open TELEMAC-MASCARET The mathematically superior suite of solvers

Dear all,

I am trying to model wave propagation inside a gulf and use results from external sources for the current velocities. Thus I have modified couuti.f using meteo.f from telemac2d as a blueprint. I had run the opposing current test case and verified that the subroutine works fine and populates velocity values in the results.

However, when I am trying to model a field case I notice that the velocity values are always 0. I suspect that it has to do with the available space allocated for the Spatial coordinates X,Y (for example I use UTM coordinates, that have 10 digits while the examples in telemac (windtxy) have only 1 digit.

Could you be so kind to help locate if this is indeed the problem? Here is the test case we.tl/t-9CTb0cFH8B

Kind regards,
Andreas

Hello

Apparently in your file Thess_curr_UTM.txt you have only one time step
and in couuti you refer to more than one time step :
TC2=AT_CUR(K)
with K=2 or more.

Hello Thiery,

I have tried to add more time points, but to no avail the results are still zero. I also tried to apply this subroutine to the Opposing current test case with only one time step and the current values of the Thess_curr_UTM.txt file, i just modified the X,Y of the file (etc. 2,3 ; 3,3 as in the geometry file of the test case) and the model run just fine. That is why i suspect it has to do with the Coordinates of the external current file.

Regards,
Andreas

Because you put NP=NUMPOINT instead of NP=NUMSTA, then when you try to interpolate in FASP you only consider the first point so you never find any correct point for the nodes of your mesh.
After having corrected that, are you sure you want to use currents of 1e-3 or 1e-4. That is what i see in your file Thess_curr_UTM.txt.

Regards.

Thank you Thierry I completely missed that.
However, even after this change the results are zero. I even put for a test two points inside the mesh and still the values are not interpolated. It puzzles me completely.Attached is the test file I used.

Kind regards,
Andreas

actually i set the times at 100 sec and 200 sec in the above file,to coincide with the time step interval

I actually located the error.

When using CONSIDERATION OF TIDE = YES, the transfer rates computed at the propagation step are time depended when considering variable water levels.

Due to the domain being quite large, the solution does not converge, since it requires too many number of sub-iterations even at the first time step. So I disabled the keywords TIDAL WATER LEVEL FILE and TIDAL WATER LEVEL FILE FORMAT. What happens is that even though the current is read from my file, the velocities are set to zero since after the call to Couuti.f the subroutine anamar.f is called which by defaults sets the velocities to zero.

Does anyone know a workaround for this without resorting to disable the calls to anamar.f?
Maybe if i use binary files, to avoid using couuti.f, maruti.f, the simulation will converge faster?

Regards,
Andreas