open TELEMAC-MASCARET The mathematically superior suite of solvers

You are right, I will add it for version 6.2, it was only announced in the update file of version 6.1.
5 is like 4, but the depth used in the formula is a virtual depth computed with the lowest elevation in the boundary, it tends to concentrate the flow in the main channel.

JMH

Hello,

I am trying to impose a discharge at the upstream boudary of a simple trapezoid channel but I am meeting a problem. Big water depths and velocity appear on the banks while they must stay dry.

I am using the following config :
BANCS DECOUVRANTS = VRAI
OPTION DE TRAITEMENT DES BANCS DECOUVRANTS = 1
TRAITEMENT DES HAUTEURS NEGATIVES = 2
PROFILS DE VITESSE : 0;4

I thought it was an instability with "VELOCITY PROFILE = 4", because to me, to increase depth leads to increase velocity and to increase velocity leads to increase depth.

So, I changed the BORD subroutine so that the velocity profil is computed with the averaged water elevation along the boundary and it seems to have improve a bit.

But in fact, the water elevation is still slowly increasing on some of the dry nodes while there is no velocity.

Does anybody have an idea of a way to get a clean flat free surface in my boundary?

Thank you

Thomas

Hi Thomas,

If your problem is still not fixed, can you send us your model with the steering file, geometry and boundary conditions.

My best regards
Riadh

Hi Riadh,

my problem is still not fixed indeed.. A zip file containing my steering, geometry and boundary conditions files is attached to this post.


Thomas

Hello,

You can use also PROFILS DE VITESSE : 0;5 which is a variant of profile 4 based on the minimum elevation at the entrance. Anyway such inflow conditions with dry zones are not really well posed, as you have supercritical and fluvial regime on a single boundary (and they do not request the same boundary conditions). A solution consists of doing like in physical models : an extra artificial section upstream with deeper waters, so that all the entrance has a fluvial regime.

With best regards,

Jean-Michel Hervouet

I will try to use PROFILS DE VITESSE 5 later, I am still using the v6p0 for the moment.

About the 2nd solution : this is nearly the same as having a shorter liquid boundary that is only composed of the points that are never dry (the point in the bottom of my trapezoid section), isn't it?

Thomas

Yes, that can do it, but you may have a recirculation, as it will do like the jet in a larger canal.

JMH