open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi,

I would like to add a weir in my river simulation.
I tried to put one but I have weird results.
When I run the simulation without the weir, it's working but when I add the weir, it's like if some waves were generated next to it with high velocities.
Can someone have a look at my files and tell me if my problem comes from the mesh or from the parameters that I chose ?

Thanks in advance
Marie

Hello Marie,
Here some remarks about your model:
1-The implementation of the weir is correct
2-your previous (restart) file is a bit strange, you have velocity vectors normal to the river direction,
3- you have a big restriction in the main river bed immediately downstream of the weir, this could generate (physical) perturbations (high velocities)
4- immediately after this restriction, you have a 90° bend, this will generate interfering waves like the one observed in the CADAM test N2. This phenomenon induces high water depths at the bend location which will cause the wettinng of the domain located in its vicinity.
5- The bathymetry around the weir (which is very low) allows the water to flow substantially around the weir, which creates big recirculation and negative pressures downstream, this is normal in your case, if you want to restrict the flow through the weir you have to add dykes to prevent flow upstream of the weir.

I hope that it was helpful

Riadh ATA

Dear Riadh,

Thanks a lot for your remarks.

I changed the mesh next to the weir, and I cleaned the bathymetry around the weir to have a proper channel without this big restriction.
I generated the restart file with fudaa to be able to run a first calculation, and the result of this calculation will be the restart file for the real simulation. I removed these velocity vectors normal to the river direction.
And now the results are looking better :)

However I still have a few questions:
1 – I saw in the weir example ( 037) that in the weir file, the “Point cote 1” were corresponding to the downstream points of the weir, and the “point cote 2” to the weir's upstream points. Is it important or can we reverse them ?
2 – Can you have a look at my new mesh next to the weir and tell me if it's a proper mesh or not (in comparison with the previous one that was more tight next to the weir) ?
3- Unfortunately this 90° bend does exist on the river, and the bad news for me is that there is also a bridge between the weir and the bend. Do you think I should tighten the mesh in this area ? Where can I find the CADAM test N2 you spoke about ?

Thanks a lot for your kind attention to my case.
Best Regards
Marie

Dear Marie,

1-For your first question, in a general and very usual case, modeler knows the flow direction (in rivers and channels), thus side 1 will be the updstream and side 2 the downstream of the weir regarding the flow direction. However, in a physical point of view, a weir is nothing but a geometric "bump" which behaves symmetrically without any priviliged direction. For our case, in Telemac, the weir has not predefined direction and works in both directions. The flow direction is automatically detected by computing the water depths in both sides of the weir.
2- I had a glance on your new mesh and it looks much better, However, you did not give the boundary conditions file so that I can run the case.
3- Thee is no problem to add a bridge close to the bend downstream of the weir. You need only to care about the way to include all these singularities in your model. I remarked that you have coarsened the mesh in this area, depending on how you will add the bridge, it would be worth to refine a little bit this area. If you are sure that the water elevation will never reach the lowest point of the bridge deck, you can just add the piers in the mesh (in this case, more refined mesh would be better), otherwise, you need to add it as a singularity as for the weir.

A last recommandation, don't adapt your model, exlcusively, in order to make it working better. You have first, to reproduce the real case with its real hydraulic structures.

I hope that this will help you.
With my best regards,

Riadh

Dear Riadh,

Thank you again for your answers, it helps me a lot !
I attach all the required files to run the calculation. Don't look at the flood at the "end" of the river, I didn't yet add in the mesh a secondary channel that is parallel to the river (so it's a mess, but I take problems one by one).
On the results, there are some high velocities on the weir at the beginning but then it's ok (I think). And we can see the mess caused by the 90° bend.

Regarding the bridge, the bridge deck is really low and it will be reached easily during a flood, so I'd rather try to use the new bridge file of Telemac. With this new bridge definition, can the water go above the bridge ?

I don't have so much profiles of the river, so the bathymetry of the minor river bed is not so exact between the profiles (I interpolated it as best I could). That's why I could clean the bathymetry trying to make it look like more "physically probable".

Best regards
Marie

Hello Marie,
The bridge solution in Telemac was introduced mainly to represent small sturtures (less or equal to the mesh size) in order to reproduce the flow that occurs under the deck (in charge or in free surface). In this case, the bridge is included with its upper elevation and the flows goes through it (like a culvert) and obove it if there is enough water.

With my best regards,

Riadh