open TELEMAC-MASCARET The mathematically superior suite of solvers

Dear all,

I have a question regarding the way the boundary fluxes are computed and/or printed in the listing file when using FV (scheme 5 in my case).

I notice for example that the values printed in the listing file does not correspond exactely to the prescribed values, with some % of difference. For example, a prescribed inflow of 8,9 m3/s gives 8,81 m3/s as shown in the listing file. Same set up with FE give exactely the same values, as expected.

Is the actual inflow actually 8,81 m3/s inside the model or is the difference due to the way the fluxes at boundaries are computed for listing outputs?

Thank you in advance,

Best regards
PL

PS: I thought I read a similar post in the past but could not find the thread now. Sorry if this question has already been answered.

Hi PL,

there was a discussion some years ago:


www.opentelemac.org/index.php/kunena/16-...tions?start=10#10278

maybe it helps.

Best regards

Leo

Thanks Leo, it is actually this thread I read in the past.
I will look at it in full (it takes a bit more than 5 minutes!) and will hopefully find the answers I am looking for.

I just wonder if things have changed since then? Did you notice anything?

Thanks in advance,
PL

Hi PL,

I gave up and wrote my own FVM codes with cell centered methods.

I have not tested the boundary conditions and the schemes in Telemac since 2014. But I know that they improved the schemes e.g. there is a new hydrostatic reconstruction available and some improvements in the HLLC solver.

I hope that Riadh knows how you can solve the boundary condition problem.


Best regards,

Leo