[peterchen]

HiI'm simulating the flow between 2 rotating cylinders. I use the rng k-epsilon-model and defined a velocity inlet and pressure outlet, the Fluid is defined as water. The wall is defined as a moving wall, but if I increase the rotation to 2000rpm or higher,the swirl velocity drops to almost zero between the two cylinders or the residuals remain high and oscillating (maybe because I have reversed flow on the outlet?)The problem occurs in 2D and 3D.Is there a possibility to stop the reversed flow? I already tried a different mesh or slowly decreasing the under-relaxation factors.regards,peter