open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi,
I am using TELEMAC with a dischage gage as domain boundary condition.
My simulation works perfectly fine under a windows environnement (I have v5p9 release on the computer) but it shows some issues under Linux (version6). I get the following error :
" ERREUR DANS LE FICHIER DES COURBES DE TARAGE". Here is my "ctar" file :
I don't know if it's because of the difference of version (but I can't find anything new on discharge gage on version 6 user manual) or it can be a problem with the linux installation of TELEMAC ?
Please note that I also have a "fliq" (fichier des frontières liquides) file for the other boundary and TELEMAC reads it perfectly on both Windows and Linux.

Do you have any leads on that matter ? I enclosed my files .

King regards.
JM

Hi JM
Did you check there was no problem of CR/LF on the linux version?

Hi,
yes, I created a whole new ctar file under Linux then I tried to edit the working fliq file to match discharge gage syntax (all lines are ending by a LF though), but nothing new happened.
I'm wondering if it is not a problem with the Linux installation although everything but the ctar file is working.

Hi
I had a short look into your file.
Maybe you could delete the last line (which is empty).

Here is the answer we got from Dr A. Stoukov (ENSEEIHT) :

Le problème vient des compilateurs fortran GNU (gfortran ou g95) qui gèrent le end-of-file à sa façon,
différente de celle du compilateur Intel (chez MM) ou du CVF (dans les salles Windows).
En effet,dans read_fic_curves.f, le résultat de l'instruction
READ(NFIC,FMT='(A)',END=1000,ERR=999) LIGNE
n'amène pas le programme au label 1000 quand le fichier est lu jusqu'à la fin mais le programme va au label 999.
Telemac a été développé avec Visual DEC Fortran, le prédécesseur de CVF et donc il se comporte bien quand on utilise les compilateurs Intel et CVF .

Pour remédier au problème de lecture de la courbe de tarage avec le gfortran sous Linux, il suffit d'ajouter une ligne supplémentaire avec un #
# COURBE DE TARAGE FRONTIERE 2
#
Q(2) Z(2)
m3/s m
38.39 89.31
275.6 90.65
591.8 91.21
1052. 91.51
1291. 91.99
2250. 93.22
3131. 94.05
4800. 95.72
#

I won't translate the whole email, but adding a comment line at the end of the file solved the problem.

Thank you!

Louis Poirel
X-ENSEEIHT student
(working with original poster on this project)

Hello,

Thank's for this answer. A solution would be putting also ERR=1000 in the READ statement, though this is a bit dangerous. I'll look at that.

Jean-Michel Hervouet