Mesh Refinement - Code Saturne
- Robin Borland
- Topic Author
- Offline
- Senior Member
-
Less
More
- Posts: 48
- Thank you received: 0
16 years 11 months ago #2025
by Robin Borland
Mesh Refinement - Code Saturne was created by Robin Borland
I have a question on mesh refinement, I am trying to use Code Saturne to look at the air flow around a building. When I build the geometry and perform a mesh on the solid part, Saturne solves without any problems. To get mesh refinement, I have been trying to build a 'compound' in the geometry module, and then create a sub mesh. Any time that I have tried this, I get errors from the Saturne solver eg.
@ @@ ATTENTION PHASE 1, FACE 110289, PROP.No1 0, TYPE CL 0
@ @@ ATTENTION PHASE 1, FACE 110300, PROP.No1 0, TYPE CL 0
@ @@ ATTENTION PHASE 1, FACE 110306, PROP.No1 0, TYPE CL 0
@ @@ ATTENTION PHASE 1, FACE 110315, PROP.No1 0, TYPE CL 0
@ @@ ATTENTION PHASE 1, FACE 110324, PROP.No1 0, TYPE CL 0
@
@
@
@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
@
@ @@ ATTENTION : ARRET LORS DE LA VERIFICATION DES COND. LIM.
@ *********
@ CONDITIONS AUX LIMITES INCORRECTES OU INCOMPLETES
@
@ La valeur du type ITYPFB des conditions aux limites doit
@ etre superieure ou egale a 1
@ et inferieure ou egale a 200
@ Une ou plusieurs erreurs sont listees ci-dessus.
@
@ Le calcul ne sera pas execute.
@
@ Verifier le codage du sous-programme de definition des
@ aux limites.
@
I have tried to model/mesh several different ways, but any time that I mesh a compound the error occurs.
What is the preferred method of obtaining mesh refinement in Code Saturne when using the Salome Interface ?
See the attached example - I have deleted the mesh to save space - when the geometry is not a compound, I can get a solution - when I use a compound in order to get mesh refinement, I get the errors. I need the refinement
Thanks
Robin
PS I sent an email to Code Saturne support a few days ago and did not get a response - is the support email adress active ?
@ @@ ATTENTION PHASE 1, FACE 110289, PROP.No1 0, TYPE CL 0
@ @@ ATTENTION PHASE 1, FACE 110300, PROP.No1 0, TYPE CL 0
@ @@ ATTENTION PHASE 1, FACE 110306, PROP.No1 0, TYPE CL 0
@ @@ ATTENTION PHASE 1, FACE 110315, PROP.No1 0, TYPE CL 0
@ @@ ATTENTION PHASE 1, FACE 110324, PROP.No1 0, TYPE CL 0
@
@
@
@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
@
@ @@ ATTENTION : ARRET LORS DE LA VERIFICATION DES COND. LIM.
@ *********
@ CONDITIONS AUX LIMITES INCORRECTES OU INCOMPLETES
@
@ La valeur du type ITYPFB des conditions aux limites doit
@ etre superieure ou egale a 1
@ et inferieure ou egale a 200
@ Une ou plusieurs erreurs sont listees ci-dessus.
@
@ Le calcul ne sera pas execute.
@
@ Verifier le codage du sous-programme de definition des
@ aux limites.
@
I have tried to model/mesh several different ways, but any time that I mesh a compound the error occurs.
What is the preferred method of obtaining mesh refinement in Code Saturne when using the Salome Interface ?
See the attached example - I have deleted the mesh to save space - when the geometry is not a compound, I can get a solution - when I use a compound in order to get mesh refinement, I get the errors. I need the refinement
Thanks
Robin
PS I sent an email to Code Saturne support a few days ago and did not get a response - is the support email adress active ?
Attachment bldg_test_nomesh.bz2 not found
- Robin Borland
- Topic Author
- Offline
- Senior Member
-
Less
More
- Posts: 48
- Thank you received: 0
16 years 11 months ago #2035
by Robin Borland
Replied by Robin Borland on topic Re:Mesh Refinement - Code Saturne
Here is an update on this:-
Code Saturne support did respond to my question, I have attached a copy of the response.
Apparently, when you create a volume sub mesh for refinement, Salome will add surface elements around it. Before you export the mesh to a MED file, these surface elements need to be isolated and deleted. There are two ways to do this:
1. In the geometry module create a volume group for the refined volume. Then create a surface group, based on this volume, call it "internal_surfaces". Back in the mesh module, crete an element group based on these surfaces, then view them and delete them.
OR
2. Create a group of all surfaces in the model in the geometry module. Create a mesh group based on this, then use the mesh subtract operations, to subtract the "walls, inlet outlet, symm" mesh groups from it. You should get the same result as in "1".
After doing this, I still did have errors due to the fact that my refined group lies on the symmetry plane. When I ran the pre-processor per the Code Saturne support email, the face on the symmetry plane corresponding to the refinement got removed. When I get more experienced, I will figure out a way to resolve this.
To resolve I re-modeled with the refinement volume totally encapsulated within the symmetry face - which works. I also ran the model without symmetry - this also works, since I can isolate and delete the internal faces.
If anyone figures out how to resolve the symmetry issue, or uses the geometric selection criteria as mentioned in the support email, I would be interested in getting details.
Code Saturne support did respond to my question, I have attached a copy of the response.
Apparently, when you create a volume sub mesh for refinement, Salome will add surface elements around it. Before you export the mesh to a MED file, these surface elements need to be isolated and deleted. There are two ways to do this:
1. In the geometry module create a volume group for the refined volume. Then create a surface group, based on this volume, call it "internal_surfaces". Back in the mesh module, crete an element group based on these surfaces, then view them and delete them.
OR
2. Create a group of all surfaces in the model in the geometry module. Create a mesh group based on this, then use the mesh subtract operations, to subtract the "walls, inlet outlet, symm" mesh groups from it. You should get the same result as in "1".
After doing this, I still did have errors due to the fact that my refined group lies on the symmetry plane. When I ran the pre-processor per the Code Saturne support email, the face on the symmetry plane corresponding to the refinement got removed. When I get more experienced, I will figure out a way to resolve this.
To resolve I re-modeled with the refinement volume totally encapsulated within the symmetry face - which works. I also ran the model without symmetry - this also works, since I can isolate and delete the internal faces.
If anyone figures out how to resolve the symmetry issue, or uses the geometric selection criteria as mentioned in the support email, I would be interested in getting details.
Attachment R_f____Saturne_Error_Message.zip not found
Moderators: catux
Time to create page: 0.135 seconds