ABSTRACT
Bed shear stress that is relevant for bed-load transport calculation is compared for a calibrated 1D and 2D numerical model for different channel geometries. With focus on the wall friction effect for different aspect ratios it is illustrated that for the prediction of cross-sectional averaged longitudinal bed-profiles a 2D model is not more precise per se than a 1D model. The 1D approach considers wall friction effects by the hydraulic bed radius for any aspect ratio. The 2D approach is capable to adequately consider wall roughness effects only by the use of a turbulence model. Otherwise bed shear stress will be overestimated in compact channels. Herein an algebraic turbulence model is applied that introduces additional uncertainties and unknown variables to be calibrated. Numerical 1D and 2D model simulations in compact channels are compared with flume data focusing on the lateral velocity and bed-shear stress distribution. Furthermore, the expansion of a narrow channel to a widened section is investigated. The results of the simulations using the algebraic turbulence model in combination with adjusted wall roughness are in good agreement with the results of the 1D simulation and experimental data.
