I've drawn up a turbine in a section of pipe and am attempting to get the torque and damping characteristics at various flows and rotational velocities.  If the entire system is stationary, the simulations run fine, however, if I introduce a local rotating region, the simulation crashed before it is completed.  If I run it with low mesh resolution, it eventually crashes on a 'high mach number' problem.  With high resolution, it crashes after a single iteration. 

This happens at higher angular velocities (35 rads/s), lower (5 rad/s), and even if the region is stationary (0 rads/s).  I have tried this with out wall boundary conditions, as well as with the wall boundary conditions set as stationary stator walls.  I have also do this is higher flows 1 m/s, and lower flows (0.05 m/s), and with low and high pressure differentiation, and yet I constantly have the same failure. 

Has this been a common problem, is there a simple setting perhaps that I am missing that is causing this?  Or is this something more complex.  I can make more information available about my simulation set up, but am hoping that perhaps its something simple, because it makes no sense that runs if there is no rotating region, but crashes once I introduce a rotating region, even if angular velocity is set to 0 rads/s.

Thanks for your help,

James Kelly