open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi everyone,

does anybody know how can I edit the mesh, so that these 2 points (red circles) are connected?

Thanks in advance

Hi
There is no direct solution.
The only one I could imagine is to save the mesh as a .t3s file and then manually add a triangle with the 3 nodes.
Then open again the .t3s file and create the selafin mesh...

Regards

Thanks, I solved the problem by manually adding a triangle with the 3 nodes (exact coordinates as the mesh). Then create a mesh from the 3 points of the triangle. Finally i merged the 2 meshes. Hopefully it will cause no problem by simulation. But at least right now it looks fine.

My mesh looks fine but when I create a boundary file for the mesh. The boundary will not be identified correctly. Do you have a solution for that?

Hi Annatrinh,
Post your .bc2 file here so we can have a look at the problem.
Cheers... Martin

I uploaded the mesh and the bc2 files.
Thank you very much in advance.

drive.google.com/drive/folders/1Uu4TfgAV...nEb9-EFeMuyAGjtlnMYP

Edit: The location is around 294590/ 5628840.

Hi,
The problem with creating the boundary conditions file is because you seem to have added the element twice. See the attached image.
Cheers... Martin

Incidentally there are other problems with your mesh.
It will be worth your while to scan the mesh carefully for other anomalous elements. In particular, there are many over-constrained and edge-spanning elements. There are also many orphan elements/element groups that will be disconnected from flow.
I've attached a file containing an analysis of your mesh for your information.

Hope this helps.. Martin

Thank you very much!
The duplicates were the reason!

The orphan elements/element groups that will be disconnected from flow are on purpose.

But I still have a question regarding to the over constrained and edge-spanning elements. What are the disadvantages of these? Though I could easily fix them using FUDAA PrePro, I only observe that if I have many over constrained elements, the simulation in parallel mode will not work, but as long as I use the normal one (without parallel), the simulation goes fine.

Perhaps someone more up-to-date than I can verify this but I was under the impression that TELEMAC wasn't able to properly model flow in elements did not contain at least one internal node.

Cheers... Martin