Ok, seems to be working now. I'm modelling a rotary hydrocyclone. Currently I'm looking at tangential inlet slots vs vanes to give the swirl component pre-rotating section. The simulation with the slots does not have this geometry issue and was very stable and converged well, the one with the vanes was not. I've attached two tangential velocity plots and the area of concern is at the far end by the cone, which I thought could be geometry related as they were meshed identically. Also, I noticed when I go to create the plane the centre is not at 0,0...also attached. Perhaps I should put these as my frame motion coordinates?