open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

I am doing my master thesis on implementing a flocculation model (PBE) in TELEMAC3d under supervision of Qilong. Seeing as the structure of the (coupled) differential equations is very similar to the 3D equation of tracers, we decided to investigate this more in the source code.

On line 2381 in Telemac3d.f main subroutine the advection-diffusion of tracers is solved using subroutine CVDF3D (line 2430). We would like to know if this call will solve the entire tracer equation (see attachment) or only part of it. In general we would like to know how the (entire) tracer equation is solved in Telemac3d, and which subroutines are used in this process.

If possible, some general information on how (non-zero) source terms are prepared and dealt with, and where this happens in the source code (we are working with implicit source terms as equations in the PBE) would be very welcome.


Thank you very much in advance

Hello,

Yes, the tracer equation is entirely solved in CVDF3D (which stands for ConVection DiFfusion in 3D), except when advection is done with the method of characteristics, in which case advection is done before. The source terms for tracers can be set in the subroutine SOURCE_TRAC, there is an example of radioactive decay.
A very important thing to know is that advection is done in a transformed mesh (with a transformed vertical velocity W*) and this does the relocation to take into account the movement of the mesh. Then diffusion is done entirely in the mesh at the end of the time step.
Depending on the advection solver chosen, the source terms are treated at the level of advection or in diffusion.

With best regards,

Jean-Michel Hervouet

Thank you very much for your quick response.

If the call to CVDF3D.f is done when we set "YAWCHU = true" and use "SLVD = SLVDSE", will this call solve the advection-diffusion equation of suspended sediments? (see attachment). I need to make sure the settling velocity is included in the equation. Are there any more keywords I need to change in the call to CVDF3D.f to make sure it solves the advection-diffusion of suspended sediments, or are the above 2 the only ones? (when I look through the code I think these 2 will suffice).

Thanks in advance.

Hello,

Yes, there should be a priori no other thing to do. In version 7.0 there is a new keyword ADVECTION-DIFFUSION SCHEME WITH SETTLING VELOCITY
The old treatment corresponds to 0, and option 1 is rather finite volume minded and splits horizontal and vertical diffusion, to group in an extra step vertical diffusion and settling velocity.

With best regards,

Jean-Michel Hervouet