1

CHAPTER 1

INTRODUCTION

1.1 VIBRATION

Vibration is the motion of a body or system of connected bodies

displaced from a position of equilibrium periodical or random in time domain.
Machinery of all types commonly produces vibrations to indicate its operating
conditions. Most of the vibrations are undesirable in machines and structures
because they produce energy losses, increased stresses and bearing loads and
wear. Consequently, they affect vehicle dynamics and absorb energy from the
system.

Rotating machinery parts need careful balancing in order to prevent

premature damage from vibrations. Vibration analysis is used to detect the
causes of vibration and it is useful for predictive maintenance. Vibration is
not a steady one and it varies from instrument to instrument and time to time.
Even a same instrument will sense and produce different sets of vibration.
Hence there is no fixed method in measuring and controlling vibration. An
analysis must be done on a vibration body before framing a method to control
or minimize it.

1.2 NEED FOR VIBRATION ANALYSIS

Vibration analysis is found to be a significant approach in the

industries for analyzing the performance of rotating machinery. Vibration
textures vary signal patterns in time and frequency domain, continuous and
 2

random. They are complex and carry the impact of certain faults in the
rotating machinery.

The measurement of vibration becomes necessary due to the

increasing demand for of higher productivity, power consumption, economic
design with higher operating speeds and assurance for the safety margins.
Any shift in the natural frequency will indicate premature failure and a need
for maintenance. Vibration is caused due to bearing faults, angular and
parallel misalignments, loose joints and improper mounting.

Vibration analysis is further useful in identifying the natural

frequencies to avoid resonance and to design vibration isolator to prevent
structural instability. Vibration profile is useful for condition diagnostics and
maintenance of the rotating machinery. In rotating machinery, predictive
maintenance is significant to address the plant needs.

1.3 TYPES OF MACHINE PREDICTIVE MAINTENANCE

TECHNIQUES

The following are the major predictive maintenance techniques for

plant machinery, for health monitoring and optimum operating conditions of
the machinery:

x Vibration Analysis

x Ultrasonic techniques

x Thermography

x Visual Inspection

x Non Destructive Analysis

 3

Vibration analysis is found to be one of the dominant and predictive

maintenance techniques. It ranges from budding problems to catastrophic
failure of the rotating machinery. Its leads to acceptance testing, quality
control, loose part detection, noise control, leak detection, machine design and
engineering.

Table 1.1 below presents, a few equipment and processes which

generally require vibration analysis and monitoring of its continuous
operations.

Table 1.1 Equipment and Processes monitored by Vibration Analysis

Centrifugal Reciprocating Continuous Process

Pumps Pumps Hot and Cold strip lines
Compressors Compressors Paper machines
Blowers / Fans Engines Continuous casters
Motors / Cylinders Printing machines
Generators
Ball mills / Rotary Hobbing machines Process flow lines
dryers (Chemical, Petroleum etc.,)
Mixers Metal working
machines
Gearboxes Temper mills

The table makes it clear that vibration is found to be a significant

parameter and it needs to be monitored and control out of these different
machinery operations and plant process. From different manufacturing sectors
which use various machinery, it is understood that under optimal operating
conditions, high productivity with low cost design can be achieved. The
majority of the mechanical forces are transferred by rotating machinery.
Health monitoring of these is important and vibration analysis is one among
 4

the predictive maintenance techniques. It plays a major role in the efficiency

of the equipment listed in the table.

1.4 VIBRATION SIGNAL

Vibration is a repeatable oscillatory motion. It is a manifestation of

any of the defects like unbalanced force, joint or alignment. Vibration
increases bearing load and decreases bearing life. Regular vibration
monitoring/analysis can predict fault, shorten shutdown time and decrease
costs due to sudden failure.

A vibration is a periodic motion or one that repeats itself after a

certain interval of time. This time interval is referred to as the period of the
vibration, ‘T’. The inverse of the period, f = 1/T, is called as frequency of
vibration.

Vibration can be expressed in Equation (1.1) as,

X = Xo sin (Zt) (1.1)

where

X = Vibration displacement (m)

Xo = Maximum displacement or amplitude (m)

Z = Angular frequency (rad/sec)

t = Time (s)

Since vibration signals are complex, continuous and random in time

and frequency domain, often vibration signals are understood by spectrum
analysis where the extraction of the frequency information in the signal
composition is possible.
 5

1.5 FREQUENCY SPECTRUM

The frequency domain signal or frequency spectrum is a plot of the

amplitude of vibration response versus the frequency. It can be derived by
using the digital Fast Fourier transform analysis of the time waveform. The
frequency spectrum provides valuable information about the condition of the
machine.

The vibration response of a machine is governed by its components

assembly, mounting, and installation. Thus the vibration characteristics of any
machine are somewhat unique to that particular machine. Hence the vibration
spectrum can be considered as the vibration signature of that machine.

Figure 1.1. depicts the common spectrum for some basic signals. It
indicates time and frequency domain content in a signal and also is related to
the machinery faults with respect to frequency spectrum.

Waveform Time signal Spectrum Machine fault

Sine Unbalance

Rectangular
Looseness

Shock
Rotor Rub
Impulse

Amplitude Eccentricity
modulated
Sine wave
Cracking

Figure 1.1 Time signal and Spectrum of common signal

 6

1.6 COMMON LIMITS FOR VIBRATION

Rotating machinery such as electric motors of 1000 to 2000 rpm,

Generators, Centrifugal fans less than 600 rpm, blowers, gear boxes etc suffer
from vibrations. Until certain limits of vibration these machinery will not
normally cause problem within their vibration limits. The machinery will
have normal vibration limits in terms of displacement from 2 to 3 mm at
1 rpm, velocity from 0.2 to 0.3 mm/s and in terms of acceleration 0.5 to
10 mm/s2.

Based on the force of action, vibration displacement varies, and

hence the acceleration. By using proper sensory unit, all these three
parameters are evaluated. Normally blowers, pumps, centrifugal fans are
known for their vibration. It is limited through proper mounting.

1.7 VIBRATION ANALYSIS TECHNIQUES

Common vibration analysis techniques for studying vibration

signature of rotating machinery are enlisted below:

1. Raw time waveform analysis

2. Spectrum analysis

(i) Linear spectrum

(ii) Auto power spectrum

(iii) Frequency spectrum by FFT

3. Speed trials

4. Impact testing

5. Synchronous averaging
 7

6. Operating deflection shape (ODS) analysis

7. Spectral and other analytical techniques

8. Finite element analysis

1.8 VIBRATION STANDARDS

Based on the common limits for vibration, vibration has been

categorized by certain ISO standards for different kinds of machinery.
Vibration standards commonly followed in industries are presented in
Table 1.2. It is useful for monitoring different types of rotating machinery.

Table 1.2 ISO Standard for large and medium sized machinery

The above Table 1.2 displays the categories of vibration with

different colours for large machines to medium size machinery as standard.
The size of the machine is rated in terms of power.

Also common standards for different vibration severity criteria for

the general machinery turning from 600 to 12000 rpm and slow speed large
 8

machinery turning from 600 to 1200 rpm have been classified in Table 1.3
and Table 1.4.

Table 1.3 Vibration Severity criteria (Recommended for General

Machinery Turning from 600 to 12000 RPM, based on ISO
IS 2372)

Table 1.4 Vibration Severity criteria (Recommended for Slow Speed

Large Machinery Turning from 600 to 1200 RPM, based on
ISO IS 3945)
 9

Permissible vibrations for large induction motors are classified in

Table 1.5 which represents the rated speed and its corresponding
displacement. It is understood that vibration parameters like displacement is
related to rotating machinery parameters.

Table 1.5 Permissible Vibration for Large Induction Motors

(Based on NEMA MG 1-20.52)

S.No RPM Range Displacement amplitude (in)

1 3000 – 4000 0.001
2 1500 – 2999 0.0015
3 1000 – 1499 0.002
4 999 and below 0.0025

These standards are applicable in the no-load operation of AC

motors, at rated voltage and nominal frequency, nominal speed and operating
temperature.

1.9 SPECIFIC GOALS FOR VIBRATION ANALYSIS

From the standards and operating conditions and maintenance of the

rotating machinery in the industry, the measurement of vibration becomes
necessary due to the following reasons:

1. Increasing demand for higher productivity in any observed

shift in the natural frequencies or other vibration
characteristics will indicate the need for maintenance of the
machine.
 10

2. Measurement of the natural frequencies of a machine is useful

in selecting the operational speeds of nearby machinery to
avoid resonant conditions.

3. Theoretically computed characteristics of a machine may be

different from the actual values due to assumptions made in
the analysis.

4. Measurement of frequencies of vibration and the forces

developed are necessary in the design and operation of active
vibration isolation system.

5. In many applications, the survivability of a machine in a

specified vibration environment is to be determined. If the
machine can perform the expected task even after completion
of testing under the specified vibration environment, it is
expected to survive the specified condition.

6. Measurement of input and the resulting output vibration

characteristics of a system help in identifying the system in
terms of its mass, stiffness, and damping.

1.10 COMMON MACHINERY FAULTS

There are many faults observed during the operating condition of

the machinery in which vibration plays a set role as the resultant of faults.
Rotaing machinery of speeds less than an rpm will suffer from beats, oil
whirl, looseness and resonance. The machinery which is operated just above 1
rpm with certain harmonics is disturbed by unbalance, misalignment,
eccentricity, bent shaft etc. Medium frequency machinery is found to be
confrented with misalignment, looseness and wear whereas high frequency
machinery is intervened by cavitation and bearing problems. Hence vibration
 11

from different machinery is to be analysed as the indication of budding faults

in the machinery.

Amplitudes during beats will appear for a while and vanish

immediately. It will not persist like the signals related to unbalance and
misalignment. Faults like bent shaft, bearing damage will introduce certain
axial forces during vibration. When gears are operated under high frequency,
side bands in the spectrum will indicate the vibration composition. Looseness
and eccentricity will produce shock pulse at the point of contact only.
Sometimes thery are perceived when the load increases on the machinery.

1.11 ADVANTAGES OF VIBRATION ANALYSIS

Here are some common advantages of analyzing vibration through

proper methods and instruments:

1. Determination of machine dynamics and condition of the

machine.

2. Predicting the feasibility and capacity of using the machinery.

3. Cost involved in maintenance, service and their scheduling.

4. Monitoring the health, early detection of mechanical fatigue

and breakdown of plant machinery.

5. Forecasting gradual development of new vibrations to detect

wear at the earliest stages.

1.12 OBJECTIVES OF THE RESEARCH

The objective of this research is to develop an enhanced virtual

instrument system with intelligent monitoring and analysis techniques to
estimate the performance characteristics and to determine the operating
 12

conditions. It is proposed to develop a control scheme for predictive

maintenance.

The objectives of the research that fulfill the need for the analysis
are as follows:

1. To design an enhanced virtual instrument system for monitoring

and analysis with network based automation features.

2. To develop a fuzzy logic based classification system for

classifying rotating machinery vibrations.

3. To formulate an Artificial Neural Network based fault

identification system for identifying the specified faults for
predictive maintenance.

4. To simulate suitable control scheme with PID controller and

evaluate its tuning parameters.

1.13 ORGANISATION OF THE THESIS

The thesis is organized with the following chapters to explain the

scope and results obtained through this research.

Chapter 1 narrates the need for analysis with rotating machinery,

vibration standards and limits, common machinery faults and usefulness of
vibration analysis.

Chapter 2 describes the previous research review, which were

carried out in relation to the scope of this research. Sub- divisions are
categorized based on the condition monitoring of the rotating machinery,
 13

analysis of vibrations, Fuzzy logic and Neural network based vibration

analysis and vibration control.

Chapter 3 presents a LabVIEW based Virtual instrument system

with a performance compatibility with conventional instruments along with
various analytical methods. The methods analyse the vibration signals to
enhance the research in a focused way. It further discusses the energy loss
calculation during vibration. It explains the web enabled measurement system
as VI server to carry out the analysis over the network. Some specifications of
the signal interfaces and other hardware are also presented for information.

Chapter 4 explains fuzzy logic based vibration classification system

for different frequency categories and presents them with diagrammatic
illustration.

Chapter 5 delineates Artificial Neural Network based fault

identification for predictive maintenance of rotating machinery. Here, a
comparative study has been highlighted to show that the proposed method is
more efficient than the previous methods discussed in the literature review.

Chapter 6 is dedicated to a discussion of vibration control to explain

a PID controller with auto tuning parameters derived from fuzzy rule base.
The results are compared to show that the proposed research is enhanced
approach.

Chapter 7 consists of the conclusion, suggestions and scope for

further research.