open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi every one,

I was trying to simulate dam breaching by coupling Tel 3D and Sisyphe. as it is shown in the attached pics, result doesnt look reasonable and the fluccuation near the crest of the dam is very high;however, i tried with Tel 2D and I didnt have such a problem and simulated very well. I would like to use telemac 3D instead of 2D for the same model, but I dont know what is wrong in my parameters. I attach my steering files into this post and hope to get some idea.

Thanks,
AMANJ

Hello,

The time step is very small, is it a scale model ?

Do you have suspended sediment? I find strange the solver for sediment = 1 (it will not work in some options with settling velocity).

There is nothing on diffusion in your steering file, so you have probably a very low constant diffusion, this is to investigate to remove the fluctuations.

With a scale model, beware the keyword:

THRESHOLD FOR VISCOSITY CORRECTION ON TIDAL FLATS

which is set by default at 0.2, a high value for a scale model.

With best regards,

Jean-Michel Hervouet

Hello dear Jean-Michel Hervouet,

This is laboratory scale model. In Telemac 2D I tried with the same time step and everything went right. In Telemac 3D even i tried with larger time step such as 0.1 or even 1 but didnt work. Here I just considered bedload.
The strange thing I have noticed here is that In Telemac 2D the flow start overtopping through the pilot channel at the middle of the crest; however, in Telemac 3D overtopping is happen over all crest and this is not realistic.

I will test those parameters you option and I will see the result.

Thank you so much!
Have a nice day,
AMANJ

dear Jean-Michel Hervouet,

I tried many parameters, but still I have unrealistic and unpreditable result from telemac 3D. I tried with telemac 2D coupling with Sisyphe, it works very well. However; I didnt change the sisyphe parameters and I coupled with telemac 3D, the result is too bad. I uploaded some results here with steering file.

I dont know what make this high instabilities at the u/s crest of the dam.I was predicting the flow start overtopping through pilot channel at the middle of the crest just like in the 2D case, but here not, its overtopping over all the crest. The last to pics from 2D result. Note that this is laboratory small scale, height is 30cm, width 1.5m.



Thanks
AMANJ

the last two pics from 2D result

Hello,

Actually the :

THRESHOLD FOR VISCOSITY CORRECTION ON TIDAL FLATS

should be set to a very low number, like 0.01 in your case.

You can try also a constant viscosity model(model 1) on the horizontal with an horizontal coefficient set at a value from 0.1 to 1, and mixing length on the vertical (model 2).

and to avoid oscillations, try

FREE SURFACE GRADIENT COMPATIBILITY : 0.

With best regards,

Jean-Michel Hervouet

Hello dear Jean-Michel Hervouet,

Thank you so much for always fast reply and helping us for solving our issues in modelling. Actually, before I see your new reply, I tried to run my hydrodynamic 3D telemac model without coupling the sisyphe. The model is working well. So, I believe the problem is in coupling with sisyphe. I did before with telemac 2d and everything was fine. But now with Telemac 3d it gives me problem as you saw in the previous post. What parameters can affect and create this instabilities? Now I am running the hydrodynamic model without sisyphe with the new advices you gave me in your last respond and waiting for result. I have another question relating to installing Telemac, is it possible having single and parallel telemac under windows at the same time or we can have just compiles parallel or single processing at the same time? this is for installing Telemac!why i am asking because so far I am running under single processing, however; I wish to run under parallel condition as well.

Have a wonderful time!

AMANJ

Hello,

No problem for mixing parallel and serial, I always do that. The executable files for parallelism also work with one processor.

After looking again at your parameter file, I found something that could well explain your problem:

TURBULENCE MODEL FOR THE BOTTOM =1

is smooth friction regime and should be changed to rough regime (value = 2).

If you found this value in an example we will change it also. It is very rare to have a smooth regime, except for testing analytical solutions.

With best regards,

Jean-Michel Hervouet

Dear Jean-Michel Hervouet,

I need your help. I tried what you suggested and is working perfectly when only Telemac 3D. Once I coupled with Sisyphe using internal coupling, doesnt work and I couldnt find the reason. I attached some images for coupling and one video without coupling.

Looking forward to hear from you,
Sincerely,
AMANJ

steering files
Link for video file
www.dropbox.com/s/pegyeppdnyto248/video.swf?dl=0