How to obtain the grid and geometry files for the T Junction tutorial

[Debashis Basu]

Thank You Very Much.
Debashis

[Debashis Basu]

Hello,
I am trying to start the tutorial problem on T-JUNCTION. Which is the grid to be used ? Is it
Noyau/ncs-1.3.3/doc/TUTORIAL/TEST_CASES/STRATIFIED_JUNCTION/MAILLAGE/sn_total.des
or as explained in the tutorial (Version 1.3.2 tutorial)
downcomer.des.
The downcomer.des is located in
 Noyau/ncs-1.3.3/doc/TUTORIAL/TEST_CASES/FULL_DOMAIN/MAILLAGE/downcomer.des
I have created the T_JUNCTION directory using
cree_sat -etude T_JUNCTION CAS1
There is a MAILLAGE directory without any grid file. I need to copy the grid file. Kindly let me know which is the correct grid file for this T_JUNCTION problem ? According to the tutorial it is downcomer.des.
Thanks and Best Regards,
Debashis
 
 
Previously Yvan Fournier wrote:

Hello,
When you do move to GAMBIT, choose the "GAMBIT neutral" format when you export the mesh (avoid .unv, which does not allow pyramids, or .msh, which is the FLUENT mesh format, different from Gmsh's .msh format which Code_Saturne handles).
Boundary conditions defined in GAMBIT will appear as groups in Code_Saturne, so you will need to re-specify the types to Code_Saturne, but the names should stay the same (you will then use names in the same way you use numbers with the T-junction tutorial meshes).
Also, I would recommend avoiding white spaces in any boundary condition/group name : Code_Saturne can handle them by enclosing the string in quotes or using escape "\" characters before spaces, but it makes things unnecessarily complicated.
Best regards,
  Yvan

[Debashis Basu]

Hello,
I am trying to start the tutorial problem on T-JUNCTION. Which is the grid to be used ? Is it
Noyau/ncs-1.3.3/doc/TUTORIAL/TEST_CASES/STRATIFIED_JUNCTION/MAILLAGE/sn_total.des
or as explained in the tutorial (Version 1.3.2 tutorial)
downcomer.des.
The downcomer.des is located in
 Noyau/ncs-1.3.3/doc/TUTORIAL/TEST_CASES/FULL_DOMAIN/MAILLAGE/downcomer.des
I have created the T_JUNCTION directory using
cree_sat -etude T_JUNCTION CAS1
There is a MAILLAGE directory without any grid file. I need to copy the grid file. Kindly let me know which is the correct grid file for this T_JUNCTION problem ? According to the tutorial it is downcomer.des.
Thanks and Best Regards,
 
Debashis

[Yvan Fournier]

Hello,
I would suspect downcomer.des is a leftover from another example, and trust the files more than the text.
Short answer: use sn_total.des
Best regards,
  Yvan

[Debashis Basu]

Thank You. But will the zone names and boundary conditions match as explained in the tutorial.
Could you please suggest some tutorial problems where instructions (authetnticated text) and the mesh are also available ?
That would be really helpful.
One problem from start to finish with instructions step by step would be really helpful for me.
Hopefully, the zone names and boundary conditions will match for sn_total.des .
Regards,
Debashis
Previously Yvan Fournier wrote:

Hello,
I would suspect downcomer.des is a leftover from another example, and trust the files more than the text.
Short answer: use sn_total.des
Best regards,
  Yvan

[Debashis Basu]

Hello Yvan,
I applied the command ecs -sc -m sn_total.des -ensight
It generated the following files
chr.geo
CHR.case
Which one of this I should use to see the grid.
Kindly let me know.
Thanks and Best Regards,
Debashis
 
 
Previously Yvan Fournier wrote:

Hello,
The only current tool I know of which can handle .des files (besides Code_Saturne) is Simail, the (old) mesh generator used to build these meshes. You may convert them to a postprocessing output using the "ecs" preprocessor, with a command such as "ecs -sc -m downcomer.des -ensight" to see one mesh at a time, or "ecs -sc -m *.des -join -ensight" to see all, but if you are not already very familiar with the code,I would recommend usign the GUI's mesh checking option to handles this for you.
Do not forget to activate mesh joining inthe GUI as you have 3 meshes (the "-join" in the command-line above is for this).
I know TECPLOT may import EnSight Gold files, but I don't know how good at it it is (can it handle the polyhedral cells that result from mesh joining, and which do not appear inthe preprocessor-generated output, but do appear in the Kernel-generated output ?). If you do try this, do tell us if it works well.
Otherwise, both EnSight and ParaView work fine (use the "extract blocks filter to separate pieces of the mesh in ParaView; in EnSight, parts appear in a tree, so they are obvious).
Best regards,
  Yvan

[Yvan Fournier]

Hello,

Alexandre answered in another thread, but the short answer is CHR.case (which is an index linking to chr.geo).

In any case, the file open menus from EnSight, Visit, or ParaView probably only suggest the "correct" option, so in case of doubt, all you need to do is try both.

Best regards,

  Yvan

[Debashis Basu]

Thanks a lot. I have another question. Does grids generated in ANSYS workbench can be used in SATURNE ? Kindly let me know. Also what is the significance of the "localization" parameter in boundary condition ? How do I know what is the localization number for any geometry.

Best regards,

Debashis

Previously Yvan Fournier wrote:

Hello,

Alexandre answered in another thread, but the short answer is CHR.case (which is an index linking to chr.geo).

In any case, the file open menus from EnSight, Visit, or ParaView probably only suggest the "correct" option, so in case of doubt, all you need to do is try both.

Best regards,
 Yvan

[Yvan Fournier]

Hello,

Meshes generated in ANSYS workbench may be used in Code_Saturne if they are exported in CGNS format
(unv is OK if you do not have Pyramids, but CGNS is preferred).

Be careful if you use ANSYS workbench 13: its CGNS export seems to produce CGNS-3 (still in beta) files instead of CGNS-2.5 files, even when CGNS 2.5 is requested. This works and has been tested with Code_Saturne 2.0.1 (to be released in January) if built with CGNS 3, but won't work with prior versions (CGNS-3 broke some compatibility to add new eleemnt types).

The "localization" should read "location", and is described in the "selection" section of the user documentation (p19 or whereabouts if I remember correctly). In this tutorial example, boundary face groupes have numbers instead of names due to the format which is used.

Best regards,

  Yvan

[Debashis Basu]

Thank You. I am currently using SATURNE version 1.3.3. If I use ANSYS workbench for grid generation, what CGNS format should I save the grid ? Kindly let me know. Also is there is any document / tutorial available that shows how you label the grid different zones in SATURNE. It is not very clear in the tutorial.

I am able to do post process in TECPLOT 360 (2009). I ran the stratified junction tutorial case and saved the result in ENSIGHT format. TECPLOT 360 (2009) can load ENSIGHT format files. I was able to plot the results in TECPLOT. I am attaching the figures that I got for certain variables and also the grid. The grid was of course provided in the tutorial case.

If I can get some clear idea about mesh generation in ANSYS workbench and setting up the zones and boundary conditions, it will be good for me.

Best Regards,

Sincerely,

Debashis Basu
 
Previously Yvan Fournier wrote:

Hello,
Meshes generated in ANSYS workbench may be used in Code_Saturne if they are exported in CGNS format
(unv is OK if you do not have Pyramids, but CGNS is preferred).

Be careful if you use ANSYS workbench 13: its CGNS export seems to produce CGNS-3 (still in beta) files instead of CGNS-2.5 files, even when CGNS 2.5 is requested. This works and has been tested with Code_Saturne 2.0.1 (to be released in January) if built with CGNS 3, but won't work with prior versions (CGNS-3 broke some compatibility to add new eleemnt types).

The "localization" should read "location", and is described in the "selection" section of the user documentation (p19 or whereabouts if I remember correctly). In this tutorial example, boundary face groupes have numbers instead of names due to the format which is used.

Best regards,

  Yvan