[SSSERB13]

Dear Friends,

I am running a VAWT case using pimpleFoam with moving mesh (AMI). The pre-processing (mesh, physics, BCs, solvers) is apparently correct and the solution is running well. However, the results are strange on the interface side. We can clearly see the velocity gradients between the stationary and rotating regions that are not physically accepted. Please find attached an image to illustrate that.

I have run a propeller case before using the same setup and the results were much better. We don't see that apparent interface.

Have you seen such kind of behavior before? Any idea how I can improve this?

Thank you for your help.

[HPE]

it seems that there is no inflow into the rotating region. personally, I have never seen this. are you entirely sure that the setup for the boundary conditions is correct? can you double check it with the propeller case, e.g. any difference to the propeller tutorial setup and steps to establish the setup? can you create your VAWT mesh inside the propeller tutorial, i.e. meshing VAWT instead of the propeller?

[SSSERB13]

Originally Posted by HPE

it seems that there is no inflow into the rotating region. personally, I have never seen this. are you entirely sure that the setup for the boundary conditions is correct? can you double check it with the propeller case, e.g. any difference to the propeller tutorial setup and steps to establish the setup? can you create your VAWT mesh inside the propeller tutorial, i.e. meshing VAWT instead of the propeller?

Hi HPE, thanks for answering.
Yes, I have checked the BCs and solvers, and are exactly the same (walls, openings, AMI). I am using the salome to generate the mesh. I did the same with the propeller tutorial and worked fine.
For me, it's something related to the rotating orientation. When we have the flow oriented to the rotating axis (such as the propeller case) the solution goes well. However, when having the flow transversal to the rotating axis (in the case of VAWT) this kind of thing happens. Do you have any more ideas about this?

Regards;

[HPE]

I think I understood what you have been describing, need to investigate this a bit.

Could you add more descriptive illustrations of the mesh and rotating body, or even files of case settings?

[SSSERB13]

Originally Posted by HPE

I think I understood what you have been describing, need to investigate this a bit.

Could you add more descriptive illustrations of the mesh and rotating body, or even files of case settings?

Hi,
Please, find attached the illustrations and the files (0, constant, system). The mesh size transition is not bad and the files are also configured properly (by default). Also, note that velocity discontinuity with the plot over line. It's a non-physical situation for an unsteady state. The rotation is happening but the velocity is too low inside the domain compared to the stationary.
Now, I am trying to take the rotating domain a bit bigger to see if the discontinuity becomes smoother.

Thank you for your help.

[HPE]

- is the rotating body rotates around the z-axis? (yes, I assume). could you show some orthographic/perspective view of the rotating zone - is it a cylinder, so that it can rotate?
- the origin in your input is = (0 0 0). does this correspond to the correct position for the rotating zone? maybe it is off-shoot, and the solid body tries to rotate around the wrong rotation axis.
- is there any reason why "MRF" is being used despite the solid-body motion? (I think this is the culprit). I think I have asked if you had tried propeller tutorial - but this tutorial does not have any "MRF". Could you double check if you miss out something in your trial?
- if possible, please just use the "inletOutlet" BC for U at "OUTLET" boundary. And please review your "p" BCs as well.

[SSSERB13]

Originally Posted by HPE

- is the rotating body rotates around the z-axis? (yes, I assume). could you show some orthographic/perspective view of the rotating zone - is it a cylinder, so that it can rotate?
- the origin in your input is = (0 0 0). does this correspond to the correct position for the rotating zone? maybe it is off-shoot, and the solid body tries to rotate around the wrong rotation axis.
- is there any reason why "MRF" is being used despite the solid-body motion? (I think this is the culprit). I think I have asked if you had tried propeller tutorial - but this tutorial does not have any "MRF". Could you double check if you miss out something in your trial?
- if possible, please just use the "inletOutlet" BC for U at "OUTLET" boundary. And please review your "p" BCs as well.

Hi,
You were right about the MRF file. It's working normally now.
In the beginning, I tried to run a steady case but I shifted to unsteady and leave the file there. I thought that the file within the folder would not be used since I was going to use pimpleFoam and this module is addressed to unsteady solutions. Anyway, it was a lesson learned.

Thanks a lot HPE!!