open TELEMAC-MASCARET The mathematically superior suite of solvers

Does anybody know about the feasibility of modelling submerged vertical walls in T3D? I am currently investigating the potential use of such a structure in diverting strong propeller induced currents away from environmentally sensitive areas and would love to use T3D for this study.

I noticed that similar models such as Delft3D Flow have these capabilities and was wondering if there are any fundamental reasons (which potentially apply to other packages as well) as to why this is not yet a feature of Telemac?

Cheers,
Toby

Hi,

Have a look to the following topic:
www.opentelemac.org/index.php/kunena/21-...ubmerged-walls#32197

Regards,
Laurent

Hi,

As Laurent mentioned you can have a look at the papers linked in the other similar thread.

If you don't need CFD-like precision on the results I would first try with increased head losses at the wall location to stop and divert the current. Otherwise you can try to model the wall with a gradient as large as reasonably possible (maybe 10:1?) but then you would need to heavily refine your mesh at this location. You might have issues to reach convergence though... Non-hydrostatic simulation is a must.

Good luck and please leave a feedback!

Kind regards
PL

Dear PL and Laurent,

I'm facing a similar situation trying to simulate a spillway that, at some point downstream, faces a "vertical wall". No matter how much I refine my mesh I'm not able to reproduce the expected behavior ,which is an Hydraulic jump, the flow justs "climbs" the wall and mantains it's energy/similar froude...

I suspected from the beginning that T2D would fall short, so I i'm struggling with T3D now...

As a matter of TM v8.0, are the features mentioned by laurent (2016) implemented in T3D?

Any recommendations besides non-hydrostatic, good mesh refinement to model this case?

Regards,

José D.

Hi José,

For the first approach (adding local head losses or local increase in atmospheric pressure), you can add it by yourself in the suited subroutines, by modifying local source terms as explained in PL's paper.

For the second approach (local modification of the number of vertical layers), this feature is not available in the latest version as it requires larger code modifications and controls. I don't know if it's planned, and when...

No more recommendations than PL's ones in the previous post.

Regards,
Laurent

Hello,

Laurent, thanks for the insight. So it seems that the first approach is basically the only one available...

I hope that the modifications necessary to incude such structures will be implemented soon™.

Regards,

José D.

Hello,

Yes, I think it's the only reasonable solution if you can't move to CFD.

Regarding the development of vertical structures in the mesh, I wouldn't expect something anytime soon. This has been tested in 2016 and showed that further work was needed. I haven't seen anything hapenning from EDF since then so it's maybe a dead topic?

Good luck!

Kind regards
PL