ABSTRACT
Measurements in a flume (0.6 m wide) with a gravel of median size 22 mm, a range of water discharges 0.01−0.1 m3/s and slopes of 3%,5%,7% and 9% are used, along with other flume and field data, to investigate the criteria which determine initiation of sediment transport in steep channels. The most common method for predicting initiation of transport is the Shields diagram which plots the Shields dimensionless shear stress against shear Reynolds number. However, this may not be appropriate to steep channels with small ratios of depth to sediment size because (a) the processes of sediment transport are not adequately described by shear stress considerations; and (b) the Shields diagram is poorly designed mathematically. By contrast the Schoklitsch approach, giving critical flow discharge as a function of slope, has a more suitable, process-related basis. The flume data support this analysis, indicating that the Shields diagram breaks down for steep channels (Fig. 1). The success of the Shields diagram elsewhere probably follows from its past restriction to flows with gentle slopes where variations in the Shields dimensionless shear stress are small (Fig. 2). The Schoklitsch approach explains the major variations in both field and flume data satisfactorily but effects related to bed packing and other field conditions have still to be quantified (Fig. 3). A critical discharge equation is proposed for flows with slopes > 2%.
