Hello,
is there really noone who can give me a little hint?

[gschaider]

Have you taken fvSolution/fvSchemes from a simpleFoam-tutorial-case? One favourite for diverging turbulence with steady-state-solvers (at least in my experience) is having another time-discretization than steadyState in fvSchemes.

But that is just a guess. Have you tried laminar on that mesh (just to rule out that this is a turbulence-probleme)?

Hello Bernhard,

thx for your answer. In fvSchemes I have steadystate for ddtSchemes, I think this is all which is important, isnt it?
I tried to run it as laminar, but it still didnt converge. After 409 iterations I got:

time step continuity errors : sum local = 2.58644e+26, global = 2.68736e+24, cumulative = 1.23778e+25

I assume the problem is the mesh, as it got very high values in skewness. So is it possible, that because of this high skewness, the mesh isnt able to converge in OpenFOAM?

Hello again,

sorry, I was wrong. Now I set up some NonOrthogonalCorrections and it looks like it converges. After 100 iterations the screen in paraFoam looks reasonable in some way. In the former computations with turbulence at iteration 100 I had strange values for p and U. I will check the next iterationsteps to evaluate the results.

Ok,

looks like without turbulence it is fine. Does that mean, that the mesh is proven to be able to converge and the problems are only based on the setup of turbulence?

Thanks in advance