Comparison of five depth-averaged 2-d turbulence models for river flows

In this study, five depth-averaged 2-D turbulence models for river flows, including the depth-averaged parabolic eddy viscosity model, modified mixing length model, standard k-e turbulence model, non-equilibrium k-e turbulence model and re-normalized group (RNG) k-e turbulence model, are compared in the simulation of flows around a spur-dyke, in a sudden-expanded flume and in two natural rivers. It is shown that in the two field cases where the channel geometries are simple, all five models can give generally good predictions for the main flow features. However, in the two laboratory cases where the channel geometries are complex, differences have been found among these models. The depth-averaged parabolic eddy viscosity model over-predicts the recirculation flows behind the spur-dyke and the flume expansion. The modified mixing length model gives better prediction than the depth-averaged parabolic model. The standard k-e turbulence model predicts well for the recirculation flow in the sudden-expended flume, but under-predicts the length of recirculation zone behind the spur-dyke, while the non-equilibrium and RNG k-e turbulence models provide good results for both laboratory cases.