open TELEMAC-MASCARET The mathematically superior suite of solvers

Good mornig,
The case we studied is the overflowing of a dike throw a notch in a channel (slope 3/1 after the notch).
In order to stabilize the velocity magnitude in Telemac-2D the only solution found was to use variable time step.
Unfortunatly in Telemac-3D variable time step does not exist?
So we tried very low time step (until 0.001s) but it was not effective (already waves in the velocity). We tried to change a lot of parameters without success (viscosity, turbulence model...).
Do you think it is possible to stabilize velocity with this slope?
Thank you in advance for your help.

Hello,

Did you try the non hydrostatic option ? Given the case it seems the first thing to do. I would also prefer that you work with a time step that gives a Courant number of about 1. Also set all the mass-lumping coefficients to 1. and the implicitation coefficients to 1.

I also enclose an example of parameter file provided by my colleague Kamal El Kadi, for a similar case, a dam-break on a physical model, so that you can inspire from it (it includes a coupling with Sisyphe but you can remove this).

With best regards,

Jean-Michel Hervouet

Hi!

I also tried such a thing some time ago. It worked well in the end with Jean-Michel's settings but I also had to use SUPG OPTION = 2;2.

Good luck!

Regards
PL

Hello,
Thank you for yours settings.
With v6p1 i already had some instabilities. But with v6p3, results are great.
I add i worked with 10 plans for a better result.