open TELEMAC-MASCARET The mathematically superior suite of solvers

I want to modify source terms for tracer transport equation in TELEMAC and for sediment transport in SISYPHE. I have the source code but don't know where to start. Which subroutines do I need to alter?

Hi budo

The new sisyphe user manual is now available for download. Have a look on it it will help you to start.
Since wednesday, latest telemac documents are also available.
regards

It seems that I didn't explain the problem well enough. If I want to model decay of tracer (which can be done through source term in basic tracer transport PDE), what function (or functions) have to be modified?

Regards,
Budo

Hi Budo

For Telemac-2D, you should modify difsou.f.
There is an example of decay in comment.

Best regards

Christophe

I'd probably edit CVDFTR (or CVDFTR_POS if you're using one of the schemes for tidal flats). After solving for the new tracer distribution, you could add your decay term just by updating the object called 'F'. Others might have better ideas.

You can find CVDFTR in the BIEF library, and you might find it useful to know that the actual tracer concentrations (at time t^n+1) are stored in the double precision array F%R(*).

Hope that helps!

Chris

Edit: Christophe was faster than me, and had a more standard idea. It's useful to know about that example of decay.

Thanks Chris

I suppose you're solution give the same result but maybe in a more complex way.

For new Telemac users, BIEF Library should be modified with precautions. It's a library and the sources files are shared with the other modules.

Regards

christophe

Christophe, Chris

Thank you for the advices. I've stumbled upon the "difsou.F" file and found comment before I posted question but could not understand implementation because everything is commented in French. I've tried to translate it via Google-translate and some of them are good while others don't make sense. Is there a English translation of this file?

Also, modifying source terms should be followed with adding new keywords. What is the most effective way to do this?

Regards,
Budo

Hi budo

Not yet for the translation in English.
I don't understand what you mean with adding new keywords.

Could you clarify your request?

Regards

We (HR Wallingford) have been working on an English translation of the TELEMAC source source code, which will hopefully be made available soon.

In the meantime, I had a look at the DIFSOU subroutine, and even keeping to the French version the comments are misleading.

The key part is the comment before the example that Christophe refers to:
What's misleading is the use of SMI here. This stands for (I think) "Second Member Implicit", but means the implicit part of the RHS of the tracer equation. In fact, the RHS is really SMI*T/H, where T is evaluated at the next timestep (hence 'implicit'). To add to the confusion, SMI is passed to DIFSOU, but with the name TIMP.

To define a global source(/decay) term, you need to modify TIMP, as is done in the example commented out in the DIFSOU code. Remember that whatever you put in TIMP will be multiplied by T and divided by H to give the actual source term in the tracer equation.

I hope that makes sense. I'm happy to try to explain it again in a different way if not.

Thank you Chris. Actually that part confused me in the first place and I thought it was an error in the code. AS I remember the theory, division with H results from depth-averaging. Then the coefficient -0.0001 in the given example is multiplied by H_prop because subroutine CVDFTR divides source term with H_prop. Variable TIMP is not listed in the table of variable names description but is the function argument. It is also for the N_TRAC that I presume is total number of tracers. Right?

Since I don't understand French (Google translate has very limited capabilities), I hope that English translation will be available soon.