open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi,

Thank you very much to you two for your comments.
Looking forward to the next release

PL

Dear PL,

I would not use MPM for the modelling of very fine sediment (you wrote >0.1mm), because the formula was developed for coarser sediments.
What is the reason for choosing MPM?

Best regards,
Gabi

sorry,

I meant for sediment < 0.1mm

Gabi

Hi Gabi,

No, I do not plan to use MPM for such sands. I am currently assessing what would be the best method, but I can already say that bed-load won't be computed with MPM. For the moment I use EB / Van Rijn which seem to give good results and that have a validity range of 0,2 - 2 to 3 mm.

I was thinking to combine bed-load with suspended-load. I had good results so far with EB (bed-load). I don't have that much time to make further tests by now, but I was thinking of trying to computed suspended-load with a conservative value for the settling velocity Ws (ie. inferior to the one corresponding to my D50). Does someone ever tried it? With which results?
The idea behind this was that according to Van Rijn the median diameter of the suspended particles is somewhat and logically smaller than the material's D50 (I think I read 0,5-1 x D50 but I need to check again).

I hope to have time to try this soon.

Thanks for your remark.
Best regards
PL

Hi again,

I was also using van Rijn for the simulation of a flushing process for an alpine reservoir with a slope of about 0.15%. The grain size was from 0.5mm to 70mm (finer at the weir).
Van Rijn worked best, MPM caused too less erosion and Engelund-Hansen (total load) to much compared to the measurements.
But I think it is difficult to choose a sediment transport formula without calibration or validation data.
Have you tried to combine bed-load and suspended load yet? How are the results?
Are you using multiple grain sizes?

Best regards,
Gabi

Hi again,

Thanks for your feedback, it's interesting.
I am actually testing various methods on a fictive geometry, hence without calibration data...

I just launched a couple of other simulations, let's see what will the results be

But I can already say that I observed the same trends than you : MPM gives too less erosion and EH too much (certainly due to the fact that my case is really unsteady). EB and Van Rijn give similar results.

I made some first tests with suspension with EB bed-load formula, and the results are looking fine (more erosion, development of channels in the reservoir). I must confess that these are my first simulations with sisyphe and that I still need to fully understand how to interprete the results for each variable. I was for example a bit surprised to see that I had very low suspended concentrations (0,05 g/L). I must have a closer look at that.

I am now running Van Rijn (bed-load) with suspension, and I also launched a case with a smaller settling velocity than the one computed automatically (x 0,5).

I do not have multiple grain sizes in my computations (I have a rather uniform sand (D30 = 0,07 - D50 = 0,13 - D90 = 0,25 mm). I thought of trying that later. Did it help you?

Best regards,
PL

Hi,
To be honest, no, multiple grain sizes did not help much. Stay with uniform grain size as long as you can, you will avoid a lot of trouble.
If you have some new results, I would be interested
Best regards,
Gabi