open TELEMAC-MASCARET The mathematically superior suite of solvers

Dear all,
I have a program with my fortran (attached file) that works seem correct in parallel mode with telemac 2d version v6p2. Unfortunately, when I launch my simulation with telemac2d version v6p3, there is an error (see attached file). I don't know why?
(In my fortran file, I would like to calculate the discharge of a defined section). Thank you very much for any suggestion.
Best regards
Thong NGUYEN
HCMUT, VN

Hello Thong,

The problem is what the listing says, you have an error of dimension at line 3108 of your specific Fortran file (however your file does not seem to have so many lines so I am not sure it corresponds). It is probably an index of array which is 0 on a processor because the point is not in this processor, or something like this. If it is an error in our own sources it could be that it has been corrected in the version 7.0 which is now available in its release R0 for tests (an extra question is why do you move to version 6.3 now that we have version 7.0 available?).

In all cases it means that you have compiled Telemac with an option that checks array bounds and it is rather costful. I do this just for developing, then you can remove it to go faster. If this option has been added when you moved to version 6.3, it is the reason why you did not see it before.

With best regards,

Jean-Michel Hervouet

Dear Mr. J-M Hervouet,
Thank you very much for your help. I found the error and already corrected it. There is a cluster at HCM City university of technology (faculty of informatic science). I can launch telemac2d with upto 32 processor in parallel mode. I will ask them to install the sofware when it's already (telemac2d version v7 for example). They just have updated telemac2d version v6p2 to version v6p3. One more time, thank you very much for your suggestion.
Best regards
Thong NGUYEN
HCMUT, VN

Dear Thong,

I am experiencing exactly the same error when adding sources in parallel mode (v6p3r2). Could you please tell me what was exactly the error you corrected?

Thank you,

Olivier Gourgue.

Dear Mr. Olivier Gourgue,
As Mr. J-M-Hervouet say, when I define the local number of point in parallel mode with function n1=GLOBAL_TO_LOCAL_POINT(n1,MESH) & n2=..........., the value n1 (or/and n2) =0 is returned if point n1 not in subdomain. So I check this number before (line 170). If n1 and n2 are different to 0, I will calculate the discharge of section (see my attached fortran) and do nothing if n1 (or/and n2)=0.
Best regards
Thong NGUYEN