Volume 59, Issue 1, January 1976
Index of content:
59(1976); http://dx.doi.org/10.1121/1.380819View Description Hide Description
This paper reviews the history of horns used in loudspeaker systems for audience‐type sound reproduction. It reviews the early pioneering work started around 1910 which used telephone‐type transmitters before vacuum tubeamplifiers were available. With the availability of amplifiers more powerful driver type loudspeakers were developed to operate with exponential straight and folded horns.Sound motion pictures created a demand for higher quality, and higher‐powered loudspeaker systems for both voice and music. The many stages of development of two‐way loudspeaker systems are reviewed. The limitations of each stage of development before proceeding to the next improvement is discussed. Photographs of many early horns are included as well as the more recent types.
Subject Classification: 10.60; 85.60.
59(1976); http://dx.doi.org/10.1121/1.380832View Description Hide Description
The orthonormal sets of functions known as Gaussian–Laguerre and Gaussian–Hermite functions (G–L/H) have previously been used in quantum mechanics, electromagnetic waves and laser optics. These functions satisfy a modified wave equation, valid under certain restrictive conditions appropriate to slowly diverging sound fields. The nearfield of an ultrasonic transducer can be expressed as a linear combination of modes, each described by a G–L/H function. Each mode has a slightly different phase velocity in the nearfield but has the same spreading nature and phase velocity in the farfield. Using this formulation, the transition from nearfield to farfield is readily explained.
Subject Classification: 20.55; 35.80; 20.15.
59(1976); http://dx.doi.org/10.1121/1.380844View Description Hide Description
The problem of the scattering of a Rayleigh wave by the edge of a thin surface layer of negligible inertia is solved by means of Fourier transforms and the Wiener–Hopf technique. The effect of the layer is approximated by a pair of boundary conditions applied at the surface of the half space. The condition that the traction on the e d g e of the layer vanish is explicitly enforced. Expressions are given for amplitudes of the transmitted and reflected surface waves, and the power transmission and reflection coefficients are evaluated.
Subject Classification: 20.15, 20.30, 35.54.
59(1976); http://dx.doi.org/10.1121/1.380846View Description Hide Description
Cavitation in liquid helium was induced using an ultrasonic cylindrical transducerresonant at approximately 57 kHz. Interaction of the sound field with superfluid helium during cavitation runs was studied using a low‐power second‐sound signal. This signal was observed to decay when a voltage whose amplitude increased linearly with time was applied to the ultrasonic transducer. This is interpreted as being due to the growth of turbulence in the liquid. It is contended that for the sound field to induce cavitation, the liquid has to be in a state of turbulence. It is possible that there is a threshold for additional attenuation of the second‐sound signal near the onset of cavitation. Some cavitation runs were also made in the presence of a high‐power second‐sound current. Under these circumstances turbulence was generated, either by the second‐sound field or by the associated steady heat current. This turbulence lowered the cavitation threshold.
Subject Classification:25.60; 35.32, 35.47.
59(1976); http://dx.doi.org/10.1121/1.380821View Description Hide Description
A model for the prediction of wake‐related sound generation by a single airfoil is presented which assumes that the net force fluctuation on an airfoil can be expressed in terms of the net momentum fluctuation in the near wake of the airfoil. The model predicts a forcing function for sound generation which is related to the spectra of the two‐point velocity correlations in the turbulent region near the airfoil trailing edge. Cross spectra were obtained from correlations of the longitudinal and transverse components of turbulence in the wake of a 36‐in. chord NACA series 63‐009 airfoil in the 7×10 ft tunnel at the NASA Ames Research Center using x‐probe hot‐wire sensors. Additional data were obtained from a 10‐in. chord airfoil in a free‐jet facility at the University of Washington. Based on results from these tests, a scaling model was developed using the turbulent boundary‐layer thickness as the primary length measure. The model was used to predict the sound radiated from a 2‐in. chord airfoil for which acoustical data is available with good agreement both in level and spectral shape. The single airfoil result is extended to a rotor geometry and comparisons studied for various aerodynamic parameters.
Subject Classification:28.65; 50.55.
Calculation of reverberation and average intensity of broad‐band acoustic signals in the ocean by means of the RAIBAC computer model59(1976); http://dx.doi.org/10.1121/1.380822View Description Hide Description
Based on the conventional ray‐tracing principle, a fast numerical algorithm is developed that calculates averaged propagation loss and reverberation of underwater sound. The averaging is performed by summing incoherently the contributions of individual ray bundles to each of the rectangular cells into which the range/depth plane is divided. The resulting matrix description of the average intensity field is used to calculate reverberation. The propagation‐loss algorithm is preliminarily tested against mathematical test functions and measured data.
Subject Classification: 30.20, 30.40; 85.84.
59(1976); http://dx.doi.org/10.1121/1.380823View Description Hide Description
The work presented in an earlier paper [D. Epstein, J. Geophyos. Res. 72, 3701–3710 (1967)] on the expansion of a high‐pressure gas bubble in a fluid‐saturated porous medium, with the fluid treated as incompressible, is extended by treating the fluid as slightly compressible. Two limiting cases are considered: If the particulate matter is carried along with the fluid, the porous medium acts as a dense liquid, with parameters determined by simple functions of the density and compressibility of the fluid and solid particles, and the usual equations of motion apply. However, if the relative velocity of the fluid with respect to the solid is essentially equal to the fluid velocity, the motion is assumed to be governed by a compressible, generalized Darcy equation. From this equation a pair of integrodifferential equations for the bubble radius as a function of time are derived. The form of the radius‐time curve depends on a nondimensional resistance factor m? = (1/2) β?/c, a combination of quantities characteristic of the fluid, the porous medium, and the gas bubble. The transition regime between ideal and Darcy flow is determined as a function of this parameter. For very small values, the bubble pulsates about the equilibrium raius ?. With increasing m?, the ratio of radiated acoustic energy to energy dissipated decreases, until the Darcy limit is reached, where bubble pulsation no longer occurs, and the expansion radius approaches ? monotonically.
Subject Classification: 30.60; 25.60.
59(1976); http://dx.doi.org/10.1121/1.380824View Description Hide Description
The connection between creeping wave and flat surface wavetheory is established by investigating the limit of acoustic scattering from a solid elastic cylinder imbedded in a fluid, as its radius tends to infinity. After applying the Watson–Sommerfeld transformation to the scattering solution, it is shown analytically that the asymptotic expressions for the residue terms (creeping waves) in the Whispering Gallery mode series combine to form separately both the longitudinal and the transverse lateral waves on a flat surface. The cylindrical Rayleigh wave tends individually towards the flat Rayleigh wave, while the Franz and Stoneley waves disappear.
Subject Classification: 30.35; 35.54; 20.30.
59(1976); http://dx.doi.org/10.1121/1.380825View Description Hide Description
By using both continuous and pulsed insonification of a fixed volume at 160 kHz, measurements of scattering strength, scatterer target strengths, and reverberation spectra were made at a single location near San Diego at depths up to 90 m. Scattering strength and density of scatterers decrease slightly with depth while the short‐term fluctuation increases somewhat. Dimensions of the scatterers are estimated to be of the order of millimeters. The scatterers apparently do not migrate vertically over a 24‐hour period and reverberation spectra indicate that they are nonmoving or very slowly moving obstacles carried along with the current. Application of the apparatus in the detection and measurement of turbulence at considerable ranges is considered, and analyses indicate that turbulent speeds of the order of centimeters per second might be observed at ranges of approximately 100 m with a transmitted power of a few hundred watts.
Subject Classification: 30.40; 80.20; 28.60.
59(1976); http://dx.doi.org/10.1121/1.380826View Description Hide Description
Acoustic reflection coefficients for reflection from ocean bottoms are sometimes calculated for postulated models of the bottom structure. A published example, using a model based on geophysical data, is examined with attention to possible simplification and to the physical links between model and computer output. It is shown that calculated results at small grazing angles have a physical cause and can be well approximated by a simpler model. It appears, however, that at larger angles the results are not physically reasonable. The cause is that a commonly used ’’pseudolinear’’ gradient of sound speed is unphysical in a half space, because at a certain finite depth the sound speed becomes infinite, and imaginary at greater depths. A physically reasonable profile of sound speed is proposed as a substitute and expressions for the reflection coefficient are derived in the approximation of vanishing absorption. At that stage, the expressions are considerably simpler than those of the published example. An extension incorporating absorption is planned.
Subject Classification: 30.20, 30.30.
59(1976); http://dx.doi.org/10.1121/1.380827View Description Hide Description
We report in this paper some results on ultrasonicabsorption and velocity of selected low loss liquids and mixtures which may improve the resolving power of the scanning acoustic microscope. We have investigated the effect of electrolytes as a solvent in water which serve to lower the value of α/f 2 of water. We also report on our results with hydrogen peroxide and carbon disulfide for these two liquids should allow us to improve the resolution by more than a factor of 2.
Subject Classification: 35.24.
59(1976); http://dx.doi.org/10.1121/1.380828View Description Hide Description
Westinghouse is developing two different types of sonic cameras for real time imaging in turbid water. A 2‐MHz side‐look sonic camera has been built for viewing the ocean bottom at angles from 35° to 55°. The plane of the bottom is mapped into an image plane behind the lens. Ninety‐eight line transducers are used in the image plane to form 98 simultaneous beams. Range resolution is determined by time, as in a side‐look sonar, and angular resolution is determined by the lens. A 25‐cm lens obtains resolutions of 1.5 to 6 cm at ranges from 3 to 12 m. A circular‐scan camera has been built to operate at 3 MHz. This camera maps a plane orthogonal to the direction of view into the image plane behind the sonic lens. The image is swept out by a row of 128 transducers rotating in the image plane. A 27‐cm lens makes it possible to achieve resolutions of 1 to 3 cm at ranges from 3 to 9 m. Lens data applicable to both cameras is given and initial pictures taken with both types of cameras are shown.
Subject Classification: 35.80; 30.80; 35.65.
59(1976); http://dx.doi.org/10.1121/1.380829View Description Hide Description
A simple frequency equation is derived for the in‐plane vibration of thick rings, taking account of the effects of transverse shear and rotatory inertia. The remarkable accuracy of the expression is demonstrated by comparison with experimental data.
Subject Classification: 40.26.
59(1976); http://dx.doi.org/10.1121/1.380830View Description Hide Description
The changes in the resonance frequencies of circular crystal plates subject to initial bending stresses are studied. The plate is flexed as a cantilever clamped near the edge and stressed by a transverse, concentrated force applied at a point diametrically opposite the support. A system of plate equations which takes into account the nonlinear effects due to the large, initial deflection and to the third‐order elastic stiffness coefficients in the stress–strain relations is employed. In obtaining the initial fields caused by bending, strain components are assumed to be small, but large gradient of plate deflection and large rotations of the plate element about x 1 and x 3 axes are permissble by retaining their quadratic terms in the strain‐displacement relations. The changes of resonance frequencies of fundamental thickness‐shear vibrations are calculated for rotated Y cuts of quartz. The calculated frequency changes are compared with experimental values as a function of the force position, the azimuth angle ψ, and also as a function of the magnitude of thr force P.
Subject Classification :40.24.
59(1976); http://dx.doi.org/10.1121/1.380831View Description Hide Description
Range‐distributed targets are often characterized as arrays of point scatterers. This conceptualization corresponds to a target impulse response that is a sequence of impulses. A physically meaningful generalization of this description is obtained by modeling an impulse response in terms of pulse doublets, impulses, step functions, ramp functions, etc., i.e., as a superposition of delayed impulses a n d differentiated or integrated impulses. A sonar system for generalized target characterization is derived in this paper. Signals and receivers are specified for (1) optimum estimation of reflector parameters and (2) detection of a known target in clutter. The derived signals correspond closely with some of the waveforms that are used by dolphins and bats for echolocation. The corresponding echo analyzers are similar to psychoacousticmodels of the mammalian hearingsystem.
Subject Classification: 60.20, 60.50; 30.82; 80.50.
59(1976); http://dx.doi.org/10.1121/1.380840View Description Hide Description
A software‐oriented adaptive null‐processing technique for underwater acoustic arrays has been developed and applied to experimental data obtained from a 20×20 square array. This method, which involves spatial homodyning, is holographic in nature, with important similarities to both DICANNE and a recent holographic enhancement technique. The approach is described and evaluated, compared with other methods, and results obtained from its application to experimental underwater data are presented. The technique removes the side lobes from a source image as well as the main lobe in the null direction; a striking example of this in an underwater experiment is given.
Subject Classification: 60.30; 35.65; 30.82.
59(1976); http://dx.doi.org/10.1121/1.380841View Description Hide Description
Average discharge rate of single auditory‐nerve fibers in cats was measured in response to one‐ and two‐tone stimuli. One component (the ’’suppressor tone’’) of each two‐tone stimulus was at a frequency (f 2) which produced two‐tone suppression at some stimulus levels. The other component (excitor tone) produced an increase in rate above the spontaneous rate when presented alone. Fractional response was defined as the driven rate to the two‐tone stimulus divided by the driven rate to the excitor alone. Fractional response is thus a quantitative measure of the amount of suppression produced by a suppressor tone. A number of qualitative differences were found in the dependence of fractional response for f 2≳CF and f 2<CF. For suppressor tone frequencies greater than CF, fractional response depends only on the ratio of suppressor to excitor levels (P 2/P 1) for a range of excitor levels (P 1). For P 1 large enough to drive a unit into saturation, fractional response increases with P 1. For f 2<CF, however, fractional response is a monotonic decreasing function of P 2 and P 2/P 1; it is also a monotonic decreasing function of P 1 for P 2/P 1 fixed. Consistent with these results if the tone level ratio P 2/P 1 is fixed, rate is a monotonic increasing function of the overall level of a two‐tone stimulus for f 2≳CF; for f 2<CF, rate is typically a nonmonotonic function of overall level. For f 2≳CF, slopes of (log) fractional response versus (log) P 2/P 1 curves are a monotonic decreasing function of suppressor frequency f 2. For f 2<CF, on the other hand, slope does not depend on f 2. If f 2 is fixed, and excitor‐tone frequency (f 1) varied, fractional response decreases with f 1 for f 2≳CF; for f 2<CF, fractional response is independent of f 1.
Subject Classification: 65.42, 65.59.
59(1976); http://dx.doi.org/10.1121/1.380842View Description Hide Description
A model of hair cell excitation that involves mechanical impact between the hair cells and the tectorial membrane has been proposed by Crane [J. Acoust. Soc. Am. 40, 1147–1154 (1966); J. Acoust. Soc. Am. 51, 508–514 (1972)]. The model predicts that if the impact is unilateral, biasing of the basilar membrane toward or away from the tectorial membrane will significantly affect the generation of nonlinear combination tones. An f 2−f 1 envelope tone synchronized with the f 1 and f 2 primaries was used to achieve such biasing. The experiment reported here investigates the effect of the phase of the f 2−f 1tone on the prominence of the beating of a 2f 1−f 2 combination tone (generated by input primaries f 1 and f 2) and a probe tone that differs in frequency by a few hertz from the frequency of the combination tone. The input thus consists of four explicit tones:f 1, f 2, f 2−f 1, and (2f 1−f 2)+δ. It was found that the f 2−f 1tone can have a large effect on the level of the 2f 1−f 2 combination tone. The phases of the f 2−f 1tone that lead to a maximum and minimum level of combination tone are approximately 180° apart, and there can be almost total discrimination between those two phase conditions.
Subject Classification: 65.56,  65.28, 65.60.
59(1976); http://dx.doi.org/10.1121/1.380843View Description Hide Description
In previous work (e.g., Durlach and Braida, 1969; Braida and Durlach, 1972), we have implicitly assumed that sensitivity in absolute identification of sound intensity is independent of the payoff matrix. There are a number of reasons, however, to question the validity of this assumption. For example, if the payoff matrix causes the subject to focus his attention on a particular subrange of stimuli at the expense of the remaining stimuli, the sensitivity between pairs of stimuli in this subrange might be increased relative to the sensitivity between pairs of stimuli outside this range. This paper reports the results of a series of experiments designed to test the above‐mentioned assumption. On the whole, the results are consistent with this assumption. The only significant exception is that, for one of the five subjects tested, the total sensitivity varies with the payoff matrix (in a manner that appears unrelated to the specific properties of the payoff matrix). Also, the observed changes in response bias associated with the changes in payoff matrix are generally in the same direction as, but somewhat smaller than, the changes in bias that would occur in an optimal Bayes receiver.
Subject Classification; 65.75, 65.50.
59(1976); http://dx.doi.org/10.1121/1.380845View Description Hide Description
The effects of direct variation in static middle earpressure were studied in the acute guinea pig. Pressure differences of from −500 to +500 mm of H2O were examined. Pure tones were delivered to the test ear in a closed system and the sound pressure levels monitored with a condensermicrophone and probe tube situated 2 mm from the tympanic membrane. The cochlear microphonic was recorded from a round window electrode. Both the relative impedance change and cochlear microphonic were noted at various middle earpressure levels. The animals were perfused with Prussian blue and cochleas stained with osmic acid and prepared for histopathological study. It was found that negative pressure causes greater changes on impedance and transmission than positive pressure. Increase in positive pressure was found to cause a monotonic increase in input impedance and decrease in transmission. For negative pressures, the effect was nonmonotonic. Observed changes in transmission with both positive and negative pressures could be accounted for only in part by changes in input impedance of the middle ear. Morphological findings showed some discrete dilation of the vessels on the malleus and the tympanic membrane.
Subject Classification: 65.24, 65.40.