open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi all,
We are trying to use BlueKenue for the first time as mesh generator (we used to do that with Matisse before), and it often happens that it produces meshes with flat triangle (i.e. with two nodes at the same place). Such mesh produces an error launching T2D (GEOELT: ELEMENT 3685 : DETERMINANT NEGATIF).
I don't remember that Matisse have never produces such corrupted mesh, and we want to know if there is something to do to avoid that in Blue Kenue.
To be more precise, I can say that we are modelling a bridge over a river with very small elements around piers, and flat elements apears in this area.
Thank you for your help.
Y.

I would have to see what your input to the mesher looks like.
Please post the t3m file.
This might happen if you overconstrain the mesher by providing duplicate hard points or hard points colocated with geometry supplied in hardlines.

Cheers... Martin

This is one of the very first mesh we produce. The boundary between the mesh and the sub-mesh is not fine, but the problem appear in another place, inside the small elements area, where there are no constraints on the mesh except bridge piers.
Thanks.

Hi,

There are a number of places where the mesher is over constrained.
There are a number or trouble areas along the edges where you supplied hardlines that "almost" coincide with the outline or crossed the outline. Also the "small elements area" overlaps the submeshes so the mesher isn't free to "grow" the elements.

I've made some adjustments and included the results and updated t3m file here.

Hope this helps... Martin

Ok, I think this will help us.
Thanks.
Y.

Hey,

I have a similar problem with Bluekenue.
I have a lot of flat mesh...like you can see on this picture
If you can take a look to my t3m file I don't understand where I'm wrong, I use BlueKenue for the first time as mesh generator too

Thank you for your help.

Regards

Marine

Hello,
The only thing I've found to avoid this problem is to remove all hard line in area where problem appear and to use an edge lenght greatest than one.
YV.

Thanks
But I have no hard line... I have just soft line it's the reason why the result worry me

Marine

Hi Marine,
I've remarked that you have some density area with value less than 1 in your t3m file. We have generally constated that mesh inconsistency appear in these areas.
I don't know if this is a general rule but for us, we used to have density greater than 1 to avoid mesh problems.
Y.

Hi Yann,

Ok thank you. I supposed this small density area could be a problem... But I have no choice because I modelized street which are narrow so I need to have small meshes.
I am just disappointed by Bluekenue because I had a lot of problem like superimposition of meshs, it don't follow well my outline,... I don't know if I use well Bluekenue but it is a test for me because I see on the forum a lot of people who use Bluekenue instead of Matisse...
So I'll back with Matisse :)

Thank for your help

Regards

Marine