Volume 70, Issue S1, November 1981
Index of content:
- PROGRAM OF THE 102ND MEETING OF THE ACOUSTICAL SOCIETY OF AMERICA
- Session A. Underwater Acoustics I: Propagation
- Contributed Papers
70(1981); http://dx.doi.org/10.1121/1.2018749View Description Hide Description
Several new methods which handle sound speed and density discontinuities have recently been used to solve parabolic equations for underwater sound propagation. The methods are similar in that each involves a matrix whose eigenvalues are, for horizontally stratified media, the usual eigenvalues found by normal mode programs. This paper compares the accuracy of the eigenvalues computed by four different matrix methods currently in use: (1) a finite difference method, (2) a cubic spline method, (3) a linear finite element method, and (4) a cubic finite element method. Some implications of using these matrix methods to calculate sound propagation via the parabolic approximation will be discussed.
70(1981); http://dx.doi.org/10.1121/1.2018750View Description Hide Description
A computermodel for predicting propagation loss based on an implicit finite‐difference solution of the parabolic wave equation is introduced. Prominent features of this model are: (1) Generality: Capable of solving a wide range of range‐dependent and range‐independent problems. Automatic handling of the horizontal interface and arbitrary irregular bottom boundary conditions. Capable of handling multiple layers and an artificial bottom. (2) Easy to use: User's efforts are minimized. (3) Portability: Programs are written in FORTRAN and can be transferred easily from one computer to another. (4) Flexibility: A number of options are included. (5) Efficiency: A unique solution is produced with accuracy. Solutions of problems under different environments will be discussed.
70(1981); http://dx.doi.org/10.1121/1.2018751View Description Hide Description
A theory of mode disappearance at a critical depth has previously been developed for propagation in a Pekeris waveguide of slowly varying depth [A. D. Pierce, submitted to J. Acoust. Soc. Am., July 1981; preliminary results reported in J. Acoust. Soc. Am. Suppl. 1 69, S69 (1981)]. Present paper extends the theory to instances where the sound speed and density are not constant in the water layer. The key assumption, partly justified by the method of matched asymptotic expansions, is that the impedance at the water‐bottom interface, as seen by disturbances in the bottom fluid, is the same (before the critical range) as predicted by the adiabatic mode theory. Beyond the critical range, it is found by analytic continuation. Disturbance in the bottom is given in terms of Airy functions of complex argument. Solution predicts a cylindrically spreading (as r −1/2) wave of narrow beamwidth, which originates from the transition region and which propagates obliquely downward into the bottom fluid. The width of the beam and the angle it makes with the interface are relatively simple functions of the environmental parameters. The shape of the modal profile beyond the critical range does not vary substantially, but the amplitude decreases asymptotically as 1/x 5/2.
70(1981); http://dx.doi.org/10.1121/1.2018752View Description Hide Description
The reflection of pulsed acoustic plane waves from ocean sedimentary layers is studied using a stochastic transport theory originally introduced by Barabanenkov et al. [lzv. VUZ., Radiofiz. 15, 1852 (1972)] and subsequently extended to the two‐frequency context needed for the study of pulse propagation by Besieris and Kohler [Proceedings of the Symposium on Multiple Scattering and Waves in Random Media, edited by P. L. Chow, W. E. Kohler, and G. C. Papanicolaou (North‐Holland, Amsterdam, 1981)]. The sediments are assumed to constitute a random medium in which the density and sound speed undergo small, highly laminated (pancake‐like) fluctuations. Although the problem is formulated in a general context, the predictions of the theory are fully evaluated only in the special case of normal incidence and no refracting profile. However, even with these approximations, some reasonable qualitative agreement of theoretical predictions with measured data (Hatteras abyssal plains) is achieved. [Work supported by ONR Contract No. N00014‐76‐C‐0056 and NAUSEA Code 63D.]
70(1981); http://dx.doi.org/10.1121/1.2018806View Description Hide Description
There are two basic problems associated with performing high‐frequency deep‐water normal mode calculations. First, in deep water the number of modes is approximately of the order of the frequency, and lengthy calculations must therefore be performed. The second problem is that numerical instabilities occur because of the evanescent nature of the lower modes which are “trapped” in the deep sound channel. Because of these problems, normal mode calculations have typically been done only for frequencies less than a few hundred hertz in deep water. A method has been developed to perform such calculations in the multi‐kilohertz region in deep water of the order of 5 km. The results from this technique have been favorably compared with infinite frequency solutions. The wave theory calculations presented, therefore, provide a set of test cases for possible future numerically more efficient approximations.
Low‐frequency acoustics in shallow water: Experimental and theoretical propagation and noise studies70(1981); http://dx.doi.org/10.1121/1.2018807View Description Hide Description
Digital three‐component ocean‐bottom seismometers (OBS) with radio link have been developed to study extremely low‐frequency sound propagation (about 1–15 cps) in coastal waters (mainly below the cutoff of the shallow‐water duct). They have been deployed successfully in conjunction with variable depth hydrophones during several sea trials in the Mediterranean. By recording the hydroacoustic and seismic signals from small underwater explosions and by the power and cross spectra of ambient and ship‐induced noise, it has been demonstrated that the infrasonic energy is mainly transmitted via a seismicinterfacewave. In general this propagation mode may be termed a modified Scholte wave as it is guided by the acoustically most significant interface between water column (including liquid‐type sediments) and solid basement. The very particular features of this wave type are investigated in detail with the help of different software packages, and a theoretical study has been made using a seismic fast field program to model the geoacoustic waveguide.
70(1981); http://dx.doi.org/10.1121/1.2018808View Description Hide Description
Resonance coupling between a horizontally propagating atmospheric wave and Rayleigh waves in the unconsolidated sediment layer occurs at frequencies for which the phase velocity of the dispersive Rayleigh wave is equal to the velocity of sound in air. Measurements of resonance coupling of atmospheric explosive signals into sediments have been reported in numerous investigations. Besides the resonance of the fundamental Rayleigh mode, for which theory was developed by Press and Ewing, higher frequency lines have also generally been reported. Resonance coupling of atmospheric signals above shallow water into sediments has also been observed. A working hypothesis to account for the observed multiple resonances and the effect of a thin water layer on recoupling resonant air‐coupled Rayleigh waves was developed. The theoretical calculations, based on the FFP wave theoretical model developed by Kutschale, reveal that high‐frequency lines are due to normal moderesonances and that significant coupling may occur only when the water depth is less than about one quarter‐wavelength.
70(1981); http://dx.doi.org/10.1121/1.2018809View Description Hide Description
The fact that a normal mode's group velocity may be calculated, without invoking a finite difference approximation involving a second eigenmode computation, does not appear to be widely recognized. The reciprocity relation is typically employed in the derivation of a group velocity formula in a way analogous to the derivation of normal mode attenuations. This paper obtains these results easily using the adjointness property and inner product appropriate to the problem. The complementarity of the reciprocity and adjoint methods is discussed. [Work sponsored by Naval Ocean Research and Development Activity.]
Precise measurement of frequency disperson and experimental tests of the scaling law for group velocities of different modes70(1981); http://dx.doi.org/10.1121/1.2018810View Description Hide Description
The time‐frequency structure of wideband pulses propagated to long ranges in the sound channel is investigated experimentally. A method is presented to measure precisely the frequency dispersion with respect to time for each mode. The method is based on a saddle point approximation for the power spectra of a series of segments of the received time series. It is found that if the length of the time segment lies within certain bounds, the spectrum shape can be predicted and should be peaked at the mode frequencies. Results of the analysis for typical shot data in the Arctic Ocean, Shallow Water, and SOFAR Channel will be presented. This method improves the accuracy of the mode frequency measurement by roughly a factor of six over the sonogram method. With the improved precision, one can check experimentally the validity of the scaling law for group velocities of different modes, i.e., U n (ω) = U m (αω) where α is the scaling factor determined partly by the mode number n and m. Some useful implications of the scaling law will be discussed.
70(1981); http://dx.doi.org/10.1121/1.2018811View Description Hide Description
A computer program for computing the received time signature of explosive sound propagated in a surface duet is described. A synthesized pressure‐time waveform for the explosion is combined with Fourier transform techniques to calculate the source spectrum, which then is converted with the complex frequency response of the surface duct, as predicted by virtual mode theory [F. M. Labianca, J. Acoust. Soc. Am. 53, 1137–1147 (1973)] to obtain the spectrum of the received signal. The key here is the ability to rapidly evaluate the propagation loss at several hundred frequencies. The received shot time signature is found by an inverse FFT of this spectrum. The model is particularly useful for exploring the dispersion of the propagating modes and the interference effects of the source bubble pulses. A mode‐by‐mode decomposition of the received signal is also readily obtained. Results are presented for various shot conditions as well as for source signals other than those arising from shots.
70(1981); http://dx.doi.org/10.1121/1.2018812View Description Hide Description
While measuring ship‐radiated noise levels in the Inland Sea of Japan, broadband interference patterns were observed for ranges less than 1500 m. The water depth was 42 m with little bottom relief. The water column was well mixed with practically no thermocline. The sea state was nearly calm. For these conditions interference patterns could be observed at acoustic frequencies of 200–2000 Hz, which varied with source‐receiver geometry. A ray‐tracing program was developed to explain these patterns assuming no acoustic refraction and a simplified bottom loss algorithm. In addition to the Lloyd mirroreffect, the patterns were reconstructed from those paths which encountered one bottom reflection. The reconstructed patterns have intensity variations, frequency dependence, and range dependence quite similar to the observed data.
70(1981); http://dx.doi.org/10.1121/1.2018857View Description Hide Description
In many ocean environments, the major range‐dependent feature is the variation in water depth with range. There are presently several wave theoretic means of modeling transmission loss in range‐dependent environments. These include various implementations of the parabolic approximation, adiabatic normal‐mode theory, and solutions to the coupled mode equations. In this paper, we consider sound transmission in an environment in which the coupling between modes plays a dominant role in propagation. Numerical results are obtained using four different transmission loss models, the results are compared, and the limitations of each model discussed. [Work supported by NAVSEA.]
70(1981); http://dx.doi.org/10.1121/1.2018858View Description Hide Description
The standard approach to the problem of determining the effects of ocean variability, i.e., internal waves, on system performance is to measure or calculate the mutual coherence function for the acoustic field since it enters directly into measures of system performance such as array gain. Because of difficulties in determining this quantity, we have taken a different approach. Rather than attempting to calculate, e.g., array gain, we seek only to bound it. Consequently, we calculate a limit on the degradation in system performance. The advantage of this approach is that the limit depends only on the mean pressure field, a quantity easier to calculate than the mutual coherence function. The calculation of the mean field does require, however, the use of a generalization of the Markov approximation appropriate when diffracting and refracting effects are present. With this approximation the calculation of the mean field can be carried out with a slightly modified version of the parabolic‐equation technique of Tappert and Hardin. A case study is made for propagation in the Western North Atlantic.
70(1981); http://dx.doi.org/10.1121/1.2018859View Description Hide Description
Recent numerical studies of upslope propagation in a wedge‐shaped ocean with penetrable bottom [F. Jensen and W. Kuperman, J. Acoust. Soc. Am. 67, 1564–1566 (1980)], have revealed that intermode coupling is negligible for small bottom slopes but that each range‐dependent adiabatic trapped mode field is strongly perturbed when that mode passes through cutoff. This observation has revived interest in extending the description of an adiabatic mode field through the transition region from trapped to radiating [A.D. Pierce, J. Acoust. Soc. Am. Suppl. 1 69, S69 (1981)]. We present a procedure of inherent general validity for range‐dependent waveguides whereby an initial ray‐acoustic field undergoing multiple reflection is converted into local mode fields by Poisson summation. Before the conversion, a plane‐wave spectrum is fitted to the ray family, and the resulting Poisson‐transformed spectral integrals, after asymptotic evaluation by the stationary phase method, are found to produce the conventional adiabatic trapped modes downslope from their cutoff points. As mode cutoff is approached, the simple stationary phase procedure must be modified, and the transformed spectral integrals become “canonical integrals” that trace the transition of a mode uniformly from the trapped to the radiating (leaky) regime. Results are shown for the transition function and its connection to the trapped and leaky mode fields downslope and upslope from the cutoff region. [Work supported by ONR Ocean Acoustics Branch.]
70(1981); http://dx.doi.org/10.1121/1.2018860View Description Hide Description
By the method of resolvents or characteristic Green's functions [L. B. Felsen and N. Marcuvitz, Radiation and Scattering of Waves (Prentice‐Hall, Englewood Cliffs, N J, 1973), Chap. 3], one may construct general contour integral representations of Green's functions for separable range‐independent and range‐dependent ocean profiles. From these general integral representations, one may derive a variety of alternative sound field representations, including those involving normal (discrete and continuous) modes, leaky modes, conventional and generalized rays, hybrid ray‐mode combinations, etc. The method cannot be applied rigorously to nonseparable problems such as a wedge‐shaped ocean with penetrable bottom. However, for small bottom slopes, we have managed to construct an approximate characteristic Green's function integral based on the ray‐cycle invariant and on the requirement that the resulting normal mode representation arising from residues at the resonance poles yields the properly normalized and symmetrized trapped adiabatic modes downslope from their cutoff points. For upslope propagation, as a mode approaches cutoff, its resonance pole approaches a branch point and therefore couples strongly to its radiation field. In this transition region, the Green's function may be reduced asymptotically to a canonical integral which is discussed in detail. Beyond the transition region, the previously trapped mode pole becomes a leaky mode pole. The resulting formulation is compared with recent results obtained by other techniques (A.D. Pierce, J. Acoust. Soc. Am., to be published; J. M. Arnold and L. B. Felsen, this meeting). It is also shown how ray, ray‐mode, and other relevant field representations can be obtained from the original integral. [Work supported by ONR Ocean Acoustics Branch.]
- Session B. Architectural Acoustics I and Noise I: Community Noise Impact of Outdoor Amphitheaters
- Invited Papers
70(1981); http://dx.doi.org/10.1121/1.2018903View Description Hide Description
The Universal Studios Amphitheatre is a 5317‐seat open‐air amphitheater located in the vicinity of several residential communities. When the Amphitheatre began its program of summer evening rock concerts, noise complaints were received from the neighbors. A shielding wall was built in an attempt to contain concert noise, but complaints continued. An evaluation of noise levels in areas from which complaints were received was begun in 1973. The program was extended in 1974 to an evaluation of various techniques to contain the noise.Loudspeaker panels and baffles for exit doors were designed and installed. In 1973, a program of noise data collection was begun. Procedures were developed for data collection and for providing information to establish interior sound level limits. This data collection and evaluation continued for eight summer concert seasons from 1973 through 1980. In 1981, work began to completely enclose the Amphitheatre so it could be used year‐round. The development of the noise control program and the eight‐year data collection program is reviewed and summarized.
70(1981); http://dx.doi.org/10.1121/1.2018904View Description Hide Description
A 3800‐seat amphitheater offers amplified popular music concerts within hilly and densely populated terrain. Measuredsound levels within the amphitheater and the residential community from 0.5 to 1.25 miles away are presented. Music levels are compared to neighborhood ambient levels and complaint records. Additional attenuation‐versus‐distance data for rough terrain are offered as well as the noise reduction expected from a fabric enclosure.
70(1981); http://dx.doi.org/10.1121/1.2018905View Description Hide Description
The introduction of popular music concerts to the Tanglewood Music Shed, Lenox, MA, brought noise complaints and legal action against the Boston Symphony Orchestra (BSO); several noise control approaches were investigated, with mixed results. Property‐line noise measurements indicated that some popular music concerts produced significantly higher A‐weighted sound levels than BSO or Boston Pops Orchestra concerts. Closing the side of the shed in the direction of the complainant's property provided significant noise reduction but was unacceptable for practical reasons. A temporary property‐line barrier was constructed and tested for noise reduction effectiveness. Flanking by trees in the high‐frequency region and loss of existing ground‐effect attenuation in the 500‐Hz region prevented the 15‐ft barrier from providing more than a few decibels of overall noise reduction. The complainant issued criteria for acceptable sound levels, which provided guidelines for a sound level control program instituted for the popular music concerts. Sound levels were controlled at the sound control board, requiring cooperation from the performing group. Success was dependent on the degree of cooperation received and the nature of the performing group.
70(1981); http://dx.doi.org/10.1121/1.2018906View Description Hide Description
Technology, economics, and social attitudes have spawned a phenomenon known as “the rock concert.” Audiences of three thousand to three‐hundred thousand attend these events daily in theaters, coliseums, amphitheaters, pavilions, stadiums, race tracks, and parks throughout the United States. The stage stacks used to amplify these events produce sound pressure levels of over 115 dB 4 ft from the loudspeaker assemblies. In enclosed buildings, these levels, though potentially dangerous to those exposed, rarely produce problems related to violations of community noise standards. However, performances in amphitheaters, music pavilions, and park settings are another matter and complaints are mounting regarding the noise pollution created by these concerts. This paper discusses the physical and political problems related to finding sites for new facilities in urban and suburban locations. Experiences in Concord, California and Dallas, Texas are described with examples of solutions developed to meet local ordinances and reduce annoyance factors.
- Session C. Physical Acoustics I and Underwater Acoustics II: Mathematical Methods in Scattering
70(1981); http://dx.doi.org/10.1121/1.2018953View Description Hide Description
Development of a comprehensive computer model to estimate long‐range, low‐frequency reverberation must incorporate models of scattering from rough oceansurfaces. This paper reviews ongoing reverberation research, incorporation of developed segments of the scattering process into a prediction model, and identifies additional theoretical studies which would expand reverberation estimation capabilities. At‐sea measurements of mean reverberation spectra have shown that surface reverberation is Doppler shifted from the carrier frequency by an amount predictable by treating the sea surface as a moving diffraction grating. Reverberation spectra also show significant spectral spreading about both the doppler‐shifted and carrier (bottom reverberation) components. Forward scattering at the sea surface of signals returning from long‐range accounts for the carrier spread, but underestimates the spread of the Doppler‐shifted component. Scatteringmodels to estimate this additional spread and to compute efficiently backscattered amplitudes will complete the mean reverberation estimation. Recommended analytic developments include extensions down to 50 Hz and to grazing angles below 10°; scattering dependence on the orientation to a directional sea surface; and spatial, temporal coherence and fluctuation estimates.