Vortex initiation detection and evolution analysis on a pitching airfoil under reverse flow conditions

Abstract

Dynamic stall under reverse flow conditions exhibits distinct aerodynamic behaviors compared to conventional stall, yet the timing of critical vortex events remains poorly understood. This study establishes a pressure-based method for detecting the initiation of the dynamic stall vortex (DSV) and trailing-edge vortex (TEV) for an airfoil under reverse flow conditions. Due to the dominance of the DSV and TEV structures on the surface pressure distributions, the proposed method provides a simple yet robust means of identifying key vortex events using single-port pressure measurements at the leading and trailing edges, thereby eliminating the need for information on the full surface pressure. Validation against the proper orthogonal decomposition method confirms the accuracy of the approach: the proposed method detects DSV and TEV initiation times with averaged relative deviations of 2.4% and 1.0% of the cycle period, respectively. While existing methods, such as the spatial distribution coefficient of pressure (SDCP) and the modulated location of peak pressure, face limitations in the reverse flow regime, the proposed method offers a robust alternative to the modified SDCP and high-order central moment of pressure methods, which require full surface instrumentation. The application of the proposed method reveals that the reduced frequency k is a crucial parameter governing DSV evolution, which proceeds through two distinct phases: an initial separation phase dominated by boundary layer separation and a major convection phase strongly influenced by reduced frequency. In the first phase, increasing k results in a milder evolution rate specifically for cases with DSV-dominated flow. In the second phase, however, the evolution rate decreases significantly with increasing k across all cases, leading to stronger vortices due to prolonged vorticity accumulation. For a low pitching amplitude of 5°, the evolution rate remains low for both phases regardless of k. In contrast, the mean angle of attack acts primarily as a phase-shift parameter, advancing or delaying the DSV life cycle without altering its intrinsic evolution dynamics.