Abstract: This paper gives an overview on the current status of vibration-based methods for
Structural Health Monitoring. All these methods have in common that a structural change due to a damage results in a more or less pronounced change of the dynamic behavior. The use of modal information is discussed, as well as the direct use of forced and ambient vibrations. From this information, different strategies can be deduced which depend on the type of measurement data (time/frequency domain) but also on the frequency spectrum. The incorporation of actuation and sensing devices into the structure leads to modern concepts of Smart Structural Health Monitoring. Examples from civil and aerospace engineering show the applicability of these methods.
Abstract: Many machine faults, such as local defects in bearings and gears, manifest themselves in vibration signals as a series of impulsive events. Kurtosis is a measure of the impulsiveness of a signal, and spectral kurtosis (SK) gives an indication of how the kurtosis (of a bandpass filtered signal) varies with frequency. This not only gives an indication of the frequency bands to be processed, but can also be used to generate a filter to extract the most impulsive part of a signal. The first step in calculating SK is to perform a time/frequency decomposition of the signal, and then calculate the kurtosis for each frequency line. The paper compares the original STFT (short time Fourier transform) with wavelet analysis for the time/frequency decomposition, and for determining the optimum combination of centre frequency and bandwidth for maximizing the SK. The paper also describes how the SK can be enhanced by “prewhitening” the signal using an autoregressive (AR) model, this sometimes revealing an incipient fault at a much earlier stage.
Abstract: This paper addresses the fundamentals of the acoustic emission effect associated with fatigue and stress corrosion cracking in metals. It considers the microstructure of cracks and the magnitude of the different types of physical event that can occur at the crack tip during plastic deformation and stable crack growth. Expressions are given for the threshold plastic zone size ‘Dl’ at which local fracture instability occurs and the stress-wave displacement amplitude as a function of distance ‘ui(r)’ for the different wave types ‘i’ produced during crack extension. Dispersion of the stress-wave and its convolution into an electrical burst signal at the sensor is considered together with the choice of appropriate sensing frequency. A methodology is described for correcting the measured signal amplitude for attenuation in the structure and for determining the maximum sensor spacing for the detection and location of events of a specified magnitude ‘Mae’ similar to the Richter scale. Case studies are presented to illustrate the extensive database now available on acoustic emission from crack growth in metallic structures and the technical and commercial benefits to be gained from an acoustic emission based inspection strategy. The applications considered are:
• Fatigue crack growth in the node joints of offshore structures,
• Stress corrosion cracking in platform flow lines.
Abstract: Structural damage detection and monitoring is one of the major maintenance activities in transportation, processing and civil engineering. Current procedures are based on scheduled inspections which are often time/labour consuming and expensive. Guided ultrasonic waves offer the ability of inspecting large structures with a small number of transducers. Recent developments in smart sensor technologies allow for integration of these transducers with monitored structures. This is associated with a new design philosophy leading to more efficient and economically attractive structures. The paper briefly discusses various damage detection methods based on structural, ultrasonic and guided ultrasonic waves. The focus is on recent research advances in damage monitoring techniques, smart sensor technologies and signal processing.
Abstract: An improved method to identify the crack location and size is presented which takes advantages of wavelet finite element (WFE). The important property of wavelet analysis is the capability to represent functions in a dynamic multiscale manner, so solution with WFE enables a hierarchical approximation to the exact solution. WFE has good ability in modal analysis for singularity problems like a cracked beam. The crack in a beam is modeled with WFE and represented as a rotational spring. The additional flexibility caused by crack in its vicinity is evaluated according to linear and elastic fracture mechanics theory. The WFE stiffness matrix of the crack is constructed and the algorithm for crack identification through the use of vibration-based inspection (VBI) is developed. With the accurate natural frequencies obtained from the transient signal measured, graphs of crack equivalent stiffness versus crack location are plotted, by providing the first three natural frequencies as an input. The intersection of the three curves gives the crack location and size. Experimental studies of cracked shafts are presented to demonstrate the accuracy of the method. The error in identification of crack location and size are both less than 2%. This study provides the new method for the diagnosis of incipient small crack.
Abstract: In case of mechanical system health monitoring, a need to develop normal-knowledge
based novelty detection techniques is increasing. The negative selection algorithm, which is inspired from the operation mechanism of human immune system, is one of such approaches. Our approach is to apply the idea for the anomaly detection in the vibration time series of the rotor system. A real-valued negative selection algorithm based on Euclidean distance, as well as cosine similarity, has been implemented. By means of adding the corresponding coverage radius to each antibody elements, the detection efficiency of each antibody element is increased. The detection efficiency is evaluated with simulated data as well as vibration signal sampled from one rotor system. The results indicate that the algorithm can efficiently detect the anomaly in time series data. Moreover, the number of detectors in antibody set is less enough for potential application in online signal monitoring.
Abstract: Machine vibration signal has been used in fault detection and diagnosis. Modulation and non-stationarity existing in the signal generated by a faulty gearbox present challenges to effective fault detection. Hilbert transform has the ability to address the modulation issue. This paper outlines a novel fault detection method called Hilbert & TT-transform (HTT-transform) which combines Hilbert transform and TT-transform obtained from the inverse Fourier transform of the S-transform. The principle of the proposed method is to analyze the modulating signal created by a faulty gear using a time-time representation. The method has the advantage of providing a new way of localizing the time features of the modulating signal around a particular point on the time axis through scaled windows. It is verified with simulated signals and real gearbox vibration signals. The results obtained by CWT, S-transform, TT- transform, and HTT-transform are compared. They show that utilizing the proposed method can improve the effectiveness of gearbox fault detection.
Abstract: The demodulation analysis has been extensively used for gear diagnosis. However these techniques mainly deal with the amplitude-modulated signal instead of the frequency-modulated signal. Due to the symmetrical phase relationship of the sidebands, the amplitude-demodulated methods are not suitable for the frequency-modulated signal. This paper introduces the theory of cyclostationary processes as a powerful frequency-demodulation tool for the diagnosis of gears. The Cyclic Autocorrelation Function (CAF) is an important second-order cyclic statistics and acts as an efficient parameter to the frequency-demodulated analysis. In this paper, the CAF of frequency-modulated signal is deduced carefully. Through the discussion of frequency feature of the CAF slice at different cyclic frequency, two useful conclusions have been arrived about the frequency-demodulation. Firstly, the CAF slice at even multiples of the modulator-frequency can demodulate the frequency-modulated signal directly. Secondly, the amplitude-demodulated methods are suitable for the CAF slice of frequency-modulated signal at some special cyclic frequencies, which are equal to odd multiples of the modulator-frequency or close to the double carrier-frequency. These features of the CAF slice mentioned above overcome the invalidation of amplitude-demodulated methods for the frequency-modulated signal and increase it’s application range in engineering. Application in simulated and experimental data from a gear rig verifies the effectiveness of the frequency-demodulated method based on cyclostationarity.
Abstract: Vibration signals acquired from a gearbox usually are complex, and it is difficult to
detect the symptoms of an inherent fault in a gearbox. In this paper, an adaptive redundant second generation wavelet (ARSGW) based on second generation wavelet (SGW) is developed. It adopts data-based optimization algorithm to design the initial prediction operator and update operator at each scale. The initial operators are interpolated with zero, and then the redundant prediction operator and update operator are obtained. The splitting step in ARSGW is removed, the approximation signal at each scale is predicted and updated with redundant prediction operator and update operator directly, and the length of approximation signal and detail signal at every scale remains the same, ARSGW eliminates translation variance of SGW. Since the redundant prediction operator and update operator lock on to the dominant structure of the signal, ARSGW can well reveal the characteristics of the signal in time domain. ARSGW is found to be very effective in detection of symptoms from the vibration signal of a large air compressor gearbox with impact rub fault. SGW is also used to analyze the same signal for comparison, no modulation signals and periodic impulses appear at any scale.
Abstract: Two alternative software tools for damage identification are presented. The first tool,
developed on the basis of the Virtual Distortion Method (VDM), takes advantage of an analytical formulation of the damage identification problem. Consequently, gradient-based optimization method is applied to solve the resulting dynamic inverse problem in time domain. Finite element model of the structure is necessary for the VDM approach. The second tool utilizes the Case-Based Reasoning (CBR) for damage identification. This method consists in i) extracting principal features of the response signal by wavelet transform, ii) creating a base of representative damage cases, iii) organizing and training the base by neural networks, and finally iv) retrieving and adapting a new case (possible damage) by similarity criteria. Basic description of both approaches is given. A comparison of numerical effectiveness, in terms of accuracy and computational time, is provided for a simple beam structure. Advantages and weaknesses of each approach are highlighted.