Repository navigation
Mesh Partitioning #1264
|
I am working on a calculation where I would like to easily partition different sections of the mesh to solve different problems. Looking at Ferrite-FEM/FerriteGmsh.jl#26, it doesn't appear that importing nodesets from Gmsh works yet. Has anyone else done something similar that I haven't been able to find yet? One (less-desirable) option is to write a code to determine which nodes need to have a separate local matrix and vector provided (since this can be done), but it would be more convenient if I could just label the mesh and then read in the elements that need to be treated differently. |
Replies: 2 comments 3 replies
|
Wouldn't it be simpler to work with a cell set instead of nodeset? See for example https://ferrite-fem.github.io/Ferrite.jl/stable/tutorials/porous_media/ |
|
This code snippet from a colleague (who probably used an older version of FerriteGmsh) worked at some point (at least) Ferrite-FEM/FerriteGmsh.jl#26. It was used to import nodesets for boundary conditions. I think the code should work, but it needs some polishing. Do you want to mark nodes in order to constrain the interface layers? |
Wouldn't it be simpler to work with a cell set instead of nodeset? See for example https://ferrite-fem.github.io/Ferrite.jl/stable/tutorials/porous_media/