open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

I'm working on an hydro-sedimentary model and I would like to test the impact of a uniform and non-uniform particle size on my simulation.I launch my simulation in parallel. When I launch telemac which is coupled with sisyphe in the case of uniform particle size everything goes correctly.
However when I launch the simulation with non-uniform particle size, the calculation stops after 1 days of simulation with this error message and I don't understand why.

Has anyone ever had this problem and could help me?

Thanks,
Claire

Hi
You're running your simulation in parallel and the problem was not located in the master processor...
You should look into the slave listing (PEXXX-XXX.log file in the temp directory)
You could also run the model in sequential mode.

Regards

Hi,

Thanks for your reply !
I found the problem :

DIMINUER DT OU AUGMENTER ELAY0
EVOL= -0.12000000000000050 ELAY0= 0.12000000000000000

I have to change the value of the active layer thickness or decrease the DT. Do you know why by decreasing the time step my problem will be solved?

Best regards,
Claire

Because the evolution will be lower with a smaller time step, then you will not emptying the active layer in one time step...

Regards

Dear Claire,

we had such errors as well. I doubt that you will fix the problem if you use a smaller time step (but nevertheless try this first!) Because it looks like a numerical problem. The 2 numbers are nearly the same, only the 2 last digits are different, this is numerics.

A trick could be to change the stop in layer.f not directly at the layer thickness but a very little higher.
If you are interested I can send you a modified layer.f. Just send me an email to This email address is being protected from spambots. You need JavaScript enabled to view it. and your used telemac version.

Another idea could be changing your compiler options to less optimised ones. But this is only a guess I never tried it.

Best regards,
Rebekka

Dear Rebekka,

The simulation runs with a smaller time step but I think that I have an other problem..

First, for time step = 1 the calculation was running but I had a numerical problem. A "wave" is created in the direction of the flow which means that I had a numerical scheme problem. So I reduced the time step to 0.5 which solved my problem but another "wave" appears in the direction perpendicular to the flow. Do you know what can cause this and how to fix it?

Best regards,
Claire

Finally this is not a numerical problem, it comes from the bathymetry.