open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi there,

I am doing a model with TELEMAC 2D (version v5p9).

As you can see on the attached file, my water level in river bed is 33m, my embankment is 35m ... but water spill over.
The water "stick" the embankment ...

I already tried to untick the "bancs decouvrant" option, but the problem is still there.

So I have some flooded area which should not be flooded ... is it a bug from the TELEMAC 2D? only the version 5p9?

Thanks

I forget to say that my embankment is well represented with 2 cells (5m width).

Cheers

Hello,

What you see is the interpolation of the post-processor, so the free surface taken into account is not what you see here. However points at the summit of a dyke may be influenced by neighbours which are wet, so that they get a very small amount of water. Generally this very small amount of water falls back and this is no problem, unless you have a single point at the top of your dyke, then it can fall back on the other side. So it is highly recommended to have several dry points at the top of a dyke, the dry points close to water having a slope that enables gravity to act. For example if you have 2 cells on the top of the dyke, it would be good to have a middle point higher than the two others. If your embankments have vertical sides and always stay dry, they would be rather treated as islands.

Note that there is now in version 6.1 an option :

TREATMENT OF NEGATIVE DEPTHS : 2

that ensures exact mass conservation and positivity of depths. It brings an improvement on this problem but so far we still cannot ensure that a dyke is not flooded if it has too few dry points.

With best regards,

Jean-Michel Hervouet

Thank you for your reply.

The problem is that my model is already big (200 000 points), I can't really do a more detailed dyke.

Also, for other scenario (higher discharge), I don't have this problem of spill. At the same location, for a higher water level, there is no water in the floodplain ...

Here is a 2nd exemple ... as you can see, even if the embankment is wide, the water from the river (on the left) spill.

Hello,

Such quantities are very strange. I would suggest that you move to version 6.1 and its new treatment of negative depths, and then we look more precisely at what you do.

JMH

Hi,

The wetting/drying algorithm in Telemac induces, in your case, the wetting of node i (see figure attached). Thus, in the next time step node j will be wetted also. That's why you will have a kind of non physical overspilling.
To overcome this kind of problems, follow the suggestion of Jean Michel, i.e. modelize the dyke as a chineese hat. I attached an example of dyke with the bathymetry of every node.
I hope that I was clear.
Kind regards.
Riadh
dyke_overspill.pdf

Let me try with jpg :)
dyke_overspill_01.jpg