open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

I'm trying to inject about 34 000 sources in a TELEMAC 2D model to simulate the runoff due to the increase of groundwater level above the ground. I wonder if someone has ever tried this kind of simulation and if there is a limit of source number in TELEMAC 2D ?

Best regards,
Rho

Hello,

This is certainly not the way to do this, point sources are rather a small number of pipes bringing water to a few points. What you plan to do is closer to bringing water on every point, something rather like rain. You could declare it as rain and see how it is treated in subroutine prosou.f, where the rain is transformed in an array that can have variations in space. Changing directly the right-hand side of the continuity equation is also possible (array SMH), unless you have tracers, in which case it is more difficult to implement.

With best regards,

Jean-Michel Hervouet

Hello,

Thank you for your explanation. However, I'm not that expert in Telemac and I don't know coding in FORTRAN. Do you have an instance of modified subroutine prosou.f ? Or an exemple of the array SMH ? For information, I don't have any tracers.

In the meantime, I run an other simulation with only 5 sources and 20 cm of initial water in most of my modele to use it as a hotstart later. The simulation stopped after only 6 hours of simulation over 7 days with this message :
"@streamline::scaract:: the number of traceback interface crossings igen > 99
plante : arret du programme apres erreur
sortie de PVM : appel de P_EXIT"
The error message is also in attachment. Do you know what this mean ?

Thank you in advance for your help,
Best regards,
Rho

Hello,


This error generally occurs when the velocity becomes very high, so probably the model is crashing somewhere.

As for modifying prosou you can inspire for the rain, the mm/day rain is transformed into m/s, and the right-hand side of the continuity equation, called SMH, is in m/s. So where the rain is added to SMH in prosou, you can add to SMH a value in m/s corresponding to your source term. Then the discharge entered at every point will be this value in m/s multiplied by VOLU2D which is the integral of test functions, more precisely VOLU2D%R(I) for a point I, and VOLU2D is an area, so it will do m3/s, a discharge.

With best regards,

Jean-Michel Hervouet