| Popis: |
This research work is aimed at developing a coupled surface-groundwater flow model which can be used to simulate both surface and groundwater flow at the swash zone. The coupled model is then used to investigate the effects of seepage on swash hydrodynamics as well as morphodynamics. The surface flow model was originally developed by Briganti et al. (2012), which solved a system of equations consisting of the Nonlinear Shallow Water Equations and the bed-evolution (Exner) equation with bed shear stress computed using a boundary layer model without seepage developed in Briganti et al. (2011). In this work, a groundwater flow model which solves Laplace's equation following the approach of Li and Barry (2000) is incorporated into the surface flow model, which allows computation of seepage into the bed (infiltration) and out of it (exfiltration). The seepage is then included into the boundary layer models to incorporate the effects of seepage on the bed shear stress. To assess the performance of the surface flow model, dam-break cases are simulated and compared against analytical and quasi-analytical solutions from literature. Firstly, the dam-break case on a fixed bed is simulated and compared against Ritter solution (Stoker, 1957) and then the dam-break case on a mobile bed is verified against Zhu (2012)'s quasi-analytical Riemann solver. Both models show good agreement with their respective reference results. Subsequently, the verification of the groundwater flow model is conducted by simulating phreatic surface flow through a rectangular dam and comparing the results against those of Kazemzadeh-Parsi and Daneshmand (2012). Next, the coupled surface-groundwater flow model is validated by reproducing surface and groundwater flow in the prototype-scale BARDEX II experiment. Firstly, the groundwater flow cases (higher and lower lagoon levels than the initial sea level) without surface water waves are simulated. The comparison of time-averaged numerical phreatic surface elevations against the experimental data shows excellent agreement. Next, the surface water waves are included and the simulations are repeated for the previous two cases. The groundwater comparisons again yield good agreement and the hydrodynamics of the surface waves show reasonably close agreement. Increase in exfiltration is observed to result in an increase in boundary layer thickness, which subsequently results in smaller velocity gradients and a decrease in bed shear stress using exfiltration included BBL model of Cheng and Chiew (1998). Conversely, the increase in infiltration causes a decrease in boundary layer thickness, which results in an increase in bed shear stress using infiltration included BBL model of Chen and Chiew (2004). The model results also show that the boundary layer effect by infiltration is opposed by the 'continuity effect' in the swash zone (Baldock and Nielsen, 2009). The model results show that an increase in infiltration rates is observed to increase slip velocity, and also compares well against the empirical equation derived in Chen and Chiew (2004). Furthermore, the rate of increase (decrease) of bed shear stress due to infiltration (exfiltration) compares favourably against the empirical trend line of Nielsen et al. (2001) and experimental data of Conley (1993). Additionally, the boundary layer model bed shear stress is compared against single swash event bed shear stress results from Kikkert et al. (2013) experiment and shows reasonably good agreement. The boundary layer models can be used to account for seepage effects on bed shear stressfor a larger range of ventilation parameters than Nielsen et al. (2001), which would improve morphodynamical modelling on permeable beds in the swash zone. Finally, the performance of the coupled surface-groundwater model is further investigated by simulating the BARDEX II experiment with a mobile bed. The swash zone water depth compares well with the BARDEX II experimental results. Although the corresponding dataset for velocity is shown to be rather unreliable during backwash, during uprush, the comparison is very close. Using both Meyer-Peter-Müller (MPM) and Grass sediment transport models, similar morphodynamical patterns are observed. The bed change comparisons against experimental results show that the model predicts the same order as well as the same pattern of erosion. However, deposition in the upper swash zone is not predicted by the model which could be due to the presence of significant amounts of suspended sediment which would lead to onshore sediment transport (Pritchard and Hogg, 2005, Zhu and Dodd, 2015) which is not accounted for in the simplified numerical model. The model is shown to be robust and flexible and it is capable of simulating both surface and groundwater flow simultaneously on fixed or evolving bed. |
