ABSTRACT
Transverse exchange processes of mass and momentum in floodplain regions of channels are of primary importance regarding the sediment transport and riverbank stability. The presence of large horizontal coherent structures (LHCSs) at the interface of the floodplain and main channel regions may contribute up to 90% the amount of transverse momentum exchange between these areas. Although many momentum exchange models have been proposed and developed, their applicability in different circumstances is still unclear as their validity is usually restricted to a narrowly ranging experiment data set. In order to obtain more insight, two unique laboratory experiments of a shallow flow field in a floodplain channel with and without vegetation have been conducted. One small scale experiment was conducted at the TU Delft Water Lab. Another large-scale experiment of floodplain vegetated channel has been conducted at the Korea Institute of Civil Engineering and Building Technology - River Experiment Center (KICT-REC). The experimental data has been used to verify state-of-the-art momentum exchange models. As the limitations of these models were analyzed, a new eddy viscosity model based on the occurrence of LHCSs was proposed and validated using a variety of experimental data sets. A numerical model mimicking physical models was constructed. The experimental results were compared with the numerical results, showing the capacity of the new eddy viscosity model. Furthermore, the experimental results confirm the presence of LHCSs in a large-scale experiment. The LHCSs have the length of about 15m, which is one order of magnitude larger than that observed in the small-scale experiment.
