ABSTRACT
Groynes are used for river training and providing navigation depth. Groynes add resistance, increasing the water level, which can be hazardous during floods. To reduce groyne resistance, different optimizations of the groyne geometry are considered. Often, numerical modelling is applied to assess the effects of such measures. Therefore, we investigate the performance of several numerical models in representing the flow over groynes, by comparing to laboratory measurements. Low- and high-resolution, two- and three-dimensional, as well as hydrostatic and non-hydrostatic computations are considered. Using a coarse-grid, two-dimensional model, with subgrid weir-losses for representing the groynes, the discharge over the groynes is strongly over-estimated. Better agreement is obtained using the non-hydrostatic, three-dimensional model. Due to the over-prediction of the discharge over the groynes also the effect of groyne modification is over-estimated. Therefore, coarse-grid, two-dimensional simulations, with the present subgrid weir-loss formulation, may not be appropriate for modelling the flow over (modified) groynes.
