|
ABSTRACT A fully three-dimensional non-hydrostatic model was applied to compute the flow and morphodynamic processes
in a laboratory flume. The laboratory experiments provided data for bed deformation and grain
sorting processes in a 180° channel bend. The data were used for validating the numerical model
applying three different sediment transport approaches. A formula for uniform sediments was compared
with two advanced non-uniform sediment transport formulas that considered the interaction between
individual grain sizes by defining hiding-exposure approaches. Figure 2 shows an
example of a three-dimensional view of the calculated bed deformation in the 180° channel bend.
Results showed that the three-dimensional model is able to predict bed deformations and sorting processes in the laboratory bend with the correct
magnitude and trend when using the default configuration. However, the computed results considerably
overestimated the bar evolution. Computed results improved significantly when applying any of the
two non-uniform sediment transport approaches. The best results were obtained when the absolute
values of the hiding-exposure exponents were reduced for both approaches. The results of the non-
uniform approaches showed less sorting intensity and smaller bar heights, both giving a better fit to
the measured data. The computed results of the non-uniform approaches in comparison with the default
configuration represent the physical mechanisms of the hiding-exposure effect well. The superior
agreement with compared to the default approach shows that these mechanisms ocour in nature and that
they significantly influence the bed deformation and sorting processes. Both non-uniform approaches
predicted the total sediment transport with satisfactory agreement.
|
