[sai193]

I am trying to model a rotor in steady state using a mrf, but the mrf-zone intersects other parts of the geometry because of their proximity with the rotor.

To get around this I was wondering if overset meshing can be used.

In my case (looking at the rotor in isolation), I have this cylindrical overset-zone, with an MRF applied to it (P1).

Mapping background mesh (P2)
Mapping rotor/front mesh (P3)

The overSimpleFoam simulation seems to run fine but the forces are way off (the viscous forces predictions especially are too big).

Does anyone know if this is a feasible approach?

P1.jpg

P2.jpg

P3.jpg

[ozi]

Did you able to find a solution?

Originally Posted by sai193

I am trying to model a rotor in steady state using a mrf, but the mrf-zone intersects other parts of the geometry because of their proximity with the rotor.

To get around this I was wondering if overset meshing can be used.

In my case (looking at the rotor in isolation), I have this cylindrical overset-zone, with an MRF applied to it (P1).

Mapping background mesh (P2)
Mapping rotor/front mesh (P3)

The overSimpleFoam simulation seems to run fine but the forces are way off (the viscous forces predictions especially are too big).

Does anyone know if this is a feasible approach?

Attachment 80323

Attachment 80324

Attachment 80325

[<nico>]

Hello,

I have a similar issue where I want to apply overset meshing with an MRF or basically I want to get a steady state solution.

In my case I simulate a rotor-stator system. Because of the tools are like „merged“ into each other, it is not easily possible to use a simple MRF approach with simpleFoam. From the tutorials I was able to run an unsteady simulation with overPimpleFoam. Of course this takes a long time which is why I want to use overSimpleFoam.

My solver collapses already after 2 iterations. What I can see is, that my time step continuity error rapidly increases and my residuals of my pressure immediately go down to 10e-5.

In my boundaries I work with noSlip for my velocity conditions on the rotor. Due to MRF this is correct right?
I also tried with movingWall. Here the solver was running but with unrealistic high force values. But it didn’t crash.

Do I have to consider any special settings when working with MRF and overset?

Thank you and best regards,
Nico

[JBeilke]

There is never a need to use overset for pumps etc. It is always possible to have a rotating part and a fixed part without overlap.

[<nico>]

That’s true. But especially for a rotor-stator application where you not only have the rotor but the stator as another tool, you have to define a quite complex MRF zone. You cannot use an easy geometry like a cylinder if I am correct. Or how would you proceed here?

I am not an expert in geometry design which is why I first try to use easier geometries. Of course, if this is the only way, then I need to create this zone as an stl file.

But even if it is not the best way, can I use an overset mesh with MRF?

Thank you for supporting!

[JBeilke]

I would use my CAD program and create the interface there. It can be a little bit tricky when you do it the first time but overset is no option you should consider.

[JBeilke]

Originally Posted by <nico>

But even if it is not the best way, can I use an overset mesh with MRF?

MRF alone is something that might or might not give you a useful result. You never know in advance for a new configuration. Thats why I only use MRF to create an initial solution for a transient simulation with rotating mesh.

MRF + overset ==> no idea

[<nico>]

Ok thank you! Then I will try and go ahead with the CAD design.