open TELEMAC-MASCARET The mathematically superior suite of solvers

Dear ALl

Do you know why for the same cas study simulation in sisyphe after a simulation of telemac2d

in V5P9 with time step = 60 et number of iteration = 180 the simulation duration is 10800s = 3h and

in V6P0 with the same time step and the same number of iteration the simulation duration is 9000s = 2h30

????

thanks
Philippe

Hello again,

However, if you do chaining, i.e. computation with Sisyphe after computation with Telemac-2D, I would strongly recommend that you do a real coupling. This is very important, the updating of velocity and depth is better done by a real Saint-Venant solver. Moreover if you have suspended sediment, the mass conservation of sediment depends on the continuity equation done for Saint-Venant, which means that if you do chaining, you will never have conservation of suspended sediment.

Have a good week-end,

Jean-Michel

OK, I misunderstood the meaning of the message. We have a look and report...

JMH

Dear All

I found where is the mistake : I dont know if it is me or the code ....

When I use Sisyphe in external coupling with an hydrodynamic file from telemac , Sisyphe V6P0 does not read my time step in the CAS file of SISYPHE but use as time step the GRAPHIC PRINTOUT PERIOD of the telemac run ???

Do you know why ??

Best regards
Philippe

Hello Philippe,

This is probably not so. Actually when Sisyphe is used in coupling, the time-step is monitored by Telemac-2D, and you have to use the coupling period if you want to have a larger time-step in Sisyphe. The time-step given in Sisyphe parameter file is just ignored. Same thing for the number of processors.

For the problem of 2h30 or 3h00, we are looking at it. A first opinion given by another user (PL) is that it is only a problem in the listing. However there seems to be an error when sub-iterations are asked.

With best regards,

Jean-Michel Hervouet

Hello Jean-Michel

I AM NOT USING SISYPHE IN COUPLING with telemac2d but the time step is even ignored IT WAS NOT THE CASE IN THE V5P9 and I am not asking for sub iteration

Best regards
Philippe

Hello Philippe,

Yes, I have checked that what you say is right, if not what you expect. If you prefer the previous implementation, you can add sisyphe.f in your Fortran file and in the first call to bief_suite, you can suppress the sequence ,DT=DT at the end of the sequence of arguments, the first DT is the name of an optional argument, and the second one is the time step in Sisyphe. The previous implementation has been commented out about ten lines above.
You can phone to CV tomorrow to discuss further this question.

With best regards,

Jean-Michel Hervouet