Skip to main content
main-content
Top

About this book

This book comprises the selected contributions from the 2nd World Congress on Condition Monitoring (WCCM 2019), held in Singapore in December 2019. The contents focus on digitalisation for condition monitoring with the emergence of the fourth industrial revolution (Industry 4.0) and the Industrial Internet-of-Things (IIoT). The book covers latest research findings in the areas of condition monitoring, structural health monitoring, and non-destructive testing which are relevant for many sectors including aerospace, automotive, civil, oil and gas, marine, and manufacturing industries. Different monitoring systems and non-destructive testing methods are discussed to avoid failures, increase lifespans, and reduce maintenance costs of equipment and machinery. The broad scope of the contents will make this book interesting for academics and professionals working in the areas of non-destructive evaluation and condition monitoring.

Table of Contents

Frontmatter

Signal Processing: Fault Diagnosis for Complex System Monitoring

Frontmatter

Detection and Localization of a Gear Fault Using Automatic Continuous Monitoring of the Modulation Functions

In the context of automatic and preventive condition monitoring of rotating machines, this paper presents a case study of a naturally-worn parallel straight gear by monitoring the evolution of the modulation functions. The Hilbert demodulation is automatically performed considering only the frequency content of the signals detected by the AStrion software. The gear has been worn over 3000 h with a constant axial load. A particular focus is set on the amplitude modulation function in order to assess its efficiency to characterize both the severity of the wear and the most worn part of the gear. The results are confronted with on-site observation of the teeth. For this purpose, the evolution of both amplitude and phase modulations over several meshing harmonics are compared, as well as demodulation on both original and residual signals. Indicators to automatically classify the wear are discussed.

X. Laval, N. Martin, P. Bellemain, C. Mailhes

Multi-harmonic Demodulation for Instantaneous Speed Estimation Using Maximum Likelihood Harmonic Weighting

This paper details a novel way to estimate the instantaneous angular speed of a rotating shaft directly from its vibration signal measurement. Speed estimation through single-harmonic phase demodulation has proven itself already multiple times in the past to be a fairly reliable way to obtain accurate instantaneous angular speed estimations. However, if the chosen harmonic experiences any disturbances such as crossing orders or noise increases, the end result can be severely compromised. The technique in this paper utilizes the phase of multiple harmonics in the signal simultaneously to obtain an estimate of the instantaneous speed. The novelty lies in the simultaneous phase demodulation formulation and in the way the phases of the harmonics are combined using a maximum likelihood estimation approach. To examine the robustness of this technique, this paper investigates the potential benefits and hindrances of using a multi-harmonic demodulation approach as compared to a single-harmonic approach. Simulation and experimental results show that a significant increase in estimation accuracy can be gained by employing more information contained within the phases of the harmonics in the vibration signal.

C. Peeters, J. Antoni, J. Helsen

A Variational Bayesian Approach for Order Tracking in Rotating Machines

Order tracking is important in rotating machine signal processing. It consists in tracking the complex envelope of phase varying sinusoids associated with distinct rotating mechanical elements. The resulting magnitudes and phases are usually used for vibration level control and system resonance identification. A common approach to tackle this issue is the Vold-Kalman filter. However, the performance of this filter is highly conditioned by the choice of some tuning parameters that control its selectivity. In this paper, a new order tracking method is proposed to jointly estimate the targeted envelopes and the optimal values of these parameters in an automatic manner. To do so, the order tracking problem is formulated within the Bayesian framework. In details, prior distributions are affected to the target variables (including the tuning parameters and the noise variance); posterior distributions are then derived and estimated using a bariational Bayesian approximation algorithm. Results on simulated and real helicopter vibration data demonstrate the effectiveness of the proposed approach.

Y. Marnissi, D. Abboud, M. Elbadaoui

Reconstructing Speed from Vibration Response to Detect Wind Turbine Faults

Speed information is often necessary to detect certain failure modes in wind turbines, such as unbalance, looseness, misalignment, and gear related problems. This is normally obtained from tacho sensors targeting certain shafts, e.g. generator shaft or main shaft. Tacho pulses can be used to track rotational speed of certain components as well as to perform angular resampling. Most importantly, knowing machine speed is necessary for alarming speed dependent failure modes. Correct speed information may be unavailable due to broken tacho sensor and incorrect setting of the sensor. Consequently, many failure modes cannot be detected when the information is not available. Unable to detect these failure modes early may cause unexpected secondary damages that leads to unexpected downtime of the wind turbines and loss of production. Maintenance work on correcting speed sensor is costly, especially for offshore wind turbines, and may result in downtime. Furthermore, condition monitoring system (CMS) of wind turbines is significantly cheaper without speed sensors; considering a wind farm normally consists of many wind turbines. Therefore, it is important to acquire speed information without the need for having a tacho sensor. This paper presents an industrial implementation of speed reconstruction based on vibration response to detect speed-related faults, such as unbalance and gear faults. Speed is determined by frequency demodulation of response vibration signals. Hilbert transform is used to extract the instantaneous frequency, where the derivative of the analytic signal is done in the frequency domain. Using data from wind turbines monitored by Brüel & Kjær Vibro, we show this method can detect speed-related faults as early as when the speed information is available through real speed sensor. The method was able to detect unbalance, misalignment and looseness on wind turbine generators, as well as gear-related problems on high speed stage, intermediate stage, and planetary stage gearboxes. This offers the possibility of reducing the cost of wind turbines CMS significantly.

K. Marhadi, D. Saputra

Non-regular Sampling and Compressive Sensing for Gearbox Monitoring

Vibration-based condition monitoring is a powerful field to achieve preventive maintenance and control the efficiency of mechanical systems. The common monitoring scheme consists of capturing vibration signals at a sufficiently high sampling frequency to consider the dynamic and kinematic properties of the machine, and then condition indicators are constructed and saved or sent to a ground-based surveillance station. A major drawback of such an approach is that these indicators may not be sufficient to establish an accurate diagnosis, and the signal itself is needed in various situations. As the frequency rate of vibration signals and, consequently, their size highly depends on the highest frequency to be monitored, it is generally set high in order to respect the Nyquist theorem. As a consequence, sending these data can be compromised because of the limited bandwidth of the transmission channel. This paper proposes a non-regular sampling strategy via compressive sensing to significantly reduce the sampling frequency (below the Nyquist limit) and the data size, while preserving enough information for diagnosis. It is based on the hypothesis that the measured signal consists of a sum of sinusoids corrupted by some random Gaussian noises. The signal recovery problem from random samples is then formulated through an $${{\ell }}_{{{2}}} - {{\ell }}_{{{1}}}$$ ℓ 2 - ℓ 1 optimization problem, which is solved via the iterative shrinking-thresholding algorithm. At last, the potentiality of the suggested approach is demonstrated on real vibration signals measured from a spur gearbox with a spalling fault that progresses over 12 days.

C. Parellier, T. Bovkun, T. Denimal, J. Gehin, D. Abboud, Y. Marnissi

Change Detection in Smart Grids Using Dynamic Mixtures of t-Distributions

The analysis of massive amounts of data collected via smart grids can contribute to manage more efficiently energy production and water resources. Within this framework, a method is proposed in this article to detect changes in panel data from smart electricity and water networks. The adopted approach is based on a representation of the data by clusters whose occurrence probability varies in the course of time. Because of the cyclical nature of the studied data, these probabilities are designed to be piecewise periodic functions whose break points reflect the intended changes. A mixture of t-distributions with a hierarchical structure forms the basis of our proposal. This model was also chosen for its robustness properties. The weights of the mixture, at the top of the hierarchy, constitute the change detectors. At the lowest hierarchy level, the mixture weights are designed to model the periodic dynamics of the data. The incremental strategy adopted for parameter estimation makes it possible to process large amounts of data. The practical relevance of the proposed method is illustrated through realistic urban data.

A. Samé, M. L. Abadi, L. Oukhellou

Unbalance Detection and Prediction in Induction Machine Using Recursive PCA and Wiener Process

This paper focuses on the detection of unbalance in induction machines and the prediction of its dynamic evolution based on the analysis of the current signals. The proposed method is based on the combination of recursive Principal Component Analysis (PCA) and a Wiener process. The recursive PCA is processed on the different features computed from the current signals in order to choose the most relevant ones. This way the fault detection is processed with the only knowledge of the healthy state of the machine. A linear Wiener process is then used to model the behavior of PCA components and predict their evolution over time in order to estimate the dynamic of the tracked fault along with the remaining useful life (RUL). The proposed method is applied on real data from a 5.5 kW induction machine with three different levels of unbalance and obtains very promising results.

A. Picot, J. Régnier, P. Maussion

Automatic Fault Diagnosis in Wind Turbine Applications

A wind turbine may have any—out of hundreds—types of bearing on its generator or other components. The bearing type is often unknown until a possible defect occurs or until a diagnostic engineer performs a diagnosis. We identify faults on different wind turbine components by matching their fault frequencies with frequency harmonics and sidebands found in the spectrum analysis. It is crucial to know the fault frequencies, such as ball pass inner and outer race defect frequencies, before the diagnosis. Lack of expertise and experience in identifying bearing type and frequency may result in an incorrect diagnosis. Moreover, not knowing the bearing frequency leads to the inability to track bearing fault development continuously. In this paper, we propose automatic diagnosis, a novel method to identify prominent peaks above carpet level in power spectrum analysis, and then label them for potential faults. Prominent peaks are noticeable spectral peaks above the noise floor, and we calculate the noise floor using median filtering. Once we extract the peaks, we implement priority labeling. The fittest bearing frequency is intelligently estimated using approximate peak matching and envelope of bandpass segment. We use the estimated frequency to identify, quantify, and track early and late bearing faults without knowing the bearing type. Using data from wind turbines monitored by Brüel & Kjær Vibro, we show that automatic diagnosis accurately detects faults, such as ball pass outer race fault, ball spin frequency fault, generator imbalance, and gear fault. We expect automatic diagnosis to reduce manual identification of fault frequencies. Implementing automatic diagnosis could significantly reduce the cost of simultaneously monitoring the condition of hundreds, or even thousands, of wind turbines.

D. Saputra, K. Marhadi

Blade Monitoring in Turbomachines Using Strain Measurements

Blade rubbing is an undesirable but unavoidable phenomenon in turbomachinaries. It decreases the engine performance and can lead to blades or/and stator damages. Thus, blade monitoring is of high importance not only to ensure security, but also to optimize the engine efficiency. In particular, the detection and the identification of the instants of contacts between the blades and the stator are a valuable information that can be used through a preventive maintenance scheme. This paper proposes a new detection approach based on strain gauge measurements. Since the rubbing phenomenon creates a wave shock that deform the structure, these latter can be efficiently measured by a strain gauge mounted on the stator. However, the strain signal is generally corrupted with other mechanical and electromagnetic interferences. To tackle this issue, the authors propose a signal separation technique based on modeling the blades’ signal as cyclo-non-stationary. Also, the use of the synchronous fitting is proposed to separate blades’ component from environmental noise. The proposed technique is validated on real signals captured from a strain gauge mounted on the stator of a turbomachine.

D. Abboud, M. Elbadaoui, N. Tableau

Measuring the Effectiveness of Descriptors to Indicate Faults in Wind Turbines

Over the years, numerous descriptors have been proposed to indicate the health and various failure modes of wind turbine components. Description and mathematical interpretation of descriptors, such as kurtosis, crest factor, peak to peak, mesh, and residual RMS, have been shown in several works to detect bearing and gear faults. However, there are only a few works that evaluate the effectiveness of various descriptors in detecting faults. This paper proposes Descriptor Ranking, a novel machine learning application that measures descriptors’ effectiveness to indicate different failure modes accurately. The proposed application uses different feature selection algorithms to rank the importance of the descriptors. Results show that Descriptor Ranking can correctly rank the importance of any calculable descriptors. Also, we evaluate the performance of different feature selection algorithms, such as random forest regressor and logistic regression. Given a balanced set of vibration data with apparent faults and no faults, Descriptor Ranking visualizes different descriptors’ effectifeness ranks, including newly proposed descriptors, in detecting different failure modes. Knowing the effectiveness of a descriptor in indicating specific failure modes is essential to optimize the condition monitoring strategy of wind turbines or any other machinery. Having multiple descriptors to identify the same failure mode can be avoided. Hence, we can reduce false alarms, and faults can be detected more accurately. With fewer false alarms, monitoring the condition of many wind turbines is going to be more efficient.

D. Saputra, K. Marhadi

Gearbox Condition Monitoring Using Sparse Filtering and Parameterized Time–Frequency Analysis

It is challenging to monitor the condition of a gearbox being operating under variable conditions. Considering that numerous variable conditions have the non-stationarity nature, a time–frequency analysis (TFA) approach is very suited to analyze the frequency components of interest. As an emerging TFA method, parameterized TFA utilizes further parameters to parameterize kernel functions, so a complex signal can be characterized more accurately. However, it is an issue to estimate the instantaneous frequency with the interference of background noise. This paper presents a method to solve this problem by using the sparse filtering method, as the sparse filtering is capable of enhancing a particular desired feature to achieve the desired effect. With the sparse filtering, the parameterized TFA is used to achieve signal decomposition and clarify the signal spectrum structure. Finally, these interferential frequencies are extracted and removed, and the processed signals are used to monitor the gearbox. The availability of this method was verified by simulated and experimental signal analysis in gearbox condition monitoring.

Shiyang Wang, Zhen Liu, Qingbo He

Rotating Machinery Fault Diagnosis with Weighted Variational Manifold Learning

The transient characteristics attributed to partial fault of rotating machinery are of technical evidences for health diagnosis, which are unavoidably overwhelmed by much noise disturbance in complex modulation bands. Usually, only the optimal component with a selected scale is filtrated and then extracted to be the evidence for fault diagnosis, which may ignore some important features distributed in the other scales. Focusing on the feature distribution in a whole-band, this paper proposed a novel weighted variational manifold learning (WVML) for transient features reinforcement and fault diagnosis in the view of multi-scale analysis. First, a series of intrinsic mode functions, containing different modulated frequency information, are automatically obtained via an improved variational mode decomposition (IVMD) approach, where a signal difference average criterion is introduced to determine the appropriate decomposition level. Time–frequency manifold (TFM) learning is referred to then sequentially mine their intrinsic structures in the corresponding time–frequency domain. Therewith, the reinforced signal for each variational mode can be achieved via phase preserve and a series of inverse transforms in a suppression of the in-band noise. Combining the reinforced mode signals with different weight coefficients, the reconstructed waveform can be finally synthesized. In these manner, the desired features of all scales can be soundly reinforced with mode manifolds mined and dynamically weighted in a self-learning way. The consequence of proposed WVML scheme is verified through experimental gear defective signal and result reveals this method yields an excellent accuracy in health diagnosis of rotating machinery.

Quanchang Li, Xiaoxi Ding, Wenbin Huang, Yimin Shao

Recent Advances on Anomaly Detection for Industrial Applications

Frontmatter

A Data-Driven Approach to Anomaly Detection and Health Monitoring for Artificial Satellites

In the operation of artificial satellites, overall health monitoring of the system and detection of symptoms of potential troubles are very important. For about twenty years, we have studied the data-driven health monitoring methods for artificial satellites, in which machine learning techniques are applied to the house-keeping data of satellites. In this paper, we review the basic of anomaly detection based on unsupervised learning and some results in early studies and then discuss the lessons learned and challenges to be tackled in the data-driven health monitoring for artificial satellites.

Takehisa Yairi, Yusuke Fukushima, Chun Fui Liew, Yuki Sakai, Yukihito Yamaguchi

Robust Linear Regression and Anomaly Detection in the Presence of Poisson Noise Using Expectation-Propagation

This paper presents a family of approximate Bayesian methods for joint anomaly detection and linear regression in the presence of non-Gaussian noise. Robust anomaly detection using non-convex sparsity-promoting regularization terms is generally challenging, in particular when additional uncertainty measures about the estimation process are needed, e.g., posterior probabilities of anomaly presence. The problem becomes even more challenging in the presence of non-Gaussian, (e.g., Poisson distributed), additional constraints on the regression coefficients (e.g., positivity) and when the anomalies present complex structures (e.g., structured sparsity). Uncertainty quantification is classically addressed using Bayesian methods. Specifically, Monte Carlo methods are the preferred tools to handle complex models. Unfortunately, such simulation methods suffer from a significant computational cost and are thus not scalable for fast inference in high dimensional problems. In this paper, we thus propose fast alternatives based on Expectation-Propagation (EP) methods, which aim at approximating complex distributions by more tractable models to simplify the inference process. The main problem addressed in this paper is linear regression and (sparse) anomaly detection in the presence of noisy measurements. The aim of this paper is to demonstrate the potential benefits and assess the performance of such EP-based methods. The results obtained illustrate that approximate methods can provide satisfactory results with a reasonable computational cost. It is important to note that the proposed methods are sufficiently generic to be used in other applications involving condition monitoring.

Yoann Altmann, Dan Yao, Stephen McLaughlin, Mike E. Davies

Spacecraft Health Monitoring Using a Weighted Sparse Decomposition

In space operations, spacecraft health monitoring and failure prevention are major issues. This important task can be handled by monitoring housekeeping telemetry time series using anomaly detection (AD) techniques. The success of machine learning methods makes them attractive for AD in telemetry via a semi-supervised learning. Semi-supervised learning consists of learning a reference model from past telemetry acquired without anomalies in the so-called learning step. In a second step referred to as test step, most recent telemetry time-series are compared to this reference model in order to detect potential anomalies. This paper presents an extension of an existing AD method based on a sparse decomposition of test signals on a dictionary of normal patterns. The proposed method has the advantage of accounting for possible relationships between different telemetry parameters and can integrate external information via appropriate weights that allow detection performance to be improved. After recalling the main steps of an existing AD method based on a sparse decomposition Pilastre et al (Sign Proc, 2019 [1]) for multivariate telemetry data, we investigate a weighted version of this method referred to as W-ADDICT that allows external information to be included in the detection step. Some representative results obtained using an anomaly dataset composed of actual anomalies that occurred on several satellites show the interest of the proposed weighting strategy using external information obtained from the correlation coefficient between the tested data and its decomposition on the dictionary.

Barbara Pilastre, Jean-Yves Tourneret, Loïc Boussouf, Stéphane D’escrivan

When Anomalies Meet Aero-elasticity: An Aerospace Industry Illustration of Fault Detection Challenges

This paper aims at describing some industrial aerospace applications requiring anomaly detection. After introducing the industrial context and needs, the paper investigates two examples of oscillation/vibration detection associated with aero-elastic phenomena: (i) the first example related to aircraft structural design optimization is characterized by very stringent constraints on detection time; (ii) the second example linked to equipment condition monitoring has more relaxed constraints in terms of detection requirements.

P. Goupil, J.-Y. Tourneret, S. Urbano, E. Chaumette

Anomaly Detection with Discrete Minimax Classifier for Imbalanced Dataset or Uncertain Class Proportions

Supervised classification is becoming a useful tool in condition monitoring. For example, from a set of measured features (temperature, vibration, etc.), we might be interested in predicting machine failures. Given some costs of class misclassifications and a set of labeled learning samples, this task aims to fit a decision rule by minimizing the empirical risk of misclassifications. However, learning a classifier when the class proportions of the training set are uncertain, and might differ from the unknown state of nature, may increase the misclassification risk when classifying some test samples. This drawback can also occur when dealing with imbalanced datasets, which is common in condition monitoring. To make a decision rule robust with respect to the class proportions, a solution is to learn a decision rule which minimizes the maximum class-conditional risk. Such a decision rule is called a minimax classifier. This paper studies the minimax classifier for classifying discrete or discretized features between several classes. Our algorithm is applied to a real condition monitoring database.

Cyprien Gilet, Lionel Fillatre

A Comparative Study of Statistical-Based Change Detection Methods for Multidimensional and Multitemporal SAR Images

This paper addresses the problem of activity monitoring through change detection in a time series of multidimensional Synthetic Aperture Radar (SAR) images. Thanks to SAR sensors’ all-weather and all-illumination acquisitions capabilities, this technology has become widely popular in recent decades when it concerns the monitoring of large areas. As a consequence, a plethora of methodologies to process the increasing amount of data has emerged. In order to present a clear picture of available techniques from a practical standpoint, the current paper aims at presenting an overview of statistical-based methodologies which are adapted to the processing of noisy and multidimensional data obtained from the latest generation of sensors. To tackle the various big data challenges, namely the problems of missing data, outliers/corrupted data, hetergenenous data, robust alternatives are studied in the statistics and signal processing community. In peculiar, we investigate the use of advanced robust approaches considering non-Gaussian modeling which appear to be better suited to handle high-resolution heterogeneous images. To illustrate the attractiveness of the different methodologies presented, a comparative study on real high-resolution data has been realized. From this study, it appears that robust methodologies enjoy better detection performance through a complexity trade-off with regards to other non-robust alternatives.

Ammar Mian, Frédéric Pascal

Bearing Anomaly Detection Based on Generative Adversarial Network

It is challenging to detect incipient fault of a bearing in a rotating machine being operated under complex conditions. Current intelligent fault diagnosis models require massive historical data of bearing monitoring signals under different health states and the corresponding state labels. However, in some cases, only the samples when the bearing is under healthy condition are available. To deal with the scenario when anomalous samples are absent, this paper proposes a bearing anomaly detection method based on generative adversarial network (GAN). Specifically, only the normal samples are used to train the proposed GAN model, by which the signal data distribution of normal samples is learned. Then, the newly acquired signals with unknown health states are taken as the input data of the trained model. Residual loss of each new sample is calculated by comparing the dissimilarity between the generated sample and the input sample. Because the trained model is optimized by the normal samples, the residual losses of the anomalous samples will be large due to the different data distribution from that of the normal samples. Finally, the bearing anomaly is detected when the residual loss of newly acquired sample increase quickly during continuous inspection of the bearing health condition. Life-circle vibration signals of a bearing measured from a run-to-failure test are used to validate the proposed method. The result indicates that the bearing anomaly can be detected earlier by the proposed GAN model than the traditional method based on signal time-domain statistical criterion, which is of great significance for incipient weak fault detection of mechanical systems.

Jun Dai, Jun Wang, Weiguo Huang, Changqing Shen, Juanjuan Shi, Xingxing Jiang, Zhongkui Zhu

Classification/Analysis for Fault or Wear Diagnosis

Frontmatter

Specific Small Digital-Twin for Turbofan Engines

Ten years ago, we develop a numeric state vector to allow us following the performance of a small fleet of turbofan engines. As engine manufacturer with “flight by the hour” new contracts going, we needed to improve our technology; hence, building analytic digital-twin solutions. This improvement allows to monitor the engines as well as the behavior of the airlines. It becomes possible to identify use classes to help design systems better adapted to our clients. The main difficulty to upgrade this technology was to manage huge amount of data. At the time, the only observations we had were small 4 Kb snapshots broadcasted by satellite link. And even with such small amount of observations, the mathematics were limited to an evolving buffer of one year and a half of flights for a ten aircrafts small company. Today, we download full flight temporal measurements reaching more than 2 Gb per flight and we want to manage our fleet of more than 35,000 engines. Moreover, the life duration of the engine parts increases and the first shop visit is now more than three years away. The solution was to use new cluster technology and adapt algorithms. A main element is to be able to reduce the observation dimension hence allowing mathematic computation, but specifically for each class of behavior. The compression algorithm is now running an auto-encoder neural network on GPU and the cluster analysis still uses a self-organizing map but deployed in parallel on a cluster of computers. We propose to present this new improvement and conclude with our view about operational digital-twins.

J. Lacaille

Fault Diagnosis of Single-Stage Bevel Gearbox by Energy Operator and J48 Algorithm

The main objective of this paper is to detect and classify the faults in a single-stage bevel gearbox. In this paper, energy operator (EO) is combined with the spectrum and cepstrum analysis is used for detection of faults and for classification purpose, J48 algorithm is used. Usually, the vibration signal is contaminated with noises which suppress the fault frequencies in spectrum analysis. So, a filter-free, simple, and efficient EO method is used to reduce noise and interferences of vibration signal coming from the gearbox test rig. All the experiments are carried out by varying the speed and load under three health conditions of gearbox such as healthy, chipped, and missing tooth. The vibration signal is preprocessed by EO, which is further analyzed by spectrum and cepstrum analysis in MATLAB to detect the location of faults. J48 decision tree algorithm is used to select the dominant feature among 12 statistical features which are extracted from the processed signal and based on these chosen features; the overall classification accuracy is approx. 89% with a precision of 0.917. The obtained results show that the presented method can detect and classify the faults of gearbox efficiently.

V. Kumar, A. K. Verma, S. Sarangi

An Efficient Multi-scale-Based Multi-fractal Analysis Method to Extract Weak Signals for Gearbox Fault Diagnosis

Weak fault signals are always embedded in mass vibration noise in many gear systems, thus making difficulty in gear fault diagnosis. In order to extract weak fault signals, a new multi-scale-based multi-fractal analysis (MMA) method is introduced in this paper, which is based on classical multi-fractal detrended fluctuation analysis (MFDFA) framework. Firstly, the Hurst surface features are utilized to describe the characteristics with multifractal of the vibration signal, which have been proved to be sensitive to the dynamical responses of the various gear faults. Secondly, a moving fitting window is added to the MFDFA framework to sweep through all the range of the scales, and then obtain final multi-scale features, whose purpose is to magnify those features in some important scales and weaken the rest scales. In addition, other techniques, such as the distance-based feature selection and the random forest (RF) classifier, are also introduced into the gearbox fault diagnosis procedure to verify the effectiveness of extracted features for differentiating various gear states. Experiments using the Qianpeng testbed (QT) prove that the MMA method can effectively extract weak signals, and has higher diagnostic accuracy than other algorithms, such as empirical mode decomposition (EMD), wavelet transform (WT), and classical MFDFA.

Ruqiang Yan, Fei Shen, Hongxing Tao

A Novel Method of PEM Fuel Cell Fault Diagnosis Based on Signal-to-Image Conversion

This paper proposes an image processing-based approach for the fault diagnosis of polymer electrolyte membrane (PEM) fuel cells. As more abundant information is contained in the image than that that in 1D signal, features representing PEM fuel cell faults could be better highlighted with the image. Experimental data from a PEM fuel cell system at different states, including flooding and dehydration scenarios, is used to validate the proposed method. By converting the PEM fuel cell voltage signal into a 2D grey image, several features are extracted from the image, their performance in discriminating different PEM fuel cell states is investigated, and two optimal features are determined for fault diagnosis. Moreover, the diagnostic performance of optimal features from grey image is compared with features from PEM fuel cell voltage. Results demonstrate that better diagnostic performance could be obtained with the proposed method.

Zhongyong Liu, Weitao Pan, Yousif Yahia Ahmed Abuker, Lei Mao

Fault Classification of Polymer Electrolyte Membrane Fuel Cell System Based on Empirical Mode Decomposition

This paper investigates the empirical mode decomposition (EMD)-based feature selection technique in identifying polymer electrolyte membrane fuel cell water management issue. In the study, EMD is applied to fuel cell voltage signals at various states, including fuel cell flooding and membrane drying out, from which intrinsic mode functions (IMFs) are obtained. Sensitivity index is used to determine the IMFs more sensitive to fuel cell fault, which are then used for the fault diagnosis. The effectiveness of the selected IMFs in discriminating flooding and drying out scenarios is validated using experimental data from a PEM fuel cell system. The results show that different faults can be distinguished effectively. From the findings, a generalized feature extraction and selection technique can be provided for fault diagnosis in practical PEM fuel cell applications.

Yousif Yahia Ahmed Abuker, Zhongyoug Liu, Lisa Jackson, Lei Mao

Fleet-Oriented Pattern Mining Combined with Time Series Signature Extraction for Understanding of Wind Farm Response to Storm Conditions

Offshore wind turbine installations are rapidly spreading around Europe and all over the world. These turbines are typically installed in large wind farms combining turbines of the same type. Farm owners target maximal performance of the farm in general and particularly predictability of behaviour. The latter is getting increasingly important since offshore wind farms are being managed more and more as conventional power plants driven by the electricity market supply and demand considerations. The context of zero subsidy farms exposes farm operators to fluctuations in electricity market prices. As such, deep understanding of farm behaviour is essential to come up with a good strategy to deal with these fluctuations. This paper focusses on the automated extraction of farm-wide response to storm conditions. The input data for the analysis are status logs and SCADA 1-second data. The status logs record the important turbine controller events. Typically, they consist of a number, a time of occurrence, and a time of de-activation. The number is linked to a detailed description. The SCADA data consists of time series of the most important sensors in the turbine: power produced, RPM, wind speed,… The advantage of the 1-sec data over the traditional 10-minute averages is that the dynamic event content is much more preserved. Data of several offshore wind farms is used in the analysis to have a solid dataset. In total, 5 years of data of more than 50 turbines is used. We show a novel farm-wide pattern mining approach that extracts events occurring for multiple turbines in the same time period. This allows us to identify those events that are predominantly driven by global wind excitations (e.g., gusts) or grid events (e.g., low voltage ride through). From the extracted events, we lift out the storm conditions. For these conditions, a further investigation of the time series data is done. Using event detection algorithms, we extract the signatures of the stop events that each turbine is performing from the time series data. We show that the extreme change in wind speed and wind direction leads to an excessive misalignment of the turbines in the farm, followed by a stop of those turbines. The extracted patterns are compared to the time signatures to show their correlation and complementarity. As such, the typical turbine response to this event is identified. This can serve as input for identification of novel controller approaches by the farm owner and turbine manufacturer to deal with this problem.

P.-J. Daems, L. Feremans, T. Verstraeten, B. Cule, B. Goethals, J. Helsen

Efficiency and Durability of Spur Gears with Surface Coatings and Surface Texturing

Experimental results on mechanical efficiency and durability of spur gears are presented. Reducing the power losses and increasing the reliability of gears are fundamnetal aspects of the design of transmissions. In order to analyze the tribological performance of innovative coatings and laser texturing, three different types of spur gears were tested: standard carburized gears, WC/C-coated carburized gears, and carburized gears with laser texture pattern. Tests have been carried out through a power recirculating test rig equipped with a single-stage transmission; power losses were evaluated by analyzing the torque-meters signals; several pictures of the teeth were taken at scheduled times to monitor the wear progression. In conclusion, results are presented and discussed.

Giovanni Iarriccio, Antonio Zippo, Marco Barbieri, Francesco Pellicano

System Model for Condition Monitoring and Prognostic

Frontmatter

Wheel-Rail Contact Forces Monitoring Based on Space Fixed-Point Strain of Wheel Web and DIC Technology

Condition monitoring of wheel-rail contact forces has great significance for evaluating the dynamic performance of railway vehicles. Instrumented wheelset method has disadvantages that strain gages are complicated in arrangement, wheel-rail forces decoupling is difficult, and the accuracy is low. Combined with the digital image correlation (DIC) technology, this paper introduces a novel non-contact monitoring method based on space fixed-point strain on the surface of wheel web. Through the finite element modeling and simulation analysis of wheelset, it is found that the strain at some space fixed-points on the surface of wheel web is linear with lateral/vertical forces. Based on this conclusion, the numerical equation between single space fixed-point strain and the lateral/vertical forces is established. Two point numerical equations can be decoupled to obtain the lateral and vertical force. The space fixed-point strain can be obtained by non-contact using the mature DIC technology and then achieve the simple and accurate measurement method of wheel-rail contact forces. It is verified by FEM simulation results that under the premise of using DIC technology to get space fixed-point strain of wheel web; the force monitoring method introduced in this paper can obtain the wheel-rail contact forces accurately and simply, which has important theoretical guiding significance for wheel-rail contact forces monitoring.

Xiaochao Wang, Zhenggang Lu, Juyao Wei, Yang He

Experimental Validation of the Blade Excitation in a Shaft Vibration Signals

Monitoring of blade vibrations is an important part in the maintenance of the rotating machine and its proper operation. The signal analysis of blade vibration plays an important role in detection of blade structure change. Early detection can avoid unnecessary losses. This is important especially for blades in the last stage of low-pressure turbine. The popular method used for blade vibration monitoring is blade-tip timing, which is not installed standardly yet due to the higher installation costs. An alternative method for blade vibration monitoring is based on relative shaft vibration signal analysis. This signal is measured by standardly installed instrumentation thus any additional sensor installation is not necessary. Recently, a few papers were published showing that rotor oscillations contain the information about the blade vibrations. However, the principle of the blade oscillations transition into the shaft vibrations was not described in detail so far. Explanation of main principle of rotor excitation by rotating blades is described in this paper. The described theory is experimentally validated using a machinery fault simulator. The experiment was based on the excitation of the forced blade vibrations by the piezo elements installed on the bladed disc, which was mounted on the simulator shaft. Excitation of piezo elements was set in an appropriate form to be able to clarify the principle and validate theory which was described in this paper.

V. Vasicek, J. Liska, J. Strnad, J. Jakl

Seismic Risk Evaluation by Fragility Curves Using Metamodel Methods

This study proposed an efficient and reliable fragility estimation by metamodel methods. To evaluate seismic risk, it is important to account for ground motions and structural parameter uncertainties. Nevertheless, seismic fragility analysis with uncertainties is impractical since it requires many time-consuming nonlinear dynamic simulations. To this end, an efficient approach based on metamodel in conjunction with Monte Carlo simulation is proposed to develop fragility curves. Several metamodel methods such as kriging, response surface method (RSM), and radial basis function (RBF) are investigated for this purpose. Optimum Latin hypercube design is used to generate “space filling” samples for nonlinear dynamic analysis under seismic excitations. A metamodel is constructed based on these design of experiment samples which include the input parameters and output fragility results. Kriging and RBF are better than the RSM metamodel. The fragility curves can be generated based on metamodel by considering the randomness of earthquake ground motions and uncertainties in material properties. Finally, seismic risk is evaluated by fragility curves. The computation time is significantly reduced by applying the metamodel method with acceptable accuracy. The proposed methodology also avoids nonlinear dynamic non-convergent problems. Kriging and RBF methods complement each other and are able to accurately evaluate fragility curves.

H. Z. Yang, C. G. Koh

Reliability Centred Maintenance (RCM) Assessment of Rail Wear Degradation

Rail wear degradation durability performance is an important concern that affects railway track maintenance and strategies to mitigate and prevent costly events of rail steel failures. An understanding of wear degradation will enable predictions of rail wear degradation trends, providing guidance for effective rail maintenance planning. In this work, an approach to predict rail steel wear degradation based on rail wear records is developed. The approach is based on statistical modelling of wear degradation data from maintenance records. For each track type and radius, available historical wear measurements are extrapolated to wear limits (as time-to-failure), and suitable continuous probability distributions are used to fit these failures. The low rail head wear and high rail side wear of sharp curves deemed most critical are emphasized, while differentiating the different track foundation types (viaduct or tunnel). Predictions can be made for the degradation trend of rails, providing guidance for effective rail maintenance planning. As an example, for R300 (sharp curve radius of 300 m) high rails in the tunnel sections, rail wear failures follow an increasing failure rate Weibull distribution, with scale parameter of about 20 years. This indicates that a large proportion of rails will fail within that stipulated time frame, and hence, it is mandatory for these rails to be replaced in advance prior to failure.

H. J. Hoh, J. L. Wang, A. S. Nellian, J. H. L. Pang

Prognostic of Rolling Element Bearings Based on Early-Stage Developing Faults

Rolling element bearing (REB) failure is one of the general damages in rotating machinery. In this manner, the correct prediction of remaining useful life (RUL) of REB is a crucial challenge to move forward the unwavering quality of the machines. One of the main difficulties in implementing data-driven methods for RUL prediction is to choose proper features that represent real damage progression. In this article, by using the outcomes of frequency analysis through the envelope method, the initiated/existed defects on the ball bearings are identified. Also, new features based on developing faults of ball bearings are recommended to estimate RUL. Early-stage faults in ball bearings usually include inner race, outer race, ball and cage failing. These features represent the sharing of each failure mode in failure. By calculating the severity of any failure mode, the contribution of each mode can be considered as the input to an artificial neural network. Also, the wavelet transform is used to choose an appropriate frequency band for filtering the vibration signal. The laboratory data of the ball bearing accelerated life (PROGNOSTIA) are used to confirm the method. To random changes reduction in recorded vibration data, which is primary in real-life experiments, a preprocessing calculation is connected to the raw data. The results obtained by using new features show a more accurate estimation of the bearings’ RUL and enhanced prediction capability of the proposed method. Also, results indicate that if the contribution of each failure mode is considered as the input of the neural network, then RUL is predicted more precisely.

M. Hosseini Yazdi, M. Behzad, S. Khodaygan

Recent Technologies for Smart and Safe Systems

Frontmatter

Fault-Tolerant Actuation Architectures for Unmanned Aerial Vehicles

Rising civilian applications that make use of unmanned aerial vehicles (UAVs) demand crucial precautions to minimize safety hazards. Future UAVs are expected to incorporate fault-tolerant architectures for critical on-board systems to ensure compliance with airworthiness certification. Reliability reports of in-service UAVs showed that flight control actuators are among the highest root-causes of UAVs mishaps. In this paper, the current state-of-the-art actuation architectures for UAVs are reviewed to identify technical requirements for certification. This work is part of a TEMA-UAV research project aimed at developing certifiable fault-Tolerant Electro-Mechanical Actuators for future UAVs.

M. A. A. Ismail, C. Bosch, S. Wiedemann, A. Bierig

Towards Certifiable Fault-Tolerant Actuation Architectures for UAVs

There is an increasing demand for integrating unmanned aerial vehicles (UAVs) into civilian airspaces. Consequently, onboard systems should be evaluated for possible threats to human life. This paper discusses future certification requirements for flight control actuators. A scheme is proposed for evaluating reliability requirements for flight control electro-mechanical actuators (EMAs) considering different flight control configurations. This work is part of the TEMA-UAV project, which aims at developing certifiable fault-tolerant actuation for future UAVs.

C. Bosch, M. A. A. Ismail, S. Wiedemann, M. Hajek

Development of a Patrol System Using Augmented Reality and 360-Degree Camera

From the viewpoints of safety plant operation and improvement of utilization rate of equipment, reduction of the human error in inspection work and efficient maintenance engineering and inspection work are needed. Therefore, we developed a maintenance support tool using augmented reality (AR) technology and 360-degree camera. Furthermore, the self-localization technology using Visual Simultaneous Localization and Mapping (Visual SLAM) technology is adopted. The inspection work and data management which linked with localization data with Visual SLAM technology are the feature of this system. These technologies help users identify equipment, assist users’ recognition and judgment of the object conditions and effective inspection data management.

Kenji Osaki, Tomomi Oshima, Naoya Sakamoto, Tetsuro Aikawa, Yuya Nishi, Tomokazu Kaneko, Makoto Hatakeyama, Mitsuaki Yoshida

Research on Evaluation Method of Safety Integrity for Safety-Related Electric Control System of Amusement Ride

The safety-related electric control system (SRECS) of amusement ride as an important system ensures the safe operation of the device. Its safety integrity determines whether the SRECS can perform safety functions reliably. At present, the professional safety integrity evaluation of the SRECS for amusement ride is still blank all over the world. Based on the analysis of the structural type and movement characteristics of the amusement rides, this paper develops a safety integrity evaluation method suitable for the SRECS of amusement rides with reference to the international function safety standards of safety-related system. First step is to identify the safety-related control function (SRCF) of SRECS combined the failure case database, and then to do the risk assessment of SRCF considering the failure severity and possibility, such as death, injury, and high altitude detention. The safety integrity level requirements of ride (RSILr) can be obtained by risk matrix. The four reliability indicators of the designed control loop are analyzed, and the safety integrity level (RSIL) of the control loop is obtained by combining the safety integrity histogram. Finally, the effectiveness of the safety integrity of the control system is confirmed by tests.

Weike Song, Peng Cheng, Ran Liu

Study on Warning System of Transportable Pressure Equipment from Regulatory Perspective

Accidents on transportable pressure equipment happen occasionally and some of them were due to weak supervision on its using and maintenance processes. In order to reduce this kind of accidents, a warning system from regulatory perspective is designed and studied in this paper. The warning system involves several significant processes of the transportable pressure equipment including design, manufacture, using, filling, and maintenance. First, the risks were identified and analyzed in all processes. Second, the index system was established and the weight of each index was calculated by using AHP and entropy methods. Third, the warning level classification and the waning model were illuminated in this study. The warning system would have a certain reference value to the safety management on transportable pressure equipment.

Z. R. Yang, Z. Z. Chen, Z. W. Wu, P. Yu

Industry 4.0: Why Machine Learning Matters?

Machine Learning is at the forefront of the Industry 4.0 revolution. Both the amount and complexity of data generated by sensors make interpretation and ingestion of data beyond human capability. It is impossible for engineers to optimise an extremely complex production line and to understand which unit or machine is responsible for quality or productivity degradation! It is extremely difficult for engineers to process monitoring or inspection data as it requires a protocol, a trained and certified engineer as well as experience! It is extremely difficult to guarantee the quality of every single product particularly at high production rates! Artificial Intelligence can help answering the above questions. Indeed, machine learning can be used for predictive or condition-based maintenance. Even without knowing the design of the machine (i.e., gearbox stages, bearing design, etc.), a machine learning algorithm is capable to monitor deviation of monitoring sensors features compared to a healthy state. Machine learning can be used to monitor the quality of production by allowing the automation of the quality control process. Monitoring additive manufacturing process to detect defects during printing and allow mitigation through real-time machining and reprinting of the defective area. Ensuring the quality of very complex and sensitive production processes such as MEMS, electronic wafers, solar cells or OLED screens. Brunel Innovation Centre (BIC) is working on developing algorithms combining statistical, signal/image processing for features extraction and deep learning for automated defect recognition for quality control and for predictive maintenance. Brunel Innovation Centre is also working on integrating those technologies into the digital twin.

T. H. Gan, J. Kanfoud, H. Nedunuri, A. Amini, G. Feng

Condition Monitoring of Electromechanical Systems

Frontmatter

A Method for Online Monitoring of Rotor–Stator Rub in Steam Turbines

In recent years, the operation of steam turbines has been associated with the occurrence of rotor–stator rub. This is because the clearances in the turbine flow paths decrease in an effort to increase the turbine efficiency. At present, detailed rubbing diagnostics is implemented as offline analysis of measured data. The detection of the rotor–stator rub is based on analysis of vibration signals in time domain when a machine operator monitors the overall level of vibration or the vector of rotation frequency contained in the measured vibration signal. This paper deals with design and implementation of the system for online monitoring of rotor–stator rub implemented as the rub advanced monitoring system (RAMS), which, in addition to detection phase, provides also localization of the place of contact occurrence, based on additional knowledge of the rotor geometry. Except the contact on the shaft, the contact between the blade tips and the stator seal can also be detected and localized. The contact is in that case accompanied by periodic impacts whose frequency for example corresponds to the product of the number of swirl brakes above blade tips and the rotation frequency. With this information obtained from standard rotor vibration signals, the rotor operation can be diagnosed in more detail and, if necessary, the turbine operation can be adapted to the current situation.

J. Liska, J. Jakl

Implementation of Instantaneous Power Spectrum Analysis for Diagnosis of Three-Phase Induction Motor Faults Under Mechanical Load Oscillations; Case Study of Mobarakeh Steel Company, Iran

Induction motors are the most commonly used prime movers in various industrial applications. Their vast amount of implementation necessitates a thorough condition monitoring framework to assure their well-being, reduction of costs, increasing their lifetimes and avoiding unwanted stoppages. Intensive research efforts have been focused on motor current signature analysis (MCSA), the technique which utilizes the spectral analysis of the stator current and aims to detect electrical and mechanical defects. Although MCSA is one of the most common and efficient techniques in condition monitoring of induction motors, several drawbacks hinder their applicability and efficiency in certain conditions; one of the most important of which is when the motor faces oscillating loads where fault-related sidebands in MCSA spectrum are misunderstood with the ones due to load oscillation and hence it leads to false indications . In this paper, a new fault detection technique focusing on active and reactive instantaneous power signatures rather than stator current is proposed for the diagnosis of faults in three-phase induction motors and prevention of false indications due to overlapping fault sideband components with those pertained to load oscillations. As a specific case study, a large 2 MW motor in steel production process in Mobarakeh Steel Company, Iran, with a relatively oscillating load is studied. A common fault in induction motors, a broken rotor bar is simulated for verification of the results. Computer simulation is performed using MATLAB based on the proposed case study and both approaches are compared in both healthy and faulty situations, highlighting the advantages of the proposed method.

S. Mani, M. Kafil, H. Ahmadi

Analysis of Unbalanced Magnetic Pull in a Multi-physic Model of Induction Machine with an Eccentric Rotor

2nd World Congress on Condition Monitoring (Singapore 2–5 December 2019)

Based on a multi-physic model of an asynchronous electrical machine with a strong electro-magneto-mechanical interaction, the Unbalanced Magnetic Pull (UMP) generated from the uneven air-gap due to the rotor eccentricity is calculated and discussed. The proposed model includes strong couplings between mechanical, magnetical, and electrical behaviors, particularly, the relationship between angle sampling and time sampling about the rotor position which introduces the instantaneous angular speed. Simulations with different input static eccentricity values are investigated in rated operation, and the effects of UMP on the dynamic behavior of the whole system are analyzed. Vibration analysis is performed at the supports to extract information in vibration signals for eccentricity monitoring.

Xiaowen Li, Adeline Bourdon, Didier Rémond, Samuel Koechlin, Dany Prieto

Bearings Fault Diagnosis in Electromechanical System Using Transient Motor Current Signature Analysis

The bearing faults may lead to expensive and catastrophic failures, which affect the reliability of power drivetrain in electromechanical system. To reduce impact of the bearing failures, accurate analysis and timely diagnosis are required for improved reliability of the electromechanical systems. In this work, an enhanced transient current signature analysis has been investigated for aircraft application using permanent magnet synchronous motor with bearing defects. The motor current under steady-state and transient conditions is acquired from an experimental test-rig with bearing defects at different loading and speed levels. The acquired signals are first investigated using frequency domain analysis and then compared with the time–frequency domain analysis such as wavelet analysis. The discrete wavelet transform is used to analyze on the calculated residual current for the bearing fault diagnosis. The proposed time–frequency-based technique is able to provide useful features that characterize the condition of the bearings on transient current. Further, back-propagation neural network is used over the calculated features which distinguishes the defective bearing from the healthy bearing with high accuracy.

Alok Verma, Haja Kuthubudeen, Viswanathan Vaiyapuri, Sivakumar Nadarajan, Narasimalu Srikanth, Amit Gupta

SHM/NDT Technologies and Applications

Frontmatter

Assessment of RC Bridge Decks with Active/Passive NDT

We are facing a lot of infrastructures damaged seriously. In consideration with restriction of the maintenance budget, strategy to prolong their service as long as possible has been generally adopted. To maintain those structures reasonably, current condition of the structures shall be properly evaluated by ideal inspections, for example, using NDT. The authors have thus been studying NDT utilizing elastic wave approaches to quantify the damage of infrastructures. In the paper, periodic monitoring and continuous monitoring with elastic wave approaches are demonstrated respectively. Specifically elastic wave- or AE- tomography is applied to the existing concrete bridge deck to interpret the damage distribution with their characteristic parameters. As for the continuous monitoring, newly developed super acoustic (SA) sensors and a four channels FPGA-based edge computing unit with event-driven technology are demonstrated. And to introduce the NDT into in-situ maintenance program, essential issues are summarized. As for a key technology of remote monitoring, several fundamental aspects of laser-induced elastic waves are experimentally studied.

T. Shiotani, H. Asaue, K. Hashimoto, K. Watabe

Estimation of In-Plane Strain Field Using Computer Vision to Capture Local Damages in Bridge Structures

The monitoring of variation in strain field near the ends of girder and bearings in steel plate girder bridges to capture local damages, is of a paramount importance for condition assessment of its service life. It is well-known that, the thickness reduction due to corrosion and failure at bearings, possibly will decrease the necessary functional requirements of structure. By understanding the computer vision technique, it is implemented for a typical bending test under quasi-static loading to aluminum plate girder bridge model specimen at laboratory level, for estimating in-plane strain field variation near end of girder. In this research work, the experiments are carried out using low-cost camera and two-dimensional digital image correlation (2D-DIC) technique is employed for analysis. The experiment is repeated under two different loading conditions and change in boundary conditions, to capture small responses with noise in it. The in-plane strain field patterns at different time step corresponding to the applied quasi-static load can be clearly understood from 2D-DIC results. The obtained results are in good agreement with the strain gauge measurements. The residual strain can also be captured at the unloading condition. From the strain captured at the Web section, it is between 15 and 35μ which is very low strain value. Hence, the computer vision camera can capture the low strain variations which corresponds to local damages in steel plate girder bridge structures through 2D-DIC technique.

T. J. Saravanan, M. Nishio

Wireless Strain Sensors for Dynamic Stress Analysis of Rail Steel Structural Integrity

Fatigue fractures were observed at the web-to-head and web-to-foot interface regions of rail steel thermite weld joint on curve tracks in a mass rapid transit tunnel. To investigate these failures, onsite wireless strain measurement was conducted to record the train-induced stress spectrum at a thermite weld joint and the results were presented in this paper. Dynamic finite element (FE) analysis was then conducted using the thermite weld and rail steel structural model under train-induced loading conditions and is validated by the strain gage sensor measurements. The train-induced stress at critical locations of thermite weld was investigated. Combining the strain measurements and FE analysis results, the service load stress spectrum using the variable amplitude strain analysis spectrum at the stress concentration site was obtained and fatigue life for a critical location of the thermite weld was predicted using the cumulative damage summation method. This research provides a fatigue mechanics-based approach for rail steel asset life prediction methodology.

X. P. Shi, Z. F. Liu, K. S. Tsang, H. J. Hoh, Y. Liu, Kelvin Tan Kaijun, Phua Song Yong, Kelvin, J. H. L. Pang

Purpose and Practical Capabilities of the Metal Magnetic Memory Method

This paper presents the purpose and practical capabilities of metal magnetic memory (MMM) methods. MMM is an advanced Non-destructive Testing (NDT) method developed through accumulation of Self-Magnetic Leakage Field (SMLF) and its distribution analysis in Stress Concentration Zones (SCZs) area. The SCZs is a major source of damages mechanism development and can be detected in express-control mode using specified MMM sensor, scanning devices, and software. It can be performed without require surface preparation and artificial magnetization where it utilizes residual magnetization formed naturally during product fabrication and throughout the operation. MMM method when combined with other NDT methods such as ultrasound dramatically improves inspection result efficiency. Presently, MMM methods are widely employed in base metal and weld joints inspection. For inaccessible objects such as buried pipelines, Non-Contact Magnetometric Diagnostics (NCMD) will be employed with the same fundamentals used by the MMM method. Therefore, it is proposed that MMM is best suited for SCZs detection in feasible diagnosis of equipment and structure and at once enables the assessment of actual stress–strain state to be carried out for all structural elements.

A. A. Dubov, S. M. Kolokolnikov, M. N. Baharin, M. F. M. Yunoh, N. M. Roslin

Development and Application of Inner Detector in Drilling Riser Auxiliary Lines Based on Magnetic Memory Testing

Drilling riser auxiliary lines are critical components of deep-water floating drilling platform, which mainly consist of choke line, kill line, hydraulic line, boost line, etc. This paper discusses an inner detector in drilling riser auxiliary lines based on metal magnetic memory testing. In contrast to conventional detect methods currently employed, the prominent advantage of the detector is that it can scan the inner wall directly by the pull of a winch, with no need to disassemble the buoyancy modules. This way, the time and cost of testing would be shortened considerably. The detector is designed with two sensor arrays assembled in staggered pattern, and each array consists of six tunnel magnetoresistance probes radially arranged from the canter of the shaft, much like spokes on a wagon wheel, ensuring 100% coverage of the inner wall. The raw data acquired from the front and rear sensor arrays can be aligned automatically by the alignment algorithms of imaging software for further processing. For determination of the detector characteristics, the laboratory test was conducted with 3.5″ tubing that were modified with a series of prefabricated groove and dome-shaped artificial defects. And result shows that it can present the damage sharply. Besides, field test also confirms the design has features of reasonable structure and stable performance.

Xiangyuan Liu, Jianchun Fan, Wei Zhou, Shujie Liu

Defect Recognition and Classification Techniques for Multi-layer Tubular Structures in Oil and Gas Wells by Using Pulsed Eddy Current Testing

Pulsed eddy current (PEC) method has attracted researchers’ attention in the detection of multi-layer tubular structure defects in oil and gas wells because of its advantages of non-destructive testing, not being affected by non-ferromagnetic interference. In this paper, the detection experiment of the casing with prefabricated defects is carried out, and the PEC signal analysis method for defect recognition and classification of multi-layer tubular structures is studied. To deal with the problem that it is difficult to extract time-domain features, a modified principal component analysis method is proposed. Combining the statistical features extracted by the method and frequency-domain features, the combined features are obtained, and the random forest algorithm is introduced to classify the defects. This paper demonstrates an effective method for the detection of multi-layer tubular structure defects in oil and gas wells.

Hang Jin, Hua Huang, Zhaohe Yang, Tianyu Ding, Pingjie Huang, Banteng Liu, Dibo Hou

Corrected Phase Reconstruction in Diffraction Phase Microscopy System

This paper presents an approach for correcting phase reconstruction from the interferogram Diffraction phase microscopy (DPM) system if there exist artefacts around the defect features due to noises. By averaging local region around the artefacts of the unwrapped phase using a sliding window with iterative approach, the correct phase information can be retrieved. The reconstructed phase images after correction step demonstrate the ability of the proposed approach in extracting the correct phase data from the system.

V. P. Bui, S. Gorelik, Z. E. Ooi, Q. Wang, C. E. Png

Study on Fatigue Damage of Automotive Aluminum Alloy Sheet Based on CT Scanning

Fatigue damage is the damage accumulation process under the repetitive cycle load, which is accompanied by degradation of material properties. It is great significant to study the fatigue life of aluminum alloy in the field of engineering application. The damage characteristics of 6061 aluminum alloy are studied by three-dimensional reconstruction and finite element method to estimate low-cycle fatigue damage based on the X-ray computer tomography. The results show that the estimated damage variables are accordant with those measured experimentally. Moreover, the three-dimensional reconstruction method combined with the finite element analysis method to calculate the material damage can meet the engineering requirements.

Ya li Yang, Hao Chen, Jie Shen, Yongfang Li

IoT-Web-Based Integrated Wireless Sensory Framework for Non-destructive Monitoring and Evaluation of On-Site Concrete Conditions

This research combines the use of IoT-web-based application in monitoring the real-time conditions of concrete, to accurately evaluate the maturity of concrete samples from its fresh state, accurately predict the strength of concrete among other properties, and it provides remote access to the outputs in real-time. The predicted strength of concrete was acquired using various maturity functions based on the temperature-time history of concrete samples. Two cement types of fresh concrete were prepared, ordinary portland cement (OPC) and blast furnace slag (BB) cement, at water-cement ratio 50%. Immediately, after fresh concrete samples were cast into molds and well vibrated, the monitoring of internal average temperature commenced, and these specimens were kept under controlled curing condition over a 28-day period, and the temperature data were logged directly into the database by the IoT sensors via a mobile internet hotspot connection. Graphical outputs of temperature, concrete maturity and strength data for both concrete types were monitored remotely; presented on the programmed web-based application and was accessible for download on any internet-connected device.

A. O. Olaniyi, T. Watanabe, K. Umeda, C. Hashimoto

A Study of Atmospheric Corrosive Environment in the North East Line Tunnel

This project was conducted to study the effect of the tunnel atmosphere on uniform corrosion of metals. Metal coupons and environment monitoring systems (ECMs) were deployed at buffer areas of different stations along the North East Line to measure the rate of corrosion. Concurrently, temperature and relative humidity were also measured through the ECM. Results from one year of monitoring are presented in this paper.

L. T. Tan, S. M. Goh, H. Y. Gan, Y. L. Foo

Acoustic Emission

Frontmatter

Fatigue Crack Growth Monitoring Using Acoustic Emission: A Case Study on Railway Axles

Railway axles are safety-critical components, which are designed to be safe and reliable under standard operation. However, unexpected service conditions, impact events and issues missed in maintenance can affect their lifetime. Non-destructive testing (NDT) methods employed to assure structural integrity are expensive and disruptive requiring axles to be taken out of service for inspection. Therefore, there is need for more efficient, real-time monitoring approach. This paper presents an overview of key challenges facing successful in-service implementation of an acoustic emission (AE) condition monitoring system. This is followed by an experimental study of AE data captured during a mechanical test of a railway axle. The data is clustered using a self-organising map (SOM) to differentiate damage signals from other sources. The results presented show consistency with the crack growth measured using traditional NDT techniques.

R. Marks, A. Stancu, S. Soua

Acoustic Emission Monitoring for Bending Failure of Laminated Glass Used on Glass Water Slide

In recent years, a new kind of amusement ride called glass water slide has been rapidly increasing. As a key component of this ride, laminated glass may be subject to all kinds of applied loads, which impose safety risks. In addition, the test data and evaluation method for laminated glass used on glass water slide is not enough. To address this issue, a method combining acoustic emission (AE) with digital image correlation (DIC) is used to monitor the damage and deformation under three-point bend loading conditions. The results show that the lower tempered glass is broken when loaded to the maximum bending load (about 16kN), then the load dropped rapidly. Afterward, the bending load is restored to about 8kN, at the moment, the upper tempered glass is broken, which indicated the tempered glass completely failure. For the laminated glass with bonding defects, more AE signals with higher amplitude value are generated during the load process. The strain concentration and severe displacement deformation are observed when the upper and lower parts of the glass fail. The distribution characteristics of the displacement and strain fields can be well correlated with the amplitude and frequency distribution of the AE signals, revealing the whole process of the damage evolution of the laminated glass. It is suggested that the complementary technology combining AE with DIC is effective for damage monitoring and evaluation of the laminated glass.

Ran Liu, Gongtian Shen, Yong Zhang, Wei Zhou, Pengfei Zhang

Source Location of Acoustic Emission for Anisotropic Material Utilizing Artificial Intelligence (WCCM2019)

AE testing (AT) is one of the nondestructive inspection methods using elastic waves (AE) generated by cracks or deformation of materials. The location of sources of AE is determined by arrival time differences to the AE sensors attached on the structures (AE source location). In this study, only one three-axis accelerometer is used for two-dimensional source location on anisotropic thin plate as an alternative to attaching a number of general AE sensors to the plate. As a result, arrival time differences that required for traditional AE source location algorighm cannot be obtained. Convolution neural network (CNN) which is one of the AI is utlilized to realize source location in this study. Even used only one accelerometer, source location accuracy is better than that by using a number of AE sensors.

Y. Mizutani, N. Inagaki, K. R. Kholish, Q. Zhu, A. Todoroki

Standard Development and Application on Acoustic Emission Condition Monitoring of Amusement Device

Rotating structure, as the most important part of the amusement device, is difficult to be disassembled after its installation. Acoustic emission (AE) technology is an effective tool for condition monitoring and fault diagnosis of rotating machinery. Acoustic emission tests for non-defective and defective rolling bearings were studied. The analysis and evaluation method of AE condition monitoring on amusement devices were introduced. Application test on a rotating amusement device was carried out. The results indicated that the AE technique can be applied to the condition monitoring and fault diagnosis of the amusement devices.

J. J. Zhang, G. T. Shen, Z. W. Wu, Y. L.Yuan

Fault Diagnosis of Pitch Bearings on Wind Turbine Based on Acoustic Emission Method

Pitch bearing is one of the critical components of wind turbines. If it broke down, it would lead to serious phenomena, such as falling leaves and falling towers. Therefore, the fault diagnosis of the pitch system bearing is important. Acoustic emission technology has been widely applied to the bearings fault diagnosis. It could capture the dynamic information of the pitch bearing when the bearing generated microscopic deformation and crack propagation. A fault diagnosis method of pitch bearing based on acoustic emission technology is proposed in this paper. Firstly, the acoustic emission signal with obvious cracking characteristics is selected from collecting signals by the kurtosis value. Then, the selected acoustic emission signal is processed by the wavelet spectrum theory to extract the fault feature and identify the fault pattern of pitch bearings. Lastly, an online monitoring system of pitch bearings based on acoustic emission method is developed. The real application results showed that the system could online monitor the work condition of the pitch bearing. The crack fault of pitch bearing was identified by wavelet packet transform analysis and time–frequency spectrum analysis under continuous wavelet transform.

Guang-hai Li, Yang ming An, Tao Guo, Guang-quan Hou

Ultrasonic Testing

Frontmatter

In Situ Condition Monitoring of High-Speed Rail Tracks Using Diffuse Ultrasonic Waves: From Theory to Applications

Real-time condition monitoring is a critical step to warrant the integrity of rail tracks in bourgeoning high-speed railway (HSR) industry. Nevertheless, existing damage identification, condition monitoring and structural health monitoring (SHM) approaches, despite their proven effectiveness in laboratory demonstration, are restricted from in-situ implementation in engineering practice. By leveraging authors’ continued endeavours, an in situ health and condition monitoring framework, using actively generated diffuse ultrasonic waves (DUWs) and a benchmark-free condition-contrasting algorithm, has been developed and deployed. Fatigue cracks in the tracks show unique contact behaviours under different conditions of external loads and further disturb DUW propagation, and the crack growth induced by external loads can also alternate DUW propagation. By contrasting DUW propagation traits, fatigue cracks in rail tracks can be characterised quantitatively and the holistic condition of the tracks can be evaluated in a real-time manner. Compared with guided wave- or acoustic emission-based methods, the DUW-driven inspection philosophy exhibits immunity to ambient noise and measurement uncertainty, less dependence on baseline signals, and high robustness in atrocious engineering conditions. Conformance tests are performed on rail tracks, in which the evolution of fatigue damage is monitored continuously and quantitatively, demonstrating effectiveness, reliability and robustness of DUW-driven condition monitoring towards HSR applications.

K. Wang, W. Cao, Z. Su

Theoretical Analysis of Scattering Amplitude Calculation of Torsional Wave on a Pipe

The use of guided wave is widely used for inspection of long-range pipelines due to its propagation characteristic with low attenuation. Also, recently structural health monitoring (SHM) and condition monitoring (CM) is made a huge paradigm shift on inspection and monitoring of large structures. However, the most important issue of structure monitoring is quantitative defect sizing and interpretation from the signal. In this research work, the elastodynamic reciprocity is applied to calculate scattered wave amplitude calculation. The reciprocity is a well-known theorem to formulate the complicated scattering problem in a simpler calculation. The torsional wave mode on a cylindrical structure is explored by the reciprocity theorem. The scattered amplitude is presented as a function of incident wave frequency and defect size. The defects on the surface with an inclined angle are formulated to analyze the relation of defect angle and wavelength. The wave field is obtained by using the reciprocity theorem and a superposition technique. The theoretical result can provide signal interpretation for quantitative defect sizing. The closed-form solution of torsional wave scattering amplitude can be used for quantitative signal interpretation. It is expected to be used in the analysis of structure condition by SHM and CM approach as baseline data.

Jaesun Lee, Jan D. Achenbach, Younho Cho

Detection of a Micro-crack on a Rotating Steel Shaft Using Noncontact Ultrasonic Modulation Measurement

Detection of a micro-crack on a rotating structure is limited because conventional non-destructive testing (NDT) is developed based on contact-type transducers. By using air-coupled ultrasonic NDT system, micro-crack detection on a rotating steel shaft is studied. Through air-coupled ultrasonic system, ultrasonic-guided waves at two distinctive frequencies are generated. Then, their ultrasonic modulation at the sum and difference of input frequencies are generated due to micro-crack. Finally, the spectral correlation (SC) of ultrasonic modulation is computed for diagnosis of micro-crack. This is the first study for the detection of a micro-level fatigue crack on a rotating steel shaft using noncontact ultrasonic NDT system. The developed system is experimentally validated using the steel shaft through cyclic loading tests. At each cyclic loading case, the result was compared with conventional penetrant test and metallographic analysis.

I. Jeon, P. Liu, H. Sohn

Measurement of Axial Force of Bolted Structures Based on Ultrasonic Testing and Metal Magnetic Memory Testing

Bolted structures are widely utilized in flange sealing of the petrochemical high-pressure facilities, measurement of axial force of bolted structures can contribute to monitor flange sealing status. In this paper, static tensile tests are conducted in laboratory. The stress distribution of bolt with different axial tensile load is analyzed by using finite element (FE) method. Two non-destructive testing (NDT) methods including ultrasonic testing and metal magnetic memory testing are used to measure axial force of two bolt materials (low-carbon steel 4.8 and medium-carbon steel 8.8 bolts) during the whole tensile process. For the ultrasonic testing, the test results show that there is a stable linear correlation between the axial force of the bolt and the ultrasonic time of flight (TOF), and the TOF measurement is upgraded by using cross-correlation analysis technique. For the metal magnetic memory testing, magnetic signals increase with the increase of axial force because of the effect of stress-magnetization. Furthermore, magnetic memory characteristic parameters have a good linear relationship with axial force within the scope of a certain stress. Therefore, metal magnetic memory testing can be used as a novel method to measure axial force of bolted structures. Finally, the advantages and drawbacks of these two NDT methods are analyzed, which can provide a reference for the engineering application of effective NDT methods and technologies in bolt axial force measurement.

Siqi Yang, Laibin Zhang, Jianchun Fan

Ultrasonic Testing of Anchor Bolts

Anchor bolts play a crucial role in supporting the power cables of the overhead catenary system (OCS) in underground train tunnels. With exposure to humid conditions and water seepage in the tunnel from wet weathers, the anchor bolts can experience corrosion and damage at an accelerated rate. As anchor bolts are embedded into the concrete tunnel, and their conditions are unknown without proper inspection of their internal bodies. This paper aims to propose an inspection process to allow technicians to perform ultrasonic testing (UT) on anchor bolts during scheduled maintenance works in the tunnel. Experimental studies using ultrasonic waves were carried out on the anchor bolts of different integrity conditions. Analytical studies conducted were then used to determine the tolerance level and decision matrix to categorize the different conditions of anchor bolts. The outcome of this study shows that there is a feasible approach to perform UT inspections on the anchor bolts.

Y. M. Tong, A. R. K. Rajkumar, H. G. Chew, C. K. Liew

Weld Inspection of Piping Elbow Using Flexible PAUT Probe

In order to optimize the operation of power generation equipment, structures, and facilities, and to maintain safety and reliability for a long period of time, non-destructive testing (NDT) has to be conducted to secure the structural integrity. Among the various NDT methods, radiographic testing (RT) is mainly used for the production/installation and scheduled maintenance of power generation facilities according to relevant codes and specifications, and they mostly have a significant effect on process delay and economic loss that occur due to duplicate application of RT and ultrasonic testing (UT). Therefore, the phased array ultrasonic testing (PAUT) is actively introduced and utilized as a new volumetric examination technique, which can reduce the use of the RT as much as possible and replace the RT in the industrial field. The PAUT is capable of high-speed inspection because not only it can control the injection of the ultrasonic beam electronically, but it can also easily test specimens with complicated shapes, which are difficult to be detected with the conventional UT. It is not easy to obtain objective and reliable testing data because the probe contact is difficult when the PAUT is used to examine the welded part with curved shapes such as a piping elbow welding part. Keywords: Please list a maximum of 6 keywords or phrases. Therefore, in this study, instead of using a wedge, which was essentially used for the inspection of the pipe with curvature structures in the conventional studies, a 5 MHz, 32 Ch Linear flexible PAUT probe that can examine a piping elbow welding part was designed and manufactured, and the usefulness was verified. In order to verify the usefulness, welding specimens were manufactured according to the ASME standard and artificial defects were processed. Finally, the possibility to overcome the limitations of the existing PAUT technologies and to improve the reliability of test results were confirmed through and signal acquisition and analysis.

S. J. Lim, T. S. Park, M. Y. Choi, I. K. Park

Laser Ultrasonic Imaging of Wavefield Spatial Gradients for Damage Detection

Laser ultrasonic techniques (LUTs) are increasingly used in non-destructive evaluation (NDE) applications because they can provide full-field ultrasonic wavefield data with high resolution in a three-dimensional (3D) space–time domain. This paper investigates the properties of the wavefield spatial gradient as a detector and localizer for defects that distort the local wavefield. A laser ultrasonic interrogation method based on a 527-nm Q-switched laser scanning system was used to interrogate 3D ultrasonic signals in a 3-mm aluminum plate. The plate was tested with and without multiple artificial scatterers to test the proposed method. During the scanning process, the ultrasonic waves were generated by the pulsed laser at an n-by-m grid of sampling points and recorded by a single fixed PZT transducer. The whole scan generates a 3D matrix which allows the accessibility of showing the wave propagation over time in movie form with a high spatial resolution. After the scanning process, the full-field ultrasonic data of the plate without and with artificial damage were obtained, respectively. Then, spatial gradient vectors were computed from each frame of the full-field ultrasonic data along the time axis. For a space-domain frame $$\left\{ {U_{i} } \right\}^{n \times m}$$ U i n × m at a time $$i \in t$$ i ∈ t , the gradient vectors were determined by two components, $$\left\{ {G_{i,x} } \right\}^{n \times m}$$ G i , x n × m and $$\left\{ {G_{i,y} } \right\}^{n \times m}$$ G i , y n × m , which reflected the spatial derivatives of every pixel with respect to the position coordinates. Once the gradient vectors of the full-field ultrasonic data (intact and damage cases) were determined, the spatial gradient ultrasonic wavefield imaging was generated with the gradient orientation profile $$\left\{ {\theta_{i} } \right\}^{n \times m}$$ θ i n × m and the gradient magnitude profile $$\left\{ {A_{i} } \right\}^{n \times m}$$ A i n × m as the features used for damage detection. The wavefield spatial gradients of the defect-free case were used as a baseline to compare with the wavefield spatial gradients of the damaged case. The gradient profiles of the wavefield spatial gradients could detect and localize the damage by the inconsistency of the wave pattern in the 2D space domain. The wavefield spatial gradients also showed the ability to remove the complexity of identifying the damage among multi-mode waveform. Future work will be focusing on applying the algorithm to complex structures such as CFRP composites.

S. Y. Chong, Z. Wu, M. D. Todd

Guided Waves Testing

Frontmatter

Guided Ultrasonic Waves in Glass Laminate Aluminium Reinforced Epoxy

Knowledge on the structural health of hybrid composites that consist of laminated layers of thin metal sheets and fibre reinforced adhesive films are of increasing importance due to growing use in aircraft industry. For this purpose, we propose damage detection techniques based on the propagation behaviour of guided ultrasonic waves. Therefore, it is crucial to know the characteristic effect of fibre metal laminates on the wave propagation. Accompanying, in this study, we investigate experimentally and numerically how guided ultrasonic waves propagate in a fibre metal laminate.

M. Rennoch, M. Koerdt, A. S. Herrmann, N. Rauter, R. Lammering

Ultrasonic Guided Waves in Unidirectional Fiber-Reinforced Composite Plates

The article is concerned with the motion of Lamb-type guided waves in a unidirectional fiber-reinforced composite under ultrasonic sources. This kind of composite is much stiffer along its axis than across its transverse direction and therefore shows orthotropic material properties. The dispersion relation of free guided waves propagating in a unidirectional composite plate is derived that results in dispersion curves. Using reciprocity in elastodynamics, closed-form solutions of elastic wave motions subjected to time-harmonic loads in a unidirectional fiber-reinforced composite are in a simple manner computed. The obtained computations can be very useful for developing ultrasound-based methods for nondestructive evaluation of composite structures.

Ductho Le, Jaesun Lee, Younho Cho, Duy Kien Dao, Truong Giang Nguyen, Haidang Phan

Mode Selectivity and Frequency Dependence of Guided Waves Generated by Piezoelectric Wafer Transducers in Rebars Embedded in Concrete

Guided waves have the potential to recognise the corrosion damage in reinforced concrete (RC) at its nascent stages. The latest approaches to generate guided waves deploy contact acoustic transducers and piezoelectric wafer transducers (PWTs). Experimental studies show that excitation of guided waves using contact transducers causes multiple longitudinal guided wave modes in embedded rebars and these modes can be used to assess various features of the damage. In contrast, the excitation with piezoelectric transducers introduces only a dominant wave mode. In this study, mode selectivity of guided waves in embedded rebars is investigated. Finite element simulations are adopted to calculate the response of the RC beam to PWT excitation and are verified with the experimental signal. The frequency dependence of the dominant guided wave mode is studied by examining the displacement profile of the RC beam. The displacement pattern is compared with the mode shape of various wave modes that are calculated using theoretical dispersion curves to recognise the possible wave modes. It is found that the lower frequencies cause F(1, 1) mode, whereas the higher frequencies trigger higher-order flexural modes. The results of this work assist in selecting the appropriate frequency to assess damages of debonding or diameter reduction.

Rajeshwara Chary Sriramadasu, Ye Lu, Sauvik Banerjee

How Temperature Variations Affect the Propagation Behaviour of Guided Ultrasonic Waves in Different Materials

As guided ultrasonic waves are sensitive to changes in the acoustic impedance of the structure, knowledge of the waves behaviour with respect to different materials and varying temperatures is crucial for reliable damage detection. Hence, we present how specific wave characteristics, such as velocity and amplitude, behave for typical materials at various frequencies and temperatures, and how this knowledge can be integrated into a damage detection technique. We also provide an analysis of the thickness of the adhesive layer since we have experimentally observed a strong temperature-related effect on guided wave signals.

M. Rennoch, J. Moll, M. Koerdt, A. S. Herrmann, T. Wandowski, W. M. Ostachowicz

Nonlinearity Study of Damaged Spring Materials Using Guided Wave

Nonlinear ultrasonic techniques that can detect micro crack or defects before use of ultrasonic nonlinearity are emerging as a way to predict the life of materials. Guided wave propagates according to the shape of the material, thus providing great potential for non-destructive evaluation applications. However, in the non-destructive evaluation method using nonlinear guided waves, it is difficult to get necessary information due to the dispersion and multi-mode characteristics of guided wave. In this study, the nonlinearity of spring materials and the degree of fatigue was investigated using nonlinear guided wave. The specimens were prepared with different degree of fatigue through Nakamura fatigue test. The guided wave has generated a signal at the point that 460 mm far from the left end of the spring specimen and was received at a distance of 50 mm from the transmitted point. In addition, the experiment carried out for six sections at intervals of 10 mm. The signal was generated using a tone buster (RPR-4000), and the vicinity of the point, where the fatigue load was largest (the middle point of the specimen), was investigated. This study presents a quantitative evaluation method for the health of spring materials using nonlinear guided wave. Histological examinations, such as SEM imaging, can provide more accurate and meaningful results, and further, studies such as these must identify the cause.

Jeongnam Kim, Yonghee Lee, Sungun Lee, Younho Cho

Transducer Arrangement for Baseline-Free Damage Detection Methods With Guided Waves

Structural Health Monitoring (SHM) refers to methods for the early detection of damage in components in order to initiate planned repair. Acoustic methods using guided waves are widely established for the monitoring of large-area, thin-walled structures. Current SHM approaches based on active ultrasonic techniques use the knowledge of a known reference condition of a considered transmit–receive path, with which all subsequent conditions are compared, in order to obtain a statement about the local component condition. In most cases, these methods evaluate amplitude or phase changes with respect to the reference signal. The dependency of the signal evaluations on environmental influences such as temperature or humidity changes makes it necessary to explicitly consider these influences through statistical analyses and background knowledge on the material side or to provide different reference states known as baselines for different environmental conditions. The subject of this study is the development of baseline-free methods. The focus is on the identification of the damage interaction of the guided waves and the mode identification by suitable transducer arrangements for the determination of structural damage on reciprocal sensor transmission paths. For this purpose, parameterized simulation experiments were performed to investigate different arrangements of piezoelectric transducers for wave mode identification and to validate their results in laboratory measurements. It could be shown that stacked transducers can be used for wavemode identification. Further more an additional effect of phase shifts in between two wavemodes was observed using a piezoelectric ring transducer with two electrode pairs.

U. Lieske, T. Gaul, L. Schubert

Mode Switchable Guided Elastic Wave Transducer Based on Piezoelectric Fibre Patches

An improved transducer for shear horizontal guided waves is presented. Two ideas are combined: (1) the generation of pure shear tractions by stapling two piezoelectric fiber patches and (2) matching the excitation pattern with the spatial-time behaviour of the intended wave. The spatial pattern is determined by the structure of the interdigital electrodes. One possible realization of the electrode layout is selected, and the performance of this type of transducers is demonstrated by FEM modelling of the generated wave field. The results are discussed and scheduled further work which is indicated.

Y. Kim, K. Tschöke, L. Schubert, B. Köhler

Lamb Wave Excitation Using a Flexible Laser Ultrasound Transducer for Structural Health Monitoring

Using Lamb waves for detection is a desirable method for the structural health monitoring (SHM) and has been widely used in plate-like structure. In this paper, an originally flexible ultrasonic transducer with high photo-acoustic conversion efficiency was proposed by using candle soot nanoparticles (CSNPs) and polydimethylsiloxane (PDMS). Experiments results demonstrate that the developed transducer can generate a longitudinal wave in the aluminum block with the amplitude of 10 times to that of the conventional laser ultrasound technique, demonstrating the high energy conversion efficiency of the proposed transducer. Wedge-shape transducers were developed to excited Lamb wave in the 1.5 mm aluminum substrate by the oblique incident method. The specific A0 and S0 modes of the Lamb wave with the central frequency of 647 kHz were successfully excited in the aluminum plate of 1.5 mm thickness. A0 mode Lamb wave was selected for damage detection application. According to the synthetic aperture focusing imaging technique, a delay-and-sum signal processing method was adopted for damage location in the plate. A 3.5 mm defect was well imaged and the results demonstrate that the developed flexible photo-acoustic transducer can be a promising method for the SHM.

Wei Li, Jitao Xiong, Wenbin Huang

Lamb Wave Propagation of S0 and A0 Modes in Unidirectional CFRP Plate for Condition Monitoring: Comparison of FEM Simulation with Experiments

CFRP is a material widely used in a variety of industries, including aerospace, wind blade, architecture, automobile and other fields of engineering. Ultrasonic waves are widely used for condition monitoring (CM) of CFRP structures, so quantitative understanding of wave propagation on anisotropic material is essential. Conceptual ambiguities were found in previous researches and were clarified by introducing new perspective. FEM simulation was carried out for Lamb wave propagation on unidirectional CFRP plate. The wave speed of S0 mode from the simulation showed good agreement with that from experiment, which guarantees the accuracy of the simulation. For the simulation results, S0 and A0 mode wave can be separated by adding and subtracting displacements at upper and lower surface. Group velocity of S0 mode is highly affected by the propagating direction whereas A0 showed lower dependence, and this was shown in our simulation data. Also, an unexpected wave was observed in the simulation and was identified as SH0 mode.

Subin Kim, Joonseo Song, Seungmin Kim, Younho Cho, Young H. Kim

Thermographic Testing

Frontmatter

Three-Dimensional Reconstruction of Leaked Gas Cloud Image Based on Computed Tomography Processing of Multiple Optical Paths Infrared Measurement Data

The current gas leakage source detection was conducted by the human senses and experience. The development of remote gas leakage source detection system is required. In this research, infrared camera was used to detect gas leakage. The gas can be detected by the absorption of infrared rays by the gas, and the infrared rays are emitted from the gas itself. Three-dimensional reconstruction of leaked gas cloud was performed to identify the gas leakage source. The 3D reconstruction of leaked gas cloud was performed by the multiple optical paths of infrared measurement and inverse tomography analysis. Algebraic reconstruction techniques (ART) method was applied to reconstruction. In the experiments, the gas concentration distribution was simulated by the arrangement of gas cells. It was found that the gas concentration distribution composed by gas cell could be estimated by infrared images obtained with few optical path and ART method.

M. Uchida, D. Shiozawa, T. Sakagami, S. Kubo

Experimental Analysis on Fault Prediction of AC-Contactor by Infrared Thermography

Aiming at the problems of poor contact, aging and dampness of elevator AC-contactors, the infrared thermal imaging technology is proposed to detect the faults of elevator AC-contactors. By simulating the operation of elevator, an experimental platform is built to realize real-time monitoring of the operation of contactor. Through a lot of experiments, the infrared thermal characteristic maps of contactors under dust, loosening of terminals, oxidation of contacts and overload operation are obtained, which can better detect the abnormal situation. At the same time, in order to predict the remaining life of the contactor in advance, accelerated life experiment was designed, and the experimental data were analyzed by the median iteration method and the least square method. The reliability function of contactor under experimental temperature is obtained by Weibull distribution model. The 90% reliability is taken as the residual life of the contactor at this temperature, and the relationship between the residual life of the contactor and the temperature is obtained by curve fitting. It can be used to predict the residual life of contactors at different operating temperatures. The experimental results can provide basis for inspection of elevator maintenance personnel.

Yue Yu, Wei Li, Chao Ye, Bin Hu

Research on the Detecting Method of Motor Running State of Amusement Device Based on Infrared Thermography Technology

The motor is the source power of the amusement device. Once a failure occurs, it will not only cause damage to itself, but also lead to high-altitude retention and even casualties. The traditional method of monitoring the operating state by installing sensors is costly and complicated to implement, and a large number of existing motors cannot be installed. Therefore, there is an urgent need for quick and convenient technical means and methods for detecting the operating state of the motor. To address this issue, this study simulated various operating conditions of the amusement devices and conducted comparative tests of different load rates, different duty ratios, and different cycle durations. By comparing the temperature rise data collected by the embedded sensor in the motor with the temperature rise data collected by the external infrared thermography of the motor, the main factors affecting the difference of temperature rise between the inner and outer parts of the motor casing are obtained, and the correction value of the temperature rise difference between the inside and outside of the casing is also obtained. A way of testing motor running state by infrared thermography is proposed, which can be applied in site detection with ease.

C. Ye, Y. Yu, J. J. Zhang, R. Liu, G. T. Shen

Application of Ultrasound Thermometry to Condition Monitoring of Heated Materials

Monitoring of thermal information of heated media is important for ensuring the reliability of high temperature processes in various industries. Ultrasound thermometry, which is a temperature measurement method by ultrasound, is expected as a potential candidate for such monitoring. In this work, an effective ultrasound thermometry utilizing both pulse-echo measurements and unsteady heat conduction analysis is applied to real-time monitoring of temperature profiles and heat fluxes of a heated steel. This method provides non-invasive measurements of such thermal information. To examine the feasibility of the method to condition monitoring, the internal temperature profile of a steel plate whose back surface is being heated, and cooled is monitored every 70 ms. Furthermore, the ultrasonic method is utilized to monitor heat fluxes of bonded materials of aluminum and copper, and reasonable results on transient variation in heat flux are then obtained. Thus, it is considered that the ultrasonic method could be applicable to condition monitoring of various heated materials.

I. Ihara, R. Sawada, Y. Ogawa, Y. Kawano

Gas Leak Source Identification by Inverse Problem Analysis of Infrared Measurement Data

With the retirement of skilled workers and the aging of gas plant facilities, interest in the development of automated and remote gas detection methods is increasing. Therefore, a gas leak detection method using infrared camera has been proposed. Invisible gas can be detected from infrared absorption of gas and the emitted infrared ray from the gas itself. The piping of a gas plant is very complicated, so it is very useful that the gas leak source can be identified. In this study, the gas leak source identification is performed by the inverse problem analysis using the infrared measurement data. The least residual method is one of the typical inverse problem analysis methods, and it estimates the position of the gas leak source from the sequential infrared images of the gas cloud distribution and flow. It was found from numerical simulation and laboratory experiments that it is possible to estimate the position of the gas leak source even when the leak source is hidden by piping in the infrared image.

H. Nishimura, D. Shiozawa, T. Sakagami, S. Kubo
Additional information

Premium Partners

    Image Credits