Non-conformal meshes using cfMesh
 

Hello folks,

I am running a FSI (Fluid-Structure Interaction) case, which the fluid is confined inside the solid. There is plane of symmetry on the right hand side and this is a 2D case. I am using two very simple cfMesh dictionaries (the simplest in fact). Here they are:

Mesh:
https://www.dropbox.com/s/o2pmzigzek...lMesh.jpg?dl=0

https://www.dropbox.com/s/o2pmzigzek...lMesh.jpg?dl=0

I would like them to be conformal to with other at the interface. Does anyone know how to do it?

Ps: Using cfMesh 1.1.2 form official site https://sourceforge.net/projects/cfmesh/

Hi !

If you use two different geometries and two different meshing procedures, there is not much you can do about having them perfectly conformal during at mesh generation...

BUT if you give two identical refinements on two identical surfaces (with surfaceRefinement for instance), then the two meshes will be similar enough, and then you can conform them with stitchMesh.

Another option is to use a single mesh procedure, and then separate the two zones.

Good luck !

Thank you for your help!

I have looked at the documentation of stitchMesh and I could not see how to use in my case, since I need two separeted meshes in order to run the solver. If have understood correctly, after using it I would end up with a single mesh, am I wrong?

Are you talking about SplitMeshRegions? If yes, would the procedure be like this?
1) Generate the mesh for the whole domain(using cfMesh of course) and, using stl files, refine the internal boundaries.
2) Use SplitMeshRegions. But for my case I would have a problem with the symmetry boundary of the fluid. This one has two boundaries, FSI boundary and symmetry boundary. Using SplitMeshRegions I would end up with only one boundary.

Too much complicated :)
I think I'll return to this problem later. I thinking on another possibility, i.e., changing the code of cfMesh to support the use of a reference mesh created by previously cfMesh procedure. Something like this:
1) The same algorithm of cfMesh, but at the moment of the removing internal cells (in this case the fluid domain) the code would save this removed cell into file A.
2) Run for cfMesh for the fluid, but now use a base mesh specified by A.