Session G7: Flow Control: Separation and Reattachment

Spatially Distributed Forcing for Boundary Layer Separation Control on a Wall Mounted Hump

Numerous successful efforts on controlling flow separation have been demonstrated using spatially distributed actuators. These include both steady and unsteady forcing from discrete locations in the vicinity of separation. Despite this, there are many open questions on the actual flow control mechanism. A canonical hump model is used to investigate these physics in a subsonic wind tunnel. Reynolds number independence is achieved above 0.72x10\textasciicircum 6 and testing is performed up to 2.2x10\textasciicircum 6. The efficacy of discrete steady jets is studied as a function of spacing, momentum coefficient, velocity ratio and mass flux. Highly-resolved surface pressure data for the controlled flow are compared to an inviscid solution establishing a figure of merit. Results indicate the inviscid limit is reached for a momentum coefficient of 1{\%} with actuator spacing of 0.5{\%} span. A comparison of steady discrete jets with sweeping jets actuators of equivalent cross-sectional area is undertaken. Surface flow visualization and PIV are employed to extract detailed information on the baseline and controlled flow field. This importance of establishing critical baseline features is also discussed with respect to establishing proper boundary conditions for accompanying numerical simulations.   

Exploring active flow control for efficient control of separation on an Ahmed model

Active flow control is applied to an Ahmed model with a rear slant angle of $25^{\circ}$, where a typical flow field consists of a three-dimensional separation region on the rear slant of the bluff body. Linear arrays of discrete microjets, previously proven to effectively control this separation, are investigated further. A principal aim of this experimental study is to examine the sensitivity of control as the actuator location is shifted with respect to the separation location. Aerodynamic force and surface pressure measurements, combined with the velocity field obtained using particle image velocimetry, provide a measure of control efficacy and insight into the interaction of jet arrays with the local flow field, including the separating shear layer. An energy balance is conducted to characterize control efficiency for multiple positions over a range of microjet array blowing conditions. Results show that moving the actuator array further into the separation region requires higher microjet momentum to obtain a desired aerodynamic benefit. An empirical relationship is also developed for determining the required jet velocity as a function of position by relating the jet penetration distance to local flow features and length scales.   

Proportional feedback control of flow over a hemisphere

In the present study, a proportional feedback control is applied to laminar flow over a hemisphere at Re = 300 to reduce its lift fluctuations. As a control input, blowing/suction is distributed on the surface of hemisphere before the separation, and its strength is linearly proportional to the transverse velocity at a sensing location in the centerline of the wake. To determine the optimal sensing location, we introduce a correlation function between the lift force and the time derivative of sensing velocity. The optimal proportional gains for the proportional control are obtained for the sensing locations considered. It is shown that the present control successfully attenuates the velocity fluctuations at the sensing location, resulting in the reduction of lift fluctuations of hemisphere.   

Active Flow Control on a Generic Trapezoidal Wing Planform

Fluidic oscillators are employed to increase the lift and improve longitudinal stability of a generic trapezoidal wing having aspect ratio of 1.15 and taper ratio of 0.27. Actuation is applied along the flap hinge which spans the entire wing and is parallel to the trailing edge. Experiments are conducted at a Reynolds number of $1.7\times10^6$ for a wide range of incidence ($-8^\circ$o to $24^\circ$) and flap deflection angles ($0^\circ$ to $75^\circ$). Baseline flow on the deflected flap is directed inboard prior to boundary layer separation, but changes to outboard with increasing incidence and flap deflection. The attached spanwise flow can be redirected using a sparse distribution of fluidic oscillators acting as a fluidic fence. However, the majority of lift enhancement and pitch break improvement is accomplished using a more dense distribution of actuators which attaches separated flow to the flap. Integral force and moment results are supported by surface flow visualization, pressure sensitive paint and PIV which reveal unique flow features such as a hinge vortex analogous to the leading edge vortex on a forward swept wing and the possible existence of an absolute instability in a plane parallel to the highly deflected flap.   

Active Control of Airfoil Boundary Layer Separation and Wake using Ns-DBD Plasma Actuators

Nanosecond pulse driven dielectric barrier discharge (ns-DBD) plasma actuators are employed to control boundary layer separation and the wake of a NACA 0012 airfoil having aspect ratio of three. Ns-DBD plasma actuators are known to operate via a thermal mechanism in contrast to ac-DBDs which are momentum-based devices. Nominally 2D forcing is applied to the airfoil leading edge with pulse energy of 0.35~mJ/cm. Experiments are conducted at a Reynolds number of $0.74\times10^6$ primarily at $18^\circ$ incidence which is well within the stalled regime. Baseline and controlled flow fields are studied using surface pressure measurements, constant temperature anemometry (CTA) and PIV. Forcing at a dimensionless frequency of $F^+=fc/U_\infty=1.14$ results in reattachment of nominally separated flow to the airfoil surface. Lower frequency forcing is less optimal for separation control, but produces strong fluctuations in the wake which are intended for use in the study of vortex body interaction in the future. Actuation below $F^+=0.23$ shows behavior consistent with an impulse-like response while forcing in the range $0.23 [Preview Abstract]

Passive Boundary Layer Separation Control on a NACA2415 Airfoil at High Reynolds Numbers

The design and analysis of a passive flow control system for a NACA2415 airfoil is undertaken. There exists a vast body of knowledge on airfoil boundary layer control with the use of controlled mass flux, but there is little work investigating passive mass flux-based methods. A simple duct system that uses the upper surface pressure gradient to force blowing near the leading edge and suction near the trailing edge is proposed and evaluated. 2D RANS analyses at $Re_{c} \approx 1.27\times10^{6}$ were used to generate potential configurations for experimental tests. Initial computational results suggest drag reductions of approximately $2-7\%$ as well as lift increases of $4-5\%$ at $\alpha = 10.0^{\circ}$ and $\alpha = 12.5^{\circ}$. A carbon composite-aluminum structure model that implements the most effective configurations, according to the CFD predictions, has been designed and fabricated. Experiments are being performed to evaluate the CFD results and the feasibility the duct system.   

Flow Physics of Synthetic Jet Interactions on a Sweptback Model with a Control Surface

Active flow control using synthetic jets can be used on aerodynamic surfaces to improve performance and increase fuel efficiency. The flowfield resulting from the interaction of the jets with a separated crossflow with a spanwise component must be understood to determine actuator spacing for aircraft integration. The current and previous work showed adjacent synthetic jets located upstream of a control surface hingeline on a sweptback model interact with each other under certain conditions. Whether these interactions are constructive or destructive is dependent on the spanwise spacing of the jets, the severity of separation over the control surface, and the magnitude of the spanwise flow. Measuring and understanding the detailed flow physics of the flow structures emanating from the synthetic jet orifices and their interactions with adjacent jets of varying spacings is the focus of this work. Wind tunnel experiments were conducted at the Rensselaer Polytechnic Institute Subsonic Wind Tunnel using stereo particle image velocimetry (SPIV) and pressure measurements to study the effect that varying the spanwise spacing has on the overall performance. Initial SPIV data gave insight into defining and understanding the mechanisms behind the beneficial or detrimental jets interactions.   

Nonlinear optimal control of bypass transition in a boundary layer flow

Bypass transition is observed in a flat-plate boundary-layer flow when high levels of free stream turbulence are present. This scenario is characterized by the formation of streamwise elongated streaks inside the boundary layer, their break down into turbulent spots and eventually fully turbulent flow. In the current work, we perform DNS simulations of control of bypass transition in a zero-pressure-gradient boundary layer. A non-linear optimal control algorithm is developed that employs the direct-adjoint approach to minimise a quadratic cost function based on the deviation from the Blasius velocity profile. Using the Lagrange variational approach, the distribution of the blowing/suction control velocity is found by solving iteratively the non-linear Navier-Stokes and its adjoint equations in a forward/backward loop. The optimisation is performed over a finite time horizon during which the Lagrange functional is to be minimised. Large values of optimisation horizon result in instability of the adjoint equations. The results show that the controller is able to reduce the turbulent kinetic energy of the flow in the region where the objective function is defined and the velocity profile is seen to approach the Blasius solution. Significant drag reduction is also achieved.   

The Dynamics of Controlled Flow Separation within a Diverter Duct Diffuser

The evolution and receptivity to fluidic actuation of the flow separation within a rectangular, constant-width, diffuser that is branched off of a primary channel is investigated experimentally at speeds up to M $=$ 0.4. The coupling between the diffuser's adverse pressure gradient and the internal separation that constricts nearly half of the flow passage through the duct is controlled using a spanwise array of fluidic actuators on the surface upstream of the diffuser's inlet plane. The dynamics of the separating surface vorticity layer in the absence and presence of actuation are investigated using high-speed particle image velocimetry combined with surface pressure measurements and total pressure distributions at the primary channel's exit plane. It is shown that the actuation significantly alters the incipient dynamics of the separating vorticity layer as the characteristic cross stream scales of the boundary layer upstream of separation and of the ensuing vorticity concentrations within the separated flow increase progressively with actuation level. It is argued that the dissipative (high frequency) actuation alters the balance between large- and small-scale motions near separation by intensifying the large-scale motions and limiting the small-scale dynamics. Controlling separation within the diffuser duct also has a profound effect on the global flow. In the presence of actuation, the mass flow rate in the primary duct increases 10{\%} while the fraction of the diverted mass flow rate in the diffuser increases by more than 45{\%} at 0.7{\%} actuation mass fraction.