In an effort to increase machining efficiency and minimize costs, research into tool condition monitoring (TCM) systems has focused on developing methods to allow for unmanned machining. For drilling processes, such systems typically use indirect approaches to monitoring the tool condition by measuring spindle torque and feed force as well as vibrations including acoustic emission (AE – mechanical vibrations faster than 100 kHz). This project aimed to advance the state-of-the-art in the area of TCM by developing a method to detect sudden tool failures in large diameter (> 25 mm) indexable insert drills. This project was a continuation of the research conducted by Mr. R. Griffin (a former MSc student), who developed a model capable of predicting long term wear trends in indexable insert drills [1]. Notably, his model was unable to react to sudden tool breakage due to tool chipping, which was addressed by this project as presented in this thesis. In order to develop and train models able to detect sudden tool failure, an experiment was developed and installed in the field of the industry partner of this project. The experiment’s main feature was a pair of AE sensors added to the existing torque and force sensors. On this setup, experiments were conducted by drilling 2251 holes in workpieces using indexable insert drills with or without the insert breaking. When drilling holes without the insert breaking, the holes were named as good ones; and when drilling holes with the insert breaking they were named as bad holes. During the drilling process, data was collected from current sensors attached to the spindle motor and feed motor as well as from an AE sensor on the spindle and on the workpiece. From the signals from the spindle motor current and feed motor current sensors, algorithms were developed to identify and divide the signals of drilling a hole into different sections of the drilling cycle (i.e. entrance, steady-state, exit, etc.). Steady-state time-domain features were extracted from the sensor signals measured for all holes drilled in the experiments and the extracted features were used to train and test the classifier models. These models were cross validated to determine which type of model was the best fit for the drilling data collected. The results from the classifier models show that most of the classifiers tested have the ability to identify sudden tool breakage based on the data recorded in the present study, with varying degrees of success. The naïve Bayes classifier was able to detect the most failures but suffered from a large number of falsely detected failures. Both the classification tree and linear discriminant analysis classifiers had lower failure detection rates than the naïve Bayes classifier, but did not suffer from the same amount of false positives; as such, these two classifiers had higher overall classification rates than the naïve Bayes. These results suggest that classification tree and linear discriminant analysis methods are better suited for the drilling application and that the time-domain features should be complemented by others, such as the features extracted from the frequency domain, to accurately diagnose the tool condition. Future research should focus on extracting frequency and time-frequency domain features as these features might contain more information on tool condition. In addition, methods of examining features at the entrance and exit of the holes should be investigated as these two points in the drilling cycle are the most prone to sudden tool failure.

The purpose of this research is to reduce maintenance resources and improve Escravos–Gas–to–Liquids plant availability (EGTL) in Escravos, Nigeria using wireless condition monitoring. Secondary to the above is to justify the use of this technology over other conventional condition monitoring methods in petrochemical plants with specific reference to cost, reliability and security of the system. Wireless and continuous condition monitoring provides the means to evaluate current conditions of equipment and detect abnormalities. It allows for corrective measures to be taken to prevent upcoming failures. Continuous monitoring and event recording provides information on the energized equipment's response to normal and emergency conditions. Wireless/remote monitoring helps to coordinate equipment specifications and ratings, determine the real limits of the monitored equipment and optimize facility operations. Bentley N, (2005). Using wireless techniques eliminate any need for special cables and wires with lower installation costs if compared to other types of condition monitoring systems. In addition to this, wireless condition monitoring works well under difficult conditions in strategically important locations. The Escravos gas–to–liquid plant in Nigeria, located in a remote and offshore area where accommodation and space for offices is a factor for monitoring plant conditions in every office, is a typical example. Wireless technology for condition monitoring of energized equipment is applicable to both standalone and remote systems. In the research work of Meyer and Brambley (2002), they characterized the current problem with regards to cost effectiveness and availability of wireless condition monitoring. Maintenance of rotating equipment provides probability estimates of the total impact of the problem, cost implication of plant equipment maintenance and describes a generic system in which these developing technologies are used to provide real–time wireless/remote condition monitoring for rotating main air compressor (MAC) units and their components as a case study. Costs with today’s technology are provided and future costs are estimated, showing that benefits will greatly exceed costs in many cases, particularly if low–cost wireless monitoring is used. With management trends such as “re–engineering” and “downsizing” of the available workforce, wireless condition–monitoring of critical machines has been given more importance as a way to ensure quality production with fewer personnel. Wireless condition–monitoring using inexpensive wireless communication technology frees up existing plant maintenance personnel work on machines that are signaling problems and focusing the maintenance efforts away from attempting to work on a large population of machines to only those machines requiring immediate attention. Lloyd and Buddy (200) suggested that Point–to–point wireless data transmission systems, an excellent example of recent technological advances in communication systems, are now practical and cost–effective for industrial use. While both complex infrastructures and complex protocols are required for cellular communications, non– cellular communication systems, such as the point–to–point wireless data transmission system example, require no elaborate infrastructure. Limited research was done on the immediate benefits of implementing wireless condition monitoring systems in plants. All papers on the subject have been drawn up by manufacturers of such equipment. This research will thus also deliver a "third–party" perspective on the effectiveness of such devices, justifying their impact on data gathering security, cost and reliability. / Thesis (M.Ing. (Development and Management Engineering))--North-West University, Potchefstroom Campus, 2012.

The purpose of this research is to reduce maintenance resources and improve Escravos–Gas–to–Liquids plant availability (EGTL) in Escravos, Nigeria using wireless condition monitoring. Secondary to the above is to justify the use of this technology over other conventional condition monitoring methods in petrochemical plants with specific reference to cost, reliability and security of the system. Wireless and continuous condition monitoring provides the means to evaluate current conditions of equipment and detect abnormalities. It allows for corrective measures to be taken to prevent upcoming failures. Continuous monitoring and event recording provides information on the energized equipment's response to normal and emergency conditions. Wireless/remote monitoring helps to coordinate equipment specifications and ratings, determine the real limits of the monitored equipment and optimize facility operations. Bentley N, (2005). Using wireless techniques eliminate any need for special cables and wires with lower installation costs if compared to other types of condition monitoring systems. In addition to this, wireless condition monitoring works well under difficult conditions in strategically important locations. The Escravos gas–to–liquid plant in Nigeria, located in a remote and offshore area where accommodation and space for offices is a factor for monitoring plant conditions in every office, is a typical example. Wireless technology for condition monitoring of energized equipment is applicable to both standalone and remote systems. In the research work of Meyer and Brambley (2002), they characterized the current problem with regards to cost effectiveness and availability of wireless condition monitoring. Maintenance of rotating equipment provides probability estimates of the total impact of the problem, cost implication of plant equipment maintenance and describes a generic system in which these developing technologies are used to provide real–time wireless/remote condition monitoring for rotating main air compressor (MAC) units and their components as a case study. Costs with today’s technology are provided and future costs are estimated, showing that benefits will greatly exceed costs in many cases, particularly if low–cost wireless monitoring is used. With management trends such as “re–engineering” and “downsizing” of the available workforce, wireless condition–monitoring of critical machines has been given more importance as a way to ensure quality production with fewer personnel. Wireless condition–monitoring using inexpensive wireless communication technology frees up existing plant maintenance personnel work on machines that are signaling problems and focusing the maintenance efforts away from attempting to work on a large population of machines to only those machines requiring immediate attention. Lloyd and Buddy (200) suggested that Point–to–point wireless data transmission systems, an excellent example of recent technological advances in communication systems, are now practical and cost–effective for industrial use. While both complex infrastructures and complex protocols are required for cellular communications, non– cellular communication systems, such as the point–to–point wireless data transmission system example, require no elaborate infrastructure. Limited research was done on the immediate benefits of implementing wireless condition monitoring systems in plants. All papers on the subject have been drawn up by manufacturers of such equipment. This research will thus also deliver a "third–party" perspective on the effectiveness of such devices, justifying their impact on data gathering security, cost and reliability. / Thesis (M.Ing. (Development and Management Engineering))--North-West University, Potchefstroom Campus, 2012.

Transformers form a very important part of the power transmissions and distribution network; as they are responsible for the transfer of electrical energy from the power generation sites onto the transmission lines and finally to the distribution stage. Dry-type and oil-filled single-phase transformers, either alone or as a part of three-phase banks, are used extensively in the power distribution network, ultimately providing power to the domestic consumers. Any faults in the single-phase transformers leading to power outages or catastrophic power systems failures cause huge loss of capital, property and in some cases even human casualties. Gradual deterioration of the electrical winding insulation ultimately leads to inter-turn winding short circuit faults; which account for a significant proportion of all transformer failures. Incipient stages of inter-turn winding faults have negligible impact on the terminal voltages and currents of transformers; thus these faults often go undetected by the traditional differential relay based protection mechanisms. By the time, the faults manifest themselves into severe winding short-circuit faults consequently forcing the differential relays to operate for tripping the circuit breakers; a significant part of the transformer windings and core may get extensively damaged. Over the years, various techniques have been developed for detecting and studying inter-turn winding faults; however their practical implementation involves quite a few challenges such as high cost, lack of reliability, low accuracy and need for mounting additional equipment inside the transformer casing. Additionally, none of the existing techniques are suitable for online and real-time condition monitoring of the transformers. This absence of any proven technique to detect incipient levels of inter-turn winding faults in single-phase transformers has motivated the research of this thesis. In the thesis, firstly, a non-invasive technique for modeling single-phase transformers has been developed which is based solely on the terminal measurements of voltages and currents. The effects of transformer core saturation, non-linearity, hysteresis are incorporated in the model by considering a time-varying magnetizing inductance comprising of any desired number of harmonic components. The coefficients of the magnetizing inductance are computed from the instantaneous values of flux linkage and magnetizing current over one complete cycle. The model is found to replicate the behaviour of the single-phase transformer with an extremely high level of accuracy, under any load conditions for healthy as well as faulty operations. Detailed simulation and experiment based studies have been performed for corroborating the effectiveness of the proposed terminal measurement based modeling technique not only in detecting incipient stages of inter-turn winding faults (involving less than 1% of the turns) but also in estimating fault severity. Also, a non-invasive, online and real-time implementation of the proposed inter-turn winding fault detection technique for continuous monitoring of the transformer health has been suggested. Firstly, with the experimentally acquired primary line voltage and line current data of the healthy transformer, a healthy no-load model of the transformer is generated. Next, a healthy estimated indicator value, computed from this model under the given input voltage condition, is compared with the actual indicator value for detecting the presence of an inter-turn winding fault. It involves minimum hardware (only two current sensors and one voltage sensor), low memory requirements and low computational complexity and thus holds a good promise for practical applications. Further discussion is made on the possible challenges for realizing the proposed fault diagnostic technique in the industry and suitable recommendations have been made for further improvement. / Graduate / 0544 / bhowmick@uvic.ca

This research presents a complete review of signal processing techniques used, today, in vibration based industrial condition monitoring and diagnostics. It also introduces two novel techniques to this field, namely: the Kolmogorov-Smirnov test and Volterra series, which have not yet been applied to vibration based condition monitoring. The first technique, the Kolmogorov-Smirnov test, relies on a statistical comparison of the cumulative probability distribution functions (CDF) from two time series. It must be emphasised that this is not a moment technique, and it uses the whole CDF, in the comparison process. The second tool suggested in this research is the Volterra series. This is a non-linear signal processing technique, which can be used to model a time series. The parameters of this model are used for condition monitoring applications. Finally, this work also presents a comprehensive comparative study between these new methods and the existing techniques. This study is based on results from numerical and experimental applications of each technique here discussed. The concluding remarks include suggestions on how the novel techniques proposed here can be improved.

Condition monitoring of critical machine tool components and machining processes is a key factor to increase the availability of the machine tool and achieving a more robust machining process. Failures in the machining process and machine tool components may also have negative effects on the final produced part. Instabilities in machining processes also shortens the life time of the cutting edges and machine tool. The condition monitoring system may utilise information from several sources to facilitate the detection of instabilities in the machining process. To avoid additional complexity to the machining system the use of internal sensors is considered. The focus in this thesis has been to investigate if information related to the machining process can be extracted directly from the internal sensors of the machine tool. The main contibutions of this work is a further understanding of the direct response from both linear and angular position encoders due the variations in the machining process. The analysis of the response from unbalance testing of turn tables and two types of milling processes, i.e. disc-milling and slot-milling, is presented. It is shown that operational frequencies, such as cutter frequency and tooth-passing frequency, can be extracted from both active and inactive machine axes, but the response from an active machine axis involves a more complex analysis. Various methods for the analysis of the responses in time domain, frequency domain and phase space are presented. / QC 20100518

The ability to forecast machinery failure is vital to reducing maintenance costs, operation downtime and safety hazards. Recent advances in condition monitoring technologies have given rise to a number of prognostic models for forecasting machinery health based on condition data. Although these models have aided the advancement of the discipline, they have made only a limited contribution to developing an effective machinery health prognostic system. The literature review indicates that there is not yet a prognostic model that directly models and fully utilises suspended condition histories (which are very common in practice since organisations rarely allow their assets to run to failure); that effectively integrates population characteristics into prognostics for longer-range prediction in a probabilistic sense; which deduces the non-linear relationship between measured condition data and actual asset health; and which involves minimal assumptions and requirements. This work presents a novel approach to addressing the above-mentioned challenges. The proposed model consists of a feed-forward neural network, the training targets of which are asset survival probabilities estimated using a variation of the Kaplan-Meier estimator and a degradation-based failure probability density estimator. The adapted Kaplan-Meier estimator is able to model the actual survival status of individual failed units and estimate the survival probability of individual suspended units. The degradation-based failure probability density estimator, on the other hand, extracts population characteristics and computes conditional reliability from available condition histories instead of from reliability data. The estimated survival probability and the relevant condition histories are respectively presented as “training target” and “training input” to the neural network. The trained network is capable of estimating the future survival curve of a unit when a series of condition indices are inputted. Although the concept proposed may be applied to the prognosis of various machine components, rolling element bearings were chosen as the research object because rolling element bearing failure is one of the foremost causes of machinery breakdowns. Computer simulated and industry case study data were used to compare the prognostic performance of the proposed model and four control models, namely: two feed-forward neural networks with the same training function and structure as the proposed model, but neglected suspended histories; a time series prediction recurrent neural network; and a traditional Weibull distribution model. The results support the assertion that the proposed model performs better than the other four models and that it produces adaptive prediction outputs with useful representation of survival probabilities. This work presents a compelling concept for non-parametric data-driven prognosis, and for utilising available asset condition information more fully and accurately. It demonstrates that machinery health can indeed be forecasted. The proposed prognostic technique, together with ongoing advances in sensors and data-fusion techniques, and increasingly comprehensive databases of asset condition data, holds the promise for increased asset availability, maintenance cost effectiveness, operational safety and – ultimately – organisation competitiveness.

Vibration-based fault diagnosis (FD) with a simple spectrum can be complex, especially when considering FD of rotating machinery with multiple bearings like a multi-stage turbine. Various studies have sought to better interpret fault spectra, but the process remains equivocal. Consequently, it has been accepted that the simple spectra requires support from additional techniques, such as orbit analysis. But even orbit analysis can be inconclusive. Though promising, attempts at developing viable methods that rival the failure coverage of spectrum analysis without gaining computational complexity remain protracted. Interestingly, few researchers have developed FD methods for transient machine operation, however, these have proven to be involved. Current practices limit vibration data to a single machine, which usually requires a large unique data history. However, if sharing of data between similar machines with different foundations was possible, the need for unique histories would be mitigated. From readily available works, this has not been encountered. Therefore, a simple but robust vibration-based approach is warranted. In light of this, a novel on-bearing vibration response integration approach for condition monitoring of shaft-related faults irrespective of speed and foundation type is proposed in the present study. Vibration data are acquired at different speeds for: a baseline, unbalance, bow, crack, looseness, misalignment, and rub conditions on three laboratory rigs with dynamically different foundations, namely: rigid, flexible support 1 (FS1) and flexible support 2 (FS2). Testing is done on the rigid rig set up first, then FS1, and afterwards FS2. Common vibration features are computed from the measured data to be input to the proposed approach for further processing. First, the proposed approach is developed through its application to a machine at a steady speed in a novel Single-speed FD technique which exploits a single vibration sensor per bearing and fusion of features from different bearings for FD. Initially, vibration features are supplemented with bearing temperature readings with improved classification compared to vibration features alone. However, it is observed that temperature readings are insensitive to faults on the FS1 and FS2 rigs, when compared to vibration features, which are standardised for consistent classification on the different rigs tested. Thus, temperature is not included as a final feature. The observed fault classifications on the different rigs at different speeds with the standardised vibration features are encouraging. Thereafter, a novel Unified Multi-speed FD technique that is based on the initial proposed approach and which works by fusion of vibration features from different bearings at different speeds in a single analysis step for FD is proposed. Experiments on the different rigs repeatedly show the novel Multi-speed technique to be suitable for transient machine operation. Then, a novel generic Multi-foundation Technique (also based on the proposed approach) that allows sharing of vibration data of a wide range of fault conditions between two similarly configured machines with similar speed operation but different foundations is implemented to further mitigate data requirements in the FD process. Observations made with the rigs during steady and transient speed tests show this technique is applicable in situations where data history is available on one machine but lacking on the other. Comparison of experimental results with results obtained from theoretical simulations indicates the approach is consistent. Thus, the proposed approach has the potential for practical considerations.

New methods for monitoring the condition of roller element bearings in rotating machinery offer possibilities to reduce repair- and maintenance costs, and reduced use of environmentally harmful lubricants. One such method is sparse representation of vibration signals using matching pursuit with dictionary learning, which so far has been tested on PCs with data from controlled tests. Further testing requires a platform capable of signal processing and control in more realistic experiments. This thesis focuses on the integration of a hybrid CPU-FPGA hardware system with a 16-bit analog-to-digital converter and an oil pump, granting the possibility of collecting real-time data, executing the algorithm in closed loop and supplying lubrication to the machine under test, if need be. The aforementioned algorithm is implemented in a Zynq-7000 System-on-Chip and the analog-to-digital converter as well as the pump motor controller are integrated. This platform enables portable operation of the matching pursuit with dictionary learning in the field under a larger variety of environmental and operational conditions, conditions which might prove difficult to reproduce in a laboratory setup. The platform developed throughout this project can collect data using the analog-to-digital converter and operations can be performed on that data in both the CPU and the FPGA. A test of the system function at a sampling rate of 5 kHz is presented and the input and output are verified to function correctly.

The main focus of this study is to improve the profitability, availability and condition monitoring of Liquefied Natural Gas (LNG) Floating Production Storage and Offloading platforms (FPSOs). Propane pre-cooled, mixed refrigerant (C3MR) liquefaction is the key process in the production of LNG on FPSOs. LNG liquefaction system equipment has the highest failure rates among the other FPSO equipment, and thus the highest maintenance cost. Improvements in the profitability, availability and condition monitoring were made in two ways: firstly, by making recommendations for the use of redundancy in order to improve system reliability (and hence availability); and secondly, by developing an effective condition-monitoring algorithm that can be used as part of a condition-based maintenance system. C3MR liquefaction system reliability modelling was undertaken using the time-dependent Markov approach. Four different system options were studied, with varying degrees of redundancy. The results of the reliability analysis indicated that the introduction of a standby liquefaction system could be the best option for liquefaction plants in terms of reliability, availability and profitability; this is because the annual profits of medium-sized FPSOs (3MTPA) were estimated to increase by approximately US$296 million, rising from about US$1,190 million to US$1,485.98 million, if redundancy were implemented. The cost-benefit analysis results were based on the average LNG prices (US$500/ton) in 2013 and 2014. Typically, centrifugal turbines, compressors and blowers are the main items of equipment in LNG liquefaction plants. Because centrifugal equipment tops the FPSO equipment failure list, a Condition Monitoring (CM) system for such equipment was proposed and tested to reduce maintenance and shutdown costs, and also to reduce flaring. The proposed CM system was based on a novel FFT-based segmentation, feature selection and fault identification algorithm. A 20 HP industrial air compressor system with a rotational speed of 15,650 RPM was utilised to experimentally emulate five different typical centrifugal equipment machine conditions in the laboratory; this involved training and testing the proposed algorithm with a total of 105 datasets. The fault diagnosis performance of the algorithm was compared with other methods, namely standard FFT classifiers and Neural Network. A sensitivity analysis was performed in order to determine the effect of the time length and position of the signals on the diagnostic performance of the proposed fault identification algorithm. The algorithm was also checked for its ability to identify machine degradation using datasets for which the algorithm was not trained. Moreover, a characterisation table that prioritises the different fault detection techniques and signal features for the diagnosis of centrifugal equipment faults, was introduced to determine the best fault identification technique and signal feature. The results suggested that the proposed automated feature selection and fault identification algorithm is effective and competitive as it yielded a fault identification performance of 100% in 3.5 seconds only in comparison to 57.2 seconds for NN. The sensitivity analysis showed that the algorithm is robust as its fault identification performance was affected by neither the time length nor the position of signals. The characterisation study demonstrated the effectiveness of the AE spectral feature technique over the fault identification techniques and signal features tested in the course of diagnosing centrifugal equipment faults. Moreover, the algorithm performed well in the identification of machine degradation. In summary, the results of this study indicate that the proposed two-pronged approach has the potential to yield a highly reliable LNG liquefaction system with significantly improved availability and profitability profiles.