Abstract
Data have been collected for regime conditions in trapezoidal, rigid-sided channels of laboratory scale. Immediate analysis suggested that surface width and mean depth of the main channel are interdependent and that the effects of bank slope, and of bed sediment load and size on the cross-sectional geometry are less important. There was indication that a log-normal relation held between non-dimensional width and non-dimensional depth. The combination of the new laboratory data with existing data, particularly from prototype scale canals and rivers, allowed analysis of the cross-sectional geometry, using the log-normal premise, through the spectrum of scale encompassed by the representative data treated. A functional form has been obtained for the relationship between channel surface breadth and mean depth of main channel. For channel slope, a numerical correlation with discharge has been obtained. In both cases the Froude number of the main channel is the correlating factor. These relationships show reasonable consistency when tested against the laboratory and prototype data, and allow the formulation of possible geometries for average regime conditions once a characteristic discharge has been selected. The predictive capacities of existing formulae for sediment transport were compared for the conditions under which the new data were obtained and the results are outlined. The principal conclusions are that there is no discontinuity between laboratory and field circumstances and that the standard practices regarding regime channel design may be unduly restrictive. Nevertheless, the traditional regime rules are compatible with the new findings over portions of he spectrum of scale of regime channels.