open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi,
I have been trying to conduct a landslide generated wave simulation using Telemac2D.
I've tried various FV schemes as well as Boussinesq but the characteristic of the wave seems to be different than that from the lab experiments (see figure below). I was able to get similar wave characteristic using Flow3D as well as using Telemac3D in non-hydrostatic mode. Both models require significant computation power and time. I am hoping to be able to conduct the simulation in Telemac2D instead. Any help will be greatly appreciated.

Hi
Could we have more details on the way you manage the landslide in the simulation and all information about your model...
Regards

Hi,

The figure from the initial post was a comparison between Flow3D, Telemac and lab experience.

I am currently comparing Flow3D and Telemac for a different case. The landslide was simulated by a sudden release of water on a bank. The mesh element was 5 m with timestep of 0.01 sec.

The first figure (slide_probes) shows the landslide location and the locations selected to compare between the models. The second figure (slide) shows the surface elevation 10 seconds after the slide. The third figure (probes) shows the comparison of surface elevations between the two models at 3 probe locations. The black line is FLOW3D and the red line is from Telemac. The 3 panels on the left are from Telemac2D and the 3 panels on the right are from Telemac3D (non-hydrostatic).

As the figure shows, the wave from Flow3D and Telemac3d (non-hydrostatic) has a more of a symmetrical form whereas the wave from Telemac2D is more assymetrical.
I have also tried running the Telemac3D first to generate the initial wave and then continued the model using Telemac2D. The resulting waves are still assymetrical.






Edwin

Hello

You should expect that Telemac-2D will not give similar results that those of Telamc3d. For such kind of applciation you should use a model with 2 major properties;
1 - dispersive model : and in this case Telemac3d is fully dispersive. Telemac 2D can retrieve partilly the dispersive character if you use Boussinesq model.
2 - high order schemes: and in this case, you can use Telemac 2D cith kinetic schemes second order in space. Telemac2D and Teemac-3D with finite elements are theoretically also second order (P1 Finite elements) but several parameters degradates the order and we retrive in most times 1 order.

I hope that this helps.
with my kind regards

Riadh

Hi Riah,

Thanks for the reply and explanation. Based on the lab experiment and FLOW3D results, it looks like Telemac3D (non-hydrostatic) will better simulate the landslide generated wave.

Edwin

For those who are interested, I found that there is a discussion on the use of Saint-Venant equation, Boussinesq equations, Navier-Stokes hydrostatic and non-hydrostatic equations for solitary wave in the book "Hydrodynamic of Free Surface Flows" by Jean-Michel Hervouet.