[taalf]

Hello all,

I am trying to set periodic conditions on a SU2 4.0.0 case.

My periodic marker:

MARKER_PERIODIC= ( periodic2, periodic1, 0.0, 0.0, 0.0, 0.0, 90.0, 0.0, 0.0, 0.0, 0.0 )

(I inverted periodic2 and periodic1 to see, but no change)

- I tried to launch SU2_MSH and to use the output mesh.
- I tried to use radians instead of degrees...
- I am sure the center of the periodic rotation is at (0,0,0) and that the angle is 90 degrees from periodic1 to periodic2 around the y axis.
- I have looked through the documentation and found nothing to help.
- The discussions of this forum are about older versions of SU2 and didn't help either.

I have tried lots of things, but nothing works. The flow diverges always as in the enclosed screenshot.

Replacing periodic boundaries by walls make it converge (so, the problem comes from periodical conditions).

Any help would be very much appreciated!...

Best regards,

William

[taalf]

After investigation, it seems that the right process is as following:

1- Launch SU2_MSH.

This creates a file mesh_out.su2 (for example).

2- Use this mesh for the CFD computation.

I wasn't able to complete this in 3D with the 4.0.0 version.

In the best case, the computation started (maybe in good conditions), but SU2 was not able to write the solutions files (I got a segmentation fault). This problem appeared only when using the output mesh from the SU2_MSH command.

So I came back to SU2 3.2.3 (which I am not able to compile with parallel support...) Simple 3D test cases worked fine.

Best regards,

William

[albega]

Could you send the mesh in cgns or su2 format and the configuration file?