open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello everyone,

I am setting up a 3D hydrodynamic model with Sea Level open sea boundary condition. In order to get rid of superficial flow (eddies) in the vicinity of the ill posed boundary, I aim to introduce the spatially varying horizontal viscosity coefficient field with high values of visc. coeff. along the open boundary; just to have some head loss there.
Do you have any idea to implement this idea in Telemac3D? I assume that the viscosity field should be introduced via the geometry file.
I see that in Telemac2D it can be done through modification of CORVIS.f which receives the coordinates of a polygon with a new viscosity coefficient.

Kind regards,
Mostafa

P.S. In my case study, it is not possible to move the open boundary to deeper areas. Even smoothing the bathy data along the open boundary didn't help. I also have tried high bed roughness values. But it affects also the water level predictions near the open boundary.

Hello Mostafa,

You can implement it directly in TELEMAC3D and TELEMAC3D_INIT subroutines, just before or just after the call to VISCLIP and change VISCVI variable.

Hope this helps,

Chi-Tuan

Hej Mostafa,

Is your question still valid? I managed to implement a similar feature for a similar issue a while ago and could look up the code if necessery. In general the approach follows these steps:

1. Add a 'VISC' variable in your geometry file. I implemented this as a boolean (1 / 0). And read this parameter to one of the private arrays in the .cas file.
2. Update the viscosity algorithm to overwrite the viscosity to some manual value for all grid cells that have VISC==1.

Results were as expected, but they did not fully resolve the issue that I had. It should also be noted that bathymetric changes within this viscosity box should be prevented.

@Pham: Thank you. I will try to implement it in the mentioned subroutines.

@jurjendejong: Yes, my question is still valid! The method you have mentioned is very elegant; in the way that we can quickly test different values of viscosity at the open boundary only by changing its value in the Fortran file; without the need to modify the geometry file each time. If it is possible, could you please share that piece of Fortran code?

Please find attached my modification of the vissma.f routine in TELEMAC3D V8P4. Note that you might need to copy this code to another viscosity routine if you are not using Smagorinksy.


Also include in the .cas file:
Please share your experiences when testing this approach!

Thank you very much! It will be a great help.

Hello Mostafa,

Have succeeded in solving this implementation?

There is a new feature available in next release v8p5 so-called "soft boundaries" which may solve your issue of superficial flow (eddies) in the vicinity of the ill posed boundary.
You can try it by using the main release or the branch v8p5 (which is the basis of what will be available in new release v8p5r0).

Please find attached a description of the new feature in the TELEMAC-3D user manual (also available for TELEMAC-2D) and the TELEMAC-2D validation manual.

Hope this helps,

Chi-Tuan

Hey Pham,

Many thanks for providing additional information on how to resolve the jet flow issue. I read the PDFs and also reviewed the relevant piece of code in bord3d.f, based on the current v8p5 branch version on GitLab. Since we intend to maintain the Telemac version of our project, I might consider implementing it in v8p4.

Kind regards,
Mostafa