Electrical
Content
JOURNAL OF INFORMATION, KNOWLEDGE AND RESEARCH IN
ELECTRICAL ENGINEERING
COMMONLY OCCURRING FAULTS IN THREEPHASE INDUCTION MOTORS – CAUSES, EFFECTS
AND DETECTION - A REVIEW
1
1
DR. SHASHI RAJ KAPOOR
Department of Electrical Engineering, University College of Engineering, Rajasthan
Technical University (RTU), Kota (RAJ.), India.
[email protected]
ABSTRACT: This paper expounds an elementary delineation of the tangible verities and occurrences
associated to induction motors and induction motor faults. Besides, it explicates the underlying tenet, causes
and consequences of frequently occurring faults in induction motors. The paper attempts to discuss different
types of induction motor faults, their causes and detection techniques. It is found that the detection techniques
which evaluate the dynamic behaviour of the signal (such as Wavelet Transform analysis) are best suited for the
purpose.
Keywords— Fault detection and identification (FDI), Mechanical and electrical faults, Condition
monitoring.
I: INTRODUCTION
Induction motors are intricate electro-mechanical
devices that are widely used in industrial processes
and commercial installations. Such motors have
extensive usage in consequence of their sturdy
edifice, unpretentious installation, undemanding
control strategies, and adaptability to various
industrial applications. Furthermore, induction
motors may plausibly be fed from an unswerving
sinusoidal power supply or by an inverter fed
variable frequency drive. Most of the faults in the
induction motors may be detected in the nascent
stages so as to prevent untimely failures [3]-[7]. This
paper addresses induction motor faults, their causeseffects and various detection techniques that are used
for incipient fault detection in three-phase induction
motors.
II: MAJOR
MOTORS
DAMAGE TYPES
OF
INDUCTION
The probable reasons for induction motor impairment
can be a multiplicity of factors such as filth and dirt
instigated temperature intensification; unwarranted
vibrations due to faulty bearings; thermal stresses due
to rotor rub, rotor skewing, and end ring heating;
mechanical stresses due to air gap eccentricity,
frequent startups, rapid acceleration and deceleration;
long persistent overload conditions; transient torques
due to faulty bearings, poor supply quality, and
unbalanced stator phase winding; flaws in
manufacture or design; imperfect installation;
deterioration due to abrasion, erosion and aging.
Most of the time there are copious factors that beget
motor breakdown. The most unequivocal basis for
motor breakdown is damage of the bearings or
winding or rotor but the paramount rationale that is
often an attribute to such failures is overheating
prompted through dirt, filth and grime.The literature
point out that majority of the failures in the threephase induction motors are mechanical in nature such
as bearing faults, misalignment or eccentricity faults
and balance related faults [1], [8]. The commonly
occurring electrically detectable Induction motor
faults are as follows [1], [4]-[7], [14]-[15], [31]:
(A) Unbalanced Supply Condition:
The unbalanced supply condition [15] is the most
commonly occurring electrical anomaly in the
Induction motors. The unbalanced supply condition
might be due to the poor power quality resulting from
unequal phase voltage and/or currents, poor or
damaged terminal contacts to motor, unbalanced load
distribution on one or more phases, faulty power
factor correcting equipment installed on one or more
phases etc. The unbalanced supply fault although
external to motor, might result in damage to motor
foundation, bent shafts, rotor eccentricity, damaged
bearings, and burnt motor terminals. These fault
conditions, under extreme situations, might result in
severe and objectionable motor vibrations, oscillating
torque, and unsymmetrical phase currents. The phase
unbalanced condition causes increase in current in
one or more phases and might result in blown fuse in
affected phase circuit. This subsequently turns up into
single phasing of the motor which need necessarily
be checked to avoid severe motor damage.
(B) Broken Rotor Bars:
Outsized squirrel cage induction motors are
assembled using copper rotor bars and end-rings,
whereas small and medium sized cage motors are by
ISSN: 0975 – 6736| NOV 12 TO OCT 13 | VOLUME – 02, ISSUE - 02
Page 178
JOURNAL OF INFORMATION, KNOWLEDGE AND RESEARCH IN
ELECTRICAL ENGINEERING
and large crafted using die-cast aluminum rotor and
end rings. Therefore, there is preponderance of diecast aluminum rotors in the induction motors on
account of its distinctive light weight and low cost.
These die-cast rotors bring in numerous
predicaments, such as rotor eccentricity and
temperature instigated softening of bars and endrings. The alternative approach to fabricate the cage
rotors is by using brazed or welded copper rotor bars.
The rotor failures get trigger off by an assortment of
diverse stresses and strains that appear in the rotor.
Stresses can be electromagnetic, thermal, residual,
dynamic, environmental and mechanical [2]. The
foremost reason for the rotor failure typically resides
in flaws related to imperfect casting or substandard
jointing during manufacturing. Whilst the rotor bar is
slighter than the bar slot, slot harmonics will surface
that subsequently initiate radial advance of the bar,
particularly during starting-up, accelerating and
braking periods. This might beget frailty
consequently ensuing in fractured and ruptured rotor
bar(s). Thermal stress is an additional quandary,
occurring when the bar cannot progress
longitudinally in the rotor slot. Sustained motor
overloads and frequent starts, acceleration and
deceleration contribute to cultivate substantial
currents that bring forth considerable mechanical and
thermal stresses, consequently prompting damage to
the rotor and the stator as well. Even so, the induction
motor rotor faults by and large onset from a
minuscule crevice or a high resistivity spot in the
rotor bar. This spot shows temperature intensification
which worsens the damage until the rotor bar is
absolutely conked out. Subsequent to a solitary rotor
bar breakage, the rotor current reassigns to the other
co-existing healthy bar(s) causing the over current
and ultimately a cluster of broken bars. Another
frequently occurring source for the rotor failures is
the over-current typically due to rotor clog up
condition of the motor. It is on the whole not
beneficial or feasible to mend the rotor.
A broken rotor bar brings about assorted outcomes of
concern in induction motors. A documented
consequence of a broken rotor bar is the advent of the
purported sideband components [4, 9, 10]. Bar failure
progression that induces sideband frequencies in
current and power spectrum of the induction motors
involves an intricate mechanism. The precursors of
rotor faults encompass the twofold slip frequency
side bands flanking the supply frequency in a
frequency spectrum of the stator current. The lower
side band component to the left of the fundamental is
prompted by electrical and magnetic disparity in the
rotor cage of an induction motor [9], whilst the lower
sideband component to the right is attributable to
ensuing speed undulations on account of the varying
air-gap torque [4, 16]. The frequencies of these
sidebands, fSB, can be formulated as:
(1)
Where k = harmonic index = 1, 2, 3, ----s = per unit slip
p = number of fundamental pole-pairs
fe = fundamental supply frequency
On account of the configuration of the windings in an
induction motor, only the frequency components with
harmonics k =1, 5, 7, 11, 13, and so forth will figure
in the current. Then, the equation (1) can be
transformed as:
(2)
For harmonic index k = 1, the equation (2) gets
further altered as:
(3)
The spectral sideband components are availed of to a
large extent for induction motor fault classification
[9, 10, 20, 29, 48]. At high voltage or low inertia, the
upper sideband takes the limelight owing to speed
undulations churned out of the defective bar. Air-gap
space harmonics brought into being due to tooth or
core saturation influence the upper spectral sideband
component. Other correlated effects of broken rotor
bars that are used for motor fault diagnosis and
detection purposes include speed undulations [16],
torque ripples [19], stator power swings [24], and
stator current envelopes [49].
(C) Damaged Bearing Faults:
Four major types of bearing faults [3] are material
deterioration in inner race, outer race, cage, and ball
defects. The bearing faults can be classified into
cyclic faults and non-cyclic faults. Cyclic faults
emerge when the rolling element and the race or the
rolling element cage of the bearing passes through
the point of flaw. The deep scratches in a race or in a
rolling element are a type of cyclic fault. The material
abrasion, qualitative degradation of the lubricant due
to contaminants, insufficient lubrication, slither and
skid amongst the movable bearing components cause
mutilation of the contact areas, which is a non-cyclic
fault category. The bearing defects induce nonstationary and fault specific frequency components in
the stator current and the engendered vibrations.
a)
Cage Defect:
The bearing cage of a ball bearing upholds the balls
at evenly balanced positions and assists the confined
rolling of the balls along the raceways. When the
motor shaft is rotating, the bearing cage revolves at a
constant angular velocity that is mean of the inner
and outer race angular velocities. The cage angular
velocity can be exploited to work out the value of
dominant fault frequency due to cage defect, fCD as
given below:
(4)
ISSN: 0975 – 6736| NOV 12 TO OCT 13 | VOLUME – 02, ISSUE - 02
Page 179
JOURNAL OF INFORMATION, KNOWLEDGE AND RESEARCH IN
ELECTRICAL ENGINEERING
Where ωi = Angular speed of the inner race in
RPM
ωo= Angular speed of the outer race in RPM
D = Pitch Diameter
d = Ball Diameter
Φ = Ball contact angle
ri = Inner race radius, ro = Outer race radius
In case of motors, the outer race is connected to the
casing that is stationary. The inner race and the shaft
are mounted together and both rotate at the same
angular speed. Accordingly, it can be assumed that:
ωo = 0 and
ωi = ω r
(5)
Where ωr = Rotor angular speed in RPM
Incorporating the above mentioned assumption as
depicted in equation (5) fetches equation (6) as given
below:
(6)
Empirically, the cage defect fundamental frequency
for a ball bearing with six to twelve balls in it is
given as:
(7)
Where
b)
Inner Race Defect:
The inner race defect frequency, fIRD, depends on the
rate at which bearing balls pass through the point of
defect on the inner race. Each ball move across the
flaw point at a rate that is proportional to the
difference of angular speed of the cage and inner
race. The fault frequency of the inner race defect is
also proportional to the number of balls in the
bearing. The fundamental fault frequency for the
inner race defect (for conditions ωo = 0 and ωi = ωr )
is given as:
(8)
Where the used variables have definition as
mentioned along with eqn. (4)
The empirical formula for inner race defect fault
frequency of a ball bearing consisting of six to twelve
balls in it is given as:
(9)
Where ωrs is as mentioned along with eqn. (7)
c)
Outer Race Defect:
The outer race defect frequency, fORD, depends on the
rate at which bearing balls cross the point of defect
on the outer race. Each ball move across the point of
defect at a rate that is proportional to the difference of
angular speed of the cage and outer race. The fault
frequency of the outer race defect is also proportional
to the number of balls in the bearing. The
fundamental fault frequency corresponding to the
inner race defect is given as:
(10)
Where the used variables have definition mentioned
along with eqn. (4)
Using equation (5), i.e. ωo = 0 and ωi = ωr (Where ωr
= Rotor angular speed in RPM), yields:
fODR = (0.4).n.ωrs
(11)
Where ωrs is as mentioned in eqn. (7)
(D) Inter-Turn Short Circuits:
In Induction motors the Inter-turn short circuits occur
as short circuits between turns of one phase, or
between turns of two phases, or between turns of all
phases, or between winding conductors and the stator
core. Inter-turn short circuits between turns of same
phase, winding short circuits, short circuits between
winding and stator core, short circuits between
different phases is usually caused by stator voltage
transients and abrasion. Burning of winding
insulation and complete winding short circuits of all
phase windings are usually caused by motor
overloads and blocked rotor, stator energising by subrated voltage, overrated voltage power supplies,
frequent starts and rotation reversals. Inter-turn short
circuits are also caused due to voltage transients
because of the successive reflections resulting from
the cable connection between induction motor and
induction motor drives.
Inter-turn short circuits in stator windings constitute a
category of faults that is most common in induction
motors. Typically, short circuits in stator windings
occur between turns of one phase, or between turns
of two phases, or between turns of all phases.
Moreover, short circuits between winding conductors
and the stator core also occur. The different types of
winding faults are summarizes below as follows [37]:
Inter-turn short circuits between turns of the same
phase, winding short circuits, short circuits between
winding and stator core, short circuits on the
connections, and short circuits between phases are
usually caused by stator voltage transients and
abrasion.
Stator faults originate in the stator core or in the
stator windings. Stator winding faults can be due to
several different reasons. The insulation damage can
be, for example, due to impact damage during
installation, movement due to repeated starting, slack
core laminations, thermal damage due to over current
and finally due to thermal aging. The stator winding
faults, in a case of a low voltage induction motor, are
often not repaired. If repaired, the machine is taken to
a special service establishment or to a manufacturer
for re-winding. Stator faults are indicated by
analysing e.g. the phase unbalance of the stator
currents or the axial leakage flux. The unbalance is
calculated with the aid of symmetrical components.
The negative sequence current or impedance is used
as a fault indicator.
(E) Air gap eccentricity:
ISSN: 0975 – 6736| NOV 12 TO OCT 13 | VOLUME – 02, ISSUE - 02
Page 180
JOURNAL OF INFORMATION, KNOWLEDGE AND RESEARCH IN
ELECTRICAL ENGINEERING
The presence of static and dynamic eccentricity can
be detected by monitoring the behaviour of the
fundamental sidebands of the supply frequency. The
frequencies of interest are given as:
(12)
Where fe = supply frequency
s = per unit slip
m = 1, 2, 3, --P = number of poles
The air gap eccentricity is also detected by
monitoring the behavior of sidebands of the slot
frequencies in stator current. The sideband
frequencies of interest are given as:
(13)
Where k = 1, 2, 3, ---------, R (R = number of rotor
slots)
nd = order of the rotating eccentricity
nw = order of the stator MMF harmonics
III: INDUCTION MOTOR FAULT MONITORING
The features used in fault detection methods are
intended to classify the motor condition as faulty and
healthy and also to identify the motor fault type. The
detection technique is intended to identify electrically
detectable faults in the induction motors and also
identify the fault severity. The classifier technique
classifies the Induction motor as either healthy or
faulty. The fault features are extracted from a suitable
motor parameter. Good number condition-monitoring
propositions have concerted explicitly with sensing
related failure methodologies. All of the currently
accessible techniques entail the user to acquire some
competence in making a distinction amongst normal
operating condition from a prospective fiasco.
Ideally, it is aspired to create a diagnostic procedure
that endow with a clear inference of machine health
in minimum time through processing of minimal
measurable inputs [8]. Various different parameters
such as temperature, current, voltage, vibration, flux
and acoustics have been in force for monitoring
electrical machines.
Good number condition-monitoring propositions
have concerted explicitly with sensing related failure
methodologies. All of the currently accessible
techniques entail the user to acquire some
competence in making a distinction amongst normal
operating condition from a prospective fiasco.
Ideally, it is aspired to create a diagnostic procedure
that endow with a clear inference of machine health
in minimum time through processing of minimal
measurable inputs [8]. Expediency, consistency, and
sensitivity are the basis of sensor signals. The
existing methods of condition monitoring of
electrical machines are [6], [7], [14]:
1. Noise Monitoring:
Acoustic noise from air gap eccentricity in induction
motors can be used for fault detection. Noise
monitoring is accomplished by measuring and
analyzing the acoustic noise spectrum. However, the
application of noise measurements in a plant is not
practical because of the noisy background from other
machines operating in the vicinity.
2. Torque Monitoring [25]:
Almost all motor faults cause harmonics with specific
frequencies in the air gap torque. However, air gap
torque cannot be measured directly. From the input
terminals, the instantaneous power includes the
charging and discharging energy in the windings.
Therefore, the instantaneous power cannot represent
the instantaneous torque. From the output terminals,
the rotor, shaft, and mechanical load of a rotating
machine constitute a torsional spring system that has
its own natural frequency. The attenuations of the
components of air gap torque transmitted through the
torsional spring system are different for different
harmonic orders of torque components. Generally,
the waveform of the air gap torque curve is different
from the torque measured at the shaft. The fault
condition can be identified by monitoring the specific
harmonics in the air gap torque.
The air gap torque in terms of measurable motor
terminal quantities is given as:
(14)
Where iA, iB , and iC are three-phase line currents of
an induction motor,
VCA and VAB are line-to-line voltages
r is half of the line-to-line resistance
p is the number of pole pairs
Frequencies of major torque harmonics associated
with the certain defects in induction motors are as
follows:
Under normal operation:
Angular frequency of torque = 0
With a single-phase stator:
Angular frequency of torque = -2ωs
With a single-phase rotor:
Angular frequency of torque = 2sωs
Where ωs is the supply frequency in rad/s, and s is the
slip. Therefore, the fault condition can be identified
by monitoring the special harmonics in the air gap
torque.
3. Flux Monitoring:
Air gap flux of induction motors contains rich
harmonics. A flux monitoring scheme can give
reliable and accurate information about electrical
machine conditions. Any change in air gap, winding,
voltage, and current can be reflected in the harmonic
spectra. The change of air gap flux is a function of
static eccentricity. Air gap flux can be measured by
search coils installed in the stator core. Because of
the enclosed structure of induction motors, this
operation requires the disconnection of induction
ISSN: 0975 – 6736| NOV 12 TO OCT 13 | VOLUME – 02, ISSUE - 02
Page 181
JOURNAL OF INFORMATION, KNOWLEDGE AND RESEARCH IN
ELECTRICAL ENGINEERING
motors from the main supply before dismantling. As
such, this is neither practical nor economical for the
motors that are in service.
4. Vibration Analysis:
Vibration monitoring [5], [21]-[24] is one of the
oldest condition monitoring techniques and is widely
used to detect mechanical faults such as bearing
failures or mechanical imbalance. A piezo-electric
transducer providing a voltage signal proportional to
acceleration is often used. This acceleration signal
can be integrated to give the velocity or position.
5. Current Monitoring:
One of the, most economically attractive technology
in motor condition monitoring is stator current
monitoring [11]-[20]. The stator current of an
induction motor is readily available since it is used to
protect machines from destructive over-currents,
ground current, etc. Therefore, current monitoring is
a sensor less detection method that can be
implemented without any extra hardware.
The vibration monitoring [5], [21-24] is one of the
widely used techniques for detection of electrical and
mechanical faults in electrical machines. The stator
current monitoring [11]-[20] is relatively a recent
technique that is fast gaining importance due to its
non-invasive nature, cost-effectiveness, preciseness
in analysis, easy and efficient signal processing, and
convenient installation [34]. The stator current
monitoring and analysis can provide as much
information as the vibration monitoring provides
[14], [22].
The motor stator current analysis is based on the fact
that the stator and rotor faults are ultimately reflected
in the stator current and in most of the cases cause
unbalanced phase currents or introduces a distinct
signature in the stator current. The motor stator
current analysis being non-intrusive, comparatively
simple to implement, load and operating point
independent - prove out to possess an advantageous
edge over the other analysis techniques.
Various stator current analysis methods that are
commonly employed for fault detection in Induction
motors are as shown below:
Motor stator current analysis:
Time domain techniques
Statistical parameter analysis –
Root Mean Square (RMS), mean, variance, crest
factor, skewness, kurtosis etc
Time synchronous averaging based
methods – Time synchronous averaged (TSA) signal,
Residual signal analysis (RSA) etc.
Spatial angular vector (SAV)
analysis
Negative sequence component
analysis
Filter
based
methods–
Demodulation, enveloping etc.
Stochastic methods – Thresholding,
Autoregression based methods
Frequency based techniques
(STFT)
Spectrum analysis
Fast fourier transform (FFT)
Discrete fourier transform (DFT)
Short term fourier transform
Power cepstrum analysis
Time-frequency based techniques
Spectrogram analysis
Wigner distribution (WD) analysis
Continuous wavelet transform
analysis
Discrete wavelet pocket analysis
Time-averaged wavelet spectrum
(TAWS)
Time-frequency-scale
(TMS)
domain
The commonly employed artificial intelligence based
interpreting techniques are as follows [15,17,18]:
Neural network based inference
Feed forward networks
Multi layer perceptron
Radial basis function
(RBF)
Principal
component
analysis (PCA)
Recurrent networks
Kohonen Self-organising maps
(SOM)
Fuzzy set based inference
Expert system based inference
Neuro-fuzzy computing based inference
Adaptive Neuro-fuzzy inference
Fuzzy expert system based inference
Bayesian classifier
Support vector machines (SVM)
IV: Causes of Faults in Induction motors
The reason for induction motor impairment [2] can
be a multiplicity of factors such as filth and dirt
instigated temperature intensification; unwarranted
vibrations due to faulty bearings; thermal stresses due
to rotor rub, rotor skewing, and end ring heating;
mechanical stresses due to air gap eccentricity,
frequent startups, rapid acceleration and deceleration;
long persistent overload conditions; transient torques
due to faulty bearings, poor supply quality, and
unbalanced stator phase winding; flaws in
manufacture or design; imperfect installation;
deterioration due to abrasion, erosion and aging.
Most of the time there are copious factors that beget
motor breakdown. The most unequivocal basis for
motor breakdown is damage of the bearings or
winding or rotor but the paramount rationale that is
often an attribute to such failures is overheating
prompted through dirt, filth and grime.
ISSN: 0975 – 6736| NOV 12 TO OCT 13 | VOLUME – 02, ISSUE - 02
Page 182
JOURNAL OF INFORMATION, KNOWLEDGE AND RESEARCH IN
ELECTRICAL ENGINEERING
The rotor failures get trigger off by an assortment of
diverse stresses and strains that appear in the rotor.
Stresses can be electromagnetic, thermal, residual,
dynamic, environmental and mechanical [2], [22].
The foremost reason for the rotor failure typically
resides in flaws related to imperfect casting or
substandard jointing during manufacturing. Whilst
the rotor bar is slighter than the bar slot, slot
harmonics will surface that subsequently initiate
radial advance of the bar, particularly during startingup, accelerating and braking periods. This might
beget frailty consequently ensuing in fractured and
ruptured rotor bar(s). Thermal stress is another
quandary, occurring when the bar cannot progress
longitudinally in the rotor slot. Sustained motor
overloads and frequent starts, acceleration and
deceleration contribute to cultivate substantial
currents that bring forth considerable mechanical and
thermal stresses, consequently prompting rotor and
the stator damage. Even so, the induction motor rotor
faults by and large onset from a minuscule crevice or
a high resistivity spot in the rotor bar. This spot
shows temperature intensification which worsens the
damage until the rotor bar is absolutely conked out.
Subsequent to a solitary rotor bar breakage, the rotor
current reassigns to the other co-existing healthy
bar(s) causing the over current and ultimately a
cluster of broken bars. Another frequently occurring
source for the rotor failures is the over-current
typically due to rotor clog up condition of the motor.
It is on the whole not beneficial or feasible to mend
the rotor.
Bearing faults are one of the prominent causes that
instigate the failures in induction motors [1, 8]. The
indications of bearing mutilation are bumpy running
with jerks, abridged exactness of producing shaft
movement with very less rotational tolerance and an
atypical noise. The bearing faults in the electrical
motors can on the whole be attributed to the material
attrition and abrasion at contact points, ageing and
exhaustion, misalignment, sustained load overstresses and pressure induced welding. The erosion,
attrition and abrasion of the material crop up as a
result of presence of contaminants and impurities,
friction induced overheating and subsequent
occurrence of hot spots on the bearing balls or rolls,
inner and outer races. The duration of bearing life
span is also reliant on the quality of the lubricant,
stresses due to mechanical load and electrical starts.
The frequent start ups and rapid accelerating and
decelerating periods cause repeated over stressing of
the bearings. Besides unfeigned overloading, the
lumber on the bearing in induction motors can be due
to improper alignment or rotor unbalance, and sooner
or later bring forth a state of the bearing that is not
appropriate for unblemished operation of the
machine.
Under ideal state of affairs, the rotor potential is
believed to be zero which is not practically the case.
A potential relative to the ground emerge at the rotor
due to inequality in phase capacitances. The rotor
voltage set off a difference of potential crosswise the
bearings. This leads to a current flowing through the
bearing and brings about an alteration in the chemical
composition of the lubricant, consequently resulting
in degradation of the quality of the lubricant. This
further brings in abrasion in the bearing and may
sometimes set in electrical discharges between inner
and outer races, eventually leading to the inopportune
bearing failures. Stator faults originate in the stator
core or in the stator windings. Stator winding faults
can be due to several different reasons. The insulation
damage can be due to impact damage during
installation, movement due to repeated starting, slack
core laminations, thermal damage due to over current
and due to thermal aging. Stator faults are indicated
by perceiving parameters such as phase unbalance of
the stator currents or the axial leakage flux or the
vibration content of the motor etc.
V: Schematic block diagram for fault detection in
Induction motors
The motor under test (MUT) is made to run as linefed induction motor or an inverter fed adjustable
speed induction drive (ASID), as the case may be, by
feeding an apposite 3-phase supply. The data
acquisition and analysis software (the popular ones
are dSPACE-Matlab or National Instrument’s DAQ –
Matlab etc.) along with the pertinent control card is
used to develop the data acquisition setup. The input
signals are processed and analysed through suitable
signal processing and analysis tools so as to extract
fault features and present them in a format acceptable
to the successive classifier stage. The interpreter
stage takes a decision on the basis of fault features
presented to it as regards the motor health status. The
general schematic of an experimental setup for fault
detection in three-phase induction motors is as shown
in fig.1 below:
Data
Acquisition
Hardware and
software
Data
Acquisition and
Control
Hall Effect
current and
voltage
Transducers
Three-phase
Inverter/On-Off
Arrangement
Speed
Measur
ement
3-Φ Induction
motor
ISSN: 0975 – 6736| NOV 12 TO OCT 13 | VOLUME – 02, ISSUE - 02
Page 183
JOURNAL OF INFORMATION, KNOWLEDGE AND RESEARCH IN
ELECTRICAL ENGINEERING
Australasian
Universities
Power
-------,
2003/itee.uq.edu.au.
[8]
O’Donnell P., ”Report of Large Motor
Loading
Reliability
Survey of Industrial and Commercial
arrangement
Installations”, Part I and II, Motor Reliability
Working Group, IEEE Industry Application Society
Fig. 1: General schematic of the experimental setup (IAS), IEEE Transaction on Industry Applications, Pp
853-872, July/August 1985; Part III, Motor
Reliability Working Group, IEEE Industry
Application Society (IAS), IEEE Transaction on
VI: CONCLUSION
Industry Applications, Pp 153-158, January/February
The brief review of the commonly occurring
1985.
induction motor faults has been presented. The
[9]
Arthur N. and J. Penman, “ Induction
relevant references have been mentioned for ready
Machine Condition Monitoring with higher order
reference. The conventional and recent signal
spectra”, IEEE transaction on Industrial Applications,
analysis, fault extraction and interpreter techniques
Vol.47, No.5, Pp. 1031-1041, October 2000.
have been discussed. The outcome and conclusions
[10]
Arthur N., J. Penman, A. McLean, and A.
buoy to support the author’s assumption that it is
Parsons, “Induction Machine Condition Monitoring
possible to employ certain new techniques so as to
with higher order spectra, Part II: Variable frequency
improve the consistency of existing, well reported,
operation and automated diagnosis”, IECON’98,
and established strategies for fault detection.
Proceedings of the 24th Annual Conference of the
Furthermore the techniques that are apt in processing
IEEE Industrial Electronics Society, Pp. 1895-1900,
non-stationary signal analysis have a definite edge
Vol. 3, 31 Aug.-4 Sept. 1998 .
over the conventionally employed techniques based
[11]
Barendse P. S. and P.Pillay, “The Detection
upon current monitoring alone. It is worth to be
of Unbalanced Faults in Inverter-Fed Induction
mentioned that recent works in motor fault detection
Machines”, IEEE---------, 2007.
are anchored in analysis of stator current based on
[12]
Bellini A., G. Fraceschini, and C. Tassoni,
time-frequency (or time-scale) methods.
“Monitoring of Induction Machines by Maximum
Covarience Method for Frequency Tracking”,
REFRENCES
Conference Records of the 2004 IEEE Industry
[1]
Albrecht P.F., J.C. Appiarius, and D.K.
Applications Conference, 39th Annual Meeting, Vol.
Sharma,”Assessment of the reliability of motors in
2, Pp. 743-749, 3-7 Oct. 2004.
utility applications-Updated,” IEEE Transactions on
[13]
Bellini A., F. Filippetti, G. Franceschini, T.
Energy conversion, Vol. 20, Pp. 719-729, 2005.
J. Sobczyk and C. Tassoni, “Diagnosis of Induction
[2]
Bonnett A. H. and George C. Soukup,
Machines by d-q and i.s.c Rotor Models”, IEEE
“Cause and Analysis of Stator and Rotor Failures in
International Symposium on Diagnostics for Electric
Three-Phase Squirrel-Cage Induction Motors”, IEEE
Machines, Power Electronics and Drives, Vienna,
Transactions on Industry Applications, Vol. 28, No. 4,
Austria, 7-9 September 2005.
July/August 1992.
[14]
Benbouzid M.E.H.,” A Review of Induction
[3]
Devaney M. J. and L. Eren, “Detecting
Motors Signature Analysis as a medium of faults
Motor Bearing Faults”, IEEE Instrumentation and
detection,” IEEE Transactions on Industrial
Measurenent Magazine, Pp. 30-50, December 2004.
Electronics, Vol. 47, No. 5, Oct. 2000.
[4]
Faiz J. and B. M. Ebrahimi, “Mixed Fault
[15]
Benbouzid M.E.H., Mighelle Vieira, and
Diagnosis in Three-phase Squirrel-Cage Induction
Céline Theys, “Induction motors’ Faults Detection
Motor Using Analysis of Airgap Magnetic Field”,
and Localization Using Stator Current Advanced
Progress in Electro-magnetics Research, PIER 64,
Signal processing Techniques”, IEEE Transactions on
Pp. 239-355, 2006.
Power Electronics, Vol. 14, No.1, Pp. 14-22, January
[5]
Han Y. and Y. H. Song, “Condition
1999.
Monitoring Techniques for Electrical Equipment-A
[16]
Bonnett A. H. and George C. Soukup,
Literature Survey”, IEEE Transaction on Power
“Cause and Analysis of Stator and Rotor Failures in
Delivery, Vol. 18, No. 1, January 2003.
Three-Phase Squirrel-Cage Induction Motors”, IEEE
[6]
Nandi S., H.A. Toliyat, and X. Li,”
Transactions on Industry Applications, Vol. 28, No. 4,
Condition Monitoring and Fault Diagnosis of
July/August 1992.
Electrical Motors- A Review,” Proceedings of 34th
[17]
Ilonen J., J. K. Kamarainen, T. Lindh, J.
IAS annual meeting on Industry Applications, IEEE,
Ahola, and H. Kalaviainen, ”Diagnosis tool for motor
Vol. 3, Pp. 197-204, 3-7 Oct 1999.
condition monitoring”, IEEE Transaction on
[7]
Sin M. L., W. L. Soong, and N. Ertugrul,
Industrial Applications, Vol. 41, No.4, Pp. 963-967,
“Induction Machine On-line Condition Monitoring
2005.
and Fault Diagnosis – A survey”, Proceedings of the
[18]
Kim K. and A. G. Parlos. “Reducing the
Impact of False Alarms in Induction Motor Fault
ISSN: 0975 – 6736| NOV 12 TO OCT 13 | VOLUME – 02, ISSUE - 02
Page 184
JOURNAL OF INFORMATION, KNOWLEDGE AND RESEARCH IN
ELECTRICAL ENGINEERING
Daignosis”, Journal of Dynamic Systems,
Measurement, and Control, Transactions of the
ASME, Vol. 125, Pp. 80- 95, March 2003.
[19]
Nejjari H. and M. E. H. Benbouzid,
“Application of Fuzzy Logic to Induction Motors
Condition Monitoring”, IEEE Power Engineering
Review, Pp. 52-54, June 1999.
[20]
Schoen R. R. and T. G. Habetler,
”Evaluation and Implementation of a System to
Eliminate Arbitrary Load Effects in Current-Based
Monitoring of Induction machines”, IEEE
Transaction on Industrial Applications, Vol. 33, No.
6, Pp. 1571-77, Nov/Dec. 1997.
[21]
Dorrell D. G., W. T. Thomson, and S.
Roach. “Analysis of air gap flux, current and
vibration signals as a function of the combination of
static and dynamic air gap eccentricity in 3-phase
induction motors”, IEEE transaction on Industrial
Applications, Vol.33, No.1, Pp. 24-34, Jan./Feb 1997.
[22]
Riley C. M., B. K. Lin, T. G. Habetler, and
R. R. Schoen, “Stator Current Harmonics And Their
Causal Vibrations: A Preliminary Investigation of
Sensorless Vibration Monitoring Applications”, IEEE
Transaction on Industry applications, Vol. 35, no. 1,
Pp. 94-99, Jan/Feb 1999.
[23]
Riley C. M., B. K. Lin, T. G. Habetler, and
R. R. Schoen, “A Method for Sensorless On-Line
Vibration Monitoring of Induction Machines”, IEEE
Transaction on Industry applications, Vol. 34, no. 6,
Pp. 1240-45, Nov/Dec 1998.
[24]
Riley C. M., B. K. Lin, T. G. Habetler, and
R. R. Schoen, “A Method for Sensorless On-Line
Vibration Monitoring of Induction Machines”, IEEE
--------, Pp. 201-207, 1997.
[25]
Hsu John S., “Monitoring of Defects in
Induction Motors Through Air-Gap Torque
Observation”, IEEE Transactions on Industry
Applications, Vol. 31, no.5, Pp. 1016-1021,
September/October 1995.
[26]
Khan M. A. S. K., Tawfik S. Radwan, and
M.Azizuur Rahman, “Real –Time Implementation of
Wavelet Packet Transform-Based Diagnosis and
Protection of Three-Phase Induction Motors”,IEEE
Transactions on Energy Conversion, Vol. 22, No.3,
Pp. September 2007.
[27]
Kim K. and A. G. Parlos. “Induction Motor
Fault Diagnosis based on Neuropredictors and
Wavelet Signal Processing”, IEEE transaction on
Mechatronics, Vol. 7, No. 2, Pp. 201- 219, June 2002.
[28]
Magnago F. and A. Abur,” Fault location
using wavelets”, IEEE Transactions on Power
Delivery, 13(4), 1475-1480, 1998.
[29]
Zanardelli W. G., E. G. Strangas, H. K.
Khalil, J. M. Miller, “Wavelet Based Methods for the
Prognosis of Mechanical and Electrical Failures in
Electric Motors”, Mechanical Systems and Signal
Processing (Elsavier), Pp 411-426, 2005.
[30]
Zanardelli W. G., and E. G. Strangas,
“Methods to Identify Intermittent Electrical and
Mechanical Faults in Permanent Magnet AC Drives
Based on Wavelet Analysis”, IEEE Conference on
Vehicle Power and Propulsion, 7-9 Sept. 2005.
[31]
Awadallah M. A. and M. M. Marcos,
“Application of AI Tools in Fault Diagnosis of
Electrical Machines and Drives – An Overview”,
IEEE Transactions on Energy Conversion, Vol. 18,
No. 2, June 2003.
[32]
Chow Mo-yuen, Peter M. Mangum, and Sui
Oi Yee, “A Neural Network Approach to Real-Time
Condition Monitoring of Induction Motors”, IEEE
Transactions on Industrial Electronics, Vol. 38, No. 6,
December 1991.
[33]
Eren L., A. Karahoca and M. J. Devaney,
“Neural Network Based Bearing Fault Detection”,
Instrumentation and Measurement Technology
Conference, Como, Italy, 18-20 May 2004, Pp. 16571660.
[34]
Han Tian, Bo-suk Yang, Won-Ho Choi, and
Jae-Sik Kim, “Fault Diagnosis System of Induction
Motors Based on Neural Network and Genetic
Algorithm Using Stator Current
Signals”,
International Journal of Rotating Machinery, Pp. 113, Volume 2006.
[35]
Lehtoranta J. and H. N. Koivo, “Fault
Diagnosis of Induction Motors with Dynamical
Neural Networks”, IEEE International Conference on
Systems, Man and Cybernetics, Pp. 2979-2984, Vol.
3, Oct. 2005.
[36]
Nejjari H. and M. E. H. Benbouzid,
“Application of Fuzzy Logic to Induction Motors
Condition Monitoring”, IEEE Power Engineering
Review, Pp. 52-54, June 1999.
ISSN: 0975 – 6736| NOV 12 TO OCT 13 | VOLUME – 02, ISSUE - 02
Page 185
ELECTRICAL ENGINEERING
COMMONLY OCCURRING FAULTS IN THREEPHASE INDUCTION MOTORS – CAUSES, EFFECTS
AND DETECTION - A REVIEW
1
1
DR. SHASHI RAJ KAPOOR
Department of Electrical Engineering, University College of Engineering, Rajasthan
Technical University (RTU), Kota (RAJ.), India.
[email protected]
ABSTRACT: This paper expounds an elementary delineation of the tangible verities and occurrences
associated to induction motors and induction motor faults. Besides, it explicates the underlying tenet, causes
and consequences of frequently occurring faults in induction motors. The paper attempts to discuss different
types of induction motor faults, their causes and detection techniques. It is found that the detection techniques
which evaluate the dynamic behaviour of the signal (such as Wavelet Transform analysis) are best suited for the
purpose.
Keywords— Fault detection and identification (FDI), Mechanical and electrical faults, Condition
monitoring.
I: INTRODUCTION
Induction motors are intricate electro-mechanical
devices that are widely used in industrial processes
and commercial installations. Such motors have
extensive usage in consequence of their sturdy
edifice, unpretentious installation, undemanding
control strategies, and adaptability to various
industrial applications. Furthermore, induction
motors may plausibly be fed from an unswerving
sinusoidal power supply or by an inverter fed
variable frequency drive. Most of the faults in the
induction motors may be detected in the nascent
stages so as to prevent untimely failures [3]-[7]. This
paper addresses induction motor faults, their causeseffects and various detection techniques that are used
for incipient fault detection in three-phase induction
motors.
II: MAJOR
MOTORS
DAMAGE TYPES
OF
INDUCTION
The probable reasons for induction motor impairment
can be a multiplicity of factors such as filth and dirt
instigated temperature intensification; unwarranted
vibrations due to faulty bearings; thermal stresses due
to rotor rub, rotor skewing, and end ring heating;
mechanical stresses due to air gap eccentricity,
frequent startups, rapid acceleration and deceleration;
long persistent overload conditions; transient torques
due to faulty bearings, poor supply quality, and
unbalanced stator phase winding; flaws in
manufacture or design; imperfect installation;
deterioration due to abrasion, erosion and aging.
Most of the time there are copious factors that beget
motor breakdown. The most unequivocal basis for
motor breakdown is damage of the bearings or
winding or rotor but the paramount rationale that is
often an attribute to such failures is overheating
prompted through dirt, filth and grime.The literature
point out that majority of the failures in the threephase induction motors are mechanical in nature such
as bearing faults, misalignment or eccentricity faults
and balance related faults [1], [8]. The commonly
occurring electrically detectable Induction motor
faults are as follows [1], [4]-[7], [14]-[15], [31]:
(A) Unbalanced Supply Condition:
The unbalanced supply condition [15] is the most
commonly occurring electrical anomaly in the
Induction motors. The unbalanced supply condition
might be due to the poor power quality resulting from
unequal phase voltage and/or currents, poor or
damaged terminal contacts to motor, unbalanced load
distribution on one or more phases, faulty power
factor correcting equipment installed on one or more
phases etc. The unbalanced supply fault although
external to motor, might result in damage to motor
foundation, bent shafts, rotor eccentricity, damaged
bearings, and burnt motor terminals. These fault
conditions, under extreme situations, might result in
severe and objectionable motor vibrations, oscillating
torque, and unsymmetrical phase currents. The phase
unbalanced condition causes increase in current in
one or more phases and might result in blown fuse in
affected phase circuit. This subsequently turns up into
single phasing of the motor which need necessarily
be checked to avoid severe motor damage.
(B) Broken Rotor Bars:
Outsized squirrel cage induction motors are
assembled using copper rotor bars and end-rings,
whereas small and medium sized cage motors are by
ISSN: 0975 – 6736| NOV 12 TO OCT 13 | VOLUME – 02, ISSUE - 02
Page 178
JOURNAL OF INFORMATION, KNOWLEDGE AND RESEARCH IN
ELECTRICAL ENGINEERING
and large crafted using die-cast aluminum rotor and
end rings. Therefore, there is preponderance of diecast aluminum rotors in the induction motors on
account of its distinctive light weight and low cost.
These die-cast rotors bring in numerous
predicaments, such as rotor eccentricity and
temperature instigated softening of bars and endrings. The alternative approach to fabricate the cage
rotors is by using brazed or welded copper rotor bars.
The rotor failures get trigger off by an assortment of
diverse stresses and strains that appear in the rotor.
Stresses can be electromagnetic, thermal, residual,
dynamic, environmental and mechanical [2]. The
foremost reason for the rotor failure typically resides
in flaws related to imperfect casting or substandard
jointing during manufacturing. Whilst the rotor bar is
slighter than the bar slot, slot harmonics will surface
that subsequently initiate radial advance of the bar,
particularly during starting-up, accelerating and
braking periods. This might beget frailty
consequently ensuing in fractured and ruptured rotor
bar(s). Thermal stress is an additional quandary,
occurring when the bar cannot progress
longitudinally in the rotor slot. Sustained motor
overloads and frequent starts, acceleration and
deceleration contribute to cultivate substantial
currents that bring forth considerable mechanical and
thermal stresses, consequently prompting damage to
the rotor and the stator as well. Even so, the induction
motor rotor faults by and large onset from a
minuscule crevice or a high resistivity spot in the
rotor bar. This spot shows temperature intensification
which worsens the damage until the rotor bar is
absolutely conked out. Subsequent to a solitary rotor
bar breakage, the rotor current reassigns to the other
co-existing healthy bar(s) causing the over current
and ultimately a cluster of broken bars. Another
frequently occurring source for the rotor failures is
the over-current typically due to rotor clog up
condition of the motor. It is on the whole not
beneficial or feasible to mend the rotor.
A broken rotor bar brings about assorted outcomes of
concern in induction motors. A documented
consequence of a broken rotor bar is the advent of the
purported sideband components [4, 9, 10]. Bar failure
progression that induces sideband frequencies in
current and power spectrum of the induction motors
involves an intricate mechanism. The precursors of
rotor faults encompass the twofold slip frequency
side bands flanking the supply frequency in a
frequency spectrum of the stator current. The lower
side band component to the left of the fundamental is
prompted by electrical and magnetic disparity in the
rotor cage of an induction motor [9], whilst the lower
sideband component to the right is attributable to
ensuing speed undulations on account of the varying
air-gap torque [4, 16]. The frequencies of these
sidebands, fSB, can be formulated as:
(1)
Where k = harmonic index = 1, 2, 3, ----s = per unit slip
p = number of fundamental pole-pairs
fe = fundamental supply frequency
On account of the configuration of the windings in an
induction motor, only the frequency components with
harmonics k =1, 5, 7, 11, 13, and so forth will figure
in the current. Then, the equation (1) can be
transformed as:
(2)
For harmonic index k = 1, the equation (2) gets
further altered as:
(3)
The spectral sideband components are availed of to a
large extent for induction motor fault classification
[9, 10, 20, 29, 48]. At high voltage or low inertia, the
upper sideband takes the limelight owing to speed
undulations churned out of the defective bar. Air-gap
space harmonics brought into being due to tooth or
core saturation influence the upper spectral sideband
component. Other correlated effects of broken rotor
bars that are used for motor fault diagnosis and
detection purposes include speed undulations [16],
torque ripples [19], stator power swings [24], and
stator current envelopes [49].
(C) Damaged Bearing Faults:
Four major types of bearing faults [3] are material
deterioration in inner race, outer race, cage, and ball
defects. The bearing faults can be classified into
cyclic faults and non-cyclic faults. Cyclic faults
emerge when the rolling element and the race or the
rolling element cage of the bearing passes through
the point of flaw. The deep scratches in a race or in a
rolling element are a type of cyclic fault. The material
abrasion, qualitative degradation of the lubricant due
to contaminants, insufficient lubrication, slither and
skid amongst the movable bearing components cause
mutilation of the contact areas, which is a non-cyclic
fault category. The bearing defects induce nonstationary and fault specific frequency components in
the stator current and the engendered vibrations.
a)
Cage Defect:
The bearing cage of a ball bearing upholds the balls
at evenly balanced positions and assists the confined
rolling of the balls along the raceways. When the
motor shaft is rotating, the bearing cage revolves at a
constant angular velocity that is mean of the inner
and outer race angular velocities. The cage angular
velocity can be exploited to work out the value of
dominant fault frequency due to cage defect, fCD as
given below:
(4)
ISSN: 0975 – 6736| NOV 12 TO OCT 13 | VOLUME – 02, ISSUE - 02
Page 179
JOURNAL OF INFORMATION, KNOWLEDGE AND RESEARCH IN
ELECTRICAL ENGINEERING
Where ωi = Angular speed of the inner race in
RPM
ωo= Angular speed of the outer race in RPM
D = Pitch Diameter
d = Ball Diameter
Φ = Ball contact angle
ri = Inner race radius, ro = Outer race radius
In case of motors, the outer race is connected to the
casing that is stationary. The inner race and the shaft
are mounted together and both rotate at the same
angular speed. Accordingly, it can be assumed that:
ωo = 0 and
ωi = ω r
(5)
Where ωr = Rotor angular speed in RPM
Incorporating the above mentioned assumption as
depicted in equation (5) fetches equation (6) as given
below:
(6)
Empirically, the cage defect fundamental frequency
for a ball bearing with six to twelve balls in it is
given as:
(7)
Where
b)
Inner Race Defect:
The inner race defect frequency, fIRD, depends on the
rate at which bearing balls pass through the point of
defect on the inner race. Each ball move across the
flaw point at a rate that is proportional to the
difference of angular speed of the cage and inner
race. The fault frequency of the inner race defect is
also proportional to the number of balls in the
bearing. The fundamental fault frequency for the
inner race defect (for conditions ωo = 0 and ωi = ωr )
is given as:
(8)
Where the used variables have definition as
mentioned along with eqn. (4)
The empirical formula for inner race defect fault
frequency of a ball bearing consisting of six to twelve
balls in it is given as:
(9)
Where ωrs is as mentioned along with eqn. (7)
c)
Outer Race Defect:
The outer race defect frequency, fORD, depends on the
rate at which bearing balls cross the point of defect
on the outer race. Each ball move across the point of
defect at a rate that is proportional to the difference of
angular speed of the cage and outer race. The fault
frequency of the outer race defect is also proportional
to the number of balls in the bearing. The
fundamental fault frequency corresponding to the
inner race defect is given as:
(10)
Where the used variables have definition mentioned
along with eqn. (4)
Using equation (5), i.e. ωo = 0 and ωi = ωr (Where ωr
= Rotor angular speed in RPM), yields:
fODR = (0.4).n.ωrs
(11)
Where ωrs is as mentioned in eqn. (7)
(D) Inter-Turn Short Circuits:
In Induction motors the Inter-turn short circuits occur
as short circuits between turns of one phase, or
between turns of two phases, or between turns of all
phases, or between winding conductors and the stator
core. Inter-turn short circuits between turns of same
phase, winding short circuits, short circuits between
winding and stator core, short circuits between
different phases is usually caused by stator voltage
transients and abrasion. Burning of winding
insulation and complete winding short circuits of all
phase windings are usually caused by motor
overloads and blocked rotor, stator energising by subrated voltage, overrated voltage power supplies,
frequent starts and rotation reversals. Inter-turn short
circuits are also caused due to voltage transients
because of the successive reflections resulting from
the cable connection between induction motor and
induction motor drives.
Inter-turn short circuits in stator windings constitute a
category of faults that is most common in induction
motors. Typically, short circuits in stator windings
occur between turns of one phase, or between turns
of two phases, or between turns of all phases.
Moreover, short circuits between winding conductors
and the stator core also occur. The different types of
winding faults are summarizes below as follows [37]:
Inter-turn short circuits between turns of the same
phase, winding short circuits, short circuits between
winding and stator core, short circuits on the
connections, and short circuits between phases are
usually caused by stator voltage transients and
abrasion.
Stator faults originate in the stator core or in the
stator windings. Stator winding faults can be due to
several different reasons. The insulation damage can
be, for example, due to impact damage during
installation, movement due to repeated starting, slack
core laminations, thermal damage due to over current
and finally due to thermal aging. The stator winding
faults, in a case of a low voltage induction motor, are
often not repaired. If repaired, the machine is taken to
a special service establishment or to a manufacturer
for re-winding. Stator faults are indicated by
analysing e.g. the phase unbalance of the stator
currents or the axial leakage flux. The unbalance is
calculated with the aid of symmetrical components.
The negative sequence current or impedance is used
as a fault indicator.
(E) Air gap eccentricity:
ISSN: 0975 – 6736| NOV 12 TO OCT 13 | VOLUME – 02, ISSUE - 02
Page 180
JOURNAL OF INFORMATION, KNOWLEDGE AND RESEARCH IN
ELECTRICAL ENGINEERING
The presence of static and dynamic eccentricity can
be detected by monitoring the behaviour of the
fundamental sidebands of the supply frequency. The
frequencies of interest are given as:
(12)
Where fe = supply frequency
s = per unit slip
m = 1, 2, 3, --P = number of poles
The air gap eccentricity is also detected by
monitoring the behavior of sidebands of the slot
frequencies in stator current. The sideband
frequencies of interest are given as:
(13)
Where k = 1, 2, 3, ---------, R (R = number of rotor
slots)
nd = order of the rotating eccentricity
nw = order of the stator MMF harmonics
III: INDUCTION MOTOR FAULT MONITORING
The features used in fault detection methods are
intended to classify the motor condition as faulty and
healthy and also to identify the motor fault type. The
detection technique is intended to identify electrically
detectable faults in the induction motors and also
identify the fault severity. The classifier technique
classifies the Induction motor as either healthy or
faulty. The fault features are extracted from a suitable
motor parameter. Good number condition-monitoring
propositions have concerted explicitly with sensing
related failure methodologies. All of the currently
accessible techniques entail the user to acquire some
competence in making a distinction amongst normal
operating condition from a prospective fiasco.
Ideally, it is aspired to create a diagnostic procedure
that endow with a clear inference of machine health
in minimum time through processing of minimal
measurable inputs [8]. Various different parameters
such as temperature, current, voltage, vibration, flux
and acoustics have been in force for monitoring
electrical machines.
Good number condition-monitoring propositions
have concerted explicitly with sensing related failure
methodologies. All of the currently accessible
techniques entail the user to acquire some
competence in making a distinction amongst normal
operating condition from a prospective fiasco.
Ideally, it is aspired to create a diagnostic procedure
that endow with a clear inference of machine health
in minimum time through processing of minimal
measurable inputs [8]. Expediency, consistency, and
sensitivity are the basis of sensor signals. The
existing methods of condition monitoring of
electrical machines are [6], [7], [14]:
1. Noise Monitoring:
Acoustic noise from air gap eccentricity in induction
motors can be used for fault detection. Noise
monitoring is accomplished by measuring and
analyzing the acoustic noise spectrum. However, the
application of noise measurements in a plant is not
practical because of the noisy background from other
machines operating in the vicinity.
2. Torque Monitoring [25]:
Almost all motor faults cause harmonics with specific
frequencies in the air gap torque. However, air gap
torque cannot be measured directly. From the input
terminals, the instantaneous power includes the
charging and discharging energy in the windings.
Therefore, the instantaneous power cannot represent
the instantaneous torque. From the output terminals,
the rotor, shaft, and mechanical load of a rotating
machine constitute a torsional spring system that has
its own natural frequency. The attenuations of the
components of air gap torque transmitted through the
torsional spring system are different for different
harmonic orders of torque components. Generally,
the waveform of the air gap torque curve is different
from the torque measured at the shaft. The fault
condition can be identified by monitoring the specific
harmonics in the air gap torque.
The air gap torque in terms of measurable motor
terminal quantities is given as:
(14)
Where iA, iB , and iC are three-phase line currents of
an induction motor,
VCA and VAB are line-to-line voltages
r is half of the line-to-line resistance
p is the number of pole pairs
Frequencies of major torque harmonics associated
with the certain defects in induction motors are as
follows:
Under normal operation:
Angular frequency of torque = 0
With a single-phase stator:
Angular frequency of torque = -2ωs
With a single-phase rotor:
Angular frequency of torque = 2sωs
Where ωs is the supply frequency in rad/s, and s is the
slip. Therefore, the fault condition can be identified
by monitoring the special harmonics in the air gap
torque.
3. Flux Monitoring:
Air gap flux of induction motors contains rich
harmonics. A flux monitoring scheme can give
reliable and accurate information about electrical
machine conditions. Any change in air gap, winding,
voltage, and current can be reflected in the harmonic
spectra. The change of air gap flux is a function of
static eccentricity. Air gap flux can be measured by
search coils installed in the stator core. Because of
the enclosed structure of induction motors, this
operation requires the disconnection of induction
ISSN: 0975 – 6736| NOV 12 TO OCT 13 | VOLUME – 02, ISSUE - 02
Page 181
JOURNAL OF INFORMATION, KNOWLEDGE AND RESEARCH IN
ELECTRICAL ENGINEERING
motors from the main supply before dismantling. As
such, this is neither practical nor economical for the
motors that are in service.
4. Vibration Analysis:
Vibration monitoring [5], [21]-[24] is one of the
oldest condition monitoring techniques and is widely
used to detect mechanical faults such as bearing
failures or mechanical imbalance. A piezo-electric
transducer providing a voltage signal proportional to
acceleration is often used. This acceleration signal
can be integrated to give the velocity or position.
5. Current Monitoring:
One of the, most economically attractive technology
in motor condition monitoring is stator current
monitoring [11]-[20]. The stator current of an
induction motor is readily available since it is used to
protect machines from destructive over-currents,
ground current, etc. Therefore, current monitoring is
a sensor less detection method that can be
implemented without any extra hardware.
The vibration monitoring [5], [21-24] is one of the
widely used techniques for detection of electrical and
mechanical faults in electrical machines. The stator
current monitoring [11]-[20] is relatively a recent
technique that is fast gaining importance due to its
non-invasive nature, cost-effectiveness, preciseness
in analysis, easy and efficient signal processing, and
convenient installation [34]. The stator current
monitoring and analysis can provide as much
information as the vibration monitoring provides
[14], [22].
The motor stator current analysis is based on the fact
that the stator and rotor faults are ultimately reflected
in the stator current and in most of the cases cause
unbalanced phase currents or introduces a distinct
signature in the stator current. The motor stator
current analysis being non-intrusive, comparatively
simple to implement, load and operating point
independent - prove out to possess an advantageous
edge over the other analysis techniques.
Various stator current analysis methods that are
commonly employed for fault detection in Induction
motors are as shown below:
Motor stator current analysis:
Time domain techniques
Statistical parameter analysis –
Root Mean Square (RMS), mean, variance, crest
factor, skewness, kurtosis etc
Time synchronous averaging based
methods – Time synchronous averaged (TSA) signal,
Residual signal analysis (RSA) etc.
Spatial angular vector (SAV)
analysis
Negative sequence component
analysis
Filter
based
methods–
Demodulation, enveloping etc.
Stochastic methods – Thresholding,
Autoregression based methods
Frequency based techniques
(STFT)
Spectrum analysis
Fast fourier transform (FFT)
Discrete fourier transform (DFT)
Short term fourier transform
Power cepstrum analysis
Time-frequency based techniques
Spectrogram analysis
Wigner distribution (WD) analysis
Continuous wavelet transform
analysis
Discrete wavelet pocket analysis
Time-averaged wavelet spectrum
(TAWS)
Time-frequency-scale
(TMS)
domain
The commonly employed artificial intelligence based
interpreting techniques are as follows [15,17,18]:
Neural network based inference
Feed forward networks
Multi layer perceptron
Radial basis function
(RBF)
Principal
component
analysis (PCA)
Recurrent networks
Kohonen Self-organising maps
(SOM)
Fuzzy set based inference
Expert system based inference
Neuro-fuzzy computing based inference
Adaptive Neuro-fuzzy inference
Fuzzy expert system based inference
Bayesian classifier
Support vector machines (SVM)
IV: Causes of Faults in Induction motors
The reason for induction motor impairment [2] can
be a multiplicity of factors such as filth and dirt
instigated temperature intensification; unwarranted
vibrations due to faulty bearings; thermal stresses due
to rotor rub, rotor skewing, and end ring heating;
mechanical stresses due to air gap eccentricity,
frequent startups, rapid acceleration and deceleration;
long persistent overload conditions; transient torques
due to faulty bearings, poor supply quality, and
unbalanced stator phase winding; flaws in
manufacture or design; imperfect installation;
deterioration due to abrasion, erosion and aging.
Most of the time there are copious factors that beget
motor breakdown. The most unequivocal basis for
motor breakdown is damage of the bearings or
winding or rotor but the paramount rationale that is
often an attribute to such failures is overheating
prompted through dirt, filth and grime.
ISSN: 0975 – 6736| NOV 12 TO OCT 13 | VOLUME – 02, ISSUE - 02
Page 182
JOURNAL OF INFORMATION, KNOWLEDGE AND RESEARCH IN
ELECTRICAL ENGINEERING
The rotor failures get trigger off by an assortment of
diverse stresses and strains that appear in the rotor.
Stresses can be electromagnetic, thermal, residual,
dynamic, environmental and mechanical [2], [22].
The foremost reason for the rotor failure typically
resides in flaws related to imperfect casting or
substandard jointing during manufacturing. Whilst
the rotor bar is slighter than the bar slot, slot
harmonics will surface that subsequently initiate
radial advance of the bar, particularly during startingup, accelerating and braking periods. This might
beget frailty consequently ensuing in fractured and
ruptured rotor bar(s). Thermal stress is another
quandary, occurring when the bar cannot progress
longitudinally in the rotor slot. Sustained motor
overloads and frequent starts, acceleration and
deceleration contribute to cultivate substantial
currents that bring forth considerable mechanical and
thermal stresses, consequently prompting rotor and
the stator damage. Even so, the induction motor rotor
faults by and large onset from a minuscule crevice or
a high resistivity spot in the rotor bar. This spot
shows temperature intensification which worsens the
damage until the rotor bar is absolutely conked out.
Subsequent to a solitary rotor bar breakage, the rotor
current reassigns to the other co-existing healthy
bar(s) causing the over current and ultimately a
cluster of broken bars. Another frequently occurring
source for the rotor failures is the over-current
typically due to rotor clog up condition of the motor.
It is on the whole not beneficial or feasible to mend
the rotor.
Bearing faults are one of the prominent causes that
instigate the failures in induction motors [1, 8]. The
indications of bearing mutilation are bumpy running
with jerks, abridged exactness of producing shaft
movement with very less rotational tolerance and an
atypical noise. The bearing faults in the electrical
motors can on the whole be attributed to the material
attrition and abrasion at contact points, ageing and
exhaustion, misalignment, sustained load overstresses and pressure induced welding. The erosion,
attrition and abrasion of the material crop up as a
result of presence of contaminants and impurities,
friction induced overheating and subsequent
occurrence of hot spots on the bearing balls or rolls,
inner and outer races. The duration of bearing life
span is also reliant on the quality of the lubricant,
stresses due to mechanical load and electrical starts.
The frequent start ups and rapid accelerating and
decelerating periods cause repeated over stressing of
the bearings. Besides unfeigned overloading, the
lumber on the bearing in induction motors can be due
to improper alignment or rotor unbalance, and sooner
or later bring forth a state of the bearing that is not
appropriate for unblemished operation of the
machine.
Under ideal state of affairs, the rotor potential is
believed to be zero which is not practically the case.
A potential relative to the ground emerge at the rotor
due to inequality in phase capacitances. The rotor
voltage set off a difference of potential crosswise the
bearings. This leads to a current flowing through the
bearing and brings about an alteration in the chemical
composition of the lubricant, consequently resulting
in degradation of the quality of the lubricant. This
further brings in abrasion in the bearing and may
sometimes set in electrical discharges between inner
and outer races, eventually leading to the inopportune
bearing failures. Stator faults originate in the stator
core or in the stator windings. Stator winding faults
can be due to several different reasons. The insulation
damage can be due to impact damage during
installation, movement due to repeated starting, slack
core laminations, thermal damage due to over current
and due to thermal aging. Stator faults are indicated
by perceiving parameters such as phase unbalance of
the stator currents or the axial leakage flux or the
vibration content of the motor etc.
V: Schematic block diagram for fault detection in
Induction motors
The motor under test (MUT) is made to run as linefed induction motor or an inverter fed adjustable
speed induction drive (ASID), as the case may be, by
feeding an apposite 3-phase supply. The data
acquisition and analysis software (the popular ones
are dSPACE-Matlab or National Instrument’s DAQ –
Matlab etc.) along with the pertinent control card is
used to develop the data acquisition setup. The input
signals are processed and analysed through suitable
signal processing and analysis tools so as to extract
fault features and present them in a format acceptable
to the successive classifier stage. The interpreter
stage takes a decision on the basis of fault features
presented to it as regards the motor health status. The
general schematic of an experimental setup for fault
detection in three-phase induction motors is as shown
in fig.1 below:
Data
Acquisition
Hardware and
software
Data
Acquisition and
Control
Hall Effect
current and
voltage
Transducers
Three-phase
Inverter/On-Off
Arrangement
Speed
Measur
ement
3-Φ Induction
motor
ISSN: 0975 – 6736| NOV 12 TO OCT 13 | VOLUME – 02, ISSUE - 02
Page 183
JOURNAL OF INFORMATION, KNOWLEDGE AND RESEARCH IN
ELECTRICAL ENGINEERING
Australasian
Universities
Power
-------,
2003/itee.uq.edu.au.
[8]
O’Donnell P., ”Report of Large Motor
Loading
Reliability
Survey of Industrial and Commercial
arrangement
Installations”, Part I and II, Motor Reliability
Working Group, IEEE Industry Application Society
Fig. 1: General schematic of the experimental setup (IAS), IEEE Transaction on Industry Applications, Pp
853-872, July/August 1985; Part III, Motor
Reliability Working Group, IEEE Industry
Application Society (IAS), IEEE Transaction on
VI: CONCLUSION
Industry Applications, Pp 153-158, January/February
The brief review of the commonly occurring
1985.
induction motor faults has been presented. The
[9]
Arthur N. and J. Penman, “ Induction
relevant references have been mentioned for ready
Machine Condition Monitoring with higher order
reference. The conventional and recent signal
spectra”, IEEE transaction on Industrial Applications,
analysis, fault extraction and interpreter techniques
Vol.47, No.5, Pp. 1031-1041, October 2000.
have been discussed. The outcome and conclusions
[10]
Arthur N., J. Penman, A. McLean, and A.
buoy to support the author’s assumption that it is
Parsons, “Induction Machine Condition Monitoring
possible to employ certain new techniques so as to
with higher order spectra, Part II: Variable frequency
improve the consistency of existing, well reported,
operation and automated diagnosis”, IECON’98,
and established strategies for fault detection.
Proceedings of the 24th Annual Conference of the
Furthermore the techniques that are apt in processing
IEEE Industrial Electronics Society, Pp. 1895-1900,
non-stationary signal analysis have a definite edge
Vol. 3, 31 Aug.-4 Sept. 1998 .
over the conventionally employed techniques based
[11]
Barendse P. S. and P.Pillay, “The Detection
upon current monitoring alone. It is worth to be
of Unbalanced Faults in Inverter-Fed Induction
mentioned that recent works in motor fault detection
Machines”, IEEE---------, 2007.
are anchored in analysis of stator current based on
[12]
Bellini A., G. Fraceschini, and C. Tassoni,
time-frequency (or time-scale) methods.
“Monitoring of Induction Machines by Maximum
Covarience Method for Frequency Tracking”,
REFRENCES
Conference Records of the 2004 IEEE Industry
[1]
Albrecht P.F., J.C. Appiarius, and D.K.
Applications Conference, 39th Annual Meeting, Vol.
Sharma,”Assessment of the reliability of motors in
2, Pp. 743-749, 3-7 Oct. 2004.
utility applications-Updated,” IEEE Transactions on
[13]
Bellini A., F. Filippetti, G. Franceschini, T.
Energy conversion, Vol. 20, Pp. 719-729, 2005.
J. Sobczyk and C. Tassoni, “Diagnosis of Induction
[2]
Bonnett A. H. and George C. Soukup,
Machines by d-q and i.s.c Rotor Models”, IEEE
“Cause and Analysis of Stator and Rotor Failures in
International Symposium on Diagnostics for Electric
Three-Phase Squirrel-Cage Induction Motors”, IEEE
Machines, Power Electronics and Drives, Vienna,
Transactions on Industry Applications, Vol. 28, No. 4,
Austria, 7-9 September 2005.
July/August 1992.
[14]
Benbouzid M.E.H.,” A Review of Induction
[3]
Devaney M. J. and L. Eren, “Detecting
Motors Signature Analysis as a medium of faults
Motor Bearing Faults”, IEEE Instrumentation and
detection,” IEEE Transactions on Industrial
Measurenent Magazine, Pp. 30-50, December 2004.
Electronics, Vol. 47, No. 5, Oct. 2000.
[4]
Faiz J. and B. M. Ebrahimi, “Mixed Fault
[15]
Benbouzid M.E.H., Mighelle Vieira, and
Diagnosis in Three-phase Squirrel-Cage Induction
Céline Theys, “Induction motors’ Faults Detection
Motor Using Analysis of Airgap Magnetic Field”,
and Localization Using Stator Current Advanced
Progress in Electro-magnetics Research, PIER 64,
Signal processing Techniques”, IEEE Transactions on
Pp. 239-355, 2006.
Power Electronics, Vol. 14, No.1, Pp. 14-22, January
[5]
Han Y. and Y. H. Song, “Condition
1999.
Monitoring Techniques for Electrical Equipment-A
[16]
Bonnett A. H. and George C. Soukup,
Literature Survey”, IEEE Transaction on Power
“Cause and Analysis of Stator and Rotor Failures in
Delivery, Vol. 18, No. 1, January 2003.
Three-Phase Squirrel-Cage Induction Motors”, IEEE
[6]
Nandi S., H.A. Toliyat, and X. Li,”
Transactions on Industry Applications, Vol. 28, No. 4,
Condition Monitoring and Fault Diagnosis of
July/August 1992.
Electrical Motors- A Review,” Proceedings of 34th
[17]
Ilonen J., J. K. Kamarainen, T. Lindh, J.
IAS annual meeting on Industry Applications, IEEE,
Ahola, and H. Kalaviainen, ”Diagnosis tool for motor
Vol. 3, Pp. 197-204, 3-7 Oct 1999.
condition monitoring”, IEEE Transaction on
[7]
Sin M. L., W. L. Soong, and N. Ertugrul,
Industrial Applications, Vol. 41, No.4, Pp. 963-967,
“Induction Machine On-line Condition Monitoring
2005.
and Fault Diagnosis – A survey”, Proceedings of the
[18]
Kim K. and A. G. Parlos. “Reducing the
Impact of False Alarms in Induction Motor Fault
ISSN: 0975 – 6736| NOV 12 TO OCT 13 | VOLUME – 02, ISSUE - 02
Page 184
JOURNAL OF INFORMATION, KNOWLEDGE AND RESEARCH IN
ELECTRICAL ENGINEERING
Daignosis”, Journal of Dynamic Systems,
Measurement, and Control, Transactions of the
ASME, Vol. 125, Pp. 80- 95, March 2003.
[19]
Nejjari H. and M. E. H. Benbouzid,
“Application of Fuzzy Logic to Induction Motors
Condition Monitoring”, IEEE Power Engineering
Review, Pp. 52-54, June 1999.
[20]
Schoen R. R. and T. G. Habetler,
”Evaluation and Implementation of a System to
Eliminate Arbitrary Load Effects in Current-Based
Monitoring of Induction machines”, IEEE
Transaction on Industrial Applications, Vol. 33, No.
6, Pp. 1571-77, Nov/Dec. 1997.
[21]
Dorrell D. G., W. T. Thomson, and S.
Roach. “Analysis of air gap flux, current and
vibration signals as a function of the combination of
static and dynamic air gap eccentricity in 3-phase
induction motors”, IEEE transaction on Industrial
Applications, Vol.33, No.1, Pp. 24-34, Jan./Feb 1997.
[22]
Riley C. M., B. K. Lin, T. G. Habetler, and
R. R. Schoen, “Stator Current Harmonics And Their
Causal Vibrations: A Preliminary Investigation of
Sensorless Vibration Monitoring Applications”, IEEE
Transaction on Industry applications, Vol. 35, no. 1,
Pp. 94-99, Jan/Feb 1999.
[23]
Riley C. M., B. K. Lin, T. G. Habetler, and
R. R. Schoen, “A Method for Sensorless On-Line
Vibration Monitoring of Induction Machines”, IEEE
Transaction on Industry applications, Vol. 34, no. 6,
Pp. 1240-45, Nov/Dec 1998.
[24]
Riley C. M., B. K. Lin, T. G. Habetler, and
R. R. Schoen, “A Method for Sensorless On-Line
Vibration Monitoring of Induction Machines”, IEEE
--------, Pp. 201-207, 1997.
[25]
Hsu John S., “Monitoring of Defects in
Induction Motors Through Air-Gap Torque
Observation”, IEEE Transactions on Industry
Applications, Vol. 31, no.5, Pp. 1016-1021,
September/October 1995.
[26]
Khan M. A. S. K., Tawfik S. Radwan, and
M.Azizuur Rahman, “Real –Time Implementation of
Wavelet Packet Transform-Based Diagnosis and
Protection of Three-Phase Induction Motors”,IEEE
Transactions on Energy Conversion, Vol. 22, No.3,
Pp. September 2007.
[27]
Kim K. and A. G. Parlos. “Induction Motor
Fault Diagnosis based on Neuropredictors and
Wavelet Signal Processing”, IEEE transaction on
Mechatronics, Vol. 7, No. 2, Pp. 201- 219, June 2002.
[28]
Magnago F. and A. Abur,” Fault location
using wavelets”, IEEE Transactions on Power
Delivery, 13(4), 1475-1480, 1998.
[29]
Zanardelli W. G., E. G. Strangas, H. K.
Khalil, J. M. Miller, “Wavelet Based Methods for the
Prognosis of Mechanical and Electrical Failures in
Electric Motors”, Mechanical Systems and Signal
Processing (Elsavier), Pp 411-426, 2005.
[30]
Zanardelli W. G., and E. G. Strangas,
“Methods to Identify Intermittent Electrical and
Mechanical Faults in Permanent Magnet AC Drives
Based on Wavelet Analysis”, IEEE Conference on
Vehicle Power and Propulsion, 7-9 Sept. 2005.
[31]
Awadallah M. A. and M. M. Marcos,
“Application of AI Tools in Fault Diagnosis of
Electrical Machines and Drives – An Overview”,
IEEE Transactions on Energy Conversion, Vol. 18,
No. 2, June 2003.
[32]
Chow Mo-yuen, Peter M. Mangum, and Sui
Oi Yee, “A Neural Network Approach to Real-Time
Condition Monitoring of Induction Motors”, IEEE
Transactions on Industrial Electronics, Vol. 38, No. 6,
December 1991.
[33]
Eren L., A. Karahoca and M. J. Devaney,
“Neural Network Based Bearing Fault Detection”,
Instrumentation and Measurement Technology
Conference, Como, Italy, 18-20 May 2004, Pp. 16571660.
[34]
Han Tian, Bo-suk Yang, Won-Ho Choi, and
Jae-Sik Kim, “Fault Diagnosis System of Induction
Motors Based on Neural Network and Genetic
Algorithm Using Stator Current
Signals”,
International Journal of Rotating Machinery, Pp. 113, Volume 2006.
[35]
Lehtoranta J. and H. N. Koivo, “Fault
Diagnosis of Induction Motors with Dynamical
Neural Networks”, IEEE International Conference on
Systems, Man and Cybernetics, Pp. 2979-2984, Vol.
3, Oct. 2005.
[36]
Nejjari H. and M. E. H. Benbouzid,
“Application of Fuzzy Logic to Induction Motors
Condition Monitoring”, IEEE Power Engineering
Review, Pp. 52-54, June 1999.
ISSN: 0975 – 6736| NOV 12 TO OCT 13 | VOLUME – 02, ISSUE - 02
Page 185
