How to handle the interface between rotor and stator?

MarcoC501 · May 12, 2023, 12:49

Dear all,

I have recently used OpenFOAM (v2012) for simulating the Rotor37, and the results with rhoSimpleFoam are similar to those from the literature and from CFX (ANSYS).

I am now trying to add up the stator and I would like to have some advice on how I should handle the rotor-stator interface.

From what I have seen there's no available package for the mixing-plane option (for steady state RANS simulations), do you confirm that?

For this reason I thought about using rhoPimpleDyMFoam, so directly go unsteady and introduce a moving mesh. The main problem here is related to the matching of the interface: should the two patches of interface between rotor and stator (set with cyclicAMI) match perfectly? For instance, in CFX the pitch ratio can be in the ratio 0.6-1.5, but is this still valid for OpenFOAM?

Moreover, feel free to suggest any example of simulations of this type from the literature (I could not find much related to the OpenFOAM I am using). I would really appreciate that.

Thanks a lot
Best,
Marco

MarcoC501 · May 17, 2023, 05:42

Dear all,

I have figured it out.
I installed foam-extend 5.0 and here there are some routines capable of dealing with the interface problem.

I have been trying both mixingPlaneand ggiinterface, using steadyCompressibleMRFFoam and steadyUniversalMRFFoam.

Is rhoSimpleFoam supported for these two routines?
That's because the convergence is not so good with the PISO algorithm used in the tutorial (e.g. pressure residuals do not go down), at the limit of divergence, and I do not know if that's a problem of solver or BCs.

The stage is transonic, do you maybe have suggestions on what solver to use for foam-extend?

I really appreciate your help and support!

Rahmat1 · May 17, 2023, 06:57

Marco,
I did model an axial fan installed in a wind tunnel using simplefoam and results were far as expected. Pressure across the fan was negative instead of a gain. Is it possible to share your model? Thanks.

MarcoC501 · May 17, 2023, 07:06

Which version of OpenFOAM are you using?
There should be some tutorials on that in the $FOAM_tutorials folder.
May be related to your model?

For me, it is PIMPLE algorithm (in steadyCompressibleMRFFoam) is diverging. I also have looking at the BCs, they are the same as the ones I set when using rhoSimpleFoam (with slight modifications)...

May 12, 2023, 12:49	How to handle the interface between rotor and stator?	#1
MarcoC501 Member MC Join Date: Apr 2021 Posts: 43 Rep Power: 5	Dear all, I have recently used OpenFOAM (v2012) for simulating the Rotor37, and the results with rhoSimpleFoam are similar to those from the literature and from CFX (ANSYS). I am now trying to add up the stator and I would like to have some advice on how I should handle the rotor-stator interface. From what I have seen there's no available package for the mixing-plane option (for steady state RANS simulations), do you confirm that? For this reason I thought about using rhoPimpleDyMFoam, so directly go unsteady and introduce a moving mesh. The main problem here is related to the matching of the interface: should the two patches of interface between rotor and stator (set with cyclicAMI) match perfectly? For instance, in CFX the pitch ratio can be in the ratio 0.6-1.5, but is this still valid for OpenFOAM? Moreover, feel free to suggest any example of simulations of this type from the literature (I could not find much related to the OpenFOAM I am using). I would really appreciate that. Thanks a lot Best, Marco Rahmat1 likes this.

May 17, 2023, 05:42		#2
MarcoC501 Member MC Join Date: Apr 2021 Posts: 43 Rep Power: 5	Dear all, I have figured it out. I installed foam-extend 5.0 and here there are some routines capable of dealing with the interface problem. I have been trying both mixingPlaneand ggiinterface, using steadyCompressibleMRFFoam and steadyUniversalMRFFoam. Is rhoSimpleFoam supported for these two routines? That's because the convergence is not so good with the PISO algorithm used in the tutorial (e.g. pressure residuals do not go down), at the limit of divergence, and I do not know if that's a problem of solver or BCs. The stage is transonic, do you maybe have suggestions on what solver to use for foam-extend? I really appreciate your help and support!

May 17, 2023, 06:57		#3
Rahmat1 New Member Rahmat Join Date: Mar 2023 Posts: 16 Rep Power: 3	Marco, I did model an axial fan installed in a wind tunnel using simplefoam and results were far as expected. Pressure across the fan was negative instead of a gain. Is it possible to share your model? Thanks.

May 17, 2023, 07:06		#4
MarcoC501 Member MC Join Date: Apr 2021 Posts: 43 Rep Power: 5	Which version of OpenFOAM are you using? There should be some tutorials on that in the $FOAM_tutorials folder. May be related to your model? For me, it is PIMPLE algorithm (in steadyCompressibleMRFFoam) is diverging. I also have looking at the BCs, they are the same as the ones I set when using rhoSimpleFoam (with slight modifications)...