open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

I do not have much experience with TELEMAC.
My aim is to obtain a steady state in a reservoir. The reservoir is 140 kilometers long, one side is limited by a dam and I have about 31000 elements. The flowrate at one boundary is 2060 m³/s and the elevation is 302.8 m (min z=260 m). The bathymetrie is very complex and I have no values to validate.
In one area I get always big velocities with alternating directions, which I cannot understand. I calculated already for 50 days but I do not get a steady state (although at the two boundaries). I also changed some numerical parameters, but didn't get sufficient results.
Maybe somebody with more experiences can give me some advice. I have attached the steering file and also the geometry and boundary file I used so far. I also attached a picture to see the alternating velocities.


Many thanks in advance

Hi
One basic remark on your configuration. As you start from a constant level, it could be great if you impose a varying in time discharge, starting from 0 and growing slowly to 2060 m3/s.
This could avoid some shocks and maybe solve your problem

Hope this helps

Hi,

I just had a quick look at your cas file, and you have VELOCITY DIFFUSIVITY = 0. I don't know if that can explain your problem but it is worth trying with higher values.

Regards
PL

Hello,

That's exactly what I was about to say, try :

VELOCITY DIFFUSIVITY : 1. or 2.

and see what happens (provided that on the points with strange velocity, the bathymetry is correct).

Regards,

Jean-Michel Hervouet

Hi

Some other hints and remarks:
1- do not run simulation for a so long duration (> 4 million time steps !), instead, run the simulation several separated times using restart option. This allows you to control the evolution of the solution and to avoid useless extra long time of simulation.
2- You've chosen a constant level initial condition that corresponds to the highest point in the model. This means that you have a huge domain completely filled with water! and thus, you need a very long time to empty it and to retrieve a physically acceptable steady state. try to start with lower elevation, this will help to reduce CPU time and to relax numerically the simulation.
3- I've had a glence on your mesh and it seems that bathymetry is not coherent especially in the vicinity of the area given by x= 9010000 and x= 9006000 and y=542000 and y = 540000. The bathymetry looks relatively high which may choke the flow at this region. to verify!

I hope that this helps

with my best regards,

Riadh