An investigation of the ability of a numerical model to predict the cross sectional shape of an alluvial channel

摘要

The research analyses the behaviour of the flow in an open channel with self-formed banksides, with the purpose of exploring the ability of a numerical model to predict such geometries. The strategy consists in contrast a numerical model with physical model data. With respect to the numerical model, it is divided in two main parts, the first one describing the flow and the second one defining the cross sectional geometry. For the first part, a quasi 2D flow structure was selected to model the flow, i.e. the Shiono and Knight Model (SKM) (Shiono and Knight, 1991), due to its simplicity and flexibility. It allows users to incorporate the variation of friction factor, (f), secondary flow gradient, Γ, and dimensionless eddy viscosity, (λ), across the section. In order to calculate the bankside geometry, the Yu and Knight Model (1998) was chosen, because it relates the equilibrium of particles on the boundary to the shear stress, (τ)(0), distribution. udWith respect to the physical model, the shape of a self-formed bankside has been reproduced and assessed in a tilting flume, in order to identify its flow pattern by measuring velocity and shear stress. Such data was used for calibration and validation of the numerical model. The cross section was inspired in the bankside obtained by Ikeda (1981), fitting it into a flume 46cm wide. The experiment consists of testing the channel by three different slopes, three depths and two surfaces (smooth and rough), mapping velocity and measuring shear stress on the boundary across the section. The novelty of the work is to improve the flow estimation for this type of cross section, incorporating the secondary flow, and subsequently enhancing the approximation to the geometry that will be formed in alluvial channels.