open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi

I've been trying to reproduce small waves (5 to 10 cm) generated by a sudden variation of outflow near a boundary condition.

I tried different turbulence model, but i just can't seem to reproduce the observed shock in simulated water level measurments. No matter the turbulence model, the simulated wave's amplitude and oscillation duration is too small compared to observation.

Does somebody have already tried to reproduce this phenomenon with T2D ?

Any practical guidelines would help if it's possible to reproduce with T2D...(numerical and physical keywords values suggestions etc...)

Thanks

Mathieu

Hello

to reproduce this kind of waves, you need to use less diffusive advection schemes and to turn off all diffusive terms. It is not a matter of turbulence model, in my opinion.
You can try to use the characteristic scheme (MOC), this has the advantage to be completely implicit with no stability restriction. Furthermore, if your problem is not too large, you can use the second order kinetic scheme with finite volume kernel. The WAF scheme also can give satisfactory results also.

I hope that this helps,

With my best regards,

Riadh ATA

Hello Riadh,

Thanks for suggestion, i'll give it a try.

Hello,

You must also check the kind of waves you observe, if it is non linear waves it could be that they cannot be reproduced by Saint-Venant equations. Otherwise another important parameter in the case of waves is implicitation :

IMPLICITATION FOR DEPTH

and

IMPLICITATION FOR VELOCITY

must be as close to 0.5 as possible (this is the limit for stability so reducing the time step may be necessary, or playing with the keyword FREE SURFACE GRADIENT COMPATIBILITY).

With best regards,

Jean-Michel Hervouet