open TELEMAC-MASCARET The mathematically superior suite of solvers

Dear friends,

I am facing this strange problem and stuck in it.

When I run t2d simulation on a given mesh, the simulation is Ok.

However, when I try to simulate the coupled t2d and sisyphe with same mesh, the program compiles but the simulation stops at the first time step with the error message 'out_telemac2d.exe stopped working'.

I tried to debug with Visual Studio, it says 'Unhandled exception at 0x00007FFB19061CD0 (ntdll.dll) in out_telemac2d.exe: 0xC0000374: A heap has been corrupted (parameters: 0x00007FFB1909ED40)'.

Can anyone please suggest how can this problem be resolved?

Thank you.

Regards,
Saroj

Hello Saroj,

I am not sure what could be happening.
Could you please attach the files required? In order to run the simulation by myself.

Best Regards,
José Andrés.

Dear Jose,

Thank you for your reply.

Please find the attached files of the simulation.


Regards,
Saroj

Hello Saroj,

I can't reproduce the error above by running your case, but I have NaN values appearing at the first iteration of the coupled simulation.

The problem comes from your sisyphe boundary conditions (8th column of your sisyphe boundary conditions file which deals with the bottom elevation conditions on the boundary nodes, ie LIEBOR).
LIEBOR=2 means that you don't allow bottom evolutions on these nodes (your actual setting).
LIEBOR=4 means that you let the model computing the bottom evolution.
LIEBOR=5 means that you prescribe the bottom evolution.

NaN values appear on these boundary nodes because you're not allowing bottom elevations (LIEBOR=2 wall conditions) that lead to an incompatibility during the resolution of mass conservation over these boundary nodes. If you set LIEBOR=4 (free bottom evolutions), your case runs fine.

I advise you to read the section 1.2.2 of sisyphe user manual on this topics! It's the same principle than for telemac2d but for sediment quantities.

Your sisyphe steering file is also weird. It doesn't mention which processes you're considering (bedload, transport formula, Shields parameter, suspended load...), low density value of 1410kg/m3 (without porosity), morphological factor? ...

I hope it helps.

Regards,
Laurent

Dear Laurent,

Thank you for looking into this problem. As you suggested, the problem was in the sisyphe boundary condition file. After changing LIEBOR=4, the simulation now runs fine.
I was actually thinking that the problem is related to some memory allocations in the T2D source code. I didn't think of the boundary condition.
I now have read the topic on boundary conditions in T2D and SISYPHE manual.

Thank you.

Regards,
Saroj