scispace - formally typeset
Search or ask a question

Showing papers in "Journal of Nondestructive Evaluation in 1992"


Journal ArticleDOI
TL;DR: In this paper, it is shown that the ratio between the normal and transverse interfacial stiffnesses can be used to classify the interface imperfections, e.g., kissing, partial, and slip bonds.
Abstract: Ultrasonic reflection measurements from material interfaces are commonly used to detect and quantitatively characterize boundary imperfections of different kinds. Either shear or longitudinal waves can be used to assess the degree of the interface imperfection in acoustical terms. On the other hand, the evaluation of this data in terms of strength-related mechanical properties requiresa priori knowledge of the physical nature of the imperfection. It is shown in this paper that the ratio between the normal and transverse interfacial stiffnesses can be used to classify the interface imperfection. This ratio is readily measured, e.g., by comparing the longitudinal and shear reflection coefficients at normal incidence. Both theoretical and experimental results indicate that different types of imperfections, such as kissing, partial, and slip bonds, can be distinguished by this simple technique.

256 citations


Journal ArticleDOI
TL;DR: In this article, a review of the literature on elastic inclusion problems is presented, with special emphasis on the development of interface conditions modeling different types of interface layer, and various systems of boundary integral equations over the interface are derived.
Abstract: Elastic waves are scattered by an elastic inclusion. The interface between the inclusion and the surrounding material is imperfect: the displacement and traction vectors on one side of the interface are assumed to be linearly related to both the displacement vector and the traction vector on the other side of the interface. The literature on such inclusion problems is reviewed, with special emphasis on the development of interface conditions modeling different types of interface layer. Inclusion problems are formulated mathematically, and uniqueness theorems are proved. Finally, various systems of boundary integral equations over the interface are derived.

76 citations


Journal ArticleDOI
TL;DR: In this article, a configuration of transducers together with a self-calibrating measurement technique is proposed to investigate the reflection and transmission of surface waves by a surface-breaking or near surface defect.
Abstract: A configuration of transducers together with a self-calibrating measurement technique is proposed to investigate the reflection and transmission of surface waves by a surface-breaking or near surface defect. By means of this technique, the ratio of the reflection and transmission coefficients (R/T and/orT/R) can be obtained in a reliable and accurate manner. The reflection and transmission of surface waves for oblique incidence on a surface breaking crack is investigated in detail. Information onT/R for the latter case can be used to determine the depth of the crack. The experimental measurements ofT/R show excellent agreement with theoretical results.

70 citations


Journal ArticleDOI
TL;DR: In this article, the authors used a pair of transducers in a combination of pulse-echo and through-transmission modes to determine material attenuation in composites. But they did not consider the transfer functions of the transducers and specimen surfaces.
Abstract: Ultrasonic techniques utilizing a pair of transducers in a combination of pulse-echo and through-transmission modes were developed. Two types of methods were discussed for determination of material attenuation in composites: direct or absolute methods for materials with low signal loss, and indirect or relative methods for materials with higher signal loss. In all cases, transfer functions of the transducers and specimen surfaces were taken into consideration so that the measurement system is self-calibrated. Void content was measured by means of microscopic image analysis of photomicrographs of the specimen cross-sections and a correlation of material attenuation with porosity was established.

59 citations


Journal ArticleDOI
TL;DR: In this article, a Lamb wave-based approach is proposed for detecting interfacial weakness between an adhesive and adherent in a three-layered medium with imperfections on the individual interfaces.
Abstract: Utilization of specific Lamb wave modes with special cross-sectional wave structures is proposed for the detection of interfacial weakness between an adhesive and adherent. Such an approach is based on plate wave behavior in a three-layered medium with imperfections on the individual interfaces. Selection of appropriate modes and frequencies for adhesion weakness detection is obtained by numerical analysis of the dispersion relations and comparisons of the dispersive curves for perfect, welded and imperfect, smooth boundary conditions. The inspection parameters were evaluated by an analysis of displacement, stress, and power distributions across the three-layered asymmetric adhesive structure. Special selection criteria are proposed. Utilization of dispersive curves do not always provide expected increased sensitivity as in the cross-sectional field distribution approach.

51 citations


Journal ArticleDOI
TL;DR: In this article, the ultrasonic reflection from, mode conversion at, and/or transmission through these imperfect interfaces, are reported and interpreted in terms of a quasi-static model and an independent scattering model.
Abstract: Model specimens are prepared, each of which may be viewed as two sections of similar material joined imperfectly at a planar interface. Measurements of the ultrasonic reflection from, mode conversion at, and/or transmission through these imperfect interfaces, are reported. The interface structures include distributions of pores, contacts, and inclusions. Included are both near-periodic and random cases. As the frequency is increased, the measured reflection coefficients generally show an initially linear increase from zero, followed by a maximum which may exhibit multiple peaks, and a subsequent decay. These results are interpreted in terms of a quasi-static model and an independent scattering model for ultrasonic interactions with imperfect interfaces.

49 citations


Journal ArticleDOI
TL;DR: In this paper, a rational way of obtaining boundary conditions that are exact to first order in the layer thickness is described, and numerical comparisons are performed between these "exact" first order boundary conditions, the commonly used spring boundary conditions and the exact solution.
Abstract: In scattering problems for time-harmonic elastic waves, thin elastic layers are often of interest, e.g., in laminates. Various ways of substituting such layers by some effective boundary conditions have been proposed, and these are briefly reviewed. A rational way of obtaining boundary conditions that are exact to first order in the layer thickness is then described. For a thin spherical layer numerical comparisons are performed between these “exact” first order boundary conditions, the commonly used spring boundary conditions and the exact solution, and it is shown that the “exact” boundary conditions are far superior to the spring boundary conditions in most situations. A drawback with the “exact” boundary conditions is that they are quite complicated.

26 citations


Journal ArticleDOI
TL;DR: A probabilistic neural network is used here to classify flaws in weldments from their ultrasonic scattering signatures, and it is shown that such a network is both simple to construct and fast to train.
Abstract: A probabilistic neural network is used here to classify flaws in weldments from their ultrasonic scattering signatures. It is shown that such a network is both simple to construct and fast to train. Probabilistic nets are also shown to be able to exhibit the high performance of other neural networks, such as feed forward nets trained via back-propagation, while possessing important advantages of speed, explicitness of their architecture, and physical meaning of their outputs. Probabilistic nets are also demonstrated to have performance equal to common statistical approaches, such as theK-nearest neighbor method, while retaining their unique advantages.

26 citations


Journal ArticleDOI
TL;DR: In this paper, first and second order asymptotic boundary conditions are introduced to model a thin anisotropic layer between two generally aisotropic solids, and the results are illustrated by calculation of the interface wave velocity as a function of normalized layer thickness and angle of propagation.
Abstract: First and second order asymptotic boundary conditions are introduced to model a thin anisotropic layer between two generally anisotropic solids. Such boundary conditions can be used to describe wave interaction with a solid-solid imperfect anisotropic interface. The wave solutions for the second order boundary conditions satisfy energy balance and give zero scattering from a homogeneous substrate/layer/substrate system. They couple the in-plane and out-of-plane stresses and displacements on the interface even for isotropic substrates. Interface imperfections are modeled by an interfacial multiphase orthotropic layer with effective elastic properties. This model determines the transfer matrix which includes interfacial stiffness and inertial and coupling terms. The present results are a generalization of previous work valid for either an isotropic viscoelastic layer or an orthotropic layer with a plane of symmetry coinciding with the wave incident plane. The problem of localization of interface waves is considered. It is shown that the conditions for the existence of such interface waves are less restrictive than those for Stoneley waves. The results are illustrated by calculation of the interface wave velocity as a function of normalized layer thickness and angle of propagation. The applicability of the asymptotic boundary conditions is analyzed by comparison with an exact solution for an interfacial anisotropic layer. It is shown that the asymptotic boundary conditions are applicable not only for small thickness-to-wavelength ratios, but for much broader frequency ranges than one might expect. The existence of symmetric and SH-type interface waves is also discussed.

25 citations


Journal ArticleDOI
TL;DR: In this article, the authors examined the magnetic field produced by small flaws in a two-dimensional, conducting plate carrying an otherwise-uniform current, and they used a conjugate function approach to calculate the current and voltage distributions about the circular and elliptical flaws in the conducting plate.
Abstract: We examine the magnetic field produced by small flaws in a two-dimensional, conducting plate carrying an otherwise-uniform current. We use a conjugate function approach to calculate the current and voltage distributions about the circular and elliptical flaws in the conducting plate, and examine the dependence of the normal component of the magnetic field upon distance, hole size, elliptical eccentricity, and elliptical orientation. We show that when the field is calculated, far from the hole, the field falls off as 1/z 3, wherez is the distance above the plate, and as 1/r 2, wherer is the distance from the center of the hole to the observation point. We also show that for circular and elliptical flaws, the normal component of the magnetic field in the far-field region is linearly related to the area of the flaw.

15 citations


Journal ArticleDOI
TL;DR: In this paper, a general theory of scattering by thin layers included in a continuous matrix is given, which can account for general inhomogeneities and interfacial roughness provided that the root mean square thickness is small compared with the wavelength.
Abstract: In a recent article the author gave a general theory of scattering by thin layers included in a continuous matrix. Such layers may be thought of as plates, shells, coatings, or interface layers depending on the particular physical context. It was shown that, in the general case, the scattering may be described in terms of the stress and momentum polarizations in the layer and that when the latter is thin compared with the wavelength of the incident sound, the through-thickness average polarizations satisfy certain boundary integral equations. In this article we develop the theory in the context of imperfect interfaces such as those which may occur in adhesive and diffusion bonds. The advantage of this approach is that it can account for general inhomogeneities and interfacial roughness provided that the root mean square thickness is small compared with the wavelength.

Journal ArticleDOI
TL;DR: In this paper, a numerical boundary element method was used to find the far-field scattering amplitudes of a single defect for a normally incident plane wave, and the normal incidence reflection coefficient for a planar distribution of such defects was obtained from the independent scattering model.
Abstract: A numerical technique for obtaining interface reflection coefficients for imperfect bonds between similar materials for a wide range of distributed defects is developed. A numerical boundary element method is utilized to find the far-field scattering amplitudes of a single defect for a normally incident plane wave. Then, the normal incidence reflection coefficient for a planar distribution of such defects is obtained from the independent scattering model. As a validation, the reflection coefficients are compared to the quasi-static model results where the latter are available. This establishes the basis for one application of the new model, the determination of spring constants which are not available. Other applications of the model, including studies of the response at frequencies beyond the quasi-static limit, the ratio of longitudinal to transverse wave reflectivities, and the effects of selected multiple scattering are discussed.

Journal ArticleDOI
TL;DR: In this paper, an acoustic emission (AE) technique is used to detect the fiber fractures and locate the fiber breaks and determine both the fiber failure stresses as well as the fiber fragmentation lengths while the test is in progress.
Abstract: In the single-fiber-composite (SFC) test, a fiber imbedded in a matrix is loaded in tension, resulting in a fragmentation of the fiber. In the conventional version of this test, the final fiber fragmentation length distribution is used with a micro-mechanical model to determine the average fiber/matrix interfacial shear stress. In the enhanced version of this test, one also determines the applied stress at each fiber fracture, and from this, one can evaluate the strength of the fiber at short gage lengths. In our measurement system, we utilize an acoustic emission (AE) technique to detect the fiber fractures and to locate the fiber breaks and so determine both the fiber failure stresses as well as the fiber fragmentation lengths while the test is in progress. Critical to the success of this test is a broadband AE system that utilizes point-like AE sensors, procedures for evaluatingin situ, the wavespeed of the first wave arrival and signal processing techniques for determining the arrival time of this signal as precisely as possible for a broad range of wave shapes. Here we describe the application of such an enhanced SFC test procedure to investigate the failure of a Nicalon™ fiber in an epoxy matrix.

Journal ArticleDOI
TL;DR: In this article, the authors used the complex source point (CSP) method for modeling quasi-Gaussian beams and solved the reflection problems rigorously by wavenumber spectral decomposition.
Abstract: Nonspecular reflection plays an important role in acoustic beam interaction with fluid-immersed elastic media. Such anomalous reflection is attributed to the strong interaction which occurs when the incident beam is phase-matched to one of the leaky waves supported by the structure. The properties of the incident beam as well as those of the interface geometry exert a marked influence on the observed nonspecular return. Previous investigations have been limited to rather special beam and interface conditions. The present study removes many of these limitations by allowing for arbitrarily collimated beams incident on plane and (cylindrically) curved layered geometries as well as simultaneous excitation of multiple leaky waves. By use of the complex source point (CSP) method for modeling quasi-Gaussian beams, the reflection problems are solved rigorously by wavenumber spectral decomposition. They are then reduced by asymptotic techniques to yield physically meaningful wavefield contributions, which explain the phenomenology and also allow efficient computation. The accuracy of the CSP asymptotic algorithms, and that of more restrictive conventional algorithms, is assessed by comparison with purely numerically generated reference data. The results establish the accuracy and versatility of the CSP strategy for a broad range of beam-interface conditions. While the present study is for two-dimensional problems, the method has also been extended to the three dimensional case. The data base generated by this method is the first step toward developing a strategy for extracting from data information about the interface conditions.

Journal ArticleDOI
TL;DR: Guided Rayleigh-Lamb waves in a laminated composite plate with interlaminar bond layers have been studied in this article, and the predictions for the dynamic response of the plate based on the (massless) spring model of the bond layer have been compared with those using the exact multilayer formulation and those of a plate without interface bond layers.
Abstract: Guided Rayleigh-Lamb waves in a laminated composite plate with interlaminar bond layers have been studied in this paper The predictions for the dynamic response of the plate based on the (mass-less) spring model of the bond layer have been compared with those using the exact multilayer formulation and those of a plate without interface bond layers It is found that all three model predictions agree well at low frequencies when the bond layer stiffnesses are not too low As the stiffnesses get smaller, however, the response changes drastically even at low frequencies These results would be useful in predicting weakening of the bond

Journal ArticleDOI
TL;DR: In this paper, a model of a compressional ultrasonic transducer is implemented into a T-matrix method-based solution to a crack scattering problem, where the probe can act both as a receiver and as a transmitter, and it is modeled as an acoustic piston-like source.
Abstract: A model of a compressional ultrasonic transducer is implemented into a T-matrix method-based solution to a crack scattering problem. The probe can act both as a receiver and as a transmitter, and it is modeled as an acoustic piston-like source. A previous solution by the null field approach is applied and is here used to model a crack that is partly closed due to an external background pressure. Numerical calculations of the signal response when the crack is penny-shaped are performed and compared with results from a program based on the Geometrical Theory of Diffraction, and the agreement is generally found to be very good.

Journal ArticleDOI
TL;DR: In this article, the authors provided a theoretical framework to study the nonlinear viscoelastic behavior of adhesive layers in shear and applied it to the stress-strain behavior of adhesives.
Abstract: The purpose of this paper is to provide a theoretical framework to study the nonlinear viscoelastic behavior of adhesive layers. The stress-strain behavior of the adhesive material in shear is assumed to have a nonlinear elastic part and a linear viscoelastic part, i.e., σyx=f(eyx)+Kmeyx. To measuref(eyx) andKm by an ultrasonic technique, it is proposed to prestress the adhesive layer in the nonlinear range and to superimpose a small amplitude ultrasonic signal. The reflected field is used to obtain the wave speed and hencedf(eyx)/deyx as a function of the pre-stress. The viscoelastic parameter is obtained from the amplitude decay. By repeating this procedure for various values of prestress and using numerical integration, the stress-strain behavior of the adhesive layer can be reconstructed.

Journal ArticleDOI
TL;DR: In this paper, it is shown that errors in estimating the flaw distribution variance affect both the scattering amplitude mean and variance, and that the distribution has a bimodal character at high frequency.
Abstract: Probabilistic approaches to flaw detection, classification, or characterization often assume prior knowledge of the flaw distribution. It is implicit that there is a scattering amplitude distribution associated with the flaw distribution. In a number of previously published probabilistic analyses, it has been assumed that scattering amplitude is an uncorrelated, Gaussian random variable with zero mean and known variance. In the work reported here, these assumptions are evaluated for the case of a lognormal distribution of spherical flaws. The correlation, mean, variance, and nature of the scattering amplitude distribution are considered as a function of frequency and as a function of the breadth of the assumed flaw distribution. It is shown for the assumed flaw distributions that scattering amplitude is not uncorrelated and does not have zero mean. It is shown that errors in estimating the flaw distribution variance affect both the scattering amplitude mean and variance. Using both analytical and numerical procedures, the scattering amplitude distribution is shown to be lognormal at long wavelength for a lognormal distribution of spherical scatterers. At high frequency, the distribution is shown to have a bimodal character.

Journal ArticleDOI
TL;DR: In this paper, a random speckle modulation technique was proposed to improve the optical sensitivity of the interferometer on the surface of a diffuse object by moving a large number of very bright speckles around at an appropriate speed.
Abstract: In spite of its obvious advantages over conventional contact and immersion techniques, laser interferometry has not yet become a practical tool in ultrasonic nondestructive evaluation since its sensitivity is insufficient in most practical applications. Part of the problem is that the maximum signal-to-noise ratio often cited in scientific publications and manufacturers' specifications cannot be maintained on ordinary diffusely reflecting surfaces. Although these surfaces reflect a fair amount (5–50%) of the incident laser light, this energy is randomly distributed among a large number of bright speckles. Unless the detector happens to see one of these bright speckles, the interferometer's signal-to-noise ratio will be much lower than the optimum. This adverse effect is almost completely eliminated by the suggested random speckle modulation technique. The conventional interferometric technique was modified to assure random occurrence of a few very bright speckles and to move the whole speckle pattern around at an appropriate speed. Random but frequent bright flashes detected from the surface of the specimen resulted. The bright periods are 0.1 ms or longer, sufficient to trigger the ultrasonic pulser and detect the transmitted signals before the flash subsides. As much as 5–10 times improvement of the optical sensitivity was achieved by this novel approach and close to maximum signal-to-noise ratio was maintained everywhere on the surface of a diffuse object.

Journal ArticleDOI
TL;DR: In this article, the early-time asymptotics of the reflected wave are given exactly by the independent scattering model for waves that are normally incident on the bond plane, and theoretical arguments for the plausibility of the hypothesis are presented.
Abstract: The specular reflection of ultrasound from defect covered bond planes is analyzed in the time-domain using the independent scattering model for incident plane waves. We hypothesize that the early-time asymptotics of the reflected wave are given exactly by the independent scattering model for waves that are normally incident on the bond plane. For non-normal incidence, a more restricted result is available for reflected longitudinal waves. We present theoretical arguments for the plausibility of the hypothesis. Experimental measurements made on two sets of model bond planes test and support the hypothesis. Our motivations are as follows. An effort is underway to develop nondestructive methods for estimating the integrity of metal-metal bonds. These methods primarily focus on the reflection of ultrasound in the long wavelength limit, where one can estimate an effective elastic constant. People have characterized the overall quality of the bond in terms of this elastic constant. However, one cannot in this way infer more detailed information such as the average size of the defects or their area fraction. Higher frequency probes, which do provide more detailed information, have been studied in the independent scattering model. Consequently, we have (1) extended the independent scattering model to the time-domain, and (2) shown that it is asymptotically correct for early-time reflections from a defect covered plane (the bond plane). We expect that time-domain methods, based on the analysis presented in this paper, will form the basis for a more rigorous technique for determining the area fraction and the average defect size.

Journal ArticleDOI
TL;DR: In this article, a theoretical study of the interaction of ultrasonic waves with the rough solid-solid interface is presented, where reflection and transmission coefficients for longitudinal and shear coherent waves are calculated as a function of the angle of incidence within the framework of a second order perturbation theory.
Abstract: An imperfect interface between two dissimilar materials is modeled by a random interface profile. A theoretical study of the interaction of ultrasonic waves with the rough solid-solid interface is presented. The reflection and transmission coefficients for longitudinal and shear coherent waves are calculated as a function of the angle of incidence within the framework of a second order perturbation theory. The effects of the statistical interface parameters, as well as the interface spectral density on the scattered fields, are investigated. These results are used to determine the roughness-induced attenuation of the coherent fields as a function of the above parameters. In addition, the relation between the incoherent part of the scattering cross-section, and interface roughness is examined.

Journal ArticleDOI
TL;DR: In this article, the transient response of a random medium as predicted by two methods, a perturbation scheme and a causal approach, is investigated by direct computation of a pulse propagation.
Abstract: The paper investigates the transient response of a random medium as predicted by two methods, a perturbation scheme and a causal approach. It is shown by direct computation of a pulse propagation that the perturbation scheme yields an “advanced” response inconsistent with the physical realizability of the medium. The causal model as well as a modified perturbation scheme exhibit a physically reasonable behavior.

Journal ArticleDOI
TL;DR: In this paper, the authors used double-exposure holography on the detection and assessment of plates containing locally thinned regions as artificial defects, where the artificial defect is programmed by thinning locally a plate to various depths with a milling cutter.
Abstract: This paper describes the use of double-exposure holography on the detection and assessment of plates containing locally thinned regions as artificial defects. The artificial defect is programmed by thinning locally a plate to various depths with a milling cutter. During testing, the plate, illuminated by an expanded laser beam, is clamped along its periphery and an incremental uniform pressure applied between two exposures on the holographic plate. Local thinning is revealed from anomaly in the reconstructed fringe pattern—the boundary of the anomalous fringes is larger than the boundary of the actual artificial defect, and the amount of thinning is deduced from the number of fringes within the boundary. Artificial defects with eccentric local thinning are more readily revealed from the reconstructed fringe pattern than the artificial centric defects. A simple theoretical model is developed so that when the experimental fringe-counts are compared with the theoretical ones, any discrepancy between them reveals not only the presence of centric and eccentric artificial flaws but also their sizes and depths.

Journal ArticleDOI
TL;DR: In this paper, a low frequency (0.5-2.5 MHz) acoustic microscope has been fabricated where the microscope lens has been replaced by two ultrasonic transducers with cylindrical concave faces.
Abstract: In recent years, acoustic microscopy has been found to be very useful for characterizing engineering as well as biologic materials. With the present state of knowledge on acoustic microscopy, one can obtain the surface wave velocity of a homogeneous specimen or coating thickness of a coated material and produce images of near surface internal defects and inhomogeneities in a specimen. Applications of acoustic microscopy for obtaining material properties of anisotropic specimens and detecting material defects at a greater depth are meager because commercially available acoustic microscopes are insensitive to direction-dependent material properties and they, in general, have poor penetration properties because of high operating frequencies. Recently at the University of Arizona an unconventional low frequency (0.5–2.5 MHz) acoustic microscope has been fabricated where the microscope lens has been replaced by two ultrasonic transducers with cylindrical concave faces; one works as a transmitter and the other one works as a receiver. Using this arrangement, it has been found that it is possible to detect internal damages in a material and identify material anisotropy in fiber-reinforced composite plates. These capabilities of the microscope are demonstrated in this paper by presenting some experimental results along with theoretical justifications. Then pattern recognition techniques are used to solve the inverse problem, that is, to predict the type of material defect from reflected acoustic signals.