open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello all,

I am currently working on the coupling between TELEMAC 2D and MASCARET with the OpenPALM code coupler. I have decomposed TELEMAC and MASCARET in 3 subroutines each one dedicated to : initialisation, run, finalisation. After the initialisation step the algorithm is based on a loop where TELEMAC and MASCARET are launched several times with the same initial condition but with different boundary conditions until the ending condition of the algorithm is reached.

The TELEMAC run step corresponds to the time loop of the telemac2d.f subroutine. My question is : what are the TELEMAC variables that I need to save at the beginning of the TELEMAC run step so that I can restart from this initial state later ? I guess those variables are U, V, H, T, AK and EP, do you confirm ?

I hope I am clear enough.

Hello,

This should be OK on Telemac side. Note that the keyword INITIAL GUESS FOR H (ORDRER DU TIR INITIAL POUR H) may give differences if set to 1 (default) because in this case the previous increment of depth is kept and used as an initial guess for the next increment. During a restart this increment is not saved and is initially set to 0 so it makes a difference. A solution to see no difference in the behaviour is to set : INITIAL GUESS FOR H = 0. Only in Telemac-3D we have programmed a restart mode that ensures the same initial guesses for the linear solvers.

With best regards,

Jean-Michel Hervouet

Thank you JMH for this quick answer.

But I have still a basic question. I have defined some BIEF_OBJ objects (UO, VO, HO...) to save the initial values of U, V, H... and in my TELEMAC run step routine I wrote "UO = U" to save U in UO for example. But it gives me strange results. Is there a specific command to do this with BIF_OBJ ? I trid the os.f subroutine but this is only for vectors.

Regards

I finally understood what was the problem, I guess I was a bit tired yesterday... I wrote a subroutine that makes two BIEF_OBJ identical and now it works ! Can you tell me if a similar subroutine already exists in TELEMAC ? If so I would prefer to use it.

Regards

Hello,

This subroutine is precisely OS with a call :

CALL OS('X=Y ',X=..., Y=...)

You can refer to the guide for programming in the Telemac system which is available on this site.

With best regards,

Jean-Michel Hervouet