International Journal of Latest Trends in Engineering and Technology (IJLTET)

Fault Diagnosis of Gear by Vibration Analysis

Aditendra Jaiswal
M Tech Final year Mechanical Department
Priyadarshini College of Engineering Nagpur, India

Dr. K.S. Zakiuddin

HOD and Professor Mechanical Department
Priyadarshini College of Engineering Nagpur, India

V.V. Shukla
Assistant Professor Mechanical Department
Priyadarshini College of Engineering Nagpur, India

Abstract - Unexpected production delays cause substantial costs in many branches of industry. A machine consists of
many potentially defective elements, such as gears, bearings, shafts etc. During operation, each element contributes to
the overall vibration of the machine. Each specific machine element has its own vibration signature. The purpose of
machine vibration monitoring is to use the modified vibration signature to detect, localize, diagnose and
prognosticate the defective gear so that relevant maintenance can be planned. When a defect e.g., a fatigue crack,
appears in an element, e.g., a gear, its signature is modified. The dynamic response of gears is a matter of concern
because of noise generation and dynamic loads. The present study is focused on the different modes failures observed
in gears which are detected by using vibration based monitoring technique Fast Fourier Transform. A test rig is
prepared to carry out the analysis by simulating faults in the gear and generating spectrums of specific faults which
are compared with the spectrum of a new gear indicating severity and specificity of fault. The experimental setup has
been designed to facilitate generation of spectrum for different pairs of new and defective gears. The modified
spectrum from combination of new and defective gears can be attributed to variation from the defective gear, thus
identifying the defective gear in the pair. In this paper the empirical data generated by the experimental setup for
both new and defective gear for a gear pair used in an automotive transmission is analysed and discussed.

Keywords - Vibration based condition monitoring, Fast Fourier Transform, Modes of Gear failure.

I. INTRODUCTION

Transmission is an important subsystem of engineering systems such as power, automotive gear trains,
manufacturing machinery etc. Effectiveness of such systems is dependent on their continued availability through
reliability of subsystems. Maintenance of equipment and machines has graduated from breakdown to condition
based predictive maintenance. Right information at the right time is crucial aspect of condition-based
maintenance strategy in industries. Condition-based maintenance (CBM) consists of real time monitoring of
condition of machine systems, identifying faults before catastrophic breakdown occurs. Condition monitoring
based maintenance improves the availability by predicting faults facilitating scheduling and planning of
maintenance activities, defective components like gears can be withdrawn from service before their condition
impairs performance. Gears are critical part of power transmission system. Numerous condition monitoring
(CM) and diagnostics methodologies are utilized to identify the machine faults to take corrective action. A
number of condition monitoring techniques have emerged such as wear debris analysis, visual inspection
through strobe lights, vibration signature analysis, lubricant signature analysis, noise signature analysis, and
temperature monitoring, with the use of appropriate sensors /transducers, different signal conditioning, and
analyzing hardware /software. The use of vibration analysis as one of the fundamental tools for condition
monitoring has been developed extensively over last six decades. Failure mode analysis of gears obtained from
published industry data shows the following failures Crack, wear, pitting, corrosion, scoring, plastic flow, tooth
breakage. Failure mode analysis of gears obtained from published industry data shows the following failures
Crack , wear, pitting, corrosion, scoring, plastic flow, tooth breakage . The paper attempts to collect vibration
spectrum of five different simulated gear defects through design of an experimental setup for comparison with
the reference signals. A reusable set up was therefore conceptualized and designed for multiple gear fault

Vol. 3 Issue 1 September 2013 26 ISSN: 2278-621X

 International Journal of Latest Trends in Engineering and Technology (IJLTET)

simulation. To control speed of the motor it was connected to variable frequency drive (VFD). Accelerometer
probes to measure the time domain response were attached magnetically to measure the horizontal and vertical
vibration signals generated on the bearing housing of pinion shaft. Accelerometers were used to detect vibration
and transmitting the signal to the FFT equipment. Time domain signals of vibration of good and defective gears
were obtained through the experimental set up and transformed to frequency domain through FFT. The
frequency domain spectrum have been analyzed several patterns have emerged from the analysis of the
spectrums obtained in the experiment.

Methods of Thermal Wear Acoustic Visual Vibration analysis

analysis Debris Emission inspection
analysis
Analysis Strobe light
Spectro analysis
meter &
Oil

Out of balance

Misalignment

Bent shaft

Ball Brg

Journal Brg

Mechanical
Looseness

Mechanical
Rubbing

Noise
Cracking

Leaking valve
seats

Gear Damage

Table 1 Applicability of diagnostic techniques to machine fault.

II. SELECTION AND SIMULATION OF GEAR FAULTS

WEAR- A surface phenomena in which layers of metals are removed or worn away more or less
uniformly from contacting the surfaces of gear teeth
.

Vol. 3 Issue 1 September 2013 27 ISSN: 2278-621X

 International Journal of Latest Trends in Engineering and Technology (IJLTET)

Wear describes a loss or removal of material of gear flanks. In terms of gear failure, it is more a
deterioration of a gear profile, for instance, a damage of a tooth layer. Adhesive and abrasive wear are
important modes of wear.

PITTING- A surface fatigue failure which occurs when the endurance limit of the material is
exceeded, a failure of this nature depends on surface contact stress and number of stress cycles. The
condition is characterised by small pits 1/64 to 1/32 in dia. For the analysis the gear used had spalling
type of fault which is a sub type of pitting seen in hardened gears. SPALLING is a term used to describe a
large or massive area where surface material has broken away from the tooth. In through-hardened and softer
material, it appears to be a massing of many overlapping or interconnected large pits in one locality. It is
similar to destructive pitting except that pits are usually larger in diameter and quite shallow.

FRACTURE- Failure caused by breakage of whole tooth or substantial portion of tooth this can result from
overload or more commonly by, cyclic stressing of the gear tooth beyond the endurance limit of the
material.

CORROSION- It is a surface phenomena observed on the surface of the gear. In some cases when the
lubricant or other fluid in the system comes in contact with the surface of the gear material it may rust
this is known as corrosion.

x The gears used in this study are spur gears of Kinetic scooters.
x The larger driving gear is connected to motor and has 36 teeth. Two sets of gears used. Material of the
gear is SAE8620.
x Smaller gear is the driven gear having 18 teeth. Five sets of gears used. Material of the gear is
SAE8620.

III. COMPONENTS OF TEST RIG

Figure 1 Flowchart of Experimental setup

x This whole experiment consists of a three phase induction motor and single stage spur gear
arrangement.
x Two jaw flexible coupling is use to connect the motor shaft and the driver shaft
x The driver shaft is shaft 1 on which gear is mounted denoted as (G1) which is coupled with the motor
by a two jaw flexible coupling.
x The present study the motor is operated at 700 rpm. Therefore speed of the first gear is 700.With a
step-up ratio of 1: 2, the speed of the pinion shaft 1400 rpm.
x Parallel shaft is shaft 2 on which pinion a small gear is mounted denoted as (P1). .
x The assembly of the test rig is as per the requirement of the analysis different faults have to be
simulated in gears so that respective vibration spectrums can be generated.

Vol. 3 Issue 1 September 2013 28 ISSN: 2278-621X

 International Journal of Latest Trends in Engineering and Technology (IJLTET)

Figure 2 Gear new condition Figure 3 Pinion new condition Figure 4 Faulted gear

Figure 5 Full view of experimental setup Figure 6 FFT setup with all accessories

Working and Procedure

As power is supplied the gear start to rotate and mesh with each other the speed of the gear can be varied with
the use of Variable frequency drive (VFD). The operational speeds at which the reading has to be taken are 700
rpm and 1400 rpm of shaft 1 as the velocity ratio of gear pair is 2 so the rotational speed of shaft 2 is 1400 rpm
and 2800 rpm respectively. There are two sets of gears G1 one is new and another one is worn out with crack on
tooth. Similarly there are five sets if pinion P1 one is new and the others are with default as mentioned above.
Firstly the new gears are mounted on the shaft and vibration spectrums are generated at 700 and 1400 rpm speed
which are denoted as reference spectrums. These spectrums generated will be used as the reference spectrums.
New pinion is replaced by a faulty pinion in which dominant defect is wear and spectrum is generated. In this
way the procedure is repeated for all the combination of faulty and new gears. The signal is generated by
accelerometer which is processed in the FFT setup to generate vibration spectrums. Time domain and frequency
domain spectrums are generated which are compared to find that due to fault what amplitude changes are
incurred. Accelerometers are placed in horizontal and vertical position of the bearing case to obtain two signals.

Vibration spectrums of new gears are listed below taken at shaft 1 and shaft 2 at rotational speed of 700 rpm
and 1400 rpm respectively.
M/C1 - Project Machine - 2
M/C1 - Project Machine - 2 PCE M2 -G1V Shaft 01 Inboard Vertical
PCE M2 -G1H Shaft 01 Inboard Horizontal 1.8
PK Velocity in mm/Sec

3.5
Route Spectrum
Route Spectrum 1.5
PK Velocity in mm/Sec

15-May-13 17:19:13
3.0 15-May-13 17:19:13 OVERALL= 2.49 V-DG
2.5 OVERALL= 4.58 V-DG 1.2 PK = 2.46
PK = 4.53 LOAD = .0
2.0 LOAD = .0 0.9
RPM = 697. (11.62 Hz)
1.5 RPM = 659. (10.98 Hz)
0.6
1.0
0.3
0.5
0
0
0 20000 40000 60000
0 20000 40000 60000
Frequency in CPM
Frequency in CPM
2.0
3 Route Waveform
Route Waveform 1.5 FAULT
Acceleration in G-s

2 15-May-13 17:19:13
15-May-13 17:19:13
Acceleration in G-s

FAULT 1.0 RMS = .2756

RMS = .3736 ALERT
1 PK(+/-) = 1.25/1.42 0.5 PK(+/-) = .7974/.8655
ALERT
CRESTF= 3.80 CRESTF= 3.14
0 0
-0.5
-1
ALERT -1.0
ALERT
-2 -1.5
FAULT
FAULT
-3 -2.0
Freq: 658.59 Freq: 697.24
0 50 100 150 200 250 300 350 400 0 50 100 150 200 250 300 350 400 Ordr: 1.000
Ordr: 1.000
Time in mSecs Spec: .592 Time in mSecs Spec: .01250

Shaft 1 new gear at 700 horizontal Shaft 1 new gear at 700 vertical

Vol. 3 Issue 1 September 2013 29 ISSN: 2278-621X

 International Journal of Latest Trends in Engineering and Technology (IJLTET)

M/C1 - Project Machine - 2

PCE M2 -G3H Shaft 02 Inboard Horizontal M/C1 - Project Machine - 2
3.5 PCE M2 -G3V Shaft 02 Inboard Vertical
Route Spectrum

PK Velocity in mm/Sec
3.0 1.2

PK Velocity in mm/Sec
15-May-13 17:20:14 Route Spectrum
2.5 OVERALL= 4.81 V-DG 15-May-13 17:20:14
0.9 OVERALL= 2.83 V-DG
PK = 4.70
2.0 LOAD = .0 PK = 2.65
RPM = 663. (11.06 Hz) LOAD = .0
1.5 0.6
RPM = 643. (10.71 Hz)
1.0
0.3
0.5
0
0
0 20000 40000 60000
0 20000 40000 60000
Frequency in CPM
Frequency in CPM
3
Route Waveform 3
Route Waveform
2 15-May-13 17:20:14
Acceleration in G-s

FAULT 2 15-May-13 17:20:14

Acceleration in G-s
RMS = .4958 FAULT
1 RMS = .3762
ALERT PK(+/-) = 1.58/2.02 1 PK(+/-) = 1.36/1.84
ALERT
CRESTF= 4.08 CRESTF= 4.89
0 0
-1 -1
ALERT ALERT
-2 -2
FAULT FAULT
-3 -3
Freq: 663.33 Freq: 642.89
0 50 100 150 200 250 300 350 400 Ordr: 1.000 0 50 100 150 200 250 300 350 400 Ordr: 1.000
Time in mSecs Spec: .678 Time in mSecs Spec: .102

Shaft 2 new pinion at 700 horizontal Shaft 2 new pinion at 700 vertical
M/C1 - Project Machine 2.1 M/C1 - Project Machine 2.1
PCE-2.1 -G1H Shaft 01 Inboard Horizontal PCE-2.1 -G1H Shaft 01 Inboard Horizontal
10 10

PK Velocity in mm/Sec
Route Spectrum Route Spectrum
PK Velocity in mm/Sec

8 15-May-13 17:21:45 8 15-May-13 17:21:45

OVERALL= 11.74 V-DG OVERALL= 11.74 V-DG
PK = 11.58 6 PK = 11.58
6 LOAD = 100.0
LOAD = 100.0
RPM = 1398. (23.31 Hz) RPM = 1398. (23.31 Hz)
4 4

2 2

0 0

0 20000 40000 60000 0 20000 40000 60000

Frequency in CPM Frequency in CPM

3 Route Waveform 3 Route Waveform

15-May-13 17:21:45

Acceleration in G-s
15-May-13 17:21:45
Acceleration in G-s

2 2
FAULT RMS = .8694 FAULT RMS = .8694
1 PK(+/-) = 3.32/4.01 1 PK(+/-) = 3.32/4.01
ALERT ALERT
0 CRESTF= 4.62 0 CRESTF= 4.62
-1 -1
ALERT ALERT
-2 -2
FAULT FAULT
-3 -3
-4 -4
-5 -5
Freq: 1398.4 Freq: 1398.4
0 50 100 150 200 Ordr: 1.000 0 50 100 150 200 Ordr: 1.000
Time in mSecs Spec: 1.830 Time in mSecs Spec: 1.830


Shaft 1 new gear at 1400 horizontal Shaft 1 new gear at 1400 vertical

M/C1 - Project Machine 2.1 M/C1 - Project Machine 2.1

PCE-2.1 -G3H Shaft 02 Inboard Horizontal PCE-2.1 -G3V Shaft 02 Inboard Vertical
10 4.0
PK Velocity in mm/Sec

PK Velocity in mm/Sec

Route Spectrum 3.5 Route Spectrum

8 15-May-13 17:21:01 15-May-13 17:21:01
OVERALL= 13.47 V-DG 3.0 OVERALL= 7.19 V-DG
6 PK = 13.02 2.5 PK = 6.52
LOAD = 100.0 LOAD = 100.0
2.0
RPM = 1399. (23.32 Hz) RPM = 1401. (23.35 Hz)
4 1.5
1.0
2
0.5
0 0
0 20000 40000 60000 0 20000 40000 60000
Frequency in CPM Frequency in CPM
6 6
Route Waveform Route Waveform
15-May-13 17:21:01 15-May-13 17:21:01
Acceleration in G-s

Acceleration in G-s

3 RMS = 1.38 3 RMS = 1.18

FAULT PK(+/-) = 5.39/4.71 FAULT PK(+/-) = 5.15/4.95
ALERT CRESTF= 3.90 ALERT CRESTF= 4.38
0 0
ALERT ALERT
-3 FAULT -3 FAULT

-6 -6
Freq: 1399.2 Freq: 1400.9
0 50 100 150 200 Ordr: 1.000 0 50 100 150 200 Ordr: 1.000
Time in mSecs Spec: 2.642 Time in mSecs Spec: .133


Shaft 2 new pinion at 1400 horizontal Shaft 2 new pinion at 1400 vertical

IV. ANALYSIS OF SPECTRUM

Spectrum is studied by seeing the amplitude of the gear mesh frequency, 1X 2X rpm frequency and the running
speed sidebands of the gear mesh frequencies.
Gear mesh frequency (GMF) - angular speed of shaft* No of teeth on gear
GMF 1 = 1 * GMF
GMF 2 = 2 * GMF
GMF 3 = 3 * GMF
1X rpm - X is the angular speed of shaft
2 X rpm - X is the angular speed of shaft
Crest factor peak value / root mean square value

Vol. 3 Issue 1 September 2013 30 ISSN: 2278-621X

 International Journal of Latest Trends in Engineering and Technology (IJLTET)

Table 2 Amplitude values when the condition of gear is new

New condition of 1x 2x 1GM 2GM OVERALL MAX CF

gear AMP
S 1 700 H 0.7 1.3 0.2 0.05 4.58 2.5 3.80
S 1 700 V 0.2 1.25 0.3 0.05 2.49 1.45 3.14
S 2 700 H 0.7 1.2 0.3 0.05 4.81 2.5 4.08
S 2 700 V 0.7 1.25 0.2 0.1 2.83 0.8 4.83
S 1 1400 H 2 3.5 0.05 NA 11.74 7 4.62
S 1 1400 V 1.2 1 0.1 NA 5.72 2.7 4.93
S 2 1400 H 4 8 0.05 NA 13.47 7.5 3.90
S 2 1400 V 1 2.8 0.1 NA 7.19 2.8 4.38

From the vibration spectrums generated of new gear, values of the amplitude are noted which are then compared
to the amplitude values of the faulty gears to find the deviation in condition and observe change in vibration
spectrum.

V. CONCLUSION
x It is observed from the study of various results. There are changes in the vibration spectrum of defective
gears as compared to spectrum of reference gear (new) gear.
The deviation in spectrum of different defective gears is as follows:
x Wear - It is observed by analyzing the spectrums that in most cases at 1X RPM frequency amplitude
decreases at 2X RPM amplitude increases or decreases mostly no change or decrease in amplitude is
observed at GMF 1 and no change is observed at GMF 2. Change in side band frequencies is observed.
x Pitting- It is observed by analyzing the spectrums that in most cases at 1X RPM frequency amplitude
decreases at 2X RPM amplitude decreases, increase in amplitude is observed at GMF 1 and no change or
decrease is observed at GMF 2. Change in side band frequencies is observed.
x Corrosion- It is observed by analyzing the spectrums that in most cases at 1X RPM frequency amplitude
decreases at 2X RPM amplitude increases mostly no change or increase in amplitude is observed at GMF 1
and amplitude decreases or no change is observed at GMF 2. Change in side band frequencies is observed.
x Crack Gear- It is observed by analyzing the spectrums that in most cases at 1X RPM frequency amplitude
increases at 2X RPM amplitude increases, mostly no change or decrease in amplitude is observed at GMF 1
and no change or increase and decrease in amplitude is observed at GMF 2. Change in side band
frequencies is observed.
x Crack Pinion- It is observed by analyzing the spectrums that in most cases at 1X RPM frequency amplitude
increases at 2X RPM amplitude increase or decrease in amplitude, increase in amplitude is observed at
GMF 1 and amplitude decreases or no change is observed at GMF 2. Change in side band frequencies is
observed.

REFERENCES

[1] Pratesh Jayaswal, A.K.Wadhwani and K. B.Mulchandani Machine Fault Signature Analysis.
[2] C.R.Brodersen T.E.Hancock and D.M.Johnson Integrated plant temperature measurement using heat-sensitive paint and colour image
analysis.
[3] Hasan Ozturk Gearbox health monitoring and fault detection using vibration analysis.
[4] Kobra Heidarbeigi, Hojat Ahmadi, M.Omid Adaptive Vibration Condition Monitoring Techniques for Local Tooth Damage in
Gearbox. .
[5] Abouel-seoud, S., El-morsy, M. and Saad, A. A Laboratory Apparatus for Investigation of Vibration Performance of Wind Turbine
Planetary Gearbox Journal of Recent Research, Vol. 3, Issue 11, 2011.

Vol. 3 Issue 1 September 2013 31 ISSN: 2278-621X

 International Journal of Latest Trends in Engineering and Technology (IJLTET)

[6] Cesar San Martin, Edgar Estupinan, Daniel San Martin. A Methodology for the Detection and Diagnostic of Localized Faults in
Gears and Rolling Bearings systems.
[7] J. Lin and M. J. Zuo Gearbox Fault Diagnosis using Adaptive Wavelet Filter
[8] Lester E. Alban: Systematic Analysis of Gear Failure. (1985). ISBN: 0-87170-200
[9] E.E. Shipley: Failure Modes in Gears,ct 1973 (Detroit), American Society of Mechanical Engineers]
[10] Prof. K.Gopinath & Prof. M.M.Mayuram Gear Failure Indian Institute of Technology Madras.
[11] G. DALPIAZ, A. RIVOLA and R. RUBINI Gear Fault Monitoring comparison of Vibration Analysis Techniques.
[12] Ebrahim Ebrahimi Fault diagnosis of Spur gear using vibration analysis Journal of American Science, 2012;8(1).

Vol. 3 Issue 1 September 2013 32 ISSN: 2278-621X