open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello all,

I am troubleshooting an error with a case that involves imposing spectra from an offshore model to a coastal one (similar to the impose_spectra example). I have pinpointed the issue at line 222, trunk version):where division by zero occurs. The problem originates at line 210:In my spectrum file BC_X(IPOIN) is never equal to 0.0 so BC_NF stays 0 as originally declared in line 204. Here is the list of the BC_coordinates:


Is there a problem with my spectrum file structure? I do not have a means to visualise the spectrum it is not easy to troubleshoot. Can anyone provide more insight into the issue? Any help would be greatly appreciated.

Best Regards,
Costas

PS. I cannot attach the spectrum file (probable because it is quite large ~6mb), so I will need to find another way to upload.

Download link for the spectrum file (OneDrive):
1drv.ms/u/s!AlgdD0M4jdq2jDSm7dgfoOoRZc3I
I hope it works,
Costas

Hello Costas,

You are right (once again) there is a problem. We assumed that there is always an angle of 0 degree in the discretisation of theta but it can also be epsilon (your case) so i will change BC_X.eq.0 by abs(BC_X).lt.epsilon which is better in any case.

thank you

best regards

Thierry

Hello Thierry,

I applied your tentative solution (using eps=1.D-06) and it works. This issue puzzled me for days as the case initially worked. Thank you for your insight and help.

Regards,
Costas