I have tried to examine the effect of sediment-induced stratification on open channel flows with cohesive sediment. The flow is a steady homogeneous open channel flow. Flocculation and hindered settling processes are incorporated into the model using a simple algebraic formula.
Sediment is introduced into the flow by bottom erosion. By choosing a series of bed erodibility constant, the sediment-induced stratification varies from sub-saturation to super-saturation. Two experiments are shown in the figure below. The upper plane (a-d) is under sub-saturation condition. The lower plane (e-h) represents the case under super-saturation condition.
The velocity and SSC show very large value above the bottom, particularly, under super-saturation condition. I don’t know the abnormal value is computation error or bottom boundary problem. Does it mean the two-equation turbulence model can not correctly represent bottom laminar layer?
1) the settling velocity algorithm requires a constant settling value in the vertical. This value can vary for each sediment class, and can vary in x and y directoins. So I am not sure how you modified the code to deal with the variation in ws based on local (?) concentration.
2) for the top row of sub-saturated:
The vel and ssc look ok. The Eddy visc has a strange kink near the bed, not sure what that is from. The ws varies over the vertical, but the PPM model can not deal with this.
3) For super saturated, the SSC has a large value near the bed. This must be from bottom BC. If you are pumping too much sed into the water column, then it can not mix upwards fast enough.
4) The ws is a function of ssc, so the high values nera the bed must be from the ssc.
5) the 2 equation models do not resolve the laminar sublayer. I do not think your model is resolving that layer either, at the scales in the plot.
Thanks a lot, John. I simply used the local sediment concentration in the calculation of settling velocity which may cause this problem. When constant settling velocities are used, we don't have the abnormal bottom boundary.
