open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello all,

I am currently working on the coupling between telemac and mascaret and I want to impose H, U and V as boundary conditions for telemac through the subroutine bord.F. If I have postive water level on the liquid boundary that I consider then every thing is fine and I get the results I want to. But when I do a test case where the free surface on the liquid boundary varies from 0.1m to 0.0m (the bottom of the geometry is at 0.0m) then when all points on the liquid boundary are close to 0m then the program enters the subroutine charac.f but never gets out. I put some flags all around this routine and I shown that when all the points on the liquid boundary are close to 0 then the program enters the subroutine scaract of the module streamline but never gets out. I don't understand what is the problem and I am a bit disapointed. Can you please help me ?

Hello,

When the Froude number becomes greater than 1 as the depth decreases, something should change in your boundary conditions or the problem is not well posed. Scaract is certainly not the problem, and you can choose another advection scheme to be sure, it will probably also crash.

With best regards,

Jean-Michel Hervouet

Hello JMH and thank you for your answer,

My previous question was a bit tactless. Anyway, the current implementation for the boundary conditions is as follows : if H(K) for the K-th point of the liquid boundary if greater than a given threshold T then the boundary condition type is turned to "5 6 6" and I precribe H, U and V on this point. On the contrary if H(K)<T then the boundary condition type is turned to "5 4 4" and I prescribe H. Now if T<=1.d-2 then I have the problem I exposed in my previous message. If T>=1.5d-2 I have not this problem anymore.

Do you have any other suggestions of what I should prescribe as boundary conditions when the Froude number becomes greater than 1 ?

Other question : is there a variable in telemac that gives the Froude number in each point of the domain ?

Regards

Hello,

Late answer... yes, to have the Froude number just add F in the list of variables for graphic printouts.

Regards,

JMH

Hello and thank you for your answer.

I must precise that I solved the freezing problem that I mentionned in my first message by reducing the time step from 1s. to 0.1s.

Regards.