open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi,

I try to make a tomawac simulation but i was an error when i put the keyworld 'Consideration of propagation'.

I thought that it was my mesh which was so big (108000 points). I create a new mesh with 66000 points but the problem is always here.

If i don't put the propagation i can make my simulation but when i put the propagation, the run fails after the first iteration when it calls semim.f.

I would like to know what is the propagation?? What are the differences if i put the propagation or if i don't put it?

I use tomawac v6p2 and i have the manual 6.1. Are there new keywords to define the propagation in the tomawac v6p2 but which are not in the manual 6.1?

Here, you can see my error which appears when i put the propagation:
================================================================================
TEMPS : 0.0000 SECONDES 0IEME ITERATION

---

RETOUR DE IMPR
APPEL DE LIMWAC
RETOUR DE LIMWAC
APPEL DE INIPHY
RETOUR DE INIPHY
APPEL DE PRELEO_MPI
RETOUR DE PRELEO_MPI
APPEL DE DUMP2D
RETOUR DE DUMP2D
APPEL DE BIEF_DESIMP
RETOUR DE BIEF_DESIMP
APPEL DE ECRSPE
RETOUR DE ECRSPE

---

CALCUL DU CHAMP CONVECTEUR ET REMONTEE DES
CARACTERISTIQUES
APPEL DE PREPRO 1
FREQUENCE 1 , NOMBRE DE SOUS PAS RUNGE KUTTA : 136
FREQUENCE 2 , NOMBRE DE SOUS PAS RUNGE KUTTA : 124
FREQUENCE 3 , NOMBRE DE SOUS PAS RUNGE KUTTA : 113
FREQUENCE 4 , NOMBRE DE SOUS PAS RUNGE KUTTA : 102
FREQUENCE 5 , NOMBRE DE SOUS PAS RUNGE KUTTA : 93
FREQUENCE 6 , NOMBRE DE SOUS PAS RUNGE KUTTA : 84
FREQUENCE 7 , NOMBRE DE SOUS PAS RUNGE KUTTA : 77
FREQUENCE 8 , NOMBRE DE SOUS PAS RUNGE KUTTA : 70
FREQUENCE 9 , NOMBRE DE SOUS PAS RUNGE KUTTA : 63
FREQUENCE 10 , NOMBRE DE SOUS PAS RUNGE KUTTA : 58
FREQUENCE 11 , NOMBRE DE SOUS PAS RUNGE KUTTA : 52
FREQUENCE 12 , NOMBRE DE SOUS PAS RUNGE KUTTA : 47
FREQUENCE 13 , NOMBRE DE SOUS PAS RUNGE KUTTA : 43
FREQUENCE 14 , NOMBRE DE SOUS PAS RUNGE KUTTA : 39
FREQUENCE 15 , NOMBRE DE SOUS PAS RUNGE KUTTA : 36
RETOUR DE PREPRO 1
APPEL DE LIMWAC
RETOUR DE LIMWAC
APPEL DE PROPA
RETOUR DE PROPA
APPEL DE SEMIMP
RETOUR DE SEMIMP
APPEL DE LIMWAC
RETOUR DE LIMWAC
APPEL DE PROPA
RETOUR DE PROPA
APPEL DE SEMIMP
RETOUR DE SEMIMP
APPEL DE LIMWAC
RETOUR DE LIMWAC
APPEL DE PROPA
RETOUR DE PROPA
APPEL DE SEMIMP
_____________
runcode::main:
/jeliazovski/Documents/CONGO/test_flo/tomawac:
|runCAS: fail to run
| /usr/bin/mpiexec -wdir /jeliazovski/Documents/CONGO/test_flo/tomawac/cas_test.txt_2013-02-08-10h40min14s -n 1 --host nemo /jeliazovski/Documents/CONGO/test_flo/tomawac/cas_test.txt_2013-02-08-10h40min14s/out_Test_bis

I put my steering file in attachments.

Thanks for your help
Regards

Florian

I have an other hypothesis to explain my error.
I try to put the propagation (TRUE) but i don't put my WIND FILE and the simulation runs.

Perhaps, my error comes of my wind file.
I build a wind file with the WAM format.
Someone can look my wind file and say me if it exists an error, please?

Thanks

Florian

Hello,

The format of your file could be the cause of your problem indeed. You cannot expect TELEMAC to know about all formats in use (WAM is one of many), just like we cannot expect everybody to comply with the TELEMAC files.

I would therefore suggest you try with a proper (TELEMAC) file format and let us know if you still have a problem.

Please note that models of a few million elements is now the norm. The limit is only down to the computer you use (and whether you can do parallel).

Good luck.
Sebastien.

Sebastien,

Thanks for your answer.
In the tomawac manual, it exists 4 different formats.I use the first: WAM-Cycle4- formatted file. I just wanted to know if someone had already used this format in tomawac simulation.
Apparently,my question was stupid.

Sorry

Florian

Hello,

Actually in the next version and for the sake of parallelism, the requested format will be the Telemac format (called serafin or sometimes selafin, the format is given at the end of the "Guide to programming in the Telemac system" which is available in this website) and the mesh will have to be the same. As a matter of fact in parallel this file will be split into subdomains, as for the geometry file.

In the current version WAM-cycle 4 files are indeed allowed, though I am not sure that there is a test case checking this, and parallelism is not very well taken into account, as all the file is considered for a projection on a sub-domain, which is not optimal.

Regards,

JMH

Thank you Jean-Michel,

indeed, after double checking, Florian seem to have the correct file format (Appendix 8 of the TOMAWAC user Manual) and the correct settings in your CAS file for the current version of TOMAWAC.

So at this stage I can only refer your question to other colleagues -- who ever has tried wind with WAM formats..

Sebastien.

Sebastien,

Thanks for your help.
I try to change my wind file in order to find a solution.

I wait the opinion of your colleagues.

Regards

Florian

Sebastien,

Do you have some news about my wind file which is in WAM format?

Good afternoon

Regards
Florian

Hello,

There were former test-cases with the WAM-4 wind file format, but I cannot find any recent one, that would still work. I can try to have a look at your wind file to see what is the problem when SEMIMP uses your wind file (semimp is the routine where the source and dissipation terms are calculated). As this option was not tested recently, there might be some bugs.
Otherwise, as Jean-Michel said, you should try to put your wind in Selafin format : for the next version of TOMAWAC, the actual problems with the wind file format will be solved but only the selafin format will be allowed.

Regards,

Elodie

Elodie,

Since friday, i try to put my wind file in SELAFIN Format as it says in tomawac manual but i have some difficulties to do this. Have you got an example or a model to help me? Is it the same format that the boundary conditions file?

To create my wind file in WAM format, i write a script where i respect all the formats which are indicated in the manuel. I don't understand the problem when SEMINMP uses my wind file.
When i put the keyword "PRISE EN COMPTE DE LA PROPAGATION" with a stationnary wind, the simulation has not bugs. When i put my wind file and i don't put the keyword "PRISE EN COMPTE DE LA PROPAGATION", there are no bugs. The simulation fails when i try to put my wind file and the keyword. I put the debugger and i see that it fails when it calls SEMIMP.

You can test my wind file?


I have another observation. Tomawac fails when i try to run in parallel mode. But when i put tlemeac2d and tomawac together, i can work in parallel mode and there are not problems during the simulation. Is this normal?


Regards

Florian