Simulation of a curved flume bed-load experiment

The mathematical model for river bend morphology, as developed by Olesen, for bed-load transport is discussed, by comparing the results with some new experimental data. The model consists of a two-dimensional depth-averaged flow model together with a sediment balance and can be used to compute the dynamic river bend morphology. The main purpose of this investigation was to develop the data acquisition procedures for a suspended-load experiment in a curved flume by means of an experiment with bed-load only. The experiment provided some new data to test the present morphological model. In this report the results of the experiment are discussed. A remarkable point is that the experimental data show a persistent harmonic longitudinal perturbation of the bed topography. The computations, however, show a strongly damped bed topography, even for unrealistic parameter combinations. The computed depth at the point-bar and the corresponding pool is made to correspond with the experimental data, by a significant change of the submodel for the direction of the bed-load transport and the transport model itself. Probable causes for the discrepancy between the data and the computation are the small width-depth ratio and the low transport rate of the experiment (close to the initiation of sediment motion). The small width-depth ratio is an unfavourable circumstance in view of the assumptions used in the depth-averaged flow model. The low transport rate requires a modified transport model. Nonetheless a satisfying explanation for the bed oscillation, as observed in the experiment could not be found