Optimized Mesh Zoning for Dam Bottom Outlet CFD Simulations Using Richardson Extrapolation and Grid Convergence Analysis

Accurate numerical prediction of hydraulic behavior in dam bottom outlets is essential for the safe operation and optimal design of hydraulic infrastructure. However, CFD simulations of bottom outlets face a key challenge: strong hydraulic gradients occur in localized regions near gate openings, while much milder flow variations exist in upstream and downstream zones. Conventional grid convergence studies typically apply uniform mesh refinement across the entire computational domain, which often results in excessive computational cost without proportional improvements in solution accuracy. To address this challenge, this study proposes a systematic mesh‐zoning methodology based on Richardson extrapolation and Grid Convergence Index (GCI) analysis to identify locally optimal mesh resolutions. Six mesh scenarios were evaluated using the realizable k–ε turbulence model in ANSYS Fluent, and key hydraulic parameters—velocity, pressure, and turbulence kinetic energy (TKE)—were sampled along a 31.2 m computational centerline at 200 locations. By integrating monotonic ternary mesh combinations, continuum‐value extrapolation, and a frequency‐based convergence assessment, the proposed approach determines spatial regions that require high mesh resolution while allowing coarser grids in areas with weaker gradients. The results indicate that Mesh #2 provides the optimal resolution in the critical gate region, whereas progressively coarser meshes are sufficient in upstream and downstream zones. The optimized zoned grid reduces the total computational cell count by ∼35%–40% compared with a uniformly refined mesh while maintaining deviations below about 2%–3% in discharge coefficient, wall pressure distribution, and key flow‐field parameters. The main contribution of this study is the development of a Richardson‐frequency‐based mesh zoning framework that enables reliable mesh‐independent CFD solutions while significantly improving computational efficiency for hydraulic structures with localized flow gradients.