interFoam with irregular Mesh
 

Hallo,

Is it possible to use interFoam with a tetrahedral mesh or is it only designed for hexahedral meshes.

I can get a simple case to run with tetra, but it crashes after a while. With the same geometry and input conditions, but with a hex mesh it works fine.

All the examples and other work I see of interFoam is with hex meshes.

If someone can please shed some light I would appreciate it.

Kind Regards
Luther

Hi Luther

From the point of view of OpenFOAM, I cannot see any limitations in interFoam, hence it must be a problem with your setup, which makes it unstable with tets.
If you could describe your problem in more details, e.g. with figures, I might be able to help.

Best regards,

Niels

Hi Luther,

also from my experience interFOAM works much better with hex-meshes than with tet-meshes. I think one reason is that the element growth rate in tet meshes can be rather high. So I made better experience with tet meshes providing a smoother transition (propagation rate). But this may result in rather large meshes...

Ulrich

Thank you Niels,

I will have a thorough look at my setup again and if the problem stil persists I will mail you the case and you can see if you find anything wrong.

Kind Regards
Luther

Hi Luther,

I met the same problem when using the interFoam solver. The result was ok with hex-meshes. When I used tet-meshes for the same case, the courant number became bigger and bigger, then the calculation stopped.

Maybe, it is caused by the non-uniform tetrahedral mesh. I will pay more attention to it.

If there are any new developments, please let me know.

Kind Regards
Jitao

Hi Luther,

I use almost every time interFoam for my work. At first, I tried it only with hex mesh, and had no issue.
Then I tried the same cases with tet mesh. At first it crashed, but since I pay more attention in meshing, I do not face this kind of problems. So be carefull in meshing and there is no reason your simulation crashes because of mesh (I think...)

Regards

Julien

My experience with non-orthogonal grids is to make sure you use corrected schemes!

You can also try to increase the number of non orthogonal corrector loops (nOrthogonalCorrectors).

Don't waste time with tetrahedral grids for the interFoam class solvers by playing with correctors, schemes and etc. I wasted weeks trying to get things to work with tets and results were far from satisfactory = the effort is not worth it. Just go with hexes if you can.

Thank you. I have the same experience. Do you perhaps know about any literature where I can read up on the cause of this phenomenon?

I have posted some literature reference on this problem here:
http://www.cfd-online.com/Forums/ope...-than-tet.html
It's post #9 ...