open TELEMAC-MASCARET The mathematically superior suite of solvers

Dear all,

We were testing the new binary boundary file functionality in 3D v7p2r0, and we found that only a small number of boundary points were actually given values, and the other boundary points received zero velocity and tracer values. After some code-digging, we found that the problem may be due to line 227 in read_bin_frliq.f:
Sometimes, there are still non-zero boundary points after the first point with NUMLIQ%I(IPTFR).EQ.0 and so those are skipped. If we change the EXIT command to cycle, i.e.then everything works alright.

We generated our binary boundary file using the Pythonscript convertToBND.py from v7p2r0.

Anyone else experienced this issue?

Hello,

Yes you're right about this, I will correct this bug on the trunk.
When I made the tests, the solid boundary points were always at the end of the list so I did not notice this mistake. Thank you for reporting it.

Best regards,
Agnès

Hello,

I would like to know if this feature is available in parallel mode ?
I have some troubles with partel to split the BND file...

Regards,
Laurent

Hello,

Normally this feature should work in parallel. Which version of TELEMAC are you using exactly?

Cheers,
Agnès

Hello Agnès,

Thank you first for your reactivity.
I've tried both with v7p2r1 and v7p3r0.
I use "convertHYCOM.py" and "convertToBND.py" scripts provided with the distribution, I have the same problem than this described in this topic :
www.opentelemac.org/index.php/kunena/21-...dary-data-file#28671

Can you please provide the expected right format of the binary boundary file or a sample of file which can be split by partel ?

Best regards,
Laurent Leballeur

Hello,

Could you send me your boundary file so I can see what it looks like?

Agnès

Hello,

You can use the boundary file provided by Luz (shortoutput.slf) in this topic:
www.opentelemac.org/index.php/kunena/21-...dary-data-file#28671

I also tried to re-generate it by myself with:
convertHYCOM.py --from 2012-01-10 --stop 2012-01-12 --bl 50,-4 --tr 51,-3 -r hycom_test.slf
convertToBND.py DS.cli bathy_Dart_fin.slf t3d_hycom_test.slf out-bound.slf
The resulting file is not the same (number of vertical levels) but still not partition by partel.


I noticed a possible inversion between 3rd and 4th column of the connectivity table and maybe another strange thing in the order of the horizontal points for each layer.

Thank you for your help.
Laurent

Hello,
I think that you forgot to convert the hycom file to spherical coordinates. There is also a script available for this.

Cheers,
Agnès