open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

I'm trying to use NEXT COMPUTATION in Tomawac.

I have made some tests with the validation case "whirl current" where some commented line are dealing with NEXT COMPUTATION.

When I use a GLOBAL RESULT FILE as PREVIOUS COMPUTATION FILE, I have this error :



I was not sure of the type of file to use so I have also tried with a GLOBAL RESULT FILE BINARY as PREVIOUS COMPUTATION FILE BINARY and I have this error :



Next I have read some "old" post in the forum and if I have well understood, the conclusion was NEXT COMPUTATION isn't implemented in Tomawac. But this old post were about previous versions (V6P0 & V6P1) :
www.opentelemac.org/index.php/kunena/19-...bal-result-file#5558
www.opentelemac.org/index.php/kunena/19-...arallel-version#3235
I have also read a more recent post, but there is no conclusion :
www.opentelemac.org/index.php/kunena/19-...-limwac?limitstart=0

So is here my question, is NEXT COMPUTATION is really implemented in Tomawac? If yes, how to use it? I mean I have tried with the validation case but what I did wrong?

Regards

Hi Nicolas,

I never got a straight answer to my case, although when I first posted my problem, I had issues with my own fortran file and couldn't address straight away the problem with the Tomawac code.

It has been a while, but the outcome of my personal study is that NEXT COMPUTATION in Tomawac does not function properly and spent a lot of time trying (unsuccessfully) to figure out what was wrong and concluded that the best compromise is to set as initial conditions, the boundary condition of the end of the previous computation. Of course, my domain is quite small so it does not make a big difference.

I hope that a Tomawac developer can shed some light into this issue.

Best Regards,
Costas

Hello,

It will be looked at next Monday. Officially the restart should work, the file to use as PREVIOUS COMPUTATION FILE is indeed the GLOBAL RESULT FILE, and it is a Selafin file with a 3D mesh containing the wave action energy F, it must work also in parallel.

Regards,

JMH

Hello

I have tested a next computation on the whirl_current test case. And found two possible problems.
During the first computation you have to mention the two lines

SORTIE GLOBALE A LA FIN : OUI
FICHIER DES RESULTATS GLOBAUX = 'tom_whirl.pre'

otherwise the file might be empty

Then when you want to make the next computation,

SUITE DE CALCUL : OUI
FICHIER DU CALCUL PRECEDENT : tom_whirl.pre
FORMAT DU FICHIER DU CALCUL PRECEDENT = 'SERAFIND'

The default for that keyword is not correct, it will be changed in next version.

regards

Thierry

Hello Thierry,

I didn't have the correct FORMAT DU FICHIER DU CALCUL PRECEDENT.
But when I make the correction for this key word, it doesn't work and I have the next error :



Something is crashing in LECSUI.f. I have put some flags in LECSUI.f and it seems the block IF(COURAN) THEN is at the origin of the problem.
Do you have the same mistake (reading your message I think that no)? What do you think about it?

Regards

Hello Nicolas,

I don't have the same misstake.
Which version are you using ? Could you upload all your files that i can be sure that we are doing the same calculation ?

regards

Thierry

I'm using the V6P3 version.

Here all the files I'm using :



Regards

I have just tested on another computer, there is the same mistake, but not at the same time (in the Fortran). Here the error :



Very strange...

Regards

Seeing your file i wonder if you did the case in 2 step
During the first step (run) you should build a file called tom_whirl.pre
During that step the lines
/SUITE DE CALCUL : OUI
/FICHIER DU CALCUL PRECEDENT : tom_whirl.pre
/FORMAT DU FICHIER DU CALCUL PRECEDENT : 'SERAFIND'
should be commented

and during a second step (run) the other lines should be commented

/SORTIE GLOBALE A LA FIN : OUI
/FICHIER DES RESULTATS GLOBAUX = 'tom_whirl.pre'

Is that what you did ?

Doing that the run works with my version V7P0.

I will try with your version to see what's happening

regards

T.

Yes I did the two steps. The first one was successful and I send you the second one.

Nevertheless, I keep this keywords in the second step :
SORTIE GLOBALE A LA FIN : OUI
FICHIER DES RESULTATS GLOBAUX = 'tom_whirl_2.pre'
Because I would like to use again next computation. I have a long time to compute and I want to cut in shorter runs.

Regards.