open TELEMAC-MASCARET The mathematically superior suite of solvers

Dear all,

I am trying to model a coastal area with a river discharge. Incoming water has salinity=0 (LITBOR=5) but temperature of this fresh water should be the same as it is in the modelling area (LITBOR=4). How it can be implemented? Thank you.

Best regards
Alexander

Hello,

It cannot be really implemented, as it is against the nature of equations, and you would get immediately errors of mass. It is the hydrodynamic conditions that should say what is the condition on tracers. What you can do is prescribe on the border a value taken inside the domain, this would be to do in subroutine bord3d.f where the values of TABORL are set. TABORL is the prescribed values of tracers on boundaries. This is easy in scalar mode, a bit more difficult in parallel.

With best regards,

Jean-Michel Hervouet

Dear Jean-Michel,

Thank you very much for the prompt answer. Sorry, but can I ask why it is against nature of equations? Why the Neumann boundary condition for one tracer cannot coexist with the Dirichlet boundary condition for another tracer on the same open boundary?

Best regards
Alexander

Hello,

It would raise a problem with the advection terms. It is true that diffusion terms would not be hindered because the Neumann boundary condition is natural for them. On the contrary the advection solvers use the fluxes at the upstream boundary and multiply them by the locally prescribed value of tracer to get the total mass of trcer that enters the domain, so you need an upstream condition because everything depends on this condition. If the tracer is free here, or given with a loose condition like Neumann, its value is somewhat unpredictable.

With best regards,

Jean-Michel Hervouet