open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello.

I would like to know to definite the mesh transformation, including how to choose the value for the keyword and how to specify the values of each level if mesh transformation is not 1, especially for bathemetry with great gradient, as that in my case.

Please see the attached geo file.

The 2d case with the same geo file and bc file ran very well, while the 3d case didn't.
In my case ,
mesh transformation=2
and
ZSTAR%R(1)=0.0D0
ZSTAR%R(2)=0.2D0
ZSTAR%R(3)=0.4D0
ZSTAR%R(4)=0.6D0
ZSTAR%R(5)=0.8D0
ZSTAR%R(NPLAN)=1.0D0

I'm not sure it's right or not.

Could anyone be nice to help me with this question?

Best wishes

Hong

Hello Hong,

If you want to use the option 2 for the keyword MESH TRANFORMATION, you have to give this value in your steering file + uncomment the following lines in the CONDIM subroutine:
DO IPLAN = 1,NPLAN
TRANSF_PLANE%I(IPLAN)=2
ENDDO
and give the distribution as you have written.

Do not forget to read the User Manual (for example subsection 4.1). It really helps!

Hope this helps,

Chi-Tuan

Thank you,pham.

Actually, I read the manual and did as it said.

However,the results from telemac3d is not as good as those from telemac2d, especially the velocities in some places of the channel.

Could you please see the attached result files and give me some advice to improve the simulation?

T2DRES is result from 2d simulation ; while the other 2 results files are generated by 3d simulation.

Best wishes

Hong

Hello, pham.

The results files are too big to upload.


Would you please give me an email box?


Sorry for the inconvenience.


Hong

Could anybody help me with this?

The case is very important and urgent for me.

Hi,

It is difficult to analyze the problem given limited information available on the posting. Can you elaborate a bit more?

Edwin

I would like to upload the results file to make you clear about what I said.

But they are too big to upload

Hello,everyone.

PLease see the attatched some pictures about the results and help me analysize why the t3d results seem wired.

Thanks in advance.

In those files,geo.jpg is geo file;
t3dpv.jpg is the time series of velocity at some point;
UVW.jpg is 3d velocity;
UV2D.jpg is velocity from telemac2d;
and UV3D.jpg is 2d velocity from telemac3d;
In UV3d, transverse veloties at some points can be seen.I can't figure out why that happens.

Hello,

Also check the free surface (have you spurious oscillations locally) and the depth associated to these velocities (does this occurs on tidal flats ?). Are the numerical parameters (time step, implicitation coefficients, free surface gradient compatibility) the same than in 2D ?

Regards,

JMH

Thank you,jmhervouet.

The numerical parameters (time step, implicitation coefficients, free surface gradient compatibility) are the same in 2D and 3D.

As you see, the banks of the channel are very high(z=10m), while the depth is small(1.5m or so).

Therefore,which kind of definition of vertical levels is most suitable?