Freestream Turbulence Conditions Research Articles

Simultaneous measurement of the distributed longitudinal strain and velocity field for a cantilevered cylinder exposed to turbulent cross flow

The structural response of a cantilevered cylinder under free-stream conditions both with low and high turbulence intensity generated by turbulence-generating grids is analysed with concurrent measurements of the velocity field and the distributed strain. The thin-walled cylinder, with a diameter (D) of 50 mm, is mounted in a water flume as a cantilevered beam supported at one end, with 95% of its body submerged and exposed to cross flow yielding a Reynolds number of Re = 25000. We employ a novel combination of simultaneous particle image velocimetry and distributed strain measurements using Rayleigh backscattering fibre optic sensors. These sensors are embedded onto the surface of the cylinder to measure the experienced strain (ε\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\varepsilon $$\\end{document}) of the structure along its spanwise direction, covering both the windward and leeward faces of the cylinder. The sensors fine spatial resolution allows us to discern the influence of the flow on the structural response of the cylinder in two distinct regions of the structure: upstream and downstream of the mean separation location. This differentiation allows us to isolate the local effects introduced by the free-stream conditions on the loading events over the body from the global force generated by the vortex shedding and other coherent motions present within the flow. Distinguishing between these direct and indirect effects helps determine which is more relevant for fatigue-life cycle analysis. The cross power spectral density between the fluctuating velocity field and the strain reveals that the load is dominated by the vortex shedding, and this relationship is intensified with the introduction of free-stream turbulence. It also helps to discern the different dynamics imposed by the two free-stream turbulence conditions.

Effects of integral length scale variations on the stall characteristics of a wing at high free-stream turbulence conditions

The effect of variations in the integral length scale of incoming free-stream turbulence on a NACA0012 wing is investigated with the use of force, moment and particle image velocimetry measurements. At a chord-based Reynolds number ( $Re = U_\infty c/\nu$ where c is the chord length, $U_\infty$ is the free-stream velocity and $\nu$ is the kinematic viscosity) of $2\times 10^5$ , an active grid generates turbulence intensities of 15 % at normalised integral length scales ranging from 0.5 $c$ to 1 $c$ . The introduction of turbulence improves the time-averaged performance characteristics of the wing by delaying stall and increasing the peak lift coefficient. It is found that for half-chord integral length scales, the magnitude of the fluctuations in forces and moments is larger than that of full-chord integral length scales, as the former amplifies the naturally occurring unsteadiness in the flow (when there is no free-stream turbulence). The increase in magnitude is ascribed to a larger density of smaller-scale vortices within the separated flow and wake region of the wing.

Mass-Heat Coupling in Film Cooling Experiments—The Influence of Temperature and Turbulence on the Dufour Effect

Abstract The use of foreign gases in laboratory film cooling experiments is attractive since variable density ratios can be achieved with coolant-to-freestream temperature ratios near unity, often reducing the cost and difficulty of the experimental campaign. In adiabatic effectiveness experiments employing pressure sensitive paint along with the mass transfer analogy to heat transfer, isothermal surfaces are often an experimental requirement. Furthermore, low-temperature laboratory experiments using thermal techniques often employ relatively close matches between the coolant and freestream temperatures. Using foreign gases, however, introduces off-diagonal couplings of heat and mass transport, which can produce unexpected results in film cooling experiments. In particular, the Dufour effect—also called the diffusion-thermo effect—which is the transfer of thermal energy by mass transfer processes, can manifest in surface temperatures that break the traditional bounds of thermal adiabatic effectiveness experiments: outside the upper and lower bounds of the coolant and freestream temperature. Beyond the expected confusion for the researcher, this effect can also be detrimental to those that assume that matching the coolant and freestream temperatures are the necessary and sufficient conditions to ensure isothermal surface conditions in traditional pressure sensitive paint experiments. In this work, the influence of cooling gas selection, experimental temperature, and experimental freestream turbulence conditions are explored on a simulated leading edge with compound injection from a cylindrical cooling hole. Air, argon, carbon dioxide, helium, and nitrogen coolants were analyzed due to their use in prior film cooling studies. The Dufour effect was found to be significant when using helium as the coolant, though temperature separation was also observed in argon and carbon dioxide cases. Additionally, elevated experiment temperatures generally increased temperature separation. Finally, high freestream turbulence intensity was found to reduce, but not eliminate, the Dufour effect in helium experiments.

Effects of free-stream turbulence on motion-induced aerodynamic forces of a long-span flexible flat roof using LES simulations

This study aims to investigate the effects of free-stream turbulence on motion-induced aerodynamic forces of a long-span flexible flat roof. Large-eddy simulations (LES) of both rigid and forced vibrating roofs under different free-stream turbulence conditions are performed to obtain the aerodynamic forces and flow field data simultaneously. Firstly, the LES simulations of the rigid and vibrating roofs are validated against the reference experiments. And then, characteristics of the distributed pressures on vibrating roofs under free-stream flow turbulence conditions with different turbulence intensities are elucidated. Moreover, the motion-induced forces acting on the vibrating roof under different turbulence conditions are identified and compared, in which the power spectra of wind forces and aerodynamic damping and stiffness are examined. Furthermore, the underlying flow mechanism of the turbulence effects on the motion-induced forces is investigated based on POD analysis of the flow around the vibrating roof. It is found that dual effects exist between the influences of free-stream turbulence and roof motion on the aerodynamic forces, the motion-induced aerodynamic forces on the roof would be greatly decreased with the increase of free-stream turbulence, and the motion-induced effects become less dominant under high free-stream turbulence. The outcomes of this study would facilitate our understanding of the role played by the motion-induced aerodynamic forces and improve the wind-resistant design of long-span roof structures for engineering practices.

Freestream Turbulence Effects on the Aerodynamics of an Oscillating Square Cylinder at the Resonant Frequency

Flow past a bluff body in freestream turbulence can substantially change the flow behaviour compared to that in smooth inflow. This paper presents the study of wake flow and aerodynamics of an oscillating square cylinder at the resonant frequency in freestream turbulence, with the integral length not greater than the cylinder side and the turbulence intensity not greater than 10%. Large eddy simulations (LES) in the Cartesian grid using the Immersed Boundary Method (IBM) technique embedded in a FVM solver, together with an efficient synthetic turbulent inflow generator implemented in an in-house parallel FORTRAN code are used for the study. The results are compared with those for smooth inflow, and relevant data published in the literature. The key findings are: the freestream turbulence conditions evidently reduces the local turbulent scales and fluctuations in the shear layer compared to in smooth flow, as small scale freestream turbulence breaks down cylinder-generated larger scale eddies and weakens them; but does not evidently affect the vortex shedding frequency, or the length of the recirculation region behind the cylinder. This suggests negligible change of drag coefficient compared to in smooth inflow. Moreover, this is because the vortex shedding is dominated by the forced oscillation at the resonance frequency, and the turbulence intensity is small.

Large-Eddy Simulation of the Boundary Layer Development in a Low-Pressure Turbine Cascade With Passive Flow Control

The boundary layer development on a low-pressure turbine blade surface modified by recessed dimples, U-grooves, and rectangular grooves has been investigated through the large-eddy simulations. The simulations are performed at a Reynolds number of 50,000 (based on the inlet velocity and axial chord length) and extremely low freestream turbulence conditions. The characteristic parameters of the boundary layer are used to estimate the development of the boundary layer, and spectral analysis has also been performed to identify the dominant frequency of shedding vortices. The results of simulations indicate that three surface modifications all reduce the profile losses by restraining the separation bubble size. However, the grooves and dimples show different mechanisms in inhibiting laminar separation. Grooves tend to promote the formation of spanwise vortices, which is more difficult to break into turbulence. A high-speed shedding vortex is induced by the particular 3D structure of dimple, and its shedding frequency is nearly twice the Kelvin–Helmholtz (K-H) instability frequency. The interaction between the shedding vortices and the K-H vortices promotes the breakdown process of the spanwise vortices, which leads to an earlier transition of the boundary layer at a low disturbance level. The current study reveals the different mechanisms of dimples and grooves and shows the great potential of dimples for flow control in low-pressure turbines. Besides, the flow structures inside the dimples with adverse pressure gradients are also explored.

Boundary-layer evolution over long wind farms

The structure of the internal boundary layer above long wind farms is investigated experimentally. The transfer of kinetic energy from the region above the farm is dominated by the turbulent flux of momentum together with the displacement of kinetic energy operated by the mean vertical velocity: these two have comparable magnitude along the farm opposite to the infinite-farm case. The integration of the energy equation in the vertical highlighted the key role of the energy flux, and how that is balanced by the growth of the internal boundary layer in terms of energy thickness with a small role of the dissipation. The mean velocity profiles seem to follow a universal structure in terms of velocity deficit, while the Reynolds stress does not follow the same scaling structure. Finally, a spectral analysis along the farm identified the leading dynamics determining the turbulent activity: while behind the first row the signature of the tip vortices is dominant, already after the second row their coherency is lost and a single broadband peak, associated with wake meandering, is present until the end of the farm. The streamwise velocity peak is associated with a nearly constant Strouhal number weakly dependent on the farm layout and free stream turbulence condition. A reasonable agreement of the velocity spectra is observed when the latter are normalised by the velocity variance and integral time scale: nevertheless the spectra show clear anisotropy at the large scales and even the small scales remain anisotropic in the inertial subrange.

Large-Eddy Simulation of Variable Speed Power Turbine Cascade With Inflow Turbulence

Abstract Numerical results are presented from the National Aeronautics and Space Administration (NASA) Glenn Research Center's in-house turbomachinery code Glenn-HT applied to the variable-speed power turbine (VSPT) experiment at the NASA Transonic Turbine Blade Cascade Facility. The main goal of this paper is to implement a digital filtering method to generate turbulence upstream and a subgrid model (localized dynamic k-equation model (LDKM)) in the framework of large-eddy simulation (LES) in order to investigate the effect of inflow turbulence on the transition seen in the VSPT experimental data at the cruise condition (incidence angle of 40 deg and Tu = 0.5%, 5%, 10%, and 15%). Although the boundary layer on the suction side and pressure side of the blades is initially laminar due to favorable pressure gradient, the laminar flow can transition to turbulent flow past a separation zone on the suction side or by natural or bypass transition. This process determines the total pressure losses in the wake. Therefore, it is desirable to develop a reliable prediction tool to accurately capture the transition mechanism in blade rows operated under the conditions of low Reynolds number and at a variety of freestream turbulence conditions. Our numerical studies reveal that the location of separation is rather insensitive to the level of Tu; however, the effect of increasing Tu seems to be in reducing the size and ultimately suppressing the separation bubble. In addition, we performed spectral analysis to identify the peak frequencies in the region where the separation bubble is formed, which provides valuable insights into the transition/separation mechanism.

Pilot workload investigation for rotorcraft operation in low-altitude atmospheric turbulence

This paper aims to investigate pilot workload for rotorcraft operations in low-altitude atmospheric turbulence. A novel low-altitude turbulence model derived through the pre-warped Tustin transformation is used to modify the classic distributed turbulence model. The distributed model is then validated against the von Kármán spectra and is integrated into a flight simulation environment that consists of a high-fidelity nonlinear flight dynamics model and a multi-loop compensatory pilot model. The simulation responses and pilot controls in turbulence have been compared against flight test data for operations in bluff-body and freestream turbulence conditions. The effect of terrain roughness on pilot workload is analyzed. The results show that low-altitude turbulence becomes more intense and has smaller length scales over rougher terrain for the same conditions of the reference wind speed and altitude. Moreover, the effect of terrain roughness can increase pilot workload with increasing airspeed and decreasing flight altitude.

Effect of freestream turbulence on recovery factor of a thermocouple probe and its consequences

The temperature recovery factor of a thermocouple probe is important when being used to measure fluid temperature of a high-speed flow. Typically, the recovery factor is assumed to be dependent on the probe geometry and Prandtl number of the fluid, and invariant with freestream turbulence. In the present study, recovery factor of a butt-joint thermocouple probe is determined for two freestream turbulence conditions (0.25% and 7%). Results show that the recovery factor decreases with an increase in the freestream turbulence, which can lead to significant errors if not taken into consideration. The consequences of not taking into account the effect of freestream turbulence on the probe recovery factor is demonstrated with an application to energy separation in the wake of a circular cylinder.

Experimental Model-Based Estimation and Control of Natural Tollmien–Schlichting Waves

A compensator strategy based on the linearized Navier–Stokes equations, aimed to suppress Tollmien–Schlichting waves naturally occurring in low freestream turbulence conditions, is experimentally demonstrated. Experiments have been conducted on a flat-plate geometry under the influence of an externally imposed adverse pressure gradient. A Kalman filter is used to estimate the effect of upstream disturbances based on pressure information from a single wall-embedded microphone. This information is used by the controller to cancel the incoming perturbations using a surface dielectric barrier discharge plasma actuator. The estimates obtained from the Kalman filter are compared with experimental data obtained using particle image velocimetry. It is shown that the Kalman filter is able to estimate the spatiotemporal behavior of the perturbation field even though it relies on a priori assumptions on the upstream disturbance environment. The performance of the controller and its robustness to varying freestream velocities are analyzed in comparison with open-loop continuous forcing. It is shown that the controller is able to additionally reduce the fluctuating pressure power at the most amplified frequencies with more than one order of magnitude and is capable of reducing the standard deviation of the downstream sensor signal between 30 and 60% for a range of offdesign freestream velocities.

Combined effects of unsteady wake and free-stream turbulence on turbine blade film cooling with laid-back fan-shaped holes using PSP technique

Detailed film cooling effectiveness distribution for a gas turbine blade under the effects of unsteady wakes and oncoming free-stream turbulence intensities was obtained using pressure sensitive paint (PSP) technique. Tests were performed on a linear cascade at Reynolds number of 3.85 × 105 based on the blade chord at cascade exit. Upstream unsteady wakes were simulated using a spoke-wheel type wake generator. The test blade has three rows of compound angled cylindrical film holes at the leading edge, five rows of laid-back fan-shaped holes on the pressure surface and three rows of laid-back fan-shaped holes on the suction surface. The wake Strouhal number was varied from 0 to 0.36 and three mass flux ratios were determined. The oncoming free-stream turbulence intensities are 2.7% and 26.9%, respectively. Results show that the effect of the mass flux ratio on the film cooling effectiveness decreases under the high turbulence intensity and unsteady wake conditions. In most regions of the blade surface, the film cooling effectiveness decreases with the increase of wake Strouhal number, and the free-stream turbulence superimposed on the unsteady wake reduces the film cooling effectiveness further. The effect of the unsteady wake decreases under the high free-stream turbulence conditions.

Simulations of Slot Film-Cooling With Freestream Acceleration and Turbulence

Slot film cooling in an accelerating boundary layer with high freestream turbulence is studied numerically using large eddy simulations (LES). Calculations are done for a symmetrical leading edge geometry with the slot fed by a plenum populated with pin fins. The synthetic eddy method is used to generate different levels of turbulence and length scales at the inflow cross-plane. Calculations are done for a Reynolds number of 250,000 and freestream turbulence levels of 0.7%, 3.5%, 7.8%, and 13.7% to predict both film cooling effectiveness and heat transfer coefficient over the test surface. These conditions correspond to the experimental measurements of (Busche, M. L., Kingery, J. E., and Ames, F. E., 2014, “Slot Film Cooling in an Accelerating Boundary Layer With High Free-Stream Turbulence,” ASME Paper No. GT2014-25360.) Numerical results show good agreement with measurements and show the observed decay of thermal effectiveness and increase of Stanton number with turbulence intensity. Velocity and turbulence exiting the slot are nonuniform laterally due to the presence of pin fins in the plenum feeding the slot which creates a nonuniform surface temperature distribution. No transition to fully turbulent boundary layer is observed throughout the numerical domain. However, freestream turbulence increases wall shear stress downstream driving the velocity profiles toward the turbulent profile and counteracts the laminarizing effects of the favorable pressure gradient. The effective Prandtl number decreases with freestream turbulence. The temperature profiles deviate from the self-similar profile measured under low freestream turbulence condition, reflecting the role of the increased diffusivity in the boundary layer at higher freestream turbulence.

Experimental investigation of the influence of freestream turbulence on an anti-vortex film cooling hole

Film cooling is used in gas turbines to thermally protect combustor and turbine hot section components by creating a layer of relatively cooler air to insulate the components from the hot freestream gases. This relatively cooler air is taken from upstream in the compressor section at a loss to the engine efficiency and therefore must be used as effectively as possible. A novel anti-vortex hole (AVH) geometry has been investigated experimentally through a transient infrared thermography technique to study the film cooling effectiveness and heat transfer coefficient by varying blowing ratio and freestream turbulence intensity. The AVH geometry is designed with two secondary holes stemming from a main cooling hole that attempts to diffuse the coolant jet and mitigate the vorticity produced by conventional straight holes. The AVH geometry data collected showed improved cooling performance at low freestream turbulence intensities compared to conventional straight holes. Three turbulence intensities of 1, 7.5, and 11.7% were investigated for the AVH geometry at blowing ratios of 0.5, 1.0, 1.5, and 2.0 for a total of twelve different conditions. Results showed that higher freestream turbulence conditions were beneficial to the performance of the AVH. Increasing the blowing ratio at all turbulence levels also improved film cooling effectiveness, both span-averaged and on the centerline. The highest performing case was at a turbulence intensity of 7.5% and a blowing ratio of 2.0. The 11.7% cases outperformed the 1% cases, but it appears that at the 11.7% cases the higher freestream turbulence reduces the performance of the secondary holes compared to the 7.5% case. Increasing the blowing ratio and turbulence intensity will result in a higher heat transfer coefficient, which must be taken into account for future designs and overall reduction of heat load.

Recognition of coherent structures in the boundary layer of a low-pressure-turbine blade for different free-stream turbulence intensity levels

Particle Image Velocimetry (PIV) has been adopted to analyze the instantaneous flow field developing on a high-lift turbine blade profile operating under low and elevated free-stream turbulence conditions (FSTI). Results reported in the paper allow us to analyze the dynamics leading to transition and separation of the suction side boundary layer, looking to generation, propagation and breakdown of coherent structures observed in the two different FSTI cases. To this end, measurements have been performed in two orthogonal planes. Results obtained in the blade-to-blade plane allow the detailed characterization of the propagation of Kelvin–Helmholtz (KH) rolls generating, at low FSTI condition, as a consequence of a non-reattaching separation. Otherwise, data in the wall-parallel plane allow recognizing the presence of three-dimensional disuniformities induced at high FSTI by low and high speed streaks (Klebanoff mode). The sinuous breakdown of boundary layer streaks generates other complex three-dimensional coherent structures such as hairpin or cane-like vortices that induce transition. Proper Orthogonal Decomposition (POD) has been adopted to in depth characterize these structures, thus further explaining the mechanisms through which the free-stream turbulence intensity modify the transition/separation processes of the suction side boundary layer of an highly loaded low pressure turbine blade.

Numerical Simulation of Heat Transfer Coefficient on Turbine Blade using Intermittency Factor Equation

This paper deals with the numerical simulation of fluid flow and heat transfer coefficient on a gas turbine blade. The mathematical model of flow is based on the Reynolds Averaged Navier-Stokes (RANS) equations completed with the formulation of transition model by using an intermittency transport equation. An intermittency based transition model, which is critical for invoking transition onset according to Abu-Ghannam and Shaw correlation, is implemented into the proven Reynolds-Averaged N-S (RANS) solver. The intermittent behavior of the transitional flow is incorporated into the computation by modifying the eddy viscosity μt, obtained from a turbulence model. Wilcox [4] low Reynolds k−ω turbulence model is employed to calculate the eddy viscosity and others turbulent quantities. This model is designated for the prediction of flow transition under the influence of freestream turbulence (FST). The current model is applied to prediction of a modern high pressure turbine experiment, and detail comparisons of the computational results with the experimental data are presented. The model has been shown to be capable of predicting the high pressure turbine flow transition under a variety of Reynolds number and free stream turbulence conditions.

Effect of free-stream turbulence on flow characteristics over a transversely-grooved surface

An experimental investigation has been performed to study the free-stream turbulence influence on the flow over a transversely-grooved surface. Free-stream flow at 8.1m/s with turbulence intensities of 0.5% and 4.4% have been provided over smooth and grooved surfaces. Mean velocity profiles, fluctuating velocity moments and turbulence length scales are studied using data obtained from hot-wire anemometry measurements. Introducing quasi-isotropic free-stream turbulence of 4.4% to the flow resulted in fuller velocity profiles, increased boundary layer thickness and significantly augmented streamwise turbulence intensity throughout the boundary layer in flows over both smooth and grooved surfaces. Probability density functions of the velocity fluctuation show that interaction of quasi-isotropic free-stream turbulence with wall turbulence leads to nearly-isotropic turbulence in a small portion of the boundary layer in the near-wall region; while the remaining boundary layer experiences an anisotropic turbulence. In other words, with the introduction of additional turbulence generated by the grooves, the momentum of the fluid particles is increased, resulting in better flow separation resistance. Accordingly, the flow over a grooved surface showed an increase in turbulence intensity, and hence in energy dissipation rate. This is amplified under free-stream turbulence conditions as a result of the interaction of groove-generated turbulence with free-stream turbulence in the boundary layer. The energy cascade span as indicated by the integral to Kolmogorov ratio is also widened by the grooves.

Momentum and Thermal Boundary Layer Development on an Internally Cooled Turbine Vane

Recent advances in computing power have made conjugate heat transfer simulations of turbine components increasingly popular; however, limited experimental data exist with which to evaluate these simulations. The primary parameter used to evaluate simulations is often the external surface temperature distribution, or overall effectiveness. In this paper, the overlying momentum and thermal boundary layers at various streamwise positions around a conducting, internally cooled simulated turbine vane were measured under low (Tu = 0.5%) and high (Tu = 20%) freestream turbulence conditions. Furthermore, experimental results were compared to computational predictions. In regions where a favorable pressure gradient existed, the thermal boundary layer was found to be significantly thicker than the accompanying momentum boundary layer. Elevated freestream turbulence had the effect of thickening the thermal boundary layer much more effectively than the momentum boundary layer over the entire vane. These data are valuable in understanding the conjugate heat transfer effects on the vane as well as serving as a tool for computational code evaluation.

Separation control on high lift low-pressure turbine airfoils using pulsed vortex generator jets

Boundary layer separation control has been studied using vortex generator jets (VGJs) on a very high lift, low-pressure turbine airfoil. Experiments were done under low (0.6%) freestream turbulence conditions on a linear cascade in a low speed wind tunnel. Pressure surveys on the airfoil surface and downstream total pressure loss surveys were documented. Instantaneous velocity profile measurements were acquired in the suction surface boundary layer. Cases were considered at Reynolds numbers (based on the suction surface length and the nominal exit velocity from the cascade) of 25000 and 50000. Jet pulsing frequency and duty cycle were varied. In cases without flow control, the boundary layer separated and did not reattach. With the VGJs, separation control was achieved. At sufficiently high pulsing frequencies, separation control was possible with low jet velocities and 10% duty cycle. At lower frequencies, a 50% duty cycle helped by separating the disturbances associated with the jets turning on and turning off, thereby doubling the frequency of separation control events above the pulsing frequency. Phase averaged velocity profiles and wavelet spectra of the velocity show the VGJ disturbance causes the boundary layer to reattach, but that it can re-separate between disturbances. When the disturbances occur at high enough frequency, the time available for separation is reduced, and the separation bubble remains closed at all times.

Effect of Unsteady Wakes on Boundary Layer Separation on a Very High Lift Low Pressure Turbine Airfoil

Boundary layer separation has been studied on a very high lift, low pressure turbine airfoil in the presence of unsteady wakes. Experiments were done under low (0.6%) and high (4%) freestream turbulence conditions on a linear cascade in a low speed wind tunnel. Wakes were produced from moving rods upstream of the cascade. Flow coefficients were varied from 0.35 to 1.4 and wake spacing was varied from one to two blade spacings, resulting in dimensionless wake passing frequencies F=fLj-te/Uave (f is the frequency, Lj-te is the length of the adverse pressure gradient region on the suction surface of the airfoils, and Uave is the average freestream velocity) ranging from 0.14 to 0.56. Pressure surveys on the airfoil surface and downstream total pressure loss surveys were documented. Instantaneous velocity profile measurements were acquired in the suction surface boundary layer and downstream of the cascade. Cases were considered at Reynolds numbers (based on the suction surface length and the nominal exit velocity from the cascade) of 25,000 and 50,000. In cases without wakes, the boundary layer separated and did not reattach. With wakes, separation was largely suppressed, particularly if the wake passing frequency was sufficiently high. At lower frequencies the boundary layer separated between wakes. Background freestream turbulence had some effect on separation, but its role was secondary to the wake effect.