Alan B. Cain

AERO-OPTICAL CHARACTERISTICS OF COMPRESSIBLE, SUBSONIC TURBULENT BOUNDARY LAYERS

An extensive experimental study of optical aberrations due to propagation through fully-developed turbulent boundary layers at high subsonic Mach numbers was performed. Time-resolved, high- bandwidth, direct optical measurements of the dynamic aberrations were made using a Malley probe. The probe was used to obtain the convective speeds of the optically-significant turbulence structures and to measure the optical path differences. Measurements were made over a range of boundary layer thicknesses and Mach numbers. Optical distortions were found to scale linearly with boundary layer thickness and freestream density, and to go as the square of the freestream Mach number.

The Optical Environment of a Cylindrical Turret with a Flat Window and the Impact of Passive Control Devices

Optical aberrations over a cylindrical turret with a flat window were measured using a 2-D wavefront sensor and a Malley probe as a function of laser beam elevation angle. It was found that depending on the window back-facing angle the flow either had a weak separation bubble followed by a reattached boundary layer or a strong separation with a large recirculation region behind the cylindrical turret. It was found that optical aberrations were high in the case of a weak separation bubble and at large look-back elevation angles. Different passive devices placed upstream from the turret were studied for their effectiveness in improving the optical-propagation environment. The results are presented and discussed.

Aero-Optical Environment Around a Cylindrical Turret with a Flat Window

Optical aberrations over a cylindrical turret with a flat window were measured using a two-dimensional wavefront sensor and a Malley probe, as a function of laser beam elevation angle. Topology of the flow around the cylinder was extensively studied using hot wires. It was found that, depending on the window elevation angle, the flow either had a weak separation bubble followed by a reattached boundary layer or a strong separation with a large recirculation region behind the cylindrical turret. Optical aberrations were shown to be consistent with flow topology. It was found that optical aberrations were high around an elevation angle of 100 deg and at large lookback elevation angles.

Similarities and Differences in Aero-Optical Structure over Cylindrical and Hemispherical Turrets with a Flat Window

This paper discusses similarities and differences for aft-looking angles of the aerooptical environment over flat windows on a three-dimensional hemispherical turret and two-dimensional cylindrical turret. Both fluid-mechanic and aero-optical data for the baseline flows over the turrets and with two types of passive flow devices are presented and discussed for relatively-high, subsonic Mach numbers. Although the flow around the hemispherical turret was found to be far-more complex than over the cylindrical turret, the flow physics over the windows themselves were shown to be very similar; this similarity afforded an opportunity to study the flow over the window in more detail for the cylindrical turret and to easily examine a range of flow devices for improving its aero-optical environment, selecting two devices, presented here, for further study on the hemispherical turret. For both turret configurations the most aberrating flow environments were cased by separated flows. For small aft-looking angles the separated flows reattach over the windows and the devices were shown to be effective in preventing the separation altogether and greatly improving the aero-optical environment. Once fully separated, neither device improved the environment.

Jet Impingement Tone Suppression Using Powered Resonance Tubes

This work is an experimental study of high subsonic jet impingement tone suppression. We begin by documenting the characteristics of the impingement tone for various Mach numbers and standoff (nozzle exit to ground plate) distances. The results revealed frequency staging and the presence of two types of impingement tones. A novel feature of our work is the use of four miniature high-frequency actuators known as powered resonance tubes that were located circumferentially around the main jet nozzle. The powered resonance tubes were capable of producing high amplitude acoustic excitation over a range of frequencies, up to 17.5 kHz. Our target excitation frequency range was about 3-5 times that of the natural flow instability. Using high-frequency excitation, tonal suppression levels as high as 20 dB and broadband suppression levels as high as 5-10 dB were obtained. The mass addition rate from the powered resonance tubes was of the order of 2% of the mass flow rate from the main jet. Mass flow reductions could be obtained under conditions when the powered resonance tube resonated strongly. Our results suggest that appropriately designed miniature powered resonance tube actuators have potential for use in flow control applications.

Characterization of High-Frequency Excitation of a Wake by Simulation

Insights into the effects of high-frequency forcing on free shear layer evolution are gained through analysis of several direct numerical simulations. High-frequency forcing of a fully turbulent plane wake results in only a weak transient effect. On the other hand, significant changes in the developed turbulent state may result when high-frequency forcing is applied to a transitional wake. The impacts of varying the characteristics of the high-frequency forcing are examined, particularly, the streamwise wavenumber band in which forcing is applied and the initial amplitude of the forcing. The high-frequency excitation is found to increase the dissipation rate of turbulent kinetic energy, to reduce the turbulent kinetic energy production rate, and to reduce the turbulent kinetic energy suppression increases with forcing amplitude once a threshold level has been reached. For a given initial forcing energy, the largest reduction in turbulent kinetic energy density was achieved by forcing wavenumbers that are about two to three times the neutral wavenumber determined from linear stability theory.

Aero-Optical Measurements of High-Mach Supersonic Boundary Layers

Aero-optical measurements of a boundary layer at high supersonic speeds of M = 3.0 and 4.3 were performed in the Trisonic Wind Tunnel at the US Air Force Academy. Overall levels of aero-optical distortions, convective speeds and the aperture functions were calculated from the data. It was found that the statistics of the aero-optical distortions were similar to subsonic boundary layers. Several modified models were shown to properly predict aerooptical levels of boundary layers up to M = 5. As a separate experiment, a wedge model was placed inside the tunnel to study the aero-optical effects of the attached oblique shock. Aerooptical structures with an abnormally-low convective speed were observed in the laser beam traversing the oblique shock around the wedge. These structures were speculated to be related to convecting unsteady distortions imposed on the oblique shock by the naturallyvibrating wedge.

Modeling supersonic jet screech - Differential entrainment and amplitude effects

A research program for predicting supersonic jet screech is motivated by the need for reduced military aircraft maintenance cost and acoustic signature. An interactive computer design tool with a turn-around on the order of seconds is required for screening proposed propulsion exhaust systems for the occurrence of screech. Simple models of the elemental processes of screech and a preliminary interactive screech analysis tool have been developed. The previous results showed that increasing flight speed increases the number of resonant screech frequencies, tending to an infinite number of such resonant peaks as sonic flight is approached. In the present work a differential shear layer spreading model is developed based on conservation of mass and momentum. The effect of differential spreading is shown to change the number of shock cells by more than a factor of two in some cases, thus producing a major impact on screech predictions. The differential spreading also shows promise in explaining temperature effects that were previously unexplainable. Finally, a process for the prediction of screech amplitude is described and results show remarkable agreement in the preliminary comparison with experimental data.

Mechanism of Vorticity Generation in Plasma Streamwise Vortex Generators

An experimental investigation into the mechanism of streamwise vorticity generation in an array of plasma streamwise vortex generators is presented. The array is flush mounted to a flat plate on which a nominally zero pressure gradient turbulent boundary layer develops upstream. The investigation is focused on characterization of the influence of freestream velocity, applied peak-to-peak voltage, length of the active electrode, and spanwise interelectrode spacing on streamwise vorticity generation. It is shown that the actuator creates wall-normal vorticity and reorients it into the streamwise direction. In addition, spanwise boundary-layer vorticity is reoriented into the streamwise direction. Scaling relations based on the vorticity transport equation are obtained and experimentally validated. These provide guidance for optimizing the actuators for particular flow control applications.

Management of Wind Tunnel Performance Data Using Neural Networks

This paper describes the use of neural networks to organize and manipulate wind-tunnel performance data into formats that are useful to tunnel operators. The neural networks were trained and tested using a mathematical model that was developed from basic fluidmechanic equations, and calibrated to accurately simulate the behavior of a subsonic, closedcircuit wind tunnel. The ability of the trained neural networks to extract secondary results such as the model drag area from a database composed of typically-measured wind-tunnel control and status parameters, is illustrated. Methods to identify the parameters required in the neural networks are also discussed.