open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi!

I am running some TELEMAC simulations to study the discharge of wastewater of a nuclear power plant. There are 4 exit channels. In one of the scenarios a channel is out of order for maintenance, so no water is flowing through. In my mesh I interrupted that channel, but in the results of my simulation I see the creation of a standing wave in the beginning of the closed off channel. This standing wave causes flow fluctuations in the whole system.

I already performed these measures:

- I let the flow gradually build up to avoid sudden shocks
- See CAS file in the attachment for used keywords

I also notice some strange velocity peaks at the sides of the inflow boundary (the water flows from down to up).

I hope someone can help me :)

Kind regards and thank you!

Tim

Hello,

In one similar case I remember that we had also small standing waves and they were observed in reality. Anyway I find that your time-step is very small, you can certainly increase it.

You can do the following changes :

Remove H CLIPPING and MINIMUM VALUE OF DEPTH (slightly negative depths are not a problem, if they are add : TREATMENT OF NEGATIVE DEPTHS : 2)

SUPG OPTION : 0;0;0;0 (2 duplicates the effect of FREE SURFACE GRADIENT COMPATIBILITY, if you have spurious oscillations, decrease it, down to 0).

The two keywords IMPLICITATION : you can out 1., going close to 0.5 is important for short waves.

OPTION FOR LIQUID BOUNDARIES : 2 I do not understand why you do a Thompson boundary condition (generally it is for open conditions offshore)

You can also try a good old constant diffusion instead of Smagorinsky.

The solver can be conjugate gradient (1) instead of GMRES (7) which is used for the primitive equations option (TREATMENT OF LINEAR SYSTEM = 1). Solver 1 is much faster for easy problems.

I hope this helps, regards,

Jean-Michel Hervouet