open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello!
I'm new here and a beginner with Telemac 2D. I have some difficulty to have satisfactory results with my modelization.
It is a large reservoir ended by a spillway made of several gated weirs. My objective is to know the velocity distribution (champ de vitesse?) around the spillway intake in a steady state.
As the detail of the flow through the spillway itself is not very interesting for me, i though it would be better not to include it in the model : i've "cut" it just before the weir to avoid sharp bathimetry gradient. i had in mind to impose a water depth and velocity matching the critical depth and velocity.

I've tried several types of boundary conditions (mainly upstream water level and downstream water elevation and velocity), several time step (down to 0.02s) and reduce my mesh size at the vicinity of the spillway (1 m) and each time i obtain the same result: the velocity distribution is rather good, but the water elevations are weird : it goes sharply up and down for several meters from the spillway boundary (and it seems steady).
If Anyone has some hints for me there...

Hello Odile,

The first thing that comes to mind is that your artificial downstream boundary should be at a position where the flow is still sub-critical, otherwise your boundary condition is not well posed and you will observe a discrepancy between what is computed by the program and what you try to impose. Could that be the point ?

Regards,

Jean-Michel Hervouet

Thanks for your quick reply!
I think i have make sure of that by doing some trials where slightly higher water level and lower velocity than the critical ones are imposed: same unexpected water level variations. And at the same time Froude stay below 1 (0.89).
My initial condition is a constant water level on the whole domain (392 masl), and then the level is lowered on the DS boundary down to 387 masl. This variation is important, could it be the reason?
Meanwhile, i will try to impose a gentle variation of level and velocity : reach the final one in 100 s for instance.

Regards

Hello,

Yes, it could be that you generate a wave that reflects on the downstream boundary, so a longer time to lower the boundary can do it!

Regards,

JMH

I'm having the same problem even with a slow BC variation (reaching final condition in 550s)...
Would have used different boundary conditions? or specific calculation settings? (i'm using all default values except for the TYPE DE TRAITEMENT where i've specified "équation d'onde").
Regards,

Hello,

The boundary conditions at the exit should be 5 4 4 in the boundary conditions file. You could post your parameter file, so that we can see exactly what you do (default values are left as such for compatibility, but are not the best).
TYPE DE TRAITEMENT is not a key word but TRAITEMENT DU SYSTEME LINEAIRE : 2 is the most efficient option (SOLVEUR to 1 in this case).

Regards,

JMH

Hello,
I'm sending you my files so that you can advice me on a better configuration.
Best regards,