open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi!

I am running a morphological simulation in a river using bed-load and suspension (Van Rijn's formulae and with a morphological factor).
I use a coupled T2D-Sisyphe model and I previously ran the hydrodynamics with T2D alone to check the validity of the results. I model a constant river discharge, so it is ok to use a morphological factor.

I use a time-step of 1s which gives me a maximum CFL number of 0,9-1,0 in T2D. But when running Sisyphe I have a maximum CFL for suspension of about 10. My coupling period is 1.
In the log I see that I have rather high volume losses for suspension, which I believe are related to my high CFL number:

*** FINAL BALANCE FOR TRACER 1 ***
INITIAL QUANTITY OF 1 : 0.000000 M3
FINAL QUANTITY : 30.50841 M3
QUANTITY ENTERED THROUGH LIQ. BND. : 104994.9 M3
TOTAL MASS OF DEPOSIT : 63647.92 M3
TOTAL QUANTITY LOST : 41316.50 M3

Are there any numerical optimizations that would allow to lower my CFL for suspension? Should the CFL for suspension also be around 0,8-1,0 as for hydrodynamics? I did not try (yet...) to run with a time step lowered to 0,1 s since I don't have enough time for such a test (my initial run takes 35h).

Finally, I initially tried to run a first version using Engelund & Hansen (formula no. 30, with suspension turned off) but got a very unstable run already on the first time steps. That seemed strange since I start from a constant water level with a prograssively increasing discharge. Any tips for this formula?

I attach my sisyphe steering file.




Thank you in advance for your help!

Regards
PL

Hi PL,

Please note that the implementation of the morphological factor still has some limitations. All it does is multiply the bed change by that factor. The time is not updated and you should multiply time yourself as postprocesssing.

The CFL for suspension or bed load should not be influenced by the morphological factor.

Thank you for your reply.

Yes, this is how I understood the implementation of this factor, as described in your paper from the 2012 proceedings.

Regarding CFL / mass balance for suspension, what are the main guidelines to be followed? Do you have any tips? Is there a good example case available?

Thanks in advance!
PL