1.
1.1 Introduction
Machine condition monitoring is an important part of condition-based maintenance (CBM), which is
becoming recognized as the most efficient strategy for carrying out maintenance in a wide variety of
industries. Machines were originally ‘run to break’, which ensured maximum operating time between
shutdowns, but meant that breakdowns were occasionally catastrophic, with serious consequences for
safety, production loss and repair cost. The first response was ‘preventive maintenance’, where
maintenance is carried out at intervals such that there is a very small likelihood of failure between
repairs. However, this result in much greater use of spare parts, as well as more maintenance works
than necessary.
Condition monitoring of machinery is the measurement of various parameters related to the
mechanical condition of the machinery (such as vibration, bearing temperature, oil pressure, oil
debris, and performance), which makes it possible to determine whether the machinery is in good or
bad mechanical condition. If the mechanical condition is bad, then condition monitoring makes it
possible to determine the cause of the problem.1,2 Condition monitoring is used in conjunction with
predictive maintenance, i.e., maintenance of machinery based on an indication that a problem is about
to occur. In many plants predictive maintenance is replacing run-to-breakdown maintenance and
preventive maintenance (in which mechanical parts are replaced periodically at fixed time intervals
regardless of the machinery’s mechanical condition). Predictive maintenance of machinery:
 Avoids unexpected catastrophic breakdowns with expensive or dangerous consequences.
 Reduces the number of overhauls on machines to a minimum, thereby reducing maintenance
costs.
 Eliminates unnecessary interventions with the consequent risk of introducing faults on
smoothly operating machines.
 Allows spare parts to be ordered in time and thus eliminates costly inventories.
 Reduces the intervention time, thereby minimizing production loss. Because the fault to be
repaired is known in advance, overhauls can be scheduled when most convenient.
This chapter describes the use of vibration measurements for monitoring the condition of machinery.
Vibration is the parameter which can be used to predict the broadest range of faults in machinery most
successfully. This description includes:
 Selection of an appropriate type of monitoring system (permanent or periodic)
 Establishment of a condition monitoring program
 Fault detection
 Spectrum interpretation and fault diagnosis
 Special analysis techniques
 Trend analysis
 The use of computers in condition monitoring programs.
Condition Monitoring Methods
Condition monitoring is based on being able to monitor the current condition and predict the future
condition of machines while in operation. Thus it means that information must be obtained externally
about internal effects while the machines are in operation. The two main techniques for obtaining
information about internal conditions are:

2.
1. Vibration Analysis. A machine in standard condition has a certain vibration signature. Fault
development changes that signature in a way that can be related to the fault. This has given rise to the
term ‘mechanical signature analysis’ [9].
2. Lubricant Analysis. The lubricant also carries information from the inside to the outside of
operating machines in the form of wear particles, chemical contaminants, and so on. Its use is mainly
confined to circulating oil lubricating systems, although some analysis can be carried out on grease
lubricants. Each of these is discussed in a little more detail in the following, along with a couple of
other methods, performance analysis and thermography, that have more specialized applications.
1.3.1 Vibration Analysis
Even in good condition, machines generate vibrations. Many such vibrations are directly linked to
periodic events in the machine’s operation, such as rotating shafts, meshing gear teeth, rotating
electric fields, and so on. The frequency with which such events repeat often gives a direct indication
of the source and thus many powerful diagnostic techniques are based on frequency analysis. Some
vibrations are due to events that are not completely phase locked to shaft rotations, such as
combustion in IC (internal combustion) engines, but where a fixed number of combustion events
occur each engine cycle, even though not completely repeatable. As will be seen, this can even be an
advantage, as it allows such phenomena to be separated from perfectly periodic ones. Other vibrations
are linked to fluid flow, as in pumps and gas turbines, and these also have particular, quite often
unique, characteristics. The term ‘vibration’ can be interpreted in different ways, however, and one of
the purposes of this chapter is to clarify the differences between them and the various transducers
used to convert the vibration into electrical signals that can be recorded and analysed. One immediate
difference is between the absolute vibration of machine housing and the relative vibration between a
shaft and the housing, in particular where the bearing separating the two is a fluid film or journal
bearing. Both types of vibration measurement are used extensively in machine condition monitoring,
so it is important to understand the different information they provide.
Another type of vibration which carries diagnostic information is torsional vibration, that is, angular
velocity fluctuations of the shafts and components such as gears and rotor discs. All three types of
vibration are discussed in this chapter, and the rest of the book is devoted to analysing the resulting
vibration signals, though overwhelmingly from accelerometers (acceleration transducers) mounted on
the machine casing. It should perhaps be mentioned that a related technique, based on measurement of
acoustic emission (AE), has received some attention and is still being studied. The name derives from
high-frequency solid-borne rather than airborne acoustic signals from developing cracks and other
permanent deformation, bursts of stress waves being emitted as the crack grows, but not necessarily
otherwise. The frequency range for metallic components is typically 100 kHz to 1MHz, this being
detected by piezoelectric transducers attached to the surface.
One of the first applications to machine diagnostics was to detection of cracks in rotor components
(shafts and blades) in steam turbines, initiated by the Electric Power Research Institute (EPRI) in the
USA [10]. Even though EPRI claimed some success in detecting such faults on the external housing
of fluid film bearings, the application does not appear to have been developed further. AE monitoring
of gear fault development was reported in [11], where it was compared with vibration monitoring. The
conclusion was that indications of crack initiation were occasionally detected a day earlier (in a 14
day test) than symptoms in the vibration signals, but the latter persisted because they were due to the
presence of actual spalls, while the AE was only present during crack growth. Because of the
extremely high sampling rate required for AE, huge amounts of data would have to be collected to

3.
capture the rare burst events, unless recording were based on event triggering. In [12], AE signals are
compared with vibration signals (and oil analysis) for gear fault diagnostics and prognostics, but the
AE sensors had to be mounted on the rotating components and signals extracted via slip rings.
Because of the difficulty of application of AE monitoring to machine condition monitoring, it is not
discussed further in this book, although new developments may change the situation.
1.3.2 Oil Analysis
This can once again be divided into a number of different categories:
1. Chip Detectors. Filters and magnetic plugs are designed to retain chips and other debris in
circulating lubricant systems and these are analysed for quantity, type, shape, size, and so on.
Alternatively, suspended particles can be detected in flow past a window.
2. Spectrographic Oil Analysis Procedures (SOAP). Here, the lubricant is sampled at regular
intervals and subjected to spectrographic chemical analysis. Detection of trace elements can tell of
wear of special materials such as alloying elements in special steels, white metal or bronze bearings,
and so on. Another case applies to oil from engine crankcases, where the presence of water leaks can
be indicated by a growth in NaCl or other chemicals coming from the cooling water. Oil analysis also
includes analysis of wear debris, contaminants and additives, and measurement of viscosity and
degradation. Simpler devices measure total iron content.
3. Ferrography. This represents the microscopic investigation and analysis of debris retained
magnetically (hence the name) but which can contain non-magnetic particles caught up with the
magnetic ones. Quantity, shape and size of the wear particles are all important factors in pointing to
the type and location of failure. Successful use of oil analysis requires that oil sampling, changing and
top-up procedures are all well defined and documented. It is much more difficult to apply lubricant
analysis to grease lubricated machines, but grease sampling kits are now available to make the process
more reliable.
1.3.3 Performance Analysis
With certain types of machines, performance analysis (e.g. stage efficiency) is an effective way of
determining whether a machine is functioning correctly. One example is given by reciprocating
compressors, where changes in suction pressure can point to filter blockage, valve leakage could
cause reductions in volumetric efficiency, and so on. Another is in gas turbine engines, where there
are many permanently mounted transducers for process parameters such as temperatures, pressures
and flow rates, and it is possible to calculate various efficiencies and compare them with the normal
condition, so-called ‘flow path analysis’. With modern IC engine control systems, for example for
diesel locos, electronic injection control means that the fuel supply to a particular cylinder can be cut
off and the resulting drop in power compared with the theoretical.
1.3.4 Thermography
Sensitive instruments are now available for remotely measuring even small temperature changes, in
particular in comparison with a standard condition. At this time, thermography is used principally in
quasi-static situations, such as with electrical switchboards, to detect local hot spots, and to detect
faulty refractory linings in containers for hot fluids such as molten metal. So-called ‘hot box
detectors’ have been used to detect faulty bearings in rail vehicles, by measuring the temperature of
bearings on trains passing the wayside monitoring point. These are not very efficient, as they must not
be separated by more than 50 km or so, because a substantial rise in temperature of a bearing only
occurs in the last stages of life, essentially when ‘rolling’ elements are sliding. Monitoring based on
vibration and/or acoustic measurements appears to give much more advance warning of impending
failure.

4.
1.4 Types and Benefits of Vibration Analysis
1.4.1 Benefits Compared with Other Methods
Vibration analysis is by far the most prevalent method for machine condition monitoring because it
has a number of advantages compared with the other methods. It reacts immediately to change and
can therefore be used for permanent as well as intermittent monitoring. With oil analysis for example,
several days often elapse between the collection of samples and their analysis, although some online
systems do exist. Also in comparison with oil analysis, vibration analysis is more likely to point to the
actual faulty component, as many bearings, for example, will contain metals with the same chemical
composition, whereas only the faulty one will exhibit increased vibration. Most importantly, many
powerful signal processing techniques can be applied to vibration signals to extract even very weak
fault indications from noise and other masking signals.
REVIEW on literature survey
GEORGE B. FOSTER “Recent Developments in Machine Vibration Monitoring”. Electronic
vibration monitoring devices are spreading to a wide variety of industrial machinery applications.
Present monitor designs are capable of predictive signalling of impending mechanical problems in
many classes of prime movers and loads. The established limits of allowable vibration are in the
process of change, both in units of measurement as well as values. Velocity measurement is gaining
acceptance over mils and g units. Manufacturers of machinery are incorporating vibration levels into
their warranty conditions. Noncontact devices for measuring shaft vibration are undergoing rapid
development. Their need is well established in large journal bearing machines and holds promise for
additional applications in axial turbine designs. Initial experience is being gained with computer trend
monitoring. Coupled with procedures of mechanical impedance analysis, this data gives promise of
achieving a greatly improved reliability factor in industrial process machinery, as well as permitting
closer design of future equipment.
1986
P.A.L. Ham. “Trends and future scope in the monitoring of large steam turbine generators” Current
practices in the monitoring of large steam turbine generators are briefly discussed, consideration being
given to the traditional range of turbine supervisory equipment, and the more extended facilities
which are sometimes now associated with rotating machinery, such as vibration monitoring, together
with the more generalised data logging systems now specified by some Utilities. Consideration is
given to the possible range of parameters and equipment areas which may now be incorporated into a
monitoring scheme, and attention is drawn to the advances in display technology and operator
interfaces which are now possible at moderate cost. In a concluding section, a range of monitoring
functions which could be of wide general application in the field of steam turbine generators is
discussed.
1993
Israel E. Alguindigue, Anna Loskiewicz-Buczak, and Robert E. Uhrig “Monitoring and Diagnosis of
Rolling Element Bearings Using Artificial Neural Networks” Vibration monitoring of components in
manufacturing plants involves the collection of vibration data from plant components and detailed
analysis to detect features that reflect the operational state of the machinery. The analysis leads to the
identification of potential failures and their causes and makes it possible to perform efficient
preventive maintenance. This paper documents our work on the design of a vibration monitoring
methodology for rolling element bearings (REB) based on neural network technology. This
technology provides an attractive complement to traditional vibration analysis because of the potential

5.
of neural networks to operate in real-time mode and to handle data that may be distorted or noisy. The
significance of this work relies of the fact that REB failures are responsible for a large fraction of the
malfunctions in manufacturing equipment. The technique enhances traditional vibration analysis and
provides a means of automating the monitoring and diagnosis of vibrating equipment.
1999
Gary Y. Yen Kuo-Chung Lin “Wavelet Packet Feature Extraction for Vibration Monitoring”.
Condition monitoring of dynamic systems based on vibration signatures has generally relied upon
Fourier based analysis as a means of translating vibration signals in time domain into the frequency
domain. However, Fourier analysis provided a poor representation of signals well localized in time. In
this case, it is difficult to detect and identify the signal pattern from their expansion coefficients
because the information is diluted across the whole basis. The wavelet packet transform is introduced
as an alternative means of extracting time frequency information from vibration signature. Moreover,
with the aid of statistical based feature selection criteria, a lot of feature components containing little
discriminant information could be discarded resulting in a feature subset with reduced number of
parameters. This significantly reduces the long training time that is often associated with neural
network classifier and increases the generalization ability of the neural network classifier.
2004
Lingam Kanth “Preventive Maintenance of Spring Switchgear Applying Vibration Diagnostic Tool-
A Malaysian Experience”. Broad ranges of spring switchgears have been extensively used in most of
Malaysian substations. Condition monitoring of hydraulic and spring switchgears have been a critical
problem since the conventional methods applied failed to alert prior occurrence of faults. Recent
routine reports of TNB Malaysia highlight an occurrence of 20% failures over a period of 5 year. This
has lead to greater loss and high expenditure in replacing these switchgears. One of the major
contributing factors in failure of hydraulic switchgears is occurrence of mechanical faults, which the
conventional method of monitoring failed to detect and locate. However the recent vibration testing
prevents and minimizes the breakdowns due to mechanical faults. This current vibration analysis is a
non-invasive diagnostic tool applied in comprehensive monitoring of HV switchgears. This method of
investigation is well ventured in many countries in Europe and USA but not yet in Malaysia. TNB
Malaysia is taking the stepping stone is venturing this technology in line with our objective of seeking
predictive maintenance, cost effective and reliable diagnostic tool. This paper will explore in depth
the possible mechanical faults that contribute to spring switchgears failures applying vibration
analysis. This paper will further interpret the results of vibration testing on 118 spring switchgears
132 kV. This preliminary study involves interpreting the vibration signal during the event of close and
open only to better understand the movement of the mechanical component during .operation time
enabling us to distinguish which component contributes to the fault of the switchgear.
2005
E. F. Simas F., L. A. L. de Almeida and Antonio C. de C. Lima. “Vibration Monitoring of
On-Load Tap Changers Using a Genetic Algorithm” On Load Tap Changers (OLTC) are widely used
for voltage regulation in electricity networks. Non-invasive vibration methods for condition
monitoring of internal electrical contacts have been recently proposed in the literature. The vibration
signals emitted during the tap changes are usually recorded and post-processed using spectral analysis
and some pattern classier technique. To reduce the complexity of the classier, a new technique based
on Genetic Algorithm is proposed in this paper. A description of the data acquisition system and the
corresponding collected experimental data are presented. The proposed technique is detailed and
preliminary experimental results are discussed.
2006
Theodor D. Popescu “New Approach for Machine Vibration Analysis and health monitoring”
The paper presents a new approach for machine vibration analysis and health monitoring combining
blind source separation (BSS) and change detection in source signals. So, the problem is translated
from the space of the measurements to the space of independent sources, where the reduced number
of components will simplify the monitoring problem and where the change detection methods will be

6.
applied for scalar signals. The approach has been tested in simulation and the assessment on a real
machine is presented in the last part of the paper.
2007
John Demcko & John Velotta “Generator End Turn Vibration Monitoring a Case Study”. APS and
WEC have long been exploiting both preventive and predictive maintenance technologies as a means
of increasing generation reliability. This paper documents the dea1s of the successful application of
one such technology, which facilitated operating both steam turbine generators (STGs) at Redhawk
Power Plant through the summer 2003 peak as well as monitoring the performance of subsequent
modifications.
2008
B. Sreejith, A.K. Verma and A. Srividya “Fault diagnosis of rolling element bearing using time-
domain features and neural networks” Rolling element bearings are critical mechanical components
in rotating machinery. Fault detection and diagnosis in the early stages of damage is necessary to
prevent their malfunctioning and failure during operation. Vibration monitoring is the most widely
used and cost-effective monitoring technique to detect, locate and distinguish faults in rolling element
bearings. This paper presents an algorithm using feed forward neural network for automated diagnosis
of localized faults in rolling element bearings. Normal negative log-likelihood value and kurtosis
value extracted from time-domain vibration signals are used as input features for the neural network.
Trained neural networks are able to classify different states of the bearing with 100% accuracy. The
proposed procedure requires only a few input features, resulting in simple preprocessing and faster
training. Effectiveness of the proposed method is illustrated using the bearing vibration data obtained
experimentally.
Pang Peilin, Ding Guangbin “Wavelet-based Diagnostic Model for Rotating Machinery Subject to
Vibration Monitoring”. This paper proposes a new diagnosis method based on the wavelet transform
with fuzzy theory in order to improve the limitation of applying traditional fault diagnosis method to
the diagnosis of multi-concurrent vibrant faults of turbo-generator sets. To increase the signal-noise-
ratio, a novel method based on the statistic rule is brought forward to determine the threshold of each
order of wavelet space and the decomposition level adaptively. The binary discrete wavelet transform
is used to acquire effective eigenvectors. The fuzzy diagnosis equation based on correlation matrix is
used to classify the fault modes. The network structure is obtained by establishing the fault diagnosis
model of turbo-generator set and using the improved least squares algorithm. Also the robustness of
fault diagnosis equation is discussed in this paper. The faults are input into the trained diagnosis
equation by means of choosing enough samples to train the fault diagnosis equation and the
information representing. The type of fault can be determined according to the output result. The
experiment results show that multi-concurrent fault for stator temperature fluctuation and rotor
vibration can be diagnosed effectively by this new method and the diagnosis result is correct.
M.Todd, S.D.J.McArthur, G.M.West, J.R.McDonald, S.J.Shaw. J.A.Hart “The design of a
decision support system for the vibration monitoring of turbine generators”. Condition Monitoring
(CM) systems monitor the health of expensive plant items such as turbine generators. They interpret
turbine parameters by signalling an alarm when pre-defined limits are breached. Often these alarms
have no further operational consequence but still require investigation by an expert. This is a time
consuming and laborious process due to the volume of data interpreted for each alarm. In order to
reduce the burden of alarm assessment, a Decision Support System (DSS) is proposed. The DSS will
feature a Routine Alarm Assessment (RAA) module which provides an initial analysis of the alarms,
highlighting those with no further operational consequence and enabling the expert to focus on those
which indicate a genuine problem with the turbine. The structured approach taken to capture and
document the expert knowledge on RAA along with the generation of a module specification and the
selected IS techniques are outlined. The implementation of an RAA prototype is discussed along with
how this will act as a foundation for a full alarm interpretation and fault diagnostic system.