I'm working on a 2D pure radial centrifugal compressor inviscid simulation, to get a good initialization for a turbulent simulation. Initializing the solution from the inlet I had convergence troubles, also adopting Courant 0.1 and a 9 multigrid level. The residuals never go down 10-1 and I'm using a single moving reference frame cause I'm going to simulate just the rotor. Just after initialization I see that the relative velocity vectors at the outlet are to blent (I expected a normal-to-outlet exit), causing very soon reverse flow problems. I also move the outlet far from the blade (half of the chord). Is it possible that the problem is a wrong initialization and in this case how can I choose a better initialization? Thank you very much for the attention.
|All times are GMT -4. The time now is 22:15.|