Compact directional acoustic sensor using a multi-fiber optical probe.

A compact directional acoustic sensor is described which uses a two-fiber optical probe, a light emitting diode (LED), a photo-diode detector, and a slender cylindrical cantilever to the end of which is attached an optical reflector. Acoustically induced transverse displacement of the cantilever tip modulates the light reflected by it into the collection fiber, which conveys the light to a photo-detector. Directional sensitivity is achieved through the dependence of the collected light on the cosine of the angle between a line through the centers of the two fibers and the cantilever tip displacement (the sound direction). The sensor requires relatively low power, and its LED source has low levels of 1/f noise. These attributes make it a good choice for remote low frequency applications requiring long operating lifetimes. An analytic model of the acoustic response of the cantilever is constructed, which is partially verified using a finite element model and experimentally validated using measurements of the acoustic response in air. The model is used to predict to what extent and over what frequency band that response depends upon the acoustically generated flow (drag) force [Yuan et al., IEEE Sensor J. 8, 1114-1117 (2008)].

Bistatic, above-critical angle scattering measurements of fully buried unexploded ordnance (UXO) and clutter.

Laboratory grade bistatic scattering measurements are conducted in order to examine the acoustic response of realistic fully buried unexploded ordnance (UXO) from above-critical angle insonification, between 2 and 40 kHz. A 127 mm diameter rocket UXO, a 155 mm diameter artillery shell, a natural rock of approximately the same size, and a cinder block are fully buried in water-saturated medium grained sand (mean grain diameter, 240 µm) at depths of 10 cm below the water-sediment interface. A two-dimensional array of bistatic scattering measurements is generated synthetically by scanning a single hydrophone in steps of 3 cm over a 1 m × 1 m patch directly above the targets at a height of 20 cm above the water-sediment interface. Three-dimensional volumetric acoustic images generated from the return waveforms reveal scattering components attributed to geometric and elastic scattering, as well as multiple-scattering interactions of returns between the sediment-water interface and the buried objects. The far-field target strength of the objects is estimated through extrapolation of the angular spectrum. Agreement is found between experimental data and simulated data generated from a finite-element-based, three-dimensional time-harmonic model (2-25 kHz). Separation of the measured UXO from the clutter objects is demonstrated through exploitation of structural-acoustics-based features.

Forward scatter target strength extraction in a marine environment (L).

A 48 m rail with a moving receiver was used to measure forward scattering from a spherical shell lying on the bottom in the Gulf of Mexico. The target was mid-way between the source and rail, on a line from the source bisecting the rail. The major obstacle to the measurement of forward scattering is the much stronger source signal which overlaps the scattered signal in space and time. Here, forward scattered target strength is obtained by processing the received signals using a wavenumber filter to remove the incident wave. The result compares favorably to that obtained from numerical predictions.

Structural-acoustic modeling for three-dimensional freefield and littoral environments with verification and validation.

This paper describes a high-order, finite-element-based, three-dimensional time-harmonic model for large-scale exterior structural-acoustics problems. It is applicable to both freefield and littoral environments. For the freefield case, the infinite exterior is treated as a homogeneous linear acoustic medium. For littoral applications, the water or air and the sediment domains are each treated as linear homogeneous, semi-infinite half-spaces with piecewise-constant properties. Both domains admit complex-valued wave speeds to enable the inclusion of damping. The finite element formulation uses a variational statement which naturally incorporates the transmission-condition at the water or air-sediment interface. The truncation of the infinite exterior is realized using an infinite-element for the freefield case, and the perfectly-matched-layer approximation for littoral applications. Computation of the farfield quantities is done based on an integral representation which, for the littoral cases, uses efficient approximations for the appropriate Green's function. Numerical computations are presented for a series of progressively more complex problems, and are used to verify the model against analytic and other numerical solutions and validate it based on the experimental data for scattering from elastic scatterers as measured in freefield and sediment pool laboratory facilities.

Exploiting forward scattering for detecting submerged proud/half-buried unexploded ordnance.

Laboratory underwater bistatic scattering measurements are reported for free, proud, and half-buried unexploded ordnances for 0 degrees and 90 degrees source angles. Forward echoes are larger than backscattered returns, and half burial significantly decreases the latter but not the former. Results agree with analytic predictions borrowed from radar. The forward echo and source signal are separated by measurements made with and without the target, a method not possible in a target search. For this, a method is described that uses knowledge of the source location and the hyperbolic character in time-cross range of the signals received at points along a line.

Bistatic scattering from submerged unexploded ordnance lying on a sediment.

The broadband bistatic target strengths (TSs) of two submerged unexploded ordnance (UXO) targets have been measured in the NRL sediment pool facility. The targets-a 5 in. rocket and a 155 mm projectile-were among the targets whose monostatic TSs were measured and reported previously by the authors. Bistatic TS measurements were made for 0 degrees (target front) and 90 degrees (target side) incident source directions, and include both backscattered and forward scattered echo angles over a complete 360 degrees with the targets placed proud of the sediment surface. For the two source angles used, each target exhibits two strong highlights: a backscattered specular-like echo and a forward scattered response. The TS levels of the former are shown to agree reasonably well with predictions, based on scattering from rigid disks and cylinders, while the levels of the latter with predictions from radar cross section models, based on simple geometric optics appropriately modified. The bistatic TS levels observed for the proud case provide comparable or higher levels of broadband TS relative to free-field monostatic measurements. It is concluded that access to bistatic echo information in operations aimed at detecting submerged UXO targets could provide an important capability.

A numerical study of defect detection in a plaster dome ceiling using structural acoustics.

A numerical study is carried out to evaluate the effectiveness of using measured surface displacements resulting from acoustic speaker excitation to detect and localize flaws in a domed, plaster ceiling. The response of the structure to an incident acoustic pressure is obtained at four frequencies between 100 and 400 Hz using a parallel h-p structural acoustic finite element-based code. Three ceiling conditions are modeled: the pristine ceiling considered rigidly attached to the domed-shape support, partial detachment of a segment of the plaster layer from the support, and an interior pocket of plaster deconsolidation modeled as a heavy fluid. Spatial maps of the normal displacement resulting from speaker excitation are interpreted with the help of predictions based on static analysis. It is found that acoustic speaker excitation can provide displacement levels readily detected by commercially available laser Doppler vibrometer systems. Further, it is concluded that for 1 in. thick plaster layers, detachment sizes as small as 4 cm are detectable by direct observation of the measured displacement maps. Finally, spatial structure differences are observed in the displacement maps beneath the two defect types, which may provide a wavenumber-based feature useful for distinguishing plaster detachment from other defects such as deconsolidation.

Conformal Fourier wavenumber decompositions on continuous differentiable surfaces.

In this manuscript, a method is introduced for the evaluation of Fourier wavenumber decompositions on C(1) vibrating surfaces for spatial-spectral analysis. Whereas typical Fourier analysis is restricted to geometries that are separable for meaningful interpretations of the corresponding wave motion, this approach allows for conformal spectral analysis along curved surfaces. This is accomplished by restricting the wavevectors to lie within the local tangent to the surface and to be spatially dependent. The theoretical development is presented and it is demonstrated that commonly utilized kernels appropriate for some simple geometries can be recovered. Additionally, this approach is applied in the analysis of the vibration and radiation of a point driven, fluid loaded cone, where the displacements and pressures have been obtained using the finite element method.

Broadband acoustic scattering measurements of underwater unexploded ordnance (UXO).

In order to evaluate the potential for detection and identification of underwater unexploded ordnance (UXO) by exploiting their structural acoustic response, we carried out broadband monostatic scattering measurements over a full 360 degrees on UXO's (two mortar rounds, an artillery shell, and a rocket warhead) and false targets (a cinder block and a large rock). The measurement band, 1-140 kHz, includes a low frequency structural acoustics region in which the wavelengths are comparable to or larger than the target characteristic dimensions. In general, there are aspects that provide relatively high target strength levels ( approximately -10 to -15 dB), and from our experience the targets should be detectable in this structural acoustics band in most acoustic environments. The rigid body scattering was also calculated for one UXO in order to highlight the measured scattering features involving elastic responses. The broadband scattering data should be able to support feature-based separation of UXO versus false targets and identification of various classes of UXO as well.

Determination and analysis of guided wave propagation using magnetic resonance elastography.

We present a novel extension of standard magnetic resonance elastography (MRE) measurement and analysis methods, which is applicable in cases where the medium is characterized by waveguides or fiber bundles (i.e., muscle) leading to constrained propagation of elastic waves. As a demonstration of this new method, MRI is utilized to identify the pathways of the individual fibers of a stalk of celery, and 3D MRE is then performed throughout the volume containing the celery fibers for a measurement of the displacements. A Helmholtz decomposition is performed permitting a separation of the displacements into longitudinal and transverse components, and an application of a hybrid Radon transform permits a spectral decomposition of wave propagation along the fibers. Dot product projections between these elastic displacements measured in the global coordinate system and three Frenet vectors representing the tangent and two corresponding orthogonal vectors along any particular fiber orientation yield the displacement contributions to wave propagation along the fiber as if it were a waveguide. A sliding window spatial Fourier transform is then performed along the length of each fiber to obtain dispersion images that portray space-wavenumber profiles. Therefore, this method can permit localized tracking and characterization of wave types, velocities, and coupling along arbitrarily oriented fibers.

Detection and localization of inclusions in plates using inversion of point actuated surface displacements.

A numerical simulation is carried out demonstrating the use of plate surface vibration measurements for detecting and locating inclusions within the structure. A finite element code is used to calculate normal surface displacement for both steel and mortar plates subjected to a monochromatic point force. The data is generated for the homogeneous plate and the identical plate within which exists a small rectangular inclusion. It is observed that when the elastic modulus of the inclusion is orders of magnitude lower than the base material, resonances of the inclusion produce large local displacements that are readily observed in the raw displacement data. For more modest moduli differences, there are no such directly observable effects. In this case, three inverse algorithms are used to process the displacement data. The first two are local inversion techniques that each yield a spatial map of the elastic modulus normalized by density. These algorithms successfully detect and localize the inclusion based on its modulus difference from that of the base plate. The third technique uses a form of the inhomogeneous equation of motion to obtain the induced force distribution connected with the inclusion. The spatial mapping of this force also successfully detects and localizes the inclusion.

Evaluation of a material parameter extraction algorithm using MRI-based displacement measurements.

The performance of an inversion algorithm is investigated when applied to measured displacement data for a determination of the material parameters /sup /spl lambda/+2/spl mu////sub /spl rho// (longitudinal wave velocity squared) and /sup /spl mu////sub /spl rho// (shear wave velocity squared) throughout an inhomogeneous test phantom. The vector displacement components throughout a test phantom subject to monochromatic shear excitation measured in time using magnetic resonance imaging (MRI) were temporally Fourier transformed to extract the component of monochromatic excitation, and the data was delivered to the inversion algorithm. A series of inversions is presented demonstrating the effects of subsequent wavenumber filtering, polarization selection, and variation in the size of the incremental volume elements. The resulting performance is assessed, and recommendations for future efforts are discussed.

On the noninvasive determination of material parameters from a knowledge of elastic displacements theory and numerical simulation.

In this paper, we present a formulation and numerical simulation for the noninvasive determination of the material parameter ratios (i.e., the Lame parameters divided by density) of an isotropic, inhomogenous elastic medium subject to time harmonic vibration. Given a knowledge of the displacements throughout the medium, a novel implementation of a variational formulation is used to determine the ratios lambda/rho and mu/rho. A theoretical formulation is presented and validated using numerical data obtained from a finite element method. The results indicate that the method may be applied locally within an inhomogenous medium, and that the corresponding material parameters can be recovered to a high degree of accuracy. In addition, the method does not appear to be subject to the typical wavelength constraints of previous methods.

Planar flexible fiber-optic interferometric acoustic sensor.

A planar flexible fiber-optic interferometric acoustic sensor has been developed by wrapping optimized single-mode fibers in a planar spiral form and then embedding the fiber in a thin polyurethane layer. Both the acoustic and the acceleration responses have been found to compare favorably with those of a planar polyvinylidene fluoride sensor of similar geometry.

Microbend fiber-optic sensor.

Intensity modulation induced by microbending in multimode fibers is considered as a transduction mechanism for detecting environmental changes such as pressure, temperature, acceleration, and magnetic and electric fields. A generic microbend sensor has been defined and studied, and its components, such as sensing fiber, light source, optical fiber leads, and detector, have been examined and optimized. Finally, the generic microbend sensor has been tested demonstrating good performance.

Dynamic thermal response of single-mode optical fiber for interferometric sensors.

The dynamic temperature phase sensitivity of a three-layer optical fiber is calculated for unjacketed as well as Al- and Hytrel-coated fibers. The calculations include both the variation of the refractive index with temperature and the thermally induced axial and radial strains. The calculated phase sensitivity indicates that it is currently possible to measure a 1-microdegree C temperature change at frequencies exceeding 50 kHz with 1 cm of a metal coated optical fiber.

Acoustic desensitization of single-mode fibers utilizing nickel coatings.

The pressure sensitivity of the phase of light propagating in a single-mode fiber coated with a thin nickel jacket is determined both analytically and experimentally. The measured acoustic response of the fiber is found to be 1 order of magnitude lower than that of the bare fiber, in agreement with analytical predictions. The technique thus appears to be a promising way for desensitizing optical-fiber leads for use with fiber-optic sensors.

Application of fiber-optic sensors to the detection of air-acoustic signals.

The application of interferometric fiber-optic sensors to the detection of air-acoustic signals is discussed. Differences between air-acoustic and hydroacoustic transduction mechanisms are expected to provide enhanced sensitivity in the former case, and low-frequency measurements confirm this enhancement for the case of an unjacketed fiber loop.

Sensitive, high-speed thermometry using optical fibers.

The use of aluminum-jacketed optical fibers for sensitive, high-speed thermometry is demonstrated. Sensitivity of the fiber to thermal variations is observed at frequencies up to 30 kHz. Although the presence of configurationdependent mechanical resonances between 10 and 22 kHz precludes a simple analysis of the thermal response at these frequencies, the response between 5 Hz and 4 kHz is in reasonable agreement with analytic results. Assuming a minimum-detectable phase change of 10(-6) rad, the observed sensitivity at 4 kHz corresponds to a minimum detectable temperature variation of 1 microK for a 1-cm length of fiber. This performance represents orders-of-magnitude improvement in both sensitivity and frequency response relative to those of commercially available temperature-smeasuring systems.

Magneto-optic coupling coefficient for fiber interferometric sensors.

The magneto-optic sensitivity (coupling coefficient) has been measured as a function of frequency from 15 Hz to 3 kHz for a fiber interferometer incorporating a magnetostrictive nickel toroid. The frequency response of the coupling coefficient was found to be essentially flat from 15 Hz to approximately 600 Hz, where eddy-current losses become significant for the sample tested.