Session M16: Flow Control: Separation

Interaction of a Synthetic Jet Actuator with a Severely Separated Crossflow

A coordinated experimental/computational study of synthetic jet-based flow control on a vertical tail/rudder assembly has been carried out on a 1/19th scale model operating at 30 degree rudder deflection, 0 degree side slip, and 20m/s free-stream flow. Under these conditions a very strong span-wise separated flow develops over the rudder surface for a majority of its span. Twelve synthetic jets were distributed across the span of the vertical tail just upstream of the rudder hinge-line to determine their ability to reduce flow separation and thereby increase the side force production; to extend the rudder effectiveness. Experiments were completed for the baseline case (i.e. no jets blowing) and for cases where 1, 6, and 12 jets were activated. RANS and DDES computations were completed to match these four experiments. While some experimental results for the same geometry have been previously reported, more detailed results concerning the experiments and their comparison to the DDES computations for the baseline and 1 jet active cases are reported here. Specifically, this effort focuses on the near-jet flow and the phase-averaged vortical structures produced by a single jet interacting with a severely separated, turbulent cross-flow.   

Synthetic Jet Interactions with Flows of Varying Separation Severity and Spanwise Flow Magnitude

Flow physics associated with the interactions of synthetic jet actuators with a highly three-dimensional separated flow over a flapped airfoil were investigated experimentally and analyzed using stereo particle image velocimetry (SPIV) and surface pressure data. Increased understanding of active flow control devices in flows which are representative of airplane wings or tails can lead to actuator placement (i.e., chordwise location, spanwise spacing) with the greatest beneficial effect on performance. An array of discrete synthetic jets was located just upstream of the control surface hingeline and operated at a blowing ratio of 1 and non-dimensional frequency of 48. Detailed flowfield measurements over the control surface were conducted, where the airfoil's sweep angle and the control surface deflection angle were fixed at 20\textdegree . Focus was placed on the local and global flowfields as spanwise actuator spacing was varied. Moreover, surface pressure measurement for several sweep angles, control surface deflection angles, and angles of attack were also performed. Actuation \quad resulted in an overall separation reduction and a dependence of local flowfield details (i.e. separation severity, spanwise flow magnitude, flow structures, and jet trajectory) on spanwise jet spacing.   

Incorporating swirl effects into the coefficient of momentum for separation control

Addition of swirl in flow control has been known to enhance suppression of separation over airfoils at high angles of attack. Utilizing large eddy simulations, the present open-loop control study examines the influence of wall-normal and angular momentum injections in mitigating separation over a NACA0012 airfoil at $\alpha = 9^\circ$ and $Re = 23,000$. We introduce these swirling jets near the separation point with wall-normal momentum and swirl independently prescribed through velocity boundary conditions. The changes to the flow from control are examined and the corresponding lift enhancement and drag reduction are assessed as a function of the two velocity components. Since the standard coefficient of momentum does not consider swirling effects, we extend its definition to incorporate both the wall-normal momentum and swirl to quantify the overall flow control effectiveness. We are able to observe a trend in lift force enhancement over this single modified coefficient of momentum (that is dependent on the non-dimensional jet velocity ratio and swirl number). Moreover, we are able to identify a critical value for the modified momentum coefficient and categorize controlled flows into separated, transitional, and attached flows.   

Controlled Aerodynamic Loads on an Airfoil in Coupled Pitch/Plunge by Transitory Regulation of Trapped Vorticity

The aerodynamic loads on an airfoil moving in coupled, time-periodic pitch-plunge beyond the static stall margin are controlled using transitory regulation of trapped vorticity concentrations. ~Actuation is effected by a spanwise array of integrated miniature chemical (combustion based) impulse actuators that are triggered intermittently during the airfoil's motion and have a characteristic time scale that is an order of magnitude shorter than the airfoil's convective time scale. ~Each actuation pulse effects momentary interruption and suspension of the vorticity flux with sufficient control authority to alter the airfoil's global aerodynamic characteristics throughout its motion cycle. ~The effects of the actuation are assessed using time-dependent measurements of the lift and pitching moment coupled with time-resolved particle image velocimetry over the airfoil and in its near wake that is acquired phased-locked to its motion. ~It is shown that while the presence of the pitch-coupled plunge delays lift and moment stall during upstroke, it also delays flow reattachment during the downstroke and results in significant degradation of the pitch stability. ~These aerodynamic shortcomings are mitigated using superposition of a limited number of pulses that are staged during the pitch/plunge cycle and lead to enhancement of cycle lift and pitch stability, and reduces the cycle hysteresis and peak pitching moment.   

Separated Flow Control with Actuated Membrane Wings

By perturbing shear layer instabilities, some level of control over highly separated flows can be established, as has been demonstrated on rigid wings using synthetic jet actuators or acoustic excitation. Here, we demonstrate similar phenomena using sinusoidal actuation of a dielectric membrane wing. The effect of actuation on lift is examined as a function of freestream velocity (5-25 m/s), angle of attack (10$^\circ$-40$^\circ$), and actuation frequency ($0.1 [Preview Abstract]

Separation Dynamics of Controlled Internal Flow in an Adverse Pressure Gradient

The effects of fluidic actuation on the dynamic evolution of aggressive internal flow separation is investigated at speeds up to M $=$ 0.4 within a constant-width diffuser branching off of a primary flow duct. It is shown that a spanwise array of fluidic actuators upstream of the separation actively controls the flow constriction (and losses) within the diffuser and consequently the local pressure gradient at its entrance. The effectiveness of the actuation, as may be measured by the increased flow rate that is diverted through the diffuser, scales with its flow rate coefficient. In the presence of actuation (0.7{\%} mass fraction), the mass flow rate in the primary duct increases by 10{\%} while the fraction of the diverted mass flow rate in the diffuser increases by more than 45{\%}. The flow dynamics near separation in the absence and presence of actuation are characterized using high speed particle image velocimetry and analyzed using proper orthogonal and spectral decompositions. In particular, the spectral contents of the incipient boundary layer separation are compared in the absence and presence of actuation with emphasis on the changes in local dynamics near separation as the characteristic cross stream scale of the boundary layer increases with separation delay.   

Active Flow Control in an Aggressive Transonic Diffuser

A diffuser exchanges upstream kinetic energy for higher downstream static pressure by increasing duct cross-sectional area. The resulting stream-wise and span-wise pressure gradients promote extensive separation in many diffuser configurations. The present computational work evaluates active flow control strategies for separation control in an asymmetric, aggressive diffuser of rectangular cross-section at inlet Mach $\sim$ 0.7 and Re $\sim$ 2.19M. Corner suction is used to suppress secondary flows, and steady/unsteady tangential blowing controls separation on both the single ramped face and the opposite flat face. We explore results from both Spalart-Allmaras RANS and DDES turbulence modeling frameworks; the former is found to miss key physics of the flow control mechanisms. Simulated baseline, steady, and unsteady blowing performance is validated against experimental data.