Sand dispersion in a laboratory flume
Date
1968
Authors
Yang, Tsung, author
Shen, H. W., advisor
Simons, Daryl B., committee member
Richardson, Everett V., committee member
Sandborn, Virgil A., committee member
Mielke, Paul W., committee member
Journal Title
Journal ISSN
Volume Title
Abstract
This study is concerned mainly with the longitudinal dispersion of sand particles along the bed of an alluvial channel under conditions of steady, uniform flow. Attention is focused on developing a general one-dimensional stochastic model to describe and predict the longitudinal dispersion process. The method of approach used by Sayre and Conover (1967) for a two-dimensional stochastic model, which described the movement of sand particles along an alluvial bed, is adapted here for the development of a general one-dimensional stochastic model. The parameters used in this general one-dimensional stochastic model can be obtained either from longitudinal dispersion and transport data, or from bed configuration data, or from a combination of both. The statistical analysis of ripple bed configurations indicates that the distribution of bed elevation closely follows a normal distribution, and may possess the ergodic property. The Aris moment equations are used to solve the problem of sand dispersion along an alluvial bed as a special case of the problem of dispersion of suspended sand particles near the bed. The Aris moment equations used in this study are modified forms of the conservation of mass equations for the transport, deposition, and re-entrainment of suspended sediment. When appropriate initial and boundary conditions are used, there is excellent agreement between solutions of the Aris moment equation and results given by the general one—dimensional stochastic model. Fine, medium, and coarse sized radioactive sand grains were used as tracer particles in experiments at two different flow conditions, namely, ripple and dune conditions. In spite of the irregularities of the experimental longitudinal dispersion curves caused by the irregularities of the bed configurations, the mean longitudinal displacement and the variance of the longitudinal distribution of the tracer particles were found to increase linearly with time, as required by the stochastic model. The shape of the experimental longitudinal dispersion curves could also be fairly well represented by the general one-dimensional stochastic model.
Description
Rights Access
Subject
Channels (Hydraulic engineering)