open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello dear,

I am trying to simulate planar erosion of dike. The sediment used in the experiment is fine sand 0.24mm D50. As per attached image shows that in the experimental dike, there is no deposition downstream of the dike and all sediment washout, however, in sisyphe created deposition downstream of the dam which is not realistic. My question is where to look for depostion in sisyphe code? I noticed also the boundary condition has big impact on depositon as I left as free boundary 444 in the boundary file CLI.

Thanks,

AMANJ

Dear Amanj,
Deposition, as well as erosion, are strongly related with the (channel) transport capacity, and therefore linked with the flow and sediment conditions.
Probably, you are using either a transport formula that needs to be calibrated (for example, through the modification of the coefficient for the Meyer-Peter and Muller formula) or you need to play around with another sediment transport formula.
By the way, if your flow condition is subcritical, why are you using 4 4 4 in the cli file for the hydrodynamics BCs?

Best wishes,

Pablo

Hello Dear Pablo,

I have supercritical flow downstream of the dike. I have notice that those transport formula programmed in sisyphe are behaving well with medium to coarse sand, but with fine sand such as 0.25mm D50 not. The problem is downstream outlet is too short only 0.25m and would like to washout (pass) all sediment flux rather than accumulate and deposit right behind the open boundary. I have also noticed boundary condition and prescription has great impact on deposition. What would be the best solution? We played with almost all parameters but still can't avoid deposition. Our goal here is zero deposition downstream and only erosion should take place during flow condition. how to impose on sisyphe to not consider deposition or just simply ignore it or all erosion put in solid transport flux?

Hi Amanj,
OK.
Why don't you extend the limit of your model to get rid of the outflow boundary condition, that probably is perturbing the results? It's possible for your case to do that?

Cheers,

Pablo