open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi,

I have a model of a large estuary and for the seaside boundary condition I have tidal measurements from the two sides of the mouth. The total difference is more or less 15 cm between them.

For the left bank I want to use the measurement done on the left bank and on the right bank the same. Further I used an average of both to use it as a water level boundary condition in the middle of the mouth (the seaside boundary condition lies at the mouth of the estuary). I imported the data and got three curves, which I assigned to the apropriate places. I had thus three different boundary conditions on one big boundary (=the estuary mouth)

The problem is that I noticed that telemac uses only one of them over the entire boundary. Is it possible to use three different water levels on the same big boundary? How should I implement this?

regards,

Sven

Hi Sven
The only way for you to use 3 different curves is to create 3 different liquid boundaries on your model (That implicate you have solid boundaries between the liquid boundaries).
Otherwise, you could manage the water level imposition on the different boundaries by fortran programming in the subroutine bord.f
With this solution, you can manage the spatial interpolation between your measurements and the location or your boundary nodes

Regards

Dear Sven,

Actually the subroutine BORD.F is fairly complicated because it combines a number of possibilities : boundary conditions given in boundary conditions file, given by key-words, variations in time given in liquid boundaries file, etc.

In fact the role of BORD.F is only to give the prescribed values in case of Dirichlet boundary conditions. These values are contained in UBOR, VBOR and HBOR for the 2 components of velocity and the depth. People who deal with complicated conditions, like time and space variation of free surface read in a file, do not bother with this and re-program the subroutine BORD.F.

However in your case you can do it by programming function SL. This function providing the free surface elevation at boundaries is called in BORD.F if you use the keyword PRESCRIBED ELEVATIONS = ..... (see lines 127 to 144 in BORD.F, and cases when SL is called).
By default SL takes the values given by the PRESCRIBED ELEVATIONS, i.e. one value per liquid boundary (array ), this is probably why you see that Telemac-2D takes only one value. Then you must copy SL into your user FORTRAN FILE and modify it. In the arguments you have the global number of the point, and you can use AT as the time (it is seen in SL because it is in DECLARATIONS_TELEMAC2D). Then you can prescribe any elevation as a function of the global number of the point and the time, by just giving the elevation SL at the end of the function, regardless of what is done before (which basically treats the case of values given in a file). It would be a little bit more difficult in parallel, but a first step would be to get what you want in scalar mode. So please first try this and we'll see,

With best regards,

Jean-Michel Hervouet

Dear Sven,

I f you want to venture a bit more in the advanced use of TELEMAC, you can find a more detailed description of how to use SL.f with a tenporal- and spatial- variation of your level boundary, and that would work in parallel, please refer to the post #749, answering Steve Renaud on the subject of "Spatial and temporal varying boundary conditions".

There are details of Fortran code and of LIQUID BOUNDARY FILE setup.

Sébastien.