open TELEMAC-MASCARET The mathematically superior suite of solvers

This is a huge regression compared to 3.3.4.
It appears that a T3Mesher in 3.12.31 with submeshes deplicates the nodes and elements of the submeshes.




In above screencapture, black is the submesh, red is the generated mesh
See that BK identifies duplicated elements
See also the mess near the outline of submesh

On 3.3.4, it was not perfect but at least no duplicated elements

Thank you for taking the time to share your experience with BK 3.12.31 beta.

I can't seem to post any detailed response... I keep getting blocked. Know that I will look into this.

Kindly... Alan

Can you share a simple example that I can use to exactly replicate your observation? Doesn't have to be your private data... just something I can use on my side.

Kindly, Alan

I will release 3.12.32 - beta today with all of the accumulated changes I've made over and above 3.12.31 - beta. This way we will be closer to using the same code. Maybe my changes have already fixed this fault. We'll see...

In any case, please try again with the newer BK and let me know...

Kindly, Alan

Please upgrade to the most recent version. As of 25 April 2025, it is BK 3.12.32 - beta

Kindly, Alan

3.12.32 does not solve the issue.

See sample data: cloud.nicodet.fr/s/x7HSJ4dDmeQ4Hnc

Thanks for the link to example data. I'm not (yet) able to download what you shared due to being blocked on my side. I will investigate.

In the meantime, once I'm able to download your data, what steps did you take after loading the data? If you shared a mesh generator file, then I assume you only generated the mesh. Thanks for confirming...

Kindly, Alan

Thank you for sharing your *.t3m file. I was able to download it from home.

On first blush, I can see that all of your density polygons do NOT have values. In order for the meshing to work with densities, the values need to be defined (i.e. something other than a NO_DATA_VALUE). Otherwise there is no point to add "Zones_densite_Nive" into the meshing process...

How did you create "Zones_densite_Nive"? In any case, you will likely need to correct it.

Please share a corrected *.t3m file with me if that does not solve the fault.

Kindly, Alan

Here is the updated t3m: cloud.nicodet.fr/s/TBbGcH9orB4B74E

Was a mistake (forget to set the correct attribute).
Initial issue still happens

Thank you for the updated data example. I tried with BK 3.3.4 and seem to see a different submesh than you. Here I've coloured it green.

Also, I can clearly see that BK 3.3.4 is creating an invalid mesh with the data you shared. It will take me some time to figure out the easiest way to fix this. Stay tuned...

Kindly, Alan