In my simulation, I have gas being injected tangentially into a cylinder full of water and I want to observe the free surface after steady state is reached. I've used the "Free Surface" option in Pro-am, but the calculation either stops due to divergence or the end result doesn't look what I expect for steady state. Any suggestions on which settings to change to improve the Free Surface function? Thanks in advance.

It must be transient. Example like you (1636KB): http://www3.msiu.ru/~murash-v/widsh.avi

Is the initial volume of gas conserved in this example?

It does not seem to be. This happens in comet as well, there is a loss.

Have you done a graph plot total mass of gas versus time?

volume or mass? I have compressible gas

The volume of gas isn't conserved. I realized that I made a mistake in my original description. The cylinder is not completely full of water, but mostly (maybe 80%). There is an outlet that the two phase mixture can leave through.

Grid refinement & time step.

try also to use an UD scheme for the sim...you can restart later with a MARS scheme if you want to be correct with the results... And try also to initialize the sim via initfi.f, not by auto-initialization...if you can do that you'd probably get a more "stable" simulation

Thanks everyone.