Turbulence modelling for depth-averaged velocity and boundary shear stress of a dense rigid grass bed open channel

Abstract

The present research focusses on a comparison of experimental and numerical approaches for flow over fixed artificial rigid grass bed channels. Various flow parameters like longitudinal velocity, depth-averaged velocity (DAV), boundary shear stress (BSS) and secondary current are analysed and compared with seven numerical models: standard, realizable and renormalization group (RNG) k–ε models and standard, shear stress transport (SST), generalized k–ω (GEKO) and Baseline (BSL) k–ω models. To evaluate the strength of the seven applied models, the error analysis has been performed. It is found that the RNG k–ε and SST k–ω models provided better results for both the DAV and BSS prediction, but the RNG k–ε model is found to be the most suitable for predicting the DAV and the SST k–ω model for BSS as compared to the other models. For the longitudinal velocity profiles, both the RNG k–ε and SST k–ω models are found to provide good agreement with experimental results at the centre of the channel, whereas the SST k–ω model is more accurate near the wall. Overall, the SST k–ω model has predicted the results with good accuracy for all the flow parameters considered in the present study. HIGHLIGHTS A comparative study of all the seven sub-models pertaining to the k–ω and k–ε groups was performed.; DAV and BSS profiles are presented using the discussed turbulence models, CES and compared with the experimental results.; Statistical error analysis is performed.; The RNG k–ε model estimated the depth-averaged velocity more accurately, whereas the SST k–ω model is found to be more accurate in predicting the boundary shear stress.;

Keywords

• ansys fluent
• experimental results
• grass bed
• k–ε and k–ω models
• statistical error analysis