open TELEMAC-MASCARET The mathematically superior suite of solvers

I have been trying to run a simulation in an estuatry. However, after 230seconds I continue to get the error:

PROPAGATION AND DIFFUSION WITH WAVE EQUATION
TRID3D: SYSTEM ILL-DEFINED
PRECONDITIONING 17 IMPOSSIBLE

From previous post in this forum, I have set the "preconditions" to 2 and have tried reducing the time-step but the error persist.

Please find attached my files.

Ps: I am running the TELEMAC3D on v7p2r3

Hi,

Do you get the same error with version V7P3? Before the crash, do the simulation results look fine or are there places where there are strange velocities, tracer or water depth values?

Cheers,
Agnès

Hello Agnes,

I do not have v7p3 running on my PC. However, I get strange velocity results.

Discharge is only recorded at Boundary 2. Boundary 1 and 3 show -0.000000 m3/s all through the simulation until the crash.

Ps: Boundary 1 and 3 condition = "flow and tracers" while Boundary 2 condition = "elevation and tracers"

Cheers.

Hi,

We have a known issue regarding the use of tracers that influence the density and tidal flats. You are maybe facing this problem but we do not have a fix for this bug right now. To see if this is the case, you could try to modify the mesh so as not to have tidal flats anymore, and see whether you get the same crash or not.

Agnès

Ok Agnes, I will look into that.
Thank you.

Assuming I take out the tracers and use discharge and elevation as my boundary condition, will I still face this issue?

If you take out the active scalar you should also avoid the crash, if you're really facing the issue I mentioned.
Agnès

You could also try to use the keyword OPTION FOR THE TREATMENT OF TIDAL FLATS=2 to see if it prevents the crash.
Agnès

Thank you Agnes. I will explore these options and feedback.
Chijioke.

Hello Agnes,

I still haven't been able to complete the simulation.

I have removed tracers from the boundary conditions. Also, adding "OPTION FOR THE TREATMENT OF TIDAL FLATS=2" stops the simulation from running.

Hello,

Ok, I learned later that the option for the treatment of tidal flats = 2 is less good than the option 1 so you should definitely stay with option 1. I think you have to remove the active tracers from the simulation. Now I see that your density law is 0 so the sediment does not modify the density. I thought your problem was related to this but if it is not the case, I suggest you look at the places where the crash happens: does the mesh look good in these areas?

Agnès