Stewart A. L. Glegg

Jet engine noise source location: The polar correlation technique

The objective of this paper is to introduce a new method of source location for use in aero-engine noise research. The technique, termed “polar correlation”, employs an array of far field microphones which are normally located on a polar arc centred on the jet nozzle. It is demonstrated that measurements of the cross power spectral density of signals from such an array can be employed to obtain both the location and relative strengths of an axially distributed array of sound sources, for any frequency of interest. For the modern generation of high by-pass ratio engine this might include a measurement of the relative contribution to the far field noise emanating from the fan inlet, by-pass duct and hot core exhaust as well as indicating that portion of the jet flow which is the principal contributor to the frequency considered. The potential of this method is illustrated by examples of test results for full size jet engines and model jets. Certain fundamental and philosophical difficulties involved in the general use of this and other far field source location methods are also reviewed is some detail.

Broadband Self Noise from a Ducted Fan

This paper describes the prediction of broadband self noise from ducted fans. The source mechanism is assumed to be the interaction of the turbulent boundary layer with the trailing edges of the fan blades. The source levels are obtained from the measurements of self noise from isolated blades by Brooks, Pope and Marcolini and these are corrected to give the in duct sound power from a high solidity fan. It has been found that the blade surface pressures are not uncorrelated on each fan blade and corrections must be included for the scattering from multiple trailing edges. A method is introduced for coupling the modes in a circular duct to the modes of a linear cascade so that the sound power is not singular at the mode cut on frequencies. Results show that the in duct sound power scales with the fifth power of the fan speed at low Mach numbers, but this changes to the sixth power or greater at high Mach numbers. The angle of attack of the blade increases the self noise as 2.4 dB per degree and there are significant increases in low frequency noise when blade stall occurs.

The prediction of broadband noise from wind turbines

Abstract This paper describes a broadband noise prediction scheme for wind turbines. The source mechanisms included in the method are unsteady lift noise, unsteady thickness noise, trailing edge noise and the noise from separated flow. Special methods have been developed to model the inflow turbulence from the atmospheric boundary layer and acoustic radiation to the geometric near field of the rotor. Predictions are compared with measurements on 20 m and 80 m diameter wind turbines. The results show that the turbulence length scale in the atmospheric boundary layer is too large to give the measured noise levels. Very good agreement is obtained between predictions and measurements if the turbulence length scale is taken to be equal to the blade chord.

The far-field sound from rough-wall boundary layers

The noise radiated by a turbulent boundary layer over a rough wall is shown to be characterized by a dipole surface source that, if the surface pressure is spatially homogeneous, can be specified by a convolution integral combining the surface pressure wavenumber spectrum and the wavenumber spectrum of the surface roughness slope. For random roughness elements with almost vertical sides, the surface slope has a wavenumber white spectrum and the radiated sound is directly proportional to the surface pressure spectrum multiplied by an acoustic efficiency factor (koh)2, where ko is the acoustic wavenumber and h is the geometrical r.m.s. roughness height. The theoretical result agrees with the roughness noise measurements of Smith et al. (Smith et al. 2008 AIAA paper no. 2008-2904) for all flow speeds and roughness element heights (in wall units) of kg+<34. The theory can also be applied to wavy walls and this provides a new method for measuring the wavenumber spectrum of surface pressure fluctuations.

Two-point descriptions of wake turbulence with application to noise prediction

Measurements of the four-dimensional two-point correlation tensor of a fully developed airfoil wake are presented. These data allow examination of the characteristic eddies of the turbulence through proper orthogonal decomposition and linear stochastic estimation. A simple generic technique has been developed to extrapolate the correlation tensor function from the Reynolds stress field based on the hypothesis that its form is largely determined by the constraints imposed by inhomogeneity and continuity. Estimates for the plane wake compare favorably with measurements, and, interestingly, proper orthogonal modes and characteristic eddy structures inferred from the estimates are similar to those obtained from measurements. This implies that these measures of the correlation function are largely determined by fairly basic physical information, suggesting some simplification of the turbulence-modeling problem for applications such as aeroacoustics.

Perpendicular blade vortex interaction

The interaction between a streamwise vortex and a spanwise blade was studied in incompressible flow for blade-vortex separations between ±1/8 chord. Three-component velocity and turbulence measurements were made from 4 chord lengths upstream to 15 chord lengths downstream of the blade using miniature four-sensor hot-wire probes. The interaction of the vortex with the blade causes an almost immediate loss in vortex core circulation. Downstream of the blade the core becomes embedded in the blade wake and begins to grow rapidly. Core radius increases and peak tangential velocity decreases but circulation remains roughly constant. True turbulence levels within the core are much larger downstream than upstream of the blade. Outside of the core the interaction produces a substantial region of turbulent flow associated with the blade and vortex generator wakes. Overall, perpendicular blade vortex interaction substantially alters the flow and produces a much larger and more intense region of turbulent flow than that presented by the undisturbed vortex. These results have significant implications for the prediction of both impulsive and broadband helicopter noise.

Effect of centerbody scattering on propeller noise

This paper describes how the effect of acoustic scattering from the hub or centerbody of a propeller will affect the far-field noise levels. A simple correction to Gutins formula for steady loading noise is given. This is a maximum for the lower harmonics but has a negligible effect on the higher frequency components that are important subjectively. The case of a blade vortex interaction is also considered, and centerbody scattering is shown to have a significant effect on the acoustic far field.

Comparison between theory and model scale measurements of three‐dimensional sound propagation in a shear supporting penetrable wedge

Laboratory experiments have been carried out to investigate underwater sound propagation over a sloping bottom which can support shear. The model of the seafloor consisted of a sediment layer (modeled experimentally by epoxy) over a hard substrate made of concrete. The sound field was measured in both the across slope and downslope directions for various frequencies and bottom slopes. The experimental results have been compared with a new source image theory for sound propagation in an ocean wedge with an elastic (shear supporting) bottom and have shown very good agreement for both downslope and across slope propagation. In particular the effect of shear waves on across slope propagation has been demonstrated both theoretically and experimentally. Further, the effect of a sediment layer overlying a semi‐infinite substrate has been successfully incorporated into the source image theory for comparison with the experimental measurements.

Noise from a Rotor Ingesting a Planar Turbulent Boundary Layer

This study concerns rotor noise generated by the ingestion of anisotropic, inhomogenous turbulence produced by a planar turbulent boundary layer. Far field noise was recorded at multiple receiving angles and blade wake profiles were measured for advance ratios ranging J=0.48 to 1.44. A novel method to infer the turbulence characteristics from the measured far field noise is presented. The time-frequency distribution of the recorded noise is used to estimate the streamwise scale of the turbulence while the time-averaged spectra are used to estimate the lateral scale. Results show that the estimated streamwise scale doubles over the measured range of advance ratios and the lateral scale remains approximately constant. This suggests a complex distortion of the turbulence due to the sheared flow and presence of the wall, a topic addressed in a companion paper (Glegg et al., 2013).

Determination of noise source heights, part II: Measurement of the equivalent source height of highway vehicles

Abstract This is the second of two companion papers which describe the measurement of the equivalent point source height on highway vehicles. The first paper describes the measurement method, and the second its application to highway vehicles. This paper discusses the measurements on moving highway vehicles. First it is shown how the measurement method is suitable for measuring the equivalent point source height for application to noise barrier design. The measurement of noise source height on 36 heavy trucks, 37 medium vehicles and 25 small vehicles are then described. The results give the equivalent point source height as a function of frequency for each type of vehicle and identify the contributions of engine noise and tire noise. The measured source height was found to be ∼ 1·2 m for heavy vehicles, ∼0·7 m for medium vehicles and ∼0·6 m for small vehicles. These heights are significantly different from those currently used to design noise barriers, and suggest that further work should be undertaken to assess the implications of these results.