open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello.

I'm running a long simulation of tracer transport. It seems that I have to compute every hydrodynamic step in order to calculate tracer transport (coupling). Since hydrodynamic solution is steady-state, there is no need to do this. Since my tracer transport problem is solved for a period of few months and can be extended to years, is there a way to decouple the tracer transport from hydrodynamics?

Budo

Hello Budo,

Yes there is a frequent request to do the tracer transport without recomputing hydrodynamics, but it is not yet done. The main reasons are :

* the steady state is never exactly steady.
* saving on a file would give huge quantities of data.
* saving only certain time steps would raise problems of mass conservation (all fluxes have to be summed)
* some algorithms need to do the hydrodynamics and tracer transport at the same time (kinetic schemes for example).

For Telemac-2D you can however skip the call to the subroutine propag (in telemac2d.f) after some time, I have already done this to have a steady rotating velocity field (in my case it was initialised in the subroutine condin). It will then stop to recompute hydrodynamics. If you feel like doing this...

Regards,

Jean-Michel Hervouet