open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi

I'm working on a Telemac2D model (v8p4) with a larger than normal number of sources, 180 sources with 3 time varying tracers. Following advice on previous posts I've modified both debsce.f and trsce.f to have greater than the standard maximum number of sources of 99. I've recompiled Telemac after modifying declarations_special.F after modifying the length of MAXKEYWORD = 3000 to read the greater number of sources, coordinates etc in the cas file. And I've also modified read_fic_sources.f to change the length of SIZELIGN = 6500 to read the full number of columns in my time series for the sources.

At first I thought this was all successfully running, as the listings was showing that it was reading the source files and listing both the flow rates and all three tracers for the 180 sources correctly. However, the modelled crashed following extreme temperatures coming from one of the sources. Where I should have had temperatures in the low teens, it was coming out in values closer to the hundreds. Also I realised this particular source shouldn't have been discharging anyway as the flow rate should have been zero at the start of the model.

My inclination is despite modifying the fortran to read these extremely wide text files, there's potentially other keywords that need modifying somewhere, and its not applying the right source terms to the right sources.

Does anyone have any suggestions?

I've attached the fortran plus an example of the sources text file.

Many thanks

David

Update:

Think I've found the problem. Nothing to do with the code modifications or size of the file.

Looks like where I've defined the coordinates of the source points, rather than a global number, some of the sources are close together and where the code finds the nearest point in the mesh to apply the source, its added multiple sources to the same node. It doesn't appear to like that and when I move the sources to adjacent nodes the problem disappears.

Conducting further tests, but that appears to be the solution so far.

Hello David,

Indeed, if you give coordinates of sources, TELEMAC-2D will find the closest node to apply sources flowrates. So if coordinates are in the neigbour of the same node, the sources may be confused.

What about the distribution of your punctual sources? Can they be included inside a polygon?
In that case, do not forget sources by region feature in TELEMAC-2D (see SOURCE REGIONS DATA FILE keyword).

Chi-Tuan