open TELEMAC-MASCARET The mathematically superior suite of solvers

I could attach all the files (the wind file is to large). In attachment you can find the mesh, boundary conditions and other files.

Kind regards
Joren

Could you joint at least a small part of the wind file (just to allow us to run few time steps)
Regards

I have split the 1.3GB windfile and try to attach it the with other files you might possibly need but the file is to large. Is there an e-mail address I can send them to?

Thanks in advance for your time

After having a look into your steering file and the fortran file I think i've found the problem...
In the limwac subroutine, you try to use a slf file: spectral_bc_bcg_10min.slf

As this file is not described in the steering file, the file is not partitionned so in parallel, each processor try to open this file and this is impossible, particularly on windows...
this explain why it's work on sequentiel and not in parallel.
You should probably declare the file in the steering file, probably in the keyword BINARY FILE 1.
Then during the preparation of the computation, this file will be splitted in the same way than the subdomain and there will no longer have concurrent access to the same file...

Hope this helps

Regards

Hey Coulet

Thanks for your help so far. I learned a lot from it. I can however still not find a solution. If your theorie is correct and a 'BINARY DATA FILE 1' is missing from the steering file it would not be possible for the old grid to run using the same files. I tried adding a liquid boundary in the old network (to see if multiple boundaries could be the problem) and this also worked.

Kind regards
Joren

Hello

this looks more like a problem of coordinates of sources.
In parallel, source coordinates must absolutly be coincidant with the coordinates of one node of the mesh.
Verify this point and I hope that this will fix your issue.

kind regards

Riadh