[Jun]

Hi there:

I am trying to simulate a rotating cruciform shape basket in a cylinder shape container. Please see the attachment for more details.

I have trouble to make right choice for the boundary.

First, I try to model the cylinder liquid body as rotating frame and set the basket surface as wall boundary. The problem was no converge was achieved. The domain imbalance is always high (~ 200%)

Second, I try to model the basket as immersed solid and the model converged after around 40 iterations (RMS 1e-4), but the velocity of liquid in the cylinder does not make sense: almost still everywhere, except a little bit speed on the tips of the basket.

I think the basket is too thick to be treated as blade, is that correct?

One possible reason for the converging is that I have very fine prism layers at the basket surface: total height 0.3; layers 50; initial layer height ~ 1e-6. The fine prism layers are needed to capure the very thin boundary layer (2e-5 thickness).

Please help me with a right direction to model this thing.

I appreciate your feedback.

Jun