I am trying to simulated a geometry were a cylindrical domain rotates inside another domain. I model the momentum transfer through the interface I have used MFR and also SM, however in both cases a non-physical behavior occurs at the interface. Has anyone face such a problem? Is there any suggestion?

What do you mean with non physical behaviour? Do you mean in the variable Velocity? This is the velocity relative to domain motion. indeed you will find a discontinuity. If you look to "Velocity in Stn frame", the discontinuity will dissappear...

Gert-Jan

www.bunova.nl

It actually happens with standard frame velocity. The system I am modelling is a moving t impeller inside a domain. With non physical behaviour I meant the following: Physically, the velocity monotonically should decreases from the tip of the impeller and be zero at the walls. This is what happen with the simulated results before interface is reached, However at radius values close to the interface (3 elements before it), velocity becomes constant. Three elements after it is still constant and afterwards it start to decrease again. thanks for your quick answer

Hmmm, can you share a figure?

Gert-Jan

Could you please send my an email to send you relevant pictures? Thanks

What's the type of interface you use? Frozen rotor, Stage, or transient rotor-stator?

In the result you have seen, I have used frozen rotor. although the use of sliding mesh ( transient rotor-stator) have not eliminated the problem