open TELEMAC-MASCARET The mathematically superior suite of solvers

What are your flow velocities at the boundary? As Jose suggested: What are the predicted concentrations if you use equilibrium concentrations at the boundary?

Based on the trail, you impose coarse sand of 1mm at 28 g/l. with a discharge of 480 m3/s. This means 13440 kg or 5m3 of sand enters every second. This seems quite a lot and will require a large flow velocity to keep in suspension.

plot also the shear velocity (or shear stress). The ratio between this value and the settling velocity can give an estimation of how much sediment would be transported by suspension or bedload

The good old Hjulström curve can give a rough idea of the behavior of your system.

It seems that the phenomenological laws that you are imposing to close the sediment transport are not compatible with the expected behavior.

best wishes

pablo

Thank you both for your answers.
mafknaapen wrote:
My flow velocities seem to be between 0.05 and 0.5 m/s at the boundary. Which seems quite slow to me...

How could I get the predicted concentrations? The sédiments don't seem to stop at the input by using the keyword EQUILIBRIUM INFLOW CONCENTRATION, but how can I see the concentrations calculated by Sisyphe with this keyword?
Pablo wrote:
I can plot the shear stress, and I have the settling velocity. What are the critical values of the ratio between these two values?

Thanks in advance!

You have CS1 and CS2 defined as output paremeters, so your output file will contain the depth averaged concentration of fraction 1 and fraction 2.

I you say that with EQUILIBRIUM INFLOW CONCENTRATION the sediment 'travels further', I suspect that you just have too much sediment imposed with your time varying boundary condition. flows of up to 0.5m/s won't carry 28g/l.
The same amount of sediment is transported down the channel with the flow, but with he prescribed sediment load, this amount is dwarfed by the massive amount of sediment depositing very close to the entrance (note you are not giving us the values of the colour scales in your figures).

Thanks.

I have put CS1 and CS2 in the keyword VARIABLES FOR GRAPHIC PRINTOUTS but my old fudaa version doesn't seem to recognize them...

OK, so there is probably a too great amount of sediment on the input. I'll have to double check my initial datas then.

Sorry, I did forget to put the scales on my figures: I don't have those results with me at the moment but I remember the bathymetry went from around 100m at the initial state to something closer to 500m (if not 1000m) after 5000 seconds of simulation.

Thanks for your answer.

Hello,
Add them (CS1 and CS2) directly into Sisyphe steering file using a text editor.
If you plot the ratio between U* and WS, you will be able to identify the zones where the sediment is prone to deposit. What are the critical values of the ratio? I am not sure how to know it in a certainly way.
However, Pablo answer me a question about it two months ago.

I hope it helps.

Cheers,
José Andrés.

Hi guys,
An approximate criterion for the onset of suspension is:
u* = w
where u* is the shear velocity and w is the particle settling velocity.
For values of u* less than w, there should be no suspension, and for
values of u* greater than w, some of the sediment should be traveling as
suspended load.

Be aware, this is VEEEERYYY approximative...

Best wishes,

Pablo

Thanks.
jose2kk wrote:
Well I did that, but fudaa is also the tool I use to plot my results and the CSi variable are nowhere to be seen when I open the results file with it.
Pablo wrote:
Thanks for the general idea! I'll check that out.

Hello all,

So, after reducing the desired input, I finally got to make the simulation run yesterday!

However, I am trying to run it again today but it just doesn't want to...

I am a bit lost. I have set an input of 0.25 g/L with the same size (0.001m) of sediment, and yet it looks like very few sediments enter the domain. Sisyphe ends up telling me only 1m3 entered the domain, and this amount doesn't change between a simulation of 1000 s, one of 2000 s and one of 6000 s...

Does anyone have any idea?
I have attached below the steering file (new_sis), the fortran file used to set the input (conlit_inverse) and the boundary file (new_sis_conlim).

Hello all,

So, after reducing the desired input, I finally got to make the simulation run yesterday!

However, I am trying to run it again today but it just doesn't want to...

I am a bit lost. I have set an input of 0.25 g/L with the same size (0.001m) of sediment, and yet it looks like very few sediments enter the domain. Sisyphe ends up telling me only 1m3 entered the domain, and this amount doesn't change between a simulation of 1000 s, one of 2000 s and one of 6000 s...

Does anyone have any idea?
I have attached below the steering file (new_sis), the fortran file used to set the input (conlit_inverse) and the boundary file (new_sis_conlim).