Unsteady modelling - AMG divergence + acoustics

Neville

Hi all

I have been trying to do an unsteady analysis or a fairly complicated 3d geometry.. As I want to use the Acoustics module, I can only use either the LES, DNS or RANS turbulence models.

The main problem is that is I initialise the problem, then I always have the problem that it will not iterate dur to AMG dtecting divergence. I have yet found a solution for this problem.

When I do not initialise (i.e. I start of from ther solution of a steady solution) I do not have the above problem, it iterates and converges well for the first 2-3 time steps but after that it diverges badly and has no hope of converging - dire starits. I have tried LES and RANS with no success so far (I had the problem using K-epsilon)

If anybody could help I would appreciate it - also if anybody knows anything about the acoustics module could you please reply (having difficulty sice begining of the year )

Thanks Neville