open TELEMAC-MASCARET The mathematically superior suite of solvers

Good morning everyone,
I'm writing here to ask your opinion about the results I get from a simulation.
I'm using Telemac 2D to simulate a flooding event in a urban area.
In this area the river flows under a bridge and then it passes under a covered area: I modelled these structures with the tube file.
The mean length of the river bed is about 60m and the discharge that should pass through should be about 1000 m3/s.
During my simulation the river floods the surrounded area also with low discharge of about 200 m3/s.
I know it is hard to identify the reason with so basic information but I would appreciate every your comment and suggestion.
Liquid boundary file, tube data file and the .cas file are attached.
Figure 1 gives an idea of the studied area.
geometry file is to heavy to be attached here.
Thank you in advance
Daniele

Hi
Hard to say but maybe you could plot a profile along the river...

Regards

thank you,
I attach some picture of the bottom focused on the river area. (Figure2, 3, 4, 5).










Zooming on the area of the bridge and showing the velocities vector with BlueKenue, I get what is shown in Figure6.





the velocity vectors are really high. the points with high values are the outlet point of the bridge (first 8 lines), so it is related to the tubes file. Could it be the problem? if yes, how can I get real velocity?

Regards,

Daniele

hi,

When you say "I modelled these structures with the tube file."... can you explain why do you chose this strategy ?

You can find other posts on bridge modelling strategy.

Another point, try to reduce the element distorsion between your channel mesh and your general mesh.

I don't really understand why do you reproduce buildings in your geometry. Maybe you can replace them by closed vertical walls.

Please read this basic section.

Hello abernard,
I chose to model the two structures with the tube file because I treated them like bridges. To model a bridge, it is suggested in the manual to treat them like tubes, and so create a tube file. That allows me to have both pressured and free surface cases.

I think that buildings do not cause any problem and for that reason I did not replace them with vertical walls.

I think the problem could be in the tube file. Despite I've read a lot of the previous tasks about this strategy, I'm not able to find my errors.


I used a "8 points" channel mesh because of bridges. They have 4 arches each. So each arches is represented by 2 channel points. I decided to do that for the following reason:
to have a channel mesh with more nodes means to create too many tubes. So I do not know how to attribute the relative head losses to each tube. I know that also with 8 nodes it is not realistic, in fact they should be 4. By the way I consider negligible this difference for the moment.


best regards.

Daniele