Perform Condition Monitoring Research Articles

Csapágyvizsgáló berendezés funkcióstruktúrái

A csapágyak állapotfelmérése kiemelt jelentőségű témakör. A csapágyak kísérleti vizsgálata többféleképpen történhet. Laboratóriumi körülmények között vizsgálóberendezés szükséges a csapágyak állapotfelméréséhez. Jelen cikk egy olyan eszköz kifejlesztésére fókuszál, amely alkalmas csapágyak fárasztására, valamint a mérési vizsgálatok elvégzésére egyaránt.

Normalisation of the amplitude modulation caused by time-varying operating conditions for condition monitoring

Performing condition monitoring under time-varying operating conditions is challenging. The time-varying operating conditions result in amplitude and frequency modulation which mask the presence of incipient damage and make it difficult to distinguish between changes in the condition of the machine and changes in its operating conditions. In this work, the benefits of normalising the amplitude modulation caused by the varying operating conditions for condition monitoring are illustrated and a method is proposed to perform this normalisation. It is shown that the proposed method can be used as a pre-processing methodology for deterministic-random separation, it can be used to detect incipient damage and it can be used to reliably estimate the severity of the damage under time-varying operating conditions as well. The proposed method is investigated on numerical gearbox data and experimental gearbox data, where its benefits for condition monitoring under time-varying operating conditions are shown.

Comparing Data Sources for Condition Monitoring Suitability

In our global competitive markets, industrial companies are forced to strive for the best achievable economic efficiency. This is valid for the whole product life cycle. For machine manufacturers, this often means exploring new and better ways of product maintenance, e.g. moving from scheduled or reactive maintenance towards condition-based maintenance in order to reduce the total cost of ownership. In machines with gearboxes, vibration picked up at the machine housing often serves as input signal for performing condition monitoring. However, other measurable variables such as the loaded transmission error exist, too. Even for one physical dimension such as acceleration, many different types of sensors with different properties are available. This work provides an insight on the suitability of different kinds of measurements and sensors for gear damage detection. It takes into account the specific properties of various kinds of gear damage as well as the operating point of the gearbox to be monitored. Based on the physical properties of real sensors, adequate sensor equipment for condition monitoring is selected.

A pre-processing methodology to enhance novel information for rotating machine diagnostics

Abstract Many sophisticated signal analysis techniques are developed to efficiently detect, localise and trend damage in rotating machine components such as bearings and gears for example. However, these techniques are generally applied without effectively incorporating historical information when performing condition monitoring. It is possible to enhance the performance of the analysis techniques by incorporating historical data from a machine in a reference condition. In this paper, a methodology is proposed to extract a novel signal i.e. a signal that contains information that is not present in the historical reference data, from a vibration signal. This is performed by utilising the available historical data. Sophisticated signal analysis techniques can subsequently be used on the novel vibration signal to diagnose the machine. The benefits of the methodology are illustrated on data, generated from phenomenological gearbox model data and experimental gearbox data, by utilising advanced techniques based on cyclostationary analysis. The results indicate that the novel vibration signal is more sensitive to damage, which highlights its potential as a pre-processing technique for rotating machine applications where historical data are available.

Diagnosis of Bearing Outer Race Faults Using a Low-Cost Non-Contact Method with Advanced Wavelet Transforms

This study analyses the vibration of Squirrel Cage Induction Motor (SCIM) with bearing faults using high-frequency signal (Handheld UWB radar) and Software Phase Locked Loop algorithm (SPLL). The contact or non-contact methods perform condition monitoring of the SCIM. The proposed method is a non-contact technique to perform condition monitoring of the SCIM. The contact methods execute via vibration, instantaneous frequency, rotor speed and flux signals analysis; whereas non-contact methods accomplish via acoustic, current and stray flux measurement. The existing techniques suffer from the influence of adjoining electrical machines; require human expertise to mount sensors and analysing the signals. In this paper, a new, non-contact method proposed for bearing fault identification in the SCIM. The proposed method uses a high-frequency signal projected on the motor and the reflected signal captured. The signal obtained is analysed with an advanced signal processing algorithm like Rational Dilation Wavelet Transform (RDWT) to identify the faults in the SCIM. The signal energy at the fault frequency level increases from 4.72 % to 5.82 % with the increase in the number of the faults. The signal energy variation indicates the severity of the faults. From the experimental results, the bearing fault of the motor identified in the beginning stage of the fault.DOI: 10.5755/j01.eie.25.1.22735

Localised gear anomaly detection without historical data for reference density estimation

Abstract Performing condition monitoring on rotating machines which operate under fluctuating conditions remains a challenge, with robust diagnostic techniques being required for the condition inference task. Some of the developed diagnostic techniques rely on the availability of historical fault data, which is impractical or even impossible to obtain in many circumstances and therefore novelty detection techniques such as discrepancy analysis are used. However, discrepancy analysis assumes that the condition of the machine is the same throughout the signal in the model optimisation process i.e. no localised damage is present, which can pose problems if the training data unwittingly contain a component with localised damage. In this paper, an automatic procedure is proposed for diagnosing localised gear damage in the presence of fluctuating operating conditions, with no historical data being required to model the data used in the condition inference process. The continuous wavelet transform, principal component analysis and information theory are used to obtain divergence data of the gear under consideration. The divergence data are used with Bayesian data analysis techniques to automatically infer the presence of localised anomalies due to localised gear damage. The proposed technique is validated in two experimental investigations, with promising results being obtained.

Mapping of the electromagnetic environment on the railway: Condition monitoring of signalling assets

Conventional track circuit condition monitoring systems are fixed at the wayside, with each installation reporting on a single track circuit. In this work, we present a custom-built, sensitive, magnetic field detection system, which can be fitted to the underside of a rail vehicle. With this system installed, some characteristics of an operating track circuit can be monitored from the vehicle whilst it is in motion. By using appropriate analysis techniques, it is possible to identify the signatures of equipment relating to audio frequency track circuits, the topic of this work. Analysing the signatures of track circuit equipment demonstrated that there were clear differences between track circuit assets. By building on other research into the behaviours of failing track circuits, and continuing to conduct this research, the authors believe that it is possible, and beneficial, to perform condition monitoring of track circuits from low-cost equipment mounted on the train. Coupling this with advanced analysis techniques will allow predictive maintenance of track circuits with very little capital outlay.

A discrepancy analysis methodology for rolling element bearing diagnostics under variable speed conditions

Abstract Performing condition monitoring on critical machines such as gearboxes is essential to ensure that the machines operate reliably. However, many gearboxes are exposed to variable operating conditions which impede the condition inference task. Rolling element bearing component failures are important causes of gearbox failures and therefore robust bearing diagnostic techniques are required. In this paper, a rolling element bearing diagnostic methodology based on novelty detection is proposed for machines operating under variable speed conditions. The methodology uses the wavelet packet transform, order tracking and a feature modelling approach to generate a diagnostic metric in the form of a discrepancy measure. The probability distribution of the diagnostic metric, statistically conditioned on the corresponding operating conditions is estimated, whereafter the condition of the rolling bearing element is inferred. The rolling element bearing diagnostic methodology is validated on data from a phenomenological gearbox model and two experimental datasets.

On the use of temperature for online condition monitoring of geared systems – A review

Gear unit condition monitoring is a key factor for mechanical system reliability management. When they are subjected to failure, gears and bearings may generate excessive vibration, debris and heat. Vibratory, acoustic or debris analyses are proven approaches to perform condition monitoring. An alternative to those methods is to use temperature as a condition indicator to detect gearbox failure. The review focuses on condition monitoring studies which use this thermal approach. According to the failure type and the measurement method, it exists a distinction whether it is contact (e.g. thermocouple) or non-contact temperature sensor (e.g. thermography). Capabilities and limitations of this approach are discussed. It is shown that the use of temperature for condition monitoring has a clear potential as an alternative to vibratory or acoustic health monitoring.

Convolutional Neural Network Based Fault Detection for Rotating Machinery

Vibration analysis is a well-established technique for condition monitoring of rotating machines as the vibration patterns differ depending on the fault or machine condition. Currently, mainly manually-engineered features, such as the ball pass frequencies of the raceway, RMS, kurtosis an crest, are used for automatic fault detection. Unfortunately, engineering and interpreting such features requires a significant level of human expertise. To enable non-experts in vibration analysis to perform condition monitoring, the overhead of feature engineering for specific faults needs to be reduced as much as possible. Therefore, in this article we propose a feature learning model for condition monitoring based on convolutional neural networks. The goal of this approach is to autonomously learn useful features for bearing fault detection from the data itself. Several types of bearing faults such as outer-raceway faults and lubrication degradation are considered, but also healthy bearings and rotor imbalance are included. For each condition, several bearings are tested to ensure generalization of the fault-detection system. Furthermore, the feature-learning based approach is compared to a feature-engineering based approach using the same data to objectively quantify their performance. The results indicate that the feature-learning system, based on convolutional neural networks, significantly outperforms the classical feature-engineering based approach which uses manually engineered features and a random forest classifier. The former achieves an accuracy of 93.61 percent and the latter an accuracy of 87.25 percent.

Data-driven mono-component feature identification via modified nonlocal means and MEWT for mechanical drivetrain fault diagnosis

It is significant to perform condition monitoring and fault diagnosis on rolling mills in steel-making plant to ensure economic benefit. However, timely fault identification of key parts in a complicated industrial system under operating condition is still a challenging task since acquired condition signals are usually multi-modulated and inevitably mixed with strong noise. Therefore, a new data-driven mono-component identification method is proposed in this paper for diagnostic purpose. First, the modified nonlocal means algorithm (NLmeans) is proposed to reduce noise in vibration signals without destroying its original Fourier spectrum structure. During the modified NLmeans, two modifications are investigated and performed to improve denoising effect. Then, the modified empirical wavelet transform (MEWT) is applied on the de-noised signal to adaptively extract empirical mono-component modes. Finally, the modes are analyzed for mechanical fault identification based on Hilbert transform. The results show that the proposed data-driven method owns superior performance during system operation compared with the MEWT method.

An Energy-based approach to condition monitoring of industrial processes

Condition monitoring has traditionally been deployed for the monitoring of a single component, such as a rotating electrical machine. The application of condition monitoring to an entire process plant has not yet received much attention. In this work we outline the challenges associated with plant-wide condition monitoring, and explore the possibility of using energy as the parameter of interest. The use of energy mitigates the multi-domain monitoring problem, and could also provide a hierarchical view of the process plant's condition. A simple chemical reactor is considered and it is shown that the energy parameter can be used to perform condition monitoring of such a reactor. The identified parameter is also sensitive to time domain variations. This indicates that the concept of using energy flows as the parameter of interest for process-wide condition monitoring applications has merit.

Probabilistic Novelty Detection With Support Vector Machines

Novelty detection, or one-class classification, is of particular use in the analysis of high-integrity systems, in which examples of failure are rare in comparison with the number of examples of stable behaviour, such that a conventional multi-class classification approach cannot be taken. Support Vector Machines (SVMs) are a popular means of performing novelty detection, and it is conventional practice to use a train-validate-test approach, often involving cross-validation, to train the one-class SVM, and then select appropriate values for its parameters. An alternative method, used with multi-class SVMs, is to calibrate the SVM output into conditional class probabilities. A probabilistic approach offers many advantages over the conventional method, including the facility to select automatically a probabilistic novelty threshold. The contributions of this paper are (i) the development of a probabilistic calibration technique for one-class SVMs, such that on-line novelty detection may be performed in a probabilistic manner; and (ii) the demonstration of the advantages of the proposed method (in comparison to the conventional one-class SVM methodology) using case studies, in which one-class probabilistic SVMs are used to perform condition monitoring of a high-integrity industrial combustion plant, and in detecting deterioration in patient physiological condition during patient vital-sign monitoring.

Quantifying the added value of an imperfectly performing condition monitoring system—Application to a wind turbine gearbox

Implementation of a condition monitoring system is a difficult decision due to many uncertain parameters. This is certainly the case for the wind turbine industry where factors like long logistical times and weather conditions have a major influence on the economic benefit. One of the parameters that is neglected in most of the available literature is the performance of the condition monitoring system itself. In this paper a new concept for modeling this performance based on the P–F curve of different failure modes is presented. The concept is illustrated on an extensive case study for a gearbox of a wind turbine. A stochastic simulation model is constructed in order to quantify the economic added value of implementing an imperfectly performing condition monitoring system into a gearbox. This case study proves that a condition monitoring system generates an economic benefit compared to the currently applied maintenance strategy. However, the magnitude of this benefit depends strongly on the performance of the condition monitoring system.

Diagnosis of MV Power Cables for Nuclear Power Plants

The diagnosis of MV (Medium Voltage) cable used in NPPs (Nuclear Power Plants) has become one of the most urgent issues. In accordance with PSR (Periodic Safety Review) results, continuous condition monitoring of cables is also continuously required. In this paper, HFPD (High Frequency Partial Discharge) measurement and an infrared imaging thermography test were performed to check the status of 13.8 and 4.16 kV MV cables installed in the domestic power plant. The diagnosis was found to be good for the target cable. The on-line diagnostic technology to assess the condition of MV cables during the operation of a power plant will contribute to enhance the safety of NPPs. In the future, these latest technologies will be applied to perform condition monitoring of HV (High Voltage) cables in NPPs.

Thermal sensitivity of strain failure mode in forming limit diagram and development of thermal line analysis

The suitability of temperature as a parameter to determine strain failure mode in press forming is examined by thermographic evaluation of the forming limit diagram (FLD). The work reported in the present paper forms part of a research programme to develop condition monitoring technology for press forming, the highlights of which are discussed. Experimental methodologies are described. The inability to use temperature to distinguish between the left and right hand sides of the FLD is reported. A strong correlation between major strain and fracture temperature is also observed. From this, it is concluded that it may be possible to perform condition monitoring of press forming, by a technique termed 'thermal line analysis'.

Thermal sensitivity of strain failure mode in forming limit diagram and development of thermal line analysis

The suitability of temperature as a parameter to determine strain failure mode in press forming is examined by thermographic evaluation of the forming limit diagram (FLD). The work reported in the present paper forms part of a research programme to develop condition monitoring technology for press forming, the highlights of which are discussed. Experimental methodologies are described. The inability to use temperature to distinguish between the left and right hand sides of the FLD is reported. A strong correlation between major strain and fracture temperature is also observed. From this, it is concluded that it may be possible to perform condition monitoring of press forming, by a technique termed 'thermal line analysis'.

Wavelet Packet Analysis in the Condition Monitoring of Rotating Machinery

In this paper, novel methods for performing condition monitoring for power station turbine shafts are presented. The objective of this work is to investigate methods for producing accurate turbine vibration fault alarms during turbine shaft rundowns. Wavelet packet analysis is employed to extract spectral features from healthy vibration signals and the probability density functions of these features are estimated. Both Gaussian models, using Bayesian inferencing, and mixture models are employed. Preliminary results show that the more computationally expensive mixture models produce more accurate density estimates and hence more reliable fault alarms.

Gas-turbine condition monitoring using qualitative model-based diagnosis

Gas turbines are critical to the operation of most industrial plants, and their associated maintenance costs can be extremely high. To reduce those costs and increase the availability of their gas turbines, plant operators have for many years relied on routine preventative maintenance-routinely checking and solving small problems before they grow into major ones. Recently, however, the power industry has moved sharply toward condition-based maintenance and monitoring. In this approach, intelligent computerized systems monitor gas turbines to establish maintenance needs based on the turbine's condition rather than on a fixed number of operating hours. By integrating several AI technologies-including qualitative model-based reasoning-the Tiger system significantly cuts costs and improves performance by using control-system information to perform condition monitoring for gas-turbine engines.