Piervincenzo Rizzo

Propagation of ultrasonic guided waves in lap-shear adhesive joints: Case of incident a0 Lamb wave

This paper deals with the propagation of ultrasonic guided waves in adhesively bonded lap-shear joints. The topic is relevant to bond inspection by ultrasonic testing. Specifically, the propagation of the lowest-order, antisymmetric a0 mode through the joint is examined. An important aspect is the mode conversion at the boundaries between the single-plate adherents and the multilayer overlap. The a0 strength of transmission is studied for three different bond states in aluminum joints, namely a fully cured adhesive bond, a poorly cured adhesive bond, and a slip bond. Theoretical predictions indicate that the dispersive behavior of the guided waves in the multilayer overlap is highly dependent on bond state. Experimental tests are conducted in lap-shear joints by a hybrid, broadband laser/air-coupled ultrasonic setup in a through-transmission configuration. The Gabor wavelet transform is employed to extract energy transmission coefficients in the 100 kHz 1.4 MHz range for the three different bond states examined. The cross-sectional mode shapes of the guided waves are shown to have a substantial role in the energy transfer through the joint.

A Nonlinear Acoustic Technique for Crack Detection in Metallic Structures

A crack detection technique based on nonlinear acoustics is investigated in this study. Acoustic waves at a chosen frequency are generated using an actuating lead zirconate titanate (PZT) transducer, and they travel through the target structure before being received by a sensing PZT wafer. Unlike an undamaged medium, a cracked medium exhibits high acoustic nonlinearity which is manifested as harmonics in the power spectrum of the received signal. Experimental results also indicate that the harmonic components increase nonlinearly in magnitude with increasing amplitude of the input signal. The proposed technique identifies the presence of cracks by looking at the two aforementioned features: harmonics and their nonlinear relationship to the input amplitude. The effectiveness of the technique has been tested on aluminum and steel specimens. The behavior of these nonlinear features as crack propagates in the steel beam has also been studied.

Reference-free damage detection by means of wavelet transform and empirical mode decomposition applied to Lamb waves

Guided ultrasonic waves are increasingly used in all those structural health monitoring applications that benefit from built-in transduction, moderately large inspection ranges, and high sensitivity to small flaws. This article describes a monitoring system based on the generation and detection of the guided ultrasonic waves from an array of sparse transducers. In a round-robin manner, ultrasonic waves are generated and measured from all possible different pairs of excitation and sensing transducers. The ultrasonic signals are then processed using continuous wavelet transform and empirical mode decomposition to extract few damage-sensitive features that enable the detection and localization of damage. With respect to most of the existing guided ultrasonic wave–based methods, the proposed approach does not require to record data from a pristine structure (baseline data), and damage is inferred by examining the selected features obtained from all the possible combinations of actuator–sensor pairs of the array. In this study, the method is validated using commercial finite element software to model the presence of 10 ultrasonic transducers bonded onto an aluminum plate. The results are promising and ongoing studies are focusing on the experimental validation and the application to other waveguides.

Defect Classification in Pipes by Neural Networks Using Multiple Guided Ultrasonic Wave Features Extracted After Wavelet Processing

This paper casts pipe inspection by ultrasonic guided waves in a feature extraction and automatic classification framework. The specific defect under investigation is a small notch cut in an ASTM-A53-F steel pipe at depths ranging from 1% to 17% of the pipe cross-sectional area. A semi-analytical finite element method is first used to model wave propagation in the pipe. In the experiment, reflection measurements are taken and six features are extracted from the discrete wavelet decomposition of the raw signals and from the Hilbert and Fourier transforms of the reconstructed signals. A six-dimensional damage index is then constructed, and it is fed to an artificial neural network that classifies the size and the location of the notch. Overall, the wavelet-based multifeature analysis demonstrates good classification performance and robustness against noise and changes in some of the operating parameters.

An unsupervised learning algorithm for fatigue crack detection in waveguides

Ultrasonic guided waves (UGWs) are a useful tool in structural health monitoring (SHM) applications that can benefit from built-in transduction, moderately large inspection ranges, and high sensitivity to small flaws. This paper describes an SHM method based on UGWs and outlier analysis devoted to the detection and quantification of fatigue cracks in structural waveguides. The method combines the advantages of UGWs with the outcomes of the discrete wavelet transform (DWT) to extract defect-sensitive features aimed at performing a multivariate diagnosis of damage. In particular, the DWT is exploited to generate a set of relevant wavelet coefficients to construct a uni-dimensional or multi-dimensional damage index vector. The vector is fed to an outlier analysis to detect anomalous structural states. The general framework presented in this paper is applied to the detection of fatigue cracks in a steel beam. The probing hardware consists of a National Instruments PXI platform that controls the generation and detection of the ultrasonic signals by means of piezoelectric transducers made of lead zirconate titanate. The effectiveness of the proposed approach to diagnose the presence of defects as small as a few per cent of the waveguide cross-sectional area is demonstrated.

Feature Extraction for Defect Detection in Strands by Guided Ultrasonic Waves

Multi-wire strands are widely used in civil structures as tensioning members in prestressed concrete, cable-stayed and suspension bridges. In recent years, various researchers including the authors have investigated guided ultrasonic waves as a tool for monitoring the health of the strands. Magnetostrictive ultrasonic transducers have been successfully used to excite and detect the guided waves in these components. This study improves the general guided-wave technique with magnetostrictive transducers for the detection and sizing of defects in strands. The improvement consists of extracting sensitive and robust features of the wave signals, and then using these features to construct a damage index (D.I.). The proposed D.I. is unaffected by accidental variations in the excitation power or by changes in the electromechanical coupling efficiency of the transducers. The specific defect under investigation is a notch cut at varying depths in a strand that is subjected to a typical operational load. Features extracted after discrete wavelet processing of the wave signals result in a D.I. that is robust against noise and is linearly related to the notch depth in a logarithmic scale. In particular, the variance of the signal’s wavelet coefficients results in the largest sensitivity to notch depth.

Semi-Supervised Multiresolution Classification Using Adaptive Graph Filtering With Application to Indirect Bridge Structural Health Monitoring

We present a multiresolution classification framework with semi-supervised learning on graphs with application to the indirect bridge structural health monitoring. Classification in real-world applications faces two main challenges: reliable features can be hard to extract and few labeled signals are available for training. We propose a novel classification framework to address these problems: we use a multiresolution framework to deal with nonstationarities in the signals and extract features in each localized time-frequency region and semi-supervised learning to train on both labeled and unlabeled signals. We further propose an adaptive graph filter for semi-supervised classification that allows for classifying unlabeled as well as unseen signals and for correcting mislabeled signals. We validate the proposed framework on indirect bridge structural health monitoring and show that it performs significantly better than previous approaches.

Ultrasonic guided waves for nondestructive evaluation/structural health monitoring of trusses

Ultrasonic guided waves (UGWs) are particularly effective in those nondestructive evaluation and structural health monitoring applications that benefit from built-in transduction, moderately large inspection ranges and high sensitivity to small flaws. This paper describes a method to detect cracks in large trusses that combines the advantages of UGWs with the extraction of defect-sensitive features to perform a multivariate diagnosis of damage. The proposed algorithm was applied to the guided waves propagating along one of the main chords of a dismantled overhead sign support structure. The probing hardware consisted of a data acquisition system that controlled the generation and detection of ultrasonic signals by means of piezoelectric transducers made of lead zirconate titanate. The effectiveness of the proposed approach to diagnose the presence of an artificial defect around the welded joint between one main chord and a diagonal member of the truss structure is explained.

Structural Health Monitoring of Immersed Structures by Means of Guided Ultrasonic Waves

Guided ultrasonic waves (GUWs) are increasingly considered in the nondestructive evaluation and structural health monitoring of engineering systems that benefit from built-in transduction, moderately large inspection ranges, and high sensitivity to small flaws. Sometimes, owing to the kind of system being inspected, a non-contact approach for the generation and detection of GUWs is desired. This article presents an initial study of the feasibility of using a hybrid laser/immersion transducer system for the detection of damage in submerged structures. A pulsed laser was used for the generation of stress waves in an aluminum plate immersed in water, which were detected by a pair of conventional immersion transducers. The detected time waveforms were processed using the joint time-frequency analysis of the Gabor wavelet transform to extract information about the velocity and the attenuation of the propagating modes. Damage was simulated by devising a rectangular notch and a small circle on the face of the plate exposed to the probing system. The study shows promising results and may pave the road toward an innovative approach to the non-contact inspection/monitoring of underwater structures.

Highly Nonlinear Waves’ Sensor Technology for Highway Infrastructures

This paper describes preliminary results towards the development of an innovative NDE/SHM scheme for material characterization and defect detection based on the generation of highly nonlinear solitary waves (HNSWs). HNSWs are stress waves that can form and travel in highly nonlinear systems (i.e. granular, layered, fibrous or porous materials) with a finite spatial dimension independent on the wave amplitude. Compared to conventional linear waves, the generation of HNSWs does not rely on the use of electronic equipment (such as an arbitrary function generator) and on the response of piezoelectric crystals or other transduction mechanism. HNSWs possess unique tunable properties that provide a complete control over tailoring: 1) the choice of the waves width (spatial size) for defects investigation, 2) the composition of the excited train of waves (i.e. number and separation of the waves used for testing), and 3) their amplitude and velocity. HNSWs are excited onto concrete samples and steel rebar. The first pilot study of this ongoing effort between Caltech and the University of Pittsburgh is presented.