open TELEMAC-MASCARET The mathematically superior suite of solvers

For simulating the river running underground, I inserted a tube with 2 nodes of extremities one at the beginning and one at the end (as stated in the manual) and then assigned the width and height (both big enough for the applied discharge). However, I have a problem with the incomplete flow of entry (10 mm compared to 20mm runoff as an example) in the tube and a backflow as a result. In addition I used head linear loss coefficient of 1 and those of inlets and outlets of 1 as well. Where the problem could be???

Also, is it possible to add more points in the extremities to represent the tube? (for example 2 at the inlet and 2 at the outlet)

And if the river is running with a 45` angle, is there a possibility to represent that feature ? or the only option is to make a straight tube?

Thank you.

Hello

The use of Tube is recommended for locations where the flow can switch from a free surface one to a close-conduit flow. The flow is approximated coarsely as flow in a cylindric conduit. That's why, you can observe big differences between real flow and simulated one especially if the tube is very long.
To answer to your question: the tube is only between 2 points. Actually, you can introduce more than one tube, in parallel configuration (to simulate greater discharges) by connecting two (or more) close points to two (or more) ohter points in order to have several parallel tubes.

For your case, in order to be more accurate, if you are sure a priori that your flow (in the underground part) will stay always a free surface flow, it is better, in my opinion, to introduce the real shape and bathymetry of the channel and to simulate the case as if it was an open-air channel.

whith my best regards

Riadh ATA

Actually the flow is switching from a free surface to a close-conduit flow as it's not certain for the flow to remain free surface flow under extreme rainfalls. Thus a tube is a good option. The tube is long about 2200 m and not going straight (changing its course with a 45 degree after 1300 m). I used a width of 60 m (same width of the river channel) and a height of 4 m which is big enough to let water in. However the difference between the open air channel and this tube is really big. There is around 50 percent loss of discharge every time. Did you mean by the difference because we are moving from rectangular to cylindrical? And what can be done to avoid that?

In this case, you need to use more than one tube. very probably the flow has reached the maximum capacity of the tube. Try with 2 or 3 tubes and tell us what it gives.

Riadh

Okay then. I guess it's applicable. However, i didn't understand what you meant with connecting close points together?

And is there a possibility to add a 3rd point L3 with elevation z3 to represent a bending tunnel? (I know you've said only 2 points but just to confirm)

Hatem

I mean, you can use more than 1 tube each tube links a couple of points.
no, you can't add intermediate point, for that you need to change the formula in the subrotuine buse.f (folder telemac2d) and this seems to be tricky.

regards
Riadh