Issue with Neumann boundary conditions on .inp mesh created with gmsh #18499
-
Hi, I have a mesh created with gmsh with a pretty simple geometry, but I'm having trouble using it with Neumann BCs. The above is the raw/unedited .inp export from gmsh- when I used this I got the error:
So I manually deleted all the line and surface elements, and left only the volume element. I applied Dirichlet conditions on one of the "physical groups" I created, which is a surface of one side, and it ran fine. The problem I'm facing now is applying Neumann condition to the same surface- I get:
So I added that argument to the Mesh block, and nothing changed. I also thought it could be the element sets created by gmsh, and since I deleted all the surface elements, I should delete the surface elsets too (and leave only the nodesets). Doing that gave me the error:
I also tried remaking the mesh and instead of defining the surface as the boundary, I'd define the lines that make up the surface as the boundary (in the attached file, these are named "left_tract" and "right_tract"). I did that and tried in MOOSE but just got a segmentation fault with no backtrace. |
Beta Was this translation helpful? Give feedback.
Replies: 2 comments 4 replies
-
@richmondodufisan you can attach files here directly without resorting to sketchy looking download sites... See here: dike_mesh.inp.gz It would be more useful to see you gmsh script that generated that file. |
Beta Was this translation helpful? Give feedback.
-
Dear @richmondodufisan I paste below a simple 3D .geo file with definitions of physical surface. You can compare with your .geo file and see if there are differences. Hope this helps. Best Regards,
|
Beta Was this translation helpful? Give feedback.
Dear @richmondodufisan
I paste below a simple 3D .geo file with definitions of physical surface.
This was working for me some time ago and I could define the boundary conditions in the moose input file
by referring to the names of the physical surfaces, such as "bottom", "top", et cetera.
You can compare with your .geo file and see if there are differences.
Or maybe the problem is your gmsh version, not all of versions generate the same .msh files.
Hope this helps.
Best Regards,
Nicolò