See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/308803405

Bearing condition monitoring using vibration signature

Conference Paper · February 2012

CITATIONS READS

6 89

4 authors, including:

Gopinath Gurunathan

3 PUBLICATIONS   6 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Development of reaction arm for safety critical application View project

Process Improvement & Proving of Heat Treatment Furnace by process capability and CPK value View project

All content following this page was uploaded by Gopinath Gurunathan on 04 October 2016.

The user has requested enhancement of the downloaded file.

 Bearing condition monitoring using vibration signature
G.Gopinath*1, K.Ganesan2, M.Senthil Kumar3 S.Senthil Kumar4
1, 2,3Department
of Mechanical Engineering, P S G College of Technology, Coimbatore-641004, India
4
Aeronautical department Karpagam Institute of Technology, India
* Corresponding Author Email: [email protected]
Abstract— Mechanical vibration is dynamic phenomena, i.e. which have been used in industry so far in order to reduce the
their intensity varies with time. Most failures of rotating cost, can be grouped into categories, which are worth
machinery are rooting in the damage of rolling element mentioning. Different definitions and grouping can be found
bearings. The widespread applications of rolling element in different resources, but maintenance methods are here
bearings in both industry and commercial life require divided into three: run-to-failure, preventive and predictive
advanced technologies to efficiently and effectively monitor maintenance.
their health status. In practice, it is very difficult to avoid 2. Predictive maintenance
vibration during machine running conditions. Therefore, to Predictive maintenance technique (fig 1) helps to determine
know the level of damage on bearing due to machine running the condition of in-service equipment in order to predict when
conditions, vibration measurement should be carried out by maintenance should be performed. This approach offers cost
suitable sensors/transducers. Many problems arising in motor savings over routine or time-based preventive maintenance
operations are linked to bearing faults. because tasks are performed only when warranted. The main
In this paper, bearing vibration frequency features are value of predicted maintenance is to allow convenient
discussed for motor bearing fault diagnosis. This paper then scheduling of corrective maintenance, and to prevent
presents an approach for motor rolling bearing fault diagnosis unexpected equipment failures. The key is "the right
using neural networks and time/frequency-domain bearing information in the right time".
vibration analysis. Vibration simulation is used to assist in the
design of various motor rolling bearing fault diagnosis
strategies.Record the vibration spectrum, specify the peaks
corresponds to the bearing components, Record each
component peak and frequency. By using the software and the
standard limits, determine the trend of each peak. Determine
the bearing state (good –need service –need change).

Keywords ; Bearing vibration, fault diagnosis, frequency

domain, time domain, velocity RMS value.
Fig 1 Predictive maintenance
1. INTRODUCTION

The problems encountered in rotating machineries are 3. Bearing Condition monitoring

of great concern to designers and maintenance engineers. It Due to the close relationship between motor system
has been observed by the various researchers on several development and bearing assembly performance, it is difficult
occasions that rotating machinery stability conditions can to imagine the progress of modern rotating machinery without
change the operating conditions between the driver and the consideration of the wide application of bearings. In addition,
driven machines. The mechanical coupling is used mainly to the faults arising in motors are often linked with bearing
transmit torque from the driver to the driven machine. The faults. In many instances, the accuracy of the instruments and
two connected machines can go under misalignment, shaft devices used to monitor and control the motor system is
bow, unbalance, etc., where severe misalignment can cause highly dependent on the dynamic performance of bearings.
high vibration to the machinery assembly. Also, condition Bearing vibration can generate noise and degrade the quality
monitoring (CM) equipment is used extensively in the energy, of a product line which is driven by a motor system. Heavy
petrochemical, cement, steel, paper, and pulp industries. In bearing vibration can even cause the entire motor system to
addition, these industries are constantly under economic function incorrectly, resulting in downtime for the system and
pressure to reduce costs while increasing service and economic loss to the customer. Proper monitoring of bearing
productivity. Hence, CM is increasingly becoming an vibration levels in a motor system is highly cost effective in
important asset management tool. Consequently, end users are minimizing maintenance downtime both by providing advance
looking to online, continuous CM systems to replace fixed- warning and lead time to prepare appropriate corrective
interval scheduled maintenance practices. . The methods, actions, and by ensuring that the system does not deteriorate

PSG College of Technology

to a condition where emergency action is required. Thus, it is type of energy, usually an electric current or voltage.
important to include bearing vibration diagnosis into the Commonly used transducers are velocity pickups,
scheme of motor system fault diagnosis. accelerometers and Eddy current or proximity probes. Each
Signals from vibration sensors are usually measured and type of transducer has distinct advantages for certain
compared with reference measurements in order to interpret applications.
bearing conditions. The methods used to analyze these signals 4.1Accelerometer:
include probabilistic analysis, frequency analysis [1], time- Accelerometers are inertial measurement devices
domain analysis [4] and finite-element analysis . Among these that convert mechanical motion into a voltage signal. The
methods, the frequency analysis approach is the most popular signal is proportional to the vibration’s acceleration using the
one. This popularity is most probably due to the availability of piezoelectric principle. Inertial measurement devices measure
Fourier transform technique, as characteristics of vibration motion relative to a mass. This follows Newton’s third law of
signals are more easily noticed in the frequency domain rather motion: body acting on another will result in an equal and
than in the time domain. The frequency analysis technique opposite reaction on the first.
involves frequency analysis of the vibration signal and further Accelerometers are the most popular
processing of the resulting spectrum to obtain clearly defined transducers used for rotating machinery applications. They are
diagnosis information. rugged, compact, lightweight transducers with a wide
frequency response range. Accelerometers are extensively
used in many condition-monitoring applications. Components
such as rolling element bearings or gear sets generate high
vibration frequencies when defective. Machines with these
components should be monitored with accelerometers
(Accelerometer NA 5431 fig no 6, 7). The installation of an
accelerometer must carefully be considered for an accurate
and reliable measurement. Accelerometers are designed for
mounting on machine cases. This can provide continuous or
periodic sensing of absolute case motion (vibration relative to
Fig. 2. General flow of signals in a typical motor bearing fault
free space) in terms of acceleration.
detection process.
5. THEORETICAL CALCULATIONS
4 Data acquisition
5.1 DETERMINATION OF SHAFT DIAMETER
With data acquisition, we take the first steps into the domain
of practical vibration analysis. It includes the following main 1. Motor Power (P) = 1 Hp
tasks: 12
• Collection of machinery vibration P = 62.16 W
• Conversion of the vibration signal to an electrical signal 2. Speed (N) = 5000 rpm (Motor)
• Transformation of the electrical signal to its components Shaft material – Mild steel
• Providing information and documentation related to 3. Length (L) = 1000 mm
vibration data. 4. Disc weight (W) = 1 kg
The above entails the entire hardware of the vibration analysis W = 10 N
system fig 3. 5. Shear stress (Ss) = 45 N/mm2
It includes transducers, electronic instruments that store and 6. Bending stress (Sb) = 70 N/mm2
analyse data, the software that assist in vibration analysis, Minor shock loads
record keeping and documentation. kb = 2 kt = 1.5 (From PSG design data book)
kb = combined shock and factor applied to bending moment
kt = combined shock and factor applied to twisting moment
7. Torque (T) = 60 P
Data 2 N
acquisition
system T = 60  62.16
2   5000
T = 118.9 Nmm
8. Moment (M) = WL
DAQ 9234 4
M = 10  1000
4
M = 2500 Nmm
9. Combined Moment (Me) = 1 
Fig 3 Data acquisition system 2 
2  2500  2  2500) 2

 (1.5  118.9 
2


1
To measure machinery or structural vibration, a transducer d >  32  M b  3

or a vibration pickup is used. A transducer is a device that  

  S b 
converts one type of energy, such as vibration, into a different

PSG College of Technology

 1

d >  32  5001.58  3
  70 
 
d > 9mm
From that calculation diameter is greater than 9mm but
standard size of bearing is 12mm. so selected as shaft
diameter.
5.2 CALCULATION FOR NATURAL FREQUENCY Fig .4 .Design model Fig.5 .Fabricated model
AND AMPLITUDES 6 .Bearing vibration signature
As per the design model fabrication of test rig is
ready to measure the vibration signals. In the test rig (Fig 7)
accelerometer is placed in the bearing mounting head and the
frequency response, velocity rms value, displacement value
has been taken. In the vibration signals has been taken from
the coupling end-B2 (fig 6) and non-drive end-B3 (fig 7)

Young's modulus (E) = 2.1 x 105 N/mm2 (for mild steel)

Rotor mass (m) = 1 kg
Shaft diameter (d) = 12 mm
Center of gravity of the shaft (g) = 0.02 mm
Speed (N) = 5000 rpm
Bearing span (l) = 500mm
Stiffness (k) =  48  E  I  Fig 6 Fig 7 Accelerometer NA 5431
 l3 
  4 
Accelerometer Position
=  48  (2.1  10 )  ( 64  12 ) 
5

 
 500 3 
 

K = 82.081N/mm
Natural frequency (ωn) = K
M
= 82081
1
= 286 rad/sec = 45.54 Hz Fig 8 - LABVIEW Program
Angular velocity ω = 2   N
60
= 2   5000
60
= 524.38 rad/sec
This implies that ω > ωn, applied frequency is greater than the
natural frequency so shaft is safe condition.
5.3 SPECIFICATIONS
Motor
Motor Torque (T) = 118.9 Nmm
Motor Power (P) = 1/12 HP Fig 9 - Data acquisition system with Test Rig
Natural Frequency (ωn) = 46HZ
Shaft Frequency spectrum has been taken in coupling and non-drive
Diameter (d) = 12 mm end bearing at three various speeds (750 RPM, 1000 RPM and
Density (ρ) = 7800 kg/mm3 1250 RPM). The frequency spectrum for 0 mm coupling,
Modulus of Elasticity (E) = 2.1 x 105 N/mm2 mass and unbalance placed at middle from coupling end and
Stiffness (k) = 80.281 N/mm non-drive end. Frequency spectrum of coupling end bearing is
Disc Mass (m) =1kg shown in Fig 10,11,12,13.

PSG College of Technology

 Non drive End (unbalanced mass) bearing

2NUT - 1000 RPM 3NUT - 1000 RPM

Fig 10-750 RPM Fig 11 - 1000 RPM

4NUT - 1000 RPM

Fig 18

B2 - Displacement spectrum with Mass at Middle

With reference to the displacement plot for the configuration
Fig 12 - 1250 RPM Fig 13 - Static Position of location of the mass at middle. The frequency spectrum has
been taken from coupling end. Maximum value of
Frequency spectrum of Non-drive end bearing (0 mm displacement has been plotted at various conditions Table no
coupling) at various speeds (750, 1000, 1250 RPM) is shown 1 and also tabulated.
in Fig 14,15,16. Table 1
S.No Speed Condition Displacement Displacement Displacement
(RPM) 1.5 mm 0 mm Coupling 1 mm Coupling
Coupling

With mass 886gm 0.00006 0.00005 0.00008

1 750 With unbalance of 32gm (1) 0.00007 0.00008 0.0001

With unbalance of 64gm (2) 0.00011 0.0001 0.00013

With unbalance of 96gm (3) 0.00009 0.0001 0.00015

Fig 14 – 750 RPM Fig 15 – 1000 RPM With balance of 128gm (4) 0.00006 0.00008 0.00008

With balance of 64gm (1&3) 0.00007 0.00008 0.00005

With mass 886 gm 0.00008 0.00006 0.0001
2 1000
With unbalance of 32gm (1) 0.00014 0.00015 0.00013
With unbalance of 64gm (2) 0.00013 0.0002 0.0002
With unbalance of 96gm (3) 0.00012 0.00012 0.00015

With balance of 128gm (4) 0.00042 0.00016 0.0001

With balance of 64gm (1&3) 0.0001 0.00007 0.0001

With mass 886 gm 0.00024 0.00004 0.00035

3 1250
With unbalance of 32gm (1) 0.00017 0.00018 0.0002
Fig 16 – 1250 RPM
With unbalance of 64gm (2) 0.00028 0.0006 0.0004

For the unbalanced mass coupling end (Fig 6)and non-drive With unbalance of 96gm (3) 0.00032 0.00015 0.00042
end bearing reading (Fig 7) has been taken as shown in table I. With balance of 128gm (4) 0.00015 0.0002 0.0004
For various unbalanced mass (Fig 17) reading taken out and
With balance of 64gm (1&3) 0.00018 0.00015 0.00012
listed.
With Unbalanced Mass

Unbalanced mass 0.032 Kg

Mass 0.886 Kg

With Unbalanced Mass – 2Nut

From the table it has been inferred as the unbalance varies
With Unbalanced Mass – 3Nut the value of displacement predominantly increases.
Unbalanced mass 0.064 Kg
Unbalanced mass 0.096 Kg

Mass 0.886 Kg B3 - Displacement spectrum with Mass at Middle

With reference to the displacement plot for the configuration
of location of the mass at middle. The frequency spectrum has
been taken from Non-drive end. Maximum value of
Fig 17 Unbalanced Mass displacement has been plotted at various conditions Table no
2 and also tabulated.

PSG College of Technology

 Table 4
Table 2 S.No Speed Condition Velocity(RMS) Velocity Velocity(RM
(RPM) 0 mm coupling (RMS) S)
1 mm 1.5mm
S.No Speed Condition Displaceme Displacement Displaceme Coupling Coupling
(RPM) nt 1mm Coupling nt
0mm 1.5 mm With mass 886gm 0.015 0.009 0.014
Coupling Coupling 1 750
With unbalance of 32gm (1) 0.016 0.0175 0.025
With mass 886gm 0.00008 0.00004 0.0001
1 750 With unbalance of 64gm (2) 0.03 0.03 0.03
With unbalance of 32gm (1) 0.00005 0.00007 0.0002
With unbalance of 64gm (2) 0.0002 0.0002 0.00025 With unbalance of 96gm (3) 0.018 0.03 0.028
With unbalance of 96gm (3) 0.00006 0.00012 0.0002 With balance of 128gm (4) 0.012 0.0175 0.02
With balance of 128gm (4) 0.00006 0.0001 0.00015
With balance of 64gm (1&3) 0.0001 0.0001 0.00013 With balance of 64gm (1&3) 0.02 0.015 0.0225
With mass 886 gm 0.00007 0.0001 0.00014 With mass 886 gm 0.016 0.0125 0.025
2 1000 With unbalance of 32gm (1) 0.000125 0.00015 0.0003 2 1000
With unbalance of 64gm (2) 0.00015 0.00022 0.00035 With unbalance of 32gm (1) 0.025 0.024 0.05
With unbalance of 96gm (3) 0.00012 0.00014 0.0003
With balance of 128gm (4) 0.00007 0.00015 0.00016 With unbalance of 64gm (2) 0.02 0.025 0.07
With balance of 64gm (1&3) 0.00012 0.0002 0.00015
With unbalance of 96gm (3) 0.025 0.036 0.06
With mass 886 gm 0.0001 0.0001 0.0004
3 1250 With unbalance of 32gm (1) 0.000125 0.00012 0.0004 With balance of 128gm (4) 0.02 0.025 0.035
With unbalance of 64gm (2) 0.0007 0.0006 0.00042
With balance of 64gm (1&3) 0.03 0.035 0.03
With unbalance of 96gm (3) 0.00015 0.0004 0.00045
With balance of 128gm (4) 0.00025 0.0004 0.0003 With mass 886 gm 0.025 0.022 0.06
With balance of 64gm (1&3) 0.00015 0.00022 0.0003 3 1250
With unbalance of 32gm (1) 0.022 0.04 0.07

With unbalance of 64gm (2) 0.12 0.1 0.08

From the table it has been inferred as the unbalance varies the With unbalance of 96gm (3) 0.03 0.05 0.07

value of displacement predominantly increases. With balance of 128gm (4) 0.04 0.06 0.06

7.2 Velocity RMS value With balance of 64gm (1&3) 0.032 0.04 0.05

B2 – Velocity (RMS) with Mass at Middle

With reference to the velocity plot for the configuration of 7. Results and Discussion
location of the mass at middle. The RMS (Velocity) has been 7.1 Displacement spectrum
taken from coupling end. Maximum value of velocity has The displacement plot for the various unbalance &
been plotted at various conditions table no 3 and also configuration of location of the mass the spectrum is plotted &
tabulated. tabulated. The frequency spectrum has been taken from
Table 3 coupling end and non drive end also. Maximum value of
S.No Speed
(RPM)
Condition Velocity(RMS)
0 mm
Velocity
(RMS)
Velocity(RMS)
1.5 mm
displacement has been tabulated in the Table no 1,2.From the
coupling 1mm
Coupling
Coupling values as the unbalance varies the value of displacement
With mass 886gm 0.009 0.02 0.01 predominantly increases.
1 750
With unbalance of 32gm (1) 0.012 0.02 0.008
With unbalance of 64gm (2) 0.012 0.02 0.016
With unbalance of 96gm (3)
With balance of 128gm (4)
0.013
0.012
0.022
0.016
0.016
0.009
7.2 Velocity RMS value
With balance of 64gm (1&3) 0.015 0.012 0.012
With mass 886 gm 0.01 0.02 0.012 The RMS velocity is measured for the various configuration
2 1000
With unbalance of 32gm (1) 0.025 0.018 0.016 and location of the mass at middle. The RMS (Velocity) has
With unbalance of 64gm (2) 0.023 0.023 0.02
With unbalance of 96gm (3) 0.015 0.022 0.015
been taken from coupling end and non drive end. Maximum
With balance of 128gm (4) 0.025 0.0175 0.009 value of velocity has been tabulated in Table no 3,4.
With balance of 64gm (1&3) 0.016 0.02 0.008
With mass 886 gm 0.018 0.05 0.04
3 1250 With unbalance of 32gm (1) 0.028 0.025 0.035 7.3 Acceleration Spectrum
With unbalance of 64gm (2) 0.05 0.08 0.03
With unbalance of 96gm (3) 0.018 0.03 0.038 The various acceleration spectrums are illustrated in the Figs
With balance of 128gm (4) 0.03 0.09 0.05 10 – 16 for various speeds. Figs 10 -13 illustrates the spectrum
With balance of 64gm (1&3) 0.035 0.028 0.04
at the coupling end bearing location. Figs 14-16 illustrate the
spectrum at the non drive end location.
B3 – Velocity (RMS) with Mass at Middle
8. Conclusion. By using the vibration severity chart refers the
With reference to the velocity plot for the configuration of
coupling end and non-drive end bearing vibration velocity rms
location of the mass at middle. The RMS (Velocity) has been
value indicate the good, satisfactory, unsatisfactory and
taken from Non-drive end. Maximum value of velocity has
unacceptable region clearly.
been plotted at various conditions table no 4 and also
tabulated.
9. Future work; Evaluate the health condition of the bearing,
Develop a neural-network-based model for predicting bearing
failure. Model the vibration signals using artificial neural
networks to estimate failure times of a degrading bearing in
order to make better decisions regarding scheduling of
maintenance activities and replacement strategies.

PSG College of Technology

10. References

1. Sebastian Villwock and Henning Zoubek Student Member,

IEEE, Mario Pacas Senior Member IEEE, University of
Siegen ‘’ Rolling Bearing Condition Monitoring Based on
Frequency Response Analysis’’ Institute of Power Electronics
and Electrical Drives Holderlinstrasse 3, IEEE 2007 vol no
4.1.0

2. Wei Zhou, Member, IEEE, Thomas G. Habetler, Fellow,

IEEE, and Ronald G. Harley, Fellow, IEEE ‘’Bearing
Condition Monitoring Methods for Electric Machines: A
General Review ‘’ 1-4244-1 062-2/07/ ©2007 IEEE

3 R. Keith Mobley Boston Oxford “Root Cause Failure

Analysis, & Vibration Fundamentals “A member of the Reed
Elsevier group 1999 by Butterworth–Heinemann.
Vibration fundamentals book part2.

4. T. Hoshi Technical Research, Toyohashi, Japan ‘’ Damage

Monitoring of Ball Bearing’’ 2005.

5. Runqing Huanga,_, Lifeng Xia, Xinglin Lib, C. Richard

Liuc, Hai Qiud, Jay Lee a Department of Industrial
Engineering & Management, Shanghai Jiao Tong University
‘’Residual life predictions for ball bearings based on self-
organizing map and back propagation neural network
methods’’ Received 23 January 2005;accepted 21 November
2005.

6. Changting Wang, Student Member, IEEE, and Robert X.

Gao, Senior Member, ‘’A Virtual Instrumentation System for
Integrated Bearing Condition Monitoring‘’ IEEE
TRANSACTIONS ON INSTRUMENTATION AND
MEASUREMENT, VOL. 49, NO. 2, APRIL 2000.

7. N. Tandon_, G.S. Yadava, K.M. Ramakrishna ITMME

Centre, ‘’A comparison of some condition monitoring
techniques for the detection of defect in induction motor ball
bearings’’ Indian Institute of Technology, Hauz Khas, New
Delhi 110016, India accepted 5 August 2005

8. Zeki Kiral∗ , Hira Karagu¨lle Department of Mechanical

Engineering, Dokuz Eylul University, Engineering Faculty,
35100 Izmir, Turkey ‘’Simulation and analysis of vibration
signals generated by rolling element bearing with defects’’
accepted 13 January 2003 .

9. Jason R. Stack, Member, IEEE, Thomas G. Habetler,

Fellow, IEEE, and Ronald G. Harley, Fellow, IEEE ‘’
Experimentally Generating Faults in Rolling Element
Bearings Via Shaft Current’’ IEEE TRANSACTIONS ON
INDUSTRY APPLICATIONS, VOL. 41, NO. 1,
JANUARY/FEBRUARY 2005.

PSG College of Technology

View publication stats