[oskar]

Looking for meshing or solver settings help!


I have some local mesh issues (like to the left of the screenshot), but overall I think I'm getting the mesh to work ok.
However I am running into numeric instability with simpleFoam at cell size jump transitions. At least I'm assuming that's what's happening here.
Short of running a limiter on the absolute calculation values (which I don't think simpleFoam can do inside the solver?), I think my main remedy would be to smooth the cell size transition, but this is of course not really possible with hex cells.


Anyone here have suggestions on how to either tweak the solver settings, snappyHexMesh settings, or switch to a better tet based mesher, to get better stability?

[oskar]

Note to future self when I encounter this same issue again:
snappyHexMesh isn't really capable of generating higher quality meshes. Only a pure tet generator (at least for now) is capable of reducing cell volume jumps and providing a high quality mesh.



However, hex mesh simulation can be stabilized by adding nNonOrthogonalCorrectors (try about 3-6 for my typical case meshes) and reducing relaxation values to achieve stable min and max values in the zone.