open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

for a complete description of the Selafin format, you should have a look at the TELEMAC-2D user manuel, Appendix 3. A Selafin file is a binary file, it is not the same format as the boundary file.

I will have a look at your wind file : can I test your case with the parameter and geometry file you sent me last week?

Regards,

Elodie

Elodie,

Thanks for yours indications for the SELAFIN format.

I send you my different files

Regards

Florian

Hello,

I tried to run your case, and these are my observations :
- the program stops in semimp, in part 6, where the high frequency tail is treated : the problem is due to wrong values of USNEW (=NAN) that give wrong values of FM1...
- Indeed, USNEW has very strange values, because the reading of your wind file does not work well.
- I am now looking at the values that are read in lecdoi from your wind file and there is indeed a problem.

I will try to see where is the problem in your wind file and in its reading, but in the same time you can also try to do some prints in lecdoi to see from where it comes.

One other thing : I don't understand why the keyword VENT STATIONNAIRE in you parameter file is equal to TRUE. Your wind evolves in time!

Regards,

Elodie

I complete a bit my answer : in fact, the reading of your wind file in lecdoi.f works well. Problems happen when the interpolation with the routine fasp.f is done... It returns -1.D6 for all wind component.

Elodie

Hello again,

I found something : i just realised that your wind data are given in (Longitude, Latitude) coordinates, but your domain uses other coordinates! (maybe Lambert?) I don't see how fasp.f can do the interpolation in these conditions...

Lastly : your total simulation time is 10000 s (less than 3 hours) and your wind file has a 3 hours time step and its duration is many days : is that normal??

Regards,

Elodie

Elodie,

Thanks for all.

Whatever the total simulation time, it stops.

My wind data works with longitude, latitude. My domains is defined in X/Y UTM. Perhaps, i can change the first line of my wind file to define NLIG, NCOL, XMIN and XMAX, YMIN and YMAX in X/Y UTM.
In the tomawac manual, in the appendix 8, there is a description of the WAM-Cycle4 format where we can see that they use longitude or latitude.
Perhaps,i can make a new mesh with a domain which is defined in longitude latitude?

In the routine fasp.f, the values -1.D6 is the water depth. What is the link with the wind velocities?

I made a simulation where i don't put my wind file and i define a stationnary wind. i forgot to crash this keyword.

Hello,

as you said, WAM-Cycle4 format uses Longitude, Latitude coordinates : I am not sure you can change that but I let you check (maybe it's not really an obligation, but you have to see what it implies). Otherwise, you will have to define your mesh in Longitude, Latitude. Be cautious : when you work in spherical co-ordinates, the value of the logical keyword SPHERICAL COORDINATES should then be set to "TRUE". The co-ordinates are then expressed in degrees.
Another solution would be to use the subroutine venuti.f to program by yourself your wind file reading and interpolation.

The routine fasp.f interpolates bathymetric data or whatever else we want to interpolate with the same interpolation method.

Regards,

Elodie

Hello,

Thanks for your help.
I try to write the interpolation in the routine venuti.f.
If it's so hard, i will make a new mesh.

Regards,
Florian

Hi,

With a debugger, i can see where my run fails in semimp.f.
In this subroutine, the first error comes where we threat the high frequencies differently (part 6 of the routine, line 714). Apparently, IP, JP and IFF are not defined. When i see at the beginning of the routine, where IP, JP and IFF are defined, i can see that the error is caused by the number 1 in "IP = 1,NPOIN2", or JP= 1,NPLAN or IFF=1,NF
There is a problem with the number 1.

Have you got an idea why? Is it an error in the steering file?

Thanks florian

I already spent some time to look at your files and answer your questions :
- The program stops in semimp, in part 6, where the high frequency tail is treated : the problem is due to wrong values of USNEW (=NAN) that give wrong values of FM1, etc.
- Indeed, USNEW has bad values (including NAN), because the interpolation of your wind file does not work well.
- Your wind data are given in (Longitude, Latitude) coordinates, but your domain uses other coordinates! I don't see how fasp.f can do the interpolation in these conditions...

Did you try to modify your wind or geometry file??

Elodie