Hello,
thank you both for the feedback.
We will present this case at the user club this year!
At the moment I have not much time for further explanations, but we simulated this case:
Coleman, S., Andrews, D., and Webby, M. (2002). ”Overtopping Breaching of Noncohesive Homogeneous Embankments.” J. Hydraul. Eng., 128(9), 829–838, 2002.
and we match the experimental results in time very well with some adapted numerical parameters.
Yes, we created a steep water race with a basin downstream. This worked very well since we could get rid of possible influences of the outflow boundary on the base of the dam.
First tries with the outflow boundary at the downstream base of the dam (as in the experiment) worked well for only hydrodynamics since it is supercritical flow but we got a lot of sediment accumulation which influenced the results. The sediments did not find the way out. Probable explanation: slope correction is turned off at boundaries?
Some conclusions:
The sediment slide algorithm implemented in Sisyphe works very well.
The transverse deviation factor beta in the Talmon formula plays a large role. Values from 0.85 - 3.5 reported in the literature (Talmon, Van Rijn, Schoonen (TU Delft), Zanke) are generally too high and lead to a too slow erosion process. However these values are from experimental results with much slower sediment transport processes and there is not much in the literature about experiments with sediments & fast flow & transverse deviation. So we calibrated the value down to 0.40.
Best regards,
Clemens
