By using an experimental setup yielding the acoustical Two-Port properties of earplugs. the attenuation of three fundamentally different types of earplugs is predicted. The setup used is based on molds of the ear canals of a real person, and the Two-Port of each earplug is numerically coupled to an acoustical model of the ear. The predicted attenuation is compared to the subjective attenuation (REAT) measured on the same person. For one type of earplug the results are modified to compensate for the lack of resilience in the artificial ear canals. The results of repeated measurements are encouraging. Not only does the objective method yield attenuation data similar to the REAT, it also seems to reflect some of the variations given by subjective measurements.
The acousto-optic technique to evaluate coated substrates, presented by Devolder, Wevers, Demeester and Leroy [1], uses the phase dependence of the reflected beam on the coating parameters to measure the thickness of thin coatings. In this theoretical paper, we have used Fourier analysis and inhomogeneous wave theory to perform numerical simulations of the phase shift. In the region of small layer thicknesses, the results of the two models agree to within one degree. Furthermore, it is shown that the phase sensitivity can be significantly improved by using higher frequencies and wider beams. The predictions of the phase shift are consistent with plane wave analysis when the ratio of the beamwidth to the sound wavelength becomes small. Measuring more to the right of the maximum of the non-specular lobe also results in higher phase shifts. Finally, we investigate the dependence of the phase shift on the material parameters of the coating. The results presented in this paper can be used to optimize the sensitivity of the acousto-optic nondestructive testing (NDT) technique for thin coating characterization.
In order to reduce computational complexity and memory requirements for three-dimensional acoustical analysis, using the boundary element method (BEM), a new method for steady-state sound fields is developed based on the fast multipole algorithm. This method, called the Fast Multipole BEM (FMBEM), drastically accelerates an iterative solution of large-scale linear systems, without composing the dense influence coefficient matrices used for the conventional BEM. An efficient computational scheme is presented for the boundary integral equations in the basic form and in the normal derivative form, where the fast multipole algorithm is introduced over the multiple levels, by employing a concept of cells clustering boundary elements and hierarchical cell structure. Theoretical analysis shows how efficiently the FMBEM can reduce the computational complexity and the memory requirements compared with the BEM, in two typical problems assuming uniform distributions of nodes within a space and on its surface, respectively. A numerical test solving an acoustic tube problem confirms the validity of this method with respect to computational accuracy and efficiency.
The paper deals with an inversion approach based on modal travel time measurements for the estimation of water column and bottom properties in shallow water, by acoustic means. The modal structure of the acoustic field measured at a single hydrophone is defined on the basis of group velocity predictions at the central frequency of a broadband signal (tomographic signal). In the proposed approach, the assumption of good a-priori knowledge of the parameters to be recovered which is normally applied in modal-travel time inversion schemes, is waived and a relatively wide search space is considered. Following an identification process to assign modal arrivals to the peaks of the measured signal, a matching procedure is adopted based on the minimization of a cost function relating the travel time differences of the identified peaks with respect to the modal travel times for the same peaks. This procedure leads to a set of environments, which are considered close to the actual one, among which a reference environment for the application of a linear inversion procedure is chosen on the basis of the convergence of the local search. Thus, the inversion procedure adopted is hybrid in the sense that an optimization procedure for the estimation of the most probable reference environment is combined with a linear inversion method to estimate the unknown parameters. The approach can be applied in shallow water tomographic experiments involving a single source-receiver pair and leads to the simultaneous estimation of water and sediment sound velocities.
A new monaural method for the suppression of late room reverberation from speech signals, based on spectral subtraction, is presented. The problem of reverberation suppression differs from classical speech de-noising in that the "reverberation noise" is non stationary. In this paper, the use of a novel estimator of the non-stationary reverberation-noise power spectrum, based on a statistical model of late reverberation, is presented. The algorithm is tested on real reverberated signals. The performances for different RIRs with T-T ranging from 0.34 s to 1.7 s consistently show significant noise reduction with little signal distortion. Moreover, when used as a front end to an automatic speech recognition system, the algorithm brings about dramatic improvements in terms of automatic speech recognition scores in various reverberant environments.
In this paper, we study the effect of atmospheric turbulence on sound propagation in the low atmosphere and of its influence on the localization of vehicles. In the course of experiments, the average and fluctuating meteorological magnitudes of the low atmosphere have been measured in order to have a better knowledge of the stability state, and of the rates of thermal and kinematic fluctuations of the atmosphere. The localization antenna is made up of two overlapping linear antennas, each composed of 8 microphones. The whole acoustic and meteorological data have been measured simultaneously and recorded in situ. From this information, an average value analysis has allowed us to layout the vertical speed profiles of the wind and temperature by means of the Monin-Obukhov theory of similarity. In three different cases of wave refraction, these profiles have been used for a model of propagation based on the method of rays in order to calculate the sound paths and for a calculation code of the acoustic level (PE) allowing us to know the propagation conditions of the wave between the source and the antenna. Then the localization results have been analyzed together with thermal and kinematic fluctuation results of the low atmosphere. A narrow relationship between the precision of the localization results and eigenvalues of the correlation matrix of acoustic signals has been established. The degradation of these eigenvalues is a direct consequence of the thermal and the kinematic fluctuation rates of the low atmosphere.
A recent paper by D. Hohenwarter and R Jelinek (Acustica - acta acustica 86 (2000) 1-14) states that the refraction law known in the literature for a sound ray in a two-dimensional stratified moving medium is approximate. In that paper, using Fermat's principle of least time, a refraction law for a sound ray is derived which is claimed to be different from those known in the literature. In the present paper, we show that, in fact, the refraction law obtained by D. Hohenwarter and R Jelinek coincides with those known in the literature. However, these authors were most likely the first to derive the refraction law for a sound ray in a moving medium by using Fermat's principle of least time.
Operation of a Lamb wave device, using an interdigital transducer deposited on a bilayer substrate with spatially varying thickness ratio, is investigated both theoretically and experimentally to understand the characteristics of ultrasound radiation into water that accompanies Lamb wave propagation. The bilayer is composed of a tapered piezoelectric ceramic plate and a tapered acrylic plate and the interface boundary surface of the two plates is gradient in the wave propagation direction, retaining a constant total thickness. Acrylic is used because its acoustic coupling with water is strong. A new calculation model for Lamb wave propagation in the water-loaded bilayer is established and the Lamb wave velocities as well as the radiation efficiencies of the compressional waves in water are presented for various gradient values of the interface boundary. The phase velocities increase or decrease only slightly as the gradient grows, but the radiation efficiency increases drastically with gradient. Experiments confirm that the compressional waves are emitted in two directions from a Lamb wave device and that the radiation in one direction is stronger than in the other, showing the possibility of designing a unidirectional transducer.
The ultrasonic non-destructive testing and more particularly the characterization of material flaws requires a good knowledge of ultrasonic field interaction with flaws. The aim of this work is to model the interaction of an ultrasonic beam with a plane flaw to simulate a fatigue crack in a solid in bidimensional geometry. The model takes into account all the sequences: Emission - Propagation - Transmission through a plane interface - Interaction with the flaw Retro-propagation towards the transducer - Reception. This model predicts the echographic signal stemming from the diffracted field by both edges of the crack in order to have profound understanding and better interpretaion of the experimental results.
A new approach of hearing aids characteristics measurement based on signal processing algorithms is proposed. Actually the pure tones tests used in classical methods are not convenient for digital devices and cannot reach some digital hearing aids properties like speech signal / noise ratio. This new method is founded on the modelisation of hearing aid by an ARMAX filter as the transfers function of the device. From this function are defined the response curve and its variations with different inputs. But other properties and signal / noise ratio can be discovered by this method. In this first approach the authors undertook a validation of the method with stationary signals like white noise and an analogic hearing aid. An essential study of the number of coefficients was made, and the response curves obtained by modelisation was compared with the curve given by classical methods.
The transmission coefficient for a single plane wave sent through a thin infinite homogeneous isotropic plate separating two fluid media has been derived by Cremer in 1942. Cremer's classical formula is converted into two distinct compact representations revealing the genuine dependencies on material parameters including damping. It is pointed out that the calculation of the diffuse sound field transmission loss is by no means trivial, and a safe and efficient method of calculation is presented. Furthermore, the asymptotical behaviour of the diffuse sound field transmission loss is worked out for both high and low frequencies, generalising Cremer's results with respect to arbitrary maximum incidence angle and general damping. With an adequate expansion a simple analytic estimate for the maximum of the transmission loss and its position has been found. An approximate formula valid for all frequencies is constructed on the basis of the low- and high-frequency expansions. Analytical approximations for all local extremal values and their positions have been generated by a numerical procedure. Finally generalisations for anisotropic plates and distinct fluid media are described. The combination of the suggested exact method with standard integrators enables safe calculation of the transmission loss of anisotropic plates including third-octave or octave frequency averages.
Nowadays, passive mufflers are commonly used to reduce ventilation noise. However, the pratical work conditions of these devices, here in aeronautic domain, involve new calculation mean for performance estimation of finite length dissipative mufflers. The previsional calculation is based on the modal theory development for cylindrical lined ducts, in the presence of a uniform air flow. The important point of the calculation is performed by the numerical resolution of transverse boundary conditions, from the normal acoustic impedance modelling of the absorbant structure. The realised testing stand gives us the fine analysis of the the attenuation with an air flow, by measuring the insertion loss. This experimental study brings us a validation of the theoretical tendances.
In order to retain the scaling properties of the time domain solution for the beam equation, local longitudinal strain must be considered, giving rise to rotational viscosity and viscous damping. The formulation is formally equivalent to hysteretic damping. Generalisation to two-dimensional beams, or plates, is suggested.
A system is introduced with the purpose of simplifying the basic standard ISO 717-1 for rating of sound insulation in buildings and of building elements. The core idea is that the whole rating procedure can be accomplished by using only two basic weighted sound reduction indices. The difference between these indices plays an important role, because all other indices defined in ISO 717-1 that depend on the incident spectrum can be approximated using one of the basic indices and the defined difference only. The system is based on so-called spectrum indices defined as a C- and A-weighted level difference of the spectrum considered. Sound insulations have been found to exist that are unequivocal functions of spectrum index. General representations for such sound insulations are introduced, and they are shown to be very effective for estimating a weighted sound reduction index for any spectrum. A given collection of spectra is shown to give accurate results. The basic rating mechanisms to be used in requirements and for rating of sound insulation of building elements are based on the new indices and the defined difference. The rating for octave band measurements is based on reference curve methods also in the case of incident traffic noise spectra. The estimation procedure introduced in the study also has applications for estimating A-weighted sound pressure levels when hearing protectors are worn.
This study reports on some practical experiments with subjective hearing in simulated workplace situations. A methodological approach for sound quality evaluation is developed. Here two experiments are described which investigate the influence of room geometry and absorber placement on the perception of loudness, roughness and sharpness under conditions where the room volume, the sound level, and the reverberation time were kept constant. The results presented in this paper directly describe simulated (virtual) situations and relate perceptions to objective quality parameters respectively calculated on a mathematical model. The deviations across different signals are small, both for the subjective and the objective results, but they are in line, showing this approach worth of trying.
The acousto-optical interaction of a plane monochromatic light wave with a plane sound pulse, consisting of a continuous frequency spectrum, is considered. A new integro-differential equation, describing the phenomenon, is derived. This equation is firstly solved in the Raman-Nath regime. A general expression for the intensity in the optical near field (Fresnel zone) is given. For the case of a weak sound field and close to the sound column, it is shown that the light intensity, as a function of time, is proportional to the second derivative of the sound pulse signal. Secondly this equation is solved beyond the Raman-Nath regime. The obtained expressions can he very useful for optical near field measurements of sound pulses with a discrete or continuous spectrum, that consists out of higher frequencies than in the restricted Raman-Nath case.
