[attachment=10246]

Abstract
The meshing of gears produce a sound .This sound will normally vary for a perfect gear, a gear with one
or more broken tooth and a gear that has undergone wear and tear. Similarly the sound of a chain drive
will be different when the chain has slackness. The sound difference is captured by a microphone or piezo-
electric crystal and fed into the computer to generate frequency-amplitude graphs in real time. These
graphs serve as ideal references for finding fault of similar gear box without actually opening it. With the
pre-recorded data, a microcontroller also analyses and gives an indication on the type of error on to a 7
segment display.

Introduction
This is a method of fault detection of gear box
without actually opening it. The sound and
vibration of gear meshing is different for a set of
perfectly meshing gears, one that has one or
more broken teeth and one that has wear and
tear. An experimental gear box is set up to
demonstrate the same. Three sets of gears are
fabricated and one pair is set for reference. A
tooth is machined off from the second set of
gears to induce a tooth fracture. Overall tooth
thickness is reduced for the third set of gears and
it acts as an induced wear and tear. Two
microphones are used to capture the sound of
gear box and ambient sound respectively. The
ambient sound is captured to remove the ambient
content from the main microphone. The signal is
amplified and fed into a computer through the
‘line in’ port. A graph is generated in real time
on the computer. The graph for perfect gear,
fractured gear and wear induced gear is noted.
These acts as reference for similar gear box’s
fault analysis.
Pillars Of Reference
Perfect gear
It acts as a reference to compare the faulty gears
.
Tooth fractured gear
This type of gear gives periodic disturbances. As
the fracture rotates, a graph will repeat its peak
or peaks in regular intervals. The interval
depends on the number of tooth fractured. A
time dependent amplitude graph is used.
Gear with Induced wear and tear:
When a wear is formed eroding away the tooth,
the average tooth thickness reduces, causing
clattering sound. This has an increase in
frequency as well as amplitude with reference to
the perfect gear. A graph of frequency with
respect to amplitude is taken for analysis.
Classification Of Components
• Mechanical components
• Electrical components
• Electronic components
Mechanical Components
The two major mechanical components are:
• A fabricated gearbox
• Chain drive with sprocket
Gearbox
The experimental gear box is made from an
angle-iron frame welded together. The Shafts of
the gears are mounted on bearing blocks. One
Shaft permits relative parallel motion with the
other. This allows the gears to be shifted. Six
gears of equal dimensions are used. One set is
the perfect gear, the second set is the one that
has a fracture and the third set has a wear
induced in it. The gears can be shifted to each
position for each fault. Grooves are provided on
one shaft to hold the shaft to one place during
drive.

Chain & Sprocket Drive
A chain and sprocket is setup for experiment on
the same frame housing the gearbox. The small
sprocket drives the bigger one via a chain. The
set up is such that the slackness can be adjusted.
This is done with the help of a chain adjuster
used in motorcycles. The graph is analyzed for
all slackness of the chain and the results are
noted. An indicator is made using
microcontroller that indicates when it is time to
tighten the chain.
Electrical Components
An AC motor of 45 watts is used to drive the
shaft having the gear as well as the shaft of the
sprocket. A foot pedal with a variable resistance
controls the speed of the gear. This enables to
shift the gear during running by momentarily
releasing the foot pedal. This eliminates the use
of clutch, making the experimental setup more
economical as well as simple in design.
Electronic Components
• Analog Components
• Digital Components
Analog Components
Analog components account for all the electronic
components used in to process the signal that has
to be fed into the micro controller.
Components are
• Amplifier (Using LM741)
• Signal Peak Extraction System
• Condenser Microphone / Piezo-
Electric crystal
• Voltage Comparator (LM311)
• 7 Segment LED display
Digital Components
The microcontroller is the main digital part of
the system. The micro controller used is an
ATMEL ATMEGA8 microcontroller. The
program is written in C++ and compiled. It is
then programmed using an interface programmer
board that transfers the program to the controller
via RS232 port of the computer.


Signal Generation
Two microphones feed data from gear box and
ambient sound. The sound is first amplified
using an operational amplifier IC LM741. A
voltage comparator is used to remove the
ambient content out of the main microphone,
leaving behind the sound from the gear box or
chain drive alone. This sound is taken and the
peak is extracted using a Zener diode. This
signal is given a negative gain using the same
type amplifier. This is done to reduce the peak of
the voltage signal to a range of 1 volt.
Analysis And Detection Methodology
There are possibly three types of display that the
micro controller makes for a gear and two for
chain analysis. The display is made on a seven
segment led display. The code generated for
displaying each error on the 7 segment is pre-
defined. When encountered with cyclic or
repeated disturbances the microcontroller
displays an error indicating a fractured gear. The
analog signal from the circuit is fed into the
Analog to Digital Converter of the
microcontroller. The values of a signal at any
instant in the graph is noted and assigned to a
variable .The variable is reset to zero every 3
seconds. Peak is noted and the number of times it
is attained it noted. Thus it identifies repeating
type of disturbances. The average of amplitudes
for a perfect gear is noted. This is compared
along with frequency for the analysis of gear
having wear and tear. The same methodology is
used for fault analysis using graphs. Cyclic peaks
and change in amplitude and frequency is noted

Application
This method of gearbox fault analysis can be
used to check or predict the fault of a gearbox
without actually opening it. This can be used for
similar type of gearbox without re-calibration. It
can be used in production plants to check large
number of gearboxes as they are being
manufactured before leaving the industry. It can
be used in service centers where the same type of
gearbox is serviced.
The analysis on a chain can be made as an
indicator in the instrument panel of a motorcycle
so that the customer is informed the time for
tightening the chain