关键词: 动态失速, 数值模拟, 气动特性, 验证与确认, 计算流体力学

Abstract: Dynamic stall is a common unsteady aerodynamic phenomenon encountered by rotorcraft and wind turbines, and its numerical simulation is of significant value for engineering prediction and mechanism analysis. However, current research predominantly compares a single simulation with experimental data and lacks an in-depth analysis of the discrepancies caused by various numerical parameters, leading to ambiguous reliability boundaries and limiting the capability of Computational Fluid Dynamics (CFD) to transition from a reproductive to a predictive tool. Focusing on the pitching oscillation of a NACA0012 airfoil in a deep dynamic stall state, the effects of epistemic uncertainty introduced by key modeling choices were evaluated and quantified. The discrepancies in simulation results caused by key parameters, including grid types, turbulence models, pressure-velocity coupling methods, and boundary conditions, were systematically investigated, and the consistency of the numerical simulation with experimental data was assessed using time-history curves and hysteresis loops. The results indicate that with a proper configuration of modeling parameters, the simulation can closely approximate the experimental data, exhibiting higher fidelity, particularly during the critical stall-vortex evolution phase. Through a systematic comparative analysis of parameter combinations, a modeling reference scheme with high stability and accuracy was established, and by using interval quantification methods, the uncertainty band for the aerodynamic coefficients of the benchmark case was established. This provides practical guidance for subsequent numerical studies of dynamic stall.

Key words: dynamic stall, numerical simulation, aerodynamic characteristics, verification and validation, computational fluid dynamics