Effects of suspended sediments on river bars

Bars are large sediment deposits in river channels emerging at low flows with wavelength and height scales of the order of the channel width and of the flow depth, respectively. Bars develop due to morphodynamic instability of flat river beds. They consist of recurring sequences of erosion holes and deposition fronts. Bars create problems to river navigation, intakes, and cause bank erosion, making their morphological study and management an important part of river engineering and rehabilitation works. The study of bars should include both bed and suspended sediment processes, as rivers normally have both fine and coarse material. However, only few theoretical works focused on the effects of suspended sediment transport on bars. In particular, experiments showing the effects of suspended sediments are still lacking. This research is conducted to fill this gap. The research investigates the effects of both bed load and suspended load transport on bar characteristics based on laboratory investigation and 3D numerical simulations. The laboratory experiments were carried out in a flume with mobile bed and fixed banks with the aim of visualize the formation and characteristics of alternate bars in relation to the transport mechanism of sediment, as bed load and suspended load. The flume was 0.20 m wide, 5.0 m long and 0.25 m deep, located in TU Delft University of Technology. The flume was filled with either river-sand or plastic-sand, in succession. Both sands have identical sediment size distribution, but different density so that one is transported in suspension and the other one as bed load at the same flow conditions. A groyne obstructing half of the channel width was constructed near the upstream boundary to facilitate steady bar formation by creating a permanent flow perturbation. The results show that there are differences in both longitudinal bed slope and bars. In case of bed load dominated transport, steady alternate bars dominated the flume bed in all the runs except one, presenting a central bar in the second half of the flume (smallest discharges). Only steady alternate bars were formed in case of suspended load dominated transport. The flume experiments produced similar wavelengths for both bed load and suspended load transport with the exception that bars were damped when suspension was dominant. Finally, a 3D morophodyanamic model was setup using Delft3D to reproduce an upscaled version of the flume experiments using the similarity scaling procedure. The goal of the numerical model was to establish what type of water course is represented by the flume experiments and assess also whether the flume results can be interpreted for real rivers. Two sets of simulations were performed, one representing bed load and the other suspended load dominated cases. Each set corresponds to the upscaled version of two sets of flume experiments. For dominating bed load, the model showed the development of steady alternate bars. In case of suspended load dominated transport, the alternate bars were damped (high damping coefficient) leading to flatbed down steam of the first two alternate bars. The numerical simulations produced results that are similar to the corresponding results of the flume experiments for both bed load and suspended load dominated transport. Both the flume experiments and numerical simulations showed that bars became damped when suspended load transport was dominant. This confirmed the analytical findings of Federici & Seminara (2006) and Watanabe & Tubino (1992). Since the numerical model and flume tests produced similar results, we can reasonably state that the flume experiments can be considered representative of real river cases. The upscaling of the flume experiments has further shown that the laboratory experiments represent the behaviour of a gravel-bed river with steep slope, fixed banks and single thread, like the ones studied by Parker et al. (2007). An example of this river type is the upper Rhine river along the border between Switzerland and Liechtenstein (Jaeggi, 1984)