Amy F. Fagan

Investigation of Broadband Shock Noise from a Jet Near a Planar Surface

Many current and future aircraft designs rely on the wing or other aircraft surfaces to shield observers on the ground from the engine noise. However, the available data showing how surfaces interact with a jet to shield and/or enhance the jet noise are currently limited. Therefore, far-field noise data and background-oriented schlieren images were acquired for a round jet, operating in the overexpanded, ideally expanded, and underexpanded supersonic flow regimes, near a planar surface to investigate how airframe surfaces might affect the shock-cell structure in the jet plume and the broadband shock noise produced. These data show that broadband shock noise is produced by the relatively weak shocks far downstream of the nozzle exit; consequently, a surface will be effective at reducing broadband shock noise only if it is long enough to shield the noise produced by shocks. Furthermore, the presence of a surface very near the edge of an underexpanded jet increases the shock-cell spacing, pushing the shock c...

Background Oriented Schlieren Implementation in a Jet-Surface Interaction Test

Many current and future aircraft designs rely on the wing or other aircraft surfaces to shield the engine noise from observers on the ground. However the available data regarding how a planar surface interacts with a jet to shield and/or enhance the jet noise are currently limited. Therefore, the Jet-Surface Interaction Tests supported by NASAs Fundamental Aeronautics Programs Fixed Wing Project were undertaken to supply experimental data covering a wide range of surface geometries and positions interacting with high-speed jet flows in order to support the development of noise prediction methods. Phase 1 of the Test was conducted in the Aero-Acoustic Propulsion Laboratory at NASA Glenn Research Center and consisted of validating noise prediction schemes for a round nozzle interacting with a planar surface. Phased array data and far-field acoustic data were collected for both the shielded and reflected sides of the surface. Phase 1 results showed that the broadband shock noise was greatly reduced by the surface when the jet was operated at the over-expanded condition, however, it was unclear whether this reduction was due a change in the shock cell structure by the surface. In the present study, Background Oriented Schlieren is implemented in Phase 2 of the Jet-Surface Interaction Tests to investigate whether the planar surface interacts with the high-speed jet ow to change the shock cell structure. Background Oriented Schlieren data are acquired for under-expanded, ideally-expanded, and over-expanded ow regimes for multiple axial and radial positions of the surface at three different plate lengths. These data are analyzed with far-field noise measurements to relate the shock cell structure to the broadband shock noise produced by a jet near a surface.

Resonant Interaction of a Rectangular Jet with a Flat-Plate

A resonant interaction between a large aspect ratio rectangular jet and a flat-plate is addressed in this experimental study. The plate is placed parallel to but away from the direct path of the jet. At high subsonic conditions and for certain relative locations of the plate, the resonance accompanied by an audible tone is encountered. The trends of the tone frequency variation exhibit some similarities to, but also marked differences from, corresponding trends of the well-known edgetone phenomenon. Under the resonant condition flow visualization indicates a periodic flapping motion of the jet column. Phase-averaged Mach number data obtained near the plate’s trailing edge illustrate that the jet cross-section goes through large contortions within the period of the tone. Farther downstream a clear ‘axis switching’ takes place. These results suggest that the assumption of two-dimensionality should be viewed with caution in any analysis of the flow.

Noise from a Jet Discharged into a Duct and its Suppression

This study addresses unwanted high intensity noise sometimes encountered in engine test facilities. Model-scale experiments are conducted for a round jet discharged into a cylindrical duct. In most cases, the unwanted noise is found to be due to longitudinal resonance modes of the duct excited by the random turbulence of the jet. When the ‘preferred mode’ frequency of the jet matches a duct resonant frequency there can be a locked-in ‘super-resonance’ accompanied by a high intensity tone or ‘howl’. Various techniques are explored for suppression of the unwanted noise. Tabs placed on the ends of the duct are found ineffective; so are longitudinal fins placed inside the duct. Arod inserted perpendicular to the flow (‘howl stick’) is also found generally ineffective; however, it is effective when there is a super-resonance. By far the most effective suppression is achieved by a wire-mesh screen placed at the end of the duct. The screen not only eliminates the super-resonance but also the duct mode spectral peaks. Apparently the screen works by dampening the velocity fluctuations at the pressure node and thereby weakening the resonant condition.

Investigating the Feedback Path in a Jet-Surface Resonant Interaction

Abstract A resonant interaction between an 8:1 aspect ratio rectangular jet and flat-plates, placed parallel to the jet, is studied experimentally. For certain locations of the plate relative to the jet, the resonance takes place with a loud accompanying tone. The sound pressure level spectra are often marked by multiple peaks. The frequencies of the spectral peaks are studied as a function of the streamwise length of the plate, its relative location to the jet as well as the jet Mach number. It is demonstrated that the tones are not due to a simple feedback between the plate’s trailing edge and the nozzle’s exit; the leading edge of the plate also comes into play in the frequency selection. With parametric variation, it is found that there is an order in the most energetic spectral peaks; their frequencies cluster in distinct bands. The ‘fundamental’, i.e., the lowest frequency band is explained by an acoustic feedback involving diffraction at the plate’s leading edge. 1. Introduction

Application of a Novel Projection Focusing Schlieren System in NASA Test Facilities

A portable projection focusing schlieren system for flow visualization has been designed and built by Metrolaser, Inc. with support from NASA’s Small Business Innovative Research Program. The system is based on a lens and grid technique where a source grid is projected onto a retro-reflective screen and imaged onto a cut-off grid creating the schlieren effect. The light source is a 5 Watt Xenon flashlamp with a 1-s pulse duration, which enables instantaneous snapshots of the flow field. Density gradients or refractive index gradients in the flow cause the light rays to bend or refract. The cut-off grid, which is analogous to the “knife-edge” in a classical schlieren system, causes these disturbances to be seen as light or dark regions in the image of the flow that is captured on a camera sensor located somewhere behind the cut-off grid. Two system designs are presented: one that relays the image of the object to the CCD sensor by way of relay lenses, enabling very large fields of view but at the expense of image quality, and another that directly images the object on the CCD sensor by way of the same lens that images the projected grid onto the cut-off grid, which results in a limited field-of-view but higher quality images. A discussion of the two optical systems and images from various NASA experiments demonstrating the system capabilities are presented.

Background Oriented Schlieren Applied to Study Shock Spacing in a Screeching Circular Jet

Background oriented schlieren (BOS) is a recent development of the schlieren and shadowgraph methods. The BOS technique has the ability to provide visualizations of the density gradient in both the axial and radial directions. The resultant magnitude of the density gradients allows for comparison with shadowgraph images. This paper first compares data obtained by the BOS and shadowgraph techniques at identical conditions in a free jet. The patterns and spacing of the shock trains obtained by the two techniques are found to be consistent with one another. This provides confidence in the shock spacing measurement by the BOS technique. Due to its simpler setup, BOS is then applied to investigate the shock spacing associated with the screech phenomenon, especially during stage jumps. Screech frequencies from a 37.6 mm convergent nozzle, as a function of jet Mach number (M(sub j)), are shown to exhibit various stages. As many as eight stages are identified with the present nozzle over the range 1.0 < M(sub j) <1.7. BOS images are acquired at various screech conditions and the shock spacing is examined as a function of M(sub j).

Experimental Study of Tip Vortex Flow from a Periodically Pitched Airfoil Section

An experimental investigation of a tip vortex from a NACA0012 airfoil is conducted in a low-speed wind tunnel at a chord Reynolds number of 4x10(exp 4). Initially, data for a stationary airfoil held at various angles-of-attack (alpha) are gathered. Detailed surveys are done for two cases: alpha=10 deg with attached flow and alpha=25 deg with massive flow separation on the upper surface. Distributions of various properties are obtained using hot-wire anemometry. Data include mean velocity, streamwise vorticity and turbulent stresses at various streamwise locations. For all cases, the vortex core is seen to involve a mean velocity deficit. The deficit apparently traces to the airfoil wake, part of which gets wrapped by the tip vortex. At small alpha, the vortex is laminar within the measurement domain. The strength of the vortex increases with increasing alpha but undergoes a sudden drop around alpha (is) greater than 16 deg. The drop in peak vorticity level is accompanied by transition and a sharp rise in turbulence within the core. Data are also acquired with the airfoil pitched sinusoidally. All oscillation cases pertain to a mean alpha=15 deg while the amplitude and frequency are varied. An example of phase-averaged data for an amplitude of +/-10 deg and a reduced frequency of k=0.2 is discussed. All results are compared with available data from the literature shedding further light on the complex dynamics of the tip vortex.

Improvement in Rayleigh Scattering Measurement Accuracy

(Abstract) Spectroscopic Rayleigh scattering is an established flow diagnostic that has the ability to provide simultaneous velocity, density, and temperature measurements. The Fabry-Perot interferometer or etalon is a commonly employed instrument for resolving the spectrum of molecular Rayleigh scattered light for the purpose of evaluating these flow properties. This paper investigates the use of an acousto-optic frequency shifting device to improve measurement accuracy in Rayleigh scattering experiments at the NASA Glenn Research Center. The frequency shifting device is used as a means of shifting the incident or reference laser frequency by 1100 MHz to avoid overlap of the Rayleigh and reference signal peaks in the interference pattern used to obtain the velocity, density, and temperature measurements, and also to calibrate the free spectral range of the Fabry-Perot etalon. The measurement accuracy improvement is evaluated by comparison of Rayleigh scattering measurements acquired with and without shifting of the reference signal frequency in a 10 mm diameter subsonic nozzle flow.

An Experimental Investigation of Jet Noise from Septa(e) Nozzles

Results of an experimental study with a large aspect ratio rectangular nozzle, divided into multiple compartments or septa, as pertinent to distributed propulsion, are presented. Noise measurements at high-subsonic conditions show that the nozzle with the septa is quieter than the corresponding baseline nozzle without the septa. At relatively lower Mach numbers a high-frequency tone is heard. This is shown to be due to Karmann vortex shedding from the trailing edge of the partitions that separate a septum from the adjacent ones. Flowfield measurements for a six septa case show that the cellular flow structure, issuing from the nozzle, goes through a curious coalescence with increasing downstream distance (x) from the nozzle. Adjacent cells pair to yield a three-cell structure by x/D =2, where D is the equivalent diameter of the baseline nozzle. By about x/D =16, both the septa case and the baseline case evolve to yield axisymmetric flowfields.