open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello

Is there a way to specify different number of layers depending on the depth in a 3D mesh?? For example in cases with irregular topography (lake) there could be 10 layers in the deepest portions and only 3 layers having the same deltaZ in the the shallower areas.

Thanks

Viviana

Hello,

The number of layers is constant, but when you prescribe planes with a constant elevation that goes below the bottom they will be superimposed to the bottom, with elements withoout volume. It is what we do in the Berre lake study. It then looks like the number of layers is variable.

Regards,

JMH

Thank you very much, I was using option 3 for mesh transformation and I prescribed the elevation of the planes, but when I checked a crossection in Tecplot I saw various superimposed planes in the shallower areas and thought something is wrong. But now with your clarification that this elements have no volume I think then I have a variable number of layers depending on the depth.

Thanks again and sorry for all my questions

Viviana

Hi JMH

I have another question about the mesh transformation. I ran two 3D simulations:

1) Mesh transformation =1 and 8 layers
2) Mesh transformation =3 and 11 layers at different depths

By checking at the output file I found out that the ERROR ON THE MASS FOR THE TRACER was much smaller for the first case. In the second case I was getting very high ERRORS ON THE MASS FOR THE TRACER. I have attached the output of the two cases. I have tidal flats so I am using scheme 14 for both advection of velocity and tracer.

Also, both models stop working after 2 days of simulation because I have some problems in the boundaries and after checking at the 3D mesh in this boundaries I think that the problem is that close the boundary I have shallow areas close to the deeper mesh elements where I have imposed the boundary condition. So when I looked at the 3D mesh I saw there are very step lines that connect one element to the next (figure attach). I am still not sure if that is what is causing the problem, but I can't think in something else, because I check the water depth and is pretty deep (7m) so it is not getting dry.
Do you have any suggestions for the cases of an irregular bathymetry with shallow and depth areas connected.

Sorry for all these questions but I really don't know what the problem could be

Thank in advance

Viviana

Hi Viviana,

You seem to use the version 6.1 of TELEMAC-3D, do not you?
My first advice is to use the version 6.2 and re-run your computation and check if you still have the same problems. Do not forget to change the CONDIM subroutine as told in the topic www.opentelemac.org/index.php/kunena/21-...le-test-problem#8144

If the bathymetry interpolated in your geometry file is too irregular, you can smooth it by using the keyword NUMBER OF BOTTOM SMOOTHINGS. You can also try to modify the bathymetry with a threshold (e.g. CORFON subroutine, ZF... = MAX(ZF..., one value)) just to test if you still get the same error.

Regards,

Chi-Tuan

Hi

Thanks for your suggestions, I am using Telemac version 6.1 and I will try update it in the cluster I am using. I already changed the CONDIM subroutine to use the option where I defined the distribution of vertical planes. I am using mesh transformation=3, should I use mesh transformation=4 instead?

What I will try to do now is implement my own treatment of boundary conditions, following the suggestions you gave in post: changing-the-boundary-conditions-over-the-vertical/8145. Maybe that will improve the simulation.

Thanks again for your help and I will let you know if version 6.2 works

Viviana

Hi
I was checking in the Manual about BOTTOM SMOOTHINGS, but I couldn't find much information on how they work, and how many is recommended to use? where can I get more information??

Thanks so much

Viviana

Hi Viviana,

I have the feeling that MESH TRANSFORMATION = 4 is not allowed in version 6.2 (neither is in version 6.1).

Regards,

Chi-Tuan

Yeah, I just realized when I got an error in the simulation. Thanks so much

Viviana

Hi Viviana,

You can read the CORFON (in telemac2d), T3D_CORFON (in telemac3d) and FILTER (in bief) subroutines to see how the keyword BOTTOM SMOOTHINGS is used.
To my knowledge, there are no general rules about the number of smoothings to use. So just try to see if this improves your results.

Hope this helps,

Chi-Tuan