Abstract: River sediment transport capacity is of a key issue in suspended sediment studies. Due to the strong time-lag of fine sediment adjustment, it is difficult to obtain an equilibrium condition by the traditional recirculating flume experiments, especially for fine graded sediment. In this study, a special sand feeder and an outlet filter were implemented to the traditional experimental flume system, and experiments were carried out to approach the equilibrium transport process through sedimentation. Results were compared with traditional experiment data and the existing transport formulas, indicating: i) the implemented system can load more accurate inlet sediment for given compositions with the proposed feeder and outlet filter; ii) the non-equilibrium experiment can rather better approach equilibrium transport in a flume reach over the self-established depositional delta, where a sediment concentration, that is slighly higher than the actual transport capacity, can easily obtained without measurements; iii) results of limited runs indicate an extensively lower transport capacity than that from traditional experiments. Preliminary results reveal a somehow overestimation by the traditional formulas that calibrated using previous experiment data.

Key words: flume experiment, sediment transport capacity, equilibrium and non-equilibrium transports, sediment concentration