[Yann]

Hello all,

I'm currently experimenting with mixing planes in OpenFOAM-v2412 thanks to the port from foam-extend provided by the Turbomachinary Special Interest Group.

I'm simulating an axial fan and my air intake is basically a pipe at the center of the fan. So I end up with the axis of rotation being inside one of the faces of my mixing plane interface. (see 1st pic attached)
This is covered by the very useful documentation here:

twg_checkMesh

The standard application checkMesh was modified to run additional tests on mixingPlane patches when this application is invoked with the -allTopology command-line option.

A new topology test was added specifically for mixingPlane patches to detect faces intersecting the rotation axis of the mesh, but without any discretization point at r=0. If that situation is detected, a new configuration file for the application twg_modifyMesh will be generated in order to identify the faces in problem and specify a new discretization location for triangulation.

My issues is that the twg_checkMesh -allTopology command only detect a problem on one of the 2 patches of the mixing plane interface:

    coupledPolyPatch patches:
	Checking mixingPlane patch : mrf-interface-out_master for facet intersecting the r=0 axis.
    Checking mixingPlane patch : mrf-interface-out_slave for facet intersecting the r=0 axis.
    Checking mixingPlane patch : mrf-interface-in_master for facet intersecting the r=0 axis.
*** Not OK. MixingPlane patch mrf-interface-in_master requires a discretization point at r=0: (0 0.500578 0.44) for face index: 17434195
*** Writing a new modifyMeshDict entry for fixing this issue
    Checking mixingPlane patch : mrf-interface-in_slave for facet intersecting the r=0 axis.
    Checking mixingPlane patch : mrf-interface-side_master for facet intersecting the r=0 axis.
    Checking mixingPlane patch : mrf-interface-side_slave for facet intersecting the r=0 axis.

The mrf-interface-in_slave patch is not supposed to pass the test and I don't know why twg_checkMesh fails to detect it. master and slave patches are very similar and on this example the axis of rotation is close to the center of the face.
The following step with twg_modifyMesh is working fine and it does retriangulate the face to add a point on the axis of rotation (2nd pic).

Has anyone ever experienced this or has an idea of where is the issue?

Thanks,
Yann

[mbeaudoin]

Hello,

This looks like a bug (maybe), or a tolerance parameter issue related to this specific mesh (probably).

Would you mind sharing your mesh? It would be easier for me to check this.

Martin

[mbeaudoin]

And by the way, the best way to report those kind of issues with the TurboWG version of the mixingPlane is to use the TurboWG Ticket tracker tool: https://sourceforge.net/p/turbowg/tickets/

I will receive a message as soon as you submit your ticket. Much quicker...

Martin

Originally Posted by mbeaudoin

Hello,

This looks like a bug (maybe), or a tolerance parameter issue related to this specific mesh (probably).

Would you mind sharing your mesh? It would be easier for me to check this.

Martin

[Yann]

Hello Martin,

Thanks for your answer and sorry for the delay.
I had to put this project on hold and I just got back to it.

I wasn't sure if it was worth opening a ticket on the bug tracker as it might just be a mistake on my side.

However I followed your advice and created a ticket with a simplified case reproducing the issue here: https://sourceforge.net/p/turbowg/tickets/11/

Thanks for your help!
Yann

[mbeaudoin]

Hello Yann,

Thank you for sharing a simplified mesh replicating the problem. I'll have a look.

Martin

Originally Posted by Yann

Hello Martin,

Thanks for your answer and sorry for the delay.
I had to put this project on hold and I just got back to it.

I wasn't sure if it was worth opening a ticket on the bug tracker as it might just be a mistake on my side.

However I followed your advice and created a ticket with a simplified case reproducing the issue here: https://sourceforge.net/p/turbowg/tickets/11/

Thanks for your help!
Yann