Simulation of suspended sediment and contaminant transport in shallow water using two-dimensional depth-averaged model with unstructured meshes

摘要

With growing global populations and human activities, the problems of sediment transport and environmental hydraulics have been becoming more and more important and valued. Computational fluid dynamics is a very useful and powerful tool in those studies. In 2008, Dr. Liu developed a two-dimensional model HydroSed which solves for 2D shallow water equations and sediment bedload transport equation at Hydrosystems Laboratory, UIUC. In this thesis, the author further modifi es and improves the HydroSedv1.0 modelso that the model is able to deal with the transport of contaminant. In addition, the transport of suspended sediment is added and it is coupled with bed evolution.The 2D advection-di usion equation with source/sink terms is solved by finite volume method using Godunov scheme explicitly on unstructured meshes. Roe's approach is used to solve Riemann problem because the analytical solver is much computationally slower. Slope-limiter approaches are appliedin order to get higher accuracy. Several pure advection tests are simulated to evaluate the performance of the model and the effect of diff erent slope-limiterapproaches. Furthermore, a dye-tracer study in the Chicago Riveris simulated for verifying the model in a real-world project. The numerical results show satisfactory matches with the field measurement data. A widely used three-dimensional model EFDC is also used for comparison. Both pure advection tests and dye-tracer study demonstrate that the performance ofthe model is satisfactory.The model is applied to study the potential impacts of planned hydraulic structures in the Canar River, Ecuador. The effects to both water flow during flood and sediment transport are studied. Due to the lack of eld data in terms of water surface elevation, flow discharges and velocities in this river, the HydroSed model can help to a large extent to understand the problem and work together with a 1D model and other techniques for optimized designs.