open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi

I'm new using telemac and I would like to understand the functionalism of the culvert we can use. I made a very simple channel, with slope quite normal and a hole (I mean a hole in the mesh) in the middle. I created a culvert through this hole with the normals singular head losses, and no linear one. The levels of the tapers are equal to the level of the thalweg at the same point. The section is 5 square meters large and the input flow in the channel is 3 cm/s.

So evereything should works, but the flow through the culvert is no higher than 0,3 cm/s. Does anyone got an idea of where I am wrong !!
Thanks

Sorry the question isn't exact. I have two situations depending on parameters : either the flow is very small or it diverge to a very large number.

Hi

Here some suggestions for your case:

- You have a prescribed discharge both at the inlet and at the outlet of the flume which is not taken into account strictly by Telemac. Change the outlet condition to prescribed elevation (with a 5 4 4 condition in the cli file and by adding the keyword "Prescribed elevation").
- You use Saint-Venant-VF which is the finite volume kernel of Telemac. However you don't indicate which FV scheme neither the Courant number. The easiest way to star with, is to switch to the Finite element kernel by changing Saint-Venant-FV by Saint-Venant-FE.

Kind regards.
Riadh

Thanks

But I already try. Even if I delete all the liquid boundaries and put a source the flow is too small through the culvert.

I think the problem comes from the culvert itself !

Maybe this parametrization could be more helpful to understand my problem. The flow is imposed upstream and the elevation downstream. The water flows up in the culvert (positive flow from low extremity to high one).
I can't join the result because the file is too heavy.

Thanks

Hello,

I don't quite understand what you exactly do, especially the culvert through the hole. A culvert is simply two points in the mesh, preferably not on boundaries, otherwise it should be an open liquid boundary (on a solid boundary it would go against the no-flux condition of that boundary). The two points are treated as sources, except that their discharge is deduced from the free surfaces and the culvert characteristics. The direction of the flow goes from the point with highest free surface elevation to the other. The flow depends on the characteristics of the culvert, but cannot go from low level to high level. To prevent drying, there is a relaxation coefficient that limits the flow at the first time steps (this can explain a small flow at the beginning). I hope this can help.

Jean-Michel Hervouet

In fact the aim of my test is to model a small bridge in flood situation. I would like to superimpose a culvert and then a weir in the same hole. That could have both behaviours of small bridge in low and high flow.

That's why I am building a hole and then trying to put a culvert through it.

Maybe it isn't possible that way ??

Hello,

OK, the hole will be the weir, but you should then put the points of the culvert outside the boundary, let's say one element before and one element after the hole.

JMH

Hi

There is an increasing need for a bridge model in Telemac2D. This question was pointed out in the frame of the developper meeting last few weeks. Hopefully the development is already achieved by one of our partners and we hope include the model in the next release i.e. V6P2.
For the moment I refer you to the following post ant the answers related to it:
www.opentelemac.org/index.php?option=com...39&Itemid=89&lang=fr

With my kind regards.

Riadh

I'm really sorry to insist but this topic did not contest my answer finally !

I really would like to understand why I can't have a normal steady state flow through the culvert ! I mean I tried many different boundary conditions a none seems to work. Either the flow trough culvert stay very low or it diverge to abnormal values.

I think there is something I understood wrong in the parametrization of culvert, I did :
- Default head loss at both extremities
- No linear head loss (lambda=0)
- Levels of the tapers equal to the elevation of the thalweg at the same point
- All the rest equal to 0
- Source flow zero too

I made tests with 0.5,1 and 2s time step and each time with 200 steps. Maybe I should take into account the length of culvert in the choice of time step ??

Thanks for helping me !!!

Quentin