Acoustic Source Characterization Research Articles

Wideband RELAX and wideband CLEAN for aeroacoustic imaging.

Microphone arrays can be used for acoustic source localization and characterization in wind tunnel testing. In this paper, the wideband RELAX (WB-RELAX) and the wideband CLEAN (WB-CLEAN) algorithms are presented for aeroacoustic imaging using an acoustic array. WB-RELAX is a parametric approach that can be used efficiently for point source imaging without the sidelobe problems suffered by the delay-and-sum beamforming approaches. WB-CLEAN does not have sidelobe problems either, but it behaves more like a nonparametric approach and can be used for both point source and distributed source imaging. Moreover, neither of the algorithms suffers from the severe performance degradations encountered by the adaptive beamforming methods when the number of snapshots is small and/or the sources are highly correlated or coherent with each other. A two-step optimization procedure is used to implement the WB-RELAX and WB-CLEAN algorithms efficiently. The performance of WB-RELAX and WB-CLEAN is demonstrated by applying them to measured data obtained at the NASA Langley Quiet Flow Facility using a small aperture directional array (SADA). Somewhat surprisingly, using these approaches, not only were the parameters of the dominant source accurately determined, but a highly correlated multipath of the dominant source was also discovered.

Estimation of acoustic source strength within a cylindrical duct by inverse methods

One method for deducing the strength of an acoustic source distribution from measurements of the radiated field involves the inversion of the matrix of frequency response functions relating the field measurement points to the strengths of a number of point sources used to represent the source distribution. This method uses the singular value decomposition (SVD) as the primary analysis tool and has shown some promise. In the transformation process associated with the SVD, the strength of the discretized point source can be simply represented by the pressure field and the inversion of corresponding matrix of the frequency response function and the transformed pressures and source distribution are related by single real numbers only (i.e., the singular values). Therefore, the resolution and the accuracy of the reconstruction produced by such inversion methods will be highly dependent on how these small singular values are treated during the inversion process. In the present work, the emphasis is placed on the description of the inverse method for the investigation of the characteristics of acoustic sources within a circular duct. Firstly, an analytical model is developed of the internal field from a finite duct with open ends. An examination is presented of the capability of the inverse method in dealing with in-duct sound source problems. These guide the design of an optimal sensor array which can make the condition number as small as possible for the assumed source distribution. Also, a series of experimental demonstrations are given of the proposed in-duct sensor array for producing high-resolution information regarding in-duct sources.

On one-port characterization of noise sources in ducts by using external loads

Equivalent acoustic source characterization of duct-borne fluid machinery noise is often undertaken by interpolating the results of two-microphone pressure measurements with different external acoustic loads over a linear one-port source model. If the source is time-invariant, the one-port source characteristics can be determined by using only two external loads. This is well known as the two-load method. An extension of the two-load method for time-variant sources is also available and known as the multiple-load method. In these methods the source is treated as a ‘black-box’. This paper addresses the problem of one-port source characterization when the linear operations inherent in the ‘black-box’ are known explicitly. The equations governing the explicit one-port source models are derived and the source characteristics are shown to be measurable using only few acoustic loads. It is not the purpose of this paper to discuss the application of these models to any specific fluid machinery; however, of particular interest are the explicit source models that require only two loads. Numerical results are presented to show some features of such time-invariant and time-variant explicit one-port source models.

Implementation issues in the development of a real-time, Windows-based system to study spatial hearing

As a tool to study spatial hearing, a customizable real-time 3-D-audio and virtual acoustic environment rendering system, Sound Laboratory (SLab), was developed. Design goals for this system included flexibility, extensibility, maintainability, and a fast development cycle per experiment. SLab is a Windows application composed of five distinct software layers: Application programming interface (API), 3D projection, signal flow translation, signal routing and processing, and digital signal processing library. The acoustic scene, including listener, source, and environment characteristics, is specified in the API layer and translated in the 3-D projection layer to geometric quantities, such as range and arrival angle, for each path rendered between the source and listener. The geometry and head-related transfer function coefficients are rendered using a set of parallel finite impulse response filters and delay lines built on Intel’s Signal Processing Library. This rendering architecture provides the flexibility and the extensibility required for psychoacoustic experimentation, while the modular nature of the programming interface ensures maintainability and quick development of new experiments. The psychoacoustic parameters to be manipulated, the signal flow diagram, and the five software layers are discussed in detail, and various implementation issues are examined.

A refined four-load method for the evaluation of in-duct acoustic source characteristics

By using the four-load method, one-port source characteristics can be determined in the duct system. In this study, a modification is made to avoid the instability problems occurring in the conventional four-load method. In the present method, the acoustic power spectra are measured at the inside or outside of the duct varying the acoustic loads, and an error function based on the linear time-invariant source model is obtained. It is numerically shown that the method is less sensitive to input errors compared to the conventional four-load method. The two methods are utilized to obtain the source parameters of a loudspeaker and a centrifugal fan using various types of loads. The estimated impedance values are compared with those obtained from the direct measurement method. As a direct measurement method, the multiple sensor method is adopted using three microphones. It is observed that the conventional four-load method results in large errors, whereas the present method yields far better agreement with the actual source characteristics.

Temporal and spatial distribution of whale calls off Monterey, California

Twenty-seven hours of acoustic data were recorded from a horizontal array in deep waters off the central coast of California. These data were analyzed to determine the characteristics of diverse underwater acoustic sources. Some of the identified sources were: moored RAFOS sources at ranges of 150–1000 km, low-frequency ship and machinery noise, and sounds of biological origin. Over 400 whale calls were identified and analyzed to determine the distribution of these calls in both time and azimuth. Spectral analysis of the vocalizations indicate that most of the calls were from humpback whales. [Work supported by the United States Navy, Naval Postgraduate School, and Monterey Bay Aquarium Research Institute.]

A day in the life of a SOSUS array

Twenty seven hours of acoustic data from the Point Sur SOSUS array were recorded and analyzed to determine the characteristics of diverse underwater acoustic sources. The SNR and variability of moored RAFOS sources at ranges of 150–1000 km were estimated, as was the source level of imploding incandescent light bulbs. The measured and modeled transmission loss from the bulbs were compared out to the first convergence zone. Finally, the distribution of whale calls in both time and azimuth was determined. These projects are meant to be illustrative of potential uses of Navy array data for both scientific research and graduate education. [Work supported by the United States Navy, Naval Postgraduate School, and Monterey Bay Aquarium Research Institute.]

Finite element modelling of sheared flow effects on the radiation characteristics of acoustic sources in a circular duct

AbstractThe recent interest in propeller noise generation, stimulated by development of new propeller types for commerical propjets, has generated a need for the ability of measure the noise characteristics of propellers. However, wind tunnel noise measurements are affected by reflections from the wind tunnel walls. Computer codes predicting the free‐field noise of a propeller and its noise field in a circular wind tunnel allow validating the use of wind tunnel measurements to predict free‐field noise characteristics. A wind tunnel contains flow which is uniform in the duct axial direction, but can vary in the radial direction. It can be shown that a third‐order differential equation governs the acoustic pressure field for such a duct containing radially sheared subsonic flow. This third‐order problem is then posed as a coupled pair of equations which are second‐order in terms of acoustic density and first‐order in terms of an artificial variable which represents the effects of the flow being sheared. It is shown that this form of the problem allows a natural extension of the existing numerical solution techniques for non‐sheared flow. The sheared flow problem is presented, and a finite element method is developed to yield a solution for propeller‐type acoustic forces. The finite element code and method are refined with numerical experiments, and results are presented for a specific propeller and duct geometry. Good agreement is shown between this method and an alternate approach to the sheared flow problem using a piecewise constant representation of the velocity in the boundary layer. This validates both the numerical methods.

Characteristics of acoustic emission source in a fatigue crack: Nondestructive Testing Communications, Vol. 4, No. 1, pp. 11–21 (1988)

A model of AE sources in a cracked solid under cyclic loading is proposed. The cleavage step created locally on a cracked surface, referred to as the frictional source, provides a source in addition to the fracture event at the crack tip, i.e. the cracking source. Using this model, the AE activity from a growing crack is explained and characterized. The frictional and cracking sources can be distinguished from one another by the rise-time or slope of the AE signal waveform.

On numerical prediction of the acoustic source characteristics of an engine exhaust system

The contemporary methods for source characterization rely mainly on experiments. These methods produce inaccurate results in the low-frequency band, where the characteristics are all the more important. Moreover, the experimental methods cannot be used at the design stage. Hence, a numerical technique to obtain the source characteristics is desirable. In this paper, the pressure-time history and the mass-flux-time history obtained by means of the time-domain analysis have been used, along with the two-load method to compute the source characteristics. Two new computational methods for obtaining the source characteristics have been described. These are much simpler, and computationally more economical than the complete time-domain simulation, which makes use of the method of characteristics.

Perception of acoustic source characteristics: Walking sounds

The current study investigated the ability of subjects to perceive source characteristics (the gender of a human walker) of a naturally occurring auditory event (human walking). A number of acoustic properties were measured and subjected to statistical analyses in order to identify those properties that differentiate male and female walking footsteps. A principal component analysis on the statistical properties of the spectral energy distributions then identified two classes of information that were important in determining subject perception of the gender of a walker: (1) the spectral peak which integrates the information about the spectral central tendency of frequency and shape of the spectral peak; and (2) the contribution of high-frequency spectral components. A follow-up experiment then manipulated these spectral properties to verify their contributions in the perceptual classification of walker gender. Additionally, the effect of shoe on the gender judgement in walking sequences was assessed by having both male and female walkers wear male's shoes.

Acoustic Measurements of Men’s and Women’s Voices

Several acoustic measures of laryngeal activity were made on adult speech to help answer two questions left unresolved by previous work: (1) how each measure varies, if at all, with phonetic structure, and (2) what aspect of laryngeal activity each measure specifies. Speech samples of 15 syllables (three vowels in five prevocalic consonantal contexts) were collected from men and women at two times of the day (early morning and late afternoon). Eight measurements were made, mainly on slices extracted from the middle of the vocalic portions, and inferential and correlational statistics were applied to these measures. Results of the inferential tests indicated differences between men and women in how laryngeal adjustments are made, affecting relative amounts of vocal jitter and spectral tilt of the voicing source. In addition, the voicing and manner characteristics of the prevocalic consonant were found to affect fundamental frequency, cycle-to-cycle perturbations, and amount of aspiration noise. To a lesser extent, vowel height and front/back tongue placement also affected these acoustic source characteristics. Results of the correlational tests showed that different laryngeal mechanisms contributed differentially to signal-to-noise ratios for men and women, and these mechanisms were more greatly affected by fundamental frequency for men's samples. Finally, various acoustic measures of laryngeal noise were found to be related to the same underlying mechanism.

Characteristics of acoustic emission source in a fatigue crack : Mori Y.; Obata Y. Nondestructive testing communications, vol. 4, no. 1, pp. 11–21 (1988)

Characteristics of acoustic emission source in a fatigue crack : Mori Y.; Obata Y. Nondestructive testing communications, vol. 4, no. 1, pp. 11–21 (1988)

Analyse expérimentale des phénomènes d'interaction d'une machine et de son circuit

The interaction of a machine with its circuit may generate excessive vibrations of the components. Current studies show the experimental means and the analysis methods of the results of the experiments. The emphasis is placed on the validation of models and on the localization and characterization of acoustic sources. The results submitted are then related to laboratory tests and semi-industrial cases. The problems encounted when analysing these cases are emphasized.

Characterization of structure-borne sound sources: The source descriptor and the coupling function

An introductory, theoretical study for characterization of structural acoustic sources has been carried out. The concept source denotes any active structure with unknown internal elementary source mechanisms and unknown internal transmission paths from generation to the contact points with the receiving structure. An alternative description of the power transmission from the source to the receiver is presented in the case of one component of motion and one contact point. A general source parameter called the source descriptor is introduced. This parameter is a function of source data only. It is proportional to power and involves both the internal vibration of the source and its mobility at the contact point. From experimental measurement series, it can be emphasized that the velocity of the free source is a suitable and manageable quantity for describing the internal vibration of many structure-borne sound sources. The active power transmitted from the source structure to the receiving one is found to be the product of the source descriptor and a dimensionless function called the coupling function. This function illustrates the dynamic properties of the source-receiver interface.

Modeling wind tunnel effects on the radiation characteristics of acoustic sources

It is pointed out that the relatively high fuel economy available from propeller-driven aircraft has renewed interest in high speed, highly loaded multiple blade turboprop propulsion systems. Undesirable features related to community noise and the high intensity cabin noise have stimulated new research on the acoustic characteristics of turboprops. The present investigation has the objective to develop a mathematical model of the essential features of the radiation of acoustic disturbances from propellers in a duct and in free space in order to quantify the success with which duct testing can be expected to approximate free field conditions. In connection with the importance of source directionality, a detailed model is considered which consists of a finite element representation of the Gutin propeller theory valid in both the near and far field.

Digital signal processing in acoustics. I

Digital signal processing techniques have gained steadily in importance over the past few years in many areas of science and engineering and have transformed the character of instrumentation used in laboratory and plant. This is particularly marked in acoustics, which has both benefited from the developments in signal processing and provided significant stimulus for these developments. As a result acoustical techniques are now used in a very wide range of applications and acoustics is one area in which digital signal processing is exploited to its limits. For example, the development of fast algorithms for computing Fourier transforms and the associated developments in hardware have led to remarkable advances in the use of spectral analysis as a means of investigating the nature and characteristics of acoustic sources. Speech research has benefited considerably in this respect, and, in a rather more technological application, spectral analysis of machinery noise provides information about changes in machine condition which may indicate imminent failure. More recently the observation that human and animal muscles emit low intensity noise suggests that spectral analysis of this noise may yield information about muscle structure and performance.

A new method for characterization of downhole antennas used in geophysical probing

We have developed a technique for measuring from boreholes the radiation pattern of a high frequency electromagnetic antenna system. Three coplanar holes are required with a single transmitter in one and two identical receiving antennas in the other holes. Signal attenuation is measured along several ray paths at different angles from the source. The method does not depend on the subsurface to be electrically homogeneous but does require reflections and refractions to be negligible. Measurements in fractured granite demonstrate the utility of the technique to obtain antenna radiation patterns. A modification of this same method could be used for characterization of downhole acoustic sources.

Nearfield characteristics of parametric acoustic sources

Because it is often impractical to make measurements in the farfield of a parametric acoustic source, it is desirable to be able to predict the farfield from nearfield measurements. A nearfield theory applicable to saturation-limited as well as absorption-limited sources, previously outlined [J. Acoustic. Soc. Am. 63, 1622–1624 (1978)], has been programmed for digital computation. Growth curves for the source level and beamwidth of absorption-limited sources are presented in generalized form. The theoretical results are compared with experimental data from several sources. Sources involving substantial difference-frequency generation in the primary farfield have broader nearfield beam patterns and lower apparent source levels than in the farfield. On the other hand, when most of the generation occurs in the primary nearfield, the pattern is narrower and the apparent source level higher than in the farfield.

Radiation characteristics of acoustic sources in circular motion

This paper examines the acoustic properties of periodic unsteady rotor blade forces. An efficient computer program, which retains what is regarded as the essential radiation terms, has been developed to numerically evaluate the radiation equations for sources rotating in a circle. A variety of operating conditions and types of blade loading have been computed; of special interest are the radiation properties of impulsive blade loading at high tip speeds. These computations show an acoustic beaming effect similar to that which is radiated by helicopters in high forward speed flight.