Two-dimensional finite element river morphology model

Alluvial streams are those that flow over a movable bed that they have formed by the transport of sediment, either suspended in the water current or dragged along the bottom. Bedload represents the fraction of that sediment that is transported along the channel bottom. Since bedload transport is associated with the removal (erosion) or deposit of sediment over the bed, it is the main transport mechanism responsible for the morphological changes in alluvial channels. Channel changes made by sediment deposition increase the risk of river flooding, block navigation canals, reduce the life span of reservoirs and modify the course of streams. While erosion can undermine the foundation of built structures (e.g. bridges, pipelines, bank revetments), scour riparian habitats or induce drawdown in the water table levels, just to mention a few effects. For those reasons, bedload transport in alluvial streams is a topic of high interest for both river engineers and fluvial morphologists. In this thesis, a two-dimensional numerical model based on the Finite Element method was developed for the morphodynamic simulation of scour and deposition in straight and curved alluvial channels. The model uses unstructured meshes formed by elements of triangular shape. The model was developed by coupling a solver for the bedload sediment continuity equation to existing fixed-bed hydrodynamic models to dynamically update the channel bed elevation as sediment is transport by the flow. The model was successfully applied to reproduce experimental observation of bed degradation and aggradation along straight flumes; as well as scour and deposition along curved channels. Data from a meandering river was also applied to test the model for full-scale applications. For the tests performed, the model proved stable and accurate. There are two versions of the model. A conventional vertically averaged (VA) model and a novel vertically averaged and moment (VAM) model. The VAM is a quasi-3D flow model that provides more detail of the vertical structure of the flow than a VA model for curved channels. The model has some unique capabilities for modeling rapidly varied flows and transcritical flow (supercritical flow and hydraulic jumps) over alluvial beds. This is probably the first two-dimensional river morphology model based on unstructured triangular Finite Elements that has combined capabilities for bedload transport, secondary flow effect in bends, transverse bed slope and transcritical flow. The results of this thesis suggest that the model has the potential for future morphodynamic applications to dam-breaks, dam removal and steep mountain streams.Applied Science, Faculty ofCivil Engineering, Department ofGraduat