open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi,

I have a working TELEMAC 2D model which I am currently running using the Finite Element method however when I change the configuration file to run Finite Volume (EQUATIONS = 'SAINT-VENANT FV') I receive the following error:

ITERATION 0 TIME: 0.0000 S
T= -1.0000000000000000 OUT OF RANGE
OF THE SOURCES FILE

PLANTE: PROGRAM STOPPED AFTER AN ERROR

I have tried to trace the problem by running the debugger but still can't find the problem (log files from both failed runs attached).

The error refers to the source file but this is the same file run by the FE model which works fine. Dos anyone have any suggestions on what the problem might be?

Many thanks
Jonathan

Hello Jonathan

This is not an error linked to finite volumes. The error message tells that the time is -1 which is a bit wierd.
I think that you didn't add properly the good options related to finite volumes.
You have to add at least the following keywords
VARIABLE TIME STEP = YES
DESIRED COURANT NUMBER =0.9
FINITE VOLUME SCHEME = 1 or 5

If it does not fix the problem, please post us the case.

with my best regards

Riadh

Hi Riadh,

I've just checked and I am using those keywords. I initially thought the problem might be related to my compilation but I get the same error running on my university cluster (manually compiled) and on my laptop (compiled using the automatic installer).

My .cas, source and liquid boundary files are attached.

Many thanks
Jonathan

Hi Jonathan,

I'm not an expert on the supported features of the Finite-Volume-Schemes. But so far I know, turbulence models are not supported. So you may have many options in your cas file which are not supported/implemented.

It is easier to start with a minimum cas-file, there are some Finite-Volume test cases in the examples/telemac2d folder. I hope that this can help as a starting point.


Best regards

Leo

Hi Leo,

Thank you for your suggestion. I have removed most of the keywords (so TELEMAC will use default values) and I still get the same problem.

As the error message suggests he issue occurs when I try and use a sources file so I am at a loss as to what the problem is, considering I am following the format specified in the user manual and the file works when running a FE simulation.

Many thanks
Jonathan

Hi Jonathan,

the question is, are source files supported by the finite volume schemes?
Does the simulation run without the source file?

Best regards

Leo

Hi Leo,

Yes, the simulation does work without the sources file. You may be correct that a sources file cannot be used with the FV scheme but this isn't mentioned anywhere in the user manual.

Regards
Jonathan

Hi Jonathan,

Yes, the situation is unclear for the users. I've tested the HLLC scheme a few years ago and had some problems because I've used an unsupported friction law.

I think the FVMs are still under development and the FEMs are a better choice for daily work and more complex tasks (Tracer, Tide, Wind). However, there are some developments to include more processes in the FVMs of TELEMAC. Maybe they are already supported in the trunk version?


Best regards

Leo

Hello Leo and Jonathan

Yes, as said by Leo, a major part of source terms are not connected so far with finite volumes.
If you are interesed by a quick test, I integrated all these source terms recently to the trunk.
Hopefully all these new developments will be in next release.

With my best regards

Riadh

I have exactly the same problem on v8p0r3. The finite volumes work when setting constant WATER DISCHARGE OF SOURCES. Though as soon as I add in a SOURCES FILE I get the same error
T= -1.0000000000000000 OUT OF RANGE
OF THE SOURCES FILE

Is there any way I could modify some subroutine to make this work? Could anyone please give me a hint?

thanks,