[sasanghomi]

Dear friends,

Do you have any reference which talks about the position of frozen rotor interface? (Rotating Domain Thickness)
Should it be away from the rotating boundary?

Best Regards

[AtoHM]

The interface is a 2d region. It can be a plane where the two domains are connected. What do you mean by "Rotating Domain Thickness"?

[sasanghomi]

I did not ask about the definition of interface. Let me clarify my issue;
When it comes to the simulation of a fan or a rotating equipment, one of the methods is using MRF (Multi Reference Frame) in order to simulate the rotating domain and modeling its effects on the stationary domain. Moreover, Frozen Rotor is a type of interface that can be applied between a stationary and a rotating domain. It is responsible for transferring the momentum from rotating domain to the stationary one.
Now, my question in that how far the interface should be placed? Should it be close to the rotating domain or not? Why?

Please look at the picture attached.

BR

[Opaque]

Unlike the mixing plane model, where the interface is averaged, the frozen rotor interface is effectively a direct connection but with a frame change.

If your sketch is representative of your model, what is the need to use a frozen rotor since you should be able to model the whole within a rotating frame, and account for stationary walls using a boundary condition?

Keep in mind the frozen rotor model is a crude assumption if the interaction between the solids in the rotating domain, and the those in the stationary domain is of any importance.

In any case, you should be able to assess the effect of the location by meshing the model with the interface at different locations and compare the results. That is, verify that your solution is independent of the interface location (a mandatory task for any turbo related CFD calculation)

[sasanghomi]

Thank you for your response.

1) Is it possible to simulate a fan with a casing without having a stationary frame? (Similar to the picture attached)

2) I have heard that Frozen Rotor interface should NOT be placed close to the rotor and it should be outside the boundary layer. Do you agree with this idea?

3) Honestly, my real case is the simulation of the fluid flow inside an electrical generator and the rotor is rotating at 3000 RPM and the stator is stationary. There are flow paths from rotor surface that goes towards the stator. I am going to place the frozen rotor interface in the middle of the air gap.(the distance between rotor and stator)
I think using the sliding mesh is so time-consuming.

[Opaque]

Such geometry does not satisfy the requirement for rotating domains: boundary must be a surface of revolution.

Have you considered the mixing plane approach?

I usually stay away from the Frozen Rotor approximation. The wake behind the blades propagates too far into the downstream component which is unrealistic.

[sasanghomi]

Thank you so much for your attention.
One question;
When I use Mixed-Plane (Stage), my simulation faces an error when the interface has a face perpendicular to the axis of rotation.
I am just curious to know that what the reason is behind that? Is there any problem relevant to the averaging process?
If I rotate the spline shown in the attached picture, do you think that the result is a surface of revolution?