[Oxmox]

... I'm working on a problem about high pressure homogenization.
the calculation is run pressure based, incompressible, single phase and time-steady with the k-epsilon- or k-omega-model in 3D, but even after 50.000 iterations the solution does not converge, the residuals just keep zig-zagging and there's occasionally reversed flow in the pressure outlet. The problem seems to be the turbulent area behind the (simply circular) valve where the residuals are the biggest, there you can also see, that although the geometry is perfectly axially symmetrical, the flow is'nt... what I'd like to know is what might be the problem and how I could possibly resolve it, thanks in advance...