Two successive knickpoints, with a 10% slope were constructed 1 m apart on a sandy bed in a rectangular flume with a longitudinal slope of 0.003. Bed erosion and knickpoint migration were studied experimentally for different discharges. The performance of two grade-control structures–Newbury rock riffles (NRR) and cross-vane riffles (CVR)–were studied experimentally for the stabilization of each knickpoint. Both of the structures were successful in controlling the bed erosion; however, the NRR operated relatively better than the CVR for they could concentrate the flow at the middle part of the channel to produce more regular contours with less local erosion and bed settlement. The experiments demonstrated that the construction of a control structure was not only effective in the stabilization of a knickpoint but also retarded the migration of its neighboring counterpart.

Keywords