open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello Telemac users,

I have been running without any problem telemac and tomawac uncoupled (attached steering files), but when I couple them, I get the following error
[/CALLING PROPAG CPSTMT : FORBIDDEN CASE FOR X AND Y: X=MSEG ELEMENTS 11 AND 11 Y=AM1code] I also attached the full error message. To solve that issue, I have tried: *different types of solver *reduce the time-step (down to 1s) But that did not work. I have seen the same error on this forum, but for 3D simulations, and that was resolved by using the hydrostatic pressure. Here, the bathymetry includes some sand banks (see attached), but not sure if that could be the issue? Would be glad if someone could give me some indication on how to solve that please?[code][/CALLING PROPAG CPSTMT : FORBIDDEN CASE FOR X AND Y: X=MSEG ELEMENTS 11 AND 11 Y=AM1code]
I also attached the full error message.

To solve that issue, I have tried:
*different types of solver
*reduce the time-step (down to 1s)
But that did not work.

I have seen the same error on this forum, but for 3D simulations, and that was resolved by using the hydrostatic pressure. Here, the bathymetry includes some sand banks (see attached), but not sure if that could be the issue?

Would be glad if someone could give me some indication on how to solve that please?

Hello Claire

Sorry for this late reply. We were (and still) overloaded.

This looks to be an non-compatible choice of numerical parameters between Telemac and Tomawac. I see that you have lots of commented lines in your steering files, but also and which is strange, that Telemac is not able to read the type and the maximum size of matrix AM1. This latter remark could presume some issues with allocation at the beginning of the computation.

Did you fix the problem, meanwhile?
If not can you send us the case please.

Have a nice holidays.

with my best regards

Riadh

Hi Riadh,

And thanks a lot for your reply.
No I have not fixed the problem yet.

By amending the steering file (mainly changing the projection type), I managed to get the following error message:
"Note: The following floating-point exceptions are signalling: IEEE_UNDERFLOW_FLAG IEEE_DENORMAL". Which I understood may be linked to numerical problems or a lack of using doubles.

In case it helps, I have found again the topic dealing with a similar error:
#6369

Thanks a lot for your help, have a lovely winter break too.

Regards,
Claire

Hello Claire
and all my best wishes for this new year.

Can you please send me the following missing files (you can send them directly to my email box):
1- Gab_tide_Spinup10d.slf
2- hf.ES2008.out
3- uv.ES2008.out

kind regards

Riadh

Hello Claire

I'm replying on the forum in order to give the remaining users the the fix for this bug.
I'm attaching a subroutine that is compatible only with v7p2. You can add it to your fortran file and hopefully it will fix the issue.

Let me know

kind regards

Riadh

Hi Riadh,

Thanks a lot for your reply and the fortran scripts -sorry it took me so long to run it!

Unfortunately, I got another error, please find attached the error message.
I have tried different wave discretisations, time steps but get the same error.
And Tomawac on its own is running fine.

Any advise please?
Kind regards,
Claire

I forgot to add:
I also have an issue with the time in Tomawac: I guess the timestamp are not correctly saved and hence not correctly displayed when reading outputs with Blue-Kenue or when extracting a time-series at one node.
-I always had this error with Tomawac-

Thanks,
Claire

Hello Claire






Here attached another version of the fix. I see that Tomawac, eventhough called by Telemc2d, is not considered in the forcing of momentum equation. This is a bit strange.
please add princi.f as a fortran file (keyword FORTRAN FILE=princi.f)
I have run your case with a coupling period of 1 and it runs well but it is very costly. So I changed the period to 2 and it runs a bit quicker.
I have changed also the time step to 10 instead of 30 (also in tomawac), which are really big. If you are not satisfied, decrease it again.
What I suggest, if Tomawac runs well in standalone and if it physically relevant, include wave effect through input files rather than coupling with Tomawac.

I hope that this helps

with my best regards

Riadh

Hello

There was an error in computing the radiation in spherical coordinate, here is the routine corrected for the V7P2 version.

this routine will be in the trunk very soon

Thierry