open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi everyone,

I generated a fine mesh in BlueKenue with a maximum edge length of 1 cm, and the mesh analysis shows a minimum edge length of around 1 mm. However, when I start the simulation, I receive an error stating that the mesh is too small.

Has anyone worked with very fine meshes in TELEMAC and encountered similar issues?

Thank you!

Hi

1mm is a harcoded value in the checkmesh.f file (utils\bief).
You could add checkmesh.f into your user fortran and then bypass this error.
I don't think users built mesh with such small size...
Did you try to create a model of a small scale physical experiment?

Hope this helps

Thank you! I was curious whether there was a specific hardcoded value like this.

I'll try your tip with the checkmesh.f!

And yes, I'm trying to replicate a small-scale physical experiment, and since the error occurred, I was thinking of scaling it up a bit. But I haven't had time so far...

In a perfect world, this harcoded value should be replaced by a default value linked to a dedicated keyword...
I think this value has been chosen because in nature modeling we rarely represent details lower than 1mm.
Most of the time, this message allow users to identify a problem in their mesh (two nearest points in a constraint line for example)
Small scale physical model often suffer of distorsion problem leads by constraint on the spatial extent of the lab which is not the case of numerical model. Maybe you could think to build you model at the real scale...
Regards

Hello,
I'm replying to your post because I'm facing an issue that is very similar to yours.
I'm currently building a model for a small fish passage (the canal is about 1 m wide).
To do so, I'm creating a mesh with a very fine resolution (around 3 cm).
I've encountered some problems while generating the mesh in BlueKenue, it seems that it can't handle such small elements.
As an alternative, I tried creating the mesh using Q4TS, which works much better. However, Telemac seems to struggle with this mesh, and the output mesh appears to be different from the input, as you can see in the attached file.
Has anyone encountered a similar issue?
Any feedback on working with such fine resolution meshes would be greatly appreciated.
Have a nice day!

sounds strange...
Could you share the different files of your case?

Yes, please find my attached files (.cas, .slf, .cli and .Liq)

Some files haven't beeen uploaded so I created a Wetranfert link :

we.tl/t-VYeFpTqoze

The geometry file is not really clean with very small triangles (4 triangles had an edge length < 1cm) but there is also a problem with the bottom elevation.
Min value : -999
Max value : +371.65
Hope this helps
regards

Hi,
I never heard of Q4TS but i was able to create a mesh with 1 cm edge length with BK. The issue occurred when i tried to run it with Telemac. Disabling the Check Mesh option helped, but i had some pretty distorted cells in the mesh and therefore the results weren't amazing. I honestly just scaled up my flume.
How do you handle the vertical walls?
I'm curious if you find a solution!

This is probably related to precision in your selafin file. Did you tried with double precision (SELAFIND)?