August 28 - 30, 2013 Poitiers, France CON2E 1 ∞ a ∞ a α(t) = α0 + αP sin(ωt) k = ωc/(2U∞) Tburst Tdelay Tpulse Tstart α0 αP b α(t) = α0 + αP sin(ωt) k = ωc/(2U∞) Tburst Tdelay Tpulse Tstart α0 αP b Figure 1. The airfoil model with flow partitions and integrated combustion actuator array (a). The timing of model motion and actuation (b). For the present experiments, S c = S. UNSTEADY SEPARATION CONTROL USING SPATIALLY-COMPACT PULSED ACTUATION George T. K. Woo George W. Woodruff School of Mechanical Engineering Georgia Institute of Technology 771 Ferst Drive NW, Atlanta, Georgia 30332, USA gtkwoo@gatech.edu Ari Glezer George W. Woodruff School of Mechanical Engineering Georgia Institute of Technology 771 Ferst Drive NW, Atlanta, Georgia 30332, USA ari.glezer@me.gatech.edu ABSTRACT The dynamics of controlled 3-D transitory attachment of stalled flow over a static and dynamically pitching 2-D airfoil is investigated in wind tunnel experiments. Pulsed actuation is effected over a spanwise fraction of the separated domain on a time scale that is an order of magnitude shorter than the characteristic convective time scale using surface-integrated pulsed, combustion-driven actuator jets. The dynamics of vorticity concentrations over the airfoil and in its near wake is computed from high- resolution PIV measurements of the flow field that are obtained phase-locked to the actuation in multiple cross- stream planes. The present measurements show that transitory attachment spreads towards the outboard, unactuated flow domains and far exceeds the spanwise width of the actuator. The attachment is accompanied by the formation of 3-D vortical structures that are advected and shed into the near wake. It is shown that coupling of the pulsed actuation to the airfoil’s motion enhances the control authority of 3-D unsteady separation and can significantly mitigate the adverse effects of dynamic stall on the unsteady lift and pitching moment. EXPERIMENTAL SETUP AND PROCEDURES Transitory, 3-D attachment induced by pulsed actuation is investigated experimentally over a 2-D NACA-4415 airfoil (c = 457 mm, S ≈ 890 mm). The model is partitioned into three sections by streamwise acrylic fences that are installed symmetrically about the airfoil’s center plane z = 0 (Figure 1a). The center segment is instrumented with a spanwise array of seven, individually addressable combustion-based (COMPACT) jet actuators located at x/c = 0.15 symmetrically about z = 0 within the center segment while the two outboard segments are unactuated. In the present experiments, S c = S and the seven actuators are triggered synchronously producing high velocity pulsed jets (span S a ≈ 0.21S c ) that emanate from the airfoil upper surface via a rectangular orifice (0.19 x 20 mm). The airfoil model is mounted on a 2-DOF (pitch and plunge) traverse. The time-dependent lift force, C L (t), and pitching moment, C M (t), are measured independently using built-in load cells and a torque sensor. For the dynamic experiments, the model is oscillating about its pitch axis defined by α(t) = α 0 + α P sin(ωt) where α 0 is the nominal average angle of attack, α P is the oscillation amplitude, and k = ωc/2U ∞ is the reduced frequency. Figure 1b shows a schematic representation of the oscillatory motion and actuation timing. The experiments are conducted in an open-return wind tunnel with a test section measuring approximately 0.9 x 0.9 m. The free stream velocity is U ∞ = 20 m/s (Re c = 570,000) and the convective time scale of the flow over the airfoil is T conv ≈ 25 msec. The flow about the airfoil and in its near wake is characterized using phase-locked, high-speed particle image velocimetry (PIV) in several cross-stream planes (constant z) using two synchronized CMOS cameras (1280 x 800 pixels). The flow is seeded with micron-size fog particles and is illuminated using a double-pulse Nd- YLF laser. Sets of PIV images are captured at a sequence of predetermined time delays relative to the actuation trigger and model trajectory.