Convergence of Continuity

Dodi

Hi everyone!

I'm trying to model 3D flow with a sudden contraction. Initailly I kept on getting error messages "Error: divergence detected in AMG solver: pressure correction" but after making my structured grid more uniform the problem dissapeared, however therefore the continuity residual doesn't want to converge below 3.6 e-10 (the other residuals are doing fine). It seems that increasing the mesh density lowers the value where the continuity flattens off but I cant go any denser anymore as otherwise I'll get a too small Y+ value (am using the standard wall function). Here's a list of things that I already tried to do: - change pressure-velocity coupling to coupled & simplec - lowering the under relaxation factors - changing the discretisation to 2nd order schemes

Non of these steps however lowered my continuity residual - does anyone have an idea why that is?? Or what I can do?

Thanks for any help! Dorit