[Daniel Frederik]

Hello Fluiders,

I simulated 3 Domains as shown in the picture. i know it's kinda funny cause it doesnt simulate anything useful. Anyways the problem is interesting. The two Domains of both ends are stationnary, the one in the middle rotaes. Now actually the mesh is rotating but since there is nothing in that middle section that could make the fluid rotate it is quite interesting to see that the streamlines show a rotation in the middle section. Does anyone have an idea what that is due to?

[Opaque]

What boundary condition for velocity are you using on the outer wall ?

No Slip in the rotating frame, or in the stationary frame ?

[Daniel Frederik]

I'm using No Slip on both. Maybe it is some sort of a local coordinate system that makes it just appear that way.

[Opaque]

That seems to be the problem.

If you are in the rotating frame, a stationary wall counter rotates; otherwise, you are imposing a rotating wall in the stationary frame.

• No Slip stationary wall in the rotating frame = No Slip Rotating wall in the stationary frame

• No Slip counter-rotating wall in the rotating frame = No Slip stationary wall in the stationary frame.

Either will be consistent with the walls in the upstream and downstream domains, and you should not see the rotation in the stream lines (barring discretization errors)

[Daniel Frederik]

Thank you for the suggestion. I tryed so but had no succes. I doubt that it has to do with the walls. The domains are in different referential frames, a rotating wall would just affect the boundary layer not the whole fluid.
But why would cfx post not show the movement of the fluid in the absolute coordinate system?

[Daniel Frederik]

Here the same problem without rotating walls, only a rotating domain Inside a stationary.
The pictures show the streamlines of each domain. There again in the rotating domain the streamlines are not straight. Why??
If you for instance want to simulate a Wind turbine you would put that one into the rotating domain, and the turbine would affect the fluid flow, but since there is no obstacle in this case...

[ghorrocks]

Don't forget you need to plot velocity in stn frame, not just velocity.

[Daniel Frederik]

Thank you very much!