solid primitive. A grip manipulator is any grip manipulator created manually or by TD Direct. Only one entity and one grip manipulator can be selected at a time. When an entity and grip manipulator have been selected, the Change Grip dialog opens with two drop-down lists. Select Mode. This list consists of Key Point and Parameter, the grip point types that can be controlled by grip manipulators. See Section 2.10.3 for more information about grip point types. Select Grip. This list contains the grip points of the selected type in Select Mode that are available for the selected object. When OK is selected, the is prompted again to continue connecting grip points to grip manipulators. Important: To end the selection cycle, the must select <Esc>. <Enter> or right-click will not exit out of the command.

Thermal Models

4-107

4.15.3

Update • Command: RcUpdateGripManipulators • Ribbon: Thermal > Grips > Update • Icon:

This command updates all grip manipulators in case a change was not captured.

4.16 4.16.1

Network Functions Merge Coincident Nodes Note: When thermal connection must be modeled between surfaces or solids, the can use either: merge nodes or ors (see “ors” on page 4-74). For finite conductances, consider using ors instead. For very large conductances or perfect (infinite conductance), use merge nodes, if possible.

The Thermal > FD/FEM Network > Merge Coincident Nodes command will automatically merge the nodes that are within the input tolerance of each other. This is typically done to thermally connect two surfaces or solids into one contiguous object. When the command is issued, the will be prompted to select the nodes to be considered for merging. A dialog will open to allow the to input the tolerance and then choose which node will be kept at each location. The options for the nodes to keep are: First Selected, Smallest Node ID, Largest Node ID, Lesser Submodel Name, and Greater Submodel Name. The Lesser and Greater Submodel Name options refer to the alphabetical order: A is less than Z. The program will then highlight the areas of coincident nodes with a red cross, if the nodes are in the same submodel, or a yellow cross, if in different submodels, and prompt the to accept or cancel the merge. To clear the crosses, use View > Regen or the Reset Thermal Desktop Graphics icon. Nodes on the same surface, solid, or element are prevented from being merged. When merging nodes to two surfaces or solids, the is recommended to use edge nodes and align nodes as best as possible before merging nodes. Nodes can be merged between surfaces using Recession (Section 3.2.7) only if the same heat load and insulation definition (material, thickness, number of node layers, etc) are applied to all surfaces in the merge operation. When surfaces are edited or moved, tests are internally made to determine if the merge has been broken. These internal tests use the tolerance that was input during the merge command. The can also force merged nodes to be unmerged by using the Thermal > FD/FEM Network > Unmerge Coincident Nodes command. This command only works on nodes that were originally merged using the Merge Coincident Nodes command. 4-108

Thermal Models

Important: When nodes are unmerged, the node will return to the number of nodes merged, however, the node numbers remain the same. The thermal solution will appear as though the nodes are still merged unless the resequences the node IDs (see “Resequence IDs” on page 7-2) Want "Hands-On" Information? Section 20.7 "Conduction and Radiation Using Finite Elements" includes this functionality as a part of the exercise. 4.16.2

Surface Coat Free Solid FEM Faces

The command Thermal > FD/FEM Network > Surface Coat Free Solid Faces will automatically place planar elements on the free faces of the selected solids. Consider the example shown in Figure 4-68. Three brick elements are lined from end to end. Suppose the wants to for radiation off the solid elements. To do this, the could place planer element on the faces of the solids that will be radiating. The could use the Create Element command and place a planar element by selecting nodes 1,2,3,and 4 to cover the end cap. Another element would have to be created using nodes 2, 6, 7, and 3. It can quickly be seen how tedious this could be on the . Enter the Surface Coat Free Solid command. This command builds a list of all the nodes that make up a solid face in the model. The algorithm determines that face 1,2,3,4 is only used by one solid, so it knows to surface coat that face. The algorithm also determines that face 5,6,7,8 is shared by two solids, therefore it knows not to surface coat that face. 8 4

1 2 Figure 4-68

15

11

7 3

16

12

13 9

5

14 6

10

Surface Coating Example

The planar elements are created using the same nodes as the solid. The planar elements are created so that the top side is always out. The planar elements are placed in the current default analysis group with the top side active. The thickness is set to zero. Once the planar elements are created, the multiple edit dialog box will be displayed to let the specify the optical properties. The logic in the algorithm not only checks for solids that share faces, but will also check any planar elements or conics (rectangles, disk, etc.) that already reside on the face of a solid. Thus, the algorithm will not create two planar elements that lie on top of each other.

Thermal Models

4-109

Want "Hands-On" Information? Gain experience by completing tutorial exercise "Beer Can Example" on page 20-89. 4.16.3

Synchronize Element Normals

Consider a group of planar elements that are all connected. The wants to have them radiate all from the same side, so the active side is set to top. The display active preferences command is executed, and all but a few of the elements are active in the correct direction. The could manual pick the offending elements, and reverse their connectivity to make the active side correct. This could quickly become quite tedious, thus the Thermal > FD/ FEM Network > Synchronize Element Normals command. This command allows the to select the master element, followed by all of the elements that connect to it, and the program will make all of the element normals point the same direction. 4.16.4

Convert AutoCAD Surface to Nodes/Elements Note: The may wish to investigate TDMesh as an alternative method to that described in this section. TDMesh usage and guidelines are provided in the Advanced Modeling Techniques s Guide accessed by Windows Start > Programs > Thermal Desktop > s Manual - Meshing.

The Thermal > FD/FEM Network > Convert AutoCAD Surface to Nodes/Elements command allows the to convert an AutoCAD surface or mesh to Thermal Desktop nodes and finite elements. The should investigate the use of the AutoCAD rulesurf, and edgesurf commands combined with the SURFTAB1 and SURFTAB2 parameters. Meshes created by this command use finite element conductance calculations and cannot be used with the Toggle FD Mesh Nodalization command described in Section 7.8. 4.16.5

Extrude/Revolve Planar Elements Into Solids Note: This feature allows the to use exiting planar elements or surfaces to create solid finite elements. If an AutoCAD surface is available or can be created, the may find that TD Mesher will be more versatile and powerful. See the Advanced Modeling Guide in a separate document that came with the installation.

The Thermal > FD/FEM Network > Extrude/Revolve Planar Elements into Solids commands allow the generation of planar elements into solids. The is first prompted to select the solid elements. If the extrude command is executed, the is prompted to select the point to extrude from, followed by the point to extrude to. These points define the direction of extrusion and also the length, but the length can be changed on the Extrude/ Revolve Planar Elements into Solids dialog box. If the revolve command is executed, the selects two points to define the axis of rotation, and the angle of rotation is input on the dialog box. The Extrude/Revolve Planar Elements into Solids dialog box is shown in Figure 4-69. The may select how many nodes to create along the path and may also

4-110

Thermal Models

select to delete the planar elements used to create the solids.

Figure 4-69

Extrude/Revolve Planar Elements into Solids Dialog Box

The may also select to have uneven breakdowns which allows for varying thicknesses and material properties (select the Uneven Breakdowns radio button). If a thickness of zero is input, then the program will input a planar element for that layer. Finally, if the has variable thicknesses for a material type, they may link the node numbers to a file, and input the variable material name. This option is powerful for doing Thermal Protection System (TPS) modeling for reentry vehicles. When the command is completed, the Edit Solid Node dialog box is displayed to allow the to select the material of the generated solids. The Thermal > FD/FEM Network > Extrude Normal to Planar Elements into Solids will extrude the elements normal to the surface being extruded. Want "Hands-On" Information? Tutorial exercises "Beer Can Example" on page 20-89 and "Conduction and Radiation Using Finite Elements" on page 20-129 both use this functionality as part of the exercises. Try the exercises to see how it works when applied. 4.16.6

Map Solid Mesh Between Conics

The Thermal > FD/FEM Network > Map Solid Mesh between Conics command (the dialog box is shown in Figure 4-70) allows the to generate solid elements between conic meshes. For example, the may have a rectangle broken up 10x10 that lies in the XY plane. A second rectangle, with the same nodal breakdown 10x10 might lie parallel to the XY plane at Z equal to 1. The Map Solid Mesh command will generate 100 solid brick

Thermal Models

4-111

elements between the two rectangles. After the command is completed, the edit solid element dialog box will be displayed to allow the to select the material of the new elements. The surfaces selected for meshing must have the same number of nodes, and will most likely be using edge nodes so that the volume fully spans the space between the two surfaces. An interesting fact of this generation is to consider the mapped meshing between two conic surfaces, such as two paraboloids or a disc and a sphere (for a lens). The conics are used for the radiation of the solids, and thus retain full radiation capabilities for precise specular reflection and refraction. The effects on radiation heat transfer may be significantly different than using the faceted mesh formed by a solid.

Figure 4-70

4.16.7

Map Mesh between Conics Dialog Box

Solid Interior Faces

The FD/FEM Network > Show/Hide Solid Interior Faces command makes a calculation to determine if the faces of a solid are connected to other solids or to shell elements. If they are, then in shaded mode, they cannot be seen, and therefore only slow down the graphics. In wireframe mode with a large model, drawing the interior faces can make the graphics seem cluttered, therefore the hide interior faces function can make the wireframe model more viewable. It is important to note that the calculation is only made when the Hide Solid Interior Faces command is selected. If the adds or deletes elements, the Hide Solid Interior Faces command must be selected again or the model may appear incorrect. For example, consider a shell element being deleted off a solid face while in the Hide Solid Interior Faces mode. If the then goes into solid shaded mode, the solid will then appear to be open on that face and the will actually see the inside of the solid. Want "Hands-On" Information? Check out Section 20.7 "Conduction and Radiation Using Finite Elements" to see how this function works as a part of an exercise.

4-112

Thermal Models

5 Fluid Models FloCAD® is the module of Thermal Desktop that allows s to develop FLUINT (fluid integrator) submodels. The mechanics of building fluid submodels is very similar to that of building thermal submodels, with many commands operating on components of both types of submodels. In fact, the models are built, executed, and postprocessed simultaneously by default. A SINDA/FLUINT model may be constructed with thermal submodels, fluid submodels, or both. However, no single submodel may be created using components of both thermal and fluid submodels. Want "Hands-On" Information? For examples of working with fluid models refer to “FloCAD® Tutorials” on page 22-1. This tutorial exercises provides the with a variety of exercises focusing on the FloCAD application.

5.1

Overview

The commands to create fluid models are centralized under Thermal>Fluid Modeling. The menu choices are shown in Figure 5-1. Lumps, paths, ties, ifaces and fties are the main objects used to build fluid models, with subcategories of each of these network elements available for specific modeling purposes. These fluid network elements are automatically numbered by the program during generation of the each element. Once created, these IDs may also be changed via the edit forms. Complete submodels or any selection set can be resequenced as well. Internal checks insure that duplicate IDs cannot be entered by the . The submodel properties can be set to use any increment and starting ID value. The Thermal >Modeling Tools > Resequence Fluid IDs command can also be used to reset the network identifiers.

5.2 5.2.1

Fluid Submodels About Fluid Submodels

Fluid submodels are subdivisions of the fluid model. Unlike thermal submodels, fluid submodels cannot be subdivided arbitrarily - use one fluid submodel for each subsystem (fluid loop, air flow, etc.) containing a single working fluid (whether a single substance of a mixture) as described in Section 5.2.3. If a submodel is excluded from the solution (Section 15.2.4.1), the lumps, paths, ties, logic, and anything else included in that submodel are unavailable for that solution.

Fluid Models

5-1

Figure 5-1

5.2.2

Fluid Modeling Menu Choices

Creating Fluid Submodels

Selecting the Thermal>Fluid Modeling > Submodel Manager command displays the FLUINT Submodel Manager Form dialog box that allows fluid submodel names to be defined (see Figure 5-2). These names will then be available on pulldown lists when using the Lump Edit and other dialog boxes, as an example. Preferentially, submodel names should be defined before they are used to name fluid lumps and other network elements. However, if s manually type a new submodel name into a field while editing a FloCAD object, a dialog box will appear asking if the would like to add the submodel to the submodel manager. Submodel names may be added at any time during the creation of a model. To add a new submodel name, select the Add button and enter the name (see Figure 52). To remove all unused submodel names from the drawing database, select the Purge button.

5-2

Fluid Models

Figure 5-2

FLUINT Submodel Manager Form Dialog Box

To set a submodel as the default fluid submodel, select the name in the list box and then press the Set Default button. The default submodel will be used for all newly created fluid lumps. A submodel may be renamed by selecting the Rename button and then typing in the new name. This function will update all the lumps, paths, and ties that used the previous submodel name with the new name. The Properties button opens a dialog box where the type of fluid for the submodel is set, and the network IDs number and increment to be used can be defined. The Scan DB button will check the current drawing for all submodel names that are being used and add them to the list. This can be necessary if xrefs are used or object have been inserted into the drawing from the another drawing. Want "Hands-On" Information? Gain experience working with fluid submodels by completing tutorial exercises “Air Flow Through an Enclosure” on page 22-3 and “Manifolded Coldplate” on page 22-37. 5.2.3

Fluid Selection

Each fluid submodel can utilize one or more substances as working fluids. Selecting the Properties button in the FLUINT Submodel Manager Form dialog box brings up the Fluid Submodel Properties dialog box shown in Figure 5-3, in which the fluids can be selected or changed. The default fluid is air (as a perfect gas). The letter listed in parentheses is called the fluid constituent identifier which can be used within SINDA/FLUINT logic blocks and expressions to reference a specific fluid within a submodel. For example, the might elect to use the identifier ‘A’ for air and ‘W’ for water. The letter identifier is also is used in output tables to refer to each individual fluid species. To change from the default, select the Edit button. This selection will bring up the Fluid Edit dialog box shown in Figure 5-4. FloCAD contains a built-in library of fluids.1 These “library” fluids can all be listed by clicking on the arrow next to the Library field to display a drop-down list. (See Figure 5-5.) Fluid Models

5-3

Figure 5-3

Fluid Submodel Properties Dialog Box

Figure 5-4

Fluid Edit Dialog Box

If the has fluid property descriptions (FPROP DATA Block, as described in the SINDA/FLUINT ’s Manual) for additional fluid species,2 they can be utilized by selecting Browse For New Fluid Property File in the pulldown library menu. An attempt is made to read the file to locate the fluid constituent identifier and place it into the dialog box automatically. This integer identifier can also be changed by the , but should correspond to the number within the file (within the HEADER FPROP DATA block). It is also possible to keep all FPROP DATA files in a one or more centralized directories and only refer to them by the file name without the fully qualified path name. In that case, a file named “paths.txt” must exist in either the local directory, or in the /bin directory of the SINDA/FLUINT install path, or both. Within the “paths.txt” file a fully qualified path name to the directory where fluid properties are stored can be named within this file. Up to 10 paths will be searched from the “paths.txt” file. Refer to the SINDA/FLUINT manual

1 The built-in description of two-phase water is highly approximate. The REFPROP-based version, which can be fetched from www.crtech.com, is highly recommended as an alternative. 2 Many are available from www.crtech.com, or can be requested at that site as well.

5-4

Fluid Models

Figure 5-5

Fluid Edit Dialog Box Pulldown Library Menu

for more details. This feature allows s to easily share models without having to change the path names associated with fluid property files if they are located in the same directory by different s. It also encourages the maintenance of a single set of updated property files. The Add button in Figure 5-3 can be used to add one or more fluid species to a submodel using the same Fluid Edit dialog box as when selecting the Edit button. The number of fluids in a model should be kept to a minimum to contain computational cost and modeling difficulty. Only one condensable/volatile species is allowed per submodel. Other restrictions and fluid mixture modeling options (e.g., dissolution/evolution, reactions and combustion, etc.) can be found in the SINDA/FLUINT ’s Manual. 5.2.4

Fluid Network IDs

Lump, path, tie, iface, and ftie IDs are all automatically generated by FloCAD. The next ID to be created is listed in the Fluid Submodel Properties dialog box by selecting the Network ID’s tab as shown in Figure 5-6. Once an entity is created and an ID used, the increment determines the next ID number to be generated. These fields can be edited by the . The input values are checked against IDs currently in use to prevent duplicate IDs from being generated. If a conflict arises during generation of a new ID, the next available ID will be found by using an increment of one.

Fluid Models

5-5

Figure 5-6

5.3

Fluid Submodel Properties Dialog Box Network IDs Tab

Fluid Network Objects

Fluid submodels consist of lumps, paths, ties, ifaces and fties. Lumps are analogous to thermal nodes. A path is used to connect two lumps and transfer mass. Ties provide a thermal connection between fluid lumps and thermal nodes: ties typically model the convection between a fluid and the container or adjacent structure. Ifaces connect tanks and only tanks: since all tanks are junctions within a steady state, ifaces are not active in steady state runs. Ifaces describe how a shared wall between two tanks behaves. It tells the program that two adjacent control volumes (tanks) share a common boundary, and how their pressures and volumes are interrelated. Fties establish heat transfer directly between two lumps in the same submodel for either slow-moving flows or highly conductive fluids: situations where conduction cannot be neglected compared to advection. Pipes (Section 5.4) are higher level objects similar to SINDA/FLUINT macros, though they are really supersets of macros since they offer many more modeling options; pipes are a major feature of FloCAD. They contain lumps, paths, ties and nodes depending on the type of pipe selected. This section describes the fluid network elements, while pipes are described in Section 5.4. Path Rotation Axis objects can be created and associated with one or more paths to provide spin data to SINDA/FLUINT. These are described in Section 5.3.5. 5.3.1

Lumps

Fluid lumps may be created using the Fluid Modeling > Lump command. When invoked, the is prompted to select a location for the lump. The characteristics of a lump may be edited by selecting the lump and then choosing Thermal > Edit. This function will 5-6

Fluid Models

display the Lump Edit Form dialog box shown in Figure 5-7.

Figure 5-7

Lump Edit Form Dialog Box

Lumps may be specified to be tanks, junctions, plena or clones. The first three types of lumps are analogous to diffusion nodes, arithmetic nodes, and boundary nodes, respectively. Table 5-1 Lump Types

Lump Type

Description

Tank

A tank is a control volume with a finite volume. Tanks store and release energy and mass with time.

Junction

A junctions has zero volume and reacts instantaneously.

Fluid Models

Graphical image (wire frame and shaded)

5-7

Table 5-1 Lump Types

Lump Type Plenum

Description


A plenum may be considered to have infinite volume and hence represents an infinite source or sink of fluid at a constant temperature and pressure.

Note: While plena are infinite volume, this does not necessarily mean that the fluid in a plenum is at zero velocity: pressures and temperatures may be static or stagnant. Clone lumps are intended to provide an alternate visualization point to an existing lump in a model. By selecting the parent lump, the clone lump can be used to connect other graphical entities in a drawing as if it were the parent lump. The properties of a clone lump can not be edited. A clone lump without a parent selected will not be assigned to a particular submodel, nor will it be output to SINDA/FLUINT. The fluid in tanks and junctions by default are assumed to be moving at the velocity at which they exit: static temperature and pressure is the default. To force the fluid to be accelerated as it exits, the check box Tanks/Junctions at Zero Velocity can be checked. This turns on the SINDA/FLUINT option LSTAT=STAG, which signifies that the fluid within the tank or junction is stagnant or at least slow. Conversely, plena are assumed to be at zero velocity (LSTAT=STAG) by default. To force them to be moving (static states), check the box Plenum flowing. This designation affects the kinetic energy and choking calculations associated with fluid leaving this lump. Appropriate paths (Stubes, losses, etc.) flowing out of a stagnant lump also automatically add an acceleration loss term (equivalent to a K-factor of one). The initial thermodynamic states of lumps should be specified. Care should be used to ensure that temperatures and pressures are within the range of valid fluid properties for that submodel. If they are out of range, they will be reset at the start of SINDA/FLUINT execution. The thermodynamic state is defined by the temperature, pressure and quality. By default, the quality and temperature are used and the pressure will be overridden if it conflicts thermodynamically with temperature and quality. If the check the Pressure Priority box, then the pressure and quality are used and the temperature is overridden if it conflicts with pressure and quality. See Thermodynamics States: FLOW DATA in the SINDA/FLUINT ’s Manual for more information. A heat load can be applied directly to a lump using the Heatload field unless the lump is a plenum. The Basic Data tab has a button near the bottom that reads either Add Code or Edit Code. This button accesses the Network Element Logic as described in Section 2.10.10. 5-8

Fluid Models

The Twinned Lump Data tab allows specification of twinned lumps for modeling phasic non-equilibrium within a tank. Twinned tanks is only necessary for very specific applications and should be used judiciously. Refer to Section 3.25 of the SINDA/FLUINT ’s Manual, Version 5.1. The Additional Input tab can be used to access many of the other advanced features of SINDA/FLUINT. Multiple input lines can be used here with the contents being written directly to the SINDA/FLUINT input file after the lump information that is generated from the information supplied in the remainder of the Lump Edit dialog box. Comments can be added in these boxes by using a ‘$’ before the comment. An example would be using the additional input box to specify the compliance as ‘COMP=1.0E-5,’ or to specify inputs that contain references to SINDA/FLUINT “processor variables,” which cannot be resolved by the Thermal Desktop expression manager since their value is only valid while SINDA/ FLUINT is executing. This Additional Input field can also be used to override inputs from other parts of the dialog box. Since all contents of this box are added at the end of the lump input in the SINDA/FLUINT input file, variables such as TL, XFA, and XGA can be overridden by adding them to the box. The Constituents button will not be greyed out for submodels for which multiple constituents have been defined (see “Fluid Selection” on page 5-3). Selecting this button will bring up a dialog box to set the initial mass fractions of each constituent for the vapor and liquid phases. Care should be taken to insure that reasonable amounts are chosen. Property errors can result if the mass fractions are not consistent with the temperature and pressure of the lump. A special case exists for a model that uses an air (from the Library) and water mixture. Instead of inputting mass fractions, the may input the relative humidity of the vapor space. Calculations are performed inside of FloCAD to compute the mass fractions based on the temperature and pressure of the lump. These inputs can be overridden by using the additional input field to set the mass fractions directly. The input fields for temperature, pressure, heat load, quality, and volume will accept symbolic expressions when the double clicks in the text field. This allows the to specify inputs as mathematical expressions using Thermal Desktop symbols and algebraic, logarithmic, or trigonometric functions (for more information see section 5.5.1). The Output Expressions button allows the to specify these expressions to be explicitly ed into the SINDA/FLUINT input deck for use in dynamic calculations.3 In the shaded mode, a junction is displayed as a sphere, a tank as a box, and a plenum as a tetrahedron. In the wireframe mode, a junction becomes a circle, a tank a square, and a plenum becomes a triangle. In order to distinguish the wireframe mode shapes from their thermal counterparts, tanks and junctions have an X drawn in the middle and plena have a Y. All three shapes have a vertical pole through the center. Want "Hands-On" Information? Create and edit lumps in the tutorial exercises “Air Flow Through an Enclosure” on page 22-3, “Manifolded Coldplate” on page 22-37, and “FEM Walled Pipe” on page 22-99. 3 Since FloCAD updates cannot be made in the dynamic mode, the use of Output Expressions is strongly advised as an alternative means. In other words, perform model updates in SINDA/FLUINT’s expression system directly, instead of in dynamic calls to Thermal Desktop from SINDA/FLUINT.

Fluid Models

5-9

5.3.2

Paths

There are several types of paths that can be created to model the fluid flowing between lumps. Under the Thermal > Fluid Modeling command submenu, the path types are listed below (with descriptions and images added here, but not in the program menu): Table 5-2 Path Types

Type

Description

Loss

A valve, fitting, or other K-factor based loss

(S)Tube

A duct or duct-like age

SetFlow

A constant mass or volumetric flow

Orifice

A sharp or long orifice (or vent, leak, etc.) or a control valve

Capil

A filter, wick or other capillary age

Pump/Fan

A fan, blower, centrifugal pump, etc.

5-10


Fluid Models

Table 5-2 Path Types

Type

Description

Tabular

A -defined gain or loss based on tables

Turbine

An axial or radial turbine or turbine stage

Compress

An axial or radial compressor or compressor stage

Comppd

A positive displacement compressor


Additional subcategories exist for some of these types, and these subcategories can be changed by editing the path after it has been created. After selecting a path type, the is then prompted to select the From lump and the To lump of the path. The order of selection determines the direction of positive flow for the path: the first lump selected becomes the upstream lump, and the second lump becomes the downstream lump. (After the path has been created, the positive flow direction can be reversed by selecting the Modeling Tools > Reverse Path Direction command. The Modeling Tools > Move Path End command allows s to change the either the upstream or downstream lump for an existing path.) Once the lumps have been selected during the initial generation of the path, the has the option of selecting an existing cross section shape upon which to base the path’s flow area and (if applicable) hydraulic diameter. Press <Enter> if no shape is desired (in which case the geometry can be input separately in the Path Edit dialog box). Otherwise, a number of different AutoCAD® and Thermal Desktop entities can be used as cross sections, including 3D polylines, polylines, arcs and cylinders as a few examples. AutoCAD lines used to define cross sections must be coplanar. An open shape (e.g., a square that is missing one

Fluid Models

5-11

side) will be assumed to represent the wetted perimeter for any hydraulic diameter calculation that is needed for ducts, and FloCAD will assume that a line drawn between the endpoints of such an open shape can be used to determine the flow area.4 The path may be edited by selecting the path followed by the Thermal > Edit command. Each path type has its own edit dialog box, customized to the input necessary for that path type. The paths are all displayed as a line between the two lumps. The icon for each type of path is displayed (in 2D) in the titles of the subsections that follow. Tubes have an arrowshaped cone in the center, Losses have a butterfly shape, Pump/fans have a sphere with a cylinder along the positive flow direction, Tabular losses have an arrow shaped cone with a disk at the pointed end, and Orifices have two disks in the center. The SetFlow path has two arrow shaped cones pointing in the direction of positive flow. The Capil has two spheres, with the upstream sphere shading part of the downstream sphere for positive flow. The Turbine has a truncated cone with the smaller diameter at the upstream side for positive flow. The Compress has three spheres, with the largest sphere at the upstream side for positive flow. The Comppd has three cylinders with the largest one at the upstream side for positive flow. Preferences contains an option on the Graphics Size tab to toggle the mode which path icons are displayed. The actual flow area is used to compute a hydraulic diameter, which is in turn used as a measure of the size of the cylinder drawn between the lump in any 3D display mode. If different up and downstream flow areas (AFI/AFJ) are input, these will be depicted as well. This allows the to easily detect where flow discontinuities exist in the model, which may cause problems with kinetic energy .5 There is also an icon to allow the to switch modes with the click of the mouse. All of the path edit forms are tabbed dialogs (Figure 5-8). The first tab is always pathtype specific information. The path type tab has a button near the top that reads either Add Code or Edit Code. This button accesses the Network Element Logic as described in Section 2.10.10. Each path type also has a Flow Area tab. The flow area tab varies by type to include features that are specific to each type (e.g., Throat Area, or Hydraulic Diameter). All path types can have duplication factors set on the Flow Area tab. The duplication factors act as multipliers on the flow seen by the lump either upstream (DUPI) or downstream DUPJ) end of the path. A duplication factor is normally 1.0 such that it appears as a single path connection. A duplication factor of zero will still compute the flow for a path, but the attached lump will not see any mass change as a result of this path. The duplication factors are also used to modify the area over which the flow acts for calculations involving kinetic energy, or computing continuous flow ages for macro initialization. All path types include the Suction Data and Additional Input tabs. The Suction Data tab allows specification of phase and/or species specific suction through the path. The Additional Input tab can be used to access advanced features that not available via forms or to override form data. 4 For extreme (far from circular) cross-sectional shapes, refer to the SINDA/FLUINT DEFF or “effective diameter” option in the SINDA/FLUINT ’s Manual. DEFF is an option for Tubes and Stubes that must be specified in the Additional Input field. 5 Otherwise, the initial outputs of the SINDA/FLUINT run will contain one-time warnings about discontinuities in flow areas between adjacent paths.

5-12

Fluid Models

Figure 5-8

Path Edit Form Dialog Box

Want "Hands-On" Information? Work with paths by completing the tutorial exercises “Air Flow Through an Enclosure” on page 22-3, and “Manifolded Coldplate” on page 22-37. 5.3.2.1

Loss

The Loss Edit dialog box is shown in Figure 5-9. The minimum inputs for a loss type components consists of the flow area and the loss coefficient (“K-factor”). The FK Calculator button provides access to advanced loss and to some libraries of loss factors corresponding to common flow components. These are described in Section 5.7. The Loss Type field allows the to specify additional component types that also only require the flow area and loss coefficient to be specified. These are variations of the same “K-factor theme” such as a loss that will close if the K-factor ever becomes negative (CTLVLV, where as otherwise a negative K-factor results in a pressure rise instead of a loss), and a simplified check valve (CHKVLV, which closes for reversed flow). When other inputs are required for these other loss-type elements, an otherwise greyed edit field is provided with a description of the input required.

Fluid Models

5-13

Figure 5-9

Path Loss Edit Form Dialog Box

The Additional Input field is also available to provide access to other advanced features such as choking options. As with the equivalent field in the Lump Edit Form dialog box, comments can be added in these text boxes by typing a ‘$’ before the comment. The Select Area button allows the to change or add the cross section shape to be used for flow area calculations. To remove a shape from a path (so that the flow area may be specified numerically), select the Select Area button and press <Enter>. The Display Shape button is greyed if no shape has been selected. Throat areas may be specified as needed for choked flow calculations, otherwise they default to the same as the flow area (namely, no contraction to the throat is assumed). The input fields Loss Coefficient, Flow Area, and Throat Area (as well as any other data specific to a particular loss type) will accept symbolic expressions when the double clicks in the text field. This allows the to specify inputs as mathematical expressions using Thermal Desktop symbols and algebraic, logarithmic, or trigonometric functions (for more information see section 5.5.1). The Output Expressions button allows the to specify these expressions to be explicitly ed into the SINDA/FLUINT input deck to perform calculation updates during the solution. 5.3.2.2

(S)Tubes

The (S)Tube Edit dialog box is shown in Figure 5-10. STubes (or Short Tubes) are time independent (instantaneous, inertialess) ducts. Tubes for inertia of the fluid in the path, and hence are functions of time during transients: it takes time to accelerate or decelerate the fluid within a tube. The length of the tubes can be either specified by the ,

5-14

Fluid Models

or computed by FloCAD based on the distance between the centers of the two lumps. If only the hydraulic diameter is input (and no cross section shape is selected, as seen by a greyed Display Shape button), then the flow path is assumed to be circular. The Loss Coefficient, FK Calculator and shape selection options are identical to the Loss dialog box (Section 5.3.2.1). Similar to losses, symbolic expression used in defining (S)Tubes can be ed explicitly to the SINDA/FLUINT input deck by checking the Output Expressions button.6

Figure 5-10

Tube/STube Edit Form Dialog Box

The Additional Input field is also available to provide access to other advanced features such as choking options or pressure drop correlation choices for two-phase flow. As with the equivalent box on the Lump Edit Form dialog box, comments can be added in these boxes by typing a ‘$’ before the comment. The default for tubes and Stubes is to carry liquid and vapor flows at equal velocities: a homogeneous assumption. By selecting the Twinned Path check box, there will be matched set of two flow paths generated for visible path. One such twinned path is primarily intended to carry liquid when available, while the other carries vapor. These are referred to as the primary twin and secondary twin, respectively. There will be a separate velocity for each of these paths when two-phase flow is flowing through them, enabling slip flow cal6 Since FloCAD updates cannot be made in the dynamic mode, the use of Output Expressions is strongly advised as an alternative means. In other words, perform model updates in SINDA/FLUINT’s expression system directly, instead of in dynamic calls to Thermal Desktop from SINDA/FLUINT.

Fluid Models

5-15

culations (unequal velocities) to be performed based on the default methods (see SINDA/ FLUINT ’s Manual for details and customizations). Each twin has a separate numeric ID identifying it for use in logic or output tabulations. There is a default offset between the primary and secondary twin ID that can be set via the Fluid Submodel Manager Properties Edit dialog box. When only a single phase fluid (whether vapor or liquid) is available for transport, the secondary twin will become inactive and only the flow rate for the primary twin will be computed. The Friction and Heat Transfer effects region of the form allows specifying wall roughness factor, laminar friction multiplier, turbulent friction multiplier and radius of curvature for bends and coiled tubes. After setting the factors, multipliers, and radius of curvature, the effects can be easily turn off using the buttons on the right. Tubes and STubes also are automatically checked for connections that would constitute a continuous age, and if this condition is detected, they are automatically turned into a FLUINT duct (LINE) macro. This feature can be turned off using the Aggregate into continuous age (Ducts) check box at the bottom of the form. To visualize which lines become macros, type rcmacrolist into the command line. Each macro will be assigned a unique color. 5.3.2.3

SetFlow

The SetFlow Edit dialog is shown in Figure 5-11. The SetFlow device is used to specify a fixed flow rate in either mass or volumetric units. It can represent either a pressure gain or loss, and is often used as a boundary condition.

Figure 5-11

5-16

Set Flow Edit Dialog Box

Fluid Models

It is very common to use a symbolic expression when defining the mass or volumetric flow rate, and this expression can be ed explicitly to the SINDA/FLUINT input deck by checking the Output Expressions button.7 This allows the flow rate to be changed during the numerical solution, perhaps parametrically or as a function of temperatures or pressures. The flow area of the device should be entered so that the solution engine can compute velocity for the kinetic energy in the path. A zero entry will result in the code assuming upstream flows have stagnated. The pressure drop across this type of device will be computed by the solution to match the system resistance it encounters. For pressure versus flow type devices, a Tabular or Pump/Fan object should be used. The Additional Input field is also available to provide access to other advanced features. Comments can be added in these boxes by typing a ‘$’ before the comment. 5.3.2.4

Orifice

The Orifice Edit dialog box is shown in Figure 5-12. The Orifice path type can be used to model a square or circular orifice flow restriction, and can be generalized to model a leak, partially closed valve, etc. The must supply both the upstream flow area (AF) and the hole area (AORI). The upstream and downstream flow areas are assumed to be equal: AFI and AFJ cannot be input for this path type. The calculated K-factor (for information purposes) uses upstream dynamic head as the basis of the calculation. Compressibility effects may be optionally included in the calculation. The default calculation is to use a built-in correlation for either sharp or long orifices, as documented in the SINDA/FLUINT ’s Manual. As an alternative, the may specify either the discharge coefficient Cd or the K-factor. In all three cases. the size of the vena contracta is estimated and applied as the throat area (AFTH) for choking calculations. The orifice length to diameter ratio (ELLD) field is utilized if it is greater than zero. The ELLD modifies the correlation as described in the SINDA/FLUINT manual. An ELLD of zero represents a sharp-edged orifice. The FK Calculator button provides access to advanced loss and some predefined loss components. These are described in Section 5.7. This option is only available when the has selected the input FK value radio button. The Additional Input field is also available to provide access to other advanced features such as choking options. Comments can be added in these boxes by typing a ‘$’ before the comment.

7 Since FloCAD updates cannot be made in the dynamic mode, the use of Output Expressions is strongly advised as an alternative means. In other words, perform model updates in SINDA/FLUINT’s expression system directly, instead of in dynamic calls to Thermal Desktop from SINDA/FLUINT.

Fluid Models

5-17

Figure 5-12

5.3.2.5

Orifice Edit Form dialog box

Capillary

The Capil Edit dialog is shown in Figure 5-13. The Capil device is used where surface tension is a dominant force in a small ageway, such as a wick barrier. It may also be used on a single-phase system to represent laminar losses through a porous media such as a filter. Liquid may always through the capil connector, but vapor and gas cannot if the capillary limit (or bubble point) of the age is not exceeded and any liquid is present to block the age of vapor. All flow through the connector is assumed to be laminar because of the typically small effective diameter of such devices. There is no preferred direction, blockage may occur in either direction depending on conditions. The capillary radius (RC) is used to describe the maximum capillary pressure to be developed by the path. The capillary flow conductance (CFC) defines the effective flow resistance of the path. See the SINDA/FLUINT manual for more details, including formulae for calculating CFC based on various configurations.

5-18

Fluid Models

Figure 5-13

Capil Edit Dialog Box

A Capil connection by itself does not simulate dynamic wicking nor capillary evaporative pumping.8 5.3.2.6

Pump/Fan

The Fan/Pump Edit dialog box is shown in Figure 5-14. Three types of flow rates calculations may be specified: • simple fan or pump curve • full head versus flow map • the GPMP method for logic or expression-based “maps” To specify a fixed flow rate or expression, the SetFlow object should be used (Section 5.3.2.3). The three Pump/Fan options allow for a pressure head versus flow type input. The Head Versus Flow Options have many different forms and input options depending on which type of pump is selected. The status of the check boxes for H,G as Coefficients (COEFS) and pressure treatment - Head Defined As Total Pressure (ISTP) - also affect the input options. The Flow Area options are dependent upon the type of pump selected. The following summarizes some of the input options, but a complete description is available in the SINDA/FLUINT ’s Manual. The Treat H,G as Coefficients (COEFS) check box allows for the quasi-unitless input form to be used for head and flow. In other words, checking this box indicates that the supplied curves won’t be head versus flow, but rather head coefficient versus flow coeffi8 A special macro element called a CAPPMP serves this purpose, but is not accessible directly in FloCAD because it represents a specialized (rare) usage. CAPPMP macros can be added as insert files to the FLOW DATA blocks.

Fluid Models

5-19

Figure 5-14

Fan/Pump Edit Form Dialog Box

cient. The conversion factors may be specified to convert the units. Pumps can also elect to use either static pressures or total pressures within the supplied maps, as specified by the Head Defined as Total Pressure check box. This option requires flow area(s) to be specified. No unit conversions are performed for the pump input curves and conversion factors. Instead, all input is ed directly to SINDA/FLUINT for processing so extra care must be taken to assure that the correct and consistent units are employed. When the Simple Fan/Pump Curve option is selected, pressing the Set button brings up a dialog box to enter a single curve of the head versus flow data for the pump or fan. The units used in the head-flow curve can be selected in the drop-down lists. These units may be chosen independently of current model units. To employ the pump similarity rules instead of a full map (as outlined below), this single curve can be scaled internally by supplying a reference speed for that curve. The initial speed of the pump (SPD) may be set in -consistent units (usually RPM). When selected, the Full Head Vs Flow Map option accepts a series of curves at different speeds to define the pump or fan flow. The Add button next to the Speed field brings up the Head vs Flow Edit dialog box to enter a speed and the associated head versus flow 5-20

Fluid Models

curve. An efficiency curve at this speed is also allowed if the Full Efficiency Map input is selected in the Pump Efficiency group box. The efficiency, which otherwise defaults to unity, will be used to calculate both pump heating and hydraulic torque.9 The GPMP method allows a flow rate and a derivative to be specified as input for a pump. This mostly intended to be able to drive a pump with an external pump code, but could also be used to supply head-flow information using logic or expressions instead of tables. The Additional Input field is also available to provide access to other advanced features such as head degradation factors (cavitation, viscosity corrections, etc.) which are explained in the SINDA/FLUINT ’s Manual. As with the equivalent box on the Lump Edit Form dialog box, comments can be added in these boxes by typing a ‘$’ before the comment. 5.3.2.7

Tabular

While K-factor and other losses can be manipulated in logic or via expressions, certain component performance data is best described using interpolation of tabular-style inputs of the pressure drop versus flow. Examples include vendor data, test data, and outputs from CFD programs. The Tabular Edit dialog box is shown in Figure 5-15. The Tabular type allows the to specify flow rate versus head (or pressure drop) relationships in three different tabular (array) formats: • single array of head versus flow • table of head versus flow • table of flow versus head The units available for flow rate and pressure drop are available for selection in the Head Units and Flow Units drop-down lists. The units selected on the drop-down list specify how the data in the table is to be treated, and is independent of model units. Changing the drop-down list does not do any units conversion on the data in the table; the units can be changed after inputting data into the table. The units are also affected by the Flow Specifier, and Flow Area Specifier drop-down lists (see MREF and NAF in the SINDA/FLUINT ’s Manual). The Flow Specifier allows the flow values to be treated as a “corrected” or “equivalent” or “reference” value, which are commonly used for compressible gas flows. The appropriate reference values of temperature, pressure and gamma should be input when those fields are enabled. The Flow Area Specifier modifies the data to be treated as a function of area if desired. The Table Pressure Basis drop-down list specifies how the input head values are to be treated. If total pressure is used, the flow area must be input. The values for head are used by the code as a retarding force and hence are normally positive. Negative values can be used where appropriate to achieve a pumping force. 9 Refer to the SINDA/FLUINT manual for examples of how this torque can be used to predict shaft speeds, either by balancing torques in a steady state or by using a co-solved first order ordinary differential equation to predict the speed during a transient.

Fluid Models

5-21

Figure 5-15

Tabular Loss Edit Dialog Box

The OFAC (“other factor”) parameter is utilized for two functions. It is a -defined value that can be used to interpolate into families of head curves that are input under the Head vs. Flow Table or Flow vs. Head Table radio buttons. The value of OFAC is then used to interpolate into these tables to produce a head versus flow curve at any value of OFAC. For example, OFAC might represent the position of a valve pin, or the temperature of the inlet fluid: any important parameter that requires the use of families of curves, instead of a single curve of pressure versus flow. Negative values of OFAC will result in a closed path, which means table values of OFAC should be non-negative. By default, the Symmetric Curve check box is selected. This means that the resistance of the device is equivalent for forward and reverse flows. To be symmetric, the initial flow rate and corresponding head in the tables must be zero: the curves must cross the origin. If the map is not symmetric, then the initial flow value must be negative so that the code can compute the resistance if the flow in the path reverses. If the Head term is set to be the Martin Beta Factor, the number of sequential losses should be specified in the field supplied. This specialized option, used to model gas labyrinth seals, is fully explained in the SINDA/FLUINT manual. The Additional Input field is also available to provide access to other advanced features such as the heating and torque tables. Comments can be added in these boxes by typing a ‘$’ before the comment.

5-22

Fluid Models

5.3.2.8

Turbine

The Turbine Edit form is shown in Figure 5-16. A Turbine connector is a componentlevel model of a gas, steam, or hydraulic turbine (whether axial or radial): a turbine whose performance varies as a function of pressure difference or ratio for a given rotational speed. While a Turbine connector is normally used to represent an entire turbine containing any number of stages, it may also be used to model either a single turbine stage or a radial portion of an axial turbine or stage (e.g., hub, meanline, or tip). Such subset models may be combined in series (for single stages) and/or in parallel (for radial portions).

Figure 5-16

Turbine Edit Dialog Box

The Flow vs Head tab provides 4 choices for specifying how the flow versus “head10” information is to be input for the device. The Head Units and Flow Units selections affect how the data entered for flow versus head is treated by SINDA/FLUINT, but they are not used for any conversion process: the data entered in the arrays is itself not altered. Units must be selected to correspond to the data entered in the curves. The Single Flow vs Head Curve option uses one curve, which will be used independent of speed. 10The term “head” is used generically in this case, and includes the most common option: absolute pressure ratio.

Fluid Models

5-23

The Full Flow vs Head Map - Single Pressure Array option uses a different flow curve for each speed, but reuses a single pressure or pressure ratio array for all speeds in the speed array (list). The input for this option will prompt the to input the flow for each pressure, one flow per line, the corresponding pressure will be listed in the input prompt. The number of pressure points can exceed the number of flow values for any one speed curve; a flow does not have to be input for each pressure point at each speed. It is common, for example, for the flows at higher pressure to be absent at lower speed curves if choking behavior is contained within the map itself. The Full Flow vs Head Map option is a full map with independent flow vs. head values input for each speed. The Point Slope Method (GTURB) option allows the to specify the current flow (GTURB) based on the current value of head (HTURB) which is output by the code. The rate of change of flow relative to change in head (DGDH, the slope of the performance curve at the current head) must also be supplied. In other words, the GTURB method provides an alternative to tabular inputs. Instead, the can supply a flow and slope value at the current “head” (e.g., pressure ratio) by logic in FLOGIC 0, or by expressions, or by calls to a routine or external program. The flow-pressure map input options may also tied to the options on the Efficiency tab. If a full efficiency map is also to be input, that choice should be made early such that the head and efficiency data can be input together at the same time on the form. The Speed (SPD) input is used to set the initial speed for the turbine. Any units for the speed input are allowable as long as they are used consistently with other inputs (e.g., the speed array, torque conversion coefficient). If units of RPM (revolutions per minute) are used, an option on the Efficiency tab will enable the torque conversion constant to be automatically computed. The diameter of the turbine blade tip (DTURB) must be input if the AEFFP or ELISTR option is selected on the Efficiency tab. If the speed is in RPM, setting DTURB also allows the velocity ratio U/C output variable (UCR) to be computed in SINDA/FLUINT. The Property Weighting Factor (UPF) is used to weight the fluid properties in the upstream and downstream lumps for the computations of the path. The Table Pressure Basis (static vs. total) selection defines how the pressures in the tables are to be treated, including the Tref and Pref inputs if needed. This selection applies to all aspects of turbine calculations, including up and downstream properties, velocities, etc. The Flow Specifier (MREF) and Flow Area Specifier (NAF) selections define how the flow values in the flow vs. head tables are to be treated. The values entered in tables will be assumed be either raw or in need of correction (based on flow area, current inlet state, etc.) before being used by the program. Further details on the equations themselves can be found in the SINDA/FLUINT input manual. The MREF specifier Use Base Flow Units means that flow values are in the (GU) units selected, and no correction is required. Otherwise, for any other choice of MREF, the Tref, Pref, and Gref values for the appropriate conversion must be input defining the basis of the table values. If a conversion constant is required to get from the table flow units Gtable, to those specified by the Flow Units (GU), 5-24

Fluid Models

a GCF value should be used. GCF is not applied unless some value besides Use Base Flow Units and Flow is Mass Based has been selected. A choice of MREF of Mass Corrected Function of Tref, Pref and Gref will ungrey the checkbox for Speed Array input are Equivalent Speed. If checked, the speeds in the table are then not expected to be the actual speed. The SINDA/FLUINT manual describes the form of the equivalent speeds. The Flow Area Specifier (NAF) allows mass fluxes or velocities to be used in the tables, rather than total mass or volumetric flow rates. The Efficiency tab contains the choices available for how to specify the isentropic Efficiency of the turbine. The Constant Efficiency (EFFP) option can use an expression to set almost any kind of efficiency, though the EFFP value may also be adjusted using Fortran logic in FLOGIC 0. The Single Efficiency Curve vs. Velocity Ratio (AEFFP) assumes the speed is in RPM for calculating the velocity ratio (UCR, the ratio of tangential tip velocity to isentropic spouting or fluid jet velocity) based on the turbine diameter DTURB and the current speed and inlet conditions. Efficiency can then be input as a function of UCR, and this relationship is assumed to apply at any pressure (see also the ELISTR method below if speed variations are to be included). Input for the Full Efficiency Map and the Map of Extracted Power (PLIST) options in made from the Flow vs Head tab when entering flow data into the maps. The Efficiency vs. Velocity Ratio Interpolated against Pressure Array (ELISTR) option is like the AEFFP option except that a different efficiency-UCR curve is specified at each pressure in the pressure array. The same pressure array is used for both efficiency and the Full Flow vs Head Map - Single Pressure Array option. Editing the pressure array on either the flow-head or the efficiency tab will affect the same array. If data already exists in the tables, it will be shifted to accommodate the new pressure. All tables should then be edited to supply new values for flow and efficiency. All efficiency options (except the EFFP input) allow the value computed from the map to be adjusted for scaled by the use of a multiplier and adder, perhaps to model degradations (see SINDA/FLUINT manual). Array Multiplier (EFFM) and Array Adder (EFFA) fields accept the inputs for these two options. Checking the Torque converted from SPD in RPM box will compute the appropriate Torque Conversion Factor (TCF), but this option assumes the has input SPD in RPM. By default, the turbine device turns off the choking calculations for the path by setting MCH=0. The Additional Input box can be used to modify this choice, as an example. 5.3.2.9

Compressor

The Compress Edit main page is shown in Figure 5-17. A Compress connector is a component-level model of an axial or radial compressor: a compressor whose performance varies as a function of flow rate or pressure difference ratio for a given rotational speed. While a Compress connector is normally used to represent an entire compressor containing any number of stages, it may also be used to model either a single compressor stage or a radial portion of an axial turbine or stage (e.g., hub, meanline, or tip). Such subset models

Fluid Models

5-25

may be combined in series (for single stages) and/or in parallel (for radial portions). For positive displacement compressors (e.g., scroll, vane, or piston), see COMPPD in Section 5.3.2.10.

Figure 5-17

COMPRESS Edit Dialog Box

The Head vs Flow tab provides 3 choices for specifying how the “head11” versus flow information is to be input for the device. The Head Units and Flow Units selections affect how the data entered for head versus flow is treated by SINDA/FLUINT, but they are not used for any conversion process: the data entered in the arrays is itself not altered. Units must be selected to correspond to the data entered in the curves. The Pressure Ratio or Difference vs Flow option uses one curve for any speed input. The Full Head vs Flow Map option allows the to input a series of head versus flow curves, each at a different speed. The current speed will then be used to interpolate into the table. When this option is selected, the also has the ability to add an explicit surge line for the flow or pressure in the map.

11The term “head” is used generically in this case, and includes the most common option: absolute pressure ratio.

5-26

Fluid Models

The Point Slope Method (GCOMP) option allows the to specify the current flow (GCOMP) based on the current value of head (HCOMP) which is output by the code. The rate of change of flow relative to change in head (DGDH, the slope of the performance curve at the current head) must also be supplied. In other words, the GCOMP method provides an alternative to tabular inputs. Instead, the can supply a flow and slope value at the current “head” (e.g., pressure ratio) by logic in FLOGIC 0, or by expressions, or by calls to a routine or external program. The flow-pressure map input options may also tied to the options on the Efficiency tab. If a full efficiency map is also to be input, that choice should be made early such that the head and efficiency data can be input together at the same time on the form. The Table Pressure Basis (static vs. total) selection defines how the pressures in the tables are to be treated, including the Tref and Pref inputs if needed. This selection applies to all aspects of Compressor calculations, including up and downstream properties, velocities, etc. The Flow Specifier (MREF) and Flow Area Specifier (NAF) selections define how the flow values in the flow vs. head tables are to be treated. The values entered in tables will be assumed be either raw or in need of correction (based on flow area, current inlet state, etc.) before being used by the program. Further details on the equations themselves can be found in the SINDA/FLUINT input manual. The MREF specifier Use Base Flow Units means that flow values are in the (GU) units selected, and no correction is required. Otherwise, for any other choice of MREF, the Tref, Pref, and Gref values for the appropriate conversion must be input defining the basis of the table values. If a conversion constant is required to get from the table flow units Gtable, to those specified by the Flow Units (GU), a GCF value should be used. GCF is not applied unless some value besides Use Base Flow Units and Flow is Mass Based has been selected. The Flow Area Specifier (NAF) allows mass fluxes or velocities to be used in the tables, rather than total mass or volumetric flow rates. The Specific Speed Conversion Factor (SSCF) is the conversion factor to compute the current specific speed for the compressor (see the SINDA/FLUINT manual for more information). The Property Weighting Factor (UPF) is used to weight the fluid properties in the upstream and downstream lumps for the computations of the path. The Speed (SPD) input is used to set the initial speed for the compressor. Any units for the speed input are allowable as long as they are used consistently with other inputs (e.g., the speed array, torque conversion coefficient). If units of RPM (revolutions per minute) are used, an option on the Efficiency tab will enable the torque conversion constant to be automatically computed. The Efficiency tab contains the choices available for how to specify the isentropic efficiency of the compressor. The Constant Efficiency (EFFP) option can use an expression to set almost any kind of efficiency, though the EFFP value may also be adjusted using Fortran logic in FLOGIC 0.

Fluid Models

5-27

The Single Efficiency Curve vs. Flow (AEFFP) uses only a single curve for any speed. Input for the Full Efficiency Map and the Map of Extracted Power (PLIST) options are made from the Flow vs Head tab when entering flow data into the maps. All efficiency options except the constant (EFFP) input allow the value computed from the map to be adjusted or scaled by the use of a multiplier and adder. Array Multiplier (EFFM) and Array Adder (EFFA) fields accept the input for these options. Checking the Torque converted from SPD in RPM box will compute the appropriate Torque Conversion Factor (TCF), but assumes the has input SPD in RPM. The Compress device by default turns off the choking calculations for the path by setting MCH=0, assuming that this phenomena has been built into the provided tables. The Additional Input box can be used to modify this choice, as an example. 5.3.2.10

Compressor - Positive Displacement

The Comppd Edit main page is shown in Figure 5-18. A COMPPD connector is a component-level model of a piston, rotary or reciprocating vane, screw, or scroll compressor. These machines differ from variable-displacement compressors (see Compress in Section 5.3.2.9) in that a fixed volume of working fluid is, in principle, delivered with each revolution. While a COMPPD connector is normally used to represent an entire compressor containing any number of stages, it may also be used to model a single compressor stage. Such subset models may be combined in series for multi-stage compressors. The Head vs Flow tab provides 3 choices for specifying how the “head12” versus flow information is to be input for the device. The Head Units and Flow Units selections affect how the data entered for head versus flow is treated by SINDA/FLUINT, but they are not used for any conversion process: the data entered in the arrays is itself not altered. Units must be selected to correspond to the data entered in the curves. In a COMPPD device, the principle input is a volumetric efficiency (not to be confused with the isentropic efficiency, which is also an input for this device). While unitless, the is also required to specify the units that will result from a multiplication of this isentropic efficiency and the displacement volume and the current speed. Refer to the SINDA/FLUINT ’s Manual for formulae. The Volumetric Efficiency vs Speed option uses one curve for any speed input. The Full Volumetric Efficiency vs Head Map option allows the to input a series of volumetric efficiencies versus head curves, each at a different speed. The current speed will then be used to interpolate into the table. The full map input option is tied to the isentropic efficiency options on the Efficiency tab. If a full isentropic efficiency map is also to be input, that choice should be made early if both the volumetric efficiency and isentropic efficiency data are to be input together at the same time on the form. The Point Method (EFFV) option allows the to specify the volumetric efficiency directly (perhaps via expressions, updates in logic, etc.). 12The term “head” is used generically in this case, and includes the most common option: absolute pressure ratio.

5-28

Fluid Models

Figure 5-18

COMPPD Edit Dialog Box

The Displacement per Rotation (DISP) of the compressor, the volume swept per rotation, is input based on the current -defined length units for Thermal Desktop. If a conversion constant (GCF) is required to get from the table flow units to those specified by the Flow Units (GU),13 then the is required to make the GCF value consistent with the units chosen. Otherwise, if SPD is in RPM, the GCF conversion factor can be automatically calculated. The Property Weighting Factor (UPF) is used to weight the fluid properties in the upstream and downstream lumps for the computations of the path. The Efficiency tab contains the choices available for how to specify the isentropic efficiency of the COMPPD device. The Constant Efficiency (EFFP) option can use an expression to set almost any kind of isentropic efficiency, though the EFFP value may also be adjusted using Fortran logic in FLOGIC 0. The Single Efficiency Curve vs. Flow (AEFFP) uses only a single curve at all speeds. Input for the Full Efficiency Map and the Map of Extracted Power (PLIST) options in made from the Flow vs Head tab when entering volumetric efficiency data into the maps.

13GGU = GCF*EFFVtable*DISP*SPD

Fluid Models

5-29

All isentropic efficiency options except the constant (EFFP) option allow the value computed from the map to be adjusted or scaled by the use of a multiplier and adder. Array Multiplier (EFFM) and Array Adder (EFFA) fields accept the input for these options. Checking the Torque converted from SPD in RPM box will compute the appropriate Torque Conversion Factor (TCF), but assumes the has input SPD in RPM. The COMPPD device by default turns off the choking calculations for the path by setting MCH=0. The Additional Input box can be used to modify this choice, as an example. 5.3.3

Ties

Ties provide a method for connecting thermal submodels to fluid submodels as needed to represent convection. Only heat energy (and not fluid mass) can be transferred between the two submodel types: advection is allowed only within a single fluid submodel. The can create a connection between a group of lumps, a group of nodes or surfaces, and (optionally) a group of paths. Although one set of interconnections between groups of network elements is collectively referred to as a single “tie” in FloCAD, a set of “FLUINT ties” (each of which connects one node to one lump) will be created by a FloCAD tie. In other words, a “FloCAD tie” can be thought of as a generator of “FLUINT ties.” A tie is generated using either the Thermal > Fluid Modeling > Tie-to-Node or Tieto-Surface menu option (see Section 5.3.3.1 and Section 5.3.3.2, respectively). When creating ties, the selects the lumps to be connected by the tie followed by nodes or surfaces. The can then press <Enter> if no paths (and therefore no internal convection correlations) are needed. Otherwise, the can select one or more paths (Tubes and/or STubes only). After a tie is created, the tie may be edited by selecting the tie, followed by the Thermal > Edit command. The Tie Edit Form will look slightly different depending on whether the heat transfer is -defined or calculated. In either case, the form has a button that reads either Add Code or Edit Code. This button accesses the Network Element Logic as described in Section 2.10.10. If no paths are selected for the tie, then the heat transfer coefficient will be -defined (see “-Defined Heat Transfer” on page 5-34). If one or more paths are attached to the tie, the internal heat transfer correlations can be used to compute the heat transfer coefficient (see “Calculated Heat Transfer” on page 5-32). Ties can be created either from all nodes in the group to the closest lump, or from all lumps in the group to the closest node. This choice can be made in the Tie Generation Direction region of the Tie Edit Form. In the first case (Nodes to lumps), all nodes will be connected to the nearest lump, but some lumps might not be connected. In the second case (Lumps to nodes), all lumps will be connected to the nearest node, but some nodes might not be connected. (This is the same logic followed by the From versus To list for ors.) Since the area for ties is associated with the nodes, if a node is not connected, its area is not used either. If some of the lumps or nodes included in the FloCAD tie are moved, the FLUINT ties will be adjusted, with some nodes or some lumps perhaps becoming disconnected according to the choice of Lumps to nodes versus Nodes to lumps.

5-30

Fluid Models

The Additional Input field is also available to provide access to other advanced features like DUPL/DUPN and UAM factors. Comments can be added in these boxes by using a ‘$’ before the comment. The lump, node and path objects are all displayed in lists in the HTC Edit dialog box. The Add and Delete buttons below each list can be used to change the objects associated with the tie. If all nodes or lumps are deleted and the closes the dialog box by selecting OK, the tie will be deleted. Except for specialized Pipe ties (Section 5.4), all of the ties generated within Thermal Desktop which compute the heat transfer coefficient are of the FLUINT HTN type. The HTN tie provides heat transfer based on an upstream or downstream weighting, where the temperature of the lump is used as the bulk fluid temperature for the heat transfer. Coarse discretization of the fluid model can cause the heat transfer to be underestimated. 5.3.3.1

Ties to Nodes

To create a FloCAD tie to one or more individual nodes (as opposed to surfaces containing nodes, as described in Section 5.3.3.2), select the Fluid Modeling > Tie-to-Node command. Since no heat transfer area is associated with an individual node, the area is either computed from the nearest path (if paths are selected for the tie) or it can be input directly. The area field is used depending on the status of the Use Path Area button, or the per Area button. The Use Path Area check box can be selected when there is only one lump, one node (with no area associated with the node from a surface) and one path included in the tie. The per Area button is available for surfaces or nodes with area attached to them. Care should be taken to insure that the actual area available for heat transfer is not exaggerated when multiple nodes use the same path to define their heat transfer area. The Multiplier field is used as a scaling factor on top of the overall heat transfer coefficient. The can switch between a -defined HTC and a computed HTC by using the Type list box menu (see Section 5.3.3.4 and Section 5.3.3.3, respectively). The surface options are greyed for the tie-to-node type of tie, since that type has no area associated with the nodes. 5.3.3.2

Ties to Surfaces

To create a FloCAD tie to one or more surfaces (as opposed to individual nodes, as described in Section 5.3.3.1), select the Fluid Modeling > Tie-to-Surface command. The area of the surface is used for the heat transfer area. The per Area button is available for surfaces or nodes with area attached to them. The Multiplier field is used as a scaling factor on top of the overall heat transfer coefficient. The can switch between a -defined HTC and a computed HTC by using the Type list box menu (see Section 5.3.3.4 and Section 5.3.3.3, respectively) if paths are selected for the tie. By default, the tie connects to the top side of the surface. By double-clicking on the surface listed in the Nodes/Surfaces field, this choice can be changed to the bottom surface, or to both top and bottom surfaces. If solids are attached to the tie, then the options will be the individual faces of the solid.

Fluid Models

5-31

It should be noted that all surfaces selected will be connected to a lump via a tie, assuming the default Nodes to lumps option is selected. Even if one of the surfaces covers another surfaces, the nodes underneath are still connected: no attempt is made to resolve overlaps or geometric interference. Want "Hands-On" Information? The tutorial exercise “Air Flow Through an Enclosure” on page 22-3 creates ties to surfaces. Work through the exercise for practical knowledge. 5.3.3.3

Calculated Heat Transfer

When paths are selected and the Type is set to Calculated HTC, the Calculated Heat Transfer Coefficient Tie Edit Form dialog box is available, as shown in Figure 5-19.

Figure 5-19

5-32

Tie Edit Form Dialog Box for Calculated Heat Transfer Coefficients

Fluid Models

The internally computed heat transfer coefficient is based on a set of correlations for convective heat transfer for internal duct flow. There are correlations for both single- and two-phase flows. Flow regime mapping is utilized to pick the appropriate correlation for two-phase flow conditions. Single-phase flows utilize separate correlations for laminar, turbulent and transition flow. All of the correlations can be accessed within SINDA/FLUINT logic blocks, or overwritten or otherwise customized by the . Appendix B of the SINDA/ FLUINT ’s Manual provides greater detail on the specific correlations and how the can access or replace them with their own routines. Selecting the Advanced Coefficient Specification button within the Calculated Heat Transfer Coefficient Edit dialog box displays the Advanced Tie Heat Transfer Coefficient Equation Parameters dialog box (Figure 5-20) giving the access to all of the underlying constants used in the internal correlations. This provides a means to customize the underlying equations, such that many of the correlations available in the literature can be input on this page.

Figure 5-20

Fluid Models

Advanced Tie Heat Transfer Coefficient Equation Parameters

5-33

5.3.3.4

-Defined Heat Transfer

When no paths are selected,14 or if the wants to directly specify the heat transfer coefficient (HTC), the Defined Heat Transfer Coefficient Tie Edit Form dialog box is available, as shown in Figure 5-21.

Figure 5-21

Tie Edit Form Dialog Box for -Defined Heat Transfer Coefficients

The -defined heat transfer coefficient has two modes that are selected via the radio buttons in the Tie Heat Transfer Coefficient field.

14Tubes and STubes provide additional information required to perform forced convection calculations, including flow area, wetted perimeter, fluid velocity, etc. Without that information, the must supply the convection coefficient. Similarly, a tie to a surface provides heat transfer area information, whereas when making a tie to a node that is not associated with a surface (or surface-coated solid), the must provide the area information separately.

5-34

Fluid Models

In the first mode, when the Input Heat Transfer Coefficient*Area (UA) button is selected, a single value that includes the area and HTC is input. Three choices of how this UA is to be applied are available. First, the Use UA for each Tie will use the input value for each FLUINT tie generated. It is intended for the case where the area of each node is the same and matches the area that makes up the “A” in the UA value. Second, the Average UA over all ties option takes the UA input as the total heat transfer coefficient to be divided evenly over all the nodes selected. Each tie will have the identical value so this choice is not appropriate unless each node has the same area. The third choice, the UA spread over surfaces choice is intended for cases where the nodes or surfaces selected have differing areas. This option interprets the UA to be a total value that is apportioned to each node based on its fractional area. The multiplier can also be used to scale the input UA. If the Input Area and Heat Transfer Coefficient (AHT/UB) button is selected, a number of internal correlations are available. Separate inputs for the area (AHT) and the heat transfer coefficient (UB) are available using the Generic type of tie. The correlations and their input parameters are all described in detail in Section 7 of the SINDA/FLUINT manual. (Caution: the jet impingement correlations are for gas jets only.) The multiplier is also available for input on all types. The heat transfer coefficient input assumes that the exponent used on the temperature difference (UEDT) is equal to zero. See the SINDA/ FLUINT ’s Manual for more detail on alternative usage. 5.3.3.5

Pool Boiling Ties

To create a FloCAD Pool Boiling Tie select the Fluid Modeling > Pool Boiling Tie command (if the Tie already exists, edit the Tie and select Pool Boiling Tie for the Type). The will be prompted to select a single lump, a set of surfaces, and an optional path. The must then select a reference point for the bottom of the pool, and another point to provide the direction vector towards the top. The liquid level input on the edit form should also be based to this vector, i.e. zero liquid level should correspond to the bottom point. The DEPTH (described in the SINDA/FLUINT manual section 3.6.3.3) input for the tie will be computed as the input liquid level minus the node elevation (computed as the distance along the vector). The direction vector has a graphical depiction on the screen as an arrow, and will have grip edit points attached. If the creates a different type of tie and uses the edit form pull-down to change to a Pool Boiling Tie, the grip edits provide a mechanism to changing the direction vector location. The default location for the direction vector is a unit vector at 0,0,0 in the world coordinate system, pointing in the positive z direction. If a path has been selected for the Pool Boiling Tie, the FPORT input is used to correct the sign of the flow rate if necessary. The code assumes that positive flow rate in the path FR variable is entering the vessel. If path has been drawn as leaving the vessel, a value of FPORT=-1.0 should be used to rectify the flow rate. The radio buttons on the form should be used to signal if the path is entering at the top or bottom of the vessel. The Multiplier field is used as a scaling factor on top of the overall heat transfer coefficient. The Use Node Attached Area check box will use the area associated with each node in the drawing when checked. When unchecked it will allow the to input an area to use for each node in the Area field.

Fluid Models

5-35

The ACCM, THKL, SPR, CSF and DCH fields provide the the ability to control the heat transfer coefficient as described in the SINDA/FLUINT manual for the HTP Ties (section 3.6.3.3). 5.3.4

Lumps and Paths

A special utility exists to aid in creating a series of lumps and path in one operation. This series of lumps and paths might represent a discretized flow age.15 This utility is accessed by using the Fluid Modeling > Lumps And Paths command that results in the display of the Create Lumps and Paths dialog box, shown in Figure 5-22. The selection choices in the dialog box give the control over the start and end points for the path. Picking a point allows the to create a lump at that point. Picking a lump uses an existing lump to start or end the string of lumps and paths.

Figure 5-22

Create Lumps And Paths Dialog Box

The also specifies the number of lumps to create in the series. The number entered must take into the nature of the end points. For example, selecting Pick Point to Pick Point and then entering ‘4’ as the number of lumps will generate four new lumps and three paths, while Pick Lump to Pick Lump and four lumps will generate four new lumps and five paths. Once generated, such linear strings of lumps and paths can be attached to adjacent nodes or surfaces using the convection options detailed in Section 5.3.3.1 and Section 5.3.3.2. Want "Hands-On" Information? The tutorial exercise “Air Flow Through an Enclosure” on page 22-3 shows how this command works when applied to a model.

15The should make sure that FloCAD Pipes (Section 5.4) are not a more appropriate or useful method. The “lump and path” model construction method predates the availability of Pipes, but has been retained since it is often appropriate for flows through sections with complex walls defined by independent Thermal Desktop surfaces. For example, for air flow between two closely spaced electronics boards, a “lump and path” method is perhaps more appropriate than a Pipe.

5-36

Fluid Models

5.3.5

Path Rotation Axis

Paths normally represent stationary ages, but can also be spun about an axis, including cases where one wall rotates and the other doesn’t. Such channels are present in rotating machinery (engines, turbomachines, etc.) as leakage paths, coolant channels, or lubrication ages. Therefore, the can optionally specify an axis of rotation, the rate of rotation of the axis, and can then associate paths that will rotate about that axis. (Rotations are performed mathematically during the simulation, and are not depicted graphically.) To create the axis object, the selects Thermal > Fluid Modeling > Path Rotation Axis and, when prompted, specifies the base point and then the top point of the rotation axis. The selection order is important because the direction of rotation is based on a right hand rule for positive rotation, with the thumb pointing up the axis (from the first to second point), and the rotation being in the direction of the fingers of the right hand. The path entrance and exit angles are measured in a clockwise direction when looking down the axis from the top from the tangent vector to the path direction vector. (See figures in Section 3 of the SINDA/FLUINT ’s Manual for examples.) The axis is drawn as a blue line with an arc that shows the direction of positive rotation. Paths can be added, deleted and edited from the axis object using the Rotation Axis Edit dialog box: either select the path and then choose Thermal > Edit, or select the Edit icon in the Rotation Axis Edit dialog box is shown in Figure 5-23.

Figure 5-23

Fluid Models

Rotation Axis Edit Dialog Box

5-37

Add, Delete and Edit Path buttons are located on the right side of the dialog box. To add a path, select Add and follow the prompt to identify the path(s) to be added to the current rotation axis. To delete a path, select the path in the Paths Attached to this Axis field and then select Delete. A path may be edited by selecting the path in the Paths Attached to this Axis field, selecting Edit Path. and making the desired changes using the Path Rotation Overrides dialog box (shown in Figure 5-24).

Figure 5-24

Path Rotation Overrides Dialog Box

The inlet radius, outlet radius, and angles to the tangent vector and the axis vector are all computed from the geometry of the model. Paths not associated with a FloCAD Pipe have a direction vector from the inlet lump to the outlet lump. Paths associated with a FloCAD Pipe take the vectors from the pipe centerline direction at the inlet and outlet lump locations. These values can be overridden by the by editing the paths from within the Rotation Axis Edit dialog box. The appropriate values are output to SINDA/FLUINT in the individual path input sections as the values VAI, VAJ, VXI, VXJ, ROTR, etc. 5.3.6

IFaces

An IFace object connect two tanks (and only tanks: IFaces cannot interconnect junctions or plena). IFaces model the boundary between the two control volumes: they describe how that boundary is affected by pressure and volume changes within the lumps. IFaces can be created by preselecting selected two tanks, or being prompted by the command to select the lumps. IFaces are displayed as an arc between two lumps. The arc has a smaller radius than the arc drawn for an FTie (see below), so it extends further out from the lumps than does an FTie arc: the curvature is more pronounced. Six different types of IFaces are available via the IFace Type pulldown. The parameters specific to each type of device will appear below in the options block. The data of all types is saved even when the type is changed, such that the can change IFace types without losing data in case that decision is later revoked and the original type is restored. Where parameters are applicable to more than one type of IFace, that data is used for all applicable types. The lump id list box allows new lumps to be added and old lumps to be deleted. You cannot exit the form unless exactly two lumps are in the set.

5-38

Fluid Models

5.3.7

FTies

FTies (fluid to fluid heat transfer ties, or “fluid ties”) connect two lumps. They are used to transfer heat (but not mass) between the two lumps. FTies are needed when fluid velocities are low and/or fluid thermal conductivities are high (e.g., liquid metals), such that the usual assumption that advection overwhelms within-fluid conduction is no longer appropriate. FTies can be created by preselecting two lumps or any number of (S)tubes, or by being prompted after selecting the Thermal > Fluid Modeling > FTie menu. When multiple (S)tubes are selected, a unique (single) FTie is made from each path, and there is no subsequent dependence upon the originally selected group of paths. FTies are displayed as an arc between two lumps. The arc has a larger radius than the arc drawn for an IFace, so it does not extend as far out from the lumps as does an IFace arc: the curvature is less pronounced. Three different types of FTies are available, as selected using the radio buttons in the FTie Type box. The parameters specific to each device will appear next to the type. All three types may have a (S)tube selected instead of two lumps. The end lumps of the (S)tube will be used as the lumps for the FTie. The Defined Conductance allows the to supply a conductance. The Constant Heat Flux type allows the to compute the rate of heat transfer. The Conductance Along Tube type uses the size and shape of a (S)tube to compute the heat transfer based on the properties of the path, and the thermal conductivity of the end lumps. Since a path must be selected for this type, the lump input is becomes inaccessible. Both the lump and path boxes allow the to change or replace the currently selected lumps or path. The lump id list box allows new lumps to be added and old lumps to be deleted. You cannot exit the form unless exactly two lumps are in the set.

5.4

Pipes

Pipes are a high level representation of a fluid duct, a heat pipe, an empty pipe, or a “wire” (extruded solid) of thermal nodes. Pipes can generate nodes, conductors, lumps, paths, and ties as needed to represent the duct, heat pipe, or wire. Pipes can also be set up to apply other (non-pipe) surfaces as the pipe wall. Pipes can model: • Gas, liquid, or two-phase flows within a duct or heat exchanger, with automatic generation of appropriate ties and convection correlations. The pipe walls may or may not be included in the definition. • Constant or variable conductance (gas loaded) heat pipes. • The walls of a pipe without fluid flow. • A rectangular or circular solid cross section extruded along a centerline, perhaps representing a rod, conduit, single wire, wire bundle, etc. Pipes are primarily defined by a centerline. The centerline can be a single wireframe object of a series of wireframe objects with shared end points. The order of multiple curves determines the positive flow direction of the pipe. The ordering of multiple AutoCAD lines

Fluid Models

5-39

can be found on the Pipe Selection tab (Figure 5-25), using the AutoCAD list command, or use the Thermal Desktop Model Browser. Another hint of flow direction is provided by the FloCAD pipe ID, which appears near the upstream end of the pipe. The pipe direction can be reversed using Reverse Pipe Direction in Modeling Tools (Section 7.14). It is possible to use a single line object to define more than one pipe. One example of such usage is the modeling of concentric (typically annular) ducts. In such a case, the outer (annular) pipe must have a -defined flow area and hydraulic diameter to correctly compute the path parameters. All of the sub-elements in a pipe can be overridden by using the Override calculations by pipe check box on the individual edit dialogs. Caution should be used when using the override calculations by pipe option. First, overrides may be lost if the number of subdivisions in the pipe is changed. Second, inconsistent lengths and volumes in a pipe can lead to solution instabilities and non-physical responses. Pipes, with the exception of heat pipes, can be connected and disconnected from each other. These procedures are discussed in Section 7.16. 5.4.1

Pipe Creation

The FloCAD pipe is created by choosing the Thermal > Fluid Modeling > Pipe command, typing Ripe or selecting the appropriate icon in the Ribbon or Toolbars. The following prompts are given: • Select Line Entity(s) for Pipe Centerline - Select a preexisting, single-line object (Draw > Line, Spline, Polyline, 3D Polyline, Ellipse, or Arc) or a set of line objects connected in series (e.g, sharing end points exactly) representing the centerline of the pipe. The pipe centerline cannot be a closed curve. If the command is given with one or more line objects selected, this prompt is skipped and the selected. If multiple centerline curves are selected, but they do not share end points exactly, the pipe will not be created. • Select Line Entity for Pipe Cross Section Shape Definition - Optionally select a preexisting, closed curve (AutoCAD object) representing the perimeter of the fluid space (inner perimeter of the pipe wall). The shape curve must be closed and planar. Shape curves can be placed anywhere and in any orientation: they will be copied, relocated to the start of the centerline and placed perpendicular to the centerline, automatically. Pipes with circular or rectangular cross-sections do not require a shape curve. Selecting <ENTER> instead of selecting a shape curve is permitted. If a shape curve is not selected. the can specify a circular or rectangular cross-section, select a shape curve later, specify the flow area and hydraulic diameter, or select faces of thermal objects (FD surfaces and solids, or planar finite elements) that form the wall of the fluid flow. After the prompts, the Ripe Edit Form opens. The tabs of the Ripe Edit Form are described in the following sections.

5-40

Fluid Models

5.4.2

Pipe Selection

The first tab of the Ripe Edit Form is the Pipe Selection tab. This tab allow the to choose the pipe type, modify the pipe centerline curve selection, choose the pipe discretization method, select surfaces to define the wall of the pipe when appropriate, and add Network Element Logic to the pipe elements.

Figure 5-25

Ripe Edit Form Dialog Box Pipe Selection Tab

Fluid Pipe with Wall. This selection generates the entire thermal-fluid network associated with the pipe: it will generate lumps, paths, nodes, conductors and ties based on the pipe definition. The Ripe Edit form tabs available for this selection are: Subdivision (Section 5.4.3), Pipe Attributes (Section 5.4.4), Ties (Section 5.4.6), Node Numbering (Section 5.4.7), Radiation (Section 5.4.8), Surface (Section 5.4.9), Insulation (Section 5.4.10), and Advection (Section 5.4.11). Want "Hands-On" Information? Gain practical experience working with pipes by completing the FloCAD tutorial exercise “Manifolded Coldplate” on page 22-37. Fluid Pipe without Wall. This selection generates a fluid-only network associated with the pipe: it will generate lumps and paths based on the pipe definition. This option may be desired if either an adiabatic duct is to be modeled or if custom heat transfer and wall models will be added separately after the duct has been generated. The Ripe Edit form tabs available for this selection are: Subdivision (Section 5.4.3), Pipe Attributes (Section 5.4.4),

Fluid Models

5-41

and Surface (Section 5.4.9). Heat Pipe. This selection generates a thermal-only network associated with the pipe: it will generate nodes and conductors based on the pipe definition and will additionally generate solution logic to emulate the behavior of the heat pipe. The Ripe Edit form tabs available for this selection are: Subdivision (Section 5.4.3), Heat Pipe Data (Section 5.4.5), Node Numbering (Section 5.4.7), Radiation (Section 5.4.8), Surface (Section 5.4.9), Insulation (Section 5.4.10), and Advection (Section 5.4.11). Heat pipes utilize the SINDA/FLUINT HEATPIPE and HEATPIPE2 routines, generating most of the input for those routines automatically. (Refer to the SINDA/FLUINT ’s Manual for more details on both these key routines as well as on ing subroutines.) Heat pipe models use only nodes and conductors: no lumps, paths, nor ties are generated. Therefore, only a thermal submodel is needed. If a heat pipe is circumferentially subdivided, the HEATPIPE2 routine is used. Vapor chamber fins and 2D/3D heat pipes without gas can also be modeled, without using FloCAD pipes, as long as the vaporization and condensation film coefficients are assumed to be equal. To model such a system in Thermal Desktop, create a unique arithmetic node representing the vapor temperature, and connect it to the surface or surfaces representing the interior of the heat pipe or vapor chamber fin via a “node to surface” conductor. Unfortunately, the capacity of the pipe cannot be compared against the actual heat flows in such circumstances: no QLeff value is available. Want "Hands-On" Information? Two FloCAD tutorial exercises work with heat pipes. Refer to exercises “Heat Pipe Model” on page 22-23, and “Drawn Shape Heat Pipe” on page 22-85. Wall Only. This selection generates a thermal-only network associated with the pipe: it will generate nodes and conductors based on the pipe definition. The Ripe Edit form tabs available for this selection are: Subdivision (Section 5.4.3), Pipe Attributes (Section 5.4.4), Node Numbering (Section 5.4.7), Radiation (Section 5.4.8), Surface (Section 5.4.9), Insulation (Section 5.4.10), and Advection (Section 5.4.11). Fluid Pipe with Surfaces for Wall. This selection generates only the fluid network associated with the pipe: it will generate lumps, paths, and ties based on the pipe definition. The node and conductor definitions will be based on the surfaces defining the walls of the pipe. Special methods are used to calculate the path area and lump volumes from the objects selected. The Ripe Edit form tabs available for this selection are: Subdivision (Section 5.4.3), Pipe Attributes (Section 5.4.4), Ties (Section 5.4.6), and Surface (Section 5.4.9). See “Wall Surface IDs” on page 5-44 for specifying the surfaces used for the walls. Want "Hands-On" Information? Gain practical experience working with pipes with surfaces for wall by completing the FloCAD tutorial exercise “FEM Walled Pipe” on page 22-99.

5-42

Fluid Models

Centerline Curve IDs This field lists all centerline curves defining the pipe centerline. Additional centerline curves can be added by selecting the Add button directly below the list. Curves can be removed from the list by: selecting the curve ID in the list and selecting the Delete button directly below the list; or by selecting the Graphical Delete button and selecting the curve or curves from the graphics area. If the Pipe is defined using more than one centerline curve, then those lines must be connected end-to-end, with end-points shared exactly. If the lines are not connected end-toend when initially selected, the pipe object will not be created. After a pipe object has been created, if the lines are edited or moved such that they are no longer connected at all of the endpoints, then the pipe object will still exist but it will contain no lumps, nodes, paths or ties. If the lines are further edited or moved such that the lines are again in , the pipe will be reformed. The lines represent the centerline of the pipe. The lumps will be created along the center lines, and the paths are drawn as straight lines between the lumps. If the pipe is following a curve, or makes turns, the paths as drawn will therefore not follow the pipe, but will appear to “take a shortcut.” Despite this graphical depiction, the data for a path will still take into the actual length of the pipe centerline.

Pipe Discretization Options The pipe discretization determines the location of a fluid pipe’s lumps relative to the location of the pipe wall nodes. All fluid pipes have junctions at either end for connecting pipes together (Section 7.16). The discretization method may change the connecting junction to a pipe-defined lump. These options are not available for pipes without flow. Unless specific requirements are needed, the default Centered discretization is highly recommended. The Upstream and Downstream options allow defining the pipe for flat-front modeling: two-phase modeling of a liquid displacing a gas and “filling” the pipe or high-pressure gas displacing a liquid thereby “purging” the system. In this situation the gas-liquid interface travels as a planar surface (“front”): the liquid and gas are separated axially. For these two types of modeling, the twinned lump/path conditions upstream or downstream of the pipe must be consistent with the pipe construction. Flat-front modeling is very specialized and should not be used without understanding the background requirements. See the SINDA/ FLUINT s Manual for more information regarding flat-front modeling. Upstream. With this discretization method, the pipe starts with a pipe-defined lump and ends with a connecting junction. Any ties generated are connected to the downstream node. When this option is selected, the can choose Enable Purge Mode. This selection sets up the model for flat-front modeling in the purge mode: gas purging liquid from the pipe. Downstream. With this discretization method, the pipe starts with a connecting junction and ends with a pipe-defined lump. Any ties generated are connected to the upstream node. When this option is selected, the can choose Enable Fill or Prime Mode. This selection sets up the model for flat-front modeling in the fill mode: liquid filling a pipe.

Fluid Models

5-43

Centered (Recommended). With this discretization method, the pipe starts and ends with a connecting junction. Any ties generated are connected to the node co-located with the lump. Unless required for special circumstances, such as flat-front modeling, this is the recommended option.

Add Code... Buttons The Add Code buttons provide access for Network Elements Logic (Section 2.10.10). Network Element Logic defined in the Ripe Edit Form can reference the indirect operators for pipes as well as the indirect operators for the element type of the selected button. The indirect operators are as shown in Table 2-3 on page 49.

Wall Surface IDs When the Fluid Pipe with Surfaces for Wall option is selected, thermal model objects such as surfaces, solids, and planar elements can be selected as the wall of the flow. The buttons below the Wall Surface IDs field allow definition of the pipe walls. Add. Click on the Add button and select the surfaces/solids from the drawing. The visibility controls below the Wall Surface Ids field are used to help find and identify the objects associated with the pipe wall. The faces that are selected on each surface/solid are the only surfaces that will have ties attached. The surfaces selected should completely encircle the path centerline: no gaps in the circumference are allowed. The wall surfaces may be open at the ends, but should the code fail to find a wall at any point along the centerline where the path area is to be computed, will lead to zero path area. This means any openings for side flow must have a surface placed over them, noting that such a closure surface need not be output to SINDA/FLUINT. To determine the path flow areas, rays are shot within planes that are normal to the pipe centerline at any point where a flow area is needed. In areas where the flow centerline makes a sharp turn, a series of radial planes will be swept around the corner using an approximate radius for the curvature. Using this method, path flow areas are calculated at the end of each path, and also at the boundary of each lump. The lump volumes are computed by averaging the path areas for each half of the lump, multiplied by the length of the path. Delete. To delete an object from the list, select the object and select the Delete button below the Wall Surface IDs list. Edit. The Edit button allows the to select the faces of the surfaces and solids to which ties are connected. The options for thin shells are Top, Bottom, Inside or Outside depending on the type of shell. This options for solids are given in Section 4.4.7. Add Tie. Multiple Ties can be used for different sections of the pipe walls (e.g. inner and outer walls of annular flow). Selecting Add Tie will add another tab in the Wall Surface IDs section to specify the surfaces associated with the additional ties. Delete Tie. This button allows the to delete the currently active Tie tab. Note that all surfaces for the pipe must be included in the Tie tabs.

5-44

Fluid Models

Ties are generated to all of the faces that are selected on the wall surfaces. This means that extra surfaces can be part of the selection set, but such surfaces do not influence the tie generation. By deselecting all the faces on a surface/solid, that surface will be shown with a red background in the Ripe Edit Form dialog box. Such a surface will be used to mark the path and volume boundaries, but it will not included in tie generation. 5.4.3

Subdivision

The Subdivision tab (Figure 5-26) allows the to specify the number of axial segments of a pipe and the subdivision of the of pipe circumference.

Figure 5-26

Ripe Edit Form Dialog Box Subdivision Tab

Pipe Circumference This subdivision defines the number of nodes around the circumference of the pipe. This subdivision is only used for Fluid Pipes with Wall, Heat Pipes, and Wall Only pipes. The can choose Equal or List to define whether the circumferential subdivisions are equally or unequally spaced, respectively. The List is provided as an ascending list of fractional locations of the nodal boundaries with values between 0 and 1, not inclusive.

Pipe Length This subdivision defines the number of axial segments of the pipe, not including the connecting junctions and nodes at the ends of the pipe (see “Pipe Discretization Options” on page 5-43). The can choose Equal or List to define whether the circumferential Fluid Models

5-45

subdivisions are equally or unequally spaced, respectively. The List is provided as an ascending list of fractional locations of the nodal boundaries with values between 0 and 1, not inclusive. The starting location for this list is based on the order in which the underlying centerline object was generated. If multiple line objects were used to define the centerline, or if there is any confusion over the order, select the pipe and list in the command prompt, or use the Thermal Desktop Model Browser. Another hint of directionality is provided by the FloCAD pipe ID, which appears near the “start” end of the pipe. For Heat Pipes, a minimum of two subdivisions for Pipe Length must be used (for VCHPs, the start end of the pipe is the location of the reservoir: define the pipe from the condenser to the evaporator). The should use only as many subdivisions as required by the analysis and not the number required to capture the shape of the centerlines. The lumps will be created along the center lines and the paths are drawn as straight lines between the lumps. If the pipe is following a curve or makes turns, the paths as drawn will therefore not follow the pipe, but will appear to “take a shortcut.” Despite this graphical depiction, the data for a path will still take into the actual pipe centerline. 5.4.4

Pipe Attributes

The Pipe Attributes tab of the Ripe Edit Form controls many of the pipe features (Figure 5-27). The options on this form are:

Figure 5-27

Ripe Edit Form Dialog Box Pipe Attributes Tab

Submodel Information Fluid Submodel. The fluid submodel in which all lumps, paths and ties are placed. Cond. Submodel. The thermal submodel in which all pipe wall conductors are placed.

5-46

Fluid Models

Outer Material Properties. The material or property alias used for pipe wall node capacitance and pipe wall conductance (axial, circumferential, and radial, as required by the pipe definition). If Use Same ID’s on both sides (Section 5.4.7) is unchecked, then this is the material used for the outer nodes and a fraction of the conductances, as defined by the Inner Wall Thickness Fraction. Inner Material Properties. This option is made available by unchecking Use Same ID’s on both sides (Section 5.4.7). This defines the material or property alias used for pipe wall node capacitance and pipe wall conductance (axial, circumferential, and radial, as required by the pipe definition). Inner Wall Thickness Fraction. When Use Same ID’s on both sides (Section 5.4.7) is unchecked, the may choose different materials for the inner and outer portions of the pipe wall. The Inner Wall Thickness Fraction can be modified to specify the thickness of the inner material in proportion to the overall thickness of the wall. Lump Type. The chooses the pipe-defined lumps to be Time Dependent (TANK) or Time Independent (JUNCTION). Tube Type. The chooses the paths to be Time Dependent (TUBE) or Time Independent (STUBE). Wall Node. The chooses the wall nodes to be Diffusion or Arithmetic. If Diffusion is chosen, the nodes may still be defined as arithmetic if the material density and/or specific heat are zero, or the pipe wall thickness is zero. Twin Paths. Checking this option defines all paths in the pipe as twinned or in phasic nonequilibrium. Twin Lumps (and Ties). Checking this option defines all lumps and ties in the pipe as twinned or in phasic non-equilibrium.

Wall Geometry The Wall Shape drop-down list provides the options that are available in the Wall Geometry section of the Pipe Attributes tab. The Wall Geometry defines the flow area and hydraulic diameter for the flow and the thickness of the wall for node capacitance and conductance in the pipe wall. Use a wall thickness of zero to create wall nodes but to neglect node mass, axial conduction, radial conduction, and circumferential conduction. Use the Fluid Pipe without Wall option to model an adiabatic pipe, or to add heat transfer independently. The options and available parameters are: Standard Pipe. For a Standard Pipe, the selects the Pipe Type, the Schedule, and the Nominal size from the available drop-downs. The Inner Diameter, Outer Diameter, and Thickness will be set based on the selected standard. In the Pipe Type list, any specified materials are only in reference to the standard and not automatically transferred to the Inner and Outer Material Properties defined in the Submodel Information section of the form.

Fluid Models

5-47

Rectangular Pipe. For a Rectangular Pipe, the specifies the Width and Height of the flow area and the Thickness of the wall. The Width and the Height specify the inside dimension of the rectangle. The Thickness is applied outside of the width and height. The width and height values will then be used to calculate the correct flow area and hydraulic diameter for the fluid paths. Drawn Outline. This option will be automatically selected if a cross section shape curve is selected when creating the pipe (Section 5.4.1). If the pipe was created without a cross section shape curve, the can select this option. When the Wall Shape is changed to Drawn Outline, the will be returned to the graphics area to select a previously created shape (a closed curve or polyline) used for the shape of the flow cross section. The Line must be closed and planar. It can be drawn with any orientation in the drawing, and will be translated to the start point of the pipe centerline, and then rotated once such that the normal of the shape plane is tangent to the centerline. Grip point editing can be performed to rotate the shape to any orientation about the centerline. The center of the shape (computed as the midpoint of the start- and mid-point of the shape line) will be located at the centerline. All sections of the curve should be within direct line-of-sight to the centerline to allow calculation of the flow area and perimeter. After selection, the can change the shape curve by selecting Replace beside the Line field. The shape curve specified in the Line field defines the shape of the flow area and the Thickness defines the wall thickness. Defined Pipe. The -defined options should be used to generate a circular cross section with sizes not found in the standard tables. Inputting zero flow area (AF) will define the flow area based on a circular cross section with the hydraulic diameter (DH). Non-circular cross sections can specified by providing the flow area (AF) and hydraulic diameter (DH). The Thickness is applied outside of the flow area defined by the flow area and hydraulic diameter. Though these values will be used in the solution, the display for this type of input will be circular. This allows for quick input of a situation where the outside geometry is not critical. For full treatment of the any shape, the -drawn shape can be used.16 Defined Pipe (wall area). The -defined options should be used to generate a circular cross section with sizes not found in the standard tables. Inputting zero flow area (AF) will define the flow area based on a circular cross section with the hydraulic diameter (DH). Non-circular cross sections can specified by providing the flow area (AF) and hydraulic diameter (DH). The X-Section Area is the area of the wall cross section. The wall is applied outside of the flow area defined by the flow area and hydraulic diameter. Though these values will be used in the solution, the display for this type of input will be circular. This allows for quick input of a situation where the outside geometry is not critical. For full treatment of the any shape, the -drawn shape can be used.17

16Caution: The advanced FLUINT input DEFF (effective diameter, used for far-from-circular cross sections) is not calculated by FloCAD. 17Caution: The advanced FLUINT input DEFF (effective diameter, used for far-from-circular cross sections) is not calculated by FloCAD.

5-48

Fluid Models

5.4.5

Heat Pipe Data

The Heat Pipe Data tab is available when Heat Pipe is selected on the Pipe Selection tab. This tab allows the to define the type of heap pipe and the specifications of the heat pipe. Normally, FloCAD will create inputs automatically for the HEATPIPE routine in SINDA/FLUINT, which assumes one wall node at each axial location. Using more than one node in circumferentially in the pipe will invoke the HEATPIPE2 routine. See the SINDA/ FLUINT manual for information regarding modeling heat pipes. One of the key outputs from the heat pipe analysis is the power-length product (QLeff), which should be compared to the vendor-supplied capacity of the pipe as a function of temperature and adverse tilt to adequate design margin. The QLeff product for each heat pipe is stored in an automatically generated SINDA/FLUINT . In order to create a unique name for each heat pipe, the name is constructed by using the letters “QP” followed by the pipe object handle (which can be found after the “::” in the Model Browser). This value can be plotted or tabulated using the postprocessing modules, or can be compared with the pipe capacity using SINDA/FLUINT logic blocks. Before discussing the first section of the form, Type of Heat Pipe, the other form inputs are described below. Wall Node Type. This section allows the to choose whether the heat pipe wall nodes are modeled using Arithmetic nodes or Diffusion nodes18. Abort transients if problems encountered. When checked, the solution will stop if problems related to the heat pipe solution are encountered. Aid stability during steady state (use HTRNOD call). When checked, the solution assumes a constant saturation temperature over a steady state relaxation step. When gas is present in the heat pipe (Fixed Conductance Heat Pipe with NC Gas or Variable Conductance Heat Pipe), difficulties might arise with steady state convergence and oscillations might occur in transients causing small time steps. The heat pipe routines necessarily assume that the gas front location is constant over each interval (where “interval” means either a steady state relaxation step or a transient time step). In some cases, the gas front artificially oscillates between two or more positions. If this numerical instability is experienced, consider checking this box. Reservoir T. This is the current reservoir temperature for a Variable Conductance Heat Pipe. The temperature can be supplied by pointing to a node that represents the reservoir, or at least its average temperature.19 Pressing the Reservoir T button will prompt the to select a node that represents the reservoir. (Make sure the node is visible before initiating

18Diffusion nodes will be used if the wall material density and specific heat and wall thickness or area are all greater than zero. 19In cases where many nodes or surfaces are used to model the reservoir, the should use some method to calculate the average reservoir temperature. One possible way is to create an arithmetic node and connect it to the reservoir surfaces with an area-based, radiation conductor. The radiation conductors will be small and the arithmetic node will be the area-weighted average temperature of all of the surfaces.

Fluid Models

5-49

this action.) Note that the reservoir model itself is otherwise independent of the FloCAD pipe, so this reservoir model can be as simple or complicated as desired, although simpler is better. Wall Material. This drop-down list lists the available material properties and aliases. If the material density or specific heat is zero, then the nodes will be arithmetic regardless of the selection under Wall Node Type. Wall Shape. The options for Wall Shape are Circular Heat Pipe or Drawn Outline. The Wall Shape defines the shape or shape and size of the Vapor Core, or inner wall. For a Circular Heat Pipe, the Vapor Core Diameter must be specified. Selecting Drawn Outline requires selecting a closed, planar line or polyline that defines the shape of the inner wall of the heat pipe. The line object ID will be shown in the Line field. The shape curve can be replaced by selecting the Replace button. More information about Drawn Outlines can be found in Section 5.4.4. Vapor Core Diameter. The vapor core diameter (at the liquid/vapor interface). This will be used to calculate heat transfer “wetted” area for vaporization and condensation assuming a circular cross section [i.e. - for irregular shapes, use the formula for hydraulic diameter: 4*Area/(wetted perimeter)]. If NC Gas is present, it will also be used to calculate pipe volume. Therefore, it need not represent a physical dimension; it can be any value that produces the right area or volume. Wall Input. The Wall Input defines the method of specifying the heat pipe wall. The options are Area Input, Mass/Unit Length Input, and Thickness Input. Selecting one of these will change the field below to the appropriate input. Area Input requires the cross-sectional area of the heat pipe wall and will adjust the wall thickness to match the area based on the Wall Shape. Mass/Unit Length Input will use the input value and the material properties to set the thickness of the heat pipe wall. The Thickness Input uses the input value as the wall thickness. Condensing H. This is the heat transfer coefficient to be applied to unblocked condensing sections Evaporating H. This is the heat transfer coefficient to be applied to evaporating sections (which cannot be blocked by gas). Reservoir Volume. This is the reservoir volume for a Variable Conductance Heat Pipe. Mass of NC Gas. The mass of non-condensible gas within the pipe. Working Fluid. When defining a Fixed Conductance Heat Pipe with Non-Condensible Gas or a Variable Conductance Heat Pipe, the working fluid must be specified. The drop-down allows the to choose from library fluids or browse for a fluid property file. Gas Property. When defining a Fixed Conductance Heat Pipe with Non-Condensible Gas or a Variable Conductance Heat Pipe, the non-condensible (NC) gas must be specified. The drop-down allows the to choose from library fluids or browse for a fluid property file.

5-50

Fluid Models

Type of Heat Pipe Fixed Conductance Heat Pipe. A fixed-conductance heat pipe (FCHP, also called a CCHP for constant conductance heat pipe) with negligible non-condensible gas present. This common modeling choice neglects the presence of any gas, and therefore requires no fluid descriptions. Note that there is no assumed “condenser” or “evaporator” or “adiabatic zone.” Whether or not vaporization or condensation occurs at any one location, and to what extent, is left to the numerical solution. Fixed Conductance Heat Pipe with NC Gas. A FCHP/CCHP with non-condensible (NC) gas present, in order to represent a degraded state. Non-condensible gas can be present in a reservoirless constant conductance heat pipe because of residuals, post-sealing chemical reactions, and even working fluid dissociation (perhaps in a ionizing radiation environment). This gas generation represents the only other degradation mechanism for a heat pipe besides loss of containment (leakage) or other structural damage. When gas is present, it tends to locate at the coldest end of the pipe and block condensation there: the working fluid must diffuse through the NCG in order to condense at the wall. A few common simplifying assumptions are that (1) the gas front is flat, permitting 1D treatment, (2) condensation is completely blocked within the length covered by the NCG (i.e., diffusion is neglected), and (3) gas can only exist at one end of the pipe or the other (it may shift during the simulation, but cannot reside in the middle of the pipe). To use this option, the must specify the amount (Mass of NC Gas field) and type of gas present (Gas Property field) and must also name the working fluid (e.g., ammonia, water, methanol, etc.) in the Working Fluid field. If the required fluids are not available, they can be added using the Browse for New Fluid Property File option in the drop-down lists. Caution should be exercised to make sure the temperatures do not exceed the operating range of the working fluid, or that too much gas is present (the actual amounts are usually extremely small). Unlike a gas-free heat pipe, significant axial temperature gradients can form in the condenser end of the pipe when gas exists in the heat pipe. This means that the axial resolution (Subdivision tab) must be adequate in this region, and that axial conduction might not be negligible: a nonzero cross-sectional area, thickness or mass/length value should be specified. For gas-blocked FCHPs, the gas front location is available through a SINDA/FLUINT using the letters “GP” plus the object handle (found after the “::” in the Model Browser). An automatically generated call to the HPGLOC routine is created to set the GP for each heat pipe. Variable Conductance Heat Pipe. A Variable Conductance Heat pipe (VCHP) is similar to an FCHP with NC gas, except that gas of a known quantity has been added purposely from a reservoir sized to contain most of that gas under warm conditions. Under certain conditions, the gas volume in the reservoir can be expanded (via heaters or warm environ-

Fluid Models

5-51

ment biasing) to block a portion of the condenser, increasing the overall resistance of the heat pipe and therefore limiting the heat flowing through it. Unlike the other two options, a VCHP requires a designated condenser: the portion of the pipe near the reservoir. The reservoir itself may be heated or unheated (“cold reservoir”). The VCHP model requires the reservoir temperature (Reservoir T), the Reservoir Volume, the Mass of NC Gas, the Working Fluid and the NC Gas property. Since the volume and mass of gas are used in calculating the location of the gas front, the accuracy of these two inputs are critical. Often a VCHP will have a transition section between the reservoir and the heat pipe. If the volume of the transition section is significant, it should be included in the gas front calculations. To do this, simply define the reservoir volume on the heat pipe form as the volume of the reservoir plus the volume of the transition section. The is cautioned that significant temperature gradients can exist in the gas blocked portion of a VCHP, therefore the must ensure adequate axial resolution in the blocked zone. To properly locate the gas front it is recommended, although not required, that axial heat flow be modeled in the VCHP (i.e., a nonzero cross-sectional area, thickness or mass/ length value should be specified). For a VCHP with a heated reservoir, it is recommended that the reservoir node be thermally connected to the pipe. To represent this thermal/structural conduction, a conductor should be added between the reservoir node and the first node on the condenser. A common problem when modeling VCHPs is that the reservoir is too large or becomes too cold, or that not enough20 gas has been specified. In such cases, the working fluid collects (and possibly freezes) in the reservoir, drying out the evaporator. This condition normally represents a failure mechanism or at least an off-design scenario, and is therefore not simulated. Instead, warnings are produced and SINDA/FLUINT will abort during a transient. To avoid this abort, uncheck the Abort transients if problems encountered check box, accepting that the resulting predictions may be flawed. As was noted above, the gas location is stored in an automatically generated SINDA/ FLUINT . In order to create a unique name for each heat pipe, the name is constructed by using the letters “GP” followed by the pipe object handle (which can be found after the “::” in the Model Browser). This value can be plotted or tabulated using the postprocessing modules. Its value is the length (in units) of the gas front, as measured from the condenser/reservoir end of the pipe. If this value is negative, this means that the pipe has reversed and the absolute value of the is the distance of the gas front from the evaporator end. For VCHPs, the gas front location is available through a SINDA/FLUINT using the letters “GP” plus the object handle (found after the “::” in the Model Brower). An automatically generated call to the HPGLOC routine is created to set the GP for each heat pipe.

20Too much gas can cause solution convergence problems by blocking off too much of the heat pipe.

5-52

Fluid Models

Variable Driven Heat Pipe. Selecting a Variable Driven Heat pipe allows the to use a symbolic expression to specify the type of heat pipe. The result of the expression determines the type of heat pipe: FCHP for 0; FCHP with NC Gas for 1; and VCHP for 2. To set the values for all necessary fields, cycle through the possible types of heat pipes to set the fields. Want "Hands-On" Information? Two FloCAD tutorial exercises work with heat pipes. Refer to exercises “Heat Pipe Model” on page 22-23, and “Drawn Shape Heat Pipe” on page 22-85. 5.4.6

Ties

The Ties tab of the Ripe Edit Form (Figure 5-28) allows modifying the general behavior of tie generation. The drop-down at the top of the form has two self-explanatory options: Generate Ties and No Ties Generated. After turning off automatic generation of ties, other tie types can be created manually by the . If Generate Ties is selected, the can scale the UA (convection conductance) factor that is used for the automatically generated ties. Ties are generated to the inside surface of Fluid Pipes with Walls (including insulation applied to the inside using the Insulation tab) or the selected faces of surfaces selected for Fluid Pipe with Surfaces for Wall.

Figure 5-28

Ripe Edit Form Dialog Box Ties Tab

Entrance Heat Transfer Effects (HTCENTR) The button in the section of the form will read either Disabled Entrance Effect or Enabled Entrance Effects. This button behaves like the Enabled/Disabled option described in Section 2.10.8. When enabled, a call to HTCENTR (see SINDA/FLUINT manual) is made with the appropriate settings for the arguments. The distance from inlet is measured from the start of the pipe as defined by the centerline curves. The defines the Coefficient C in Turbulent Factor (CEE) and Exponent in Turbulent Factor (EN).

Fluid Models

5-53

5.4.7

Node Numbering

The Node Numbering tab (Figure 5-29) in the Ripe Edit Form dialog box controls the node numbering of the pipe. The first option on this form is Use Same ID’s on both sides. When this option is checked, the pipe wall will have one node in the radial direction of the pipe wall; the form displays numbering options for Both Sides; and only one material can be specified for the wall on the Pipe Attributes tab (Section 5.4.4). When the option is unchecked, the pipe wall will have two nodes in the radial direction of the pipe wall; the form displays numbering options for Top/Out and Bottom/In; and Inner and Outer material properties can be specified for the wall on the Pipe Attributes tab (Section 5.4.4).

Figure 5-29

Both Sides

Ripe Edit Form Dialog Box Node Numbering Tab

or Top/Out and Bottom/In

In this section of the form, the chooses numbering options for the body of the pipe. The body nodes define the entire wall of the pipe. Submodel. Select or type the thermal Submodel for the nodes of the pipe body.

Use Start ID. Start the node numbering of the pipe body from the specified number. Use List. Use the numbering provided in the list. The list must have a number of values equal to the product of Pipe Length subdivisions and Pipe Circumference subdivisions as specified on the Subdivision tab (Section 5.4.3).

5-54

Fluid Models

End Nodes The End Nodes are used solely for connecting pipes (Section 7.16); therefore, they are arithmetic and are not tied to the pipe fluid. The fields in this section are the same as similar fields in the section defined above. The Upstream IDs are used at the beginning of the pipe and the Downstream IDs are used at the end of the pipe as defined by the centerline curves. If the Use List option is selected for the End Nodes, only a single value of the list is displayed on the dialog box due to space requirements. The list can be scrolled using the arrow buttons. The entire list can be displayed by double clicking inside the list field, which will open a new window to edit this list. The list should contain the same number of IDs as the value of Pipe Circumference subdivisions specified on the Subdivision tab (Section 5.4.3). Note: The should that the end node IDs are not included in the list for the body nodes when the number of subdivisions is changed. Heat pipes do not have end nodes and therefore cannot be connected to other pipes.

Vapor Node When Heat Pipe is selected on the Pipe Selection tab, the Vapor Node section will be visible in place of the End Nodes section, described above. An arithmetic vapor node submodel and ID are specified in this section. The is cautioned to that each heat pipe has a unique vapor node (submodel and ID number). 5.4.8

Radiation

The Ripe Edit Form Radiation tab is the performs in the same manner as the Thin Shell edit form Radiation tab. See Section 4.3.1.3 for information on this form. 5.4.9

Surface

The Surface tab in the Ripe Edit Form allows adding a comment (Section 2.10.4) and changing or parameterizing (Section 11) the length of the pipe. When the length is changed a uniform scaling is used, holding the start point of the pipe constant and changing the length of the pipe while keeping the proportional shape. The must ensure that any connected thermal and fluid networks are scaled or relocated to match the new pipe size. 5.4.10

Insulation

The Insulation tab allows defining insulation on the inner and/or outer surface of the pipe. The Insulation form is the same as the form described in Section 4.3.1.6. As opposed to thin shell insulation, pipe insulation does for radial and area changes due to the insulation thickness. Just like thin shell insulation, extra nodes and conductors are created in the model based upon the selections made on this tab. Conductors, ors, ties heat loads, and heaters can all be attached to the insulation nodes, or directly to the wall nodes

Fluid Models

5-55

even if insulation is present. This choice is not made on the pipe form, but rather in the edit forms of the conductors (Section 4.7), ors (Section 4.8), ties (Section 5.3.3), heat loads (Section 4.9), and heaters (Section 4.10). If insulation is placed on the inside of a pipe, the thickness of that insulation is not used in the hydraulic calculations. 5.4.11

Advection

The Advection form allows defining material movement within the pipe wall (a turning pipe or extruding wall viewed from the perspective of the extruder). See Section 9.5 for more information about advection. The options on the form are: Inside to Outside. Selecting this option generates one-way conductors from the inside of the wall to the outside of the wall. This option requires the Use Same ID’s on both sides option to be unchecked on the Node Numbering tab (Section 5.4.7). A negative value defines the flow from outside to inside. Inlet Face to Outlet Face. Selecting this option generates one-way conductors from the inlet face of the pipe to the outlet of the pipe as defined by the centerline curves. A negative value defines the flow from outlet to inlet. Circumference Start to End. Selecting this option generates one-way conductors in the circumferential direction (clockwise looking at the inlet of the pipe as defined by the centerline curves. This option requires a Pipe Circumference subdivision value greater than one on the Subdivision tab (Section 5.4.3). A negative value defines the flow in the counter clockwise direction, looking at the inlet of the pipe.

5.5

System-level Heat Exchangers (HXs)

System-level heat exchanger (“HX”) elements provide a way to simulate a heat exchanger at a very high level, where the details of the heat exchanger construction are not needed as long as top-level metrics (e.g., effectiveness) adequately describe the performance. This top-level approach treats the heat exchanger as a simplified component. Such treatment is appropriate when the focus is on the system level instead of on the heat exchanger itself. These methods are only marginally appropriate when one or more sides of the heat exchanger contains two-phase flow, or when short time-scale transient responses are required, or when strong fluid or material property gradients exist (due perhaps to large pressure drops, to temperature-sensitive transport properties, etc.).

5-56

Fluid Models

If any of the above exclusions are present, or when the focus is on the design of the heat exchanger itself, then a detailed thermal/fluid model should instead be considered using specific thermal geometry and FloCAD Pipes.21 Analogously, a fin (extended surface) could be modeled at a system level as a root node with a fin efficiency, or it could be modeled at a detailed level as a FD/FE surface or solid. FloCAD HX elements cause an energy exchange to occur between two lumps, but they are very different from an FTIE. For one, the two lumps are only referenced indirectly. The HX element actually connects two FLUINT paths,22 with the heat exchange happening between the current exit lumps of those paths: if either path changes flow direction, the energy exchange will happen between a different pair of lumps. Each path therefore defines one inlet and one exit state, plus the flow rate through that side of the heat exchanger. While any path may be used to define a heat exchanger, the flow resistance (and perhaps inertia) of each age should be represented by that path. Therefore, the tube and STUBE connector choices are the most logical path types, with LOSS and TABULAR paths also being commonly used. Twinned paths should not be used, and the endpoint lumps of these paths (especially the exit lumps) should not be twinned tanks. FloCAD HX elements generate logic that includes a call to the SINDA/FLUINT HXMASTER utility. Refer to the HXMASTER documentation in Section 7 of the SINDA/ FLUINT ’s Manual for more information. Note that HXMASTER’s MODE_S is set to 1 for all FloCAD HX elements, which means that the exit lumps’ temperatures may be prescribed during each solution step using internal calls to HTRLMP/RELLMP. 5.5.1

Creating Heat Exchangers

To create a new heat exchanger (HX), choose Thermal > Fluid Modeling > Heat Exchanger, the toolbar icon in the FloCAD bar, the Heat Exchanger option in the FloCAD section of the Thermal Ribbon, or type RCHEATEX in the command line. The will be immediately prompted to select any two flow paths, after which the heat exchanger edit form will appear, as described later (Section 5.5.2). Once generated, the selection of these two paths can be revised from within the edit form. Any heat exchangers within the model can be listed in Model Browser (List > Heat Exchangers). In TD/FloCAD, end points (lumps and/or nodes) are shown as icons, and links (paths, ties, conductors, FTIEs, and IFACEs) are depicted as line segments between them. The HX element is unique in that it interconnects two links (specifically, two flow paths).

21An intermediate step for parallel and counterflow heat exchangers is to subdivide them into multiple HX elements in series. This usage is simplified by exploiting each segment’s description as a library entry (Section 5.5.5). 22These paths may be of any type (e.g., tube, SetFlow, etc.), and may be in the same or different fluid submodels.

Fluid Models

5-57

Therefore, the HX network icon is not explicitly placed in the drawing by the . Instead, FloCAD creates an icon that appears to hover between the two paths that are currently named by that HX element. If the endpoint lumps are moved, the HX icon will move too. The HX icon consists of two linked cylinders (Figure 5-30). While any unique geometry shape would suffice, this icon was chosen since it is representative of two linked flow ages. Of course, the use of two cylinders does not imply that the heat exchanger itself is anything like two parallel cylinders: it could instead be concentric ducts, a multi- shell-and-tube heat exchanger, etc. The details of the heat exchanger geometry need not be defined at the system level. This distinction between a graphical representation and the real geometry is important because the HX icon can be post-processed by the colors of the four nearest endpoint lumps’ temperatures. This representation is an abstraction, since the cylinders are not actual TD cylinders, and likely do not represent gradients between sides of the heat exchanger. Nonetheless, this usage makes it easy to visualize the inlet and outlet temperatures for each side. The “diameters” of HX icons can be resized within the Thermal > Preferences menu selection (Section 2.7.3).

Figure 5-30

5-58

System Level Heat Exchanger Icons (postprocessed)

Fluid Models

5.5.2

HX Inputs and Operating Modes

To edit an HX element, select its icon on the drawing and choose Thermal > Edit..., or locate it within the Model Browser and choose Edit. One such representation of an HX edit form is shown in Figure 5-31.

Figure 5-31

System Level Heat Exchanger Edit Form

Each HX element should be given a unique name, both to locate it more easily within the Model Browser, and to help interpret any SINDA/FLUINT error messages that it might generate. Each HX should also be given a unique prefix. See Section 5.5.3 for more details on the set of SINDA/FLUINT s that generated by each HX element. An HX element’s performance may be specified in one of seven ways: • Effectiveness: the ratio of total energy actually transferred over the maximum possible energy that theoretically could be transferred. • Overall NTU: the number of heat transfer units. NTU is a dimensionless expression of the heat exchanger’s size, and its exact definition and usage are best left to an undergraduate heat transfer textbook. If NTU is chosen as an input method,

Fluid Models

5-59

then the must also define the type of heat exchanger being used, as described later in this subsection. • Overall UA (UAtot): the overall heat transfer coefficient that represents the heat exchanger. If UAtot is chosen, then the must also define the type of heat exchanger being used, as described later in this subsection. • Path 1 Outlet Temperature: In a design or sizing mode, the does not specify how well a specific heat exchanger works, but what its intended purpose is instead such that the required effectiveness (or NTU or UAtot) can be calculated. One way to specify the design’s purpose is by prescribing the desired outcome in of an exit temperature. However, that the temperature specified may not be possible to achieve. The program will automatically limit the outcome based on the range of possible powers, from minimum (zero) to maximum (as if efficiency were unity). Furthermore, the specified temperature should not exceed the range of allowed temperatures for the working fluid. This option specifies the desired outlet temperature, in units, of the exit of Path 1. Note that if the flow rate in Path 1 reverses, the lump whose temperature is being controlled will change accordingly • Path 2 Outlet Temperature: the temperature at the outlet of Path 2 (see above). • Transferred Power: Another “design/sizing mode” is to specify the total power (heat rate) that should be transferred. Again, this power level might not be achievable, and if so the program may limit the total amount transferred to the actual possible. Note that the sign of this input is important: positive means thermal power flows from Path 1 to Path 2, and negative means the power flows from Path 2 to Path 1. • Transferred Power divided by (Thot-Tcold): The power divided by the inlet temperature difference (“Titd” in literature). Despite the re-use of the term “Power Transferred,” this input is not a design mode option, but rather a performance metric. The input is always positive, and has units of heat rate per unit degree: the same units as UAtot, in fact. Note that this input option is provided since it is popular in the electronics industry for describing heat sinks. However, its use is not recommended since it does not represent a safe or complete method of describing performance compared to the alternatives: effectiveness,23 NTU, and UAtot.24 The above values may be specified as numeric constants or expressions, and the expressions (including single definitions) can be updated within the SINDA/FLUINT run. Use the Value/Expression option for this mode. 23Divide P/Titd (the power divided by inlet temperature difference), by Cmin (the smaller of the mass flow rate times the specific heat for Path 1 and Path 2) to yield effectiveness. 24http://www.thermalfluidscentral.org/journals/index.php/Heat_Mass_Transfer/article/view/140/208

5-60

Fluid Models

Alternatively, the effectiveness, temperature, etc. can be specified as tabular functions of mass or volumetric flow rate of one or both paths, using Tables (see Section 5.5.4). If either NTU or UAtot is chosen as the method for describing the heat exchanger’s performance, then the must supply additional information concerning the type of heat exchanger (see bottom left of the HX edit form). For parallel or counterflow heat exchangers, no additional specifications are required. For cross-flow heat exchangers, the should identify which, if any, of the paths represents mixed flow (usually the choice when a single, open fluid age exists on that side) versus unmixed flow (usually because the FloCAD path represents multiple parallel ages). “Mixed” means that the fluid stream is not separated via baffles, tubes, etc. and can move laterally (perpendicular to the net flow velocity). For flow over a tube bank, for example, the tube side is unmixed but the external side is mixed. If instead the tubes were embedded in common fin plates such that the external fluid cannot mix laterally (in the tubewise direction) as it flows between each tube, then both streams are considered unmixed. For shell and tube heat exchangers, the integer number of shell es is required, and this value must be 1 or larger. The tube es will then be assumed to be twice the number of shell es (e.g., if 2 shell es are defined, the number of tube es is assumed to be 4). In other words, the tubes are assumed to enter and exit the same end of the shell; single tube es (i.e., the tubes exiting on the opposite end of the shell) are not ed. If multiple shell es are employed, note that the input or calculated NTU is the total overall NTU for the whole heat exchanger: the NTU for each shell can be calculated by the separately as the total divided by the number of shell es. If the same heat exchanger is to be used in more than one model drawing, or multiple times within a single model, then the HX Libraries feature (Section 5.5.5) is applicable. This feature is particularly helpful if a single parallel or counterflow heat exchanger has been subdivided into multiple identical sections, perhaps to better capture temperature variations or shorter time scale transient events. 5.5.3

HX s

For the seven methods used to define the FloCAD HX element’s performance, seven s are automatically generated containing the corresponding values. For example, if “myHX_” were input as the prefix, the following s would be generated and updated such that they can be postprocessed or referenced in logic or expressions: myHX_EFF ...... the current effectiveness myHX_NTU ...... the current overall NTU myHX_UAtot .... the current UAtot myHX_TOUT1 .... myHX_TOUT2 .... myHX_POW12 .... myHX_Pitd .....

Fluid Models

the current outlet temperature of Path 1 the current outlet temperature of Path 2 the current power flowing from Path 1 to 2 the current power divided by the current inlet temperature difference 5-61

Six of these s will produce new information, but the one that corresponds to the selected input value will be somewhat redundant: an echo of the input value. Nonetheless, for tabular inputs, this “input echo” provides a mechanism to track and the results of the array interpolation and application of scaling factors (see Section 5.5.4). When the outlet temperature or transferred power is specified (“design mode”), the output effectiveness, NTU, etc. provide the sizing data required for the heat exchanger. Since the type of heat exchanger is not a required input unless either NTU or UAtot were specified as input, please note that the output values of NTU and UAtot will correspond to an assumed type: a counterflow heat exchanger.25 5.5.4

Working with Tabular HX Inputs

Occasionally, performance data for a specific heat exchanger is available as a table of values, perhaps effectiveness or NTU as a function of the volumetric flow rate of one or both paths. To choose this option, select Table instead of Value/Expression, and the portion of the HX form to the right becomes available, as shown by the example in Figure 5-32. The

Figure 5-32

Table Data on Heat Exchanger Edit Form

would then choose whether the table is bivariate (a function of both path’s mass or volumetric flow rate) or doublet (a function of just one path’s mass or volumetric flow rate). The choice of units (mass or volumetric flow rate) is available for each path independently. An optional scale factor is available for each path that can include unit conversions if the model-level choices do not correspond to the table data. The path’s current flow rate will be divided by this value before the interpolation is performed during the SINDA/FLUINT run. The units of the table itself are determined by the Specified by: choice, and again if those model-level units are not applicable, a scale factor that can include unit conversions is also available. All interpolated table values will be multiplied by this scale factor before being used by SINDA/FLUINT. It might be also appropriate to use this table output scale 25For different types of heat exchangers, refer to the CMINMAX and EFF2NTU utilities in SINDA/FLUINT.

5-62

Fluid Models

factor as an uncertainty factor on the table itself, perhaps multiplying the table output by a symbol or which varies from 0.9 to 1.1 between runs, for example. A scale factor less than unity may also be used as a first approximation of fouling. Once the units and table type are selected, if any available tables ( Arrays) have already been input that match those specifications, they will appear in Table Name: pulldown and can be selected. Otherwise, choose Add Table to create a new table. In the Array form that appears, the will be prompted to provide a name and array ID, and perhaps a submodel into which the array should be placed. For FloCAD HX, the appropriate portion of the form (either Doublet Array or Bivariate Array) will be available as shown in Figure 5-33 (for the bivariate array example started in Figure 5-32). Do not change the type or units of the array within this form, or it will break the connection to the HX element since the units and type will no longer match.

Figure 5-33

Array Edit Form Showing Double and Bivariate Options

Pressing the Edit... button goes to the appropriate Table Input form (Bivariate or Doublet), with a bivariate form (continuing the same example) shown in Figure 5-34. Notice that the choices made in the original HX Element form (i.e., units, which path is X, which path is Y) have been carried down to this form, as shown in the instructions at the top. If the data is being copied and pasted from another source (perhaps an Excel file) and that source does not match the format specified, stop now and go back and correct the heat exchanger form, otherwise either the data will be used incorrectly or the connection between the table and the HX element will be lost due to the mismatch in units or table type. Otherwise, as long as the format, units, and type are not changed, the data itself can be changed by either choosing Edit Table from the heat exchanger form, or by accessing the array from the Model Browser, or as a Logic Object within the Logic Manager.

Fluid Models

5-63

Figure 5-34

5.5.5

Example Bivariate Array of Effectiveness vs. Path Volumetric Flow Rates

Working with HX Libraries

Once an HX description has been completed, it may be saved and re-used in the same model, or in a different one. Collections of such descriptions, or Libraries, may be created such that you can choose from a list of pre-defined HX elements when creating a new heat exchanger. Libraries are especially helpful if the subdivides a parallel or counterflow heat exchanger into smaller identical segments in order to better capture gradients in time or temperature. To save an HX definition to the library: 1. Edit or create a heat exchanger. 2. Complete the definition of the heat exchanger. 3. Type a name into the Name field. The name should be unique. This is the name that will appear in the Model Browser. 4. Select a library or create one using the Browse button. 5. Select the Add HX to Library button and provide a library listing name. This is the name that will appear in the Library Selection drop-down. If the name is already in use the can change the name for a new listing or repeat the OK button to replace the current listing. The library is stored in a XML file type that just contains ASCII text. The command RcHxLib will bring up a form that allows you to see what the library contains and rename the items in the library. To use a heat exchanger from a library: 1. Edit or create a heat exchanger. 2. Select a library or create one using the Browse button.

5-64

Fluid Models

3. Select the desired heat exchanger from the library using the Library Selection drop-down list. 4. Select Use Selection.

Figure 5-35

5.6

Library Section of the Heat Exchanger Edit Form

Capillary Evaporator Pumps (CAPPMP)

Capillary pumps, only applicable in two-phase single-constituent submodels, are used to model capillary pumping or condensation through a porous structure. This macro generates a junction between two NULL connectors: the simplest way to think of a CAPPMP model is a CAPIL connector with a junction in the middle. The presence of the junction allows heat to be added or removed from a CAPPMP model, whereas a CAPIL connector is adiabatic like any path. This junction has no other physical meaning, and the state of the junction is largely meaningless. If the heat rate into the junction is zero, the CAPPMP behaves exactly like a CAPIL connector. Unlike a CAPIL, the CAPPMP is direction sensitive. If the heat load into the device is positive, the CAPPMP simulates an evaporator-pump. Liquid at the designated liquid end (i.e., not the designated vapor end) can be evaporated and forced into the designated vapor end, performing pumping by capillary action. If the pressure drop is too big or a capillary interface cannot otherwise exist, the device will deprime. This will generate error messages if the flow rate through the junction suddenly drops yet the heat load continues to apply. If the junction heat load is negative, any vapor present in the designated vapor side will be condensed, and any liquid will be allowed to to the liquid side if it is at a lower pressure. Because the junction within a CAPPMP macro is really only a location into which heat is added or subtracted, and has no physical significance, the heat rate into the tie is not simply the product of the UA coefficient times the temperature difference between the node and the junction. When primed, the saturation temperature is used instead of the junction temperature. When deprimed, the tie will spontaneously jump to the designated vapor lump and become an HTU tie with the same UA value. This usually eliminates the error messages that

Fluid Models

5-65

would result when a CAPPMP deprimes with a constant heat rate still being applied. If the CAPPMP reprimes, the tie will return to the central junction, and once again become an HTM tie. In either mode, the UA value can be modified. For more information on model capillary pumps, guidelines, restrictions, alternative, etc. please see Section 3.9.3 of the SINDA/FLUINT ’s Manual.

5.6.1

Creating a Capillary Pump

1. Issue Capillary Pump command by preferred method: • Command: RcCapPmp 2. Select the lump from which flow is expected. 3. Select the lump to which flow is expected. 4. Select a node for the tie, if required. After completing the selections the CAPPMP objects will be drawn.(Figure 5-36)

Figure 5-36

5-66

Capillary Pump (CAPPMP) Display

Fluid Models

While the CAPPMP object appears to have two paths, a junction and a tie, these items are only created in the SINDA/FLUINT input files. These items are not individually selectable or editable within Thermal Desktop. 5.6.2

Editing a Capillary Pump

1. Select the Capillary Pump device. 2. Issue the RcEditAny command to open the CAPPMP Macro Edit Form dialog. (Figure 5-37)

Figure 5-37

Fluid Models

CAPPMP Macro Edit Form dialog

5-67

Enabled/Disabled. See Section 2.10.8. CAPPMP Id. This is the unique macro ID. This number is automatically generated but can be overridden by the . Submodel. This shows the fluid submodel of the lumps selected when creating the CAPPMP. Comment. See Section 2.10.4 Upstream lump. This shows the ID of the From lump selected when creating the CAPPMP. The lump can be changed by selecting the Reselect button immediately to the right and choosing a different lump. Downstream Lump. This shows the ID of the To lump selected when creating the CAPPMP. The lump can be changed by selecting the Reselect button immediately to the right and choosing a different lump. Tie Node (none for no tie). This shows the ID of the tie node selected when creating the CAPPMP. The node can be changed by selecting the Reselect button immediately to the right and choosing a different node. Upstream/Downstream Vapor Lump. “vapor” side: CAPPMP endpoint lump (ulid or dlid) at which heat transfer occurs (location of deprimed tie for TIE option) Add Code. See Section 2.10.10

CAPPMP Data Capillary Flow Conductance (CFC). CFC is the capillary flow conductance through the entire age. Capillary Radius (RC). The effective capillary radius of the meniscus if it were two-dimensional. For circular tubules, RC is simply one half of the diameter. For a uniform (isotropic) wick, the capillary radius should be available from test data, vendor data, or estimates. For a thin slot, RC is equal to the slot width. Non-positive RC values are interpreted as perfect capillary devices capable of withstanding or producing an infinite pressure gradient. This is useful for preliminary design work or for controlling steady-state solutions to prevent deprime. Capillary Priming Dryness, Upper Quality Limit (XVH). XVH and XVL represent the limits of a hysteresis cycle. An adjacent lump with a quality higher than the current value (XVH or XVL, depending on past history) is considered dry enough to permit the device to prime, or perhaps too dry to permit priming, depending on whether the lump has higher or lower pressure than the opposite lump. Capillary Priming Dryness, Lower Quality Limit (XVL). XVH and XVL represent the limits of a hysteresis cycle. An adjacent lump with a quality higher than the current value (XVH or XVL, depending on past history) is considered dry enough to permit the device to prime, or perhaps too dry to permit priming, depending on whether the lump has higher or lower pressure than the opposite lump. Temperature. This is the initial temperature for the junction to be created.

5-68

Fluid Models

Pressure. This is the initial pressure for the junction to be created. Heatload. This is the heat load applied to the junction of the capillary pump. In general it is recommend this be left as zero and the tie option be used, as discussed in Section 5.6.1. Quality. This is the initial quality for the junction to be created. The radio button to the left can be used to select quality over void fraction. Void Fraction. This is the initial void fraction for the junction to be created. The radio button to the left can be used to select void fraction over quality. Initial Flow Rate (FR). This is the initial flow rate in the capillary pump. Tie UA. The initial heat transfer conductance for the tie (if tied to a thermal node). Generated Junction Id. This is ID number for the junction which will be created as part of the macro. This number is automatically generated but can be overridden by the . Generated Path 1 Id. This is ID number for the path from the upstream lump to the generated junction which will be created as part of the macro. This number is automatically generated but can be overridden by the . Generated Path 2 Id. This is ID number for the path from the generated junction to the downstream node which will be created as part of the macro. This number is automatically generated but can be overridden by the . Generated Tie Id. This is ID number for the tie which will be created as part of the macro. This number is automatically generated but can be overridden by the .

Additional Input The Additional Input allows the to add in advanced features as text in SINDA/ FLUINT input file format.

5.7

FK Calculator

The FK calculator provides access to advanced loss control features and the database of internal components the can select to automatically compute the additional K-factor losses (fractions of dynamic head) in most types of paths. Keep in mind that K-factors are appropriate for turbulent flow only. The FK Calculator dialog box, shown in Figure 5-38, contains a summary of the current path. The Advanced Effects field allows the to turn off inclusion of any path into a SINDA/FLUINT duct (LINE or HX) macro. Whereas most paths do not take into the velocities of paths up or downstream of themselves, duct macros treat groups of paths as a continuous flow age in which velocity gradients are significant. This distinction enables FLUINT to perform additional momentum calculations including automatic calculation of AC and FG factors.

Fluid Models

5-69

Figure 5-38

FK Calculator Dialog Box

FloCAD will check all tubes and stubes to see if they can be aggregated into a duct macro. This is usually true for Pipe paths, but for other paths they must meet a list of considerations in order to qualify as a macro. First, the set of paths under consideration must be either all tubes or all stubes, but not a mixture of the two. Furthermore, all lumps between those paths must be of the same type: either tanks or junctions, but again not a mix of the two. The paths must all be continuous in the defined flow direction, i.e. the downstream end of one path connects to the upstream end of the next path. Finally, the paths must all be continuous in flow area (within a tolerance of 1% for any non-zero value). The flow area is based on the input values for AF, or AFI/AFJ, and multiplied by the duplication factors (DUPI/DUPJ). The downstream area of each path is compared to the upstream area of the next path to determine if they are to be considered as a continuous duct. The rcMacroList command can be issued at the command line to display a text listing of the lumps and paths in each macro. The display will also color those paths and lumps that are visible in each macro with the same color. The post-processing color bars are used to set the colors for each macro. Turning on lump and path numbering will show the specific macro Id to which each lump and path are associated. The macro Id is found by taking the fractional part of the number and multiplying by 10000. For example, if the display shows 1.021, the lump or path would be a member of macro number 210. The whole number is used to set the color, such that consecutive macros are much less likely to be the same color. Refer to the SINDA/FLUINT ’s Manual for more details on duct macros. If the inlet and outlet area of the path are different (i.e., AFI and AFJ are unequal), the loss term associated with the expansion or contraction can be automatically added in the logic by selecting the check box in the Advanced Effects box. The wall roughness fraction, defined as the ratio of the characteristic length of wall roughness to the effective diameter (which is usually the same as the hydraulic diameter), can be specified by the .

5-70

Fluid Models

The Add Loss button displays the Component Addition dialog box shown in Figure 5-39 from which components can be selected to compute the K-factor for the current path.

Figure 5-39

FK Calculator Component Addition Dialog Box

The K-factors are based on equations provided in the references noted in parentheses: Crane, Idelchik, or Miller. Making a selection from the list of components can be accomplished by double clicking on a component, or by highlighting the component of interest and then clicking on the Select button. This results in the display of the Governing Equation dialog box, which contains the equation or value for that component loss factor as shown in Figure 5-40.

Figure 5-40

FK Calculator Governing Equation Dialog Box

Some components have variables that the can specify to define how the K-factor is to be calculated. Examples of variables include the ratio of valve throat diameter to inlet diameter, the angle at which a valve reduces, or the angle of the flow to the component. The underlying correlations are typically valid only at Reynolds numbers high enough such that the loss is independent of further increases in the Reynolds number. Specific details can be found by consulting the references cited for each component. The losses are all evaluated based on the current path. Effects of path diameter are automatically taken into . If Fluid Models

5-71

the diameter of the path is subsequently changed, the losses will also be updated to reflect the new diameter. The quantity specified for each component is multiplied by its individual K-factor to arrive at the total FK for each path. Keep in mind that K-factor losses are based on turbulent flow and become meaningless in laminar flow.

5-72

Fluid Models

6 External Heating Environments and Orbits Thermal Desktop provides a complete external heating environment definition and viewing facility. Selecting Orbit from the top level Thermal menu will produce the cascading menu shown in Figure 6-1. Throughout this section, the “external heating environment”, “heating rate case” and “orbit” will be used interchangeably, even though only a few of the external heating environments are truly orbits. Orbit definitions are stored in the drawing file along with the model geometry.

Figure 6-1 Orbit Menu

New heating rate cases are created by selecting Thermal > Orbit > Manage Orbits, which then displays the Heating Rate Case Manager dialog box, shown in Figure 6-2. Multiple orbit definitions can be created and saved under defined names using the Add button. If after working with the orbits, and you do not want switch to orbit display mode, select the Done button instead of the Display Orbit button when exiting the dialog.

External Heating Environments and Orbits

6-1

Figure 6-2

Heating Rate Case Manager Dialog

All menu operations using orbits are performed on the current orbit. Select an orbit from the list in the Heat Rate Case Manager dialog, then select the Set Current button to set the current orbit. The current orbit will be edited when Orbit > Edit Current Orbit is selected. The current orbit will be used when heating rate calculations are performed using commands from the menu, and for Model Checks > View Model from Sun/Planet. The current orbit will be displayed whenever any of the Orbit commands are chosen (except for Orbit Display Off). The orbit may be viewed from preset points such as the vernal equinox, orbit normal, Sun, planet north pole, etc. using the options under Orbit > View From. Orbits can be imported directly from other Thermal Desktop models or from files created by selecting the Export button. Import and export functions are discussed in Section 2.10.12. Thermal > Orbit > Display Preferences can be used to toggle items such as planetary

shadow, orbit path, vehicle coordinate systems, etc., on and off in the graphical display. Scaling factors may also be set to adjust the length of the shadow and the size of the vehicle axes (Section 6.2.1). After orbits are defined, and a heating rate environment case computed, heating rate data may be output for use by a thermal analyzer. For more information, see Section 10.3. Want "Hands-On" Information? Chapter 21 "RadCAD® Tutorials" includes exercises that work with orbits. Complete one or more of the following exercises to learn how to apply Thermal Desktop’s orbit commands: Chapter 21.3 "Importing a TRASYS Model and Using Articulators"; Chapter 21.4 "Orbital Heating Rates"; and Chapter 21.5 "Simple Satellite".

6-2


6.1

Defining Heating Environments

A new orbit may be created by selecting Thermal > Orbit > Manage Orbits and then clicking the Add button located on the Heating Rate Case Manager dialog box. The Create New External Environment dialog box shown in Figure 6-3 will be displayed. The Create New External Environment dialog box allows the to define the orbit name and also to specify the type of orbit. The type of external heating can be: • Basic Orbit • Keplerian Orbit • Planetary Latitude, Longitude, Altitude list • Vector List • Free Molecular Heating using a Vector List • Free Molecular Heating with Reference Orbit

Figure 6-3

6.1.1

Create New External Heating Environment Dialog Box - Create New Orbit

Heating Environment Forms

Many of the heating rate case types share common inputs. For example, in most of the heating rate types, the vehicle may be spun around an axis. Likewise, most heating environments specify planetary, solar, and albedo properties. The following sections describe sub-forms (on tabs) used in the dialog boxes for the various heating environments. The tabs described are commonly shared, but may or may not be available for a particular heating rate environment. Following the description of the commonly used input tabs, each heating environment type will be presented in more detail.


6-3

6.1.1.1

Orientation Tab

The orientation of the spacecraft is specified using the Orientation tab, shown in Figure 6-4. This tab is available for Basic, Keplerian, or Planetary environments. In the orbit view, the vehicle’s axes are displayed in color with red used for the X axis, green used for the Y axis, and blue used for the Z axis (XYZ-RGB).

Figure 6-4

Specifying the Orientation of the Spacecraft in Basic and Keplerian Orbits

Pointing and Additional Constraint These two regions allows aligning the entire system using the World Coordinate System. Axis. The positive or negative Z axis can be specified to point directly at a target. Optionally, another constraint can be applied by orienting a second axis. The chosen constraint axis (+X, -X, +Y or -Y) will be rotated about the pointing axis to align as close as possible to the additional constraint target. Choosing N/A reverts to the methods used in versions 5.6 and earlier. If N/A is chosen for the additional constraint for a Basic or Keplerian orbit, the X axis will lie in the plane of the orbit, oriented towards the direction of motion. The Y axis is formed from the X and Z to complete a right handed coordinate system. Specifically, the X axis is formed by the cross product of the Z axis and the orbit normal. If the Z axis and the orbit normal are collinear, the X axis is computed from the cross product of the orbit plane Y axis and the system Z axis. If N/A is chosen for a Planetary environment, the +X axis will be oriented to the East when the +Z axis points to zenith, and to the West when the -Z axis points to Zenith. Specifically, the X axis is computed from the cross product of the North pole vector (plan-

6-4


etary Z axis) with the WCS Z axis. If the Z axis already is pointing North (or South), the X axis is computed from the cross product of planetary coordinate system Y axis and the Z axis. For example, when the vehicle Latitude is +90, the X axis will be aligned with the planet coordinate system X; when the vehicle Latitude is -90, the X axis will be opposed to (in the opposite direction of) the planet coordinate system X. Nadir/Zenith. The Nadir is center of the orbital central body and will be an option in the Basic and Keplerian orbits. Zenith is normal to the planetary surface and will be an option for the Planetary environment (Section 6.1.4). Sun. The center of the Sun. Star. A point in the celestial sphere defined by the Right Ascension, measured from the Vernal Equinox, and the Declination. Velocity vector. The tangent to the orbit path or East for Planetary environment (Section 6.1.4).

Orientation Override Align to Celestial Coordinate System. When checked, the system will be aligned to the Celestial coordinate system. Any pointing and constraint options are ignored. This option is useful when orbital maneuvers are complex or if the spacecraft orientation is defined by data calculated in an external program. The Additional Rotations can be used to perform the maneuvers from a fixed coordinate system.

Additional Rotations After the vehicle coordinate system is aligned using one of the pointing operations, additional rotations may be applied by using the input fields in this region. Use the dropdown lists to specify the axis of rotation. The rotations are applied in order from top to bottom. Expressions may also be used to vary the spacecraft orientation during the orbit. For example, if the model is slow spinning, say one or two revolutions per orbit, then a way to adjust the orientation would be to define an expression for the axis of rotation (see Section 2.10.7 for more information regarding expressions). The expression could simply be “hrMeanAnom” for a single revolution per orbit or “5*hrMeanAnom” for 5 revolutions per orbit. “hrMeanAnom” is an internally-generated symbol based on the current heating environment. See Section 11.1.4 for more internally-generated symbols. 6.1.1.2

Positions

The locations in orbit, or positions, for which heating rates will be calculated are specified using the Positions tab. The Orbit dialog box, shown in Figure 6-5, allows the number of positions to be specified using an equal interval breakdown, or by inputting an arbitrary list of angular positions. The dialog has the following options.


6-5

Use Equal Increments. By selecting this option, the can specify the Start and End angles and the number of equal Increments. The number of primary calculations will be the number of Increments plus one (to capture the start and end positions). Use Positions. This option allows the to generate a list of specific position angles for calculations. Each entry can be followed by <Enter> to move to the next line. The columnar list can be copied and pasted from Excel. Update List. This button will update the Use Positions list with the positions calculated from the Use Equal Increments settings. Shadow Crossings. The Shadow Crossings can be automatically included for Basic and Keplerian orbits. With the Automatically Include check box checked, calculations will be made at the specified Entry and Exit positions. The positions are not added to the positions list, but are displayed in the orbit display. At each shadow crossing position, two calculations are made: one fully illuminated and one fully shadowed. The two heating rates at each crossing position are separated in the time arrays by a small fraction of the orbit period to provide the transient effect. Note that the penumbra and antumbra are not included in the heating rate calculations as they include atmospheric effects that are not included in the calculations. Position Angles. The angles may be input using True Anomaly or Mean Anomaly. True anomaly is the angle, in degrees, formed between the line ing through the center of the planet and the periapsis and the line ing through the center of the planet and the location in orbit. Mean anomaly is scaled to the period of the orbit (360 degrees equals one orbital period) and can be used to specify calculation positions based on time. Thermal Desktop keeps track of the illumination state of each orbit position when generating data for SINDA/FLUINT. Within the planetary shadow, Thermal Desktop will automatically skip albedo calculations, since heating from albedo will be zero. On the sun side of the terminator, solar plus planetshine heating rates are output. On the shade side, only planetshine data is generated. Planetshine heating rate data is considered the same over the small interval used for defining the transition across the terminator. Outputting the data in this fashion will cause SINDA/FLUINT to correctly interpolate heating rates across the shadow boundary in a near step change fashion. Orbit positions are shown in the graphics display with a green arrow at the first position, and a red arrow at the last position. 6.1.1.3

Planetary Data

The planet, moon or star about which the satellite orbits is described by the data input on the Planetary Data tab on the Orbit dialog box shown in Figure 6-6. Parameters may be input directly or set from default values1 using the Reset to dropdown list. The units for planetary data, with the exception of Gravitational Mass, are specified from the current units (see “Units” on page 2-25). Radius of Planet. The mean radius of the planet. 1Wertz, James R., “Spacecraft Attitude Determination and Control”, D. Reidel Publishing Company, Holland, 1995.

6-6


Figure 6-5

Specifying Positions in Orbit for Heating Rate Calculations

Gravitational Mass (GM). The product of the gravitational constant (G) and the planet mass (M). This may also be referred to as the standard gravitational parameter in some references. The Gravitational Mass is always input in km^3/sec^2. Inclination of Equator. This is the angle between the planet’s equatorial plane and its ecliptic plane. Sidereal Period. This is the time for the planet to complete one revolution when viewed from a fixed inertial position.

Figure 6-6

Specification of Planetary Data For an Orbit


6-7

Mean Solar Day. This is the time for the planet to complete one revolution with respect to the Sun. Color. This button allows the to select the color of the sphere used to represent the planet in orbit view. Reset to. The may change the planet or reset the values to default values by selecting a new planet in the Reset to field drop-down list. Selecting the Sun in the Reset to field creates a heliocentric orbit.

Heliocentric Orbit Heliocentric orbits have special rules for the Solar, Albedo, and IR Planetshine data forms in the following section; see Section 6.1.1.4, Section 6.1.1.5, and Section 6.1.1.6, respectively, for more information. Since the Sun becomes the orbital central body, it becomes the “Planet” for purposes of orientation, trackers, and orbit views. 6.1.1.4

Solar Data

Values in the Solar, Albedo, and IR Planetshine tabs are used for outputting orbital heating rate data to SINDA/FLUINT and for generating post processing displays. These values may be changed after heating rates have been computed, and new SINDA/FLUINT data generated by re-outputting orbital heating rates (see “Calculating and Outputting Environmental Heating Rates” on page 10-32). The Solar tab is shown in Figure 6-7. The value of the solar flux may be input as a value or as a table of flux vs. time. The flux value is the solar flux incident upon the vehicle, including diffusion due to distance from the Sun and any atmospheric attenuation.

Figure 6-7

Solar Data Input Tab

If you wish to model the Sun as a finite disc, enter a value greater than zero for the angle subtended by the Sun. The default value of zero for Subtended Angle means that all the solar rays are parallel to the Sun vector. Any surface that is edge-on to the Sun will calculate a zero value for the solar direct incident heating. When working with highly sensitive optical

6-8


equipment, the may wish to for the true angle subtended by the Sun. Using a subtended angle will also cause soft shadows, as penumbra and umbra effects of surfaces2 are captured.

Heliocentric Orbit For heliocentric orbits with the Sun as the orbital central body (selected on the Planet Data tab, Section 6.1.1.3), the Sun is modeled as a finite object with a given surface flux or temperature. The solar surface flux or temperature is set using the IR Planetshine tab, Section 6.1.1.6. The fields on this tab should be set to zero. 6.1.1.5

Albedo Data

Albedo is the energy reflected off of the planet surface to the vehicle. The fraction of the solar energy that is reflected off of the surface is called the albedo factor, but is often abbreviated to simply “albedo”. Albedo input is unitless and represents the fraction of solar energy that is reflected from the planet. If you are given the albedo in of a flux, simply divide it by the solar flux to calculate the input value. The albedo factor may be input as a numerical value, expression, a table of values vs. time, or a table of values vs. latitude/longitude locations on the planet. The input form is shown in Figure 6-8.

Figure 6-8

Albedo Input Tab

Albedo values that are input with respect to latitude and longitude must span a range of -180 (W) to +180 (E) for the longitude, and -90 (S) to +90 (N) for the latitude (Figure 69). Values at the -180 and +180 locations must be identical. Care must be taken to orient the right ascension of the prime meridian to the correct value. The planet will be rotated with orbit time for the calculations using the sidereal period for calculation of albedo. If it

2 The penumba and antumbra effects are not captured for the planet. The partial shadow of the planet includes atmospheric effects and refraction that are not included in the RadCAD calculations.


6-9

is desired not to rotate the planet, use a value of zero for the sidereal period. Values may be verified by plotting in color on the planet surface. See Section 6.2.1 "Orbit Display Preferences" on page 6-29 for more information on displaying the values in color. Note: Albedo should not be defined as a function of latitude and longitude for Basic orbits due to the generic nature of those orbits.

Figure 6-9

Albedo Bivariate Table Input Example

Heliocentric Orbit For heliocentric orbits, Albedo should be set to zero. 6.1.1.6

IR Planetshine Data

IR Planetshine may be input as a numerical value or expression, a table vs. time, or a table vs. latitude and longitude. The values may be input as a flux or a black-body temperature (the emissivity of the planet surface is assumed to be 1). The values, and the associated units, will be converted appropriately when the Input Mode radio buttons are switched. The IR Planetshine input tab is shown in Figure 6-10. The ability for planetshine radiation to be a function of the sun and dark sides is also available. When selected, the can input a different temperature (or flux) for the Sun and dark sides. IR planetshine values that are input with respect to latitude and longitude must span a range of -180 (W) to +180 (E) for the longitude, and -90 (S) to +90 (N) for the latitude. Values at the -180 and +180 locations must be identical. Values may be verified by plotting in color on the planet surface. See Section 6.2.1 "Orbit Display Preferences" on page 6-29 for more information on displaying the values in color. 6-10


Figure 6-10

IR Planetshine Input Tab

Please note that if values for the Sun/Dark side option or for the lat/long option are changed, RadCAD must re-shoot rays for this calculation. This is different than described previously where simply changing the solar flux or albedo value does not require recalculation. For values of albedo and planetshine that do not vary over the planet surface, each ray may be assigned a unit energy, and the result may be scaled after calculations are completed. However, for data that varies over the planet surface, each ray must be weighted according to the location on the planet. In this case, the final result cannot be rescaled with a change in data and must be recomputed. If latitude/longitude data is input, an option exists to define the reference planetshine coordinate system. If the Planet Coordinate System radio button is selected, the coordinate system is defined as the +Z axis extending through the north pole, and the +X axis extending through the prime meridian (longitude = 0). This is the traditional latitude/longitude coordinate system and is useful if temperature or flux data has been computed for the planet surface in an external model or obtained from measured data. Note: IR Planetshine should not be defined as a function of latitude and longitude for Basic orbits due to the generic nature of those orbits. The Subsolar Coordinate System is defined with the +Z axis towards the Sun and the +X axis in the equatorial plane. In this coordinate system, the Sun is positioned at a latitude of +90 degrees. All latitudes less than zero are in the shade. This coordinate system is useful for inputting data computed from empirical or functional calculations based on the angle from the subsolar location. For IR planetshine calculations using the latitude/longitude option, the IR data is not rotated with orbit time. That is, the temperature distribution remains fixed inertially at the time = 0 position. An example showing the planet temperature input as a table using the Subsolar Coordinate System option is shown in Figure 6-11.


6-11

Figure 6-11

IR Planetshine Analysis with Planet Temperature as a Function of Lat/Long

Heliocentric Orbit When a heliocentric orbit is used, the albedo and solar flux are set to zero and the Sun is modeled as a black body with a given temperature, or flux, listed under IR Planetshine. A value for flux must be defined at the Sun’s surface (not incident to the vehicle); the flux will be diffused based on the distance from the Sun provided in the orbit data. Use the following relationship to convert a required incident flux to a Solar-surface flux:

where the G’s are the fluxes and the R’s are the radii of the sun surface and required location. The required location radius is the sum of the orbit altitude and the Solar radius. Even though the IR Planetshine tab is used to input the data for the Sun, heliocentric orbits use the solar optical properties for absorbing, transmitting and reflecting the incident radiation.

6-12


6.1.1.7

ASHRAE Data

The ASHRAE3 tab is available for the Planetary Latitude, Longitude, Altitude environment. The ASHRAE tab allow the use of the ASHRAE Atmospheric Extinction model, also known as the ASHRAE Clear Sky Model. This ASHRAE model was developed for use in fenestration (building window) analysis and is therefore meant to be used for ground-based terrestrial applications. The coefficients can probably be adjusted to match other solar attenuation models.

Figure 6-12

ASHRAE Atmospheric Extinction Modeling

To use the ASHRAE model, check the box at the top of the form and provide the values for the fields. The total solar load is calculated as two components: direct solar and diffuse solar. Each component has coefficients and/or factors use to calculate the component. The available fields for the Direct Solar Component are: • Extinction Coefficient - determines the attenuated solar flux by reducing the exoatmospheric solar flux, defined on the Solar data form (Section 6.1.1.4), as it travels through different lengths of atmosphere based on the Sun and planet locations. The extinction coefficient is used for both direct and diffuse components. Value >= 0 (no extinction). • Cloudiness Fraction - defines the fraction of attenuated solar flux that is not direct. 0 (100% of attenuated flux is applied as direct) <= value <= 1 (no direct solar). The available field for the Diffuse Solar Component is: • Fraction of Solar Scattering - defines the fraction of attenuated solar flux that is scattered. 0 (no diffuse solar) <= value <= 1 (solar). 3 ASHRAE - American Society of Heating, Refrigerating and Air-Conditioning Engineers


6-13

Typical values for these fields can be found in the ASHRAE Handbook or other resources. In the environmental heating calculations, the direct solar component flux is used for the directional solar load and the diffuse solar component is used for the diffuse sky solar, which is evenly distributed above the planet horizon. The diffuse solar component overrides any inputs on the Diffuse Sky Solar form. The direct and diffuse solar component fluxes are additive. Therefore, setting the fraction of solar scattering to 1 will effectively double the attenuated flux. 6.1.1.8

Spin

Heating rates for a spinning spacecraft can be specified using the Spin tab, shown in Figure 6-13. Check the Spin vehicle about spin axis: check box and enter the spin vector in the input fields provided to enable spin. The vector is in the Vehicle Coordinate System after all rotations from the Orientation tab have been applied. In orbit view, the fast-spin axis is shown with the Vehicle Coordinate System as a yellow cylinder.

Figure 6-13

Specifying a Spinning Spacecraft

The spin rate assumes that temperature variations over one spin revolution of the satellite can be ignored. Rays fired toward the heating rate sources are randomly and uniformly rotated about the spin axis, which avoids the need to calculate heating rates for a discrete number of spin angles. Typically, this would be applied for a fast spinning spacecraft. If a spacecraft was spinning at a much lower rate, then the may wish to program that rate in an expression for a rotation on the Orientation tab (Section 6.1.1.1).

6-14


6.1.2

Keplerian Orbit

The dialog box for defining a Keplerian orbit is shown in Figure 6-14.

Figure 6-14

Keplerian Heating Rate Case Dialog Box

The Heating Rate Case dialog box for the orbit is organized with tabs which provide the means for defining the shape of the orbit, planet related data, the orientation of the vehicle, the number of orbit positions for heating rate calculations, and optional spin. The Keplerian Orbit tab shown in Figure 6-14 contains input fields to define the shape of the orbit and its orientation relative to the planet and Sun. Any two of the four parameters on the right side of the dialog box may be used to define the shape of the orbit. The remaining two parameters may be calculated by selecting the Update button. An example of an orbit is shown in Figure 6-15. The following inputs and definitions are used to specify the orbit: • Orbit Inclination - The angle between the orbit plane and the equatorial plane. The angle is formed between the north pole of the planet, and the orbit normal vector. Inclinations less than 90 degrees will have the satellite traveling west-to-east at the ascending node. Inclinations greater than 90, or retrograde orbits, will have the satellite traveling east-to-west at the ascending node. Input is in degrees between 0 and 180. • Ascending Node - The point at which the orbit crosses the equatorial plane, traveling south to north. The line between the ascending node, center of the planet, and the descending node is called the line of nodes. • Right Ascension (R.A.) - The right ascension is the angle measured in the equatorial plane from the Vernal Equinox. The angle is measured counter-clockwise


6-15

North Pole

Orbit Normal

Vernal Equinox

Subsolar point on Planet Periapsis

Ascending Node

Subsolar Point on Orbit

Figure 6-15

RadCAD® Orbit Graphical Display

using the right-hand rule about the north pole of the planet. The input is in degrees. • R.A. of Ascending Node - The right ascension of the ascending node. • Argument of Periapsis - The angle formed between the line connecting the center of the planet to the ascending node and the line connecting the center of the planet to point in the orbit closest to the planet (the periapsis). The angle is measured positive in the direction of motion of the satellite. The orbit begins at the periapsis point: this is Time = 0. See Section 10.1.4.1 "Radiation Advanced Logic" on page 10-24 to offset the orbit time for the solution. • R.A. of Sun - The right ascension of the Sun. The declination (angle from the equatorial plane, positive towards the north pole) of the Sun is automatically computed using information entered on the Planetary Data tab. A R.A.of Sun equal to 6-16


zero degrees represents the vernal equinox. A value of 90 degrees places the Sun at the max declination and represents the summer solstice for the northern hemisphere. At value of 180 degrees represents the autumnal equinox. A value of 270 degrees represents winter in the northern hemisphere and the Sun is located at the minimum declination (the negative value of the Inclination of the Equator). • R.A. of Prime Meridian - The right ascension of the prime meridian of the planet. In GeoLatLong heating environments, the Prime Meridian defines the subsolar longitude. The difference between the R.A. of Prime Meridian and R.A. of Sun defines the subsolar longitude at time zero. If the R.A. of Prime Meridian and R.A.of the Sun are equal, the zero Longitude will be at solar noon (subsolar) at time equal to 0. • Minimum Altitude - The distance between the planet surface and the satellite at the periapsis point. • Maximum Altitude - The distance between the planet surface and the satellite at the apoapsis (farthest away) point. • Eccentricity - The ratio between half the distance between foci and the semimajor axis of the orbit ellipse. • Period - The time to complete one revolution of the orbit. If the Earth is being orbited, the right ascension of the Sun and the prime meridian may be computed by selecting the Calc RA Sun/Prime Meridian button. Enter the desired date and time, and select the Calculate button to compute the right ascensions. Select the Update button to transfer the values to the Keplerian Orbit input form. The date and time used for this calculation would most likely be the mid-point of the solution since the RA Sun and RA Prime Meridian do not progress during the solution.


6-17

6.1.3

Basic Orbit

The Basic Orbit tab is used for defining a circular orbit and is shown in Figure 6-16. A basic orbit is defined with beta angle and altitude. The beta angle is the angle between the solar vector and the orbital plane. Note that the beta angle is a function of the orbit inclination, the R.A. of the Ascending Node, the planet’s equatorial inclination, and the R.A. of the Sun.

Figure 6-16

Input Form to Define a Basic Orbit

The Basic orbit is simply a special case of the Keplerian orbit with the R.A. of the Sun set to zero, and the R.A. of the Ascending Node set to 90. The inclination then represents the beta angle. Want "Hands-On" Information? Create and view basic orbits in Section 21.3 "Importing a TRASYS Model and Using Articulators" on page 21-35 and Section 21.5 "Simple Satellite" on page 21-71. 6.1.4

Planetary Latitude, Longitude, Altitude

The Heating Rate Case dialog box for defining a Planetary Latitude, Longitude, and Altitude case (Lat/Long Input tab) is shown in Figure 6-17. All of the other tabs for defining this environment are the same as used for defining a Keplerian Orbit with the exception of the addition of two new heating sources, Diffuse Sky Solar and Diffuse Sky IR, and the addition of ground emissivity for IR calculations. Vehicle positions for a series of time values are entered using planet-centered values for latitude, longitude and altitude. The altitude is measured from the surface of a perfect sphere using the planet’s radius. The longitude is positive to the East and negative to the West. The coordinate system is positioned such that the Z axis is pointing away from the surface of the planet (up or zenith), the X axis is pointing east, and the Y axis is pointing north. An additional rotation counter-clockwise about the Z axis may also be input. The data in the field may be space, comma or tab delimited. You may also cut and paste data from Excel or other applications.

6-18


Figure 6-17

Heating Rate Case Dialog Box for Geo Latitude, Longitude, Altitude

The correct terminology of the latitude input is “Geocentric”. Geocentric latitude is the angle between a line starting at the center of the planet, to where the vehicle is, and the equatorial plane. Geographic or geodetic latitude is the angle between the local zenith and the equatorial plane. They differ because of the oblateness of the planet being analyzed. Note that for all of the calculations the planet is assumed to be a perfect sphere. Although the term “Geo” is used here for the purpose of explanation, the description applies to all planets or moons being analyzed. If the vehicle is stationary on the planet’s surface, simply use the same latitude and longitude values for each time point. If values are omitted on a line, the last values input will be repeated. Values omitted for the first line will be zero. At least one value representing the time must be input for each line of data. The Sun and the prime meridian for the planet are positioned using the right ascension input boxes at the top of the dialog box. If the planet is set to Earth, these values may be computed directly from a given date and time using the Calc R.A. Sun/Prime Meridian button. The Astronomical Calculator dialog box will appear that allows the specification of a date and time (Greenwich Mean Time). Selecting Calculate and then Update on this dialog box will copy the computed values to the Start Date/Time field on the Lat/Long Input tab. Since this type of Heating Rate Case does not follow an elliptical orbit about the planet, automatic calculation of the planetary shadow entry is not available. Therefore the must take responsibility to ensure the day/night transient is captured to the desired accuracy. When modeling on a planet with an atmosphere, both direct and diffuse components of radiation energy are present. The Solar tab is used to specify the direct solar energy at the surface of the planet for a receiving surface perpendicular to the direction of the Sun. It should for atmospheric attenuation. The diffuse component of solar energy is specified with the Diffuse Sky Solar tab, as shown in Figure 6-18.


6-19

Figure 6-18

Diffuse Sky Solar Input Tab

The diffuse component of solar radiation is uniform over the view of the sky. The may for varying degrees of cloud cover by specifying the appropriate amount of energy for the direct and diffuse components of solar radiation. Modeling on a planet surface with an atmosphere also produces a diffuse component of infrared radiation. The Diffuse Sky IR tab, shown in Figure 6-19, is used to specify the data for this heating source.

Figure 6-19

Diffuse Sky IR Input Tab

The diffuse sky IR component may be specified using a temperature or a flux. The flux is uniform over the view of the sky. For both diffuse sky solar and IR, the view of the sky is dependent on altitude. The higher the altitude, the greater the view of the sky. When modeling a planetary surface heating rate case, six separate heating rate sources may be selected to be computed, post processed, and output to SINDA/FLUINT. Like an orbit, the direct solar, albedo, and planetshine sources may be computed and output. In addition, diffuse sky solar, diffuse sky albedo, and diffuse sky IR components may be computed, post processed and output. See Section 6.1.4.2 "Ground Modeling on a Planetary

6-20


Surface" on page 6-22 for more information. Note: Be sure to check all heating rate sources on the Radiation Analysis Data Control tab (Section 10.1.1) Note that a diffuse sky albedo tab is not present. The parameters for computing this heating rate source are taken from the Diffuse Sky Solar tab and the Albedo tab. 6.1.4.1

Radk Calculations with Planetary Surface Modeling

Because all of the external heating sources, both in the solar and IR wavelengths, are computed as heating rates, the calculation of the infrared radiation exchange factors needs to specify a “Space” node with a temperature of absolute zero, 0 K or 0 oR (set the space node temperature on the Radk Output form, Section 10.1.3.1). In this respect, the framework for performing an analysis on a planet surface with or without an atmosphere is the same as calculating heat loads for an orbiting spacecraft. Only a single radk job in the infrared spectrum is required. Although it may seem strange to have the thermal model radiate to an environmental temperature of absolute zero when it is on a planet surface blanketed by a sky with an effective temperature far above absolute zero, it is indeed necessary for proper heat balance. For a spacecraft in orbit, the view to space is not blocked by the planet when computing infrared radiation exchange factors (radks). Likewise, since the external heat sources in the IR spectrum for a planetary surface model (the planet and the sky) are computed as heating rates, the radiation exchange factors must radiate to a sink at zero temperature. For example, if the IR radiation to the sky were computed with radks, rather than an external heating source, we would see an equation of the form: Qnetsky-i = Ksky-i * ( Tsky4 - Ti4 )

Eq. 1

That is, the net heat absorbed by surface i is the difference between what is absorbed from the sky and what is radiated to the sky: Qnetsky-i = Ksky-i * Tsky4 - Ksky-i * Ti4

Eq. 2

Because the absorbed diffuse sky IR component is computed as a heat rate, and added as an impressed heat load in SINDA/FLUINT, the radk from surface i must radiate to a sink at zero degrees to maintain the proper heat balance: Qnetsky-i = Qabs_skyi - Ksky-i * Ti4

Eq. 3

Qnetsky-i = Qabs_skyi + Ksky-i * ( Tspace4 - Ti4)

Eq. 4

It can be seen that Tspace must be zero for equation 4 to equal equation 3. Radiation exchange can be viewed as two separate “one-way” processes, one a process of absorption and the other a process of emission. A radk conveniently expresses the net exchange of energy between these two one-way processes. However, if the absorbed energy is computed externally, then the radiation emitted by the surface must still be ed for. Using a radk to a zero temperature sink models the energy radiated by the surface. External Heating Environments and Orbits

6-21

6.1.4.2

Ground Modeling on a Planetary Surface

It should be noted that the heating rate calculations involving emission or reflection of energy from the idealized planet surface do not take into shadowing of the vehicle on the planet surface. That is, the planet surface uniformly reflects solar energy and uniformly radiates infrared energy, no matter how the thermal model casts shadows on the ideal planetary surface. In reality, when a shadow is cast on the ground, that portion of the ground does not reflect solar energy, and most likely becomes cooler, also reducing the infrared emission from that location. The temperature change of the planet surface due to shadowing cannot be computed ahead of time as a heating rate calculation, the ground must be included in the thermal model and its temperature computed in SINDA/FLUINT to capture these effects. If it is considered important to capture the effect of shadows on reducing reflected solar energy, then it is most likely important to capture the effect of the reduced temperature. If the ground comes to thermal equilibrium, the net energy leaving the surface as either reflected energy or emitted infrared energy is equal to the incoming solar energy. That is, at steady state, whatever energy is not reflected is radiated as infrared energy. If the ground has an albedo factor of 0.5, as much energy leaves the ground as reflected solar as it does in the infrared spectrum. Nothing special is required when the ground is included as part of the thermal model as far as computing radks or heating rates. Conceptually, you should consider the ground as simply an extension of the thermal model, and only model that part of the ground that is necessary for the desired level of accuracy. It is not necessary for the ground to extend to the horizon. Where the ground ends, the heating rate calculations model the ideal planetary surface. Include geometry that models the ground in the same analysis group as used to compute radks and heating rates for the thermal model. Consider that a view factor between adjacent sides of a cube is approximately 0.2. That is, only about 20% of the energy that leaves one side of a cube hits another side. Radiation effects fall off with the square of the distance, so if the distance between the two surfaces is doubled, the view factor is diminished by a fourth. The effects of a shadow cast far from the model are not as important as a shadow at the base where the thermal model rests on the ground. As a rule of thumb, consider an initial approach to modeling the ground with a radius of two to three characteristic dimensions of the thermal model. If the model is tall, consider the dimension near the ground as the characteristic dimension. For example, use the diameter of a tall cylinder rather than the height. All energy is ed for no matter how large or small a supplied ground model is used. In the region where the ground is modeled, the vehicle does not see the planet. Heating rates calculated explicitly (IR from the ground, direct solar albedo, and diffuse solar albedo) are blocked by the ground surfaces. That energy is ed for by radk exchange with the ground surface, and from reflected direct solar and reflected diffuse sky solar from the ground surface. The ground should be considered an extension of the thermal model necessary to capture the effects of shadowing.

6-22


In the far field, the vehicle receives reflected solar and IR radiation from the ideal planetary surface. This also avoids the problems associated with using a large ground plane. When the heating rate calculation uses the ideal planetary surface, rays are shot from the surface to the planet. Every ray that leaves a surface contributes to the albedo calculation (whether direct or diffuse sky). When a supplied ground surface is used, albedo is simulated by the reflected energy from the ground surface. The reflected energy is computed when the direct absorbed energy is computed for the ground surface, and those rays are then reflected from the ground surface. If the ground surface is large, billions of rays may have to be shot from the ground surface in order that enough of the reflected rays hit the vehicle to produce a low error in the calculation. Most of the rays reflected from a large ground surface do not hit the vehicle, they are reflected back to the sky. Therefore, the most accurate and efficient model will only consider as much of the ground that is necessary to capture the effect of shadows on reducing reflected solar energy and ground temperature. Far field albedo and IR calculations are efficiently handled by the ideal planetary surface, and near field effects are accurately captured by the ground model. IR energy radiated from the planetary surface differs slightly for the Planetary Surface Option as compared to a basic or Keplerian orbit. An additional option is available to include the reflected component of IR sky radiation if the emissivity of the planet surface is less than 1.0. For a typical basic or Keplerian orbit, there is no diffuse sky IR component. The IR flux from the planet is expressed either directly as a flux value or as an equivalent temperature. The equivalent temperature is not necessarily the actual temperature of the planet surface, it is a convenient value to represent the flux leaving the planet surface: Qplanet_IR = s * Tplanet4 Planetary surface temperatures represent an equivalent blackbody temperature and are converted internally to a flux by RadCAD when computing heating rates. However, for the Planetary Surface heating rate case, the emissivity of the idealized planet surface (ground) may be specified. This option is only available when a temperature is used, not for a flux input. When the flux from the ground is specified, it is assumed that it represents the combination of what is emitted from the ground, and what is reflected from the diffuse sky IR component. Likewise, if the Equivalent Black Body option is chosen for temperature input, the temperature is assumed to be that which a calorimeter pointing at the ground would read. The equivalent black body temperature represents the total IR leaving the ground, which consists of both emission and reflected components. Often, data for a site will contain actual historical temperature readings of the ground surface, not equivalent black body representations. In this case, the my select the Actual with Emissivity option for temperature input. The total flux leaving the ground will be computed by RadCAD taking into both the emissivity of the ground, and the value specified by the diffuse sky IR component: Qplanet_IR = eground *s * Tplanet4 + (1-eground) *s * Tsky4


6-23

If the Diffuse Sky IR input is a flux, it will be converted to a temperature for this calculation. If either the planet temperature or the sky temperature is a function of time, the value used for the IR flux from the ground will be computed as a function of time. The form to input Ground IR data is shown in Figure 6-20.

Figure 6-20

IR Planetshine (Ground IR) for Lat/Long case may specify emissivity

Please note that for all calculation and output options, “Ground IR” is the same as “IR Planetshine”. It is renamed in the Planetary Surface heating rate case to emphasis that the ground emissivity option is only available for planetary surface heating, and only with a temperature input. 6.1.5

Vector List

The Heating Rate Case dialog box (Vector Input tab) for defining a Vector List of Sun/ planet orientations is shown in Figure 6-21. This example shows an object on the planet surface (distance ratio is approximately 1) with the solar vector moving from the horizon to vertical. The simply inputs the time, the vector to the Sun, the vector to the planet, and the ratio of the vehicle’s distance from the planet center to the planet radius. The magnitudes of the vectors must be greater than zero and the distance ratio must be greater than one. The input vectors are unit vectors from the spacecraft to the source (sun/planet). The data in the field may be space, comma or tab delimited. You may also cut and paste data from Excel or other applications. This option can be used for calculations of a spacecraft in an arbitrary trajectory, or as a generic interface to external orbit generation programs. All of the other tabs for defining this environment are the same as used for defining a Keplerian Orbit.

6-24


Figure 6-21

Heating Rate Case Dialog Box for Vector List

If a heliocentric orbit is defined by setting the planet data to the Sun, the planet vector is used to show the vector to the Sun, since the Sun is the object being orbited. The solar vector is not used, but still must have a magnitude greater than zero. The graphical orbit display, shown in Figure 6-22 for the vectors provided in Figure 6-21, is different for this option, since there is not enough information defined to display the positions with reference to the planet. The graphical display will show the vehicle’s world coordinate system and will have lines to show the solar and planet vectors: heavy arrows show the current position and lines show the other positions defined in the orbit. The solar vectors are colored yellow and the planet vectors, light blue when the vehicle is illuminated by the Sun. The solar vectors are magenta and the planet vectors, dark blue when the vehicle is shadowed by the planet. Displaying the vehicle in orbit is also different for the vector list option. Instead of showing the analysis geometry relative to the planet, the geometry will be displayed in its world (central) coordinate system and the solar and planet vectors for the current position will be indicated by drawing the vectors as solid shaded arrows. Viewing the orbit from preset vantage points such the ascending node and orbit normal are undefined for the vector list option and will default to one of the principle coordinate axes. 6.1.6

Free Molecular Heating Vector List

The Heating Rate Case dialog box’s Free Molecular Heating tab is used to input data to define free molecular heating and is shown in Figure 6-23. The inputs are simply time, the normalized velocity vector in the vehicle’s coordinate system, followed by the imposed flux at that time. A normalized Solar vector and planet vector are also required for tracker


6-25

Figure 6-22

Graphical Orbit Display for Vector List

positioning; if trackers are not being used, any non-zero vector may be input for each. The data in the field may be space, comma or tab delimited. You may also cut and paste data from Excel or other applications. The free molecular heating calculations do not use any optical properties for the calculation, since the imposed heating is specified directly. The amount of heating depends on the projected area in the direction of the velocity vector. The graphical display for the free molecular heating vector list is the velocity vector represented by a yellow arrow. Displaying the vehicle in orbit and stepping through the orbit positions will display the velocity vector at successive times. Use the IR Planetshine check box for the heating rate calculations, post processing and output to SINDA/FLUINT dialogs.

6-26


Figure 6-23

6.1.7

Heating Rate Case Dialog Box for Free Molecular Heating

Free Molecular Heating with Reference Orbit

Similar to the vector list option, free molecular heating with a reference orbit is also available. In this case, the velocity vectors are supplied by an orbit specified using the drop down menu shown in Figure 6-24.

Figure 6-24

Free Molecular Heating with a Reference Orbit Input Form

The other input for this option is a table of flux vs. time. This table is edited by selecting the Edit Free Molecular Flux Table button.


6-27

The graphical display for this heating rate type shows the vehicle in the reference orbit. Selecting the graphical display of the orbit and then using the list command will list the velocity vector information for the orbit. Use the IR Planetshine check box for the heating rate calculations, post processing and output to SINDA/FLUINT dialogs. 6.1.8

Celestial Coordinate System Location and Orientation

This heating environment is defined in reference to the celestial coordinate system. A generic trajectory can be specified using the Celestial Coordinate System Location/Orientation heating rate case. When this environment is created through the Heating Rate Case Manager or edited, the Heating Rate Case: Celestial dialog opens (Figure 6-25). The vehicle trajectory will be displayed around the planet for verification and post processing.

Figure 6-25 Celestial Coordinate System Location and Orientation Input tab of the Heating Rate Case: Celestial dialog.

Right Ascension Definitions Specified. Selecting Specified allows the to specify the Sun’s location by Right Ascension of the Sun (R.A. of Sun) and the location over the planet by Right Ascension of the Prime Meridian (R.A. of Prime Meridian). The R.A. of Sun and the R.A. of Prime Meridian can be defined as symbols or symbol expressions. The R.A. of Sun and R.A. of Prime Meridian can be calculated from a date and time by selecting Calculate From Date. Using the Calculate From Date option gives static values for R.A. of Sun and R.A. of Prime Meridian; to change the date using symbols, select the Use Date/Time option. Use Date/Time. Selecting Use Date/Time allows specifying individual parameters (year, month, day, hour, minute, second) of the date and time. These parameters can be defined using symbols and symbol expressions.

6-28


Celestial Coordinate System Location and Orientation Table The table must include at least one row. Values may be space or comma delimited and can be brought in from external sources (spreadsheets) using cut and paste. If a few values are repeated, the data should be entered as comma delimited with two sequential commas representing the repeated value. The following line would repeat the X and Rot1 values from the previous line. 100, , 500, 0, , 90, 45

Time. Each row is associated with a time in model units. Each row represents a new heating rate calculation position with the SINDA solution interpolating between calculations for the heating rate at the current solution time. Rows can be fully repeated by only providing a new time. X, Y, and Z. The X, Y, and Z locations are specified relative to the celestial coordinate system (CCS) in the current orbit length units. Rot1, Rot2, and Rot 3. Three rotations can be defined about the system coordinate system axes. The order of the rotations are provided in the Rotation Axes region of the form. Without rotations, the system coordinate system will be centered at the X, Y, Z location and aligned with the celestial coordinate system.

Rotation Axes The Rotation Axes associate the system coordinate system (World Coordinate System or WCS) to the Rot1, Rot2, and Rot3 columns in the

Heating Rate Case Available Tabs The tabs available for this heating rate case are: • Planetary Data • Solar • Albedo • IR Planetshine • Fast Spin

6.2 6.2.1

Displaying Heating Environments Orbit Display Preferences

The visual appearance of the orbit display can be modified by selecting Orbit > Display Preferences. The Orbit Display Preferences dialog box, shown in Figure 6-26, allows visual items to be toggled on and off, and permits size factors and colors to be set for certain items.


6-29

Planet These options apply to the display of the planet. Show Planet. When checked, the planet will be displayed in the orbit. Auto Hide. When checked, the planet is automatically hidden when the model position is behind the planet from the s perspective. Show Continents (Earth). This option only applies when Earth is selected as the planet. A basic orbit is too general for this option to have any meaning; e.g. a zero Beta angle will only be displayed as equatorial. Sun Lighting Effects. When checked, the planet and vehicle display will be illuminated from a source along the solar vector. This is only for display purposes and does not affect the calculations. When checked, the following options determine the objects that are displayed Shadow Grid. Wire frame representation of shadow. Shadow Cone. Shaded surface representation of shadow. Solar Vector/Terminator Markers. Two lines, one drawn from the center of the planet directly in the direction of the Sun, the other through the subsolar point on the orbit. The lines are drawn in yellow along with small circles on the planet surface and at the orbit subsolar point. Celestial Coordinate System. A right handed coordinate system with one axis through the north pole (Z), one through the Vernal Equinox (X), and one pointing at the summer solstice (Y). The Vernal Equinox is labeled with the symbol for Aries. Text Label. Orbit Path. Wire frame representation of orbital path. Heading Line. The projection of the orbit path onto the planet’s surface. Not to be confused with ground track, which also takes into the rotation of the planet. Orbit Plane. Shaded surface representation of orbit ellipse. Orbit Positions. Triads representing the orientation and location of the spacecraft for orbital heating rate calculations. The X axis is drawn in red, the Y in green, and the Z in blue. A green arrow is drawn at the first orbit position, and a red arrow is drawn at the last orbit position. Focal Coordinate System. A coordinate system with the origin at the center of the planet and the X axis ing through the periapsis. The Y axis is at a right angle to the X axis in the plane of the orbit. The Z axis is normal to the orbit. A small red circle is drawn around the orbit normal vector at the planet’s surface. The size and colors of a few select items can be changed by selecting the Size/Colors tab on the Orbit Display Preferences dialog box (see Figure 6-26).

6-30


Figure 6-26

Orbit Display Preferences Dialog Box Visibility Tab

Figure 6-27

Orbit Display Preferences Dialog Box Size/Color Preferences Tab

Colors may be set for the shadow cone, orbit path, and heading line by selecting the Set Color button next to Solar Path Color, Orbit Path Color or Heading Line Color options. The Set Color dialog box (Figure 6-28) will appear that will allow a color to be selected from a palette, entered by name, or color index number. The size of the shadow, solar vector lines, and orbit position triads may be set. The size of the shadow is input in planet radii, the other scale factors are arbitrary.


6-31

Figure 6-28

6.2.2

Set Color Dialog Box

Viewing From Preset Vantage Points

The orbit may be viewed from a variety of preset vantage points, such as: • Sun • Star • Subsolar point • Planet North • Vernal Equinox • Ascending Node • Orbit Normal The commands to set the view are Thermal > Orbit > View From > *. Selecting one of the viewing points will update the display with the new view. The graphics are automatically scaled to fit the graphics area. Want "Hands-On" Information? Tutorial exercise “Importing a TRASYS Model and Using Articulators” on page -35 views a model and orbit from the sun. 6.2.3

Displaying the Vehicle in the Orbit

The vehicle may be displayed on the orbit by selecting Thermal > Orbit > View Vehicle > Set Orbit Position. Once selected, the View Vehicle In Orbit dialog box shown in Figure 6-29 will allow the to input the position and to scale the model. The Model Scale Factor will scale the model to be more visible. The Model Translation Factor repositions the model image proportionally to the orbit radius. When viewing the vehicle in orbit, it will display all surfaces in the default analysis group.

6-32


Figure 6-29

View Vehicle In Orbit (Preferences) Dialog Box

The may select to show the vehicle at a single position or at multiple positions. If a single position is selected, then the can also input the desired positions. To input multiple positions, selecting the Set Positions button will bring up the View Vehicle Positions dialog box where the may select the desired positions to be displayed. When Animate is selected, the Continuous Cycle Dialog dialog box will be displayed (see “Animate Through Time” on page 17-22) allowing input for the number of cycles. This option also can aid in making a movie (AVI) of the spacecraft in orbit. The Continuous Cycle Dialog dialog box is shown in Figure 6-30.


6-33

Figure 6-30

Continuous Cycle Dialog Box

Once the model has been displayed on the orbit, the may step through the positions by selecting Thermal > Orbit > View Vehicle > Next Position or Previous Position. A helpful hint to cycle through multiple positions is to simply select Next Position and then use the AutoCAD® repeat last command by clicking the right mouse button. Want "Hands-On" Information? Tutorial exercises “Importing a TRASYS Model and Using Articulators” on page -35 and “Orbital Heating Rates” on page -53 use the Orbit > View Vehicle command. 6.2.4

Color by Albedo/Planet

Albedo factors, IR Flux, or IR Temperature may be displayed on the planet’s surface in color. Albedo and planetshine data may now be input as a single value, an expression, a table versus time, or as a table versus latitude and longitude of the planet.

6-34


If the command is selected, the display will switch to a layout. Data will always be plotted, whether a single value, an array vs. time, or data vs. latitude and longitude. An example of albedo plotted on the planet surface is shown in Figure 6-31.

Figure 6-31

Albedo as a Function of Latitude and Longitude


6-35

6-36


7

Modeling Tools

Modeling tools are commands to assist the in building a model. These commands allow the to correct graphics problems, resequence IDs, correspond nodes, show IDs, and change visibility of objects.

7.1

Reset Thermal Desktop Graphics

• Icon: • Command: rcReset • Menu: n/a • Ribbon: Thermal > Common > Reset TD to Model Space • Toolbar: Modeling Tools This command switches back to the Model tab from layouts (Section 17.1.1) and also switches back to model view from orbit view (Section 6.2). This command also regenerates all graphics and can be useful for correcting graphics problems.

7.2

Regen Shade and Wireframe

• Icon: • Command: rcTouchAll • Menu: n/a • Ribbon: Thermal > Common > Regen All Thermal Desktop Objects • Toolbar: Modeling Tools This command regenerates the graphics and is useful for correcting graphics problems in layouts (Section 17.1.1) and orbit view (Section 6.2).

Modeling Tools

7-1

7.3

Resequence IDs

A collection of Thermal Desktop surfaces or nodes may have their Node IDs resequenced using the Thermal > Modeling Tools > Resequence IDs menu choice. After issuing the command, the is prompted to Select entity(s) for Node ID Resequencing. The can then select any number of nodes and/or any number of Thermal Desktop surfaces or solids (primitives). When entity selection is complete, the Resequence Node IDs dialog box, shown in Figure 7-1, is opened.

Figure 7-1

Resequence Node IDs dialog box

The Resequence Node IDs dialog box has the following fields and options. Resequence nodes in Submodel. This drop-down list allows the to select which submodel contains the nodes to be resequenced. Only submodels present in the selected entities are listed. Note that only the nodes within the selection set and within the selected submodel will be resequenced. Starting node number. This field specifies the starting ID of the resequenced numbers. This must be a positive integer within the limitations of SINDA/FLUINT node IDs. The nodes are resequenced in the selection order if nodes are selected and in surface order if surfaces are selected. Note: The selection order when using selection boxes (drawing a box around the selection set) cannot be certain. Node number increment. The increment added to the previous resequenced number. This value can be any integer within the limits of SINDA/FLUINT node IDs. Specifying a value of zero will use the starting node number for all nodes being resequenced. Specifying a negative number will decrement subsequent node IDs. Add increment to existing node number. When this option is checked, the node number increment is added to the current node number. The starting node number is ignored and the field is grayed out.

7-2

Modeling Tools

Want "Hands-On" Information? Use the Resequence IDs commands in the following tutorial exercises: “Simple Meshing Methods” on page 20-57, “Conduction and Radiation Using Finite Elements” on page 20129, and “Manifolded Coldplate” on page 22-37.

7.4

Resequence Fluid IDs

A collection of fluid network objects may have their IDs resequenced using the Thermal > Modeling Tools > Resequence Fluid IDs menu choice. After issuing the command, the is prompted to Select Fluid entity(s) for ID Resequencing. The can then select any number of lumps, paths, ties, ifaces, or fties. When entity selection is complete, the Resequence Fluid Network IDs dialog box, shown in Figure 7-2, is opened.

Figure 7-2

Resequence Fluid Network IDs dialog box

The Resequence Fluid Network IDs dialog box has the following fields and options. Resequence Entities in Submodel. This drop-down list allows the to select which fluid submodel contains the entities to be resequenced. Only submodels present in the selected entities are listed. Note that only the entities within the selection set and within the selected submodel will be resequenced. Starting entity1 number. This field specifies the starting ID of the resequenced numbers. This must be a positive integer within the limitations of SINDA/FLUINT fluid IDs. The fluid entities are resequenced in the selection order. 1 “Entity” can mean lump. path, tie, iface, or ftie.

Modeling Tools

7-3

Note: The selection order when using selection boxes (drawing a box around the selection set) cannot be certain. Entity1 number increment. The increment added to the previous resequenced number. This value can be any positive or negative integer within the limits of SINDA/FLUINT fluid IDs. A value of zero is not permitted. Specifying a negative number will decrement subsequent IDs. For fluid entities, if a new ID will be the same as the ID of another object of the same type, the ID will continue to be changed by the increment until the new ID is unique for that type of object. Add increment to existing number. When this option is checked, the number increment is added to the current number for each entity. The starting entity number is ignored and those fields are grayed out.

7.5

Node Correspondence

One or more nodes may be mapped into a different node name by using the Correspondence Manager dialog box. This dialog box, shown in Figure 7-3, is invoked using the Thermal > Modeling Tools > Node Correspondence menu choice.

Figure 7-3

Thermal Desktop Node Correspondence Manager Dialog Box

Correspondence data is made up of “Thermal Desktop nodes” and “SINDA/FLUINT nodes”. The SINDA/FLUINT (S/F) nodes are the output of the correspondence operation. The Thermal Desktop (RC) nodes are the input. Correspondence data is applied at the time SINDA/FLUINT data is generated from Thermal Desktop calculations. Whenever radks or heating rates are output, any RC nodes that are corresponded to a S/F node will automatically have data summed together and output using the S/F node name. Nodes that are not explicitly

7-4

Modeling Tools

corresponded to a S/F node are output directly using the RC node name. Correspondence data may be changed and new SINDA/FLUINT data generated without rerunning radiation calculations (using the Thermal > Radiation Calculations > Output SINDA/FLUINT Radks command). Correspondence data may be used to simply rename a node to a different submodel and ID, or to group nodes together for model simplification. A common use of correspondence data is to simplify a model that was nodalized to meet the requirement of uniform illumination when using radiosity based methods. For example, suppose that only the gradient along the length of a cylinder is of interest thermally, but its participation in the radiation environment is significant. Using RadCAD’s Monte Carlo methods, the cylinder can be nodalized along just the axial direction and accurate results will be obtained, since RadCAD’s ray tracing method does not require the condition of uniform illumination. If the faster progressive radiosity method is used, care must be taken to subdivide the cylinder along the circumferential direction as well, since a node that curves 360 degrees cannot be illuminated uniformly by other nodes. This is a property of all radiosity based methods and is a result of the fact that the view factor is used to compute the energy exchange between nodes. By definition, the view factor assumes that the energy leaving a surface is diffuse and uniform. In order for the energy leaving a surface to be diffuse and uniform, it must emit uniformly (be isothermal) and must reflect uniformly. To reflect uniformly, it must be illuminated uniformly. Radiosity methods assume that all energy is reflected and emitted uniformly. Another way of thinking about the radiosity process is to visualize that any radiation impinging on a node will be forced to reflect uniformly from the surface of that node. If solar energy illuminates one side of the cylinder that has a node that spans the entire circumference, the reflected energy will be distributed uniformly over the surface of the node, and the cylinder will appear to be transparent, since energy will be reflected from the far side of the cylinder as well as the illuminated near side. This problem can be avoided by subdividing the cylinder into an appropriate number of circumferential nodes for radiosity based radiation calculations, and then combining the results back together for SINDA/ FLUINT using the Correspondence Manager. The Correspondence Manager (CM) displays all submodels defined by the Submodel Manager (see Section 4.2). If a new submodel name is desired to be used for S/F output nodes, it must first be defined using the Submodel Manager. SINDA/FLUINT output nodes are created by selecting a submodel in the Correspondence Manager tree display and then selecting the Add S/F Node button. The Enter a Node Number dialog box will appear (Figure 7-4) allowing the specification of a node ID. An S/F Node icon will appear beneath the submodel indicating the submodel and node ID of the output node. Thermal Desktop nodes are mapped to the S/F output node by highlighting the Correspondence Manager S/F icon in the tree display, and then selecting the Add RC Node button. The Enter RadCAD Node Names dialog box shown in Figure 7-5 will appear to define the RC nodes to be corresponded to the S/F node.

Modeling Tools

7-5

Figure 7-4

Enter a Node Name Dialog Box

Figure 7-5

Enter RadCAD Node Names Dialog Box

A single node may be corresponded by selecting the submodel from the drop-down list and entering the ID in the Start ID field, leaving the other fields blank. If an End ID and an ID Increment is entered, a range of RC nodes will be corresponded. Nodes from different submodels may be corresponded to the same S/F node, just select Add RC Node as many times as necessary. The RC nodes that make up the S/F output node are shown as icons beneath the S/F Node icon. The tree display may be collapsed, expanded, and scrolled to locate desired correspondence data. Thermal Desktop and SINDA/FLUINT nodes may be removed by highlighting the choice in the tree display and selecting the Delete button. Selecting the Display button on the Correspondence Manager will update the graphics display area with the selected nodes highlighted in red, with the rest of the model appearing grey. If a submodel icon is chosen, the entire submodel will be highlighted. If an S/F icon is chosen, all of the RC nodes that make up the S/F node will be highlighted. Individual RC nodes will be highlighted if an RC icon is selected. Correspondence data may be turned on or off using the Correspondence On and Off radio buttons. If correspondence data is on, all SINDA/FLUINT data generated by RadCAD will use the correspondence data. Display of node submodels and IDs using the model checking commands will also display the S/F node name, followed by an asterisk so that it is visually evident that this is a correspondence node. Turning correspondence data off will ignore the data for all operations. Selecting the Correspondence Manager Clear button will delete all the entities in the tree. The Explode button will take data in the tree and go through the model and change the node numbers to be the proper number based on the tree. The Export button will export

7-6

Modeling Tools

the correspondence data to a new file. The Import button will read in a correspondence file. The import and export file are simple ASCII files that external programs can create if necessary.

7.6

Make AutoCAD Group

The Thermal > Modeling Tools > Make AutoCAD Group > From Thermal Objects command allows the to quickly put thermal objects into an AutoCAD group. The can select a group of objects from the graphics area and then the filter will be displayed. The filtering criteria allows the to be more specific about the objects that are selected. (see “Toggle Selection Filter” on page 7-11) The program first tests to see if a group of the input name already exists. If it does, than a new name is requested. A second test is made to determine if the name is valid. AutoCAD groups are not allowed to have blanks in the names or any special characters. The Thermal > Modeling Tools > Make AutoCAD Group > From Radiation Analysis Group menu choice is used to create an AutoCAD group consisting of the surfaces that are in the analysis group. The Thermal > Modeling Tools > Make AutoCAD Group > From Submodel menu choice creates a generic AutoCAD group consisting of all of the surfaces that have at least one node in the selected submodel. The Thermal > Modeling Tools > Make AutoCAD Group > From Planar Element using Normals menu choice creates a generic AutoCAD group consisting of all of the continuous elements considering the change in normals from subsequent elements. With the command, the will select a single quad or tri element. The will then input the maximum normal angle between coning elements. The program will then calculate all the elements in a group that meet the criteria. The purpose of this command is a way to provide a way to select continuous elements that may be difficult to select any other way. For example, a tube going through a solid. It may be difficult to select the elements on the inside of the tube that go into the solid. Groups can then be used at any of the AutoCAD selection set prompts. For example, all of the entities in a selected group may be changed to the same color. Enter the AutoCAD command chprop color and at the select surfaces prompt enter the command “group” followed by the name of the desired AutoCAD group. This has the same effect as manually selecting all of the surfaces in the group using the picking device. The group names may also be used to aid in the selection of surfaces for node ID resequencing, and in moving surfaces to different layers for better visualization. Please refer to the AutoCAD documentation for more information on AutoCAD groups and selection set operations.

Modeling Tools

7-7

7.7

Align UCS to Surface

Two coordinate systems for defining geometry are available, the World Coordinate System (WCS), which always remains fixed, and a Coordinate System (UCS), that can be defined as desired. UCSs can be positioned using a variety of techniques and saved under defined names. Point data that is entered in response to a command is, by default, input relative to the current UCS. (Please review the AutoCAD documentation for more information on UCSs.) To facilitate model building, the UCS may be aligned to the origin of any Thermal Desktop surface by selecting the surface and then selecting Thermal > Modeling Tools > Align UCS to Surface. The UCS may be displayed at the lower left corner of the screen or at the actual origin location.To display the UCS, use the AutoCAD View > Display > UCS Icon > On menu choice. To show the UCS at the actual origin, rather than the lower left corner of the display, use the View > Display > UCS Icon > Origin menu choice.

7.8

Toggle FD Mesh Nodalization

The Thermal > Modeling Tools > Toggle FD Mesh Nodalization command allows the to change the nodalization scheme used by finite difference surfaces that were converted from AutoCAD geometry (Thermal > FD/FEM Surfaces > From AutoCAD Surface, Section 4.3.11). By default, these surfaces are represented by a single node. Each facet of the converted surface may be a separate node with finite difference conductance approximations by using the Toggle FD Mesh Nodalization functionality. This is also discussed in Section 4.3.11 on page 4-34. This command cannot be used on finite element meshes. Want "Hands-On" Information? See this command used in Section 20.4 "Simple Meshing Methods"

7.9

Reverse Connectivity of Planar Elements/Meshes

The Thermal > Modeling Tools > Reverse Connectivity of Planar Elements/Meshes command will reverse the order of the nodes of an element (vertices of a mesh). That process will then change which side of the surface is the top side.

7-8

Modeling Tools

7.10

Shift Connectivity of a Planar Element/Rectangle

The Thermal > Modeling Tools > Shift Connectivity of a Planar Element/Rectangle changes the node order for a planar element such that the first edge (the edge between the first and second nodes) shifts to the right about the element normal. If a Thermal Desktop rectangle is selected, then the origin of the rectangle is shifted to the location of the next vertex to the right, about the surface normal. Only one element or rectangle may be selected at a time.

7.11

Convert Finite Difference to Finite Elements

The Thermal > Modeling Tools > Convert Finite Difference to Finite Elements command will convert finite difference surfaces and solids to finite elements. The thermal properties of the finite difference objects are also transferred in the conversion, such as the material and the thickness. Finite element solids created by this command will be surface coated to allow participation in radiation.The should be aware that the number of nodes from the finite difference surface is not necessarily the same upon the conversion to finite elements. The will most likely need to merge the nodes and resequence the node IDs for the converted nodes. This command does not have a reverse command.

7.12

Split Quad Elements into Tri Elements

The Thermal > Modeling Tools > Split Quad Elements into Tri Elements command creates an edge between two opposing corner nodes in a quad element. The quad element is then replaced by the two tri elements.

7.13

Refine Elements

The Thermal > Modeling Tools > Refine Elements command divides each edge of the selected elements into two edges with a node at the midpoint. Elements are then formed using all of the nodes. A quad element becomes four quad elements and a tri element becomes four tri elements.

Modeling Tools

7-9

7.14

Reverse Path/Pipe/Axis Direction

After a path, pipe, rotation axis or Node-to-Node Conductor has been created, the positive flow or rotation direction can be reversed by selecting the Modeling Tools > Reverse Path Direction command.

7.15

Move Path End

The Modeling Tools > Move Path End command allows s to change the either the upstream or downstream lump for an existing path.

7.16

Connect/Disconnect Pipe

FloCAD pipes representing ducts or wires (but not heatpipes2) can be ed together by use the RcConnectPipe command or the Connect Pipe icon. Selecting two lumps at the same location will merge the lumps into one: the extra end lump will be deleted. The nodes at the end of a pipe will be similarly merged if the pipe also has a wall model. The RcDisconnectPipe command (or Disconnect Pipe icon) will unhook pipes that have been merged. Unpredictable behavior may occur if the lines were moved after having been ed. Since only one end lump exists in connected pipes, the newly created lump (necessary to terminate the now disconnected pipes) may exist on either of the center lines of the disconnected pipes. Furthermore, actual path lengths will be computed that may not match the drawn line objects. The RcCheckPipes command can be used to display information about lines for which this may be a problem.

7.17

Show/Clear Path Area

Commands to visually display calculated flow areas can be displayed using the rcShowPathArea command or by selecting Thermal > Model Checks > Show Path Area from the menus. Figure 7-6 shows the results on an example rectangular cross section made up of Thermal Desktop rectangles. The planes containing the rays that are shot from the centerline are shown on the screen in red. Incomplete areas, or places with a single ray, will indicate a location where zero flow area has been computed due to a ray that failed to strike a surface when shot from the centerline at that axial location. The RcClearPathArea command will clear these lines from the screen. 2 Heat pipes (Section 5.4.5) cannot be ed: ing heat pipes would not generate a single larger heat pipe.

7-10

Modeling Tools

Figure 7-6

7.18

Example Pipe with Surfaces for Wall Showing Path Areas

Toggle Selection Filter

The Object Selection Filter dialog box (Figure 7-7) appears when the selects dissimilar objects for editing and issues the Edit command or when the selects multiple objects and issues a Thermal Desktop commands with the Thermal Desktop filter toggled “on”. The Object Selection Filter dialog box allows the to “filter” a selection set based on various input criteria. The first criteria to select is the type of object (Node, Surface, Solid Element, and such). Further filtering can be on a property of the objects, such as the submodel, thermophysical properties, etc. (Note that when the dialog box first appears, the contents of the Select Type to filter field will vary with each model depending upon the objects included in the selection set.) The Select Type to filter field on the left side of the Object Selection Filter dialog box lists the types of objects found in the selection set. The must select the type of object(s) to be included in the filtering. Multiple types can be selected by holding down the Shift key during the selection process. The Additional Criteria fields available on the right side of the dialog box will vary depending on the type of object(s) selected on the left side. Additional criteria may include Submodel, Material, Node Type, Analysis Group, Optics, and more.

Modeling Tools

7-11

If the selected item is a Surface/Planer Element, the Other criteria option is also available. Using Other criteria, the may select Normal Angle to Point or Normal Parallel to Vector. For the Normal Angle to Point option, the angle is measured between the surface normal and the vector from the base of the normal and the given point. If one of these options is selected, then the is prompted to input the required values after OK is selected on the Object Selection Filter dialog box. More options for selecting objects are provided by the Make AutoCAD Group command (see Section 7.6 ’Make AutoCAD Group’ on page 7-7). If the selected type is Nodes, Other criteria include Not Connected to Anything, Not Connected to Geometry, or Within Distance From a Point. With the last option, the is prompted to select a point, and then the distance from that point. The can force the Object Selection Filter dialog box to display if only one type of object is selected by toggling the selection filter to “on” (see “Toggle Selection Filter” on page 7-11). Toggling the selection filter “on” also allows filtering for other Thermal Desktop commands such as toggling visibility. Commands such as Erase or Move, which are AutoCAD based commands, do not recognize the Thermal Desktop filter. The Thermal > Modeling Tools > Toggle Selection Filter command toggles the Object Selection filter on and off. If the filter is on, then a command that prompts the for a selection set, such as Surface Coat Free Faces, will display the filter after the has input the selection set. The filter will allow the to place additional criteria to the selection set. For example, the may really only wish to surface coat solids made out of aluminum. With the filter on, the Surface Coat Free Faces command can be issued, the can select all of the model, and then input aluminum in the Material Filter field (Figure 7-7), and then only aluminum surfaces will be considered for surface coating.

Figure 7-7

7-12

Object Selection Filter Dialog Box

Modeling Tools

7.19

Synchronize Node Layer

The Thermal > Modeling Tools > Synchronize Node Layer command will move the nodes to the same layer as that of the surface that has been selected by the . This command makes it easier for the to develop a good scheme of layer management.

7.20

Turn Visibility Off/On/Undo

The Thermal > Modeling Tools > Turn Visibility Off/On commands allow the to turn the visibility of objects off and on. The must be careful in the use of these commands because the commands are only for the graphics, the objects still exist and are used for all calculations. The must also be aware that for an object to be visible, the layer it resides on must also be turned on. The Thermal > Modeling Tools > Undo Turn Visibility Off command will turn the visibility back on for the previous selection set that was turned off. Turning the visibility of an object on can be a difficult task since the cannot pick the object in the graphics area. It may be easier for the to use AutoCAD groups for selection or typing in ALL at the selection prompt will turn the visibility on for all the objects (provided the layer they exist on is also on). Want "Hands-On" Information? Most of the tutorial exercises utilize the object visibility commands. The exercises in Section 20.2 "Setting Up a Template Drawing" are a good way to gain a solid understanding of how working with the visibility commands can be used to simply a job.

7.21

Turn Node Numbers Off/On

The Thermal > Modeling Tools > Turn Numbers Off/On commands allow the to display the node number on the model. The node numbers are displayed as a function of the node. The node layer must be visible to see the node numbers. The size of the number can be controlled by the Graphics Size Preference dialog box (see “Graphics Size” on page 2-29). The can also control whether or not the submodel name is displayed with the node number. Want "Hands-On" Information? Practice using this command in tutorial exercise “Manifolded Coldplate” on page 22-37.

Modeling Tools

7-13

7.22

Copy Properties From Master

The Thermal > Modeling Tools > Copy Properties from Master command allows the to copy the properties from one object to other objects of the same type. For instance, the has a surface that has all the properties assigned correctly (i.e. optical properties, thickness, material, etc.) and would like to have all the same properties copied to other surfaces. This command allows the to accomplish this function. The command does not write over the node number or the name of an object.

7-14

Modeling Tools

8

Model Checks

Before calculations are performed, Thermal > Model Checks commands may be used to the thermal model data. The following type of checks may be made: • Displaying Active Sides (Section 8.1) • Color by Property Value (Section 8.2) • Viewing the Model From the Sun/Planet (Section 8.3) • List Duplicate Nodes (Section 8.4) • Show Free Edges (Section 8.5) • Check Pipe Connectivity (Section 8.6) • Display /or Markers (Section 8.7) • Calculate Mass (Section 8.8) • Output Analysis Group Summary (Section 8.9) • Output Node Optical Property Summary (Section 8.10) • Check Overlapping Surfaces (Section 8.11)

8.1

Displaying Active Sides

The Thermal > Model Checks > Active Display Preferences menu choice displays the Display Preferences dialog box shown in Figure 8-1. When displaying active sides, all Thermal Desktop surfaces are set to colors indicating the active sides for the selected analysis group. The colors are shown when the AutoCAD® shade or render command is executed. The may first select what is to be displayed. The default is to display the active sides for the current radiation analysis group. This same functionality may also be used to display which surfaces have insulation nodes, which surfaces have area conductance applied, which is the top side, or which side is active for a Tag Set (Section 2.5). The colors are as follows: • Green: Indicates that the side is active (or top), and that the opposite side of the surface is inactive (or bottom). • Light blue (cyan): Indicates that the side is inactive (bottom), and that the opposite side is active (top). • Dark blue: Indicates that both sides are inactive. • Yellow: Indicates that both sides are active. Model Checks

8-1

• Red: Indicates that this surface is not in the current default radiation analysis group. Active side information may also be shown as arrows. The may select to be prompted to select the surfaces for which the arrows will be displayed. An example of the active

Figure 8-1

Active Side Display Preferences dialog

side display using arrows is shown in Figure 8-2. The active sides of surfaces are visually indicated on the drawing (using the method selected in the Display Preferences dialog box) by selecting the Display button on the Display Preferences form or by using the command Thermal > Model Checks > Display Active Sides. The Thermal > Model Checks > Display Active Sides command will display the active sides of the default analysis group. The default analysis group is set by the drop down beside the Analysis Group radio button or in the Radiation Analysis Group dialog (Section 4.1). If the Prompt for subset of surfaces option is set in the Surface Selection field, a prompt will appear at the command line asking for a selection set of surfaces. Use AutoCAD selection methods to select the set of surfaces in the analysis group for which active sides are to be verified. Individual surfaces may be selected to reduce visual clutter when working with large models.

8-2

Model Checks

In the arrow display modes, arrows will be drawn emanating from the active sides of the surface. Only one active side arrow is drawn per active side node. The curved surface shown in Figure 8-2 consists of individual nodes, one per facet. If the curved surface were a single node, only one active side designator would be drawn on the middle facet of the mesh.

Figure 8-2

Active Side Verification for an Analysis Group Using Arrows

Arrows will be correctly hidden by the AutoCAD hide command. The size of the arrow may be controlled by the Preferences dialog box Graphics Size tab (see “Graphics Size” on page 2-29). To remove the active side designators from the display screen, select Thermal > Model Checks > Active Sides Off. It is not necessary to clear the active side display when displaying a different analysis group or if the type of designator is changed. Selecting Thermal > Model Checks > Display Active Sides will automatically clear existing designators before drawing new ones. Want "Hands-On" Information? Tutorial exercises “Radks for Parallel Plates” on page 21-3 and “Space Station Oct Tree Example” on page 21-23 utilize the Display Active Sides command.

Model Checks

8-3

8.2

Color by Property Value

The commands accessible from the Thermal > Model Checks > Color by Property Value > * provide various ways to color the model based on different properties. Various optical property, material property and conductance values may by verified on the model. The K-star and E-star options refer to the thermal conductivity and effective emittance of insulation on a surface respectively. For example, if Optical Properties > IR Emissivity is chosen, the geometric model will be updated with color values indicating the emissivity values assigned to the surface using the currently loaded optical property database. A color bar will be shown to indicate the value of the colors used. More information about modifying the color bar can be found in Section 17.1.2.1. Note: Only global values will be displayed. Symbol and property overrides defined in the Case Set Manager cannot be checked by this method. Want "Hands-On" Information? See this command used in “Beer Can Example” on page 20-89.

8.3

Viewing the Model From the Sun/Planet

If an orbit is currently defined, the model may be viewed from the vantage point of the Sun, the planet center or the current view. This command will update the orientation of articulators that are defined based on the orbit position. The menu choice Thermal > Model Checks > View Model From Sun/Planet > Set Orbit Position/Location will display the View Vehicle Setup dialog box shown in Figure 8-3. This dialog box is used to set the orbit position and to specify the vantage point as the Sun or the planet center. An option also exists to automatically execute the shade command after the orbit position is incremented or decremented with Thermal > Model Checks > View Model From Planet/Sun > Next Position or Thermal > Model Checks > View Model From Planet/ Sun > Previous Position. Tip: Use the <Enter> key to repeat the last Next Position command to cycle through the orbit. The current position is listed in the text window at the bottom of the main screen. The geometry will be oriented such that the North pole of the planet appears to be “up”. Want "Hands-On" Information? Tutorial exercises “Importing a TRASYS Model and Using Articulators” on page 21-35, and “Orbital Maneuvers” on page 21-87 both use this functionality.

8-4

Model Checks

Figure 8-3

8.4

View Vehicle Setup

List Duplicate Nodes

The Thermal > Model Checks > List Duplicate Nodes command scans the database and finds nodes that have the same Submodel name and Node ID. Two nodes that have the same submodel name and ID are considered to have the same temperature. The must determine if this is correct for their model or not. Please note that nodes that are selected to be Clone nodes are not considered in the List Duplicate Nodes function.

8.5

Show Free Edges

The Thermal > Model Checks > Show Free Edges command prompts the to select finite elements to be checked. These elements are checked against all the other elements in the model and the edges that are not connected to anything are drawn in red. This visual verification of the model allows the to quickly determine if a model is set up properly. If an edge is drawn in red, but it visually looks to the that the edge is connected to another edge, then there are duplicate nodes that lie on top of each along the edge. Want "Hands-On" Information? See how this command is used in a practical exercise - refer to the tutorial exercise “Conduction and Radiation Using Finite Elements” on page 20-129.

Model Checks

8-5

8.6

Check Pipe Connectivity

The Thermal > Model Checks > Check Pipe Connectivity command checks for lumps that are connecting two pipes, but not located at the intersection of the two pipes. When two pipes that have been connected using the Thermal > Modeling Tools > Connect Pipes command, they are ed by combining one lump from each pipe into a single lump. This lump must be located at the intersection of the two pipes. The single lump is then an object in both of the pipes, and should be physically located at the intersection of the two pipes based on the discretization of the pipes. If the pipes are moved, or more lumps are created, the lump which is at this intersection may no longer be in the correct position. This command will list the lumps for which this situation occurs. If the SINDA/FLUINT model is generated and run, the lumps and paths created will all be connected, but the geometry may not be correct.

8.7

Display /or Markers

The Thermal > Model Checks > Show /ors Markers commands allows the to see how the program calculates the ors between surfaces. Each of these commands will prompt the to select the surface with or the or. The Show Markers command will create an AutoCAD point at each sampling point for the . A red point means no was found and a yellow point means that was made. The Show or Markers command will create a yellow AutoCAD line from the test point (integration interval) to the node to which it is connected for the point algorithm or a yellow line normal to the test point to the surface with which it intersects for the ray trace algorithm. If no connection is made, then a red AutoCAD point is created at the sample point. or markers are simply graphical representations showing the connection of test points: they do not represent a distance for the conductance calculation nor do they represent individual conductors (test points of one node connected to the same node are integrated to form a single conductor between the node pair). The Thermal > Model Checks > Clear /or Marker command will delete the markers created by the above commands. The markers are all placed on the RADCAD_RAYS layer. Note the clear command will delete all entities on the RADCAD_RAYS layer and will remove the layer when it is completed. An alternative to using the Display or Markers command is to use the Model Browser. After selecting a or or ors in the Model Browser, the Display or Markers command can be accessed through the right-click context menu.

8-6

Model Checks

Want "Hands-On" Information? Tutorial exercises “Circuit Board Conduction Example” on page 20-67, “or Example” on page 20-171, and “Drawn Shape Heat Pipe” on page 22-85 use the Display /or Markers commands.

8.8

Calculate Mass

The Thermal > Model Checks > Calculate Mass command will calculate the mass of the model. The command first calculates the mass of the surfaces and solids. If Generate Nodes and Conductors is on for the entity, then the mass is calculated for each node in the solid as long as the Node has not been overridden. The mass for all the solids is then summed and output per Node Submodel. The command next cycles through all the nodes and determines if they are specified to be diffusion nodes with a material. The mass is summed for each of these nodes and then output. Please note that the mass of a diffusion node that does not have a material defined cannot be calculated and thus is not output by the program.

8.9

Output Analysis Group Summary

The Thermal > Model Checks > Output Analysis Group Summary command will write out active side information for each node and each analysis group. The data will be written to a file named “AnalysisGroupSummary.xls”. This file is a simple tab delimited file that will easily load into Excel, Word, or Powerpoint.

8.10

Output Node Optical Property Summary

The Thermal > Model Checks > Output Node Optical Property Summary command will write out active side information for each node and each analysis group. The data will be written to a file named “NodeOpticalPropSummary.xls”. This file is a simple tab delimited file that will easily load into Excel, Word, or Powerpoint.

8.11

Check Overlapping Surfaces

The Thermal > Model Checks > Check Overlapping Surfaces command provides the capability to test the current default radiation analysis group for overlapping surfaces. Overlapping surfaces can cause incorrect results in Monte Carlo Ray Tracing. When the

Model Checks

8-7

command is issued, the Check Overlapping Surfaces dialog box shown in Figure 8-4 is displayed.

Figure 8-4 Check Overlapping Surfaces Input Dialog Box

The algorithm used to find overlapping surfaces is the same that is used to generate conductance. Basically, points are generated on each surface and the other surfaces in the model are tested to see if the point lies in their volume. For the Check Overlapping Surfaces test, the thickness of the 2D surfaces is made very small so that effectively only the surfaces are used for the tests. The test is performed only on the default Analysis Group (Section 4.1). The default Analysis Group can be changed using the drop down at the top of the form. The can control the tolerance of the test. This number should be slightly above zero, but should remain a small number in respect to the size of the average surface in the model. The integration intervals determines how many points are tested. A value of 10, will test 100 points per node. The Oct-Cell Subdivisions are used to speed up the time of the tests, but do not affect the results. And finally, the can input a percentage that must be achieved for a warning to be issued. Once the tests have been performed, a log file is created in the working directory and an AutoCAD group is created containing any overlapping surfaces. The group name will be OVERLAP_analysisgroup, where analysisgroup is the name of the analysis group used in the calculations. The can use the model browser to list the groups, isolate which surfaces overlap, and then adjust them properly. Want "Hands-On" Information? See this command used in tutorial exercise “Space Station Oct Tree Example” on page 21-23.

8.12

Check Elements • Command: RcCheckElements

8-8

Model Checks

• Ribbon: Thermal > Model Checks > Check Elements

The Check Elements opens the Element Quality Check dialog (Figure 8-4). When OK is selected, Thermal Desktop will compare all finite elements with the quality values provided and place all elements that do not meet the criteria into the AutoCAD group BADELEMS. This group can be displayed by using the Model Browser to List by Groups, selecting the BADELEMS group, and then selecting Display Only.

Figure 8-5 Element Quality Check Dialog Box

Interior Angles These are measurements of interior angles. Min Allowed. Elements with an internal angle less than this value will be included in BADELEMS. Max Allowed. Elements with an internal angle greater than this value will be included in BADELEMS.

Skew Skew is a non-dimensional parameter ranging from 0 (ideal) to 1 (degenerate). The value is calculated by comparing the area or volume of the element to that of an ideal element in the same circumcircle or circumsphere: a circle or sphere that contains all of the vertices of a planar or solid element, respectively. Max Allowed. Elements with a calculated skew greater than this value will be included in BADELEMS.

Model Checks

8-9

8-10

Model Checks

9 Conductance and Capacitance Calculations and Controls

9.1

Capacitance Calculations

For nodes attached to a surface, a thermophysical material is defined on the Thin Shell Data dialog box CondCap tab (see “Conductance/Capacitance Tab” on page 4-15) for the surface. The nodal capacitance is calculated by multiplying the area of the node times the thickness times the specific heat times the density. The capacitance may be constant or temperature varying. If the value is calculated to be zero (i.e. the surface thickness, specific heat, or the density are zero), the node will be output as an arithmetic node. For doublesided surfaces (see Section 9.2.1), the capacitance for one face of the surface is calculated using the density, specific heat and thickness of the face material and the density, specific heat and half of the thickness of the separation material. Two nodes are included in the thickness direction. If a node is attached to an element, the capacitance is calculated from the element material and volume. The volume of a planar element is calculated from the area and the thickness. The can use the nodal boundaries graphics functions in Preferences to get an idea of how the node on a finite element entity is represented (see “Graphics Size” on page 2-29). If a node is not attached to a surface or element, then the capacitance is specified directly by the in the Node dialog box (see “Nodes” on page 4-62). These options are also available for nodes attached to surfaces if the Override calculations by elements/surfaces check box is checked. 9.1.1

Double-Sided Surfaces

Double-sided surfaces, or gradient surfaces, are finite difference surfaces with different node numbers for each side of the surface. This allows a temperature to be calculated for each side of the surface. Double-sided surfaces are created by unchecking the Use same ID’s on both sides check box on the Numbering tab of the surface edit form (Section 4.3.1.2 Numbering Tab) and by specifying the materials and material thickness on the Cond/Cap tab (Section 4.3.1.4 Conductance/Capacitance Tab). For double-sided surfaces, two nodes are created in the thickness direction. The capacitance of each node is based on: • the corresponding face thickness and face material specific heat and density • half of the separation thickness and the separation material specific heat and density

Conductance and Capacitance Calculations and Controls

9-1

Note: Double-sided surfaces are typically used when a low conductivity material separates two higher conductivity materials. This would be typical of a foam-core sandwich or perhaps a honeycomb sandwich. If the gradient through the thickness is due to a low conductivity material on one or both sides of a high conductivity material, then the Insulation tab (Section 4.3.1.6) should be used instead. If the plate is sufficiently thick or the conductivity through the thickness sufficiently low to produce a significant gradient, a better option may be to use a finite difference solid (Section 4.4) or solid finite elements (Section 4.5).

9.2

Conductance Calculations

For finite difference surfaces, such as rectangles and cones, the conductors between the nodal regions on a surface are output to SINDA/FLUINT using a finite difference formulation. The Galerkin partial differential equation is used to solve the conductance between nodes of a finite element. The equation set representing the heat transfer between nodes is output in SINDA/FLUINT conductor format. Conductors between the same node pairs are added together, if they are of the same type (constant or temperature-varying conductivity). It should be noted that an individual conductor generated by Thermal Desktop for a finite element does not represent the heat transfer between the two nodes referenced by the conductor. The heat transfer between two nodes is represented by all conductors within the element. For a complete description of the calculation of element conductivity, please refer to “The Finite Element Method and Thermal Desktop”, that can be found at the CRTech web site (www.crtech.com) under “Resources”. 9.2.1

Double-Sided Surfaces

Double-sided surfaces, or gradient surfaces, are finite difference surfaces with different node numbers for each side of the surface. This allows a gradient to be calculated through the thickness of the surface. Double-sided surfaces are created by unchecking the Use same ID’s on both sides check box on the Numbering tab of the surface edit form (Section 4.3.1.2 Numbering Tab) and by specifying the materials and material thickness on the Cond/Cap tab (Section 4.3.1.4 Conductance/Capacitance Tab). For double-sided surfaces, the two nodes in the thickness direction are connected by a linear conductor whose conductance is calculated from the conductivity of the separation material, the area of the nodes (the face of the surface), and the thickness of the separation material. If the separation material is defined with an effective emissivity (see Section 3.1.1), the a radiation conductor will be used in parallel to the linear conductor. The radiation

9-2


conductor will be the effective emissivity multiplied by the nodal area. If the separation thickness is set to zero and an effective emissivity is not defined, then the two nodes in the thickness direction will be thermally isolated. The conduction between the nodes on one face of the surface is based on the conductivity of the face material, the thickness of the face material, the node width and the distance between nodes. Note that the lateral conduction does not include the separation material. Note: Double-sided surfaces are typically used when a low conductivity material separates two higher conductivity materials. This would be typical of a foam-core sandwich or perhaps a honeycomb sandwich. If the gradient through the thickness is due to a low conductivity material on one or both sides of a high conductivity material, then the Insulation tab (Section 4.3.1.6) should be used instead. If the plate is sufficiently thick or the conductivity through the thickness sufficiently low to produce a significant gradient, a better option may be to use a finite difference solid (Section 4.4) or solid finite elements (Section 4.5).

9.3

Area Calculations

Area conductance calculations are made when the faces of objects are used for defining the thermal interfaces between objects. There are two methods for specifying area conductance: surface/solid dialog box tab and ors. Although there are some similarities between the two types, ors have many more options including how the area conductance is calculated. 9.3.1

Surface/Solid Area

If the specifies an area conductance for a surface or solid in the tab of the respective Thin Shell Data dialog box (Section 4.3.1.5 or Section 4.4.1.5), each node on the face is broken into regions. The number of regions is set by integration intervals (see “Set Cond/Cap Parameters Dialog Box” on page 9-18) which is the number of interval in each direction. The centers of these regions are then checked to find any surfaces or solid elements that are within the boundaries. If a surface or solid is found, a conductor is generated between the node with the area condition and the node located near the center of the region. The conductor is calculated by taking the area of the region


9-3

multiplied by the factor input on the tab of the Thin Shell Data dialog box (see “ Conductance Tab” on page 4-16 or " Tab" on page 4-48). Conductors connecting the same two nodes are summed before being output to SINDA/FLUINT.

centers of integration regions are computed using surface thickness

surface with area condition (graphical surface shown in black) chip

board

Figure 9-1

Thicknesses of surfaces are used for area calculations

A node on a surface with area will be connected to a node on another surface or solid if a center of a integration region is located on or inside of another surface. The thickness of potential surfaces to which a connection could be made is taken into for the test. For example, suppose an area condition is placed on the bottom side of a chip as shown in Figure 9-1. The area of the nodes on the bottom side of the chip are subdivided into a number of smaller interval regions. The center of each region is tested to see if it lies on or in another surface or solid. The thickness of the chip is taken into when generating the integration regions if the Exterior radio button is selected next to Generate area at:. If an area condition is placed on the bottom side of the chip, the centers of the integration regions will be displaced from the graphically displayed surface by one half the thickness of the surface in the direction normal to the bottom side of the surface. Locations of integration regions for double sided surfaces are shown in Figure 9-2. top face

top side locations

graphically displayed surface core material bottom face bottom side locations

Figure 9-2

9-4

Locations of integration regions for double sided surfaces


Figure 9-1 shows the top side of the board making with the bottom side of the chip due to the thicknesses of each surface. The recommended approach is to position geometry such that their theoretical surfaces, when considering the thickness, each other. will be successful if the surface with the condition lies inside of a surface, but care should be taken when using double sided surfaces so that the test points lie on the correct half of the surface to which a connection is desired. If a conductor is made to a surface that has double sided node numbering, then the location of the center of the interval region relative to the mid-plane of the surface to which the connection is made will determine which side’s nodes are used for the conductor. The connection will be made to the appropriate nodes on proper side of the surface. If a surface has insulation on it, the connection is always made between surface nodes, not the insulation nodes. 9.3.2

Face or

For a “Faces” type of or, test points are created in each direction on each nodal area of the From surfaces. The number of test points in each direction is determined by the Integration Intervals value on the or edit form (see “ors” on page 4-74). The test points are generated using the surface thickness, much like the regions used for surface/solid area discussed in the previous section, if Apply Surface Thickness To Test Points is selected. For the calculation method, the can select between the point algorithm and the ray trace algorithm. The point algorithm consists of testing each test point to see if the test point lies in the volume of one of the To surfaces. If the test point does not lie in a volume, then the algorithm determines which face of all of the To surfaces is closest to the point. The algorithm searches in all directions. If the closest surface is outside of the tolerance specified in the or form, then no connection will be made for that test point. The point algorithm is very accurate and uses points to find the closest surface in the To list. As the From and To lists increase in size, however, this algorithm can be slow. When this happens, it is recommended that the consider the “Ray Trace” algorithm. The Ray Trace algorithm shoots rays normal (perpendicular) to the From surface, at the test points, and whatever is intersected is deemed the closest node. When the surfaces are parallel, this algorithm works very well and is extremely fast. Note that the tolerance is checked as the length of the rays between the surfaces. With this algorithm, s should make sure the proper Top or Bottom side is correctly input. Note: Thermal Desktop checks for coplanar surfaces only when performing radiation calculations that are run through the Case Set Manager. Using the Ray Trace algorithm with a model that has overlapping surfaces will result in erroneous results. If the is not going to be running radiation calculations, or just wants to check for overlapping surfaces, the command Thermal > Model Checks > Check overlapping surfaces can be used.


9-5

In order to provide a faster mechanism for the point algorithm than checking each object, the Max Check Objects field can be used to speed up the process for the point algorithm. When the value specified is less than the number of objects in the To list, a special procedure is used to find the objects to check the distance from the current integration point. A number of points are computed on each object in the To list. These points are stored in a tree that can be quickly searched to locate points nearest to the test point. By only looking at the objects with the nearest point, the integration time can be 1-2 orders of magnitude faster than checking each object. When the geometry of the objects in the To list is well behaved, the results can be identical with checking all objects. But if there are objects which occupy the same space or poke out through the surface of another object, the conductors generated can be different depending on the number of objects checked. By specifying a number of Max Check Objects greater than the number of To objects, all objects are tested during the integration. The number of points placed in the tree for each object is set at 5 in each direction with points on each face for 2-dimensional objects. The thickness of surfaces are utilized in the determination of the surface points, except for pipes which are treated as linear objects along their length. Once connections are made from test points on the From surfaces to nodes on the To surfaces, the connections are integrated using the test points. The area integral from the test points of a node on a From object to a node on a To object is used to create a conductor between the two nodes. The integral is multiplied by the Conduction Coefficient, and the material conductivity if that option is selected, to create the conductance value for SINDA/ FLUINT. Note that any conductance from the node to an objects face is not included in the node-to-node conductance. The Conduction Coefficient Scaling is applied along an entire surface using the test points.

9.4

Edge Calculations

Much like area conductance, edge conductance calculations are made when the edges of 2-D objects are used for defining the thermal interfaces between objects. There are two methods for specifying edge conductance: surface/solid dialog box tab and ors. Although there are some similarities between the two types, ors have many more options including how the conductance is calculated. 9.4.1

Surface Edge

If edge is specified for an edge of a surface (Section 4.3.1.5), that edge is broken into integration intervals segments for each node on that edge. The centers of these segments are then tested against the rest of the model to find what other nodes lie within the volumes of these nodes. The rules for making a connection are the same as those for area calculations. The value of the conductor will be the length of the interval segment multiplied by the factor input on the tab of the Thin Shell Data dialog box (Section 4.3.1.5).

9-6


A per-length value is input for edge rather than a per-area value in order to make the edge independent from the thickness specified for the surface. The thickness of the surface may not be the same at the edges as in the center. For example, a honeycomb may only have one face sheet extended for connection at the edges. An area coefficient may be converted to a line coefficient by multiplying by the thickness of the surface at the ing edge. 9.4.2

Edge or

For an “Edges” type of or, test points are created along nodal edges coincident with the active edges of the From surfaces. The number of test points along each node is determined by the Integration Intervals value on the or edit form (see “ors” on page 4-74). For the calculation method, the can only select the point algorithm. The point algorithm consists of testing each test point to see if the test point lies in the volume of one of the To surfaces. If the test point does not lie in a volume, then the algorithm determines which face of all of the To surfaces is closest to the point. The algorithm searches in all directions. If the closest surface is outside of the tolerance specified in the or form, then no connection will be made for that test point. The point algorithm is very accurate and uses points to find the closest surface in the To list. As the From and To lists increase in size, however, this algorithm can be slow. In order to provide a faster mechanism for the point algorithm than checking each object, the Max Check Objects field can be used to speed up the process for the point algorithm. When the value specified is less than the number of objects in the To list, a special procedure is used to find the objects to check the distance from the current integration point. A number of points are computed on each object in the To list. These points are stored in a tree that can be quickly searched to locate points nearest to the test point. By only looking at the objects with the nearest point, the integration time can be 1-2 orders of magnitude faster than checking each object. When the geometry of the objects in the To list is well behaved, the results can be identical with checking all objects. But if there are objects which occupy the same space or poke out through the surface of another object, the conductors generated can be different depending on the number of objects checked. By specifying a number of Max Check Objects greater than the number of To objects, all objects are tested during the integration. The number of points placed in the tree for each object is set at 5 in each direction with points on each face for 2-dimensional objects. The thickness of surfaces are utilized in the determination of the surface points, except for pipes which are treated as linear objects along their length. Once connections are made from test points on the From surfaces to nodes on the To surfaces, the connections are integrated using the test points. The length integral from the test points of a node on a From object to a node on a To object is used to create a conductor between the two nodes. The integral is multiplied by the Conduction Coefficient, and the material conductivity if that option is selected, to create the conductance value for SINDA/ FLUINT. Note that any conductance from the node to an objects edge is not included in the node-to-node conductance. The Conduction Coefficient Scaling is applied along an entire surface using the test points.


9-7

9.5

Material Flow (Advection) Options

Thermal Desktop objects can be relocated (rotated, translated, etc.), perhaps even dynamically during a transient analysis. Loaves of bread entering and exiting an oven, or ingots moving through a furnace, or a sprocket or roller bearing rotating slowly could all be modeled as such discrete, commanded movements. However, such rotations and/or translations have no meaning within a steady-state analysis, and can become cumbersome in a transient solution if the motion is continuous or even if it is fast compared to the time scale of the event. For example, what if the focus of the analysis is not the heating transient of a single ingot, but the effects on the furnace of a steady stream of in-flowing cold ingots? Other examples include temperature profiles within roller or ball bearings, along conveyor belts, and the simulation of continuous annealing of sheet metal, drying cardboard, solidifying glass, etc. Some components such as rotating disk heat exchangers intrinsically rely on continuous motion of solids to achieve their performance. The Thermal Desktop material flow options provide a means of defining steady or unsteady radial, axial, or circumferential movement of materials that are to be superimposed upon the conduction, radiation, , and convection that might also be occurring. To model the continuous flow of a sheet of metal, an FD solid brick can be generated with appropriate material properties, nodal resolution, radiation, convection off the surface, and with other surfaces (perhaps rollers or guides). The fact that the sheet is moving from the left to the right (say) is modeled by adding a velocity to the Advection tab of the FD Solid Edit form (Figure 9-3). Figure 9-4 shows the steady-state temperature profiles that might appear in a roller underneath such a hot sheet of metal; the outer radius of this roller (modeled as an FD solid cylinder) has also been assigned an advection velocity matching that of the sheet at the point they touch.

Figure 9-3

9-8

FD Solid Edit Advection Tab for Specifying Material Flow Velocities


Figure 9-4

Thin-walled Roller ing a Hot Surface at the Top

The material flow options generate one-way SINDA conductors to represent the advection term, which is superimposed upon the rest of the solution. In fact, for FD solids and FloCAD pipes, a pair of such conductors (in opposing directions) is generated in case the velocity is reversed (becomes negative): the velocity can be either constant or variable. One-way conductors represent a fast-solving way1 to represent material flows, which unlike fluid flow is characterized by uniform velocity profiles: the motion of the otherwise solid material is represented as a flow. Therefore, the material for the advection term is the same as the material used for the thermal capacitance term. In certain cases, such as flow through porous media, the solid is truly stationary and a fluid of uniform velocity of gas or fluid es through it, with heat exchange between fluid and solid assumed to be infinite at every point since one node represents both. Currently, this situation must be modeled using superimposed solids, one with advection (the fluid) and one without (the solid), and the nodes of the two objects merged. Otherwise, if heat exchange is not perfect (e.g., low heat transfer area per unit volume), consider a concurrently solved FloCAD/FLUINT representation of the fluid. Another restriction is related to the use of expressions for velocities. If rotation or translation rates are to vary during a SINDA/FLUINT run (perhaps for parametric variations, or to transient changes), then a symbol-containing expression can be used for velocity, and the symbols and expression should be output to SINDA/FLUINT. The derived SINDA/ FLUINT expressions for the one-way advection conductances will then permit variation of velocities during the run, without the need to call Thermal Desktop dynamically for recalculation of the advection term. For those reasons, the submodel containing those advection

1 One-way conductors will result in SINDA/FLUINT’s energy balance criteria reporting “STABLE BUT UNBALANCED” energy flows, which is to be considered normal. Refer to the SINDA/FLUINT ’s Manual for details.


9-9

conductances should be designated as static, not dynamic, during a dynamic run. (Use the Submodels... button on the Dynamic tab of the Case Set Manager to make a submodel static.) If the dimensions of the FD solid are to change during a dynamic run, then the other conductors generated by TD will also have to be updated dynamically. If this is the case, leave the submodel as dynamic and uncheck the “Output Expressions to SINDA” option for the advection conductors, such that they too are updated during the dynamic mode. Four types of Thermal Desktop objects are applicable for material flow advection: FD solids, FloCAD “wires” (pipes without fluid), Conductors, and ors. While the basic use of these entities is described elsewhere in this manual, the following sections describe their use in application to material flow problems. Material flow velocities can be assigned in any or all of the three principal directions for FD solid cones, cylinders, bricks, and spheres (e.g., along the X axis of a brick, rotation about the centerline of a cylinder, radially in a hollow sphere) on the Advection tab. Axial flow from a cylinder to a truncated cone (representing an expanding or reducing section) can be ed using a or. Even more elaborate shapes and paths can be generated using FloCAD “wires” (wall-only pipes), perhaps built on complex polyline centerlines and with arbitrary cross sections. Such wires which might represent curved sections such as cylindrical material conduits, bent extrusions, or rectangular conveyor belts. The ends of wires can be ed to other wires or to FD solids using the or options as well. A complex series of objects representing the continuous flow of a material through various shapes can therefore be built (Figure 9-5 as an example). This series can be openended (an inlet or outlet, as shown in Figure 9-5) or closed (such as a complete model of a conveyor belt that returns to its starting point). For open systems, note that an upstream (entrance) boundary condition must be defined, perhaps using flow from a boundary node to the first surface via a Conductor. However, the downstream (exit) boundary condition need not be defined. Note that, unless the material flow is fast and/or the material thermal conductivity is low, then both one-way Conductors/ors and two-way Conductors/ors must be generated when solids/pipes are to be ed to each other. The one-way Conductor/ or described material advection as described in Section 9.5.3 Using Conductors to Represent Material Flow and Section 9.5.4 Using ors to Connect Material Flows, whereas the two-way Conductor/or describes normal conduction, which has been described elsewhere in this manual. 9.5.1

Solid Finite Difference Object

Advection, or material flow velocities, can be defined in any or all of the three principle directions of bricks, cylinders, cones, or spheres as defined below.

9-10


Figure 9-5 3D Material Flow Example with FD cones, cylinders, FloCAD Pipes, and ors (not visible)

9.5.1.1

Solid Brick

For an FD solid brick (Section 4.4.2 Solid Brick), material flow velocities along one or more axes can be defined in the Advection tab. Usually, only one such velocity (e.g., along the brick’s X axis: Xmin to Xmax as positive) is defined. For example, if a sheet of moving material were being modeled, the velocity in the direction of position motion can be defined. The advection conductances will be generated according to each node’s flow area (perpendicular to the velocity) based on the current resolution. If the nodal resolution changes, so too will the advection conductors.


9-11

9.5.1.2

Solid Cylinder

For an FD solid cylinder (see “Solid Cylinder” on page 4-52), material flow velocities for one or more directions can be defined in the Advection tab. One common usage is a velocity along the axial direction (Hmin to Hmax), representing flows along the centerline, with one disk face being an inlet and the other an outlet. Another common usage is the circumferential direction (Gmin to Gmax), which represents rotation about the centerline (Z axis), and is commonly used to represent rollers or disks. For this case, note that the velocity is defined at the outer radius, such that if it matches the velocity of a ing sheet with no slip, for example: the same velocity can be specified in both forms. If a rotation rate is defined (e.g., RPM), it must be multiplied by the circumference, with appropriate conversion factors for time units, in order to specify the tangential velocity. This outer velocity will be reduced radially within the cylinder to represent solid body rotation. Negative velocities are commonly needed to reverse the rotation direction without redefining the cylinder. A less common option is radial flow (Rmin to Rmax): a velocity can be specified at the inner radius of a hollow cylinder (e.g., a tube). If the inner radius is zero (e.g., a solid cylinder), this radial velocity will be ignored. Otherwise, the radial velocity will decrease as it is extended radially to the nodes at the outside of the cylinder, representing conserved mass flow (i.e., an implicit assumption of incompressible flow is made). Usually, only one such velocity is defined. However, there is no restriction on defining multiple velocities, which will then be superimposed. One possible example is the definition of both an axial and circumferential term, representing perhaps a helical screw motion. 9.5.1.3

Solid Sphere

Normally, only a rotating sphere (see “Solid Sphere” on page 4-55), whether hollow or not, can be represented by specifying a tangential velocity at the outer radius (meaning the equator, whether the sphere spans this equator or not). Like the corresponding option on a cone or cylinder, this velocity (Gmin to Gmax) represents solid body rotation about the Z axis. Negative velocities are commonly needed to reverse the rotation direction without redefining the sphere. For hollow spheres, a radial term (Rmin to Rmax) can be added, defined as uniform at the Rmin surface. If the inner radius is zero, the velocity will be ignored. As with a cone or cylinder, this radial velocity will decrease as it is extended radially to the outside of the sphere, representing conserved mass flow (i.e., an implicit assumption of incompressible flow). For truncated spheres, an axial (Bmin to Bmax, from north to south pole) velocity term can be added. This velocity is defined as uniform at the Bmin surface. If Bmin is zero, any input axial velocity will be ignored.

9-12


9.5.1.4

Solid Cone

Truncated cones (see “Solid Cone” on page 4-52) are useful for representing changes in axial flow area, as depicted in Figure 9-5, with either solid cylinders or FloCAD wires at one or both ends connected via a or. Thus, the axial term (Hmin to Hmax, positive along the z axis) is the most common. This velocity is defined at the Hmin surface, and will grow or shrink as needed along the cone to represent conserved mass flow (incompressible flow). If Hmin is zero (a pointed cone), any axial velocity will be ignored. Tapered roller bearings can be represented by a cone using a rotational term (Gmin to Gmax). As with the cylinder, this rotation is specified as a tangential velocity at the maximum radius. If a rotation rate is defined (e.g., RPM), it must be multiplied by the circumference at the widest radius, with appropriate conversion factors for time units, in order to specify the tangential velocity. This outer velocity will be reduced radially to represent solid body rotation about the Z axis. Negative velocities are commonly needed to reverse the rotation direction without redefining the cone. Flow may also exist radially, from Rmin to Rmax. As with the cylinder and sphere, this velocity is defined to be uniform at the Rmin surface, and is ignored for a solid cone. 9.5.2

FloCAD “Wires” (Wall-only Pipes)

As documented in Section 5.4 "Pipes" on page 5-39, FloCAD pipes allow a circular (solid or hollow), rectangular (again, solid or hollow), or arbitrary cross-section to be extruded along a complicated centerline. Analogous to a cylinder, material flow velocities can be assign in the axial, circumferential, or radial directions. With the great variety of cross sections and centerline definitions that are possible, FloCAD pipes represent a truly powerful means of specifying complex material flows along the centerline (e.g., conveyor belts, extrusions, etc. along the Hmin to Hmax direction). The only major restriction is that the cross section is constant along each pipe, though pipes can be ed end to end to other pipes (of different cross section) or to FD solids using ors. A minor restriction is that the radial resolution of a pipe wall (unlike the circumferential or axial resolution) is limited, unlike that of an FD cone or cylinder. The term FloCAD “wire” refers to a pipe with no fluid: a wall-only pipe. While this is the most common option when modeling material flow, there is actually no restriction against applying wall motion to a pipe with fluid flow, including a heat pipe with a moving (translating and/or rotating wall). Possible applications might include transpiration in porous pipe walls, or counterflow of cooling fluid into a hollow extrusion. The analogies with cylinders (Section 9.5.1.2) mean that most of that discussion need not be repeated here. The exceptions to this analogy include the handling of radial and circumferential flows for non-circular cross sections, which are circumstances but require detailed explanation nonetheless. A FloCAD pipe uses either one or two nodes radially, though the number of axial or circumferential nodes is -defined.


9-13

When a non-circular cross section is assigned a tangential velocity (Gmin to Gmax), true solid body rotation is not simulated. Instead, the velocity is taken to be a local value perpendicular to the thickness, as if the wall material were circulating instead of rotating (Figure 9-6).

Figure 9-6 The Special Case of Circumferential Material Flow in Noncircular FloCAD Pipe Wall Cross Sections

When a material flow velocity is assigned radially (Rmin to Rmax), the velocity is defined at the inner (Rmin) surface and ratioed by the outer to inner surface area to calculate the velocity at the outer (Rmax) surface. This is consistent with incompressible flow, even though the term “radius” cannot be applied to a rectangular cross section, for example. For highly resolved circumferential models of highly non-circular cross sections, this calculation is very approximate (consider radial flow near the corners of Figure 9-6, compared to that at the sides). While the Advection tab is available for other types of FloCAD pipes (pipes with walls, for example), the convection calculations do not currently for the relative pipe movement. 9.5.3

Using Conductors to Represent Material Flow

Single (node to node) one-way conductors (see “Function of Temperature Difference Conductor Type” on page 4-73) can be used to represent material flow that does not need to be represented geometrically, perhaps as needed to move material between sections that do need to be represented geometrically using the solids or pipes. To specify a single advection conductor, select the One way conductor button on the Conductor edit form, but don’t select the Use material option since the specific heat () of the material is needed, not the more commonly applied thermal conductivity. Instead, specify (perhaps via expression) the conductance as the mass flow rate multiplied by the specific heat, which has units of power per degree. Output the expression to SINDA/FLUINT if the flow rate is to vary during the run.2

2 However, if a dynamic run is to be made in which the area of the upstream surface (on a surface-to-node conductor) is to change, do not output the expression to SINDA/FLUINT, and leave the submodel as dynamic (vs. static).

9-14


The “from” and “to” define the positive direction for material flow. If the flow is ever to reverse during a SINDA/FLUINT run, an equal an opposite conductor will need to be added, with expressions controlling which conductor is active, and which is zeroed.3 A one-way generic conductor type (Section 4.7.1) can also be generated when flowing from a single node (perhaps the boundary condition at the inlet) to the entrance face of an FD solid or FloCAD pipe, which is particularly common as an inlet boundary condition. When the conductor is edited, the One way conductor checkbox must be checked. The Per Area option may be chosen if the flow is specified as a mass flux (mass/time/area) rather than an absolute flow (mass/time). This convenient choice allows the input expression to be written in of velocity consistent with the options in the Advection tabs of solids and pipes: velocity*density*. Unfortunately, a surface-to-node conductor does not currently exist in Thermal Desktop: no option is currently available to reverse the direction of the resulting one-way conductor to represent flow out of the surface and into the node. Such a connection is not normally needed at an exit (or a reversed-flow entrance), so the only problem arises if the material flow is to collapse to a single node/conductor stream at the outlet of some solid or pipe, perhaps later to expand back to a solid or pipe that is further downstream.4 To overcome this limitation, create a normal (two-way) node-to-surface conductor from the exit surface to an arithmetic node using a very small fake conductance (say, 1.0e-10). This node will then represent the average temperature of the surface without affecting the gradients within that surface, and so it can be used as the upstream node for a one-way conductor carrying material from the exit surface (which is not affected by material flow downstream of it). 9.5.4

Using ors to Connect Material Flows

Connecting the downstream surface of a solid or pipe to the surface of a downstream solid or pipe is a common need, met by ors (see “ors” on page 4-74). ors were used to connect the cones to the cylinders and pipes in Figure 9-5, for example. Analogous to node-to-surface conductors (see “Using Conductors to Represent Material Flow” on page 9-14), the One Way option is selected. By default, the mass flux per unit area is specified, which allows the input expression to be written in of velocity consistent with the options in the Advection tabs of solids and pipes: velocity*density*. Otherwise, select the Use Absolute Conductance option in order to specify the total conductance as the mass flow rate times the material specific heat (massflow*). Output the expression to SINDA/FLUINT if the flow rate is to vary during the run.5 The “from” and “to” define the positive direction for material flow. If the flow is ever to reverse during a SINDA/FLUINT run, an equal and opposite or will need to be added, with expressions defining which conductances are active and which are zeroed. 3 For example, “G = (vel > 0)? vel*dens*Area* : 0” for the forward flow conductor, and “G = (vel < 0)? abs(vel)*dens*Area* : 0” for the reverse flow conductor. 4 An example is provided in the next subsection: perfect mixing assumed between solids or pipes of differing cross section. 5 However, if a dynamic run is to be made in which the area of the upstream surface is to change, do not output the expression to SINDA/FLUINT, and leave the submodel as dynamic (vs. static).


9-15

When mating two bricks, or two cylinders, or a cylinder to a cone, it is strongly recommended that the nodalizations align... that the same resolution is used in both faces. However, this is not always possible (e.g., a FloCAD pipe end ed to a cone face), so care should be taken to make sure that the resulting turbulent-like “cross-flow” (caused by fact that flow from one upstream node will arrive at more than one downstream node, and vice versa) is tolerable. When the flow areas do not match (e.g., a rectangular cross section flows to a circular cross section), some amount of mixing is inevitable. A or may the most appropriate choice because the nodes on the fringes of one surface (e.g., the corners of a square) might not be appropriately connected if a uniform velocity is required. In such a case, consider a perfect mixing assumption: flowing into and out of an intermediate node using two nodeto-surface conductors back to back. (In this case, note that the small conductance and arithmetic node work-around must be used for the upstream “surface to node” conductor, as described in Section 9.5.3 Using Conductors to Represent Material Flow).

9.6

Super Network

The super network feature of Thermal Desktop is a unique and powerful tool used to reduce complexity in a thermal model. Using the method, a regular thermal network is separated into a sub-network and a super-network. Using the node editing dialog box, nodes may be specified to be in the sub-network. If a model contains nodes in a sub-network, the nodes not in the sub-network are referred to as being in the super-network. Only the super-network is sent to SINDA/FLUINT for calculations. Thermal Desktop automatically distributes the thermal mass of the nodes in the sub-network into the nodes in the super-network. The effect of conduction, radiation, and heat loads in the sub-network are also distributed into the super-network. Thermal Desktop retains information about the sub-network so that temperatures for nodes in the sub-network can be recreated from temperature data of the super-network. When SINDA/FLUINT results are post-processed, temperature data is automatically calculated for the sub-network. Thus it appears that SINDA/FLUINT is performing calculations on the full model, when in fact it is performing calculations on the simplified super-network. Super/sub networks are useful for reducing the complexity of complicated geometries. For example, consider a complicated finite element representation of a bracket that may have been imported from a stress model. Even though the bracket may be modeled with holes to save weight, and with extra flanges for stiffening, only the conduction through the bulk of the bracket is of interest thermally. In this case, the interior nodes of the bracket may be specified as being in the sub-network. A few nodes at the mounting locations of the bracket are left as is, which places them in the super-network. When the Thermal > Cond/Cap Calculations > Output SINDA/FLUINT Cond/Cap command is selected, Thermal Desktop will perform calculations to generate mapping information between the sub and super networks. This information will be used to distribute

9-16


the sub-network into the super-network. Only nodes in the super-network will be in the thermal model sent to SINDA/FLUINT. The effect of thermal mass, conduction, heat loads, and heaters present in the sub-network will all be factored into the super-network. If radiation conductors or orbital heating rates are being used, then the Output SINDA/ FLUINT Cond/Cap command must be issued before radiation conductors and orbital heating rates are output to SINDA/FLUINT. Radiation conductors and orbital heating rates may have been computed before calculating the sub-to-super network mapping information, but they must be re-output to SINDA/FLUINT after the sub-to-super network mapping information has been created in order to factor the sub-network into the super-network. Using the super network will make the SINDA/FLUINT calculations more efficient, since the temperature calculations will be done on a simplified thermal model. This can be important if many case scenarios or trade studies are to be performed. It should be noted however, that the specification of nodes to be in the sub-network has no effect on the efficiency of radiation calculations. Radiation calculations are performed with the original geometry, only the SINDA/FLUINT output is modified to reduce the sub-network from the thermal model. Consider the previous example of the bracket. If radiation calculations are to be performed, then an alternative representation of the geometry should be built for radiation. For example, the holes in the bracket could be modeled by a single surface with an effective transmissivity to for the effect of the holes. Thus, two representations would exist for the complex part, a simplified geometric representation for radiation, and the use of the sub-network to reduce the complexity of the conduction/capacitance network. The sub-to-super network mapping information is automatically generated by performing a series of steady state solutions for each of the super-nodes in the model. Using superposition, the temperatures of the sub-network can be recreated for any set of super-network temperatures. Steady state temperatures computed for the full thermal model will compare exactly to the temperatures calculated using the super-network and recreated sub-network.

9.7

Conductance Capacitance Parameters

The Thermal > Cond/Cap Calculations > Set Cond/Cap Parameters command displays the Set Cond/Cap Parameters dialog box and allows control for when conduction and capacitance calculations are made. The Set Cond/Cap Parameters dialog box


9-17

is shown in Figure 9-1.

Figure 9-1

Set Cond/Cap Parameters Dialog Box

SINDA/FLUINT file. When creating the node and conductor data for use outside of a Case Set (Section 15), the SINDA/FLUINT file is the name of the file to which the data is written when the Output SINDA/FLUINT Conp/Cap command is issued. This filename is not used for Case Sets. integration intervals. This value determines the number of test points for each direction when calculating conductance as defined on the thermal data forms (Section 4.3.1.5). For edge , this is the number of test point per node edge in the . For face , the value in this field is squared: there will be this many test points in each of the u and v directions or each node on the face included in the . Oct-tree Subdivisions. This value is the number of Oct tree subdivisions used for accelerating the calculations. This parameter is only used for edge or area as defined on thermal data forms (Section 4.3.1.5). The Oct tree subdivisions will not change the calculated results, but increasing the value to 5 or 6 may arrive at the results faster. Ignore lateral conduction in solid elements. This option can be used to eliminate cross conductors when the height of an element is small when compared to the width. This option will reduce the thermal network calculated by SINDA/FLUINT which should speed up the SINDA/FLUINT calculations. The height and width directions are based on the order in which the nodes are selected to define the solid element. For a brick, pick four nodes for the base, then four nodes for the top. The height is defined as the distance from the center of the base face to the center of the top face. The width used for the comparison is the average of the distance between the centers of the left-right faces and the front-back faces.

9-18


The or Restarts check box allows the to control if or restarts will be used. If the calculations for a or take longer than 5 seconds, then the program will store the inputs to the or and the results of the calculations to a file in the working directory. When the program goes to recalculate the or, the file will be opened, and if none of the inputs have changed, the results from the last run will be used. Please note that the inputs stored are not just from the or, but also from the geometry in the From and To sets of the or. There is also logic to test if just the or value has changed, and if it has, the stored results are scaled to be correct. Unchecking the or Linear Tri Spreading check box disables the default method of connecting tri-elements when using ors. By default, ors from tri-elements create weighted conductors for all three nodes. Without linear tri spreading, each triangular element is subdivided into three sections, one associated with each node. Advanced s might want to investigate the effect of unchecking this box. The may also specify that nodes that have been selected to be in the sub-network be distributed into the super-network. If the Use super-network option is not selected, the sub/super-networks are not created and a regular SINDA/FLUINT model is output.

9.8

Output SINDA/FLUINT Cond/Cap

Nodal capacitance and conductance thermal network values can be output to SINDA/ FLUINT with the Thermal > Cond/Cap Calculations > Output SINDA/FLUINT Cond/ Cap command. All output will be in the -defined units (Section 2.7.1) or the selected system of units for a FloCAD model. When this command is issued, Thermal Desktop will calculate conductors and capacitances for surfaces and elements, conductors for surface (Section 4.3.1.5) and ors (Section 4.8), generate logic for network elements such as heaters, and perform super-network calculations (if requested). Typically, the SINDA/FLUINT cond/cap data is generated as part of Case Sets (Section 15.2.2); this command is useful for reviewing the generated information or ing the node and conductor data to someone for use directly in a SINDA/FLUINT input file. Want "Hands-On" Information? Reference tutorial exercises "Circuit Board Conduction Example" on page 20-67, and "Dynamic SINDA Example" on page 20-201.


9-19

9-20


10 Radiation Calculations and Controls

10.1

Radiation Calculations and Output to SINDA/FLUINT

The Thermal > Radiation Calculations commands allow run-time and output parameters to be set, the initiation of either a radk, dialog box factor, or heating rate calculation run, and the output of data for manual input to SINDA/FLUINT. The nodes and positions to be calculated may also be specified. Calculations are performed for the default radiation analysis group, which is set using Thermal > Radiation Analysis Groups (see “Radiation Analysis Groups” on page 4-1). For automatic inclusion of radiation calculations in SINDA/ FLUINT, the radiation calculation parameters must be applied within the Case Set Manager (see “Case Set - Radiation Tab” on page 15-4). RadCAD® uses a stochastic integration technique (often called “Monte Carlo”) for computing radks, dialog box factors, and heating rates. Rays are emitted from each node and “traced” around the geometry. The rays simulate the effect of a “bundle” of photons. When a ray strikes another surface, energy is decremented from the ray and absorbed by the struck surface. The ray is then reflected or transmitted, according to the optical properties on the surface. RadCAD also has the option to compute radiation exchange factors from view factor data (view factors previously computed using ray tracing). A unique progressive radiosity algorithm is used to iteratively compute radks. The method optimizes calculations for those view factors that contribute the most to the energy balances for each node. The currently loaded optical properties are used, allowing radks for different optical property files to be computed using the same view factor matrix. The method does not require the view factor matrix to be normalized, since normalization is inherent in the raytracing and progressive radiosity algorithm. To compare using Monte Carlo methods for calculation of Radks versus using a radiosity method from factors, consider a simple cylinder as an example. Suppose that only the gradient along the length of a cylinder is of interest thermally, but its participation in the radiation environment is significant. Using RadCAD’s Monte Carlo methods, the cylinder can be nodalized along just the axial direction and accurate results will be obtained, since RadCAD’s raytracing method does not require the condition of uniform illumination. If the faster progressive radiosity method is used, care must be taken to subdivide the cylinder along the circumferential direction as well, since a node that curves 360 degrees cannot be illuminated uniformly by other nodes. This is a property of all radiosity based methods and is a result of the fact that the view factor is used to compute the energy exchange between nodes.

Radiation Calculations and Controls

10-1

By definition, the view factor assumes that the energy leaving a surface is diffuse and uniform. In order for the energy leaving a surface to be diffuse and uniform, it must emit uniformly (be isothermal) and must reflect uniformly. To reflect uniformly, it must be illuminated uniformly. Radiosity methods assume that all energy is reflected and emitted uniformly. Another way of thinking about the radiosity process is to visualize that any radiation impinging on a node will be forced to reflect uniformly from the surface of that node. If solar energy illuminates one side of the cylinder that has a node that spans the entire circumference, the reflected energy will be distributed uniformly over the surface of the node, and the cylinder will appear to be transparent, since energy will be reflected from the far side of the cylinder as well as the illuminated near side. This problem can be avoided by subdividing the cylinder into an appropriate number of circumferential nodes for radiosity based radiation calculations, and then combining the results back together for SINDA/ FLUINT using the Correspondence Manager. Heating rates may be computed using Monte Carlo, or by using a fast direct incident method coupled with the progressive radiosity method for computing reflected components. Direct heating rate computation also has an automatic error option, which adjusts the number of rays fired per node to meet a specified error tolerance. SINDA/FLUINT data may be output any time after calculations have been completed. Initial conductor ID, space node, and filtering parameters can be changed and a new set of data output without having to rerun calculations. Want "Hands-On" Information? Check out tutorial exercises "Radks for Parallel Plates" on page 21-3, "Space Station Oct Tree Example" on page 21-23, and "Orbital Heating Rates" on page 21-53. 10.1.1

Setting Control Parameters

Choosing Thermal > Radiation Calculations > Set Radiation Analysis Data invokes the Radiation Analysis Data dialog box shown in Figure 10-2. These forms can also be accessed from the Case Set Manager by Adding or Editing a radiation task on the Radiation Tasks tab in the Case Set dialog (Section 15.2.1). Rays Per Node. The maximum number of rays shot per node in the analysis group (Section 4.1). See Section 10.1.1.1 for more information. Desired Percent Error (or Weighted Error or Total Absorbed Error). Amount of error desired for calculations. Rays will be shot until the estimated error is below this value or until the maximym number of rays per node has been reached. See Section 10.1.1.2 for more information. Rays Before Initial Error Check. Number of rays shot before checking the estimated error. See Section 10.1.1.2 for more information. Energy Cutoff Fraction. A fraction of initial ray energy. When the ray energy drops below this factor, the ray will be absorbed or reflected statistically. See Section 10.1.1.3 for more information.

10-2


Figure 10-2

Dialog Box for Specifying Run-time Parameters for RadCAD® Calculations

Heating Rate Sources The check boxes in this region determine the heating rate sources used for heating rate calculations. If a box is unchecked then the heating rates for that source are ignored for calculations and output. With orbits centered about a planet or the Moon, solar optical properties are used for Solar, Albedo, Diffuse Sky Solar, and Diffuse Sky Albedo calculations and IR optical properties are used for Planetshine and Diffuse Sky IR calculations. If an orbit is heliocentric, then Solar and Albedo calculations are ignored and Planetshine uses solar optical properties for the calculations. Only Planetary Lat/Long/Alt heating environments (Section 6.1.4) use the Diffuse sources.


10-3

Nodes All. If All is selected, then the radiation calculations are performed for all nodes in the analysis group (Section 4.1). List. If List is selected the can list nodes in the analysis group that will shoot rays. The list format is described in Section 2.10.13. Validity of node names entered on the dialog box is checked when computations begin.

Positions All. If All is selected, then heating rate calculations are performed for all positions in the heating environment (Section 6.1.1.2). List. If List is selected the can list heating environment positions for which calculations will be performed.

Add and Delete The Add and Delete buttons will add or delete calculation sets within the current radiation task. The calculation sets appear as tabs beside the Default tab. Calculations are performed for each calculation set and the results are combined for output. For example, the may want to shoot 5000 rays for the whole model, but would like to shoot 10000 additional rays for a specific submodel PGA. Shooting 5000 rays would be accomplished by defining the Default tab as shown in Figure 10-2. Selecting the Add button, a second tab will appear. On the new tab, input 10000 for the Rays per Node, select List for the nodes, and type in PGA.* for the node list. Adding calculation sets is a useful way for decreasing the error for a select set of nodes without shooting as many rays for the entire model. In addition, radks from a single node to the rest of the model may be computed in order to simulate the node as a heating source. For example, setting the node to a sink temperature corresponding to the emissive power leaving the surface can be used to model heating lamps. 10.1.1.1

Rays Per Node

The number of rays desired to be emitted from each node is entered in the Rays per node field (Figure 10-2). The question often arises of how many rays should be shot for a given problem. Detailed statistical approaches are possible for estimating the error based on the random variations of each ray firing and are covered briefly here. The optimum approach relies on a basic understanding of the error estimating techniques and, more importantly, an understanding of the expected results of the thermal model. Radks are computed by keeping track of the energy emitted from a node and the energy deposited into each node during the course of tracing a ray around the geometric model. The ratio of the energy absorbed by a node to the energy emitted for a single ray can be considered as an estimate of the radiation exchange factor, or in statistical , an experimental trial. The average, or mean of these trials gives the final result.

10-4


The basic approach to error analysis is to estimate the variance in the mean of the sample of ray firings. This variance, in conjunction with a confidence interval approach1, may be used to generate a percentage error number. The variance of the mean of the energy tallies can be estimated in two ways. One method involves maintaining not only the sum of the energy absorbed, but also the sum of the square of the energy absorbed. These quantities are used in the standard formula for sample variance. However, this approach incurs a substantial overhead in maintaining an N by N matrix of squared energy tallies. A simpler approach is to consider the ray tallies as a discrete distribution. That is, for an individual ray, the energy is either completely absorbed by a node or it is completely reflected (or transmitted). The discrete distribution approach usually gives an error estimate that is higher that what would be obtained if information on individual ray trials were maintained. Since the discrete distribution approach is simple and gives a conservative estimate of the error, it will be used here. The following formula yields an estimate of the percentage error of an individual radiation exchange factor using a confidence interval of 90%:

1 – B ij Error ij = 1.65 ------------------ 100 N rays Bij

If the exact error were known, then so would the exact answer. All we can be 100% confident of is that the true answer lies between plus and minus infinity. A more practical statement of estimated error can be made using confidence intervals. A confidence interval of 90% may be interpreted as: “if the problem were performed many times, then 9 times out 10 the actual error in the calculation will fall within the bounds estimated by the above equation.” A 90% confidence interval is recommended because it matches well with the actual error produced when comparing results to analytical solutions. The Bij is the fraction of energy that leaves node i and is absorbed by node j by all possible reflection paths. The sum of Bij’s for a node is equal to unity. The Bij multiplied by the surface emissivity and area becomes the standard radiation conductor input to SINDA/ FLUINT. A Bij of one half means that node j absorbed half of the energy emitted by node i. The error equation highlights two important facts. One is that the error is inversely proportional to the square root of the number of rays shot. The number of rays will need to be quadrupled to cut the error in half. The other is that the error depends on the magnitude of the Bij. Large Bij’s are the result of capturing a large number of emitted rays, and hence the statistical sample population is much larger than a smaller Bij, yielding a lower variance and error estimate. 1 Kreyszig, E., Advanced Engineering Mathematics. 4th ed., Chapter 20, New York: John Wiley & Sons, Inc., 1979.


10-5

A simple rule of thumb in deciding the number of rays to shoot is to consider the “average” Bij in a problem. If for a particular node, the “average” Bij is one half, shooting 1000 rays gives an average error of about 5%. If you know that a particular radiation coupling is critical to the accuracy of the thermal simulation, say a small cryogenic surface looking at a small hot surface, then modify the error estimate based on expected values of Bij for the critical component. Another item to consider that is often overlooked is that energy is conserved globally for a node. The sum of the energy that is deposited in all of the nodes is exactly equal to the amount of energy that is emitted. For example, consider a small rectangular region “looking” at a much larger rectangle as shown in Figure 10-3. Suppose the small rectangle consists of one node and the larger rectangle is subdivided into many small nodes. If a small number of rays are shot, then the error in the individual radks from the node on the small rectangle to each of the small nodes on the large rectangle will be high. However, the nature of the process ensures that if one radk is too low, another will be too high. Globally, the error is lower than when considering the error in individual radks.

Figure 10-3

Rays Shot From a Small Rectangle to Nodes on a Large Rectangle

The error in the sum of the radks from the node on the small rectangle to all of the nodes on the large rectangle is much lower than what would be expected by looking at individual radks. As the summed Bij increases by summing the radks from node i to multiple j’s, the summed error drops according to the error equation. A larger Bij has a lower error. In practice, what needs to be considered is the dominance of conduction or radiation in the problem. If the large rectangle in the aforementioned example has a very high conductance, then the large rectangle will have close to a uniform temperature. Too much energy that is transported by radiation (because of error) to one small node will be conducted to a node that has too little energy transported to it by radiation. If the conductance is high, the temperature gradients due to the error in the radks are low. The large rectangle behaves thermally as if one large radk existed between the small node and the entire large rectangle. Error in the individual radks is not as important as the error to the entire rectangle, since the rectangle will run at essentially a uniform temperature and the effective radk is the average radk. On the other hand, if there was no conduction between the nodes on the large 10-6


rectangle, then the error in each of the individual radks becomes more important. The average temperature of the large rectangle, even though there is variation between nodes, is still more accurate than individual nodal temperatures, because energy is conserved globally. A node that is too hot is balanced by a node that is too cold. The most useful advice follows from the most basic principle of thermal analysis: you should have some idea of the what the temperature solution should be before the numerical analysis is performed. Look at areas of the model that should have a symmetrical temperatures. Estimate the error in the model by looking at the temperature difference between regions that should be the same, then shoot more rays to bring the temperatures together within an acceptable range. Look at areas of the model that should have temperatures that vary monotonically, for example, the temperature should drop smoothly as you move along a heat fin. If the results show an erratic distribution, then shoot more rays. It should be ed that the error in the “bulk” temperature of the model is much less than individual components. This allows you to shoot a small number of rays to get a “peek” at the answers in a short time. Because RadCAD allows restarts, more rays can be shot at a later time to refine the accuracy of the simulation. The final thing to be considered is that rays are shot from both nodes i and j, resulting in two estimates of the radk between them. Using reciprocity, the results can be combined to yield a result that has less error than either individual radk. For example, if the areas and emissivities of the two nodes are the same, then Bij equals Bji and combining the ray tallies yields a result that has an error reduced by the square root of two. If the areas of the two nodes differ significantly, the errors could also be significantly different. If the large rectangle in the example shown in Figure 10-3 was a single node, then the error in the radk from the small node to the large node will be much less than radk computed for the large node to the small node. The errors differ because a larger percentage of the rays from the small node strike the large node, giving more samples to reduce the error in the average. Most of the rays from the large node miss the small node, giving a smaller sample size with a higher error. RadCAD takes this factor into when combining Bij and Bji to compute the final radk for SINDA/FLUINT. The ray tallies are weighted by their areas and error estimates, and always produce a radk that has less error than either individual estimate. The main point being presented here is that one should not rely too much on mathematical formulas for estimating the error in radiation exchange factors. Rather, be familiar with the expected results, and look for variations in the results that indicate an inadequate number of rays. If it is critical that accuracy be demonstrated, run a series of cases with an increasing number of rays until the temperature results no longer change. A few of these exercises will quickly enable one to obtain a “feel” for what an adequate number of rays are for a variety of situations. The total sum of the radks/view factors for a node can also be used to estimate error. The total sum should equal the area multiplied by the emissivity. The SINDA/FLUINT output file lists data at the end of the file useful in determining the accuracy of the results.


10-7

The total sum for each node is listed as a percentage of maximum. Values above 95% for all nodes should give acceptable results. Values significantly lower than this usually indicate exchange with inactive surfaces. This approach to error analysis also applies to computing orbital heating rates using the monte carlo method. For heating rates, rays are shot from the node to the source. If the ray is not absorbed by intervening surfaces, the ray is reversed, direct energy is deposited into the node and the ray is reflected. The traversal of the ray for the reflected energy is the same as traversing the ray for radk calculations. Both the error in the direct heating component and the reflected components depend on the number of rays shot. Using Thermal Desktop’s postprocessing functions, the heating rates can be examined in color on the geometric model. Similarly to analyzing error in radks, look for regions in the model that should have symmetric or smoothly varying heating rates. Increase the number of rays until the heating rates show the expected trends. RadCAD also contains a direct incident heating rate option that uses the view factor matrix along with RadCAD’s progressive radiosity algorithm to compute total absorbed heating rates. Direct components are computed using monte carlo, but rays are not reflected, since the reflected components are derived from the view factor matrix. In this case, the error can be easily estimated during the direct energy computations, since only one parameter is being computed. If the error parameters are used, the rays per node parameter gives the maximum number of rays that will be shot per node. Computations will terminate whenever the error tolerance is met, or if rays per node have been shot. 10.1.1.2

Automatic Error Control

The next two parameters are for specifying automatic error control for radk, view factor, and heating rate calculations (refer to Figure 10-2). Note that the error control has a different meaning in radk calculations than it does for heatrate calculations. The value for the field is percent: a value of 10 means 10% error. The forms in Thermal Desktop have slightly different wording for the error control field: Percent Error Desired, Weighted Error and Total Absorbed Error are each used depending on how the Radiation Analysis Data form was opened and for which analysis the form is meant. For radk calculations, rays will be fired until a weighted percent error is obtained for each node, or until the maximum number of rays per node have been fired. RadCAD uses a combination of a confidence interval technique and a weighting scheme based on the relative value of the radks, view factors, or heating rates to estimate the error (to avoid wasting U time computing insignificantly small values down to a low error value). For heatrate calculations, the error criteria is based on the total absorbed value for the source being calculated. The program will first shoot rays for the selected nodes based on the value input for “Rays Before Initial Error Check”. Once that has been done, the program will calculate the total absorbed for all the surfaces and will determine which surfaces are not in the error criteria input. For those surface, the program will figure out if the error for those surfaces is based on the direct energy to those nodes, or in some reflected energy from another node. RadCAD will then shoot more rays for nodes with direct illumination that are

10-8


causing the error criteria not to be met, with a maximum being the “Rays Per Node” input. Only directly illuminated nodes are used since only rays that “see” the source directly are propagated through the model. If a value greater than zero is input for the Percent Error Desired, automatic error tolerance checking will be enabled. Rays will be shot until the desired error tolerance is achieved, or until rays per node rays have been shot. To ensure that a reasonable number of rays have been fired before performing an error estimate, the Rays Before Initial Error Check should be set to a number that provides a good basis for the solution; too small a number is not advised. A minimum of 2000 is recommended for Rays Before Initial Error Check. For example, suppose a flat surface is shaded by an intervening surface by 40%. If the error was computed after only two rays, a reasonable probability exists that both rays will see the source (or both be blocked). Since there is no variation in the samples, the error estimate will be zero. On the other hand, if there is no intervening shading, only one ray is necessary for zero error. The default value of 200 rays is a good compromise, but if your model contains little shading effects, the value may be lowered. Partially obscured curved surfaces may need more rays to ensure a good starting sample, perhaps on the order of 1000 rays. A built in estimator is used to predict the number of rays necessary for the desired error tolerance so as not to use excessive U time performing weighted error calculations. The error tolerance for radk calculations only applies to the current calculation run. Rays shot for previous calculations will not be used in the estimation. The general rule of thumb regarding error applies: to halve the error, quadruple the number of rays. If a 5% error run is executed four times, the result will be equivalent to a single 2.5% error run. Please note that Error Control should not be used when the has defined a fast spinning surface. This would conflict with the fast spin averaging and give incorrect results. (see “Setting Fast Spin Parameters” on page 10-29) 10.1.1.3

Ray Cutoff

A ray that is emitted from a node will continue to propagate around the model until its energy level reaches zero. A ray’s energy reaches zero when it strikes a surface that is perfectly black or escapes to “space.” If a ray is emitted in an enclosed geometry, with surface absorptivities less than 1.0, the ray energy will be continually decremented and will never reach zero. One possible way of preventing an infinite loop like this would be to simply ignore ray energies once they reach some low level. However, errors can compound to an unacceptable level. Fortunately, there is a better method. The absorptivity of a surface can be viewed in two ways with respect to propagating ray energy. The first way is to use the absorptivity to decrement the ray energy. The second way is to view the absorptivity as the probability that all of a particular ray’s energy will be absorbed for this surface strike. Both views are consistent with a Monte Carlo approach. In the limit of an infinite number of rays, both techniques produce the same result. The effects are somewhat different when comparing a finite number of rays used for each technique, or more importantly, when the same amount of U time is used for each technique.


10-9

For an equal amount of U time, the decrementing technique will distribute the error more uniformly among all of the surfaces in the model. The probability approach (where a ray is completely absorbed or reflected) tends to distribute the error more towards surfaces farther away from the surface that emits the ray when emissivities are high, since rays are more likely to be terminated early. The Energy Cutoff Fraction specifies at what fraction of the original energy level the ray energy absorption scheme changes from partial energy absorption to an all or nothing scheme. This hybrid method ensures that all ray energy is absorbed by the radiation model, without getting stuck in an infinite loop. If the Energy Cutoff Fraction is set to 1.0, the all or nothing method is always used. If the parameter is set to 0.0, the ray energy is always partially absorbed (be careful!). In general, the effect of this parameter is minimal and can therefore be left at the default setting. The advanced may be able to optimize U time by careful selection of this parameter, but it is a more of a safeguard against rays getting stuck in an infinite loop. Most of the time rays are terminated normally by escaping the radiation model or by hitting a black surface. For enclosures with very low emissivity, a higher cutoff fraction may reduce U time. The Energy Cutoff Fraction is used only during Monte Carlo calculations of Radks or Heating Rates. It is not used for the calculations of form factors. 10.1.2

Setting Advanced Control Parameters

There are four basic advanced run parameters for controlling radk runs, available by selecting the Advanced Control tab on the Radiation Analysis Data dialog box. These parameters are shown in Figure 10-4. Before changing any of these parameters, especially the bottom 3, please have a complete understanding of what you are actually doing. The first set of parameters (Oct Cells) are for Oct cell control. Please see “Oct-Tree Parameters” on page 10-13 for a complete discussion. The second set of parameters is the Random Number Seed Control field. The option Use unique random number seed at the start of calculations causes the random number seed to be a different value each time calculations are performed (using the time of day). This allows restarts to improve the accuracy of calculations. If the same seed was used, the rays would retrace the same paths and the results would not change. When using the following two options, be sure to return to random number seeds (this option) before resuming normal calculations! Occasionally, however, consistency in the ray tracing is desired. If the difference between two cases is of importance (e.g. different optical properties), the seed can be reset to the same value before each calculation begins by selecting the Use same random number seed sequence at start of calculations radio button. The idea being that the error due to the random nature of the calculation method can be subtracted out from the two cases, if the random error is essentially the same. This method will have the best results with the Direct/VF heating rate method, since the optical properties can affect the ray generation sequence (via the ray cutoff, probabilities to transmit or reflect, etc.). 10-10


Figure 10-4

Radiation Analysis Data Dialog Box Advanced Control Tab

In the case of articulating radks, consistency in the non-articulating regions can reduced the number of time-varying radks. For this behavior, select Use same random number seed sequence at start of every node. Note: The use of parallel processors, while beneficial to the solution speed, may affect the behavior of reusing random seeds. The third set of parameters (Nodalization Schemes) is the node name mode. This parameter changes the nodalization used for calculations. The primary purpose of this input is to allow for radiosity methods to be used on double sided surfaces. Take a look at the example shown in Figure 10-5. The model consists of two triangular elements, which are active on both sides. In the General node name mode, this model is 4 nodes on the database. The top and bottom sides are combined into a single database location for each node and nodes 2 and 3 on the left and right elements are also combined. This is perfectly valid when calculating Monte Carlo (MC) Radks and Heating Rates, since the


10-11

reflections from surfaces are made from the intersection point of the surface and the incoming ray. This nice little technique saves a lot of disk space and memory for MC Radks and HR’s, but does not work as well when radiosity methods are to be employed for active both surfaces. The assumption of radiosity is that the incident energy is equally distributed about the surface receiving the energy. Since the General mode sums these the top and bottom (along with like node numbers), energy that hits the top side of a surface is being distributed about both sides for radiosity methods. 2

4 1 3 Figure 10-5

Advanced Run Parameters - Node Name Mode

This radiosity problem is alleviated with the Specific mode. This model is 12 nodes on the database (6 for each element) in the specific mode. By storing 12 different values for the energy, 6 per element (3 for each side, both sides), the radiosity method will properly reflect energy from the side and surface that receives the energy. The Specific mode works fine, but quickly increases the number of nodes on the database, which in turn increases the size of the file and the amount of memory required to perform the analysis. The Top-Bottom Specific mode is meant to use less nodes, but s must determine if their model can take advantage of this option. In the Top-Bottom Specific mode, this model would be 8 nodes on the database (nodes 2 and 3 would be combined for the left and right elements.). This method works fine as long as there is not another element coming off of nodes 2 and 3. If there is the additional element from 2 to 3, then the Specific mode would have to be used. The fourth set of parameters (Radk Calculation Spectrum-Used for Modeling Lamps) allow the to specify the spectrum used for the calculation of Radks. The can select to use the Solar spectrum if a solar source is present in their model. A good example of this might be a solar lamp used for thermal vacuum testing. This option should be used in conjunction with the “Output Radks as Heating Rates” option on the Radk Output page. The last parameter, Wavelength Dependent Properties, allows banded radiation analysis to for wavelength-dependent optical properties (non-grey optical properties). Checking this option and selecting Edit opens the Wavelength Bands for Calculations dialog (Figure 10-6). The inputs on this dialog determine the wavelength bands used for calculating non-grey radiation.

10-12


Figure 10-6

10.1.2.1

Wavelength Bands for Calculations

Oct-Tree Parameters

The oct-tree method2 is a technique to significantly reduce run times by eliminating unnecessary ray/surface intersection tests. CRTech has enhanced this method with a unique approach to traversing rays through the oct-tree structure. The approach optimizes testing as the rays near edges and corners of the cells, yielding improvements over previous oct-tree methods. In a Monte Carlo approach, a ray is emitted from a surface and the task is to find which surface the ray hits. Since the closest surface must be found, all surfaces in the model must be tested. Time to solve a model grows exponentially with model size. This single factor has limited the use of Monte Carlo methods as compared to view factor / radiosity methods.

2 Panczak, T., “A Fast, Linear Time, Monte Carlo Radiation Interchange Program Utilizing Adaptive Spatially Coherent Subdivision”, Proceedings of the Sixth International Conference on Numerical Methods in Thermal Problems, p. 702, Swansea, U.K., July 3-7 1989, Pineridge Press, Swansea.


10-13

In the oct-tree method, the geometry of the radiation model is partitioned into separate boxes, or “cells”. The “oct” part of the method is named for the way in which the geometry is partitioned. Initially the geometry is completely enclosed by a single box. This box is subdivided into eight (“oct”) smaller boxes by cutting along the three principle planes of the larger box (see figure Figure 10-7). Each of the smaller cells is then examined to see how many surfaces it contains. If a cell contains too many surfaces (as set by the ), it is subdivided again in the same manner.

Figure 10-7

Oct-cell bounding volume subdivided into eight smaller cells

The subdivision process continues until each of the cells contains only a small number of surfaces. The cells “adapt” to regions of the model that are geometrically complex by subdividing into smaller cells. Regions of the model that contain a small number of surfaces are bound by larger cells. The hierarchical arrangement of “oct-cells” is called an “oct-tree”. The goal is to partition the geometry into separate, contiguous regions, each with a small number of surfaces. When a ray is first emitted, only surfaces in the cell containing the ray origination point are tested. If no intersection is found, the ray is propagated into the next cell along the ray’s path. Surfaces in this cell are then tested for intersection. This process continues until the ray strikes a surface or leaves the geometry entirely (escapes to space). Instead of testing all surfaces in the model, only surfaces in the immediate vicinity of the ray’s path are tested. The effect is to change the time complexity of the problem from exponential to closer to linear. Doubling the number of surfaces in a problem only doubles the U time required, instead of quadrupling it. Two parameters are used to control the subdivision process, Max Oct-tree subdivisions and Max surfaces per cell (see Figure 10-4). Each cell will be subdivided until it contains no more than Max surfaces per cell, or until the cell has been subdivided Max Oct-tree subdivisions. Imagine a point in a geometric radiation model where three surfaces intersect, such as the corner of a cube. If it is desired to have no more than two surfaces per cell, then cells would subdivide forever at the corner of the cube. The Max Oct-tree subdivisions parameter prevents this infinite loop condition. 10-14


Since it costs U time to determine the next cell along the ray’s path, there is a trade off between the depth of the tree (number of subdivisions) and the maximum number of surfaces desired for each cell. Subdividing the tree too finely will start to increase U time because more time is spent moving the ray from cell to cell than is eliminated by reducing surface intersection tests. In general, the acceleration due to using oct-cells is a weak function of both parameters. Almost any oct-tree will provide dramatic improvement compared to an unaccelerated run. However, a few rules of thumb may be employed. Each time the tree is subdivided, the cell size is halved. A tree that has been subdivided 10 times consists of cells that are the size of a single pixel on a high resolution monitor (1/1024 of the original size)! A tree subdivided five times consists of cells that are 1/32 of the model’s size, with a total possibility of over 32,000 cells. This is usually an adequate number of subdivisions. Occasionally models will warrant more subdivision, but more than 7 levels usually shows little improvement, and may start to increase run times. The Max surfaces per cell parameter should be set by considering how many surfaces intersect at a single point in the model. A truss structure for example, could consist of many surfaces intersecting at a common point. Set the Max surfaces per cell parameter to be one or two more than the maximum number of surfaces that can be found in a small region (a region smaller than the smallest desired oct-cell). This will prevent the oct-cells from subdividing too finely. Again, the default parameters are normally adequate, and will reduce U time significantly. Models with extreme scales (that is, a large model with small pockets of high complexity) may warrant a larger number of subdivisions. Models with regions that have many surfaces intersecting at a point (or contained within a small region) may benefit with a larger number for Max surfaces per cell.

Optimize Cells To find the optimum Oct-Cell setting for an analysis group, select Thermal > Radiation Calculations > Optimize Cells; the Optimize Cells form opens (Figure 10-8). Optimum Oct-Cell settings are dependent on the analysis group, so the must select an analysis group from the Analysis Group drop-down list. The then specifies a lower (From) and upper (To) limit for subdivision and surfaces per cell in the Vary the Subdivisions and Vary the Surfaces Per Cell fields, respectively. The specifies the increment for each variation. The increment must be greater than zero, but the may specify the same From and To values to prevent variation of that parameter. The must also specify how many rays to shoot for the optimization. This value will be a relatively small number (the default is on the order of 200). RadCAD adjusts the Oct-Cell parameters, shoots the requested number of rays for each combination and records the times for each step in the radiation calculation process. The random seed for ray tracing is reset for each Oct-Cell parameter combination to limit the variation of the ray tracing time due to different ray propagations. The results of the optimization are written to the command line and a text file in the working directory, OptimizeCells.txt (an example is shown in Figure 10-9).


10-15

Figure 10-8

Optimize Cells Dialog Box

The first section of the file provides the times for Database setup I/O time, Ray tally I/ O time and Total I/O time. These times are not affected by the Oct Cell parameters. The second section of the file provides the oct-tree generation time for each pair of octcell parameters. This time is dependent on the oct-cell parameters but independent of the number of rays. The third section of the file provides the maximum number of surfaces in any cell in the oct-tree. A decrease in this value by an order of magnitude or more should significantly reduce the ray tracing time. The last section of the file provides the ray tracing time for the requested number of rays. This value will scale with the number of rays used for the solution calculations. These numbers are good indications, but could be affected by other applications running on the same computer. Adjusting the number of rays for cell optimization such that the ray tracing times are greater than about 30 seconds will provide the best comparisons and estimates.

10-16


Figure 10-9

Text output for Optimize Cells, OptimizeCells.txt

If increasing the number of subdivisions decreases the maximum number of surfaces per cell (the third section of the OptimizeCells.txt file) by an order of magnitude or more then the higher number of subdivisions should significantly increase the speed of the run. Radiation Calculations and Controls

10-17

If that number changes by a small amount, then subdividing the model with Oct-Cells has little effect. Ideally, the maximum surfaces in any one cell should be less than 100, therefore, small models gain very little with Oct-Cells. Want "Hands-On" Information? Set Oct-Cell parameters and use the Optimize Cells command in "Space Station Oct Tree Example" on page 21-23. 10.1.3

Setting Radk Output Parameters

The end product of RadCAD is one or more input files to be used with SINDA/FLUINT. Two tabs on the Radiation Analysis Data dialog box are used to control the output of radk data, one for constant output, and another for Radks versus time. The two tabs, Radk Output and Radk Time Vary Output, shown in Figure 10-10 allow the specification of parameters used in the generation of SINDA/FLUINT input files. The data is output in the current units (see “Units” on page 2-25). 10.1.3.1

Radk Output Tab

The Radk Output Tab (Figure 10-10) allows specification of file names for form factor and radk files, as well as, parameters for filtering and numbering radks. SINDA/FLUINT files may be created automatically after calculations are performed, or manually at any time. Different numbering or cutoff factors may be used without having to recalculate data. RadCAD does not impose any particular constraints for the usage of SINDA/FLUINT submodels. The output file produced by a radk run may be placed in a separate submodel or combined with radiation conductors generated from a different analysis group. The file simply consists of the node-to-node radiation conductors for the nodes contained in the analysis group that was analyzed. For radk output, the file consists of node-to-node radiation conductors for all of the nodes contained in the analysis group being analyzed. The file explicitly uses the submodel name for each node ID, since an analysis group may contain nodes in different submodels. The conductor ID’s start with Initial Conductor ID and are incremented by one. If the radks are for variable geometry, the radk file name will be appended with an ‘l’ for logic data and an ‘a’ for array data. Radks to space are output to the input Space Node Submodel ID. When the node definitions are output for SINDA via the Cond/Cap command (see “Output SINDA/FLUINT Cond/Cap” on page 9-19), a test is performed to determine if a graphical node for the space node exists. If the node does not exist, then a boundary node with a temperature defined in the Space Node Temperature field is created. The space node temperature defaults to 2.73 K (4.91 oR) 3. RadCAD also contains a parameter, the Bij/Fij Cutoff Factor, used to reduce the number of SINDA/FLUINT radiation conductors, thereby reducing SINDA/FLUINT runtime. 3) In previous versions of Thermal Desktop, the default for the space node temperature was Absolute Zero: we have made this change to be consistent with space physics.

10-18


Figure 10-10

Radiation Analysis Data Dialog Box Radk Output Tabs

Recall that the Bij is the fraction of energy emitted by node i that is absorbed by node j, via all possible direct and reflected paths. Because we are assuming grey behavior, the Bij’s are also reciprocal: eiAiBij equals ejAjBji. Rays are shot from both node i and node j, and are combined to yield a final radiation conductor for SINDA/FLUINT (see “Rays Per Node” on page 10-4). A radiation conductor between nodes i and j is considered inconsequential when it is less than a certain fraction of both node’s radiative energy balance. The Bij represents that fraction. When a radk has both a Bij and a Bji less than the Bij/Fij Cutoff Factor, it is culled from the SINDA/FLUINT input file.


10-19

A few rules of thumb can be followed when selecting a cutoff parameter. The more nodes in the problem, the smaller the cutoff fraction should be, since the individual Bij’s tend to be smaller. A good place to start is one divided by ten times the number of nodes in the Radiation Analysis Group. The estimated spread in Bij’s for the nodes in the problem must also be considered. For example, consider the extreme case of the inside of a sphere with equal sized nodes. In this case, all Bij’s between nodes should be exactly the same (1/n, where n is the number of nodes). A Bij of zero should be used (or less than 1/n), since all Bij’s are equally important. Picking a cutoff greater than 1/n will remove all radks (ignoring variations due to statistical error). At the other extreme, for models whose temperatures are dominated by a few large nodes, a higher cutoff fraction can be employed. A small SINDA/FLUINT run can also be performed to that excessive culling has not been done. Perform a limited transient run with both a complete radk file and one that has been filtered. If the results do not differ significantly, perform the longer transient simulation with the filtered set of radks. Data written by RadCAD at the end of the SINDA/FLUINT input file can also be used to determine if excessive culling has been performed. The total sum of radks for each node is indicated as a percentage of the maximum sum. Percentage sums that are too low can indicate excessive culling of radks. Care should be taken since low sums can also mean views to inactive surfaces or an insufficient number of rays being shot. Overall, percentage sums above 95% indicate that calculations and culling were carried out satisfactorily. The List summary if %kept is off by more than field can be used to filter the output so that only erroneous data values may be printed. Also note that a complete summary of the radk data is written to an Excel file of the same name as the radk file, just with the .xls extension. This excel file has all the data for all the nodes and within excel, it is easy to sort the data versus the various parameters. A second parameter, Output Bij’s Until Bijsum >=, used with Bij Cutoff Factor allows more control over which radks are output to SINDA. An example of how this parameter works is for the to increase the Bij Cutoff Factor to say, 0.05. All radks greater than the cutoff factor are output to SINDA. The sums of these are added together. For the nodes whose sum is less than the input value for Output Bij’s Until Bijsum >=, additional Bij’s will be output that are smaller than the cutoff factor. This additional Bij’s will be added until the sum reaches the desired value. The goal of this parameter is to generate less radks for SINDA, but give the a confidence that most of the radiation energy is being ed for. The Output 1-BijSum to ERN Node parameter will output a radk to the ern node that is equal to 1.0 minus the sum of all the other Bij’s that have been output to SINDA. If using the Case Set Manager, the ERN node will automatically be output as an arithmetic node in the SINDA model. If you would like it to be a diffusion node or something else, create the node in the model and then the Case Set Manager will not create that node.

10-20


The Disable Radks to Space check box stops the program from outputting any radks to space. This is useful for situation where the analysis group is entirely enclosed, however, there might not be surfaces in the groups that actually complete the enclosure. Note that any radks to space are still written to the .k file, but that they are simply comments and not part of any temperature calculations. The check box to Output as Heating Rates with the Edit function allows the modeling of heating sources. The source can be constant or a function of time. The must specify the submodel.id for the source. Likewise, the can shoot rays only from the source if desired. If the source is in the solar spectrum, the needs to change the spectrum on the Advanced Control tab (see “Setting Advanced Control Parameters” on page 10-10). To define a radk job in the Case Set Manager, select the Radks radio button and the radiation analysis group on the Job tab, as described in "Case Set - Radiation Tab" on page 15-4. 10.1.3.2

Radk Time Vary Output Tab

The Radk Time Vary Output tab (Figure 10-11) contains the parameters used to control the output of time-varying radks. Time-varying radks (also referred to as articulating radks) can only be calculated if an orbit has been defined since they are based on the orbit time, not SINDA/FLUINT solution time. Articulating radks are typically, although not necessarily, calculated when a tracker has been defined, but may also be necessary if translation or rotation of objects or assemblies has been defined using the orbital internal symbols (Table 11-2 on page 11-9). The Variable Radk Determination Factor is used to determine if a radk is time varying or if it is constant. The model must have an orbit defined and have articulators for this factor to be used. The process to determine if a radk is time varying is as follows: 1) The time average radk is calculated. 2) The average radk is compared to the radks at each time in the orbit, if the radk at any time is off by the determination factor percentage, then the radk is deemed time varying. In the *.k file, radks that are determined to be time varying will be tagged with ‘TV’ after the ‘$’ and be fore the Bij value. Radks without the ‘TV’ tag are considered to be constant. For articulating radks, the can control the array formats to be Binary or ASCII. The binary format makes SINDA run a little faster, especially in the pre-processor. The ASCII version uses SINDA array data to hold the time varying radks. The ASCII file has an ‘a’ appended to the file name (typically ‘ka’), and the binary file has ‘ab’ appended to the file name (typically ‘kab’). Both files are always output, but SINDA only uses one of them depending on which is selected (Binary or ASCII). Finally, the Advanced Logic Control fields allows the more control over when the radks are loaded into SINDA. Please see “Radiation Advanced Logic” on page 10-24. To define an articulating radk job in the Case Set Manager, select the Articulation Radks radio button, the radiation analysis group and the orbit on the Job tab, as described in "Case Set - Radiation Tab" on page 15-4.


10-21

Figure 10-11

10.1.4

Radiation Analysis Data Dialog Box Radk Time Vary Output Tab

Setting Heating Rate Output Parameters

The end product of RadCAD is one or more input files to be used with SINDA/FLUINT. The Heatrate Output tab, shown in Figure 10-12, allows the specification of parameters used in the generation of SINDA/FLUINT input files. The data is output in the current units (see “Units” on page 2-25). Names for orbital heating rate files and submodels are specified using this dialog box. SINDA/FLUINT files may be created automatically after calculations are performed through the Case Set Manager, or manually at any time using the Thermal > Radiation Calculations submenu. The manual mode is provided so that different numbering or cutoff

10-22


Figure 10-12

Radiation Analysis Data Dialog Box Heatrate Output Tab

factions may be used without having to recalculate data. Orbital constants such as the solar flux, albedo, and planet temperature may also be changed (using Thermal > Orbit > Edit Current Orbit) and new SINDA/FLUINT data generated without recomputing heating rates. RadCAD does not impose any particular constraints for the usage of SINDA/FLUINT submodels. The output file produced by a heating rate run may be placed in a separate submodel. For heating rate data, the filename provided in the Output Filename field will be appended with an ‘l’ for the logic data and an ‘a’ for array data. The may specify the starting array ID in the S/F Starting Array ID field.


10-23

The selection of Output Format, determines whether the heat loads are applied using the compact and versatile LOADQ subroutine or a series of DA11MC (refer to the SINDA/ FLUINT ’s Manual) subroutine calls. The LOADQ subroutine is equivalent to a series of DA11MC subroutine calls for a transient solution and equivalent to applying the orbitalaverage heat load for a steady state solution. The Combine SAP arrays into a single array check box allows the to decide if the solar, Albedo and planetshine heat loads will be stored in a single array for each node, if checked, or in separate arrays for each node. The Output as fluxes check box will output an area array and fluxes for each node, if checked, instead of absolute heat loads. The sources region allows the to choose which source loads are output. Obviously, if the source was not selected on the control tab for calculation and does not already exist in the database, the heat loads cannot be output. The tracker data file provides a text file describing the transformations of any tracker active for the calculations. The Output HR Symbols to SINDA check box will include s in SINDA/FLUINT set to the values (or time-interpolated values) of certain heating rate symbols (see "Internally-Generated, Heating-Rate Symbols" on page 11-9) in the HRL file. Finally, the Advanced Logic button allows the to have more control over when the heating rates are loaded. Please see “Radiation Advanced Logic” on page 10-24. To define a heating rate job in the Case Set Manager, select the Heating Rates radio button, the radiation analysis group and the orbit on the Job tab, as described in "Case Set - Radiation Tab" on page 15-4. 10.1.4.1

Radiation Advanced Logic

This Advanced Logic dialog box (Figure 10-13) allows the a little more control for when heating rates or articulated radks are loaded. The first input parameter is for a time offset. This time offset allows the to manipulate the interpolation time for cases when the SINDA time does not match up directly with the orbit start time (such as converting from local time to GMT). The offset is subtracted from the current SINDA time and the result is used to interpolate on the heating rate arrays, therefore, a positive offset will use heating rates at an earlier orbit time and vice versa. The second set of parameters allows the

10-24


to only load heating rates during a specified start and stop time. Finally, the can input their own special logic before or after the logic that is automatically output by RadCAD.

Figure 10-13

10.1.4.2

Radiation Analysis Data Heatrate Output Tab, Advanced Logic

LOADQ

LOADQ is SINDA subroutine that simplifies the SINDA input file. The LOADQ logic performs one of two functions: • for transient solutions or steady state solutions where DTIMES > 0, it is equivalent to a DA11MC call interpolating the heating rate arrays with the SINDA variable TIMEM • for steady state solutions where DTIMES = 0, it applies the time averaged heating rate Thermal Desktop writes the LOADQ call into the heating rate logic file, *.hrl. Thermal Desktop writes the data in the array file, *.hra, with the node submodels and numbers associated with the heating rates stored in CARRAY type data. The SINDA/FLUINT input file written by Thermal Desktop uses the macro-command INSERT to include these two files into the model. Calling Sequence: CALL LOADQ('SMN',NCA1,N,AN(IC),AA(IC),AS(IC),AT(IC), AF(IC)) where: • 'SMN' - submodel name that contains the array data • NCA1 - the first CARRAY number containing the node submodels and numbers • N - total number of CARRAYs containing the node submodels and numbers


10-25

• AN - singlet array for storing the relative node numbers of the form smn.An or

An; used internally by LOADQ • AA - singlet array of node areas if heating rates are output as flux; set to zero if

heating rates are total absorbed; of the form smn.An or An • AS - singlet array of time-averaged heating rates of the form smn.An or An • AT - singlet array of time of the form smn.An or An • AF - singlet array of heating rate scale factor(s) of the form smn.An or An; single

value if Solar, Albedo and Planetshine heating rates are combined into a single array; three scale factors for Solar, Albedo and Planetshine, respectively, if heating rate arrays are separated. Example: C The following example shows the LOADQ call and examples of the C arrays written by Thermal Desktop. C HEADER VARIABLES 0, C0A2 C LOADQ called within VARIABLES 0 since the heat load is time dependent CALL LOADQ('C0A2',1,3,A2,A3,A1,A5,A4) C HEADER ARRAY DATA, C0A2 C Average heating rates 1= 5.360949e+001,5.473296e+001,5.553302e+001,5.460822e+001 5.374372e+001,5.348048e+001,5.452574e+001,5.557201e+001 5.453088e+001,5.344599e+001,5.344553e+001,5.450615e+001 5.553493e+001,5.450872e+001,5.357235e+001,5.348794e+001 5.444569e+001,5.548737e+001,5.454376e+001,5.354758e+001 5.363795e+001,5.461206e+001,5.553643e+001,5.466795e+001 5.369263e+001,1.366805e+003,1.366226e+003,1.365810e+003 1.908515e+002,1.920125e+002,1.893623e+002,1.983020e+002 7.783840e+002,1.981422e+002,7.787539e+002,1.684397e+003

10-26


C Relative node storage for LOADQ 2 = SPACE, 36 C Dummy array for LOADQ (contains node areas when heating rates C are provided as fluxes) 3 = SPACE, 36 C Scale factors used by LOADQ 4 = 1. C Time Array 5= 0.0,4.730820e+002,9.461640e+002,1.419250e+003 1.764227e+003,1.767633e+003,1.892330e+003,2.365410e+003 2.838490e+003,3.311570e+003,3.784650e+003,3.909347e+003 3.912753e+003,4.257740e+003,4.730820e+003,5.203900e+003 5.676980e+003 C solar albedo planetshine - MAIN.5 Area = 0.160000 Avg = 4.977036 C 19.500551 29.131907 C Errors: 2.4, 2.2, 1.8, 2.3, 0.6, 2.4, 2.4, 2.4, 2.4, 2.4 C 2.3, 2.4 0.6, 2.2, 1.8, 2.2, 2.4 6= 9.073225e+001,8.281288e+001,6.044160e+001,3.084614e+001 1.101144e+002,2.901368e+001,2.914097e+001,2.914097e+001 2.931786e+001,2.905791e+001,2.932999e+001,2.905738e+001 1.101581e+002,3.089951e+001,6.007341e+001,8.247578e+001 9.066256e+001 C solar albedo planetshine - MAIN.6 Area = 0.160000 Avg = 4.977036 C 19.979088 29.776833 C Errors: 2.2, 2.0, 1.7, 2.1, 0.6, 2.2, 2.2, 2.2, 2.2, 2.2 C 2.2, 2.2 0.6, 2.1, 1.7, 2.0, 2.2 7= 9.340993e+001,8.491866e+001,6.160230e+001,3.132634e+001 1.109409e+002,2.984016e+001,2.976217e+001,2.976217e+001 2.985216e+001,2.973822e+001,2.985990e+001,2.981175e+001 1.109125e+002,3.142107e+001,6.121961e+001,8.483404e+001 9.295551e+001 ... C Remaining heating rate arrays with monotonically increasing numbers C HEADER CARRAY DATA, C0A2 1 = MAIN, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20 2 = 21, 22, 23, 24, 25, 26, 27, 28, 29, 31, 32, 33, 34, 35, 36 3 = 10001, 10002, 10003, 10004, 10030

10.1.5

Setting Ray Plot Options

The may graphically plot the calculated rays on the model. This option may be turned on by selecting the Ray Plot tab in the Radiation Analysis Data dialog box (Figure 10-14). Caution must be exercised when using this feature because shooting many rays for all the nodes in the model will simply clutter up the graphics area. This feature should be used in conjunction with the Nodes List (Control tab) to limit the number of rays on the screen (see “Setting Control Parameters” on page 10-2). The may select to plot rays to space, rays from heating sources, and ray reflections between surfaces. The length of ray may be input for rays that go to space and for rays that are incident from heating sources. The ray is colored based on the value of its energy. A red


10-27

Figure 10-14

Radiation Analysis Data Dialog Box Ray Plot Tab

ray has energy equal to 1 and a dark blue ray has energy equal to 0. The rays to space and from the source may be controlled with the parameter LTSCALE. A larger value of LTSCALE will make the ray dotted (or invisible if too large), while a smaller value will make the ray a solid line. The different line types enables the to easily distinguish the rays between surfaces and the rays that go to space. The may clear the rays on the plot by using the Thermal > Radiation Calculations > Clear Ray Plot function. Once the selects the type of rays to plot (space, source, or reflections between surfaces), additional criteria may be selected so that only rays that hit a specific node or rays that only hit an inactive node may be displayed. These options can make it very easy to determine how you can get a radk from one surface to another, or how a surface sees an inactive node. When plotting heating source rays, only the rays that see the source will be plotted. Rays that intersect another surface before seeing the source are not plotted.

10-28


At the bottom of the dialog box, the can select to Output Average Ray, which can be used to output the average ray length of view factor calculations. A field also exist for the to input the file name that the data is written to. 10.1.6

Setting Fast Spin Parameters

If the radiation job is defined from the Case Set Manager (see Section 15.2.1), the can specify the parameters for a fast spinning or translating surface. In order to do the fast spin or translation calculations for radks or heating rates, an assembly must be set up with one or more of its rotations or translations parameterized. That parameter name is then input in the Radiation Analysis Data dialog box Spin tab (Figure 10-15) as the symbol name. The can then set the start and end values as well as the number of increments. Rays will be shot for each increment, and the results are then averaged before output. Be careful not to shoot too many rays.

Figure 10-15

Radiation Analysis Data Dialog Box Spin Tab

The check box for Inclusive Start and Stop Values works well if the is trying to model a translation. Consider the case with Starting Value = 1, and Ending Value = 21, with increments = 5. Without Inclusive Start and Stop Values selected (not checked), the program would run values at 1, 5, 9, 13, and 17. With the option selected (checked), the program will run values at 1, 6, 11, 16, and 21. Leaving the option unchecked for full rotations prevents duplicate calculations at the first and last points which would be the same (e.g. 0 and 360 degrees)


10-29

The also has the option of inputting a list of symbol values to do calculations (select Do Fast Spin Calculations and then select Defined List). This would facilitate an object that spins or translates non uniformly or oscillates with ascending and descending values. 10.1.7

Disabling Specific Trackers

If the radiation job is defined in the Case Set Manager and is either articulating radks or heating rates, the will see the Trackers tab shown Figure 10-16. The Trackers tab allows the to disable trackers for this particular radiation job. Please note that if a tracker has already been disabled globally (by editing the tracker), then there is no need to disable it on this dialog box.

Figure 10-16

10.1.8

Radiation Analysis Data Dialog Box Trackers Tab

Overlapping Surfaces Checks

In Radiation Analysis Data in the Case Set Information, the Overlap tab allows the to set overlapping surfaces checks to be run for each radiation task. If the geometry does not change from case to case and an overlapping surfaces check has been performed by the (Section 8.11), then this option can be unchecked.

10-30


10.2

Calculating and Outputting Form Factors and Radks

When Thermal > Radiation Calculations > Calc Form Factors, Thermal > Radiation Calculations > Calc Radks From VF, or Thermal > Radiation Calculations > Calc Radks Ray Trace is selected, a message box will appear confirming the name of the internal database to be used to store the ray tally information. The database name is constructed using the default analysis group and the currently loaded optical property file. A database named “analysis_group_name.rcf” will be created for form factors, and a database named “analysis_group_name-optical_property_file_name.rck” will be created for radks. The ray tally database is used for restarting calculations and (for radks) is processed after calculations to generate a SINDA/FLUINT input file. The form factor ray tally database is used in subsequent radiosity based radk and heating rate computations. Radks are computed directly using Calc Radks Ray Trace. The data is saved in a RCK database that is then used to generate the SINDA/FLUINT input deck. View factors may also be computed using ray tracing for conversion to radks using Calc Radks From VF. Calc Radks From VF must be preceeded by Calc Form Factors. The progressive radiosity method generates an RCK file from an RCF database. Radks computed using Monte Carlo methods may use any type of optical properties, including specular reflection and specular transmission. Radks computed using progressive radiosity are restricted to diffuse reflection and diffuse transmission. Continuation of calculations is possible using the Monte Carlo method. For Calc View Factors and Calc Radks Ray Trace, if the ray tally database already exists, RadCAD will examine the contents of it to see if it is possible to continue calculations. If the number of SINDA/FLUINT nodes and their names have not changed, then it is possible to add to the data already present. For example, more rays could be shot to increase accuracy. Or geometry could be repositioned and rays shot to average the results over a number of different physical configurations (effectively a motion blur). If a restart is possible, a message box will pop up asking if you would like to append the results of this run to the existing database or continue with a brand new database. The append option will add ray tally data to the existing database (for example, to shoot more rays for better accuracy). If the new database option is selected, the old data will be deleted and a new database created. Make a choice and select OK to begin calculations. If a restart is not possible, a dialog box will appear asking if the old database is to be deleted and a new one created. Selecting OK will begin calculations with a new database. Selecting Cancel will cancel the calculation operation. A new radk ray tally database is created from an existing view factor ray tally database each time radks are computed using Calc Radks From VF. Because the radk matrix is derived directly from the view factor matrix in a single step, “restarting” is not applicable. When Calc Radks From VF is performed for an analysis group, RadCAD searches for an existing analysis_group_name.RCF database. Using this database and the currently loaded optical properties, an analysis_group_name-optical_property_file_name.RCK database is created. This database is then used to produce the SINDA/FLUINT radk input file. The


10-31

progressive radiosity algorithm used to compute radks for view factors is extremely efficient and the time to produce radks will usually be insignificant compared to the time used to compute view factors. The may stop calculations by selecting the <Esc> key. It is always a good idea to issue a save command before calculations are initiated. This will save your work in case an unexpected power failure occurs (say for an overnight run). Results to the ray tally database are saved after each node is calculated, so if a machine goes down for some reason, the remaining nodes may be calculated using the Set Nodes to Calculate dialog box. Radks and form factors are output automatically after calculations if the Generate SINDA/FLUINT input file after calculations option is checked on the set run parameters dialog box, or whenever Thermal > Radiation Calculations > Output Area*FijFile or Thermal > Radiation Calculations > Output SINDA/FLUINT Radks are selected.

10.3

Calculating and Outputting Environmental Heating Rates

When Thermal > Radiation Calculations > Calc Heating Rates Direct/VF, or Thermal > Radiation Calculations > Calc Heating Rates Ray Trace is selected, a message box will appear confirming the name of the internal database to be used to store the ray tally information. The database name is constructed using the default analysis group, the currently loaded optical property file, and the current orbit (see “External Heating Environments and Orbits” on page 6-1). Thermal > Radiation Calculations > Calc Heating Rates Direct/ VF must be preceeded by Thermal > Radiation Calculations > Calc Form Factors. A database named “analysis_group_name-property_file_name-orbit_name.rch” will be created for orbital heating rates. The ray tally database is used for restarting calculations and is processed after calculations to generate a SINDA/FLUINT input file. Continuation of calculations is possible for both heating rate methods. If a database exists, RadCAD will examine the contents to see if it is possible to continue calculations. If the number of SINDA/FLUINT nodes and their names have not changed, and the number of orbit positions and orbit times are the same, then it is possible to add to the data already present. For example, more rays could be shot to increase accuracy. Or geometry could be repositioned and rays shot to average the results over a number of different physical configurations (effectively a motion blur). If a restart is possible, a message box will pop up asking if you would like to append the results of this run to the existing database or continue with a brand new database. The append option will add ray tally data to the existing database (for example, to shoot more rays for better accuracy). If the new database option is selected, the old data will be deleted and a new database created. Make a choice and select OK to begin calculations. If a restart is not possible, a dialog box will appear asking if the old database is to be deleted and create a new one created. Selecting OK will begin calculations with a new database. Selecting Cancel will cancel the operation. 10-32


Calculating heating rates using the ray tracing method allows specular properties to be modeled, and allows more flexible choices in nodalization (since uniform illumination is not required). RadCAD automatically compresses and decompresses the database files needed for the full monte carlo based ray tracing method, resulting in greatly reduced disk space usage by the ray tally files. Calculations are performed by first generating a random location on the nodal area. Then a ray is fired towards the source randomly distributed within the solid angle of the source. If the ray can reach the source unimpeded (including specular transmissivity), the ray is “reversed”. Energy is deposited in the node according to its optical properties, and then the ray is propagated to compute the reflected energy. The flux from a source is based on the values provided in the Heating Environment’s Planet Data (Section 6.1.1.3: Planetary Data on page 6-6). For the Albedo, the flux is a function of the Albedo, the Solar flux and the location on the planet where the ray intersects (based on the Solar vector and planet normal). For the Direct/VF method, a Monte Carlo technique is used to compute the direct incident energy only. The form factor ray tally database is then used to compute the reflected energy using CRTech’s unique progressive radiosity algorithm. (If more rays have been shot to increase the accuracy of the form factor database, new heating rates can be computed by performing a restart and shooting zero heating rate rays). The sum of the direct absorbed and reflected components is computed to yield a total absorbed heating rate for each node. The combination of the unique oct-cell algorithm, automatic error tolerance, and the innovative progressive radiosity algorithm make RadCAD’s Direct/VF method extremely efficient. Orbital heating rates are output automatically after calculations if the Generate SINDA/ FLUINT input file after calculations option is checked on the set run parameters dialog box, or whenever Thermal > Radiation Calculations > Output SINDA/FLUINT Heatrates is selected. The current working directory will be searched for a heating rate database that corresponds to the default analysis group, the currently loaded optical property file, and the current orbit. The orbital heating rates are output to SINDA/FLUINT in two forms: average and position dependant. When a steady-state solution is called in SINDA/FLUINT, the average heating rate (time-weighted average of all positions in the heating rate case) is used in the solution. When a transient solution is called (or a steady-state solution with control constant DTIMES greater than 0), the position-dependant values are interpolated using a cyclical, linear interpolation based on the solution time.


10-33

10-34


11

Parameterization

Thermal analysis is typically performed using a point design approach, where a single model is analyzed one analysis case at a time. Changes to the system design are analyzed by updating the thermal radiation and conduction models manually, which can become tedious and error-prone. The parametric features of Thermal Desktop allow the thermal model to be characterized by a set of design variables that are easily modified to reflect system-level design changes. Geometric features, optical and material properties, initial temperatures, -defined thermal masses and conductances, and orbital elements may be specified using -defined variables and expressions. Furthermore, these variables may be automatically modified by SINDA/FLUINT’s optimization capabilities in order to satisfy -defined design goals, or for correlating thermal models to test data (available with SINDA/FLUINT 4.3 and higher).

11.1

Symbols

Symbols are -defined variables in Thermal Desktop. A Thermal Desktop model is parameterized by using symbols alone or in mathematical expressions anywhere that the expression editor (Section 2.10.7) is available. Symbols may be defined as values or mathematical expressions based on other symbols. Unless specifically used to create SINDA s (Section 11.2), symbols are used for Thermal Desktop calculations. After the model has been parameterized, the may make sweeping changes to the model by changing the values of the symbols using the Symbol Manager. The use of symbolic expressions can greatly reduce the effort in generating new thermal models as the result of system design changes. 11.1.1

Symbol Manager

The Symbol Manager, shown in Figure 11-1, is accessed through the Thermal > Symbol Manager menu. The Symbol Manager allows symbols to be defined and organized. The Symbol Manager displays the list of symbols, which can be organized into Symbol Groups. Each group is shown on an individual tabbed field, and the symbols in that group are displayed in alphabetical order, along with the expression and its evaluated result. Most of the following functions can be performed with several symbols selected. See Section 11.1.1.3 for information about the behavior when multiple symbols are selected.

Parameterization

11-1

Figure 11-1 Thermal Desktop Symbol Manager New Symbol Name. This field is used to type in the name of a new symbol. To open the symbol expression editor (Section 11.1.1.1) and define the new symbol click the Add button, or hit <ENTER>.

Edit. Select the Edit button to open the symbol expression editor for the selected symbol. A symbol can also be edited by double clicking in the symbol’s row. Copy. The Copy button allows copying the definition of a selected symbol to a new name. In the Copy form, the has the option to Edit After OK Selected. When checked the symbol expression editor will be opened after the OK is selected. Rename. The Rename button allows changing the name of the selected symbol. The name change will be propagated through the entire Thermal Desktop model with the exception of text fields; expressions and value fields referencing the symbol will be updated. In the Rename form, the has the option to Edit After OK Selected. When checked the symbol expression editor will be opened after the OK is selected. See Section 11.1.1.3 for renaming multiple symbols. Important: Renaming symbols will change the names within value fields and drop-downs, but symbol names in text forms and logic will not be affected. Delete. The Delete button will delete the selected symbol or symbols. There will be no confirmation. While the symbol is deleted from the list, it will remain in expressions and the expression will be in error, returning a value of 1. The may prefer to use Purge, below.

11-2

Parameterization

Purge. The purge button will evaluate whether or not symbols are used in expressions. If any symbols are not used, the Groups, Symbols not used. Select to Purge. dialog will be opened listing all unused symbols. At the top of the dialog, the can type in a symbol name to search for that symbol name if the list is long. The must select any of the symbols that are to be purged and the select the Delete button. Find. The Find button opens the Find Symbol dialog. The can select to search for Name, Expression, or Value. The Name field is a drop down with all of the symbol names. By typing a letter, the list will jump to the first symbol name starting with that letter. When OK is selected the Symbol Manager will reopen with the symbol group containing the desired symbol and the symbol selected. The Expression field lists the symbol expressions in alphabetical order. In addition to the expressions, each line shows the number of times the expression is used and the symbol names that use that expression. Selecting OK will open a text window listing all symbols matching the search criteria. The information in the text window is also saved to a file named SymbolFindResults.txt. The Value field lists the value of the evaluated expressions, the number of symbols that have the value, and the symbols that have the value. The values are listed in alphanumeric order so typing in a 2 will jump to the first value that starts with a 2 and all other values that start with 2 will follow. Selecting OK will open a text window listing all symbols matching the search criteria. The information in the text window is also saved to a file named SymbolFindResults.txt. Import. The Import button performs an import as described in Section 2.4.1. Export. The Export button performs an export as described in Section 2.4.1. Done. Selecting Done will close the Symbol Manager and will apply symbol value changes to the model. 11.1.1.1

Symbol Expression Editor

When a new variable is defined, or an existing variable is edited, the symbolic Expression Editor dialog box shown in Figure 11-2 is displayed

Note: The symbolic Expression Editor dialog box and its use are similar to the input-field Expression Editor described in Section 2.10.7, but there are some slight differences. The first field on the from is named for the symbol being edited. The expression for the variable is input in this field. The expressions follow the same rules and have the same capabilities as described in Section 2.10.7. To add a symbol to the expression without typing the symbol name, right-click in the expression field, hover the cursor over the symbol group name (Section 11.1.1.2) and then choose the symbol from the list of symbol names. Description. A descriptive comment may be entered in the Description field.

Parameterization

11-3

Figure 11-2 Expression Editor for Defining the Value of a Symbol

Symbol Type in Thermal Desktop. Symbols can be one of three types: double, a numerical value; array, a series of numerical values with one value per line; and string, a file name. The type is selected by choosing from the Symbol Type in Thermal Desktop drop-down list. For more information regarding array symbol types see Section 11.1.3. String symbols are currently only used for the Boundary Condition Mapper described in Section 4.13. Group. The Group drop-down list provides a list of existing groups that can be selected for the symbol being edited. Groups are used for organizing symbols in the Symbol Manager. Symbol groups and their creation is described in Section 11.1.1.2. Control Symbol Output to SINDA . The Control Symbol Output to SINDA button allows the to define how the symbol is exported to SINDA/FLUINT as a . When selected, Output SINDA dialog opens. The options on the dialog are: Always Output Thermal Desktop Symbol as SINDA . When checked this

option will create a in SINDA. The value of the will be determined by the selection of the next three options, the symbols overrides in the Case Set Symbols tab (Section 15.2.8), and the s definitions on the Case Set SINDA tab (Section 15.2.4), respectively (this means that changes made in the s definitions override all other changes and overrides). Output Resultant Value. If selected the will be defined as the calculated

value of the expression and will be assumed to be a fixed value. Output Symbol Expression. If selected the will be defined as the expression

provided in the symbol expression editor. Any other symbols used in the expression

11-4

Parameterization

will also be used to create s regardless of the selections made for their respective Output SINDA dialogs. Output Integer. When checked, the generated by this symbol will be gener-

ated as an integer and will be treated as such in the SINDA/FLUINT logic. See Section 11.2 for more information regarding the use of symbols and s. Check consistent usage of units when in expressions. When checked, Thermal Desktop will compare the units of the fields in which symbols are used to be sure the symbol is used consistently. This prevents the same symbol from being used for length and angle fields. If the symbol is a multiplier or other unitless value that may be used in many different situations, the may wish to uncheck the box.

Figure 11-3 Symbol Output SINDA Dialog

11.1.1.2

Symbol Groups

Symbol groups are an organizational tool for symbols. Symbols are placed in groups and the groups are shown as tabs on the Symbol Manager dialog.The default group name is general. The orbital group may also already exist; see Section 11.1.4 for more information on the orbital group. Symbol groups are created by right-clicking on any existing group tabs and selecting Add New Group. A new tab will appear with the name group1 (or group2, group3, etc). A symbol group can be made active and the symbols in the group displayed by selecting the group tab. The active symbol group can be renamed by right-clicking any of the group tabs and selecting Rename Group. The active symbol group can be deleted by right-clicking on any group tab and selecting Delete Group. Only empty symbol groups can be deleted. Any empty symbol groups will be deleted when the Symbol Manager is closed. Symbols are added to groups by editing the symbols and selecting the desired symbol group in the Group field (Section 11.1.1.1).

Parameterization

11-5

11.1.1.3

Selecting Multiple Symbols

When multiple symbols are selected, some of the functions on the right side of the Symbol Manager provide special operations. Multiple symbols can be selected in the Symbol Manager by <Shift>-clicking a series of symbols or -clicking individual symbols. Edit. The only change to the Symbol Edit - Multi Edit Mode form is that the expression field is not named. Multi Edit Mode is most useful to change the group name of multiple symbols at the same time. Multi Edit Mode works just like regular Thermal Desktop multi edit in that only fields that are changed will be updated for the edited symbols. Copy. The Copy Multiple Symbols form will copy the selected symbols and create symbols with modified names using the following options: Append String to each existing symbol. The string that is typed into the String

field is appended to the names of the selected symbols to form the names of the new symbols. Replace String.The string that is typed into the Existing String field will be replaced with the string in the Replacement String field to form the names of the new symbols. If the Existing String does not exist in a symbol, that symbol will not be copied.

Rename. The Multiple Rename form will change the name of the selected symbols using the following options: Append String to each existing symbol. The string that is typed into the String

field is appended to the names of the selected symbols to form the names of the new symbols. Replace String.The string that is typed into the Existing String field will be replaced with the string in the Replacement String field to form the names of the new symbols. If the Existing String does not exist in a symbol, that symbol will not be copied.

Important: Renaming symbols will change the names within value fields and drop-downs, but symbol names in text forms and logic will not be affected. 11.1.2

Built-In Symbols

The Thermal Desktop Symbol Manager contains a number of built-in symbol definitions that may be used in an expression. These built-in symbol definitions are not listed in the Symbol Manager's display area, but are always available. These symbols are shown in Table 11-1.

11-6

Parameterization

Table 11-1 Predefined Symbols

Name

Value

Description

pi

3.141592653589739

sbcon

0.1712E-08 BTU/h-ft2-R4

Stefan Boltzmann Constant

sbconsi

5.6693E-08 W/m2-K4

Stefan Boltzmann in SI units

dtor

0.017453292519943

Degrees to Radians

rtod

57.2957795131

Radians to Degrees

cmtoin

0.3937007874

Cm. to inches

mtoft

3.280839895

Meters to feet

intocm

2.54

Inches to cm.

fttom

0.3048

Ft. to meters

inhgtoat

3.342E-02

In. Hg to Atmospheres

psitoat

6.804E-02

PSI to Atmospheres

jtobtu

9.480E-04

Joule to BTU

btuhrwat

2.930722E-01

BTU/Hr to Watt

lbf3kgm3

16.01846

lb/ft3 to kg/m3

lbf2n

4.44822

lbf to newton

lb2kg

0.4535924

lb. to kg.

kgs2lbh

7.936641E+03

kg/sec to lb/hr

psi2pa

6.894757E+03

psi to pascals

grav

32.1725*3600.0**2 ft/hr2

gravity in FLUINT ENG units

gravsi

9.80621 m/sec2

gravity in SI units

gasr

1545.0 ft-lbf/lbmol-R

gas constant

gasrsi

8314.23 J/kgmol-K

gas constant in SI units

ft3tom3

2.831685E-02

cubic feet to cubic meters

wmk2bhfr

1.73073

Watt/m-K to BTU/hr-ft-R

#length_eng

varies

length of current pipe in feet; only available in pipes

#length_si

varies

length of current pipe in meters; only available in pipes

Parameterization

---

11-7

11.1.3

Array-Based Symbols

Symbols can be defined as a singlet array by selecting array - enter 1 value per line in the Symbol Type in Thermal Desktop drop-down menu at the bottom of the Expression Editor dialog box (Figure 11-2). The “interp” function (Section 2.10.7.2) can be used to interpolate between two array symbols. An array can be created by selecting the array check box, and data. The data should be entered with one data point on each line. For example, create a symbol array called time_array and input the data to be 0, 100, 200, 300, 400, 500 as shown in Figure 11-4.

Figure 11-4 Defining a Symbol as an Array

Create a second array called data_array where the data is 1, 2, 10, 20, 100, 200. Recall that the input is one data point per line. Finally, the “interp” function can be used to interpolate between the two arrays. To interpolate between them, enter interp(time_array, data_array, 5.). The value 5 represents the time to interpolate on. The value itself can be another symbol, such as hrTime, or also an expression. Please note that array-based symbols may not be sent to SINDA as s and that the ‘interp’ function is not defined in SINDA. For array interpolation within SINDA, use Array Interpolation within the Logic Options Manager (Section 12.1) or entity-based time or temperature varying properties. Want "Hands-On" Information? For examples of working with arraybased symbols refer to the tutorial exercise “Orbital Maneuvers” on page 21-87.

11-8

Parameterization

11.1.4

Internally-Generated, Heating-Rate Symbols

When a heating environment has been defined in a model, Thermal Desktop will automatically create several symbols that define part of the orbit, as well as the status of the currently calculating or displayed position. This allows complicated orbital maneuvers, or any other parameterization, to be programmed as a function of orbital position. These symbols are listed in Table 11-2. s can reference these symbols to program the behavior of trackers, assemblies, and vehicle rotations of an orbit. These symbols, however, are not -definable (do not modify them directly) and are not updated in SINDA. While the symbols are not updated in SINDA, the time-dependent values can be accessed from SINDA logic by selecting the Output HR Symbols to SINDA check box on the Heatrate Output tab of the heatrate task in the Case Set Manager (see Section 10.1.4) Table 11-2 Environmental Heating Symbols

Name

Description

hrBetaAngle

Beta angle of the current orbit

hrEccen

Eccentricity of the current orbit

hrIllum

Illumination state of the current position in orbit. 0-In Sun, 1-At Shadow Entry, 2At Shadow Exit, 3-In Shade

hrMeanAnom

Mean Anomaly of the current position

hrPeriod

Period of the current orbit

hrPlanetX, Y, and Z

Vector to the planet center in the model WCS

hrPos

Current orbital position

hrShadowEntry

Shadow Entry Angle of the current orbit

hrShadowExit

Shadow Exit Angle of the current orbit

hrSpeed

Speed of spacecraft in current position

hrSunX, Y, and Z

Vector to the Sun in the model WCS

hrTime

Time at the current position

hrTimeShadowEntry

Time of shadow entry for the current orbit

hrTimeShadowExit

Time of shadow exit for the current orbit

hrTrueAnom

True Anomaly of the current position

hrVelocityX, Y, and Z

Velocity vector in the model WCS

Parameterization

11-9

Want "Hands-On" Information? For examples of working with symbols and the Symbol Manager refer to the tutorial exercises “Parameterizing for a Common Input” on page 20-187, “Beer Can Example” on page 20-89, “Dynamic SINDA Example” on page 20-201, and “FEM Walled Pipe” on page 22-99.

11.2

Using Symbols and s

Symbols are used within Thermal Desktop only (see Section 11.1.1)—symbols are not available in SINDA/FLUINT. Therefore, any symbol will be a set value during the SINDA/ FLUINT run. s can be accessed by the during the SINDA/FLUINT run. When a defines a symbol it is assigned a global value when defined. This global value may be a number or an expression. The global values can be viewed by choosing Thermal > Symbol Manager. If the wants to change this value for a specific case, this is best done in the Case Set Manager on the Symbol tab. The can override the global value for the specific case. In most of the Thermal Desktop edit dialog boxes input values can be set to be expressions and symbols by double clicking in the associated field. The can then enter an expression that is numeric or can enter an expression that uses symbols (see Section 2.10.7). If the then selects the Output expression to SINDA check box (displays a check mark in the box), Thermal Desktop will keep the symbol in the field definition and automatically add the symbol to the SINDA/FLUINT s. If the has chosen to override the global value in the Case Set Manager, the override value will be output to SINDA/FLUINT. If the Output Expression to SINDA option is greyed out, then the field is required for Thermal Desktop or RadCAD calculations (geometry dimensions, optical properties, etc.) and should be updated only within Thermal Desktop as a symbol. A second method of creating SINDA s from symbols is on the Symbol Manager Expression Editor form. The Control Symbol Output to SINDA allows the to determine if the symbol is always output as a , what is used to define the (value or expression), and how the will be used in SINDA logic (single precision, double precision, or integer). The final option to output a symbol as a in SINDA/FLUINT is from within the Case Set Manager on the SINDA tab by choosing in the Global S/F Inputs (see “Global S/F Inputs” on page 15-15). This displays the SINDA Variable Definition dialog box. The can double click a Global Symbol on the left which brings up the Variable dialog box. The can then choose Use Global Value to assign the global symbol value to the ; choose Value to input a new value or SINDA/ FLUINT expression; or choose Use Symbol Expression to use the actual text of the symbol expression to define the . The can add a comment that will appear in the SINDA/

11-10

Parameterization

FLUINT Header Data block. The Output As option allows the to determine how the new will be treated in SINDA/FLUINT logic. The Disable button comments out the . Important: Warning to the : If you have selected an output expression in an Edit dialog box and have also used the Variable dialog box, the value of the will be defined in the SINDA/FLUINT Header Data block as the value in the Variable dialog box, not the value from the Symbols definition.

Parameterization

11-11

11-12

Parameterization

12 Logic Manager The Logic Manager contains utilities to aid the in customizing the solution of a model. The objects are intended to add functionality to other parts of Thermal Desktop without complicating the interface for each object with redundant options. Figure 121 shows the Logic Manager. Objects are created from the pull-down list at the upper right and added to the list of Objects in the model. The order of the objects in the list determines the execution order of each individual object within any given logic block. Some objects, such as the PID controller, may generate code in a number of different places. To provide pre- or post-operations on the logic, multiple objects can be placed in sequence in the list. For more information regarding the location options for Logic Options (“Code performed in”) see Section 2.10.11. The Logic Manger form has the following fields and options.

Figure 12-1

Logic Manager Edit Form

All Logic Objects. The main field of the Logic Manager is a tree containing the Logic Objects that have been created. The Logic Manager has a simple, tree-based format allowing single-level grouping of logic objects (like the one named Logic in Figure 12-1). The Logic Objects can be rearranged by clicking and dragging to obtain the desired order and grouping. Logic objects which have been disabled (Section 2.10.8) will have a preceding icon in the tree with black and red stripes. Select Object type to Create. Choose the Logic Object type to create from the drop-down list. After selecting from the drop-down list, select the Create button to open the dialog for that Object type. Each Object type and the associated dialog are described in the following sections:

Logic Manager

12-1

• Array Interpolation - Section 12.1 • Bivariate Array Interpolation: Section 12.2 • PID Controller: Section 12.3 • Text Input HEADER/SUBROUTINE: Section 12.4 • Equation of Motion - Linear: Section 12.5.1 • Equation of Motion - Angular: Section 12.5.2 • Equation of Motion - Shaft: Section 12.5.3 • Data Logger Compare: Section 12.6 • Array: Section 12.7 • COMPLQ/WAVLIM: Section 12.8 Edit. Opens the appropriate form for editing the selected Logic Object. See the Logic Object type in the following sections for the description of the individual forms. If a group is selected for editing, the Logic Objects are edited sequentially. Group. Creates a group for organization of the Logic Objects tree. Logic Objects can be dragged into the group within the Logic Object tree. Delete. Deletes the selected Logic Object. Import. Imports Logic Objects directly from another DWG file or from an exported Logic Object See Section 2.10.12. Export. Exports the selected Logic Objects so they can be imported to another model. See Section 2.10.12. For Logic Objects with forms, selections allow the to choose a Thermal Desktop Symbol (Section 11.1). The reason the term is used is that Logic Objects are only active within SINDA/FLUINT. Thermal Desktop, when generating the SINDA/ FLUINT input file, will create a using the symbol name for any symbol selected in a Logic Object form. While the logic manager is a convenient method of adding logic to the model, care must be taken when dealing with units. FORTRAN logic must be in the system of units used for the solution: for thermal-only models, the system of units for the solution is the system of units defined in the Units section of the Thermal > Preferences: Units form; for fluid models, the system of units for the solution is the system of units chosen in the Output Units For FLUINT Models section of the Thermal > Preferences: Units form. For the other Logic Object forms, units can be chosen for conversions, however, these conversions apply to model units, not units assigned in the expression editor.

12-2

Logic Manager

12.1

Array Interpolation

The Array Interpolation object allows the to input x,y data with unit types (Pressure, Temperature, Heat Rate, etc.) defined by the . The input values should be made in the current -defined units. The array will be output to SINDA/FLUINT in the same units in which the model is running, performing any conversions necessary to do this. The Array Interpolation edit form is shown in Figure 12-2. The independent variable can be set to use the current model time (TIMEN), a selected by the from the Thermal Desktop Symbols, or any text string. The text string is most likely to be some model state variable like a flow rate (FLOW.FR23) or temperature (MAIN.T200). This value should be in the model units. The output value is a selected chosen from the Thermal Desktop Symbols. If the is used as an expression which is output to SINDA/FLUINT, e.g. a boundary node temperature, SINDA/FLUINT will automatically update the boundary node after the interpolation. The can control the timing and placement of the logic block where the interpolation occurs by selecting the logic block from the Interpolation performed in drop-down list. The submodel can be selected if the one of the variable/flogic blocks has been selected, but the GLOBAL option ensures the code will be run no matter which submodels are included in the solution. The type of interpolation to perform is selected from one of the radio button choices in the Array Data block. The specifics actions of the interpolation routines can be found in the SINDA/ FLUINT manual. The multiplier for the output variable is used to scale the result and is applied to the value interpolated from the array data and is included in the returned result. The option Limit time step at input time points forces time steps at the independent array values to ensure short-duration ev3ents are not skipped. Important: For cyclical interpolations, a value of 0 (zero) in the Period field will set the cyclical period to the final value of the independent array. This also means that an expression in the Period field that is initially equivalent to zero cannot be written to SINDA; instead, the last value in the independent array will be used. An example of the usage follows. The current units in Thermal Desktop are time in minutes, and pressure in bar. If the selects Time as the independent variable and pressure as the dependent variable. He would enter data in the form in minutes and bar. If then SINDA/FLUINT model runs in ENG units, the input data would be converted to hours and psi. The input variable TIMEN would be in hours, and the output variable from the interpolation would be in psi. The X Id and Y Id fields allow the to specify the array number for use in SINDA.

Logic Manager

12-3

Figure 12-2

12.2

Array Interpolation Object Edit Dialog

Bivariate Array Interpolation

The Bivariate Interpolation Object provides the ability to input data tables that is a function of x and y, i.e. Z = F(x,y). The edit form is shown in Figure 12-3. The interpolation uses linear interpolation into the table. The has the choice of selecting Time (which will use the TIMEN variable within SINDA/FLUINT), a , or a text string for the independent variables of X and Y. The text string is most likely to be some model state variable like a flow rate (FLOW.FR23) or temperature (MAIN.T200). This value should be in the model units. The output value is a selected . If the is used as an expression which is output to SINDA/FLUINT, e.g. a boundary node temperature, SINDA/ FLUINT will automatically update the boundary node after the interpolation. The can control the timing and placement of the logic block where the interpolation occurs by selecting the logic block from the Interpolation performed in drop-down list. The submodel must be selected if the one of the variable/flogic blocks has been selected.specifics actions of the interpolation routines can be found in the SINDA/FLUINT manual. The multiplier for the output variable is used to scale the result. It is applied to the value interpolated from the array data and is included in the returned result.

12-4

Logic Manager

An example of the usage follows. The current units in Thermal Desktop are time in minutes, and pressure in bar and temperature in Kelvin. If the selects Time and Temperature as the independent variables and pressure as the dependent variable. He would enter data in the form in minutes, Kelvin and bar. If then SINDA/FLUINT model runs in ENG units with Rankine for temperature, the input data would be converted to hours and psi and Rankine. The input variable TIMEN would be in hours, a temperature in Rankine, and the output variable from the interpolation would be in psi.

Figure 12-3

12.3

Bivariate Interpolation Object Edit Dialog

PID Controller

A PID (Proportional-Integral-Differential) Controller object simulates a controller that can be used within the simulation. The edit form is shown in Figure 12-4. The object generates all of the code within the SINDA/FLUINT logic to initialize and execute the controller in both steady state and transient runs. Because the code generated runs within the SINDA/FLUINT logic, all inputs to the routines called must be in the units SINDA/

Logic Manager

12-5

FLUINT is using. The control constants (GP, GI, and GD) are input in the current units within Thermal Desktop, and converted to SINDA/FLUINT units for the logic. The Setpoint (SP) can be set with a constant, or text block. It should be in SINDA/FLUINT units. No conversions are performed on either the or the text string. The text string can be used to specify model variables like MAIN.T100. The units type of the setpoint and control variable should be selected so that the controller constants units can be determined.Changing units after inputting expressions or data for the control constants can affect the value as the base units and conversions may have changed. Edit the expressions to insure correct units are still selected. The Process Variable Input can also be either a or a text string. A text string can be a model state variable like FLOW.TL55. The process variable units are the same as the setpoint. The PID Controller Gain Constants should be set after the units have been selected for the setpoint and control variable. These inputs can be expressions or constants. The values are placed in a call to the PIDSET routine which is performed once at the beginning of the run. If expressions are output to SINDA/FLUINT, they will also be set every iteration or timestep so that time-varying constants can be used, or Solver runs made to optimize settings. The PID controller can be set to run in steady state calculations by leaving the Run in Steady State box checked. The PID controller will run in transient calculations if the Run in Transient box is checked. Note: To turn off the PID controller, use the Enable/Disable button at the top of the form (Section 2.10.8). Do not uncheck both Run in Steady State and Run in Transient to turn off the controller. Checking the Discrete Interval Controller box allows the to control how often the controller will be called during transients. The Steady State Timestep is used at each iteration to control the PID changes. Because the steady state values may change drastically from timestep to timestep, unpredictable results may occur. Smaller time steps will generally be better. A Discrete Interval Timestep will be used to call the PID controller at that interval. This is done by setting the Submodel to be named after the PID, and placing the call to the controller logic within the Output Calls block for that submodel. The simulation time steps will then be controlled to be not more than that interval. Without requesting a discrete interval controller, the controller logic is called on every timestep. Checking the Prevent Integral Windup box signals the controller logic to prevent the integral term from building up past the level where the control variable hits the limits. For example, if the CV is limited to the range zero to one, Gi is negative, Gp and Gd are zero, and the process variable is continuously greater than the setpoint, the error term will drive the integral term to be much greater than 1.0. If the integral term is allowed to grow beyond one, the process variable would need to be below the setpoint for a long time before the CV term would drop below one. If the windup is prevented, as soon as the process variable drops below the setpoint, the CV would also begin to drop below one. The Control Variable limits can be set even is windup is not prevented.

12-6

Logic Manager

The PID controller will attempt to initialize the integral term such that the first call to the PID controller will yield a CV output that matches the initial value of CV. It does this via a call to PIDINIT placed in the logic. The example in Section 12.3.1 demonstrates the usage of the PID object.

Figure 12-4

12.3.1

PID Controller Object Edit Dialog

PID Controller Example

Double click on the file tankControl.dwg located in the Samples/TankControl folder. In this example, a simple flow loop has a PID controller used to adjust the flow through a control valve (FLOW.FR3) which regulates the flow through a heater (FLOW.TL4) to control the system temperature (FLOW.TL5). Tank 5 is the large system volume of water that is being controlled. The total flow around the loop is set by the Setflow device in path 1. The boundary node provides a heat sink via some natural convection ties. The boundary node temperature is driven by a Tamb. Tamb is set via the Logic Manager using an Array Interpolation Object to drive the Tamb . Figure 12-1 shows the list of objects for this example. The order of the objects in the lists determine the order that the code will execute within any given logic block, within a given routine. Objects in different submodels may not run in the same order as the order in which submodels is called will be important as well. This example converges better in the steady state case if the steady state timestep given is longer. Too small a timestep results in the small changes to the valve stem position, and the model can be considered converged before it has iterated enough times to reach a

Logic Manager

12-7

good steady state. TPID, the PID timestep, is set from s TPIDSS and TPIDTRAN for steady state and transient values that are different. Several of the s have been added to the model via the Case Set Manager->Sinda-> page. The Solver Run to Find PID Controller Constants is the first case in the Case Set Manager. This case does a steady state to get initial conditions. Then runs DSCANH and SOLVER to find the best PID control constants. The Single Run to PID case does just a single transient case to the results. The can set the design values found from the first case into the 3 s in the Case Set Manager->Sinda->s page. In order to help the Solver find answers over a very wide range of timescales, the exponents for the actual PID constants are used as the Design variables, GPTOW, GITOW and GDTOW.

12.4

Text Input HEADER/SUBROUTINE

The Text Input HEADER/SUBROUTINE object in the Logic Manager allows the to input code that executes during the SINDA/FLUINT solution or adds data to the model. The s selects the submodel with which the logic is associated and selects the location for the logic. The is responsible for ensuring the text input agrees with the units used for the solution (Section 2.7.1). All code in these objects that are placed in logic blocks will be translated by the SINDA/ FLUINT preprocessor. All entries in the object must be valid FORTRAN code after the preprocessing phase. Placing an F in the first column with turn off preprocessing for that line. Comments can start in column one with a “C”, or after a “$” in any other column. Keep in mind that statements must start in or after the seventh column, the sixth column is reserved for line continuation, and the statement number can be placed in columns 2-5. The example in Section 12.3.1 demonstrates the usage of the Text Input object. The edit form, shown in, Figure 12-5, has the following fields and options. Enabled/Disabled. See Section 2.10.8 Comment. The Comment field will be copied into the SINDA/FLUINT file containing the text as comments. The first line of the Comment field will be included in the Logic Manager tree. Submodel. The Submodel drop-down allows the s to associate the logic with a submodel. If the submodel for the logic is not built (Section 15.2.4) then the logic will not be included in the solution. Asg the submodel as GLOBAL will include the logic independent of submodels included in the solution. Code placed in. The Code placed in drop-down allows the to select the location of the text. For guidelines on logic location see Section 2.10.11 and the SINDA/FLUINT ’s Manual. the drop-down list changes depending on the submodel selected in the Submodel drop-down list:

12-8

Logic Manager

Figure 12-5

Fortran Code Object Edit Dialog

• GLOBAL submodel selected allows creation of a SUBROUTINE DATA block in addition to standard logic blocks • A thermal submodel selected allows creation of NODE DATA, CONDUCTOR DATA, ARRAY DATA, or CARRAY DATA blocks in addition to standard logic blocks • A fluid submodel selected allows creation of FLUID DATA, ARRAY DATA, or CARRAY DATA blocks in addition to standard logic blocks Refer to the SINDA/FLUINT manual for information regarding the formats for the DATA blocks. Declarations. The Declarations field is for code that must be placed at the beginning of the logic location for declarations. Double-clicking in the text field will open the advanced text editor (Section 2.10.14). Code. The Code field is for the main logic that is being generated. Double-clicking in the text field will open the advanced text editor (Section 2.10.14).

12.5

Equations of Motion

The Equations of Motion allow the appropriate differential equations of motion to be co-solved with the thermal/fluid solution.

Logic Manager

12-9

12.5.1

Linear

The Equation of Motion - Linear option provides set-up for solving the differential equation related to linear motion. The inputs are: • A - Mass • B - Friction or damping • C - Spring constant • D - Applied force The can optionally specify initial conditions of position, velocity and acceleration. The outputs which are placed in the specified s are position, velocity, and acceleration. 12.5.2

Angular

The Equation of Motion - Angular option provides set-up for solving the differential equation related to angular motion. The inputs are: • A - Moment of inertia • B - Friction or damping • C - Spring constant • D - Applied torque The can optionally specify initial conditions of angle, angular velocity and angular acceleration. The outputs, which are placed in the specified s, are angle, angular velocity, and angular acceleration. 12.5.3

Shaft

The Equation of Motion - Shaft option provides set-up for solving the differential equation related to shaft motion. The inputs are: • B - Friction or damping • C - Spring constant • D - Applied torque The can optionally specify initial conditions of rotational speed and rotational acceleration. The outputs, which are placed in the specified s, are rotational speed and rotational acceleration.

12-10

Logic Manager

12.6

Data Logger Compare

The Data Logger Compare (DLC) logic object can be used to compare the model solution to a dataset, perhaps from a test or another solution. When used in conjunction with the SINDA/FLUINT Solver, the DLC can provide the comparison for automatically correlating the thermal model to test data or another model. To create a DLC object in the Logic Objects Manager, select Data Logger Compare from the Select Object To Create drop-down list and select Create. The Data Logger Compare dialog appears (Figure 12-6). On the dialog are: • Enable/Disable button

Section 2.10.8

• Comment field

Section 2.10.4

• Filename for input

Section 12.6.1

• Filename for output

Section 12.6.2

• Variable Prefix String

Section 12.6.3

• Type of comparison to perform Section 12.6.4 • P-Program button

Figure 12-6

Section 2.10.9

Data Logger Compare Dialog

A model can contain multiple DLC’s. Each DLC is numbered in the order that it was created. That number is indicated in the title of the edit dialog (e.g. Data Logger Compare 1, Data Logger Compare 2, etc.). The number of the DLC is used in the s that are created by the DLC. Data Logger Compares can be paired with Measures (Section 13) to make comparisons with temperatures at specific locations instead of nodes that may not be located where the test measurements or comparison measurements were made.

Logic Manager

12-11

12.6.1

Data Logger Compare Input File

The input file (examples in Table 12-1 and Table 12-2) uses the format described below. There are three ways to enter a filename: 1. the filename can be typed into the Filename for input field 2. the drop-down list can be used to choose files with a *.dlc extension in the working directory or to browse for a filename with any extension or in any direction 3. the ‘P’ button can be used to reference a string symbol (Section 11.1). The input file can be comma, space or tab delimited. First line - Contains the names of the solution values. These can be in the form of a name or a processor variable name. The exception to this is when a transient solution is being compared at several time points and the first column is time: the name for time is only a place-holder and can be “Time” or “TIMEN” (anything for that matter, but Time or TIMEN make sense). Submodel-specific processor variables must be in the submodel.Xn format, where X is a variable name like T or PL, and n is the node or lump number (or tie, path, etc.). Second line - Contains the units of the data on the following lines. The code attempts to translate these by using the internal designators for units. The same format as found on dialog boxes should be used for the same units type. Examples are: K, R, C or F for temperature; Pa or PSI for pressure; and Btu/sec/in^2 for a heat flux. Data values will be converted to model units, as defined on the Preferences dialog (Section 2.7.1), before the SINDA/FLUINT input files are generated. Remaining lines - Contain the test values to be compared to the solution values named in the first line. If the data is time-dependent, the time is provided in the first column. For time-dependent data, the solution will be forced to compare at the time points for which data is available.

Table 12-1 Data Logger Compare Input File for single comparison (steady state)

12.6.2

Output File

An optional output file can be designated by: 1. typing the filename into the Filename for output field 2. using the drop-down list to select a file in the working directory with the extension *.dlo or browse for a file with any extension in any directory

12-12

Logic Manager

Table 12-2 Data Logger Compare Input File for transient comparison

3. selecting the ‘P’ button to specify a string symbol (Section 11.1) containing the path and filename. The output file named in the Filename for output field will contain TIMEN, LOOPCT, the model variables listed in the input file, and the output symbol value for that call. Units will be in the current SINDA/FLUINT units. A comparison report will be generated at the end of the calculation for single comparison point runs (usually steady state). If the Filename for output field is left empty, no output will be generated. 12.6.3

Variable Prefix String

The DLC logic will generate 3 s for each DLC. The names and descriptions are: • vpsActive# - The value of this determines if comparisons for the associated DLC are calculated. The default value is 1 (active), but can be set by the to 0 (inactive) in logic. For logic purposes, vpsActive# is an integer. • vpsCount# - The value of this is automatically updated to show the (total number of variables compared)*(number of comparisons per variable). This value is used to calculate mean errors, as in root-mean-squared. For logic purposes, vpsCount# is a real value. • vpsError# - The error as defined by the Type of comparison to perform dropdown list (Section 12.6.4). For logic purposes, vpsError# is a real value. Some situations that would require setting vpsActive# = 0 in logic are: - A steady state solution is used to obtain initial conditions when the actual test and correlation solution are transient. During the steady state solution, the DLC is inactivated.

Logic Manager

12-13

- DLCs are defined for different test conditions and the correlations will be made sequentially in the same run. Each DLC is activated and deactivated when appropriate. Note: Disabling the DLC using the Enable/Disable button prevents the DLC logic from being written to the SINDA/FLUINT input files: this cannot be changed during the solution. Using the vpsActive# allows the DLC to be activated and deactivated during the solution in logic. In the names above, vps is the optional string entered into the Variable Prefix String field and # is the number of the DLC as described at the beginning of Section 12.6. Therefore, if ‘test1’ is entered into the variable prefix string for DLC 2, the s would be ‘test1Active2’, ‘test1Count2’ and ‘test1Error2’. Because the strings ‘Active’, ‘Error’ and ‘Count’ are added to the variable prefix string along with the DLC number, the variable prefix string plus the number of digits in the DLC number should not exceed 26 characters: 32-character name length limit for s minus 6 characters for the string ‘Active’. Underscores are the only non-alphanumeric character allowed, but must not be the leading character of the variable prefix string. 12.6.4

Comparisons

The comparison method for the DLC is designated in the Type of comparison to perform drop-down list. The options are listed below with a brief description or the function name for the COMPARE subroutine described in Section 7 of the SINDA/FLUINT manual. More information can be found in Section 5.10.3 of the SINDA/FLUINT manual.

• Time-weighted root of sum of squared values of the error1 - This method allows for uneven time points in the DLC input file, but otherwise works like the Root-Mean-Square error below. This method cannot provide a comparison report in the output file described in Section 12.6.2. • Sum of squared values of the errors - ‘SUMSQR’ • Root-Mean-Square error2 - ‘RMSERR’ • Absolute value of the maximum error3 - ‘MAXERR’ • Average of absolute values of the errors2 - ‘AVGABS’ • Sum of absolute values of the errors - ‘SUMABS’ • Sum of cubes of absolute values of the errors - ‘SUMCUBE’ • Average value of raw errors3 - ‘AVGERR’

1 Recommended to be used as an OBJECT value to be minimized. 2 Recommended to be used as an OBJECT value to be minimized. 3 Not recommended to be used as an OBJECT value to be minimized. In the case of Absolute value of the maximum error, refer to MINIMAX in the SINDA/FLUINT manual.

12-14

Logic Manager

• Standard deviation of error for above2 - ‘STDDEV’ calculating the standard deviation about the average value of raw errors A subroutine is generated for each DLC, called DALOG#, where # is the number of the DLC in the logic manager. There is one argument for the subroutine: an integer called FLAG. This indicates what that particular call should do. Most comparisons are performed in the VARIABLE 2 block of the GLOBAL submodel. The exception is the initial condition comparison performed from the OUTPUT CALLS block. For transient test comparisons, the routine keeps track of the next time to perform a comparison and will force time step sizes to ensure a comparison at the next data time point. It then checks for the times to match and does a comparison. For input files containing a single comparison point, every call to OUTPUT CALLS will generate a comparison but no time variable. The OUTPUT CALLS comparisons will not sum into the output variable for steady state solutions. All transient runs will sum into the output variable. • FLAG = 0 - Initialization and opening of the output file. Resets the next comparison time for transients to the first value in the input file. • FLAG = 1 - Indicates a call in an OUTPUT CALLS block. This call makes a comparison for the initial time of a transient solution. • FLAG = 2 - Indicates a call in VARIABLES 2 block. Most of the comparisons are done here. • FLAG = 4 - Termination - closing the output file. 12.6.5

DLC Implementation for Test Correlation

The DLC can be used simply to gather comparison information between two solutions or a solution and test data. However, the real power lies in combining it with the SINDA/ FLUINT Solver. Below are the steps that must be taken to combine one or more DLCs with the Solver. Use of the Solver and terminology is covered in the SINDA/FLUINT manual. • Call SOLVER from OPERATIONS block - access OPERATIONS block from the SINDA tab of the Case Set Information as described in Section • Define uncertainties as design variables - Uncertainties are the most likely causes for disagreement between a test and a model. Common uncertainties are, but are not limited to: heat transfer coefficients, emissivities, actual power supplied, and conductances. Design variables are set by double-clicking Design in the SOLVER section of the Dynamic tab in the Case Set Information (Section 15.2.5). • Check the box for OBJECT set From Data Loggers, if desired, in the CONTROL information of the Solver section on the Dynamic tab in the Case Set Information (Section 15.2.5). When this option is checked, logic will be added to the solution to combine the error from all DLCs. As an alternative, the can define the OBJECT based on an arbitrary function of the DLC errors. When using the OBJECT set from Data Loggers option, the comparison method should be one of

Logic Manager

12-15

the methods indicated as recommended for use as OBJECT in the previous section. • Define the PROCEDURE - The PROCEDURE is a logic block used to update the solution for a new set of uncertainties (or design variables). At the least it can be a call for a steady state solution. When a transient solution is used, the is required to reset the initial conditions of the solution using the RESPAR subroutine and also reset the data loggers using ‘CALL DALOGn(0)’, where n is the DLC number, for each DLC used in the solution. If any of the uncertainties or design variables require updates to internal conductors (finite difference or finite element), TD/RC node capacitances, or radiation calculations, the Dynamic SINDA should be used by checking the box Use Dynamic SINDA under the Dynamic SINDA Options section of the Dynamic tab in the Case Set Information (Section 15.2.5). To not use dynamic SINDA, the uncertainties should be applied as symbol expressions with the Output Above Expression To SINDA option checked (Section 11.2) or in some other way output the symbols and all related expressions to SINDA.

12.7

Array

The can specify Arrays (ARRAY DATA blocks) in the Array dialog (Figure 12-7). Arrays will be placed in the specified submodel with the specified ID. If GLOBAL is selected as the submodel, then the Array will be added to the TDPROPS submodel along with thermophysical properties. Singlet Array. This array is a columnar list of values in the units defined in Array Units. Doublet Array. This array is a two-column list of values with the independent value in the first column (X) and the dependent value in the second column (Y). The array input form is described in Section 2.10.1. Bivariate Array. A bivariate array is data that is based on two independent variables (X and Y). The input form for bivariate arrays is described in Section 2.10.2. Thermophysical Property. This option allows asg a specific array number to a property defined in the Thermophysical Property database (Section 3.2). By selecting the Property name and the Data type, the Array ID specified at the top of the form is assigned to the property data. This feature is useful for converting legacy models when the node and conductor definitions are not defined by Thermal Desktop or to ensure the proper reference by logic.

12-16

Logic Manager

Figure 12-7

12.8

Array Object Edit Dialog

COMPLIQ/WAVLIM

The COMPLQ and WAVLIM Logic Objects provide options that enable water hammer and acoustic wave modeling. The form () includes the following fields and options: Enable/Disable. Section 2.10.8. Comment. Section 2.10.4. Symbol Manager. Provides access to the Symbol Manager (Section 11.1.1).

Logic Manager

12-17

Submodel. The fluid submodel to which the COMPLQ and WAVLIM subroutines will be applied. Add COMPLQ call to add compressibility to incompressible fluids. Checking this box will add the appropriate subroutine call to add compliance to all liquid-filled tanks in the selected submodel. The compliance value is set in the Additional Compliance added to all field. For more information on this feature, look up COMPLQ in the SINDA/FLUINT s Manual. Add WAVLIM call to limit time step for fast hydrodynamic transients. Checking this box will add the appropriate subroutine call to calculate an estimated limit to the time step to capture of transients necessary for watter hammer and acoustic wave modeling. The maximum time step returned by the WAVLIM subroutine is multiplied by the value specified in the Safety factor for additional time step size field. Decreasing the time step by applying a safety factor between 0 and 1, helps ensure that the time step is adequately limited. The safety factor should be reduced until the results do not change between runs. For more information on this feature, look up WAVLIM in the SINDA/FLUINT s Manual.

Figure 12-8

12-18

Compressibility and Wave Limit Calls Dialog

Logic Manager

13

Measures

Measures are devices for estimating information within the Thermal Desktop model when a calculation point (node) is not in the location of interest, the Measures functions are available through the Measures menu (Figure 13-1) and Measures toolbar. Global visibility (Section 2.7.3) of temperature measures is controlled with Nodes. The visibility of the nodes associated with the measures is controlled with TD/RC Nodes.

Figure 13-1

13.1

Measures menu

Temperature Measure

Temperature measures act as thermocouples for a thermal model. Temperatures can be estimated for a location even when a node does not exist at the location. When Measures > Temperature Measure is selected, the is asked for a location for the measure. When the location is selected, the Edit Temperature Measure dialog (Figure 13-2) opens. A temperature measure assumes a temperature interpolated or extrapolated using nodes of a surface, solid, or finite element. The surface associated with the measure is determined based on location and tolerances (Section 13.3). Temperature measures will interpolate between the faces of a double-sided surface. Temperature measures are shown in Figure 13-3. The Measure on top is on the surface and the one on the side is offset and is measuring the insulation temperature. After Measures are created, the Measures > Update Measures command (Section 13.4) must be completed. Name. The name for the temperature measure. The name of a measure is displayed at the top of the measure. The name is replaced by the measure value while postprocessing. The name is displayed in the Model Browser. Size. The radius of the temperature measure entity. The size should be set to allow the best visibility.

Measures

13-1

Figure 13-2

Edit Temperature Measure dialog

Figure 13-3 Temperature Measure (on surface [top] and offset from surface measuring insulation temperature [right])

13-2

Measures

Sinda Interface Output . When checked, a with the name specified in Name is created in SINDA and the interpolation calculation is used as the expression for the . Output Node. When this check box is checked a node will be created in SINDA using the Submodel and ID provided in the form. By default the created node will be a boundary node. Alternatively, checking the Use Conductor and Thermal Capacitance check box will create a node that participates in the solution and is connected to the nodes of the surface using conductors proportioned based on the interpolation factors and the Conductor Value provided on the form. The node will be a diffusion node if the Thermal Capacitance field is set to a positive value or an arithmetic node if the Thermal Capacitance field is set to zero.

Connection to Model Test All TD Entities. If selected, the Measure will test all Thermal Desktop objects in the model and will map to (interpolate data from) the first object found within the tolerance (Section 13.3). Test AutoCAD Group. If selected, the Measure will test only objects in the AutoCAD group named in the field and will map to (interpolate data from) the first object found within the tolerance (Section 13.3). This can speed up temperature measure updates for very large models and can prevent accidental connection to the wrong Thermal Desktop objects. Connect to Outermost Nodes of Insulation (if found on Surface). When checked, the Measure will interpolate the temperature from the outermost insulation nodes, if they are on the mapped object.

13.2

Temperature Measures from File

The Measures > Temperature Measures from File command (RcMeasureTempFromFile) imports measure locations from a file. After issuing the command, the Measures Temperatures Input File dialog opens. In this form the Input File is the file defining the Measures as described below. Using the drop-down allows browsing for the file. The importer will either Use World Coordinate System (WCS) or Use Current Coordinate System when reading the Measure X, Y, Z locations from the input file. After Measures are created, the Measures > Update Measures command (Section 13.4) must be completed. For a complete description of the file format see Section C.1.

Measures

13-3

13.3

Mapper Tolerance

The Mapper Tolerance command allows the to set a tolerance or a progressive series or tolerances. The tolerances are used to for slight differences in the location of the measure and the entities to be measured. A progressive series will check smaller tolerances first followed by larger tolerances. A series of tolerances is listed in the input field one value per line.

13.4

Update Measures

The Update Measures command forces the measures to update their values. For new measures, this command determines the surface from which the measure will calculate its value and the interpolation expression used for the calculation. Postprocessing automatically updates the measures. When a measure finds an entity within tolerance, a line is drawn from the point of the measure to the surface and a cross is placed on the surface. If the measure is defined to obtain its temperature from an insulation node, then the cross has a circle on it. The command line will show the results of mapping the measures: how many measures were successfully mapped at each tolerance level and the total number of measures mapped compared to the total number of measures.

13.5

Snap Measure to Mapped Entity

The Snap Measure to Mapped Entity command repositions the measure to be normal to the surface being measured. To snap the measure to the negative normal (the bottom of a rectangle or the inside of a surface of revolution) then the measure must be placed on that side of the surface, offset by a small amount (less than the maximum tolerance).

13-4

Measures

14

Modeling with TD Mesher

The Thermal Desktop meshing utility, TD Mesher (“TDMesh”), is a basic meshing tool developed to bridge the gap between building models using native (finite difference) surfaces and solids, and generating a finite element mesh with a fully-featured meshing tool such as TD Direct. TDMesh can be used for quickly creating simple 2D or 3D finite-element meshes using CAD geometry that is either created within AutoCAD or imported into AutoCAD. TDMesh is a feature available within the core Thermal Desktop product. While TDMesh is a powerful mesher, the following requirements exceed its capabilities, requiring TD Direct instead: • Assembly meshing - combining multiple parts into a single mesh, including handling of merging nodes and matching meshes. • Automated mesh and network updates - automatically updating the mesh, node types, heat loads, conductors and ors, etc, based on changes to the geometry or mesh controls • Localized mesh controls - specifying mesh controls for edges or faces that are finer than the overall settings, or that benefit from different types of elements • Super Parametric Elements - Super-P elements provide curved edges and faces to provide accurate representation of the surface area and volume of the geometry TDMesh is recommended if the part geometry is too complex to be represented by Thermal Desktop finite difference surfaces or solids, but is not so complex (or likely to change) that it requires TD Direct instead. Native finite difference or “conic” surfaces are preferred when the geometry matches these surfaces, since the surface curvature is mathematically precise, rather than approximated by flat elements. Native surfaces also allow for a coarser nodal breakdown while still maintaining geometric fidelity (when a coarse breakdown is acceptable for accuracy). However, for geometry that does not fit well with the standard set of conic primitives, TDMesh provides a convenient and quick method for generating thermal models. While TDMesh provides the ability to model geometry that is not easily described by the standard set of conic primitives and finite difference solids, overly complex or defective geometry is problematic as well. First, it might not be successfully meshed by TDMesh. Or, if the complex object is successfully meshed, the resulting model might be slow to solve and perhaps cumbersome to use (e.g., when asg boundary conditions) because of the enormous number of elements. Therefore removing any details not required by the thermal analysis (small holes, fillets, chamfers, etc.), and using a simplified starting point (e.g., a thickened or extruded mid-plane instead of a thin solid) will be well worth the effort. TDMesh is not the starting point: it should be invoked after appropriate preparations have been made (see the Advanced Modeling Guide in a separate volume).


14-1

For arbitrary surface and solid geometry, TDMesh creates triangular surface elements and tetrahedral solid elements. More advanced techniques than such “tri-tet” meshes are available in TD Direct. For this reason, and for cases where TDMesh fails to accommodate a geometric detail that cannot be adequately simplified, it should be noted that some meshing needs will be best met by TD Direct or by third-party mesh preprocessors. In these situations, the analyst should refer to the section on importing finite element nodes and elements (see Section 18.2.3 in this manual and the Advanced Modeling Guide in a separate volume), or to the TD Direct documentation (TD Direct s Manual in a separate volume). TDMesh (Section 14.2) will produce triangular elements on an surface (e.g., AutoCAD region or surface). An AutoCAD region is a planar surface that is constructed out of bounding wireframes. The bounding wireframes may be straight lines, arcs, or splines. Regions may be modified by boolean operations, such as subtracting a circlular region from a larger region to create a hole. An AutoCAD surface can be a simple or complex surface created by lofting, extruding or revolving an open wireframe, or by exploding a solid. AutoCAD solids can be created in many ways including boolean operations, lofting, extruding, and revolving surfaces or closed wireframes (the starting vertex is coincident with the ending vertex). When TDMesh is applied to an arbitrary solid (3D) object, the will have the choice of generating only triangular surface elements, or generating only tetrahedral solid elements, or both. TDMesh may also be applied to AutoCAD polygon meshes. An AutoCAD polygon mesh is a collection of vertices and faces used for geometric display. Polygon mesh faces are either three or four sided polygons. An AutoCAD polygon mesh can be specified manually by creating vertices and faces, by extruding or revolving curves, or by fitting a surface between four boundary curves (e.g., the AutoCAD command EDGESURF). AutoCAD polygon meshes are the original way surfaces were geometrically represented in AutoCAD, before the recent introduction of advanced NURB based surface and solid modeling. When TDMesh is applied to polygon meshes, the FEM mesh generated matches the vertices and faces of the polygon mesh, no further subdivision is performed. Triangular or quadrilateral elements are created, depending on if the face is a three or four sided polygon. TDMesh also s extrude and revolve operations for regions, arbitrary surfaces, and polygon meshes. Again, the final output may consist of only surface elements, or only solid elements, or both. However, because of the nature of the extrusion or revolution process, the surface elements are not always triangular (and in fact, they are often quadrilaterals on side surfaces), and the solid elements are not always tetrahedrons (instead, they are often pentahedrons). Furthermore, unlike meshing an arbitrary solid using TDMesh, certain options are available when extruding or revolving: the ability to selectively generate and customize key surfaces. This capability facilitates application of boundary conditions (e.g., radiation, convection) and model reduction via the exploitation of symmetry. In the following discussion, a part is a CAD surface or solid that is the basis of the mesh to be generated. A “preview mesh” (or simply preview) is the geometric mesh made from and corresponding to the part. The geometric mesh is called a “preview” because it is simply a set of points and lines, and is not yet a thermal model. The may examine the preview to be sure it meets the needs of the thermal analysis. If acceptable, the preview can be used

14-2


as the basis of generating a TD finite element model (FEM mesh). If the part is modified, and/or meshing parameters are modified (e.g., finer resolution is chosen), then the preview and the FEM mesh may be easily regenerated.

14.1

Final Preparation for Meshing

The Advanced Modeling Guide describes how to build CAD parts in AutoCAD,1 SpaceClaim, and other software. It also describes how to use imported CAD parts and how to prepare parts before or after importing such that sensible and efficient thermal models are possible. This section therefore does not describe geometric preparations, since they are covered in prior sections. Instead, the preparations for efficient display of meshes are listed. Meshes tend to be geometrically complex to draw, and so tips for accelerating their display and refresh rates are well worth noting. First, consider turning off the visibility of Thermal Desktop nodes by unchecking the TD/RC Nodes button on the Graphics Visibility tab of the Preferences option as described

in Section 2.7.3. The locations at which nodes exist will be obvious as the vertices of the resulting mesh, and rarely will the need to select and edit FEM mesh nodes directly. Second, consider turning off AutoCAD’s selection preview. By default, when the cursor es over a mesh, it will be highlighted as a prompt. However, highlighting forces the redraw of every line in the mesh, which can be considerable even if some lines are hidden, per the options described in Section 14.2.5.1. To avoid this effect, in AutoCAD use the Tools > Options > Selection option (or type OPTIONS at the command prompt and choose the Selection tab), then turn off Selection Preview (at least “When no command is active”). The preview may be displayed as a wireframe outline of just the edges of the mesh, a wireframe of the elements with internal faces hidden, a wireframe of internal and external elements, a solid shaded view of just the external faces, or a solid shaded view of both internal and external elements. Different display options may be selected for post processing mode and in regular model mode. See Section 14.2.5.1 for more information. The default mode when model building is to show a wireframe of internal and external elements, and the default mode in post processing is to show a solid shaded view of the external faces. When in model building mode, after the resolution of the mesh has been examined and deemed acceptable, the display option may be changed to show a wireframe of just the edges to increase graphics performance. The preview will postprocess data in postprocessing mode. Therefore, the nodes and elements that have been created from the preview may be hidden for better graphics performance.

1 At the very least, the reader should be familiar with creating AutoCAD regions and solids, and with the Union, Subtraction, and Intersect commands before continuing.


14-3

14.2

TDMesh

TDMesh is used to create triangular finite element meshes of AutoCAD regions or arbitrary AutoCAD surfaces. It may also be applied to arbitrary solids, with the option of creating only triangular surface elements, or creating only tetrahedral solid elements, or both simultaneously (which is the default). TDMesh can also generate FEM meshes from AutoCAD polygon mesh surfaces. Once a region, surface, or solid (“the part”) has been constructed, or imported and simplified if necessary, TDMesh can be invoked. As will be shown, the generation of a thermal model from a geometric object is a multi-step process: • Generation of the geometric mesh (preview mesh, or simply preview) from the part, with the creation of the corresponding Mesh Controller • Assignment of surface and/or solid element thermal material properties and boundary conditions (e.g., insulation, radiation). • Generation of thermal elements and nodes (FEM mesh) from the preview Usually, these three steps should be performed in the order listed above. However, a significant feature of the TDMesh design is that any of the three steps can be revisited at any time using the Mesh Controller. In fact, the underlying part can be moved or modified as well. Of course, some modifications may require the deletion and regeneration of the FEM mesh (TD nodes, etc.) as part of the revision process, but the effort required to reconstruct a new FEM mesh has been minimized. Once a preview (geometric mesh), thermal properties, and FEM mesh have all been created, the can revisit earlier decisions without losing all of their work. As with any TD object, node, surface, and solid properties can be altered at any time, either at the element level or at the Mesh Controller level (which will override any element-level customizations). The preview mesh itself can be revised as needed (perhaps to increase or decrease mesh resolution). While this step will necessarily require deletion and regeneration of the FEM mesh (and therefore any element-level customizations), any node, surface, or solid properties that have been defined using the Mesh Controller will be preserved such that a new FEM mesh can be immediately generated using the revised preview mesh. However, please also note that certain boundary conditions such as heat loads, ors, etc. will be lost if the FEM mesh is deleted or regenerated. Modifying the underlying part (surface or solid) will require that the preview be regenerated, and this in turn requires that the FEM mesh be regenerated as well. The Mesh Controller manages the regeneration of the preview, the deletion of the previous FEM mesh, and the creation of the new FEM mesh with the previously used node, surface, and solid properties, making such modifications easy to accomplish.

14-4


14.2.1

Generating a Preview (Mesh Controller)

To construct a mesh of a planar region, surface, or solid, type TDMESH at the command prompt and then select the part to be meshed. This action invokes the creation of a new Mesh Controller object. The Mesh Controller provides access for later editing and further mesh construction steps that can be deferred if needed. The Mesh Controller dialog will appear (Figure 14-1).

Geometric (Preview) Mesh Control

Geometric Mesh Status

Element, Properties Control

(Grey for 2D parts)

FEM Mesh Control

Figure 14-1

FEM Mesh Status

Thermal Desktop Mesh Controller (surface, no mesh yet generated)

Selecting Generate Preview... invokes the following form (Figure 14-2), which controls the generation of the geometric mesh or preview. The dialog is shown for a mesh controller used on a 2D part. The options for adjusting the mesh resolution are covered in Section 14.2.2.

Mesh Resolution Control

(Grey for 2D parts)

Figure 14-2

Mesh Generation Options for a Surface Part


14-5

When Generate Preview... is invoked for a solid part, the following form (Figure 143) appears. Again, mesh resolution controls are described in Section 14.2.2. The two options that become active at the bottom of the form when the Mesh Controller is applied to a solid will now be described.

Preview Mesh Resolution Control Generate surface mesh Generate solid mesh

Figure 14-3

Mesh Generation Options for a Solid Part

If the requires a thermal model of the surface (to which radiation, convection, etc. can be applied), check the Generate Surface Mesh Preview button. If the thermal response within the solid is required, also check Generate Solid Mesh Preview. (Having both options checked is equivalent to “surface coating” a solid FE mesh.) It is possible to neglect either the solid preview or the surface preview, which means that one of the corresponding FEM meshes will not be generated in later steps: only an uncoated solid or only a hollow surface model is needed. For example, the surface preview mesh can be skipped if no surface area effects such as radiation, convection, or will be needed (which would be unusual). More commonly, the solid geometric mesh can be skipped if a “hollow” object is desired, keeping in mind that a mathematical (not displayed) thickness can be assigned to any TD surface. When a new Mesh Controller is invoked, the controller is placed on a layer named “TDFEM_MC_label.id” and the underlying part (surface or solid) is placed on a layer named “TDFEM_PRT_label.id” (with visibility turned off), where label is the optional -assigned label (Section 14.2.5.1) and id is the TD-designated Mesh Controller ID (e.g., “348” or “1A4”). Section 14.2.5.2 describes how to use these layers to easily control the visibility of each component of a TD mesh. If a Mesh Controller is deleted, the part will be restored to its original layer. 14.2.2

Controlling Mesh Resolution

The methods for controlling the resolution of meshes will be provided by example. A 2D planar example will be described in detail. This discussion is also applicable for a 2D mesh on an arbitrary curved surface, or 3D mesh on an arbitrary solid. For the TDMesh Extrude and Revolve options (Section 14.3 and Section 14.4, respectively), the mesh resolution is applied first to the base surface object, and then extruded or revolved into a solid.

14-6


The example surface to be meshed is displayed in Figure 14-4. Note that this drawing has units of inches, with the largest dimension of the part shown being approximately 1/2 inch.

Figure 14-4

Example Surface to be Meshed

The default options for the mesh preview appear in Figure 14-2. Since the part has no curved edges, the “Max Turning Angle” is irrelevant for now (it will be described later). Therefore, the resolution of the resulting mesh is controlled by one of two options shown at the top of that form as selected by the radio buttons: • The fraction of the maximum dimension of the part. • The absolute size of the largest element length along any edge. Note that the default is 1 meter (39.37 inches). This is usually not a good starting value, and should be modified appropriately by the . Normally, the Fraction of Max Dimension option should be used, as it automatically adjusts the size of the element to the size of the part. The alternative, the Absolute Size option, is intended to help make meshes that directly correspond to other meshed parts, such that subsequent merging operations are facilitated because nodes will be coincident. For example, the Absolute Size option could be used to ensure that edges of a box (whose faces are meshed as separate parts) can be merged since the nodes in each face will then be coincident along the common edges. (Note that changing the resolution of either part’s mesh will then force the to repeat the merge operation.)


14-7

The default value for the relative mesh fraction is 0.1, meaning that 10% of the largest dimension of the object will dictate the element size. For the part shown above (Figure 144), the default settings result in the following mesh (Figure 14-5):

Figure 14-5

Example Surface, Default Resolution (Fraction = 0.1)

A Mesh Controller has been created by generating this preview mesh, with the tag visible to the right of the surface (“1D53” in this case). This controller can be relabeled using the Set Label... button in Figure 14-1, as described in Section 14.2.5.1. The Mesh Controller can be accessed in the Model Browser, or edited directly in the drawing as can any other TD object. Note that the size of the thin protrusion at the right of the part has not dictated the size of the mesh, although smaller-than-average elements might be required to deal with the sharp corner and the thin region itself. Instead, the 10% fraction has been applied to the longest length. The most coarse mesh that can be created for this object, with a fraction of 100% (1.0), is shown in Figure 14-6. Even though the fraction calls for one element along the longest length, other internal parameters override this in order to capture detail and provide a smooth progression of element sizes. Each edge will have at least one element. Figure 14-7 shows a mesh that is more finely resolved than the default: a fractional resolution of 0.03 (3%). Note that the number of nodes that will be generated rises approximately with the square of the number of elements along the edge. In other words, the model size will be proportional to 1/F2, where F is the fraction employed. For a solid part, the model size grows even faster: 1/F3. This means a FEM model of a solid object meshed with F=0.01 will be roughly three orders of magnitude times larger than one meshed with F=0.1. The solution costs for radiation calculations and for transient thermal analyses will rise even faster: requiring a total growth of perhaps six orders of magnitude in solution times! Small fractions should be used sparingly, and values below 0.01 should be avoided if possible. Often, what is important to thermal analyses is a mesh fine enough to adequately represent total volume and surface area, and meshes fine enough to resolve detailed internal temperature gradients are often a luxury rather than a necessity. 14-8


Figure 14-6

Example Surface, Most Coarse Possible (Fraction = 1.0)

Figure 14-7

Example Surface, Fine Resolution (Fraction = 0.03)

In fact, if the fractional or absolute control parameter is too small, the mesher itself may take an inordinate amount of U time to compute the preview mesh. If a too-small resolution parameter is entered accidentally, use the escape key <ESC> to abort mesh construction. To illustrate the effects of the Max Turning Angle control, a circular hole will be cut in the surface presented in Figure 14-4, resulting in the following default mesh (Figure 14-8). Although some of the turning angles around the hole are smaller than the default value of 45 degrees (especially as needed to encom the thin section above the circle), the maximum turning angle is visible at the right section of the circle’s circumference. The “turning angle” is the angle that a straight line segment “turns” into the next straight line segment as the segments pave a 1D curve.


14-9

Figure 14-8

Example Surface with Hole, Default Resolution (Max Turning Angle = 45)

To better visualize the effects of this resolution parameter, Figure 14-9 shows the effects of both increasing and decreasing the Max Turning Angle parameter from the default value.

Figure 14-9

Example Surface with Hole, Max Angle = 90 (Left) and 22.5 (Right)

Another option for controlling mesh resolution is adding break points to edges. An edge that is defined by three points will be subdivided by the mesher into more elements than an edge defined by two points. Likewise, a face that is subdivided will be meshed differently than a face that is not, even if the subdivisions are in the same plane. Faces of solid geometry may be subdivided using the imprint command (Modify > Solid Editing > Imprint Edges). To ensure nodes are created along a certain line or curve on a solids face, create the line or curve and imprint it on the solid. To ensure a node is created at a specific location, create a line that ends at the point and imprint that line onto the surface.

14-10


14.2.3

Slivers, Cracks, and other Problems in Parts

Numerical inaccuracies and truncations in the geometry engines used to build representations of complex solid modeling parts are not uncommon. “Slivers,” “cracks,” and other defects can cause problems for meshers, which cannot assume that such mistakes were accidental. Figure 14-10 presents an example of a crack introduced in the previous example. In

Figure 14-10

Irregular Mesh Density Indicates a Possible Problem in the Part

this case, the mesher does not fail per se, but rather returns a unusual distribution of element sizes that should draw the ’s attention toward a possible problem in the underlying part.


14-11

If the defect is severe enough, the mesher can fail mathematically preventing the mesh from being drawn at all. Please CRTech for any questions regarding meshing failures. 14.2.4

Generating the FEM Mesh

The bottom two thirds of the form depicted in Figure 14-1 control the generation of nodes and elements. The middle third (described in Section 14.2.4.1) is used to define thermal submodels, numbering, initial temperatures, material properties, radiation, , insulation, and other thermal properties. The bottom third (described in Section 14.2.4.2) invokes the generation of the FEM mesh based on the preview (geometric mesh) and on the thermal properties. 14.2.4.1

Nodes and Properties

Thermal properties may be assigned within the Mesh Controller (middle third of Figure 14-1) independent of the preview mesh, and independent of whether or not the FEM mesh has yet been generated. In other words, customizations made in these locations are preserved as long as the Mesh Controller itself is not deleted. Newly created FEM meshes will use these thermal properties. If a FEM mesh exists and the thermal properties are edited, the FEM mesh will be updated accordingly. Edit Node Properties... By default, a new thermal submodel named “Sid” will be created for each part’s mesh, where id is the TD-assigned designation for the Mesh Controller (e.g., S99, S1c45, etc.), and the nodes will be numbered starting with 1. The can alter these choices, along with the initial temperature of the nodes, using the TD FEM Mesh Node Properties form (Figure 14-11). These designations will be preserved if either the preview or the FEM mesh is regenerated as long as the Mesh Controller is not deleted.

Figure 14-11

FEM Mesh Node Properties Form

Edit Surface Properties... allows the to set the properties of all the 2D surface

elements that are generated, either from a part that is a surface, or from the surface mesh generated on the outsides of a solid. This is the same form used for editing triangular or quadrilateral elements in Thermal Desktop. Optical Properties, Analysis Groups, Conduction/Capacitance data (including non-displayed thickness) etc. can all be set in this form. If the FEM mesh has already been generated, then this form edits all of those surfaces. If the

14-12


FEM mesh has not yet been generated, or if it has been deleted, then this form specifies the properties for any newly created surface elements generated from the preview mesh in the future. Edit Solid Properties... allows the to set the properties of all the 3D solid elements

that are generated. This is the same form used for editing solid elements in Thermal Desktop. This form can be used to set the materials, material orienters, conductor submodels, and density/conductivity scaling factors. If the FEM mesh has already been created, this form modifies the parameters of those solid elements. If the solid elements have not yet been created, or if they have been deleted, then this form specifies the properties for any newly created elements that will be generated from the preview mesh. 14.2.4.2

Generating FEM Mesh from Preview

Generate TD FEM Mesh from Preview button will create Thermal Desktop nodes and elements from the preview mesh. These are the network objects that Thermal Desktop will use for thermal calculations. Note that these FEM objects will be placed on a new layer whose name is based on the object ID and label of the Mesh Controller. Surfaces are placed on a layer named “TDFEM_2D_label.id,” and solids are placed on “TDFEM_3D_label.id,” where label is the optional -assigned label (Section 14.2.5.1) and id is the TD-designated Mesh Controller ID (e.g., “714” or “5C4”). Section 14.2.5.2 describes how to use these layers to easily control the visibility of each component of a TD mesh.

Each TD mesh will also be placed into a unique AutoCAD group named according to the same scheme used for layers. Up to three such groups can exist for a single Mesh Controller: one for the nodes (“0D”), one for the surface elements (“2D”) and one for the solid elements (“3D”), as described in more detail in Section 14.2.5.2. If the label is changed, corresponding layer and group names are automatically updated. Delete TD FEM MESH provides a means of deleting a FEM mesh, though this step is

not necessary in order to regenerate a mesh: deletion of the FEM mesh is performed automatically if the preview mesh is regenerated. Release from Controller allows a generated FEM mesh to become uncoupled from its

Mesh Controller. This is an advanced option that should be used with caution. The uncoupled FEM mesh can no longer be accessed through this or any other Mesh Controller: it is now a stand-alone set of FEM elements that must be edited separately. An uncoupled mesh should be moved into a different layer and AutoCAD group, since otherwise confusion or conflict can arise if the Mesh Controller is used to regenerate a new and independent FEM mesh. One possible use is of this option is to employ the same object to create multiple meshes, though it is better to start with multiple copies of the underlying part. Another use is to freeze the configuration of a FEM mesh such that subsequent changes within the Mesh Controller do not affect it. Nodes created on TD finite difference surfaces and solids cannot be moved or deleted directly, they are under the control of the surface or solid. Likewise with a mesh controller. Nodes cannot be moved or deleted directly by the , they must be deleted using the controller, or moved using the controller. (When moving a controller, it is recommended to


14-13

move the underlying part simultaneously. Otherwise, if the mesh is regenerated, it will be placed back at the location of the part.) Individual nodes can be moved or deleted once the mesh is released from the controller. 14.2.5

Viewing the Mesh

TDMesh uses an underlying part (CAD surface or solid), a geometric mesh or preview, and finally the surface and solid FEM elements themselves. Section 14.1 contained suggestions for preparations to avoid excessive time the software spends redrawing meshes and elements. This section describes how to modify the visibility and display of the generated objects after they have been created. 14.2.5.1

Controlling the Mesh Display

Set Label... The Mesh Controller label can be optionally defined (or redefined) using this button on the TD FEM Mesher dialog (Figure 14-1). This label will be used as part of the layers and AutoCAD group names used to contain the preview mesh, the part, and the TD FEM elements. Display Preferences... Using this button on the TD FEM Mesher dialog, the preview mesh display can be customized using the controls shown on the left side of Figure 14-12. These options are especially useful for viewing solid geometric meshes. Note that the preview’s visibility can be turned off at the AutoCAD layer level, as described in Section 14.2.5.2. Preview Display Control

TDMesh Part Visibility Preview Visibility Surface Visibility Solid Visibility Master Visibility

Figure 14-12

14.2.5.2

Editing Mesh Display, and Controlling Visibility using the Toolbar

Accessing and Controlling Visibility of Parts, Previews, and FEM Meshes

If a solid is meshed and both the surfaces and interior portions are used as the basis for generating FEM meshes, then at least 4 layers and 3 AutoCAD groups will have been generated automatically. For example, if such a solid were meshed and the resulting Mesh 14-14


Controller ID was assigned by Thermal Desktop to be “8BE,” and if the had assigned the label name “myblock,” then the following AutoCAD layers and groups would have been created: Layers: TDFEM_PRT_MYBLOCK.8BE... containing the part (CAD surface, solid) TDFEM_MC_MYBLOCK.8BE... containing the Mesh Controller TDFEM_2D_MYBLOCK.8BE... containing the 2D (surface) elements TDFEM_3D_MYBLOCK.8BE... containing the 3D (solid) elements Groups: TDFEM_0D_MYBLOCK.8BE... containing the thermal nodes TDFEM_2D_MYBLOCK.8BE... containing the 2D (surface) elements TDFEM_3D_MYBLOCK.8BE... containing the 3D (solid) elements The part layer is automatically rendered invisible when the Mesh Controller is invoked with the tdMesh command (and by having generated a preview). Otherwise, the visibilities of these layers can be individually controlled using AutoCAD techniques, or they can be controlled using the TD toolbar options shown at the right of Figure 14-12. For example, a common choice is to turn off visibility of all 3D solid layers (TD_FEM3D …) so that only surfaces are visible for a coated 3D model. AutoCAD groups can also be used to help perform selection operations. 14.2.6

Editing the Mesh using the Mesh Controller

Once a Mesh Controller has been created by generating a preview, it can be accessed by selecting anywhere on the preview, including the ID (for example, the “1D53” as shown at the right in Figure 14-5) and editing the object, or by locating the Mesh Controller in the Model Browser (List->Meshers). The Mesh Controller will be saved along with the drawing even if no FEM mesh yet exists, such that this task can be completed later. Also, as described in prior sections, any of the modeling decisions (e.g., mesh resolution, material properties, etc.) can be revisited at any time using an existing Mesh Controller. However, some operations will force the deletion of the FEM mesh (e.g., changing preview mesh resolution), which in turn might require some customizations (e.g., assignment of heaters, convection, etc.) to be repeated. 14.2.7

Moving the Mesh

To move (using the AutoCAD Move command) a mesh created using TDMesh, the best practice is to move the part and the Mesh Controller together. The nodes of a TDMesh cannot be moved away from the Mesh Controller much like a node cannot be separated from its TD surface. If the Mesh Controller is moved alone, then if the mesh is regenerated the mesh will be realigned with the part.


14-15

14.2.8

Copying the Mesh

A mesh can be copied by selecting the Mesh Controller alone using either / or the AutoCAD Copy command. A copy of the part, preview and mesh will be placed wherever the selects. New layers (as described in Section 14.2.5.2) will also be created. 14.2.9

Attaching the Mesh to an Articulator

Meshes generated using TDMesh can be attached to articulators (assemblies or trackers). The Mesh Controller is the only mesher entity (part, preview or mesh) that must be attached to an articulator for the mesh to move with the articulator. Attaching only the nodes and elements or only the part will not cause the mesh to move with the articulator. A best practice, however, would be to attach both the Mesh Controller and the part to the articulator. This ensures that the mesh remains aligned appropriately with the articulator. If only the Mesh Controller were attached to the articulator and the articulator repositioned, then regenerating the mesh from the preview would cause the mesh to realign with the part even though the mesh would still be attached to the articulator.

14.3

TDMesh Extrude

For meshing arbitrary solid CAD parts, TDMesh described in Section 14.2 is applicable. Arbitrary CAD surfaces can also be meshed, and if they are sufficiently thin, they can be modeled as planar elements whose thickness (and perhaps insulation properties) can be assigned for thermal response (even though the thickness and insulation are not depicted geometrically). For objects that are thick enough or thermally resistant enough such that through-thickness temperature gradients are important, or when a definite geometric depiction of their overall thickness is required, the TD Mesh Extrude option may be applicable. Many types of solids are most efficiently formed by extruding an arbitrary CAD surface through a length, with the mesh generated on that original surface also extruded at the same time. The resulting mesh is no longer a tri-tet mesh. Figure 14-13 shows a pentagon (the red base part) that has been meshed in the original plane, then extruded into three solid layers. The same triangular surface mesh exists at all four planes between these three layers, but at the outsides the surface elements are quadrilaterals, and pentahedral elements (extruded triangles) are used to represent the 3D elements. In addition to using higher-level finite elements, an extruded mesh has several important advantages over a 3D tetrahedral mesh that would result if the full TDMesh had been applied to the final (extruded) solid. All of these advantages relate to the fact that the key surfaces in the resulting mesh can be generated and addressed separately, which greatly facilitates application of boundary conditions that vary according to the face. This capability is con-

14-16


Figure 14-13

Example Extrusion: Pentagon Extruded into 3 Layers

trasted with that of the full TDMesh, which can only generate and address all the surfaces as a single entity. In other words, the full TDMesh recognizes only one outer surface, and it can make no distinctions based on “faces” (which might not even exist, as is true in the case of a sphere). The usage of the TD Mesh Extruder is almost identical to that of the full TDMesh explained in Section 14.2, so this section will only highlight important differences. To use the mesh extruder, type tdmeshextrude in the command prompt, then highlight an AutoCAD region or appropriate CAD surface. Figure 14-14 shows the resulting form. Note that the base, end, and side surfaces can be edited separately. “Base” refers to the surface formed by the original CAD part (surface), “end” refers to the analogous surface formed at the far end of the extrusion (opposite the base, and perpendicular to the direction of extrusion), and “side” collectively refers to all other surfaces (parallel to the direction of extrusion). Selecting Generate Preview... invokes the preview parameter options shown in Figure 14-15. This form is similar that of the full 3D mesher in of mesh resolution adjustments, but it also contains fields for defining the distance to extrude in model units (can be positive or negative), and the number of layers to employ. The number of layers can be specified to be broken down in equal size layers, or a parametric list of boundaries may be input. The parametric distance is 0 at the base and 1 at the end of the extrusion. Intermediate positions are specified as number between 0 and 1. Since the first and last boundaries will always exist, they are not input. Only the intermediate layer boundary locations need to be supplied. New elements will be extruded in a direction normal to the surface elements. “Layers” refers to the number of solid sections. The total number of copies of the base surface mesh, including the original base mesh, will be one plus the number of solid layers requested.


14-17

Base, end, and side surfaces can be customized separately

Figure 14-14

TD Mesh Extruder and Revolver Form

Base, end, and side surfaces can be generated or omitted separately

Figure 14-15

14-18

TD Mesh Extruder Parameters


These two values can be redefined at any time, but doing so will force a regeneration of the preview mesh, and this action will in turn cause the deletion of any FEM meshes that have been created based on those previews; the same effects that are caused by changing mesh resolution within the base part. This form is similar that of the full 3D mesher in of mesh resolution adjustments, but it differs in that the can selectively omit the preview mesh for any or all of the three surfaces (base, end, side) associated with the extrusion. This means the mesh extruder can create open shells (e.g., sides only) and other objects that cannot be created using the full TDMesh. For extrusion and revolving operations, an option exists to give each copy of the base mesh layer the same node IDs. For an extrusion, this would make a three dimensional thick finite element mesh, but the resulting network of nodes and conductors would be two dimensional in SINDA/FLUINT. This may be a convenient way of modeling thick parts that perhaps must attach to other parts on both sides, but it is desired to keep only a two dimensional resolution. Likewise for revolved parts, Using the same node ID’s on each layer simulates a two dimensional, radially symmetric part. Select the Repeat Base ID’s for 2D Symmetry option on the Mesh Node Properties form, as shown in Figure 14-16.

Figure 14-16

14.4

Option for Creating a 2D Symmetric Part When Extruding or Revolving

TDMesh Revolve

For meshing arbitrary solid CAD parts, TDMesh described in Section 14.2 is applicable. However, many types of solids can be formed by revolving an arbitrary CAD surface around an axis, with the mesh generated on that original surface also revolved along the arc. The resulting mesh is no longer a tri-tet mesh. Figure 14-17 shows a rectangle with three holes (the red base part) that has been meshed in the original plane, then revolved 90 degrees into six solid layers subtending 15 degrees each. The same triangular surface mesh exists at all seven planes between these six solid layers, but the sides the surface elements are quadrilaterals, and pentahedral elements (extruded triangles) are used for the solid portions of the model.2


14-19

Figure 14-17

Example Revolution: 3-holed Rectangle, 90 degrees with 6 layers

In addition to using higher-level finite elements, a revolved mesh has several important advantages over a 3D tetrahedral mesh that would result if the full TDMesh had been applied to the final (revolved) solid. All of these advantages relate to the fact that the key surfaces in the resulting mesh can be addressed separately, which greatly facilitates application of boundary conditions that vary according to the face. This capability is contrasted with that of the full TDMesh, which can only generate and address all the surfaces as a single entity, making no distinctions based on “faces” (which might not even exist, as in the example of a sphere). In addition, revolving a base mesh can model the same solid geometry with less elements than if it were meshed with tetrahedrons. The usage of the TD Mesh Revolver is almost identical to that of the full TDMesh explained in Section 14.2, so this section will only highlight important differences. To use the mesh revolver, type tdmeshrevolve in the command prompt, then select an AutoCAD region or appropriate surface, and then select an axis of revolution. Figure 1414 shows the resulting form (which is identical to that of the mesh extruder described in Section 14.3). Note that the base, end, and side surfaces can be edited separately. “Base” refers to the surface formed by the original part, “end” refers to the analogous surface formed at the far end of the revolution (in a plane containing the axis of revolution), and “side” collectively refers to all other surfaces.3 2 Revolving about an axis that touches one side of the part, or that is coincident with it, is legal. Such a revolution results in some side surface elements that may not be quadrilaterals, and in some solid elements that may not be pentahedrons.

14-20


Selecting Generate Preview... invokes the preview parameter options shown in Figure 14-18. This form is similar that of the full 3D mesher in of mesh resolution adjustments, but it also contains fields for defining the number of degrees to revolve (can be positive or negative), and the number of layers to employ. “Layers” refers to solid sections, such that the number of copies of the base surface mesh, including the original base surface mesh, will be one plus the number of solid layers. For revolutions equal to 360 degrees, the base and end meshes are automatically merged to form a continuous part. These two values can be redefined at any time, but doing so will force a regeneration of the preview mesh, and this will in turn cause the deletion of any FEM meshes that have been created based on those previews; the same effects that are caused by changing mesh resolution within the base part. The can selectively omit the preview mesh for any or all of the three surfaces (base, end, side) associated with the revolved part. This means the mesh revolver can create open shells (e.g., sides only) and other objects that cannot be created using the full TDMesh.

Base, end, and side surfaces can be generated or omitted separately

Figure 14-18

TD Mesh Revolver Parameters

If a full 360 degree revolution is performed, then the first and last surfaces are coincident. Only one set of nodes will be generated at this location, and the should not activate the properties of the base and end surfaces (e.g., radiation, insulation): only the side surface is now applicable. 3 The “side” includes the internal toroidal shapes in the example shown in Figure also considered “sides.”


14-17: internal surfaces are

14-21

The node IDs used for the base layer may be repeated for each copy of the base layer by checking the Repeat Base ID’s for 2D Symmetry option on the Mesh Node Properties form, as shown in Figure 14-16. The essentially creates a 3D representation of a 2D radially symmetric configuration. Node and conductor data for each layer is combined together before being sent to SINDA/FLUINT for analysis, and is computationally a 2D radial symmetric problem inside of SINDA/FLUINT. This will improve computational efficiency where such symmetry is applicable.

14-22


15 Case Set Manager The primary purpose of the Case Set Manager is to allow the to set up different thermal analysis cases and to make calculations from radiation calculations to creating and running the SINDA/FLUINT model to postprocessing temperatures with the click of a single button. By default, when the Run Selected Case button is clicked, Thermal Desktop will first perform the radiation and heating rate tasks for the current Case Set, compute and output the nodes and conductors, build and run a SINDA/FLUINT model, and finally, display the temperature results mapped onto the thermal model in color.

15.1

Managing Case Sets

The may set up multiple Case Sets and each can be a different analysis for the current model. Each Case Set may have different start and stop times for transient runs, different SINDA logic, different property databases or aliases, or even different symbol values, which might modify the geometry. Cases sets can be organized into groups for organization and simplified selection for editing or running multiple cases. The Case Set Manager dialog box, shown in Figure 15-1, is displayed by selecting Thermal > Case Set Manager or its corresponding toolbar icon. The Case Sets are listed in a tree format and can be rearranged by dragging and dropping. Right-click functionality provides access to the Manage Case Sets commands described below. Run n Selected Case(s) When this button is clicked, the case sets selected in the Case Set tree or all case sets in the selected case set groups will be executed. OK Cancel Help

Manage Case Sets Add. Allows the to create a new Case Set, provide a Case Set name, provide the prefix for SINDA input and output files, assign the Case Set to a group, and choose where the Case Set is created in relation to any existing Case Sets. Alternatively, the may create a separator in the Case Set tree.

Case Set Manager

15-1

Figure 15-1

Case Set Manager Dialog Box

Copy button. Provides the same capabilities as the Add button, with the exception of al-

lowing a separator to be created. The is given a choice of maintaining unique radiation filenames or using the same radiation filenames as in the original Case Set. Edit button. Display the properties of the selected Case Set or Sets (double-clicking a Case

Set name will also perform this action). Editing multiple Case Sets will modify all fields changed within the edit. The Change Name/Group button allows the to change the name of a Case Set or a group. When changing the name of a Case Set, the can also change the prefix of the SINDA files. When Edit is selected in the Case Set Manager dialog box, a Case Set Information dialog box specific to the Case Set is displayed. The Case Set Information dialog box consists of nine tabs, each of which are described in detail later in the chapter: • Radiation Tasks • Calculations • Output

15-2

Case Set Manager

• SINDA • Dynamic • Advanced • Props • Symbols • Comments The Compare button is available when two Case Sets are selected. The resulting window lists the differences between the two Case Sets. The Import and Export buttons allow the to copy Case Sets from one drawing file to another. Case set properties are discussed below. See Section 2.10.12 "Import and Export Buttons" on page 2-52 for more information. In the Options section of the form, the can select to save the drawing file when Run Case is selected. The Run with lower system priority option allows the to specify to run the job with a lower system priority. Using this option gives the other processes on the computer, such as checking E-mail or writing a document, a higher priority so as to have better response time. When Run Case is selected, the program will run the duplicate nodes check ("List Duplicate Nodes" on page 8-5). The drop-down list gives the several options on how to deal with duplicate nodes if they are found. The default option is to prompt the to renumber the nodes. The second option is to allow duplicate nodes in the model. And the third option is to automatically renumber the duplicate ids. The second and third options do not prompt for input. Multiple Case Sets may be selected using and/or <Shift> or by selecting a group name. The Run 1 Selected Case button will change to Run X Selected Cases when multiple items are selected, where X is the number of selected cases. If the Run Case button is selected, then all of the selected cases will be solved sequentially. This functionality gives the the ability to perform ‘batch’ style operations. Want "Hands-On" Information? Many of the tutorial exercises use the Case Set Manager. Complete one or all of the following exercises as a means to get some “hands-on” experience: • In Chapter 20.2 "Setting Up a Template Drawing": "Model Browser Example" on page 20-41; "Beer Can Example" on page 20-89; "Conduction and Radiation Using Finite Elements" on page 20-129; "Mapping Temperatures From a Coarse Thermal Model to a Detailed NASTRAN Model" on page 20-157; and "Dynamic SINDA Example" on page 20-201. • In Chapter 21 "RadCAD® Tutorials": "Orbital Heating Rates" on page 2153. • In Chapter 22 "FloCAD® Tutorials": "Air Flow Through an Enclosure" on page 22-3; "Heat Pipe Model" on page 22-23; "Manifolded Coldplate" on page 22-37; and "FEM Walled Pipe" on page 22-99.

Case Set Manager

15-3

15.2 15.2.1

Editing Case Sets Case Set - Radiation Tab

The Case Set radiation tasks (Radiation tab) are shown in Figure 15-2. The information contained on this tab is used to define which analysis groups and which orbits are to be used to calculate radiation conductors and heating rates. The Add function has intelligent defaults so that the output name and output submodel for that case will not only be valid names, but will also be unique. Options on the form are:

Figure 15-2

Case Set Dialog Box Radiation Task Tab

Add. Add a new radiation task by opening the Radiation Analysis Data dialog (Section 15.2.1.1). Copy. Create a duplicate radiation task from the selected radiation tasks. The should be careful with the Copy command, since the output files and submodels will have the same name. These names should be changed to avoid potential errors when running multiple Case Sets. Delete. Delete the selected radiation task(s) from the list.

15-4

Case Set Manager

Edit. Edit the selected radiation task(s) from the list. Selecting this option opens the Radiation Analysis Data dialog (Section 15.2.1.1).

Options The radiation options on the right side of the Radiation tab allow the to inform the program what to do in case a database for that job is found. A database is specified by the Radiation Analysis Group name, the Orbit name (if required), and the Optical Properties database name. Re-use calculated data is valid, otherwise recalculate. The default is to reuse the calculated data if it has determined that no changes have been made that invalidate the data. The program is very intelligent in this regard and knows if the geometry has changed, or optical properties, or node numbers. For instance, this option is smart enough to know that if a solar optical property is changed that the heating rates need to be recalculated, but the radks do not need to be. These options are also smart enough to know if only the output needs to be regenerated in case the has changed a parameter like the file name, Bij cutoff factor, etc. Recalculate data (current database will be replaced). The second option is to always recalculate the data. This option works well for small models, but can really slow down the for larger runs that do not need to be recalculated. Add rays to database, if possible, otherwise recalculate (accuracy of current database will be refined). The third option is to Add rays to the existing database. Always reuse database (no testing performed). The last option allows the to force the program to reuse the data. Important: The option to always reuse the radiation calculation database ignores changes to the model. Therefore it should be used with extreme caution. 15.2.1.1

Radiation Analysis Data

When the Add button is selected, or an existing radiation task and the Edit button are selected, then the Radiation Analysis Data dialog box is displayed. On the first tab, the Job tab (Figure 15-3), the defines the Calculation Type, the Analysis Group and the Orbit (if necessary for the Calculation Type), the Calculation Method, and the text to Add to Database Name.

Calculation Type • Radks - surface-to-surface radiation exchange • Heating Rates - heating rates due to heating environments (requires Orbit) • Articulating Radks - surface-to-surface radiation exchange for articulating geometries (requires Orbit) • Free Molecular Conduction - free molecular conduction calculation between fixed surfaces

Case Set Manager

15-5

• Articulating Free Molecular Conduction - free molecular conduction between articulating surfaces (requires Orbit)

Figure 15-3

Radiation Analysis Data Dialog Box Job Tab

• View Factors - black-body view factors for use prior to calculations using radiosity method. Using this option for heating rates allows specifying a maximum number of rays for the view factor calculation while having a different number of rays for the heating sources. • Articulating View Factors - black-body view factors for articulating geometry (requires Orbit) for use prior to calculation using radiosity method.

Analysis Group. The selects the desired Radiation Analysis Group (Section 4.1) from the drop-down list. Orbit. The selects the desired heating environment for the radiation task. The orbit selection is available for Heating Rate, Articulating Radks, Articulating Free Molecular Conduction, and Articulating View Factors calculation types.

Calculation Method Monte Carlo. The radiation calculations use Monte Carlo ray tracing. 15-6

Case Set Manager

Progressive Radiosity. The radiation calculations use Thermal Desktop Progressive Radiosity calculations to calculate radiation interchange factors and indirect environmental heating from view factors. The view factors must have been previously calculated. Add to Database Name. Choosing an option, other than None, for Add to Database Name will make the radiation calculation database unique to the solution. The remaining tabs on the form are the same that are used with the Thermal > Radiation Calculations > Set Radiation Analysis Data command, however only tabs specific to the type of calculation for the chosen radiation job will be shown. See Section 10.1 "Radiation Calculations and Output to SINDA/FLUINT" on page 10-1 for description of the tabs. Note: Any values entered using the Thermal > Radiation Calculations > Set Radiation Analysis Data command are not transferred to the Radiation Task Properties in the Case Set Manager, and vice versa. 15.2.2

Case Set - Calculations Tab

The Calculations tab (Figure 15-4) allows the to set up a SINDA/FLUINT model and have it launched through the Case Set Manager.

Figure 15-4

Case Set Dialog Box Calculations Tab

Case Set Manager

15-7

15.2.2.1

SINDA Model Options

The can choose any, all or none of the options listed in the SINDA Model Options field. If none of the options are selected then the case will only perform any radiation tasks listed on the Radiation Tasks tab (Section 15.2.1). The options are: • Generate Cond/Cap File: Calculates and outputs the NODE DATA, CONDUCTOR DATA, DATA and logic for heaters, heat loads, logic objects and data logger comparisons for the model. The SINDA/FLUINT input file (next paragraph) will reference the file named in the field. If the option is unchecked, the conductance and capacitance information will not be updated to match changes in the Thermal Desktop model. • Build SINDA Input File: Generates the primary input file for SINDA/FLUINT. This file typically includes the OPTIONS, CONTROL, OPERATIONS, and OUTPUT CALLS blocks. If unchecked and Run SINDA Model is checked, SINDA will use the filename listed in this field. • Run SINDA Model: Run the SINDA/FLUINT model listed under Build SINDA Input File. • Post Process SINDA Results: Postprocesses the SINDA/FLUINT results at the specified time. Results can be in either the SAVE file format or the Compressed Solution Results (CSR) directory format. • Execute Mapping to Stress: Maps data to existing Data Mappers. This option is only available when postprocessing is selected. The Output File Options determines if the mapping output goes directly to the Mapper’s output file (None), if the Mapper output file is appended with the Case Set name or SINDA file name, or if the Mapper output file is written to the Run directory chosen on the Advanced tab (Section 15.2.6). • Generate Log File: Generates a log file that contains all of the information written to the AutoCAD® status window. 15.2.2.2

Solution Type

The may direct the type of solution SINDA/FLUINT uses to calculate the steady state or transient response. The options in the Solution type region defines the SINDA/ FLUINT OPERATIONS logic for basic solution combinations. If more logic is required for the OPERATIONS block, the can input that from the SINDA tab (see “Thermal Inputs” on page 15-18). Once the OPERATIONS block on the SINDA tab is defined, the Steady State, Transient, and parametric inputs on this tab are no longer used, and the must input the steady or transient calls. Any combination of solution types can be selected, however, if steady state or transient are not selected a solution will not be solved.

15-8

Case Set Manager

Steady State For steady state solutions, the SINDA/FLUINT routine STEADY is used. The options for a steady state solution are: • Before Transient (if selected): When used with heating environments, this option will use the orbital average heating. Other time-varying conditions, such as boundary node temperatures will use the Initial Value. If a transient solution has been selected, it will be executed after the steady state solution. • At Each Orbit Time Position: When used with a heating environment, a steady state solution will be executed at each orbit positions. This is useful for low capacitance models. • After Transient: This option is the same as the first option, but the steady state solution will be executed after the transient solution. • Before & After Transient: This option is the same as the first option, but a steady state solution will be executed before and after the transient solution.

Transient For transient solutions, the SINDA/FLUINT solution routine TRANSIENT is used. The may override these defaults by defining a custom Operations block (see “Case Set SINDA Tab” on page 15-13). If both Steady State and Transient are selected, then the should select when the Steady State should run: before the transient, after the transient, or before and after the transient. The Calculations tab is shown in Figure 15-4. The must input the Start Time and End Time if a transient run is selected. For more information on the various solution types available, please review the SINDA/FLUINT ’s Manual. Note: If the Solution Type section is greyed out, then the Operations Block has been modified. Use "Case Set - SINDA Tab" on page 15-13 to edit the Operation Block.

Parametric A Parametric solution option is also available. This option varies the selected symbol using even increments over the specified range. Initial conditions are saved and a restart option is used to insure each run starts from the same conditions. This option automatically enables the dynamic SINDA/FLUINT mode. With this option, the must select the Edit button to access the Parametric Input dialog box shown in Figure 15-5. Select the symbol to vary during the parametric along with its starting and ending values, as well as the number of increments in the parametric. The has the option of exporting the symbol as a SINDA/ FLUINT to logic references, if any. As an alternative, the can elect to perform a design sweep. This is similar to the parametric option but is designed to aid in trouble shooting or investigating trends for a model which is set up using design variables within the Solver or reliability module (see “Case Set - Dynamic Tab” on page 15-19). This option performs a sweep of the selected design variable from the lower to upper limit over the specified “number of cases” using even increments. For more information on the design sweep option, review subroutine DVSWEEP in the SINDA/FLUINT ’s Manual. Case Set Manager

15-9

Figure 15-5

Parametric Input Dialog Box

Restart The Restart option allows continuing a previous solution. The model structure must be identical in of all objects within the model. Nothing can be added or removed. To continue a previous run, check the desired solution routine (if not defining in the OPERATIONS block), check the Restart option and select the Edit button: the Restart Edit dialog will open (Figure 15-6). In the Results drop-down, select the results location from the previous solution. If the desired results are not already in the list, the can choose to browse for a Save file or a CSR directory. Select the Select Time button to choose the solution time from which to start. If the new solution is meant to extend the previous solution and not just finish an incomplete run (perhaps due to a power failure), check the Add New TIMEND value box and type the new end time in the appropriate field. The can place the restart code into a subroutine by checking the Create Restart Code in Subroutine... box. With this option checked, the must add a CALL statement to the subroutine named in the associated field (default name is Restar). This CALL statement would likely be in a -written OPERATIONS block. If any values (such as boundary conditions) need to be changed for the restart solution, those values can be changed in the Additional input for code placed after RESTAR call field.

15-10

Case Set Manager

Figure 15-6

15.2.2.3

Restart Edit Dialog box

Convergence Criteria

The may also specify the convergence criteria used for the solution. Below is a brief description of the fields; more detail can be found by searching for the uppercase in the SINDA/FLUINT s Manual. • Max Steady State Iterations (NLOOPS): Maximum number of iterations for steady state solution. • Max Transient Iterations (NLOOPT):Maximum number of iterations for a tranient time step. • Max Temperature Change (DRLXCA/ARLXCA): Maximum allowable tem-

perature change per iteration. DRLXCA is applied to diffusion nodes and ARLXCA is applied to arithmetic nodes.The value is used for both diffusion and arithmetic nodes in both transient and steady state solutions.

Case Set Manager

15-11

• System Level Energy Balance (EBALSA): Represents an acceptable relative energy balance of the system. This convergence criterion is used for steady state solutions. • Nodal Level Energy Balance (EBALNA): Represents an acceptable absolute energy balance of all nodes in each submodel. This criterion is used for steady state solutions. The default value of zero means that the nodal energy balance is skipped.

Review the SINDA/FLUINT s Manual for guidelines on setting the convergence criteria. 15.2.3

Case Set - Output Tab

The Output tab allows the to control the types of output that are written during the SINDA/FLUINT run. This tab is shown in Figure 15-7. For transient runs, the may define the Output Increment at the top left of the dialog box. If Output Increment is set to “0”, then the actual output increment becomes TIMEND*0.01, where TIMEND is the solution end time. The can also control the submodel name used for the output in the drop-down list located at the top right corner. If the pull-down says (AUTO), then the GLOBAL submodel will be used for the output calls. The GLOBAL submodel is included in all solutions regardless of built submodels. Text Output. SINDA/FLUINT will write out ASCII data to the output filename. For transient cases, this data will be output at the supplied interval. For steady state cases, the output will be before and after the steady state solution is performed. The exception to this is the Heat Map output which can produce large amounts of data; it is only performed at the end of the steady state or transient run. Output for Color Postprocessing and XY plots. The solution results for color postprocessing directly on the geometry and XY plots are stored in either a Save file or a CSR (compressed solution results) directory. Both of these formats are binary data and the format is selected on the SINDA tab of the Preferences form (Section 2.7.4). The must specify the name for this data and the type of data that will be on this file. This data is written to the file at the end of the steady state solution and at every transient output interval.

15-12

Case Set Manager

Figure 15-7

15.2.4

Case Set Dialog Box Output Tab

Case Set - SINDA Tab

The SINDA tab allows the to define the various control, logic and data blocks that are available in SINDA/FLUINT (the program). This tab also allows the to control which submodels are built in SINDA/FLUINT. The SINDA tab lists the four inputs. • Global S/F Inputs: The first column contains the global data and logic blocks for Options, Control, , Operations, Subroutine, and Other. Double click the block name to edit information in that block. • Thermal Inputs: The second column will list the submodel names that are in the Submodel manager. Double click on the submodel name to access buttons for the submodel-specific blocks. • Fluint Inputs: The third column is for the Fluid submodels. Double click on the fluid submodel name to access the list of submodel-specific blocks. • Insert Filenames: The s lists files, one per line, that are included in the SINDA/FLUINT input file. The Insert files can contain any logic or data block but does not assume where that block belongs. Therefore the must include

Case Set Manager

15-13

the HEADER line with the data in the files. Once a dialog box for a data or logic block has data entered into it, the list on the SINDA tab will have a ‘*’ at the end of the name. This makes it very easy to determine where additional logic and data has been added.

Figure 15-8

15.2.4.1

Case Set Information Dialog Box SINDA Tab

Build Submodels

By default all submodels in the Submodel Manager of Thermal Desktop are built in SINDA/FLUINT. Selecting the Build Submodels button on the SINDA tab displays the SINDA Build Statement dialog box which tells the program which submodel not to build (see Figure 15-9). This capability will allow the to manipulate various items in the model that they may want from run-to-run, such as heating rates, heaters, or conductors, as well as the nodes for sections of the model. To exclude a submodel from the SINDA calculations, select the Defined radio button. Next, select the submodels that are not to be built. The list on the left is a list of all the submodels that are currently in the Submodel manager. Please understand that submodels in the list on the left are not the ones that are put on the Build card in SINDA//FLUINT. The build card is made from all of the submodels that are found when the Generate Cond/Cap 15-14

Case Set Manager

Figure 15-9

SINDA Build Statement Dialog Box

function is performed, as well as the radiation submodels, and also any submodels that have some kind of input from the SINDA tab. Any submodels that are in the OTHER block, or INSERTed into the model will not be on the BUILD card unless they have some definition from the three methods just discussed. Selecting Written Build Statement requires that a BUILD card be manually added to the Operations block in some manner.

Global S/F Inputs This field lists the global blocks for SINDA/FLUINT: OPTIONS, CONTROL, , OPERATIONS and SUBROUTINE (OTHER is a non-specific text input option where the must specify the block in which the text will be included). Double clicking on any of these block names opens a sub-form for the DATA blocks (OPTIONS, CONTROL and ) or a text editor for the logic blocks (OPERATIONS, SUBROUTINE and OTHER). Most fields in the data block sub-forms are explained in the SINDA/FLUINT ’s Manual. The OPTION and CONTROL data block sub-forms also provides an Additional Input field within which any data block entry may be over-ridden using the guidelines and formats provided in the SINDA/FLUINT ’s Manual. One exception to this is the Solution Type on the Thermal tab of the CONTROL data sub-form. The default is for Thermal Desktop to Auto Determine based on #nodes which means: if the Case Set contains 1,000 nodes or less then the Single Matrix method

Case Set Manager

15-15

Figure 15-10

Sinda Variables Definition Sub Form

will be used (MATMET = 2); if the Case Set contains more than 1,000 nodes then the AMGCG Single Matrix will be used (MATMET = 12). The options for Solution Type, along with their CONTROL variable settings, are: • AMG-CG Single Matrix: MATMET = 12 • AMG-CG Matrix per Submodel: MATMET = 11 • Iterative: MATMET = 0. The input field for ITERXT is written to SINDA/FLUINT • Iterative - Aggressive SOR: MATMET = 0; ITERXT = 1. The input field for ITERXT is overridden • Iterative - Classical SOR: MATMET = 0; ITERXT = 2. The input field for ITERXT is overridden • Single Matrix: MATMET = 2 • Matrix per Submodel: MATMET = 1

If the plans to change the value of MATMET during the solution to any of the matrix methods (MATMET not equal to zero), then the Solution Type must be initially set to one of the matrix methods. The Sinda Variables Definition sub-form (see Figure 15-10) allows Thermal Desktop symbols to be converted into SINDA/FLUINT s or new s to be created. Since names are limited to 32 characters, symbols that are used to create s must also have names limited to 32 characters. To convert a Thermal Desktop symbol to a , the selects the symbol name on the left and clicks the right arrow. To create a that is not based on a Thermal Desktop symbol, the types the 15-16

Case Set Manager

name into the field at the lower left of the Sinda Variables Definition sub-form and clicks the Add button. In both situations, the Variable dialog box (Figure 15-

Figure 15-11

Variables Dialog Box

11) opens. The then selects the Initial Value option: Use Global Value, Value, or Use Symbol Expression. The Use Global Value option gives the in SINDA/ FLUINT the current value of the symbol. The Value option allows the to type in any legal SINDA/FLUINT expression; this overwrites any value or expression for the symbol. The Use Symbol Expression uses the symbol’s exact expression definition as the SINDA/FLUINT definition of the ; any symbols used in the symbol expression must be converted to a . When a symbol is converted to a , the value or expression used in SINDA/FLUINT is obtained as follows: • Global value or expression given in the Symbol Manager is overwritten by • Symbol Override (Section 15.2.8 "Case Set - Symbols Tab" on page 15-28) is overwritten by • Value If a symbol override exists, it will be used unless the chooses to use a Value. If a is being created, only the Value option can be used. For more information ing symbols and SINDA/FLUINT s, see Section 11.2 "Using Symbols and s" on page 11-10. The Integer checkbox determines if the is used in SINDA/FLUINT logic as a double-precision number (unchecked) or an integer (checked). The Disabled check box will disable the use of that in SINDA/FLUINT.

Case Set Manager

15-17

With the exception of OPERATIONS, the LOGIC block edit windows will be empty if -typed text has not been previously entered. The Operations block will have ThermalDesktop-generated text based on the solution type selected on the Calculations tab (see “Case Set - Calculations Tab” on page 15-7). If the adds text into the Operations block window, the Solution Type section of the Calculations tab will be greyed out. To regain access to Solution Type, delete all text in the Operations block window. Double-clicking on OPERATIONS, SUBROUTINE or OTHER will open the advanced text editor. More information about this text editor is provided in Section 2.10.14. When writing logic blocks, the must be extremely careful to write the logic in the same units that SINDA/FLUINT is running in. If no FloCAD objects exist, those units will be the same as the current Thermal Desktop units (see “Units” on page 2-25). If a FloCAD model exists, then the units will be either strict SI or ENG units. Important: Changing Thermal Desktop units may invalidate all -written logic.

Thermal Inputs This field lists thermal submodels. Double-clicking on a submodel name will open a window with buttons for the submodel-specific LOGIC and DATA blocks. The thermal submodel-specific LOGIC blocks are Variables 0, Variables 1, Variables 2, and Output. The thermal submodel-specific data blocks are Node, Conductor, Control, Array, and Carray. Clicking on any of these buttons opens the advanced text editor window (Section 2.10.14). The text entered into logic block edit windows must be the FORTRAN-like format described in the SINDA/FLUINT ’s Manual. The text entered into data block edit windows must be in the format appropriate for that specific data block as described in the SINDA/FLUINT ’s Manual. When writing logic blocks, the must be extremely careful to write the logic in the same units that SINDA/FLUINT is running in. If no FloCAD objects exist, those units will be the same as the current Thermal Desktop units (see “Units” on page 2-25). If a FloCAD model exists, then the units will be either strict SI or ENG units. Important: Changing Thermal Desktop units may invalidate all -written logic.

Fluint Inputs This field lists fluid submodels. Double-clicking on a submodel name will open a window with buttons for the submodel-specific LOGIC and DATA blocks. The fluid submodelspecific LOGIC blocks are Flogic 0, Flogic 1, Flogic 2, and Output. The fluid submodelspecific data blocks are Flow, Control, Array, and Carray. These blocks are described in the SINDA/FLUINT ’s Manual and some, such as Flogic 1, have very strict usage limits and the must be familiar with those limits before attempting to use such logic or data blocks.

15-18

Case Set Manager

Clicking on any of these buttons opens the advanced text editor window (Section 2.10.14). The text entered into logic block edit windows must be the FORTRAN-like format described in the SINDA/FLUINT ’s Manual. The test entered into data block edit windows must be in the format appropriate for that specific data block as described in the SINDA/FLUINT ’s Manual. When writing logic blocks, the must be extremely careful to write the logic in the same units that SINDA/FLUINT is running in. If no FloCAD objects exist, those units will be the same as the current Thermal Desktop units (see “Units” on page 2-25). If a FloCAD model exists, then the units will be either strict SI or ENG units. Important: Changing Thermal Desktop units may invalidate all -written logic. 15.2.5

Case Set - Dynamic Tab

The Dynamic tab (Figure 15-12) is an interface that allows SINDA/FLUINT to call back to Thermal Desktop during the SINDA/FLUINT run. The simplest use of the capability is to show the temperatures on the model as SINDA/FLUINT is calculating. More advanced uses of this capability is to change parameters in the Thermal Desktop model, such as locations of objects, optical properties, conductivities, and orbital information, and have Thermal Desktop recompute the conductors, radks, heating rates, and then reload them into SINDA/FLUINT to reach a design goal. The Dynamic tab has four parts: Dynamic SINDA Options, Solver Data, Reliability Data, and Set Dynamic/Static Submodels. The Dynamic SINDA Option Use Dynamic SINDA checkbox enables the dynamic link between SINDA and Thermal Desktop allowing information to be ed from SINDA back to Thermal Desktop. Reset Symbols To Original Values, when checked, will reset any symbols updated through Dynamic SINDA to their original values. This can be useful if the final solution may not a viable design. When Dynamic SINDA is run, however, two files are added to the working directory: originalSymbols.sym and dynamicSymbols.sym. These files can be imported at any time to the symbol manager (Section 11.1) to change the symbol values to the initial or final values, respectively. The Show Temps While Calculating option, when checked, will update the Thermal Desktop screen at every output interval with the temperature contours. The Set Dynamic/Static Submodels button on the form allows selection of submodels that will participate (dynamic submodels) or not participate (static submodels) in Dynamic SINDA. Only dynamic submodels can be updated though the Dynamic SINDA interface. Anything defined with an expression with the option Output Expression to SINDA (Section 2.10.7) checked must be in a static submodel. The Dynamic tab also provides input to the SINDA Solver and Reliability modules. For a detailed discussion on these capabilities, please see Section 16.1 "Dynamic SINDA/ Thermal Desktop Interface" on page 16-1.

Case Set Manager

15-19

Figure 15-12

Case Set Information Dialog Box Dynamic Tab

It is highly recommended that the review the Advanced Design Features section of the SINDA/FLUINT ’s Manual. This section will details the difference between Design Variables and Constraint Variables, as well as giving insight to all of the control parameters for the Solver. 15.2.6

Case Set - Advanced Tab

The Advanced tab (Figure 15-13) allows the to specify the initial temperatures to a data point from a previous run and it also allows the to run the model in a different directory than the DWG file. If the wishes to set up a Case Set to use a currently non-existent result for Case Sets to be run in series, the can choose an existing result to define the Restart time as First or Last. After the Restart time is set, name of the yet-to-be-created results (Save file or CSR directory) can be typed into the name field instead of using the drop-down browse feature. The referenced results must be created before the current Case Set is run.

15-20

Case Set Manager

Figure 15-13

Case Set Dialog Box Advanced Tab

Set Initial Temperatures Read From Results. When checked, this option allows the to set the initial temperature

for nodes and initial values for s from the results of a previous solution. The can either type the name of a Save file or Compressed Solution Results (CSR) directory, select the name from the drop-down list, or select Browse to choose the Save file or CSR directory to use. Set Time and Nodes. Selecting this button opens the Inital Temperature dialog (Section

15.2.6.1). Time. This shows the Restart Time selection from the Initial Temperature dialog (Section 15.2.6.1) and the time for that selection if the Disable loading and displaying actual time on this page box is unchecked (see below). The temperatures are set in the SINDA/FLUINT OPERATIONS block using the LOADT subroutine.

Case Set Manager

15-21

Set Initial Lump States The Set Initial State function sets the temperature, pressure and quality of the selected lumps. If lumps are twinned in both models, the void fraction will be set in the current model, but the total twin tank volume will be unchanged. Therefore the total twinned tank volume can be different in the two models, but the void fraction will be set. Read From Results. When checked, this option allows the to set the initial states for lumps and paths from the results of a previous solution. The can either type the name of a Save file or Compressed Solution Results (CSR) directory, select the name from the drop-down list, or select Browse to choose the Save file or CSR directory to use. Set Time and Lumps. Selecting this button opens the Lump State Initialization From Results dialog (Section 15.2.6.2).

Time. This shows the Restart Time selection from the Lump State Initialization From Results dialog (Section 15.2.6.1) and the time for that selection if the Disable loading and displaying actual time on this page box is unchecked (see below). The lump initial conditions are set in the SINDA/FLUINT OPERATIONS block using the FLUIDINI subroutine.

Disable loading and displaying actual time on this page When checked, the time is not displayed with the Restart Time option in the Time field, described above.

Run in Defined Directory When checked, this option allows the job to be solved in a separate folder than the DWG file. This feature can be extremely useful when the wants to keep their DWG files on the server, but have the radiation calculations and SINDA/FLUINT output on the local machine. Please also keep in mind that radiation calculations will go much faster when run locally than run off of a server. Directory. When Run in Defined Directory is checked, the must enter a directory path name. If the directory does not exist, the is prompted to have it created.

Compiler/Linker Option Additions The compiler and linker options allow the to enter compiler and linker arguments for use with the Intel Visual Compiler, only. These strings are added to the existing options that are set in the infon* file that is used. These options do not allow changing existing options, but is a great way to add a library to the linker command. The strings can be used with Batch mode. Errors will only be detected by the compiler/linker. Therefore the should look in the messages*.txt file if any errors are reported. Compiler Additions. Strings containing compiler options. See Intel Visual Compiler documentation.

15-22

Case Set Manager

Linker Additions. Strings containing linker options. See Intel Visual Compiler documentation. 15.2.6.1

Initial Temperature Dialog

When the Set Time and Nodes button is selected in the Set Initial Temperatures region on the Case Set Advanced tab, the Initial Temperature dialog is opened. (Figure 15-14)

Figure 15-14

Initial Temperature Dialog

Restart Time Select the time of the selected solution for setting initial conditions in the current solution. Last time. When selected, the last time in the solution will be used. First time. When selected, the first time in the solution will be used.

Case Set Manager

15-23

Input Index. When selected, the may type the record number in the solution that represents the initial condition. If the solution being used for initial conditions already exists, the may click on the Select button to choose from a list of record numbers and times.

Set Temps and s Set Temps For. This list contains node IDs, in submodel.ID format, that will be initialized from the existing solution. By default, this list will contain all nodes in the referenced solution unless the solution does not exist. For nodes in this list that should not be initialized based on the solution, select the node and click on the right-facing arrow to move the node ID to the Do NOT Set Temps For list. Do NOT Set Temps For. This list contains node IDs, in submodel.ID format, that will not be initialized from the existing solution. By default, this list will be blank. For nodes in this list that should be initialized based on the solution, select the node and click on the leftfacing arrow to move the node ID to the Set Temps For list. An entire submodel can be excluded from initialization by typing the submodel name in the field just below the Set Temps For list and selecting the arrow immediately to the right of the field. If the solution does not exist and the Set Temps For list is empty, nodes can be excluded from the initialization by typing the submodel.ID in the field below the Set Temps For list and selecting the arrow immediately to the right of that field. Exclude Boundary Nodes. When checked, any boundary nodes in the current model will not be initialized based on the previous solution regardless of the list they are in. Set s For. This list contains names of s that will be initialized from the previous solution. By default, this list will be blank. For s in this list that should not be initialized based on the solution, select the name and click on the right-facing arrow to move the name to the Do NOT Set s For list. If the solution does not exist, the bottom three lists and field will not be labeled, but can be added to the Set s For list by typing the name into the bottom-most field on the right and selecting the arrow immediately to the left of that field. Do NOT Set s For. This list contains names of s that will not be initialized from the existing solution. By default, this list will contain all s in the referenced solution unless the solution does not exist. For s in this list that should be initialized based on the solution, select the name and click on the left-facing arrow to move the name to the Set s For list. 15.2.6.2

Lump State Initialization From Results Dialog

When the Set Time and Nodes button is selected in the Set Initial Lump States region on the Case Set Advanced tab, the Lump State Initialization From Results dialog is opened. (Figure 15-14)

15-24

Case Set Manager

Figure 15-15

Lump State Initialization From Results Dialog

Restart Time Select the time of the selected solution for setting initial conditions in the current solution. Last time. When selected, the last time in the solution will be used. First time. When selected, the first time in the solution will be used. Input Index. When selected, the may type the record number in the solution that represents the initial condition. If the solution being used for initial conditions already exists, the may click on the Select button to choose from a list of record numbers and times.

Set Lumps and Paths Lumps that are initialized from results will have the following values set from the results:

Case Set Manager

15-25

• TL • PL • XL • AL • DL • HL • XF • XG • MF • PPG

Tanks that are twinned in both the model and the results will be initialized with the void fraction but the total volume of the twins will not be changed in the model. Paths will only have FR set from results. Lumps Set From Results. This list contains lump IDs, in submodel.ID format, that will be initialized from the existing solution. By default, this list will contain all lumps in the referenced solution unless the solution does not exist. For lumps in this list that should not be initialized from the results, select the lump and click on the right-facing arrow to move the lump ID to the Lumps Excluded From Set list. Lumps Excluded From Set. This list contains lump IDs, in submodel.ID format, that will not be initialized from the selected results. By default, this list will be blank. For lumps in this list that should be initialized based on the solution, select the lumps and click on the left-facing arrow to move the lump ID to the Lumps Set from Results list. Entire submodels can be excluded for lumps by typing a fluid submodel name in the field below the Lumps Set From Results list and selecting the arrow immediately to the right of the submodel name field. This prevents overwriting conditions that are unique to the current solution. Paths Set From Results. This list contains path IDs, in submodel.ID format, that will be initialized from the existing solution. By default, this list will contain all paths in the referenced solution unless the solution does not exist. For paths in this list that should not be initialized based on the solution, select the path and click on the right-facing arrow to move the path ID to the Paths Excluded From Set list. Paths Excluded From Set. This list contains path IDs, in submodel.ID format, that will be initialized from the existing solution. By default, this list will be blank. For paths in this list that should not be initialized based on the solution, select the path and click on the rightfacing arrow to move the path ID to the Paths Set From Results list.

15-26

Case Set Manager

15.2.7

Case Set - Property Database (Props) Tab

All Case Sets will use the default property database as specified in Section 3.1 and Section 3.2. The Case Set Information dialog box Props (Property Database) tab (Figure 15-16) allows the to define different Optical and Thermophysical databases for a specific Case Set.

Figure 15-16

Case Set Information Dialog Box Props (Property Database) Tab

Optical Properties Alias. The Alias button allows the to change properties assigned to alias names (Section 3.2.4). Override current database. With the Override current database box checked, the optical properties assigned for radiation calculations will use the property values in the database shown beside Filename instead of the database associated with the DWG file (Section 3.1). The referenced database must contain the names of all optical properties used in the model. A database is referenced by selecting the RCO file after selecting the Browse button.

Thermophysical Properties Alias. The Alias button allows the to change properties assigned to alias names (Section 3.1.3). Override current database. With the Override current database box checked, the thermophysical properties assigned for conductance and capacitance calculations will use the property values in the database shown beside Filename instead of the database associated with the DWG file (Section 3.2). The referenced database must contain the names of all thermophysical properties used in the model. A database is referenced by selecting the TDP

Case Set Manager

15-27

file after selecting the Browse button. 15.2.8

Case Set - Symbols Tab

The Symbols tab (Figure 15-17) allows the to redefine symbols (Section 11.1) for a Case Set. This makes it very easy to parameterize a model and to set up different runs in the same DWG file.

Figure 15-17

Case Set Information Dialog Box Symbols Tab

To change the definition of a symbol, the selects the symbol in the Global List and then selects the right arrow in the dialog box. Alternately, the may double click on the symbol in the Global List. Once this is done, an Expression Editor (Section 2.10.7) dialog box will appear displaying the current symbol value. The can enter the new symbol value or expression in that dialog box. After selecting OK, the symbol will be listed on the right in the Override List. If symbol in the Override List is transferred to SINDA as a (on the SINDA tab) and the is provided a value, then the ’s value is used in the SINDA/FLUINT analysis. Symbols are used within Thermal Desktop only - symbols are not output to SINDA/ FLUINT. Therefore, any symbol will be a set value during the SINDA/FLUINT run. s are output to SINDA/FLUINT in the HEADER DATA block (see Section 2.8 of the SINDA/FLUINT ’s Manual), and, therefore, can be accessed by the during the SINDA/FLUINT run and can be included in the postprocessing file. For more information ing symbols and SINDA/FLUINT s, see Section 11.2 "Using Symbols and s" on page 11-10.

15-28

Case Set Manager

Drive Symbols From Excel The Drive Symbols From Excel allows the to define many parametric runs with one Case Set and an Excel spreadsheet. When this button is selected, the Drive Symbols for Multiple Cases dialog (Figure 15-18) opens. To access this functionality, the Enabled

Figure 15-18

Drive Symbols for Multiple Cases dialog

checkbox must be checked. Excel File Input Definitions. This section defines the spreadsheet and range in the spreadsheet containing the symbol definitions. Choose the Excel file from the Excel file dropdown. Choose the sheet containing the symbol values from the Sheet drop-down. Choose either Auto Determine Used Range or Input Range to define the range of cells containing the symbols and values. The Test button will create a text file showing the cells that will be read from the Excel file. The Excel file cell range will be in the following order: Table 15-1 Excel file cells for Drive Symbols From Excel

symbolA

symbolB

symbolC

symbolD

symbolE

value1A

value1B

value1C

value1D

value1E

value2A

value2B

value2C

value2D

value2E

The Excel cells contain the names of symbols to change in the first row and each row after would contain the values for the solutions, one solution per row. Table 15-1 would set values for five symbols (one for each column) and generate two cases (one for each row below the symbol names). Run Time Options. The option to Put the files for each case in a separate directory creates a directory for each row in the Excel range; all solution files, including radiation databases, are written to the appropriate directory. The directories will be named Run_0, Run_1, etc. This option provides the cleanest file structure, but generates the most files since the radiation databases are generated for each case.

Case Set Manager

15-29

The option to Make SINDA file names unique generates all solution files within the working directory (location of the DWG file). With the Make RADCAD file names unique checked, the radiation files for SINDA are made unique for each case. This combination creates almost as many files as the separate directory option, but radiation databases can be reused if subsequent cases have identical radiation conditions. The option to Make SINDA file names unique, when Make RADCAD file names unique is unchecked, keeps the fewest files, but the RADCAD files are rewritten for each case. This option will also reuse radiation databases when possible. To reuse radiation databases most efficiently, the order of the symbol changes must be carefully planned to avoid changing symbols that will affect the radiation calculations in as long a sequence as possible. Excel Output Definitions. When Enabled, the Excel Output Definitions allows the to define symbols that are written back to an Excel file. If the same Excel file is selected under both Excel File Input Definitions and Excel Output Definitions, a different Sheet should be selected. The symbols to be written out are typed into the Symbols field, one name per line. 15.2.9

Case Set - Comments Tab

The Comments tab (Figure 15-19) allows the to input comments about a Case Set. Adding comments is a way to add details about the case without having to make the Case Set name long enough to be fully explanatory. With added comments, details about the case are easy to view.

Figure 15-19

15-30

Case Set Information Dialog Box Comments Tab

Case Set Manager

16 Running SINDA/FLUINT

16.1

Dynamic SINDA/Thermal Desktop Interface

The capability to have SINDA change a value and then have Thermal Desktop recalculate radiation jobs and conductance/capacitance exists and is called Dynamic SINDA. This capability is useful for letting the SINDA Solver dynamically solve your problem for you. There are several steps for the to do to make this happen. All of these steps involve defining the proper items in the SINDA data of the Case Set Manager. 1. The must select Use Dynamic Sinda on the Case Set Information dialog box Dynamic tab. This tells Thermal Desktop to start the dynamic connection between SINDA and Thermal Desktop. 2. The must define the Thermal Desktop symbols that will be changed as design variables. 3. The must write their own logic to have SINDA call back to Thermal Desktop. This logic would typically be in the Operations block, but could also be in the Procedure block if the Solver is being used. 4. It is extremely important that the does not touch or close Thermal Desktop while the SINDA is still running. Performing any command in Thermal Desktop will cause the program to crash. 16.1.1

Subroutine Calls from SINDA to Thermal Desktop

16.1.1.1

TDOBJ

This routine returns the value of OBJECT to Thermal Desktop for display in the status window. This routine should be called after OBJECT has been calculated. 16.1.1.2

TDSETDES

The routine TDSETDES will send all of the design variables and their current values to Thermal Desktop. This command will also make a call to TDUPDATE. This routine has no arguments. Using this routine is equivalent to the separate calls to TDSETREG for each design variable.

Running SINDA/FLUINT

16-1

16.1.1.3

TDSETRAN

This routine is the same as TDSETDES, except it returns all the random variables back to Thermal Desktop. 16.1.1.4

TDSETREG

This routine allows the to set a single /symbol in Thermal Desktop. A sample command would be: CALL TDSETREG(‘XLEN’, 4.3)

Once all the s have been set, this command should be followed with a command to TDUPDATE. TDUPDATE will update all the symbol changes in Thermal Desktop. 16.1.1.5

TDSETREGINT

This routine allows the to set a single /symbol in Thermal Desktop from an integer-type . A sample command would be: CALL TDSETREG(‘XCOUNT’, 4)

Once all the s have been set, this command should be followed with a command to TDUPDATE. TDUPDATE will update all the symbol changes in Thermal Desktop. 16.1.1.6

TDSETREGSTR

This routine allows the to set a single /symbol in Thermal Desktop from a character string, limited to 95 characters. A sample command would be: CALL TDSETREG(‘BCM_file_name’, ‘hot_CFD_data.inp’)

If the character string is longer than 95 characters, the string will be truncated at 95 characters and a message will be written to the SINDA/FLUINT log file. Once all the s have been set, this command should be followed with a command to TDUPDATE. TDUPDATE will update all the symbol changes in Thermal Desktop. 16.1.1.7

TDUPDATE

This command will cause Thermal Desktop to update all the entities that use symbols that have changed. This command should be called after the TDSETREG commands have been called. This command is called internally from TDSETDES. 16.1.1.8

TDCASE

The TDCASE routine will cause Thermal Desktop to calculate the radiation jobs and to output Cond/Cap data for the current case. This command has no argument. This command should be input after any commands to update symbols has been performed.

16-2


If the desired to only run specific radiation or CondCap jobs in the Case Set, then they could call TDCASE2, which allows arguments: CALL TDCASE2(‘rad0 rad3 cc’)

This command would simply run the first and fourth radiation jobs in the current Case Set, along with con/cap calculations. A call to TDCASE will run all the jobs in the Case Set. 16.1.1.9

TDSETALO/TDSETALT

These command are used to change either Optical (TDSETALO) or Thermophysical (TDSETALT) property aliases. The command is simply the alias name followed by new property. Please note that both aliases and property names are case dependent in Thermal Desktop. Sample commands are: CALL TDSETALO(‘coating’, ‘whitepaint’) CALL TDSETALT(‘material’, ‘aluminum’)

16.1.1.10 TDCMD This is the generic interface command. This command has only a single argument that is a character string. All of the other commands actually reference this command. The actual commands are: TDSETREG(‘xlen’, 1.2)

TDCMD(‘set xlen 1.2’)

TDCASE

TDCMD(‘case’)

TDOBJ

TDCMD(‘object 55.5’)

TDUPDATE

TDCMD(‘update’)

16.1.1.11 TDCMD - Mapping Commands There are several command available to help with mapping via the COM interface. These commands are: mapnastran This command tells the program to do the mapping of the data to the nastran file. The inputs are the input filename followed by the output filename. A third argument can be input to set the tolerance. If the third argument is left off, then the program will use the tolerance that is currently stored on the drawing file, which may be variable tolerancing. setmapset

This command has one argument and tells the program to set the current map set via the input name. This command should be input before the mapnastran command described above. Please note that if the name is not found in the list, then the entire model will be used for the map set.

setmapconstanttol This command has one argument and that is the tolerance to be used by the mapping command. This value is assumed to be in the current units of the DWG file.


16-3

setmapvariabletol This command is used to set the tolerancing of the next mapping command. For this command, the will simply enter as many values as they desire. The values should be decreasing in order, and should be separated by a space. The command to map the temperatures to a NASTRAN model might look like this: CALL TDCMD(‘mapnastran inputfile.txt outputfile.txt .001’)

The last parameter is the tolerance. The input tolerance is assumed to be in the same units as the model. Note that if the tolerance is not input on the command, then the program will use the current tolerance in the program, which may be variable tolerancing. 16.1.1.12 TDCMD - OUTPUT An additional command has been added to output messages to the status window. This command has the following form: CALL TDCMD(‘output This message is output to the window’)

Actual usage might look something like (make sure the ‘+’ sign is in column 6): F

CHARACTER*72 TEMP

F

WRITE(TEMP,*) ‘OUTPUT T1 = ‘, + MAIN.T1

F

CALL TDCMD(TEMP)

16.1.1.13 TDCMD - Postprocessing There are several commands that allow the to control postprocessing functionality. These commands are: post process Used to postprocess a simple ascii input file. The input is a simple filename. An additional option is to input an integer that will serve as a time second delay. ppsavefile

Used to postprocess a SINDA save file. An additional option is to input an integer that will serve as a time second delay.

ppsavefile


ppsavefile


ppsavefile


The ‘post process’ command tells Thermal Desktop to postprocess a generic ascii input file. This command is not normally called by the , but is called by the DUMPT routine that exists in SINDA. The command is:

16-4


CALL TDCMD(‘post process file.dat 2’)

16.1.1.14 TDCMD - START CASE The interface command provides the ability to start a job in the Case Set Manager. This is done from the ISIGHT interface. If the case name is not input, the program will run the current case. The subroutine call has the following form: CALL TDCMD(‘startcase casename’)

16.1.1.15 TDCMD - Send command to AutoCAD This command allows the to send a command to the AutoCAD command language. The subroutine call has the following form: F

CALL TDCMD('acadcommand "zoom" "extents"')

The command is in the AutoLisp format which means the actual command sent to AutoCAD would be: (command “zoom” “extents”)\n

The parenthesis, command, and carriage return on the command are automatically added by the program before the command is sent to AutoCAD. 16.1.1.16 DUMPT This command writes the temperatures in an ASCII format to the input file (first argument). The may also put a time delay (second argument in seconds) to slow down the process for picture capturing. This routine has also makes the call to Thermal Desktop to display the temperatures. The form of this subroutine is: CALL DUMPT(‘dynamictemps.dat’,0)

16.1.1.17 TDGVALUE/TDHRVALUE The routines TDGVALUE and TDHRVALUE provide a method for the to access a conductor or heat flow through a conductor while they are running in the dynamic mode. Each call has 3 arguments and will return a value of -1 in the third argument if the conductor does not exist. Sample calls are: CALL TDGVALUE(‘MAIN’, 10, XTEST) CALL TDHRVALUE(‘MAIN’, 10, XTEST)

16.1.1.18 TDGETSYM The routine TDGETSYMBOL will return the value of a symbol in Thermal Desktop back to SINDA. An example of where this routine could be used is for orbit parameters, such as the period, and shadow entry/exit times. A sample call is: CALL TDGETSYM(‘hrPeriod’, XTEST)


16-5

16.1.2

Sample Dynamic Calls

While no two problems are exactly alike, the following sections shows some sample calls. 16.1.2.1

Solver Sample Calls

Once the has defined the design variables and constraints, they must also construct the OPERATIONS and the PROCEDURE blocks. A sample OPERATION block might look like: CALL SOLVER CALL DESTAB

A sample PROCEDURE block might look like: CALL TDSETDES $ Send design variables to Thermal Desktop CALL TDCASE $Make Thermal Desktop recalc radiation/condcap CALL STEADY $ Solve temperatures OBJECT = MAIN.T6 $ Maximizing the temperature of Node 6 CALL TDOBJ $ Send OBJECT back to be displayed

16.1.2.2

Parameter Sample Dynamic Calls

If the only wants to vary a simple parameter, then a sample OPERATION block might look like: DO 10 BETA = 0, 90, 1 CALL TDSETREG(‘BETA’, BETA) CALL TDUPDATE CALL TDCASE CALL STEADY TIMEN = BETA CALL SAVE( ‘T’, 0 ) 10

CONTINUE

In this case, BETA must be a Thermal Desktop symbol and also a SINDA . This logic would then vary the BETA angle from zero to ninety, in increments of 1. The TDSETREG command tells Thermal Desktop to change BETA to the new value. TDUPDATE causes Thermal Desktop to update the model from the new values, including updating the orbit for the new beta angle. TDCASE causes Thermal Desktop to recalculate the heating rates for the new orbit. STEADY causes SINDA to calculate the new temperatures. The TIMEN and CALL SAVE lines cause the data to be written to a Save file for later postprocessing. Setting TIMEN equal to BETA is a trick so that the X-axis value changes (otherwise, all the data on the file would be at TIMEN=0).

16-6


16.2

Batch Mode Processing of Case Sets

For those s who would like to run Thermal Desktop cases in a batch mode, the following are commands and examples for setting up the run. To run a model in a batch mode, 2 files must be created on the harddrive using some kind of text edit program such as Notepad. The first file must be a batch (.bat) file, and is used to start up AutoCAD; the second file contains the commands to perform in AutoCAD that will run the desired cases and then exit AutoCAD. Available Commands to Run Case Sets rcRunAllCase Sets

Runs all Case Sets.

rcRunSelectedCaseSets

Runs the Case Sets that were selected the last time the model was saved.

rcRunInputCaseSets

This input requires the to type in zero based numbers for the Case Sets ie 0,2,6

AutoCAD Exit Command rcExitAutoCAD

Command to exit AutoCAD

Samples of Files The first file will be named “run.bat” and will contain the following 2 lines that are in the box below. Each line contains 3 pieces of critical information. •

The first piece is the location of the AutoCAD executable on the computer. It may be necessary to use Windows Find File to find the proper location of the AutoCAD executable, acad.exe, on the computer.

•

The second piece of information is the DWG file that is to be opened.

•

The third piece of information is the /b filename.

In the example “/b commands” means the file on the harddrive named “commands.scr” contains the commands used to run specific Case Sets. The run.bat file and the commands.scr file should be located in the same directory.


16-7

Samples of Files "c:\Program Files\AutoDesk\AutoCAD 2013\acad.exe"

"c:\Program Files\Cullimore and Ring\Thermal Desktop\Tutorials\Thermal Desktop\beercan\Completed\beercan.dwg" /b commands "c:\Program Files\AutoDesk\AutoCAD 2013\acad.exe" "c:\Pro-

gram Files\Cullimore and Ring\Thermal Desktop\Tutorials\Thermal Desktop\heatpipe\Completed\heatpipe.dwg" /b commands Note: If the file cannot be found try removing the DWG extension from the filename. An example of the command file, commands.scr, is shown in the box below. In this case, all of the Case Sets that are defined will be run, and when they are finished, the program will open up another drawing file and then run all of its Case Sets. The script ends with AutoCAD being exited. rcRunAllCaseSets (command "_open" "c:\\Program Files\\Cullimore and Ring\\Thermal Desktop\\Tutorials\\Thermal Desktop\\finiteElement\\Completed\\fe1.dwg") rcRunAllCaseSets rcExitAutoCAD Important: The above are for example purposes only. CRTech does not recommend running models directly in the Program Files directory.

16-8


17 Postprocessing Postprocessing, examining calculation results, in Thermal Desktop can be performed through: • Color Postprocessing - Section 17.1 - displaying calculation values on graphical objects as colors or shades • X-Y Plotting - Section 17.2 - displaying calculation values on an X-Y plot, usually as a function of time • Query Node - Section 17.3 - obtaining calculation values by selecting graphical objects • Results Queries - Section 17.4 - generating summary files based on one or more results files.

17.1

Color Postprocessing

After calculations are performed, the results may be viewed using Thermal Desktop’s postprocessing facility. View factors, radks, heating rates, SINDA/FLUINT results, and generic text files can all be postprocessed directly on the graphical entities. Multiple color bars may be displayed so that data from different objects may be examined simultaneously or objects in different viewports can have different scales. For example, temperatures may be displayed for thermal nodes at the same time that pressures are displayed for fluid lumps. Figure 17-1 shows a combined fluid/thermal model with results displayed using all four available color bars. Node values are shown as contours, by default, with colors interpolated between node locations on a continuous surface or solid. For finite difference objects, the contouring can be turned off such that the entire nodal area or volume displays the node value. This setting is found in Section 2.7.2. Thermal Desktop saves information about data to be postprocessed in a postprocessing dataset. The dataset stores the file name containing the data, the type of data to be plotted, the current time value if appropriate, and a comment to describe the data. The actual data to be plotted remains external to the postprocessing dataset. Postprocessing datasets are named by the and are stored in the drawing file.

Postprocessing

17-1

Status bar

Space Mode Button (MODEL or PAPER)

Figure 17-1 Postprocessing with Four Color Bars for Node, Lump, Path and Tie Data

17.1.1

Layouts

Layouts are tabbed windows in AutoCAD for arranging viewports and color bars (Section 17.1.2). By default a new drawing has two layouts named Layout1 (or TD PP for preexisting models) and Layout2, which can be renamed. The tabs are found at the bottom of the graphics area along with a tab for Model (the Model tab is the default working space), as seen in Figure 17-2.

Figure 17-2

Layout Tabs and the Model Tab

Right-click on any layout tab or the Model tab to access a menu to • Create a new layout (it is recommended that the number of layouts be minimized) • Delete a layout

17-2

Postprocessing

• Rename a layout • Move or copy a layout • Activate the previous layout • Activate the Model Tab • Hide the Layout and Model tabs Activating the model tab can also be done using Reset Thermal Desktop Graphics command (Section 7.1) or Thermal > Post Processing > Post Processing Off. Hiding the Model and layout tabs can allow a slightly larger graphics area. The Model and previous layout are displayed as buttons in the status bar at the bottom of the window. Clicking the third button will display a preview of all layouts and the Model view. Rightclicking the layout button allows turning the tabs back on and the buttons off.

Figure 17-3 hidden

Layout, Model and Quick View Layouts buttons in status bar when tabs are

Thermal Desktop will switch to the most recently viewed layout automatically when a postprocessing command is issued.

Paper Space and Model Space on a Layout Paper space is the overall layout including viewports and color bars (Section 17.1.2). In order to move viewports or color bars or resize them, the must switch to paper space. Switch to paper space by double-clicking anywhere in a layout outside of a viewport. Model space is the area within a viewport. The model space in a viewport behaves just like the Model tab for zooming, panning, rotating and selecting parts of the model. To switch to the model space of a viewport, double-click within a viewport or press Ctrl+R to cycle through the existing viewports. While in paper space mode, zoom and pan functions work on all objects (viewports and color bars) on the layout. To zoom-to-fit all objects on the layout without switching to paper space, use the Zoom Paper Space Command: • Command: rcZoomPaperSpace • Menu: Thermal > Post Processing > Zoom Paper Space • Ribbon: Thermal > Post Processing > Zoom Paper Space

Postprocessing

17-3

17.1.2

Viewports and Color Bars

Viewports in layouts (Section 17.1.1) can be considered frames for views of the model.

Viewports can be created on layers and the layers can be used to control the color and visibility of the viewport frame (not the model view in the viewport). The default postprocessing layout has one viewport. Multiple viewports can have independent views of the model and can have different visibility of entities using VP Freeze in the AutoCAD layers. Color bars are objects in layouts (Section 17.1.1) that show the color scale for postpro-

cessed data. Once created, color bars can be moved and resized while in paper space (Section 17.1.1) by selecting and moving the grip points. A color bar will be vertical when the height is greater than the width and horizontal when the width is greater than the height, as seen in Figure 17-4. Different color bars may be for different types of objects (nodes, lumps, paths) or for different viewports. In Figure 17-4, the top, horizontal color bar is associated with the left viewport and the right, vertical color bar is associated with the right viewport. If a model is postprocessed without setting up any layouts, viewports, or color bars, the default configuration will be a single viewport and a single node temperature color bar that is auto-scaled to the temperatures of the displayed objects.

Figure 17-4

17-4

Layout with multiple viewports and color bars

Postprocessing

17.1.2.1

Editing Viewports and Color Bars

• Icon: • Command: rcColorBar • Menu: Thermal > Post Processing > Edit Layout ColorBar/Viewports • Ribbon: Thermal > Post processing > Edit Layout Colorbars/Viewports • Toolbar: Post Processing When this command is issued, the Edit Color Bars and Viewports on Layout: dialog opens (Figure 17-5). The dialog title will be followed by the name of the layout for which the color bars and viewports will be edited. If the Model tab was active, then the last viewed layout will be used.

Figure 17-5

Edit Color Bar and Viewports dialog: Color Bar (node) tab

Color Bars (Node, Lump, Path, Tie, FTie, and IFace tabs) To edit the color bar options, select the tab for the color bar to be edited (Figure 17-5). In the graphics area, the color bar associated with the selected tab will be highlighted. The following options can be set: Shading Scale Type. Color scale can use Color or shades of Grey

Postprocessing

17-5

Num. Shades. This value is the number of bands in the color bar range. The maximum value is 15. > Max Color. Select this button to set the color of objects that have a postprocessed value above the range of the color bar. < Min Color. Select this button to set the color of objects that have a postprocessed value below the range of the color bar. Data Range Auto Scaling. The drop-down list specifies the auto-scaling of the color bar. The list includes: • Off - Input For Min/Max: The color scale is determined by the values in the Min Data Value and Max Data Value fields. • On - Program Calculates Visible Min/Max: The color scale is determined by the values of visible objects. Either nodes or surfaces can be visible for the node color bar scale. • Max Value Only - Sets Min: The maximum value of the color scale is determined from the visible objects of that color bar’s type. The minimum value is set in the Min Data Value field. • Min Value Only - Sets Max: The minimum value of the color scale is determined from the visible objects of that color bar’s type. The maximum value is set in the Max Data Value field. Note: When autoscaling, the node color bar determines the scale based on nodal temperatures: when color contours (Section 2.7.3) and centered nodes (Section 4.3.1.1) are used, the temperature contours will possibly be extrapolated beyond the scale determined by the node temperatures. Min Data Value. This value specifies the minimum value of the color bar range when autoscaling is Off or Max Value Only. Max Data Value. This value specifies the minimum value of the color bar range when autoscaling is Off or Min Value Only. Use Log Scale. When checked, the color bar will use a log scale. Text Label. Text typed into this field is added to the color bar label. Label Position. The selection of Top, Bottom, or Side determines the relative location of the label to the color bar. Label Justify. The selection of Left or Right determines the justification of the label. Number Orientation. The selection of Along or Perpendicular determines if the numbers are along the length of the color bar or perpendicular to it.

17-6

Postprocessing

Append dataset information to label (time, type,...). When checked, the data type and time will be appended to the Label. Show File Name. When checked, the file name of the results file will be displayed with the Label. Significant Digits. This value determines the number of significant digits displayed on the color bar. Scale Text. This value determines the size of the label and values text with respect to the default size. The default size is scaled to the width of a vertical color bar or the height of a horizontal color bar. Track Areas. When checked the calculated surface area for each color band will be displayed. This option is only available for node color bars. Track Volumes. When checked the calculated volume for each color band will be displayed. This option is only available for node color bars. Disable Data Value Display. When checked, the color bar values will not be displayed. This is useful for obfuscating solution values for presentations. Miscellaneous options Visible. When checked, the color bar will be visible. If a color bar is not visible, then objects of that color bar’s type will be drawn based on the viewport settings (Section 17.1.2.1). Active for all viewports that do not have a color bar assigned to them. When this option is selected, the color bar will be active for any viewport that does not have a color bar specifically assigned to it. Active for specified viewports. When this option is selected, the can select the Add button to select the viewport for which the current color bar will be active. The Delete button will remove viewports from the list.

Viewports (VP tabs) To edit the viewport options, select the tab for the desired viewport (Figure 17-6). In the layout, the viewport associated with the selected tab will be marked with a large, red asterisk. On the VP tabs, following options can be set: View From. Selection in the drop-down list determines the view of the model in the viewport. The list includes: • Current View - model remains in the view orientation selected by the . • Sun - model is seen as viewed from the Sun • Planet - model is seen as viewed from the Planet

Postprocessing

17-7

Figure 17-6

Edit Color Bar and Viewports dialog: Viewport (VP) tab

Display Model in Heating Environment. When checked and if the dataset (Section 17.1.3) has an environment selected, the environment (typically the planet) will be shown in the viewport with the model in the appropriate position and colored for postprocessing. The environment display uses the Orbit Display Preferences (Section 6.2.1). Advancing through time in the dataset moves the model to the time-appropriate position. Zoom Extents. When checked the viewport will be zoomed to fit all visible objects when the display is updated. Draw objects that do not have colorbars as. The selection in the drop-down determines how objects without a color bar are displayed. The list includes: • Wireframe - Grey • Shaded - AutoCAD Color Override current dataset plot insulation nodes. When checked, the selection in the dropdown will be used for the current viewport in place of the insulation display selection in the dataset (Section 17.1.3). The list includes: • Do Not Plot Insulation Nodes - surfaces with insulation will show the color for the underlying surface and not the insulation temperature • Plot insulation on both sides - surfaces with insulation on either side will show the color for the insulation or the surface if insulation does not exist • Plot insulation on top side - surfaces with insulation on the top side will show

17-8

Postprocessing

the color for the insulation on the top side; and the color for the surface on the bottom side or if insulation does not exist on the top side • Plot insulation on bot side - surfaces with insulation on the bottom side will show the color for the insulation on the bottom side; and the color for the surface on the top side or if insulation does not exist on the bottom side Viewports can be moved or resized by switching to paper space (Section 17.1.1), selecting the frame of the viewport, and moving the grips at the corners. 17.1.2.2

Creating Viewports and Color Bars Note: While AutoCAD has its own command for creating layout viewports (MVIEW), this manual will focus on the Thermal Desktop commands for creating and editing layouts since the Thermal Desktop commands provide functions specific to postprocessing.

Color Bars By default, a color bar of each type already exists, but only the node color bar is visible. Create an additional color bar by selecting Create Colorbar button. If a fluid model exits, a Select Color Bar Type dialog will open allowing the selection of the color bar type. The Edit Color Bars and Viewports dialog is hidden, the paper space is activated and the following prompts are given: >: Pick or enter lower left corner: >: Pick or enter point for top right corner:

Pick or enter the requested corner points and the Edit Color Bars and Viewports dialog will reopen with an additional tab for the new color bar.

Viewports Create a viewport by selecting the Create Viewport button in the Edit Color Bars and Viewports dialog (Section 17.1.2.1). The Edit Color Bars and Viewports dialog is hidden, the paper space will be activated and the following prompts are given: >: Pick or enter lower left corner: >: Pick or enter point for top right corner:

Pick or enter the requested corner points and the Edit Color Bars and Viewports dialog will reopen with an additional VP tab. 17.1.2.3

Cycling Viewports

• Command: +R The above key combination will switch from one viewport to the next. The can also double-click within a viewport to make it active.

Postprocessing

17-9

17.1.2.4

Cycling Color Bars

• Icon: • Command: rpCycleColorBar • Menu: Thermal > Post Processing > Cycle Color Bars • Ribbon: Thermal > Post processing > Cycle Color Bars • Toolbar: Post Processing When issued, this command cycles between the difference types of color bars. Only one of each type (node, lump, etc.) of color bars can be defined for the current layout. If Smart Color Bar Cycling (Section 17.1.3.3) is on, then only those objects related to the active color bar are visible. Smart color bar cycling will only display one color bar at a time. 17.1.2.5

Reset Viewports and Color Bars

• Command: RpResetPaperSpace • Menu: Thermal > Post Processing > Reset Color Bars and PP Viewports • Ribbon: Thermal > Post processing > Reset Color Bars and PP Viewports When this command is issued, all color bars and viewports will be removed and replaced by a single color bar and a single viewport. All color bar and viewport settings will be returned to their default condition. 17.1.3

Postprocessing Datasets

Postprocessing datasets define the information to be postprocessed. A dataset includes a reference to a file containing the values to be postprocessed, time or record to be displayed, values to be displayed, a description of the dataset, and other options based on the type of dataset. 17.1.3.1

Display Current Dataset

• Icon: • Command: rcColorBar • Menu: Thermal > Post Processing > Display Current Dataset • Ribbon: Thermal > Post processing > Display Current Set • Toolbar: Post Processing

17-10

Postprocessing

When this command is issued, the current dataset will be displayed. If the Model tab is active, the previously active layout will be activated. This command can be automatically issued when a solution is completed by selecting the Post Process SINDA Results option on the Case Set Calculation tab (Section 15.2.2). 17.1.3.2

Edit Current Dataset

• Icon: • Command: rcEditDataset • Menu: Thermal > Post Processing > Edit Current Dataset • Ribbon: Thermal > Post processing > Edit Current Dataset • Toolbar: Post Processing When this command is issued, the Set Dataset Properties dialog opens for the current dataset (Section 17.1.3.3), if a dataset has been created. The dialog will be specific to the dataset type selected when the dataset was created. For the dialog description see the section for the dataset type: • SINDA/FLUINT Dataset - Section 17.1.3.4 • Text File Dataset - Section 17.1.3.5 • Text Transient Dataset - Section 17.1.3.6 • Radks or Form Factors Dataset - Section 17.1.3.7 • Heating Rates Dataset - Section 17.1.3.8 17.1.3.3

Manage Datasets

• Icon: • Command: rcDataset • Menu: Thermal > Post Processing > Manage Datasets • Ribbon: Thermal > Post processing > Manager • Toolbar: Post Processing When this command is issued, the Postprocessing Datasets dialog is opened (Figure 17-7). The dialog has the following features: Current Dataset. The name of the dataset currently loaded into memory. The default name for datasets generated from case sets is the SINDA/FLUINT solution file name. Dataset list. The list of all datasets that have been created. Selected Dataset Description. The Comment field of the dataset selected in the Dataset list.

Postprocessing

17-11

Figure 17-7

Postprocessing Datasets dialog

Add New. Select this button to add a new dataset to the dataset list. When selected, the Postprocessing Dataset Source Selection dialog will open (Figure 17-8). In this dialog, the can name the dataset in the Postprocessing set name field and select the Data Source. Descriptions of the data sources are provided in the following sections: • SINDA/FLUINT Dataset - Section 17.1.3.4 • Text File Dataset - Section 17.1.3.5 • Text Transient Dataset - Section 17.1.3.6 • Radks or Form Factors Dataset - Section 17.1.3.7 • Heating Rates Dataset - Section 17.1.3.8 • SindaWorks XML Dataset - Section 17.1.3.9 • Compare Data Sets - Section 17.1.3.10 • Heat Flux Between Nodes Dataset - Section 17.1.3.11 Set Current. This button places the dataset selected in the dataset list into memory. The current dataset is the dataset displayed when the Display Current Dataset command (Section 17.1.3.1) is issued and the dataset edited using the Edit Current Dataset command (Section 17.1.3.2). Delete. This button deletes the dataset selected in the dataset list. Rename. This button allows the to rename the dataset selected in the dataset list.

17-12

Postprocessing

Figure 17-8

Postprocessing Dataset Source Selection Dialog

Edit. This button opens the Set Dataset Properties dialog. See the section for the dataset type for the description of the dialog. Close. When this button in selected, the current dataset will be displayed in the active layout or the most recently active layout. 17.1.3.4

SINDA/FLUINT Dataset

Add New If SINDA/FLUINT Compressed Solution Results (CSR) Directory or Save File is selected in the Post Processing Dataset Source Selection dialog (Section 17.1.3.3), a SINDA/ FLUINT Results Selection dialog box appears (Figure 17-9). Results. The results file name can be typed into this field, a file in the working directory selected from the drop-down list, or the can select one of the Browse options to search for the Save file or CSR directory.

Postprocessing

17-13

Source Type. The selected the operating system on which the solution file was generated: PC, UNIX, or VAX

Figure 17-9

Dataset Creation - SINDA/FLUINT Results Selection

Edit The Set Sinda Dataset Properties dialog (Figure 17-10) appears after a SINDA/ FLUINT dataset is added or edited (Section 17.1.3.2).

Figure 17-10

17-14

Set Sinda Dataset Properties Dialog Box

Postprocessing

Select a Time/Record. For transient solutions or multiple steady state solutions (like parametric or dynamic solutions), the must select the solution time (left column) or record number (right column) in this list. Nodes. This drop-down list determines the value to be shown for nodes and surfaces. One of the options is Tie Data. This option will display the heat rate of ties on the nodes to which they are attached. Plot Insulation. This drop-down list allows the to choose the default behavior for coloring the model based on the temperature of insulation (Section 4.3.1.6 or Section 4.4.1.7). The setting here can be overridden in each viewport (Section 17.1.2.1). The list includes: • Do Not Plot Insulation Nodes - surfaces with insulation will show the color for the underlying surface and not the insulation temperature • Plot insulation on both sides - surfaces with insulation on either side will show the color for the insulation or the surface if insulation does not exist • Plot insulation on top side - surfaces with insulation on the top side will show the color for the insulation on the top side; and the color for the surface on the bottom side or if insulation does not exist on the top side • Plot insulation on bot side - surfaces with insulation on the bottom side will show the color for the insulation on the bottom side; and the color for the surface on the top side or if insulation does not exist on the bottom side Insulation in postprocessing only includes insulation on thermal desktop primitives and finite elements. Insulation on pipes is controlled with Global Graphics Visibility described in Section 2.7.2. Show Cond/or Temp Diff. When checked, the model is changed to wireframe and conductors and ors are colored based on the temperature difference across them. Environment. When checked, the environment selected in the drop-down list can optionally be displayed in viewports along with the colored model. The viewport option is described in Section 17.1.2.1. Offset Time. The offset time the difference between the solution time and the orbit time. Multiple time options. If more than one time or record are selected, this drop-down allows the following options: • Difference - the difference between the last time/record and the first time/ record selected • ABS(Difference) - the absolute value of the difference between the first and last selected time/records • Maximum - the maximum value for all selected time/records • Minimum - the minimum value for all selected time/records Lumps. This drop-down list specifies the values shown for fluid lumps. Paths. This drop-down list specifies the values shown for fluid paths. Postprocessing

17-15

Ties. This drop-down list specifies the values shown for fluid ties. FTies. This drop-down list specifies the values shown for fluid fTies. IFaces. This drop-down list specifies the values shown for fluid iFaces. Smart Color Bar Cycling. When checked, the Cycle Color Bars command () will switch to the next color bar type (node, lump, path, etc.) and hide all objects except for those related to the newly active color bar. Link Lump To Node Colorbar for Temperatures. When checked, the lump temperature color bar scale will be matched to the node temperature color bar scale. Comment. This field allows input regarding the current dataset. The field is pre-filled with file information, but changing the file name does not change the dataset file. 17.1.3.5

Text File Dataset

If a text file dataset is created, the Text Data File Selection dialog box will appear in which to enter the text file name (Figure 17-11).

Figure 17-11

Text Data File Selection Dialog Box

The format of the file should be in two columns, with the first column listing the name of the entity, followed by a data value. The name should be in submodel.ID format for nodes, submodel.LUMP.ID for lumps, submodel.PATH.ID for paths, and submodel.TIE.ID for ties. The Text Data File Selection dialog box will be displayed, where a descriptive comment may be entered. The may also select to plot the data for an insulation node. A sample file might look like: MAIN.1 10. MAIN.2 20. MAIN.3 40. MAIN.LUMP.2 33. 17.1.3.6

Text Transient Dataset

If a text transient dataset is created, the Text Data File Selection dialog box (Figure 17-11) will appear in which to enter the text file name. A second dialog box will be displayed where the times on the file will be listed and the should select the time to display.

17-16

Postprocessing

The transient text data format is very simple. In summary, it should be the total number of nodes, followed by a list of the nodes, and then a time point, followed by the data for each node at that time point. The data can be one data point per line, or data can exist on the same line if separated by a space. If a node is input without a submodel name, it will be assumed to be in submodel MAIN. A sample text transient file: 3 USU.69 USU.70 USU.71 1. 53.7093946669 53.7093946669 53.2207743327 2. 60. 65. 70. 17.1.3.7

Radks or Form Factors Dataset

If a form factor or radk postprocessing dataset is created, the Directory Select dialog box showing available databases is displayed. The Set FF/Radk Dataset Properties di-

Figure 17-12

Directory Select Dialog Box

alog box, shown in Figure 17-13, will be displayed after the database name is confirmed.

Postprocessing

17-17

Figure 17-13

Set FF (Form Factor) and Radk Dataset Properties Dialog Box

The form factor/radk postprocessing Set FF/Radk Dataset Properties dialog box allows a variety data to be postprocessed. Data for all nodes may be displayed, or data from a single node. The exchange factor to space is area*Fij for form factor datasets and the radk to space for radk datasets. The Bij for radk datasets is the radk divided by the area and the hemispherical emissivity. Bij’s scale from 0 to 1.0 and indicate the fraction of energy that leaves node i and is absorbed by node j. Plotting the Bij value from a given node gives a good feel for the energy transported from a node, since the area of the node is factored out of the color display and results are presented as normalized fractions. Another interesting item to plot is the Bji to a given node. This gives a color plot that shows what it would look if we could see the energy leaving the node being plotted and illuminating the rest of the model. The plot shows the (normalized) energy per unit area absorbed by the receiving nodes. (The energy to node j from node i is given by Area_i * emissivity_i * Bij. Dividing by the area of node j gives the absorbed flux from node i. Since Ai*Bij equals AjBji, Bji (times emissivity_i) gives the absorbed flux from node i. The plot shows the absorbed flux as if node i were a black body. (To imagine illumination, imagine the color bar scale to be reversed, i.e., low absorption equals high reflection)

17-18

Postprocessing

The node for plotting single node data may be entered directly in the Single node data Node field, or the Select button can be clicked to bring up a tree based node name browser. 17.1.3.8

Heating Rates Dataset

If a heating rate dataset is created, the Directory Select dialog box (Figure 17-12) showing the available databases will be presented. The Set HR Dataset Properties dialog box, shown in Figure 17-14, will be displayed after the database name is confirmed.

Figure 17-14

Heating Rate Postprocessing Dataset Dialog Box

The orbit position to plot may be selected, along with the heating rate source components. More than one source may be selected and data will be summed for the output. The following items may be plotted: • Total Absorbed - total power absorbed from selected source(s) • Direct Incident - power directly incident from selected source(s) • Number of Rays - number of rays shot for each node for the selected source(s) • Time Average Total Absorbed - the total absorbed averaged with time over the entire orbit for the selected source(s) • Direct Absorbed - power absorbed directly from the selected source(s) Postprocessing

17-19

• Reflected Absorbed - power absorbed from the selected source(s) after reflecting off of another surface • Direct Error - the estimated error of the direct illumination • Absorbed Error - the estimated total error including direct and reflected energy Dividing by area will plot fluxes. 17.1.3.9

SindaWorks XML Dataset

[Obsolete] 17.1.3.10 Compare Data Sets Comparing datasets allows selecting two datasets for comparison. The first dataset in the list is subtracted from the second dataset in the list. Each of the datasets can be edited to change the information that is compared (e.g. - node Q instead of node T). The resulting comparison will make the most sense when the datasets have the same information. This option will compare whatever is included in the datasets. This means that if two times have been selected in each dataset for a difference, then the results of the comparison will be the difference of the difference. 17.1.3.11 Heat Flux Between Nodes Dataset A heat flux between nodes dataset allows showing color contours for heat flow from one set of nodes to another. Conductors must be included in the dataset. When this dataset is created, a Save file or CSR directory is selected. The the Set Heat Flux Dataset Properties dialog opens (Figure 17-15). The fields in this dialog are: Select a Time/Record. Choose the time or record number to display. From Nodes/Submodels/ALL/*[] enabled. Enter node IDs or Submodels for nodes in the first set. See Section 2.10.13 for allowable formats. To Nodes/Submodels/Lumps/ALL/*[] enabled. Enter node IDs, lump IDs, or Submodels for nodes in the second set. See Section 2.10.13 for allowable formats. Linear Conductors. When checked, the calculations will include heat flow through linear conductors. Radiation Conductors. When checked, the calculations will include heat flow through radiation conductors. Fluid Ties. When checked, the calculations will include heat flow through fluid ties. Divide by area. When selected, the must include a text file providing areas for the nodes in the From list. Area files are automatically created with radk calculations.

17-20

Postprocessing

Insulation Nodes. This drop-down list determines if insulation nodes will be used in the calculations. Show From Node Data. When checked, the heat flow of the From nodes will be displayed with the selected sign convention. Show To Node/Lump Data. When checked, the heat flow of the To nodes and lumps will be displayed with the selected sign convention.

Figure 17-15

17.1.4

Set Heat Flux Dataset Properties dialog

Postprocessing Through Time

The Thermal > Post Processing > Next Time function will change the data to the next time in the postprocessing dataset. This allows the to march around a defined orbit, or to step through a SINDA results. The Thermal > Post Processing > Next Time and Position command will also change the data to the next time, but will also change the orientation of the geometry if articulation has been specified. The Previous Time commands act the same as the Next Time commands, except the time is stepped backwards.

Postprocessing

17-21

17.1.5

Animate Through Time

The Thermal > Post Processing > Animate Through Time function allows the program to postprocess from the selected Start Time to the selected End Time for the current postprocessing dataset. The Continuous Cycle Dialog dialog box, shown in Figure 1716, allows the to specify either a number of cycles or a continuous cycle. A delay may be input, in case the program is cycling through too fast.

Figure 17-16

Continuous Cycle Dialog Box - Animate Through Time

The final option on the dialog box is the Movie Option. The two options for creating movies are Create AVI file and Send F4 After Update. The first option creates an AVI file from the animation. The Send F4 After Update instructs the program send an function key after each frame is displayed. Using this option in conjunction with a tool like HyperCam allows for the easy creation of movies. The steps to create the movie with HyperCam are: 1. 2. 3. 4. 5. 6. 7.

Have AutoCAD up and running. Start up HyperCam. From HyperCam, select the AutoCAD region or window to be captured. From HyperCam, Select Start Paused. From AutoCAD, Thermal > Post Processing > Animate through time. Select Send F4 After Each Update and OK. Select Stop Recording from HyperCam.

When the command is run, the max and min color bar values will be printed when the command ends. The color bar may be modified and auto-scaling turned off, setting the min and max values to the proper values, and the movie remade. This will keep the color bar constant. 17-22

Postprocessing

Want "Hands-On" Information? Tutorial exercise "Beer Can Example" on page 20-89 uses this command. 17.1.6

Saving and Printing Pictures

There are many different ways of obtaining a file of the image. The use of some third party screen capture software is recommended. Two such software applications are HyperSnap (www.hyperionics.com) and LView. HyperSnap will allow the to select the window or a region of the screen and then save it to a gif, tiff, or other formats. The Thermal > Utilities > Capture Graphics Area command will save the graphics area to a bitmap file titled ScreenCapture#.bmp in the current directory. The program will determine the lowest # that it can use to make the picture so as to not overwrite any existing file. When this command is issued, the bitmap is also pasted into the system clipboard, so Microsoft Word or Powerpoint could be opened and the pictured pasted directly into one of those applications. Another function that the may find useful is the function (hold down the and keys) that will copy the active window into the cut/paste buffer. The may then open Word or some such program and paste a copy of the window into it. And finally, as a caution, the AutoCAD Export and Tools > Display Image command will not capture the color bar for a postprocessed picture. 17.1.7

Cutting Planes

A cutting plane can allow visualization of results within a solid object, finite difference or finite element. 17.1.7.1

Cutting Plane Creation

Cutting planes are created by selecting Thermal > Post Processing > Cutting Plane. Once the command has been issued, the selects a rectangular area in manner similar to creating an FD rectangle. The values requires in the prompts are: • the origin • a point for +X and X size • a point to set the XY plane and Y size. 17.1.7.2

Cutting Plane Editing

After the cutting plane has been defined or the cutting plane has been selected and edited, the Cutting Plane form appears (Figure 17-17). The fields on the form are:

Postprocessing

17-23

Characteristic Length. The value determines the resolution of the results (typically temperature) mapping: a smaller number will produce finer mapping. Decrease the default value only if the cutting plane contours appear coarser than expected as a smaller value will take longer to update. Solid Tag Set. A cutting plane will map results for all solid objects through which its primary location (not the translated display, see Section 17.1.7.3) cuts. The drop-down allows selecting a Domain Tag Set (Section 2.5) of solid objects for mapping onto the Cutting Plane. Selecting a Domain Tag Set will map only the results from the solid objects included in the selected Domain Tag Set. Draw Unmapped Areas. The check box determines if the entire cutting plane rectangle is drawn (checked) or if only the portion that has mapped results is drawn (unchecked). The mapped areas will display temperature contours; any region of the cutting plane rectangle that extends beyond the boundaries of the objects being mapped will be displayed as gray if the unmapped area is drawn. OK & Map. Closes the Cutting Plane dialog and maps the results onto the cutting plane. Once the Cutting Plane is created, the cutting plane dialog can be accessed by editing the cutting plane (RcEdit). The Cutting Plane size and location can be changed by selecting the cutting plane and using the grip points to adjust the cutting plane.

Figure 17-17

17.1.7.3

Cutting Plane dialog

Cutting Plane Grip Points

Cutting planes have nine grip points: • Move Origin - translate the cutting plane to a new location keeping the same orientation

17-24

Postprocessing

• Stretch X Length - stretch the cutting plane in its own X direction, as defined when it was created • Aim X rotating about Y - aim the cutting plane’s X axis by rotating about the cutting planes’s Y axis • Aim X rotating about Z - aim the cutting plane’s X axis by rotating about the cutting planes’s Z axis (normal to the cutting plane) • Stretch Y Length - stretch the cutting plane in its own Y direction, as defined when it was created • Aim Y rotating about X - aim the cutting plane’s Y axis by rotating about the cutting planes’s X axis • Aim Y rotating about Z • Aim Z axis • Translate color contour display - shift the location of the color contour display away from the object being mapped (Figure 17-18) The Cutting Plane size, location, orientation and display location can be adjusted by selecting the Cutting Plane, clicking on a grip point and selecting a new location for the grip point.

Figure 17-18 Cutting Plane display location offset using Translate color contour display grip point.

Postprocessing

17-25

17.2 17.2.1

X-Y Plotting Plot Data versus Time

The Thermal > Post Processing > X-Y Plot Data vs. Time command plots transient, or parametric, results in the EZXY Plotter. Performing X-Y plot commands will open EZXY Plotter and prompt for objects to be plotted, unless items were preselected. The plots have many options and export features that can be explored using the on-line help provided on the EZXY Plotter forms. EZXY Plotter will plot whatever the currently selected dataset for postprocessing contains. Temperatures, heating rates, or even radiation conductors vs. time for articulating geometry may be plotted. The X-Y Plotter also accepts a generic ASCII format for making plots of data calculated by written programs or SINDA/FLUINT logic. Parametric steady-state results will be plotted as a function of record number since the time value will always be zero. The X axis values can be changed to any value stored in the results through the EZXY Plotter interface. The X-Y plot versus time command is also available through the Model Browser (Section 2.4) as either a menu selection or an icon. Want "Hands-On" Information? The two tutorial exercises "Beer Can Example" on page 20-89, and "Manifolded Coldplate" on page 22-37 use this command. 17.2.2

Plot Pipe Data versus Length

The Thermal > Post Processing > X-Y Plot Pipe Data vs. Length command plots currently postprocessed lump information as a function of pipe length in a temporary X-Y plot window. After issuing the command, the will be prompted to select pipes. Plot options are available through a right-click contextual menu.

17.3

Query Node

The Thermal > Post Processing > Query Node command will print out the data values of the nodes that are selected by the . These values are printed to the text output window. Another way to see these value is to use the model browser (see “Model Browser” on page 2-8).

17-26

Postprocessing

17.4

Results Queries

Several commands can be used to query existing save (*.sav) files. These commands operate only on existing results (CSR Directories or Save files) and do not require the model geometry. Therefore, it is possible to use these commands in a blank drawing file and simply point to the desired results. Most of these commands generate ASCII files, but use XLS extensions. The files can be opened in text editors. Some text editors, however, will not open the files since they attempt to read the file as XLS format due to the extension. Note: Results only contain the data selected on the right side of the Output tab of the Case Set information form (Section 15.2.3) or in explicit logic statements written by the . If the data required by the following results query commands are not in the results, then the commands will not provide the desired information. Note: The data in the results is limited to the output interval specified in the Output tab of the Case Set information form (Section 15.2.3) or in explicit logic statements written by the . Therefore, transient events between output intervals will not be captured by these commands. While smaller output intervals will generate larger files, -written logic may help capture transient events of interest while limiting the size or the results. 17.4.1

Find Results Min Max

The Thermal > Post Processing > Find Results Min Max command searches all selected results and creates three comma-separated-variable ASCII files (with XLS extension). The first file, ‘SubmodelIdMinMin.xls’, lists each node in the results and for each node shows: the minimum and maximum temperature; the times at which minimum and maximum temperatures occur; the results in which the minimum and maximum temperatures occur; the minimum and maximum time-integrated temperatures; and the results corresponding to the minimum and maximum integrated temperatures. The second file, ‘SubmodelMinMax.xls’, lists each submodel in the results and for each submodel shows: the minimum and maximum temperature; the times at which minimum and maximum temperatures occur; the node IDs of corresponding to the minimum and maximum temperatures; the results in which the minimum and maximum temperatures occur; the minimum and maximum time-integrated temperatures; The node IDs corresponding to the minimum and maximum integrated temperatures; and the results corresponding to the minimum and maximum integrated temperatures.

Postprocessing

17-27

The third file, ‘MinMax.xls’, lists each in the results and for each shows: the minimum and maximum value; the times at which minimum and maximum values occur; the results in which the minimum and maximum values occur; the minimum and maximum time-integrated values; and the results corresponding to the minimum and maximum integrated values. Temperatures and s must be output to the results. 17.4.2

TSINK from Results

The Thermal > Post Processing > TSINK from Results (rcTsink) command performs the same function as the TSINK and TSINK1 commands in SINDA/FLUINT. For information regarding the calculations made, see the SINDA/FLUINT s Manual. TSINK is useful for detailed modeling. For example, if a subsystem must be modeled with more detail than provided in the system-level model, then the subsystem would have sink temperatures and conductors calculated from the system-level model’s results. In the subsystem model the sink temperatures and conductors will act as boundary conditions and replicate the behavior of the overall system. Issuing this command opens a TSINK Manager that lists TSINK calculators, groups of results and TSINK sets. TSINK sets are the nodes for which sink temperatures will be calculated. Adding or editing a calculator will open a TSINK Calculator form. The options in the TSINK calculator form are: • Name - the name of the calculator • Add Files - add results from the current directory or the parent directory • Delete Selected - delete selected results from the list • Add - add dataset • Copy - copy datasets • Delete - delete datasets • Edit - edit datasets • Output File Prepend Designator - text string to be added into output file names • Generate conductor Output - checking this option generates conductors from each node in the dataset to its sink node; leaving unchecked does not generate conductors (not recommended) • Add Dynamic Radks named SaveFileName*.k,kab - dynamic conductors are not stored in the results: checking this option will include dynamic radks (radiation conductors) in the calculations; other dynamic conductors will not be included in the calculations Adding or editing a dataset opens the TSINK Options form. The calculations are based on the information on this form.

17-28

Postprocessing

• TSINK nodes - nodes in submodel.ID format for which sink temperatures will be calculated; with one item in the list the calculation will behave like the TSINK1 routine in SINDA. • Excluded Nodes/Submodels - nodes and submodels which will be excluded from the sink temperature calculations; equivalent to the Keeplist in the SINDA TSINK routines. Multiple nodes can be specified using wildcards (Section 2.10.13). For entire submodels, only the submodel name can be given. This list can be blank; this is equivalent to providing ‘NONE’ in the keeplist argument of TSINK or TSINK1. • Include Applied Q - when checked, nodal heat loads in the results will be included in the sink temperature calculations; equivalent to Qm = 1 in SINDA TSINK routines • Generate Single Sink Node - when checked only one sink temperature will be calculated; when unchecked each node in the TSINK Nodes list will have its own sink temperature and conductor (unchecked will provide the more accurate response); this is equivalent to setting the mode of the TSINK routine to ‘1’. • Ignore Connections of the TSINK Nodes List - when checked the nodes in the TSINK Nodes list are excluded from the sink temperature calculations. • Type of conductors - choosing Radiation will create TSINKs and corresponding radiation conductors for all radiation connections in the results (equivalent to mode = R in the SINDA TSINK routines); choosing Linear will create TSINKs and corresponding linear conductors for all linear connections in the results (equivalent to mode = L in the SINDA TSINK routines); choosing Both Radiation and Linear will create TSINKs and corresponding linearized conductors for all connections in the results (equivalent to mode = both in the SINDA TSINK routines); as in the SINDA TSINK routines, a better subsystem model would be created using a TSINK calculation for each linear and radiation than a single TSINK with both radiation and linear since the latter is linearized When the calculations are performed, a set of ASCII files with XLS extensions are created. For each results set in the TSINK Calculator Save Files list a file named results*Tsink.xls where results is the name of the results (Save file or CSR directory) and * is the Output File Prepend Designator, if any. This file contains the time and the conductance and sink temperature for each node in the TSINK Nodes list. Three more files are also generated: TsinkSubmodelIdMinMax.xls, TsinkSubmodelIdMinMaxDelta.xls and TsinkIntegratedValuesSubmodelIdMinMax.xls. If an Output File Prepend Designator is given, each of these files will be preceded by that string. The first file contains the TSINK nodes in the first column followed by the minimum sink temperature for that node, the maximum sink temperature for that node, the time of minimum sink temperature, the time of maximum sink temperature, the results containing the minimum sink temperature, and the file containing the maximum sink temperature. The second file contains the TSINK nodes in the first column followed by the minimum range of the sink

Postprocessing

17-29

temperature within the results, the maximum range of the sink temperature within the results, the file containing the minimum range, and the fie containing the maximum range. The third file contains the TSINK nodes, the minimum and maximum time integrated sink temperatures, and the results corresponding to those averages. Temperatures, applied heat loads, and conductances must be written to the results. 17.4.3

QFLOW from Results

The Thermal > Post Processing > QFLOW from Results command provides the heat flow between two sets of nodes. Issuing this command opens a QFLOW Manager that lists QFLOW calculators, groups of results and QFLOW sets. QFLOW sets are the node group pairs for which heat flows will be calculated. Adding or editing a calculator will open a QFLOW Calculator form. The options on the QFLOW Calculator form are the same as those of the TSINK Calculator form in the previous section. Adding or editing a dataset opens the QFLOW Attributes form. In each field, the lists nodes (in submodel.id form) or submodels to be added to each group. Multiple nodes can be specified using wildcards (Section 2.10.13). In the To list, ‘ALL’ can be used to refer to all submodels. For entire submodels, only the submodel name should be given (do not use a wildcard like MAIN.*) When the QFLOW is calculated an ASCII file, with XLS extension, for each results set in the QFLOW calculator is created in the working directory. The file name will be resultsQflow.xls where results is the name of the results set (Save file or CSR directory). In each file, the columns will be: • Time • Qtot - total heat flow • Qlin - heat flow by linear conductors • Qrad- heat flow by radiation conductors The heat flow columns will be repeated for each QFLOW set requested in the input order. Temperatures and conductances must be written to the results. 17.4.4

Analyze Heaters from Results

The Thermal > Post Processing > Analyze Heaters from Results command searches the selected results for heater information. If heaters exist in the results, a single ASCII file, with XLS extension, is created: HeaterSummary.xls. The columns in the file are: • Name - the heater name includes the logic submodel; the append string, if provided; and the internally generated alpha-numeric label. • Max Temp Delta - the maximum temperature range among the heated nodes at any output time

17-30

Postprocessing

• Max Total Power (TP*) - the maximum time-integrated power (energy) for a solution • Max Instantaneous Power (PO*) - the maximum applied power • Peak Density - the Max Instantaneous Power divided by the applied surface area • Max Temp - the maximum temperature of the heated nodes • Min Temp - the minimum temperature of the heated nodes • Max Delta Max Node - the node providing the higher temperature for the Max Temp Delta • Max Delta Min Node - the node providing the lower temperature for the Max Temp Delta • Max Delta Time - the time of the Max Temp Delta • Max Delta File - the file in which the Max Temp Delta was found • Max Total Power File - the file in which the Max Total Power was found • Max Instantaneous Time - the first time at which the Max Instantaneous Power was reached • Max Instantaneous Power File - the file in which the Max Instantaneous Power was found • Max Node Name - the node providing the Max Temp • Max Time - the time at which the Max Temp was reached • Max Save - the file in which the Max Temp was found • Min Node - the node providing the Min Temp • Min Time - the time at which the Min Temp was reached • Min Save - the file in which the Min Temp was found Temperatures and s must be written to the results. 17.4.5

Write Results Data to Text

The Thermal > Post Processing > Write Results Data to Text command will write the requested data to an ASCII file with XLS extension. When the command is issued, the Save to Text Manager is opened (Figure 17-19). Here the can set up multiple save sets (defined tasks) for concurrent operations. In the Save to Text Manager, the can select the desired defined tasks and select Calculate to output the requested information for all selected save sets.

Postprocessing

17-31

Figure 17-19 have title)

Save to Text Manager (left - save set has title; right - save set does not

A save set consists of an optional name, a set of results, a list of nodes (Section 2.10.13) or s, a file prepend designator and a selection of the node data to save. The name is used in the Save Text Manager. The prepend designator is added to the front of the output file names which are otherwise the node ID (in the form submodel_ID#) or name with the XLS extension. Once save sets are created, the files can be written by selecting the calculate button on the Save to Text form or in the Save to Text Manager. In the Save to Text Manager, the desired save sets must be selected. The requested information (s, temperature, capacitance, or heat load) must be written to the results.

17-32

Postprocessing

18

Data Exchange

Thermal Desktop provides the capability to import thermal geometric models, finite element models, and circuit board design models to allow reuse of data already created in other applications. The may also write out node data to a text file, “map” data to an external model or set of location points, or export a thermal finite difference or finite element model for use in other applications. Mapping allows data from the Thermal Desktop model to be appropriately distributed on a geometrically similar external model, even if the nodes and elements are not aligned. The Autodesk interface (AutoCAD) also provides some geometry import capabilities. Importing models or geometry can be a great way to take advantage of work already completed in other tools. However, directly importing a model intended for a purpose other than thermal analysis can be costly: both in model maintenance and solution time and convergence. The sections below discuss how these import operations are performed. Please refer the Thermal Desktop Advanced Modeling Techniques ’s Guide (go to Start > Programs > Thermal Desktop > ’s Manual - Meshing) for guidance on importing models or geometry from other applications and the recommended preparation of the models or geometry before they are imported.

18.1

Import Geometry Note: Often, raw CAD geometry is not suitable for meshing and is often too complex even for referencing as scaffolding for Thermal Desktop surfaces and solids. Defeaturing and simplifying the geometry as much as possible using the native CAD tool or SpaceClaim can improve the utility of the imported geometry. See the Thermal Desktop Advanced Modeling Guide (Windows:Start > Programs > Thermal Desktop > s Manual Meshing) for more information.

Starting with AutoCAD 2012, many geometry formats can be imported. Some of the formats available for import though Menu:File > Import or by typing the IMPORT command: • 3D Studio (*.3ds) • ACIS (*.sat) • CATIA V4 (*.model; *.session; *.emp; *.dlv3) • CATIA V5 (*.CATPart; *.CATProduct) • IGES (*.igs; *.iges)

Data Exchange

18-1

• JT (*.jt) • NX (*.prt) • Parasolid binary and text(*.x_b; *.x_t) • Pro/ENGINEER (*.prt*; *.asm*) • Pro/ENGINEER Granite (*.g) • Pro/ENGINEER Neutral (*.neu*) • Rhino (*.3dm) • SolidWorks (*.prt; *.sldprt; *.asm; *.sldasm) • STEP (*.ste; *.stp; *.step) The imported geometry will typically be imported as a block. If the imported geometry is to be meshed using TD Mesher, the block will need to be exploded, first. See Import Other File Formats in the AutoCAD help for more information. If AutoCAD 2012 or later is not available the following sections describe how to import ACIS, IGES and STEP files using older versions of AutoCAD. 18.1.1

ACIS

AutoCAD is capable of importing geometry objects stored in *.sat (ASCII) files. The files must be in ACIS version 7. ACIS files can be imported using one of three methods: •

Command: ACISIN

•

Menu (option 1): File > Import (AutoCAD 2012 or later)

•

Menu (option 2): Insert > ACIS

The import command converts the model to a body object or to 3D solids and regions if the body is a true solid or a true region. Once imported, the geometry may be used to snap Thermal Desktop objects to the correct dimensions or may be meshed using TD Mesher as described in the Thermal Desktop Advanced Modeling Guide in a separate volume. 18.1.2

IGES

18.1.2.1

AutoCAD 2011 and Earlier

Importing IGES information is significantly different than all the other Import and Export options. The IGES translator is written by Autodesk and is available through AutoCAD Mechanical (bundled with Autodesk Inventor Suite). s who purchase CAD functionality along with Thermal Desktop have access to the IGES translator. The IGES translator is accessed typing the command IGESIN in the command line.

18-2

Data Exchange

IGES import functionality only imports geometrical information such as lines, points, arcs, circles, etc. This information may then be used to construct Thermal Desktop surfaces, solids, and elements using these entities and the key points on these entities. IGES import is a way to get CAD data from another CAD system such as PRO/ENGINEER, I-deas, CATIA, UNIGRAPHICS and others. The must EXPORT an IGES file from those CAD systems. The export command on most CAD systems will contain options for the . Generally, it is best to try several different options to see what works best for the system being used once it is imported into AutoCAD. Some of the options may produce so much information, that the system may not be able to handle it in of memory and graphics limitations. When exporting from I-deas, some s have determined that it is best to set the Surface Types to Export option to none. This option will export just the wireframe geometry which can easily be used as reference for building Thermal Desktop surfaces. For more information on IGES translation, select the Help button on the IGESIN translation form. 18.1.2.2

AutoCAD 2012 and higher

Starting with AutoCAD 2012, IGES geometry can be imported using •

Menu:File > Import

•

Command: IMPORT

The imported geometry will typically be imported as a block. If the imported geometry is to be meshed using TD Mesher, the block will need to be exploded, first. See Import Other File Formats in the AutoCAD help for more information. 18.1.3

STEP

STEP import is a way to get CAD data from another CAD system such as PRO/ENGINEER, I-deas, CATIA, UNIGRAPHICS and others. The must EXPORT a STEP file from those CAD systems. The export command on most CAD systems will contain options for the . Generally, it is best to try several different options to see what works best for the system being used once it is imported into AutoCAD. Some of the options may produce so much information, that the system may not be able to handle it in of memory and graphics limitations. 18.1.3.1

AutoCAD 2011 and Earlier

Importing of STEP information is significantly different than all the other Import and Export options. The STEP translator is written by Autodesk and is available through AutoCAD Mechanical (bundled with Autodesk Inventor Suite). s who purchase CAD functionality along with Thermal Desktop have access to the STEP translator. The STEP translator is accessed typing the command STEPIN in the command line.

Data Exchange

18-3

The AutoCAD STEP Translator enables data exchange using STEP AP203, which is called Configuration Control of 3D Design Data. For the purposes of Thermal Desktop, AP203 provides B-rep solids, surfaces and wire frames. For more information on STEP translation, select the Help button on the STEPIN translation form. 18.1.3.2

AutoCAD 2012 and higher

Starting with AutoCAD 2012, STEP geometry can be imported using •

Menu:File > Import

•

Command: IMPORT

The imported geometry will typically be imported as a block. If the imported geometry is to be meshed using TD Mesher, the block will need to be exploded, first. See Import Other File Formats in the AutoCAD help for more information. 18.1.4

AutoCAD Block Import

AutoCAD DWG files, including those created from Thermal Desktop, may be inserted into other drawing files using the Insert > Block menu command. The resulting dialog box provides the means to specify the insertion point and define scaling factors for the file, or block, being inserted. Once the block is inserted, typing EXPLODE, or selecting Modify > Explode, will separate the block into its individual entities. The may also choose to explode the block from the Insert Block dialog box. Want "Hands-On" Information? Get some experience importing blocks in Section 20.10 "Parameterizing for a Common Input" on page 20-187.

18.2

Import Models

When models are imported, the type of model being imported will determine the information that comes in to Thermal Desktop with the model. As a general rule, the external model and Thermal Desktop should be in the same system of units. Also, if the model has boundary conditions applied, that they have been imported: not all import options will import boundary conditions. Sometimes finite element models are imported with duplicate elements, multiple elements sharing the same set of nodes or grid points. If this happens, Thermal Desktop commands, such as Surface Coat Free Solid Faces, may not work properly. It is recommended that the run the command rcCheckElements after importing a finite element model. If duplicate elements exist, Thermal Desktop creates two AutoCAD groups, DUPELEMS and DUPELEMS2. The DUPELEMS group contains the elements considered to be dupli-

18-4

Data Exchange

cates. The DUPELEMS2 group contains elements considered to be the original elements. If more than two elements share the same set of nodes, then the DUPELEMS group will contain all but one of those duplicates and DUPELEMS2 will still contain only one of the duplicates. Therefore, all elements in DUPELEMS should be deleted. The rcCheckElements command is run before running a case. If imported finite elements appear to be degenerate (multiple nodes in the same location, for example), those elements will be placed in a group named BADELEMS. Using the Model Browser (see Section 2.4 "Model Browser" on page 2-8) to display only that particular group will allow the to find the bad element and correct it or delete it. Thermal Desktop is limited to linear finite elements. Some non-linear elements are imported, but will be converted to a similar linear element. If a specific type of element does not import into Thermal Desktop, please send a sample model with that element type to [email protected] so we can add it to the element types we import. 18.2.1

Thermal Desktop

The Insert > Block command is used to import one DWG file into another, as in the case of merging two models. The imported DWG file can be AutoCAD geometry, an entire Thermal Desktop model or a subset of a Thermal Desktop model created with the WBLOCK command (Section 18.4.1). All layers and objects in the inserted DWG file are brought into the new file. To access the radiation analysis groups and submodels, the Scan DB button should be selected on the respective forms (Section 4.1 and Section 4.2, respectively). Case sets, orbits, logic objects and thermal submodels with comments can be subsequently imported using the methods described in Section 2.10.12. Properties must be imported to the local property database using the import function on the optical and thermophysical property forms (Section 3.1.1 and Section 3.2.2, respectively). Be sure the Thermal Desktop units (Section 2.7.1) are the same in both models before inserting the block. The objects in the inserted DWG file are brought in as a block (a collection of objects and their properties). Once the block is placed in the appropriate location, the Modify > Explode command can be used to convert the block into individual elements. If articulators were used in the original model, the must issue the Thermal > Articulators > Toggle Global Activation command before issuing the Modify > Explode command. After the explode command the should issue the Toggle Global Activation command again to re-activate the articulators. 18.2.2

Thermal Geometric Radiation Models

Thermal Desktop can import geometric models from legacy radiation analysis codes. In most cases, these codes do not the generation of conduction and capacitance data. However, the Thermal Desktop surfaces that are created can be edited to add the additional information if desired. In some formats units are recognized, but it is good practice to set Thermal Desktop units to match the file to be imported before the import operation.

Data Exchange

18-5

18.2.2.1

TRASYS

TRASYS models and correspondence are imported into Thermal Desktop by selecting Thermal > Import > TRASYS. This command displays the TRASYS Import Options dialog box shown in Figure 18-1.

Figure 18-1 TRASYS Import Options Dialog Box

The TRASYS Import Options dialog box is used to select the input file. The can also select the units of the TRASYS model on this dialog box. When OK is selected the TRASYS preprocessor included with Thermal Desktop is run and the TRASYS model is displayed in the graphics viewport. The imported model is placed in the current default analysis group. The model is ready for radiation exchange, view factor calculations, and heating rate calculations. TRASYS cone, cylinder, disk, rectangle, sphere, paraboloid, torus, and ogive are created directly as custom Thermal Desktop surfaces. Only surface data and MODPR subroutine calls are imported, orbit data as well as other information in the Operations block is ignored. The conversion process also converts the modify properties, MODPR, subroutine calls that may be placed in the operations data block. Successful conversion of these calls is output to the screen. Each TRASYS BCS is also placed on a unique layer using the BCS name prefixed by BCS_. AutoCAD® allows the visibility of each layer to be controlled so that individual or combinations of BCS’s may be viewed. The Freeze layer option is used to change the visibility of the layer. The Freeze column uses two symbols - a sun icon when layer visibility is On and a snowflake icon when layer visibility is Off, or frozen. Using the Freeze column to turn a layer’s visibility On or Off provides the with the fastest system performance.

18-6

Data Exchange

Want "Hands-On" Information? Import a TRASYS file in the tutorial exercise “Importing a TRASYS Model and Using Articulators” on page 21-35. 18.2.2.2

TSS

The command Thermal > Import > TSS imports TSS models to Thermal Desktop. The dialog box associated with this command is shown in Figure 18-2.

Figure 18-2

TSS Import Dialog Box

The TSS assembly structure is preserved using Thermal Desktop Assemblies. The coordinate systems of the assemblies are placed on the layer TSS_ASSEMBLIES. Thaw this layer and turn it on in order to select the assembly coordinate systems for modification. The Thermal Desktop Model Browser (see “Model Browser” on page 2-8) may be used to perform operations typically performed using the TSS assembly tree. Select List > Articulators from the Model Browser menu to update the browser tree with the assembly hierarchy. Select Option > Always Trace Children so that the visibility of entire assemblies may be turned on and off by selecting the assembly. Thermal Desktop does not recognize TSS boolean operations. If the TSS model has parts created using boolean operations, then those parts must be removed from the TSS model, the model imported into Thermal Desktop and removed parts rebuilt in Thermal Desktop. 18.2.2.3

NEVADA

Thermal Desktop will import Nevada input files by selecting the Thermal > Import > Nevada command. Certain caveats apply to this operation. Only those surface types ed by Thermal Desktop are imported (i.e., hyperboloid and mask nodes are not imported). Nevada cones, cylinders, circular ellipses (disks), spheres, and paraboloids are created directly as custom Thermal Desktop surfaces. Quads, triangles and boxes are created as meshes.

Data Exchange

18-7

After the input is selected and the translator run, the imported model is placed in the analysis group named “Nevada”. The surfaces are created in a “Nevada” submodel and the surface properties are placed in the file Nevada.rco. The model is ready for radiation exchange, view factor calculations, and heating rate calculations. As they are read, the property numbers are made into property names. These names are Nev-## where ## is the original property number. This name appears as the property name in the property database as well as in the surface property reference. 18.2.2.4

STEP TAS 5.2

Thermal Desktop will import files written to the STEP-TAS, version 5.2, standard. Thermal Desktop surfaces and optical properties will be imported. 18.2.3

Finite Element Model

Thermal Desktop can import finite element models from various formats (and an FD equivalent representation from I-Deas). FE models may be imported under the control of a persistent Mesh Importer entity or as independent nodes and elements. Using the Mesh Importer is highly recommended. This entity exists as a graphical object in the drawing and may be used to manipulate the entire imported FE model. 18.2.3.1

Import FE models with a Mesh Importer

The Mesh Importer entity manages the nodes and elements that are created for the thermal model. The Mesh Importer works with NASTRAN, ANSYS, FEMAP, and I-deas meshes, as well as meshes generated by TD Direct1. As many Mesh Importers may be created as desired, each managing the import of an FE model contained in an external file. It allows a reimport of the external model with a simple button click and automatically manages the removal of the existing model and the creation of the new model. If groups are defined in the external FE model, connections to the mesh by other parts of the thermal model can be retained upon reimport by the use of Tag Sets. For more information about creating and using Tag Sets please see Section 2.5. The Mesh Importer allows the individual nodes and elements to be hidden, and the representation displayed by an efficient single entity. Efficient postprocessing is also managed by the Mesh Importer entity. The entire model may also be moved and rotated using the Mesh Importer. A new Mesh Importer is created using the text command TdFEMeshImporter or the menu command Thermal->Import->Create FE Mesh Importer as shown in Figure 183. The form shown in Figure 18-4 will appear. This dialog allows the specification of the format of the input file, then name of the input file, and options controlling the import operation. 1 Using the FE Mesh Importer to import a TD Direct Mesh eliminates the need for a SpaceClaim or TD Direct license, but does not provide the ability to communicate with TD Direct.

18-8

Data Exchange

Figure 18-3

Menu choice to create a new Mesh Importer

The Format drop down menu allows the selection of either NASTRAN, FEMAP, ANSYS, or IDEAS import. Depending on the choice of input format, the Input File combo box will show available files in the current working directory and subdirectories of the current working directory. A file path may be directly typed into this field, or the “Browse...” choice, as shown in Figure 18-5, may be selected to bring up a standard file selection dialog. The external FE model may imported as a thermal model ready for analysis, or as just a graphical representation. If the external format provides boundary conditions, the check box will control whether or not those boundary conditions are imported into Thermal Desktop. The graphical only representation may be useful for importing a structural model, perhaps to serve as the basis for constructing a simpler thermal model. The model may be imported relative to the World Coordinate System, or to the currently placed Coordinate System. The Mesh Importer also allows the mesh to be moved or rotated as a unit after it has been imported. Some formats provide for the specification of units, but it is good practice to set Thermal Desktop to the same units as used in the input file.

Data Exchange

18-9

Figure 18-4

Figure 18-5

Mesh Importer creation dialog

Using the file browser feature of the Input File combo box

When OK is selected, a Mesh Importer will be created, along with nodes and elements. The Mesh Importer provides methods to more efficiently visualize the mesh and post processed results for the mesh. If desired, the Mesh Importer provides methods to easily reimport or reposition the mesh.

18-10

Data Exchange

A name may be given to the mesh using the Set Label... button. This label is displayed in the graphics window as text attached to the far right of the mesh. The label will also appear in the Model Browser. The default label is “FEMESH”, but may be changed to any text string. The label is also used in the layer names to which the Mesh Importer and the nodes and elements are assigned. The layer names are automatically updated when the Mesh Importer label is changed. The visibility of layers is controlled in the same manner as a Mesher entity. The Mesh Importer functions very similar to a Mesher object, the difference being that the mesh is specified in an external file rather than generated from native geometry. Please see the Thermal Desktop Advanced Modeling Techniques ’s Guide (in a separate volume) for more information on controlling the visibility of the displayer, nodes, 2D elements, and 3D elements associated with the imported mesh. The Mesh Importer has all the features of a Mesh Displayer, and are accessed using the Display Preferences... button. The mesh may be represented as wireframe or solid, with or without internal elements hidden, and also as an outline view to provide a very efficient graphical representation (see Section 2.10.15). When a FE model is imported, regular Thermal Desktop nodes, surface elements and solid elements are created along with the Mesh Importer entity. The Mesh Importer manages the nodes and elements. Individual nodes that are managed by a Mesh Importer may not be deleted or repositioned. A useful analogy is how nodes on a finite difference surface are managed. When, say a cylinder, is created, Thermal Desktop nodes are displayed on the surface. Those nodes may not be deleted or repositioned independently of the surface, they are controlled by the surface that owns them. Likewise, the Mesh Importer controls the lifetime and position of the nodes that it manages. Nodes may be edited and properties, submodel, and ID number changed in the same way as nodes managed by a Thermal Desktop surface. Any transformation done on a Mesh Importer will be done on all the nodes and elements that it controls. The Mesh Importer may be scaled, rotated, or moved and the entire imported FE mesh will be transformed in the same manner. The Mesh Importer may be attached to an assembly or a tracker. Select the Mesh Importer to be attached to an assembly or tracker, not the collection of nodes that it controls. The format and name of the input file is retained by the Mesh Importer. If the external file has been changed, Thermal Desktop can be updated with the new FE model by clicking the Reimport button. All existing nodes and elements are deleted and replaced with new nodes and elements specified in the input file. The Mesh Importer also contains a feature that enables reimported FE meshes to retain their connection to the rest of the thermal model. Objects such as heat loads, ors, and conductors are normally deleted if the nodes and elements to which they connect are deleted. If a finite element model is updated by deleting and reimporting, those connections will be lost. The solution is to use Tag Sets (see “Domain Tag Sets” on page 2-21) for defining connections to the FE model.

Data Exchange

18-11

Tag Sets are a named collection of objects of a particular type, such as nodes, lumps, edges, surfaces, or solids. Tag Sets also hold additional information about the objects they contain, such as particular sides or edges of surfaces. For example, a Tag Set consisting of faces will contain surfaces along with the specification of which sides belong in the set. If the imported FE model format allows the ability to specify groups of entities (for example the SET3 command for NASTRAN), these FE entities will be placed in a Tag Set and made available for use by other Thermal Desktop objects. Network elements that reference Tag Sets are updated to reference the actual entities when a Case Set is launched. Deleting the objects that are in a Tag Set does not delete the reference to the Tag Set that is used by objects such as heat loads, heaters, and ors. The objects in the Tag Set may be deleted and replaced without affecting the lifetime of the objects that reference the Tag Set. If it is known that the external FE model will be updated a number of times (to perhaps a trade study or design optimization), then it is worthwhile to define those groups of elements and/or nodes in the FE model that will be used to connect to the rest of the thermal model. Make the connections using the Tag Set names rather than the actual entities, and the model will automatically remain valid when the FE model is reimported. Two other functions are provided by the Mesh Importer: deleting nodes and elements of the mesh, and releasing the mesh from the importer. Selecting the Delete Imported Mesh button will delete the Thermal Desktop nodes and elements that were created on import. The Mesh Importer will remain, but will be essentially a graphics only representation. Selecting the Release from Controller button will disassociate the nodes and elements form the Mesh Importer. The Mesh Importer will now be a graphical only representation, but the nodes and elements will still exist in the thermal model as independent entities. Those entities will now be able to be deleted and repositioned independently if desired. Please note that if a mesh is released, and then reimported, two meshes will exist in the model. These functions are provided for completeness of possible operations on an imported mesh, but in general will not be typically used. 18.2.3.2

FEMAP Neutral File Note: This section provides the capabilities and limitations to importing FEMAP files either with the Mesh Importer (Section 18.2.3.1) or the Thermal > Import > FEMAP command. Due to the capabilities of the Mesh Importer, importing without the Mesh Importer is discouraged.

Thermal Desktop will import FEMAP Neutral files by selecting Thermal > Import > FEMAP ascii neutral(v10.2) command. The file must be an ASCII Neutral file in version 8.0 format for FEMAP. Please note that only the meshed data, nodes and finite elements, are imported. Geometry data is not imported. When Thermal > Import > FEMAP ascii neutral(v10.2) is selected, the FE Model Units dialog box appears. This informational dialog box cautions the about units used in the model and suggests a means to enhance graphical display performance. When OK is

18-12

Data Exchange

selected in the FE Model Units dialog box, the Open dialog box that allows the to select the ASCII Neutral file name (*.neu) appears. Once a file is chosen and OK is selected, the model will be read and displayed. 18.2.3.3

I-deas FEM Note: This section provides the capabilities and limitations to importing I-deas FEM files either with the Mesh Importer (Section 18.2.3.1) or the Thermal > Import > I-deas FEM command. Due to the capabilities of the Mesh Importer, importing without the Mesh Importer is discouraged.

Thermal Desktop will import an I-deas universal finite element model. Only the nodes, elements, material properties, and thicknesses are imported. The Thermal > Import > Ideas FEM command brings up the Open dialog box that allows the to select the universal file name (*.unv). Once a file is selected, the model will be read and displayed. The model will be imported into the I-deas analysis group. Parabolic elements will be converted to linear. If TMG boundary conditions are defined in the universal file, Thermal Desktop will attempt to read them in and apply them on the Thermal Desktop model. s must be very cautious of this approach due to the different methodologies used by TMG and Thermal Desktop. The should all boundary conditions in the TMG model and determine if they have been implemented correctly in the Thermal Desktop model. Finally, the must compare the temperatures between the model to determine if the import has worked properly. 18.2.3.4

I-deas FD

Thermal Desktop will import an I-deas universal finite element model in the finite difference surface mode used by I-deas TMG. Only the nodes, planar elements, material properties, and thicknesses are imported. This allows the Thermal Desktop to quickly calculate radiation conductors for a TMG model. The Thermal > Import > I-deas FD command brings up the Open dialog box that allows the to select the universal file name (*.unv). Once OK is selected, the model will be read and displayed. 18.2.3.5

NASTRAN Note: This section provides the capabilities and limitations to importing NATRAN files either with the Mesh Importer (Section 18.2.3.1) or the Thermal > Import > Nastran command. Due to the capabilities of the Mesh Importer, importing without the Mesh Importer is discouraged.

Thermal Desktop will import Nastran input files by selecting the Thermal > Import > Nastran command. The importer can import the following element types:

Data Exchange

18-13

• CBAR • CONROD • CROD • CTRIA3 • CQUAD4 • CTETRA • CHEXA • ENTA Please note that while element types CROD, CONROD, and CBAR are recognized by the importer, a conductor is used for these elements if an area is available. The importer will extract thermal property information from material property entries MAT1 through MAT4. The Thermal Desktop name for the extracted thermal property will be M## where ## is the original Nastran material number. Grid points, when attached to a recognized element, are turned into Thermal Desktop nodes. Free grid points, either because they were free in the Nastran model, or those that were only attached to unrecognized elements, are not imported into Thermal Desktop. Imported nodes are then connected into the same elements that existed in the Nastran model. If an element references a material ID that wasn't of type MAT[1-4] it is created with the default Thermal Desktop properties. Two node elements (CBAR) are created as nodes and a conductor. The conductor is a plain SINDA conductor that uses the cross sectional area of the bar as the A term in KA/L. In most instances, the conductor value is set to zero since the CBAR definition does not set the conductivity. When Thermal > Import > Nastran is selected, the NASTRAN Model Import Options dialog box opens (Figure 18-6). Input File. The drop-down list will list all BDF files in the DWG directory. The also has the option to browse for the NASTRAN file. Import FE model as a thermal model. When selected the FE model will be imported as nodes and elements. Import FE model as graphics only. When selected the FE mesh will be imported as a graphical opject. This will not allow running a solution using the FE mesh. Submodel. The drop-down list allows the to select the submodel in which to place the imported nodes and elements. The can optionally type in a new submodel name. Layer. The drop-down lists all existing layers. The selected layer will be used to store all imported objects unless the Put Elements on Layer by Physical ID box is checked. If the model has coordinate systems, then the items will be put on layers that are appended with the coordinate system name. For example, if the inputs MAP as the input layer, items in rectangular coordinate system 52 will be placed on layer MAP-RECT52.

18-14

Data Exchange

Figure 18-6

NASTRAN Model Import Options Dialog Box

Convert Thermal Boundary Conditions. When checked, the importer attempts to convert some heat loads and boundary nodes that are defined in the NASTRAN file. Complete boundary condition importation is not possible, so the must be very careful. Put Elements on Layer by Physical ID. When checked, the imported model will be placed on layers based on physical ID of the elements. For example, all solid elements and their nodes will be placed on a new layer for solid elements. Create groups from PATRAN neutral file. When checked, a PATRAN neutral file containing named components (Packet Type 21) will be read and the matching nodes and elements will be placed into AutoCAD groups with the same name as the PATRAN named component. The can select a file from the drop-down or browse for an unlisted file. The can use the new groups to create domain tag sets for applying heat loads, conductors or ors, or for multi-editing sets of nodes or elements. Import a NASTRAN file in the tutorial exercise “Mapping Temperatures From a Coarse Thermal Model to a Detailed NASTRAN Model” on page 20-157.

Data Exchange

18-15

18.2.3.6

ANSYS® Note: This section provides the capabilities and limitations to importing ANSYS files either with the Mesh Importer (Section 18.2.3.1) or the Thermal > Import > ANSYS command. Due to the capabilities of the Mesh Importer, importing without the Mesh Importer is discouraged.

The Thermal > Import > ANSYS command will import the finite element data of the ANSYS CDB file. The CDB file must be the BLOCKed format of this file. The importer will read in the finite elements, as well as the thicknesses of the two dimensional elements. Non-linear elements will be imported as linear. Table 18-1 lists the types of elements currently imported from ANSYS. Table 18-1 Imported ANSYS Elements

Element Type

Description

3

18-16

42

2-D Structural Plane

45

3-D Structural Solid

55

2-D Thermal Plane

57

Thermal Shell

70

3-D Thermal Solid

82

2-D 8-Node Structural Solid

87

3-D 10-Node Tet Thermal Solid

91

Nonlinear Layered Structural Shell

92

3-D 10-Node Tet Structural Solid

93

8-Node Structural Shell

95

3-D 20-Node Structural Solid

99

Linear Layered Structural Shell

142

3-D Fluid-Thermal

Data Exchange

Table 18-1 Imported ANSYS Elements

Element Type

Description

152

3-D Thermal Surface Effect

154

3-D Structural Surface Effect

183

Quadratic Quads

200

Parabolic Hex

Material property names will be imported, as well as constant density, specific heat, and conductivity. If the model contains specific heat and conductivity data, the must set the Thermal Desktop units to be the same as the ANSYS file so that these can be imported properly. If the model does not have specific heat and conductivity data, then the units do not need to be set. The Import ANSYS dialog box will be displayed for file name input (Figure 18-7). The Import FE model as graphics only option imports a graphical representation of the FE model. When the new graphical object is selected and edited, viewing options are provided (e.g. - wireframe outline) for improved visibility. The graphical object can also be used for snap points for building model geometry. Note that most ANSYS stress models do not contain specific heat and conductivity, so the should always make a review of the material properties once the import has occurred.

Figure 18-7

18.2.3.7

Import ANSYS Dialog Box

STEP-209

STEP-209 is the finite element standard of STEP. Finite element information may be imported using this standard. Material property names are transferred in the process, but the actual thermal properties of capacitance and conductance are lost. Radiation properties are not transferred. Select Thermal > Import > STEP-209 to initiate this function. Data Exchange

18-17

18.2.4

ANSYS® Iceboard®/TASPCB/BetaSoft

Simplified representations of printed circuit boards created using ANSYS Iceboard software (formerly known as Harvard Thermal TASPCB) and also BetaSoft may be imported into Thermal Desktop for system-level analysis. After constructing a detailed model of a circuit board and components using ANSYS Iceboard software, two files are exported from ANSYS Iceboard software. The first file is an equivalent thermal representation of the circuit board. Using ANSYS Iceboard software, the circuit board is divided into a set of rectangular patches. The equivalent principle conductivities (Kx, Ky, and Kz) are computed for each patch, taking into all of the details of the circuit board, such as traces, ground planes, and vias. The second file exported from ANSYS Iceboard software consists of component definitions. Each component contains information on how it is bonded to the circuit board, its size and location, and a simplified network to compute case and junction temperatures. Heat loads are also exported. These two files are imported into Thermal Desktop using the command Thermal > Import > TASPCB. This command will invoke the TASPBC Import dialog box shown in Figure 18-8.

Figure 18-8

TASPBC Import Dialog Box

The Effective PCB Properties File and the Component Model File are the files exported from ANSYS Iceboard software. The Board ID field is used to create the submodels used for the nodes on the board and components. If more than one circuit board is to be imported, specify a different ID for each board. An option is available to import the board using one node through the thickness of the board (select the Use Same Node ID’s on both sides option), or using two nodes through the thickness (uncheck the Use Air Gap Conduction option). The component model contains information on the conduction between components and the board, with some conduction computed from the air gap between the component and the board. This conduction connection may be eliminated by unchecking Use Air Gap Conduction, for example, if the board will be used in a vacuum.

18-18

Data Exchange

The board will be imported into Thermal Desktop at the current UCS location, using the 2D coordinates specified in the original ANSYS Iceboard definition. Units conversion is automatically done. An example board is shown in Figure 18-9.

Figure 18-9

Circuit board Imported From TASPCB

Four layers are created for each imported board. There are separate layers for the board, the underlying conduction network, the components, and the junction nodes with their associated heat loads. The board layer contains finite element representations for each patch. The elements are only used for radiation and convection, since the thermal node/conductor network is computed using the equivalent Kx, Ky, and Kz conductivities for each patch. The elements are set to be active in the current radiation analysis group. Optical properties are created for the board, to be filled in with appropriate values by the . The layer name for the board will be BXX_PCB_board where “xx” is the board ID. The conduction network for the board is created using nodes and conductors in a traditional finite difference grid, using the equivalent thermal properties in each local board patch. The layer containing the equivalent conduction network is named BXX_PCB_cond and is turned off by default. The network for the example board is shown in Figure 18-10. The components are represented as rectangles. Similar components (same reference designator) are all placed in a submodel with the name constructed from the board ID and the component reference designator specified in the imported file. The component surfaces are made active in the current radiation group. Depending on the bonding method (ball grid, leads,...), Area and/or Edge is automatically specified to thermally connect the component to the circuit board. The components are placed on layer BXX_PCB_comp.

Data Exchange

18-19

Figure 18-10

Equivalent Anisotropic Network for Detailed Circuit Board

Nodes and conductors representing the component junction and thermal connection to the component case for each component are placed on layer BXX_PCB_junc, which is turned off by default. Heat loads specified for each component are also placed on this layer. Figure 18-11 shows the junction nodes and heat loads for the example board.

Figure 18-11 nent

18-20

Junction nodes, conduction to case, and heat loads for each compo-

Data Exchange

The imported board contains all of the thermal information necessary for analysis and may be analyzed by itself, or included into a system-level model. Components may be relocated in Thermal Desktop for optimization purposes, but a final analysis will still need to be performed as changing the location will also affect the traces and therefore the equivalent board properties. Additional conduction, for example, to simulate a thicker ground plane, may be added to the finite elements used to model the board. Adjust the material property created for the board to include additional conduction effects. The steady state results for the example board connected to a chassis using wedge locks on the left and right sides, and placed in a cold radiation environment are shown in Figure 18-12.

Figure 18-12

Data Exchange

Analysis results computed using circuit board imported from TASPCB

18-21

18.3 18.3.1

Export Data Write Node Information

The Thermal > Export > Write Node Information displays the Export Node Info dialog box and allows the to write out node locations (in WCS or current UCS coordinates), current post processed values, and surface areas. This dialog box also gives the the option to write the data to the screen or to a file. The format of the file is simply Submodel.ID followed by the requested data. One node is output per line. If two nodes share the same id, then if the node locations are being output, the nodes will be in the list twice. If node locations are not output, then the node will only be in the list once. When selected the Export for each time point in current PP Data button outputs the data for every time point of the current post processed dataset. When this button is selected along with the Current Post Processed Data button, the program will cycle through each time point in the current post processed dataset and write that data to a file. The file name will be increment by one for each time point. The Export Node Info dialog box is shown in Figure 18-13.

Figure 18-13

18-22

Export Node Info Dialog Box

Data Exchange

18.3.2

Postprocessing Data Mapper

The Postprocessing Data Mapper is an improvement to the Map Data commands (see Section 18.3.3). It is intended to supersede the Map Data commands, however, the Map Data commands will be retained to facilitate the transition to this new approach. The Data Mapper is a graphical entity that is saved in the drawing database. After it is created, the data mapper can be repositioned using translations and rotations to ensure the structural model is aligned with the thermal model preventing the need to create the two models in the same coordinate system. The Data Mapper entity is selected and edited just like any other Thermal Desktop entity. An editing dialog will appear that contains buttons and fields to control and execute the mapping function. When mapping data, insulation temperatures (see “Insulation Tab” on page 4-17) are not included in the mapping. The mapping is intended to be used for mapping to structural models and the insulation is typically not associated with the structural model. If the insulation temperature is required to be mapped, the insulation must be explicitly modeled using additional surfaces, solids, or elements instead of the insulation tab on the edit form. By default, the data mapper only uses the grid points of the NASTRAN model. The only limitation on the element types is when gradients are being exported: gradients are exported to CTRIA3 and CQUAD4 elements, only. The Data Mapper only maps temperatures from geometric objects: finite elements and Thermal Desktop solids and surfaces. A Data Mapper will map data when the edits the mapper and selects Exit & Map. If several mappers are in a model, the can request all mappers to be updated by typing the command tdMapAllMappers. The Case Set Manager allows updating all mappers when a solution is completed (Section 15.2.2.1). Want "Hands-On" Information? A tutorial on mapping can be found in “Mapping Temperatures From a Coarse Thermal Model to a Detailed NASTRAN Model” on page 20-157. 18.3.2.1

Creating a Postprocessing Data Mapper

The Postprocessing Data Mapper maps data values in the current post processing dataset to a finite element model. The finite element model may be in NASTRAN, ANSYS, FEMAP, or I-DEAS format. As many data mappers as needed may be created and saved in the drawing database. A postprocessing data mapper is created using the command TdMapperPPToFEM or the menu choice Thermal->Export->Post Processing Data Mapper. Invoking the command to create a post processing data mapper brings up the dialog shown in Figure 18-14. The fields on the form are as follows: Format. A drop-down list to select the source of the mesh (NASTRAN, ANSYS, FEMAP

or I-DEAS)

Data Exchange

18-23

Input File. A drop-down list listing any standard-extension files in the working directory

(the directory containing the DWG file) that match the source selected in the Format field. The may also choose to browse for the input file. Input Units. The dimension units used in the input file. This allows the mesh to be converted

to the thermal model units. Use World Coordinate System (WCS) and Use current Coordinate System (UCS) radio buttons. Allows to choose which coordinate system to use for importing

the mesh. Further translations and rotations can be made after the data mapper is created to ensure the structural and thermal models are aligned.

Figure 18-14

Post Processing Data Mapper Input File Dialog

A graphical representation of the finite element model will be displayed. The data mapper may be repositioned using grip editing operations, or the AutoCAD MOVE, ROTATE, or ALIGN commands (see AutoCAD Help). 18.3.2.2

Editing a Postprocessing Data Mapper

After the data mapper has been created, or whenever it is edited, the dialog shown in Figure 18-15 is displayed. The dialog allows the to control the mapping process and the visual representation of the display of the finite element model.l The dialog is subdivided into sections as follows: Comment. A field for a descriptive, multi-line comment. The first line of the comment will

appear in the model browser.

Input/Output Input File. The name of the current input file. If an update to the finite element model to which thermal data is being mapped is obtained, it can be reimported using the Reimport

button. The Data Mapper will automatically reimport the finite element model using the file name and format that was specified when it was originally created. Any translations, rotations, and scaling that had been applied to the data mapper are retained and the graphical display will be updated with the new model. All current map settings, group associations, and tolerances are also retained. 18-24

Data Exchange

Figure 18-15

Post Processing Data Mapper Editing Dialog

Output Format. A drop-down list containing the format of the structural solver: NASTRAN

or ANSYS. Output File. The name of the file where the structural temperatures will be written. The may choose to browse for an existing file by using the Browse button.

Data to Map Map post processing dataset at current displayed time. Radio button to map a single

point in time or a steady state solution. Map post processing dataset for all times in dataset. Radio button to map a transient

solution for all output intervals Map for selected times. Radio button to map a transient solution at selected times. The Edit Selected Times button opens a window listing all times available for mapping. Output element / gradient data for planar elements (TEMPP1). Check box to calculate

element average temperatures and gradients (dT/dx) in solid objects or double-sided surfaces (Section 4.3.1.2) and apply to structural models using TEMPP1 cards. Currently, this capa-

Data Exchange

18-25

bility is limited to NASTRAN CTRIA3 and CQUAD4 elements in the mesh. When this box is checked, the TEMPP1 cards will be written to the Output file, named above, with the *.grad extension. Grid temperatures will always be mapped. Output element / gradient data for 1D elements (TEMPRB). Check box to calculate element average temperatures and gradients (dT/dx) for 1D elements and apply to structural models using TEMPRB cards. Currently, this capability is limited to CBAR, CBEAM, CONROD and CROD NASTRAN elements.

Preferences Display Preferences. A button that allows choosing the display mode of the Mesh Dis-

player. The options on the form are described in Section 2.10.15.

Data Mapping and Gradient Mapping Temperature mapping uses the selections in the Data Mapping section; gradient mapping, chosen by selecting the Output gradients normal to planar elements check box, uses the selections in the Gradient Mapping section. The radio buttons and drop-downs have similar definitions. Use all entities in thermal model. Uses all items in the thermal model for mapping tem-

peratures. The first thermal model item found within the tolerance will be used for applying the temperature or gradient Use AutoCAD group. Uses only items within the specified AutoCAD group (Section

19.7.1) to map temperatures or gradients. This provides a level of accuracy not provided with tolerance: mapping temperatures of a bracket that is mounted to a plate can ensure that only the bracket temperatures are captured. This options maps the temperatures of the selected group to the entire structural model. It is not necessary to partition the mapping process for speed, only to guarantee that certain nodal points in the external finite element model map only to selected thermal entities. Associate Groups. Allows the to associate AutoCAD groups (Section 19.7.1) with

groups defined in the structural model file. This provides the part-by-part accuracy of the previous option without requiring a separate mapper for each part. It is not necessary to partition the mapping process for speed, only to guarantee that certain nodal points in the external finite element model map only to selected thermal entities. When using group associations for data mapping, the external finite element model must contain groups of nodal points, not elements2. The corresponding AutoCAD group of thermal entities must contain thermal surfaces, planar thermal elements, or solid thermal elements. When mapping gradients using group associations, the external finite element model must contain groups of planar elements. Group associations are specified using the Edit Group Associations... button and is shown in Figure 18-16. Only structural grid groups will be displayed for data mapping and only structural element groups will be displayed for gradient mapping. 2For NASTRAN, groups are defined using the SET1 and the SET3 commands. SET1 defines groups of grid points. SET3 can be used for grids or elements by setting the DES parameter to GRID for grid points and ELEM for elements. Only grid points can be used for temperature mapping; element groups are used for gradient mapping.

18-26

Data Exchange

Figure 18-16 Input form for associating groups in the external model with groups in the thermal model

Select a group defined in the external finite element model using the left drop-down list and an AutoCAD group of thermal entities in the right drop-down list. Selecting the Update/ Add button will add the group association to the list. If the structural group is already in the association list, it will be updated with the new thermal AutoCAD group. Highlighting an association in the list, and selecting Change will set the Structural Group Name dropdown menu to the structural group in the association and the AutoCAD Group Name to the thermal group in the association. Control will be set to the AutoCAD Group Name drop down. Select a new thermal AutoCAD group and then select Update/Add to update the association. The Delete button will remove an association from the association list. All nodal points not explicitly specified in a defined group in the external finite element file are placed in the group NotInAnyDefinedGroup. This group may be used as any other group. The Use all thermal entities for structural groups that do not specify an AutoCAD group option will test all thermal entities for nodal points not explicitly put in the association list. If this option is not checked, mapping will not be performed for nodal points not in groups explicitly added to the association list. Group associations may be verified by using the Display Only button. The thermal group is isolated in the graphics display along with the nodal points of the structural group. Display All will restore the display and make all entities visible. Use structural thickness for gradient calculation. When unchecked, the thickness of the Thermal Desktop objects will be used to determine the temperature gradient (dT/dx); when checked, the thickness of the structural element will be used to determine the gradient, ignoring the thickness of Thermal Desktop objects.

Data Exchange

18-27

Tolerance Mapping Tolerance. This button that opens a tolerance list. A series of tolerances may be specified and is highly recommended. If the data mapper finds a thermal entity within the tolerance specified, it will use it. It does not check all entities to find the closest thermal entity, any thermal entity within the tolerance could be used for the mapping operation. Using a series of tolerances will enable a much faster and more accurate mapping operation than if a sorting operation is performed. The first tolerance should be a value of zero. Subsequent tolerances should be progressively larger up to the largest expected deviation between the thermal model and the structural model. The more tolerance levels used, the more accurate the mapping will be. The data mapper will map all nodal points with the first tolerance. If there are any nodal points that fail the first mapping operation, the operation will be repeated with the failed points using the next tolerance. This operation is repeated until either all external nodal points are mapped, or the list of tolerances is exhausted. If the Create graphic markers for points that fail to map option is checked, red cross marks are placed at the locations of the external nodal points that failed to find a suitable thermal entity. These marks can be cleared using the Thermal->Reset Graphics command. 18.3.2.3

Postprocessing Data Mapper Operation

The mapping is performed by taking each nodal point in the external finite element model (and the centers of planar elements when mapping gradients) and finding which thermal entity contains the point within the tolerance specified. If the point lies within a solid finite element, data is interpolated using the shape functions of the element. For surfaces and planar elements, the point is mapped into the parametric space of the entity and interpolation is performed. Thicknesses of surfaces and planar elements are used to determine if the point is within the thermal entity. That is, a point will map to a thermal surface if the point is within the thickness of the surface plus the tolerance specified. A new feature of the Data Mapper is that data is interpolated within double sided surfaces using the thickness of the surface. That is, if a nodal point specified in the external finite element model lies within the thickness of the thermal surface, the data will be interpolated from the top side to the bottom side as well as in the parametric directions of the surface. Structural elements that use the through thickness gradient option should be located at the mid plane of the surface. The nodal points will then obtain the average temperature of the top and bottom sides at that parametric location, and the center point will obtain the gradient. If the point is outside the thickness of the surface, through thickness extrapolation is not performed. The point will have the temperature of the top side or the bottom side of the surface. Data in the parametric directions, however, are extrapolated beyond the boundaries of the surface. Data for solid and planar thermal elements are also extrapolated beyond the boundaries of the surface if the point lies within the tolerance specified. Extrapolation is linear near the element, but exponentially approaches the average temperature of the element in the far field. This prevents wild extrapolations for external nodal points that are grossly misaligned with the thermal model. 18-28

Data Exchange

18.3.2.4

ing Results of the Mapping Operation

Another feature of the Data Mapper is that the visual representation of the external finite element model is also post processed with the data that was mapped. The Data Mapper saves interpolation factors, so that as the post processing dataset is advanced through time points, the Data Mapper will display the values that were mapped on to it without having to remap the nodal points. This allows the to examine the values for correctness before sending them downstream for further analysis. The Display Preferences... button controls the appearance of the Data Mapper. The options on the form are described in Section 2.10.15. 18.3.2.5

Additional Output Files

In addition to the output file specified on the editing form, an number of additional files are also output. These files contain useful information about the mapping process. The file MapInterpFactorsGridPoints.txt and MapInterpFactorsGradients.txt contain the interpolation factors computed by the data mapper. Each line will contain an equation that relates the ID of the finite element nodal point to one or more interpolation factors multiplied by the name of the Thermal Desktop node. This can be used to examine and the correctness of the mapping operation, or to facilitate other automated data translation functionality implemented as a custom application by the . Another possible application for the interpolation data is its use for automatic data correlation. If the locations of thermal couples for a thermal test are input as finite element grid point locations (element definitions are not necessary), the output will contain all the information needed to compute the thermal couple temperatures in a SINDA/FLUINT run. Inserting appropriate “T” characters will produce valid logic for use in S/F VARIABLES calls. The files MapSummaryGripPoints.txt and MapSummaryGradients.txt lists the status of the mapping operation for each nodal point in the external finite element model. Each line lists the ID of the finite element nodal point followed by its x, y, z location. The remainder of the line lists the status of the mapping operation. If the data mapper failed to find a Thermal Desktop entity to obtain data, the line will list “Did not map”. If the mapping operation was successful for that location, the line will list Cond/Cap submodel of the Thermal Desktop entity that was used for mapping along with the tolerance and the name of the entity. Locations that fail to map are listed at the beginning of the file. The files Mapper_ _map.log and Mapper_ _gradient_map.log, where is the AutoCAD database identifier for the data mapper, echo the output to the screen during the mapping operation. It will list the status of the mapping operation as each successive tolerance is processed. Finally, in addition to the data output in the requested finite element format, additional ASCII text files are created. The file names are Temps.dat and TempsTransient.dat, where is the name of the requested finite elData Exchange

18-29

ement format for output (i.e. NASTRAN, ANSYS). The Temps.dat file contains the label of the external finite element node and the data that was mapped. The TempsTransient file contains transient data in Thermal Desktop text file post processing format. These files may be read back into Thermal Desktop when doing thermal model to thermal model mapping, or may be used by further custom applications developed by the . 18.3.2.6

Tips for Better Mapping

The biggest factor for successful mapping of thermal data onto a structural model is the physical congruence between the two models. The benefit in effort spent up front in collaborating with the stress analyst to develop similar modeling abstractions that will facilitate the mapping process can not be over emphasized. However, such ideal conditions do not always exist. For areas of the model that do not align well, using group associations can at least prevent unwanted areas of the thermal model to be mapped to the structural model. Isolating thermal and structural components for mapping using group associations will more likely allow the desired thermal entities to be used for mapping. Using a progressive set of mapping tolerances will also aid in selecting the closest thermal entities to each of the structural model nodal locations. Groups for the structural model are defined in the structural model input file. AutoCAD groups are used to define the thermal groups. Groups may be created using the AutoCAD group command in conjunction with any selection set technique (please see the “group” command in the AutoCAD help). Thermal Desktop also has a variety of commands to help create AutoCAD groups, for example, creating a group from all entities that are in a particular SINDA/FLUINT submodel. For representations that are grossly misaligned, auxiliary thermal geometry may be created. The data mapper uses the data displayed during post processing for mapping. The thermal entities displayed are not required to have been used in the analysis. For example, a thermal finite element brick can be created, that has a dummy material property that has the conductivity and the density set to zero. This will prevent any conduction or thermal capacitance to be generated for the S/F input file, the geometry will only used for display. If the nodes on the brick element are all given the same node ID, and temperatures for that node have been calculated by S/F, then that brick defines a zone where any structural location that falls within that region will receive the temperature of that node. Likewise, nodes on one side of the brick could be given a particular thermal node ID, and the nodes on the other side a different thermal node ID. Now the brick defines a region where the gradient between two temperatures exist. Any combination of node naming and finite element shapes can be used to define any arbitrary temperature field. Any Thermal Desktop entity can be created and used for auxiliary mapping, it is only the node ID’s that matter. For each thermal entity, the post processing operation looks up the thermal node name and data from the S/F results, whether or not it was used to create the data. For example, a rectangle can be created with the same nodalization and node numbering as a cylinder used for analysis. If this rectangle is not in any radiation analysis group, and the Cond/Cap option 18-30

Data Exchange

is unchecked, it will not be part of any thermal analysis. However, it will display the results of the temperatures computed for the cylinder. It allows a different view of the data, allowing the to see all sides of the cylinder at the same time. This technique can be used to better visualize thermal results, as well as a technique for aligning thermal and structural models. If necessary, a completely different drawing may be used for the post processing operation. The may make a copy of the drawing used for analysis, and modify it as necessary for better alignment with the stress model. For example, surfaces can be resized and repositioned to align with the stress model. This might create an unacceptable thermal model, but the copy is used only for the mapping operation. The results are obtained from the original thermal model. In the copy of the analysis drawing, create a post processing dataset using the S/F results created during the analysis of the original model. (If the drawing is copied after the analysis, this post processing dataset will already exist in the copy). Using a separate drawing also has the benefit that dummy material properties do not need to be defined, since no analysis is performed with the auxiliary mapping model. The thermal entities may be modified in the copy, and auxiliary geometry may be added as necessary to produce an accurate mapping of thermal data to the structural model. 18.3.3

Map Data Commands

The map data commands provide the functionality of taking any 3D point and determining what the data value (often temperatures) at that point will be. The program compares the point against the volumes of each entity in the model. When it finds an entity that contains the testing point, the program will then use the shape functions of that entity to determine the precise value for the test point. Shape functions are used for finite elements, while interpolation is used for finite difference surfaces. This capability is often used to provide temperatures from the thermal model to a stress model that has nodes in different locations. This capability can also be used to convert AeroHeating rates from a CFD model to your thermal model. The has a couple of options for control over the mapping. First, the can input tolerance for the mapping. This tolerance can either be constant or be variable. Since the program finds the FIRST surface/solid within the tolerance, the should not use large tolerances. It is highly recommended that variable tolerance be used, and that the first value in the list be zero, and then have the tolerance gradually increase. As an example, if a has set a zero tolerance then Thermal Desktop will take the location of the structural node and search the thermal model to find if it lies within a thermal volume. A thermal volume could be the volume associated with a finite element; a volume can also be a Thermal Desktop surface with a thickness associated to it. A Thermal Desktop surface is displayed in the program where the mid-plane of the surface lies. The volume associated to the surface is determined by taking 1/2 of the thickness in both directions from the mid-plane. When a structural point is found to lie inside a thermal volume, the temperature value at that point is the mapped value. As the increases the tolerance, the dimensions of each entity is increased in all directions, so adjacent volumes will start to overlap, and as the

Data Exchange

18-31

tolerance increases, points may map to different volumes than when a zero tolerance is used. The program will go through all the points with smallest tolerance, and will then increase the tolerance for the points that do no map with the tolerance at the previous smaller value. By default, the map data commands use the current post processing dataset as the data values for the map. The can also specify to map all the times for the current post processing dataset. The can also control which entities are considered for the mapping. This can aid in restricting which entities are used for the mapping. Finally, the can have the program create points for any point that does not map to anything within the tolerance. These points are always created on the RADCAD_RAYS layer, and may be removed from the model using the Thermal > Radiation Calculations > Clear Ray Plot command or by using the RcReset command (green Reset Thermal Desktop Graphics icon). Also note that the default AutoCAD drawing of a point is a cross hair, and these points might be better observed by changing PDMODE to 31. Please consult the AutoCAD manual for a discussion of PDMODE and PDSIZE. A summary of the command is always printed in the mapSummary.txt file. This file will list the points that were mapped from, their coordinates, the tolerance used, the conductor submodel of the surface mapped to, and finally the description of the surface that is mapped to. The can use this file to figure how any point in the model mapped and to which surface or solid that was mapped to. When mapping data, insulation temperatures (see “Insulation Tab” on page 4-17) are not included in the mapping. The mapping is intended to be used for mapping to structural models and the insulation is typically not associated with the structural model. If the insulation temperature is required to be mapped, the insulation must be explicitly modeled using additional surfaces, solids, or elements instead of the insulation tab on the edit form. 18.3.3.1

Map Data to Locations

The Thermal > Export > Map Data to Locations command will display the Map Data to External Model dialog box and map the node locations of a input file to the post processed data on the current model. The results are output to the specified file. The Map Data to External Model dialog box is shown in Figure 18-17. The format of the file for Map Data to Locations is simply node name followed by x, y, and z locations. Each field must be separated by a space or tab. Such as: MAIN.1

1.

2.

3.

MAIN.2

.1

.2

.4

18.3.3.2

Map Data to NASTRAN Model Note: This command is superceded by the postprocessing data mapper (Section 18.3.2).

18-32

Data Exchange

Figure 18-17 Map Data to External Model (Locations) Dialog Box

The Thermal > Export > Map Data to NASTRAN command will display the Map Data to External Model dialog box and map the node locations of a input file to the post processed data on the current model. The results are output to the specified file. The input dialog box associated with the Thermal > Export > Map Data to NASTRAN command is shown in Figure 18-18. After the has mapped to an external model, they can then take the temperature boundary conditions and include (cut and paste) them into the proper place in the NASTRAN model file, and then run NASTRAN to perform thermal stress calculations. If all the time points are mapped, the data must control the NASTRAN temperatures with the use of case ids (SID). 18.3.3.3

Map Data to ANSYS Model Note: This command is superceded by the postprocessing data mapper (Section 18.3.2).

Data Exchange

18-33

Figure 18-18 ANSYS)

Map Data to External Model Dialog Boxes (NASTRAN and

The Thermal > Export > Map Data to ANSYS command will display the Map Data to External Model dialog box and map the node locations of a input file to the post processed data on the current model. The results are output to the specified file. The input dialog box associated with the Thermal > Export > Map Data to ANSYS command is shown in Figure 18-18. After the has mapped to an external model, they can then take the temperature boundary conditions and include (cut and paste) them into the proper place in the ANSYS model file, and then run ANSYS to perform thermal stress calculations. If all the time points are mapped, the data must control the ANSYS temperatures with the use of case ids (SID).

18-34

Data Exchange

18.4 18.4.1

Export Models Export Portion of Thermal Desktop Model

A subset of a Thermal Desktop model can be exported for use by itself or for insertion into another Thermal Desktop model. This is accomplished using the WBLOCK command. The WBLOCK command allows to to select the objects that will be written to a new drawing file. This new drawing file can then be opened or inserted into another model. Important: On the Write Block dialog box, do not change the Insert units. The proper steps for this method are: • Select all of the objects that are to be copied. Ensure visibility is on for all layers contain objects to be copied and the global visibility of the nodes, ors, heat loads, etc. is on. • Use the WBLOCK command to write the selected graphical objects to a new file name. • the units of the original model. • Open the DWG file in which the block will be written. • Set the units to match the original file. • Use Insert > Block to insert the objects copied from the original file. • If the Explode option was not selected on the Insert block form, then type explode or select Modify > Explode.

The newly created DWG file can be opened as a Thermal Desktop model or inserted into another model using the process described in Section 18.2.1. This method only exports the Thermal Model objects (surfaces, solids, finite elements, etc.) and network elements (heat loads, conductors, ors, etc.). To use Case Sets, symbols, properties, Tag Sets, etc., use the import and export buttons on the appropriate manager forms (Section 2.10.12). Note: Network elements that reference Domain Tag Sets cannot be included in WBLOCK operations. (see Section 2.5) 18.4.2

TRASYS

The Thermal > Export > TRASYS command will write out a TRASYS model of the current default analysis group. The is first prompted for the TRASYS file name with the display of a Save As dialog box. When a file name is input and OK selected the TRASYS Export dialog box appears (Figure 18-19). The TRASYS exporter exports all of the surfaces associated with an analysis group. The export process creates one TRASYS BCS and one TRASYS submodel for each submodel in Thermal Desktop. Since TRASYS does not allow analysis groups, the must output each analysis group independently. Data Exchange

18-35

Figure 18-19

TRASYS Export Dialog Box

Each of the custom entities is exported as its equivalent TRASYS primitive. Entities which are represented as a mesh are exported as a series of triangles. The export function automatically generates correspondence data that will recombine these triangles back into node numbers that match those in Thermal Desktop. Triangles are used to avoid the automatic generation of node numbers in TRASYS polygons, as these node numbers might conflict with node numbers in Thermal Desktop. Thermal Desktop correspondence data is also exported to the TRASYS model. There are several options for the TRASYS export. The first two are flags for how the Submodel data is output for surfaces and correspondence. The Consider Geometric Special Cases check box attempts to find Thermal Desktop 4 sided polygons that are really rectangles or trapezoids and output them to TRASYS as such. This can reduce the number of nodes in a TRASYS model. The Write Symbols For Optical Properties check box will define symbols for the optical properties which makes it easier for the to change the properties in the TRASYS model. The Same Point Tolerance input field is used to test 4 sided polygons to make sure that two of the points are not coincident. If two points are less than the input value, then the polygon will be output as a triangle. 18.4.3

TSS

The command Thermal > Export > TSS exports a TSS model from Thermal Desktop. The dialog box associated with this command is shown in Figure 18-20. Thermal Desktop will write the TSS geometry file, *.testium, in the -defined system of units, but the TSS

18-36

Data Exchange

Figure 18-20

TSS Import and Export Dialog Boxes

material file, *.tssma, will be written in MKS units. When the exported model is imported into TSS, TSS will convert the materials from MKS to the current units in the condcap application output. 18.4.4

STEP TAS 5.2

Thermal Desktop will export files to the STEP-TAS, version 5.2, standard. Thermal Desktop surfaces and optical properties and model units will be exported. 18.4.5

STEP-209

STEP-209 is the finite element standard of STEP. Finite element information may be exported using this standard. Material property names are transferred in the process, but the actual thermal properties of capacitance and conductance are lost. Radiation properties are also not transferred. Select Thermal > Export > STEP-209 to initiate this function. A NASTRAN-format temperature file is exported for all finite elements in the model along with the finite element information if solution data is present. 18.4.6

NASTRAN

Using the Thermal > Export > NASTRAN command exports all finite elements in the thermal model into a NASTRAN BDF file. This command only exports the mesh and the material properties.

Data Exchange

18-37

18.5

Export Geometry

The command TdConvertToAcad will convert all finite elements and finite difference objects to AutoCAD polyface mesh. This format allows the geometry to be written to a DWG or DXF file and opened with any application that can import DWG or DXF with polyface mesh. When importing into SpaceClaim, select AutoCAD files, select the Options button, and select Solid Models for “Insert polyface mesh... into 3D as:”.

18.6

Link to TD Direct

CRTech TD Direct is an add-in to SpaceClaim. SpaceClaim allows the to generate, import, clean, heal, and modify CAD geometry without the limitations of a history tree. CRTech TD Direct adds features to SpaceClaim to allow linking Thermal Desktop with SpaceClaim. TD Direct allows marking the geometry in SpaceClaim indicating mesh controls, submodel names, thermophysical and optical property names, face thicknesses, surface insulation, analysis groups, and domains (which are imported into Thermal Desktop as domain tag sets, Section 2.5). With TD Direct, selecting the command Thermal > SpaceClaim > Create SpaceClaim Link allows the to select a SpaceClaim document and import the geometry, as scaffolding or for meshing with TD Mesher, and/or import a finite element mesh of the geometry. As indicated by the command name, the connection is a link and the Thermal Desktop and SpaceClaim documents can be synchronized to update the geometry and/or mesh from SpaceClaim or send values of “driving dimensions” (SpaceClaim geometry parameters) from Thermal Desktop back to SpaceClaim. The synchronization can be performed for individual links by editing the SpaceClaim importer and selecting Synchronize; the synchronization can be performed for all SpaceClaim links by selecting Thermal > SpaceClaim > Synchronize SpaceClaim Links. Details of the use of TD Direct are found in the CRTech TD Direct ’s Guide.

18-38

Data Exchange

19 Interfacing with AutoCAD®

19.1

AutoCAD Versions

Thermal Desktop works with AutoCAD versions 2010 through 2014. As long as the same version of Thermal Desktop is used (or the model has been saved to an older version of Thermal Desktop, Section 2.8.8), older versions of AutoCAD can be used to open a Thermal Desktop model. To do this, select File > Save As and choose the desired version of AutoCAD from the file type option. To always save to a specific version of AutoCAD, select Tools > Options, select the Open and Save tab, and choose the AutoCAD version under Save As.

19.2

Running Thermal Desktop with AutoCAD Mechanical Note: Autodesk Mechanical Desktop is no longer compatible with Thermal Desktop since the last release of Mechanical Desktop was with version 2009 and Thermal Desktop only works with AutoCAD 2010 and later. Although Mechanical Desktop files cannot be saved as AutoCAD files, basic AutoCAD will open files created with Mechanical Desktop.

Thermal Desktop runs fine under the Mechanical extensions of AutoCAD. s usually purchase the Mechanical extensions in order to have access to the IGES and/or STEP importers1. There are two drawbacks to using AutoCAD Mechanical. The first major drawback to using the Mechanical extensions is that they take quite a bit of time to load the DWG file. The second drawback to using the Mechanical extensions is that the drop-down lists are significantly different than just using regular AutoCAD. Because of the menu change and the load time, CRTech recommends that s who have purchased the Mechanical extensions of AutoCAD just run under regular AutoCAD, which is fully installed at the same time as the Mechanical add-ons. Follow these steps to run in vanilla AutoCAD: 1. Use Start > Programs > Autodesk > AutoCAD Mechanical 20xx > AutoCAD 20xx 2. At this point, AutoCAD should come up. Check to see if the Thermal menu is in 1Starting with AutoCAD 2012, basic AutoCAD will import IGES and STEP. See Section 18.1.

Interfacing with AutoCAD®

19-1

the menus. If you do not see the Thermal menu, then CRTech. 3. If you double click on a DWG file, it should now come up in regular AutoCAD and you should see the Thermal menu. You should notice that AutoCAD loaded much faster than before, the menus will be different and will match the Thermal Desktop tutorials. 4. Note that even though you are using regular AutoCAD, you can import IGES and STEP by typing in the command (it may not be in the menus). “IGESIN” is the command for IGES, and “STEPIN” is the command for STEP. See Section 18.1.2 and Section 18.1.3, respectively. 19.2.1

AutoCAD Mechanical 2D Meshing Capability

AutoCAD Mechanical has a built-in 2D meshing capability, but TD Mesher should be used instead. TD Mesher allows definition of mesh size, node properties, and surface and/ or solid element properties in one form. See the Advanced Modeling Guide provided in a separate volume. (Windows: Start > Programs > Thermal Desktop > s Manual - Meshing.)

19.3 19.3.1

Interface Menus and Toolbars

The menus can be displayed or hidden by changing the AutoCAD system variable MENUBAR. A value of 0 hides the menu bar and a value of 1 displays the menu bar. Alternatively, AutoCAD workspaces (see “workspaces” in AutoCAD Help) can be set up to display or hide menus, along with other settings. AutoCAD provides some pre-defined workspaces, available under Menu: Tools > Workspaces. Of the pre-defined workspaces: the AutoCAD Classic workspace displays the menu bar by default; the 2D Drafting and Annotation and 3D Modeling workspaces do not display the menu bar by default. To turn specific toolbars on and off, go to Menu: Tools > Toolbars. Thermal Desktop toolbars are under RADCAD53 and TDMESHER53. See Section 2.1 for more information on the menus and toolbars specific to Thermal Desktop. 19.3.2

Ribbons

AutoCAD has introduced ribbons (tabbed groupings of icons and commands at the top of the interface) as the default command interface. The has the choice of using ribbons, menus and toolbars, or a combination. For information on Thermal Desktop ribbon tabs see Section 2.2,

19-2


Figure 19-1 Ribbons and Workspace Switching

19.4

Graphics Settings

Most of the graphics settings are set automatically by Thermal Desktop. The may turn off the automatic initialization of Thermal Desktop from the Thermal > Preferences, Advanced tab. The command, 3dConfig, accessible from Thermal > Utilities > Set Graphics pulldown will bring up the dialog box shown in Figure 19-2,. Thermal Desktop will set all the values on this page, except the Dynamic Tessellation slider controls, and the Hardware/ Software Settings. If the wishes to override the items that Thermal Desktop automatically sets, the “Automatic System Graphics Configuration” checkbox from Thermal > Preferences, Advanced tab must be deselected (see “Advanced Preferences” on page 232). The Hardware/Software settings and Dynamic Tessellation sliders are accessed using the Manual Tune button on the right side of Figure 19-2. After selecting the Manual Tune button, the Manual Performance Tuning dialog appears (Figure 19-3). On this dialog, the should choose Reset to Recommended Values to allow AutoCAD to select the setting appropriate for your graphics card. The next two sections are only necessary if graphics performance is still not satisfactory. Interfacing with AutoCAD®

19-3

Figure 19-2

3D Graphics System Configuration Dialog

Figure 19-3

Manual Performance Tuning dialog

19-4


19.4.1

Acceleration

The selection here can speed up your graphics significantly, if you have a good graphics card in your system. Please note that it may also be in your best speed interests to the latest driver for you graphics card and install it on your system. Autodesk products will not allow hardware acceleration (accessed through the Manual Tune button shown in Figure 19-2) if the graphics driver is not up to date or tested for Autodesk products. 19.4.2

Dynamic Tessellation

Dynamic Tessellation is selected by a checkbox in the third of the Manual Performance Tuning dialog. In the Dynamic Tessallation , the two slider bars control a couple of items in the display of surfaces. As the slider bars are farther to the right, a curved surface will use more facets to draw the curve, making a more realistic image at the expense of memory and maybe wall clock speed. A lower value might cause a circle to look more like an octagon because of the smaller number of facets used. This is detailed in Figure 19-4.

Higher facetization Figure 19-4

8 facets

Facetization Example

In order for Thermal Desktop to show different colors on different sides of the surfaces for double sided postprocessing and showing active radiation sides, the surfaces are actually drawn twice and separated by a small distance. As you move the slider bars to the right, this distance will become smaller, and hopefully your graphics card can handle these small deviations. If the graphics show bleeding, as in Figure 19-5, you may need to move the slider bars to the left to make this distance a little bigger (issue an ‘rctouchall’ command to force the graphics to regenerate after each change). If dynamic tessellation is not on, the only way to fix the bleeding is to render the model.


19-5

Figure 19-5

Graphics bleeding problem

These sliders will also affect how many polygons will be used to draw the conic surfaces of Thermal Desktop. If your model has many offset paraboloids, ellipsoids, or elliptic cones, lowering the slider values might speed up the graphics, but also might make the surfaces appear not as smooth. The Number of tessellations to cache configures the system according to memory and performance requirements. The 3D cache stores at least one tessellation. Increasing the number improves performance, but requires more memory. When this option is set to 1, the tessellation for all viewports is the same; some objects in the drawing may be regenerated as you zoom in and out. Setting this option to 2 or more is useful when you have more than one viewport with different views.

19.5

Speed Issues (Wall Clock and U)

There are many different things that can affect the speed of the program. Issues might arise just working with a model inside of AutoCAD. Other issues might be in the generation of Cond/Cap data, while another might be radiation calculations. Finally, you might want to speed up your SINDA run. Details of each section can be found below. 1. The number one problem with speed that people find is usually due to running models across a network. Usually, the situation is that the files on the server are backed up, but the local machine is not. The best way around this problem is to leave the DWG file on the server, but to run your radiation and SINDA jobs on the local machine. This can easily be done from the Case Set Manager (see “Case Set Advanced Tab” on page 15-20). Some s have reported models running 10 times faster when run locally versus running on the server (of course this is a function of your network). 2. If the graphics seem to be slow, make sure you have hardware acceleration turned on (Section 2.8.5). Also, just turning off the display of nodes can significantly speed up the graphics (Section 2.7.3). Finally, the may wish to disable “Auto-

19-6


matic Regens” (Section 2.7.5), which should be a last resort and may cause other abnormalities, such as nodes and text sizing not updating properly. 3. Large models can slow down the Model Browser, and this can be speed up by turning off the Auto Update options (Section 2.4.4). 4. Another common slow down is when the runs out of memory (RAM) or disk space on either the DWG file disk or their own local temporary disk. The AutoCAD command, Tools > Inquiry > Status, can be used to determine if you are low on disk space or memory. There is not much you can do if you are low on memory, except to purchase more or try to reduce the size of your model. The temporary disk is used to keep track of undo buffer, and in a big model, it can get rather large fairly quickly. You can control this using the “Undo Control None” command. This will turn off undo recording. It must be entered each time you load the model, as it is not saved from session to session. 5. To speed up radiation calculations, for both radks and heating rates, make sure you have optimized the Oct Cells, especially the Subdivisions parameter. (see “OctTree Parameters” on page 10-13 and also see “Space Station Oct Tree Example” on page -23). 6. Using Radiation Analysis Groups can improve the speed of radiation calculations by limiting the ray intersection tests to only those surfaces that can possibly exchange radiation with the emitting surface. 7. If surfaces can see no other surfaces (100% view to space), adding a radiation conductor is much more efficient than shooting rays for that surface. 8. For radiation calculations, the should consider using the %error options for speed improvement (see “Automatic Error Control” on page 10-8). 9. Shoot only as many rays as necessary. 10. If you are using ors (see “ors” on page 4-74), you can speed up the Cond/Cap output by lowering the or’s Integration Intervals. For an area , the default value of 10 will generate 100 test points. If the surfaces being ed line up to have the same nodal breakdown, the exact same results may be obtained by changing the Integration Intervals to 1, and the calculations would be performed 100 times faster. 11. If you have applied to surfaces in your model using the tab on the thin shell or solid edit forms, you may find that the model runs faster by using ors (Section 4.8). 12. Speeding up SINDA runs is much more complicated. The first issue to look at is the solution method. Using the matrix solution(MATMET=1 or 2) on large models


19-7

can slow things significantly. Changing to an iterative solution (MATMET = 0) or the AMG-CG method (MATMET = 12) may be much faster. 13. Speeding up a SINDA transient solution can also been effected by the solution method described above, but also the time step used can directly affect the solution time. When using the TRANSIENT routine, CRTech always recommends that the let the program set the time step (DTIMEI=0). This method tracks the temperature changes for nodes and adjust the time steps accordingly. What can slow it down is that it will only take a time step of 1000 times the CSGMIN (capacitance/ sum of conductors). Thus, models with small capacitance and/or large conductors will slow down transient runs. The easiest thing to do for nodes with small capacitance is to edit the node, select “Override calculations by surfaces/elements”, and then make the node arithmetic. If you have large conductors in your model, the best thing to do would be to merge the nodes between the conductors and then do away with the conductor if this is possible.

19.6

Forcing the graphics to update

In wireframe mode, the can always force the graphics to update by issuing a ‘regen’ command. This command does not cause the model to regenerate if the program is currently displaying in solid shaded mode. The command to force a regen in solid shaded mode is ‘rctouchall’.

19.7

Useful AutoCAD Features

There are some basic AutoCAD options that can be used to organize models and to speed up the interaction of large models. In addition to this section, please see “Speed Issues (Wall Clock and U)” on page 19-6. 19.7.1

Groups

A group is a saved set of objects that can be selected together or separately as needed. Groups are created by choosing Thermal > Modeling Tools > Make AutoCAD Group (see Section 7.6) or typing the AutoCAD command GROUP (see AutoCAD Help). To use groups, when the command prompt reads Select Objects (or something similar) type ‘G’ or ‘Group’ and a group name will be requested. Group names can be up to 31 characters long and can include letters, numbers, and the special characters dollar sign ($), hyphen (-), and underscore (_), but not spaces. The name is converted to uppercase characters. Some commands provide a list of group names with upper case letters providing the shortest input required.

19-8


An alternative to using the AutoCAD keywords GROUP or G, the can type GRP when it is listed as an optional response (listed as [GRP] at the end of a prompt to select items). When GRP is entered, a dialog will open allowing the to select from a list of groups that exist. Groups are automatically created for meshes created with TD Mesher (see Windows Start > Programs > Thermal Desktop > s Manual > Meshing), for stray nodes (page 462), for duplicated finite elements (page 18-4), and for bad finite elements (page 18-5). For large models with groups, make sure the Tools > Options, Selection tab, Object Grouping has been disabled. When this is on, selecting in AutoCAD can be significantly slowed. 19.7.2

Layers

Layers provide visual organization of models. Between setting a color for each layer and setting the visibility of each layer in the Layer Manager (accessed by Format > Layers...), the can distinguish sections of the model from other section. Of course, large models should be set up with layers. A layer name can include up to 255 characters (double-byte or alphanumeric): letters, numbers, spaces, and several special characters. Layer names cannot include the following characters: <>/\“:;?*|=‘ The ‘CURRENT’ layer is the layer on which all new items are drawn. The current layer can be set in the Layer Manager or by choosing a new layer in the layer drop-down menu with nothing selected. The layer on which an object resides can be changed by one of three ways: • select the item(s) and choose a new layer in the layer drop-down menu • select the item(s) and choose Modify > Properties • select and right-click the item(s) in the Model Browser and choose a layer in the Model Browser layer dropdown menu.

It is important that the ‘FREEZE’ icon (a sun for thawed and a snowflake for frozen) is used to turn a layer visibility off instead of solely selecting the ‘LIGHTBULB’ icon. If the ‘FREEZE’ icon is not turned off, then when ‘ALL’ is used as the selection set, either by the or internally by AutoCAD, then items that are not visible are still considered part of the selection set. This can significantly slow down the zoom all, zoom extends, and 3dforbit commands. In addition, it is possible to have the ‘CURRENT’ layer ‘OFF’, but not frozen: this creates new objects that are not visible, which can be confusing. A layer must be set to ‘ON’ and ‘THAW’ for items on the layer to be visible. Important: Do not lock layers if those layers contain any portion of the thermal model. Layers containing only reference AutoCAD geometry may be locked to prevent modifying the geometry.


19-9

19.7.3

Undo control

By default, AutoCAD keeps an undo list for everything that has changed in your model. This is done by writing to a file in your temporary directory. With large models, this file can fill up a small disk and cause problems. Selecting Thermal > Utilities > Toggle Undo Recording disables or enables the undo command and the recording of commands to undo.

19.8

Working with External References

AutoCAD has an External Reference Manager, command XREF, that allows a model to load data from a second dwg file. The externally referenced drawing file is then part of the current drawing, but cannot be changed in the current drawing. This capability allows the to build their models in pieces and then have them assembled in a master model. Note that this capability is defaulted to off. The must turn on Loading of External References from the Thermal > Preferences, Advanced Page in order for Thermal Desktop to load this modeling information. When a drawing is externally referenced, only items that have graphical entities are loaded into the master model. This means that the following information is meaningless to the master model: Optical/Thermophysical property names and aliases, Correspondence Data, Orbits, Case Set Manager, and Symbols. The Optical/Thermophysical properties names of the externally referenced surfaces must be defined in the master model, as also correspondence data. The use of symbols between the two models is very interesting. If a symbol is defined in both the externally referenced model and in the master model, then the master model will update the externally referenced model to use that symbol value for calculations in the master model. If a symbol in the externally referenced model does not exist in the master model, that symbol must be exported from the Symbol Manager in the XREFed model and imported through the Symbol Manager in the master model. Want "Hands-On" Information? For an example of working with symbols in external references refer to the tutorial exercises “Parameterizing for a Common Input” on page -187. With the current implementation, trackers in the externally referenced model are not updated during run time. However, it is allowable to attach the externally referenced model to a tracker that is in the master model. Note that both the master model and the externally referenced models should have the same Model Length units and that the should not change the Model Length units of the master model once the external references have been loaded.

19-10


20 Tutorials This chapter presents information and sample problems to explore some of Thermal Desktop’s features. Before beginning the tutorials, AutoCAD®, Thermal Desktop, and SINDA/FLUINT must be installed. Refer to the AutoCAD and Thermal Desktop installation instructions to complete these installations if the software is not already installed. Tutorial files may be found in the \Tutorials\Thermal Desktop subdirectory which is located in the Thermal Desktop installation directory (usually C:\Program Files\Cullimore And Ring\Thermal Desktop). The \Tutorials\Thermal Desktop subdirectory includes additional subdirectories, one for each of the tutorials covered in the subsequent sections of this chapter. For some of the tutorials, a completed tutorial file is included in the \completed directory. It is recommended that the \Tutorials\Thermal Desktop directory be copied to the ’s own working area before beginning the tutorials. This ensures a copy of the original tutorial files will be available for use by other s at a later time. The tutorials attempt to introduce a variety of concepts. Most often there is more than one way to accomplish the “task” of a tutorial. Different techniques are shown in an attempt to expose the to the variety of features present in Thermal Desktop. In particular, there are many ways of selecting objects, and both noun-verb (object selection followed by a command) and verb-noun (a command followed by object selection) order is ed for most commands. Identical tasks may be performed differently to illustrate a particular feature. It is also strongly suggested that each new take the time to work through each of the tutorials (in this chapter as well as the other tutorial chapters: see “RadCAD® Tutorials” on page 21-1 and see “FloCAD® Tutorials” on page 22-1). Although completing the tutorials requires time, doing so will provide the with enough knowledge and skills to allow the to quickly become effective and efficient while working in Thermal Desktop. There are eleven tutorials in this chapter, as follows: • Section 20.1: Getting Started on page 20-5 • Section 20.2: Setting Up a Template Drawing on page 20-35 • Section 20.3: Model Browser Example on page 20-41 • Section 20.4: Simple Meshing Methods on page 20-57 • Section 20.5: Circuit Board Conduction Example on page 20-67 • Section 20.6: Beer Can Example on page 20-89 • Section 20.7: Conduction and Radiation Using Finite Elements on page 20-129 • Section 20.8: Mapping Temperatures From a Coarse Thermal Model to a Detailed NASTRAN Model on page 20-157 • Section 20.9: or Example on page 20-171

Tutorials

20-1

• Section 20.10: Parameterizing for a Common Input on page 20-187 • Section 20.11: Dynamic SINDA Example on page 20-201 Functionality in the tutorials reflects the most recent version of Thermal Desktop and AutoCAD. Every effort has been made to present the with the exact system prompts, messages and commands, and dialog boxes the will see when using Thermal Desktop. Please note that occasionally minor changes may be made between application releases and, as such, small changes to wording in such items may be experienced. The “how to” functionality will not change. Most of the tutorials are written in a two-column format. The left column provides specific instructions and the right column provides a general description of the steps, provides example views of the graphics area, and sometimes provides background information and alternative methods for accomplishing the steps. Reading through the information in the right column before completing the steps on the left will be much more beneficial to the than autonomously completing the steps in the left column. Typographical conventions used in the examples are as follows: Typographical Conventions •

Prompts, instructions and other lines of text directed to the that appear in the Command Line area are shown in Courier type font. For example: Specify base point or displacement:

•

Actions requiring input by the will be numbered and shown in italic font:

1. Pick a point or enter coordinates. •

response, meaning text which is to be typed by the exactly as shown, is written in bold Arial type font. The may be prompted to input, or type, information into the Command Line area, into a dialog box or some other text box. For example: Command:

3dface

or combined with a numbered action requiring input: 2. Type 9x in the Command line.

20-2

Tutorials

Typographical Conventions (Continued) •

Menu functions are shown in bold Arial type font. Selecting commands from a cascading menu is shown starting with the top level choice, and with a right facing angle bracket (>) used to show selections on the cascading menus: Thermal > Surfaces/Solids > Disk Thermal > Surfaces/Solids > Disk represents the clicking on, or selecting, the menu selections, beginning with Thermal, located on the menu bar:

Figure 20-1

•

Submenu Example

Dialog box names (and any included tabs), field names, button names and such will be shown in bold font as a means to differentiate them from the surrounding text.

Tutorials

20-3

Typographical Conventions (Continued) •

Pressing the Enter key is implied after entering text in response to a command prompt. The symbol <Enter> represents pressing the Enter key when no text is to be entered but an Enter key entry is required. Anytime the is required to press a function or other keyboard key, the key will be shown in bold Arial type font and enclosed in angle brackets. For example, if the is to press function key F2, the key is shown as follows: Many menu functions have the option of using shortcut keyboard commands comprised of holding down a key such as the Control (Ctrl) key and then holding down an additional key. If a menu has such an option, the shortcut keyboard command may be included in the exercise options. The keys to be pressed will be shown in bold Arial type font and enclosed in angle brackets. For example, if the wants to copy text the Copy function can be performed by highlighting the text and clicking on Edit on the menu bar followed by clicking on Copy. The Copy function has a shortcut keyboard command Control key and the C key, or:

•

When an icon exists to perform the same function as a menu command, it will be displayed as: or Thermal > Surfaces/Solids > Disk. The can either click on the icon on the tool bar or select the displayed sequence of menu picks from the menu. Note: Icons may change. If the Icon does not appear in the toolbar, use the menus. The current icon will appear next to the menu command in later versions of AutoCAD. In many instances, the has the option of selecting an icon, making a menu selection or typing in a command in the Command Line area. In these instances, the may use any of the options. If the is required to click on an icon and there is no corresponding menu choice (for example Reset Thermal Desktop Graphics), the icon and the name of the icon will be displayed: Reset Thermal Desktop Graphics icon.

•

If a command is mis-typed, press the Escape key, or <Esc>, to cancel the command in progress. Edit > Undo or may be used to undo an action. <Esc> also empties the selection set.

20-4

Tutorials

20.1

Getting Started

The Getting Started section of the tutorial is broken into several different sections. These sections are: •

Interface

•

Graphical Objects

•

Grip Points

•

Selection

•

Pan/Zoom/Rotate

•

Shading/Wireframe

•

Layers

•

Colors

This section offers information and, in some cases, instructs the to perform an action as a means to familiarize the with Thermal Desktop functions. Although Thermal Desktop can operate with AutoCAD versions 2010 and later, the instructions in this tutorial will be updated to the most recent versions of AutoCAD. If you are using an older version, please be aware that AutoCAD commands and icons may be slightly different. The proper commands and techniques for your version can be found in the AutoCAD Help under the Help menu.

20-5

20.1.1

Starting AutoCAD for the First Time

This section provides the steps to start AutoCAD for the first few times. Once a Thermal Desktop template is created (Section 20.2), these steps will no longer be followed. Instead, the Thermal Desktop template will be copied to a new directory, renamed and double-clicked to start a new model. Starting AutoCAD for the First Time At this point, you should start up AutoCAD. Important: The process described below should only be used for this tutorial and the tutorial in Section 20.2 in which a template drawing file (file extension DWG) will be created. In future tutorials or modeling when an existing model is not available, the template drawing file should be copied to the desired folder, renamed and the file icon double-clicked to start Thermal Desktop. 1. From the Windows desktop go to Start > All Programs and look for AutoDesk. Follow that to the version of AutoCAD installed, and then follow that to the executable (i.e. Start > All Programs > AutoDesk > AutoCAD 2015 - English > AutoCAD 2015 - English). Important: If it is not possible to complete the Getting Started drawing (drawing1.dwg) in the same sitting, exit Thermal Desktop (File > Exit) and respond No when prompted to save the drawing. If the files is saved, it will be saved to the ’s My Documents directory.

20-6

Starting AutoCAD for the First Time If using a trial version of AutoCAD, the Your Autodesk Trial window will open. 2. Select Continue Trial.

20-7

Starting AutoCAD for the First Time After the Your Autodesk Trial dialog or it you are not using a trial, the New Tab will be shown.

The Create option on the New Tab can be used to start a new drawing, open files, or open recent documents.

3. Select the Create option at the bottom of the New Tab.

The Learn option at the bottom has links to overviews of new AutoCAD features, videos for getting started in AutoCAD, tips, and links to online resources.

4. Select Start Drawing (highlighted below) in the Getting Started button. The window changes to a grid and the toolbars and ribbons become active.

5. Type MENUBAR command The Command line will read: Command: MENUBAR Enter new value for MENUBAR <1>:

6. Enter 1 if the value is <0>

20-8

Note: After using AutoCAD for Thermal Desktop, the New Tab will no longer be available.

Make sure the menu bar is turned on.

Starting AutoCAD for the First Time 7. Type RibbonClose if you do not think you will use the Ribbon interface.

Turn off ribbons if they are not desired.

8. Select Tools > Options

Set up the interface to replicate the tutorial descriptions.

9. Select the Display tab, and click on the Colors button at the bottom of the Window Element region.

Ribbons are a relatively new interface, but fully functional. The tutorials do not refer specifically to the ribbons, but referenced icons can be found on either toolbars or ribbons.

The Drawing Window Colors dialog opens. 10. In the Drawing Window Colors dialog, check to see that 3D parallel projection is selected in the Context field and Uniform background is selected in the Interface element field. In the Color field choose Black. 11. Select Apply & Close to return to the Options dialog 12. Select the Preferences tab.

14. Check the box Turn on time-sensitive right-click.

Right Click customization allows setting a time duration for different behaviors of a right-click. A quick right-click is the same as hitting <ENTER>; a longer rightclick opens a context menu.

15. Select Apply & Close to return to the Options dialog

You may return to this if the click duration is too sensitive.

13. Select the Right-click Customization button.

16. Choose OK to close the Options dialog. 17. Continue to the next section or select File > Exit. Do not save the file.

20-9

20.1.2

Interface

This section provides an overview of the Interface detailing icons, Thermal menu commands, and various components of the main AutoCAD window. Interface 1. Select Thermal > About Thermal Desktop... The About Thermal Desktop/RadCAD/FloCAD dialog box appears displaying version, license, author and CRTech information. 2. Select OK on the About Thermal Desktop/RadCAD/FloCAD dialog box.

20-10

The About Thermal Desktop screen displays which version of the software is installed and the available licenses.

Interface (Continued) The next image shows the opening default screen with the exception that the drawing background is white instead of black. The change in color is to facilitate the reproduction of this manual. Much of the screen utilizes standard Windows elements such as title bars, a main menu bar and various toolbars. The Thermal menu contains all of the commands specific to Thermal Desktop. The other menu options located on the menu bar contain commands created by AutoCAD, many of which are standard Windows commands. The AutoCAD commands (MOVE, COPY, ARRAY, etc.) work on the Thermal Desktop objects. As with other Windows based applications, most of the commonly used menu commands have associated toolbars and ribbons displaying icons corresponding to menu commands. Toolbars and ribbons provide quick access for the to perform a command. There is no difference between using a menu versus a toolbar icon versus a ribbon command icon to perform a command. AutoCAD shows the icon image next to the associated menu command for reference. If the toolbars are not laid out the same as in the next image, then they can be automatically be rearranged by selecting Thermal > Utilities > Reset Thermal Desktop Toolbars. The toolbars are arranged in three primary groups. •

The toolbars located at the top of the screen are used to open and save files, manipulate layers, and also to rotate/pan/zoom a view.

•

The toolbars located on the left of the screen are used for creating entities such as surfaces, nodes, lumps, and conductors.

•

The toolbars located on the right of the screen are used to edit the entities, as well as change what is viewed such as active sides and postprocessing.

To learn what each icon on the toolbar represents, position the cursor over an icon and then stop moving the mouse. When the cursor pauses over an icon a short text description called a tool tip is displayed. At the lower left of the graphics area, the UCS ( Coordinate System) icon is displayed. All points input and displayed are in this coordinate system. Also located at the left end of the status bar, the current location (X, Y or X, Y, Z) of the cursor in the drawing area is displayed. As the cursor is moved in the drawing area, the coordinates change as the cursor moves. This update can be very handy when selecting points on the screen. If the current location is not visible, select the Customization icon at the far right of the status bar to select Coordinates and place a check mark beside it.

20-11

Interface (Continued) The primary features of the AutoCAD interface are labeled below.

At the bottom of graphics area is the Command Line area. The Command Line area is where the will be prompted to type in a command or value(s) and will see various messages. While the can only see a few lines of text in that area, the entire text area can be displayed by selecting the function key. 3. Press to open the AutoCAD Text Window. Use this window to view commands and actions performed in the current session. 4. Press again to close the window. 5. Continue to the next section or select File > Exit.

20-12

20.1.3

Graphical Objects

This section introduces the to some general Thermal Desktop functionality. Graphical Objects

1.

or Thermal > Surfaces/Solids > Rectangle.

Create an arbitrary rectangle by following the steps noted to the left. The actual coordinates are not important.

If not continuing from Section 20.1.2, the Thermal Desktop splash screen will appear briefly. The Command Line area should show: Command: _RcRectangle Origin point <0.0.0>:

2. Click on a point in the lower left of the graphics area with the cursor. The Command Line should now read: Point for +X axis and X-size <@1,0,0>:

3. Click on a second point in the graphics area that is to the right of the first point. The Command Line should now read: Point to set XY plane and Ysize <@0,1,0>:

4. Click on a third point in the graphics area that is above the second point.

20-13

Graphical Objects (Continued) The Thin Shell Data dialog box appears.

When viewing the rectangle, solid lines are visible around the outside of the rectangle. In the center of the rectangle is a node. Dashed lines from the sides of the rectangles to the nodes may or may not be visible. These dashed lines represent the node locations.

5. Select OK to close the dialog box without making any changes. The rectangle is created.

Dashed Lines

Solid Lines

Node

Rectangle

20-14

Graphical Objects (Continued) The rectangle in the previous figure has arrows pointing to dashed lines and solid lines. The solid lines represent the node boundaries, and the dashed lines represent the lines that the node resides on. The can control how many dots are used to represent the dashed lines with the LTSCALE command. When the geometry gets very complicated, sometimes it is easier to select a surface by clicking on the dots instead of a boundary that might be shared with another object. 6. Type LTSCALE. The Command Line should now read: Enter new linetype scale factor <1.0000>:

7. Type .1

The number displayed in < > after issuing the command LTSCALE is the current setting for LTSCALE. Entering a smaller number will display more dots in the dashed lines, while a larger number will display less dots.

Finer dashed lines, or possibly solid lines, appear. 8. Press + to undo the previous command. 9. Continue to the next section of select File > Exit.

20-15

20.1.4

Selecting Objects

The selection of objects is probably the most difficult new things to learn when working with a CAD program and this section introduces the to some basic selection concepts. Selecting Objects There are two types of selection, pre-selection and post selection. •

Pre-selection means one or more objects are selected and then a command such as Thermal > Edit or Modify > Copy is issued.

•

Post selection means a command is issued and then one or more objects are selected. The cursor changes shape (box) and waits for the to select the object(s) to be affected and click the right mouse button. Post selection only works if nothing is selected when the command is issued.

The easiest way to make sure nothing is selected before you issue a command is to press the <Esc> key. Pressing <Esc> resets the selection set to empty. The cursor appears as crosshairs with a square box, called a pick box, in the middle of the cursor (Figure 20-2). When selecting an object, position the cursor over the object to be selected. The item positioned within the pick box is the object that will be selected when the mouse is clicked. In the left-most graphic in Figure 20-2, the pick box includes the solid line of a rectangle. When the cursor is paused over an object, the tool tip associated with the object displays what is being selected. The tool tip shows the type of object, followed by an object identifier (format ::OBJECTID), which is unique for each object. Note: If the tool tip does not display, select Tools > Options, Preferences and make sure Display hyperlink tool tip is selected (check mark in the box). Pick Box

Figure 20-2

20-16

Getting Started - Selection

Selecting Objects (Continued) Still working with Figure 20-2, the cursor is positioned over the node which is located in the center of the rectangle. If the cursor is positioned over the node as shown in the center graphic, the tool tip still shows the rectangle as the object. This is because there are two items in the pick box - the dashed lines of the rectangle, and the solid circles from the node. The has limited control over what will be selected in this instance; however, there are two methods to assist in choosing the node: •

Method 1: The easiest method is to move the cursor so that only the node is in the pick box, as shown in the example on the farthest right of Figure 20-2.

•

Method 2: The second method is to select the rectangle, then select Tools >Draw Order > Send to Back. This will make the node be the first item selected when the rectangle and the node are both in the pick box.

A third option is to hold down the key, which will cycle through the items available in the pick box. This last option can be tedious and difficult to learn, but may be worth investigating as an additional option. The size of the pick box can be changed by typing PICKBOX in the Command line and changing the input value. The default value of 3 is often too small to be useful; many s prefer a value of 5.

20-17

Selecting Objects (Continued) Selecting objects one at a time is a common occurrence but there are many times when a may want to select multiple objects without having to go through the process of selecting and performing an operation on each object individually. Multiple objects can be selected by first drawing a box around a group of objects to select them and then performing an operation. As simple as this seems, there are different methods to encircle some objects for selection, and these different methods will result in different selection sets. Using Figure 20-3 as an example, consider the situation of an Ellipse, Disk, and a Line. Drawing a box around all three of the items will result in all three being selected. The selection set will be different when the drawing box crosses the lines of some of the items as shown with the solid black line. If the draws the box by picking in the upper left corner (#1) and then dragging the cursor to the lower right (#2), only the items completely enclosed in the box will be selected. In this case, only the disk is selected. If the reverses the order of drawing the box by picking in the lower right corner (#2) first and then dragging the cursor to #1, the items enclosed and crossing the box will be selected. In the case shown in Figure 20-3, all three of the of the items will be selected. In newer versions of AutoCAD, the selection windows are color coded to graphically show which method is being performed. The drawing box method of selecting multiple objects will be used in a future tutorial, the “Beer Can Example” on page 20-89. 1

2 Figure 20-3

20-18

Getting Started - Selection Boxes

20.1.5

Grip Points

This section demonstrates the use of grip points to edit a surface. Grip points provide for an immediate, interactive editing of objects without entering any specific commands on the command line. Grip editing is a quick and easy way to modify an object. Grip Points When an object is selected with the mouse, the object’s solid lines change into dashed lines and grip points (small squares) become visible. The grip points’ color is often preset to blue, but the can control grip point color and size by selecting Tools > Options... and then clicking on the Selection tab in the Options window, as shown below.

Figure 20-4

Options Selection Tab

Grip points provide an easy form of editing an object.

20-19

Grip Points (Continued) 1. If not already opened, launch Thermal Desktop and create a rectangle as shown in the previous tutorial (See Section 20.1.3). 2. Select the rectangle by positioning the cursor on one of the solid lines making up the rectangle and clicking the mouse or using one of the other selection methods previously discussed. The solid lines change to dashed lines and grip points appear at various points on the rectangle. 3. Position the cursor over one of the grip points. A tool tip associated with the grip point appears. The tool tip at the end of the X axis of the rectangle (Stretch X Length) instructs the that the grip point associated with that tool tip can be used to modify the X length the rectangle. 4. Click on the grip point referenced above. The grip point changes color and the tool tip Endpoint is displayed. 5. Move the cursor to a new point and click the mouse button. 6. View the change to the rectangle. 7. Select key to cancel the operation. The rectangle returns to its previous dimensions and is deselected (no longer highlighted). Note: The <Esc> key can also be used to deselect an object. 8. Position the cursor on the other grip points on the rectangle to view the tool tip messages.

20-20

20.1.6

Pan, Zoom, Rotate, and Views

This tutorial demonstrates several methods of changing the view on the screen. The view can be zoomed in for a closer view of an object or zoomed out for a higher-level view of an object or a drawing. The view can be panned - moved to the right, left, up or down - which will change the view on the screen without zooming in or out. Changing the view of a drawing or a specific object can assist the when creating detail Pan, Zoom, Rotate & Views 1. If not already opened, launch Thermal Desktop and create one or more rectangles as shown in the previous tutorial (See Section 20.1.5. 2.

or View > Zoom > Realtime. The cursor changes into a small magnifying glass with plus and minus signs.

3. Hold down the left mouse button and drag the cursor up and down on the screen. The view zooms in (closer, more detail) as the mouse is moved upward; the view zooms out (higher level, less detail) as the mouse is moved downward.

Use this function to enlarge the view of the objects on the screen (zoom in, more magnification) or decrease the view of the objects (zoom out, less magnification) as the mouse is moved. Note that when Zoom Realtime is initiated, the message Press ESC or ENTER to exit, or right-click to display shortcut menu appears in the Command Line area.

to look to the Command Line for additional information.

4. When finished, right mouse click and select Exit from the drop down menu

20-21

Pan, Zoom, Rotate & Views (Continued) or View > Pan > Realtime.

5.

The cursor changes into a small hand. 6. Hold down the left mouse button and drag the cursor up and down and sideto-side on the screen. The view moves with the mouse movement. 7. Leave the view on the screen so that some of the object(s) are off to the side or not visible.

Use this function to change view on the screen by shifting the view up, down, left or right as the mouse is moved. The magnification level does not change (no zoom in or zoom out occurs). Note that when Pan Realtime is initiated, the message Press ESC or ENTER to exit, or right-click to display shortcut menu appears in the Command Line area.

to look to the Command Line for additional information.

8. When finished, right mouse click and select Exit from the drop down menu 9.

or View > Zoom > Extents.

10. All of the objects in the drawing are moved back into view.

20-22

Adjusts the view on the screen by zooming out until all objects in the drawing are visible. Use this function if a desired object or area is no longer within the drawing area because of panning or zooming. This function will quickly return to the complete drawing area view.

Pan, Zoom, Rotate & Views (Continued) 11.

or View > Zoom > Window. The Command Line should now read: Specify first corner:

12. Pick on a point that represents the first corner of a rectangle that will enclose the object/area to be viewed. The Command Line should now read: Specify first corner: Specify opposite corner:

13. Pick at point diagonally opposite the first point. A box is drawn around the area between the two points as the cursor moves. The view zooms in around the area enclosed by the rectangular area defined by the two points.

14.

or View > Zoom > Previous.

Use this function to quickly zoom into a specific object or area in the drawing area. This command has the define a rectangle that encloses the portion of the drawing area to be viewed by clicking on two points. The first point sets a first “corner” of the rectangle; the second point determines the opposite, diagonal corner of the area to be viewed. Note that when zooming and panning, the right mouse menu can be used to switch between the various zooming and panning options. When Zoom > Window is initiated from the right mouse menu, the cursor changes into a pointer with a filled-in box attached to it. Move the cursor to a point that represents the first corner of the area and then hold down the left mouse button and drag the mouse to a point opposite the first point. Release the mouse button. Use this function to quickly return to the last zoom/panned view.

The view in the drawing area returns to the last view in the zoom/pan sequence. 15.


20-23

Pan, Zoom, Rotate & Views (Continued) 16. Select View > Orbit > Free Orbit. An arcball appears on the screen. The arcball is a large circle, or sphere. There are four smaller circles at the quandrant points on the arcball. 17. Position the cursor within the arcball. The cursor changes shape (two arrows circling a sphere). 18. Hold down the left mouse button and drag the cursor within the arcball. The rectangle (or all objects if more than one object is in the drawing area) rotates in all directions. Watch the USC icon as the cursor moves. 19. Release the mouse button. Position the cursor outside of the arcball. The cursor changes shape (an arrow circling a sphere). 20. Hold down the left mouse button and drag the cursor outside of the arcball. The object rotates around an axis a the center of the arcball. 21. Release the mouse button. Position the cursor on one of the quadrant circles. The cursor changes shape (an arrow elliptically circling a sphere). This function changes the vertical and/or horizontal rotations of the object(s). 22. Hold down the left mouse button and drag the cursor away from the quadrant circle.

20-24

Older versions of AutoCAD may be View > 3DOrbit Use this function to 3D view a drawing. When activated, a circle, called an arcball, appears around the selected object. As the selects points and drags the mouse, the view of the selected object changes so the sees the object from whatever angles the desires (vertically and horizontally). In the Thermal Desktop toolbar at the top of the graphics area the icon appears as this:

Pan, Zoom, Rotate & Views (Continued) The object(s) rotate around a horizontal axis (if the 3:00 or 9:00 quadrant circles were selected) or around a vertical axis (if the 12:00 or 6:00 quadrant circles were selected).

A special right mouse menu offers additional 3D view options.

23. Release the mouse button and right click to display the right mouse menu.

The 3D Orbit right mouse menu offers zoom and pan commands as well as various shading and projection commands. Depending on the version of AutoCAD, clicking on Other Navigational Modes displays the submenu shown above 24. Press <Esc> three times to close right mouse menus and end the 3d Orbit command.

20-25

Pan, Zoom, Rotate & Views (Continued) 25. Click on View > 3D Views. The 3D Views submenu appears.

Many predefined orientations are shown under View > 3D views. Note: If the view on the screen is not what is desired, try changing the view using the 3D View commands. The can also create and store views of particular areas of a drawing with the View > Named Views function. View > Named Views allows a to quickly return to a previously designated area of a drawing.

26. Select View > 3D Views > SW Isometric. The view of the rectangle (or other objects in the drawing area) changes to that of the SW Isometric view. 27. Select View > 3D Views > Top. The view of the rectangle (or other objects in the drawing area) changes to a top-down view of the rectangle. 28. Repeat additional view options as desired.

20-26

The flyout toolbar under the icon provides the same options as View > 3D views. A flyout toolbar expands to show more options when the left mouse button is held while pointing at the icon. While still holding the mouse button, the cursor can be moved to the desired icon.

20.1.7

Shading/Wireframe Views

The Shade and Wireframe View commands offer the different ways to view models. Since AutoCAD versions vary and Thermal Desktop Shading & Wireframe Views The Shade and Wireframe commands are found under the View > Visual Styles submenu.

The first three items in the sub-menu are 2D Wireframe, 3D Wireframe, and 3D Hidden. The differences between the three can be subtle, but each has its uses. • 2D Wireframe. Displays the objects using lines and curves to represent the boundaries. Raster and OLE objects, linetypes, and lineweights are visible. Thermal Desktop network icons (nodes, lumps, paths, etc) are shown as 2D icons. • 3D Wireframe. Displays the objects using lines and curves to represent the boundaries. • 3D Hidden. Displays the objects using 3D wireframe representation and hides lines representing back faces.

20-27

Shading & Wireframe Views (Continued) In the menu, there are two options for shaded modes: Realistic and Conceptual • Realistic. Shades the objects and smooths the edges between polygon faces. Materials that you have attached to the objects are displayed. • Conceptual. Shades the objects and smooths the edges between polygon faces. Shading uses the Gooch face style, a transition between cool and warm colors rather than dark to light. The effect is less realistic, but it can make the details of the model easier to see. Toolbar icons exist for the 2D/3D Wireframe, Hidden, and Thermal shaded commands. The Thermal shaded option is similar to the Realistic mode.

These icons are in the top toolbar as shown above. The last option in the Visual Styles menu is the Visual Style Manager. In the Visual Style Manager, the is able to make changes to the visual styles or create new visual styles. Figure 20-5 is an example of an object viewed using the six visual styles.

Figure 20-5

20-28

Visual Style Examples

20.1.8

Layers

Everything in a drawing is associated with a layer. Layers are separate drawing areas, one on top of the other, and are used to organize and to manipulate what is currently being viewed in a model including the color, linetype and lineweight of an object. Layers are a good way to group and display related objects in a drawing or model. When a new object is created, it is placed on the Current layer. Layer properties such as Name and Color are managed in the Layer Properties Manager dialog box and by using additional Layer controls. Figure 20-6, shown below, shows the Layer Properties Manager (the Layer Properties Manager changes from version to version of AutoCAD, but the basic behavior remains the same). Note: Changing object layers will also be discussed later in the tutorial. Every model has a layer numbered 0. Layer 0 is system generated and cannot be deleted or renamed. When a model is first created, Layer 0 is also created and unless the creates one or more additional layers, Layer 0 is the layer where the model components reside. With some versions of AutoCAD, another layer, named ASHADE, is system generated and internally used by AutoCAD to control the lighting settings of the objects. The ASHADE layer is locked by default and entities cannot be changed or added. Locking a layer, which prohibits entities residing on that layer from being created or changed, is discouraged (other than the ASHADE layer which is locked by default). Important: CRTech highly recommends that the does NOT lock layers, as odd results can occur. The next section of this tutorial creates a new layer and familiarizes the with the Layer Properties Manager. Layers 1. If not already opened, launch Thermal Desktop and create one or more rectangles as shown in the previous tutorial (See Section 20.1.5 if no objects are displayed. Create a new layer to be called layer1. 2.

or Format > Layer or type layer in the Command Line.

20-29

Layers (Continued)

Figure 20-6

Getting Started - Layer Properties Manager

The Layer Properties Manager appears. The two default system-generated layers (0 and ASHADE) are listed. 3. Select

New layer.

A new line is added for Layer1.

20-30

The ASHADE layer may or may not exist. Note: A layer (other than layers 0 and ASHADE) may be renamed by double clicking on the layer name to highlight it and then typing in a new name.

Layers (Continued) 4. Select the sun icon in the Freeze column of the new Layer1. The icon changes to a snowflake and the visibility of the layer is turned “Off”. 5. Select the color icon for layer 0 (currently White). The Select Color dialog box appears. 6. Select Red from the standard colors selections to change layer 0’s value to red (top left). 7. Select OK to close the Select Color dialog box. The Color value for layer 0 should display the color red in the Layer Properties Manager.

Next to the layer names are two columns with icons underneath the column headings. •

On

•

Freeze

These icons control whether or not objects residing on a layer are displayed or not. If either the On or Freeze icons are off (darkened light bulb icon or a snowflake icon), then objects on the layer will not be seen graphically. If the Freeze icon is Off (snowflake), then the layer will not be included when ALL is entered at the select prompt. In addition, the current layer cannot be frozen.

Since new objects are created on the current layer, turning the current layer off 8. Close the Layer Properties Manager. can lead to confusion (newly created objects are not visible). To avoid this confusion, CRTech recommends only using the Freeze option. When the Layer Properties Manager is closed, note the objects within the graphical screen have changed to red.

20-31

Layers (Continued) 9.

or Modify > Properties or type properties in the Command area or press <1>.

The Properties dialog box appears. Note that layer 0 is the current layer. 10. Select a rectangle on the screen. The Properties dialog box changes to show a list of property values.

20-32

Every graphical object has properties associated with them, some of which can be changed by the as needed. Double-clicking on an object will bring up the Properties window. This form is modeless and can stay up while working in AutoCAD.

Layers (Continued) The Layer drop-down list, shown to the left, can be used to manipulate some layer properties without having to open the Layer Properties Manager. These properties are: 11. Click on the Layer drop-down list arrow and select Layer1. A Thermal Desktop/AutoCAD dialog box appears stating 1 object changed to a frozen or off layer and removed from the selection set. that Layer1 was turned off in the Layer Properties Manager. 12. Click OK to close the Thermal Desktop/AutoCAD dialog box. The rectangle is no longer visible on the screen. 13. Click on the Layer drop-down list arrow and select the Freeze icon (snowflake) for Layer1 and click anywhere in the Graphics area. The rectangle previously selected reappears on the screen, in grey, Layer1’s default color. 14. Click on the X in the upper right corner of the Properties dialog box to close it.

•

Turn a layer On of Off

•

Freeze or thaw in ALL viewports

•

Freeze of thaw in current viewports

•

Lock or Unlock a layer

•

Color of layer

Once a rectangle is selected, the Color box says ByLayer. Color of an entity can be changed from the By Layer setting. When changing to a different layer, if no objects are selected and the layer is changed, then the selected layer becomes the new “current” layer and all new objects will be created on that layer. If one or more objects are selected when the layer is changed using the Layer dropdown list then the selected objects will be moved to the newly selected layer. When an object’s layer is changed to one that is not visible, a warning dialog box appears. Also note that when the layer of an object is changed, the node associated with the object is also changed to the same layer.

20-33

20.1.9

Colors Colors

Just as each object has a layer associated with it, an object has a color associated with it. The default color for objects is set to Color By Layer, which means the color of an object is the same as the color defined for the layer, which is set in the Layer Properties Manager, from the Layer drop-down list or by clicking on Color in the Properties dialog box and selecting a color from the drop-down list as noted above in Section 20.1.8.

If changing an object’s color using the Color drop-down list, shown above, while no objects are selected, then the color selected will become the default color for newly created objects. If an object’s color is changed using the Color drop-down list when one or more objects are selected, the selected items will be assigned the newly selected color. All colors that are assigned to objects will be overridden by Thermal Desktop when postprocessing of data or Display Active Sides is performed. 1. Select File > Exit. A Thermal Desktop/AutoCAD dialog box appears asking to save the drawing changes. 2. Select No.

20-34

Exit Thermal Desktop and respond no when prompted to save the drawing.

20.2

Setting Up a Template Drawing

Prerequisites: • Section 20.1.1: Starting AutoCAD for the First Time on page 20-6 A template is something that can be used as a pattern for future projects. In Thermal Desktop, a template drawing file can be created and used as a foundation for creating new Thermal Desktop drawings. When creating a template it is important to consider how the template will be used, if the template will be used by only one or by more than one, what preferences should be associated with the template and other such considerations. In this tutorial, a template drawing file (file extension DWG) will be created. The template will be used in other tutorials and may be used as a permanent template for use in future thermal analysis tasks. Creating and then using a template drawing starts a new analysis with an environment set to pre-determined preferences, helping to keep drawings consistent and easy to use. Before beginning a new thermal analysis task, copy the template to a directory chosen to store the work for that particular task or project. Once a copy is made, rename the copied template drawing to a meaningful name related to the thermal task or project. Then start Thermal Desktop by double-clicking on the renamed drawing file. Thermal Desktop will start with the preferences saved in the template, and the current working directory will be set to the directory which contains the drawing file. All database files created by Thermal Desktop for view factors, radiation conductors, and orbital heating rates will be placed in this working directory. Important: If Thermal Desktop is not launched using the drawing file for the task and instead is started from the AutoCAD Icon, the current working directory will be the ’s Documents directory. This is not recommended, since multiple thermal analysis tasks may overwrite each other’s database files.

20-35

The following exercise creates a template file with preferences set to be convenient for working with Thermal Desktop. The is encouraged to experiment with the available options as a means to discover which settings are compatible to each individual’s work style. Setting Up a Template Thermal Desktop Drawing File Important: The process described below should only be used when creating a template. In future tutorials or modeling when an existing model is not available, a template drawing file should be copied to the desired folder, renamed and the file icon double-clicked to start Thermal Desktop. 1. From the Windows desktop go to Start > All Programs and look for AutoDesk. Follow that path to the version of AutoCAD installed, and then follow that to the executable (i.e. Start > All Programs > AutoDesk > AutoCAD 2015 - English > AutoCAD 2015 - English). Important: If it is not possible to complete the Getting Started drawing (drawing1.dwg) in the same sitting, exit Thermal Desktop (File > Exit) and respond No when prompted to save the drawing. To save the drawing, perform a Save As and change to a directory other than the initial system installation directory.

20-36

Setting Up a Template Thermal Desktop Drawing File The AutoCAD window appears.

2. Select Start Drawing under Getting Started to the upper left. The window changes to a grid and the toolbars and ribbons become active. 3. View > Visual Styles > 2D Wireframe to be consistent with the images in this tutorial. 4. Select File > Save As. The Save As dialog box appears. 5. Change the name of Drawing1.dwg to thermal in the \Tutorials\Thermal Desktop\template directory. 6. Select Save. The drawing area title bar is updated with the new name of the drawing, thermal.

Be sure to save the files in your copy of the tutorials directory. This save occurs before any Thermal Desktop commands are issued, therefore, at this point the file is strictly an AutoCAD file. Note: If your models may be opened with older versions of AutoCAD, choose the oldest version that may be used. Thermal Desktop works with AutoCAD versions 2010 through 2015.

20-37

Setting Up a Template Thermal Desktop Drawing File 7. Select Thermal > Utilities > Reset Thermal Desktop Toolbars to position the Thermal Desktop toolbars in the default locations. If you did not turn off Ribbons 8. Select View > 3D Views > SW Isometric. Note the UCS icon changes to reflect the view. 9. View the menu selection View > Display > UCS Icon> Origin.

The Thermal Desktop splash screen will be displayed when the first Thermal Desktop command is issued. If the model had contained Thermal Desktop objects, the splash screen would have appeared upon opening the file. Selecting View > 3D Views > SW Isometric here sets an isometric view parameter for the new template. This step locates the Coordinate System (UCS) display icon at the origin of the model, rather than in the lower left corner of the screen. If the UCS origin is not in the display area, then the UCS will be displayed in the lower left corner.

10. Confirm a check mark is displayed next to Origin. If a check mark is not visible, click the left mouse button on Origin to select the option. Otherwise left-click anywhere in the Thermal Desktop window or hit <ESC>.

20-38

Setting Up a Template Thermal Desktop Drawing File Note: Depending upon the experience and preferences of the , before saving the template, the toolbars on the main window may be rearranged. The default arrangement of toolbars reduces the horizontal size of the viewport. The toolbars can be moved to the sides of the viewport to produce a viewport with a more square aspect ratio. Put the cursor in the border of a toolbar (not on an icon) and drag by holding the left mouse button down to the new desired location. As a toolbar is moved close to the sides of the screen, the toolbar will automatically dock to that side. The Tools > Toolbars menu choice may also be used to turn specific toolbars on and off and to customize the icons in each toolbar. See the AutoCAD help for more details on customizing toolbars. It is helpful to remove infrequently used icons from some toolbars so that more toolbars may be displayed without taking up excessive screen space. 11. Select File > Exit. 12. Select Yes to save changes to the template file.

Exiting AutoCAD without preceding the Exit command with a Save will bring up a dialog to save any changes.

Thermal Desktop is closed. 13. Open the \Tutorials\Thermal Desktop\template folder and examine the contents. The \completed folder is a folder included in many tutorials. It contains the dwg file the would have after finishing the tutorial. The thermal.bak file is a backup file generated by AutoCAD when a Save is performed. This file is the previous version of the template.dwg file and can be used by changing the extension from .bak to .dwg in case the main file becomes corrupted. The thermal.dwg file is a model file created by this tutorial. This file will be copied to other directories to use as a starting point for other tutorials and models. The RcOptics.rco and TdThermo.tdp files are optical and thermophysical properties databases, respectively, that are automatically generated if they do not already exist, unless the model points to databases in specific file locations. As experience is gained with Thermal Desktop and some of the advanced options, create a new template file with different parameters such as Units and other property settings as desired. A may find it useful to have a template for SI units, a template for English units, or possibly a template with certain default settings (Section 2.6). Important: to create a copy and rename the template before beginning design work to insure the template remains in its original state. If copying the template at the directory/file level by dragging the template file icon to the working directory, be sure to hold the key down when dragging the template drawing file so that the template is copied, rather than moved.

20-39

20-40

20.3

Model Browser Example

What will be learned: • Overview of Thermal Desktop’s Model Browser • Using Model Browser for model troubleshooting and review • Asg material properties • Basic use of Case Set Manager • Obtaining steady state and transient solutions in Thermal Desktop • Calculating heat flow between submodels • Using Model Browser to create XY plots against time for transient solutions This tutorial demonstrates some of the capabilities of Thermal Desktop’s Model Browser. The example model for this tutorial is very simple but the capabilities extend very well to larger models. Want to Learn More? Refer to Section 2.4 "Model Browser" on page 2-8 in the ’s Manual for more information about the Model Browser. Model Browser Example 1. Double click on the file ModelBrowser.dwg located in the Tutorials\Thermal Desktop\ModelBrowser folder. Thermal Desktop opens with the ModelBrowser drawing on the screen.

Figure 20-7

Initial View

20-41

Model Browser Example (Continued) Thermal Desktop’s Model Browser can be used to view information about a model. A modeless window (can be resized and minimized) will list model data based on the type of data to be selected. The window is divided into two frames: the tree frame and the output frame. The default is to list by Submodel and ID. The may select the List menu (within the Model Browser window) to see what types of objects are available for listing. The choices are: • Submodel.Id • Non Graphical Objects • Analysis Group • Optical Props • Thermo Props • Surfaces/Solids • /ors/TECs • Assemblies/Trackers • Grip Manipulators • Conductors • Heaters • Heatloads • Orienters • Pressures • Measurement Points • Fluid Submodel.Id • Paths • Ties • Pipes • Macros • Rotation Axes • IFaces • FTies • Heat Exchangers • CAPPMP’s • TD Direct Importers • Meshers/Mesh Importers • Mesh Displayers/PP Mapper/BCM/Cutting Plane • Symbols • Domain Tag Sets • Groups • Layers The can manipulate the AutoCAD graphics by simply making the graphics window active by clicking anywhere on the main AutoCAD window and then performing operations in that window.

20-42

Model Browser Example (Continued) The can determine what has been selected by looking in the output frame. The output frame will detail how many items have been selected and their type, the visibility state, the layers that the objects reside on, and additional data for the selected items. As items in the tree frame of the Model Browser are selected, additional information about the selected item is displayed in the lower portion of the window. Single clicking on an item results in high-level information about the selected item being displayed below. Double clicking on an item in the tree frame of the Model Browser results in the expansion of the data tree and more detailed information being displayed. Display the Model Browser. 2.

or Thermal > Model Browser. The Model Browser window appears on the left side of the screen.

20-43

Model Browser Example (Continued) 3. Click on AAAA in the tree. The display at the bottom of the Model Browser changes to show only the objects associated with AAAA.

As individual items, in this case submodel AAAA, are selected, the display area at the bottom of the Model Browser changes to reflect the components of the selected item. AAAA includes: 5 objects selected 2 TD/RC Nodes 1 Node 1 boundary 1 surface 1 conductor

4. Click on Submodel Node Tree located in the main view area of the Model Browser.

When Submodel Node Tree is selected, all of the submodels that make up the model are selected, and displayed in the output frame. 20 objects selected 11 TD/RC Nodes 1 Node 1 boundary 4 surfaces 3 conductors 1 heat load

Note: Use the scroll bar on the right of the output frame to view additional data.

20-44

Model Browser Example (Continued) 5. Double click on submodel AAAA.

The tree expands to show nodes, 1, 2, and 3 are associated with submodel AAAA. An ‘A’, ‘B’, or ‘D’ in the icon for a node indicates that the node is Arithmetic, Boundary, or Diffusion, respectively. A circle indicates that the node definition is obtained from a surface, solid, or finite element.

6. Select node 1. Note: Select by single clicking on the object with the left mouse button. 7. Double click on node 1. Note: Expand by double-clicking on the object with the left mouse button or click on the plus sign (+) to the left of the object. 8. Select AAAA.1::47.

Node 1 consists of a node and a surface.

The tree underneath node 1 is expanded and AAAA.1::47 is displayed. Items with the symbol :: (double colon) means the item is a graphical entity. The numbers after the :: are unique for each entity. Only the node is selected. The surface, that is below the node is not selected. When an item with a :: is selected the objects below it in the tree are not selected, by default When an entity without a :: is selected all the objects below it are selected.

9. Double click on AAAA.1::47.

The submodel AAAA tree expands again and Rect::45 is displayed underneath AAAA.1::47.

10. Select Rect[MAIN]::45.

The output frame of the Model Browser shows that Rect[MAIN]::45 is a surface and is the only object selected. MAIN is the name of the submodel for the surface’s conductors

20-45

Model Browser Example (Continued) Being a separate window, the Model Browser has its own title bar, menu bar, tool bar icons and Windows control buttons. If items are selected in the Model Browser tree, use the icons and menus in the Model Browser.

11.

or Edit > Edit on the Model Browser menu bar.

Alternatively, you can right-click on Rect[MAIN]::45 and select Edit from the menu that appears

The Thin Shell Data dialog box appears. Note: The Thin Shell Data dialog box for Rect::45 can also be displayed by double clicking on it. 12. Select the Surface tab. 13. Type Fred in the Comment field, as shown below. 14. Select OK. 15.

to rebuild the data tree and deselect the Rect-Fred::45.

Once the comment is added and OK selected, the tree “flashes” and rebuilds itself. The rectangle is renamed RectFred::45, incorporating the comment that was entered. This rebuilding capability is controlled with the Model Browser Options > Auto Update command. The Auto Update feature is useful with small models, but as models become larger, this can be time consuming. Deselect Rect-Fred::45 and rebuild the tree.

20-46

Model Browser Example (Continued)

16. Select submodel AAAA. 17.

or Display > Only on the Model Browser menu bar.

Notice the graphics in the main Thermal Desktop drawing area change to show only the selected items. Submodel AAAA is in the lower left of the drawing area. Note: It may be necessary to move the Model Browser out of the way, or minimize it, to view the drawing area.

20-47

Model Browser Example (Continued) 18. Select Submodel Node Tree.

The message Mixed Visibility for selected items appears in the lower portion of the Model Browser. This message means that although there are many objects in the submodel node tree (and they are listed in the output frame) some of the selected items in the drawing area cannot be seen by the (not visible).

19. Select Rect-Fred[MAIN]::45.

This turns visibility off for the selected item.

20.

21.

or Display > Turn Visibility Off in the Model Browser. or Display > Undo Turn Visibility Off in the Model Browser

All the entities in the drawing are now visible. This command turns on the visibility of all items on visible layers.

22. 23. Select submodel BBBB.

24.

This reverses the last visibility change. If the last change was to turn the visibility of an object(s) on, then after this command the visibility will be turned off or vice versa.

or Display > Turn Ids On on the Model Browser menu bar.

25. Right-click on submodel BBBB and select Send Selection Set to AutoCAD.

20-48

The lower portion of the Model Browser changes to display the objects associated with BBBB. The node IDs are displayed for submodel BBBB.

Notice all items is submodel BBBB are highlighted.

Model Browser Example (Continued) 26. Right-click on submodel CCCC and select Send Selection Set to AutoCAD.

The CCCC submodel is highlighted in the drawing area and grip points are displayed.

Also, these entities are now an AutoCAD selection set, so any command issued outside of the Model Browser, such as Modify >Move, will function on these selected objects after making the main window active. This capability is very powerful, but can be slow as models get larger. Because of this, the default setting is Off. 27. Hold down and select submodel DDDD.

Both CCCC and DDDD are selected.

28. Right-click on submodel DDDD and select Send Selection Set to AutoCAD.

20-49

Model Browser Example (Continued) 29.

or Thermal > Case Set Manager on the main Thermal Desktop menu/toolbar. The Case Set Manager dialog box appears.

30. Select Case Set 0 if it is not already selected. 31. Select Run 1 Selected Case. A Thermal Desktop/AutoCAD dialog box appears with a message stating Thermophysical Property DEFAULT has not been found.

The Case Set Manager is the link between Thermal Desktop and SINDA/FLUINT. Under the Case Set Manager, the can define different solution sets for the model. Case Set 0 is a simple steady-state analysis. After a Case Set is defined and selected, the Run Case button writes out the SINDA/FLUINT input files, starts SINDA/FLUINT and brings the results back into Thermal Desktop for postprocessing. The Thermophysical Property DEFAULT is used only as a name place holder and does have property definitions. Therefore, the model definition is incomplete and the solution cannot be started.

32. Select OK to close the dialog box. 33. Select List > Thermo Props on the Model Browser menu bar.

20-50

This command rebuilds the Model Browser so that objects are listed by the materials that use them.

Model Browser Example (Continued) 34. Click on the + next to DEFAULT -> to expand the list.

Change the DEFAULT material to Stainless Steel.

35. Right-click on Rect[MAIN]::62 (the surface) and select Edit. The Thin Shell Data dialog box appears. 36. Click on the Cond/Cap tab.

37. Click on the arrow next to the Material field and select Stainless Steel from the drop-down list.

Rebuilding the tree will that the DEFAULT material has been replaced.

38. Select OK to close the dialog box. 39.

or Thermal >Case Set Manager on the main Thermal Desktop menu/toolbar. Note: Rect::62 is still selected in the Model Browser.

A solution is calculated for the default conditions: steady state with no radiation. If the image below does not appear, select a different Layout tab at the bottom left of the Thermal Desktop window.

The Case Set Manager dialog box appears. 40. Select Run 1 Selected Case. A Sinda/Fluint Run Status dialog box appears confirming the successful completion of the process. 41. Select OK to close the dialog box.

20-51

Model Browser Example (Continued) 42. Select List > Submodel.Id from the Model Browser menu bar.

The submodel tree rebuilds and the window is back to its original form.

43. Select Submodel Node Tree, if not already selected.

In addition to the summary of the contents of the model, the output portion of the Model Browser includes the temperatures of the selected nodes, along with the Max and Min of the current selection set.

44. Scroll down the list in the output portion of the Model Browser look at the additional available information.

These values are the current postprocessed data. If the current postprocessed data were heat rates, then these values would be heat rates. To increase the size of the output portion of the window, the divider between the windows can be dragged. 45. Select Options > Temperatures from the Model Browser menu bar. 46. View the Output area.

47. Select Options > Output Window on Bottom from the Model Browser menu bar to deselect it (remove the check mark).

20-52

Scrolling in the text window, you’ll see the output looks more like a SINDA TPRINT.

This moves the output area to the right of the tree. The Model Browser window may be resized as desired. The example below shows the window resized horizontally and shortened.

Model Browser Example (Continued)

48. Select Submodel BBBB. 49. View the Output area. 50. Select Options > CSG from the Model Browser menu bar. 51. Select Submodel Node Tree. 52. View the Output area.

Selecting on a single submodel, such as BBBB will show the data only for that submodel. Selecting Options > CSG shows the CSG of the selected nodes, sorted in lowest to highest order. The CSG is the capacitance of a node divided by the sum of all conductances attached to it; it directly affects the timestep of the model for transient runs. In order for the CSG to work, the SINDA save file must have capacitance and conductors saved on it. The is set on the Output Tab of the Case Set Manager.

53. Select Options > Node Map from the Model Browser menu bar. 54. View the Output area.

Options > Node Map shows a SINDAlike NODMAP capability in order for the to determine how energy is transferred into a node.

20-53

Model Browser Example (Continued) 55. Select Options > Heat Map in the Model Browser. 56. Select submodel CCCC. 57. View the Output area.

The heat map shows the energy summary for energy into and out of the selected nodes. Energy between the selected nodes is not in the tabulation. If submodel CCCC is selected, the heat map implies a load of 3W and has 3W leaving into submodel BBBB. If submodels BBBB and CCCC are selected the heat map output will show the energy going from submodel BBBB into submodel AAAA, along with the 3W heat source on submodel CCCC.

58. Select Options > Heat Flow Between Submodels in the Model Browser window. The Heat Transfer Between Submodels dialog box appears. 59. Click on the arrow next to the From Submodel field and select AAAA from the drop-down list. 60. Click on the arrow next to the To Submodel field and select BBBB from the drop-down list. 61. Select OK.

20-54

The results of the heat flow analysis 3W going from submodel AAAA to submodel BBBB. The program cycles through all the nodes in submodel AAAA and sums the heat flows of all the conductors that connect to submodel BBBB.

Model Browser Example (Continued) 62.

or Thermal >Case Set Manager on the main Thermal Desktop menu/toolbar.

Case Set 1 is a transient analysis that solves for 10,000 seconds.

The Case Set Manager dialog box appears. 63. Select Case Set 1. 64. Select Run 1 Selected Case. A Sinda/Fluint Run Status dialog box appears confirming the successful completion of the process. 65. Select OK to close the SINDA/FLUINT Run Status dialog box. 66. Select Submodel Node Tree. 67.

or Edit > Plot in the Model Browser menu.

68. Minimize or close the EZXY window.

69. Close the Model Browser. 70. Select File > Exit.

With transient results, only the data for the currently postprocessed results are shown in the Model Browser output frame. The Plot command in the Model Browser creates a Data vs. Time plot of the postprocessed variable for all selected items in the model tree. Exit Thermal Desktop and save as prompted.

A Thermal Desktop/AutoCAD dialog box appears asking to save the drawing changes. 71. Select Yes.

20-55

20-56

20.4

Simple Meshing Methods

What will be learned: • Creating AutoCAD surfaces from lines • Creating finite difference surfaces from AutoCAD surfaces • Creating finite elements from AutoCAD surfaces • Resequencing node ID’s Prerequisites: • Section 20.2: Setting Up a Template Drawing on page 20-35 This example introduces some simple methods on how to create a mesh and convert it to either Thermal Desktop polygons or finite elements. Note: With the advent of TDMesh (see Advanced Modeling Techniques by opening Start > Programs > Thermal Desktop > s Manual - Meshing) the meshing techniques in this tutorial have become outdated. TDMesh provides a method of creating nodes and finite elements based on AutoCAD geometry (surfaces regions and solids). Once generated, the mesh, and all nodes and elements associated with the mesh, can be easily updated. Simple Meshing Methods Create a new folder named mesh and start with a new thermal.dwg template file. 1. Copy the template thermal.dwg file created in the first tutorial to the \mesh directory just created. Note: Be sure to hold the key down if dragging the template file icon to the new directory so that the file is copied, rather than moved. 2. Rename the copied template file to mesh. 3. Start Thermal Desktop by double clicking on the mesh drawing file icon in the mesh directory. 4. Select View > 3D Views > Top.

The UCS icon reflects the new orientation.

20-57

Simple Meshing Methods (Continued) or Draw > Line.

5.

The Command line should now read: Command: _line Specify first point:

6. Type 0,0 in the Command line. The Command line should now read: Specify next point or [Undo]:


8. Press <Enter> to complete the command. The first line displays in the drawing area. or Draw > Line.

9.

The Command line should now read: Command: _line Specify first point:



12. Press <Enter>. A second line displays in the drawing area. 13.

or select View > Zoom > Extents

20-58

The exercise begins by drawing two lines and drawing a ruled surface between the two. These lines could also be arcs or polylines if desired. The RULESURF command draws 3D surfaces between two objects: point and line; line and line; or arc and line. There are three related commands, one of which will be used later in this exercise: EDGESURF: Draws a 3D polygon mesh bordered by 4 edges. REVSURF: Draws a 3D surface of revolution. TABSURF: Draws a 3D tabulated surface.

Simple Meshing Methods (Continued) 14.

or type RULESURF in the Command line. The Command line should now read:

A mesh has been created between the 2 lines. This mesh could then be converted to Thermal Desktop polygons or finite elements.

Select first defining curve:

15. Select the first line. Note: Select the line at either the top or the bottom of the line. The Command line should now read: Select second defining curve:

16. Select the second line. Note: Select the second line at same end of the line at which the first line was selected. Connecting lines, a mesh, appear between the two lines.

Figure 20-8

Mesh Created

Note: If the lines looked crossed, then different ends of the lines were selected. Undo ) and redo the command selecting near the X axis for both of the lines.

Figure 20-9

Incorrect Mesh

20-59


or select Thermal > Surfaces/ Solids > From AutoCAD Surface. The Command line should now read: Select entity for adding thermal model data:

Make a set of polygons from the new entity. After completing these steps, notice that while there are 6 polygons, only a single node represents all of them.

18. Select a point on the mesh. The Command line should now read: Select entity for adding thermal model data:

19. Press <Enter>. The Thin Shell Data dialog box appears. 20. Select OK to close the dialog box without making any changes. A node appears on the mesh.

20-60

Figure 20-10

Polygon Created

By default, these surfaces are represented by a single node. Each facet of the converted surface may be a separate node by using the Toggle Mesh Nodalization functionality, the next functionality to be covered.

Simple Meshing Methods (Continued) 21. Select Thermal > Modeling Tools > Toggle FD Mesh Nodalization.

At the completion of these steps, there are now 6 nodes representing the polygons.

The Command line should now read: Select FD Meshes to toggle nodalization:

22. Draw a box around the entity to select the polygons. The Command line should now read: Select FD Meshes to toggle nodalization:

23. Press <Enter>.

Figure 20-11

Additional Polygons

Additional nodes display. Thermal Desktop’s Modeling Tools > Toggle Mesh Nodalization command allows the to change the nodalization scheme used by surfaces that were converted to the Thermal Desktop from AutoCAD geometry. A Thermal Desktop surface created from an AutoCAD surface initially contains one nodal region with the same or separate node IDs on each side. Mesh surfaces may be converted to one nodal region per mesh facet by using the Modeling Tools > Toggle Mesh Nodalization command (performing the command a second time will change the mesh back to one node per side).

20-61

Simple Meshing Methods (Continued) 24. Select the mesh to highlight it in the drawing area.

Delete the mesh and use the SURFTAB1 command to change the polygons from 6 to 3.

or Modify > Erase.

SURFTAB1 controls how many polygons that will be displayed when the mesh is created.

25.

The mesh is deleted and original two lines are displayed. 26. Type SURFTAB1 in the Command line. The Command line should now read: Enter new value for SURFTAB1 <6>:

27. Type 3 in the Command line. Note: The new value for SURFTAB1 is 3. To this, press and view the command line remarks. 28.

or type RULESURF in the Command line.

Use RULESURF to remesh and create 3 polygons.

The Command line should now read: Select first defining curve:

29. Select the first line. The Command line should now read: Select first defining curve:

30. Select the second line.

Figure 20-12

After Rulesurf

Three polygons are created. 31. Select the mesh to highlight it in the drawing area. or Modify > Erase.

32.

The mesh is deleted and original two lines are displayed.

20-62

Delete the mesh again.

Simple Meshing Methods (Continued) Draw another set of lines. 33.

or Draw >Line. The Command line should now read: Command: _line Specify first point:


Figure 20-13

Another Set of Lines


36. Press <Enter>. A line connecting the first two lines along the X axis displays in the drawing area. 37.

or Draw >Line. The Command line should now read: Command: _line Specify first point:



40. Press <Enter>. A fourth line displays in the drawing area.

20-63


or type EDGESURF in the Command line. The Command line should now read:

These steps create a 3x6 mesh. The SURFTAB2 parameter could be changed to change the 6 breakdown to something else.

Select object 1 for surface edge:

42. Select the first line. Note: Lines may be selected in either clockwise or counterclockwise direction. The Command line should now read: Select object 2 for surface edge:

43. Select the second line. The Command line should now read: Select object 3 for surface edge:

44. Select the third line. The Command line should now read: Select object 4 for surface edge:

45. Select the fourth line. The 3x6 mesh is created.

20-64

Figure 20-14

3x6 Mesh

The first and third lines selected use the SURFTAB1 parameter and the second and fourth lines, the SURFTAB2 paramenter. Therefore, if the top or bottom lines are selected first, the mesh will have 3 divisions from left to right and six divisions top to bottom.

Simple Meshing Methods (Continued) 46. Select Thermal > FD/FEM Network > Convert AutoCAD Surface to Nodes/Elements. The Command line should now read:

With finite elements, the geometry is bounded by the nodes, which generally gives a much better conduction model than converting to polygons.

Select entity for adding thermal model data:

47. Select the mesh to highlight it. The Command line should now read: Select entity for adding thermal model data:

48. Press <Enter>. The Thin Shell Data - Multiple Surface/Element Edit Mode dialog box appears.

Figure 20-15

Convert to Finite Elements

49. Select OK to close the dialog box without making any changes. A Thermal Desktop/AutoCAD dialog box appears confirming nothing was changed in the Thin Shell Data Multiple Surface/Element Edit Mode dialog box. 50. Select OK. Additional nodes are added to the mesh.

20-65


or select Thermal > Modeling Tools > Resequencing ID’s.

Resequence the node IDs so they are unique.

The Command line should now read: Select entity(s) for Node ID Resequencing:

52. Draw a selection box around the mesh. Note: Or type all in the Command line. The Command line should now read: Select entity(s) for Node ID Resequencing:

53. Press <Enter>. The Resequencing Node IDs dialog box appears.

54. Select OK. 55. Select File > Exit. A Thermal Desktop/AutoCAD dialog box appears asking to save the drawing changes. 56. Select Yes.

20-66

Exit Thermal Desktop and save as prompted.

20.5

Circuit Board Conduction Example

What will be learned: • Overview of how Thermal Desktop works • Creating material properties • Creating Thermal Desktop objects • Changing object colors • Applying an edge or • Applying a face or • Checking ors • Outputting SINDA/FLUINT conductance and capacitance data for review Prerequisites: • Section 20.2 - Setting Up a Template Drawing In this exercise, a circuit board will be mounted to an aluminum face. The circuit board will have a chip placed on it. Circuit Board Example 1. Copy the template thermal.dwg file created in the first tutorial to the \Tutorials\Thermal Desktop\board directory. Note: Be sure to hold the key down if dragging the template file icon to the new directory so that the file is copied, rather than moved. 2. Rename the copied template file to board. 3. Start Thermal Desktop by double clicking on the board drawing file icon in the board directory. 4.

or View > Visual Styles > 2D Wireframe to ensure consistency with the images in this tutorial.

20-67

Circuit Board Example (Continued) 5.

or Thermal > Thermophysical Properties > Edit Property Data. The Edit Thermophysical Properties dialog box appears.

20-68

This part of the exercise defines the thermophysical properties for aluminum, fr4, and the chip. The default units for Thermal Desktop are SI. These properties have the units of: •

W/m/K for conductance

•

J/Kg/K for specific heat

•

kg/m^3 for density.

Circuit Board Example (Continued) 6. Type Aluminum in the New property to add field. 7. Select the Add button. The Thermophysical Properties dialog box appears.

Define properties for Aluminum. Conductivity = 237 W/m/K Specific heat = 900 J/kg/K Density = 2702 kg/m^3

8. Highlight the current value in the Conductivity field and type 237. 9. Highlight the current value in the Specific Heat field and type 900. 10. Highlight the current value in the Density field and type 2702. 11. Select OK. The Edit Thermophysical Properties dialog box reappears with Aluminum and the above values displayed in the main property/description field.

20-69

Circuit Board Example (Continued) 12. Type fr4 2 oz copper in the New property to add field. 13. Select the Add button. The Thermophysical Properties dialog box appears. 14. Highlight the current value in the Conductivity field and type 17.7

Define properties for fr4 2 oz copper. Conductivity = 17.7 W/m/K Specific heat = 0 J/kg/K Density = 0 kg/m^3 Zero values (0) for specific heat, density, or object thickness (covered later) will make the nodes arithmetic.

15. Highlight the current value in the Specific Heat field and type 0. 16. Highlight the current value in the Density field and type 0. 17. Select OK. The Edit Thermophysical Properties dialog box reappears with fr4 2 oz copper and the above values displayed in the main property/description field. 18. Type chip in the New property to add field. 19. Select the Add button. The Thermophysical Properties dialog box appears. 20. Highlight the current value in the Conductivity field and type 0. 21. Highlight the current value in the Specific Heat field and type 837.32 22. Highlight the current value in the Density field and type 2000.

Define properties for chip. Conductivity = 0 W/m/K Specific heat = 837.32 J/kg/K Density = 2000 kg/m^3 The chip will be a single node. Therefore the chip’s internal conductivity is not used for these calculations, so any value can be used. If the chip were represented by multiple nodes, internal conduction could be ignored by either setting conductivity of the material to 0 or setting the object thickness to 0 (covered later). However, setting the thickness to 0 will also create arithmetic nodes. These values are approximations that are not really indicative of any particular chip.

20-70

Circuit Board Example (Continued) 23. Select OK to close the Thermophysical Properties dialog box.

Close the Thermophysical Properties dialog box.

The Edit Thermophysical Properties dialog box reappears with chip and the above values displayed in the main property/description field. 24. Select OK to close the Edit Thermophysical Properties dialog box. The model is to be built in inches. 25.

or Thermal > Preferences. The Preferences dialog box appears.

Notice that the energy units are in Joules, time in seconds, thus the energy rate units are Watts. The properties set earlier are automatically converted to inch (In) units.

26. Select the Units tab if not already displayed. 27. Click on the arrow next to the Model Length field and select in (inches) from the drop-down list. 28. Select OK to close the Preferences dialog box.

20-71


or Thermal > Surfaces/Solids > Rectangle. The Command line area should show: Command: _RcRectangle Origin point <0,0,0>:

30. Type 0,0.

Input the appropriate data to make the aluminum plate with the proper nodalization. A rectangle will be created for the aluminum plate. Origin = 0,0 X length = 6 Y length = 3

Note: , <Enter> is implied after typing input.

Centered nodes

The Command line should now read:

3 subdivisions in the Y direction

Point for +X axis and X-size <@1,0,0>:

31. Type 6,0. The Command line should now read: Point to set XY plane and Ysize <@0,1,0>:

32. Type 0,3. The Thin Shell Data dialog box appears.

33. Click on the Subdivision tab if not already displayed. 34. Leave Centered Nodes selected. 35. Highlight the current value in the Xdirection Equal field and type 6. 36. Highlight the current value in the Ydirection Equal field and type 3.

20-72

6 subdivisions in the X direction When values are typed, they will automatically be placed into the command line or next to the cursor; there is no need to physically click into the Command line area. Inputs and/or prompts may appear next to the pointer when typed: this is referred to as Dynamic Input. This option can be changed with Tools > Drafting Settings on the Dynamic Input tab.

Circuit Board Example (Continued) 37. Click on the Cond/Cap tab. Text fields and information specific to conductance/capacitance is displayed. 38. Click on the arrow next to the Material field and select Aluminum from the drop-down list. 39. Highlight the current value in the Thickness field and type .05

Generate Nodes And Conductors must be selected to change the material and the thickness. Set the material as Aluminum with a thickness of 0.05 in. At the end of these steps, the screen should look similar to the view below:

40. Select OK to close the Thin Shell Data dialog box.

Figure 20-16

41.


Aluminum Plate

If OK is selected too soon, select the rectangle and choose Thermal > Edit... to reenter the edit dialog.

20-73


or Thermal > Surfaces/Solids > Rectangle. The Command line area should show: Command: _RcRectangle

This part of the exercise creates the circuit board. The @ sign input tells the program to input a point relative to the last point input. In this example, the @5.5,0 is the same as typing 5.75,1.5.

Origin point <0,0,0>:

43. Type 0.25,1.5 The Command line should now read: Point for +X axis and X-size <@1,0,0>:

44. Type @5.5,0. The Command line should now read: Point to set XY plane and Ysize <@0,1,0>:

45. Type @0,0,3. The Thin Shell Data dialog box appears. 46. Click on the Cond/Cap tab if not already displayed.

20-74

Additional information for the circuit board is to be added.

Circuit Board Example (Continued) 47. Highlight the current value in the Cond Submodel field and type board. 48. Click on the arrow next to the Material field and select fr4 2 oz copper from the drop-down list.

When inputting board for the Cond submodel, the word board must be typed in. In the next step, on in the Numbering tab, board will have been added to the pulldown list.

49. Highlight the current value in the Thickness field and type .03

20-75

Circuit Board Example (Continued) 50. Click on the Subdivision tab. A Thermal Desktop/AutoCAD dialog box appears confirming the addition of BOARD to the Submodel list.

Thermal Desktop does not calculate the conductance from nodes to the edges of surfaces connected by or ors. Therefore, that conductance should be included in the value or the better method, used here, is to use edge nodes.

51. Select Yes. The Subdivision tab information is displayed. 52. Select Edge Nodes 53. Highlight the current value in the Xdirection Equal field and type 6. 54. Highlight the current value in the Ydirection Equal field and type 4. 55. Click on the Numbering tab. The Numbering tab displays. 56. Click on the arrow next to the Submodel field and select BOARD from the drop-down list. 57. Select OK to close the Thin Shell Data dialog box. 58.

When finished, the model should look similar to the drawing below (the colors will be changed shortly). Notice the half cylinder along the X axis shows that resistance has been applied along that surface.


Figure 20-17 Board on Aluminum Plate

20-76

Circuit Board Example (Continued) 59. Type LTSCALE. The Command line should now read:

Change the Linetype scale factor so the edges of the new plate are more visible.

LTSCALE Enter new linetype scale factor <39.3701>:

60. Type 0.5 61.


62. Select the Graphics Visibility tab.

The rectangle displayed in the drawing area is divided into six units long the X axis and three along the Y axis. There are small circles in the center of each unit. The small circles are the nodes. The node display is to be turned off. Notice that the nodes do not line up with the aluminum plate. This step in the exercise turns off the node display starting from the Thermal menu.

63. Click on TD/RC Nodes to deselect it (remove the check mark from the box). 64. Select OK to close the Preferences dialog box. 65. The nodes are no longer displayed.

Note that a much faster way to toggle the display of nodes is to use the Toggle TD/RC Node Visibility icon located at the lower the right side of the screen. Note: Some tool bar icons may not be visible within the Thermal Desktop window borders. As with basic Windows functionality, tool bars can be moved and docked as desired by the .

20-77

Circuit Board Example (Continued) 66. Click on the newly created circuit board (the vertical rectangle) to select it. 67. Select the Color drop-down list (showing ByLayer and select Green

20-78

Change the color of the circuit board. The default color of the circuit board is the color of the current layer (in this case the current color is white). The color is being overridden to change the circuit board to green.

Circuit Board Example (Continued) Create a chip on the circuit board. 68.

or Thermal > Surfaces/Solids > Rectangle.

The chip is purposely being placed so that it overlaps the nodes on the board.

The Command line area should show:

Instead of inputting the points, the points could be snapped to the drawing if desired.

Command: _RcRectangle Origin point <0,0,0>:

69. Type 1.5,1.5,1.5 The Command line should now read: Point for +X axis and X-size <@1,0,0>:

Hint: <Shift> and click the right mouse button to access a menu for different types of snap points: select Node to snap to the center of a surface.

70. Type @1,0 The Command line should now read: Point to set XY plane and Ysize <@0,1,0>:

71. Type @0,0,1 The Thin Shell Data dialog box appears. 72. Click on the Numbering tab. 73. Highlight the current value in the Submodel field and type chip. 74. Click on the Cond/Cap tab. A Thermal Desktop/AutoCAD dialog box appears confirming the addition of CHIP to the Submodel list. 75. Select Yes. The Cond/Cap tab information is displayed. 76. Click on the arrow next to the Cond Submodel field and select CHIP from the drop-down list.

If CHIP is not in the drop-down list, then type it in.

77. Click on the arrow next to the Material field and select CHIP from the drop-down list. 78. Highlight the current value in the Thickness field and type .1 79. Select OK to close the dialog box. 20-79

Circuit Board Example (Continued) 80. Click on the newly created chip to select it. 81. Select the Color drop-down list (showing ByLayer and select Red.

Change the color of the chip to red. It may be necessary to ZOOM in to select the chip. When finished, the model should look similar to the drawing below.

Figure 20-18 Chip on Circuit Board

82. Select the newly created chip. or Modify > Move.

83.

A Selecting Subobjects on Solids information window may open. Read the information, select the option if you wish, and close the window The Command line should now read:

It is good modeling practice to move the chip off of the board. The distance of the offset shall be set to half of the sums of the thickness of the board (.03) and the chip (.1), which equals .065. The direction is in the negative Y direction because that is the direction in the current Coordinate System (UCS).

Specify base point or [Displacement] :

84. Type 0,-0.065,0 in the Command line. The Command line should now read: Specify second point or <use first point as displacement>:

85. Press <Enter>. The chip is moved away from the circuit board.

20-80

that <ENTER> is assumed after typing into the command line. In actuality, <ENTER> is hit twice.


or Thermal > FD/FEM Network > or.

The board needs to conduct to the base plate. This can be accomplished by using a or.

The Command line should now read: Select faces ing from:

87. Select the green circuit board in the drawing area. The Command line should now read: Select faces ing from:

88. Press <Enter>. The Command line should now read: Select surfaces ing to:

89. Select horizontal surface in the drawing area. The Command line should now read: Select surfaces ing to:

90. Press <Enter>.

20-81

Circuit Board Example (Continued) The or dialog box appears.

91. Click on the arrow next to the Conductor Submodel field and select BOARD from the drop-down list.

The or dialog box is displayed. Once the changes have been made and OK is selected, the graphical image for a or is displayed. •

The from surface will be shown as a green arrow showing the edge selected for .

•

The to surfaces will be shown in gold and will have arrows pointing to both sides of the surfaces.

Figure 20-19

Edge or

92. Click on the From drop-down and select Edges. 93. Highlight the current value in the Conduction Coefficient field and type 5. 94. Double-click Rect[BOARD]::274 in the From (1): list box. 95. In the Select Edges dialog, uncheck Along Y at X=0 and check Along X at Y=0

96. Select OK to close the Select edges dialog. 97. Select OK to close the or dialog box.

20-82

The edge for the or is determined by the creation of the BOARD surface: the X at Y=0 edge is the first edge specified after the origin.



The chip needs to conduct to the board. This can be accomplished by using a or.


99. Select the red chip in the drawing area. The Command line should now read: Select faces ing from:

100.Press <Enter>. The Command line should now read: Select surfaces ing to:

101.Select green circuit board in the drawing area. The Command line should now read: Select surfaces ing to:

102.Press <Enter>.

20-83

Circuit Board Example (Continued) The or dialog box appears.

103.Click on the arrow next to the Conductor Submodel field and select CHIP from the drop-down list. 104.Highlight the current value in the Conduction Coefficient field and type 5. 105.Select OK to close the or dialog box.

20-84

The or dialog box is displayed. Once the changes have been made and OK is selected, the graphical image for a or is displayed. •

The from surface will be shown as a green arrow showing the side selected for .

•

The to surfaces will be shown in gold and will have arrows pointing to both sides of the surfaces.

Figure 20-20

Face or

Notice the green arrow points away from the circuit board. Since the default or has infinite tolerance and uses a point distance method, this will work fine for this case, but technically the or should be edited, and then the should double click on the from surface to change the side of . If the surface had different nodes on each side, or if the ray trace algorithm was being used for the or, then this change would be mandatory in order to get the correct results.

Circuit Board Example (Continued) 106.Select Thermal > Model Checks > Show or Markers. The Command line should now read: Select ors to display markers for:

107.Click on one of the ors (the green or gold arrows).

This part of the exercise utilizes the Show or Markers command which shows the actual calculations for the . When items are connected yellow lines are drawn from the point(s) to the node with which it is connected.

The Command line should now read: Select ors to display markers for:

108.Press <Enter>. Yellow lines representing the s are displayed. 109. or Thermal > Model Checks > Clear /or Markers. 110.Repeat or display for the other or.

Figure 20-21

or Markers

20-85

Circuit Board Example (Continued) Note: The number of test points is controlled in the or dialog box, displayed earlier when the or was being created, as the Integration Intervals. A value of 10 integration intervals means the 10*10, or 100, points will be tested. In this model, notice that the chip is connected to 2 nodes. If an object does not connect to a node, then it is displayed as a red point. To prevent an object from hooking up, the tolerance on the or dialog box must be changed from the default of infinity to a smaller value.

111. or Thermal > Model Checks > Clear /or Markers. 112.Select Thermal > Cond/Cap Calculations > Output SINDA/FLUINT Cond/Cap.

This command deletes the or marker lines. It actually deletes all the items on the “Radcad_rays” layer. This part of the exercise demonstrates the Output SINDA/FLUINT Cond/Cap command. This command outputs the conductors and capacitance for the model to the file SINDA.CC located in the same directory as the drawing, in this case in the board directory. If the SINDA.CC file is opened in a text editor three different node blocks— MAIN (aluminum base), (circuit) BOARD, and CHIP—are detailed. In the conductor data for the chip, the chip is tied to two nodes on the circuit board representing the planar area. Likewise, the circuit board is tied to several nodes on the base representing the linear conductance.

20-86

Circuit Board Example (Continued) 113.Select File > Exit. A Thermal Desktop/AutoCAD dialog box appears asking if the wants to save changes to the board.dwg.

Note: It is good practice whenever working on a computer to periodically perform File > Save commands during the course of a session to help ensure work is not lost.

114.Select Yes. The drawing is saved and Thermal Desktop is closed.

20-87

20-88

20.6

Beer Can Example

What will be learned: • Overview of how Thermal Desktop works • Creating material properties • Creating Thermal Desktop objects • Changing global visibility options • Extruding planar objects into solid elements • Surface coating free solid finite element faces • Using model checks to model development • Using of arbitrary nodes and conductors • Using AutoCAD layers to control object visibility • Using the Case Set Manager • Parameterizing a model • Creating XY time-dependent plots Prerequisites: • Section 20.2 - Setting Up a Template Drawing In this example, a beer can full of beer will be constructed. The initial temperatures will be set to something similar to a refrigerator temperature of 5C. Free convection heat transfer coefficients will be applied to the sides of the can and the top of the can. Beer Can Example 1. Copy the template thermal.dwg file created in the first tutorial to the \Tutorials\Thermal Desktop\beercan directory. Note: Be sure to hold the key down if dragging the template file icon to the new directory so that the file is copied, rather than moved. 2. Rename the copied template file to beercan. 3. Start Thermal Desktop by double clicking on the beercan drawing file icon in the beercan directory.

20-89

Beer Can Example (Continued) 4.

or select Thermal > Thermophysical Properties > Edit Property Data. The Edit Thermophysical Properties dialog box appears.

5. Type Aluminum in the New property to add field. Note: See comments in the righthand column. 6. Select the Add button. The Thermophysical Properties dialog box appears.

This part of the exercise defines the thermophysical properties for aluminum and water. If the tutorials are being performed in order from the beginning of the tutorial chapter, the will have already defined Aluminum properties in the board model. Instead of redefining the properties, the has two choices: •

First, it is possible to use the Thermal > Thermophysical Properties > Open/Create Property DB... command to open the database created in the board example. The Aluminum defined there could be used and the Water definitions added to that database. In the case, the material property will be stored in the Board tutorial folder.

•

Second, the can import the Aluminum properties from the Board tutorial database into the Beer tutorial database. Once the Edit Thermophysical Properties dialog is open, select the Import button. Open the database created in the board example and select Aluminum from the list of available properties. In the case, the material properties will be stored in the Beercan tutorial folder. Note: If a Material is already listed in the Edit Property Data dialog box but one or more of the properties is different than what is needed, double click on the material of interest. The Thermophysical Properties dialog box will appear allowing changes to be made.

20-90

Beer Can Example (Continued) 7. Highlight the current value in the Conductivity k field and type 237. 8. Highlight the current value in the Specific Heat field and type 900. 9. Highlight the current value in the Density rho field and type 2702. 10. Select OK to close the Thermophysical Properties dialog box. The Edit Thermophysical Properties dialog box reappears with Aluminum and the above values displayed in the main property/description field. 11. Type Water in the New property to add field.

As in real projects, some assumptions must be made.

12. Select the Add button. The Thermophysical Properties dialog box appears. 13. Highlight the current value in the Conductivity k field and type 0.6. 14. Highlight the current value in the Specific Heat field and type 4200. 15. Highlight the current value in the Density rho field and type 1000. 16. Select OK to close the Thermophysical Properties dialog box. The Edit Thermophysical Properties dialog box reappears with water and the above values displayed in the main property/description field. 17. Select OK to close the Edit Thermophysical Properties dialog box.

20-91

Beer Can Example (Continued) or Thermal > Preferences.

18.

The Preferences dialog box appears. 19. Select the Units tab if not already displayed. 20. Click on the arrow next to the Model Length field and select in (inches) from the drop-down list. 21. Select the Graphics Visibility tab.

22. Click on TD/RC Nodes to deselect it (remove the check mark from the box). 23. Select OK to close the Preferences dialog box.

20-92

These steps change the units for the model to inches and turns off TD/RC node visibility for all the nodes that are attached to the surface.

Beer Can Example (Continued) Create the bottom of the aluminum can. 24.

or Thermal > Surfaces/Solids > Disk. The Command line should now read: Pick or enter point for center of disk <0,0,0>:

25. Type 0,0 in the Command line. The Command line should now read: Pick or enter point for +Z axis of disk <@0,0,1>:

26. Type 0,0,1 in the Command line. The Command line should now read: Enter maximum radius or pick/enter point <1.0>:

27. Type 1.3125 in the Command line. The Command line should now read: Enter minimum radius or pick/enter point <0.0>:

28. Press <Enter>. The Command line should now read: Enter start angle or pick/ enter point <0.0>:

29. Press <Enter>. The Command line should now read: Enter end angle or pick/ enter point <360.0>:

30. Press <Enter>. The Thin Shell Data dialog box appears displaying default values. 31. Click on the Subdivision tab if not already displayed. 32. Click on the radio button next to Edge Nodes to select it (display a dot in the circle).

Using edge nodes is important since finite elements will be extruded from this disk. If centered nodes are used, then the finite elements will not fill the entire volume of the can.

20-93

Beer Can Example (Continued) 33. Highlight the current value in the Angular Equal field (subdivisions) and type 9. 34. Highlight the current value in the Radial Equal field (divisions) and type 3. 35. Click on the Cond/Cap tab. 36. Click on the arrow next to the Material field and select Aluminum from the drop-down list.

When completed, this disk will represent the bottom of the beer can. This part of the exercise sets the disk properties. Note: If the OK button is accidentally selected before switching to the Cond/Cap tab, simply select the disk and select Thermal > Edit to get back to the form.

37. Highlight the current value in the Thickness field and type .05. 38. Select OK to close the Thin Shell Data dialog box. or View > Zoom > Extents.

39.

40. Select the newly created disk. 41. Select Thermal > FD/Fem Network > Extrude Normal To Planar Elements into Solids. The Extrude/Revolve Planar Elements into Solids dialog box appears. (Next page)

20-94

The disk will be extruded into solid elements.

Beer Can Example (Continued) 42. Leave Even Breakdowns selected (dot in the circle). 43. Highlight the current value in the Total Distance field and type 4.75. 44. Highlight the current value in the Solids created along path field and type 4. 45. Highlight the current value in the ID Increment for new nodes field and type 100.

Note: After the extrusion is completed, if the geometry looks like there is a hole in the middle of the extruded solids then Edge Nodes (Subdivision tab in the Thin Shell Data dialog box) was not selected when the disk was created. Perform the following steps to make the correction: • Press < Z> to undo the extrusion. • Edit the disk to make the nodes edge nodes as follows: • Select the disk in the drawing area. • Select Thermal > Edit. • In the Thin Shell Data dialog box, select the Subdivision tab and make the corrections. Click on OK. • Return to Step 40.

46. Select OK to close the dialog box.

20-95

Beer Can Example (Continued) or Thermal > Edit.

47.

The Command line should now read:

The newly created solids must be edited to change their material to water. The properties of water are being used as an assumption of the properties of beer.

Select Objects or [Indiv]:

48. Type all in the Command line. The model in the drawing area is selected and Select Objects or [Indiv]: appears in the Command line area. 49. Press <Enter>. The Object Selection Filter dialog box appears. 50. Click on Solid Elements(64) to select it.

20-96

Only one type of object can be edited at a time. The Object Selection Filter makes it easy to select the desired object from the list.

Beer Can Example (Continued) 51. Select OK to close the selection filter and select the solid elements. The Solid Elements Attributes dialog box appears.

52. Click on the arrow next to the Material field and select Water from the drop-down list. 53. Select OK to close the dialog box. A Thermal Desktop/AutoCAD dialog box appears confirming the change. 54. Read the content of the dialog box and select Apply Changes.

20-97

Beer Can Example (Continued) 55. Select Thermal > FD/Fem Network > Surface Coat Free Solid FEM Faces. The Command line should now read: Select the solids for free face calculations:

56. Type all in the Command line. The Command line should now read: Select the solids for free face calculations:

57. Press <Enter>. The Thin Shell Data - Multiple Surface/Element Edit Mode dialog box appears. 58. Click on the Cond/Cap tab.

59. Click on the arrow next to the Material field and select Aluminum from the drop-down list.

20-98

Place the aluminum shell around the rest of the can. The solids will be surface coated to place the shell around the outer cylinder and the top. Surface coating will place a planar element using the same nodes used by the solid elements. The command is smart enough to figure out that the outside faces are not hooked up to other solids (and creates the planar element there), while the inside faces are hooked to more than one solid, so those faces are not free.

Beer Can Example (Continued) 60. Highlight the current value in the Thickness field and type .03. 61. Select OK to close the dialog box. A Thermal Desktop/AutoCAD dialog box appears confirming the change. 62. Read the content of the dialog box and select Apply Changes. 63.


20-99

Beer Can Example (Continued) or Thermal > Edit.

64.

Select Objects or [Indiv]:

appears in the Command line area 65. Type all in the Command line. Select Objects or [Indiv]:

appears in the Command line area. 66. Press <Enter> to end the selection process. The Object Selection Filter dialog box appears. 67. Click on Nodes(85) in the Select Type to Filter field to select it. 68. Select OK to close the dialog box. The Node - Multi Edit Mode dialog box appears.

69. Highlight the current value in the Initial temp field and type 278.15. 70. Select OK to close the dialog box. A Thermal Desktop/AutoCAD dialog box appears confirming the change. 71. Read the content of the dialog box and select Apply Changes.

20-100

This part of the exercise edits all of the nodes to set their initial temperatures.


or Thermal > FD/Fem Network > Node.

Create a node to connect to a convective conductor. This node will represent the ambient air temperature.

The Command line should now read: Enter location of node:

73. Type 3,0,0 in the Command line. The node appears to the right of the model.

20-101

Beer Can Example (Continued) 74. Select the newly created node. or Thermal > Edit.

75.

The Node dialog box appears.

The node will be edited to make it a boundary node and placed in submodel air. Notice that sometimes objects are selected before the command and sometimes after the command. If objects are selected before the command, then the first operation of the command uses the “pre-selected” objects if they are the right type. If a command requires objects, but nothing is selected before the command, then the command line will query for the needed objects. The temperature of the node will be defined as a symbol, making it easy to set up a second case that has different air temperature. Note: The Expression Editor is displayed when the mouse is double clicked in a field.

76. Highlight the current value in the Submodel field and type Air. 77. Click on the radio button next to Boundary in the Type field to select it (display a dot in the circle). 78. Double click in the Initial temp field. The Expression Editor dialog box appears.

20-102

Beer Can Example (Continued) 79. Select the Add Symbol button. The Symbol Manager dialog box appears.

Before using the symbol in the definition of the initial temperature, the symbol must be created.

80. Type Airtemp in the New Symbol Name field. 81. Select Add. An Expression Editor dialog box for Airtemp appears.

20-103

Beer Can Example (Continued) 82. Type 20 in the main entry field.

83. Select OK to close the Airtemp Expression Editor dialog box. The Symbol Manager dialog box is updated with the Airtemp information displayed.

20-104

Note: Symbols do not have units associated with them. When they are applied units are then assigned.

Beer Can Example (Continued) 84. Select Done in the Symbol Manager dialog. The Expression Editor dialog box reappears. 85. Click on the arrow underneath Select units for: Temperature and select C from the drop-down list.

Now that the symbol for the air temperature has been defined, the expression for the temperature of the boundary node can be created. Note: Symbols do not have units associated with them. When they are applied units are then assigned.

86. Right-click the Expression field, select General and then select Airtemp.

87. Select OK to close the Expression Editor dialog box. Note: The Initial Temp value is now in bold type and should read 293.15.

20-105

Beer Can Example (Continued) 88. Select OK to close the Node dialog box. A Thermal Desktop/AutoCAD dialog box appears asking for confirmation to add Air to the submodel list.

Note: When this portion of the exercise is completed, the shape of the node changes to designate that it is now a boundary node.

89. Select Yes. The node’s shape is changed to reflect its designation as a boundary node. 90. Select View > 3D Views > Front. The view changes. Note the UCS icon also moves to the lower left of the drawing area. 91. Type Zoom in the Command line. The Command line should now read:

Change the view from the current SW Isometric to a Front view. The view should look as follows. Note the new node in the lower right-hand corner.

All/Center/Dynamic/Extents/ Previous/Scale/Window/ :

92. Type .9x in the Command line.

Figure 20-22 Beer Can Front View

20-106

Beer Can Example (Continued) or type layer in the Command

93.

A new layer is to be created for the conductors to reside on.

line. Note: The menu selection Format > Layer may also be used. The Layer Properties Manager dialog box appears. 94.

to create a new layer. A new layer named Layer1 is added.

95. Highlight the name Layer1 (if not already highlighted) and type Conductors to change the name of this newly created layer. 96. Select the Freeze icon (sun) for the Conductors layer to freeze it (change the sun to a snowflake).

97. Close the Layer Properties Manager dialog box.

20-107


or Thermal > FD/Fem Network > Node to Surface Conductor. The Command line should now read: Select node:

99. Click on the boundary node (lower right on the screen). The Command line should now read: Select surfaces:

The next steps create the conductors and connect them to the surface. When prompted to select the surface areas on the beer can, it is important to drag-select from the top left to bottom right since selecting in the reverse direction has a different meaning in AutoCAD (see example below). 1 2

100.Select surfaces: Select from 1 to 2 as shown in to the right and as noted below: • Using the example to the right as a guide, click the left mouse outside and above the upper left corner of the surface area (1). The Command line should now read: Specify opposite corner:

• Position the mouse outside and below the opposite, lower right corner of the surface area as shown in the example and click the left mouse button (2). Note that as the mouse is moved, a box is drawn around the area. The Command line should now read: Select surfaces:

101.Press <Enter>. 102.Select the new conductor. Note: The new conductor set can be selected by picking any line of the set.

20-108

A set of eight lines (representing the conductor) from the boundary node to the surface area are displayed.

Beer Can Example (Continued) Edit the new conductor. 103.

or Thermal > Edit.

The Conductor dialog box appears. 104.Type Top Convection in the Comment field. 105.Click on the Type arrow and select Natural Convection Horizontal Flat Plate Upside from the dropdown list.

For the disk: •

area = pi*r^2

•

perimeter = pi*r*2

•

Area/Perimeter = radius/2

•

radius = 1.3125.

106.Highlight the current value in the Area/Perimeter field and type .65625.

107.Select OK to close the dialog box.

20-109

Beer Can Example (Continued) 108.Select the new conductor. 109.Click on the Layer Control drop-down in the upper right toolbars, as shown, and select Conductors.

110.Select Close to confirm the change and close the dialog box. The conductor moves to the Conductor layer, which is turned off, and disappears from the screen.

20-110

This part of the exercise moves the conductor to the Conductor layer that was frozen in the previous step. Doing this will make the display less cluttered.

Beer Can Example (Continued) 111. or Thermal > FD/Fem Network > Node To Surface Conductor. The Command line should now read: Select node:

112.Select the boundary node (lower right). The Command line should now read: Select surfaces:

113.Draw a selection box from points 1 to 2 as shown in figure to the right and as noted below: • Using the example to the right as a guide, click the left mouse at the lower right area of the surface area (1).

The drawing below shows the correct point selection order to be used for the next steps. It is important to begin in the lower right area of the beer can (first point, 1), as shown, and move the mouse to the upper left area (second point, 2). When selecting from the bottom right to the top left, any entity that is fully or partially enclosed will be included in the selection set. If the selection order is changed (point 2 and then point 1) only the items that are fully included in the box will be included in the selection set.

2

The Command line should now read: Specify opposite corner:

• Position the mouse on the opposite, upper left of the surface area as shown in to the right and click the left mouse button (2). Note that as the mouse is moved, a box is drawn around the area. The Command line should now read: Select surfaces:

1

When these steps are completed, the screen should appear similar to the example below.

114.Press <Enter>. A set of sixteen lines (representing the conductor) from the boundary node to the surface area are displayed.

20-111

Beer Can Example (Continued) 115.Select the new conductor. 116.

or Thermal > Edit.

The Conductor dialog box appears. 117.Type Side Convection in the Comment field. 118.Click on the Type arrow and select Natural Convection Vertical Cylinder - Isothermal from the dropdown list. The content of the Conductor dialog box changes to reflect the selection. 119.Highlight the current value in the Height field and type 4.75. 120.Highlight the current value in the Diameter field and type 2.625.

121.Select OK to close the Conductor dialog box.

20-112

Beer Can Example (Continued) 122. or select Thermal > Model Browser. The Model Browser appears.

As with the first conductor, this new conductor will be moved from layer 0 to the layer Conductor so that it does not clutter up the display.

123.Select List > Conductors in the Model Browser. The Model Browser tree displayes the Conductor Tree 124.Right-click on Cond-Side Convection under MAIN and select Change Layer > Conductors. The conductor moves to the Conductor layer, which is turned off, and disappears from the screen. 125.Close or minimize the Model Browser window. 126.Select Thermal > Model Checks > List Duplicate Nodes. The Command area should now show: Listing of duplicate nodes No duplicate nodes were found

Note: If the above statement does not appear in the command line, press to view the complete Command line comments.

Before any geometry building is complete, it is important to look for duplicate nodes. If any duplicate nodes are found, it may be necessary to use the Resequence IDs command to renumber them. Note: See “Resequence IDs” on page 7-2. Look at the output and see if any are found.

20-113

Beer Can Example (Continued) Turn off the display of the air node. 127.

or Thermal > Preferences.

The Preferences dialog box appears. 128.Select the Graphics Visibility tab if not already displayed. 129.Click on Defined Nodes to deselect it (remove the check mark from the box).

130.Select OK to close the dialog box. The air node disappears from the drawing area.

20-114

Alternatively, the visibility of some objects can be toggle off and on using icons in the toolbars. For Defined Nodes, it is the icon:

Beer Can Example (Continued) 131. line.

or type layer in the Command

Create a new layer called RightSide, which is where the right side of the beer can will be placed.

The Layer Properties Manager dialog box appears. 132.

to create a new layer.

A new layer named Layer1 is added to the existing layers. 133.Highlight the name Layer1 if not already highlighted. Type RightSide to change the name of this newly created layer. 134.Select the Freeze icon (sun) for the RightSide layer to turn freeze it (change the sun to a snowflake). 135.Close the Layer Properties Manager dialog box.

With some versions of AutoCAD, a layer named ASHADE is system generated and internally used by AutoCAD to control the lighting settings of the objects. The ASHADE layer is locked by default and entities cannot be changed or added.

20-115

Beer Can Example (Continued) 136.Select the right side of the beer can by drawing a selection box from points 1 to 2 as shown on the right and as noted below: • Using the example to the right as a guide, click the left mouse at the lower right area of the surface area (1).

Split the beer can into two sides so that the temperatures in the middle of the beercan can be determined later in the exercise. \ 2

Specify opposite corner: appears

in the Command line area. • Position the mouse as shown in to the right (above and to the right of the middle line) and click the left mouse button (2). Note that as the mouse is moved, a box is drawn around the area. The right side of the can is selected in the drawing area. 137.Click on the Layer Control dropdown in the upper right toolbars, as shown, and select RightSide. 138.Select Close to confirm the change and close the dialog box. 139.Close the Properties window.

20-116

1

Once the selected right side is moved to the RightSide layer, the drawing area should look similar to the example below:

Beer Can Example (Continued) 140.Select View > 3D Views > SE Isometric.

The new view should now look as follows.

The new view appears in the drawing area.

20-117

Beer Can Example (Continued) 141. or Thermal > Model Checks > Color by Property Value > Conductivity. Note: If the blue is a little dark, feel free to rotate a little bit to see if better. This is also a good time to review graphics settings (see “Graphics Settings” on page 19-3).

142. or Thermal > Model Checks > Color by Property Value Off. The model reverts back to the previous wireframe view.

20-118

This command verifies that the materials are set correctly. The picture should look similar to the view below with the aluminum being about 6 and the water being about .01. If the values are not right, edit the material property of the incorrect entities.

Beer Can Example (Continued) 143. or Thermal > Case Set Manager. The Case Set Manager dialog box appears.

The Case Set Manager changes the view from the geometric model to temperatures with the click of a button. The default process is to run a steady state case, but a transient run is what is needed here.

144.Select Edit. The Case Set Information dialog box appears. (next page)

20-119

Beer Can Example (Continued) 145.Select the Calculations tab. 146.Click on Steady State in the Solution Type field to deselect it (remove check mark from the box). 147.Select Transient in the Solution Type field to select it (place a check mark in the box). 148.Highlight the current value in the End Time field and type 3600.

20-120

On the Calculations tab, the Solution type is chosen and, since a transient analysis is desired, an end time is set.

Beer Can Example (Continued) 149.Select the Output tab. 150.Highlight the current value in the Output Increment field and type 100.

The Output Increment defines how often during the solution the chosen values will be written to the output files.

151.Select OK. The Case Set Manager dialog box is on the screen.

20-121

Beer Can Example (Continued) 152.Select Run 1 Selected Case. A SINDA/Fluint Run Status dialog box appears stating the successful completion of the processor.

When the run is complete the temperature view should look similar to the following.

153.Select OK to close the dialog box. The model changes from the geometric view.

If the solution fails, please check the air node temperature. If it is accidentally input as 20K, the solution will fail.

20-122

Beer Can Example (Continued) 154. or Thermal > Post Processing > Edit Current Dataset. The Set SINDA Dataset Properties dialog box appears.

155.Scroll down the list in the Select a Time/Record [sec] field and select 3600 (3.60e+003).

After the solve is completed, the initial temperatures are displayed on the model in the postprocessing state. Note: If the colors do not look right, please see “Graphics Settings” on page 19-3.

Figure 20-23 Beer Can Postprocessed View

156.Select OK. 157. or Thermal > Post Processing > PostProcessing Off. The model returns to the geometric view in the drawing area.

20-123

Beer Can Example (Continued) 158. or Thermal > Case Set Manager. The Case Set Manager dialog box appears. 159.Select Copy. The Copy Case Set dialog box appears.

160.Highlight the current value in the in the New Case Set Name field and type Hot Case. 161.Select OK to close the dialog box. The Case Set Manager dialog box updates to reflect Hot Case in the Case Sets field. 162.Select Hot Case. 163.Select Edit. The Editing 1 Case Set - Hot Case dialog box appears.

20-124

Create a hot case where the air temperature is 25 °C. By overriding the global definition of 20 °C with 25 °C, the new case can be run quickly and it will be able to go back to it at a later time. When this case is run, all the SINDA files will go to case1.* Once the run is finished, edit the postprocessing dataset to change to the end time.

Beer Can Example (Continued) 164.Select the Symbols tab. 165.Select Airtemp in the Global List field to highlight it. 166.Click on the right arrow located in the center of the dialog box.

The Expression Editor (Airtemp) dialog box appears. 167.Highlight the current value in the main entry field (20) and type 25. 168.Select OK to close the dialog box. The Case Set Information - Hot Case dialog box displays the change.

20-125

Beer Can Example (Continued) 169.Select OK to close the Case Set Information - Hot Case dialog box and re-display the Case Set Manager dialog box. 170.Select Run 1 Selected Case. A SINDA/Fluint Run Status dialog box appears stating the successful completion of the processor. 171.Select OK to close the dialog box. The model changes from the geometric view. 172. or Thermal > Post Processing > Edit Current Dataset. The Set SINDA Dataset Properties dialog box appears. 173.Scroll down the list in the Select a Time/Record [set] field and select 3600 (3.60e+003). 174.Select OK. 175. or Thermal > Utilities > Capture Graphics Area.

The Thermal > Utilities > Capture Graphics Background will save the current graphics window to ScreenCapture1.bmp. The program determines the lowest ScreenCapture# that it can use so as to not overwrite an existing file. For example, a second command would save to ScreenCapture2.bmp. To the graphic is saved, open the beercan directory folder and ScreenCapture1 will be included.

20-126

Beer Can Example (Continued) 176. or Thermal > PostProcessing> Animate Through Time.

This command will animate through all the times on the postprocessing file.

The Continuous Cycle Dialog dialog box appears.

177.Select OK. 178.View the screen. 179. or Thermal > Post Processing > PostProcessing Off. The model returns to the geometric view in the drawing area.

20-127

Beer Can Example (Continued) 180.Select an element in the drawing 181. or select Thermal > Post Processing > X-Y Plot Data vs. Time.

This command will bring up the external XY Plotting program. This program will plot the transient for nodes of the element that have been selected.

182.View the results.

Note: The results will be different depending upon what was selected in the drawing area. The can change the nodes displayed by selecting the Edit > Add/Edit menu command in EZXY. The nodes being displayed and any plot customization can be saved to a file that can then be brought up external to Thermal Desktop. 183.Select File > Exit. A Thermal Desktop/AutoCAD dialog box appears asking to save the drawing changes. 184.Select Yes.

20-128


20.7

Conduction and Radiation Using Finite Elements

What will be learned: • Creating planar finite elements • Extruding and revolving planar elements into solid finite elements • ing proper connections of elements • Surface coating finite element solids for radiation, area , or insulation Prerequisites: • Section 20.2 - Setting Up a Template Drawing In this example, a finite element model will be created with fixed temperatures at either end. The process begins by creating a single quad element. The AutoCAD array command will be used to create a grid of elements. The planar quad elements will be extruded and revolved into 3D solid elements. The solid elements will be surface-coated with zero-thickness planar elements to be used in a later tutorial for asg radiation properties. Temperature boundary conditions will be applied and the model will be solved by SINDA. Radiation will be optionally added to the analysis at the end of the tutorial. Finite Element Example 1. Copy the template thermal.dwg file created in the first tutorial to the \Tutorials\Thermal Desktop\finiteElement directory. Note: Be sure to hold the key down if dragging the template file icon to the new directory so that the file is copied, rather than moved. 2. Rename the copied template file to fe1.dwg. 3. Start Thermal Desktop by double clicking on the fe1 drawing file icon in the finiteElement directory.

20-129

Finite Element Example (Continued) 4.

or select Thermal > Thermophysical Properties > Edit Property Data. The Edit Thermophysical Properties dialog box appears.

5. Type Aluminum in the New property to add field. 6. Select the Add button. The Thermophysical Properties dialog box appears. 7. Highlight the current value in the Conductivity k field and type 240. 8. Select OK to close the Thermophysical Properties dialog box. The Edit Thermophysical Properties dialog box reappears with the updated Aluminum value displayed in the main property/description field. 9. Select OK to close the Edit Thermophysical Properties dialog box.

20-130

Create the property aluminum with a conductivity of 240. Only the conductivity will be entered/updated. This will be a steady state example. Depending upon where the started the tutorials, Aluminum may already exist. If so, perform Step 4, double click on Aluminum in the Edit Thermophysical Properties dialog box and move to Step 7.


or Thermal > FD/Network > Node.

This part of the exercise creates 4 nodes. When finished, the model should look similar to the example below.

The Command line should now read: Enter location of node:

11. Type 0,0 in the Command line. The first node is created at the origin. 12.

or Thermal > FD/Network > Node. The Command line should now read: Enter location of node:

13. Type 1,0 in the Command line.

Figure 20-24 Newly Created Nodes

To repeat a command, the can hit <ENTER> or right click.

The second node is created. 14.


15. Type 1,1 in the Command line. The third node is created. 16.


17. Type 0,1 in the Command line. The fourth node is created. 18.


20-131


or Thermal > FD/Fem Network > Element. The Command line should now read: Select nodes for linear element:

20. Select node 1, the node at the axis of the UCS icon. The Command line should now read:

A quad element is being created from the four new nodes. The order in which the nodes are picked is extremely important. Refer to the drawing below to select the nodes. The order follows the right hand rule to determine which side is up. For example, picking the nodes in the order 1,2,4,3 would produce a quad where the diagonals would cross.

Select nodes for linear element:

21. Select node 2, the node to the left of the first node. The Command line should now read: Select nodes for linear element:

22. Select node 3, the node above the first node. The Command line should now read: Select nodes for linear element:

Figure 20-25 Node Selection Order

When the element is created, the view should be similar to the example below:

23. Select node 4, the node to the right of the first node. The Command line should now read: Select nodes for linear element:

24. Press <Enter> to end the selection process. Lines appear on the screen connecting the four nodes. Figure 20-26

20-132

Quad Element

Finite Element Example (Continued) 25. Select the new quad element to highlight it. 26.

or Thermal > Edit.

The element is being edited to apply the material property created earlier. The thickness does not matter since this element is used for an extrusion.

The Thin Shell Data dialog box appears. 27. Select the Cond/Cap tab.

28. Click on the arrow next to the Material field and select Aluminum from the drop-down list. 29. Select OK to close the dialog box.

20-133

Finite Element Example (Continued) 30. Select the new quad element. 31. Type ARRAYCLASSIC. (If you are using AutoCAD 2010 or 2011, type ARRAY)

This part of the exercise uses the Array command to create a 4x3 grid of quad elements. When completed, the model should look similar to the view below.

The Array dialog box appears.

Figure 20-27 Array of Quad Elements

32. Select Rectangular Array if not already selected (display a dot in the circle). 33. Highlight the current value in the Rows field and type 4 if a different value is displayed. 34. Highlight the current value in the Columns field and type 3. 35. Highlight the current value in the Row Offset field and type 1 if a different value is displayed. 36. Highlight the current value in the Column Offset field and type 1 if a different value is displayed. Note: The Row Offset and Column Offset fields display as 1.0000. 37. Select OK to close the dialog box. 38.

20-134


Finite Element Example (Continued) 39. Select Thermal > Model Checks > Show Free Edges. The Command line should now read: Select the elements for free edge calculations:

40. Type all in the Command line. The Command line should now read: Select the elements for free edge calculations:

41. Press <Enter>. The grid lines turn red and the Command line area should show:

The next steps use the Show Free Edges command to determine if these nodes are properly connected. Once the Show Free Edges command is executed, notice that red lines cover the whole grid of the model. What has happened is that the Array command has copied the nodes as well as the elements and, therefore, the nodes are lying on top of each other. If the output the model at this point (after resequencing the nodes), there would be no conduction between the elements.

48 individual edges found 48 free edges found

20-135


or Thermal > FD/Fem Network > Merge Coincident Nodes. The Command line should now read: Select nodes to be merged:

43. Type all in the Command line. appears in the Command line area. Select Objects:

44. Press <Enter>. The Merge Coincident Nodes dialog box appears.

45. Highlight the current value in the Coincidence Tolerance field and type .01. 46. Select OK to close the dialog box. A Thermal Desktop/AutoCAD dialog box appears asking for confirmation of the merge. 47. Select Yes.

20-136

This group of steps merges the coincident nodes.

Finite Element Example (Continued) or type regen in the Command

48. line.

The array turns white. 49. Select Thermal > Model Checks > Show Free Edges.

The regen command is performed here to clear the screen from the previous Show Free Edges and Merge Coincident Nodes commands. The free edges are checked again and now only the outlying edges are drawn in red.

The Command line should now read: Select the elements for free edge calculations:

50. Type all in the Command line. The Command line should now read: Select the elements for free edge calculations:

51. Press <Enter>. The outside edge of the array turns red.

20-137

Finite Element Example (Continued) 52. Select Thermal > FD/FEM Network > Extrude Planar Elements into Solids. The Command line should now read: Select Planar Elements/Edge Conics for Revolve/Extrude:

53. Type all in the Command line. The Command line should now read: 12 found Select Planar Elements/Edge Conics for Revolve/Extrude:

54. Press <Enter>. The Command line should now read: Select point to extrude from:

55. Type 0,0 in the Command line. The Command line should now read: Select point to define extrude vector/distance:

56. Type 0,0,5 in the Command line. The Extrude/Revolve Planar Elements into Solids dialog box appears. 57. Leave Even Breakdowns selected (dot in the circle). 58. Highlight the current value in the Solids Created along path field and type 5. 59. Select OK to close the dialog box.

20-138

The planar elements are extruded to make bricks. The vector given provides the distance of the extrusion.

Finite Element Example (Continued) 60. Type zoom in the Command line. The Command line should now read: Specify corner of window, enter a scale factor (nX or nXP), or

The view of the model is zoomed to the extents of the drawing area and then rotated a little bit to move the model off of the isometric view. The model should look similar to the view below.

[All/Center/Dynamic/Extents/ Previous/Scale/Window/ Object] :

61. Type extents in the Command line. The view shifts to show the full array. 62. Type -vpoint in the Command line. The Command line should now read: Specify a view point or [Rotate] :

63. Type -1,-1,0.9 in the Command line.

Figure 20-28 Extruded Elements

The view of the extruded model is rotated.

20-139


or Thermal > FD/Fem Network > Hide Solid Interior Faces. The view in the drawing area shifts.

Since the model was rotated off the normal, notice that there are many double lines that can be seen. These are the edges of the interior of the model. These lines can clutter up the model, especially if the model is the meshed. The next step turns off these lines. By turning off these lines redisplays, rotations, and postprocessing of the model will process faster. The calculation to hide the interior lines is only made when the selects the command. Thus, if more geometry is added, or deleted, the may need to re-execute the command to get the proper view.

Figure 20-29

20-140

Interior Lines Hidden

Finite Element Example (Continued) 65. Select Thermal > FD/Fem Network > Revolve Planar Elements into Solids. The Command line should now read: Select Planar Elements/Edge Conics for Revolve/Extrude:

These steps revolve the planar elements. “All” can be used in the Command line for selection purposes since the nodes and the solids will be filtered out. , only the bottom face has planar elements.

66. Type all in the Command line. The Command line area should show: 12 found Select Planar Elements/Edge Conics for Revolve/Extrude:

67. Press <Enter>. The Command line should now read: Select base point to revolve from:

68. Type -3,0 in the Command line. The Command line should now read: Select point to define revolve axis:

69. Type -3,3 in the Command line. The Extrude/Revolve Planar Elements into Solids dialog box appears. 70. Leave Even Breakdowns selected. 71. Highlight the current value in the Total Distance field and type 90. 72. Highlight the current value in the Solids created along path field and type 9. 73. Select OK to close the dialog box.

20-141



Use the Hide Interior Faces command to clean up the display as needed. The model should look similar to the drawing below.

Figure 20-30 After Revolved Elements

20-142

Finite Element Example (Continued) 75. Select Thermal > Modeling Tools > Make AutoCAD group > From Thermal Objects. The Command line should now read: Select entities to make into a group:

76. Type all in the Command line. The Command line should now read: Select entities to make into a group:

The model now consists of planar element and solids. The planar elements must be deleted because their function is complete. The objects will be put into a group and then the AutoCAD Delete function to delete that group will be used. Once all the object are selected, the filter appears and allows the planar objects to be filtered from the solids and nodes.

77. Press <Enter>. The Object Selection Filter dialog box appears.

Leave the current selections as they appear: Surfaces/Planar Elements(12). 78. Select OK to close the dialog box. The Group Name Input Form dialog box appears.

79. Type plane in the Input Group Name field. 80. Select OK to close the dialog box.

20-143

Finite Element Example (Continued) or select Modify > Erase.

81.

The Command line should now read: Select objects:

82. Type group in the Command line. The Command line should now read: Enter group name:

83. Type plane in the Command line. The Command line should now read: Select objects:

84. Press <Enter>.

20-144

These steps will delete the planar elements. Note that the letter ‘g’ could be used instead of the word “group”.

Finite Element Example (Continued) 85. Select Thermal > FD/Fem Network > Surface Coat Free Solid Faces. The Command line should now read: Select the solids for free face calculations:

86. Type all in the Command line. The Command line should now read: Select the solids for free face calculations:

87. Press <Enter>. The Command line area should show: 0 free tri faces found 220 free quad faces found

and the Thin Shell Data - Multiple Surface/Element Edit Mode dialog box also appears. 88. Review the Radiation tab.

Finite element solids must be surface coated to allow the definition of radiation, area or ors, or insulation. The Surface Coat Free Solid Faces command will calculate all of the solid free faces and place a planar element on that face so that the active side is top. When the Thin Shell Data dialog box opens the Radiation tab is active. The optical property DEFAULT is defined as a black-body: the emissivity and absorptivity are both unity. The thickness of the planar element is set to zero so that it does not affect capacitance and conductance calculations. With zero thickness the material properties will not be used, however, a material must still be selected since the material DEFAULT is undefined, unlike the optical property.

89. Select the Cond/Cap tab. 90. Click on the arrow next to the Material field and select Aluminum from the drop-down list. 91. Select OK to close the dialog box.

20-145

Finite Element Example (Continued) A Thermal Desktop/AutoCAD dialog box appears asking to confirm the change. 92. Select Apply Changes to close the dialog box. 93. Select View > 3D Views > Front. 94. Type zoom in the Command line. The Command line should now read:

Change the view to make it easier to select the nodes on the ends. The nodes will be selected and changed to boundary nodes.

Specify corner of window, enter a scale factor (nX or nXP), or [All/Center/ Dynamic/Extents/Previous/ Scale/Window/Object] :

1

95. Type .8x in the Command line. 3

4 Figure 20-31 Front View

20-146

2

Finite Element Example (Continued) 96. Select from points 1 to 2 as shown in Figure 20-32 and as noted below: • Using the example to the right as a guide, click the left mouse above the upper left corner of the surface area (1). The Command line should now read:

The nodes on the ends of the object must be edited. to apply the boundary temperatures. Refer to the example shown below (Figure 20-32) when selecting the nodes, starting the selection box at point 1. After the first point is selected, select point 2. 1

Command: Specify opposite corner:

• Position the mouse on the opposite, lower right corner of the selection area as shown in the example and click the left mouse button (point 2). Note that as the mouse is moved, a box is drawn around the area. 97.

2

3

or Thermal > Edit. The Object Selection Filter dialog box appears.

98. Select Nodes[20] in the Select Type to filter field.


4 Figure 20-32 Node Selection Order

When points 1 and 2 are selected, not only are the nodes selected but the surface coated planar elements that were created earlier are also selected. When the Edit function is selected, the function determines that more than one type of entity has been selected. The Object Selection Filter dialog box is displayed. Note: The Object Selection Filter dialog box can also be displayed by selecting Thermal > Modeling Tools > Toggle Filter On.

20-147

Finite Element Example (Continued) The Node - Mulit Edit Mode dialog box appears. 100.Click on Override calculation by elements/surfaces to place a check mark in the box. The Type frame activates.

101.Click on Boundary to place a dot in the circle. 102.Highlight the current value in the Initial temp field and type 373.15. 103.Select OK to close the dialog box. A Thermal Desktop/AutoCAD dialog box appears asking for confirmation of the changes. 104.Confirm the changes to close the dialog box.

20-148

The selected nodes are changed to boundary nodes and the temperature is set. When nodes are part of an element (or surface), the definition of those nodes are obtained from the associated element (or surface). Overriding the calculation by the element (or surface) allows the to provide a new definition for the selected node or nodes.

Finite Element Example (Continued) 105.Select from points 3 to 4 as shown in Figure 20-33 and as noted below: • Click the left mouse at point 3 as shown in Figure 20-33. The Command line should now read: Command: Specify opposite corner:

• Click the left mouse at point 4 as shown in Figure 20-33.

Again, the selected nodes are changed to boundary nodes and the temperature is set. When nodes are part of an element (or surface), the definition of those nodes are obtained from the associated element (or surface). Overriding the calculation by the element (or surface) allows the to provide a new definition for the selected node or nodes. 1

106.

or Thermal > Edit.

2

The Object Selection Filter dialog box appears. 107.Select Nodes[20] in the Select Type to filter field. 108.Select OK to close the dialog box.

3

The Node - Multi Edit Mode dialog box appears. 109.Click on Override calculation by 4 elements/surfaces to place a check Figure 20-33 Node Selection Order mark in the box. The Type field activates. 110.Click on Boundary to place a dot in the circle. 111.Highlight the current value in the Initial temp field and type 273.15. 112.Select OK to close the dialog box. 113.Confirm the changes to close the Multi Edit Dialog box.

20-149

Finite Element Example (Continued) 114. or select Thermal > Modeling Tools > Resequence IDs. The Command line should now read: Select entity(s) for Node ID Resequencing:

115.Type all in the Command line. The Command line should now read: Select entity(s) for Node ID Resequencing:

116.Press <Enter>. The Resequencing Node IDs dialog box appears.

117.Leave the default values and select OK to close the dialog box. The Command line area should show: 300 nodes were changed

20-150

This part of the exercise resequences the node IDs so all nodes have a unique number.

Finite Element Example (Continued) 118. or Thermal > Case Set Manager. The Case Set Manager dialog box appears. 119.Click on Edit.

The Case Set Manager is the link from Thermal Desktop to SINDA/FLUINT. The Case Set Manager allows the to set up different thermal analysis cases, each with its own radiation calculations and parameters. Once the cases are defined, the can create SINDA models, run the analyses, and post-process the results with the click of a single button. Once parameters are set and the Run Case button is clicked, Thermal Desktop will calculate any radiation conductors and heating rates for all of the tasks set up for the current Case Set. Nodes and conductors are then computed and output. A SINDA/FLUINT model is then built and run. And finally, the temperature results are displayed mapped onto the thermal model in color. Editing a case allows the case to be defined beyond the default settings.

20-151

Finite Element Example (Continued) The Case Set Information - Case Set 0 dialog box appears.

120.Examine the tabs and the information on each tab, but do not change anything at this time. 121.Select OK to close the dialog box. The Case Set Manager dialog box reappears. 122.Click on Run Case. The Case Set program is run. When complete, the graphic output appears in the drawing area. the SINDA/Fluint Run Status dialog box appears on top of the graphic output confirming successful completion of the process. 123.Select OK to close the dialog box.

20-152

By selecting OK, the case is allowed to use the default settings: no radiation with a steady state analysis. Selecting RUN CASE will allow the Case Set Manager to output the conductors and capacitance, build the SINDA model, solve the SINDA model, and display the steady state temperatures back on the model.

Finite Element Example (Continued)

Figure 20-34

124.Select File > Save

Solution

Save the model before adding radiation.

If you have access to a RadCAD license, you may continue; otherwise, skip ahead to Step 139.

20-153

Finite Element Example (Continued) 125. or Thermal > Case Set Manager. The Case Set Manager dialog box appears. 126.Click on Copy. The Add New Case Set dialog box appears. 127.Type Radiation for the New Case Set Name 128.Select OK to close the dialog box. The Case Set Manager dialog box reappears. 129.Select Radiation Case Set and click on Edit. The Case Set Information - Radiation dialog box appears.

20-154

Finite Element Example (Continued) 130.Select the Radiation Tasks tab, if it is not already selected. 131.Select the Add button. The Radiation Analysis Data dialog box appears. 132.Confirm Radks is selected in the Calculation Type region. 133.Confirm BASE is in the Analysis Group field. 134.Confirm Monte Carlo is selected in the Calculation Method region.

An Analysis Group is a -defined group of objects which will exchange energy though radiation. The group BASE is a default and all surfaces are included unless otherwise specified. To learn more about Analysis Groups see “Radiation Analysis Groups” on page 41. To learn more about radiation analyses and calculations see “Radiation Calculations and Output to SINDA/FLUINT” on page 10-1.

135.Select OK to close the dialog box. The Case Set Information - Radiation dialog box reappears. 136.Select OK to close the dialog box. The Case Set Manager dialog box reappears. 137.Click on Run Case. The Case Set program is run. When complete, the graphic output appears in the drawing area. the SINDA/Fluint Run Status dialog box appears on top of the graphic output confirming successful completion of the process. 138.Select OK to close the dialog box.

20-155

Finite Element Example (Continued)

Figure 20-35

Solution

139.Select File > Exit. A Thermal Desktop/AutoCAD dialog box appears asking if the wants to save the changes made to the drawing. 140.Select Yes. The drawing is saved and Thermal Desktop closes.

20-156

Exit Thermal Desktop. Note: Be sure to save the changes to the file since it will be used as a starting point for another tutorial.

20.8 Mapping Temperatures From a Coarse Thermal Model to a Detailed NASTRAN Model What will be learned: • Importing a finite element mesh created outside of Thermal Desktop. • Mapping temperatures from a Thermal Desktop model to another type of model. Prerequisites: • Section 20.2 - Setting Up a Template Drawing This tutorial maps temperatures from a coarse model of a plate with two holes in it to a different mesh of the same model that has much finer detail. This situation is very common when the stress analyst has a very detailed model for a part, but the thermal model must be made coarse so as to get the part integrated into the entire thermal model. CRTech would like to give a special thanks to Jim Braley for providing the sample NASTRAN models for this tutorial. Mapping Example 1. Copy the template thermal.dwg file created in the first tutorial to the \Tutorials\Thermal Desktop\mappingExample directory. Note: Be sure to hold the key down if dragging the template file icon to the new directory so that the file is copied, rather than moved. In addition to the copied template drawing, there are two existing files in the mappingExample folder: • coarse_quad.nas • fine_quad.nas. The two files will be imported into the model during the exercise. 2. Rename the copied template file to coarse. 3. Start Thermal Desktop by double clicking on the coarse drawing file icon in the mappingExample directory. 4. Select View > 3D Views > Top. The UCS icon reflects the new orientation.

Change the view in the drawing area to the top view.

20-157

Mapping Example (Continued) 5. Select Thermal > Import > NASTRAN. 6. The FE Model Options dialog box appears.

The next step imports the coarse_quad.NAS file into the drawing. When the import process is complete, a model of a plate with two holes in it will be displayed. The model should look similar to the example below.

Figure 20-36

7. Change the input file name to be coarse_quad.NAS. 8. Change the submodel to MAIN. 9. Select OK to close the dialog box.

20-158

Imported Drawing

Mapping Example (Continued) 10. Select the nodes on the left edge of the model by drawing a selection box from points 1 to 2, as shown in the example to the right. Note: In this case, selection order is not important in that only the nodes are to be selected rather than the nodes and any other objects that may be partially enclosed by the selection box. that selecting objects from top to bottom will only select those items fully enclosed by the selection box whereas selecting objects from bottom to top will include items not fully enclosed by the selection box into the selection set. 11.

Some boundary conditions must be created so the model can be run and some gradients obtained. The nodes on the left edge of the model are to be designated as boundary nodes. 1

2 Figure 20-37

Selection Points

or Thermal > Edit. The Node - Mulit Edit Mode dialog box appears. (next page)

20-159

Mapping Example (Continued) 12. Select Override calculations by elements/surfaces to place a check mark in the box. The Type fields activate. 13. Select Boundary (place a dot in the circle).

14. Select OK to close the dialog box. A Thermal Desktop/AutoCAD dialog box appears asking for confirmation of the node changes. 15. Confirm the changes to close the dialog box. The node shapes change to show their new designation

20-160

Mapping Example (Continued) 16. Select the nodes on the right edge of the model by drawing a selection box from points 3 to 4, as shown in the example to the right.

A heat load of 10 watts is to be applied to the nodes on the right edge of the model. 3

Note: As in the selection of the nodes on the left edge, selection order (top to bottom versus bottom to top) is not important in this case as only the nodes on the right edge are being selected in this step. 17.

or select Thermal > FD/Fem Network > Heat Load on Nodes. The Heat Load Edit Form dialog box appears.

Figure 20-38

4

Selection Points

Note: Because of the top view of the model, once the heat load has been applied it appears the nodes change color from white to red. If the model is rotated to look at from another angle, small red arrows are displayed which give the illusion of the red nodes from the top view. If the model is rotated back to a top view, the nodes appear white in color again, with small spots of red.

18. Highlight the current value in the Heat Load [W] Value field and type 10. 19. Select OK to close the dialog box.

20-161

Mapping Example (Continued) or Thermal > Case Set Man-

20. ager.

The Case Set Manager dialog box appears.

21. Click on Edit. The Case Set Information dialog box appears.

20-162

Use the Case Set Manager to process a steady state solution on the problem. The primary purpose of the Case Set Manager is to allow the to set up different thermal analysis cases and to have the calculations made: from doing radiation calculations to creating and running the SINDA model to postprocessing temperatures with the click of a single button. When the Run Case button is clicked, Thermal Desktop will first calculate the radiation conductors and heating rates for all of the tasks set up for the current Case Set. Nodes and conductors are then computed and output. A SINDA model is then built and run. And finally, the temperature results are displayed mapped onto the thermal model in color. The may set up different Case Sets to be steady state or transient analyses. Each Case Set may have different start and stop times for transient runs. The can also have different SINDA Logic, property databases or aliases, or even different symbol values

Mapping Example (Continued) 22. Click on the Calculations tab.

Steady State in the Solution Type field is already selected. 23. Select OK to close the dialog box and return to the Case Set Manager dialog box. 24. Click on Run Case.

20-163

Mapping Example (Continued) The solution is processed. When complete, the data displays on the screen along with a SINDA/FLUINT Run Status dialog box confirming successful completion of the run. 25. Click on OK to close the dialog box.

Figure 20-39

20-164

Solution

Mapping Example (Continued) 26. Select Thermal > Export > Post Processing Data Mapper. The Mapper PP to XYZ Input File dialog box appears.

The stress analyst has added a more detailed model titled fine_quad.nas. and needs the temperatures for the nodes in this model in order to perform thermal stress calculations. The Map Object allows the Thermal Desktop to preview the stress model in side Thermal Desktop, align the two models, and map the postprocessed data to the stress model.

27. Choose NASTRAN in the Format field. 28. Choose fine_quad.nas in the Input File field drop down menu. 29. Ensure that Use World Coordinate System (WCS) is selected

The available format options are NASTRAN, ANSYS, FEMAP, and I-deas. The may choose to browse if the desired input file is not in the current directory. The may choose to align the stress model to the World Coordinate System (WCS) or the Coordinate System (UCS)

20-165

Mapping Example (Continued) 30. Select OK to close the dialog box. The Post Processing Data to XYZ Mapper dialog box appears.

20-166

Mapping Example (Continued) 31. Highlight the current value in the Output File field and type temps.out. 32. Under Tolerance, select Mapping Tolerance... 33. Type the following list in the Variable Tolerance form (one value per line). • • • • •

0 1e-6 2e-6 3e-6 4e-6

34. Select OK to close the dialog box. 35. Select Exit & Map to close the dialog box and map the data.

The values listed in the Variable Tolerance field will be used as tolerances for mapping the results to the structural model. If a single value is provided, that will be the only tolerance used. Progressive tolerancing (a list of increasing tolerances) will step through the values to map points that were not mapped at smaller tolerances. A progressive tolerance list is recommended over a single tolerance value. The can choose to map the data immediately using Exit & Map, or may simply Exit to manipulate the mapper graphical object (e.g. align the mapper to the thermal model)

The file temps.out contains mapped temperatures in the format necessary for them to be included into the Stress model. The engineer must cut and paste these into the proper place in the NASTRAN model in order to perform the thermal stress calculations.

20-167

Mapping Example (Continued) 36.


37. Choose List>Mesh Displayers/PP Mappers/BCM/Cutting Planes from the Model Browser menu bar. 38. Select MapperPP from the Mesh Displayers tree. or Display > Turn Visibility

39. Off.

The mapper turns off and the coarse finite element model is visible. Note the shape of the contours and the holes. 40.

or Display > Only on the Model Browser menu bar. The finite element model has been turned off (notice the nodes are gone) and the mapper is turned back on. Note that the shape of the holes has been refined, but the contours are the same.

20-168

After mapping is complete, the mapped data is displayed on the mapper. Any locations in the external file (the stress model) that did not get mapped will be displayed as grey.

Mapping Example (Continued)

Finite element thermal model based on coarse mesh

Mapper based on fine mesh

Figure 20-40

Compare Mapper to Model

41. Select File > Exit. Note: Thermal Desktop can also be exited using the Windows Close button (X) in the upper right corner of the screen. A Thermal Desktop/AutoCAD dialog box appears asking to save the drawing changes. 42. Select Yes.

20-169

20-170

20.9

or Example

What will be learned: • Using the Model Browser • Using layers to control object visibility • Using edge ors • Using face ors • Using material-based ors • Checking ors • Using Symbols and Expressions This tutorial demonstrates some of the capabilities of Thermal Desktop’s ors. The example model is a pipe with circular, plate fins and a small block-shaped sensor. The plate fins will be connected to the pipe by edge ors that could represent a braze or weld. The sensor is connected to the pipe by an area or that could represent a gasket or insulation. ors, like created with the tab in the object edit window, are used to thermally connect two objects. However, ors provide more functionality than such as: •

allowing a gap between the objects being connected

•

allowing material, radiation, one-way, and insulation defined connections

•

allowing the to select which objects will be included in the connection tests. Note: This tutorial is meant to show different capabilities of ors. To accomplish this, the best modeling options may not have been used. For example, using solid disks (with only one subdivision in the thickness direction) instead of planar disks for the fins would have eliminated having to specify the thickness of the fin for the or, but would have eliminated the practice with edge ors.

20-171

or Example 1. Double click on the file ors.dwg located in the Tutorials\Thermal Desktop\ors folder. Thermal Desktop opens with the ors drawing on the screen.

Figure 20-41

20-172

Initial View

or Example (Continued) 2.


Use model browser to set selection options and turn off objects that will not be modified during steps. The drawing should now appear similar to the view below:

Figure 20-42

3.

Using the Model Browser menu bar click on LIST to confirm that Surfaces/Solids is selected (check mark).

4.

If necessary, expand the Main tree by clicking on the + sign in front of the folder.

5.

Updated view

Click on FD Brick-Sensor.

6.

or Display > Turn Visibility Off in the Model Browser toolbar.

7.

Minimize the Model Browser window or move so it does not block the drawing.

8.

Press <ESC> to be sure the selection set is clear.

20-173


or Thermal > FD/FEM Network > or. The Command line should now read: Select faces ing from:

10. Select each of the four fins in the drawing area. The Command line should now read: Select faces ing from:

11. Press <ENTER>. The Command line should now read: Select surfaces ing to:

12. Select the pipe in the drawing area. The Command line should now read: Select surfaces ing to:

13. Press <ENTER>. The or dialog box appears. (next page) 14. Type Fins to pipe in the comment field. 15. From the drop-down list in the From field select Edges. 16. Select the Use Material check box. The Use Material field becomes active. 17. From the drop-down list in the Use Material field select Aluminum, 2024-T6.

20-174

Create edge ors to thermally connect the fins to the pipe. The connection is assumed to be an aluminum weld of thickness WeldThik. Important: ors do not calculate the conductance from the node to the edges of surfaces connected by ors. Therefore, that conductance should be included in the or coefficient or the better method is to use edge nodes. In this model, edge nodes are used for the fins and sensor. Important: In most cases, surfaces with smaller nodal areas should be chosen as the From surfaces. All integration points on a From surface are used in the testing algorithm.

or Example (Continued) 18. Double click in the Coeff field. The Expression Editor dialog box appears. Using right-click to select the symbol names from the general list, enter the following expression in the expression field: FinThickness/WeldThickness 19. Click OK to close the dialog box. The or dialog box updates to show 6 in the Coef field in bold type.

The SINDA conductor value resulting from a or is either kA/L, hA, or a Radk (i*Bij*Ai). When Use Material is selected for an edge or, the conductivity is obtained from the chosen material property and the length of the surface edge is taken from the geometry. Therefore, to complete kA/L, the Coeff value must be the thickness of the edge (FinThickness) divided by the distance in the direction of the heat flow (WeldThickness). To see definitions of the symbols being used, select the Add Symbol button while in the Expression editor.

20-175

or Example (Continued) 20. Select all four fins in the From: field. 21.

Edit From Object icon at the bottom of the or dialog box.

When the or dialog box closes, the graphical image for the ors are displayed: •

The from surfaces have green arrows on the edges (or faces) included in the or.

•

The to surfaces have gold arrows at the centroid of each face.

22. The Select Edges dialog box appears.

From To

23. Deselect all boxes labeled Not Used. 24. Select Min Radius so the box is checked. 25. Click OK to close the dialog box. The or dialog box updates to show Min Radius beside each object in the From field. 26. Click OK to close the or dialog box.

20-176

Figure 20-43

ors

or Example (Continued) 27. Thermal > Model Checks > Show or Markers. The Command line should now read:

Select ors to display markers for:

Use or markers to connection of the Fins to Pipe or. The or markers connect the integration points of the From edges to the nodes of the To surfaces.

28. Select any one of the or arrows or type All in the command line. The Command line should now read:

Select ors to display markers for: 29. Press <ENTER>. Yellow lines connecting the edges of the disks to nodes of the pipe appear. Changing to wireframe, , will make the markers more visible. 30.

toggles node visibility on.

31.

or Thermal > Model Checks > Clear /or Markers The or markers are cleared.

32.

toggles node visibility off.

Figure 20-44

or markers

Note: If the Coef field is 0, then the or markers will not be drawn. Ten or markers are drawn from the inner edges of each fin (Figure 11-58). The markers show which integration points have met the requirements of the testing algorithm. Any integration point not within the tolerance of the To surface would be marked with a red cross. When nodes are visible, note that not all pipe nodes are connected, but all inner edges of the fins are connected. Also note that To surface nodes can be connected multiple times, but test points are connected only once.

20-177

or Example (Continued) 33. Restore the Model Browser Window. 34. Select the four fins in the Model Browser tree. 35.

Turn Visibility Off icon in the Model Browser toolbar.

Create area or to thermally connect the sensor to the pipe. The connection is a per-area coefficient. The Model Browser typically minimizes to the upper left corner of the display.

The disks representing the fins disappear from the graphics window. 36. Select the Sensor. 37.

Turn On Visibility icon in the Model Browser toolbar. The brick representing the sensor appears in the graphics window.

38. Minimize the Model Browser window. 39. Select the sensor. 40.


The or command accepts preselected items (the sensor in this case) as the from object(s) and moves on to selecting To surfaces.

The Command line should now read: Select surfaces ing to:

41. Select the pipe in the drawing area. The Command line should now read: Select surfaces ing to:

42. Press <ENTER>. The or dialog box appears. 43. Type Sensor to pipe in the comment field. 44. Highlight the current value in the Tolerance field and type 0.008. 45. Highlight the current value in the Conduction Coefficient field and type 5. 46. Choose Ray Trace Algorithm in the Inputs for Connection Algorithm Section. 20-178

For the ray trace algorithm, a ray of length Tolerance is shot normal to the From surface at each test point. If the ray intersects with a To surface, the ray terminates and a connection is made. By contrast, the point algorithm checks for surfaces within tolerance in all directions.

or Example (Continued) 47. Double-click the sensor in the From field. 48. Deselect all check-boxes except YMAX. 49. Select OK to close the Select Faces dialog.

The or is limited to only the side where the actual is made. Using ray tracing and tolerance would also eliminate the non- sides, but calculation time would be required to determine this. Selecting only certain sides, prevents unnecessary calculations.

50. Select OK to close the or dialog box. A green or arrow appears from the side of the sensor brick closest to the pipe. 51. Select the green or arrow on the sensor. (text typed in the comment field will appear with the or identification when the cursor is held over a or arrow)

Use or markers to connection of the Sensor to Pipe or.

52. Thermal > Model Checks > Show or Markers. A window may open showing that some ors are only 70% complete. This is expected because of the tolerance chosen. If this window does open, close it to proceed.

20-179

or Example (Continued) 53. View > 3D Views > Top. The markers can be seen more easily. 54.

When the view is changed to Top, the graphics display should look like this:

or Thermal > Model Checks > Clear /or Markers The or markers are cleared.

55. View > 3D Views > SW Isometric. 56. Restore the Model Browser Window. 57. Select the four fins in the Model Browser tree. 58.

Turn On Visibility icon in the Model Browser toolbar. The fins appear in the graphics window.

59. Minimize the Model Browser window.

Figure 20-45 or markers with inactive test points

Since the ray trace algorithm was chosen, the markers appear as rays normal to the From surface. Test points that are not within the tolerance are shown in red.

20-180

or Example (Continued) 60. Using the Layer Drop Down Menu, select Boundary as the current layer. The Layer Drop Down Menu originally says Objects and is located next to the button. The drop down menu is accessed by clicking on the down-arrow to the right of the word Objects. 61.

or Thermal > Surfaces/Solids > Rectangle Origin point <0, 0, 0> appears in the

Create a small surface to represent the ambient air so ors can be created from the fins to the air. Ambient conditions will typically be represented by a defined node, but conductors between nodes and surfaces do not allow edge conduction. Therefore, a small surface will be created to represent convection from the edges of the fins and allow a or. Other options would be to use FD Solid Cylinders for the fins or ignore the convection from the edge of the fin.

Command Line area. 62. Type -0.1,0,0 in the Command Line, then return. Point for +X and X-size <@1, 0, 0>

appears in the Command Line area. 63. Type @0.01,0,0 in the Command Line then return. Point to set XY plane and Y-size <@0,1, 0> appears in the Command

Line area. 64. Type @0,0.01,0 in the Command Line then return. The Thin Shell Data dialog box appears. 65. Click on the Cond/Cap tab. 66. Click on Generate Cond/Cap button to open the Expression Editor 67. Type 0 into the Expression. 68. Click OK to close Expression Editor. 69. Click OK to close the Thin Shell Data dialog box. 70.

A value of zero in the expression prevents the surface properties from being generated for a SINDA file. If radiation analyses were to be performed, “Not in analysis group” should be selected for all analysis groups.

or View > Zoom > Extents

20-181

or Example (Continued) 71. Restore the Model Browser Window. 72. Expand CC Not Generated. 73. Expand Rect 74. Select Main.1 75.

If CC Not Generated is not visible, use the Rebuild Tree icon

Edit Selected Objects icon in the Model Browser toolbar. The Node Edit dialog box opens.

76. Select Override calculations by elements/surfaces (place a check mark in the box). 77. Select Boundary (place a dot in the circle). 78. Type Ambient in the Comment field. 79. Double-click in the Submodel field to highlight MAIN. 80. Type AMBIENT into the Submodel field. 81. Select OK to close the dialog box. 82. Select Yes to add submodel AMBIENT.

20-182

Override the surface node to create a boundary node. If Node Visibility is on, a triangle appears at the center of the surface. Even though the ambient node will not be generated as part of the surface, it will now be generated as a node.


or Thermal > FD/FEM Network > or. The Command line should now read:

Create convection from faces of fins using ors. Note new or graphical objects when complete.

Select faces ing from:

84. Select all four fins in the drawing area. The Command line should now read:

Here the order of selecting From and To is important. The area of the is calculated using the From surfaces.


85. Press <ENTER>. The Command line should now read: Select surfaces ing to:

86. Select the newly created small surface in the drawing area. The Command line should now read: Select surfaces ing to:

87. Press <ENTER>. The or dialog box appears. 88. Type Fin faces to ambient in the comment field. 89. Highlight the current value in the Conduction Coefficient field and type 50. 90. Select all four fins in the From: field. 91.

Edit From Object icon near the bottom of the or dialog box. The Select Faces dialog box appears.

92. Select both top and bottom (check marks should be in both boxes). 93. Click OK to close the dialog box. 94. Click OK to close the or dialog box.

20-183

or Example (Continued) 95. View > 3D Views > Front. 96. Select one of the green or arrows on the faces of the fins.

Use or markers to connection of the Fin faces to Ambient or.

97. Thermal > Model Checks > Show or Markers. 98. Read the warning message that appears. This is not a problem in this situation since the fin surface areas are needed and the area of the boundary surface is not used. 99. Close the Warning window. 100. or Thermal > Model Checks > Clear /or Markers The or markers are cleared. 101.View > 3D Views > SW Isometric.

20-184

Notice the separation between the ends of the markers and the fins in the view below: Fin Markers

Figure 20-46 thickness

ors with surface

This happens because the test points are allowing for the thickness of the fins.

or Example (Continued) 102. or Thermal > FD/FEM Network > or.

Create convection from edges of fins using ors.


103.Select all four fins in the drawing area. The Command line should now read: Select faces ing from:

104.Press <ENTER>. The Command line should now read: Select surfaces ing to:

105.Select the small surface in the drawing area. The Command line should now read: Select surfaces ing to:

106.Press <ENTER>. The or dialog box appears. 107.Type Fin edges to ambient in the comment field. 108.From the drop-down list for the From field select Edges. 109.Double-click in the Conduction Coefficient field. The Expression Editor dialog box appears, 110. Right-click to select the symbol names in the general list and enter the following expression in the Expression field 50*FinThickness 111. Click OK to close the dialog box.

Again, the From objects are used for the geometry. For an edge , the length of the edge is calculated so the thickness of the edge must be included in the coefficient.

The Conduction Coefficient field now shows 0.3.

20-185

or Example (Continued) 112. Select all four fins in the From: field. 113.

Edit From Object icon.

The Select Edges dialog box appears. 114. Deselect all checked boxes. 115. Select Max Radius so the box is checked. 116. Click OK to close the dialog box. 117. Click OK to close the or dialog box. 118. Restore the Model Browser if not currently visible 119. Expand CC Not Generated and Rect, if not already expanded. 120.Notice the or listings contain the comment typed in when the or was created. 121.Select File > Exit. A Thermal Desktop/AutoCAD dialog box appears asking to save the drawing changes.


122.Select Yes. Running the model requires a heat load or a fixed boundary temperature (different than the boundary node already defined). The can add a heat load to the pipe surface or change the pipe nodes to boundary nodes and obtain a solution. The node IDs must also be resequenced. If the runs a transient solution, the time step taken by SINDA/FLUINT will likely be very small. This is caused by the large conductance at the base of the fins: G = k*A/L. If the adds a factor to the coefficient of the fin-to-pipe or, he or she can see how the conductance of the or can affect the solution time step.

20-186

20.10

Parameterizing for a Common Input

What will be learned: • Inserting a model into another drawing file • Importing and renaming symbols Prerequisites: • Section 20.2 - Setting Up a Template Drawing Sometimes it is common that the same geometry is used over and over in the same model. This geometry can be as simple as a five or six-sided box or can be much more complicated. In this example, a simple five sided-box with centered nodes that has conductors between each of the faces has been created. The geometry has been parameterized so that the can simply change the x, y, and z sizes so that the box will automatically update when the changes the symbols. Parameterized Box Example 1. Double click on the file box.dwg located in the Tutorials\Thermal Desktop\Parameterized Box For Insert folder. Thermal Desktop opens with the drawing on the screen.

Figure 20-47

Box Drawing Initial View

20-187

Parameterized Box Example (Continued) 2. Select Thermal > Symbol Manager. The Symbol Manager dialog box appears.

Take a few moments to examine the model. Bring up the Symbol Manager. Notice the parameters for x, y, and z sizes. There is also a parameter for the thickness of the faces on the box.

3. Select Done to close the dialog box. 4. Click on the top of the box to select it. or Thermal > Edit.

5.

The Thin Shell Data dialog box appears. 6. Select the Surface tab.

Look at some of the data that make up the top of the box. After selecting the top, utilize Thermal Desktop’s Edit function to display the Thin Shell Data dialog box. Notice the lengths of the X and Y axes are parameterized (Surface tab) and that the Z translation is also programmed (Trans/ Rot tab). Take a moment to select some of the other surfaces and conductors to get an idea of how they are programmed.

7. Look at the X Max and Y Max fields.

20-188

Parameterized Box Example (Continued) 8. Select the Trans/Rot tab.

9. Look at the Translation Z field. 10. Select OK to close the dialog box. 11. Repeat the process for some of the other surfaces and conductors as desired. 12. Select File > Exit. A Thermal Desktop/AutoCAD dialog box appears asking to save the drawing changes.

Close box.dwg saving it to the current AutoCAD version and exit Thermal Desktop.

13. Select Yes. 14. Copy the template thermal.dwg file created in the first tutorial to the \Tutorials\Thermal Desktop\Parameterized Box For Insert directory. Note: Be sure to hold the key down if dragging the template file icon to the new directory so that the file is copied, rather than moved. 15. Rename the copied template file to parameter. 16. Start Thermal Desktop by double clicking on the parameter drawing file icon in the Parameterized Box For Insert directory.

20-189

Parameterized Box Example (Continued) 17. Select Insert > Block. The Insert dialog box appears. 18. Select Browse. The Select Drawing File dialog box appears with two drawings displayed in the drawing field. 19. Select box.dwg to highlight it.

Bring the box drawing into the template copy. Each will have a somewhat different image appear in the drawing area because of the insertion point selections but, in general, the drawing should appear similar to the drawing below once zoomed extents is performed.

20. Select Open. The Insert dialog box reappears with box displayed in the Name field. 21. Select Specify On-Screen in the Insertion field (place a check mark in the box) if not already selected.

Figure 20-48

22. Select OK to close the dialog box. The Command Line should now read: Specify insertion point or [Basepoint/Scale/X/Y/Z/ Rotate]:

23. Click at any point on the screen to place the box. (Some versions of AutoCAD may prompt for scaling factors at this point) A close view of a box corner appears in the drawing area.

20-190

Box Drawing

Parameterized Box Example (Continued) 24.

or Zoom > Extents.

25. Select Modify > Explode. The Command Line should now read: Select objects:

Explode the box so that it is no longer an AutoCAD Block, but are individual Thermal Desktop entities. Once Explode is performed, individual components of the box can be selected rather than only the whole box.

26. Click on the box. The whole box is selected and the Command Line should now read: Select objects:

Press <Enter>. 27. Click on the top of the box. Notice that only the top is now selected. 28. Select other surfaces as desired and press <Esc> to deselect when finished.

20-191

Parameterized Box Example (Continued) 29. Select Thermal > Symbol Manager. The Symbol Manager dialog box appears and the dialog box’s fields are empty. 30. Select Import. The Open dialog box appears. 31. Use the drop down to the right of the File name field to choose All files (*.*) 32. Select boxsymbols.sym to highlight it and then select Open. The Select the properties to import dialog box appears. 33. Select thickbox, xbox, ybox and zbox and select OK. The Symbol Manager dialog box reappears with the symbols displayed.

20-192

Use the Symbol Manager Import command to import the file boxSymbols.sym. Xbox, ybox, zbox and thickness symbols are imported. Symbols can be exported into SYM files from the symbol manager. Symbols can be imported either from a SYM file or directly from a DWG file.

Parameterized Box Example (Continued) 34. Select the 4 symbols: • Click on thickbox to highlight it and then, hold down the <Shift> key and click on zbox.

The box is to be imported multiple times, so the imported symbols must be renamed. Append _1 (underscore 1) to each of the symbol current names.

35. Select Rename on the right side of the form. The Multiple Rename dialog box appears. 36. Select the radio button beside Append String to each existing string. 37. Type _1 in the .

38. Select OK. The Symbol Manager dialog box reappears displaying the new symbol names.

20-193

Parameterized Box Example (Continued) 39. Select xbox_1. 40. Select Edit. The Expression Editor dialog box appears with the current xbox_1 information.

Edit the symbols for xbox, ybox, and zbox to be one tenth of their original values. The box changes in the drawing area to reflect the changes in size.

Note: Double clicking on a symbol also displays the Expression Editor. 41. Highlight the current value in the Expression field if not already highlighted and type 10. 42. Select OK. The changed parameters for xbox_1 are reflected. 43. Repeat the process for ybox and zbox, changing the values to 1/10 of the current value. The Symbol Manager dialog box displays the new values. 44. Select Done to close the Symbol Manager dialog box.

20-194

Figure 20-49

Edited Box

Parameterized Box Example (Continued) 45. Select Insert > Block. The Insert dialog box appears with box in the Name field highlighted.

Insert another box. Use the Insert > Block command, but this time it is not necessary to reselect the box.dwg file, simply change the Name: pull down to box and select OK. Select any point on the screen to place the box, and then explode the box.

Specify On-Screen in the Insertion field is already selected from the previous insertion. 46. Select OK to close the dialog box. A box is attached to the cursor waiting for insertion. The Command Line should now read:

Figure 20-50

Second Box Inserted

Once the second box is inserted, explode it so the individual entities.

Specify insertion point or [Basepoint/Scale/X/Y/Z/ Rotate]:

47. Click at any point on the screen to place the new box. The new box appears on the screen. Notice the difference in the sizes of the two boxes—the first box’s size was changed to 1/10 of its original size (xbox_1). 48. Select Modify > Explode. The Command Line should now read: Select objects:

49. Select the newly placed box. The Command Line should now read: Select objects:

50. Press <Enter>. The second box is exploded. 20-195

Parameterized Box Example (Continued) 51. Click on various parts of the box as desired to confirm “explosion” and press <Esc> when finished. 52. Select Thermal > Symbol Manager. The Symbol Manager dialog box reappears with the symbols renamed earlier displayed. 53. Right-click the general tab and select Add New Group to create a new Symbol Group.

A new tab named group1 is displayed. 54. Click on the general tab.

55. Select all 4 symbols. 56. Select Edit.

20-196

The symbols listed in the Symbol Manager belong to a group named general. Add a second symbol group named group1.

Parameterized Box Example (Continued) The Symbol Edit - Multi Edit Mode dialog box appears.

57. Click on the arrow next to the Group field and select group1 from the drop-down list. 58. Select OK to close the Symbol Edit Multi Edit Mode dialog box. The Symbol Manager dialog box shows the symbols are removed from the general symbol group and moved to group1. The general tab is empty. Note: Click on the group1 tab to the move

20-197

Parameterized Box Example (Continued) 59. Select the general tab if not already selected.

Add another set of symbols and append the names of the entities with _2.

60. Select Import.

Symbols can be imported directly from DWG files as well as exported symbol files.

The Open dialog box appears. 61. Select box.dwg to highlight it and then select Open. The Select the properties to import dialog box appears. 62. Select thickbox, xbox, ybox and zbox and select OK. Another set of the original symbols are imported and display in the general tab. 63. Select the 4 newly imported symbols. 64. Select Rename. The Multiple Rename dialog box appears. 65. Type _2 in the Append string field and that Append String to each existing name is selected. 66. Select OK. The Symbol Manager dialog box reappears displaying the new symbol names.

67. Select xbox_2.

20-198

Also change the values of xbox, ybox and zbox to 1/2 of the current values.

Parameterized Box Example (Continued) 68. Select Edit. The Expression Editor dialog box appears with the current xbox_2 information.

The second box reflects the change in size.

69. Highlight the current value in the Expression field if not already highlighted and type 50 (1/2 of the current value of 100). 70. Select OK to close the Expression Editor dialog box.

Figure 20-51

Second Box Edited

The edited value displays in the Symbol Manager. 71. Repeat this process for ybox and zbox, changing the current values by 1/2. The Symbol Manager dialog box reflects the changes. 72. Select Done to close the Symbol Manager. These steps may be repeated for as many boxes that are in the model. 73. Select File > Exit. A Thermal Desktop/AutoCAD dialog box appears asking to save the drawing changes.


74. Select Yes.

20-199

20-200

20.11

Dynamic SINDA Example

What will be learned: • Using symbols and expressions • Using symbol-based Articulators to reposition objects • Setting up a Dynamic SINDA solution Prerequisites: • Section 20.2 - Setting Up a Template Drawing This example uses Thermal Desktop’s Dynamic Solver interface to optimize the component (cylinder and box) locations and the thickness of the doubler plate such that the mass of the plate is minimized. Constraints will also be placed on the components such that their individual temperatures limits are not violated. The components are connected to the plate via conductance. Please reference Section 5 of the SINDA/FLUINT manual for a detailed documentation of the Advanced Design Modules such as the SINDA Solver. The exercise consists of three parts, or steps: • Step 1: Parameterizes the locations of the box and the cylinder, so that their best location can be found by the Solver interface. • Step 2: Sets up the problem in the Case Set Manager. • Step 3: Solves the problem.

20-201

Dynamic SINDA Example 1. Double click on the file dynamic.dwg located in the Tutorials\Thermal Desktop\Dynamic folder. Thermal Desktop opens with the dynamic drawing on the screen.

Figure 20-52 Dynamic SINDA Initial View

Take a few moments to examine the model. There are several layers. Notice that the cylinder and the box are currently on top of each other at the origin. Also notice the heat loads on the top of the cylinder and on a node on the box. Finally, notice that a space node has been created for radiation to the environment.

20-202

Dynamic SINDA Example (Continued) 2. Select Thermal > Symbol Manager. The Symbol Manager dialog box appears.

Create symbols for the box and the cylinder to set up to parameterize the location of the box and cylinder. Notice that much of the model is already parameterized.

3. Type xbox in the New Symbol Name field and select Add. The Expression Editor dialog box appears.

20-203

Dynamic SINDA Example (Continued) 4. Type 0.1 in the main input field and select OK to close the dialog box. The New Symbol Manager dialog box updates to display xbox in the main general list area. 5. Repeat the process to create symbols for the following: • YBOX= 0.1 • XCYL = 0.45 • YCYL = 0.15 6. Select Done to close the dialog box.

20-204

Dynamic SINDA Example (Continued) or type layer in the Command

7. line.

Use Thermal Desktop’s Layer functionality by making the box layer the current layer and turn off all other layers.

Note: The menu selection Format > Layer may also be used. The Layer Properties Manager dialog box appears.

8. Select the box layer to highlight it. and select Current

.

A green check mark appears next to the layer box and Current Layer changes from 0 to box. 9. Click the Freeze (sun) icons for all of the layers except the box layer to freeze them (change to a snowflake). 10. Close the Layer Properties Manager. Figure 20-53

Box Layer Current

20-205

Dynamic SINDA Example (Continued) 11.

or select Thermal > Articulators > Create Assembly. The Command line should now read: Enter origin of articulator:

12. Type 0,0,0 in the Command line. The Edit Assembly dialog box appears. 13. Highlight the current value in the Name field and type BOX. 14. Highlight the current value in the Size field and type 0.1.

15. Select OK. Red, green and blue lines along the X, Y and Z axes appear on the screen representing the box assembly. Assembly axes may be partly obscured by the coordinate system.

20-206

This part of the exercise creates an assembly—a collection of surfaces associated with a single coordinate system—for the box. The assembly coordinate system is displayed in the graphics area on the screen. The translations will be edited after the geometry is attached to the assembly. The order of this is very important.


or select Thermal > Articulators > Attach Geometry. The Command line should now read: Select an articulator:

17. Create a selection box around the box and coordinate system. The Command line should now read: Select objects to attach to articulator...:

18. Create a selection box around the box.

Geometry is attached to the assembly. When the assembly is modified, via a rotate or a move, the location of the surfaces attached to that assembly will also be modified. An assembly can be attached to another assembly, and the nesting can be infinitely deep. Once this occurs, when the assembly is moved, the geometry will move with it. The Assembly coordinate system (red, green and blue axes) may not be visible because of the UCS icon. Zoom extents may help visualization.

The Command line should now read: Select objects to attach to articulator...:

19. Press <Enter>. The geometry is attached. Note: The Command line reflects the change. Another way to the objects were attached is to press to view the command line text window.

20-207

Dynamic SINDA Example (Continued) 20. Draw a selection box around the box and articulator to highlight them. or Thermal > Edit.

21.

The Object Selection Filter dialog box appears. 22. Select Assembly[1] and click on OK.

Edit the box assembly. Z translation is entered so that the conductance works properly. When the editing is complete, the graphic is updated in the drawing area to show 2 axes connected by a blue line. The original assembly is at 0,0,0 while the evaluated assembly is at 0.1, 0.1, 0.005.

The Edit Assembly dialog box for BOX appears. 23. Select the Trans/Rot tab.

Figure 20-54

24. Double click in the Translation X field to display the Expression Editor dialog box.

20-208

Geometry Attached

Dynamic SINDA Example (Continued) 25. Type xbox in the Expression field and select OK to close the dialog box. Note: Upper or lower case letters may be entered. The Edit Assembly dialog box updates to show 0.1 in bold type in the Translation X field. 26. Double click in the Translation Y field to display the Expression Editor dialog box.

When an expression is used to define an input field, the resulting value of the expression is shown in bold type in the input field. If the expression has an error, the field will be highlighted in red. When these steps are completed the model should look something like this:

27. Right-click in the expression field, point to general and select ybox. Select OK to close the dialog box. The Edit Assembly dialog box updates to show 0.1 in bold type in the Translation Y field. 28. Type 0.001 in the Translation Z field. 29. Click OK to close the dialog box. The Z translation provides separation between the box and the plate for radiation calculations and the or calculations.

20-209

Dynamic SINDA Example (Continued) or type layer in the Command

30. line.

The Layer Properties Manager dialog box appears.

Turn visibility for the cylinder layer on and visibility for the box layer off. Display only the cylinder by making that layer the current layer.

31. Click on the Freeze (snow flake) icon for the cylinder layer to thaw the layer. 32. Select the cylinder layer to highlight it and select Current

.

A green check mark appears next to the layer and Current Layer changes from box to cylinder. 33. Click on the Freeze (sun) icon for the box layer to freeze the layer. 34. Close the Layer Properties Manager.

20-210

Figure 20-55

Cylinder Layer Current


or select Thermal > Articulators > Create Assembly. The Command line should now read:

Create an assembly at the origin for the cylinder. The order of attaching the items before putting in the translations is very important.

Enter origin of articulator:

36. Type 0,0,0 in the Command line. The Edit Assembly dialog box appears with the last selected tab (Trans/Rot) displayed. 37. Select the Assembly tab. 38. Highlight the current value in the Name field and type CYLINDER. 39. Highlight the current value in the Size field and type 0.1. 40. Select OK. Red, green, and blue lines along the X, Y, and Z axes appear on the screen representing the cylinder assembly. These axes may be obscured by the coordinate system.

20-211

Dynamic SINDA Example (Continued) Attach the cylinder to the assembly. 41.

or select Thermal > Articulators > Attach Geometry. The Command line should now read: Select an articulator:

42. Click on the green assembly articulator. The Command line should now read: Select objects to attach to articulator...:

43. Create a selection box around the cylinder. The Command line should now read: Select objects to attach to articulator...:

44. Press <Enter>. The geometry is attached.

20-212

Dynamic SINDA Example (Continued) 45. Draw a selection box around the everything to highlight. 46.


47. Select Assembly[1] and click on OK.

Edit the cylinder assembly. As with the box, the Z translation is entered so that the or works properly. When an expression is used to define an input field, the resulting value of the expression is shown in bold type in the input field.

The Edit Assembly dialog box for CYLINDER appears. 48. Select the Trans/Rot tab. 49. Double click in the Translation X field to display the Expression Editor dialog box. 50. Type xcyl in the Expression field and select OK to close the dialog box. Note: Upper or lower case letters may be entered. The Edit Assembly dialog box updates to show 0.45 in bold type in the Translation X field. 51. Double click in the Translation Y field to display the Expression Editor dialog box. 52. Type ycyl in the Expression field and select OK to close the dialog box. The Edit Assembly dialog box updates to show 0.15 in bold type in the Translation Y field. 53. Type 0.001 in the Translation Z field.

20-213

Dynamic SINDA Example (Continued) When the editing is complete, the graphic is updated in the drawing area to show 2 axes connected by a blue line. The original assembly is at 0, 0, 0 while the evaluated assembly is at 0.45, 0.15, 0.001. When complete, the view should look similar to the example below.

54. Click OK to close the Edit Assembly dialog box. The drawing area is updated and the cylinder moved out of the current viewing area.

Figure 20-56

Cylinder Assembly

or Zoom > Extents.

55.

or type layer in the Command

56. line.

Reactivate visibility for the box, plate and space layers so all three models and the ambient node are visible.

The Layer Properties Manager dialog box appears. 57. Click on the Freeze icons for the box, space and plate layers to thaw them. 58. Close the Layer Properties Manager.

Figure 20-57

20-214

Layer Visibility Changes

Dynamic SINDA Example (Continued) 59. Select View > 3D Views > SE Isometric

Change the orientation of the model view and edit the default Case Set.

or Thermal > Case Set Man-

60. ager.

The Case Set Information dialog box appears. 61. Select Edit. 62. Select the Radiation Tasks tab if not already visible. 63. Select Add. The Radiation Analysis Data dialog box appears.

Add a radk job and edit the properties to use the same random number seed. This will help the program give consistent results from radk run to radk run. More importantly, in a real-world analysis, the analyst should ensure that enough rays are shot to provide consistent results.

20-215

Dynamic SINDA Example (Continued) 64. Select the Advanced Control tab.

65. Select Use same random number seed sequence at start of every node in the Random Number Seed Control field (put a dot in the circle). 66. Click OK to close the Radiation Analysis Data dialog box. 67. Select the Calculations tab in the Editing Case Set dialog box.

68. Type 0.0001 in the Max Temperature Change field

20-216

Tightening up the convergence criteria will help the Solver converge on a solution.

Dynamic SINDA Example (Continued) 69. Select the Output tab in the Case Set Information dialog box. 70. Click on Temperatures in the Text Output field to deselect it (remove the check mark).

These runs can quickly generate a lot of data if these output options are left on. The temperatures will still be able to be seen while it calculates.

71. Click on Temperatures in the Output for Color Postprocessing and XY Plots field to deselect it (remove the check mark).

20-217

Dynamic SINDA Example (Continued) 72. Select the SINDA tab in the Case Set Information dialog box.

Replace the current OPERATIONS input text with CALL SOLVER. Note: CALL SOLVER is a FORTRAN program and must start in the 7th column.

73. Double click on OPERATIONS in the Global S/F Input field. The Operations Data dialog box appears.

74. Highlight the last three lines of text and type CALL SOLVER. Note: Formatted FORTRAN requires that the lines of code start after the 6th column, so the can be replaced by six or more spaces.

20-218

Dynamic SINDA Example (Continued) CALL SOLVER is all that remains as uncommented text. The ‘C’ of CALL must start in the 7th, or higher, column. 75. Select OK to close the Operations Data dialog box. 76. Select the Dynamic tab in the Case Set Information dialog box.

77. Click on Use Dynamic SINDA and Show Temps While Calculating to select them (place check marks in the boxes). Note: Leave Reset Symbols to Original Values selected.

When Use Dynamic SINDA is selected, the program opens a connection between SINDA and Thermal Desktop so that they can communicate to change the design variables.

The option to reset symbols upon completion prevents undesired Solver results from remaining in the model. The results can be obtained from the Solver output to modify the model if the chooses.

20-219

Dynamic SINDA Example (Continued) 78. Double click on Design in the Solver Data column. The Solver Design Variables dialog box appears.

To achieve the goal of minimizing the doubler plate mass, specify which parameters must be manipulated. Additional design variables and parameters will be defined, by editing several of the subroutines listed in the Solver Data field.

79. Double click on xbox in the Global Symbols field. The Define Variables dialog box appears.

80. Select Min Value to place a check mark in the box and activate the input field. 81. Type .05 in the Min Value field. 82. Select Max Value to place a check mark in the box and activate the input field. 83. Type .3 in the Max Value field. 84. Select OK.

20-220

Figure 20-58

Solver Data Field

In this problem, the components will be allowed to move in their XY locations in addition to varying the thickness of the plate. Note the minimum and maximum X and Y values are defined to prevent the components from moving off the plate and from moving past the centerline of the plate. The plate thickness must be at least 1 mil. Solver design variables for the box, the cylinder and the plate must be defined.

Dynamic SINDA Example (Continued) The Solver Design Variables dialog box reappears with the variables for xbox displayed in the Solver Design Variables field.

Use the table shown below for variable input values. OBJECT

MIN VALUE MAX VALUE

XBOX

0.05

0.3

YBOX

0.0375

0.2625

XCYL

0.3

0.57

YCYL

0.03

0.27

TPLATE

0.001

85. Repeat the process for the ybox, xcyl, ycyl and tplate. Use the Solver Design Variables table shown to the right as a reference. Note: Note that the tplate does not have a maximum value. When complete, the Solver Design Variables dialog box should look similar to the graphic below:

86. Select OK to close the Solver Design Variables dialog box. The Case Set Information dialog box is visible. An asterisk (*) is displayed next to Design in the Solver Data field to show the variables have been changed.

20-221

Dynamic SINDA Example (Continued) 87. Double click on Constraint in the Solver Data column. The Solver Constraint Variables dialog box appears.

What distinguishes a viable design from a bad design must be defined for Solver. For this problem, if the box or cylinder exceed their maximum allowable temperature, the design is invalid. This type of information must be defined as constraint data. Solver constraint variables must be defined as follows. OBJECT

The Global Symbols field is greyed out and cannot be accessed. 88. Type box.t8 in the text input field at the bottom of the dialog box and select Add--->. The Define Variables dialog box appears with box.t8 displayed in the Name field.

89. Select Max Value to place a check mark in the box and activate the input field. 90. Type 310 in the Max Value field. 91. Select OK to close the dialog box and return to the Solver Constraint Variables dialog box.

20-222

MIN VALUE MAX VALUE

BOX.T8

310

CYL.T112

340

Use the above table for variable values input.

Dynamic SINDA Example (Continued) 92. Repeat the process for the cyl.t112. Use the Constraint Variables table as a reference (for the max value). Note: If box.t8 is displayed in the Solver Design Variables dialog box input field, simply highlight the text and type cyl.t112. When complete, the Solver Constraint Variables dialog box should look similar to the graphic below:

93. Select OK to close the Solver Constraint Variables dialog box. The Case Set Information dialog box is visible. An asterisk (*) is displayed next to Constraint in the Solver Data field to show the variables have been changed.

20-223

Dynamic SINDA Example (Continued) 94. Double click on Control in the Solver Data column. The Solver Control Information dialog box appears.

95. Highlight the current value in the Maximum iterations field and type 1000. 96. Select OK to close the dialog box. The Case Set Information dialog box is visible. An asterisk (*) is displayed next to Control in the Solver Data field to show the variables have been changed.

20-224

For Solver Control, change the Maximum iterations from 100 to 500. The Solver will minimize the OBJECT which will be defined in the next few steps.

Dynamic SINDA Example (Continued) 97. Double click on Procedure in the Solver Data column. The Solver Procedure dialog box appears.

This is FORTRAN code, so all the text must begin in Column 7. For information on these Subroutine calls, please see “Subroutine Calls from SINDA to Thermal Desktop” on page 161. If you examine the default text in the window, you will notice that most of the text below is already included. Just delete the “C” from the first column. For Solver Procedure, input use the following data:

98. Highlight the current code and comments in the dialog box and type the text shown in the chart in the right hand column.

CALL TDSETDES CALL TDCASE CALL STEADY OBJECT = TPLATE CALL TDOBJ CALL REGTAB

Note: that the text must begin at column 7. (Enter 6 spaces before beginning the first line and type CALL beginning on the 7th space.)


20-225

Dynamic SINDA Example (Continued) The Case Set Information dialog box is visible. An asterisk (*) is displayed next to Procedure in the Solver Data field to show the variables have been changed. 100.Select OK to close the Case Set Information dialog box and return to the Case Set Manager dialog box.

Save the drawing and the run the case. As the model runs, notice the box and cylinder move around the drawing area. Once postprocessed, the commands will quickly follow to move the box and cylinder, thus making a hot spot on the board when the object is no longer in that location.

101.Select Save drawing before runThe Dynamic SINDA status window can ning to place a check mark in the box. be expanded to show all design variables and the object. 102.Select Run Case. Look at the output file, case.out, in the current directory. The best solution is found at the end of the file. The temperatures may be slightly greater than the maximum constraint input, but they are within the constraint violation control parameter.(Solver Control > Advanced tab).

Figure 20-59 Solution with radiation (results will be slightly different without radiation)

Since the symbols were reset to the original values, the results will look somewhat odd. To see the geometry as it was at the end of the solution, the must open the Symbol Manager and import dynamicSymbols.sym. Replace any or all of the symbols.

20-226

Dynamic SINDA Example (Continued) 103.Select File > Exit. A Thermal Desktop/AutoCAD dialog box appears asking to save the drawing changes.


104.Select Yes.

20-227

20-228

21 RadCAD® Tutorials RadCad Tutorials is a continuation of the tutorials presented in Chapter 20: Setting Up a Template Drawing. It is assumed the has completed the tutorials in that chapter before beginning the new tutorials presented in this chapter. Chapter 20: Setting Up a Template Drawing introduce the to how things work inside of Thermal Desktop, as well as the nomenclature used in the tutorials. The tutorials in this chapter all focus on setting up and analyzing models for radiation and heating rate purposes. There are seven tutorials, as follows: • Section 21.1: Radks for Parallel Plates on page 21-3 • Section 21.2: Space Station Oct Tree Example on page 21-23 • Section 21.3: Importing a TRASYS Model and Using Articulators on page 2135 • Section 21.4: Orbital Heating Rates on page 21-53 • Section 21.5: Simple Satellite on page 21-71 • Section 21.6: Orbital Maneuvers on page 21-87 Even though most of these tutorials are based on spacecraft (RadCAD originated in the aerospace industry, after all), lessons learned in each of these tutorials may be applied to other systems, as well. In other words, do not skip over a tutorial simply because it uses a spacecraft; the lesson are still useful if you plan to use RadCAD.

RadCAD® Tutorials

21-1

21-2

RadCAD® Tutorials

21.1

Radks for Parallel Plates

What will be learned: • Overview of how RadCAD works • Defining and asg Optical Properties • Overview of Radiation Calculation functionality Prerequisites: • Section 20.2 - Setting Up a Template Drawing In this example, a set of parallel plates will be constructed, and the radks between them and to space will be computed. This example is intended to give an overview of RadCAD in Thermal Desktop; later examples will delve into RadCAD concepts in more detail. The parallel plates will be identical and directly opposed to each other. Length and width will be 10 X 5 inches. The plates will be separated by 12 inches. Surface 1 will be defined as the lower pate, leaving the upper plate designated as surface 2. Overview Parallel Flat Plates 1. Copy the template thermal.dwg file created in the first tutorial to the \Tutorials\Thermal Desktop\parallel directory. Note: Be sure to hold the key down if dragging the template file icon to the new directory so that the file is copied, rather than moved. 2. Rename the copied template file to parallel. 3. Start Thermal Desktop by double clicking on the parallel drawing file icon in the parallel directory. 4.


21-3

Overview Parallel Flat Plates (Continued) 5.

or Thermal > Optical Properties > Edit Property Data. The Edit Optical Properties dialog box appears.

6. Type White Paint in the New Property to add field. 7. Select Add. The Edit Optical Properties-White Paint dialog box appears.

21-4

Define the optical property “White Paint”. Spaces are allowed in optical property names. Solar absorptivity = 0.23 Infrared emissivity = 0.8

Overview Parallel Flat Plates (Continued) 8. Type 0.23 in the Solar Absorptivity field. 9. Type 0.8 in the Infrared Emissivity field. 10. Select OK to close the Edit Optical Properties-White Paint dialog box. The Edit Optical Properties dialog box reappears and reflects the changes. 11. Select OK to close the Edit Optical Properties dialog box. 12.


The model will be built in inches. Notice that the energy units are in Joules, time in seconds, thus the energy rate units are Watts.

The Preferences dialog box appears with the Units tab displayed. 13. Click on the arrow next to the Model Length field and select in from the drop-down list. 14. Select OK to close the dialog box.

21-5


or Thermal > Surfaces/Solids > Rectangle. The Command line should now read: Origin point <0.0.0>:

16. Type 0,0 in the Command line. The Command line should now read: Point for +X axis and X-size <@1,0,0>:

17. Type 10,0 in the Command line. The Command line should now read: Point to set XY plane and Ysize <@0,1,0>:

18. Type 0,5 in the Command line. The Thin Shell Data dialog box appears.

21-6

Create a 10 x 5 square in the x-y plane at Z=0 for the bottom surface.

Overview Parallel Flat Plates (Continued) 19. Select the Radiation tab. 20. Select BASE both in the Analysis Group Name, Active Side field if not already highlighted. 21. Select Edit. The Edit Active Side dialog box appears.

22. Select Top/Out to place a dot in the circle.

Change the active side of the surface in the surface group BASE to be active on the top (+Z) side (). The active side for this surface in the surface group BASE is updated in the Thin Shell Data dialog box to show that it is now active on the top side. When these steps are completed, the screen should look similar to the example below.

Figure 21-1 Top Side Active

23. Select OK to close the dialog box. The Thin Shell Data dialog box returns with BASE top/out displayed. 24. Click on the arrow next to the Top/ Out Side Optical Property field and select White Paint from the dropdown list. 25. Click on the arrow next to the Bottom/ In Side Optical Property field and select White Paint from the dropdown list. Note: Although the bottom side is not active, the bottom side optical property is set here to facilitate copying the surface in a later step. 26. Select OK to close the Thin Shell Data dialog box. 27.


21-7




30. Click on TD/RC Nodes to deselect it (remove the check mark from the box). 31. Select OK to close the Preferences dialog box.

21-8

Notice the small ring in the center of the plate. That is the node. Turn off the display of all nodes.


or Modify > Copy. The Command line should now read: Select objects:

33. Click on any part of the rectangle.

These steps copy the bottom plate to make the upper plate, which is located 12 inches above the bottom. See the AutoCAD® help for more options for the copy command. The copy will be moved 12 inches in the Z direction.

The rectangle is selected and the Com- A second plate is created. The thermal model information entered for the first mand line should now read: plate is also copied to the second. The secSelect objects: ond plate is therefore also a Thermal Desktop surface. Press <Enter>. The Command line should now read: Specify base point or [Displacement/mOde] :

34. Type 0,0,12 in the Command line. The Command line should now read: Specify second point or <use first point as displacement>:

35. Press <Enter> to use the entered coordinates as a displacement. 36.

Figure 21-2 Second Plate Created


21-9

Overview Parallel Flat Plates (Continued) or Thermal > Edit.

37.

The Command line should now read: Select Objects or [Indiv]:

38. Click on the newly created surface (top plate). The Command line should now read: Select Objects or [Indiv]:

39. Press <Enter>. The Thin Shell Data dialog box appears with the Radiation tab displayed and the thermal model information that is assigned to the upper surface.

40. Double-click BASE top/out in the Analysis Group Name, Active Side filed. The Edit Active Side dialog box appears. 41. Select Bottom/In to place a dot in the circle. 42. Select OK to close the dialog box. The Thin Shell Data dialog box returns with BASE bottom/in displayed.

21-10

These steps change the active side of the surface in the surface group BASE to be active on the bottom (-Z) side. The active side for this surface in the surface group BASE is updated in the Thin Shell Data dialog box to show that it is now active on the bottom side.

Overview Parallel Flat Plates (Continued) 43. Click on the Numbering tab.

Change the node IDs for the upper surface.

44. Highlight the current value in the Use Start ID field and type 2. 45. Select OK to close the dialog box.

21-11

Overview Parallel Flat Plates (Continued)

47. Select Arrows in the Display field to place a dot in the button.

Set the display preferences for active side verification. Colors indicating active sides are always available with the shade command. If only colors are being displayed, the shade command will automatically be executed. The Display Active Sides command must be executed each time to update the display.

48. Select Display to close the Display Preferences dialog box.

that correct active sides have been input.

46.

or Thermal > Model Checks > Active Display Preferences. The Display Preferences dialog box appears.

Arrows appear in the drawing area to designate the active plate sides.

21-12

Overview Parallel Flat Plates (Continued) The drawing should look similar to Figure 21-3. If it does not, retrace the above steps to determine what went wrong. Once the drawing is at this stage, it is ready to calculate radks.

Figure 21-3

Active Sides for Parallel Plate

21-13

Overview Parallel Flat Plates (Continued) 49. Select Thermal > Radiation Calculations > Set Radiation Analysis Data.... The Radiation Analysis Data dialog box appears.

50. Select the Control tab if not already displayed. 51. Highlight the current value in the Rays per node field and type 100000 (one hundred thousand, no comma). 52. Select the Advance Control tab.

21-14

Sets the control parameters for calculations. Shoot lots of rays just for fun.

Overview Parallel Flat Plates (Continued) 53. Click on Use oct-tree to accelerate calculations to deselect it (remove the check mark from the box).

Oct-tree acceleration is not necessary for this small of a problem.

54. Select the Radk Output tab.

The default for List if % kept is off by more than: is set to 10%. This is so that only the surfaces with errors are printed and it makes it easy to find the errors in large models. This output will be viewed later.

55. Highlight the current value in the List summary if % kept is off by more than: field and type 0. 56. Select OK to close the Radiation Analysis Data dialog box. 57. Select Thermal > Radiation Calculations > Calc Radks Ray Trace. A Thermal Desktop/AutoCAD dialog box asking for confirmation to continue appears.

This step calculates radks for the Analysis Group Base using the Monte Carlo raytracing method. The radks will be output to the file “SINDA.K” in the working directory. Output options can be controlled with Thermal > Radiation Calculations > Set Radiation Analysis Data, Radk Output page.

58. Select OK to close the dialog box. 59. Press the key to bring up the text window and review. 60. Press the key to hide the text window.

See the Thermal Desktop s Manual Section 10.1.1.1 for a discussion of error vs. rays. Note that the text window will refer to 3 nodes yet the model has only 2 nodes. Radiation calculations automatically include a boundary node, SPACE.1.

21-15

Overview Parallel Flat Plates (Continued) 61. Select Thermal > Radiation Calculations > Calc Radks Ray Trace. A Thermal Desktop/AutoCAD dialog box asking for confirmation to continue appears. 62. Select OK. The Append/Replace Database dialog box appears.

63. Leave Append results to existing database selected. 64. Select OK.

21-16

Run the problem again. A dialog box ing the analysis group and property file appears. Since radks have already been calculated, a dialog box will appear giving the option of adding data to this database, or continuing with a brand new one. Rays shot will be cumulative for all runs if Append is selected.

Overview Parallel Flat Plates (Continued) Radk data is output in the file “SINDA.K” located in the current working directory, in this case the \Tutorials\Thermal Desktop\parallel directory. Use an editor such as the Windows Notepad to look at this file or look at the output shown below. Table 21-1 SINDA.K File for Parallel Plates HEADER CONDUCTOR DATA, MAIN C SINDA/FLUINT data created with Thermal Desktop 3.2 Beta Build 30 C Generated on Fri Jul 21 10:58:07 2000 C Generated from database BASE-RcOptics.rck C Cutoff factor 0.0010000 C Conductor units are: in^2 C (more information at end of file) C C radk format: C cond_id node_1 node_2 Area*e*Bij $ Bij Bji C -1, MAIN.1, SPACE.1, 37.163 $ 0.92909 -2, MAIN.2, SPACE.1, 37.146 $ 0.92866 -3, MAIN.1, MAIN.2, 2.7972 $ 0.069929, 0.069929 C C Summary data for nodes with Bij sums < 1.0000 or > 1.0000 C Summary data for position 1 C node area rays emiss Bij Bij Bij Weighted C sum self inact % Error C MAIN.1 50.000 200000 0.80000 1.0002 0.001 0.2 C MAIN.2 50.000 200000 0.80000 0.99980 0.001 0.2

The file lists the analysis group and optical property file used to create the data, followed by radks to space, then node-to-node radks. At the end of the file will be radks to inactive nodes (if any), radks to self, and statistics for the calculation process. The statistical summary data lists the node, area, number of rays shot, the sum of all the radks for this node, the effective emissivity, and the weighted error. The rays shot is 200,000 since the model was run twice. The effective emissivity should be equal to the emissivity input on the optical property form. If the emissivity were input as angular dependent, this quantity will be the integrated hemispherical emissivity. The output is shown above in Table 21-1: SINDA.K File for Parallel Plates. The BijSum is a useful quantity for error checking. It should be within a few hundredths of 1.0. Excessively low percentages indicate views to inactive sides or overly aggressive filtering of the smaller radk values. Rays shot from both node i and node j are used to compute the radk between i and j. Notice the BijSelf value is non-zero. The Bij is the fraction of energy leaving Surface i that is absorbed by Surface j by all possible paths. Since the emissivities are less than 1, some energy emitted by Surface 1 is reflected from Surface 2 back to Surface 1.

21-17

Overview Parallel Flat Plates (Continued) 65. Select Thermal > Radiation Calculations > Set Radiation Analysis Data.... The Radiation Analysis Data dialog box appears with the Radk Output tab displayed. 66. Select the Control tab. 67. Highlight the current value in the Rays per node field and type 20. 68. Select List in the Nodes field (place a dot in the circle). 69. Type MAIN.1 in the List input field.

21-18

To get an idea of how calculations are made, plot the calculated rays on the model. Set the number of rays to be a smaller value, since many rays will just clutter the screen. An option is to set the Maximum number of rays to plot on the Ray plot tab, which will be opened next. Also, only shoot rays from Node MAIN.1 so the results can be seen.

Overview Parallel Flat Plates (Continued) 70. Select the Ray Plot tab. 71. Click on Plot rays to space to select it (place a check mark in the box).

Choose to plot the rays to space and the rays for surface-to-surface reflection. Limit the length of the rays to 12 inches.

72. Click on Plot rays for surface-tosurface reflection to select it (place a check mark in the box). 73. Highlight the current value in the Length of “to space” and “from source” rays field and type 12.


21-19

Overview Parallel Flat Plates (Continued) 75. Type LTSCALE in the Command line. The rays to space (as well as nodal centerlines) are drawn according to the LTSThe command line should now read: CALE value. These lines are drawn by a Enter new linetype scale series of dots. The LTSCALE value deterfactor <39.3701>: mines how close to put the dots together. 76. Type 0.5. A smaller value means more dots are printed for each line. Lines appear inside the plates.

Figure 21-4 After LTSCALE Assigned

77. Select Thermal > Radiation Calculations > Calc Radks Ray Trace. A Thermal Desktop/AutoCAD dialog box asking for confirmation to continue appears. 78. Select OK to close the dialog box. The Append/Replace Database dialog box appears. 79. Leave Append selected.

Several orange lines (and possibly some blue lines) will appear representing rays. These rays are random, so they will be different every time the calculations are made. The color of the rays is a function of the energy of the ray. Energy values of 1 are red and scale down to dark blue for zero values. The color will change as energy is absorbed and the ray is reflected.

80. Select OK to close the dialog box without making any changes.

Figure 21-5 Ray Calculation Example

21-20


or Thermal > Radiation Calculations > Clear Ray Plot.

If additional rays are shot, they will be added to the rays already on the screen. Use the Clear Ray Plot command to delete the rays.

Figure 21-6 Clear Ray Display

82. Select File > Exit. A Thermal Desktop/AutoCAD dialog box appears asking to save the drawing changes.


83. Select Yes. 84. Examine the working folder Tutorials\Thermal Desktop\parallel and open the SINDA.xls file. This files provides the same information as found at the bottom of the *.k file, but in a convenient format for searching, sorting, etc. 85. Close SINDA.xls. Some additional things the might try: • Move the rectangles closer together so that multiple reflections may easily be seen. • Make the optical property of the upper surface to be transmissive and then plot the rays. Examine the difference between specular and diffuse transmissivity.

21-21

21-22

21.2

Space Station Oct Tree Example

What will be learned: • How to use oct-trees to accelerate execution time • How to postprocess radk results In this example, the use of Oct-trees in accelerating RadCAD radiation calculations will be explored. The space station model will show how changing a single oct-tree parameter can significantly decrease the amount of time required to perform radiation analyses. It is recommended that all s work through this example, even if not employed in the aerospace industry. Decreasing the computational time can be applied to any type of radiation problem. Space Station Oct Tree Example 1. Double click on the file spaceStation.dwg located in the Tutorials\Thermal Desktop\OctCells folder. Thermal Desktop opens with the spaceStation drawing on the screen.

Figure 21-7

Space Station Oct Tree Initial View

21-23

Space Station Oct Tree Example 2.

or Thermal > Model Checks > Display Active Sides.

active sides. Everything should be green and yellow. Note that some machines without decent graphics cards may show the solar s as blue. This is an artifact of color bleeding as the solar has been created in this model as two surfaces, that are separated by a small amount, and are active in opposite directions.

Figure 21-8 Display Active Sides

21-24

•

Green indicates that one side is active and the opposite side is inactive.

•

Light blue indicates that the side being looked at is inactive, and the opposite side is active.

•

Yellow indicates that both sides are active, dark blue indicates that both sides are inactive.

•

Red means that the surface is not in the analysis group currently being used and will not be used for calculations.

Space Station Oct Tree Example 3. Select Thermal > Radiation Calculations > Set Radiation Analysis Data.

Set the number of rays to 500 and note the number of Oct-tree subdivisions.

The Radiation Analysis Data dialog box appears.

4. Select the Control tab if not already displayed. 5. Highlight the value in the Rays per node field and type 500 if the current value is different.

21-25

Space Station Oct Tree Example 6. Select the Advance Control tab.

Notice the Max oct-tree subdivisions: field is set to 6. 7. Select the Radk Output tab.

8. Click on Generate SINDA/FLUINT input after calculations to deselect it (remove the check mark from the box). 9. Select OK to close the Radiation Analysis Data dialog box. 21-26

Remove the option to generate the SINDA/FLUINT input.

Space Station Oct Tree Example 10. Select Thermal > Radiation Calculations > Calc Radks Ray Trace. A Thermal Desktop/AutoCAD dialog box asking for confirmation to continue appears.

11. Select OK to close the dialog box. 12. Press to find the time to calculate the radks. 13. Press to close the text window.

Calculate radiation conductors. The default analysis group and the currently loaded optical properties will be used to calculate radks. Record the amount of time required to perform the radk calculations. This is most easily done by hitting the function key. The text window will appear. The amount of time to calculate the radks is needed. That value can vary based on U speed, the number of U’s available and the number of other applications running. Example times are: •

46 seconds on a 3.0 GHz, dual- dualcore (4-U) machine running AutoCAD 2008.

•

20 seconds on a 1.73 GHz, dual-quadcore (8-U), 64-bit machine running AutoCAD 2014 with 8 GB RAM.

14. Select Thermal > Radiation Calculations > Set Radiation Analysis Data. The Radiation Analysis Data dialog box appears with the Radk Output tab displayed. 15. Select the Advanced Control tab. 16. Highlight the current value in the Max oct-tree subdivisions field and type 7. 17. Select OK to close the dialog box.

21-27

Space Station Oct Tree Example 18. Select Thermal > Radiation Calculations > Calc Radks Ray Trace. A Thermal Desktop/AutoCAD dialog box asking for confirmation to continue appears. 19. Select OK to close the dialog box. The Append/Replace Database dialog box appears.

Since the database from the previous run already exists, the program asks if the wants to “append” or “replace” the existing database. Appending will add 500 more rays to the existing database, making it 1000 total rays. Replacing will replace the database. For this example, either selection is sufficient. Record the amount of time required to perform the radk calculations. This run should be about 30% faster.

20. Select the desired option (place a dot in the circle).

Note: The times are clock times, so the time required to respond to the Append/Replace Database form will affect the results. If your new time is slower, try again, but respond more quickly to the form that appears.

21. Select OK. 22. Press to find the time to calculate the radks. 23. Close the text window when finished reviewing. Each run shot 500 radk rays, and only one parameter was different, the Max oct-tree subdivisions. Please keep in mind that the oct-tree does not affect the answers, but only the speed at which they are arrived. The oct-tree breaks the model into smaller regions, and limits the amount of intersection tests performed. Every model has an optimal number of Max oct-tree subdivisions and Max surfaces per cell that will calculate the radiation job the fastest. CRTech has found that the subdivisions parameter affects the results much more drastically than the surfaces per cell. Some models will run 10 times faster by changing the subdivision setting. In other models, the subdivision setting does not affect the U time. Some models may run fastest with subdivisions equal to five, while others may require subdivisions equal to 9. In conclusion, the should run test cases to find the optimal parameters. An easier method to determine the optimum Oct-Cell settings is to use the Optimize Cells command as described below.

21-28

Space Station Oct Tree Example 24. Select Thermal > Radiation Calculations > Optimize Cells. The Optimize Cells form opens.

The Optimize Cells command runs a small radiation calculation using a variety of settings for the Oct-Cells. Each run uses the same random seed number to minimize the differences between the runs. The stronger variation in run time is usually caused by the number of subdivisions. For time reasons, we will only change subdivisions for this tutorial.

25. Under Vary the Subdivisions, enter 6 for From and 9 for To. 26. Under Vary the Surfaces Per Cell, enter 8 for From and 8 for To and 1 for Increment. 27. Enter 500 for Number of Rays. 28. Select OK and confirm any dialogs that appear. After the calculations are completed, a document called OptimizeCells.txt is opened.

It is important that the amount of time required for each test is substantial enough to see true run time changes, and not just U or operating system effects. Ideally, each setting should run for at least 30 seconds of U time. This can be controlled by the number of rays shot per run.

Examining the document, you will see at 3 main sections. The first section is the OctCell generation time. The second section is the maximum number of surfaces in any one cell. The third section Provides the ray tracing time. With a large change in the values in the second section, as seen from 6 to 7 subdivisions, a large benefit can be realized. With a smaller change in Surfaces per Cell, as seen from 7 to 8 subdivisions or 8 to 9 subdivisions, a smaller benefit will be realized.

21-29

Space Station Oct Tree Example

21-30

Space Station Oct Tree Example 29. Select OK to close the dialog box. A Thermal Desktop/AutoCAD dialog box asking for confirmation to continue appears. 30. Select OK. 31. Press to view test progress and results. 32. Close the text window when finished reviewing.

21-31

Space Station Oct Tree Example 33. Select Thermal > PostProcessing > Manage Datasets. The Postprocessing Datasets dialog box appears.

34. Select Add New. The Data Set Source Selection dialog box appears.

35. Type radks in the Postprocessing set name field. 36. Select the Radks radio button (place a dot in the circle). 37. Select OK to close the dialog box.

21-32

Create a new radks postprocessing dataset named radks.

Space Station Oct Tree Example The Directory Select dialog box appears.

View the calculated data using a color map. Enter a descriptive comment for the postprocessing dataset if preferred. Click directly in the edit field to enter the comment. A lower value for the radk to space for the interior or shuttle payload bay should be seen.

38. Select OK. The Set FF/Radk Dataset Properties dialog box appears. Figure 21-9 View Data Using Color Map

39. Select OK. The Postprocessing Datasets dialog box appears with radks displayed in the Current Data Set field. 40. Select Close.

21-33

Space Station Oct Tree Example Return to the normal display mode. 41.

or Thermal > Postprocessing > PostProcessing off.

Note: The drawing may be left in postprocessing mode when exiting if desired. It will be reloaded in postprocessing mode when the session is resumed.

Figure 21-10 View Normal Display Mode



43. Select Yes. Additional practice: Use the Model Checks > Check Overlapping Surfaces command to find the surfaces that might be overlapping in the same plane. Surfaces that overlap in the same plane will most likely cause problems with radiation calculations. Once the overlapping surfaces are found, use the Model Browser to isolate the overlapping nodes and try to determine what is wrong with the geometric model. In the Model Browser, select List > Groups to see groups of overlapping surfaces.

21-34

21.3

Importing a TRASYS Model and Using Articulators

What will be learned: • Importing a TRASYS input file • Working with layers • Creating trackers • Creating environmental heating rate cases In this example, a TRASYS model will be imported. Following that we will articulate the solar arrays. Thermal Desktop will automatically run the TRASYS input file through the Thermal Desktop supplied TRASYS preprocessor. If the TRASYS model has errors, a window will be displayed describing the errors and the import will be aborted. When the preprocessor is finished, an input dialog box will appear prompting for a desired display resolution. This dialog box controls the degree of fidelity to which non-Thermal Desktop curved surfaces are modeled (such as an ogive). Higher resolutions use more facets per degree of curvature. TRASYS surfaces will be seen being drawn on the viewport as they are read from the preprocessor output. After the TRASYS model is read in, the preprocessor intermediate files are automatically deleted. The status of the preprocessor run is retained in the file “TRASYS.OUT”. All imported TRASYS nodes are placed into the current analysis group. Use this analysis group to perform radk or view factor computations. The imported model has all the data necessary to begin calculations. Active side and submodel/node number data may be verified using the Thermal > Model Checks operations. Each BCS is placed on its own layer.

Importing TRASYS Files and Using Articulation 1. Double click on the file trasys.dwg located in the Tutorials\Thermal Desktop\trasys folder. Thermal Desktop opens with the trasys drawing on the screen. 2.


21-35

Importing TRASYS Files and Using Articulation 3. Select Thermal > Import > TRASYS.

Import the example TRASYS model, TMG3.INP.

The TRASYS Import Options dialog box appears.

Figure 21-11 Imported TRASYS Model

4. Click on the arrow to the right of the Input File field and select TMG3.INP from the drop-down list, if it is not already visible. 5. Select OK to close the dialog box. Invoke the Layer Property Manager. or type Layer.

6.

The Layer Properties Manager dialog box appears. Note: Format > Layer will also display the Layer Properties Manager dialog box.

21-36

Importing TRASYS Files and Using Articulation The Layer Properties Manager dialog box is shown below in Figure 21-12. All newly created entities are placed on the current active layer. Click the lightbulb icon in the On column to toggle layer visibility on and off. A layer that is turned Off is not visible on the screen. Click on the sun/snowflake icon in the Freeze column to freeze or thaw a layer, respectively. When working with large models, freezing is a better option than turning off since frozen layers are not redrawn during graphical updates. The display list is not regenerated for frozen layers. The lock icon can be used to prevent modification to any object that is on the locked layer, however, CRTech recommends leaving all -defined layers unlocked. Create New Layer Current Layer

Freeze/Thaw On/Off Color Names

Figure 21-12

Layer Properties Manager Dialog Box

21-37

Importing TRASYS Files and Using Articulation 7. Click on the Freeze (sun) icons for all of the BCS_ files except for BCS_SAPX to turn them off. The sun icons will change to look like snowflakes.

The goal of this command is to turn off the display of the model except for the solar on layer BCS_SAPX. When completed, the view on the screen should be similar to Figure 21-13.

Note: There are 19 rows of BCS_ files and 18 will be affected, leaving BCS_SAPX untouched. 8. Make sure the current layer, 0, is not turned off (remains untouched) or any new items created will not be visible. 9. Select OK to close the dialog box. or View > Zoom > Extents

10. or

• type zoom (or just the letter z) in the Command line and press <Enter>. • type extents (or just the letter e) in the Command line and press <Enter>.

21-38

Figure 21-13

Solar Visible

Note: The and <Shift> shortcut keyboard command works in this dialog box to select and change more than one layer. Note: Turning the layers off instead of freezing them are visually the same, however, the Zoom Extents command will work differently. Off layers, while not visible are still included in any “all” command (e.g. Zoom Extents). Frozen layers are not included in “all” commands.

Importing TRASYS Files and Using Articulation 11. Type pan in the Command line. The cursor changes into a small hand—this is the Pan Realtime command.

Use the pan and zoom command to position the model to look like Figure 21-14. Use the right mouse button to switch between pan and zoom.

12. Hold down the left mouse button to move the model across the drawing area. 13. Click the right mouse button to display the popup menu. 14. Select Zoom. The cursor changes to a small magnifying glass with a plus and a minus sign—this is the Zoom Realtime command. 15. Alternate between pan and zoom until the model is positioned similarly to Figure 21-14.

Figure 21-14

Geometry positioning

16. Press <Esc> to end the pan/zoom command mode.

21-39

Importing TRASYS Files and Using Articulation 17.

or Thermal > Articulators > Create Tracker.

The tracker must be placed at the point that the solar array will rotate.

The Command line should now read: Enter origin of tracker:

18. Hold down the <Shift> key and click the right mouse button to display the right mouse popup menu. 19. Select Center. The Command line should now read: Enter origin of tracker: _cen of

20. Click in/on the left-most positioned circle as shown by the arrow in Figure Figure 21-15 Geometry positioning 21-15. Notice as the cursor moves over the circles, a yellow circle Note: See “Trackers” on page 4appears. 102 for more information on trackThe Single Axis Tracker dialog box ers. appears.


21-40



Figure 21-16

23.

or Thermal > Articulators > Attach Geometry. The Command line should now read: Select an articulator:

Tracker Created

The surfaces that will rotate with the tracker must be attached to that tracker. The attach command will not attach a tracker to itself, so it is OK to select it when attaching the surfaces.

24. Select the articulator you just created The Command line should now read: Select objects to attach to articulator...: 25. Draw a box around all of the objects in the display. The Command line should now read: Select objects to attach to articulator...: 26. Press <Enter>. Note: If is selected the text window will show that 13 objects were attached to the articulator.

21-41

Importing TRASYS Files and Using Articulation Create an orbit with a beta angle of zero. 27.

or Thermal > Orbit > Manage Orbits. The Heating Rate Case Manager dialog box appears.

28. Select Add. The Create New External Heating Environ dialog box appears.

29. Type Test in the New Heating Case Name field. 30. Select OK to close the Create New External Heating Environ dialog box.

21-42

New orbits are created using the Orbit Manager. Multiple orbit definitions can be created and saved under a defined name. Orbit definitions are stored in the drawing file along with the model geometry.

Importing TRASYS Files and Using Articulation The Orbit: Test dialog box appears.

Rotate the entire model by 90 degrees about the Z axis. This will put the solar s in a position so that they can track the sun.

31. Select the Orientation tab.

Want to Learn More? More information about creating orbits may be found in Chapter 6: External Heating Environments and Orbits.

32. Highlight the current value in the Z Additional Rotations field and type 90. 33. Select OK. The Heating Rate Case Manager dialog box reappears with Test displayed in the Current Heating Rate Case field. 34. Select Display Orbit. The drawing area displays the orbit. 35. Select View > 3D Views > SE Isometric.

21-43


or Thermal > Orbit > View Vehicle > Set Orbit Position/ Prefs.

The vehicle can now be viewed in orbit. The size parameter allows the to manipulate the size of the vehicle with respect to the size of the planet.

The View Vehicle In Orbit dialog box appears.

37. Select OK to close the dialog box without making any changes. View the orbit and model from the sun. 38.

or Thermal > Orbit > View From > Sun.

Figure 21-17

21-44

View from the Sun


or Thermal > Orbit > Display Preferences.

Turn the planet off to see if the articulation is working properly in the shade.

The Orbit Display Preferences dialog box appears. 40. Click on Planet to deselect it (remove the check mark from the box). 41. Select OK.

Figure 21-18

View of Model Only

21-45


or Thermal > Orbit > View Vehicle > Next Position.

43. Press <Enter>.

The geometry will move to the next position in the orbit. Notice that the bottom solar array stays perpendicular to the sun, while the top array does not (because it is not attached to the articulator).

Figure 21-19

View Orbit Next Position

Use the right mouse button or <Enter> to step all the way around the orbit.

Figure 21-20

View Orbit Next Position

This example lines up the solar arrays for maximum solar flux only if the beta angle is set to zero. Now go back and add a second axis of rotation so that the solar arrays will line up independent of beta angle. The orientation of the current tracker will be changed so that it will for the beta angle. A second tracker will then be added that will for the movement around the planet. The original tracker will be attached to the second tracker. The order of attachment is extremely important. Thermal Desktop determines the nesting of the trackers and performs the rotation of the highest level tracker first.

21-46

Importing TRASYS Files and Using Articulation Turn off the orbit display. 44.

or Thermal > Orbit > Orbit Display Off. Work in wireframe mode.

45.

or View > Visual Styles > 2D Wireframe

Figure 21-21

46.

or Thermal > Articulators > Reset Trackers.

The articulators must be reset to the starting value. The model should look similar to the view below.

Figure 21-22

47. Select File > Save.

Wireframe View

Tracker Reset

Save the geometry often.

21-47

Importing TRASYS Files and Using Articulation 48. Type zoom. The command line area should now show:

Use Pan and Zoom to position the model back to the view shown in Figure 21-14.

Specify corner of window, enter a scale factor (nX or nXP), or [All/Center/ Dynamic/Extents/Previous/ Scale/Window/Object] :

49. Type all. Note: Use the pan and zoom icons as desired. 50. Select Thermal > Articulators > Toggle Global Activation. Articulators are now globally turned off—see confirmation of this in the Command line area.

21-48

Change the orientation of the current tracker. To do this, first turn off all trackers so that when the tracker is manipulated, the geometry will not move with it.

Importing TRASYS Files and Using Articulation 51. Select the articulator. The articulator is highlighted—the lines become dashed lines and blue grip boxes appear along the X, Y and Z axes.

Rotate the articulator 90 degrees about the Z-axis using the grip points. This redefines the articulator to for various beta angles. 2nd pt.

52. Click on the grip point at the end of the X axis (Red). Specify stretch point or [Base point/ Copy/Undo/eXit]: appears in the Com-

1st Point

mand line area. Note: A yellow box appears in the middle of the cursor when placed over the grip box. The yellow box becomes thicker when the cursor is positioned over the grip point and the grip box becomes red once it is selected. A “rubberband line” attaches the cursor to the first selected grip point and moves as the cursor moves.

Figure 21-23

Articulator Grip Editing

53. Move the cursor to the grip point on the end of the Y axis (Green) and click the left mouse button to select it. 54. Select Thermal > Articulators > Toggle Global Activation. The articulators are now globally turned back on

Manual manipulation of the articulators is now complete and they must be turned back on.

21-49


or Thermal > Articulators > Create Tracker. The command line should now read: Enter origin of tracker:

56. Hold down the <Shift> key and click the right mouse button to display the right mouse popup menu.

Create the articulator to for movement around the planet. Make the new articulator 1.5x the size of the previous articulator. When these steps are complete the model should look similar to the view below.

57. Select Center. The command line should now read: Enter origin of tracker: _cen of

58. Click in/on the left-most positioned circle as directed by the arrow shown earlier in Figure 21-15. Notice as the cursor moves over the circles, a yellow circle appears.

Figure 21-24

Second Tracker Created

The Single Axis Tracker dialog box appears. 59. Highlight the current value in the Display Size field and type 1.5. 60. Select OK to close the dialog box. or View > Zoom > Extents

61. or

• type zoom in the Command line and press <Enter>. • type extents in the Command line and press <Enter>.

Figure 21-25

21-50

Model After Second Tracker


or Thermal > Articulators > Attach Geometry.

Attach the smaller articulator to the larger one.

The command line should now read: Select an articulator:

63. Click on the new larger articulator to select it. The command line should now read: Select objects to attach to articulator...:

64. Click on the smaller articulator to select it. The command line should now read: Select objects to attach to articulator...:

65. Press <Enter>. 66.

or Thermal > Orbit > Edit Current Orbit. The Orbit: Test dialog box appears with 90 displayed in the Additional Rotations Z field.

67. Select the Basic Orbit tab.

Figure 21-26

New Orbit Angle

68. Highlight the current value in the Beta Angle field and type 30. 69. Select OK.

21-51

Importing TRASYS Files and Using Articulation 70. Display the model on the orbit and step through each position. 71. For practice, add 2 more articulators and get the second solar to track the sun. 72. Select File > Exit. A Thermal Desktop/AutoCAD dialog box appears asking to save the drawing changes. 73. Select Yes.

21-52


21.4

Orbital Heating Rates

What will be learned: • Calculating orbital heating rates. • Viewing a model in orbit. • Postprocessing heating rates. • Adjusting the color bar while in paper space. • Using the Case Set Manager to set up multiple heating rate jobs. In this exercise, orbital heating rates using Monte Carlo ray tracing will be computed. Orbital Heating Rates 1. Double click on the file satellite.dwg located in the Tutorials\Thermal Desktop\HeatingRate folder. Thermal Desktop opens with the satellite drawing on the screen.

21-53

Orbital Heating Rates (Continued) 2.



4. Type beta90 in the New Heating Case Name field. 5. Select OK to close the dialog box.

21-54

Create a new external heating environment named beta90.

Orbital Heating Rates (Continued) The Orbit: beta90 dialog box appears.

The beta angle is the angle between the vector to the sun and the orbital plane. “List” fields, like the one on the Positions tab, are edited directly. Click the cursor in the list field and use the or keys to remove text. Text in list fields may also be selected (highlighted), cut, , and pasted, .

6. Highlight the current value in the Beta Angle field and type 90. 7. Select the Orientation tab. 8. Highlight the current value in the X Additional Rotations field and type 180. 9. Select the Positions tab. 10. Check the Use Positions radio button to select it (place a dot in the circle). The list field below the button activates. 11. Delete all of the entries in the list box below the radio button. • Highlight all of the entries. • Press . 12. Type 90 on a single line. 13. Select OK to close the dialog box. The Heating Rate Case Manager dialog box reappears. 14. Select Display Orbit.

21-55


or Thermal > Orbit > Orbit Display Preferences.

The Orbit Display Preferences dialog box appears.

The single position for which calculations will be performed is shown crossing over the north pole of the planet. The coordinate system for the vehicle is shown with red used for the X axis, green used for the Y axis, and blue used for the Z axis (xyz=>rgb). The small green triangle near the origin shows the start position.

16. Select the Size/Colors tab. 17. Highlight the value in the Solar Shadow Length field and type 1 if the current value is different. 18. Highlight the value in the Orbit Position Scale field and type 3 if the current value is different. 19. Select OK. 20.

21-56

or Zoom > Extents.


or Thermal > Orbit > Orbit Display Off.

the orientation of the model by viewing it as it appears from the sun. The solar arrays should be hiding the battery. Be sure to turn the orbit display off before viewing the model, otherwise the orbit display will be repositioned, not the model.

22.

or Thermal > Model Checks > View Model From Sun/Planet > Set Orbit Position/Location. The New Vehicle Setup dialog box appears.

23. Select OK.

21-57

Orbital Heating Rates (Continued) 24. Select Thermal > Radiation Calculations > Calc Heating Rates Ray Trace. A Thermal Desktop/AutoCAD dialog box appears.

25. Select OK.

21-58

Compute orbital heating rates for solar, albedo, and planetshine using full monte carlo. A verification screen will appear showing the name of the analysis group, orbit, and optical property file to use for this calculation.

Orbital Heating Rates (Continued) 26. Select Thermal > Post Processing > Manage Datasets. The Processing Datasets dialog box appears.

Create a postprocessing set for the orbital heating rates. A confirmation screen appears as notification that the default analysis group, the current orbit, and the currently loaded optical properties will be used to create the postprocessing set. After the set is created, the orbital heating rate data may be viewed for this case even if other analysis groups, orbits, or property files are currently being used.

27. Select Add New. The Data Set Source Selection dialog box appears.

28. Type hr mc in the Postprocessing set name field. 29. Click on the Heating Rates radio button to select it (place a dot in the circle). 30. Select OK to close the dialog box.

21-59

Orbital Heating Rates (Continued) The Directory Select dialog box appears with the dataset already selected.

31. Select OK. The Set HR Dataset Properties dialog box appears.

32. Select OK.

21-60

Orbital Heating Rates (Continued) The Postprocessing Datasets dialog box reappears. 33. Select Close. 34.

or Thermal > Post Processing > Edit Layout ColorBar/Viewports.

The color bar was set to use fixed limits from the last exercise. Change to use autoscaling.

The Color Bar Settings dialog box appears. 35. Select the Auto Scaling drop-down and select On - Program Calculates Visible Min/Max

36. Select Done.

21-61

Orbital Heating Rates (Continued) 37. Select View > 3D Views > Back. 38. Select View > 3D Views > Right. 39.

or Thermal > Model Checks > View Model From Sun/Planet > Set Orbit Position/Location. The View Vehicle Setup dialog box appears.

The view is currently looking at total absorbed flux using the sum of all heating rate sources (solar, albedo, and planetshine). Look for some reflections onto the backside of the right-hand . The model checking feature may also be used to orient the model as seen from the sun or planet.

40. Select OK.

Named views may also be used, and layers can be toggled on and off to aid in examining data.

41. On the bottom status bar of the screen, click on the MODEL button. It will change to read PAPER.

Switch to paper space and adjust the position of the viewport.

42. Type zoom in the Command line. 43. Type .8x in the Command line. 44. Pick anywhere on the black box outlining the viewport to select the box. 45. Click on the lower left grip of the viewport and drag the corner towards the center of the screen so that the view port does not overlap the colorbar or the color bar label. 46. Click the left mouse button to accept the position.

47.

21-62

or Zoom > Previous.

Note: If the black box outlining the viewport is not visible, issue the RcTouchALL command to force a regeneration of the data.

Orbital Heating Rates (Continued) 48. Click on the color bar associated with the model in the drawing area to highlight it. 49. Position the cursor on the lower left grip, hold down the left mouse button and drag the color bar a little bit to the left and down.

Experiment with moving the color bar around and changing its size. After switching back to model space, zoom the viewport to reposition the model in the new viewport location.

50. Click the left mouse button when satisfied with the position of the color bar. 51. Position the cursor on the upper right grip, hold down the left mouse button and drag the cursor to the right to increase the width and text size. 52. Continue dragging the cursor up to increase the size. 53. Drag the cursor all the way over to the lower right hand corner until the width becomes greater than the height. Notice that the colorbar automatically switches to a horizontal format as the width grows greater than the height. 54. Click the left mouse button when satisfied with the size. 55. On the bottom status bar of the screen, click on the PAPER button. It will change back to read MODEL. 56.

or Zoom > Extents.

21-63

Orbital Heating Rates (Continued) 57. Select View > 3D Views > SE Isometric. 58.

or Thermal > Post Processing> Edit Current Dataset. The Set HR Dataset Properties dialog box appears.

59. Click in the check box next to Solar to deselect it (remove check mark from the box). 60. Click in the check box next to Albedo to deselect it (remove check mark from the box). 61. Make sure Planetshine is selected (check mark in the box). 62. Click on the Type Total Absorbed radio button to select it (place check mark in the box) if not already selected. 63. Select OK.

21-64

Look at some of the other heating rate data. Pressing the <Enter> key will recall the last command, which will bring up the postprocessing set editing dialog box again.

Orbital Heating Rates (Continued) 64. Select Thermal > Radiation Calculations > Set Radiation Analysis Data.... The Radiation Analysis Data dialog box appears.

Shoot some more rays for just planetshine calculations. A dialog will appear to confirm the analysis group, orbit, and optical property file. Another dialog will appear allowing the existing database to be appended, or to start with a brand new database. Append the existing database to increase the accuracy.

65. Select the Control tab if not already displayed. 66. Highlight the current value in the Set Rays per node field and type 20000. 67. Deselect Solar and Albedo in the Heating Rate Sources field (remove check marks from the boxes). 68. Select OK. 69. Select Thermal > Radiation Calculations > Calc Heating Rates Ray Trace. A Thermal Desktop/AutoCAD dialog box appears asking for confirmation to continue. 70. Select OK. The Append/Replace Database dialog box appears with Append selected. 71. Select OK. The process runs.

21-65


or Thermal > Post Processing> Edit Current Dataset. The Set HR Dataset Properties dialog box appears.

73. Select OK.

Bring up the dataset editing dialog box and select OK to reload the data. The postprocessing set always “points” to the data, it does not contain the data values. Updating the display will show the most currently computed values. that the heating on the solar arrays appears more uniform, and that heating rates in general look more symmetrical.

74. Select File > Save. 75.


76. Select Add. The Create New External Heating Environ dialog box appears. 77. Type beta30 in the New Heating Case Name field. 78. Select OK to close the dialog box. The Orbit: beta30 dialog box appears. 79. Select the Basic Orbit tab. 80. Highlight the current value in the Beta Angle field and type 30. 81. Select OK to close the dialog box.

21-66

Use the Orbit Manager to create a new basic orbit with a beta angle of 30 degrees.

Orbital Heating Rates (Continued) The Heating Rate Case Manager dialog box reappears. 82. Select Display Orbit.

or Thermal > Case Set Man-

83. ager.

The Case Set Manager dialog box appears with Case Set 0 highlighted. 84. Select Edit.

Use the Case Set Manager to set up two radiation jobs. Note: When using the Case Set Manager to calculate radiation, the Radiation Analysis Data must be defined through the Case Set Manager, not the Thermal menu.

The Case Set Information Case Set 0 dialog box appears with the Radiation tab displayed.

21-67

Orbital Heating Rates (Continued) 85. Select Add. The Radiation Analysis Data dialog box is now shown. This is similar to the one displayed earlier with the Thermal > Radiation > Set Radiation Analysis data, but with a new tab at the beginning and tabs specific to the Job selected.

This dialog box allows the to choose the calculations that will be made and to set the number of rays and other control parameters for this job. Change some parameters if desired. Please note that the output file names are programmed to be unique so that one job does not overwrite another.

21-68

Orbital Heating Rates (Continued) 86. Click on the Heating Rates radio button to select it (place dot in the circle). The Orbit field activates.

Define the radiation job to be performed. Examine the Heatrate Output tab and note the file names.

87. Select External from the Analysis Group drop-down list. 88. Select beta30 from the Orbit dropdown list. 89. Click on the Monte Carlo radio button to select it (place dot in the circle). 90. Select OK to close the dialog box. The Case Set Information Case Set 0 dialog box reappears and the Externalbeta30 job is added to the Radiation Task list. 91. Select Add.

Add another Radiation task.

The Radiation Analysis Data dialog box is now shown. 92. Click on the Heating Rates radio button to select it (place dot in the circle). The Orbit field activates. 93. Select External from the Analysis Group drop-down list. 94. Select beta90 from the Orbit dropdown list. 95. Click on the Monte Carlo radio button to select it (place dot in the circle). 96. Select OK to close the dialog box. The Case Set Information Case Set 0 dialog box reappears and the Externalbeta90 job is added to the Radiation Task list.

Examine the Heatrate Output tab and note that the file names are different than the earlier Radiation Task.

21-69

Orbital Heating Rates (Continued) 97. Select the Calculations tab. 98. Deselect the following options (remove check marks from the boxes):

Since a conduction model was not built, all of the items to set up and run SINDA can be disabled.

• Generate Cond/Cap • Build SINDA input file • Run SINDA Model • Post Process SINDA Save File 99. Select OK to close the Case Set Information Case Set 0 dialog box. The Case Set Manager dialog box reappears. 100.Select Run 1 Selected Case. The jobs are processed.

This will run both of the heating rate jobs that are defined.

Bring the Case Set Manager up again and select Run Case again. Notice that the program does not actually do any of the calculations. This is because the previous calculations are still valid. If a surface is moved, or an optical property changed, and then try the same Run Case to see the program will recalculate the required data. 101.Select File > Exit. A Thermal Desktop/AutoCAD dialog box appears asking to save the drawing changes. 102.Select Yes.

21-70


21.5

Simple Satellite

What will be learned: • Using trackers • Using insulation • Creating double sided Thermal Desktop surfaces Prerequisites: • Section 20.2 - Setting Up a Template Drawing In this example the simple satellite shown below will be built. This satellite will have two radiation analysis groups, one for internal of the box and one for the external. The five lower surfaces of the box will be coated with insulation. Conduction within the box is going to be ignored.

Double sided solar

Tracker Chip 15 W

Radiator, broken up 5x5

Insulation on these 5 sides

21-71

Simple Satellite Demo Example Start by opening the folder titled simpleSatellite. Copy the template DWG file into the demoSatellite folder. 1. Open the folder named simpleSatellite. (create it if it does not exist). 2. Copy the template thermal.dwg file created in the first tutorial to the new \Tutorials\Thermal Desktop\demoSatellite directory. Note: Be sure to hold the key down if dragging the template file icon to the new directory so that the file is copied, rather than moved. 3. Rename the copied template file to simpleSatellite. 4. Start Thermal Desktop by double clicking on the simpleSatellite drawing file icon in the simpleSatellite directory. Add three new materials: 5.

or Thermal > Thermophysical Properties > Edit Property Data. The Edit Thermophysical Properties dialog box appears. Your properties list may or may not be empty.

6. Type structure in the New property to add field. 7. Select Add. The Thermophysical Properties dialog box appears. 8. Highlight the value in the Conductivity field and enter 100. 9. Highlight the value in the Density field and enter 1000. 10. Select OK to close the dialog box. The Edit Thermophysical Properties dialog box reappears with ‘structure’ displayed in the list field. Notice that the conductivity is 100, specific heat is 1 and density is 1000.

21-72

•

structure: k = 100; = 1;  = 1000

•

MLI, 7-layer: k===0; * = 0.05

•

honeycomb: k=20; ==0; *=0.05

Simple Satellite Demo Example 11. Type MLI, 7-layer in the New property to add field. 12. Select Add. The Thermophysical Properties dialog box appears. 13. Highlight the current value in the Conductivity field and type 0. 14. Highlight the current value in the Specific Heat field and type 0. 15. Highlight the current value in the Effective Emmissivity e-star field and type 0.05. 16. Select OK. The Edit Thermophysical Properties dialog box reappears displaying MLI. 17. Select MLI, 7-layer in the property list and select the Copy button. 18. Type honeycomb in the Copy Material Property form that appears. 19. Select OK. 20. Double-click Honeycomb.

The MLI material has a conductivity of zero and an effective emissivity of 0.05. This means that when used as insulation, or as a core material for surfaces with different node numbers on each side, the conductor through the material will be radiation. If the conductivity were greater than zero and the effective emissivity were zero, the conductors through the material would all be linear. If both values were greater than zero, both radiation and linear conductors would be generated through the material. The internal properties of the insulation are given in the thermophysical properties and the surface properties are given in the optical properties. The honeycomb material has a conductivity of 20 and an effective emissivity of 0.05. This means that when used as insulation, or as a core material for surfaces with different node numbers on each side, two conductors will be created through the material: one linear and one radiation.

The Thermophysical Properties dialog box appears. 21. Highlight the current value in the Conductivity field and type 20. 22. Select OK. The Edit Thermophysical Properties dialog box reappears displaying MLI. 23. Select OK.

21-73

Simple Satellite Demo Example Define optical properties: 24.

or Thermal > Optical Properties > Edit Property Data. The Edit Optical Properties dialog box appears. Your properties list may or may not be empty.

25. Type MLI surface in the New property to add field. 26. Select Add. The Thermophysical Properties dialog box appears. 27. Set Solar Absorptivity to 0.15 28. Set Infrared Emissivity to 0.05 29. Select OK. 30. Repeat for ‘white, zinc oxide’ and ‘solar cells’ properties. 31. Select OK. to close Edit Optical Properties window.

32. Select View > 3D Views > SE Isometric. The UCS icon reflects the change.

21-74

• MLI surface:  = 0.15;  = 0.05 • White, zinc oxide:  = 0.16;  = 0.93 • solar cells:  = 0.82;  = 0.85

Any surfaces without an assigned property will have the DEFAULT property which is  = 1 and  = 1

The surface properties are given in the optical properties and the internal properties, including internal effective emissivity, are given in the thermophysical properties.

Simple Satellite Demo Example 33. Select Thermal > Radiation Analysis Groups. The Radiation Analysis Group Manager dialog box appears.

Radiation analysis groups designate isolated regions of the model for radiation calculation purposes. Two analysis groups are created for this model: • Internal • External

The default analysis group Base can remain in the model and will not affect the calculations if it is not used for any calculations.

34. Select Add. The Add Analysis Group dialog box appears.

35. Type Internal in the New radiation group name field. 36. Select OK to close the Add Analysis Group dialog box. The Radiation Analysis Group Manager dialog box reappears with Internal displayed in the Analysis Group list field.

21-75

Simple Satellite Demo Example 37. Select Add.

The Add Analysis Group dialog box appears. 38. Type External in the New radiation group name field. 39. Select OK to close the Add Analysis Group dialog box. The Radiation Analysis Group Manager dialog box reappears with the 2 new groups displayed in the Analysis Group list field. 40. Select OK to the close the dialog box.

21-76

Simple Satellite Demo Example 41.

or Thermal > Surfaces/Solids > Rectangle. The Command line should now read: Origin point <0,0,0>

42. Press <Enter>. The Command line should now read:

Create the box shown in the inital image of this tutorial by creating a series of six rectangles. Set the top side to be in the External Analysis Group and the bottom side to be in the internal group. Make sure to generate nodes and conductors and put insulation on the top side.

Point for +X axis and X-size <@1,0,0>: 2,0,0

43. Type 2,0,0 in the Command line. The Command line should now read: Point to set XY plane and Ysize <@0,1,0>:

44. Type 0,0,1 in the Command line. The Thin Shell Data dialog box appears.

45. Click on the Radiation tab. 46. Double click on External in the Analysis Group Name, Active Side field.

21-77

Simple Satellite Demo Example The Edit Active Side dialog box appears.

47. Select Top/Out (place a dot in the circle). 48. Select OK. The Thin Shell Data dialog box reappears with top/out displayed next to External. 49. Double click on Internal in the Analysis Group Name, Active Side field. The Edit Active Side dialog box appears. 50. Select Bottom/In (place a dot in the circle). 51. Select OK. The Thin Shell Data dialog box reappears with top/out displayed next to External. 52. Using the drop-down list for Top/Out under Optical Properties for Radiation Calculation, select MLI surface.

21-78

The surface was created such that the top (the +Z of the surface) is on the outside of the spacecraft.

Simple Satellite Demo Example 53. Click on the Cond/Cap tab. Generate Nodes and Conductors is already set by default. 54. Click on the arrow next to the Material field and select structure from the drop-down list. 55. Select the Insulation tab. 56. Click in the check box next to Put on top/out side to select it. The Top/Out field activates.

57. Click on the arrow next to the Material field and select MLI, 7-layer from the drop-down list. 58. Select OK to close the dialog box.

21-79

Simple Satellite Demo Example Create the top and the other sides of the box by copying the first rectangle. Do this so that the top side is always out. One option is to begin by copying the first rectangle and rotating it 90 degrees to create a new side. The rotation can be completed using the grip point Aim X Rotating About Y. A second, possibly easier, option is to use the ARRAY command (ARRAYCLASSIC command in AutoCAD 2012 and later) and create a polar array. 59. Select the newly created rectangle. 60.

(Copy). The Command line should now read: Specify base point or [Displacement/mOde] :

61. Click at the origin to set the first point. The Command line should now read: Specify base point or [Displacement/mOde] : Specify second point or <use first point as displacement>:

62. Click on bottom right corner of the existing rectangle (as currently oriented in the drawing area), on the X axis. A second rectangle appears adjacent to the first. 63. Click on the second rectangle to select it.

21-80

The steps to create the first of the 5 remaining sides is shown to the left.

Simple Satellite Demo Example 64.

(Rotate). The Command line should now read: Specify first point on axis or define axis by [Object/ Last/View/Xaxis/Yaxis/Zaxis/ 2points]:

65. Click on the point at the base of the line separating the two rectangles. The Command line should now read: [Object/Last/View/Xaxis/ Yaxis/Zaxis/2points]: Specify second point on axis:

66. Click on the point at the top of the line separating the two rectangles. The Command line should now read: Specify rotation angle or [Reference]:

67. Type 90 in the Command line. The second rectangle now displays as a side of the box. 68. Repeat the process to create the other 2 sides of the box and the top and the bottom or the box. The top and bottom of the box use the same concepts of copying and rotating. Pay attention to the angles and the axes for the top and bottom. The sides of the box are smaller than the top and the bottom so they will need to be stretched to fit—use the grip points. It is also possible to use the copy and move commands, but to keep the top sides out. It is also fine to zoom in for a closer view and rotate the view.

21-81

Simple Satellite Demo Example 69. Click on the new top side of the box to select it. or Thermal > Edit.

70.

The Thin Shell Data dialog box appears with the Insulation tab displayed. 71. Click on Put on top/out side to deselect it (remove the check mark from the box). 72. Select the Subdivision tab. 73. Highlight the current value in the Xdirection Equal field and type 5. 74. Highlight the current value in the Ydirection Equal field and type 5. 75. Leave Centered Nodes selected. 76. Select the Radiation tab. 77. Select white, zinc oxide from the Top/Out drop-down list of the Optical Properties for Radiation Calculations. 78. Select OK to close the dialog box.

21-82

The top side of the box is the radiator. Change the top side of the box to have a 5x5 breakdown and also to take the insulation off of it.

Simple Satellite Demo Example 79. Create a rectangle to represent the chip using the snap points on the radiator to place the chip as shown in figure to the right. Orient the chip such that the +Z (top) of the chip is facing the inside of the spacecraft. The Thin Shell Data dialog box appears as a part of the creation process.

Create the CHIP. Use the snap points to put it directly in the same plane as the radiator. Note: Hint — how to create a submodel — “Circuit Board Conduction Example” on page 2067.

80. Place the Node ID in the submodel CHIP. 81. Set the radiation for the top side in the Internal Analysis group. 82. Place the conductors in the CHIP submodel. 83. Set the material to structure. 84. Close the Thin Shell Data form 85. Select the chip. 86. Select Modify > Move. 87. Follow the prompts and move the chip down in the Z direction 0.01.

Offset the chip from the radiator for radiation calculations. Move the chip into the box. Note: Hint - Select a corner of the chip as a base and use @0,0,-.01 as the “to” point.

88. Create a or from the chip to the radiator with a value of 5 W/m2K and place in the CHIP submodel. 89. Create a heat load with power of 15 W on the chip surface and place in the CHIP submodel.

21-83

Simple Satellite Demo Example 90. Create a double sided solar . Note: Refer to this image on the next page for a visual reference. Pan and zoom as needed.

Create a solar and a tracker. For the solar : •

Place the surface origin at 1,0,2

•

Make the rectangle 2 m x 3 m

•

Make it a double sided surface and set initial node ID for bottom to 11 Note: Hint - Go to Numbering tab and uncheck Use Same ID’s on both sides

21-84

•

Place nodes and conductors in the SOLAR_ submodel

•

Set the separation distance to be 0.01.

•

Use the structure material for the faces and honeycomb for the separation.

•

Set the radiation analysis External group to “Both”.

•

Use the solar cells optical property for the Top and white, zinc oxide for the bottom.

•

Subdivide 3 in the y and 1 in the x.

Simple Satellite Demo Example 91. Create the tracker.

Create the tracker and modify it to be oriented as shown in the below graphic. Note: Hint, see “Trackers” on page 4-102.

92. Attach the solar to the tracker.

Attaching geometry to a tracker can be accomplished by either using the Thermal > Articulators > Attach Geometry command or by using drag-and-drop in the Tracker list of the Model Browser.

93. Create a basic orbit with default properties

Note: Hint, refer to “Basic Orbit” on page 6-18.

21-85

Simple Satellite Demo Example 94. Use the display active sides and the Model Browser to make sure the model is correct. 95. Display the Case Set Manager.

•

96. Set up and run the case.

Edit the radiation jobs to calculate • radks for the internal analysis group • articulating radks for the external analysis group (the geometry is changing over the orbit) • heating rates for the external group.

97. Select File > Exit. A Thermal Desktop/AutoCAD dialog box appears asking to save the drawing changes. 98. Select Yes.

21-86

•

Set for a steady state solution followed by a transient run of 15000 seconds.

•

Set the output increment to 100 seconds.

•

Run Case (allow node IDs to be automatically resequenced)


21.6

Orbital Maneuvers

What will be learned: • Using array-based symbols • Using expressions • Creating heating rate environments • Using heating-rate-environment symbols • Creating Thermal Desktop Finite Difference Surfaces Prerequisites: • Section 20.2 - Setting Up a Template Drawing • Using the Symbol Manager • Creating Assemblies Imagine analyzing a telescope, and at certain positions in the orbit, it is preferable to have a lens cover for the telescope open, and at other times, the lens cover closed. The following example will demonstrate how this can be done by programming an assembly to accomplish this task. Orbital Maneuvers Example 1. Copy the template thermal.dwg file created in the first tutorial to the \Tutorials\Thermal Desktop\OrbitalManeuvers directory. Note: Be sure to hold the key down if dragging the template file icon to the new directory so that the file is copied, rather than moved. 2. Rename the copied template file to maneuvers. 3. Start Thermal Desktop by double clicking on the maneuvers drawing file icon in the Orbital maneuvers directory. 4.


21-87

Orbital Maneuvers Example (Continued) 5.

or Thermal > Surfaces > Cylinder.

Create a cylinder using the default values shown in the Command line.

The Command line should read: Pick or enter point for base of cylinder <0,0,0>:

6. Press <Enter>. The Command line should read: Pick or enter point for top of cylinder: <@0,0,1>:

7. Press <Enter>. The Command line should read: Enter radius or pick/enter point <1.0>:

8. Press <Enter>. The Command line should read: Enter start angle or pick/ enter point <0.0>:

9. Press <Enter>. The Command line should read: Enter end angle or pick/ enter point <360,0>:

10. Press <Enter>. The Thin Shell Data dialog box appears. 11. Select OK to close the dialog box without making any changes.

21-88

At the current view, the cylinder appears small but notice the top and bottom of the cylinder are displayed with solid lines and a dotted line is shown around the middle. There is also a small symbol in the lower left area of the cylinder.

Orbital Maneuvers Example (Continued) or Thermal > Surfaces >

12.

Create a disk using the default values shown in the Command line.

Disk. The Command line should read: Pick or enter point for center of disk <0,0,0>:

13. Press <Enter>. The Command line should read: Pick or enter point for +Z axis of disk <@0,0,1>:

14. Press <Enter>. The Command line should read: Enter maximum radius or pick/enter point <1.0>:

15. Press <Enter>.

It is hard to see the disk but if the view was zoomed in it would be easier to see a small symbol to the right of the cylinder symbol, as well as a new dotted line.

The Command line should read: Enter minimum radius or pick/enter point <0.0>:

16. Press <Enter>. The Command line should read: Enter start angle or pick/ enter point <0.0>:

17. Press <Enter>. The Command line should read: Enter end angle or pick/ enter point <360.0>:

18. Press <Enter>.

Position the cursor on the symbols, lines and dots and tool tips will display identifying the cylinder and the disc.

The Thin Shell Data dialog box appears. 19. Select OK to close the dialog box without making any changes.

21-89

Orbital Maneuvers Example (Continued) 20. Click on the newly created disk to select it. Note: It may be hard to see the disk. it was created at the 0,0,0, origin. Position the cursor on the origin, the disc symbol and/or the dotted line outlining the disc and use the tool tip feature to find the disk. The disk is selected and blue grip points are displayed. 21. Click on one of the grip points—this makes the grip point “hot”. 22. Press and hold down the right mouse button and select Move from the popup menu. A “copy” of the disk appears and is attached to the original disk by a line that moves and changes as the cursor is moved. The Command line should read: Specify move point or [Base point/Copy/Undo/eXit]:

23. Press and hold down the right mouse button again and select Copy from the popup menu. 24. Move the cursor until the new disk image is at the top of the cylinder. The disk will “snap” to the top of the cylinder. 25. Click the left mouse button to place the copy. 26. Press <Esc> to end the copy command. 27. Press <Esc> to deselect all objects. A dotted line and a solid line is displayed representing the new disk at the top of the cylinder.

21-90

Copy the new disk to the top of the cylinder.

Orbital Maneuvers Example (Continued) Create a Basic orbit named Basic. 28.



30. Type Basic in the New Heating Case Name field. 31. Select OK to close the dialog box.

21-91

Orbital Maneuvers Example (Continued) The Orbit: Basic dialog box appears.

The Beta angle is already set at 0 so no change needs to be made. 32. Highlight the current value in the Altitude field and type 4000. 33. Select the Positions tab.

34. Highlight the current value in Increments field and type 100. 35. Select OK.

21-92

Set the orbit altitude to 4000 km and increase the number of orbit positions to 100.

Orbital Maneuvers Example (Continued) The Heating Rate Case Manager dialog box reappears with Basic displayed in the Current Heating Rate Case field.

36. Select Display Orbit. The drawing area displays the orbit.

Note: Exit the orbit view and return to a wireframe view.

The orbit engineer has specified that the telescope will be closed when the Mean Anomaly is 120 and then should reopen when the Mean Anomaly is greater than 240. Create an assembly at the location 1, 0, 1. 37.

or Thermal > Orbit > Orbit Display Off.

Note: See Section 4.14.1: Assemblies on page 4-101.

38. Create an assembly at the location 1,0,1. (Select Thermal > Articulators > Create Assembly.) 39. Attach the top disk to the assembly.

The top disk has an origin at 0, 0, 1.

40. Select the assembly.

Edit the assembly.

41.

or Thermal > Edit. The Edit Assembly dialog box appears.

42. Select the Trans/Rot tab. 43. Double click in the Rotation 2 field. The Expression Editor dialog box appears. 44. Type the following into the Expression field:

Type the following c-style conditional input: (hrMeanAnom < 120.1 || hrMeanAnom > 239.1) ? 180 : 0 hrMeanAnom is a symbol that is automatically created with heating rate environments. To avoid typos, right-click in the expression field, select orbital and choose hrMeanAnom.

(hrMeanAnom < 120.1 || hrMeanAnom > 239.1) ? 180 : 0

21-93

Orbital Maneuvers Example (Continued) 45. Click OK. 46. Click OK. 47. Select Thermal > Orbit > View Vehicle > Set Orbit Pos/Prefs.

Rotate the model a little bit to get some perspective.

The View Vehicle in Orbit dialog dialog box appears. 48. Select Animate. 49. Select OK. The Continuous Cycle Dialog dialog box appears. 50. Select OK. The orbit engineer has come back and said that a step change in the cover position is no longer acceptable. They want the cover to start to close when the mean anom is 60, and then be completely closed at mean anom = 120. On the other side, they want the cover to start opening at mean anom = 240, and then be completely open when the mean anom = 300. 51. Bring up the Symbol Manager and create a new symbol called angle_array. 52. Select ARRAY from the Type filed pulldown. 53. Type the following data into the Expression Editor angle_array field, one entry per line: 0 60 120 240 300 360

21-94

In this step, create an array of various Orbital Angles. Note: See Section 11.1.1: Symbol Manager on page 11-1 for information on symbols and the Symbol Manager.

Orbital Maneuvers Example (Continued) 54. Create a second symbol called cover_array.

In the next step, create an array for the angle of the cover.

55. Select ARRAY from the Type filed pulldown. 56. Type the following data into the Expression Editor angle_array field, one entry per line: 180 180 0 0 180 180 57. Edit the assembly, and double click in Edit Assembly dialog box Rotation 2 field.

After the arrays are created, program the assembly to rotate based on the interpolation of the arrays for the current position being analyzed.

58. Edit the expression to be:

interp is an internal function that simply does linear interpolation.

interp(angle_array, cover_array, hrMeanAnom) 59. View the model in orbit. The orbital engineer has come back and wants to run a case where the entire spacecraft is spinning about the velocity vector. The orbital engineer has details that the spacecraft will make 5 full rotation per orbit. 60. Edit the orbit. 61. Select the Orbit dialog box Orientation tab.

Edit the orbit to the orbital engineer’s specifications.

62. Double click in the first Additional Rotations field. 63. Make 5*hrMeanAnom expression. 64. View the orbit and then the model animating in orbit.

The first thing of notice is that the coordinate systems look a little funny—that is because they are rotating about the X axis. When the model is viewed in orbit, notice that the entire spacecraft is spinning as the cover is opening and closing.

21-95

Orbital Maneuvers Example (Continued) 65. Select File > Exit. A Thermal Desktop/AutoCAD dialog box appears asking to save the drawing changes. 66. Select Yes.

21-96


22 FloCAD® Tutorials This set of tutorials is focused on the FloCad application. Five example tutorials follow which show how to model air flow, heat pipes, and flow through a manifold. • Section 22.1: Air Flow Through an Enclosure on page 22-3 • Section 22.2: Heat Pipe Model on page 22-23 • Section 22.3: Manifolded Coldplate on page 22-37 • Section 22.4: Drawn Shape Heat Pipe on page 22-85 • Section 22.5: FEM Walled Pipe on page 22-99 At this point, the first three tutorials from Chapter 20: "Setting Up a Template Drawing" should have been completed. These tutorials give an overview of Thermal Desktop and how to create geometry, as well as define properties. The typographical conventions are defined there as well.

FloCAD® Tutorials

22-1

22-2

FloCAD® Tutorials

22.1

Air Flow Through an Enclosure

What will be learned: • How to create a fluid model • How to create all types of fluid elements In this example, the components necessary to model airflow through an enclosure with electronic components will be added. The exercise will start using a model of an enclosure for which the geometry has already been created. The enclosure has a flat base 30x30 cm. Three boards extend up from the base, with each containing a chip producing 25 watts of heat. There is a set of conductors connected to the base plate to simulate convection to an ambient temperature. Two flow paths through the enclosure are created. The temperature distribution in the enclosure will be computed for a given flow rate. Want to Learn More? Refer to Chapter 5: "Fluid Models" in the Thermal Desktop ’s Manual for detailed information on fluid models.z Want to Learn More?

Air Flow Example 1. Double click on the file fluid.dwg located in the Tutorials\Thermal Desktop\airflow folder. Thermal Desktop opens with the fluid drawing on the screen.

Figure 22-1

Fluid Drawing Initial View

2. Type Zoom and then 0.2 22-3

Air Flow Example (Continued) 3.

or Thermal > Fluid Modeling > Lump.

Create two lumps on one side of the circuit boards. One lump will be changed to a Plenum later in the tutorial.

The Command line now reads: Enter location of lump:

4. Type -20,15,5 in the Command line. (<ENTER> is implied when something is typed into the Command line) The Command line now reads: Command:

5. Press <Enter>. Note: <Enter> is used to repeat the last command and create a second lump for the outlet of the fan.

Figure 22-2

After Step 6

The Command line now reads: Enter location of lump:

6. Type -10,15,5 in the Command line. The Command line now reads: Command:

7.

or Thermal > Fluid Modeling > SetFlow. The Command line now reads: Select from lump:

8. Click on the 1st lump. The Command line now reads:

Create a fan between the two lumps that were just created. The default direction of positive flow in all paths will be in the direction in which the lumps are selected. The fan will be modified shortly to set the flow rate.

Select to lump:

9. Click on the 2nd lump.

Figure 22-3

22-4

After Step 9


or type Layer in the Command line or select Format > Layer. The Layer Properties Manager dialog box appears.

Turn on and off some layers to make it easier to select the points to generate the lumps and paths within the enclosure. Notice that Fluid is already the current layer (green check mark), which is where all of the fluid submodel components will be placed. The geometry was already created for convenience, but can easily be created with AutoCAD Draw commands.

11. Click on the On (lightbulb) and the Freeze (snowflake) icons of the Construction layer to turn the layer on (lighten the lightbulb and change the snowflake to a sun). 12. Click on the On (lightbulb) and the Freeze (snowflake) icons of the Flow Areas layer to turn the layer on (lighten the lightbulb and change the snowflake to a sun).

Figure 22-4

After Visibility Changes

13. Click on the Freeze (sun) icon of the Board layer to turn the layer off (snowflake). 14. Close the Layer Property Manager.

22-5


or Thermal > Fluid Modeling > Lumps and Paths. The Create Lumps and Paths dialog box appears.

Create all of the lumps and paths within each side of the enclosure with one command. Notice that the shape of line 5 follows along the edge of the boards for this flow channel. This provides the wetted perimeter for the flow area. The shape is closed to get the area by assuming there is a line between the two endpoints. The code then computes the hydraulic diameter and flow area for all four paths from this shape. This shape was created with a polyline.

16. Click on the Pick Point to Pick Point radio button if not already selected (place dot in the circle). 17. Highlight the current value in the Number of Lumps to Create field and type 5. 18. Select OK. The Command line now reads: Select start point:

19. Click at the midpoint of line 1. Note: A midpoint snap point (a triangle) should appear as the cursor moves midway on line 1. The Command line now reads: Select end point:

20. Click at the midpoint of line 2. The Command line now reads: Select Upstream Entity for Area Calculation (Enter for Specified Area):

21. Click on a point on line 5.

22-6

Figure 22-5

Step 23

New Lumps and Paths after

Air Flow Example (Continued) The Command line now reads: Select Downstream Entity for Area Calculation (Enter for Same Entity):

22. Press <Enter> to reuse the same . 23. Repeat the command picking the midpoints of lines 3 and 4, and line 6 for the shape. • Press <Enter> to recall the command.

Enter or right-click will repeat the previous command.

The Create Lumps and Paths dialog box appears. • Leave Pick Point to Pick Point

selected and 5 as the number of lumps to create. • Select OK to close the dialog box. The Command line now reads: Select start point:

• Click at the midpoint of line 3. The Command line now reads: Select end point:

• Click at the midpoint of line 4. The Command line now reads: Select Upstream Entity for Area Calculation (Enter for Specified Area):

• Click on a point of line 6. The Command line now reads: Select Downstream Entity for Area Calculation (Enter for Same Entity):

• Press <Enter>.

22-7


or type Layer in the Command line or select Format > Layer. The Layer Properties Manager dialog box appears.

25. Click on the Freeze (sun) icon of the Construction layer to turn the layer off (display a snowflake). 26. Click on the Freeze (sun) icon of the Flow Areas layer to turn the layer Figure 22-6 off (display a snowflake).

After Step 27

27. Close the Layer Properties Manager. Create the lump for the outlet plenum. 28.

or Thermal > Fluid Modeling > Lump. The Command line now reads: Enter location of lump:

29. Type 50,15,5 in the Command line.

Figure 22-7

22-8

After Step 29


or Thermal > Modeling Tools > Toggle Selection Filter. In the Command area you should see: _RcFilter Thermal Desktop filter turned on

31.

To view the lump numbers for selecting them in the following steps, set the selection filter “on”. Make sure that the Command line shows “on” after selecting the toggle command. Then use the selection filter to turn on the IDs for the lumps.

or Thermal > Modeling Tools > Turn Numbers On. The Command line now reads: Select entity(s) to display ids or [GRP]:

32. Type all in the Command line. The Command line now reads: Select entity(s) to display ids or [GRP]:

33. Press <Enter>. The Object Selection Filter dialog box appears.

Figure 22-8

Lumps with IDs

Note: If you had to delete and recreate some lumps during this tutorial, then your numbering will be slightly different. The remaining tutorial will refer to the numbers as shown above.

34. Highlight Lumps[13] in the Select type to filter field if not already selected. 35. Select OK.

22-9


or Thermal > Fluid Modeling > Loss. The Command line now reads: Select from lump:

37. Click on lump 2. Note: Click on the lump numbers rather than on the lump symbols for easier selection. The Command line now reads: Select to lump:

Now generate all of the entrance and exit paths. Use the loss coefficient to model the entrance and exit effects. The losses will be edited in a later step. As with the Pump/Fan command, the order the lumps are selected determines the direction for positive flow. Note that picking on the lump numbers with the mouse is probably the easiest way to select each lump.

38. Click on lump 3. The Command line now reads: Select Entity for Area Calculation (Enter for Specified Area):

39. Press <Enter>. 40. Use the same command to create losses from each of: Lump 2 to Lump 8; Lump 7 to Lump 13; and, Lump 12 Figure 22-9 to Lump 13. • Press <Enter>. The Command line now reads: Select from lump:

• Click on lump 2. The Command line now reads: Select to lump:

• Click on lump 8. The Command line now reads: Select Entity for Area Calculation (Enter for Specified Area):

• Press <Enter>. Lumps 2 and 8 are connected.

22-10

View after Step 40

Air Flow Example (Continued) • Press <Enter>. The Command line now reads: Select from lump:



• Press <Enter>. Lumps 7 and 13 are connected. • Press <Enter>. The Command line now reads: Select from lump:



• Press <Enter>. Lumps 12 and 13 are connected.

22-11

Air Flow Example (Continued) 41. Select View > 3D Views> Top or click TOP on the View Cube

Simplify the view for selecting.

View Cube Figure 22-10

42. Select the SetFlow (the fan created in Step 6 from Lump 1 to Lump 2) with the mouse. or Thermal > Edit.

43.

The SetFlow edit form dialog box appears.

Top View

The SetFlow looks like this:

Here the object (the SetFlow) is selected before the command is issued. In this order, the command acts on the item(s) that have already been selected. Assume the flow rate for this fan is known and select a mass flow rate device.

44. Make sure the SetFlow Data tab is selected. Highlight the current value in the Mass Flow Rate field and type 0.1. 45. Select OK.

22-12


or Thermal > Edit. The Command line now reads: Select objects or [Indiv | GRP]:

47. Type all in the Command line. The Command line now reads: Select objects or [Indiv |GRP]:

Here the command (Edit) is issued before anything is selected. In this order, the command requests that objects be selected. Use 100 cm2 for all of the inlet and outlet flow areas. Because they are all the same type of path, they can be selected using the Object Selection Filter. Also leave the FK value at 1.0 for an entrance and exit loss.

48. Press <Enter>. The Object Selection Filter dialog box appears. 49. Select Losses[4] in the Select type to filter field. 50. Select OK. The Loss edit form dialog box appears.

22-13

Air Flow Example (Continued) 51. Select the Flow Area Tab. In the Flow Area (AF) field, type 100.0 in the box. 52. Select OK. A Multi Edit Dialog box appears confirming the change. 53. Select Apply Changes. Turn off the selection filter. 54.

or Thermal > Modeling Tools > Toggle Selection Filter. In the Command area you should see: _RcFilter Thermal Desktop filter turned off

22-14

Make sure the Command line output states the filter is ‘turned off’.

Air Flow Example (Continued) 55. Select lumps 1 and 13. Note: Click on the lump numbers to insure only the lumps are selected. or Thermal > Edit.

56.

The Lump Edit Form dialog box appears.

The lumps created by default are junctions. For the model to run, the source and sink lumps must be plena. Lumps 1 and 13 can be changed together by selecting them together and changing their type in the Lump edit form dialog box. Notice that the shape of the lumps changes from a circle to a triangle. Thermal components share a shape with the fluid submodel components, but the fluid shapes also have interior lines and a vertical line normal to the plane.

57. Click the Plenum (Boundary Infinite Volume) radio button to select it (place a dot in the circle). 58. Select OK to close the dialog box. A Multi Edit Dialog box appears confirming the change. 59. Select Apply Changes. or type Layer in the Command

60. line.

The Layer Properties Manager dialog box appears. 61. Click on the Freeze (sun) icon of the Board layer to turn the layer on (display a sun). 62. Close the Layer Properties Manager.

22-15


or Thermal > Fluid Modeling > Tie to Surface. The Command line now reads: Select the lumps to be tied to or [GRP]:

64. Select lumps 3 to 7 by using a leftto-right selection box as shown to the right with the heavy line.

Create the thermal connection between the fluid submodel and the thermal submodel using a “tie”. The lumps and path sets are both selected using a left to right box (thick line) shown below. Note that the heat transfer coefficient will be computed by SINDA/FLUINT since there is a selected a path set.

The Command line now reads: Select the lumps to be tied to or [GRP]:

65. Press <Enter>. The Command line now reads: Select the surfaces for the tie or [GRP]:

66. Type GRP in the Command line.

Figure 22-11

Selecting lumps or paths

The Select Groups dialog opens 67. Select RIGHT from the Select Groups list and click OK.

As an aid in selection, the AutoCAD groups named ‘left’ and ‘right’ have In the Command area you should see: been pre-created in the DWG file by Added 4 4 found the authors. Groups can be created by Select the surfaces for the using the ‘groups’ command in Autotie or [GRP]: CAD. 68. Press <Enter>. The Command line now reads: Select the paths to be tied to (None for specified HTC) or [GRP]:

69. Select paths between lumps 3 and 7 as done previously for selecting the lumps. Figure 22-12 The Command line now reads: Select the paths to be tied to (None for specified HTC) or [GRP]:

70. Press <Enter>.

22-16

View after Step 70


or Thermal > Fluid Modeling > Tie to Surface.

The steps from the previous page are repeated for the left group

The Command line now reads: Select the lumps to be tied to or [GRP]:

72. Select lumps 8 to 12 by using a leftto-right selection box. The Command line now reads: Select the lumps to be tied to or [GRP]:

Figure 22-13

View after Step 78

73. Press <Enter>. The Command line now reads: Select the surfaces for the tie or [GRP]:

74. Type GRP in the Command line. The Select Groups window opens. 75. Select LEFT in the Select Groups list and click OK. In the Command area you should see: Added 5 5 found Select the surfaces for the tie or [GRP]:

76. Press <Enter>. The Command line now reads: Select the paths to be tied to (None for specified HTC)or [GRP]:

77. Select paths between lumps 8 and 12 by using a left-to-right selection box. The Command line now reads: Select the paths to be tied to (None for specified HTC) or [GRP]:

78. Press <Enter>.

22-17

Air Flow Example (Continued) 79. Select View > 3D Views> SW Isometric.

80. Select Thermal > Fluid Modeling > Submodel Manager. The FLUINT Submodel Manager Form dialog box appears.

The FLUINT Submodel Manager is where the fluid is set. The default is Air, so there is nothing to do for this model. In addition to the library fluids, the could define their own fluid.

81. With FLOW selected, select the Properties button. The Fluid Submodel Properties dialog box appears.

22-18

Air Flow Example (Continued) 82. Select Air (A) (if not already selected). 83. Select the Edit button. The Fluid Submodel Fluid Property List Edit dialog box appears.

84. Select OK to close the Fluid Submodel Fluid Property List Edit dialog box. 85. Select OK to close the Fluid Submodel Properties dialog box. 86. Select OK to close the FLUINT Submodel Manager Form dialog box. 87. Select File > Save.

22-19

Air Flow Example (Continued) or Thermal > Case Set Man-

88. ager.

The Case Set Manager dialog box appears. 89. Select Run 1 Selected Case. Sinda/Fluint Run Status dialog box

appears confirming the successful completion of the process and the drawing area graphics update. 90. Click OK to close the dialog box.

22-20

The Case Set Manager runs a steady state case by default, which is what is needed for this case. Therefore, no changes are required to get a solution. After the solve is completed, the final temperatures are displayed on the model in the postprocessing state. You may need to select SW Isometric view.


or Thermal>Post Processing>Edit Current Dataset. The Set Sinda Dataset Properties dialog box appears.

92. Select the drop-down menu beside each item to see the options available. 93. Click OK to close the dialog box.

94.

or Thermal > Post Processing > Cycle Color Bars.

95. Repeat Cycle Color Bars several times noting the change to the color bar and its labels.

The Set Sinda Dataset Properties window allows the to choose the Time or Record to postprocess and which data to view for different objects. Since this run was steady state and the output was written only after the solution was reached, only one record is available.

The Cycle Color Bars command can be used to cycle between the four types of color bars (Node, Lump, Path, and Tie). The color scale for each as well as the current variable being displayed can be found on the each of the individual color bars. The various views (Top, Bottom, SW Isometric, etc.) can be used to see all or a portion of the Board. To make viewing easier, objects not associated with the current color bar are drawn as gray. Smart Color Bar Cycling, applied next, hides all other objects.

22-21


or Thermal > Post Processing > Edit Current Dataset.

Smart Color Bar Cycling turns off the visibility of all objects except those associated with the active color bar.

The Set Sinda Dataset Properties dialog box appears. 97. Select the Smart Color Bar Cycling checkbox (so a check mark is displayed). 98. Click OK to close the dialog box. 99.

or Thermal > Post Processing > Cycle Color Bars.

100.Repeat Cycle Color Bars several times noting the change to the display. 101.Select File > Exit. A Thermal Desktop/AutoCAD dialog box appears asking to save the drawing changes.

With Smart Color Bar Cycling selected, the visibility of objects changes along with the color bar such that nodes and surfaces are shown with the Node color bar, lumps are shown with the Lump color bar, etc. Exit Thermal Desktop and save as prompted.

102.Select Yes. As time permits, advanced s may consider making the following changes to the model: • Change the inlet plenum or the SetFlow as a function of time (sine wave function or array) • Run the model as a transient solution.

22-22

22.2

Heat Pipe Model

What will be learned: • Creating AutoCAD lines and polylines • Creating FloCAD pipes from those lines • Modeling a fixed conductance heat pipe (CCHP/FCHP) • Using ors In this example, a heated aluminum ring will be connected to an air-cooled (20-°C ambient at 10 BTU/hr-ft2-°F) aluminum plate using an aluminum/ammonia fixed conductance heat pipe (FCHP). Radiation has been neglected, and the heated ring therefore has no other route for heat to leave other than through the heat pipe. The ring outer diameter is 8 cm, and the inner diameter is 2 cm, and it is 0.1cm thick, a value which has been parameterized as the Thermal Desktop symbol HotThk in case it must be changed. The cooling plate is 10cm wide by 15cm long by 0.5cm thick (similarly parameterized as PlateThk). The power on the disk is currently 25, as determined by current value of the symbol Power.

22-23

Heat Pipe Example 1. Double click on the file heatpipe.dwg located in the Tutorials\Thermal Desktop\heatpipe folder. Thermal Desktop opens with the heat pipe drawing on the screen.

The disk and plate have already been generated, along with the heat load on the disk and the convection environment on the plate. The units for this model are Watts, seconds, cm, and °C.

22-24

Heat Pipe Example (Continued) 2.

or type Layer in the Command line or select Format > Layer. The Layer Properties Manager dialog box appears. Note: Plates is the current layer.

3. Click on the Freeze (sun) icon of the Boundary layer to turn the layer off.

The current drawing shows the disk and plate, along with the heat load on the disk and the convection conductor on the plate. To prepare to add the heat pipe, turn off the visibility of some distracting elements. First, turn off the layer containing the convection boundary.

4. Close the dialog box.

5.



Next, turn off visibility of the heat loads. An alternative method for changing the visibility of heat loads is to use the icon:

When complete, the model should look similar to the view below:

7. Select Heat Loads/Heaters/Pressures to deselect it (remove the check mark from the box). 8. Select OK to close the dialog box.

22-25

Heat Pipe Example (Continued) 9. Select Tools > Named UCS. The UCS dialog box appears.

Use a polyline to create the heat pipe in the next several steps. Polylines can only be created in the xy plane of the current UCS. For this example a pre-prepared UCS has been created. Switch to this UCS. Change to the pre-prepared UCS in the plane of the heat pipe. Then change to a better view. The model should now look similar to the view below.

Note: World is the current UCS setting. 10. Select HeatPipe in the Current UCS field, the select the Set Current button. 11. Press OK to close the UCS dialog box. Note: Notice the UCS icon moves to a new position on the drawing. 12. Select View > 3D Views > SE Isometric.

22-26


or Draw > Polyline. The Command Line now reads: Specify start point:

14. Position the cursor at the top of the disk, which should be highlighted by a green square with the tool tip Endpoint displayed. 15. Click the left mouse button at this point (0,14,0) to start the polyline. The Command Line now reads: Specify the next point or [Arc/Halfwidth/Length/Undo/ Width]:

16. Click on the origin of the UCS (highlighted by a green “X” at 0,0,0). The Command Line now reads: Specify the next point or [Arc/Halfwidth/Length/Undo/ Width]:

Draw a line object representing the centerline of the heat pipe. Note: Practice drawing lines and polylines, under the Draw menu, off to the side of the drawing. Delete the lines when finished. In this case, use a polyline which must be in a single plane: the XY plane of the current UCS. Note: Other lines can be connected to this one as needed to create various 3D objects. With the new polyline selected (highlighted) the drawing should look similar to the view below. If the coordinates do not appear in the status bar along the bottom of the window, select the three bars at the bottom right of the Thermal Desktop window to place a check mark beside Coordinates

17. Click on the midpoint at the far end of the plate with the tool tip Midpoint, (15,0,0) on the indicator, displayed. 18. Press <Enter> to terminate the polyline.

22-27

Heat Pipe Example (Continued) 19. Click on the polyline between the disk and the plate to select it (highlight it). 20.

or Thermal > Fluid Modeling > Pipe. The Command line should now reads: Select Line Entity for Pipe Cross Section Shape Definition:

21. Press <Enter>. The Ripe Edit Form dialog box appears with the Pipe Selection tab displayed. 22. Select Heat Pipe on the current Pipe Selection tab to place a dot in the radio button (circle). Note: Notice when Heat Pipe is selected two of the tabs — Pipe Attributes and Ties — are replaced by a Heat Pipe Data tab.

22-28

Now turn the polyline into a FloCAD pipe. Specifically, a heat pipe. The Create Pipe command first requires a line entity for the center line. In this example the line is selected before the command. Commands will use preselected objects if they can. A line for the Pipe Shape Definitions will not be selected because a circular cross section is desired. If another shape was desired, the line should have been drawn before selecting the pipe icon.

Heat Pipe Example (Continued) 23. Select the Subdivision tab. 24. Highlight the current default value of 4 in the Pipe Length Equal field and type 50. 25. Select the Node Numbering tab. 26. In the Both Sides field, click on the arrow next to the Submodel field and select HEATPIPE from the dropdown list. The Vapor Node Submodel field updates to HEATPIPE to reflect the change made in the Both Sides Submodel field. 27. Highlight the current value in the Vapor Node ID field and type 1001.

Set the axial resolution to 50 and leave it as uniform for convenience. The Heatpipe submodel was created by the authors. Important: Make sure that the vapor node ID is unique—it must not be the same as any other node, including other vapor nodes used by other heat pipes. Similarly, heat pipe nodes should not be duplicated or merged with any other nodes in the model.

22-29

Heat Pipe Example (Continued) 28. Select the Heat Pipe Data tab.

29. Click on the arrow next to the Wall Material field and select Aluminum from the drop-down list. 30. Highlight the current value in the Vapor Core Diameter field and type 0.8. 31. Click on the arrow next to the Wall input field and select Area Input from the drop-down list. 32. Highlight the current value in the Wall Area field and type 0.3. 33. Select OK to close the Ripe Edit Form dialog box.

22-30

When the heat pipe set-up is complete, nodes and pipe segments will appear, evenly spaced on the centerline of the pipe.


or Thermal > FD/Fem Network > or. The Command Line now reads: Select faces ing from:

35. Select the heat pipe in the drawing area. The Command Line now reads:

Now connect the heat pipe to both the disk and the plate using a or. The choice of “from” and “to” sets for the or is critical. All nodes on the “from” objects will be ed to the nearest nodes on the “to” objects, subject to the input tolerance. Therefore, choose the pipe as the “from” object, and both of the surfaces as the “to” objects for the or.


36. Press <Enter>. The Command Line now reads: Select surfaces ing to:

37. Select the disk and then the plate (either order) in the drawing area. 38. Press <Enter>. The or dialog box appears.

22-31

Heat Pipe Example (Continued) 39. Click on the arrow next to the Conductor Submodel field and select HEATPIPE from the drop-down list. 40. Click on the arrow next to the From field and select Edges from the drop-down list. 41. Highlight the current value in the Conductor Coefficient field and type 1. 42. Highlight the current value in the Tolerance field and type 0.5. 43. Double-click pipe object in the From list.

As it is, the or will create connections in the “adiabatic” portion of the heat pipe between the disk and plate, so a tolerance must be used to prevent nodes that are too far away from each other from being linked. In other words, the lowered tolerance eliminates the undesirable s to the middle of the heat pipe.

The Select Edges dialog box appears. The edge or surface that is created from can be edited in the or dialog box From field. Double clicking on an object in the list will bring up an edit dialog box (in this case the Select Edges dialog box) with the choices for that object listed. The or appears as arrows, green for each From surface and gold for each To surface.

44. Click on Inlet Edge to deselect it (remove the check mark from the checkbox) and click on Along Pipe Length to select it (display a check mark in the check box). 45. Select OK to close the Select Edges dialog box. 46. Select OK to close the or dialog box.

22-32



48. Select the Graphics Visibility tab. 49. Select TD/RC Nodes to deselect it (remove the check mark from the box). 50. Select ors to deselect it (remove the check mark from the box). 51. Select OK to close the dialog box.

In preparation for running and postprocessing, “clean up” the drawing a little by turning off visibility of nodes and ors, and turning off the UCS origin marker. On alternative method for changing visibility of TD/RC nodes is the icon:

And for ors, the icon:

22-33

Heat Pipe Example (Continued) 52. Select Tools > Named UCS. The UCS dialog box appears.

53. Select the Settings tab.

54. Click On in the UCS Icon settings field to deselect it (remove the check mark from the checkbox). 55. Press OK to close the UCS dialog box.

22-34

This will turn the UCS icon off.

Heat Pipe Example (Continued) or Thermal > Case Set Man-

56. ager.

A steady state run has been set up in the Case Set Manager. The results should look similar to the graphic below.

The Case Set Manager dialog box appears. 57. Select Run 1 Selected Case. Sinda/Fluint Run Status dialog box

appears confirming the successful completion of the process and the drawing area graphics update. 58. Click OK to close the dialog box.

The hottest spot on the thin disk is about 88 degrees, and the coldest spot on the plate is about 45 degrees. 59. Select File > Exit. A Thermal Desktop/AutoCAD dialog box appears asking to save the drawing changes.


60. Select Yes. This tutorial demonstrated how to draw a line, turn it into a heat pipe, and connect it to the rest of the thermal model. To try some slightly more difficult problems, try these: • Instead of polyline consisting of two straight segments at right angles to each other, try repeating the above problem using a slightly more complicated but more realistic polyline: a line, a 90 degree bend (“arc”), then another line. It may be necessary to create construction points for the arc, corresponding to the desired radius. • Try adding a little gas to the heat pipe and see what the effect is. To do so, edit the pipe and change it to be a “Fixed Conductance Heat Pipe with NC-gas” (Noncondensible gas). Both the working fluid (ammonia) and the gas (use “air” for demonstration purposes) must also be specified. For a rather high value of 1.0e-4 kg (0.1 gm), the disk now reaches 107 degrees. Blockage of the pipe can be seen by the gradients in the plate (as shown below).

22-35

Figure 22-14

22-36

Larger View of Solution

22.3

Manifolded Coldplate

What will be learned: • Creating a polyline-based FloCAD pipe with smooth bends • Connecting pipes • Adding fluid network components • Accessing a -defined fluid • Defining a fan or pump curve • Postprocessing a transient run In this example, a manifolded copper “coldplate” bonded to a thick (1/2 inch) 10.5”x11.36” aluminum plate will be constructed. The working fluid is 50-50 water and ethylene glycol. This is not a library (built-in fluid), but a FLUINT FPROP file has been supplied. The plate is initially at 80°F and convects to an 80°F air temperature with a convection coefficient of 30 BTU/hr-ft2-F (stored as parametric symbol EnviroU). At time zero, 100W (symbol Power) is dissipated into the loop. However, the plate is not able to reject all of this power and still stay below the temperature limit 110°F (symbol Tlimit). At steady state, the peak temperature is about 126°F. Instead, the device is intended to provide temporary rejection capability for a surge in power by relying in part on its thermal mass. The point of the analysis it to determine how long before the device reaches its capacity: how long can it withstand the power surge?

22-37

Manifolded Coldplate Example 1. Double click on the file manifold.dwg located in the Tutorials\Thermal Desktop\manifold folder. Thermal Desktop opens with the manifold drawing on the screen.

Figure 22-15

Manifold Drawing Initial View

A FEM model of the plate and its convection environment have already been generated (but are not currently visible), along with some symbols that can be used to generate a parametric model. The units of the initial drawing are BTU, inches, seconds, °F.

Specify start point:

First, draw an AutoCAD polyline representing the outer (manifold) pipe. This involves generating lines, switching to arcs, switching back to lines, etc. all without leaving the polyline generation mode.

3. Select the point in the center of the red circle labeled 1 located in the top left corner of the drawing area.

Note: The should practice such drawing methods on the side of the drawing area, deleting trial lines.

or Draw > Polyline.

2.

The command line should now read:

This is the starting point for the polyline, and also for the first straight segment. The command line should now read: Specify next point or [Arc/ Halfwidth/Length/Undo/ Width]:

22-38

Ignore the green marks during this step.

Manifolded Coldplate Example (Continued) 4. Select the point in the center of the red circle labeled 2 located to the right of circle 1.

The final drawing should look similar to the view below.

Note: A straight line should appear between points 1 and 2. The command line should now read: Specify next point or [Arc/ Halfwidth/Length/Undo/ Width]:

5. Type a to start the arc drawing mode. Note: to press the <Enter> key after entering a command. The command line should now read: Specify endpoint of arc or [Angle/CEnter/CLose/Direction/Halfwidth/Line/Radius/ Second pt/Undo/Width]:

6. Select the point in the center of the red circle labeled 3.

Figure 22-16

New Polyline

If the final line is not correct, delete it and try again.

Note: A curved line should appear connecting points 2 and 3. The command line should now read: Specify endpoint of arc or [Angle/CEnter/CLose/Direction/Halfwidth/Line/Radius/ Second pt/Undo/Width]:

7. Type L (lower case is fine) to start the line drawing mode. The command line should now read: Specify next point or [Arc/ Halfwidth/Length/Undo/ Width]:

8. Select the point in the red circle labeled 4 located below the red circle 3. The command line should now read: Specify next point or [Arc/ Halfwidth/Length/Undo/ Width]:

22-39

Manifolded Coldplate Example (Continued) 9. Type a to start the arc drawing mode. The command line should now read: Specify endpoint of arc or [Angle/CEnter/CLose/Direction/Halfwidth/Line/Radius/ Second pt/Undo/Width]:

10. Select the point in the red circle labeled 5. The command line should now read: Specify endpoint of arc or [Angle/CEnter/CLose/Direction/Halfwidth/Line/Radius/ Second pt/Undo/Width]:

11. Type L (lower case is fine) to start the line drawing mode. The command line should now read: Specify next point or [Arc/ Halfwidth/Length/Undo/ Width]:

12. Select the point in the red circle labeled 6. 13. Press <Enter> to terminate the polyline.

22-40

Manifolded Coldplate Example (Continued) 14. Select the new line to highlight it. 15.

or Thermal > Fluid Modeling > Pipe.

Convert the line into a pipe segment, name its components, specify its materials, dimensions, resolution, etc.

The command line should now read: Select Line Entity for Pipe Cross Section Shape Definition:

16. Press <Enter>. 17. The Ripe Edit Form dialog box appears with the Pipe Selection tab displayed. Confirm that Fluid Pipe with wall is selected (the default) on the Pipe Selection tab.

22-41

Manifolded Coldplate Example (Continued) 18. Select the Subdivision tab. 19. Highlight the current default value of 4 in the Pipe Length Equal field and type 50.

22-42

Subdivide the pipe into 50 segments.

Manifolded Coldplate Example (Continued) 20. Select the Node Numbering tab. 21. In the Both Sides region, click on the arrow next to the Submodel field and select WALL from the drop-down list.

Set the submodel for the pipe wall nodes to ‘Wall’.

22. In the End Nodes region, click on the arrow next to the Submodel field and select WALL from the drop-down list.

22-43

Manifolded Coldplate Example (Continued) 23. Select the Pipe Attributes tab. 24. Click on the arrow next to the Fluid Submodel field and select Water from the drop-down list. 25. Click on the arrow next to the Cond. Submodel field and select WALL from the drop-down list. 26. Click on the arrow next to the Outer Material Properties field and select Copper from the drop-down list. 27. Click on the arrow next to the Wall Node field and select Diffusion from the drop-down list.

22-44

Set the fluid submodel for the fluid in the pipe to ‘Water’ and set the material for the pipe wall to ‘Copper’.

Manifolded Coldplate Example (Continued) 28. Click on the arrow next to the Pipe Type field and select Seamless Copper Pipe ASTM B42-66 from the drop-down list. 29. Confirm that STD is displayed in the Schedule field.

Select the type of pipe and the dimensions. When these steps are completed, the drawing should appear similar to the view below.

30. Confirm that 0.125 is displayed in the Nominal Size field. 31. Select OK to close the Ripe Edit Form dialog box.

Figure 22-17

New Pipe

22-45

Manifolded Coldplate Example (Continued) or Thermal > Preferences.

32.

The Preferences dialog box appears.

In preparation for the selection operations that will follow, turn off visibility of unnecessary items, in this case the TD/ RC nodes, paths and ties.

33. Select the Graphics Visibility tab. 34. Select TD/RC Nodes to deselect it (remove the check mark from the box). 35. Select Paths to deselect it (remove the check mark from the box). 36. Select Ties to deselect it (remove the check mark from the box). 37. Select OK to close the dialog box.

Figure 22-18

After Visibility Changes

Alternatives to using the Preferences window for changing the visibility are the icons: for TD/RC Nodes for Paths for Ties

22-46

Manifolded Coldplate Example (Continued) 38.

or Draw > Polyline. The command line should now read: Specify start point:

39. Select the lump in the green circle labeled 1A. The command line should now read: Specify next point or [Arc/ Halfwidth/Length/Undo/ Width]:

40. Select the lump in the green circle labeled 2A. The command line should now read: Specify next point or [Arc/ Halfwidth/Length/Undo/ Width]:

41. Press <Enter> to terminate the polyline.

Three lateral lines between the U-shaped pipe that was just built are to be created. The lines will start and end on the lumps that have just been generated—the lumps inside the green circles. Simple lines to generate these laterals could be used, but lines have no graphical thickness, whereas the thickness of polylines can be specified and changed as needed. While this thickness has no meaning to the FloCAD model, it is preferred by some s as a visualization tool: perhaps choosing a width equal to the pipe OD. In this case, the polylines are thinner than the OD, but thicker than a plain AutoCAD line would be, and hence easier to select with the mouse. The first line will be created in these steps. The second and third lines will be copied from the first line.

Figure 22-19

First Lateral Line

22-47

Manifolded Coldplate Example (Continued) 42. Select the new line to highlight it. 43.

or Thermal > Fluid Modeling > Pipe. The command line should now read: Select Line Entity for Pipe Cross Section Shape Definition:

44. Press <Enter>. The Ripe Edit Form dialog box appears with the Pipe Attributes tab displayed. 45. Select the Pipe Attributes tab if it is not already selected. 46. Click on the arrow next to the Fluid Submodel field and select Water from the drop-down list. 47. Click on the arrow next to the Cond. Submodel field and select WALL from the drop-down list. 48. Click on the arrow next to the Outer Material Properties field and select Copper from the drop-down list. 49. Click on the arrow next to the Wall Node field and select Diffusion from the drop-down list. 50. Click on the arrow next to the Pipe Type field and select Seamless Copper Pipe ASTM B42-66 from the drop-down list. 51. Click on the arrow next to the Schedule field and select XS from the dropdown list. 52. Confirm that 0.125 is displayed in the Nominal Size field.

22-48

Convert the first lateral line into a FloCAD Pipe using the same dimensions etc. as the previous line EXCEPT that a thicker wall pipe (different schedule) is used. That way, the laterals being created will have the same outer diameter as the manifold lines, but less flow area (smaller inner diameter) for more even flow distribution.

Manifolded Coldplate Example (Continued) 53. Select the Pipe Selection tab. 54. Confirm that Fluid Pipe with wall is selected (the default) on the Pipe Selection tab.

Choose the pipe option, set the number of subdivisions and the node submodel.

55. Select the Subdivision tab. 56. Highlight the current default value of 4 in the Pipe Length Equal field and type 15. 57. Select the Node Numbering tab. 58. In the Both Sides and End Nodes fields, click on the arrows next to the Submodel fields and select WALL from the drop-down lists. 59. Select OK to close the Ripe Edit Form dialog box.

22-49

Manifolded Coldplate Example (Continued) 60. Select the vertical pipe just created to highlight it. Note: Make sure the entire pipe is selected, and not just a single lump etc., For best results, pick the pipe ID (“2”) near the top of the pipe. The “2” may be hard to see. or Modify > Copy.

61.

Note: The COPY command has two modes, Single and Multiple. After the command is issued, the command area will read:

Copy the first pipe to create the other two identical laterals (vertical pipes). The result should look like similar to the view below.

Pick on the 2 to select the pipe

Current Settings: Copy mode = current Specify base point or [Displacement/mOde/Multiple]:

62. If Copy Mode = Single and Multiple is an option, type m into the Command line. Otherwise proceed to the next step. The command line should now read: Specify base point:

63. Select the lump in the 1A green circle. The command line should now read: Specify second point of displacement, or [Exit/Undo] <Exit>:

64. Select the lump in the 1B green circle. The command line should now read: Specify second point of displacement, or [Exit/Undo] <Exit>:

65. Select the lump in the 1C green circle. 66. Press <Enter>.

22-50

Figure 22-20

Additional Lateral Lines

Manifolded Coldplate Example (Continued) 67. Draw a box around all the items in the view to highlight them. 68.

or Thermal > Modeling Tools > Connect Pipe. The Node Merge Options dialog box appears.

The pipes that have been created just overlap—they do not share lumps and nodes where they overlap. The pipes need to be connected so fluid can flow between them and heat can flow, too. The enlarged lumps at the six common points are evidence of successful connection:

69. Select OK to close the dialog box without making any changes. The Pipe Merge Tolerance dialog box appears. Figure 22-21

Connected Pipes

70. Select OK to close the dialog box without making any changes. 71.

or Format > Layer. The Layer Properties Manager dialog box appears.

72. Select the Freeze icon on the Construction layer to turn the layer off. 73. Close the Layer Properties Manager dialog box.

22-51

Manifolded Coldplate Example (Continued) or Thermal > Preferences.

74.

The Preferences dialog box appears. 75. Select the Graphics Visibility tab if not already displayed. 76. Click on TD/RC Nodes to select it (display a check mark). 77. Press OK. 78. Draw a box around all the items in the view. 79.

or Thermal > Modeling Tools > Resequence ID’s. The Resequence Node IDs dialog box appears.

80. Highlight the current value in the Starting node number field and type 1, if a different value is displayed. 81. Select OK.

22-52

Turn node visibility back on. After the nodes are turned back on, both the nodes and the lumps will be resequenced to make sure they are all unique. An alternatives to using the Preferences window for changing the visibility is the icon: for TD/RC Nodes

Manifolded Coldplate Example (Continued) 82. Draw a box around all the items in the view. 83.

or Thermal > Modeling Tools > Resequence Fluid ID’s. The Resequence Fluid Network IDs dialog box appears.

84. Keep 1 in the Starting lump number field. 85. Select OK.

22-53

Manifolded Coldplate Example (Continued) 86. Draw a box around all the items in the view. or Thermal > Edit.

87.

The Object Selection Filter dialog box appears. 88. Select Nodes[97] in the Select Type to filter field if not already selected. 89. Select OK. The Node - Multi Edit Mode dialog box appears. 90. Highlight the current value in the Initial temp field and type 80.

91. Select OK. A Thermal Desktop/AutoCAD dialog box appears confirming the change. 92. Confirm changes. 93.

22-54

toggles the node visibility Off.

Now make sure the nodes’ temperatures are all 80°F, then turn their visibility back off.

Manifolded Coldplate Example (Continued) 94. Draw a box around all the items in the view 95.

Similarly, change the lumps’ initial temperatures to 80°F, and their initial pressure to 30 psia.


96. Select Lumps[97] in the Select Type to filter field if not already selected. 97. Select OK. The Lump Edit Form dialog box appears. 98. Highlight the current value in the Temperature field and type 80. 99. Highlight the current value in the Pressure field and type 30.

100.Select OK. A Thermal Desktop/AutoCAD dialog box appears confirming the change. 101.Confirm changes.

22-55

Manifolded Coldplate Example (Continued) 102.Select the inlet lump (upper left corner). Note: Position the cursor until the tool tip shows the lump and not the pipe. 103.

or Thermal > Edit.

The Lump Edit Form dialog box appears. 104.Highlight the current value in the Id field and type 999. 105.Double click in the Heatload field. The Expression Editor dialog box appears.

Change the inlet lump to be a tank representing the volume of the entire loop. Note: That is the lump that was in the red circle marked “1” in the upper left corner of the drawing area. The lumps in the pipes themselves have been left as the default volumeless (instantaneous) junctions. This efficient modeling decision is based on the speed with which fluid will be moved through the pipe network: the event time is much longer than the time for a particle to move through the loop, so tracking fluid itself (using tanks instead of junctions) is not warranted. Note: Once this tutorial is completed, go back and switch the junctions to tanks in all pipes and see if it makes any difference in the results. Such modeling choices (tanks vs. junctions, tubes vs. STubes, arithmetic nodes vs. diffusion nodes, etc.) must be revisited in each case. The dissipation power will also be applied to this inlet tank, and its temperature will be monitored to determine when the limit has been hit. The logic to terminate the run has already been added to FLOGIC 2 of the fluid submodel WATER, but this tank must be named 999 to match that preexisting logic. As long as the tank ID and the logic match, any ID could have been used. The logic to terminate the run based on a temperature must be performed in FLOGIC 2 since FLOGIC 2 is executed after the time step, or steady state solution, is complete.

22-56

Manifolded Coldplate Example (Continued) 106.Click on the arrow next to the Energy field and select J (joules) from the drop-down list.

Set the units for the expression

107.Right-click in the Expression field and select general > Power from the drop-down list. Note: Power is a predefined symbol. Power is displayed in the Expression field. 108.Select OK to close the Expression Editor dialog box and return to the Lump Edit Form dialog box. The Heatload field now displays 0.0947817 in bold type. The use of bold type signifies an underlying expression.

22-57

Manifolded Coldplate Example (Continued) 109.Select Override calculations by pipe near the bottom of the Lump Edit Form dialog box (display a check mark in the checkbox).

Note that the symbol for Lump 999 has changed from a circle to a square.

The Type field is activated. 110.Click on Tank in the now activated Type field to select it (display a dot in the circle). The Volume field activates. 111.Highlight the current value in the Volume field and type 20. 112.Press OK. The inlet lump (ID 999) has been defined as a tank. 113.Draw a box around all the items in the view.

Turn the numbers on for the lumps so that they will be easier to pick in a later step.

114.Select Thermal > Modeling Tools > Turn Numbers On.

Figure 22-22

22-58

Numbers Turned On



The Preferences dialog box appears.

Paths will be added to complete the fluid circuit. Make sure they can be seen before proceeding!

116.Select the Graphics Visibility tab if not already displayed. 117.Click on Paths to select it (display a check mark). 118.Press OK.

Figure 22-23

Paths Visible

22-59

Manifolded Coldplate Example (Continued)

Enter location of lump:

A loss element representing the rest of the fluid loop that will not be explicitly modeled must be added. A pump is also needed. Both such paths cannot be placed until there are lumps to which to connect them.

120.Create a new lump at the left of the drawing area, ideally between the endpoints of the U-shaped pipe (shown as 1st Lump in the graphic to the right).

A reference pressure for the closed loop, which will be a plenum connected to the loop by a short tube (STube), is also needed.

121.Press <Enter> (to repeat the last command).

The commands at the left add these two lumps and make one of them a plenum.

119. or Thermal > Fluid Modeling > Lump. The command line should now read:

The command line should now read: Enter location of lump:

122.Place the second lump to the left of the first lump (shown as 2nd Lump in the graphic to the right).

The diagram should look similar to the view below, but the exact location of the new lumps is not important.

1st Lump

2nd Lump

Figure 22-24

22-60

New Lumps

Manifolded Coldplate Example (Continued) 123.Select the newest lump on the far left. 124.

Make the lump on the far left a plenum.

or Thermal > Edit.

The Lump Edit Form dialog box appears. 125.Enter 80 in the Temperature field. 126.Enter 30 in the Pressure field. 127.Click on the Plenum radio button in the Type region to select it (display a dot in the circle).

Figure 22-25

New Plenum

128.Select OK. The lump’s symbol changes.

22-61

Manifolded Coldplate Example (Continued) 129. or Thermal > Fluid Modeling > Loss. The command line should now read:

Add a K-factor loss representing the rest of the system that is not explicitly modeled. The K-factor and corresponding flow area have been pre-defined as symbols.

Select from lump:

130.Click on the outlet of the Ushaped pipe (shown as 1st Lump on the graphic to the right).

2nd Lump

Note: It may be easier to click on the number of the lump to select it. The command line should now read:

1st Lump

Select to lump:

131.Click on the new junction (shown as 2nd Lump).

Figure 22-26

Selection Points

The command line should now read: Select Entity for Area Calculation (Enter for Specified Area):

After the loss has been created, the screen should appear similar to the view below.

132.Press <Enter>.

Figure 22-27

22-62

After Loss Created

Manifolded Coldplate Example (Continued) 133.Select the new path to highlight it. 134.

or Thermal > Edit.

The Loss dialog box appears.

135.Double click in the Loss Coefficient[FK] field. The Expression Editor dialog box appears. 136.Right-click in the Expression field and select general > Ksystem (a predefined symbol) from the dropdown list. Ksystem is displayed in the Expression field.

22-63

Manifolded Coldplate Example (Continued) 137.Select OK to close the Expression Editor dialog box. The Loss dialog box Loss Coefficient field now displays 500 in bold type. The use of bold type signifies an underlying expression. 138.Select the Flow Area tab. 139.Double click in the input field located below the Flow Area (AF) field. The Expression Editor dialog box appears. 140.Right-click in the Expression field and select general > Farea (a predefined symbol) from the drop-down list. Farea is displayed in the Expression field. 141.Select OK to close the Expression Editor dialog box. The Loss dialog box Specified field now displays 0.056 in bold type. 142.Select OK to close the Loss dialog box.

22-64

Manifolded Coldplate Example (Continued) 143. or Thermal > Fluid Modeling > Tube/STube. The command line should now read: Select from lump:

144.Select the junction in the middle left: the one at the outlet of the new Loss element. (See the point labeled 1st in the graphic to the right.) The command line should now read: Select to lump:

145.Select the plenum (triangular or tetrahedral icon) near it. (See the point labeled 2nd in the graphic to the right.)

The junction at the outlet of the new Loss element will become the inlet of the pump, which has yet to be defined. Model an ideal accumulator at this point: specify a pressure without specifying a temperature. To do this, create a connection between the plenum and the loop using an STube. The steady flowrate through this STube is zero, and transiently it will be a bit above zero only because of thermal expansions of the fluid in the loop. So the size of this STube is largely irrelevant: it only serves to make sure the plenum and the loop share a common pressure. The diagram should now look something like this.

The command line should now read: Select Upsteam Entity for Area Calculation (Enter for Specified Area):

1st

146.Press <Enter>. 147.Select the new path to highlight it. 148.

or Thermal > Edit.

2nd

The (S)Tube dialog box appears. Figure 22-28

STube

22-65

Manifolded Coldplate Example (Continued) 149.Select the Flow Area tab if it is not already selected. 150.Highlight the current value in the Specified Hydraulic Diameter

field and type 0.2. 151.Click on Hydraulic Diameter Only (DH) to select it. Note: This results in the STube being circular. 152.Select the (S)Tube Data tab. 153.Click on Input in the Length region to select it (place a dot in the radio button). 154.Highlight the current value in the newly activated Input field and type 12. Note: When Input is selected, Distance between lumps is deselected. 155.Select OK in the (S)Tube dialog box.

Specifying only the hydraulic diameter means that the path has a circular cross section. Close the loop with a pump.

156. or Thermal > Fluid Modeling > Pump/Fan. The command line should now read: Select from lump:

157.Click on the junction at the outlet of the loss element (labeled 1st on the graphic at the right).

2nd #999

1st

The command line should now read: Select to lump:

158.Click on the tank at the inlet to the U-shaped pipe (ID 999, labeled 2nd on the graphic at the right).

22-66

Figure 22-29

View After New Pump

Manifolded Coldplate Example (Continued) 159.Select the new path to highlight it. 160.

or Thermal > Edit.

The Fan/Pump dialog box appears. 161.On the Head Vs Flow tab, select Simple Fan/Pump curve (place a dot in the radio button). 162.Click on the arrow next to Head Units and select inches from the drop-down list (change from the default value of meter). 163.Click on the arrow next to Flow Units and select in^3/sec (inches^3/sec) from the drop-down list (change from the default value of meter^3/sec).

22-67

Manifolded Coldplate Example (Continued) The Tabular Input dialog box is used to The Tabular Input dialog box appears. specify the pump flowrate-head curve, in units of volumetric flow versus length. 165.Type the following in the Enter flow The finished table should look like this input field: (saving and reopening reformats the input values): 0, 216 0.048, 192 0.096, 168 0.13, 144 0.144, 126 0.163, 102 0.192, 48 0.221, 0 164.Click on the Set button.

166.Select OK to close the Tabular Input dialog box. 167.Select OK in the Fan/Pump edit dialog box. Note: If familiar with PC cutting and pasting functionality, it is possible to cut and paste the entry information from the pdf version of the ’s Manual. 168. or Thermal > Modeling Tools > Toggle Selection Filter. The Command area should now show: Command: _RCFilter Thermal Desktop filter turned on

22-68

Turn the filter on to make it easier to pick specific entities. Note: Make sure the filter has been turned on.

Manifolded Coldplate Example (Continued) 169.Draw a box around all the items in the view.

Turn off the numbers of all the lumps.

170.Select Thermal > Modeling Tools > Turn Numbers Off. The Object Selection Filter dialog box appears. 171.Select Lumps[99] in the Select Type to filter field if not already selected. 172.Select OK

Figure 22-30

Numbers Turned Off

Note: In addition to the view on the screen, press to view the text window as another way to confirm the numbers have been turned off 173. or Thermal > Modeling Tools > Toggle Selection Filter.

Note: Turn it off so it does not interfere with later selections.

The Command area should now show: Command: _RCFilter Thermal Desktop filter turned off

Turn off path visibility once more. 174.


The Preferences dialog box appears. 175.Select the Graphics Visibility tab if not already displayed. 176.Click on Paths to deselect it (remove the check mark from the box). 177.Select OK. Figure 22-31

Path Visibility Turned Off

An alternative method for toggling the visibility of paths is to select the icon:

22-69


or Format > Layer.

The Layer Properties Manager dialog box appears. 179.Select the On and Freeze symbols on the Board layer, turning them on.

The fluid loop is complete, but now the coolant pipes need to be attached to the plate: to model the saddle, bond, or milled seat as a or. First, make the board visible. Only the surface next to the pipes is needed for now.

180.Close the Layer Properties Manager dialog box.

Figure 22-32

22-70

Board Visibility On

Manifolded Coldplate Example (Continued) 181.Click on each of the 4 pipes to select them. Note: It is not necessary to use the <Shift> key for the multiple seletion. Note: Zooming in on the tops of the 3 cross pipes may make selection easier.

A or is used to make a thermal connection from the pipe to the plate. If objects are selected before the or command is given, the selected objects are used as the From objects. Otherwise the is queried for the From objects before selection of the To objects can be made.

182. or Thermal > FD/FEM Network > or. The command line should now read: Select surfaces ing to:

183.Click on the plate to select it. The command line should now read: Select surfaces ing to:

Note: The plate is now the “to” object. 184.Press <Enter>. The or dialog box appears.

22-71

Manifolded Coldplate Example (Continued) 185.Highlight the current value in the or Submodel field and select WALL from the drop-down list. 186.Highlight the current value in the From field and select EDGES from the drop-down list.

187.Double click in the Conduction Coefficient field. The Expression Editor dialog box appears. 188.Click on the arrow next to the Length field and select ft from the drop-down list (change from default value of inches). 189.Click on the arrow next to the Time field and select hr from the drop-down list (change from default of seconds).

22-72

These steps use a symbol to describe the linear conductance along the length of each pipe. Its units are not those of the default model, however, so they must be specified so Thermal Desktop can convert it automatically.

Manifolded Coldplate Example (Continued) 190.Right-click in the Expression field and select general > Pipe2Plt from the drop-down list. Pipe2Plt is displayed in the Expression field.

Note: The symbol Pipe2Plt was predefined by the author. 191.Select OK to close the Expression Editor dialog box. The or dialog box Conduction Coefficient field now displays 0.00138889 in bold type. 192.Click on the Insert button. 193.Highlight the current value in the Tolerance field and type 0.25. 194.Highlight the 4 pipes in the From field (bottom left) and then press the Edit button below the From field:

A tolerance is used to keep the parts of the U-shaped pipe that overhang the plate from connecting to the plate.

The Select Edges - Multi Edit Mode dialog box appears.

195.Click on Along Pipe Length to select it. 196.Click on Inlet Edge to deselect it.

22-73

Manifolded Coldplate Example (Continued) 197.Select OK to close the Select Edges - Multi Edit Mode dialog box.

The final drawing should look similar to the view below:

The pipes in the From field reflect the change. 198.Select OK to close the or dialog box. Small green icons at the beginning of each of the 4 pipes and a small gold cross in the center of the plate appear representing the or.

Figure 22-33

After New Conductor

Prepare for postprocessing by: 199. to toggle the visibility of the ors. The command area should now show: Command: _rcToggleorVis ors are now off

Note: If the command area shows that ors are now on, select the icon one more time.

22-74

•

turning off visibility of the ors

•

turning off visibility of the surfaces on the both sides of the plate

•

then turning back on the FEM model of the plate itself.

The piping model has been built on the back side of the aluminum plate.

Manifolded Coldplate Example (Continued) 200. or Thermal > Model Browser. The Model Browser window appears. 201.Using the Model Browser menu, click on List to confirm that Surfaces/Solids is selected (check mark).

The Model Browser is a key tool for navigating through a Thermal Desktop model: finding, editing, changing visibility, deleting, etc. It is a good idea to practice with the Model Browser. The drawing should now appear similar to the view below:

202.Expand the CC Not Generated tree by clicking on the + sign. Two rectangles appear in the list. 203.Highlight both rectangles within the Model Browser, holding down the <Shift> key to select the second rectangle. 204. on the Model Browser toolFigure 22-34 bar or select Display > Turn Visibility Off using the Model Browser menu.

Model Browser Visibility

205.Select PLATE in the list; do not expand this tree. 206. on the Model Browser toolbar or select Display > Turn Visibility On using the Model Browser menu. 207.Close the Model Browser.

22-75



The Preferences dialog box appears. 209.Select the Units tab. 210.Select ENG in the Output Units for FLUINT Models field (display a dot in the radio button). 211.Select OK.

22-76

Set SINDA/FLUINT to use its standard US Customary (English) units. Note: SINDA/FLUINT units need only be defined when a fluid submodel is used.

Manifolded Coldplate Example (Continued) 212.Select Thermal > Fluid Modeling > Submodel Manager. The FLUINT Submodel Manager dialog box appears.

Specify the -defined water-glycol file as the working fluid. This file already exists in the drawing directory, but the fluid must be named consistently (fluid ID “9050”) and point to this file.

213.Select Properties. The Fluid Submodel Properties dialog box appears.

214.Select (A) Water to highlight it if it is not already highlighted. 215.Select Edit.

22-77

Manifolded Coldplate Example (Continued) The Fluid Submodel Fluid Property List Edit dialog box appears.

216.Click in the field that reads Water and select Browse for New Fluid Property File in the dropdown menu. The Open dialog box appears. 217.Select waterglycol.inc and press the Open button. Note: Waterglycol.inc should now appear in the File Name field. 218.Select OK to close the Fluid Submodel Fluid Property List Edit dialog box. 219.Select OK to close the Fluid Submodel Properties dialog box. 220.Select OK to close the FLUINT Submodel Manager Form dialog box.

22-78

Manifolded Coldplate Example (Continued) 221. or Thermal > Case Set Manager. The Case Set Manager dialog box appears. 222.Select Run 1 Selected Case. The Duplicate Nodes Found dialog box appears.

It is time to launch the transient analysis, which has been set up already in the Case Set Manager. Note: If not familiar with this other key part of Thermal Desktop, take the time to browse. For example, review the logic that has been added to the FLOGIC2 Water submodel (found on the SINDA tab) and the values entered in the Calculations tab in Solution Type. Note: To learn more about why this logic is placed in FLOGIC2, see Section 4.1 of the SINDA/FLUINT ’s Manual.

223.Select Allow The Duplicate Ids To Remain In The Model. Note: that only PLATE.9999 is duplicated.

node

Sinda/Fluint Run Status dialog box

appears confirming the successful completion of the process and the drawing area graphics update. 224.Click OK to close the dialog box. Note: If problems are encountered during the run and the problems cannot be diagnosed, consult with the instructor if completing this exercise in class, or CRTech for help.

22-79

Manifolded Coldplate Example (Continued) 225. or Thermal > Post Processing> Edit Current Dataset. The Set Sinda Dataset Properties dialog box appears. 226.Scroll to the bottom of the Select Time/Record [sec] field to move to the end of the SAVE file and highlight the last point—the end of the run at 323 seconds. 227.Select Smart Color Bar Cycling to place a check mark in the box. 228.Select OK to close the dialog box. 229.Rotate the model to achieve a view similar to that seen in the image to the right. 230.Select or Thermal > Post Processing > Cycle Color Bars to cycle through color bars until the Node color bar is shown.

22-80

The answer to the design question is 323 seconds—that is how long before the dissipation point reaches 110°F. The resulting colored plot, however, shows graphically how well the sensible heat in the plate was utilized. The model can be rotated to match the view. Smart Color Bar Cycling will turn off the visibility of items not associated with the current color bar. As you cycle through the color bars, notice how the items’ visibilities turn on and off.

Manifolded Coldplate Example (Continued) 231.Select or Thermal > Post Processing > Cycle Color Bars to cycle through color bars until the Lump color bar is shown 232.Select the tank at the pipe inlet and the junction at the pipe outlet (loss inlet) to highlight them. Note: It is not necessary to hold down the <Shift> key to select the two objects. Also, to use the tool tips to insure the correct objects are being selected. 233.

The actual temperatures themselves can be shown on the screen as text, too. Instead of element IDs displayed in the drawing mode, in the postprocessing mode the post-processed values (temperature, in this case) are printed.

Turn Numbers On icon.

Note: The Turn Numbers On icon is located on a tool bar in the lower right area of Thermal Desktop (unless moved by the ). If the Object Selection Filter dialog Figure 22-35 box appears, highlight Lumps(2) if not already highlighted and click OK.

Temperature View

234.Select or Thermal > Post Processing > Cycle Color Bars to cycle through color bars until the Node color bar is shown

22-81

Manifolded Coldplate Example (Continued) 235.Select the tank at the pipe inlet and the junction at the pipe outlet (loss inlet) to highlight them once more. Note: It is easier to select the two lumps now by picking their labels (their temperatures). 236. or Thermal > PostProcessing > X-Y Plot Data vs. Time. 237.Return to the Thermal Desktop window when finished with the plot.

22-82

Call for a EZ-XY® plot versus time. The X-Y Plot Data vs. Time command can also be given though the Model Browser. Select the desired objects in the Model Browser and choose the Plot icon (the same as shown to the left) in the Model Browser toolbar.

Manifolded Coldplate Example (Continued) 238. or Thermal > Post Processing> Cycle Color Bars until the Path color bar is visible and paths are in the viewport.

Look at the flowrate distributions. Use Smart Color Bar Cycling to turn off the lumps and the plate, and turn on the paths. Growing the color bar (see Model/Paper manipulations above), and turning off the lump temperature numbers and turning on the path flowrate numbers yields:

Figure 22-36

New Visibility & Color Bar

The first and last paths carry the most flow, despite the fact that the far end of the plate is under utilized (due to the low temperature gradients: the heat in that leg was already mostly rejected in the manifold). Caution: this model did not include wye and tee (merging and splitting) flow losses, which can be calculated by SINDA/FLUINT but require more set up. 239.Select File > Exit. A Thermal Desktop/AutoCAD dialog box appears asking to save the drawing changes.


240.Select Yes.

22-83

22.3.1

For the advanced :

Assume that the pipe-to-plate conductance cannot be increased (i.e., the saddle design, the braze, or the press-fit into the ball end milled channel). Due to the low pressure drops in the coldplate, perhaps a one- serpentined pipe would have been a better choice. Try replacing the manifold with a single serpentined pipe—use the AutoCAD polyline again with arcs and lines, or build separate arcs and lines with common endpoints. Also try different pipe sizes and lay-outs of the serpentine. Note: Once a pipe and or are defined, just move around the pipe using the AutoCAD grip points and rerun the analysis. Since losses (other than frictional) are not implicit, the advanced may wish to see how losses at the Tees would affect the flow by using the FK Calculator in the path edit form for the paths near the Tees.

22-84

22.4

Drawn Shape Heat Pipe

What will be learned: • Creating a FloCAD heat pipe with a defined outer shape • Manipulating pipe orientation In this example, a simple heat pipe utilizing a -drawn outer shape will be constructed. The heat pipe will be 1 inch x 1/2 inch x 0.05 inches. There will be a heated plate below and a cooled plate above the heat pipe. The plates and the heat pipe will be connected via ors. The ambient is 68°F air with a convection coefficient of 30 BTU/hr-ft2-F (stored as parametric symbol EnviroU). 100W (symbol Power) is dissipated into the base plate.

Drawn Shape Heat Pipe Example 1. Double click on the file drawnShape.dwg located in the Tutorials\Thermal Desktop\drawnShape folder. Thermal Desktop opens with the drawnShape drawing on the screen (graphics area will be blank except for the coordinate system and some points displayed as crosses). A model of the plate and its convection environment have already been generated (but are not currently visible), along with some symbols that can be used to generate a parametric model. The units of the initial drawing are BTU, inches, seconds, °F. The drawing file includes some reference points that will be used for drawing a shape.

22-85

Drawn Shape Heat Pipe Example (Continued) or Draw > Polyline.

2.

The command line should now read: Specify start point:

3. Type 0.25,-0.25 into the Command line. The command line should now read: Specify next point or [Arc/ Halfwidth/Length/Undo/ Width]:

4. Type 0.25,0.75 into the Command line. The command line should now read: Specify next point or [Arc/ Halfwidth/Length/Undo/ Width]:

5. Press <Enter> to terminate the polyline. 6.

or double-click the middle mouse button to zoom extents

22-86

Draw an AutoCAD polyline representing the centerline of the pipe. A simple line will be used for the centerline. Use the coordinate to create the lines of the correct size. The drawing should look similar to the view below: If the final line is not correct, delete it and try again.

Drawn Shape Heat Pipe Example (Continued) 7.

or Draw > Polyline. The command line should now read: Specify start point:

8. Select the point (cross) on the far left. The command area should now read: Specify next point or [Arc/ Halfwidth/Length/Undo/ Width]:

9. Type a to start the arc drawing mode. The command area should now read: Specify endpoint of arc (hold Ctrl to switch direction) or [Angle/CEnter/ Direction/Halfwidth/Line/ Radius/Second pt/Undo/ Width]:

Next, draw an AutoCAD polyline representing the inner surface of the pipe. The shape can be drawn anywhere. An important feature of the shape is that the line must be closed. The last command when generating a line to be used for the drawn shape should be CLOSE (or the shorthand command, CL). Even if a point is created at the same location as the start of the line, the line should be closed. The object properties window can be used to edit a line that was not closed during the initial creation. The drawing should appear similar to the view below. If the shape line is not correct, delete it and try again.

10. Select the point just above or below the starting point to form a 90-degree arc (noting the final shape shown in the image to the right). The command line should now read: Specify endpoint of arc (hold Ctrl to switch direction) or [Angle/CEnter/ CLose/Direction/Halfwidth/ Line/Radius/Second pt/Undo/ Width]:

11. Type the letter l or line to start the line drawing mode. The command area should now read: Specify next point or [Arc/ Close/Halfwidth/Length/Undo/ Width]:

12. Select the point horizontally across from the previous point selected. The command area should now read: Specify next point or [Arc/ Close/Halfwidth/Length/Undo/ Width]:

22-87

Drawn Shape Heat Pipe Example (Continued) 13. Type a to start the arc drawing mode. The command area should now read: Specify endpoint of arc (hold Ctrl to switch direction) or [Angle/CEnter/ Direction/Halfwidth/Line/ Radius/Second pt/Undo/ Width]:

14. Select the point below or above the previous point selected to form a 180degree arc. The command area should now read: Specify endpoint of arc (hold Ctrl to switch direction) or [Angle/CEnter/ Direction/Halfwidth/Line/ Radius/Second pt/Undo/ Width]:

15. Type the letter l or line to start the line drawing mode. The command line should now read: Specify next point or [Arc/ Halfwidth/Length/Undo/ Width]:

16. Select the point horizontally to the left of the previous point. The command line should now read: Specify next point or [Arc/ Halfwidth/Length/Undo/ Width]:

17. Type a to start the arc drawing mode. The command area should now read: Specify endpoint of arc (hold Ctrl to switch direction) or [Angle/CEnter/ Direction/Halfwidth/Line/ Radius/Second pt/Undo/ Width]:

22-88

Drawn Shape Heat Pipe Example (Continued) 18. Select the first point of the loop. The command area should now read: Specify endpoint of arc (hold Ctrl to switch direction) or [Angle/CEnter/ Direction/Halfwidth/Line/ Radius/Second pt/Undo/ Width]:

19. Type CL to close the loop and end the creation. 20.

or double-click the middle mouse button to zoom extents

22-89

Drawn Shape Heat Pipe Example (Continued) 21. Select the straight line to highlight it.

Convert the lines into a pipe segment, name its components, specify its materials, dimensions, resolution, etc.

or Thermal > Fluid Modeling > Pipe.

If the centerline had not been selected first, the command line would have prompted to selected a line entity for the pipe centerline.

22.

The command line should now read: Select Line Entity for Pipe Cross Section Shape Definition:

23. Select the closed loop to highlight it. The command line should now read: Select Line Entity for Pipe Cross Section Shape Definition:

24. Press <Enter>. The Ripe Edit Form dialog box appears. 25. Select the Pipe Selection tab if not already displayed. 26. Click on Heat Pipe to select it (place a dot in the radio button). Note: notice when Heat Pipe is selected, the Pipe Attributes and Ties tabs disappear and a Heat Pipe Data tab is displayed. 27. Select the Subdivision tab. 28. Highlight the current default value of 1 in the Pipe Circumf Equal field and type 16. 29. Highlight the current default value of 4 in the Pipe Length Equal field and type 16. 30. Select the Node Numbering tab.

22-90

Drawn Shape Heat Pipe Example (Continued) 31. Highlight the current value in the Both Sides Submodel field and type CHAMBER.

When complete, the drawing should appear as:

The Vapor Nodes Submodel field automatically updates to CHAMBER. 32. Highlight the current value in the Vapor Node ID field and type 1000. 33. Select the Heat Pipe Data tab. A Thermal Desktop/AutoCAD dialog box appears asking confirmation to add CHAMBER as a Submodel.

34. Select Yes. The dialog box displays the Heat Pipe Data tab.

If the drawing is examined it will be found that a copy of the closed line has been placed at the start of the pipe. The original shape is not associated with the pipe and can be deleted if desired. Notice that the plane of the shape has been changed to be perpendicular to the start of the pipe.

35. Select Diffusion in the Wall Node Type region (display a dot in the radio button). 36. Click on the arrow next to the Wall Material field and select Copper from the drop-down list. 37. Click on the arrow next to the Wall Input field and select Thickness Input from the drop-down list. 38. Highlight the current value in the Wall Thickness field and type 0.01. 39. Select OK to close the Ripe Edit Form dialog box. 40. Box-select the original, closedloop polyline. 41.

or Modify > Erase.

Delete the original polyline. Since the original polyline was copied, it can be deleted. The polyline referenced for the centerline and the polyline at the start of the pipe cannot be deleted.

22-91

Drawn Shape Heat Pipe Example (Continued) 42. Select View > 3D Views > Front.

The initial orientation of the pipe will be a function of where the centerline was drawn. Switch the view to look at the heat pipe from the front. The view should then look similar to the view below:

43. Select the polyline drawn as the inner surface of the heat pipe.

When working with pipes it is important to understand that they are made up of several semi-independent pieces.

Note: When selected, only that inner line should become highlighted to show it is selected. 44.

or type rotate3d in the command line. The command line should now read: Specify first point on axis or define axis by [Object/ Last/View/Yaxis/Zaxis/ 2points]:

45. Type Yaxis in the Command line to begin that mode. The command line should now read: Specify a point on the Y axis <0,0,0>:

46. Select the node in the center of the heat pipe (should be 0.25, 0.25, 0.0 coordinates). The command line should now read: Specify rotation angle or [Reference]:

47. Type 90 in the Command line as the rotation angle.

22-92

•

The lines defining the centerline are a grouphat make up one piece

•

The line defining the shape is another piece

•

The pipe representation is a third piece

When selected, the heat pipe will have a grip point for the starting angle. This grip point can be used to easily orient the cross-section of the pipe. In this exercise, the ROTATE3D command will be used.

Drawn Shape Heat Pipe Example (Continued) 48. Select the heat pipe. Note: It is easiest to pick on the Pipe.1 label.

The previous commands work in the 3D world coordinate system. It is also possible to work in a 2D UCS.

49. Select Thermal > Modeling Tools > Align UCS to Surface.

First set the view to the object UCS.

50. Select the heat pipe and click on the Start Angle grip point (located at the top of the heat pipe).

Now rotate the heat pipe back to the horizontal position.

The command line should now read: Specify stretch point or [Base point/Copy/Undo/eXit]:

51. Type 0,0 Generate some to the heat pipe. 52.


Turn on the layer with the surfaces already generated. The model should now look similar to the following:

53. Select the On and Thaw icons on the Boundary layer, turning them on. 54. Close the Layer Properties Manager dialog box. 55.

or: • Type Zoom (or the letter Z) in the Command line. • Type Extents (or the letter E) in the Command line This centers the model.

22-93

Drawn Shape Heat Pipe Example (Continued) Now connect the base to the heat pipe. 56.

or Thermal > FD/FEM Network > or. The Command Line now reads: Select faces ing from:

57. Select the bottom plate as “from” surface. The Command Line now reads: Select faces ing from:

Press <Enter>. The Command Line now reads: Select surfaces ing to:

58. Select the heat pipe as “to” surface. The Command Line now reads: Select surfaces ing to:

59. Press <Enter>. The or dialog box appears. 60. Click on the arrow next to the Conductor Submodel field and select Base from the drop-down list. 61. Insure Faces is selected in the Contract From field. 62. Highlight the current value in the Conduction Coefficient field and type 0.01.

22-94

The tolerance used will insure that the base only s the lower surface of the heat pipe since the distance will only go from the surface of the base to the centerline of the heat pipe.

Drawn Shape Heat Pipe Example (Continued) 63. Click on Use Material to select it (place a check mark in the box). The Use Material input field activates. 64. Select ThermalGrease from the Use Material drop-down list. 65. Highlight the current value in the Tolerance field and type 0.04. 66. Press OK to close the or dialog box.

22-95

Drawn Shape Heat Pipe Example (Continued) 67.

or Thermal > FD/FEM Network > or. The Command Line now reads: Select faces ing from:

68. Select the top plate as “from” surface. The Command Line now reads: Select faces ing from:

Press <Enter>. The Command Line now reads: Select surfaces ing to:

69. Select the heat pipe as “to” surface. The Command Line now reads: Select surfaces ing to:

70. Press <Enter>. The or dialog box appears. 71. Click on the arrow next to the Conductor Submodel field and select Top from the drop-down list. 72. Insure Faces is selected in the Contract From field. 73. Highlight the current value in the Conduction Coefficient field and type 0.01. 74. Click on Use Material to select it (place a check mark in the box). The Use Material input field activates. 75. Select ThermalGrease from the Use Material drop-down list.

22-96

Repeat the same process for the top plate. This time changing the surface of the plate to be the bottom.

Drawn Shape Heat Pipe Example (Continued) 76. Highlight the current value in the Tolerance field and type 0.04. 77. Double click on the Rectangle in the From field. The Select Faces dialog box appears. 78. Click on Top to deselect it (remove the check mark). 79. Click on Bottom to select it (display a check mark). 80. Select OK to close the Select Faces dialog box. 81. Press OK to close the or dialog box. 82. Select Thermal > Model Checks > Show or Markers. The Command line should now read: Select ors to display markers for:

To see where the or will connect the top plate to the pipe use or markers. The drawing should now look similar to the view below:

83. Select the upper or, a green arrow from the upper plate. The Command line should now read: Select ors to display markers for:

84. Press <Enter>. Multiple yellow lines representing the or appear. 85.

or select Thermal > Model Checks > Clear or Markers.


This command clears the points from the screen. Leaving the or or markers on the screen can slow down the graphics update and obscure post-processing. Exit Thermal Desktop and save as prompted.

87. Select Yes.

22-97

22-98

22.5

FEM Walled Pipe

What will be learned: • creating an FloCAD pipe with a wall made up of existing surfaces • Using natural convection correlations • Visualize flow areas computed by pipes • postprocessing fluid models using the Model Browser In this example, a pipe utilizing a set of finite element surfaces will be constructed. There is hot air flowing through the duct. The duct contains a hole in the side that will be connected to another portion of the system (not performed in this model). The ambient is 20 °C air with natural convection computed using built-in correlations. 200 °C air is flowing through the duct.

FEM Walled Pipe Example A model of the pipe wall has already been generated, along with some symbols that can be used to generate a parametric model. The units of the initial drawing are Joules, meter, seconds, °C. 1. Double click on the file pipeDuctWall.dwg located in the Tutorials\Thermal Desktop\WalledPipe folder. Thermal Desktop opens with the pipeDuctWall drawing on the screen.

22-99

FEM Walled Pipe Example (Continued) 2.

to toggle visibility of TD/RC nodes off.

3.

to toggle visibility of Surfaces off.

22-100

Turn off the nodes and surfaces to make it easier to find the centerline.

FEM Walled Pipe Example (Continued) 4. Select the LWPolyline that is now visible. 5.

or Thermal > Fluid Modeling > Pipe. The Command line should now read: Select Line Entity for Pipe Cross Section Shape Definition:

6. Press <Enter> without selecting anything in order to use the surfaces for the wall.

An AutoCAD LWPolyline representing the centerline of the pipe has already been created on the Construction layer. The line was formed simply by selecting the points in the center of the wall section that were found at the intersection of construction lines from opposite corners of the wall, and extending the line vertically after the bend. The drawing should look similar to the view below when Step 11 is completed:

The Ripe Edit Form dialog box appears with the Pipe Selection tab displayed. 7. Select Fluid Pipe with Surfaces for Wall (display a dot in the radio button). The Wall Surfaces Ids field activates. 8. Click on the Add button to the right of the Wall Surfaces Ids field. The Command line should now read:

Figure 22-37

Pipe Wall

Select the surfaces for the pipe wall or [GRP]:

9. Type all in the Command line. The Command line should now read: Select the surfaces for the pipe wall or [GRP]:

10. Press <Enter>.

22-101

FEM Walled Pipe Example (Continued) The Ripe Edit dialog box reappears.

There will be messages in the command line area stating that the area calc failed three times at about distance of 10.22. This indicates a problem with the closure of the pipe wall at that distance from the beginning of the pipe. The next steps will demonstrate how to find and fix this problem.

11. Select OK to close the dialog box. to toggle visibility of Ties off.

12. 13.

to toggle visibility of TD/RC nodes on. to toggle visibility of Surfaces

14. on.

22-102

Turn off the display of the ties to prevent the screen from being too cluttered. The surfaces must be turned on to see the problem area.

FEM Walled Pipe Example (Continued) 15. Select Thermal > Modeling Tools > Show Path Area. The Command line should now read: Select pipe(s) to show area: or [GRP]:

The command RcShowPathArea will display the areas that have been calculated by FloCAD. The drawing should look similar to the view below:

16. Type all in the Command line. The Command line should now read: Select pipe(s) to show area: or [GRP]:

17. Press <Enter>.

Note in the vertical section the red area three down from the top is not a complete rectangle. This is because the wall of the pipe has a hole in the side of the pipe. We need to add a dummy surface here to close off the pipe. Change the shading to see the hole better. 18.

to change to the Thermal Visual Style

22-103


or select Thermal > Modeling Tools > Clear Path Area.

or Format > Layer.

20.

This will erase all rays shown for the path area calculations, and for radiation calculations.

Keep new objects with the rest of the wall.

The Layer Properties Manager dialog box appears. 21. Double-click the Wall layer to make it the current layer. Note: The current layer is the Pipe layer. 22. Close the Layer Properties Manager dialog box. 23.

or Thermal > Modeling Tools > Toggle Selection Filter.

Turning on the selection filter will make it easier to only display the nodes in the next step.

Note: Confirmation that the filter is on appears in the Command line area. If it turned the filter off, repeat the command to turn it back on. Change back to 2D wireframe. 24.

or select View > Visual Styles > 2D WireFrame.

22-104


or Thermal > Model Browser. The Model Browser window appears.

26. Using the Model Browser menu, select List > Submodel.Id. The list in the Model Browser updates. 27. Hold down the key and select nodes 112,126, 437 and 448 in the tree.

The hole is defined by 4 nodes in this model. The easiest way to close this particular hole is to create a surface using the four nodes at the corners of the hole. This sequence shows what the nodes are, but in reality, it is necessary to turn the node ids on, and then discern which nodes are needed to fill in the hole. After performing the command, the 4 nodes should be seen on the screen.

28. Using the Model Browser menu, select Display > Only. The Object Selection Filter dialog box appears. 29. Select Nodes(4). 30. Select OK. 31. Close the Model Browser. Note: The Model Browser window may be minimized instead of closed for use later in the tutorial. 32.

or Thermal > Modeling Tools > Toggle Selection Filter.

Turn the filter off so that it does not keep coming up when as commands are performed.

22-105


or select Thermal > FD/Fem Network > Element. The Command Line should now read: Select nodes for linear element or [GRP]:

34. Select First Node. The Command Line should now read: Select nodes for linear element or [GRP]:

35. Select Second Node. The Command Line should now read: Select nodes for linear element or [GRP]:

36. Select Third Node. The Command Line should now read: Select nodes for linear element or [GRP]:

37. Select Forth Node. The Command Line should now read: Select nodes for linear element or [GRP]:

38. Press <Enter>.

22-106

Be sure to pick the nodes by following the edge of the hole, either in a clockwise or a counter-clockwise order. Otherwise a quad element with diagonals that will cross be the result. It may be necessary to zoom in slightly to make picking the nodes easier and it also might be helpful to switch to wireframe mode.

FEM Walled Pipe Example (Continued) 39. Select the quad element. 40.

or Thermal > Edit.

In order for the area to not be associated with this surface used to create ties—or anywhere else in the model— turn off any nodes or conductors that may be created.

The Thin Shell Data dialog box appears. 41. Select the Cond/Cap tab. 42. Click on Generate Nodes and Conductors. The Expression Editor opens. 43. Type 0 into the Expression field. 44. Select OK to close the Expression Editor. Most fields on the Cond/Cap tab become inactive. 45. Select the Comment tab. 46. Type Dummy in the Comment field. 47. Select OK. 48.


49. Double-click the Pipe layer to make is the current layer. Note: The current layer is the Wall layer. 50. Close the Layer Properties Manager dialog box.

22-107


or Thermal > Model Browser. Note: Or maximize the Model Browser if minimized earlier.

52. Using the Model Browser menu, select List > Pipes. 53. Expand pipe.1 (click on the + sign). 54. Right-click pipe.1:: name and select Edit. The Ripe Edit Form dialog box appears. 55. Select the Pipe Selection tab if not already displayed. 56. Click on the Add button located to the right of the Wall Surfaces Ids field. The Command line should now read: Select the surfaces for the pipe wall or [GRP]:

Note: It may be necessary to move the Model Browser window to view the commands and to make a selection. 57. Select the quad element just created. 58. Press <Enter> to complete the selection of the surface. The Ripe Edit Form dialog box reappears. 59. Scroll down to the bottom of the Wall Surface Ids field.

22-108

Add this new element to the surfaces of the pipe. Using the Model Browser is the easiest way to be able to edit the pipe and add this one new surface. Leave the display as it is. Although it is fine to return visibility to many of the objects if desired, it is not necessary. After adding the Dummy surface to the list, edit the surface in the list so that no sides are active. The surface will then only be used for the area calculation, but will not have any heat transfer associated with it.

FEM Walled Pipe Example (Continued) 60. Double click on the Quad ElemDummy entry. The Select Faces dialog box appears. 61. Click on Top to deselect it (remove the check mark from the box). 62. Select OK to close the Select Faces dialog box. The Dummy element is highlighted in red to indicate that it is not connected with a tie. 63. Select OK to close the Ripe Edit Form dialog box. 64. Close or minimize the Model Browser. 65. Select Thermal > Modeling Tools > Show Path Area. The Command line should now read: Select pipe(s) to show area: or [GRP]:

Check to see that the error messages are no longer being received. It may be necessary to zoom out slightly to see all of the flow areas.

66. Type all in the Command line. The Command line should now read: Select pipe(s) to show area: or [GRP]:

67. Press <Enter>.

Clear the screen. 68.

or select Thermal > Modeling Tools > Clear Path Area.

22-109


or Thermal > Model Browser. Note: Or maximize the Model Browser if minimized earlier.

70. Expand pipe.1. 71. Expand pipe.1:: 72. Select all of the objects in the list except for Ties. Note: Use the key to select non-contiguous items. 73.

or Display > Turn Visibility On using the Model Browser tool bar or menu.

74. Close or minimize the Model Browser.

22-110

Turn on the visibility of the pipe and all of its components (except the ties).

FEM Walled Pipe Example (Continued) 75. Select lump 1 at the lower left Note: If having trouble selecting just the lump, zoom into the area, or draw a large box from left to right around only the first lump. 76.

Set up the model to run by creating a pressure drop across the duct to induce some flow.

or Thermal > Edit. The Lump Edit Form dialog box appears.

77. Select Override calculations by pipe. The Type region activates. 78. Click on the Plenum radio button to select it (display a dot in the circle). 79. Highlight the current value in the Pressure field and type 101400. 80. Double click in the Temperature field. The Expression Editor dialog box appears. 81. Right-click in the Expression field select general > flowtemp from the drop-down list. Flowtemp is displayed in the large Expression field. 82. Select OK to close the Expression Editor dialog box. 83. Select OK to close the Lump Edit Form dialog box. The lump changes shape to reflect the change.

22-111

FEM Walled Pipe Example (Continued) 84. Select lump 6 located in the upper right of the model. or Thermal > Edit.

85.

The Lump Edit Form dialog box appears. 86. Select Override calculations by pipe. The Type region activates. 87. Click on the Plenum radio button to select it (display a dot in the circle). 88. Select OK to close the Lump Edit Form dialog box. The lump changes shape to reflect the change. or Format > Layer.

89.

Next set up some heat transfer to the ambient.

The Layer Properties Manager dialog box appears. 90. Double-click the Boundary layer to make it the current layer. 91. Close the Layer Properties Manager dialog box. 92.

or Thermal > FD/Fem Network > Node. The Command line should now read: Enter location of node:

93. Type 0,5,0 in the Command line. The node displays in the upper left of the drawing area.

22-112

A node to connect a convective conductor to is needed. This node will represent the ambient air temperature.

FEM Walled Pipe Example (Continued) 94. Select the newly created node. 95.

or Thermal > Edit. The Node dialog box appears.

96. Highlight the current value in the Submodel field and type AIR.

Edit the node to make it a boundary node and place it in submodel air. Also make the temperature of the node a symbol. Note that when double clicking in a field, the Expression Editor displays.

97. Click on the Boundary radio button in the Type region to select it (display a dot in the circle). 98. Double click in the Initial Temp field. The Expression Editor dialog box appears. 99. Right-clik in the Expression field and choose general > airtemp from the context menu.

The symbol for the air temperature must be defined. Once that is done, the expression used for the temperature of the boundary node can be set.

Airtemp is displayed in the large Expression field. 100.Select OK to close the Expression Editor dialog box. 101.Select OK to close the Node dialog box.

Note that the shape of the node changes to designate a boundary node.

A Thermal Desktop/AutoCAD dialog box appears asking confirmation to add AIR to the submodel list. 102.Select Yes.

22-113

FEM Walled Pipe Example (Continued) 103. or Thermal > FD/Fem Network > Node To Surface Conductor.

A number of AutoCAD groups made up of surfaces were predefined to simplify the tutorial.

The Command line should now read: Select node:

104.Select the boundary node. The Command line should now read: Select surfaces

or [GRP]:

105.Type GRP in the Command line. The Select Groups dialog opens. 106.Select VERT from the list and select OK. The Command line should now read: Adding group VERT Added 351 351 found Select surfaces

or [GRP]:

107.Hit <ENTER> 108.Lines representing the conductor appear. 109.Select the new conductor. 110.

or Thermal > Edit.

The Conductor dialog box appears. 111.Type Vertical surfaces in the Comment field. 112.Click on the arrow to the right of the Type field and select Natural Convection Vertical Flat Plate - Isothermal from the drop-down list (change from default value of Generic). The dialog box changes to reflect the selection

22-114

The new conductor set can be selected by picking any line of the set. The number of surfaces in this selection set is large, and is only chosen to simplify this tutorial. A more limited selection can result in better estimates from the correlation by being able to better specify the correct angle of inclination, height of plate and a more uniform final surface temperature. The is encouraged to experiment with additional subdivisions.

FEM Walled Pipe Example (Continued) 113.Highlight the current value in the Height field and type 4.

Due to the variety of possible configurations and assumptions that can be made by the , Thermal Desktop does not automatically calculate geometry parameters for natural convection.

114.Select OK to close the dialog box. 115.Select the new conductor.

Turn off the conductor visibility so it does not interfere with the next step.

116. or Thermal > Modeling Tools > Turn Visibility Off.

22-115

FEM Walled Pipe Example (Continued) For the second group of surfaces. 117. or Thermal > FD/Fem Network > Node To Surface Conductor. The Command line should now read: Select node:


or [GRP]:

119.Type GRP in the Command line. The Select Groups dialog opens. 120.Select HORIUP from the list and select OK. The Command line should now read: Select surfaces

or [GRP]:

121.Press <Enter>. Lines representing the conductor appear.

22-116

FEM Walled Pipe Example (Continued) 122.Select the new conductor. 123.

or Thermal > Edit.

The Conductor dialog box appears. 124.Type Horizontal top side in the Comment field. 125.Click on the arrow to the right of the Type field and select Natural Convection Horizontal Flat Plate Upside from the drop-down list (change from default value of Generic).

The ‘Horizontal Flat Plate Upside’ option will use different natural convection relations depending on whether the surface is being heated or cooled.

The dialog box changes to reflect the selection 126.Highlight the current value in the Area/Perimeter field and type 0.7.

Area/perimeter is approximate.

127.Select OK to close the dialog box. 128.Select the new conductor.


129. or Thermal > Modeling Tools > Turn Visibility Off.

22-117

FEM Walled Pipe Example (Continued) The third group. 130. or Thermal > FD/Fem Network > Node To Surface Conductor. The Command line should now read: Select node:


or [GRP]:

132.Type GRP in the Command line. The Select Groups dialog opens. 133.Select HORIDOWN from the list and select OK. The Command line should now read: Select surfaces

or [GRP]:

134.Press <Enter>. Lines representing the conductor appear.

22-118

FEM Walled Pipe Example (Continued) 135.Select the new conductor. 136.

or Thermal > Edit.

The Conductor dialog box appears. 137.Type Bottom side in the Comment field. 138.Click on the arrow to the right of the Type field and select Natural Convection Horizontal Flat Plate Downside from the drop-down list (change from default value of Generic). The dialog box changes to reflect the selection 139.Highlight the current value in the Area/Perimeter field and type 0.7. 140.Select OK to close the dialog box. 141.Select the new conductor.


142. or Thermal > Modeling Tools > Turn Visibility Off. 143.

or Format > Layer.

Lets get ready to run by displaying the model.

The Layer Properties Manager dialog box appears. 144.Double-click the Pipe layer to make it the current layer. 145.Select the Freeze icon for the Boundary and Construction layers to turn them off. 146.Close the Layer Properties Manager dialog box.

22-119

FEM Walled Pipe Example (Continued) 147. or Thermal > Model Browser.

Turn the visibility back on for all the surfaces.

Note: Or maximize the Model Browser if minimized earlier. 148. or select Display > All using the Model Browser tool bar or menu.

149. or Thermal > Case Set Manager. The Case Set Manager dialog box appears. 150.Select Run 1 Selected Case. A Sinda/Fluint Run Status dialog box appears confirming successful completion of the processor. 151.Select OK to close the dialog box. Note: If problems are encountered during the run and it is not possible to diagnose the problem, consult with your instructor or CRTech for help.

22-120

It is time to launch the steady state analysis which is set up by default in the Case Set Manager. The node color bar should span from 142 °C to about 193 °C.

FEM Walled Pipe Example (Continued) 152. or Thermal > Model Browser.

The Model Browser can be used to find what the model solution looks like.

Note: Or maximize the Model Browser if minimized earlier. 153.Select List > Fluid Submodel.Id using the Model Browser menu. 154.Select the Flow submodel name. 155.Review the information displayed in the right field of the Model Browser. 156.Select Options > Lump Tabulation using the Model Browser menu. 157.Review the information displayed in the right field of the Model Browser. 158.Select Options > Node Map Options using the Model Browser menu.

Expand the Model Browser window to make as much of the right field visible as possible. The last column for mass error shows the flow out of the first plenum and into the last plenum This data can be selected, copied and pasted into a spreadsheet or document.

The Output Format Options dialog box appears.

159.Try the various sorting options available for the lump tabulation. Also try doing a path tabulation, and use the sorting options available for paths. 160.Close the Model Browser.

22-121

FEM Walled Pipe Example (Continued) 161.Select File > Exit. A Thermal Desktop/AutoCAD dialog box appears asking to save the drawing changes. 162.Select Yes.

22-122


Appendix A or Table

Type of

From

Conduction Coefficient

Absolute Conductance

Use Material

Radiation

MLI

EDGE :

--

--

--

--

--

--

Known conductance

edges

conductance

yes

no

no

no

Solid metal, weld or brazea

edges

area/thickness

no

weld, braze or metal

no

no

Bolted interface

edges

bolt conductance * number of bolts

yes

no

no

no

---

--

--

--

--

--

--

--

no

no

no

no

- or interface conductance FACE :

--

--

--

--

--

--

Known conductance

face

conductance

yes

no

no

no

Thermal grease

faces

interface conductanceb

no

no

no

no

Weld, solder or braze

faces

thickness

no

weld, braze or metal

no

no

A-1

Type of Bolted interface

From faces

Conduction Coefficient bolt conductance * number of bolts - or interface conductance

Absolute Conductance

Use Material

Radiation

MLI

yes

no

no

no

-

-

--

--

-

--

--

--

no

no

no

no

RTV, cured in place

faces

interface conductanceb

no

no

no

no

Interface pad/gasket

faces

conductance

yes

no

no

no

Epoxy t

faces

thickness

no

epoxy

no

no

Radiation

faces

emissivity*view factor

no

no

yes

optional

Simple IR shield

faces

1/(1/e1 + 1/e2 1)

no

no

yes

optional

a. A solid metal edge connection is best modeled using edge nodes with the same spatial resolution and merging coincident nodes. b. Interface conductances are overall conductances derived from test data encoming both the conduction through the interstitial filler material and ability of the material to whet the two interfaces.

A-2

Appendix B File Types The table below summarizes all the files used or generated by CRTech products. SINDA/FLUINT sinda.lic

product license file

*.inp

ASCII input file

*.inc

common designation for include files such as fluid properties

*.pp

preprocessor output file

*.out

ASCII output file

*.sav

binary format save file

astap.for

compilation listing

messages.txt

message and status file from SINDA/FLUINT execution

Thermal Desktop radcad.lic

product license file

*.dwg

drawing file

*.tdp

thermophysical property data base

*.rco

optical property data base

*.cc

node and conductor data include file

*.k

radiation conductor include file

*.ka

radiation conductor time dependent array include file

*.kl

radiation conductor lookup calls for time dependent array data

*.hra

heating array include file

*.hrl

lookup calls for time dependent heat array data

*.unv

FEM universal file

*.igs

IGES design data file

EZ-XY Plot Utility *.ezxy

EZ-XY plot file

B-1

B-2

Appendix C File Formats The following sections define the formats of a number of input and output files.

C.1

Measures Text Input File

The following format is used for importing groups of Temperature Measures as defined in Section 13.2. The file is a space- or comma-delimited, ASCII-format file with the following keywords for each line. Any dimensions are assumed to be in the model units. Double double-quotes (“”) represents the default keyword. Only the first four keywords (0, 1, 2, and 3) are required, but placeholders must be used for any keyword preceding a value to be entered. • Keyword 0: Name (required). Multiple words must be enclosed in double quotes. • Keyword 1: X location (required) • Keyword 2: Y location (required) • Keyword 3: Z location (required) • Keyword 4: Size - The radius of the temperature measure entity. The size should be set to allow the best visibility. Units are model units. The default is 1 unit. • Keyword 5: Output (0|1) - 0 means do not create a ; 1 means create a with the name given in Keyword 6. The default is 1. • Keyword 6: name - The name of the created by Keyword 5. The default is TC_#, where # is an automatically determined sequential digit. • Keyword 7: Output Node (0|1)- 0 means do not create a node; 1 means create a node with submodel.ID specified in Keyword 8. The default is 1. • Keyword 8: submodel.ID - The submodel and ID number for the node created by Keyword 7. The default is TC.#, where # is an automatically determined sequential digit. • Keyword 9: Use Conductor and Thermal Capacitance (0|1) - 0 means the node created by Keyword 7 is a boundary node that contains a calculated value for its temperature; 1 means that the node created by Keyword 7 will participate in the thermal solution with a conductance to the network specified in Keyword 10 and a capacitance specified in Keyword 11. The default is 0. • Keyword 10: Conductance - The value in model units of the conductance from the node created by Keyword 7 to the mapped object. The default is 1. • Keyword 11: Thermal Capacitance - The value in model units of the capacitance of the node created by Keyword 7. The default is 1.

C-1

• Keyword 12: Test Option (0|1) - 0 sets test option to Use All TD Entities; 1 sets test method to Test AutoCAD Group with the group name specified in Keyword 13. The default is 0. • Keyword 13: Group name - The group name to be used when Keyword 12 equals 1. Must be an existing AutoCAD group. • Keyword 14: Connect to Outermost Nodes of Insulation (if found on Surface) (0|1) - 0 means the insulation nodes will not be used for mapping; 1 means the insulation nodes will be used for mapping. The default is 0.

A following line of an Input File would be equivalent to the edit form shown in Figure C-1. Scooby 1, 1, 3 0.025 1 RegName 1 mySubmodel.1 0 1 1 0 "" 1

Figure C-1

C-2

Temperature measure dialog completed by Input File

C.2

C-3

C-4

Index A active sides displaying 8-1 preferences 8-1 Advanced Modeling Guide (open Start > Programs > Thermal Desktop > s Manual - Meshing) advection pipe edit form tab 5-56 advection (see material flow) alias thermophysical property 3-22 analysis groups 4-1 creating 4-3 list by, in Model Browser 2-8 anisotropic materials applying to finite elements 3-24 material orienters 3-24 ANSYS import 18-8 import capabilities 18-16 area conductance 9-3 articulator 4-100–4-105 assembly 4-101 attaching finite elements 4-101 attaching geometry 4-100 detach all geometry 4-101 detaching geometry 4-100 highlight geometry 4-101 list by, in Model Browser 2-10 TDMesh with 4-101 toggle global activation 4-105 tracker 4-102 assembly 4-101 attaching finite elements 4-101 attaching geometry 4-100 detach all geometry 4-101 detaching geometry 4-100 highlight geometry 4-101 list by, in Model Browser 2-10 See also articulator TDMesh with 4-101 toggle global activation 4-105

AutoCAD commands imprint 14-10 AutoCAD geometry meshing 14-1 AutoCAD group list by, in Model Browser 2-12 AutoCAD groups making 7-7 AutoCAD surface converting to nodes/elements 4-110 create FD surface from 4-34 meshing (see TDMesh) AutoCAD versions 19-1 Automatic AutoCAD Initialization Settings 233 Automatic Lighting Settings 2-33 Automatic Regens 2-33 Automatic System Graphics Configuration 233

B blocks, AutoCAD import 18-4 boundary condition mapper list by, in Model Browser 2-12 box finite difference 4-21 brick finite difference 4-51 brick finite element 4-59

C calculate mass 8-7 capacitance calculations 9-1 Capillary pump list by, in Model Browser 2-11 case set manager 15-1–15-30 building specific submodels 15-14 creating cases 15-1 radiation parameters 15-4 sequencing 15-1 setting run priority 15-3 setting Sinda/Fluint calculations 15-7 check symbol units preferences 2-34

coincident nodes merge 4-108 commands AutoCAD imprint 14-10 comment field 2-41 conductance calculations 9-2 conductance capacitance calculations capacitance 9-1 conductance 9-2 outputting 9-8, 9-19 parameters 9-17 super network 9-16, 9-19 conduction heat transfer defining for FD solids 4-46 defining for surfaces 4-15 conductor function of temperature difference type 473 generic type 4-70 list by, in Model Browser 2-10 natural convection type 4-71 network element logic 2-47 node-to-node 4-68 node-to-nodes 4-68 node-to-surface 4-68 primary node 4-68 target node(s) 4-68 type function of temperature difference 473 generic 4-70 natural convection 4-71 cone finite difference solid 4-52 finite difference surface 4-22 elliptic 4-30 scarfed 4-42 conductance area 9-3 defining in solid edit form 4-48 defining in surface edit form 4-16 defining through or 4-74 display 4-16 edge 9-6 integration intervals 9-3

list by, in Model Browser 2-10 setting parameters 9-3 or 4-74 list by, in Model Browser 2-10 convection using ties 5-30 convection heat transfer fluid ties to nodes 5-31 fluid ties to surfaces 5-31 forced 5-31 natural 4-71 convert AutoCAD surface to nodes/elements 4-110 correlate model to test data 12-11 correspondence data 7-4 cutting plane list by, in Model Browser 2-12 cylinder finite difference surface 4-24 elliptic 4-31 scarfed 4-42

D data exchange 18-1–18-38 data logger compare 12-11 data mapper 18-23 creating 18-23 editing 18-24 group association 18-26 tips 18-30 DEFAULT optical property 3-10 DEFAULT thermophysical property 3-10 defaults 2-24 disk 4-26 display ors 8-6 display menu, model browser 2-18 display preferences 6-2 double-sided surface capacitance 9-1 conductance 9-2 DUMPT subroutine call 16-5 duplicate nodes 8-5 dynamic Sinda accessing 16-1 accessing conductor conductance 16-5

accessing conductor heat rates 16-5 generate interface command 16-3 subroutine calls DUMPT 16-5 TDCASE 16-2 TDCMD 16-3 map NASTRAN 16-3 output 16-4 postprocess 16-4 send command to AutoCAD 16-5 start case 16-5 TDGETSYMBOL 16-5 TDGVALUE 16-5 TDHRVALUE 16-5 TDOBJ 16-1 TDSETALO 16-3 TDSETALT 16-3 TDSETDES 16-1 TDSETRAN 16-2 TDSETREG 16-2 TDUPDATE 16-2 dynamic sinda recalculating 16-2

E edge conductance 9-6 edit 2-7 edit menu, Model Browser 2-18 elements attaching to articulators 4-101 planar 4-32 See also TDMesh solid 4-58 ellipse 4-28 ellipsoid 4-29 elliptic cone 4-30 elliptic cylinder surface 4-31 environment, heating defining 6-3 error control heating rate 10-8 radk 10-8 view factor 10-8 executing changing run priority 15-3 export

case set 2-52 data 18-22–18-34 ANSYS 18-23 locations 18-32 mapper 18-23 NASTRAN 18-23 NASTRAN (with STEP-209 model) 18-37 node information 18-22 using FEMAP mesh 18-23 using I-DEAS mesh 18-23 geometry 18-38 logic 2-52 models 18-35–18-37 NASTRAN 18-37 STEP TAS 18-37 STEP-209 18-37 TRASYS 18-35 TSS 18-36 orbits 2-52 properties 2-52 radiation analysis groups 2-52 submodels 2-52 symbols 2-52 export button 2-52 expression editor 2-41 external heating environments 6-1–6-34 external references 19-10 extrude planar elements to solids 4-110

F fans 5-19 FEMAP import 18-8 import capabilities 18-12 finite difference (FD) solid brick 4-51 finite difference (FD) solids conductance data 4-46 conductance 4-48 cylinder 4-52 ellipsoid 4-53 numbering 4-46 radiation data 4-47 sphere 4-55 subdivision 4-45

finite difference surface box 4-21 sphere 4-42 finite element list by, in Model Browser 2-10 solid brick 4-59 pyramid 4-58 tetrahedron 4-58 wedge 4-59 finite elements attaching to articulators 4-101 planar 4-32 See also TDMesh solid 4-58 FK calculator 5-69 flat-front modeling 5-43 FloCAD units 2-27 fluid ID choosing the fluid 5-3 fluid lump list by, in Model Browser 2-11 fluid modeling 5-1–5-71 ducts 5-36, 5-39 fluint macro 5-6 heat pipes 5-39 macro 5-6 network objects fluint macros 5-36 loss 5-13 multiple lumps and paths 5-36 paths FK calculator 5-69 pump or fan 5-19 stubes 5-14 tabular 5-21 ties ties to nodes 5-31 to surfaces 5-31 objects 5-6 lumps 5-6 paths 5-10 setting ID and increments 5-5 overview 5-1 pipes 5-39

connecting pipes 7-10 pool boiling 5-35 submodel setting fluid ID 5-3 setting object ID and increments 5-5 submodels 5-1 fluid network ids 5-5 fluid path list by, in Model Browser 2-11 fluid rotation axis list by, in Model Browser 2-11 fluid submodel list by, in Model Browser 2-11 fluid submodels 5-1 network objects 5-6 lumps 5-6 orifice 5-17, 5-22 paths 5-10 object numbering and increments 5-5 specifying a fluid ID 5-3 fluid tie list by, in Model Browser 2-11 FLUIDINI subroutine 15-22 fluid-to-fluid tie list by, in Model Browser 2-11 form factors outputting 10-31 free edges show 8-5 ftie list by, in Model Browser 2-11 functions 2-45

G generic conductor 4-70 gradient surface capacitance 9-1 conductance 9-2 grip manipulator list by, in Model Browser 2-10 group list by, in Model Browser 2-12

H heat exchangers

list by, in Model Browser 2-11 heat load 4-81 list by, in Model Browser 2-10 heat pipe 5-39 heat pipe data pipe edit form tab 5-49 heater 4-83 in steady state 4-86 list by, in Model Browser 2-10 append string 4-87 s 4-87 sense method 4-86 transient scaling 4-86 heating environment defining 6-3 heating rates calculating 10-32 output parameters 10-21 outputting 10-32 heating, external environments 6-1–6-34 heating, free molecular 6-25 heliocentric orbit 6-8

NASTRAN 18-8 NASTRAN capabilities 18-13 NEVADA 18-7 STEP TAS 5.2 18-8 STEP-209 18-17 TASPCB 18-18 TRASYS 18-6 TSS 18-7 orbits 2-52 properties 2-52 radiation analysis groups 2-52 submodels 2-52 symbols 2-52 import button 2-52 insulation defining in the surface edit form 4-17 layers 4-20 pipe edit form tab 5-55

I

L

I-deas FD import 18-13 FEM import 18-8 FEM import capabilities 18-13 iface list by, in Model Browser 2-11 import case set 2-52 geometry 18-1 AutoCAD block 18-4 logic 2-52 models 18-4–18-21 ANSYS 18-8 ANSYS capabilities 18-16 ANSYS IceBoard 18-18 BetaSoft 18-18 FEMAP 18-8 FEMAP neutral capabilities 18-12 I-deas FD 18-13 I-deas FEM 18-8 I-deas FEM capabilities 18-13 Mesh Importer 18-8

K Keep Graphics Up To Date 2-33 Keplerian orbit 6-15

layer list by, in Model Browser 2-12 layers 19-9, 20-29 changing 19-9 current 19-9 freeze 19-9 off 19-9 on 19-9 thaw 19-9 LINE macro 5-69 load external references into radiation and Cond/Cap calculations 2-34 LOADT subroutine 15-21 logic 2-47 Logic Manager 12-1 lump list by, in Model Browser 2-11 network element logic 2-47 lumps 5-6 lumps and paths 5-36

M macro duct 5-16, 5-69 fluid modeling 5-6 macros list by, in Model Browser 2-11 map data commands 18-31–18-34 map to ANSYS (see data mapper) map to locations 18-32 map to NASTRAN (see data mapper) map solid mesh between conics 4-111 mapper data 18-23 creating 18-23 editing 18-24 group association 18-26 tips 18-30 mapping data to FEM 18-23 material flow 9-8 FloCAD wall-only pipes 9-13 solid brick 9-11 solid cone 9-13 solid cylinder 9-12 solid finite difference object 9-10 solid sphere 9-12 using conductors to represent, 9-14 using ors to connect, 9-15 material orienter 3-24 list by, in Model Browser 2-10 material properties (see thermophysical properties) 3-10 Max Legend Length 2-33 measurement point list by, in Model Browser 2-11 measures 13-1–13-4 menu 2-1 thermal 2-2 merge coincident nodes 4-108 mesh attach to articulator 14-16 controlling resolution 14-6 controlling visibility 14-14 copy 14-16 display preferences 14-14 editing 14-15 FD to finite elements 7-9

generating 14-12 move 14-15 viewing 14-14 mesh controller 14-15 controlling visibility 14-14 generating 14-5 mesh displayer list by, in Model Browser 2-12 Mesh Importer 18-8 mesh importer list by, in Model Browser 2-12 mesh nodalization 7-8 mesher extrude 14-6, 14-16, 14-19 list by, in Model Browser 2-12 revolve 14-6, 14-19, 14-22 See Advanced Modeling Guide meshing 14-1 AutoCAD geometry meshing 14-22 final preparation for 14-3 meshing between surfaces 4-111 MLI adding to surface 4-17 model browser 2-8–2-21 auto select 2-19 auto update 2-19 display menu 2-18 display options 2-18 edit menu 2-18 editing from 2-18 large models 2-21 options menu 2-19 model checks 8-1–8-8 check overlapping surfaces 8-8 color by property 8-4 display active sides 8-1 list duplicate nodes 8-5 show free edges 8-5 modeling tools 7-1–7-14 displaying node IDs 7-13 edit object 2-7 making AutoCAD groups 7-7 radiation analysis groups 7-7 submodel 7-7 thermal objects 7-7

mesh nodalization 7-8 node correspondence 7-4 node layer synchronization 7-13 object filter 7-11 object property copying 7-14 object visibility 7-13 resequence node IDs 7-2 Coordinate System alignment to surface 7-8

N NASTRAN export model to 18-37 import 18-8 import capabilities 18-13 natural convection conductor 4-71 network element logic 2-47 network elements thermoelectric cooler (TEC) 4-88 network logic 2-47 network objects ??–4-112 convert AutoCAD surface 4-110 heat load 4-81 heater 4-83 nodes listing duplicate 8-5 pressure load 4-91 synchronize element normals 4-110 thermal ors 4-74 nodes 4-62 node correspondence 7-4 node information export 18-22 node list 2-53 node numbering pipe edit form 5-54 nodes creating 4-62 list by, in Model Browser 2-8 merging coincident 4-108 node-to-node conductor 4-68 node-to-nodes conductor 4-68 node-to-surface conductor 4-68 non-graphical objects list by, in Model Browser 2-8

NORMAL optical property 3-10

O oct cell control 10-10 oct tree setting parameters 10-13 offset paraboloid 4-36 ogive 4-36 optical properties 3-1–3-10 alias 3-7 aliases 3-7, 3-9 angular dependence 3-5 built-in 3-10 create database 3-6 DEFAULT 3-10 editing 3-1 list by, in Model Browser 2-9 NORMAL 3-10 open database 3-6 opening a database 3-6 verification by coloring 8-4 options menu, model browser 2-19 orbit argument of periapsis 6-16 basic 6-18 defining 6-3 display preferences 6-29 eccentricity 6-17 Keplerian 6-15 vantage point 8-4 orbit manager 6-1–6-34 accessing 6-1 argument of periapsis 6-16 ascending node 6-15 celestial coordinate system 6-30 defining a basic orbit 6-18 defining orbits 6-3 display vehicle in orbit 6-32 view points 6-32 eccentricity 6-17 inclination of equator 6-7 Keplerian orbit 6-15 orbit inclination 6-15 orbit positions for calculation 6-5

period 6-17 right ascension 6-15 sidereal period 6-7 trajectory trajectory 6-24 using beta angles 6-18 vector list 6-24 vehicle spin parameters 6-14 vehicle time based positioning 6-18 orienter, material list by, in Model Browser 2-10 orifice 5-17, 5-22 output parameters heating rates 10-21 radiation analysis 10-18 Radks 10-1 overlapping surfaces check 8-8

P parabolic trough 4-36 paraboloid 4-38 offset 4-36 parameterization 11-1–11-11 about 11-1 built-in symbols 11-6 creating symbols 11-1 parametric runs sample set up 16-6 part controlling visibility 14-14 problems in 14-11 path list by, in Model Browser 2-11 network element logic 2-47 paths 5-10 pipe list by, in Model Browser 2-11 pipe attributed pipe edit form tab 5-46 pipe connectivity 8-6 pipes 5-6, 5-39 plotting rays 10-27 polygon 4-40 pool boiling ties 5-35 post processing. see postprocessing

postprocessing 17-1–17-32 color 17-1 coloring 12-3 data mapper 18-23 datasets 17-10 export display 17-23 multiple color bars 12-3 transients animation 17-22 step through time 17-21 xy plotting 12-4, 17-26 postprocessing mapper list by, in Model Browser 2-12 preferences active sides 8-1 advanced 2-32 AutoCAD initialization 2-33 automatic regens 2-33 check symbol units 2-34 graphic size 2-29 graphics visibility 2-27 keep graphics up to date 2-33 lighting 2-33 load external references 2-34 max legend length. 2-33 radk status 2-34 show message window 2-34 Sinda.Fluint output 2-31 system graphics 2-33 units 2-25 use simple text editor 2-34 2-25–2-34 XY plot legend labels 2-33 pressure load 4-91 list by, in Model Browser 2-10 preview controlling visibility 14-14 generating 14-5 primary node, conductor 4-68 property alias optical 3-7 property name 3-9 pumps 5-19 pyramid finite element 4-58

R

S

radiation analysis 10-1–10-33 analysis groups 4-1–4-5 automatic error control 10-8 controlling from the case set manager 15-4 controlling number of rays 10-4 fast spin parameters 10-29 method 10-1 model checking color by property 8-4 display active sides 8-1 Monte Carlo 10-1 oct cell 10-10 oct tree parameters 10-13 output parameters 10-18 output to Sinda/Fluint radiation analysis calculating 10-1 outputting form factors and radks 10-31 radiosity 10-1 ray cutoff 10-9 setting calculation parameters 10-2 show status during calculation 2-34 radiation analysis groups 4-1 creating 4-3 radiation heat transfer defining from a FD solid 4-47 defining from a surface 4-13 Radks outputting 10-31 ray cutoff radiation analysis 10-9 rays per node 10-4 ripe 5-39 rectangle 4-40 append string 2-41 heater 4-87 prefix 2-41 resequence node IDs 4-35, 7-2 resolution, mesh controlling 14-6 revolve planar elements to solids 4-110 rotation axis list by, in Model Browser 2-11 rotation of surfaces 4-21 Running 16-1

scarfed cone surface 4-42 scarfed cylinder surface 4-42 show message window 2-34 Show Status during radk calculations 2-34 Sinda/Fluint controlling output 2-31 creating Fluint models 5-1 setting calculations in case set manager 157 slivers 14-11 solid finite difference 4-44–4-56 brick 4-51 capacitance data 4-46 conductance data 4-46 cone 4-52 conductance 4-48 cylinder 4-52 ellipsoid 4-53 numbering 4-46 radiation data 4-47 sphere 4-55 subdivision 4-45 finite element brick 4-59 pyramid 4-58 surface coat 4-109 tetrahedron 4-58 wedge 4-59 list by, in Model Browser 2-10 solver dynamic update of design variables 16-1 dynamic update of random variables 16-2 dynamic update of s and symbols 16-2 sample set up 16-6 setting object 16-1 sphere finite difference surface 4-42 spin fast spin 10-29 heat rates for fast spin 6-14 slow spin 6-5 stack manager 4-20 STEP TAS

export 18-37 STEP-209 export 18-37 Stubes 5-14 subdivision pipe edit form tab 5-45 submodels building in case set manager 15-14 fluid 5-1 thermal 4-6 super network creating 9-16 surface capacitance data 4-15 conductance data 4-15 cone 4-22 conductance 4-16 creating from AutoCAD surface 4-34 disk 4-26 double-sided, capacitance 9-1 double-sided, conductance 9-2 ellipse 4-28 ellipsoid 4-29 gradient, capacitance 9-1 gradient, conductance 9-2 insulation 4-17 list by, in Model Browser 2-10 numbering 4-12 paraboloid 4-38 pipe edit form tab 5-55 polygon 4-40 radiation data 4-13 rectangle 4-40 sphere 4-42 subdivision of 4-11 visibility 2-27 surface coat free solid elements 4-109 symbol list by, in Model Browser 2-12 symbol manager defining symbols 11-1 symbols built-in 11-6 defining 11-1 parameterizing a model 11-1 units 2-42

updating all, in dynamic SINDA dynamic Sinda 16-2 synchronize element normals 4-110

T tabular 5-21 tag list by, in Model Browser 2-12 Tag Set 2-21 target node(s), conductor 4-68 TD Direct importer list by, in Model Browser 2-11 TDCASE subroutine call 16-2 TDCMD mapping commands subroutine call 16-3 output subroutine call 16-4 postprocessing subroutine call 16-4 send command to AutoCAD subroutine call 16-5 start case subroutine call 16-5 subroutine call 16-3 TDGETSYMBOL subroutine call 16-5 TDGVALUE subroutine call 16-5 TDHRVALUE subroutine call 16-5 TDMesh extrude 14-6, 14-16, 14-19 mesher 14-22 revolve 14-6, 14-19, 14-22 See Advanced Modeling Guide with articulators 4-101 TDmesh 14-1 TDOBJ subroutine call 16-1 TDSETALO subroutine call 16-3 TDSETALT subroutine call 16-3 TDSETDES

subroutine call 16-1 TDSETRAN subroutine call 16-2 TDSETREG subroutine call 16-2 TDUPDATE 16-2 TEC 4-88 list by, in Model Browser 2-10 test data correlation 12-11 tetrahedron finite element 4-58 text search Model Browser 2-14 thermal menu 2-2 thermal submodels 4-6 thermocouple list by, in Model Browser 2-11 thermoelectric cooler 4-88 list by, in Model Browser 2-10 thermophysical properties 3-10 alias 3-22 aliases 3-11 anisotropic materials 3-24 create database 3-11 editing 3-11 list by, in Model Browser 2-9 open database 3-11 verification by coloring 8-4 thermophysical property create 3-11 edit 3-11 tie list by, in Model Browser 2-11 network element logic 2-47 ties pipe edit form tab 5-53 pool boiling 5-35 ties to nodes 5-31 ties to surfaces 5-31 toolbars 2-1 tracker 4-102 attaching finite elements 4-101 attaching geometry 4-100 detach all geometry 4-101 detaching geometry 4-100 highlight geometry 4-101 list by, in Model Browser 2-10

See also articulator TDMesh with 4-101 toggle global activation 4-105 trackers arrow color 4-104 disabling 10-30 graphical representation 4-104 resetting 4-104 translation of surfaces 4-21 TRASYS export 18-35 import 18-6 TSS export 18-36 import 18-7 tutorials 20-1–??, 20-35–??, 20-41–??, 20-57– 22-122 beer can 20-89 circuit board conduction 20-67, 20-129 combined radiation and conduction using finite elements 20-129 conventions used 20-1 dynamic SINDA example 20-201 FloCAD air flow through an enclosure 22-3 drawn shape heat pipe 22-85 FEM walled pipe 22-99 heat pipe 22-23 manifolded coldplate 22-37 getting started colors 20-34 display 20-27 grip points 20-19 layers 20-29 interface 20-6, 20-10 views 20-13, 20-21 mapping temperature to a NASTRAN model 20-157 RadCAD importing TRASYS 21-35 oct tree example 21-23 orbital heating rates 21-53 radks for parallel plates 21-3 simple satellite 21-71 using articulators 21-35 simple meshing methods 20-57

template drawing 20-35

U units 2-25 FloCAD 2-27 scaling 2-26 use simple text editor 2-34 Coordinate System alignment to surface 7-8 display at origin 7-8 displaying 7-8 preferences 2-25–2-34

V visibility controlling through the model browser 212 from the Preferences dialog box 2-27 turn node IDs on/off 7-13 turn visibility on and off 2-27 turn visibility on/off/undo 7-13 visibility, controlling for parts, meshes and previews 14-14

W wedge finite element 4-59 wildcards 2-53

X XREF 19-10 XY Plot Legend Labels 2-33 xy plotting 12-4, 17-26

Thermal Desktop 2b75t

Overview 3z723u

More details 2i4a6q

Related Documents g842

Thermal Desktop 2b75t

Simplemind Desktop 3w1z35

Thermal Physics.pdf 134l67

Thermal Oil 373u6i

Thermal Analysis 5h3b2b

Thermal 2 1x26s

More Documents from "enilo" 6z5l5h

Mensajes De Antartida-02de02 4j6c54

Instructivo Rendicion Caja Chica Gde V1.0 323e2m

Thermal Desktop 2b75t