This thesis examines the dependency between total sediment transport and grain size under dam break generated swash flows. Experiments were performed in a dam break flume over sloping and horizontal beds using a mobile sand bed with median grain sizes ranging from 0.22mm to 2.65mm. The total sediment transport was measured by truncating the flume bed and collecting the total sediment transported over the edge. The experiments were designed to exclude pre-generated turbulence and pre-suspended sediment so as to focus solely on the swash flow. Due to the difficulties in obtaining the full duration of swash uprush velocities from measurements, the present study used the depth averaged velocity, predicted by the uncoupled finite volume model using shallow water solver, TUFLOWfv. The flow depth, tip celerity and the overtopping volume predicted by TUFLOWfv were calibrated against the measurements with model accuracy within a ±20% error band. The influence of coupling between the hydrodynamics and morphodynamics and total sediment transport prediction were investigated via the hydro-morphodynamic model of Postacchini et al. (2012). Fairly good agreement was achieved between experimental data and the hydro-morphodynamic model for tip celerity and total sediment transport for the uprush phase. Insignificant differences were obtained between coupled and uncoupled model runs, indicating a coupled model is not required to estimate flow kinematics in these experiments.
The magnitude and nature of the grain size dependency was inferred for different flow parameters; the initial dam depth, the integrated velocity cubed and against the predicted transport potential using the Meyer-Peter Muller (MPM) transport model and variations of that model. The data show that negative dependencies are obtained for the initial dam depth and the predicted transport potential, while positive dependencies are obtained for integrated velocity cubed. This indicates that a given initial dam depth and predicted transport potential transport less sediment as grain size increases, whereas transport increases with grain size for a given integrated velocity cubed. The negative dependencies for the predicted transport potential are consistent with Bagnold (1980; 1986). The derived dependencies based on the predicted transport potential are not altered using either fixed or time varying friction factors. The inferred dependency changes only with the incorporation of a pressure gradient, predominantly for the coarse sands on a horizontal bed. On average, the incorporation of a pressure gradient term via piezometric head into the MPM formulation reduces the predicted transport in the uprush by 4% (fine sand) to 18% (coarse sand) and increases predicted transport over a horizontal bed by 1% (fine sand) to two orders of magnitude (coarse sand). However, for the backwash, overprediction occurred even without the pressure gradient term in the model, particularly for larger grains. Lower friction factors are needed to reduce the predicted bed shear stresses and prevent the overpredictions in the backwash. It is suggested that future swash sediment transport models should incorporate the grain size effect, partly through the pressure gradient, although the importance of this correction will mainly occur for large grain sizes and for larger positive pressure gradients.
The inferred optimum transport coefficient obtained for the uprush and the horizontal bed in the MPM formulation are about 30 to 40, much higher than the standard coefficient for steady flow, and this was not due to the presence of the pre-suspended sediment. A closer value of about 13 was obtained for the backwash. The optimum transport coefficient indicates some sensitivity to grain size, suggesting that some transport processes remain unaccounted for in the model.
A new dimensional analysis has reconciled the imbalanced dimensions in Bagnold's (1980; 1986) formulation and produced a dimensionally balanced formula. It was also shown that the dimensionally balanced formula can be written to satisfy Bagnold's (1986) empirical D scaling and u scaling in the literature. A multi linear regression analysis conducted using the uprush data via the deduced dimensionless groups support the MPM transport model and suggest that the contribution of other dimensionless groups are insignificant.