Hello all,
I have attempted a simulation in TELEMAC-2D coupled with SISYPHE with multiple
grain size fractions. The problem that I experience is that my virtual channel does not evolve in time. I should definitely observe changes in bed morphology with the high steady flow discharge applied to this single-threaded, 1.5-km long, 20-m wide, river channel (50m³/s). This is the equivalent of a Spring flood (after snow melt) where I would at least expect silt, sand and small pebbles to be mobile.
In the SISYPHE .cas file, my sediment classes are defined as follows:
NUMBER OF SIZE-CLASSES OF BED MATERIAL = 10
MEAN DIAMETER OF THE SEDIMENT =
0.002;0.235;0.375;12.0;26.0;42.5;61.0;78.0;103.0;340.0
INITIAL FRACTION FOR PARTICULAR SIZE CLASS =
0.375;0.375;0.19;0.006;0.012;0.012;0.012;0.006;0.006;0.006
Overall, soil mainly consists of fines (94% sand, silt and clay) but also includes pebbles (from 12mm to 340mm in diameter). I enabled bed load transport (with Meyer-Peter and Müller or van Rijn formulae), but disabled suspension. I made a few tests, varying the number of classes and grain sizes, and my problem seems to be related to the broad range of grain size classes defined in my SISYPHE .cas file. If I use a single sediment class, or if I use multiple classes with a small range of grain sizes, the bed evolves as expected.
A quick solution would be to eliminate a few classes to reduce the range of size classes. A range between 0.002mm and 2mm works fine. Unfortunately, I cannot remove the lower or upper classes because I'm simulating a post-glacial environment, in which the presence of both fine and coarse material contributes to explaining the physical condition and the dynamics of the river channel. In my case, the natural river bed is mostly covered by pebbles, but banks consist of a mixture of fine-textured soil with a few pebbles in it.
I quickly searched through the forum for posts on this and I did not find anything relevant. Let me know if you ever experienced a situation like this one.
Regards,
Yannick
