open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,
I need please your help.
I want to make an unsteady calculation for a flood wave with a duration of 10 days.
But there is an error message, wich leads to break the calculation.

The study area has 6 open boundarys:
- 5 Boundaries with prescribed Q and
- 1 with prescribed H (the Number of this Boundary in Listing datei 3) .


What should I do to ensure that the calculation does not stop at some timestep?

look the attachement please!

Thanks a lot

Hi
as indicate, you have a problem at boundary 1 where the condition should probably be torrential (as it seems it works before)
If so, you have to also impose the water level or change locally your model to ensure a fluvial flow

regards

Hello

Your first figure can not be viewed.
The error message means that at boundary #1 probably the area around the inlet is being dry or/and the regime is going supercritical.
This depends tightly to the model bathymetry and initial conditions and several other parameters. A simple way to overcome this issue is to reduce, tentatively, the time step. Otherwise, you can smooth a little bit the bathymetry gradient at the boundary vicinity.
Finally, use velocity profile=4 or 5, this will help to better distribute the flowrate with respect to the bathymetry.

I hope that this helps

with my best regards

Riadh ATA

Thank you a lot.

the error occurred in T = 3090 s. as is shown in the attached figure'Flood Wave'. At this time there ist no large change in the flow at Boundary 1.


The Time step = 0.01s, should it be even smaller?


I have given no unsteady BC for Outlet 'Boundary 3'? but only in Cas_file defined as constant. Could this be the problem? In unsteady case ist it necessarily to give a time-dependent water depth in Outlet? or not?

The Cas_File is attached.