Convergence Problems

James

I am run a Eulerian Multiphase with a K-epsilon mixture turbulence model using the default discretization schemes for the Eulerian model and the solution will not converge. The model produces patches of high velocity that appear and move around the model which I can't get rid of, however the phase seem to be what would be expected. So far I have tried reducing the under-relaxation factors, a number of grid adaptations, using a second order scheme for volume fraction and change the outlet conditions from an outflow to a pressure outlet. None of which had any effect so I was wondering if anyone encounter such a problem and what they did about it.

Many Thanks

James