Investigation on the suitability of two-dimensional depth-averaged models for bend-flow simulation

A numerical experiment is carried out to study the suitability of two-dimensional (2D) depth-averaged modeling for bendflow simulation, in which the geometry of the studied channel is rectangular. Two commonly used 21) depth-averaged models for bend-flow simulation are considered in this study of which the bend-flow model includes the dispersion stress terms by incorporating the assumption of secondary-current velocity profile, and the conventional model neglects the dispersion stress terms. The maximum relative discrepancy of the longitudinal velocity, obtained from the comparison of these two models, is used as a criterion to judge their applicability for bend-flow simulation. The analysis of simulation results indicated that the maximum relative difference in longitudinal velocity is mainly related to the relative strength of the secondary current and the relative length of the channel. Empirical relations between the maximum relative difference in the longitudinal velocity, the relative strength of the secondary current, and the relative length of the channel for both the channel-bend region and the straight region following the bend have been established. The proposed relations provide a guideline for model users to determine the proper approach to simulate the bend-flow problem by either using the conventional model or the bend-flow model. Experimental data have been adopted herein to demonstrate the applicability and to verify the accuracy of the proposed relations.