Integrating numerical and regression methods for estimating discharge coefficients of intake structures with wingwalls in irrigation networks

Abstract

Wingwalls facilitates the smooth direction of flow towards or away from the intakes of irrigation and drainage systems, thereby diminishing turbulence and energy loss. This study integrates experimental measurements, three-dimensional computational fluid dynamics (CFD), and advanced regression techniques to investigate the effects of wingwall angles (θ) in range of (0° ÷ 25°) on discharge coefficients for two intake types: sluice gates (C_d,SG) and broad-crested weirs (C_d,BW) under free flow conditions. Laboratory experiments of two wingwall's angle cases θ = 0° and 10° are used to validate the numerical model, ensuring reliable simulation of hydraulic behavior. A series of numerical simulations of C_d is then performed to generate a comprehensive dataset for model development. Numerical results indicates that the discharge coefficient of a sluice gate tends to increase as the wingwall angle (θ) increased, particularly when the ratio of total head (H_o) to opening height (a) was high. For weirs, C_d,BW also increases with θ and peaks at θ = 20°, indicating an optimal flow contraction configuration. Besides, regression equations were formulated using nonlinear least squares optimization based on the L-BFGS-B algorithm, enabling parameter estimation under box constraints. The proposed equations, incorporating sin(θ) and relevant geometric ratios, outperform traditional empirical formulas in accuracy, with higher CC and NSE and lower RMSE and MAE values. By explicitly modeling the influence of wingwalls, this study addresses a critical gap in hydraulic design, offering reliable predictive tools for optimizing flow control structures in irrigation systems.