No description available.

Rolling element bearings are one of the most significant elements and frequently-used components in mechanical systems. Bearing fault detection and diagnosis is important for preventing productivity loss and averting catastrophic failures of mechanical systems. In industrial applications, bearing life is often difficult to predict due to different application conditions, load and speed variations, as well as maintenance practices. Therefore, reliable fault detection is necessary to ensure productive and safe operations. Vibration analysis is the most widely used method for detection and diagnosis of bearing malfunctions. A measured vibration signal from a sensor is often contaminated by noise and vibration interference components. Over the years, many methods have been developed to reveal fault signatures, and remove noise and vibration interference components. Though many vibration based methods have been proposed in the literature, the high frequency resonance (HFR) technique is one of a very few methods have received certain industrial acceptance. However, the effectiveness of the HFR methods depends, to a great extent, on some parameters such as bandwidth and centre frequency of the fault excited resonance, and window length. Proper selection these parameters is often a knowledge-demanding and time-consuming process. In particular, the filter designed based on the improperly selected bandwidth and center frequency of the fault excited resonance can filter out the true fault information and mislead the detection/diagnosis decisions. In addition, even if these parameters can be selected properly at beginning of each process, they may become invalid in a time-varying environment after a certain period of time. Hence, they may have to be re-calculated and updated, which is again a time-consuming and error-prone process. This undermines the practical significance of the above methods for online monitoring of bearing conditions. To overcome the shortcomings of existing methods, the following four non-parametric and non-filtering methods are proposed: 1. An amplitude demodulation differentiation (ADD) method, 2. A calculus enhanced energy operator (CEEO) method, 3. A higher order analytic energy operator (HO_AEO) approach, and 4. A higher order energy operator fusion (HOEO_F) technique. The proposed methods have been evaluated using both simulated and experimental data.

Rolling element bearings are one of the most significant elements and frequently-used components in mechanical systems. Bearing fault detection and diagnosis is important for preventing productivity loss and averting catastrophic failures of mechanical systems. In industrial applications, bearing life is often difficult to predict due to different application conditions, load and speed variations, as well as maintenance practices. Therefore, reliable fault detection is necessary to ensure productive and safe operations. Vibration analysis is the most widely used method for detection and diagnosis of bearing malfunctions. A measured vibration signal from a sensor is often contaminated by noise and vibration interference components. Over the years, many methods have been developed to reveal fault signatures, and remove noise and vibration interference components. Though many vibration based methods have been proposed in the literature, the high frequency resonance (HFR) technique is one of a very few methods have received certain industrial acceptance. However, the effectiveness of the HFR methods depends, to a great extent, on some parameters such as bandwidth and centre frequency of the fault excited resonance, and window length. Proper selection these parameters is often a knowledge-demanding and time-consuming process. In particular, the filter designed based on the improperly selected bandwidth and center frequency of the fault excited resonance can filter out the true fault information and mislead the detection/diagnosis decisions. In addition, even if these parameters can be selected properly at beginning of each process, they may become invalid in a time-varying environment after a certain period of time. Hence, they may have to be re-calculated and updated, which is again a time-consuming and error-prone process. This undermines the practical significance of the above methods for online monitoring of bearing conditions. To overcome the shortcomings of existing methods, the following four non-parametric and non-filtering methods are proposed: 1. An amplitude demodulation differentiation (ADD) method, 2. A calculus enhanced energy operator (CEEO) method, 3. A higher order analytic energy operator (HO_AEO) approach, and 4. A higher order energy operator fusion (HOEO_F) technique. The proposed methods have been evaluated using both simulated and experimental data.

The growing interest around offshore wind power, providing at the same time better wind conditions and fewer visual or environmental impacts, has lead many energy suppliers to consider the installation of offshore wind farms. However, the marine environment makes the installation and maintenance of wind turbines much more complicated, raising the capital and operation costs to an undesirable level and preventing the fast progression of this technology worldwide. Availability of offshore wind turbines varies between 65 and 90% depending on location, whereas onshore turbines range between 95 and 98% in most cases. In 2009, the ETI launched a research project aiming to improve economical efficiency of offshore wind farms by increasing their availability and decreasing their maintenance costs (partly through replacing corrective maintenance by preventive maintenance). This project named “Inflow” involves the development of a condition monitoring system, a system designed to monitor the state of different wind turbine components, and to analyze this data in order to determine the wind turbines overall condition at any given time, as well as its potential system ailments   This paper describes two different approaches to perform the condition monitoring of offshore wind farms, the first one involves thresholds-based analysis, while the other involves pattern recognition.

In this work, an automated statistical approach for the condition monitoring of a fluid power system based on a process sensor network is presented. In a multistep process, raw sensor data are processed by feature extraction, selection and dimensional reduction and finally mapped to discriminant functions which allow the detection and quantification of fault conditions. Experimentally obtained training data are used to evaluate the impact of temperature and different aeration levels of the hydraulic fluid on the detection of pump leakage and a degraded directional valve switching behavior. Furthermore, a robust detection of the loading state of the installed filter element and an estimation of the particle contamination level is proposed based on the same analysis concept.

Acoustic emission (AE) is one of many available technologies for condition health monitoring and diagnosis of rotating machines such as bearings. In recent years there have been many developments in the use of Acoustic Emission technology (AET) and its analysis for monitoring the condition of rotating machinery whilst in operation, particularly on high speed machinery. Unlike conventional technologies such as oil analysis, motor current signature analysis (MCSA) and vibration analysis, AET has been introduced due to its increased sensitivity in detecting the earliest stages of loss of mechanical integrity. This research presents an experimental investigation that is aimed at developing a mathematical model and experimentally validating the influence of operational variables such as film thickness, rotational speed, load, power loss, and shear stress for variations of load and speed conditions, on generation of acoustic emission in a hydrodynamic bearing. It is concluded that the power losses of the bearing are directly correlated with acoustic emission levels. With exponential law, an equation is proposed to predict power losses with reasonable accuracy from an AE signal. This experimental investigation conducted a comparative study between AE and Vibration to diagnose the rubbing at high rotational speeds in the hydrodynamic bearing. As it is the first known attempt in rotating machines. It has been concluded, that AE parameters such as amplitude, can perform as a reliable and sensitive tool for the early detection of rubbing between surfaces of a hydrodynamic bearing and high speed shaft. The application of vibration (PeakVue) analysis was introduced and compared with demodulation. The results observed from the demodulation and PeakVue techniques were similar in the rubbing simulation test. In fact, some defects on hydrodynamic bearings would not have been seen in a timely manner without the PeakVue analysis. In addition, the application of advanced signal processing and statistical methods was established to extract useful diagnostic features from the acquired AE signals in both time and frequency domain. It was also concluded that the use of different signal processing methods is often necessary to achieve meaningful diagnostic information from the signals. The outcome would largely contribute to the development of effective intelligent condition monitoring systems which can significantly reduce the cost of plant maintenance. To implement these main objectives, the Sutton test rig was modified to assess the capability of AET and vibration analysis as an effective tool for the detection of incipient defects within high speed machine components (e.g. shafts and hydrodynamic bearings). The first chapter of this thesis is an introduction to this research and briefly explains motivation and the theoretical background supporting this research. The second and third chapters, summarise the relevant literature to establish the current level of knowledge of hydrodynamic bearings and acoustic emission, respectively. Chapter 4 describes methodologies and the experimental arrangements utilized for this investigation. Chapter 5 discusses different NDT diagnosis. Chapter 6 reports on an experimental investigation applied to validate the relationship between AET on operational rotating machines, such as film thickness, speed, load, power loss, and shear stress. Chapter 7 details an investigation which compares the applicability of AE and vibration technologies in monitoring a rubbing simulation on a hydrodynamic bearing.

There is a critical need for inspection and condition monitoring of high temperature critical components such as pipelines and welds in electrical power generation and other plants operating at temperatures as high as 580°C. The high temperatures and pressures experienced in these pipelines, particularly for ageing plants lead to creep, fatigue and corrosion type defects. Safety of these plants is of paramount importance, and regular maintenance is carried out during planned outages at ambient temperatures. Ultrasonic non-destructive testing can be used to detect defects in the weld at ambient temperatures. However, at high operational temperatures, this technique cannot be applied due to the lack of high temperature transducers. This research has achieved significant advances towards enabling ultrasonic inspection and condition monitoring of high temperature critical points, by developing an ultrasonic transducer around an advanced piezoelectric single crystal material, called Gallium Orthophosphate (GaPO4), which can operate at the required temperature of 580°C. Based on its reported piezoelectric and other properties, and its commercial availability, GaPO4 was chosen as a candidate active material for application in a prototype high temperature transducer. In a series of confidence building tests with the selected piezoelectric material (electrical characterisation via the impedance method), it has been demonstrated that the GaPO4 piezoelectric elements are stable when subjected to 580°C for more than 600 hours. Ultrasonic thickness gauging has shown that GaPO4 works as a functional transducer generating and receiving ultrasound waves at 580°C for at least 360 hours. Furthermore, the sensitivity of the GaPO4 transducer to detect defects with simple geometry was successfully tested through measurements on steel blocks containing artificial defects (side-drilled holes) up to the same high temperatures. Based on the characterisation results from the impedance and ultrasonic measurements, a prototype ultrasonic transducer for operation at high temperatures has been designed and manufactured. The new ultrasonic transducer was tested in a laboratory environment using a steel calibration block, high temperature couplant, SONO 1100, and an electric furnace. In the range from ambient temperatures up to the target of 580°C, the ultrasonic transducer kept a signal-to-noise (SNR) level sufficiently high, above the threshold of 6 dB, which is high enough for practical non-destructive testing and condition monitoring.

With wind energy becoming a major source of energy, there is a pressing need to reduce all associated costs to be competitive in a market that might be fully subsidy-free in the near future. Before thousands of wind turbines were installed all over the world, research in e.g. understanding aerodynamics, developing new materials, designing better gearboxes, improving power electronics etc., helped to cut down wind turbine manufacturing costs. It might be assumed, that this would be sufficient to reduce the costs of wind energy as the resource, the wind itself, is free of costs. However, it has become clear that the operation and maintenance of wind turbines contributes significantly to the overall cost of energy. Harsh environmental conditions and the frequently remote locations of the turbines makes maintenance of wind turbines challenging. Just recently, the industry realised that a move from reactive and scheduled maintenance towards preventative or condition-based maintenance will be crucial to further reduce costs. Knowing the condition of the wind turbine is key for any optimisation of operation and maintenance. There are various possibilities to install advanced sensors and monitoring systems developed in recent years. However, these will inevitably incur new costs that need to be worthwhile and retro-fits to existing turbines might not always be feasible. In contrast, this work focuses on ways to use operational data as recorded by the turbine's Supervisory Control And Data Acquisition (SCADA) system, which is installed in all modern wind turbines for operating purposes -- without additional costs. SCADA data usually contain information about the environmental conditions (e.g. wind speed, ambient temperature), the operation of the turbine (power production, rotational speed, pitch angle) and potentially the system's health status (temperatures, vibration). These measurements are commonly recorded in ten-minutely averages and might be seen as indirect and top-level information about the turbine's condition. Firstly, this thesis discusses the use of operational data to monitor the power performance to assess the overall efficiency of wind turbines and to analyse and optimise maintenance. In a sensitivity study, the financial consequences of imperfect maintenance are evaluated based on case study data and compared with environmental effects such as blade icing. It is shown how decision-making of wind farm operators could be supported with detailed `what-if' scenario analyses. Secondly, model-based monitoring of SCADA temperatures is investigated. This approach tries to identify hidden changes in the load-dependent fluctuations of drivetrain temperatures that can potentially reveal increased degradation and possible imminent failure. A detailed comparison of machine learning regression techniques and model configurations is conducted based on data from four wind farms with varying properties. The results indicate that the detailed setup of the model is very important while the selection of the modelling technique might be less relevant than expected. Ways to establish reliable failure detection are discussed and a condition index is developed based on an ensemble of different models and anomaly measures. However, the findings also highlight that better documentation of maintenance is required to further improve data-driven condition monitoring approaches. In the next part, the capabilities of operational data are explored in a study with data from both the SCADA system and a Condition Monitoring System (CMS) based on drivetrain vibrations. Analyses of signal similarity and data clusters reveal signal relationships and potential for synergistic effects of the different data sources. An application of machine learning techniques demonstrates that the alarms of the commercial CMS can be predicted in certain cases with SCADA data alone. Finally, the benefits of having wind turbines in farms are investigated in the context of condition monitoring. Several approaches are developed to improve failure detection based on operational statistics, CMS vibrations or SCADA temperatures. It is demonstrated that utilising comparisons with neighbouring turbines might be beneficial to get earlier and more reliable warnings of imminent failures. This work has been part of the Advanced Wind Energy Systems Operation and Maintenance Expertise (AWESOME) project, a European consortium with companies, universities and research centres in the wind energy sector from Spain, Italy, Germany, Denmark, Norway and UK. Parts of this work were developed in collaboration with other fellows in the project (as marked and explained in footnotes).

This thesis examines the application of the techniques of Fourier spectrogram and wavelet analysis to a low power embedded microprocessor application in a novel railway and rollingstock monitoring system. The safe and cost effective operation of freight railways is limited by the dynamic performance of wagons running on track. A monitoring system has been proposed comprising of low cost wireless sensing devices, dubbed “Health Cards”, to be installed on every wagon in the fleet. When marshalled into a train, the devices would sense accelerations and communicate via radio network to a master system in the locomotive. The integrated system would provide online information for decision support systems. Data throughput was heavily restricted by the network architecture, so significant signal analysis was required at the device level. An electronics engineering team at Central Queensland University developed a prototype Health Card, incorporating a 27MHz microcontroller and four dual axis accelerometers. A sensing arrangement and online analysis algorithms were required to detect and categorise dynamic events while operating within the constraints of the system. Time-frequency analysis reveals the time varying frequency content of signals, making it suitable to detect and characterise transient events. With efficient algorithms such as the Fast Fourier Transform, and Fast Wavelet Transform, time-frequency analysis methods can be implemented on a low power, embedded microcontroller. This thesis examines the application of time-frequency analysis techniques to wagon body acceleration signals, for the purpose of detecting poor dynamic performance of the wagon-track system. The Fourier spectrogram is implemented on the Health Card prototype and demonstrated in the laboratory. The research and algorithms provide a foundation for ongoing development as resources become available for system testing and validation.

In a fully automated manufacturing environment, instant detection of condition state of the cutting tool is essential to the improvement of productivity and cost effectiveness. In this paper, a tool condition monitoring system (TCM) via machine learning (ML) and machine ensemble (ME) approach was developed to investigate the effectiveness of multisensor fusion when machining 4340 steel with multi-layer coated and multi-flute carbide end mill cutter. Feature- and decision-level information fusion models utilizing assorted combinations of sensors were studied against selected ML algorithms and their majority vote ensemble to classify gradual and transient tool abnormalities. The criterion for selecting the best model does not only depend on classification accuracy but also on the simplicity of the implemented system where the number of features and sensors is kept to a minimum to enhance the efficiency of the online acquisition system. In this study, 135 different features were extracted from sensory signals of force, vibration, acoustic emission and spindle power in the time and frequency domain by using data acquisition and signal processing modules. Then, these features along with machining parameters were evaluated for significance by using different feature reduction techniques. Specifically, two feature extraction methods were investigated: independent component analysis (ICA), and principal component analysis (PCA) and two feature selection methods were studied, chi square and correlation-based feature selection (CFS). For various multi-sensor fusion models, an optimal feature subset is computed. Finally, ML algorithms using support vector machine (SVM), multilayer perceptron neural networks (MLP), radial basis function neural network (RBF) and their majority voting ensemble were studied for selected features to classify not only flank wear but also breakage and chipping. In this research, it has been found that utilizing the multisensor feature fusion technique under majority vote ensemble gives the highest classification performance. In addition, SVM outperformed other ML algorithms while CFS feature selection method surpassed other reduction techniques in improving classification performance and producing optimal feature sets for different models.