Search

Condition monitoring (CM) is an effective approach to prevent accidents caused by structural damage. An online condition monitoring system for suspension is vital to vehicles’ safety and reliability, as suspension is an important subsystem of the vehicle. Average Correlation Signals based Stochastic Subspace Identification (ACS-SSI) is a new approach which has the potential to achieve online CM for vehicle suspension systems. In order to investigate the influences of possible errors, like placement of sensors and excitation amplitudes, on implementing ACS-SSI for online suspension CM, a simplified test device is developed to study the performance of identifying the most three common vibration modes of a vehicle, which are bounce, pitch and roll. A three degrees of freedom (3-DOF) model were established for the devices to highlight the effects of the errors. The study results show that the ACS-SSI is an effective method to carry out system identification even if the inputs are highly noisy and non-stationary. However, the implementation of ACS-SSI needs to take into account these potential errors in order to obtain accurate CM results

In many parts of the world car manufacturers are required by law to state the fuel economy\ud performance of their vehicles when undertaking one or more Standard Drive Cycles such as the\ud Urban and Extra Urban test cycle definitions required in Western Europe. Such Standard Drive Cycles have their roots in obscure vehicle testing undertaken decades ago in the USA, and no physical test track exists in Europe which enables the precise re-creation of these test characteristics. Furthermore, the fundamental validity of these test cycles as a means of gauging and comparing passenger car fuel economy is seldom considered. This paper documents the early stages of a research project which aims to segment the Standard Drive Cycle definitions into a series of operating characteristic windows which can then be searched for and automatically extracted from general driving behaviour of a car on the road, and thereafter concatenated to re-creating a real-world equivalent of the Standard Drive Cycle without the need for rolling road or laboratory testing.

The increasing interest in alternative fuels for CI engines continues has led to many different\ud alternative fuels which are claimed to be more economical and cleaner. However, the impacts of longterm use of the alternative fuels on the reliability and service life of CI engine have not yet been fully understood.AE technology is now becoming a widely accepted practice in the field of engine. AE signals can be used to obtain information on machine performance, offering the potential to monitor operating conditions such as fuel efficiency, combustion conditions, and lubrication, and also to detect faults.\ud In this paper, the potential impacts of alternative fuels (Fischer-Tropsch fuel, methanol-diesel blended fuel, emulsified diesel and standard diesel) on acoustic emission signals of cylinder have been investigated. The impacts of the different fuels on the wear condition of cylinder were compared and analysed using the acoustic emission monitoring at the cylinder. The results provide real-time evaluation method and effective data support for the development and application of alternative fuels.

Condition monitoring of a gearbox is a crucial activity due to its importance in power transmission for many industrial applications. Thus, there has always been a constant pressure to improve measuring techniques and analytical tools for early detection of faults in gearboxes. This study focuses on developing gearbox monitoring methods using the operating parameters obtained from machine control processes rather than the traditional measures such as vibration and acoustics. To monitor the gearbox conditions, an adaptive neuro-fuzzy inference system (ANFIS) is used to capture the nonlinear connections between the electrical motor current and control parameters such as load settings and temperatures. The predicted values generated by the ANFIS model are then compared with the measured values to indicate the abnormal condition in gearbox. Furthermore, a comparative\ud study of the results this technique and the general regression neural networks (GRNN) is also carried out. The comparison results show that the ANFIS model performs more accurately than the other model in gearbox condition monitoring and fault detection.

The rolling element bearing is a key part in many mechanical facilities and the diagnosis of its faults is\ud very important in the field of machinery health monitoring for safe and efficient operations. Currently,\ud envelope analysis is widely used to obtain the bearing defect harmonics from the envelope signal\ud spectrum for the detection and diagnosis and has shown good results in identifying incipient faults\ud occurring on the different parts of a bearing. However, a critical step in implementing envelope\ud analysis is to determine a frequency band that contains faulty bearing signal component with highest\ud signal to noise. Conventionally, the choice of the band is made by manual spectrum comparison via\ud identifying the resonance frequency where the largest change occurred. This paper presents a\ud spectral kurtosis (SK) based method to determine optimum envelope analysis parameters including\ud the filtering band and centre frequency through a short time Fourier transform (STFT). The results\ud show that the maximum amplitude of the kurtogram provides the optimal parameters for band pass\ud filter which allows both small outer race fault and large inner race fault to be determined from\ud optimised envelope spectrum.

Mechanical seals have become one of the most popular sealing systems for rotating machinery\ud because of low leakage and absence of a requirement for routine maintenance. Generally, a\ud mechanical face seal should operate with a fluid film as thin as possible, to reduce the leakage and to restrict friction and wear. Recent advances in a system of computer software based on finite element modelling and analytical approaches help in understanding of the working conditions of the mechanical face seals. This paper reviews tribological bahavior and condition monitoring of mechanical seals based on the literature of the recent years. It covers friction, wear and thermal characteristics of mechanical seals and the application of computational methods and other techniques to give good understanding of the tribological behavior and condition monitoring of seal\ud faces.

At present, regenerative active suspension is more attractive than conventional suspension of road vehicles for the improvement of ride comfort, performance, stability, passenger safety and the reduction of energy dissipation with regenerative energy. In a real application, the energy dissipation results in a reduction of the performance of the vehicle as well as high-energy consumption. This paper presents a hybrid shock absorber model with a modified shock absorber, which combines a hydraulic motor with a generator to recover the energy that would be otherwise wasted from vibrational motion of the suspension and transform it into useful electricity. The instantaneous oil pressures have been evaluated in the inlet and outlet pipelines using different sinusoidal wave excitation. The feasibility of the energy recovery features of this system are investigated by measuring the hydraulic motor shaft speed and the pressures at the inlet and outlet pipelines. The existing structure will be further optimized as future work to improve its performance.

The dielectric constant and the moisture value of the lubricating oil measured from a IC engine\ud running with different alternative fuels on line show a clear difference from the baseline diesel. This demonstrates that alternative fuels will have a noticeable impact on engine lubrication. Further studies should be carried out on clarifying the influences mechanisms and corresponding resolutions.

This poster summarises the progress which has been made from 1st April 2012 to 1st Mar 2013 in\ud fulfilling the PhD research project of condition monitoring of bearing condition monitoring based on advanced vibration data analysis For improve production efficiency fault detection and diagnosis of the critical components such as bearings in the rotating machinery including compressors, pumps, power turbines and aircrafts engines are becoming very important area of research, which helps to avoid unexpected shutdowns due to a faulty process. This poster presents data analysis techniques for rotating machinery, bearing types, failure and the\ud performance of methods for failure detection in ball bearings. Different kinds of faults have been created in the ball bearing such as (Inner race fault and Outer race fault), and a raw signal of the healthy and faulty has been acquired and recorded. The vibration signals were taken at different loads such as (0%, 25%, 50%, and 75%). Monitored vibration of motor-generator system supported by deep grove ball bearings to predict bearing failures. It successfully identified failures of the ball bearing for both outer and inner races. The interim result shows that the envelope analysis spectrum gives more diagnostic information than analysis of the time domain or frequency domain.

Biodiesel is one of the alternative fuels which is renewable, environmentally and can be used in diesel\ud engines without modifications. For non-intrusive diagnosis, this study investigates the characteristics of\ud combustion induced vibration and noise in a diesel engine fuelled with biodiesel blends under different\ud operating conditions. The relationship between the engine vibro-acoustics and in-cylinder pressure was\ud investigated based on the analysis of the measured cylinder head vibration, engine acoustics and in-cylinder\ud pressure (peak pressure, pressure rise, and pressure rise rate). It has found that the biodiesel blends result\ud in a slight increase of peak value of in-cylinder pressure when increasing the percentage of biodiesel. The\ud engine running with biodiesel blends has slightly higher vibration, which is more related to the maximum rate\ud of pressure rise which is closely related to the combustion process. In addition, the time domain RMS values\ud of vibration is relating to the maximum rate of pressure rise whereas the RMS of acoustics signals is more\ud relating to peak pressure. These show that vibro-acoustics include good information about combustion.\ud However, more advanced analysis is needed to suppress noise influences for accurately representation of\ud combustions due to different types of fuels.

The major focus of condition monitoring (CM) is in its' prognostic and fault prediction bilities, the power of which is determined by selecting the appropriate statistical techniques. Making the correct distributional and theoretical assumptions is the key to model robustness and strength of inference. Likewise a condition based rather than time interval based maintenance regime ensures near optimal performance for the duration of process operation, the ever increasing drive for environmental supremacy and the highest safety standards alongside increasing global competition forces companies to strive for the greatest possible achievements in both performance and quality. Decisions are made daily about manufacturing processes in the presence of variability, process specific knowledge allows quantification of the risks associated with various courses of action while statistics provides a common language to communicate information. A process is considered to be in\ud a state of statistical control if any variations in observed measurements can be attributed to chance variation only not assignable variability which it is feasible to detect and identify (loose belt, leaky valve etc). Thus a process under statistical control operating under the influence of common causes of variation only should be operating within the upper and lower control limits. Optimising process performance in all aspects it is essential that deviations from the norm and or mechanical faults are detected, quantified and corrected in a timely manner.\ud Condition monitoring provides a non-intrusive method of extracting data which has zero impact on the operation, (for example the analysis of signal data of reciprocating compressors) via appropriately inserted sensors, is ideal and gleans relevant information as to the state of the process under observation, enabling assessment of the current condition of the system, prediction of failures and/or remaining operating time. Through suitably established statistical models condition predictors and\ud reliability prognosis in systems is possible and of vital importance in maximising all aspects of\ud efficiency.

Gearing components play a pivotal role in most power transmission mechanisms but their application in industry consistently results in significant disruption and losses due to gear failures. This emphasises the significance of condition monitoring and fault diagnostics techniques of gear transmission systems to enhance overall safety and operational reliability in order to minimise gearbox failure rates and their associated disruption and losses to industry. In a wide range of industrial contexts, vibration analysis is extensively used for machinery condition monitoring and\ud diagnostics due to its good detection results. As vibration and acoustic noise have the same\ud generation mechanism, acoustic noise can also be used for machinery condition monitoring combined\ud with effective signal processing methods. In this paper, vibration and acoustic signals were both used to analyse the fatigue process of a gear transmission system based on synchronous vibro-acoustic measurements. In order to enhance the signal-to-noise ratio (SNR) of the measured signals, time synchronous average (TSA) technique was employed to pre-process the vibration and acoustic signals. The side band energy ratio (SER) is then extracted to predict the vibration and acoustic signals to indicate the fatigue process of the testing gearbox under different operating conditions. Further the fatigue process was detected through analysing the measured signals in the high frequency band as these are less contaminated by background reverberation interferences. The key results show that the gear transmission system can be monitored by vibration and acoustic analysis which show similar trend results for the gear fatigue process tested.

Condition monitoring (CM) is an effective strategy to maintain the performances of modern industrial machines. Currently, wired online condition monitoring has been successfully employed in large industrial machines. However, the high cost of these systems have restricted their wide applications, among which, the installation and maintenance cost of the communication cables are one of the major\ud factors. With the quick development of the wireless and electronics technology, communication cables could be replaced by wireless transmission, which brings several inherent advantages, such as low cost, convenience of installation and easy to replace or upgrade. One key issue of implementing\ud wireless communications in CM system is the selection of suitable wireless protocols. In this paper, the current popular wireless protocols and their applications in CM are investigated. IEEE802.15.4 is proved to be the most suitable protocol to establish wireless CM systems because of its low power consumption, low cost and its capability of setting up large scale networks. In the meantime, its bandwidth limitation can be overcome by a scheme of embedding more local processing. To evaluate the scheme, a prototype wireless CM system is developed to embed the algorithm of envelope analysis into the node for rolling bearing monitoring. The results show that the amount of data communication over the wireless network can be reduced by 95% while it upholds the performance to monitor the simulated bearing fault in real-time.

The purpose of this research is to develop a condition monitoring and fault diagnosis system for centrifugal pumps based on vibration measurement. An enhanced test-rig was built in the lab and equipped with the necessary high-quality sensors and equipment in order to examine and compare the efficacy of the monitoring methods. In addition, experimental study has been conducted to investigate condition monitoring system. Multiple tests have been presented to find a set of consistent parameters for fault diagnosis at this stage based on vibration signal data with healthy case as\ud baseline measurements only. The parameter of interest, e.g. vibration, was measured for different flow rates with a constant pump\ud speed of 2900 rpm. The flow rate was adjusted by the throttling valve in the discharge line stage by stage; visual observation of the flow through the transparent steel pipe was maintained. A test run consisted of recording at least 8 sets of data which covered a flow rate range from 350 l/min to 0 l/min. The data records from each test were processed in Matlab, the performance of the pump with the healthy condition has been achieved by developing a condition monitoring system, moreover faults of pump components will be detected and diagnosis in future work.

The incorrect inflation pressure in tyre affects vehicle handling, passenger comfort and braking conditions, as\ud well as reducing fuel efficiency and tyre life. To address this problem a numerical model has been\ud developed. The model has been developed with 7-DOF for a full car system. State space approach has been\ud adapted to analyze the system parameters using MATLAB programs. To validate the model, the\ud experimental work has been conducted by driving a car on the road with bumps with healthy tyres condition.\ud During the experiment the tyre pressure was controlled and the suspension vibration was measured. The\ud model has been validated fairly using an experimental data in time domain. For detail analysis, the Short-\ud Time Fourier Transform (STFT) method was used to obtain contents due to road excitation at the bumps.\ud The results show, by reducing the tyre pressure by 35%, the total vibration energy transmitted to the car is\ud increased by 27 %, 30.5 % and 28.2 % for reduction of tyre pressure of driver wheel, passenger wheel and\ud front wheels respectively. In addition, the result shows that the mathematical modelling can have a potential\ud for developing an on-line suspension condition monitoring tool in future.

Motor current signature analysis (MCSA) is a common practice in industry for finding motor faults. However, because of small modulations caused by faults and high noise contamination, it is difficult to quantify the modulation in measured signals which is dominated by the supply frequency, higher order harmonics and noise. In this paper a modulation signal bispectrum (MSB) is investigated to detect stator winding faults. This type of fault can cause high winding temperatures which may effect on current signal so motor temperature will be considered in this paper. The results show that MSB has the capability to accurately estimate modulation degrees and suppress the random and non-modulation components. The test results show that MSB has a better performance in differentiating spectrum amplitudes due to stator faults, and hence produces better diagnosis performance, compared with that of conventional power spectrum analysis.

Acoustic emission (AE) technology is one of the most established diagnostic techniques for\ud rolling bearing monitoring in rotating machinery. The application of high-frequency AE for\ud bearing diagnosis is gaining acceptance as a useful complimentary tool. This paper demonstrates\ud the use of AE measurements to investigate the AE characteristics of self-aligning journal bearings\ud under different rotational speed, radial load and lubrication condition. To undertake this task, a\ud purpose-built test rig was employed for collecting AE signals from the journal bearings. Then, the\ud collected data was processed using time domain and frequency domain analysis methods which\ud are of the most common techniques used for monitoring in AE applications. The results shown\ud that the data analysis method applied in this work is effective for characterising complicated AE\ud signals. Based on obtained results, it is concluded that the AE energy levels in high frequency\ud range higher for the higher radial load and speed condition. For different lubricant cases AE\ud energy becomes high when the viscosity is lower, which means that AE can be used to detect\ud lubrication degradation in journal bearings.

Maintaining correct inflation pressure in tires\ud helps to keep vehicle handling, passenger comfort and\ud braking at its best, as well as improving fuel efficiency and\ud tire life. Therefore it is very important that the tires are\ud correctly inflated. To address this problem, in the present\ud study experimental investigations have been carried out to\ud identify the effect of under-inflated tires on suspension\ud system performance by measuring vertical acceleration\ud (vibration) of the suspension system. The experimental work\ud has been conducted by driving a car on the road with range\ud of inflation pressure at four conditions (e.g. at standard\ud pressure (2.3bar) and at 1.5bar of passenger wheel, driver\ud wheel and front wheels). During the experiment the signals\ud of tire pressure, the suspension vibration and the car speed\ud were measured.\ud To analyze the signals, the Short-Time Fourier Transform\ud (STFT) method was used to detect the effects of tire inflation\ud on the performance of suspension. The STFT technique has\ud provided time-frequency information and the frequency\ud components of the signal with relative maximum energy\ud transmitted to the car body.\ud The results show that, when the tires were under-inflated by\ud approximately 35 percent, it was significantly increased the\ud vibration of the car body by up to 30 percent. The analysis\ud has also shown that STFT analysis is more accurate than the\ud spectrum analysis for distinguishing of the tire fault. These\ud effective measurements potentially will use to develop an online\ud condition monitoring system in future.

This paper develops a new scheme for the\ud lightweight design of heavy dump truck frames based on the\ud characteristics of dynamic responses. The dynamic response\ud is predicted using a finite element (FE) model which is\ud verified by an experimental mode analysis. The FE model is\ud then used to investigate the characteristics of dynamic\ud responses and frame weight changes with respect to the\ud mass changes of each frame component for selecting\ud significant components. An optimization is finally developed\ud for the lightweight design under constrains that maintain\ud required dynamic responses and static strength. The\ud optimization results show that the weight of frame can be\ud reduced by 8%, showing that the scheme is an effective way\ud to achieve automotive lightweight design.

Vibrations of a defective rolling bearing often\ud exhibit nonstationary and nonlinear characteristics which are\ud submerged in strong noise and interference components. Thus,\ud diagnostic feature extraction is always a challenge and has\ud aroused wide concerns for a long time. In this paper, the\ud multifractal detrended fluctuation analysis (MF-DFA) is\ud applied to uncover the multifractality buried in nonstationary\ud time series for exploring rolling bearing fault data.\ud Subsequently, a new approach for fault diagnosis is proposed\ud based on MF-DFA and Mahalanobis distance criterion. The\ud multifractality of bearing data is estimated with the\ud generalized Hurst exponent and the multifractal spectrum.\ud Five characteristic parameters which are sensitive to changes\ud of bearing fault conditions are extracted from the spectrum\ud for diagnosis of fault sizes. For benchmarking this new\ud method, the empirical mode decomposition (EMD) method is\ud also employed to analyze the same dataset. The results show\ud that MF-DFA outperforms EMD in revealing the nature of\ud rolling bearing fault data.

Condition monitoring of a gearbox is a very\ud important activity because of the importance of gearboxes in\ud power transmission in many industrial processes. Thus there\ud has always been a constant pressure to improve measuring\ud techniques and analytical tools for early detection of faults in\ud gearboxes. This study forces on developing gearbox monitoring\ud methods based on operating parameters which are available in\ud machine control processes rather than using additional\ud measurements such as vibration and acoustics used in many\ud studies. To utilise these parameters for gearbox monitoring,\ud this paper examines a model based approach in which a data\ud model has been developed using a General Regression Neural\ud Network (GRNN) to captures the nonlinear connections\ud between the electrical current of driving motor and control\ud parameters such as load settings and temperatures based on a\ud two stage helical gearbox power transmission system. Using the\ud model a direct comparison can be made between the measured\ud and predicted values to find abnormal gearbox conditions of\ud different gear tooth breakages based on a threshold setup in\ud developing the model.

The induction motor is one of the most used electric machines in the industry because of its strong and simplicity. This paper investigates the performance of conventional techniques such as sideband analysis in detecting broken rotor bars when the motor is fed from a common pulse width modulation voltage source inverter PWM-VSI drive for variable speed drives. The phase current signals are obtained under both the slip compensation mode and non-compensation mode. The spectra are compared between the healthy and faulty motors in the frequency domain. For the non-slip compensation, it has shown clear sidebands of the twice slip frequency which allows the detection of the broken bar and hence rotor faults under different speeds and higher loads. However, for slip compensation, the sideband pattern does not exist anymore and hence new features have to be investigated for fault detection when motors operate under these operating modes.

Motor current signal analysis has been an effective way for many years to monitor electrical machines. However, little research work has been reported in using this technique for monitoring variable speed drives and their downstream equipment. This paper investigates the dynamic responses of the electrical current signals measured from a variable speed drive for monitoring the faults from a downstream gearbox. An analytical study is firstly presented in the paper to show the characteristics of the current\ud signals due to load variation, fault effects and signal phase variation. Experimental study is then conducted under different gear fault conditions to explore the changes of the signals. Both conventional spectrum analysis and an amplitude modulation (AM) bispectrum representation are used to highlight the changes\ud for reliable fault detection. It has been found experimentally that mechanical faults lead to much higher increases in bispectral amplitudes compared to conventional spectra and hence that detection performance of the AM bispectrum is better when the drive operates non-slip compensation mode. For slip compensation, more accurate signal analysis techniques have to be developed to differentiate the small changes in the signals.

Reciprocating compressors are widely used in industry for various purposes and faults occurring in them can degrade performance, consume additional energy, and even cause severe damage to the machine. This paper will develop an automated approach to condition classification of a reciprocating compressor based on vibration measurements. Both the time domain and frequency domain techniques have been applied to the vibration signals and a large number of candidate features have been obtained based on previous studies. A subset selection method has then been used to configure a probabilistic neural network (PNN), with high computational efficiency, for effective fault classifications. The results show that a 95.50% correct classification between four different faulty cases is the best result when using a subset of frequency feature, whereas a 93.05% rate is the best for the subset from the time domain.

The aim of this research is to investigate microphone array techniques based condition monitoring (CM). In this paper, a cost effective microphone array configured with 4-sensors and data acquisition systems is investigated theoretically and experimentally to evaluate its performance in characterising sound sources in different acoustic environments. Time delay estimation (TDE) is used as a basic tool for calculating the time delay of the sound source arrived at each of the 4-sensors and hence the\ud localization parameters of the sound source such as azimuth, elevation and distance can be\ud calculated from the TDE information. Results are presented for experimental studies both in a semianechoic chamber, an outdoor environment and an enclosure. Comparing the results obtained under\ud different conditions, it has confirmed that the microphone array system can catch the location of the main sound source and also has ability to track and give the azimuthal trajectory of a moving target with high degree of accuracy.

This paper studies online condition monitoring of engine oil by analysing the acoustic emission (AE)\ud signals. The AE signals were measured using an AE sensor mounted on the thrust side between second and third cylinders in a four cylinder diesel engine. The signals were then filtered and\ud transformed to the frequency domain, where the amplitudes of the different frequency components of the AE waveforms were analysed and compared to the AE baseline signatures. The mean amplitudes\ud of the spectral components in the frequency band 10kHz to 50kHz were found linearly proportional to\ud the engine speed and load. It was also found that the RMS values of this frequency band were\ud affected by the oil viscosity. These results show that it is possible to use AE measurement to predict the quality of lubrication.

Engineering dataset has growing rapidly in size and diversities with data acquisition technology development in recent years. However the full use of the datasets for maximizing machine operation and design has not been investigated systematically because of the complexity of datasets and large scale of data amount. This also means that traditional statistical data analysis methods are no longer the efficient methods to obtain useful knowledge from the dataset. Therefore this study will focus on applying more advanced data minting technologies and optimizing database systems so that more accurate and efficient knowledge can be extracted from engineering datasets for machine performance and condition monitoring. For the first year study, a full understanding engineering dataset has been obtained based on the condition monitoring activities in the DERG laboratory.\ud In the mean time a review has also conducted on different techniques such as Neural networks, clustering, genetic algorithms, decision trees and support vector machines that are used widely for data mining. Moreover, to evaluate the effectiveness of using the techniques, a prototype\ud database has been developed and applied with the methods. The results obtained from this evaluation study have shown that the data mining can be efficient to obtain information for condition monitoring. But more work on developing methods to optimize both the parameters of using\ud the methods and the database organization.

Mechanical systems such as motors, pumps, engines, and turbine are all operating based on different shafts. Due to unexpected operating conditions, various faults such as cross-sectional cracks, looseness and misalignment may occur during their service life. This paper presents the analysis of collected signals from a wireless sensor and a shaft encoder and compares their responses under different degrees of shaft misalignment. The results have shown that the wireless sensor outperforms the encoder in detecting small shaft misalignment.

Biodiesel is a promising non-toxic and biodegradable alternative fuel in transport sector. Of all the biodiesel properties, specific gravity and viscosity are the most significant for the effects they have on the utilization of biodiesel fuels in unmodified engines. This paper presents models, which have been derived from experimental data, for predicting the specific gravity and dynamic viscosity of biodiesel at various temperatures and fractions. In addition a model has also been developed to predict the dynamic viscosity of the biodiesel from its specific gravity. In order to develop these models, the specific gravity and viscosity of rapeseed oil biodiesel, corn oil biodiesel and waste oil biodiesel blends (0B, 5B, 10B, 20B, 50B, 75B, and 100B) were tasted across a temperature range of 15.6oC to 90oC using EN ISO 3675:1998 and EN ISO 3104:1996 test procedures. Evaluation of the proposed models shows it has same order of accuracy as the other models published in literature.

This paper investigated, using experimental method, the suitability of acoustic emission (AE) technique for the condition monitoring of diesel engine valve faults. The clearance fault was adjusted experimentally in an exhaust valve and successfully detected and diagnosed in a Ford FSD 425 four-cylinder, four-stroke, in-line OHV, direct injection diesel engine. The effect of faulty exhaust valve clearance on engine performance was monitored and the difference between the healthy and faulty engine was observed from the recorded AE signals. The measured results from this technique show that using only time domain and frequency domain analysis of acoustic emission signals can give a superior measure of engine condition. This concludes that acoustic emission is a powerful and reliable method of detection and diagnosis of the faults in diesel engines and this is considered to be a unique approach to condition monitoring of valve performance.

The vibration signal obtained from operating machines contains information relating to machine condition as well as noise. Further processing of the signal is necessary to elicit information particularly relevant to bearing faults. Many techniques have been employed to process the vibration signals in bearing faults detection and diagnosis. Two common techniques, time domain techniques and frequency domain techniques are used in this paper to investigate bearings condition.

For gearbox condition monitoring, because the difficulty of sensor installation, it is a common practice to measure the vibration at a location far from the fault source. In addition operating the gearbox under different loads and speeds also produces the vibration signals with different components. The vibration measured in this way may be distorted significantly by the effect of signal transmission paths and the interference from other sources. The suppression of distortions is thus a key issue for the remote measurements based condition monitoring. In this paper, the influences of transducer locations and operating conditions on the vibration signal are investigated on a typical gearbox transmission system for the detection of the faults induced to the gearbox. The experimental study results show that the performance of traditional signal processing techniques is not sufficient to reveal fault detection information. However, the new feature from the spectrum of time synchronous average (TSA) signal is very effective in suppressing noise and hence leads to better detection of the local faults induced to the gear system.

Spurious noise (rattles, squeaks, etc) in the interior of car cabinets can be annoying, distractive and\ud indicative of potential performance problems. Fault finding these problems can be difficult since the\ud fault is intermittent and may not necessarily happen under test conditions. A 3 channel system that\ud can record the fault and indicate its location within the cabinet is presented here. The system\ud consists of a coincident microphone array that measures acoustic particle velocity along two\ud orthogonal axes at its location. Detection and direction of noise can be obtained in real time or\ud during post-processing using advanced signal analysis methods. Measurements inside a vehicle\ud show that the very reflective nature of the sound-filed inside a car cabinet present a major challenge\ud and that a combination of advance techniques from diagnostics engineering and room acoustics are\ud required to reliably indicate the direction of the annoying sound.

The rolling element bearing is a key part of machines. The accurate and timely diagnosis of its faults is critical for predictive maintenance. Most re-searches have focused on the fault location identification. To estimate the fault severity accurately, this paper focuses on the study of roller bearing vibration amplitude under increasing radial clearances due to inevitable wear using the modulation signal bispectrum (MSB). The experiment is carried out for bearings with two different clearances for the inner race fault and the outer race fault cases. The results show that the vibration amplitudes at fault characteristic frequencies exhibit significant changes with increasing clearances. However, the amplitudes of vibrations tend to increase with the severity of the outer race fault and decrease with the severity of the inner race fault. Therefore, it is necessary to take into account these effects in diagnosing the size of defect.

This paper investigates the dynamics of the engine driveline system for achieving accurate and online diagnostics of engine conditions using instantaneous angular speed (IAS). Based on a state space modelling approach and key structures of the system, a torsional vibration model is developed and its modal properties: modal frequency, damping ratio and shapes are evaluated. Then taking the output data from the model as the input a latest stochastic subspace identification (SSI) method was implemented and result shows that the SSI approach is sufficiently accurate to extract these modal properties. Moreover, experimental studies show that the SSI approach can correctly extract modal properties under the low frequency range of interest based on just the transient IAS data acquired during engine shutdown, which confirms the theoretical analysis and provides the basis for the subsequent IAS based diagnosis development in which modal characteristics will be taken into account.

To accurately detect and diagnose bearing faults, bispectral analysis has received more attention recently because of its unique property of noise reduction and nonlinearity extraction. Particularly this study investigates two typical bispectra: conventional bispectrum (CB) and modulation signal bispectrum (MSB) for suppressing noise influences in envelope signals and hence obtaining more accurate diagnostic features. The first component from the diagonal slice of CB results and that of the subdiagonal slices of MSB results are taken as the diagnostic features considering effective inclusion of information and easy of computations. Simulative and experimental studies show that both MSB and CB features result in good diagnostic performances but MSB may outperform CB slightly in that it shows smaller variance in attaining the feature and more sensitive to weak fault signatures. This merit of MSB may be due to that the MSB feature has more diagnostic information as it is the combination of first three harmonics, whereas the CB feature is combined from just the first two harmonics.

Envelope analysis is a widely used method for rolling element bearing fault detection. To obtain high detection\ud accuracy, it is critical to determine an optimal frequency narrowband for the envelope demodulation. However, many of the\ud schemes which are used for the narrowband selection, such as the kurtogram, can potentially produce poor detection results\ud because they are sensitive to random noise and aperiodic impulses which normally occur in practical applications. To\ud achieve the purposes of denoising and frequency band optimisation, this keynote presents a modulation signal bispectrum\ud (MSB) based robust detector for bearing fault detection. Because of its inherent noise suppression capability, the MSB\ud allows effective suppression of both the stationary random noise and discrete aperiodic noise. The high magnitude features\ud that result from the use of the MSB also enhance the modulation effects of a bearing fault and can be used to provide optimal\ud frequency bands for fault detection. A number of simulated and experimental evaluations show that the proposed method\ud produces more accurate and robust detection results for common bearing faults under a range of representative scenarios.

Different gear failure modes are strongly correlated with lubricant sta-tus, for example low oil level or starved lubrication leads to significant gear dam-ages. In order to develop an early detection and accurate diagnosis of gearbox lubricant serving conditions based on online vibration measurements, this study will investigate the effect of lubricant starvation on the gearbox vibration re-sponses. A two-stage helical industrial gearbox was tested under different lubri-cant shortage conditions. The results show that the gearbox vibration signature changes significantly with lubricant starvation, which includes more consistent increase in the amplitudes of vibration responses at meshing frequency harmon-ics and their associated sideband components. These changes correspond that vi-bration signal can be considered to normalise condition indicator of gearbox lub-ricant starvations.

Tooth Defect is a common failure mode that frequently occurs in gears. To develop successful diagnostic techniques, this study examines the capability of helical gear dynamic responses with the inclusion of different time-varying friction models, i.e. friction-free, Coulomb and elasto-hydrodynamic lubrication (EHL) models. The gear system is a 10-DOF (degree-of-freedom) vibration system, which incorporates the effects of gear pair, supporting bearings, driving motor and loading system. Moreover, it couples the transverse and torsional motions resulting from time-varying friction forces, time varying mesh stiffness excitations and different tooth breakage severities. To explore the vibration response, spectral peaks at characteristic mesh frequency and its harmonics along with their sidebands are considered in the light of the impulsive sources from tooth damages and different frictional excitation models. It has found that the sidebands exhibit significant difference between different friction models and mesh components for tooth defects. It is concluded that the frictional effect should be taken into account if it is to be an accurate method for the detection and diagnostic different tooth surface defects.

Temperatures are effective indicators of the health of many ma-chines such as the wind turbine gearboxes, bearings, engines, etc. This paper pre-sents a novel wireless temperature sensor node powered by a thermal harvester for monitoring the status of gearboxes. A thermoelectric generator module (TEG) is optimized to harvest the electrical power from a heat source such as the gear-box undergoing such monitoring. The power generation from this method is ob-tained based on temperature gradients emanated by sandwiching the TEG be-tween the two aluminum plates. One plate is exposed to the heat source and has the role of a heat collector, whereas the other plate, mounted with a low profile heat-sink, acts as a heat spreader. The harvested power is then used to power a wireless temperature node for condition monitoring, resulting in a powerless and wireless monitoring system.\ud To evaluate the system, an industrial gearbox is monitored by the designed temperature node. The node is fabricated using a TEG module; an LTC3108 DC-DC converter for boosting the voltage, a super-capacitor for energy storage and a CC2650 sensor tag for measuring the temperature of the gearbox. The temper-ature data is transferred via the Bluetooth Low Energy and then monitored using portable monitoring devices, such as a mobile phones. The results obtained show the system can provide a continuous monitoring of the temperature information.

Induction motors are the most common driver in the industry and consume\ud tremendous energy every year. Monitoring the status of a motor and its\ud downstream equipment and diagnosing faults in time not only avoids great damage\ud to mechanical systems but also allows the motor to run at optimal efficiency.\ud This paper studies the use of information from motor current signals to detect and\ud diagnose faults of a reciprocating compressor (RC) and its upstream three-phase\ud motor. The motor is applied by the RC with an oscillator torque which induces\ud additional components in measured current signals. Moreover, the current signatures\ud contain changes with the torque profiles due to different types of faults.\ud Based on these analytical studies, experimental studies were carried out for different\ud common RC faults, such as valve leakage, intercooler leakage, stator asymmetries\ud and the compounds of them. The envelope analysis of current signals\ud allows accurate demodulation of the torque profiles and thereby it can be combined\ud with overall current levels for implementing model based detections and\ud diagnosis. The results show these simulated faults can be separated under all operating\ud pressures.

Because of their high load carrying capacity and low cost, journal bearings are widely used to support\ud the rotor of industrial machinery with high loads, such as steam turbines, centrifugal compressors and\ud pumps. However, sudden catastrophic failure journal bearings can result in huge economic loss and\ud high safety problems. It is necessary to develop effective condition monitoring technologies to detect\ud and diagnose the failures at early stage and avoid such catastrophic failure. Previous researchers have\ud studied the low frequency vibration characteristics as well as the high frequency vibration and\ud acoustics emission for the early detection of journal bearing failure. However, these studies give\ud relatively little information to the vibro-acoustic characteristics of high speed self aligning journal\ud bearings. This paper focuses on the condition monitoring of high speed self aligning journal bearings\ud through vibro-acoustic analysis.

Many researchers have reported that biodiesel has comparable performance\ud and better emission over diesel fuel during steady state process. However,\ud the performance and emission of biodiesel under transient conditions, for\ud example engine start, warm up, acceleration and deceleration are less\ud reported. It is there fore very important to understand the performance and\ud emission of biodiesel under the transient conditions. In addition, the\ud performance and emission variation parameters under transient condition can\ud be used to develop a fault diagnostic model for transient system.

This paper presents a study of monitoring the oil starvation of a journal bearing based on vibration analysis. A diagnostic model is established by includ-ing asperity ploughs and collisions. These excitations are more significant as the oil level is reduced due to less oil film effect. However, it has been found by modulation signal bispectrum analysis that the instable oil whirls can affect the measured responses in the middle frequency range (3.5kHz to 5.5kHz), leading to a good detection of the instability but an inconsistent diagnosis. However, the structural resonances in the high frequency range (5.5kHz to 11kHz) can better reflect the excitations and result in a more agreeable separation of different levels under wide operating conditions.

Journal bearings are widely used to support the rotor of industrial machinery with high loads, such as steam turbines, centrifugal pump and compressors. The problem of journal bearings can cause catastrophic failures, results in huge economic loss and create high safety risks. It is necessary to develop effective condition monitoring technologies to detect and diagnose the failures at early stage. Many researchers have studied the low frequency vibration characteristics as well as the high frequency vibration and acoustics emission in the detection of journal bearing failure. However, these studies have shown relatively little information regarding to the vibro-acoustic characteristics of the self aligning journal bearings. This research focuses on understanding the full feature of vibroacoustics in association with the influence of radial, torsion load, radial load, shaft speed and lubricant viscosity properties in order to develop an more reliable way to monitoring journal bearings.

• Diesel engines are a kind of power source widely used in many fields.\ud • One of the biggest problems facing machinery today is the high cost of\ud maintenance of diesel engine because the downtime of the engine is expensive.\ud • Not only does a fault reduce the performance of the engine but also causes\ud secondary damage to other parts of the engine.\ud • The use of acoustic emission offers non-intrusive tool for engine monitoring