open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi,

since the last year we have some problems by simulation flushing events. In some simulations or in some areas of the mesh, we get "water surface striae" (see post www.opentelemac.org/index.php/kunena/16-...-surface-striae#8282), which result due to the calculated shear stress in furrows in the evolution (www.opentelemac.org/index.php/kunena/17-...n-the-evolution#5140).

I have this problem also with Telemac-3D internally coupled with Sisyphe. A water level difference of 1-2 cm gives differences of about 0.2-0.5 N/m2 in the TOB and causes differences in the evolution of 60-120cm (due to the small grain sizes ~1mm and the duration of 3 days).

I have already tried the bubble option, which helps a little bit and to change the free surface gradient (0.9 - 0.5 - 0.1), but this not much effect. It dumped the difference in evolution only about 5-10%.

Does anyone has some advice or hints?

Best regards,
Gabriele

Hello,

Yes this was reported two times already, but the keyword FREE SURFACE GRADIENT COMPATIBILITY did work. This seems then linked to the inf-sup condition and except this and the quasi-bubble element there is little possibility. An SUPG OPTION set to 2 for depth (i.e. SUPG OPTION : 0;2 for example) may also work but it may cause a loss of conservativity of tracers.

We would like to have a simple case that clearly shows this behaviour (all cases that I got so far hardly could run on my machine and took a lot of time) so that we can work more easily on it.

You can also try to use a more diffusive treatment of Exner equation: take subroutine positive_depths (library bief) in your FORTRAN FILE. In this subroutine change the line:

IOPT1=2

into:

IF(NAMECODE(1:7).EQ.'SISYPHE') THEN
IOPT1=0
ELSE
IOPT1=2
ENDIF


With best regards,

Jean-Michel Hervouet

Dear Jean-Michel,

thank you very much for the answer.
I am going to try the change of the positive_depth subroutine.
You are right, our case takes 3 days with isteady flow, so I will trys to set up a simple test case with the same configurations.

Many thanks,
Gabriele