open TELEMAC-MASCARET The mathematically superior suite of solvers

Dear Users,

I hope somebody could help me with the surface elevation problem I have encountered in the simple shape geometry wind case run. There is one open boundary with no prescribed velocities. The fluid is stationary and stratified. When wind is forced (by use of the wind file) then surface elevation is rising exponentially. Strange, but with addition of the Coriolis force it is rising much faster then without.

I have search on forum, but could not find any preiously similar problems posted. I have experimented with various parameters change, but it did not help. Funny, but I have the real bathymetry case, where velocities and surface are behaving very nice. Even when I used exactly the same steering and fortran files, that I used for the real bathymetry case, it still gave me very bad results for the simple geometry shape domain. The stratification without wind works fine. I have not run it for too long, but surface elevation was stable there. Also when I closed the boundary, the wind forcing effect on the surface elevation and velocity is fine. On attached figures, I have a case when surface and velocities are rising and falling on a very short time. If it is a necessary "model settle" period then I do not get the same on the real bathymetry case. In other cases the surface keeps rising.

I have attached the cas and fortran file, just in case. Also there are few plots: surface elevation and velocity with addition and without Coriolis, from the problem case. In case without Coriolis force the wind has been forced on the 7th day, so I cannot understand why they surface elevated at the very beginning when wind was 0. With addition of the Coriolis the wind had been forced from the first time step but surface did not settle down as in another case and kept rising. Sorry for the long and messy explanations, please let me know if you need more info. Sure, I made a silly mistake somewhere but I just cannot see it. Thanks in advance.

Kind Regards!

Violeta

Hello Violeta


I've read quickly your explanations and I think that, based on the fact that with closed boundary the simple case works well which is not the case with open boundary, there is probably a problem with your boundary conditions you are prescribing at the open boundary.
I let other forum users, more familiar with T3D, checking up your problem, meanwhile, try to verify your boundary conditions by making, for instance, a sensitivity analysis.

with my best regards
Riadh

Thank you Riadh for your reply. You are right, the problem is probably at the open boundary. In a case of the real bathymetry, I have 544/5 (prescribed H and tracer) and the same I used for this case. So it is working fine for one case but not for another. Also one person suggested that there could be problem with the elements at the open boundary. I cannot see any obvious problems, but might need to take a closer look. Thanks for advice. Sure something will work out:).

Kind Regards!

Violeta

Dear Users,

I have tried various boundary conditions, but surface elevation is still very high in a tank simulation case just with one open boundary. At the moment I have 544/5. I have also tried 566/5. In all cases the water is coming into domain constantly and piles up to lift the surface. The case with closed boundary works fine. The case with an open boundary but real bathymetry works fine as well.
As initial conditions I have Constant elevation, but I have tried others and it gave bad results as well. Also, I use OPTION FOR LIQUID BOUNDARIES =2; OPTION FOR THE TREATMENT OF TIDAL FLATS=1 and TREATMENT OF NEGATIVE DEPTHS = 2 with SCHEME FOR ADVECTION OF TRACERS =14. I have experimented with these parameters but all gave me elevation or drop in some cases.
I have redone the tank geometry file but still no improvement there. I have used MESH TRANSFORMATION = 3 and 1, but it did not help either. Strange that the open boundary with the real bathymetry gives very good results, but tank case has such a high elevation, even if I reused the same cas and fortran files.
Please, if there are any suggestion on how to improve results I would be happy to try. I have attached the fortran and cas files. Thanks in advance.

Kind Regards!

Violeta