Translation of Patran Thermal Input to SINDA

Thermal > PATQ Preference Program > Translation of Patran Thermal Input to SINDA

XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX''">XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX''">

Introduction to the SINDA Utility from Patran Thermal

When all of the Patran Thermal files (TEMPLATEDAT, MATDAT, CONDUCDAT, CONVECDAT, etc.) have been created, there are two possibilities for performing the analysis. The first option is to submit the job to Patran Thermal’s solver QTRAN. This solver has many advanced features such as adaptive time stepping with a continually updated weighting factor to vary the implicitness or explicitness of the transient analysis. The second option is to create a run-ready SINDA input file and submit the job for SINDA analysis. (The SINDA solver is not supplied by MSc.Software Corporation).

SINDA has been an industry standard for thermal analysis since the 1960s (as CINDA) and is still widely used.

The SINDA utility is designed to create input files for either a SINDA 85 code (which contains blocks of the format “HEADER NODE DATA”) or Network Analysis SINDA (also called SINDA/G with blocks of the format BCD 3NODE “DATA”). There are a wide variety of “BCD” versions of SINDA, such as CINDA, which have only minor input differences.

Some of the unique benefits of the SINDA input file created by Patran Thermal are as follows:

1. Radiation view factors are calculated as an integral part of the Patran Thermal program and are directly accessed by the SINDA utility. Typically, a conduction/convection model for SINDA is created and a separate radiation model is built for an external code (such as TRASYS, or NEVADA) which is run to calculate the radiation view factors. The results of this radiation analysis are cut and pasted into the conduction/convection SINDA input file and is often lengthy and cumbersome. This process is significantly streamlined by having all boundary conditions (including radiation) in one Patran model.

2. The SINDA input file is a mathematically exact representation of the finite element model and is converted into a “Resistor-Capacitor” network. Accuracy is significantly increased in the portion of a model where elements are not orthogonal (even slightly skewed elements).

3. Any of the 971 materials defined in the Patran Thermal material library (each contains thermal conductivity, specific heat, and density) can be automatically input into the SINDA file. All temperature dependent material properties are mapped directly into ARRAY DATA.

4. A subroutine is automatically included at the bottom of each newly created SINDA input file which allows SINDA to directly write Patran nodal results files for postprocessing. This eliminates the need for a reverse translator from SINDA to Patran.

5. All of the conductors for conduction, convection, advection (mass flow), and radiation are grouped separately for easy modification of their values with a “FAC” card.

Creating the SINDA Input File

At this point, the Patran model is completed and a neutral file exists. To create the SINDA input file, enter Patran Thermal by executing PATQ (or PTHERMAL). The PATQ menu picks, shown below, have been described earlier in Chapter 3. The following describes how these picks relate to SINDA input file creation. Upon entering PATQ, the following menu will appear.

Table C‑7 PATQ Menu


Select 2)	and Patran Thermal reads the default TEMPLATE.BIN file in Patran Thermals root directory, as well as a local TEMPLATE.DAT file which may exist in the current directory. All the Patran Thermal files (TEMPDAT, CONDUCDAT, etc.) will be created in the local directory.
Select 3)	only if VFAC boundary conditions were created for radiation in Patran. This will execute VIEW FACTOR in batch mode and create a VFRESDAT file that will be read later to create radiation resistors in the SINDA input format.
Select 4)	if material properties (k, rho, or Cp) need to be automatically read from the Patran Thermal material library and inserted into the local MATDAT file. Otherwise this selection can be skipped.
Select 6)	to access the utility menu which is also described in detail in section 3.2.2. The utility menu is where the actual translation to SINDA input occurs and is displayed on the next page.

Table C‑8 PATQ Utilities Menu


Select 15)	to create the SINDA input deck from existing Patran Thermal files created in the previous steps.

The first file read is the QINDAT file located in the local directory. The local directory is again searched for a file called APPENDSIN This file contains parameters that affect the way the SINDA input file will be created and also contains logic blocks and a subroutine that allows SINDA to write Patran nodal results files. The Fortran files are appended to the bottom of the SINDA input file which is called MODELSIN. The APPENDSIN file is discussed in detail in The APPENDSIN File, 756. If this file does not exist, then the user is prompted as to which type of SINDA file to create:

Table C‑9 SINDA Execution Format Options

After the user selects either “1” or “2”, the default APPENDSIN file will be written to the local directory for future customization of SINDA files.


Note:	It is recommended that all APPENDSIN files be deleted from the local directory before re-executing the SINDA utility if the run is being changed from steady-state to transient (or vice versa) or if the requested deck is being changed from SINDA85 to Network Analysis SINDA (or vice versa).

The SINDA utility then proceeds to read MSC Patran Thermal files and creates the appropriate SINDA input file. The following table shows a sample execution which indicates the files that are read from the local directory and then displays a summary of the total number of nodes and conductors created.

Table C‑10 Sample Run Log of SINDA Translation

The number of conductors created for the SINDA file might be less than the number of conductors created by Patran Thermal. This is done because Patran Thermal creates conductors between boundary nodes in case the desired classification needs to be changed to a nonboundary node during a transient.

If any errors occurred (such as modeling functionality) that SINDA does not support (such as wavelength-dependent radiation), an error message will be displayed underneath the line that displays which file is currently being read. All these messages are included in the PATQMSG message file.

The MODELSIN File

The resulting input file created by the SINDA utility is called MODELSIN. A sample of this file is shown below:

Table C‑11 Sample SINDA Input File

To understand how the Patran Thermal parameters are translated to SINDA, Table C‑12 describes the supported Patran Thermal to SINDA functionality. This table describes how a particular capability is implemented for a particular function. Much of this information will become clear after translating sample problem 1 for basic functionality and problem 10 which uses some of the more advanced functionality with variable heat loads and convection coefficients.

Table C‑12 Supported SINDA Translation


Patran Thermal to SINDA Functionality
Functionality	Capability		Implementation
Node Data	Constant rho or Cp		MPIDs in MATDAT have IEVAL = “CONST”.
	Constant rho, variable Cp		MPIDs in MATDAT have IEVAL = “CT” and “TABLE” or “ITABLE”.
	Constant Cp, variable rho		MPIDs in MATDAT have IEVAL = “C” and “TABLE” or “ITABLE”.
	Both rho & Cp variable		MPIDs in MATDAT have IEVAL “TABLE” or “ITABLE”.
	Both rho & Cp are massless		Advective, ambient, radiosity nodes, etc.
	Node has a fixed temp		TEMP LBC as fixed.
Conduction	Constant Conductivity		MPID in MATDAT have IEVAL = “C”.
	Temperature Variable		MPIDs in MATDAT have IEVAL “TABLE” or “ITABLE”
	Time Variable		MPIDS in MATDAT have IEVAL = “TABLE” or “IT” and ITSCAL=”T” in MATDAT.
Convection	Constant “h”		CONV LBC with an “h” value
	Spatial Variation		Use of a field
	Temp Variation		Conv Configuration 29 and ITSCAL=F, R, C, or K.
	Time Variation		Conv Configuration 29 and ITSCAL=T or MPID is “-”
Radiation	Gray with fixed emissivity		VFAC has constant emissivity in TEMPLATE.DAT.
Advection	Flowrate	Cp
	constant	constant	PFEG has fixed Mdot and MPID in MATDAT is constant.
	time var	constant	ITSCAL in MATDAT is “T” for time.
	temp var	constant	ITSCAL in MATDAT is F, R, C, or K.
	constant	time var	ITSCAL in MATDAT is “T” for time.
	time var	time var	ITSCALs are “T” for time for both Mdot and Cp.
	temp var	time var	ITSCAL is “T” for time for Cp.
	constant	temp var	ITSCAL in MATDAT is F,R,C, or K.
	time var	temp var	ITSCAL in MATDAT is “T” for time for Mdot.
	temp var	temp var	ITSCALs in MATDAT are F, R, C, or K.
Heat Source	Constant Q		HEAT LBC a fixed value or a function of Length, Area, or Volume.
	temp variable		QMACRO function (must be with one micro 8 and with arg=1).
	time variable		QMACRO function (must be with one micro 8 and with arg.=0).
Run Control	TMPZRO		Calculated from ICCALC from QINDAT.
	Steady State or Transient		IOPT from QINDAT.
	NLOOP		IMAXSS for steady state, 50 for transient.
	TSTART		TSTART from QINDAT.
	TIMEND		TSTOP from QINDAT.
	DRLXCA, ARLXCA		EPSISS for Steady State and EPSIT for Transient.
	DAMPD,DAMPA		RELAXS for steady state RELAXT for transients. If IFXRLX is 1, then the smallest of the individual RLX options will be used as the DAMPX factor.
	SIGMA		This will be put directly into CONDUCTOR DATA. If SBC is “0” and ICCALC is C or K, then SBC=5.6698E-8. If SBC is “0” and ICCALC is F or R, then SBC=0.1712e-8. Otherwise, SBC is read directly from QINDAT.
	OUTPUT		TPRINT from QINDAT.
	Initial Temperatures		TINITL is read for initial temps in QINDAT but is overwritten by TEMP LBCs in Patran.
Materials	Constant		ITSCAL=”C” in MATDAT. Lumped into capacitors and conductors.
	ARRAY DATA		ITSCAL=”T” or “IT” in MATDAT (time or temp). Independent temperature columns are converted if CCALC is different than ITSCAL in MATDAT. FACTOR is multiplied by dependent column.


Node Data	The nodes are written to the MODELSIN file in the NODE DATA block in the following order: diffusion (stores energy), arithmetic (massless), and boundary (fixed temperature). As shown in Table C‑12, a wide variety of options are possible for the NODE DATA block. An arithmetic node can be created with the NODE,#,ADD command in Patran. Nodes can have initial temperatures specified in Patran, otherwise the “TINITL” parameter in the QINDAT file is used to provide the initial temperatures. A node can also have a fixed temperature with the “-1” parameter specified in the TEMP LBC as described in Section 2.2.3.1 of this manual. Temperatures that reference macro and micro functions are not supported for this release. This functionality would be created by including the appropriate logic in the SINDA VARIABLES block. Currently, 62,400 is the maximum number of nodes per translation.
Conduction	If the MATDAT file specifies the thermal conductivity to be “Constant”, then a constant conductor will be created in the SINDA deck. Time and temperature varying conductors will be created if the material property is input into the MATDAT file in a “TABLE” (or “ITABLE”). If ITSCAL is “T” for time, then a time varying conductor (SIT for Network Analysis' SINDA) will be created. See MPID Number, Function Type, Temperature Scale, Factor and Label (p. 263) in the Patran Thermal User’s Guide Volume 1: Thermal/Hydraulic Analysis under Material Properties (Ch. 8) for a description of ITSCAL and IEVAL. There is no limit to the maximum number of conductors per translation.
Convection	Convection with a constant heat transfer coefficient is specified as described in Chapter 2 and also in sample problems 1 and 10. Problem 10 also demonstrates how convection can be specified as a function of temperature. This is done by creating a “CONV” template in the TEMPLATEDAT file which references configuration 29. This configuration points to a table of temperature versus heat transfer coefficient in the MATDAT file. Time varying convection is done in a similar way except ITSCAL for the convection MPID in the MATDAT file is “T” for time. Convection can also be created by referencing a spatial varying convection field or data entity (data line, data patch, or data hyperpatch). Other convection configurations are not supported by the SINDA utility.
Radiation	Gray body radiation with fixed emissivity is supported as in Section 2.2.3.5 with the VFAC LBC and in the users manual. These radiation resistors are used in problem 10 and in all 3 example View Factor problems. Variable emissivity and spectral radiation are not supported for the SINDA utility.
Advection	Variable and/or constant mass flow rate and specific heat are supported for advection. The hydraulic network functionality of Patran Thermal which calculates pressure drops and flow rates is not supported for the SINDA utility.
Heat Source	Heat sources are applied by using the HEAT LBC for heat flux or nodal heat source respectively. Constant or variable heat values are supported. Variable heat sources are allowed by using a table of independent variable (time or temperature) versus heat with a MACRO function with microfunction 8. This is the only microfunction supported. For temperature varying sources, “ARG” is equal to “1” and for time varying sources, “ARG” is equal to “0” as specified in MFID, Independent Variable, and Function Type (p. 323) in the Patran Thermal User’s Guide Volume 1: Thermal/Hydraulic Analysis under Microfunction Data (Ch. 8).
Run Control	The run control parameters are taken from the QINDAT file in the local directory. It is important to note that the SINDA utility will NOT look for any nodes or resistors inserted directly into the QINDAT file. It will only read these nodes and resistors from the appropriate files (nodedat, resdat, etc.) Table C‑12 shows which run control parameters are read. All other parameters in the QINDAT file are ignored.
Materials	If the IEVAL parameter of the material property (MPID) in the MATDAT file is T (or “IT”) for table, then the table will be put into ARRAY DATA. If the value of ITSCAL (also in the MATDAT file) is different than the value of ICCALC in the QINDAT file, then a conversion of the temperatures in the MATDAT file will take place to match the temperature units specified by ICCALC (the calculation units). A message in the array will note that this has taken place. Any of the thermal conductivities and specific heats in the Patran Thermal material library that are tables can be brought into ARRAY DATA. The phase change properties in the MATDAT file will be ignored by the SINDA utility.

The APPENDSIN File

The APPENDSIN file contains six parameters at the top of the file and the rest of the file contains the logic blocks which are appended to the bottom of the newly created SINDA data blocks (NODE, CONDUCTOR, etc.). These parameters can be modified to a user-specified format so that every subsequent SINDA file created will be customized.

Table C‑13 Sample APPENDSIN File

The APPENDSIN file is created for the user by the SINDA utility if a APPENDSIN file does not already exist in the local directory. If the file does exist in the local directory, then the parameters will be read by the utility and the logic blocks will be appended to the bottom of the SINDA input deck.

The six parameters are described below


SINDA Format	The choices are “BCD” or “HEADER”. This signifies that this file is created for Network Analysis’ “BCD” format SINDA or SINDA ‘85 with “HEADER” format. It is recommended that all APPENDSIN files be deleted from the local directory before re-executing the SINDA utility if the run is being changed from SINDA '85 to Network Analysis’ SINDA (or vice versa). The SINDA utility will prompt the user as to which type of SINDA file should be created next time since an APPENDSIN file no longer exists in the local directory. Since the default versions of SINDA do not support a negative shape factor (area/length) in SIV statements, a temporary value of 1.0 is assigned in the CONDUCTOR block for the negative cross-derivative terms which are recalculated in VARIABLES 1. If the SIV has been modified to accept a negative shape factor (without taking the absolute value) then the term FIXBCD or “FIXHEADER” can be used which will direct the utility to allow SIV statements with a negative shape factor. These cross-derivative conductors are the key to the inherent accuracy of Patran Thermal’s conductor network.
Submodel Name	This is the SINDA ‘85 submodel name and is only relevant for SINDA 85. This line must exist in the APPENDSIN for both types of SINDA but is ignored for Network Analysis’ SINDA.
Conductors & capacitors in arithmetic format (YES, NO)	If YES (default) is entered, then all of the information in NODE DATA and CONDUCTOR DATA blocks will be entered with arithmetic operators (typically with the multiplier symbol "*"). If NO is entered, then all values for a node or conductor entry will be entered as a single value.
Node increment	This value is the increment for each node in the Patran model. Node 1 in the Patran model will become node 2001 if the increment is set to 2000. This might be used if multiple Patran models need to be combined into one SINDA deck. This option should be used with caution since the Patran nodes must correspond to the SINDA nodes for postprocessing in Patran.
Conductor increment	This value is the increment for conductors in the SINDA input file. If the increment is set to 1000, then the first conductor in the SINDA input file will be 1001. This might be used if multiple Patran models need to be combined into one SINDA deck.
System energy balance criteria (BALENG or EBALSA)	This is the system energy balance criteria. For Network Analysis SINDA, this is the BALENG value. For SINDA 85, this is the EBALSA value. The default value is 0.01.

The remainder of the APPENDSIN file contains logic blocks for modification before being appended to the bottom of the SINDA input file. Included in these logic blocks is a SUBROUTINE block. This block contains a subroutine called “RESPAT” that allows SINDA to create a file or files with nodal results in a format that Patran can read directly. These nodal results files are of the form “NR#SIN” where # denotes the output number. The “RESPAT” routine can be called directly from SINDA much like a “TPRINT” routine is typically called for printing temperatures to the SINDA output file. By default, this subroutine call is put in EXECUTION (OPERATIONS DATA for SINDA ‘85) for steady- state runs and in OUTPUT CALLS for transient runs.