A Review of Numerical Modeling of Internal Flow, Cavitation, Erosion, and Flow-induced Vibration in Hydropower Valves
Abstract
Hydropower valves operate under transient, high-pressure conditions in which local throttling, pressure recovery, vortex evolution, and pressure pulsation govern hydraulic performance, cavitation, erosion, and flow-induced vibration. This review synthesizes recent advances in numerical modeling of these coupled phenomena and links internal-flow evolution with hydraulic response, damage initiation, and structural behavior. The evidence base comprised 90 primary studies and five contextual reviews. Twelve primary studies (13.3%) directly addressed hydropower valves or hydropower valve–system interactions, whereas 78 studies (86.7%) provided transferable mechanistic or methodological evidence from related valve and hydraulic systems, highlighting the limited availability of hydropower-specific validation. Reynolds-averaged Navier–Stokes (RANS)-based computational fluid dynamics (CFD) remains the principal engineering method for global hydraulic quantities, whereas large-eddy simulation (LES), multiphase erosion models, and fluid–structure interaction (FSI) methods are more appropriate for unsteady cavitation, particle-induced damage, and structural response. Quantitatively, one direct hydropower butterfly-valve study reported an approximately 10% mean absolute error between CFD predictions and experimental flow-field data. In a transferable three-way globe-valve study, geometric optimization reduced the simulated percentage of cavitation by 49.5%, from 34.90% to 17.62%, while the response-surface model predicted 19.06%. The former supports case-specific confidence in hydropower-valve FSI modeling, whereas the latter demonstrates transferable optimization potential rather than direct hydropower performance. Major limitations include incomplete uncertainty reporting, uneven numerical verification and model calibration, scarce prototype-scale evidence, and limited long-term multiphysics validation. Standardized benchmarks, prototype measurements, and physics-informed monitoring are required to support transferable design and life-cycle decisions.