open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi,

I am trying to run a steady moderate flow in TELEMAC. After 6 time steps ( 30 sec time step) TELEMAC is taking abnormally long time for the 7th step and after 8th it is crashing (exit with exit code 2) and giving me the following error code:

PROBLEM ON BOUNDARY NUMBER 1
GIVE A VELOCITY PROFILE IN THE
BOUNDARY CONDITION FILE OR CHECK THE WATER DEPTH

I have tried with CONSTANT NORMAL PROFILE, SQRT (DEPTH) PROFILE and even NORMAL VELOCITY GIVEN IN UBOR IN THE CONLIM FILE (Although, I just calculated the velocity from Manning's equation and it was by no means very accurate)

But I got the same error every time.

In terms of initial condition, every time I tried to start with ZERO DEPTH (dry domain) I got the same error and the program stopped abruptly.

Then I created a free water surface in Blue Kenue with the water surface elevation I measured in the field along the longitudinal profile of the river and used it as the initial condition. I did not have any U and V velocity at the initial condition.

At the start, I got a message that in the absence of initial U and V TELEMAC set the initial U and V at zero and started simulation but after 6 time steps crashed saying the same boundary condition message as mentioned before.

I am kind of lost. What I am supposed to do if I only have a constant/time varying water discharge
figure for the inlet and noting else. How to start the program with a moderate steady flow and later introduce the time curve after the models starts running smoothly.

Thanks and regards,

Joy

Hello,

This problem appears when you specify a discharge at a boundary (here 1) and that there is no water on this boundary. Then Telemac tries to find the velocity that will give the requested discharge, but as the depth is zero it is impossible. There remains to understand why here your entrance dries out after 5 iterations. Please check the initial depth and look if you have velocities at the entrance during the first 5 steps and what is their value. Also check that the 30 s time step is relevant with your mesh size (rule of thumb Courant number U*DT/DX not too large, to start with).

With best regards,

Jean-Michel Hervouet