On the Role of Modelling Approach for Simulation of Bank Erosion in Straight, Trapezoidal Channel

Lukas Vonwiller, David Vetsch, Robert M. Boes

Tuesday 30 june 2015

16:45 - 17:00h at Mississippi (level 1)

Themes: (T) Sediment management and morphodynamics, (ST) Sediment transport mechanisms and modelling

Parallel session: 7A. Sediment - Erosion

Numerical models for the simulation of morphological changes in natural rivers have to account for bank erosion. This dynamic erosion process depends on the critical shear stress of the bank material, the flow and gravity-induced lateral transport, and the bank failure mechanism. To take bank erosion in a numerical model into account, different model concepts have to be considered and combined. We use a numerical depth-averaged 2-d model to simulate bank erosion in a straight trapezoidal channel with non-cohesive material based on Ikeda’s laboratory experiments. The bed load transport model includes Ikeda’s lateral transport approach and a geometrical approach for bank failure including the definition of the angle of repose for dry and submerged material. We study the role of these modelling approaches and the sensitivity of important model parameters. Interestingly, not only is lateral transport important, but also bank failure plays a crucial role. The most sensitive parameters for bank failure and the overall progressive erosion process are the angle of repose, the pre-factor of the sediment transport formula, the pre-factor of the lateral transport approach, and the grid resolution. By calibration of these parameters the results of the numerical model agree well with the experimental data. In particular, the dynamics of the lateral erosion and the aggradation of the channel bed are well captured. However, the model results strongly depend on the discretization of the computational grid. To overcome this shortcoming, alternative approaches are discussed in this contribution.