[issue][regression from previous version] issue with layer generation with the viscous layer added hypothesis

[otaolafranc]

Hello,
there is an issue in salome 9.13, here is an example of an study done in salome 9.13 the exact same workflow, with same geometry and setup in 9.12 works correctly, where the mesh in 9.13 clearly has some issues.

here is the study (done in 9.13)
Study1.tar.gz

regards,

[c-bourcier]

Hi Franco,

Can you provide the step file and a python script? I don't reproduce the issue with a brep got from exporting your shape.
pb_viscous_layer.zip

Christophe

[otaolafranc]

Hello @c-bourcier
I don't have access to my pc right now, but are you testing in 9.13 or master? The version I was using is the 9.13 universal Linux, I did exactly the same workflow as you
Import geom, glue faces, mesh it with hexa 3d, quad mapping and number of segments and added the hypothesis for the BL, over if I remember correctly surface that will have the boundary layers (ie. Walls).when I will get my pc back I can report more on the issue but the tree of the object browser should be the same as yours. Maybe if you want a faster answer checking in that particular version if it has the issue. This came up while helping an user in the forum and he experienced the same problem.

[c-bourcier]

I confirm it works well in 9.13 with the python script and import brep.

[otaolafranc]

Strange, here is the link for the original step model https://discourse.salome-platform.org/t/adding-viscous-layer-fails-on-hexahedral-pipe-mesh/1504/13?u=franco.ota
I will also re do the python script when I have accès to my pc.
Regards.

[c-bourcier]

Hi Franco,

I reproduce the issue with the STEP file in 9.12, 9.13 and master, whereas it works well with the BREP file.

I have submitted the issue to our developers.

Christophe