open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

I am currently using a parallelized version of Telemac 3D - v7p2r2 in tidal flow conditions.

Whereas I succeed in conducting simulations with k-epsilon model on the horizontal and vertical, I cannot apply a Smagorinsky's model on the horizontal and a k-eps model on the vertical.

However, new features of version 7.2 clearly exhibits that "K-epsilon turbulence model can now be used independently on the horizontal and on the vertical".

I may have missed something. Does this option require further modification (as on numerical discretisation, advection scheme, implicitation...) ?

I would be very grateful if you could help me.

With best regards,

Nicolas.

Hello Nicolas,

Does your error happen after some simulation time or right at the start? What is the error you get?

Cheers,
Agnès

Hello Nicolas,

I ran a test case using the Smagorinski turbulence model for the horizontal and the K-epsilon for the vertical, so it does work as expected. Therefore the problem may be caused by something else in your simulation.

As Agnès said, further information is necessary to solve your problem.

Regards,

Phelype

Hello Agnes and Phelype,

Thank you for your answer. In fact, the error appears at the first iteration with the following message:
runcode:: main
|runCode: Fail to run
|$MPI_LAUNCH -np 8 /directory_name/out_simu
|~~~~~~~~~~~~~~~~
|1
|MPT ERROR: MPI_COMM_WORLD rank 3 has terminated without calling MPI_Finalize()
| aborting job
|/appli/mpt/2.15-p11386/bin/mpiexec_mpt: line 394: 50888 Killed $mpicmdline_prefix -f $paramfile

With best regards,

Nicolas

Hi,

The message:

|MPT ERROR: MPI_COMM_WORLD rank 3 has terminated without calling MPI_Finalize()

suggests the problem is in segment 3. You could check the log file for that segment. Usually it gives information there.

Kind regards,
Qilong

Nicolas,

Could you post the entire output on the screen?

What you posted is only the exit routine of python script and of the MPI implementation you are using. The actual error, if it exists, is above this.

Regards,

Phelype

Hi all,

Sorry for the delay of my answer (I was away from my office).
Indeed, the log file in segment 3 provides further information. The following message appears at the end of the log file:
FIN DE L'INITIALISATION TPXO
TRID3D : SYSTEME NON DEFINI
PRECONDITIONNEMENT 17 IMPOSSIBLE

PLANTE : ARRET DU PROGRAMME APRES ERREUR
RETURNING EXIT CODE: 2

For your information, preconditionning is set to default parameter except of diffusion of velocities: PRECONDITIONING FOR DIFFUSION OF VELOCITIES = 34. But this option, set by default, has no influence on the error message.

With best regards,

Nicolas.

Hello, Nicholas

It's telling you that your system cannot be solved using this method of preconditioning.

You can try using "DEBUGGER = 1" to discover in which part of telemac3d this error is happening, or try to change the preconditioning schemes...

Regards,

Phelype

Hello,

Try instead:

PRECONDITIONING FOR DIFFUSION OF VELOCITIES = 2

Preconditioning 17 (34 is preconditioning 2 and 17 altogether) is a direct solution on every vertical, it is used just for speed up the computation, so you can remove it to make things clearer. The problem could be due to tidal flats.

With best regards,

Jean-Michel Hervouet

Hello,

I have tried PRECONDITIONING FOR DIFFUSION OF VELOCITIES = 2.
However, it still returns me the same message:
TRID3D : SYSTEME NON DEFINI
PRECONDITIONNEMENT 17 IMPOSSIBLE
PLANTE : ARRET DU PROGRAMME APRES ERREUR
RETURNING EXIT CODE: 2

I have also investigated different options about tidal flats with no success.
Whereas tidal flats are in my model, I will try to modify the water depth to see if this problem remains.

With best regards,

Nicolas.