open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

I am trying to understand if the particle tracking module in Telemac3D includes diffusion and shear dispersion. I understand that diffusion and dispersion are both present in the turbulence models, but I don't understand if the particle tracking module itself includes these terms. Are the drogue trajectories determined based solely on advection, or is the horizontal diffusivity used in determining the particle trajectories?

Any insight or suggested references would be helpful. I have Jean Hervouet's book, but I have not found the answer to my question within this text.

Thank you,
Taylor

Hello Taylor,

I am not a specialist of this feature. I have asked questions to a colleague of mine who developed oil spill model during his PhD a few years ago.

Stochastic diffusion coefficient is available in TELEMAC-2D for particle transport (see the paragraph talking to the keyword STOCHASTIC DIFFUSION MODEL in chapter 12 Particle transport and lagrangian drifts).

It is also available for oil spill in TELEMAC-3D.
There is no such keyword in TELEMAC-3D but when calling DERIVE subroutine in TELEMAC3D subroutine (between variables NPOIN3 and SVIDE), the hard coded value 0 is used. If you change it to 1, it seems that you can activate stochastic diffusion model in 3D.

FYI, Jean-Michel Hervouet's book was written in 2007 without update since ever and the developments dealing with algae or oil spill transport were done after.

Hope this helps,

Chi-Tuan

Hi Chi-Tuan,

Thank you for your reply. I have a clarification question, do you mean that by changing the value from 0 to 1 you can activate the stochastic diffusion model in 3D for oil spill modeling only, or can this be used for drogue transport in 3D as well? I ask because my Fortran file currently does not include the subroutine DERIVE. I just have FLOT3D.

Thank you,
Taylor

Changing the value 0 to 1 as written in my previous post is to activate stochastic diffusion model when using drogues but not oilspill. You can see the condition a few lines before the call to DERIVE in telemac3d.F subroutine:

Oil spill is steered in other parts of the code.

You should include TELEMAC3D subroutine to your FORTRAN FILE (folder or file) and do the change manually.

Hope this helps,

Chi-Tuan

Thank you, Chi-Tuan,

I tried to turn on stochastic diffusion for drogues in telemac3d (version 8.4.0) and got a
"forrtl: severe (174): SIGSEGV, segmentation fault occurred". I run a model on 64 cores.
The model works perfectly without stochastic diffusion in hydrostatic and non-hydrostatic modes using versions 8.3.2 and 8.4.0. What can be the reason?

Thank you.

Regards,

Alexander

Hello Alexander,

Sorry, it is a bug, I have not tried my suggestion: if using stochastic diffusion, viscosity for velocities has to be given as argument of DERIVE subroutine and not SVIDE.
So rather than initial

you should replace by

Using TELEMAC-3D example particles seems to be OK without segmentation fault.

Hope this helps,

Chi-Tuan

Thanks a lot!

It works for me as well.