Rotate frame reference convergence problem!
 

Hello, everyone!
recently, I set a domain with Rotate frame reference, and this domain has a interface with a stational domain. The interface frame change/mixing model is transient rotor stator, and i let the domain rotate in the stational domain. moreoover, this model is multiphase model with water and air, my simulation aims at observing the water movement influenced by the rotating domain.

I got a successful simulation result when the model is simple, but when there are two rotating frame domain in the stational domain, the surface of water moves abnormally. the water far away from the rotating area bumps on.

I think that is the convergence problem, but I have reduced my timstep to 1e-5, increased my convergence loops to 200, used shear stress transport turbulence model, and aslo used the first order upwind advection scheme, the water still bumps for no particular reason.

So i want to know, is my solving derection right, is that a convergence problem, or maybe some other reasons. this problem bothers me alot, please, please tell me your opinion, thank you very much!

This is most likely to a numerically unstable simulation. The most important thing you can do to help is to improve mesh quality - multiphase models need good mesh quality. Double precision numerics and smaller time steps also help. A tighter convergence criteria and a better initial condition are other things to try.