open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello Sumit,

Can you check that you have the correct amount of water coming through the open boundaries? It might be best to look at the vectors of the depth average current (in the 2D output file) and make sure they look the same in the 2D and 3D model.

(Note that your FRICTION COEFFICIENT = 40.0 in 2D and 42.0 in 3D -- this is pretty rough)

If the model appears to be honouring the intended boundary values of water level and water flux, I suggest having vertical turbulence model = 2 (mixing length) and coefficient of vertical diffusion of velocities = 1.E-6. This will give a more realistic velocity profile.

Hope this helps.

Sébastien.

C:\UIUC\EFDC_Sim\LemontModeling\PresentationEndMarch\DepthFrm3DModel.png

C:\UIUC\EFDC_Sim\LemontModeling\PresentationEndMarch\DepthFrom2DModel.png


Hi Sebastien,

Thanks a lot for your kind reply. I do agree the value of Chezy coefficient is pretty high but that's what giving me the best results as of now. I am going to follow your suggestion in Tele-3D and see if I can get to resolve my problem.

I am also attaching two figures further explaining my problem. Finally the boundary flux seems to be allright in the 3D model.

When I have re-run Tele-3D simulation probably I will bother you again.

Best regards,
Sumit

Hey Sebastien,

Is it possible to get your email-id, I wanted to share some figures with you. if you are OK with it can you kindly mail me at This email address is being protected from spambots. You need JavaScript enabled to view it..

Best regards,
Sumit

Hello Sebastien,

Once again the links are to the figures which explains my problem

netfiles.uiuc.edu/sinha6/Telemac/DepthFr...ng?ticket=t_jlYtlpdr

netfiles.uiuc.edu/sinha6/Telemac/DepthFr...ng?ticket=t_jlYtlpdr

As of now I am running my simulations with vertical turbulence set to the mixing length model. I will let you know how it goes.

Best regards,
Sumit

mmmm ... this is puzzling. It is almost a if your upstream discharge is not getting through with the right magnitude ...

Can you try we a smaller time step (0.1sec) in 3D and a smaller accuracy on your solvers ( 10.e-7 a least) ?

Please let us know.

Hello Sebastien,

Thanks a lot for all your suggestion; I am also running one more simulation in which I have decreased the number of horizontal levels in Tele-3D from 9 to 5. My reasioning is that modeling results were closer to reality in Tele-2D hence less number of horizontal levels will keep it closer to that :). In essence I am trying everything.

Thanks once again.
Sumit

Dear Sumit

Can you send us your liquid boundaries file to have a better description of the evolution in time of your flow.
Moreover, we (at least I) need to know why the everaged water depth drops abruptly during the time interval you've mentioned. So, is it possible to have the evolution of the velocity in time.
Actually, if there is no confidentiality constraints, can you attach your cases in order to get global idea about your friction destribution, your mesh quality, your bathymetry description etc.

Kind regards
Riadh

Hello Sebastien,

This post is just to convey my thanks; I must say nothing is better than when you succeed after trying hard :). The trick that solved my problem was setting the vertical turbulence model to 2 and coefficient for vertical diffusion of velocities to 1.0E-6. Horizontal turbulence model is set to 1 and coefficient there is set to 1.0E-3. Also I am able to run with same delt for both Tele3 and Tele2 which is 0.5 sec. Next I am going to switch on the non-hydrostatic flag and see how the results change. Once again thanks to all; isn't modeling sweet when everything works out .

Best regards,
Sumit

Thank you for letting us know. Interesting finding, which will be making its way to the coming/new wiki site.

Glad to see both TELEMAC-2D and TELEMAC-3D work for you and that we have been able to help.

Sébastien.

Hi,

Thank you for this very interesting post. I nevertheless have a question to sebourban: could you please explain why you do recommend Chezy law especially (and not Chezy OR Strickler), it seems to me that they both use the depth-averaged velocity field. There might be a more "in-depth" reason that i don't manage to point out :)

Thank you in advance!

Regards,
PL