Dissertations / Theses on the topic 'Vibrational Monitoring'

Doctor of Philosophy
Department of Biological and Agricultural Engineering
Wenqiao Yuan
Biodiesel production and utilization has been increasing rapidly worldwide in recent years. A main challenge in the commercialization and public acceptance of biodiesel is its quality control. This work reports the use of infrared spectroscopy to monitor biodiesel quality through the development of models to predict (1) the blending level of biodiesel in biodiesel-diesel mixtures, (2) the fatty acid profile of biodiesel fuels derived from various lipids, and (3) the concentration of most common impurities present in biodiesel including water, glycerol, methanol and triglycerides. Regressions based on near-infrared (NIR) spectroscopy were developed for relatively inexpensive and rapid on-line measurement of the concentration and specific gravity of biodiesel-diesel blends. Methyl esters of five different oils—soybean oil, canola oil, palm oil, waste cooking oil, and coconut oil—and two different brands of commercial-grade No. 2 on-highway diesel and one brand of off-road No. 2 diesel were used in the calibration and validation processes. The predicted concentration and specific gravity of the biodiesel-diesel blends were compared with the actual values. The maximum and average root-mean-square errors of prediction (RMSEP) of biodiesel concentration were 5.2% and 2.9%, respectively, from the biodiesel type-specific regression. For the general regression, the RMSEP were 3.2% and 0.002 for biodiesel concentration and specific gravity predictions, respectively. Five different models were developed to determine the concentration of methyl palmitate (C16:0), methyl stearate (C18:0), methyl oleate (C18:1), methyl linoleate (C18:2), and methyl linolenate (18:3) present in biodiesel. Using the NIR range a set of models based on four different types of biodiesel was developed. The maximum RMSEP was 0.553% when the models were validated with biodiesel samples that were used in the calibration, however, prediction accuracy of the model under external samples was poor, therefore, a new set of models was proposed. For this case, six different types of biodiesel were used. The models developed for C18:1, C18:2 and C18:3 presented good accuracy on prediction. However, for C16:0 and C18:0, additional work was necessary to reach reasonable accuracy in prediction. Three sub models for specific ranges of concentration (low, medium, and high) were developed. The RMSEP was reduced from 2.98% to 1.51% for the C16:0 and from 2.33% to 0.56% for C18:0, when the sub-models were validated under internal and external samples. Similar procedures were followed to develop regression models based on mid infrared (MIR) spectra. The RMSEP for C16:0, C18:0, C18:1, C18:2, and C18:3 were 0.83%, 0.37%, 1.45%, 1.59%, and 0.84%, respectively. Predictions using MIR spectroscopy models were better than those obtained with NIR spectroscopy models for the C16:0 and C18:0 models. The most common impurities present in biodiesel from production processes, including methanol, free glycerol, triglycerides, and water, were determined by infrared methods using NIR and MIR spectra and partial least square regression (PLSR) methods. The models were developed in two different approaches, one was when a single impurity was present and the other was when all impurities were present. In the single impurity models, the maximum RMSEP obtained in the NIR and MIR models were 647 mg kg[superscript]-1 and 206 mg kg[superscript]-1, respectively. The models for methanol, glycerol, and water performed better using the NIR data. For the triglycerides model, MIR worked better. Only NIR data were used to develop the models for samples with all impurities. Data pre-treatment (Savitzky-Golay second derivative) was necessary to achieve reasonable accuracy in the predictions in this type of models. The maximum RMSEP was 932 mg kg[superscript]-1 presented in the model for triglycerides. The best performance was obtained in the model developed to predict methanol concentration in biodiesel with RMSEP of 177 mg kg[superscript]-1 when all listed impurities were presented. The feasibility of using NIR and MIR spectroscopy to monitor biodiesel quality was demonstrated in this work. The developed method was accurate, rapid, convenient, yet inexpensive to determine some important characteristics of biodiesel, such as biodiesel blending level in biodiesel-diesel mixtures, the fatty acid profile of biodiesel, and impurities present in the fuel.

The fundamental concepts of process analytical chemistry are presented in Chapter One, which also discusses the different approaches and advantages of performing process analysis in-situ. The principles and instrumentation of flow injection analysis are presented with particular emphasis on multi-detemiinations and process applications. Flow injection analysis with Fourier transform infrared detection is proposed as a potential on-line monitoring technique highlighting its ability for simultaneous multi-analyte determinations. The principles of chemometric data analysis are summarised in Chapter Two along with the systematic approach to data analysts used in this thesis. Algorithms for experimental design, data pre-processing, exploratory data analysis (RCA) and multivariate calibration (PGR and PLS1) are described. Chapter Three describes the development and optimisation of an automated flow injection manifold coupled with a Fourier transform infrared detector for the determination of toluene, ethyl benzene and o-xylene in an n-hexane matrix by univariate calibration. In addition, Fourier transform infrared spectra are acquired of experimentally designed synthetic mixtures of the above analytes using the automated Fl manifold, and the multivariate calibration algorithms PGR, PLS1 and PLS2 are employed to quantify individual species. A critical evaluation of spectral preprocessing algorithms is reported. The following chapter describes the development of an automated internally coupled valve flow injection manifold, with Fourier transform infrared detection, for the determination of sucrose in aqueous matrices. Chapter Four describes the optimisation of the data acquisition system and reports on the critical comparison of univariate and multivariate calibration approaches, and results are presented for the determination of sucrose in commercially available fruit juices. The on-line monitoring of the synthesis of sodium glycinate in water using Fourier transform detection is described in Chapter Five. Fourier transform infrared spectral analysis of the reaction using continuous flow analysis with real time Gram Schmidt data analysis is compared to the results acquired via an automated flow injection manifold. The use of a simplex designed calibration set to mimic spectral changes in the reaction mixture to facilitate multivariate calibration is also presented. Chapter Six compares the analytical performance of a thermostabilised attenuated total reflectance flow cell with a water cooled mid-infrared fibre optic probe. Univariate and multivariate calibration algorithms are used to process reaction data sets along with gas chromatographic reference data. A detailed discussion of proposed approaches to reduce analysis time and increase information in reaction development programs is presented. The thesis concludes with an investigation of the quantitative nature of Raman spectroscopy. Experimentally designed calibration sets containing aqueous herbicide formulation samples are analysed in glass vials, with a critical evaluation of long term instrumental stability measured in terms of reproducibility and repeatability. The effects of pre-processing Raman spectral data sets, in the absence of sample matrix signals, using nonmalisation and multiple scatter correction are discussed.

PONTIFÍCIA UNIVERSIDADE CATÓLICA DO RIO DE JANEIRO
COORDENAÇÃO DE APERFEIÇOAMENTO DO PESSOAL DE ENSINO SUPERIOR
PROGRAMA DE SUPORTE À PÓS-GRADUAÇÃO DE INSTS. DE ENSINO
Uma série de ensaios físico-químicos realizados em misturas comerciais Bx (0 por cento, 7 por cento, 20 por cento, e 100 por cento de biodiesel soja/sebo) em óleo diesel S10 e S500, bem como o desempenho de duas técnicas rápidas e ainda pouco exploradas, denominadas, espectroscopia FTIR-HATR e Raman, foram utilizadas para avaliar a estabilidade oxidativa e a hidrólise destas misturas. A adição de biodiesel ao diesel afeta negativamente a resistência ao envelhecimento das misturas resultantes. Misturas S500 são mais ácidas do que misturas S10, em concordância com o teor de água mais elevado da primeira. Testes de estabilidade oxidativa acelerada por Rancimat mostraram que os tempos de indução das amostras de B7 e B20 são maiores do que os de B100, independente do teor de enxofre do diesel. O uso prático de FTIR-HATR para caracterizar o estágio de degradação das misturas é condicionado pelo fato de existirem duas contribuições químicas para cada uma das bandas estudadas. Por outro lado, a espectroscopia Raman representa uma técnica espectroscópica muito adequada para detectar presença de insaturações das cadeias de ácidos graxos do biodiesel. Uma vez que as espectroscopias FTIR-HATR e Raman não necessitam de preparação de amostras, são técnicas rápidas e de baixo custo, e causam baixo impacto ao meio ambiente, mais atenção pode ser dada a elas.
A series of physicochemical studies performed on Brazilian commercial Bx (0 per cent, 7 per cent, 20 per cent, and 100 per cent soybean/tallow biodiesel) mixtures in S10 and S500 oil diesel, as well as the performance of two rapid and still underexplored techniques, namely, FTIR-HATR and Raman spectroscopies, to evaluate the hydrolysis and oxidative stability of these blends are reported. The addition of biodiesel to diesel affects negatively the aging resistance of the resulting blends. S500 blends are more acidic then S10 blends, in accordance with the higher water content of the former. Rancimat accelerated oxidative stability tests showed that the induction times of B7 and B20 samples are greater than that of B100, independent of the sulfur content of the diesel. The practical use of FTIR-HATR to characterize the mixtures degradation stage is conditioned by the fact that there are two chemical contributions for each of the studied bands. On the other hand, Raman spectroscopy represents a very suitable spectroscopic technique to detect the presence of unsaturations in the fatty acids chains of biodiesel. Since FTIR-HATR and Raman spectroscopies do not require sample preparation, are fast and quite low cost techniques, and cause low impact to the environment, further attention may be paid to them.

Analysis of the molecular speciation of organic compounds in solution is essential for the understanding of ionic complexation. The Raman spectroscopic technique was chosen for this purpose because it allows the identification of compounds in different states and it can give information about the molecular geometry from the analysis of the vibrational spectra. In this research the ionisation steps of relevant pharmaceutical material have been studied by means of potentiometry coupled with Raman spectroscopy; the protonation and deprotonation behaviour of the molecules were studied in different pH regions. The abundance of the different species in the Raman spectra of aqueous salicylic acid, paracetamol, citric acid and salicylaldoxime have been identified, characterised and confirmed by numerical treatment of the observed spectral data using a multiwavelength curve-fitting program. The non-destructive nature of the Raman spectroscopic technique and the success of the application of the multiwavelength curve-fitting program demonstrated in this work have offered a new dimension for the rapid identification and characterisation of pharmaceuticals in solution and have indicated the direction of further research. The work also covers the formation of novel cocrystal systems with pharmaceutically relevant materials. The existence of new cocrystals of salicylic acid-nicotinic acid, DLphenylalanine , 6-hydroxynicotinic acid, and 3,4-dihydroxybenzoic acid with oxalic acid have been identified from stoichiometric mixtures using combined techniques of Raman spectroscopy (dispersive and transmission TRS), X-ray powder diffraction and thermal analysis. Raman spectroscopy has been used to demonstrate a number of important aspects regarding the nature of the molecular interactions in the cocrystal. Cocrystals of salicylic acid - benzamide, citric acid-paracetamol and citric acid -benzamide have been identified with similar analytical approaches and structurally characterised in detail with single crystal X-ray diffraction. From these studies the high selectivity and direct micro sampling of Raman spectroscopy make it possible to identify spectral contributions from each chemical constituent by a peak wavenumber comparison of single-component spectra (API and guest individually) and the two- component sample material (API/guest), thus allowing a direct assessment of cocrystal formation to be made. Correlation of information from Raman spectra have been made to the X-ray diffraction and thermal analysis results. Transmission Raman Spectroscopy has been applied to the study cocrystals for the first time. Identification of new phases of analysis of the low wavenumber Raman bands is demonstrated to be a key advantage of the TRS technique.

Monitoring the health of vibrating gears is important to ensure proper operation especially in potentially life-threatening structures, such as helicopters, nuclear power plants, and uninterruptible power supply transitions in hospitals. The most common monitoring technique is casing mounted accelerometers to measure vibration. In contrast, during the last few years acoustic monitoring techniques have also provided a few diagnostic methods for gear failure. Current diagnostic methods to indicate improper gear behavior use either existing vibration data, recorded from defective gear systems, or modern dynamic models predicting gear failure behavior. This thesis uses dynamic models to indicate, predict, and diagnose healthy and unhealthy gear systems. Influence of Tip Relief on contact forces are introduced for a decent understanding of gear dynamics followed by evaluation of common gear failure mechanisms. Two software systems were used to model gear failure: Adams®, a vibration based software that uses a rigid-elastic model for multi-body dynamics, and LSDYNA ®, a transient dynamic finite element solver, capable of solving acoustic problems with the boundary element method. Results describe tooth loads along the line of contact with respect to different Tip Reliefs and contact ratios. Gear failure is examined using a Fast Fourier Transformation to characterize patterns that can be used to diagnose unhealthy gear systems. Agreement of experimental results validates theoretical predictions of analytical and numerical solutions of gear failure especially of tooth breakage.

Condition based maintenance is becoming increasingly popular in many industrial contexts, offering substantial savings and minimising accidental damage. When applied to rotating machinery, its most common tool is vibration analysis, which relies on well-established mathematical models rooted in the theory of cyclo-non-stationary processes. However, the extraction of diagnostic information from the real world vibration signals is a delicate task requiring the application of sophisticated signal processing techniques, tailored for specific machines operating under restricted conditions. Such difficulty in the current state of the art of vibration analysis forces the industry to apply methods with reduced diagnostic capabilities but higher adaptability. However in doing so most of the potential of vibration analysis is lost and advanced techniques become of use only for academic endeavours. The aim of this document is to reduce the gap between industrial and academic applications of condition monitoring, offering ductile and automated tools which still show high detection capabilities. Three main lines of research are presented in this document. Firstly, the implementation of stochastic resonance in an electrical circuit to enhance directly the analog signal from an accelerometer, in order to lower the computational requirements in the next digital signal processing step. Secondly, the extension of already well-established digital signal processing techniques, cepstral prewhitening and spectral kurtosis, to a wider range of operating conditions, proving their effectiveness in the case of non-stationary speeds. Thirdly, the main contribution of the thesis: the introduction of two novel techniques capable of separating the vibrations of a defective component from the overall vibrations of the machine, by means of a threshold in the amplitude spectrum. After the separation, the cyclic content of the vibration signal is extracted and the thresholded signals provide an enhanced detection. The two proposed methods, phase editing and amplitude cyclic frequency decomposition, are both intuitive and of low computational complexity, but show the same capabilities as more sophisticated state of the art techniques. Furthermore, all these tools have been successfully tested on numerically simulated signals as well as on real vibration data from different machinery, lasting from laboratory test rigs to wind turbines drive-trains and aircraft engines. So in conclusion, the proposed techniques are a promising step toward the full exploitation of condition based maintenance in industrial contexts.

Increased competition in the capital-intensive paper industry makes optimized operation to an important part of corporate strategies to reduce them overall costs. Running the machines in a reliable way contributes to higher plant availability, which is particularly important in industries where the production rate is high and a break lead to costly production downtime. To achieve high plant availability, an effective maintenance of manufacturing equipment is required. This requires that the right action is taken at the right time. The purpose of this project was to increase understanding of how operation parameters for a paper machine affect the condition of the machine. The condition is assessed by studying the vibration levels using a condition monitoring system installed on the wire section rollers. The goal was to identify tools that can help operators to run the paper machine in a harmless and cost-effective way. The study was conducted by varying three parameters: wire speed, wire tension and vacuum in the suction boxes of the wire section. Wire speed and wire tension were varied for the flat wire and the upper wire. The vacuum was only varied for the flat wire. Each parameter was varied separately and the influence on the rolls vibration levels were analyzed. The range for the different operating parameters was determined in consultation with the operating personnel. There were no clear trends of how the operation parameters affect the machine condition. However, changes in vibration levels for single rolls were found. For the flat wire the changes in vibration levels were small except for a specific roll where the vibration level dropped drastically at increased speed, increased vacuum and higher tension. Even for the upper wire the changes in vibration levels were marginal except for a specific roll, which increased tension resulted in reduced vibration level, -from 2,74mm / s to 1.56 mm / s, which is a significant difference when the alarm limit of the machine is 2.5mm / s. This is an important discovery because the rollers can be seen as vital components in a series of linked systems and their operation is necessary for the paper machine to perform its required function. A conclusion is that relatively small adjustments of operation parameters affect vibration levels, which will have an effect on the component remaining life, and hence the system availability.
Ökad konkurrens inom den kapitalintensiva pappersindustrin gör att en optimerad drift blir en viktig del i företagens strategier för att minska dem totala kostnaderna. Att köra maskinerna på ett driftsäkert sätt bidrar till högre anläggningstillgänglighet, vilket är särskilt viktigt vid industrier där produktionstakten är hög och ett avbrott leder till kostsamma produktionsbortfall. För att uppnå hög anläggningstillgänglighet krävs också ett effektivt underhåll av tillverkningsutrustningen. Detta kräver att rätt åtgärder sätts in i rätt tid. Syftet med detta examensarbete var att öka förståelsen för hur olika driftparametrar för en pappersmaskin påverkar tillståndet för maskinen. Tillståndet bedömdes genom att studera vibrationsnivåer från ett lagerövervakningssystem installerat på virapartiets valsar. Målet var att hitta redskap som kan hjälpa operatörerna att köra pappersmaskinen på ett så skonsamt och kostnadseffektivt sätt som möjligt. Studien genomfördes genom att variera tre driftparametrar: virahastighet, viraspänning och undertryck i virapartiets suglådor. Virahastighet och viraspänning varierades för både planviran och överviran. Undertryck i virapartiets suglådor varierades endast för planviran. Varje parameter varierades separat och parametrarnas påverkan på valsarnas vibrationsnivåer analyserades. Intervallet för de olika driftparametrarna bestämdes i samråd med driftspersonalen. Sett för hela maskinen gick det inte att se några tydliga trender för hur driftparametrarna påverkade tillståndet för maskinen. Däremot kunde man se förändringar i vibrationsnivå för enstaka valsar. För planviran var ändringarna i vibrationsnivåerna marginella undantaget en specifik vals där vibrationsnivån sjönk drastiskt vid ökad hastighet, ökat vacuum samt ökad spänning. Även för överviran var ändringarna i vibrationsnivåer marginella utom för en specifik vals, där ökad spänning gav minskad vibrationsnivå, -från 2,74mm/s till 1,56 mm/s vilket är en betydande skillnad då larmgränsen från maskintillverkaren är 2,5mm/s. Detta är en viktig upptäckt eftersom valsarna kan ses som vitala komponenter i ett seriekopplat system och deras funktion är nödvändig för att pappersmaskinen skall kunna utföra krävd funktion. En slutsats är att det med ganska små justeringar i driftparametrar går att påverka vibrationsnivåerna, som i sin tur påverkar komponenternas livslängd och systemets tillgänglighet.

Dans les dernières décennies, la surveillance vibratoire des machines tournantes a acquis un intérêt particulier fournissant une aide efficace pour la maintenance dans l'industrie. Aujourd'hui, de nombreuses techniques efficaces sont bien établies, ancrées sur des outils puissants offerts notamment par la théorie des processus cyclostationnaires. Cependant, toutes ces techniques reposent sur l'hypothèse d’un régime de fonctionnement (c.à.d. vitesse et/ou charge) constant ou éventuellement fluctuant d’une façon stationnaire. Malheureusement, la plupart des machines surveillées dans l'industrie opèrent sous des régimes non stationnaires afin de remplir les tâches pour lesquelles elles ont été conçues. Dans ce cas, ces techniques ne parviennent pas à analyser les signaux vibratoires produits. Ce problème a occupé la communauté scientifique dans la dernière décennie et des techniques sophistiquées de traitement du signal ont été conçues pour faire face à la variabilité du régime. Mais ces tentatives restent limitées, dispersées et généralement peu soutenues par un cadre théorique. Le principal objectif de cette thèse est de combler partiellement cette lacune sur la base d'une formalisation théorique du sujet et d’un développement systématique de nouveaux outils de traitement du signal. Dans ce travail, la non-stationnarité du régime est limitée à celle de la vitesse— c.à.d. vitesse variable et charge constante— supposée connue a priori. Afin d'atteindre cet objectif, la méthodologie adoptée consiste à étendre le cadre cyclostationnaire avec ses outils dédiés. Nous avons élaboré cette stratégie en distinguant deux types de signatures. Le premier type comprend des signaux déterministes connus comme cyclostationnaires au premier ordre. La solution proposée consiste à généraliser la classe cyclostationnaire au premier ordre à la classe cyclo-non-stationnaire au premier ordre qui comprend des signaux déterministes en vitesse variable. Le second type comprend des signaux aléatoires périodiquement corrélés connus comme cyclostationnaires au deuxième ordre. Trois visions différentes mais complémentaires ont été proposées pour traiter les variations induites par la non-stationnarité de la vitesse de fonctionnement. La première adopte une approche cyclostationnaire angle\temps, la seconde une solution basée sur l'enveloppe et la troisième une approche cyclo-non-stationnaire (au second ordre). De nombreux outils ont été conçus dont les performances ont été testées avec succès sur des signaux vibratoires réels et simulés
In the last decades, vibration-based condition monitoring of rotating machine has gained special interest providing an efficient aid for maintenance in the industry. Nowadays, many efficient techniques are well-established, rooted on powerful tools offered in particular by the theory of cyclostationary processes. However, all these techniques rely on the assump-tion of constant— or possibly fluctuating but stationary— operating regime (i.e. speed and/or load). Unfortunately, most monitored machines used in the industry operate under nonstationary regimes in order to fulfill the task for which they have been designed. In this case, these techniques fail in analyzing the produced vibration signals. This issue, therefore, has occupied the scientific committee in the last decade and some sophisticated signal processing techniques have been conceived to deal with regime variability. But these works remain limited, dispersed and generally not supported by theoretical frameworks. The principal goal of this thesis is to partially fill in this gap on the basis of a theoretical formalization of the subject and a systematic development of new dedicated signal processing tools. In this work, the nonstationarity of the regime is confined to that of the speed— i.e. variable speed and constant load, assumed to be known a priori. In order to reach this goal, the adopted methodology consists in extending the cyclostationary framework together with its dedicated tools. We have elaborated this strategy by distinguishing two types of signatures. The first type includes deterministic waveforms known as first-order cyclostationary. The proposed solution consists in generalizing the first-order cyclostationary class to the more general first-order cyclo-non-stationary class which enfolds speed-varying deterministic signals. The second type includes random periodically-correlated waveforms known as second-order cyclostationary. Three different but complementary visions have been proposed to deal with the changes induced by the nonstationarity of the operating speed. The first one adopts an angle\time cyclostationary approach, the second one adopts an envelope-based solution and the third one adopts a (second-order) cyclo-non-stationary approach. Many tools have been conceived whose performances have been successfully tested on simulated and real vibration signals

Vibration structural health monitoring (VHM) uses the vibration properties to evaluate many civil structures during the design steps, building steps and service life.The whole function, expressed by stiffness and frequencies of tensegrity structures are primarily related to the level of pre-stress. The present work investigates the possibilities to use this relation in designing, constructing and evaluating the tensegrity structures.One of the aims of the thesis was to improve the current models for resonance frequency simulation of tensegrities. This has been achieved by introducing the bending behaviour of all components, and by a one-way coupling between the axial force and the stiffness.The environmental temperature effects on vibration properties of tensegrity structures have been also  investigated. Changes in dynamic characteristics due to temperature variations were compared with the changes due to decreasing pre-tension in one of the cables. In general, it is shown that the change in structural frequencies coming from temperature changes could of several magnitude as those from damage.Coinciding natural frequencies and low stiffness are known issues of tensegrity structures. The former can be an obstacle in VHM, while the later normally limits their uses in real engineering applications. It has been shown that the optimum self-stress vector of tensegrity structures can be chosen such that their lowest natural frequency is high, and separated from others.The environmental temperature effects on vibration properties of tensegrity structures were revisited to find a solution such that the natural frequencies of the tensegrity structures are not strongly affected by the changes in the environmental temperature. An asymmetric self-stress vector can be chosen so that the criterion is fulfilled as well as possible. The level of pre-stress can also be regulated to achieve the solution. The last part of this thesis, services as a summary of the work.

QC 20160429

Significant advances in wind turbine technology have increased the need for maintenance through condition monitoring. Indeed condition monitoring techniques exist and are deployed on wind turbines across Europe and America but are limited in scope. The sensors and monitoring devices used can be very expensive to deploy, further increasing costs within the wind industry. The work outlined in this thesis primarily investigates potential low-cost alternatives in the laboratory environment using vibration-based and modal testing techniques that could be used to monitor the condition of wind turbine blades. The main contributions of this thesis are: (1) the review of vibration-based condition monitoring for changing natural frequency identification; (2) the application of low-cost piezoelectric sounders with proof mass for sensing and measuring vibrations which provide information on structural health; (3) the application of low-cost miniature Micro-Electro-Mechanical Systems (MEMS) accelerometers for detecting and measuring defects in micro wind turbine blades in laboratory experiments; (4) development of an in-service calibration technique for arbitrarily positioned MEMS accelerometers on a medium-sized wind turbine blade. This allowed for easier aligning of coordinate systems and setting the accelerometer calibration values using samples taken over a period of time; (5) laboratory validation of low-cost modal analysis techniques on a medium-sized wind turbine blade; (6) mimicked ice-loading and laboratory measurement of vibration characteristics using MEMS accelerometers on a real wind turbine blade and (7) conceptualisation and systems design of a novel embedded monitoring system that can be installed at manufacture, is self-powered, has signal processing capability and can operate remotely. By applying the conclusions of this work, which demonstrates that low-cost consumer electronics specifically MEMS accelerometers can measure the vibration characteristics of wind turbine blades, the implementation and deployment of these devices can contribute towards reducing the rising costs of condition monitoring within the wind industry.

It is often impractical to measure vibrations directly at /or close to their sources when condition monitoring gearbox systems. It is common to measure the vibration distant from the source due to limited access to the component which is to be monitored. In addition, operating the gearbox under different loads and speeds also produces vibration signals within different components. Vibration measured in this way may be significantly distorted by the effect of signal transmission paths and interference from other sources. Therefore, suppression of distortions is a key issue for remote measurement based condition monitoring. In this research work, the influences of transducer locations and operating conditions on the vibration signal have been investigated on a typical gearbox transmission system for the detection of faults induced within the gearbox. Vibration signals corresponding to a healthy (baseline) and faulty conditions on two-stage helical gearbox at various load and speed levels were recorded. The baseline vibration data were examined using conventional methods in the time, frequency and the joint time-frequency domains, and are referenced for comparison with more advanced methods. Several parameters have been proposed for monitoring gear condition locally (gearbox casing) including time, frequency, and joint time-frequency domain representation. The results show that traditional signal processing techniques were insufficient for revealing fault detection information due to the low signal to noise ratio (SNR). This research also presents a mathematical model for the simulation of vibration signals in order to further understand the source of the vibration. The model represents a two stage gear system using a suitable stiffness function to represent the forces acting between each pair of gears. Rotational stiffness and damping are also used to simulate the angular motion of the gears and shafts. Results show that the frequency spectrum of acceleration outputs from the model take the expected form with peaks at the meshing frequency and associated harmonics. Furthermore, if the stiffness function between the first pair of gears is simulated with a broken tooth, and various degrees of damage, outputs from the simulation have similar sideband effects to the signals produced in the experimental investigation. In addition, the model also demonstrates that variation of load and speed produces a corresponding effect to that seen in the experiments. Consequently, although relatively simple, the mathematical model can be used to explain vibration mechanisms in real gearbox systems used in condition monitoring. Time synchronous averaging (TSA) has been applied to the vibration signals from the gearbox to remove random noise combined with the raw signal. The angular domain signal, the order spectrum and the order-frequency presentation were used to characterise gearbox vibration in these new domains in more detail. Results obtained following TSA were compared with those obtained through conventional analysis from waveform characteristics, spectrum patterns and corresponding feature parameters under different operating loads and fault conditions. In addition, continuous wavelet transform (CWT) of TSA was also compared with the conventional CWT results of raw signals to further characterise vibrations. As part of this research study, the vibration transmission path has been estimated using the frequency response function (FRF) technique. A response based estimation method has been developed to revise the base path and adapted to operating conditions for more accurate fault estimation. Both theoretical analysis and test results showed that improved diagnosis when the path information was included in vibration signal processing and feature selection. Finally, the vibration data recorded from the two accelerometers located on the gearbox casing and motor flange were analyzed using different signal processing methods to investigate the effect of path transmission (transducer location) on the detection and diagnosis of the seeded gear tooth faults. Results from the angular domain, the order spectrum and the order-frequency analysis are presented to demonstrate use of these techniques for fault detection in gearboxes and that the effect of path transmissions can be observed on the vibration signals. Results showed that CWT of the TSA signal could be used to detect and indicate the severity of the gear damage effectively even if vibration signals originated from a remote motor flange.

Regardless of the objective sought, the conclusions drawn from seismic monitoring can only be as good as the quality of the recorded data: the importance of properly capturing relevant raw vibrational information in the first place is thus absolutely crucial. The difficulty is that blast-induced vibration monitoring is site specific and that general formulas do not apply: every situation will correspond to a unique combination of objectives, ground conditions, blast design and explosive types, and will need to be monitored accordingly. To adequately acquire all the pertinent seismic information, a number of points must be successfully addressed, such as the choice of sensors, their location, number, orientation and anchoring, the transmission of the captured signals from these gauges to the recording equipment, and the choice and set-up of the data acquisition system.
It is the purpose of this thesis to address these questions in some detail, in an attempt to provide the reader with an understanding of how all the components involved in blast-induced vibration monitoring interact, and on how the choices made at each step can significantly affect overall results. (Abstract shortened by UMI.)

PONTIFÍCIA UNIVERSIDADE CATÓLICA DO RIO DE JANEIRO
Um dos principais mecanismos de falha em risers flexíveis é o rompimento de arames da armadura de tração. A experiência tem mostrado que esta classe de dano tende a surgir primariamente na parte emersa do riser, próximo à sua terminação. A ruptura dos arames ocorre de forma progressiva, podendo ser causada por diferentes processos, tais como corrosão pelo ingresso de fluido no espaço anular entre capa e armadura, desgaste excessivo associado com o contato e atrito entre arames adjacentes ou entre as diferentes camadas metálicas da armadura, ou mesmo a presença de níveis elevados de tensões produzidas pelos carregamentos mecânicos aos quais o riser é submetido durante a operação. O deterioramento progressivo pode dar origem a defeitos localizados que agem como concentradores de tensão e levam o arame à ruptura através de um processo de fadiga. O duto flexível é capaz de manter-se em operação mesmo com alguns dos arames de suas armaduras rompidos, porém uma sequencia de rupturas pode levar à ocorrência de vazamentos ou mesmo a falhas catastróficas. O monitoramento contínuo em tempo real é uma das principais alternativas para evitar que o dano progressivo nas armaduras do riser resulte em acidentes com severas consequências econômicas e ambientais. As técnicas de monitoramento da integridade de risers flexíveis podem ser classificadas como diretas, onde é possível identificar diretamente a existência de um dano/falha, ou indiretas, em que o sistema de sensoriamento registra indicações secundárias, possivelmente consequência da falha. Esta dissertação relata o desenvolvimento de um sistema de monitoramento baseado em vibrações. Trata-se de uma técnica indireta baseada em eventos, onde no momento da ruptura do arame um sinal de vibração, distinto tanto na frequência quanto na amplitude, é detectado por acelerômetros instalados na capa polimérica externa do riser. No trabalho, são apresentados resultados de quatro ensaios em escala real que demonstraram a viabilidade do sistema para um primeiro teste de campo. Estratégias de instrumentação dos dutos e as variações dos sinais detectados são apresentadas e discutidas. Os resultados mostraram que a resposta vibratória do sinal de ruptura apresenta características específicas, garantindo uma boa confiabilidade na detecção. Porém, em se tratando de uma técnica indireta, verifica-se que quando dois sistemas de monitoramento de naturezas complementares são empregados em conjunto, as probabilidades de detecção dos eventos de rupturas aumentam significativamente.
The main failure in flexible risers is the disruption of the wires from the tensile armor layer. Experience has shown that this class of damage occurs primarily near to the top riser connector. The breaking of the wires occurs gradually and may be caused by different processes as corrosion by inflow of fluid in the annular space, excessive wear associated with the contact and friction between adjacent wires or between different riser layers. The progressive deterioration can lead to localized defects that act as stress concentrators and may break the wire from the tensile armor layer through a fatigue process. The flexible pipe is able to remain in operation even with some broken wires, but a sequence of ruptures can conduce to a catastrophic failure. Real time continuous monitoring is one of the main alternatives to prevent progressive wire damage results in an accident with severe economic and environmental consequences. This thesis describes the development of a monitoring system based on vibrations. This is an indirect technique based on events, where in the moment of the wire break, one vibration signal is registered. This signal can be distinguished both in frequency and amplitude and detected by accelerometers installed on the polymeric outer layer of the riser. We present results of four tests in real scale that demonstrated the viability of the system for an initial field test. Instrumentation strategies in riser and the variations of the signals detected are presented and discussed. The results showed that the vibrational signal has specific characteristics ensuring good detection reliability.

Over the past three decades, carbon-fibre reinforced plastics (CFRP) and glass fibre-reinforced plastics (GFRP) have been increasingly used in modern engineering designs to make composite laminated structures. This increase is due to their attractive mechanical performances and their stable physical and chemical properties. However, these composites are subjected to distinctive failure modes which are different from those of metallic alloys. These failure modes include delamination, matrix cracking and fibre breakage. Therefore, structural health monitoring (SHM) of composite laminated structures during the operational phase has become increasingly important. This thesis presents the development of vibration-based SHM approaches. A non-contact fibre optic sensor is developed for modal testing and structural health monitoring of composite laminate structures. Signal processing methods are used on the acquired modal data to produce a new damage index. The main investigations and contributions of the thesis are summarised as follows,1) A delamination detection method using additional mass loading and modal frequencies is numerically and experimentally studied. The study shows that the interaction between local inertia and delaminations affects the vibration characteristics of composite laminated beams for delaminations located at different depths. 2) A two-step delamination producing technique through mechanical pull-up is proposed and experimentally validated for composite laminated plates. The proposed technique overcomes the inadequate performance of PTFE inserts approach and shows the ability to produce both near surface and far surface delaminations at inaccessible regions from the boundaries. 3) A delamination detection approach using wavelet coefficients of the multiple-mode modal frequency curve for beam-like structures is developed. The method does not require the knowledge of the intact state nor the use of artificial noise filtering procedures.4) The proposed intact-free wavelet coefficients of modal frequency surface are further applied to two-dimensional composite laminate plate-like structures. In conjunction with the wavelet-based edge detection method in imaging processing, the proposed method shows the satisfactory performance in delamination identification and localisation for laminate plates.5) A cost-effective non-contact fibre optic displacement sensor is developed based on the theoretical model. The parameters of the sensor are calibrated following standard procedures. The sensor shows satisfactory performance in structural modal testing. 6) The application of the developed fibre optic sensor in structural health monitoring for composite laminate structures is demonstrated by experiments and its performance is compared with that of commercial sensors.

Cette thèse contient les résultats de travaux de recherche menés à SafranTech et au Laboratoire d’Analyse des Signaux et des Processus Industriels (LASPI) de l’Université de Lyon. Le sujet traité porte sur la surveillance vibratoire des transmissions de puissance aéronautique et plus particulièrement des engrenages. Traditionnellement, les vibrations sont étudiées par analyse spectrale au moyen d’une représentation du spectre de Fourier. Basé sur ces observations, les vibrations des engrenages ont été représentées par un modèle empirique multiplicatif : d’une part le signal d’engrènement, haute fréquence, et de l’autre les signaux de rotations des roues, basses fréquences. En effet, les vibrations d’engrenage présentent un spectre de raies ayant des caractéristiques similaires à celles de certains signaux de communication comme une porteuse modulée en amplitude. Dans le but de faire de la détection précoce de défauts, il est intéressant de pouvoir séparer les signaux basses fréquences du reste du signal car ils sont plus souvent porteurs de l’information de défaut. Partant de ce modèle et de cette constatation, ces travaux étudient la réponse à deux questions:1. A quel point le signal vibratoire produit lors de la rotation d’un engrenage peut-il être expliqué par la représentation sous la forme d’un produit?2. Considérant un signal, est-il possible de le reconstruire en estimant ses composantes? Et la solution est-elle unique?Pour répondre à ces questions, le modèle a été représenté sous la forme d’un problème d’optimisation. D’autre part, un nouvel outil a été défini pour représenter le spectre discret d’un signal vibratoire d’engrenage sous la forme d’une matrice de coefficients de Fourier. Ces travaux ont montré une équivalence entre le produit matriciel de deux vecteurs et la multiplication de deux signaux temporels, et permis de faire le lien entre la séparation du produit de deux signaux (démodulation) et les opérateurs de rang faible.Cette nouvelle approche de séparation et d’estimation des signaux vibratoire d’engrenage a montré des performances théoriques idéales et a permis de détecter de manière précoce les défauts de denture de signaux d’engrenage réels
This thesis contains the results of the research studies performed with SafranTech and the Laboratoire d’Analyse des Signaux et des Processus Industriels (LASPI) of the University of Lyon. The main subject focuses on vibratory surveillance of aeronautic power transmission systems and more specifically gearboxes.Usually, vibrations are investigated with spectral analysis by means of the common representation of the Fourier spectrum. Based on these observations, gearbox vibrations have been represented by an empirical product model: on one hand the meshing signal, with high frequency, and on the other hand the gears rotations signals, with low frequencies.Indeed, gearbox vibrations develop a line spectrum having similar characteristics with some communication signals, as a carrier signal modulated in amplitude. For the purpose of incipient fault detection, it is interesting to be able to separate low frequency signals as they usually convey more fault information. Based on these model and observation, this research work investigate the answer to the two following questions:1. To which point the vibration signals produced by gears rotation can be explained by the representation as a product?2. Given a signal, is it possible to rebuild it by estimating the two components? Is the solution unique?In order to answer those questions, the given model was formulated as an optimization problem. Then a new tool has been defined to represent the discrete spectrum of gearbox vibration signal as a matrix containing the Fourier coefficients. This work has proven equivalence between the two representations of the matrix product of two vectors and the temporal multiplication of two signals. Furthermore, it allowed us to link the remote fields of signal demodulation and low rank approximation.This new separation and estimation approach for gearbox vibration signals has shown theoretical interesting performances, close to the ideal and allowed us to perform efficient incipient fault detection on real gearbox vibration dataset

The aim of this work was to investigate the feasibility of detecting and locating damage in large frame structures where visual inspection would be difficult or impossible. This method is based on a vibration technique for non-destructively assessing the integrity of structures by using measurements of changes in the natural frequencies. Such measurements can be made at a single point in the structure. The method requires that initially a comprehensive theoretical vibration analysis of the structure is undertaken and from it predictions are made of changes in dynamic characteristics that will occur if each member of the structure is damaged in turn. The natural frequencies of the undamaged structure are measured, and then routinely remeasured at intervals . If a change in the natural frequencies is detected a statistical method. is used to make the best match between the measured changes in frequency and the family of theoretical predictions. This predicts the most likely damage site. The theoretical analysis was based on the finite element method. Many structures were extensively studied and a computer model was used to simulate the effect of the extent and location of the damage on natural frequencies. Only one such analysis is required for each structure to be investigated. The experimental study was conducted on small structures In the laboratory. Frequency changes were found from inertance measurements on various plane and space frames. The computational requirements of the location analysis are small and a desk-top micro computer was used. Results of this work showed that the method was successful in detecting and locating damage in the test structures.

Composite materials are in use in several applications, for example, aircraft structural components, because of their light weight and high strength. However the delamination which is one of the serious defects often develops and propagates due to vibration during the service of the structure. The presence of this defect warrants the design life of the structure and the safety. Hence the presence of such defect has to be detected in time to plan the remedial action well in advance. There are a number of methods in the literature for damage detection. They are either 'baseline free/reference free method' or using the data from the healthy structure for damage detection. However very limited vibration-based methods are available in the literature for delamination detection in composite structures. Many of these methods are just simulated studies without experimental validation. Grossly 2 kinds of the approaches have been suggested in the literature, one related to low frequency methods and other high frequency methods. In low frequency approaches, the change in the modal parameters, curvatures, etc. is compared with the healthy structure as the reference, however in the high frequency approaches, excitation of structures at higher modes of the order of few kHz or more needed with distributed sensors to map the deflection for identification of delamination. Use of high frequency methods imposes the limitations on the use of the conventional electromagnetic shaker and vibration sensors, whereas the low frequency methods may not be feasible for practical purpose because it often requires data from the healthy state which may not be available for old structures. Hence the objective of this research is to develop a novel reference-free method which can just use the vibration responses at a few lower modes using a conventional shaker and vibration sensors (accelerometers/laser vibrometers). It is believed that the delaminated layers will interact nonlinearly when excited externally. Hence this mechanism has been utilised in the numerical simulations and the experiments on the healthy and delaminated composite plates. Two methods have been developed here - first method can quickly identify the presence of the delamination when excited at just few lower modes and other method identify the location once the presence of the delamination is confirmed. In the first approach an averaged normalised RMS has been suggested and experimentally validated for this purpose. Latter the vibration data have then been analysed further to identify the location of delamination and its size. Initially, the measured acceleration responses from the composite plates have been differentiated twice to amplify the nonlinear interaction clearly in case of delaminated plate and then kurtosis was calculated at each measured location to identify the delamination location. The method has further been simplified by just using the harmonics in the measured responses to identify the location. The thesis presents the process of the development of the novel methods, details of analysis, observations and results.

To increase efficiency, shafts are made lighter and more flexible, and are designed to rotate faster to increase the system's power-to-weight ratio. The demand for higher efficiency in rotordynamic systems has led to increased susceptibility to transverse fatigue cracking of the shaft. Shaft cracks are often detected and repaired during scheduled periods of off-line maintenance. Off-line maintenance can be expensive and time consuming; on-line condition monitoring allows maintenance to be performed as-needed. However, inadequate (or a lack of) monitoring can allow rapidly propagating cracks to result in catastrophic shaft failure. It is therefore imperative to develop on-line condition monitoring techniques to detect a crack and diagnose its severity. A particularly useful method for transverse shaft crack detection/diagnosis is vibration monitoring. Detection, and especially diagnosis, of transverse fatigue cracks in rotordynamic systems has proven difficult. Whereas detection assesses only the presence of a crack, diagnosis estimates important crack parameters, such as crack depth and location. Diagnosis can provide the operator with quantitative information to assess further machinery operation. Furthermore, diagnosis provides initial conditions and predictive parameters on which to base prognostic calculations. There is a two-fold challenge for on-line diagnosis of transverse fatigue crack parameters. First, crack characterization involves specifying two important parameters: the crack's depth and location. Second, the nature of rotating machinery permits response measurement at only specific locations. Cracks are typically categorized as breathing or gaping; breathing cracks open and close with shaft rotation, while gaping cracks remain open. This work concerns the diagnosis of gaping crack parameters; the goal is to provide metrics to diagnose a crack's depth and location. To this end, a comprehensive approach is presented for modeling an overhung cracked shaft. Two linear gaping crack models are developed: a notch and a gaping fatigue crack. The notch model best approximates experimentally manufactured cracks, whereas the gaping fatigue crack model is likely more suited for real fatigue cracks. Crack diagnosis routines are established using free and forced response characteristics. Equations of motion are derived for both crack models, including excitation due to gravity and imbalance. Transfer matrix techniques are established to expediently obtain the steady-state system response. A novel transfer matrix technique, the Complex Transfer Matrix, is developed to distinguish forward and backward whirl components. The rotor's angular response is primarily employed in this work for crack detection and diagnosis. The overhung shaft induces an increased sensitivity to variations in crack depth and location. In addition, an available overhung rotordynamic experimental test rig allows for comparison of the current analytic results to previously obtained experimental results. Under the influence of gravity, the steady-state response of the cracked system includes a prominent 2X harmonic component, appearing at a frequency equal to twice the shaft speed. The magnitude of the 2X harmonic is strongly influenced by the shaft speed. A resonant response occurs when the shaft speed reaches half of a system natural frequency. This work demonstrates that the profile of the 2X harmonic versus shaft speed is a capable diagnostic tool. Identification of the 2X resonance frequency restricts the crack parameters to certain pairs of location and depth. Following this limiting process, the magnitude of the 2X harmonic is used to identify the crack's depth and location. Orbital shapes at the rotor are discussed, as are orbital modes of the shaft deflection. Quantitative results and qualitative observations are provided concerning the difficulty of crack detection and diagnosis.

Thesis: Nav. E., Massachusetts Institute of Technology, Department of Mechanical Engineering, 2015.
Thesis: S.M. in Engineering and Management, Massachusetts Institute of Technology, Engineering Systems Division, System Design and Management Program, 2015.
This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections.
Cataloged from student-submitted PDF version of thesis.
Includes bibliographical references (pages 131-133).
Navy ships require reliable information regarding their power and mechanical systems in order to perform their mission effectively. While today's shipboard systems are quite sophisticated, there are areas for improvement in monitoring individual loads, managing the loads to fit the ships mission, and continuously monitoring mechanical equipment. This thesis presents a method to continuously assess the condition of a rotating machinery system using vibration analysis during the machine's spin-down. A method to determine the thermal storage capacity of a structure, so that HVAC loads can be more effectively managed, is also explained. Finally, the potential impacts of a Non-Intrusive Load Monitor (NILM) on a ship are investigated.
by Ryan David Zachar.
Nav. E.
S.M. in Engineering and Management

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

This PhD research focuses on developing a wireless vibration condition monitoring (CM) node which allows an optimal implementation of advanced signal processing algorithms. Obviously, such a node should meet additional yet practical requirements including high robustness and low investments in achieving predictive maintenance. There are a number of wireless protocols which can be utilised to establish a wireless sensor network (WSN). Protocols like WiFi HaLow, Bluetooth low energy (BLE), ZigBee and Thread are more suitable for long-term non-critical CM battery powered nodes as they provide inherent merits like low cost, self-organising network, and low power consumption. WirelessHART and ISA100.11a provide more reliable and robust performance but their solutions are usually more expensive, thus they are more suitable for strict industrial control applications. Distributed computation can utilise the limited bandwidth of wireless network and battery life of sensor nodes more wisely. Hence it is becoming increasingly popular in wireless CM with the fast development of electronics and wireless technologies in recent years. Therefore, distributed computation is the primary focus of this research in order to develop an advanced sensor node for realising wireless networks which allow high-performance CM at minimal network traffic and economic cost. On this basis, a ZigBee-based vibration monitoring node is designed for the evaluation of embedding signal processing algorithms. A state-of-the-art Cortex-M4F processor is employed as the core processor on the wireless sensor node, which has been optimised for implementing complex signal processing algorithms at low power consumption. Meanwhile, an envelope analysis is focused on as the main intelligent technique embedded on the node due to the envelope analysis being the most effective and general method to characterise impulsive and modulating signatures. Such signatures can commonly be found on faulty signals generated by key machinery components, such as bearings, gears, turbines, and valves. Through a preliminary optimisation in implementing envelope analysis based on fast Fourier transform (FFT), an envelope spectrum of 2048 points is successfully achieved on a processor with a memory usage of 32 kB. Experimental results show that the simulated bearing faults can be clearly identified from the calculated envelope spectrum. Meanwhile, the data throughput requirement is reduced by more than 95% in comparison with the raw data transmission. To optimise the performance of the vibration monitoring node, three main techniques have been developed and validated: 1) A new data processing scheme is developed by combining three subsequent processing techniques: down-sampling, data frame overlapping and cascading. On this basis, a frequency resolution of 0.61 Hz in the envelope spectrum is achieved on the same processor. 2) The optimal band-pass filter for envelope analysis is selected by a scheme, in which the complicated fast kurtogram is implemented on the host computer for selecting optimal band-pass filter and real-time envelope analysis on the wireless sensor for extracting bearing fault features. Moreover, a frequency band of 16 kHz is analysed, which allows features to be extracted in a wide frequency band, covering a wide category of industrial applications. 3) Two new analysis methods: short-time RMS and spectral correlation algorithms are proposed for bearing fault diagnosis. They can significantly reduce the CPU usage, being over two times less and consequently much lower power consumption.

The focus of this research is on the development of vibration response-based damage detection in civil engineering structures. Modal parameter-based and model identification-based approaches have been considered. In the modal parameter-based approach, the flexibility and curvature flexibility matrices of the structure are used to identify the damage. It is shown that changes in these matrices can be related to changes in stiffness values of individual structural members. Using this relationship, a method is proposed to solve for the change in stiffness values. The application of this approach is demonstrated on the benchmark problem developed by the joint International Association of Structural Control and American Society of Civil Engineers Structural Health Monitoring task group. The proposed approach is found to be effective in identifying various damage scenarios of this benchmark problem. The effect of missing modes on the damage identification scheme is also studied. The second method for damage identification aims at identifying sudden changes in stiffness for real time applications. It is shown that the high-frequency content of the response acceleration can be used to identify the instant at which a structure suffers a sudden reduction in its stiffness value. Using the Gibb's phenomenon, it is shown why a high-pass filter can be used for identifying such damages. The application of high-pass filters is then shown in identifying sudden stiffness changes in a linear multi-degree-of-freedom system and a bilinear single degree of freedom system. The impact of measurement noise on the identification approach is also studied. The noise characteristics under which damage identification can or cannot be made are clearly identified. The issue of quantification of the stiffness reduction by this approach is also examined. It is noted that even if the time at which the reduction in stiffness happens can be identified, the quantification of damage requires the knowledge of system displacement values. In principle, such displacements can be calculated by numerical integration of the acceleration response, but the numerical integrations are known to suffer from the low frequency drift error problems. To avoid the errors introduced due to numerical integration of the acceleration response, an approach utilizing the unscented Kalman filter is developed to track the sudden changes in stiffness values. This approach is referred to as the adaptive unscented Kalman filter (AUKF) approach. The successful application of the proposed AUKF approach is shown on two multi-degree of freedom systems that experience sudden loss of stiffness values while subjected to earthquake induced base excitation.
Ph. D.

Le système de surveillance (HUMS) installé dans les hélicoptères permet d'anticiper les anomalies et de donner la possibilité d'effectuer des tâches de maintenance prédictive avant l'apparition de défauts critiques. Par ailleurs, HUMS est également destiné à détecter la propagation de défauts émergents. Ceci consiste à comparer les caractéristiques vibratoires en vol de l'hélicoptère aux caractéristiques d'un état normal prédéfini. L'inconvénient majeur de cette approche est que les caractéristiques de l'état normal sont relatives au type de l'hélicoptère et changent après les tâches de révision et de maintenance, ce qui nécessite un réapprentissage de ces caractéristiques. Cette étude présente des méthodes d'évaluation de la progression temporelle des signatures vibratoires. L'étude de l'évolution de la signature vibratoire dans le temps permet de détecter des événements comme des interventions de maintenance ou des propagations de défauts sans avoir à définir un modèle de l'état de bon fonctionnement de l'appareil. Des méthodes fondées sur des modèles paramétriques et des bancs de filtres d'analyse vibratoire ont été testées et validées. Finalement, une méthode de détection de défauts a été mise en oeuvre et a donné de meilleurs résultats que les méthodes traditionnelles utilisées
A Health and usage Monitoring System (HUMS) anticipates discrepancies in the rotorcraft drive-train, giving the operator an opportunity to perform corrective maintenance before any damage becomes critical. In addition to usage spectrum analysis, a HUMS deploys vibration monitoring as a means to detect propagating faults. This method consists in comparing in-flight vibration recordings to a normal state baseline. A recurrent problem with this approach is that this baseline is aircraft specific and subject to change between major overhauls, forcing the operator to relearn the baseline on regular bases. This study presents methods for evaluating the time-progression of the drive-train vibration signature. By studying fluctuation in vibration signature over time, it is possible to detect events such as maintenance actions and fault propagations without any aircraft specific baseline. Several progression analysis methods are tested, both parametric models and filter-banks. Finally, progression analysis is used as a basis for fault detection, and is shown to produce better results than traditional methods

Abstract Despite the large amount of research work in condition based maintenance and condition monitoring methods, there is still a need for more reliable and accurate methods. The clear evidence of that need is the continued dependence on time based maintenance, especially for critical applications such as turbomachinery and airplane engines. The lack of accurate condition monitoring systems could lead to not only the unexpected failures as well as the resulting hazards and repair costs, but also a huge waste of material and time because of unnecessary replacement due to false alarms and unnecessary repair and maintenance. Temperature change is a phenomenon that accompanies every dynamic activity in the universe. However, it has not been adequately exploited for mechanical system condition monitoring. The reason is the slow response of current temperature monitoring systems compared to other condition monitoring methods such as vibration analysis. Many references inferred that the change in temperature is not sensible until approaching the end of the monitored component life and even the whole system life (Kurfess, et al., 2006; Randall, 2011; Patrick, et al., March 7-14, 2009). On the other hand, the most commonly used condition monitoring method, i.e., vibration analysis, is not free from pitfalls. Although vibration analysis has shown success in detecting some bearing faults, for other faults like lubrication problems and gradual wear it is much less effective. Also, it does not give a reliable indication of fault severity for many types of bearing faults. The advancement of thermography as a temperature monitoring tool encourages the reconsideration of temperature monitoring for mechanical system fault detection. In addition to the improved accuracy and responsiveness, it has the advantage of non-contact monitoring which eliminates the need for complex sensor mounting and wiring especially for rotating components. Therefore, in current studies the thermography-based monitoring method is often used either as a distinct method or as a complementary tool to vibration analysis in an integrated condition monitoring system. The main objectives of this study are hence to: 1. Define heat sources in the rolling element bearings and overview two of the most famous bearing temperature calculation methods. 2. Setup a bearing test rig that is equipped with both vibration and temperature monitoring systems. 3. Develop a temperature calculation analytical model for rolling element bearing that include both friction calculation and heat transfer models. The friction calculated by the model will be compared to that calculated using the pre-defined empirical methods. The heat transfer model is used for bearing temperature calculation that will be compared to the experimental measurement using different temperature monitoring devices. 4. Propose a new in-band signal enhancement technique, based on the synchronous averaging technique, Autonomous Time Synchronous Averaging (ATSA) that does not need an angular position measuring device. The proposed method, in addition to the Spectral Kurtosis based band selection, will be used to enhance the bearing envelope analysis. 5. Propose a new method for classification of the bearing faults based on the fault severity and the strength of impulsiveness in vibration signals. It will be used for planning different types of tests using both temperature and vibration methods. 6. Develop and experimentally test a new technique to stimulate the bearing temperature transient condition. The technique is supported by the results of finite element modeling and is used for bearing temperature condition monitoring when the bearing is already running at thermal equilibrium condition.

This thesis describes the development of a measurement system for monitoring dynamic tests of civil engineering structures using long exposure motion blurred images, named LEMBI monitoring. Photogrammetry has in the past been used to monitor the static properties of laboratory samples and full-scale structures using multiple image sensors. Detecting vibrations during dynamic structural tests conventionally depends on high-speed cameras, often resulting in lower image resolutions and reduced accuracy. To overcome this limitation, the novel and radically different approach presented in this thesis has been established to take measurements from blurred images in long-exposure photos. The motion of the structure is captured in an individual motion-blurred image, alleviating the dependence on imaging speed. A bespoke algorithm is devised to determine the motion amplitude and direction of each measurement point. Utilising photogrammetric techniques, a model structure s motion with respect to different excitations is captured and its vibration envelope recreated in 3D, using the methodology developed in this thesis. The approach is tested and used to identify changes in the model s vibration response, which in turn can be related to the presence of damage or any other structural modification. The approach is also demonstrated by recording the vibration envelope of larger case studies in 2D, which includes a full-scale bridge structure, confirming the relevance of the proposed measurement approach to real civil engineering case studies. This thesis then assesses the accuracy of the measurement approach in controlled motion tests. Considerations in the design of a survey using the LEMBI approach are discussed and limitations are described. The implications of the newly developed monitoring approach to structural testing are reviewed.

The urgent need for a clean and sustainable power supply for wireless sensor nodes and low-power electronics in various monitoring systems and the Internet of Things has led to an explosion of research in substitute energy technologies. Traditional batteries are still the most widely used power source for these applications currently but have been blamed for chemical pollution, high maintenance cost, bulky volume, and limited energy capacity. Ambient energy in different forms such as vibration, movement, heat, wind, and waves otherwise wasted can be converted into usable electricity using proper transduction mechanisms to power sensors and low-power devices or charge rechargeable batteries. This dissertation focuses on the design, modeling, optimization, prototype, and testing of novel piezoelectric energy harvesters for extracting energy from human walking, bio-inspired bi-stable motion, and torsional vibration as an alternative power supply for wireless monitoring systems. To provide a sustainable power supply for health care monitoring systems, a piezoelectric footwear harvester is developed and embedded inside a shoe heel for scavenging energy from human walking. The harvester comprises of multiple 33-mode piezoelectric stacks within single-stage force amplification frames sandwiched between two heel-shaped aluminum plates taking and reallocating the dynamic force at the heel. The single-stage force amplification frame is designed and optimized to transmit, redirect, and amplify the heel-strike force to the inner piezoelectric stack. An analytical model is developed and validated to predict precisely the electromechanical coupling behavior of the harvester. A symmetric finite element model is established to facilitate the mesh of the transducer unit based on a material equivalent model that simplifies the multilayered piezoelectric stack into a bulk. The symmetric FE model is experimentally validated and used for parametric analysis of the single-stage force amplification frame for a large force amplification factor and power output. The results show that an average power output of 9.3 mW/shoe and a peak power output of 84.8 mW are experimentally achieved at the walking speed of 3.0 mph (4.8 km/h). To further improve the power output, a two-stage force amplification compliant mechanism is designed and incorporated into the footwear energy harvester, which could amplify the dynamic force at the heel twice before applied to the inner piezoelectric stacks. An average power of 34.3 mW and a peak power of 110.2 mW were obtained under the dynamic force with the amplitude of 500 N and frequency of 3 Hz. A comparison study demonstrated that the proposed two-stage piezoelectric harvester has a much larger power output than the state-of-the-art results in the literature. A novel bi-stable piezoelectric energy harvester inspired by the rapid shape transition of the Venus flytrap leaves is proposed, modeled and experimentally tested for the purpose of energy harvesting from broadband frequency vibrations. The harvester consists of a piezoelectric macro fiber composite (MFC) transducer, a tip mass, and two sub-beams with bending and twisting deformations created by in-plane pre-displacement constraints using rigid tip-mass blocks. Different from traditional ways to realize bi-stability using nonlinear magnetic forces or residual stress in laminate composites, the proposed bio-inspired bi-stable piezoelectric energy harvester takes advantage of the mutual self-constraint at the free ends of the two cantilever sub-beams with a pre-displacement. This mutual pre-displacement constraint bi-directionally curves the two sub-beams in two directions inducing higher mechanical potential energy. The nonlinear dynamics of the bio-inspired bi-stable piezoelectric energy harvester is investigated under sweeping frequency and harmonic excitations. The results show that the sub-beams of the harvester experience local vibrations, including broadband frequency components during the snap-through, which is desirable for large power output. An average power output of 0.193 mW for a load resistance of 8.2 kΩ is harvested at the excitation frequency of 10 Hz and amplitude of 4.0 g. Torsional vibration widely exists in mechanical engineering but has not yet been well exploited for energy harvesting to provide a sustainable power supply for structural health monitoring systems. A torsional vibration energy harvesting system comprised of a shaft and a shear mode piezoelectric transducer is developed in this dissertation to look into the feasibility of harvesting energy from oil drilling shaft for powering downhole sensors. A theoretical model of the torsional vibration piezoelectric energy harvester is derived and experimentally verified to be capable of characterizing the electromechanical coupling system and predicting the electrical responses. The position of the piezoelectric transducer on the surface of the shaft is parameterized by two variables that are optimized to maximize the power output. Approximate expressions of the voltage and power are derived by simplifying the theoretical model, which gives predictions in good agreement with analytical solutions. Based on the derived approximate expression, physical interpretations of the implicit relationship between the power output and the position parameters of the piezoelectric transducer are given.
Doctor of Philosophy
Wireless monitoring systems with embedded wireless sensor nodes have been widely applied in human health care, structural health monitoring, home security, environment assessment, and wild animal tracking. One distinctive advantage of wireless monitoring systems is to provide unremitting, wireless monitoring of interesting parameters, and data transmission for timely decision making. However, most of these systems are powered by traditional batteries with finite energy capacity, which need periodic replacement or recharge, resulting in high maintenance costs, interruption of service, and potential environmental pollution. On the other hand, abundant energy in different forms such as solar, wind, heat, and vibrations, diffusely exists in ambient environments surrounding wireless monitoring systems which would be otherwise wasted could be converted into usable electricity by proper energy transduction mechanisms. Energy harvesting, also referred to as energy scavenging and energy conversion, is a technology that uses different energy transduction mechanisms, including electromagnetic, photovoltaic, piezoelectric, electrostatic, triboelectric, and thermoelectric, to convert ambient energy into electricity. Compared with traditional batteries, energy harvesting could provide a continuous and sustainable power supply or directly recharge storage devices like batteries and capacitors without interrupting operation. Among these energy transduction mechanisms, piezoelectric materials have been extensively explored for small-size and low-power generation due to their merits of easy shaping, high energy density, flexible design, and low maintenance cost. Piezoelectric transducers convert mechanical energy induced by dynamic strain into electrical charges through the piezoelectric effect. This dissertation presents novel piezoelectric energy harvesters, including design, modeling, prototyping, and experimental tests for energy harvesting from human walking, broadband bi-stable nonlinear vibrations, and torsional vibrations for powering wireless monitoring systems. A piezoelectric footwear energy harvester is developed and embedded inside a shoe heel for scavenging energy from heel striking during human walking to provide a power supply for wearable sensors embedded in health monitoring systems. The footwear energy harvester consists of multiple piezoelectric stacks, force amplifiers, and two heel-shaped metal plates taking dynamic forces at the heel. The force amplifiers are designed and optimized to redirect and amplify the dynamic force transferred from the heel-shaped plates and then applied to the inner piezoelectric stacks for large power output. An analytical model and a finite model were developed to simulate the electromechanical responses of the harvester. The footwear harvester was tested on a treadmill under different walking speeds to validate the numerical models and evaluate the energy generation performance. An average power output of 9.3 mW/shoe and a peak power output of 84.8 mW are experimentally achieved at the walking speed of 3.0 mph (4.8 km/h). A two-stage force amplifier is designed later to improve the power output further. The dynamic force at the heel is amplified twice by the two-stage force amplifiers before applied to the piezoelectric stacks. An average power output of 34.3 mW and a peak power output of 110.2 mW were obtained from the harvester with the two-stage force amplifiers. A bio-inspired bi-stable piezoelectric energy harvester is designed, prototyped, and tested to harvest energy from broadband vibrations induced by animal motions and fluid flowing for the potential applications of self-powered fish telemetry tags and bird tags. The harvester consists of a piezoelectric macro fiber composite (MFC) transducer, a tip mass, and two sub-beams constrained at the free ends by in-plane pre-displacement, which bends and twists the two sub-beams and consequently creates curvatures in both length and width directions. The bi-direction curvature design makes the cantilever beam have two stable states and one unstable state, which is inspired by the Venus flytrap that could rapidly change its leaves from the open state to the close state to trap agile insects. This rapid shape transition of the Venus flytrap, similar to the vibration of the harvester from one stable state to the other, is accompanied by a large energy release that could be harvested. Detailed design steps and principles are introduced, and a prototype is fabricated to demonstrate and validate the concept. The energy harvesting performance of the harvester is evaluated at different excitation levels. Finally, a piezoelectric energy harvester is developed, analytically modeled, and validated for harvesting energy from the rotation of an oil drilling shaft to seek a continuous power supply for downhole sensors in oil drilling monitoring systems. The position of the piezoelectric transducer on the surface of the shaft is parameterized by two variables that are optimized to obtain the maximum power output. Approximate expressions of voltage and power of the torsional vibration piezoelectric energy harvester are derived from the theoretical model. The implicit relationship between the power output and the two position parameters of the transducer is revealed and physically interpreted based on the approximate power expression. Those findings offer a good reference for the practical design of the torsional vibration energy harvesting system.

Ogni impianto di produzione è caratterizzato da periodi di operatività, nei quali funziona correttamente, e da periodi di fermo, dovuti alla presenza di guasti o all’esigenza di effettuare attività volte a ristabilire il suo normale comportamento. L’obiettivo principale della funzione manutenzione è minimizzare i periodi di fermo impianto, al fine di renderlo il più disponibile possibile. Attualmente, la manutenzione basata su condizione (CBM) è una delle più politiche più efficaci adottate dalle industrie. Essa è basata sul monitoraggio di diversi parametri della macchina che ne riflettono lo stato di salute. Tra i parametri più utilizzati si trovano i segnali di vibrazione. La CBM può essere implementata attraverso quattro passi principali: raccolta dati, analisi dei segnali, diagnostica e prognostica. Tale procedura prende il nome di Prognostic Health Monitoring (PHM). La necessità di analizzare la grande mole di dati raccolta attraverso il vibration monitoring richiede l’utilizzo di metodi sviluppati nell’ambito della teoria statistica e del data mining, che si pongono l’obiettivo di riconoscere andamenti regolari all’interno di grandi insiemi di dati, al fine di generare conoscenza funzionale al processo decisionale manutentivo. In particolare, i modelli di classificazione, come alberi decisionali, algoritmi K-NN, reti neurali e Support Vector Machine, costituiscono un potente strumento per la diagnostica. Tali modelli, sulla base del PHM, vengono applicati dopo la fase di analisi dei segnali, che consiste principalmente nell’estrazione di features sia nel dominio del tempo che nel dominio tempo-frequenza. Il risultato principale ottenuto consiste nell’aver verificato un incremento delle performance, in termini di accuratezza, della classificazione dello stato di salute di un componente, dovuto all’introduzione dell’analisi nel dominio tempo-frequenza e allo sviluppo dei nuovi metodi “intelligenti”.

Thesis (Ph. D.)--University of California, San Diego, 2008.
Title from first page of PDF file (viewed February 5, 2008). Available via ProQuest Digital Dissertations. Vita. Includes bibliographical references.

The use of gearboxes for power transfer is widespread throughout industry. However, machines today are operating at higher speeds than ever before and gear wear and fatigue failures are serious and legitimate concerns. Incipient fault detection in gears has thus become the subject of intensive investigation and at this stage of development there are many competing condition monitoring methods based on vibration signal analysis. Vibration signals obtained from a gearbox were complex multi-component signals, generated by tooth meshing, gear shaft rotation, gearbox resonance vibration signatures and a substantial amount of noise. This thesis summarizes the research steps taken after a review of (i) current maintenance strategies, (ii) gearbox condition monitoring techniques, (iii) gear vibration fundamentals and (iv) common gearbox failure modes. A mathematical model of the gearbox was used to predict the effects of load and tooth breakage effects on induced vibrations. A test rig was built around an 11 kW two stage helical gearbox, designed and fabricated for experimental data collection. Simulation and experimental work was carried out for a healthy pair of gears under different loads for different drive speeds and a pair suffering from degrees of tooth breakage. Conventional methods using the time-domain of the vibration signal (RMS, kurtosis, skewness and the zero figure of merit) were used for detecting and diagnose the seeded faults. The total energy method was applied to the gear meshing frequency and its sidebands, as obtained from the FFT, to detect the presence of the faults, and the results compared with those obtained by the conventional techniques. The proposed method appears much more effective at detecting and diagnosing tooth breakage than statistical features extracted from the time-domain. Joint time-frequency domain techniques were then used to determine their effectiveness for diagnosing the seeded faults in the gearbox system when the gearbox operates under output loads and at different speeds. A comparison was made between empirical mode decomposition and smoothed pseudo Wigner-Ville distribution methods based on vibration signature. From the results obtained it appears that the empirical mode decompose technique offers a more effective and faster way to detect faults. To improve signal-to-noise ratio, a novel scheme based on adaptive noise cancellation technique with a least squares algorithms was used on the gearbox experimental vibration signals. It is concluded that this method offers the most effective way of all those tested to detect faults.

Condition based monitoring is today essential for any machine manufacturer tobe able to detect and predict faults in their machine fleet. This reduces the maintenancecost and also reduces machine downtime. In this master’s thesis twoapproaches are evaluated to detect long term vibration deviations also called vibrationanomalies in Siemens gas turbines of type SGT-800. The first is a simplerule-based approach where a series of CUSUM test are applied to several signalsin order to check if the an vibration anomaly has occurred. The secondapproach uses three common machine learning anomaly detection algorithm todetects these vibration anomalies. The machine learning algorithms evaluatedare k-means clustering , Isolation Forest and One-class SVM. This master’s thesisconclude that these vibration anomalies can be detected with these ML modelsbut also with the rule-based model with different levels of success. A set of featureswas also obtained that was the most important for detection of vibrationanomalies. This thesis also presents which of these models are the best suitedanomaly detection and would be the most appropriate for Siemens to implement.

Les méthodes d'identification dites sous-espace sont largement utilisées pour la caractérisation des modes propres et la surveillance des structures mécaniques. Elles ont fait leurs preuves pour les systèmes dont la dynamique est invariante dans le temps. Elles ne sont, toutefois, pas adaptées à des systèmes à rotors comme les hélicoptères et les éoliennes qui, de part leurs parties tournantes, sont périodiques dans le temps. Le but de cette thèse est d'étendre le champ d'application de ces méthodes à cette classe particulière de systèmes. Tout d'abord, un algorithme qui permet d'identifier certaine structure modale, dite de Floquet, est proposée. Ensuite, une étude de sensibilité est réalisée dans le but de quantifier les incertitudes, liées aux bruits ou à d'autres facteurs, sur les paramètres modaux identifiés. Enfin et partant de l'algorithme d'identification, une méthode de détection d'instabilité est développée. Cette méthode est basée sur la définition d'un résidu, fonction des paramètres modaux, et la surveillance d'un changement éventuel de ce résidu qui correspond à une déviation vers un régime instable. Ces méthodes ont été appliquées à des modèles numériques et à des données expérimentales
Subspace identification methods are widely used for caracterizing modal param-eters and for vibration monitoring of mechanical structures. They were shown powerful for the so-called linear time-invariant systems. However, they are not adapted to rotating sys-tems such as helicopters and wind turbines, which are inherently time-periodic systems. The goal of this thesis is to extend the applicability of these methods to this particular class of systems. First, a new identification algorithm is suggested. This algorithm permits to iden-tify the so-called Floquet modal structure. Then, a sensitivity study is conducted in order to quantify uncertainties, related to noises and other sources, about the identified modal param-eters. Finally and based on the suggested identification algorithm, a method for instability detection is developed. The main feature of this method is to define some residual, which is function of modal parameters, then to detect an eventual change over it which means a possible deviation toward an unstable regime. The suggested methods were applied to both numerical and experimental data