open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi everybody,

I am currently working on a lateral weir that I decided to design with :
- Telemac 2D/3D,
- a CFD software.

The weir is Creager shaped (cf. seuil latéral.jpg).
Under a same upstream head, Telemac 2D or 3D gives a 460 m3/s flow rate (for an equivalent weir coefficient of 0.39) whereas the CFD model gives a 610 m3/s flow rate (weir coefficient of about 0.5). As per the litterature, the coefficient of a creager weir is about 0.5.

I wonder if Telemac is relevant enough for this case in so far as I've got quasi-vertical faces (cf. joined image). And further more, i identified a head loss just upstream the weir for the 2D computation (blue line) and also for the non-hydrostatic 3D computation (red line).








Without this head loss, the weir coefficient would be close to 0.5...
The computations on Telemac 2D or 3D seem really stable by the way.

Do you know if there is anything to do to improve the taking into account of the vertical cells ? Does the problem come from somewhere else ?

Thank you very much for your help.

Hi everybody,

Any suggestion or any hint ?
It would be great to share experiences on this subject.

Hi,

I also use side weirs in my 3D model. they are actually flood control areas and the water can enter over a piece of the dike. The dike is represented in the bathymetry of the model. For a storm event I had a look at the discharges over the dike and I also calculated a coefficient from the results. I remember that this coefficient was also a lot lower than the coefficient we got from field measurements. So I question if the discharges entering my flood area are not too small compared to reality. But up till now I haven't really test this properly.

One way to test it is to implement a weir structure in the mesh and use equations in the code to test the difference.
It's on my todo list, but this list is getting longer every day

Kind regards,

Sven

Hi Sven,

Thanks for your reply.
I wonder if this problem is a matter of content or not. Can Telemac deal with such issues ?

What intrigues me the most is this head-loss just upstream the weir. It is really visible in 2D.

Hi

In such situation, the hypothesis of a 2d horizontal flow (vertically averaged) is wrong...
This explain the local head-loss and thus the result.

Regards

I agree.
But with a 3D simulation (non-hydrostatic), the problem remains nearly the same.
The head-loss is not so obvious but finally the upstream and downstream levels are really close.