Showing papers on "Condition monitoring published in 1999"

Condition monitoring and fault diagnosis of electrical machines-a review

Abstract: Research has picked up a fervent pace in the area of fault diagnosis of electrical machines. Like adjustable speed drives, fault prognosis has become almost indispensable. The manufacturers of these drives are now keen to include diagnostic features in the software to decrease machine down time and improve salability. Prodigious improvement in signal processing hardware and software has made this possible. Primarily, these techniques depend upon locating specific harmonic components in the line current, also known as motor current signature analysis (MCSA). These harmonic components are usually different for different types of faults. However with multiple faults or different varieties of drive schemes, MCSA can become an onerous task as different types of faults and time harmonics may end up generating similar signatures. Thus other signals such as speed, torque, noise, vibration etc., are also explored for their frequency contents. Sometimes, altogether different techniques such as thermal measurements, chemical analysis, etc., are also employed to find out the nature and the degree of the fault. Human involvement in the actual fault detection decision making is slowly being replaced by automated tools such as expert systems, neural networks, fuzzy logic based systems to name a few. Keeping in mind the need for future research, this review paper describes different types of faults and the signatures they generate and their diagnostics' schemes.

An overview of approaches to end milling tool monitoring

Abstract: The increase in awareness regarding the need to optimise manufacturing process efficiency has led to a great deal of research aimed at machine tool condition monitoring. This paper considers the application of condition monitoring techniques to the detection of cutting tool wear and breakage during the milling process. Established approaches to the problem are considered and their application to the next generation of monitoring systems is discussed. Two approaches are identified as being key to the industrial application of operational tool monitoring systems. Multiple sensor systems, which use a wide range of sensors with an increasing level of intelligence, are seen as providing long-term benefits, particularly in the field of tool wear monitoring. Such systems are being developed by a number of researchers in this area. The second approach integrates the control signals used by the machine controller into a process monitoring system which is capable of detecting tool breakage. Initial findings mainly under laboratory conditions, indicate that both these approaches can be of major benefit. It is finally argued that a combination of these approaches will ultimately lead to robust systems which can operate in an industrial environment.

The Application of Expert Systems and Neural Networks to Gas Turbine Prognostics and Diagnostics

Abstract: Condition monitoring of engine gas generators plays an essential role in airline fleet management. Adaptive diagnostic systems are becoming available that interpret measured data, furnish diagnosis of problems, provide a prognosis of engine health for planning purposes, and rank engines for scheduled maintenance. More than four hundred operations worldwide currently use versions of the first or second generation diagnostic tools. Development of a third generation system is underway which will provide additional system enhancements and combine the functions of the existing tools. Proposed enhancements include the use of artificial intelligence to automate, improve the quality of the analysis, provide timely alerts, and the use of an Internet link for collaboration. One objective of these enhancements is to have the intelligent system do more of the analysis and decision making, while continuing to support the depth of analysis currently available at experienced operations. This paper presents recent developments in technology and strategies in engine condition monitoring including: (1) application of statistical analysis and artificial neural network filters to improve data quality, (2) neural networks for trend change detection, and classification to diagnose performance change, and (3) expert systems to diagnose, provide alerts and to rank maintenance action recommendations.

Wavelet packet feature extraction for vibration monitoring

Abstract: Condition monitoring of dynamic systems based on vibration signatures has generally relied upon Fourier based analysis as a means of translating vibration signals in time domain into the frequency domain. However, Fourier analysis provided a poor representation of signals well localized in time. The wavelet packet transform is introduced as an alternative means of extracting time-frequency information from vibration signature. Moreover, with the aid of statistical based feature selection criteria, a lot of feature components containing little discriminant information could be discarded resulting in a feature subset with reduced number of parameters. This significantly reduces the long training time that is often associated with neural network classifier and increases the generalization ability of the neural network classifier.

Optimizing condition‐based maintenance decisions for equipment subject to vibration monitoring

Abstract: The paper reports the development of an optimal maintenance program based on vibration monitoring of critical bearings on machinery in the food processing industry. Statistical analysis of vibration data is undertaken using the software package EXAKT to establish the key vibration signals that are necessary for risk estimation. Once the risk curve is identified using a proportional hazards model, cost data are then blended with risk to identify the optimal maintenance program. The structure of the decision making software EXAKT is also presented. Concludes that perhaps the most important benefit of the study was the realization by maintenance management that it is possible to identify key measurements for examination at the time of vibration monitoring – thus possibly saving on inspection costs.

Tire condition monitoring system

Abstract: A tire condition monitoring system includes a monitoring device securely positioned within a tire and in electronic communication with a receiver and a cab mounted visual display. In one embodiment, the monitoring device includes a battery, an inductive pick-up coil, pressure and/or temperature sensors, a microprocessor, and a data transmitter. Each monitoring device has a unique multi-bit identification code. The monitoring device is shipped in an energy conserving dormant state until activated either by pressurization of the tire or by use of a portable hand-held wand transmitter. The hand-held transmitter assigns the monitoring device a relative tire position. After activation, the monitoring device periodically senses tire condition. This information is stored and compared to preset parameters and the last stored tire condition information. The tire information is periodically sent to the receiver and visually displayed. If the sensed tire condition information deviates from preset parameters, the sensed tire information is immediately telemetered to the receiver and an alarm is activated. During prolonged periods of vehicle and tire inactivity, the monitoring device measures and transmits less frequently to preserve power. The monitoring device automatically reactivates upon vehicle reactivity. In another embodiment, a monitoring device having a mechanical air pressure sensor, a motion detector, a battery and a transmitter is positioned within the tire. Once the tire is traveling at a predetermined velocity, power is supplied to the mechanical air pressure sensor. When the tire pressure drops below a predetermined level, the mechanical air pressure sensor provides power to the transmitter which generates a signal to a receiver and cab mounted alarm.

Dynamic Prognostic Prediction of Defect Propagation on Rolling Element Bearings

Abstract: Rolling element bearing failure is a major factor in the failure of rotating machinery. Current methods of bearing condition monitoring focus on determining any existing fault presence on a bearing as early as possible. Although a defect can be detected when it is well below the industry standard of a fatal size of 6.25 mm2 (0.01 in2), the remaining life of a bearing (the time it takes to reach the final failure size) from the point where a defect can be detected can vary substantially. As a fatal defect is detected, it is common to shut down machinery as soon as possible to avoid catastrophic consequences. Performing such an action, which usually occurs at inconvenient times, typically results in substantial time and economics losses. It is, therefore, important that the bearing's remaining life be more precisely forecasted, in a prognostic rather than diagnostic manner, so that maintenance can be optimally scheduled. Unfortunately, current bearing remaining life prediction methods have not been well dev...

Condition monitoring of low-speed rotating machinery using stress waves Part 1

Abstract: Condition monitoring through the use of vibration analysis is an established and effective technique for detecting the loss of mechanical integrity of a wide range and classification of rot...

Three state-of-the-art methods for condition monitoring

Abstract: This paper describes and compares three different state-of-the-art condition monitoring techniques: first principles, feature extraction, and neural networks. The focus of the paper is on the application of the techniques, not on the underlying theory. Each technique is described briefly and is accompanied by a discussion on how it can be applied properly. The discussion is finished with an enumeration of the advantages and disadvantages of the technique. Two condition monitoring cases, taken from the marine engineering field, are explored: condition monitoring of a diesel engine, using only the torsional vibration of the crank shaft, and condition monitoring of a compression refrigeration plant, using many different sensors. Attention is also paid to the detection of sensor malfunction and to the user interface. The experience from the cases shows that all techniques are showing promising results and can be used to provide the operator with information about the monitored machinery on a higher level. The main problem remains the acquisition of the required knowledge, either from measured data or from analysis.

Off/on-line FRA condition monitoring technique for power transformer

Abstract: This letter presents a novel and sensitive frequency response analysis (FRA) technique for off/on-line condition monitoring of expensive power apparatus. The main objective is to investigate the applicability of this predictive maintenance technique to diagnosing power transformer failures, to characterize the transformer in a frequency plane for safe operation, and to develop on-line monitoring technique. The effectiveness of this technique is demonstrated on a 2.5 MVA transformer housed at the High Voltage Laboratory of Pacific Power International (PPI) and then tested on three field transformers. The on-line technique is verified on a single-phase high voltage transformer. This study shows that FRA results are very sensitive to faults. The configuration of the measurement set up, frequency range and terminations play a greater role in getting reproducible, fault-indicating results. The interpretation of the voluminous data and analysis relevant to the faults opens a gateway to develop smart power apparatus.

Bispectral and trispectral features for machine condition diagnosis

Abstract: The application of bispectral and trispectral analysis in condition monitoring is discussed. Higher-order spectral analysis of machine vibrations for the provision of diagnostic features is investigated. Experimental work is based on vibration data collected from a small test rig subjected to bearing faults. The direct use of the entire bispectrum or trispectrum to provide diagnostic features is investigated using a variety of classification algorithms including neural networks, and this is compared with simpler power spectral and statistical feature extraction algorithms. A more detailed investigation of the higher-order spectral structure of the signals is then undertaken. This provides features which can be estimated more easily in practice and could provide diagnostic information about the machines.

Robust machine fault detection with independent component analysis and support vector data description

Abstract: We propose an approach to fault detection in rotating mechanical machines: fusion of multichannel measurements of machine vibration using independent component analysis (ICA), followed by a description of the admissible domain (part of the feature space indicative of normal machine operation) with a support vector domain description (SVDD) method. The SVDD method enables the determination of an arbitrary shaped region that comprises a target class of a dataset. In this particular application, it provides a way to quantify the compactness of the admissible class in relation to data preprocessing. Application to monitoring of a submersible pump indicates that combination of measurement channels with ICA gives improved results in fault detection, without requiring detailed prior knowledge on origin and type of the failure.

Condition monitoring of transformers in service by the low voltage impulse test method

Abstract: Low voltage impulse testing (LVI) on power transformers can be used to detect winding movement, such as that caused by short circuits or by loss of winding clamping pressure. The technique is being developed to perform tests with transformers online. The techniques are described and results presented.

Practical models for condition monitoring inspection intervals

Abstract: Markov models find optimum inspection intervals for phased deterioration of monitored complex components in a system with severe down time costs. The number of (pseudo)‐phases can be increased, but in most cases, simple models tracking actual states and their perception by the user will suffice, because of paucity of data and near‐constant rates. The matrix is cyclic; it includes renewal and regression to earlier states, simplifying solution and matching observation. An example involves roller bearings in paper mills with three phases, no defect, possible defect, and final deterioration towards failure. In the last phase, continuous monitoring is used.

Oil analysis cost‐effective machine condition monitoring technique

Abstract: This paper describes the development of the use of lubricant analytical programmes and trend analysis to optimise oil change intervals and to predict equipment failure. The various analytical methods are covered, as are the most frequently occurring lubricant applications where such condition monitoring programmes are most appropriate.

A Bayesian probabilistic approach to structural health monitoring

Abstract: Some general issues associated with online structural health monitoring are discussed. In order to address the problem of determining the existence and location of damage in the presence of uncertainties, a global model-based structural health monitoring method which utilizes Bayesian probabilistic inference is developed. The results of tests using simulated data are described.

Methods of condition monitoring and fault diagnosis for induction motors

Abstract: The study gives a synopsis over condition monitoring methods both as a diagnostic tool and as a technique for failure identification in high-voltage induction motors. Methods offault detection are reviewed and analyzed: mechanical (vibrations, shock pulses, acoustic, speed fluctuations ), electro-mechanical (currents, surges, partial discharges, leakage fluxes) and temperature-, oil particle-, gas analysis, and performance- and visual methods. The study indicates how different types of faults generate different frequency components in vibration spectra, in motor supply current, in torque and in magnetic field, and relationships between fault types and spectra have been developed. Supply current waveform analysis has been applied to demonstrate the detection of broken rotor bars and bearing faults.

Wavelet packet feature extraction for vibration monitoring

Abstract: Condition monitoring of dynamic systems based on vibration signatures has generally relied upon Fourier based analysis as a means of translating vibration signals in time domain into the frequency domain. However, Fourier analysis provided a poor representation of signals well localized in time. In this case, it is difficult to detect and identify the signal pattern from their expansion coefficients because the information is diluted across the whole basis. The wavelet packet transform is introduced as an alternative means of extracting time-frequency information from vibration signature. Moreover, with the aid of statistical based feature selection criteria, a lot of feature components containing little discriminant information could be discarded resulting in a feature subset with reduced number of parameters. This significantly reduces the long training time that is often associated with neural network classifier and increases the generalization ability of the neural network classifier.

An assessment of turbomachinery condition monitoring and failure prognosis technology

Abstract: This paper provides a broad overview of developments and progress in condition monitoring, diagnostics and failure prognosis technology applicable to high performance turbomachines. An assessment is made of current technological capabilities in this critical area. Selected maintenance philosophies, including Condition Based Maintenance, Reliability Centered Maintenance and Profit Centered Maintenance are discussed. Available diagnostic technologies applicable to condition monitoring of turbomachinery are described. Some observations on technological gaps and problems yet to be solved are presented; available information resources and organizations with active related programs are identified.

Neural-network approach for predicting hole quality in reaming

Abstract: Condition monitoring of the machining process is very important in today's precision manufacturing, especially in the case of reaming where in-process measurement of surface quality is difficult. In this paper, a new approach is presented for the condition monitoring in reaming using Artificial Neural Network. Acoustic emission, cutting force and vibration sensor data were measured during reaming operation and a multi-layer neural network was trained using these data with the conventional Back Propagation Algorithm for weight updation. Using force, vibration and acoustic emission parameters as input and Ra value of roughness, roundness error and residual stress as output, the network gave much superior results with sensor fusion.

Condition monitoring of low-speed rotating machinery using stress waves Part 2

Abstract: The mechanical integrity of rotating biological contractors (RBCs) and their bearings are vital to maintaining uninterrupted operation. Part 1 of this work presented a study of the high-fre...

Application of fuzzy logic to induction motors condition monitoring

Abstract: This letter applies fuzzy logic to induction motors condition monitoring. The preliminary results show that fuzzy logic can be used for accurate induction motors fault diagnosis if the input data are processed in an optimized way.

An integrated approach to machine fault diagnosis

Abstract: This paper introduces an integrated methodology to monitor and diagnose machine faults in complex industrial processes such as textile and fiber manufacturing facilities. The approach is generic and applicable to a variety of industrial plants that operate critical processes and may require continuous monitoring and maintenance procedures. A dual approach is pursued: high-bandwidth fault symptomatic evidence, such as vibrations, current spikes, etc., are treated via a feature extractor/neural network classifier construct; while low-bandwidth phenomena, such as temperature, pressure, corrosion, leaks, etc., are better diagnosed with a fuzzy rule base set as an expert system. The technique is illustrated with typical examples from benchmark processes common to many industrial plants.

Conditional health monitoring using vibration signatures

Abstract: Condition health monitoring of dynamic systems based on vibration signatures has generally relied upon Fourier based analysis as a means of translating vibration signals in time domain into the frequency domain. The wavelet packet transform is introduced as an alternative means of extracting time-frequency information from vibration signatures. Moreover, with the aid of statistical based feature selection criteria, many feature components containing little discriminant information could be discarded resulting in a feature subset with reduced number of parameters. This significantly reduces the long training time that is often associated with neural network classifier and increases the generalization ability of the neural network classifier. To validate the feature extraction algorithm proposed, the simulations have been performed with the benchmark data known as Westland vibration data set. The results show significant improvement when the data is subjected to various white, colored and pink noise.

Gas turbine engine and sensor fault diagnosis

Abstract: Substantial economic and even safety related gains can be achieved if effective gas turbine performance analysis is attained. During the development phase, analysis can help understand the effect on the various components and on the overall engine performance of the modifications applied. During usage, analysis plays a major role in the assessment of the health status of the engine. Both condition monitoring of operating engines and pass off tests heavily rely on the analysis. In spite of its relevance, accurate performance analysis is still difficult to achieve. A major cause of this is measurement uncertainty: gas turbine measurements are affected by noise and biases. The simultaneous presence of engine and sensor faults makes it hard to establish the actual condition of the engine components. To date, most estimation techniques used to cope with measurement uncertainty are based on Kalman filtering. This classic estimation technique, though, is definitely not effective enough. Typical Kalman filter results can be strongly misleading so that even the application of performance analysis may become questionable. The main engine manufactures, in conjunction with research teams, have devised modified Kalman filter based techniques to overcome the most common drawbacks. Nonetheless, the proposed methods are not able to produce accurate and reliable performance analysis. In the present work a different approach has been pursued and a novel method developed, which is able to quantify the performance parameter variations expressing the component faults in presence of noise and a significant number of sensor faults. The statistical basis of the method is sound: the only accepted statistical assumption regards the well known measurement noise standard deviations. The technique is based on an optimisation procedure carried out by means of a problem specific, real coded Genetic Algorithm. The optimisation based method enables to concentrate the steady state analysis on the faulty engine component(s). A clear indication is given as to which component(s) is(are) responsible for the loss of performance. The optimisation automatically carries out multiple sensor failure detection, isolation and accommodation. The noise and biases affecting the parameters setting the operating point of the engine are coped with as well. The technique has been explicitly developed for development engine test bed analysis, where the instrumentation set is usually rather comprehensive. In other diagnostic cases (pass off tests, ground based analysis of on wing engines), though, just few sensors may be present. For these situations, the standard method has been modified to perform multiple operating point analysis, whereby the amount of information is maximised by simultaneous analysis of more than a single test point. Even in this case, the results are very accurate. In the quest for techniques able to cope with measurement uncertainty, Neural Networks have been considered as well. A novel Auto-Associative Neural…