Principles and practice of hydraulic modelling of braided gravel-bed rivers

Abstract

Hydraulic modelling permits the forms and processes of braided rivers to be studied under closely monitored and controlled experimental conditions. The advantages of this approach stem from the degree of control available, the reduced physical size in the model, and the increased rate of geomorphic evolution. This paper outlines the principles of the hydraulic modelling of braided gravel-bed rivers, describes the practical limitations of this approach and compares model and prototype characteristics. Modelling procedures are based on the principles of hydraulic (dynamic) similarity. Models of braided river systems involve mobile bed modelling of complex two-phase flow. However, restrictions imposed by scaling ratios for gravitational acceleration, fluid viscosity and fluid density make it impossible to achieve full dynamic similarity, except with a length scale of unity. Therefore model experiments use approximate dynamic similarity, which to he satisfied only requires similarity of relative depth between the model and prototype. Practical limitations of flume size, sediment management and cost often limit model scales; because of the requirement of rough-turbulent flow in Froude-law models, it is seldom possible to use models smaller than I:50 linear scale ratio. Modelling is therefore usually restricted to modelling typical reaches of smaller braided rivers or a group of prototype rivers. Because prototype rivers cannot be described adequately to carry out a detailed assessment of hydraulic similarity, quantitative prediction using these models is limited. Nevertheless, hydraulic modelling is a very attractive method for investigating braided river processes, and given careful design, a wealth of valuable data and information can be obtained with relative case.