Hello,
I joined 3 meshes, made with Salome 5.1.4, in Code_Saturne 1.3.3. It is 3 cylinders of different sizes I want to join at the bottom and the top sides. I have defined group of faces for each cylinder at bottom and top side.
check_mesh -j command gave me what I expected. So I joined in the same way the meshes in the runcase, and preprocessing worked fine. But when I start a calculation, it seems that the groups of faces at the border are not merged. So I get boundary conditions errors for a lot of faces.
Should I have a part of certain group becoming fluid zone ?
For information, I use particular physics (MHD).
Thank you
Joining meshes : boundary conditions problem
Forum rules
Please read the forum usage recommendations before posting.
Please read the forum usage recommendations before posting.
-
Yvan Fournier
Re: Joining meshes : boundary conditions problem
Hello,
Particular physics should make no difference, but it would seem that the "-j" option you passed to check_mesh was not passed to the Preprocessor when using the script.
Did you check the "listpre" file to make sure the joining was done ?
If not, "-j" may not have been set in the correct place in the runcase.
Another possibility is that if you used "-j" with a "-group <g1> <g2> ... <gn>" sub-option, some of the faces are not in the correct group, and the joining is incomplete.
Best regards,
Yvan
Particular physics should make no difference, but it would seem that the "-j" option you passed to check_mesh was not passed to the Preprocessor when using the script.
Did you check the "listpre" file to make sure the joining was done ?
If not, "-j" may not have been set in the correct place in the runcase.
Another possibility is that if you used "-j" with a "-group <g1> <g2> ... <gn>" sub-option, some of the faces are not in the correct group, and the joining is incomplete.
Best regards,
Yvan
-
Alexandre Lebouvier
Re: Joining meshes : boundary conditions problem
Hi Yvan,
I attached my listpre. In the runcase, I defined "-j" as COMMAND_RC=-j
I didn't use -group sub-option. The point is : the geometrical group I try to merge have not the same size, so one part of the group is merged (I suppose) and the other part will stay a boundary. My question is : Do the groups I have to merge must have the same geometrical size ? Then in the code, I use geometrical condition on the group to separate the two parts of the group.
Thanks
I attached my listpre. In the runcase, I defined "-j" as COMMAND_RC=-j
I didn't use -group sub-option. The point is : the geometrical group I try to merge have not the same size, so one part of the group is merged (I suppose) and the other part will stay a boundary. My question is : Do the groups I have to merge must have the same geometrical size ? Then in the code, I use geometrical condition on the group to separate the two parts of the group.
Thanks
- Attachments
-
- listpre-09231129.txt
- (9.7 KiB) Downloaded 162 times
-
Yvan Fournier
Re: Joining meshes : boundary conditions problem
Hello,
No, groups do not need to be the same size; the part that is not joined remains as boundary faces, the rest becomes interior faces.
Do you know where you have boundary condition problems ? Some faces may be partially joined (split into an interior part and a boundary part), or you may be missing groups in your boundary condition definitions.
With Code_Saturne 2.0, you get a postprocessing output to show which faces have missing boundary conditions; with version 1.3, you just get face numbers...
Best regards,
Yvan
No, groups do not need to be the same size; the part that is not joined remains as boundary faces, the rest becomes interior faces.
Do you know where you have boundary condition problems ? Some faces may be partially joined (split into an interior part and a boundary part), or you may be missing groups in your boundary condition definitions.
With Code_Saturne 2.0, you get a postprocessing output to show which faces have missing boundary conditions; with version 1.3, you just get face numbers...
Best regards,
Yvan