MSC Sinda > Analysis > Introduction
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX''">XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX''">   
Introduction
 
SINDA for Patran has four action options: Analysis, Access Result, Thermal Studio and Delete. The Analysis action is used to submit the thermal analysis job to MSC Sinda. Access Result is used to import result file back into Patran GUI. Thermal Studio is to call Thermal Studio to manage .sin file and plot the results. Delete is used to remove the available job names from the analysis form, and also remove those related junk files.
When the action is Analysis, the object has two options: Entire Model and Current Group. This means user can either analyze the whole model, or a part of the thermal model. When Current Group option is chosen, only those loads whose application regions include the elements in the current group will be available.
There are three methods under the Analysis action: Translate and Run, Run Translator and Run MSC Sinda. Translate and Run option will call pat3sdb.exe, sdb2sg.exe, and the MSC Sinda solver. Run Translator will only call pat3db.exe and sdb2sg.exe to create a .sin file. Run MSC Sinda will only call MSC Sinda solver when the job name.sin file already exists.
 
 
When the action is Access Result, if the result file name is the same with the current job name, click the Apply button to import the result data back into Patran. This is the most common case especially when users want to see the temperature result data just after a model runs. Users can also select the different result file names by clicking “Select Result File…” button.
When the action is Delete, the job name and its related files will be deleted from the working directory, and also from the “Available Jobs” window. When the radiation solver is chosen, only the files related with chosen radiation code will be deleted.
Thermal Solution Setup is to set up the solver control parameters for MSC Sinda. Steady State Setup is used for steady state solvers. Users can select a steady state solver from nine options and set up control parameters for this solver. Two important parameters here are TMPZRO and SIGMA which are closely related with model unit system.
Transient State Setup is used for the transient state solver. There are 12 optional transient solvers available. There are 5 important parameters which are often modified: TIMEND, OUTPUT, DTIMEI, TMPZRO and SIGMA. TIMEND is the end time of the transient run, and OUTPUT is the time interval for MSC Sinda to output transient results. Even when the load case is a steady state case, the user can select a transient solver and set up the transient parameters to submit a transient run.
 
Combined SS and TR Setup is a combination of Steady State Setup and Transient State Setup. MSC Sinda will conduct a steady state run first, then use the steady state result as the initial temperature for a transient run.
Specify Initial Conditions is used to set the initial constant temperature, or to set the existing result (job_name.sot) as the initial temperature. If a specified temperature field is needed, the Initial Temperature load should be used.
Radiation Solver Setup is used to select a radiation solver and set up the control parameters. Currently MSC Sinda for Patran supports TRASYS, THERMICA, NEVADA or TSS. The user can select any one of these radiation solvers to work with MSC Sinda. The user needs to check “Select Radiation Solver” to activate the radiation solver options. The current active solver is marked with “(Default)”. For example, if you click the THERMICA button and set up the parameters, then click OK, the THERMICA button will be marked with “(Default)”. Further introduction on different radiation solvers is included in the workshop training materials.
For a steady state run, TIMESCALE is not really useful. For a transient state run, it will affect the time unit for the orbital heat flux and articulation view factors. Suppose you want MSC Sinda to use seconds as the time unit, for NEVADA and TRASYS, TIMESCALE is a multiplier of the Orbital Period. If TIMESCALE = 1.0, the Orbital Period in the orbital setup form should use the second unit. If TIMESCALE = 3600, the Orbital Period in the orbital setup form should use the hour unit, which is the default unit. For THERMICA, you do not have the option to input the Orbital Period in hours or seconds. It will be automatically calculated. The translator will automatically convert the time unit based on the TIMESCALE value. If you want to use seconds in MSC Sinda, TIMESCALE should be 3600.0. Please note: MSC Sinda can use other time units too, but some parameters in the Analysis form should be consistent with MSC Sinda time unit: TIMESCALE, TIMEND, TIMEO, OUTPUT, DTIMEI, SIGMA.
When the super element or primitives radiation loads are applied, three methods are developed to distribute the radiation conductors: Full, Area and Direct. The area method uses only element areas as the weights of distribution. The direct method uses geometry view factors as the weights. The full method uses more complex parameters as the weights. We will introduce these methods in detail later.
There are four MSC Sinda mode options available: Single-precision, double precision, Single-precision (Fortran), and double precision(Fortran). The options without (Fortran) do not need a Fortran compiler in the system, but users can not add Fortran commands in the MSC Sinda input file. To run the options with (Fortran), a Fortran compiler is needed, and users can add any Fortran commands in the input file. The file will be compiled and linked again before running. Double Precision mode is almost the same as standard mode, except double precision provides more accurate analysis.
Pause at end means MSC Sinda will not automatically quit after the calculation is finished. Users should press any key to continue. Keep all files will force MSC Sinda to keep all the intermediate files. This is useful for debugging or viewing the compiling situations, especially when some Fortran commands are added manually.
In the Analysis form, Output Routine Setup is used to set up parameters for MSC Sinda output. PATOUT must be checked to create a result for Patran post-procession. TPNTSN is for the MSC Sinda result, and is not really required by Patran.
Edit/Manage Files is used to edit the result file or input file, save the current settings, or manage skeleton files. Users can run MSC Sinda just after the input file is modified. Users can also load any existing .sin file and run it here by clicking Run MSC Sinda.
The skeleton file is a way to automatically add some comments into the MSC Sinda input file. Here are the steps to create the .ssk (skeleton) file and add some Fortran logic. After clicking Create Skeleton File, a current_job_name.ssk file will be generated. You can click Skeleton File(.ssk) to use the default editor, like notepad.exe to open the .ssk file. This file has only the main structure of the MSC Sinda input file (so called skeleton). You may add/modify this .ssk file by editing a node, modifying a heat source, or adding some Fortran logic….. After you click the Apply button, the translator will create a .sin file, and then call the MSC Sinda solver to run it. The MSC Sinda solver’s pre-processor will check the .ssk file first. If the jobname.ssk file exists, the pre-processor will back up the existing .sin to be .snd, and merge the .ssk file into the .sin file. MSC Sinda will finally run the newly merged .sin file. At the top of the new .sin file, you can find “BCD 3THERMAL LPCS NOSSK”, which means this .sin file has already merged the .ssk file. Next time when MSC Sinda runs this .sin file, the pre-processor will not merge the .ssk file again. You can also unmerge the .sin and .ssk file by clicking the “Unmerge sin_ssk Files” button, which will rename the .snd file back to replace the .sin file.
During the .ssk merging process, the MSC Sinda pre-processor will add the comments to the end of some data blocks, such as NODE, SOURCE, CONDUCTOR or CONSTANT blocks. If it causes multiple definitions, the definitions from .ssk will overwrite those originally in the .sin file, because the comments from .ssk will be added to the end, and therefore will be executed later. Please note: for EXECUTION, VARIABLE1, VARIABLE2, OUTPUT blocks, the comments from .ssk will totally replace those comments originally in the .sin file. For more details, please reference the Introduction (Ch. 3) in the MSC SINDA for Patran User’s Guide/G..
All the current control parameters can be saved into pat2sg.ini file by clicking Save Settings as default. For PCs, this file will be saved in P3-home directory. For UNIX, this file will be saved in user_account_directory/ sindaprefs/.
Default Solvers and Result Checking
SNSOR and SNDUFR are the default steady and transient solvers. They are good for most of the cases. Users are encouraged to check the results by the following methods.
Obtain nearly identical results with a different solver
Obtain nearly identical results with tighter convergence
Obtain nearly identical results with a smaller time step (transient)
Obtain nearly identical results with more rays or different random seeds (external radiation solvers)
About Non-Default Solvers
For non-spacecraft, non-radiation problems
Solid type model – SNSOR, SNTSM1
Very large solid type model – SSQMR, ATSQMR
Very large plate type model – SSSPM, ATSSPM
For spacecraft/radiation models
SNSOR (with user-specified DAMPD if necessary)
FWDBKL if thermo-stats are present
ATSDUF, SNTSM1 for most other cases
About the Time Step
1. The default computed time step (DTIMEU) = CSGMIN* CSGFAC. CSGMIN can be checked in the .sot file. If CSGFAC is not specified, it is internally set to 1.0. The DTIMEU in the .sot file has been truncated by the OUTPUT point.
2. In a normal sized model, CSGMIN usually is small enough for the time step which will assure a convergent transient run.
3. CSGFAC is used to adjust the time step. It is recommended to determine the best CSGFAC to the model while maintaining acceptable temperature errors.
4. If OUTPUT < CSGFAC*CSGMIN or OUTPUT < DTIMEI, then OUTPUT becomes the time step. All the OUTPUT points are automatically required to be calculated.
5. DTIMEI is the forced time step which will ignore any other factors. Sometimes it may lead to inaccurate answer if it is too large. DTIMEI does not affect the automatic time step solvers.
6. If the model size is very small, CSGMIN may be too big for the time step. A small CSGFAC or DTIMEI should be used to adjust the time step.
7. CSGFAC*CSGMIN or DTIMEI should be small enough to “catch” any details in time fields, temperature fields or orbital flux arrays.
The above are brief descriptions of the solvers, time step and accuracy. Please read the following chapter for more details.