[TReviol]

Hello Forum,

I try to simulate a Waterpump with MRFSimpleFoam. I am using OF in different versions: OF1.6-ext and OF2.1.1.
With OF1.6-ext, the Simulation with the kEpsilon Model converges fine and the results are absolutly plausible. In OF2.1.1 (same models, same solvers, same schemes, but differences in interfaces), the convergence is plausible until a point, the convergence rises suddently. I guesed an instability Problem an checked the variables: at the AMIs the value of epsilon becomes a singularity. I tried to modify the tolerance, but without any success. The Mesh isn't realy appropriate: the volumechange is to high, but 1.6-ext doesn't fail, why does OF2.1.1?
Are there experiences with such problems? Does anyone know the solution for this problem?

Thanks for the replies.
Best regards.

Thomas

[TReviol]

Hello Forum,

in the meantime, I found the solution for my problem. I optimized several aspects (modifying fvSchemes, fvSolution and remeshing the domains) but finally, a mistake in the definition of the frozen rotor was the reason:

In OF-1.6-ext, I never declared the ggi-patches as nonRotatingPatches in the file MRFZones, and never any problems occured. But in OF-2.1.1 it seems to be necessary to define the cyclicAMI-patches as nonRotatingPatches in MRFZones.

This thread gave me very helpfull hints. Now the solution of OF-1.6-ext and OF-2.1.1 seems to be nearly identical :-)

regards

Thomas