Fault Detection of Induction Motor Using Current and Vibration Monitoring

Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

International Journal of Advanced Computer Research (ISSN (print): 2249-7277 ISSN (online): 2277-7970)

Volume-3 Number-4 Issue-13 December-2013

Fault Detection of Induction Motor Using Current and Vibration Monitoring


Sumit Narwade1, Prasad Kulkarni2, C.Y.Patil3

Abstract
Induction motors are used worldwide as the
“workhorse” in industrial applications. Although,
these electromechanical devices are highly reliable,
susceptible to many types of faults. Condition
monitoring and fault diagnosis of induction motors
are of great importance in production lines. It can
significantly reduce the cost of maintenance and the
risk of unexpected failures by allowing the early
detection of potentially catastrophic faults. In this
paper I have used both vibration and motor current
signature analysis to detect the fault. The various
fault discussed in this paper are- Mechanical fault
such as bearing damage and Electrical Fault such
as unbalanced voltage supply, single phasing.
Condition monitoring, signal processing and data Figure 1: Types of Fault
analysis are the key parts of the Induction Motor
fault detection scheme. Bearing faults are the most frequent fault in Electric
Motors (41%) according to an IEEE motor reliably
study, followed by stator (37%) and rotor fault (10%)
Keywords [7].
Induction Motor, Unbalance voltage, bearing damage,
condition monitoring, FFT.
Condition monitoring of Induction motor are of great
importance in the production lines. Condition
monitoring means continuous evaluation of the health
1. Introduction of the motor throughout its service life. There are
many condition monitoring methods, including
The fault detection and protection are as historic as
vibration monitoring, thermal monitoring, chemical
Induction Motor themselves. Induction motors are monitoring, current monitoring techniques are
critical components in many industrial processes. usually applied to detect the various types of
They are very robust and highly reliable machines induction motor fault. I mainly use Current and
however they may be subjected to different types of Vibration monitoring to detect the fault. In current
faults. Failure of such Induction motor may cause
monitoring, no additional sensors are necessary. This
plant shutdown, personal injuries and waste of raw
is because the basic electrical quantities associated
material. However, induction motor faults can be with electromechanical plants such as current and
detected in an initial stage in order to prevent the voltage are readily measured by tapping into the
complete failure of an induction motor and existing voltage and current transformers that are
unexpected production costs. always installed as part of the protection system and
for stationary signal this method is beneficial. This
The main reason for the motor faults is mechanical
method uses the current spectrum of the machine for
and electrical stresses. Mechanical stresses are caused locating characteristic fault frequencies. But for non
by overloads and abrupt load changes, which may stationary signal vibration monitoring is best. In
cause bearing faults and rotor bar breakage. The vibration monitoring a vibration spectrum is used to
electrical stresses may produce stator winding short detect the fault in induction motor. Vibration and
circuits and result in a complete motor failure.
current are widely used in the industry since both
The different fault may occur in Induction motor can
allow detecting several fault conditions from distinct
be classified as follows. nature [7, 8]. Now days fault diagnosis uses modern
techniques data processing and spectral analysis
272
International Journal of Advanced Computer Research (ISSN (print): 2249-7277 ISSN (online): 2277-7970)
Volume-3 Number-4 Issue-13 December-2013

techniques. In this paper I used the Fourier spectral


analysis by FFT to extract the features, which are
sensitive to the presence of faults. In this study I used
a DAQ card to acquire a sensor signal and a
MATLAB programmed to detect the fault frequency.
Mechanical and Electrical faults print frequency
features in the spectrum of current and vibration
which can be used for fault detection by comparing
the amplitude of certain frequency components under
fault and normal conditions. Therefore, this paper is
organized as follows: first, a qualitative description Figure 3: Inner Race Fault
of the unbalance voltage condition, single phasing,
bearing damage, MCSA and vibration are presented.
Secondly, experimental are introduced to show the
signatures associated to this fault using vibration and
stator current signals in an Induction motor. [3, 6].

2. Methodology
A. Bearing Fault in Induction motor
Bearing are common elements of Induction
Machines. They are employed to permit the rotary
motion of the shaft. The bearing mainly consists of Figure 4: Outer Race Fault
two rings called the inner and outer rings. A set of
balls or rolling elements placed in raceways rotate B. Unbalance in supply voltage
inside these rings. A continued stress on the bearings A very important aspect of condition monitoring of
cause fatigue failures, usually at the inner and outer induction motor is to detect the external faults.
races of the bearings. Small pieces break loose from Unbalance in supply voltage and single phasing are
the bearing, called flaking or spalling. These failures the two types of external faults of induction motor are
result in rough running of the bearings that generates considered in my work. Unbalance in supply voltage
detectable vibrations and increased noise levels. And is created by connecting three single phase
this process is helped by other external sources transformers in three phases of the supply. Here
including contamination. Corrosion, brinelling, supply voltage in B phase is reduced to 10% less than
improper lubrication, improper installation. In some the rated voltages in the other two phases to create
case shaft voltage and current are also sources for unbalance. Unbalance voltage is nothing but unequal
bearing failure. High bearing temperature is also distribution of incoming voltages or when the phase
another reason for bearing failure. separation is not 120º. When there are unequal
The different faults that may occur in bearing as incoming voltages between the three legs of a motor,
follows the motor runs hotter. The reason why an unbalance
 Outer raceway defect motor runs hotter is because as the voltage is out of
 Inner raceway defect balance so is the winding current. A small voltage
unbalance causes a larger current unbalance, which in
 Ball defect
turns causes the motor.

Types of voltage unbalance


 Single phase under voltage
 Two phase under voltage
 Three phase under voltage
Single phase under voltage unbalance condition
arises when there is a large single phase load in the
system and it doesn‟t have enough compensation. In
this situation the voltage in that particular phase will
Figure 2: Ball bearing dimensions be lower than the other two phases.
273
International Journal of Advanced Computer Research (ISSN (print): 2249-7277 ISSN (online): 2277-7970)
Volume-3 Number-4 Issue-13 December-2013

Two phases under unbalance condition arises when


two of three phases have heavy load and don‟t have
enough compensation. In this situation those two
phases will have a higher voltage drop than third
phase. Three phases under unbalance condition arises
when the loads of three phases are too heavy and not
balanced. Because of voltage unbalance condition
resistive load are generally unaffected but it generates
excessive heat. Motor torque and speed are affected
by the voltage unbalance condition it creates
excessive noise and vibration [1].

C. Single Phasing Fault


For proper working of any three phase induction Figure 5: The process for fault diagnosis
motor, it must be connected three phase alternating
current (Ac) power supply of rated voltage and load. the next stage we use a DAQ system means Data
Once these three phase motors are started they will acquisition system to record that sensor signal in the
continue to run even if one of the three phase supply digital form. In block C we used different type of
lines gets disconnected. The loss of current is signal processing techniques (such as FFT, STFT,
described as single phasing. Single phasing is the Wavelet) to analysed that sensor signal and to extract
condition in three phase motors and transformers the feature which are sensitive to presence of fault. In
wherein the supply to one of the phases is cut off. next step we get the fault frequency of different fault
Single Phasing causes negative phase sequence by analysing the sensor signal. And finally we get
components in the voltage. Single Phasing is cause detailed information and degree of severity.
by the use of single phase protection devices such as
fuses and circuits breakers. Three phase loads should 1. Current Monitoring
be protected by devices which cause the interruption Current Park‟s Vector, Zero sequence and negative
of power to all three phases simultaneously hence a sequence current monitoring and current signature
fault occurs. Defective contacts in three phase analysis fall under category of Electrical monitoring.
breakers can also cause single phasing. It can be These methods use stator current to detect various
sometimes cause excessive noise and vibration in kinds of machine and inverter faults. In most
motors. applications, the stator current of an induction motor
is readily available since it is used to protect the
D. Condition Monitoring System machines from destructive over currents, ground
Condition monitoring means to access the actual currents etc. Therefore current monitoring is sensor
condition of motor using the measurements taken less detection method that can be implemented
while the motor is operating. without any extra hardware. [8]

I mainly use two types of condition monitoring Motor Current Signature Analysis
technique to detect different fault. Current and This method is used current spectrum to detect the
vibration monitoring. A block diagram for condition various faults. It is the online analysis of current to
monitoring for fault diagnosis is shown in figure 5. detect faults in three phase induction motor. This
method analyses the motor signal by using signal
This is the method for fault diagnosis. In this stage, processing algorithms such as FFT, STFT, and
First of all we create different fault (such as bearing Wavelet etc. MCSA techniques include parametric,
fault, unbalance voltage) due to which symmetry of nonparametric, and high-resolution spectrum analysis
motor would effect and creates fault characteristics methods. In the parametric methods, autoregressive
frequency. After that we get that fault frequency by (AR) models have been fitted with time series of the
different type of sensors (such as for vibration signal signal, and model parameters have been used to
we can use piezoelectric accelerometer, for speed compute the frequency spectrum. Furthermore,
measurement tachometer etc. ) shown in block B. In nonparametric methods are based on Fourier
transforms in order to search for periodicities of the
signal. And finally a high resolution spectrum

274
International Journal of Advanced Computer Research (ISSN (print): 2249-7277 ISSN (online): 2277-7970)
Volume-3 Number-4 Issue-13 December-2013

method corresponds to an Eigen value analysis of the considered in this work. Unbalance in supply voltage
autocorrelation matrix of the time series signal [2, 9, is created by connecting three single phase
and 10]. One of the classical and widely used transformers in three phases of the supply. Here
nonparametric spectrums method as a MCSA supply voltage in B phase is reduced to 10% less than
technique is the well-known fast Fourier transform the rated voltages in the other two phases to create
(FFT). The FFT is a simple and computationally unbalance. In this experiment I have created
efficient algorithm to compute the discrete Fourier unbalance supply by setting different voltages for
transform (DFT) of a discrete-time series function. each phase because of that reason I used three single
In my work I used the FFT technique to analyse the phase auto transformer.
signal.

2. Vibration Monitoring
All electric motors generate noise and vibration, and
analysis of produced noise and vibration can be used
to give information on the condition of motor. Even
very small amplitude of vibration of machine frame
can produce high noise. Noise and vibrations in
electric machines are caused by forces which are of
magnetic, mechanical and aerodynamic origin. For
stationary signal MCSA best but for non-stationary Figure 6: Experiment Set-up for unbalance
signal it‟s not convenient option for non-stationary voltage and single phasing.
signal vibration monitoring is generally used.
Three phases under- voltage unbalance condition
Four vibration properties are crucial to understanding After continuous analysis on that single phase at 140
and resolving the machine problems. These include volt I create a three phase under voltage condition
Amplitude, which indicate the level of severity of the where that all three voltages are too heavy and try to
measured condition. Frequency, which indicates the balance the condition. On my next analysis I still
repetition rate of the contributing source or sources of increase the voltage of that blue phase till the 160
the measured condition; Phase, which presents the volt, in this condition that voltages are still too heavy
timing relationship between two signals contributing and in this condition motor starts to burn.Fig.12
to the measured condition; Modulation, the process
by which the response amplitude at some frequency Single Phasing Fault
is varied by a lower frequency excitation response. In this experiment I have created a single phasing
With the help of this we can get detail information fault by removing the yellow phase. Because of that
about asymmetry in motor [4]. condition motor starts vibrating and noise was also
generates. In this condition because of removing one
3. Experiment Set Up 1 phase other two phase try to balance the condition. I
take FFT to extract the feature which is sensitive to
Unbalance supply voltage and Single phasing presence of fault. (Fig.13)
Analysis (MCSA Approach)
In order to diagnose unbalance supply voltage of In order to diagnose the single phasing of induction
induction motor, same laboratory test bench is used motor, same set up I have used to detect the single
as shown in Figure 6. It consists of three phase phasing fault. Single phasing is created by connecting
induction motor, current transformer, NI data a knife switch in B phase. Here one phase is opened
acquisition card 9203 and Personnel Computer with to create single phasing.
software Lab VIEW 2011.The rated data of the tested
three-phase squirrel cage induction machine were: 1 Results and Discussion
hp,415 v,1.6Amp. Data under healthy condition of The experiments have been performed to detect the
the motor were acquired at first. Then different faults unbalanced supply voltage faults in three phase
were then created in the motor and data was obtained induction motor using Lab VIEW software. Different
for the faulty conditions. Unbalance in the supply unbalanced condition. The results obtained from
voltage and single phasing are the two types of these experiments are given below.
external faults of an induction motor that are Balanced Condition
275
International Journal of Advanced Computer Research (ISSN (print): 2249-7277 ISSN (online): 2277-7970)
Volume-3 Number-4 Issue-13 December-2013

First of all I create a balanced condition; here I create


a balanced condition by setting same voltage of each
phase. Means this is nothing but a healthy condition
of motor. After that I take FFT of that balanced
condition to extract the features which are sensitive
to the presence of fault.Fig.7and8.

Unbalanced Condition
Here I create a unbalanced condition by setting
different voltage of each phase. Means this is nothing Figure 9: FFT (Unbalanced Supply)
but a faulty condition of motor. After that I take FFT
of that unbalanced condition to extract the features
which are sensitive to the presence of fault. (Fig.9
FFT) Single phase under- voltage unbalance
condition this is my analysis on the unbalance
condition, First of all I analysis on the one phase i.e.
Blue phase. Here I take lower voltage of blue phase
which is 40 volt and from this condition I concluded
that this is nothing but a single phase under
unbalance condition. In this condition the voltage in
that particular phase is lower than the other two Figure 10: Single Phase under voltage unbalance
phases, in this condition that two voltages try to
balance the condition. Fig10.

Two phases under- voltage unbalance condition


In the next step I increase the voltage of that
particular phase at 60 volt after that I still increases
up to 80 volt, and in this condition I concluded that
this is the two phase under voltage unbalance
condition where two phases have the higher voltage
drop than the third one.(Fig.11)
Figure 11: Two Phases under Voltage Unbalance

Figure 7: Time Domain Signal (Balanced Supply) Figure 12: Three Phases under Voltage Unbalance

Figure 13: FFT of single phasing The Voltage


unbalance percentage
Figure 8: FFT (Balanced Supply)
276
International Journal of Advanced Computer Research (ISSN (print): 2249-7277 ISSN (online): 2277-7970)
Volume-3 Number-4 Issue-13 December-2013

help of Electric Discharge Machine (EDM) and were


installed on motor.

Table 1: Voltage unbalances and Temp. Rise Bearing Specifications: 6203-Z SKF, deep groove
percentage ball bearing
 Numbers of balls: 8
 Pitch diameter: 1.537 inches
 Ball diameter: 0.3126 inches
The relationship of bearing vibration to the stator
current spectra can be determined by remembering
that any air gap eccentricity produces anomalies in
the air gap flux density.
Since ball bearings support the rotors, any bearing
defect will produce a radial motion between the rotor
and stator of the machine. The mechanical
displacement resulting from damaged bearing causes
the machine air gap to vary in a manner that can be
described by a combination of rotating eccentricities
moving in both directions. Due to rotating
eccentricities, the vibrations generate stator currents
at frequencies given by [5, 7].
Figure 14: Experiment set up for bearing fault
Experiment Set Up 2

Vibration Analysis Approach for bearing fault


In this study of bearing fault I have used the vibration Where:
monitoring technique to detect fault frequency. In Ball pass frequency outer ring (fBPFO)
vibration monitoring we use the vibration signal to Ball pass frequency inner ring (fBPFI)
detect fault. When carrying out vibration analysis in n = Number of balls
time domain, some simple quantities can be utilized N = rotational speed in RPM
such as root mean square (RMS), crest factor, d = Ball diameter
kurtosis and other statistical moments. If motor D = Bearing pitch diameter
bearing are faulty it affect vibration and motor β = Ball contact angle with the race
current signature pattern. First run the motor in
normal condition and acquire vibration signal. Results and Discussion
Acquire vibration signal for faulty motor bearings. 1. Normal Condition
Apply advance signal processing techniques (FFT) to In first stage I have taken the readings of healthy
extract features from the acquired vibration signal. motor because it is necessary to take these readings
for comparison point of view.Fig.15 and 18(FFT).
Bearing Fault Analysis
In experiment set up I have used 0.5 hp induction 2. Inner Race Fault Condition
motor, a piezoelectric accelerometer, a 16 channel This is the inner race fault condition. This experiment
DAT recorder and was post processed in the was performed on the no load condition. The inner
MATLAB environment. All data files are in Mat lab race fault in bearing was made by drilling a hole of
(*.mat).The bearings of Induction Motor are single 2mm diameter in its inner race. It is observed from
row, deep groove ball bearing and type 6203-Z. Each the power spectrums of motor that fault frequencies
bearing has eight balls. Experiments were conducted are not clearly visible at no-load condition because
on three bearings: one of these is undamaged while their magnitude is less. The power spectrum of faulty
two bearing were damaged. One bearing is drilled motor with 2mm hole in inner race of bearing under
through outer race with „hole diameters‟ of 2 mm no load condition is shown in Figures 16 and 19
while another one bearing drilled through inner race (FFT).
with „hole diameter‟ of 2 mm as illustrated in Figures
3 and 4. Bearings of type 6203 -Z were drilled with
277
International Journal of Advanced Computer Research (ISSN (print): 2249-7277 ISSN (online): 2277-7970)
Volume-3 Number-4 Issue-13 December-2013

3. Outer Race Fault Condition


In the next step I have created an outer race fault by
drilling a hole of 2mm diameter in its outer race. The
power spectrum of faulty motor with 2mm hole in
outer race of bearing under no load condition is
shown in Fig.17 and 20(FFT).

Table 2: Spectrum Analysis for Outer and Inner


Race

Load 0 Fig 18: Normal Condition FFT spectrum


FBPFO(Hz) 107.32
FBPFI(Hz) 115.40

Time domain signal

Fig 19: Inner Race Fault Condition FFT spectrum

Fig 15: Power Spectrum of Normal Condition

Fig 20: Outer Race Fault Condition FFT spectrum

Fig 16: Power Spectrum of Inner Race Fault 4. Conclusion and Future Work
Condition
The aim of this work is to advance the field of
condition monitoring and fault diagnosis in induction
motor operating in variety of operating conditions.
With the help of condition monitoring we can easily
avoid the critical emergency shutdown as well as
reduce the maintenance costs of motor other faults

 FBPFO and FBPFI are indirectly depending


on the loads so as the load increases the fault
Fig 17: Power Spectrum of Outer Race Fault
frequencies decreases.
Condition  With help of Unbalance condition, I got
different situation like Single phase under
Frequency Signal (FFT)
278
International Journal of Advanced Computer Research (ISSN (print): 2249-7277 ISSN (online): 2277-7970)
Volume-3 Number-4 Issue-13 December-2013

voltage, Double phase under voltage and mechanical analysis." Applied Superconductivity,
three phases under voltage. IEEE Transactions on 12, no. 1 (2002): 1511-
1514.
For large size motors, new challenges may exist for [6] Intesar Ahmed “Detection of eccentricity faults in
machine using frequency spectrum technique”,
current based fault detection. Therefore proposed International Journal of computer and electrical
techniques may be applied for the large size motors. engineering, vol.3.No.1, February, 2011.
Additional work is needed to investigate the [7] Neelam Mehla, Ratna Dahiya, “An Approach of
applicability of other signal processing tool in Condition Monitoring of Induction Motor Using
characterizing the fault signature. MCSA”.International Journal of system
applications, Engineering and Development.
Acknowledgment [8] Mehala, Neelam. "Condition monitoring and fault
diagnosis of induction motor using motor current
signature analysis." PhD diss., National Institute
I express my sincere thanks to my mentor, of Technology Kurukshetra, INDIA, 2010.
Dr.C.Y.Patil for his expert guidance, encouragement [9] Benbouzid, M. E. H., “A review of induction
and motivation during the development of this motors signature analysis as a medium for faults
project, without which this project would not have detection”, IEEE Transactions on Industrial
developed to its present form. Electronics, Vol. 47, October, No.5,pp. 984-
I am thankful to Dr. S. L. Patil, Mr.Ugale for their 993,2000.
support and help. I am also very thankful to my [10] M. E. H. Benbouzidi , M. Viera, and C. Theys,
friend Ms.Pranali Narwade, Mr.Chaitannya Supe, “Induction motors faults detection and
localization using stator current advanced signal
Ms.Sneha Bramhankar and Mr.Mayuresh Chavan and processing techniques,”IEEE Transactions on
every person of the department for helping me to Power Electronics, Vol. 14, No. 1, pp14-22, Jan.
complete my project. Last but not the least; I am 1999.
grateful for the endurance from my family who
silently supported me. My project has not been Sumit Narwade was born in Dhulia,
possible without them. India, in1986. He is a PG research
student at Department of
References Instrumentation and Control, College of
Engineering , Pune.He received B.E
degree in Instrumentation from the
[1] Biju K, Jagadanand G (Member, IEEE), Saly
North Maharashtra University in 2011
George, “Fault detection of induction motor
respectively.
using Energy and Wavelets”,Proc. of Int. Conf.
on Control, Communication and Power
Author‟s Photo Prasad Kulkarni was born in Pune,
Engineering 2010.
India 1989. He is pursuing his M.tech
[2] Norman Mariun, Mohammad Rezazadeh
degree from College of Engineering,
Mehrjou,”An Experimental Study of Induction
Pune. He received his B.E degree in
Motor Current Signature Analysis Techniques for
Instrumentation and Control from VIT,
Incipient Broken Rotor Bar Detection”,
Pune in 2011. His interest areas are
International Conference on Power Engineering,
Lab-View and FPGA.
Energy and Electrical Drives, Spain 2011.
[3] Legowski, Stanislaw F., A. H. M. Sadrul Ula, and
Author‟s Photo Dr. C.Y. Patil is a Professor at
Andrzej M. Trzynadlowski. "Instantaneous
Department of Instrumentation and
power as a medium for the signature analysis of
Control, College Of Engineering, Pune.
induction motors." Industry Applications, IEEE
He has a teaching experience of 22
Transactions on 32, no. 4 (1996): 904-909.
years. A PhD in Instrumentation, he has
[4] S. J. Oviedo,J.E. Quiroga,C. Borrás,
6 international journal, 3 national
“Experimental Evaluation of Motor Current
journal, 3 international conferences and
Signature and Vibration Analysis for Rotor
3 national conference papers to his name. He is a life
Broken Bars Detection in an Induction Motor”, Author‟sofPhoto
member Indian Society of Technical Education (ISTE),
International Conference on Power Engineering,
Instrument Society of India (ISOI) and Biomedical Society
Energy and Electrical Drives, Spain 2011.
of India. Currently working on the above research, he has
[5] Lee, Joon-Ho, Young-Hwan Lee, Dong-Hun
previously worked on Design and development of Wireless
Kim, Ki-Sik Lee, and Il-Han Park. "Dynamic
Sensor Network for Greenhouse Agriculture and Fuzzy
vibration analysis of switched reluctance motor
based energy efficient &environment friendly Air
using magnetic charge force density and
conditioner.
279

You might also like