open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

We have developed a large model (see fig 1) with 1 inflow boundary and 3 tidal water level boundaries. We have been struggling with instabilities, strange recirculations and very high velocities, especially near the water level boundaries. We have done several tests using different numerical schemes and solvers. Attached is the latest .cas file where results much more reasonable until the third day of the run. We are hoping for some advice on which options we should be using for some of the numerical schemes and solvers.

Our files are very large but can be posted to our ftp site if needed.

Thanks in advance!

Hello,

This reminds me of something ...

This may not be the solution to your problem but it often is a problem leading to numerical instabilities: make sure the corner elements of your open boundaries (the triangles that are on either end of your open boundary 544 code) are not over constrained, i.e. that no triangle has 544-544-222 at its corner. Attached is an illustration (in blue for the problem triangle).

The advise is to swap the edge (in red).




Hope this helps.
Sébastien.