Showing papers in "Ndt & E International in 1996"

An automated radiographic NDT system for weld inspection: Part I — Weld extraction

Abstract: An automated radiographic NDT system for weld inspection is developed. The entire system consists of two major components: weld extraction and flaw detection. Welds are first extracted from digitized radiographic images. The extracted welds are then processed by flaw detection algorithms to determine their qualities. This paper describes the weld extraction methodology, which forms the first part of the system. An example is used to illustrate the working of the methodology. The methodology is concluded to be effective for weld extraction after being successfully tested with 25 images.

NDT of welds: state of the art

Abstract: Major advances have been made in recent years in nondestructive testing of welded steel structures. This paper reviews the latest developments in: (i) radiographic inspection; (ii) ultrasonics; (iii) alternating current potential drop technique; (iv) measurement of residual stress, and (v) in-production weld inspection. Areas where further useful progress can be made are also identified.

Neural network classification of flaws detected by ultrasonic means

Abstract: A methodology for the automatic recognition of weld defects, detected by a P-scan ultrasonic system, has been developed within two stages in the present work. In the first stage, a selection of the shape parameters defining the pulse-echo envelope reflected from a generic flaw, and defined in the time domain, is performed by Fischer linear discriminant analysis. In the second stage the classification is carried out by a three-layered neural network trained with the backpropagation rule, where the input values are the parameters selected by the Fischer analysis. With regard to the neural network learning process, 135 real weld defects have been considered. The defects, distributed among the classes of cracks, slags of inclusion and porosity, had been previously characterized by X-ray inspection. The results obtained confirm the effectiveness of the approach in preserving the discriminant information needed for characterization by an iterative use of Fischer analysis, and in increasing the generalization properties of the layered network by an interpretation of the knowledge embedded in the generated connections and weights. The required computation time allows in-process application.

Effect of applied stresses on magnetostriction of low carbon steel

Abstract: We have measured the magnetostriction of low carbon steel specimens while applying magnetic fields, being both parallel and perpendicular to the uniaxial applied stress. An individual domain of a polycrystalline steel is elongated in its magnetization direction, which coincides with one of the crystallographic axes, 〈100〉. The magnetization appears as a result of the domain wall movement induced by the applied magnetic field. The rotation of the domain magnetization participates, when the field becomes larger. As the magnetic field is increased, the magnetostriction in the magnetization direction first increases and then shows a maximum when the rotation of the domain magnetization starts to occur. Since the stress affects the domain structure through magnetoelastic interaction, the maximum magnetostriction shows the stress dependence. The magnetoelastic interaction tends to increase the volume of the domains being either parallel to the tensile stress or perpendicular to the compressive stress. The amount of the domain wall movement, equivalently the maximum magnetostriction, is thus larger for the magnetization which is either perpendicular to the tensile stress or parallel to the compressive stress. In the experiment, the maxima of magnetostriction, up to 17 × 10−6, were measured with sufficient accuracy using semiconductor strain gauges attached to the specimens. The results support the above physical prediction. We found that the stress dependence of the maxima is almost insensitive to the relation between the rolling direction and the loading direction, which ensures that the magnetostriction method is suitable for the residual stress measurement.

Infrared thermovision of damage processes in concrete and rock

Abstract: Infrared thermography has been used as a nondestructive and noncontact technique to examine the progressive damage processes and failure mechanisms of concrete and rock specimens subjected to given static unconfined compressions and to a superimposed vibratory excitation. The parameter investigated in this paper is the heat generation due to intrinsic dissipation caused by anelasticity and/or inelasticity of the material which has been excited beyond its stable reversible limit. This useful technique allows accurate illustration of the onset of unstable crack propagation and/or flaw coalescence when increasing irreversible microcracking is activated by vibratory loading.

Quantitative estimation of material properties of porous ceramics by means of composite micromechanics and ultrasonic velocity

Abstract: This paper describes a nondestructive method for the quantitative estimation of property variations due to porosity in advanced ceramics. The method employs a composite micromechanics which accounts for the effective density and elastic stiffness of a porous composite medium with a measurement of ultrasonic velocity. When the measured velocity is coupled with the theoretically predicted velocity, the unknown pore volume fraction is solved, from which other material properties are determined. The micromechanics model based on the Mori-Tanaka theory can handle ellipsoidal pores with a certain orientation distribution. Given the zero-porosity matrix moduli and the pore aspect ratio, the oblate spheroidal theory is first applied to hot pressed silicon carbide (SiC) samples in the range of about 85–100% of theoretical density and then extended to sintered samples in the density range of 93.6–97.6%. It is shown that the bulk density and elastic modulus of porous ceramics can be estimated accurately by the proposed method.

Ultrasonic inspection of long steel pipes using Lamb waves

Abstract: An ultrasonic inspection technique using Lamb waves was evaluated to detect and determine the exact location of flaws present in long steel pipes. Since multiple modes of Lamb waves are generated due to their dispersive characteristics in the inspected pipes, the selection of a specific Lamb wave mode is very important for inspection of flaws. Experimental studies of flaw detectability with the use of each Lamb wave mode, namely, the A 0 S 0 , A 1 , and S 1 modes and their ultrasonic attenuation characteristics were conducted. The experimental results showed that the A 0 mode is the most effective for detection and exact determination of the location flaws. A lucite wedge containing a water column that generates the A 0 Lamb wave mode was developed and used in the present inspection study. It was found that the ultrasonic beam divergence after wrapping around the inspected pipe once interferes with exact determination of the location of flaws and that the maximum reflection signals are obtained when the transducer is offset axially from the straight line with the position of the flaw. The present study showed feasibilities of ultrasonic inspection with the use of Lamb waves for detection of flaws in insulated or inaccessible steel pipes.

A neural network for crack sizing trained by finite element calculations

Abstract: A neural network is combined with a self-compensating ultrasonic technique to size cracks emanating from rivet holes. The network is trained with a combination of experimental and synthetic data. A 2-D finite element (FE) method is used to generate synthetic data for scattering of ultrasound by cracks emanating from rivet holes in a thin aluminium sheet. Both the back and forward scattered fields are calculated at several positions parallel to the crack line. An experimental pulse from an ultrasonic transducer is applied as input excitation to the FE grid. The FE program generates a set of back- and forward-scattered signals for cracks of various lengths in the range; 0.25–3.00 mm with length increments of 0.25 mm. Data obtained in this way are used to train the neural network (NN) classifier that categorizes the data according to crack length. Once the network is trained, its performance is tested on self-compensated experimental data obtained by ultrasonic testing of specimens containing cracks of lengths in the range 0.50–3.00 mm with length increments of 0.50mm. The use of FE modelling to train the NN eliminates the need for a large number of experiments.

Ultrasonic through-transmission characterization of thick fibre-reinforced composites

Abstract: Thick composites pose a significant challenge to the ultrasonic nondestructive evaluation process due to the increased attenuation as well as the influence of material anisotropy. The physics of ultrasonic wave propagation in anisotropic material involves understanding such phenomena as beam skewing, material focusing/defocusing, unsymmetrical beam profiles, etc. This ultrasonic behaviour can be considered to be relatively insignificant in the nondestructive evaluation of thin anisotropic composite structures, but cannot be neglected in thick composites. In this paper, the ultrasonic characterization of thick glass-epoxy composites using an immersion through-transmission method employing a standard ultrasonic robot scanner system is discussed in detail. An inverse technique for computing material elastic constants from the acquired data is discussed. The experiment measures group velocity is a function of energy propagation angle (group angle), from which phase velocity is numerically computed as a function of phase angle. Then, the material constants are determined from phase velocity profiles using commercially available parameter identification software. Both uni-directional (five independent elastic constants) as well as cross-ply (nine independent elastic constants) were considered. Through-transmitted ultrasonic beam characteristics in thick composites are then analysed using feature extraction procedures and compared with the isotropic plexiglass structure.

The interaction of Lamb waves with delaminations in composite laminates

Abstract: The S0 Lamb mode can propagate over distances of the order of 1 m in composite laminates and so has the potential to be used in long‐range nondestructive inspection. This paper discusses the interaction of the S0 Lamb mode with delaminations. The dispersion curves and the corresponding stress and displacement mode shapes of the lower order Lamb modes are obtained analytically and the interaction of the S0 mode with delaminations at different interfaces in a composite laminate is then studied both by finite element analysis and by experiment. It is shown that the amplitude of the reflection of the S0 mode from a delamination is strongly dependent on the position of the delamination through the thickness of the laminate and that the delamination locations corresponding to the maximum and minimum reflectivity correspond to the locations of maximum and minimum shear stress across the interface in the S0 mode.

Non-contact laser ultrasonics for detecting subsurface lateral defects

Abstract: To evaluate subsurface lateral defects, non-contact and non-destructive laser ultrasonics was employed. Surface acoustic waves (SAW) were generated by a Q-switched YAG laser and monitored by a heterodyne laser interferometer. Both the source and monitoring lasers were scanned over the samples to monitor the location of defects and bond quality. Based on the fundamental works of the thin metallic foil and model specimen with an artificial subsurface slit, bond quality of the stainless foil brazed on the carbon steel was evaluated by the quantitative analysis of velocity dispersion of the Rayleigh and lowest order Lamb waves. Velocity dispersion of SAW over the well bond area well agreed with theoretical dispersion of the Rayleigh wave for multi-layer laminate, while that over the unbond area agreed with the lowest order anti-symmetric Lamb wave. Velocity dispersion of the SAW over the weak bond was calculated by inserting a slip layer which allowed displacement discontinuity across the interface. Analysis of the generalized Lamb wave was demonstrated to be promising in evaluating the bond quality as well as identifying the location of defects.

Imaging of surface-breaking concrete cracks using transient elastic waves

Abstract: This study develops an imaging method to determine the depth and inclination of a surface-breaking crack in a concrete structure. The migration method in reflection seismology is adopted to process the surface response of the concrete structure recorded in a transient elastic wave test. An image is constructed that shows the location of the crack tip. The response curves are processed further by the quadratic interpolation method to determine the location of the crack tip. Then, a solid line is superposed on the image to represent the crack. Numerical examples are presented to illustrate the effectiveness of the proposed method to display cracks of various dip angles and lengths. Two model tests verify that this method can detect real cracks successfully.

Surface temperature measurement near ambient conditions using infrared radiometers with different detection wavelength bands by applying a grey-body approximation: estimation of radiative properties for non-metal surfaces

Abstract: The visualization technique using an infrared radiometer has been widely used to estimate the two-dimensional temperature distribution on a surface. Infrared radiometry is a nondestructive remote sensing technique and is applicable to various engineering problems such as invisible flaw detection, heat transfer measurement and so on. It is important to analyse its fundamental characteristics to establish a practical technique of measuring temperature. We therefore discuss them experimentally using infrared radiometers with three detection wavelength bands. The radiative properties near ambient conditions are proposed for various non-metal surfaces, where the grey-body approximation may be applied.

Nondestructive estimation of fatigue damage for steel by Barkhausen noise analysis

Abstract: Magnetic Barkhausen noise (BHN) signals are detected when magnetized domain walls move discontinuously in a ferromagnetic material. This non-uniform motion is sensitive to the microstructures of material which vary continuously with the increase of applied stress and loading cycles. In this paper, the possibility of evaluation of residual fatigue life under cyclic loading conditions is discussed based on BHN analysis. The material used in the test is mild steel. Both static loading tests and load-controlled axial pulsating fatigue tests are performed on plain plate specimens. The BHN signals are measured for each increment of tensile load during static tests and for each interval of cyclic loading during fatigue tests. The main results obtained are as follows: (1) When the applied stress range is below the fatigue limit, the BHN signal keeps nearly constant. (2) When the applied stress range is above the fatigue limit, the ratio of the BHN peak voltage of the stress level S to 0 MPa at the Nth cycle, V ns V n 0 , varied clearly with loading cycles. (3) The procedure of residual life prediction is proposed.

Evaluation of case depth on steels by Barkhausen noise measurement

Abstract: Use of eddy currents to measure the depth of surface modified layers inferromagnetic materials has been the subject of numerous studies. This technique, however, is limited by surface permeability which considerably reduces skin depth value. The Barkhausen noise technique, used to evaluate residual stresses, has also been used to evaluate case depth in carburised steels (AISI 8620 and 9310) and ion nitrided tool steels (AISI P20, D2, and H13). A method employing digital signal processing, which includes afast Fourier transform, has been applied at frequencies rangingfrom 0 to 200 kHz. Experiments show good correlation between magnetic Barkhausen noise and hardening depth values between 100 and 1000 μm for carburised steels, and between 25 and 200 μm for nitrided steels.MST/2021

Features of ultrasonic wave propagation to identify defects in composite materials modelled by finite element method

Abstract: A two-dimensional plane strain finite element model with absorbing boundary condition has been developed to investigate the ultrasonic wave propagation in isotropic and orthotropic media. It is capable of simulating the experimental pulse echo technique to obtain A-scan data, when a short duration pulse is transmitted into a domain with or without a flaw. The flaw can either be a crack or an inclusion of different material such as a Teflon insert or a resin rich zone. After performing FFT on the A-scan data, frequency domain feature analysis is done. The study provides a guideline regarding the suitability of certain harmonics sensitive to certain types of flaw. The simulation shows other important artifacts of wave propagation such as mode conversion and scattering due to the presence of flaws.

A new stereoscopic X-ray imaging technique using a single X-ray source: Theoretical analysis

Abstract: A binocular stereoscopic X-ray imaging technique is presented which can be used for both visual inspection by human observers and also for the extraction of three-dimensional coordinate information The concept of implementing a stereoscopic X-ray imaging system to solve the problem of image interpretation is not new in itself However, what is innovative in this research is the stereoscopic imaging technique developed This is based on a rigorous design theory specifically developed for a single X-ray source and a pair of linear X-ray detector arrays

Improving the evaluation sensitivity of an ultrasonic pulse echo technique using a neural network classifier

Abstract: In this paper, the use of an artificial neural network (ANN) for classifying weak ultrasonic signals has been attempted. The limitations of using a single conventional parameter for signal detection and classification (namely peak amplitude alone) are highlighted. Use of a multi-parameter approach is suggested. The ANN used is a multi-layered, feedforward, error-backpropagation network. Results are compared with those of conventional approaches.

Simulation of eddy current signals from multiple defects

Abstract: The application of eddy current testing for structural integrity assessment of heat exchangers is a well established practice. Service induced multiple defects in heat exchanger tubes produce difficult-to-evaluate complicated impedance plane trajectories during periodic eddy current examinations. Finite element simulations carried out to analyse interactions in a two-defect system are discussed. Probable uncertainty in the interpretation of eddy current trajectories is highlighted. Systematic analysis of the results from these studies will simplify procedures for on-line evaluation of eddy current signals.

Development of an on-line evaluation system of formability in cold-rolled steel sheets using electromagnetic acoustic transducers (EMATs)

Abstract: An on-line evaluation system of r-values in cold-rolled steel sheet with EMATs has been developed. First the principle of nondestructive r-value measurement is described. Next the characteristics of a Lorentz and a magnetostrictive type EMAT are evaluated to permit an optimum sensor for on-line measurement. Based on the results, a system with a magnetostrictive type EMAT was fabricated. The capability for measuring r-values was evaluated using many samples and products in a continuous annealing line. The system was confirmed to continuously measure r-values in the production line with an accuracy of 1 σ = 0.07.

Variation of the stress dependent magnetic flux leakage signal with defect depth and flux density

Abstract: The effects of uniaxial stress on the normal (radial) component of the magnetic flux leakage (MFL) signal induced by blind-hole defects for depths of 25%, 50% and 75% of the thickness of the pipe wall were investigated with a pipe wall flux density of 1.24 T. These three defects were on the same surface as the magnetizer and sensor for the MFL signal (near side). A fourth 50% defect was on the pipe wall surface opposite the sensor (far side). Changes of as much as 47% in the MFL signal due to stresses of up to 300 MPa were observed. Increased changes in the stress dependent MFL signal were observed with increasing defect depth. Comparison of the near side and far side 50% defects indicated similar changes in the MFL pp signal as a function of stress, although the shape of the MFL signals was qualitatively different. The stress dependent MFL signal was also investigated for the near side 50% defect for pipe wall flux densities between 0.65 T and 1.24 T. A linear increase in the effects of stress on the MFL signal with increasing flux density was observed. Results demonstrated that the variation of the MFL signal with stress is primarily a bulk stress effect, although the effect of defect-induced stress concentrations upon the various MFL signals investigated could also be observed.

Determination of surface wave velocities in a prestressed anisotropic solid

Abstract: In this paper, the velocities of surface waves propagating in a prestressed anisotropic crystal are determined both theoretically and experimentally. The Barnett-Lothe's integral formalism, which is fast and efficient in determining the surface wave velocities, is extended to solve the surface wave problem of a prestressed anisotropic material. The governing equations and boundary conditions of the wave superposed on a prestressed elastic body are derived by acousto-elasticity, and the effective wave propagating constants of the finite deformed body are determined. As the effective constants are determined and utilized to replace the elastic constants in the Barnett-Lothe's integral formalism, the surface wave velocities of the prestressed anisotropic, body can be determined. In the experiment, the surface wave velocity of a magnesium oxide (MgO) single crystal with (001) orientation under compressive stress is measured. A uniaxial compression in the [100] direction is applied to the crystal, and the corresponding phase velocities of the surface wave propagating on the (001) surface are measured by the V(z) curves of a line focused scanning acoustic microscope (SAM) with a frequency 1.0 GHz. Copyright © 1996 Elsevier Science Ltd.

An infrared line scanning technique for detecting delaminations in carbon fibre tubes

Abstract: A thermal nondestructive testing method based on an infrared line scanner for cylindrical specimens is presented. A seven-layered carbon-glass fibre tube with an artificial delamination was used as a test sample. A finite difference computational model of the heating of the tube was constructed in order to simulate the measurement and evaluate the capability of detecting delaminations. The numerical results show the temperature difference caused by the delamination to be well above the practical detection limit, which was verified by the actual measurements.

Eddy current measurement of the wall thickness and conductivity of circular non-magnetic conductive tubes

Abstract: A method to determine the wall thickness and conductivity of non-magnetic tubes using eddy-current measurements is presented. The equivalent circuit impedance formulae for an infinitely long solenoid with a conductive tube inside are derived. Th is simplified theoretical model is appl led to calculate the tu be wall thickness using the measured resistance and inductance of a real long coil with a long tube inside. In order to take into account the coil's finite length, an apparent fill factor is introduced. The last parameter is determined experimentally and used as the fill factor in the theoretical model. We report experimental tests for brass and copper tubes of nominal diameters in the range 12-20 mm and nominal thickness from 0.9 to 2.5 mm. An uncertainty of the wall thickness measurements typically within 2% is obtained. Copyright (C) 1996 Elsevier Science Ltd.

Determination of thermal diffusivity distribution for three types of materials by transient thermography

Abstract: In many component applications knowledge of the two-dimensional distribution of material properties within the part is highly desirable as a quality assessment tool. The contour plotting of material properties allows identification of local regions of unacceptable variation for rejection, repair or further analysis. Transient thermography is a rapid, inexpensive, on-line, nondestructive technique that can be used to determine the two-dimensional distribution of material properties. In this paper the transient method of pulsed video thermography (PVT) is studied. Procedures for the parameter normalization and numerical method are combined with experimental data to accurately map the thermal diffusivity distribution over the entire surface of a specimen. Materials whose diffusivity ranged over three orders of magnitude were tested. The accuracy of the PVT results was evaluated by comparing local values with those obtained by standard laser flash diffusivity testing.

Integrity evaluation of SiOx film on polyethylene terapthalate by AE characterization and laser microscopy

Abstract: An attempt is made to investigate the integrity of gas barrier SiO x film on PET (polyethylene terapthalate) by AE characterization during tensile loading. As the amplitude of AE signals is very weak due to the thinness of the SiO x film and high wave attenuation of PET, a new AE monitoring system, consisting of an AE sensor with an internal amplification of 120 dB, a broadband preamplifier, a fast A/D converter and a personal computer, was used to monitor the Lamb waves emitted by the fracture and/or exfoliation of the SiO x film. Dispersion of the group velocity obtained by the wavelet transform agreed well with the theoretical group velocity of the zeroth order symmetric Lamb mode (S o ) of the PET. The S o mode Lamb wave was monitored for the specimen with SiO x films thicker than 180 nm and usef for the identification of source location. The threshold tensile strain to cause the first fracture, estimated by the developed AE monitoring, agreed well with the strain determined by a laser microscope. The estimated source amplitude of AE waves showed a fairly good relationship with the segmental crack size in the SiO x film.

Resolution improvement of ultrasonic C-scan images by deconvolution using the monostatic point-reflector spreading function (MPSF) of the transducer

Abstract: This paper describes a digital signal processing method called ‘post-focus processing’. The method reduces the transducer blurring effect and improves the resolution of ultrasonic C-scan images by simultaneously performing 3-D (2-D spatial and temporal) deconvolution using the MPSF (monostatic point-reflector spreading function) of the transducer. This method was applied under different imaging conditions, including the near-field and far-field of non-focused transducers and the defocused state of a focused transducer. In order to compare the proposed technique with other deconvolution methods, results from numerical simulations and experiments are presented. The results of numerical simulations of C-scan imaging with post-focus processing and 2-D Wiener deconvolution are reported. Experiments were conducted using post-focus processing and three known image resolution improving methods including axial-Wiener deconvolution, 2-D Wiener deconvolution and pseudo 3-D Wiener deconvolution. Results show the MPSF is a crucial blurring factor affecting the image resolution of C-scan images, and the post-focus processing method is more effective and generally applicable than the others.

Estimation of acoustic emission source waveform of fracture using a neural network

Abstract: The applicability of a neural network to acoustic emission (AE) is presented. It is shown that the shape of the simulated source waveform using piezoelectric ceramics is steplike, similar to that of mode I crack extension, and its rise-time can be varied by the resonance frequency in the thickness direction. The results imply that the simulated source can provide learning waveforms for the network. Actual AE waveforms were also acquired by conducting a tensile test of a chevron-notched graphite specimen. It was demonstrated that the appropriate source waveform associated with mode I crack extension was successfully determined by the network taught with simulated waveforms.

Detection of internal defects using phase shifting holographic interferometry

Abstract: Phase shifting holographic interferometry is shown to be a useful tool for detection of internal defects in materials by observing strain anomalies in a stressed object. The pos-t recording signal processing includes smoothing of the measured deformation by means of data polynomial fitting, a derivation process taken from computer vision edge detection algorithms, which makes use of a Gaussian derivative and a cubic spline interpolation, a second smoothing, a regular numeric derivative and final smoothing. Defects, ie weak areas of the sample, are detected in the second derivative of the sample deformation, which is proportional to the stress map of the sample surface. The second derivative provides information about the defect dimensions and its severity and location. A theoretical model forthe loading of a sample containing a defect is presented as well as results based on it. Good correlation between experiment and theory was found.