Numerical Modeling of Flow Patterns over Spillway

Abstract

The physical model is a main design tool to verify the spillway safety of small dams. The physical model contains many disadvantages, including a high construction cost and time-consuming. Therefore, the present study aims to performance assessment of numerical models as an economical alternative to physical models for simulation flow patterns over the ogee spillway. The small dam, Mandali, is selected as a case study. The Mandali dam is located 50 km east of Diyala city, northeast of Baghdad, Iraq. The Flow-3D software is used as the default numerical model to obtain the results of 3D spillway modeling. The Flow-3D is a commercial software to represent free surface flow and contains many turbulence closure models including the standard K-ε, RNG, K–ω, besides, large eddy simulation (LES) models. The present study also assesses performance of these models in simulate flow over the spillway. After mesh sensitivity analysis, a range of discharge values (389-1823 m3 /s) are tested to obtain the numerical model results including rating curve, discharge coefficients, pressure distributions, cavitation index and energy dissipation, then compared with data of the physical model of Mandali dam. According to the results, the maximum difference between simulation and physical model for discharge coefficients, piezometric pressures and relative energy losses is 6%, 4.5% and 2.7%, respectively. It is also observed that a decrease in discharge values over ogee spillway increasing the errors between the physical and numerical model. For a more accurate investigation of the numerical results of Flow-3D codes, CFD platform, Ansys -Fluent at a 2D domain and USBR data are presented. The results of the LES turbulence model and both platforms, Fluent-2D and Flow-3D, are in agreement with the physical model data. The results also indicate that numerical models give more agreement with USBR graphs better than the physical model. Finally, the CFD software, Flow-3D, can be adopted as an effective design tool for spillway of small dams, when the physical model causes greater errors, especially in low discharge and small scales.