ENVR-S335 - U7 Basic Principles of Noise Concepts

ENVR S335
Environmental Control,
Monitoring and Modeling
Unit 7
Basic principles of
noise concepts
211
OUHK Course Team
Course Development Coordinator:
Prof. K C Ho, OUHK
Developer:
Y H Yau & Associates
Instructional Designer:
Cliff Hall, OUHK
Members:
Dr Chin-wing Chan, OUHK
Dr Gordon Maxwell, OUHK (Units 4–12)
External Course Assessor

Prof. C S Poon, Hong Kong Polytechnic University
Production
Educational Technology and Development Unit
Copyright © The Open University of Hong Kong, 2003, 2009, 2011,

2014.
Reprinted 2021.
All rights reserved.

No part of this material may be reproduced in any form
by any means without permission in writing from the
President, The Open University of Hong Kong. Sale of this
material is prohibited.
The Open University of Hong Kong

Ho Man Tin, Kowloon
Hong Kong
This course material is printed on environmentally friendly paper.

Contents
Introduction 1
Basic concepts and the physical assessment of noise 4

Basic definitions and symbol conventions 5
Measurement of steady and time-varying noise 22
Basic theory of sound insulation and acoustics 29
Noise prediction models 46
Construction noise 52
Industrial noise 56
Basic principles of vibration 70

Introduction 70
Vibration source 80
Train noise 83
Industrial noise 84
References 89
Key environmental concepts 90
Feedback on activities 92
Unit 7 1
Introduction
Hong Kong is probably one of the noisiest cities in the world. An
estimated one million people are exposed to high levels of traffic noise,
making it the biggest noise problem in the region. Whether from a
stationary source or a moving vehicle, noise problems need complicated
solutions. Noise usually cannot be eliminated, only contained or moved
out of earshot. For example, the old Kai Tak airport, which affected
380,000 people (because Kowloon City residents used to live right
under the flight path) was moved to relatively isolated Chek Lap Kok
in 1998. However, roads and buildings cannot be moved so easily. It
is necessary to find other means of protecting residents from the worst
excesses of noise.
Aircraft noise for areas inside the yellow zones is at or above NEF25. NEF
stands for Noise Exposure Forecast, a noise descriptor commonly used to
describe aircraft noise. The yellow zone for Chek Lap Kok airport is largely
over the water while that of the Kai Tak airport covered some densely
populated areas such as Kowloon City.
(http://www.epd.gov.hk/epd/noise_education/web/ENG_EPD_HTML/m2/
types_4.html)
We are also exposed to different levels of pounding, roaring or shrill

noises from construction work, cargo-handling facilities, ventilation
systems, intruder alarm systems, music concerts and our neighbours.
These noises are dangerously loud and may cause permanent damage
to our hearing. Noise can also result in mood problems and adversely
affect job performance. Noise is a leading quality of life complaint
in Hong Kong. For example, cramped development has resulted in
highways running just outside our living rooms; the quiet of our homes
is interrupted by noise from our neighbours, music blaring in the next
apartment; train whistles and planes fly overhead; the Road Show plays
2 ENVR S335 Environmental Control, Monitoring and Modeling
disturbing noises along the road, which has violated the benefit of sitting
quietly in a bus; and some recreational activities such as disco, music
concerts and football matches create loud noises which interfere with
the peace and quiet of the community. Moreover, the noise we create
with vacuum cleaners, air-conditioners, television sets, hair dryers,
washing machines and other home appliances is of our own making (see
table below). In general, continued exposure to noise above 85 dBA
will eventually harm our hearing. Although many of these noises are not
loud enough to pose a hearing risk, they can increase stress and affect
our health and well-being
Home Workplace Recreation

Refrigerator (50–60) Library (40–50) Quiet residential area (45–50)
Electric toothbrush (50–60) Photocopy machine (70–85)
Washing machine (60–80) Electric drill (80–95) Heavy traffic (80–90)
Air conditioner (60–75) Ventilation fan (100–110) Noisy restaurants (70–85)
Electric shaver 60–75) Air compressor (95–110) Truck (80–90)
Vacuum cleaner (70–90) Saw (100–120) Motorcycle (90–120)
Hair dryer (65–85) Hammer on nail (100–120) Disco (100–120)
Alarm clock (70–85) Ambulance (110–125) Music concert (110–140)
Doorbell (75–90) Airplane taking off (120–130) Car horn (90–110)
Telephone ringing (70–85) Conference room (70–85) Stereo systems with
headphones (100–120)
Baby crying (70–85) Health clubs (85–100)
Television set (70–85) Football match (110–120)
Noisy toys (120–140) Cinemas (90–120)
Flushing toilet (70–90)
(Y H Yau & Associates)
Noise in Hong Kong, or in any other area, is an unavoidable by-product

of human activity. This unit introduces noise prediction models and
principles of vibration. It is based on and designed to follow on from
the noise units and relevant material in T234/T237. A brief discussion of
sound and its measurement is given at the beginning of each section.
Unit 7 3
Specifically, this unit:
• describes fundamental characteristics of transportation, construction

and industrial noise.
• explains how to perform calculations by hand involving sound

wavelengths, sound pressure levels, and sound power levels, and the
adding and averaging of decibels.
• describes measurement procedures and terminology relating to

transportation, construction and industrial noise.
• describes techniques to diagnose the relative importance of different

noise sources and transmission paths.
• describes factors affecting transportation, construction and industrial

noise, and methods of prediction.
Basic concepts and the physical

assessment of noise
Sound is a form of energy that is transmitted by pressure variations,
which the human ear can detect. When you play a musical instrument,
say a guitar, the vibrating chords set air particles into vibration and
generate pressure waves in the air. Your neighbours may then hear the
sound of the guitar when their ears perceive the pressure waves. Sound
can also travel through other media, such as liquids or solids. However,
sound cannot travel in a vacuum, since there are no air molecules
to transmit the energy. Apart from musical instruments, sound can
be produced by sources such as human vocal cords, a running car, a
vibrating loudspeaker diaphragm, industrial machine tools, and so on.
Noise also means ‘unwanted sound’. Usually, the sound of a piano is

referred to as music and is something pleasing. Depending on other
factors, the sound may be perceived as noise, which is subjective.
Factors such as the magnitude, characteristics, duration, and time
of occurrence may affect one’s subjective impression of the noise.
Different people may respond differently to the same level of noise. But
above certain levels, noise can affect everyone. It can lead to hearing
loss and mental stress and irritation. It can also interfere with daily
activities such as doing homework, watching television and talking on
the telephone during the day and disturbing sleep at night.
As you learned in this introductory section, sound and noise are often
used interchangeably. When dealing with physical properties, ‘sound’ is
preferred, because it does not carry any sense of subjectivity. Therefore,
it is better to restrict the use of ‘noise’ to unwanted sound, although this
does not always produce an infallible distinction. Before undertaking a
study of noise prediction models, it is important to be familiar with the
terminology of acoustics, the basic rules of sound propagation, and the
features of sound-level measurements. The basic principles and concepts
of noise are described in Unit 11, Sections 1 and 2 of T234/T237. Sound
can be defined as ‘pressure vibrations in the air, perceptible to our ears’.
However, we rarely appreciate all of its functions, for example:
One person’s music is often another’s noise.
Although we may enjoy listening to our own music, others may

regard the type of music we enjoy as noise, especially if it is loud. It
is important to remember here that sound does not need to be loud to
annoy; for example, the sound of a crying baby or a scratching sound on
glass can be very annoying.
However, the level of annoyance depends not only on the quality of the
sound but also on our attitude towards it. The judgement of loudness
will also depend on the time of the day; for example, a higher level of
noise will be tolerated during the daytime than at night.
Unit 7 5
Noise is an unwanted audible sensation, sound, or an intermittent

oscillation or vibration. It can disturb people from working at maximum
efficiency and enjoying sleep, which in turn leads to tension and certain
undesirable psychological effects. Even worse, the hearing system is
damaged and may cause temporary or permanent deafness. Because of
Hong Kong’s dense population, high-traffic density, many construction
sites, tall buildings reflecting and channelling sound, noise pollution is
gaining attention from the public and EPD. In 2002, the Government
handled about 12,000 noise complaints, about half of which concerned
noise from domestic neighbours. The rest were mostly against
stationary noise, such as construction sites and noisy air-conditioners
in commercial complexes or restaurants. In addition, excessive traffic
noise continues to affect one million people in Hong Kong each day.
Basic definitions and symbol conventions

From Unit 11, Section 2 of T234/T237, we understand that sound is
generated as a vibration which excites the particles in an elastic medium
to form an acoustic wave or sound wave. The sound waves then carry
energy and cause pressure variation in the air. The air particles move
back and forth along the direction of travelling. When the pressure
fluctuations in the wave reach the ear, the eardrum vibrates in direct
response, and the pressure variations are heard as sound.
Frequency
The pressure fluctuation is periodic, and a complete fluctuation in

pressure is called one cycle. The total time T (s) required to complete
one cycle of wave is termed period. The frequency, f (Hz) is defined
as the number of vibrations or pressure fluctuations per second. Their
relationship is:
Example
As an example of using this equation, the corresponding period for the
frequency of 5 Hz can be calculated as follows:
= 1/(5 Hz) = 0.20 s.

Wavelength
If we imagine a body moves with a uniform speed v, it will travel a

distance during a time interval. The product of the speed and the time
interval, Δt gives the distance travelled Δs.
Δs = v × Δt
For a sound wave travelling at the speed of propagation, c, the period

T represents the time it takes for the wave to travel the distance
represented by the wavelength λ. Substituting for this into the above
equation, we obtain:
λ=c×T
Period, T, is the total time needed for completing one cycle of wave.
The relationship with frequency, f, is
Since wavelength, λ, is the distance of a sound wave in one complete

cycle, and sound speed is the distance travelled by a wave in a unit
time, its relationship with frequency and wavelength is expressed by the
equation:
Example
A body oscillates in simple harmonic motion at 300 Hz at 21.1oC.
Calculate the wavelength of the sound wave generated by the vibrating
body.
Solution
From equation c = λf, you should observe that once the sound speed is
established for the medium, you are able to calculate the wavelength
knowing the frequency of the sound source.
TK = 273.16 + 21.1 = 294.26 K

Unit 7 7
Sound speed
In fact, speed of sound relates to the elasticity and density of the

medium. If the medium is stiff or elastic, the restoring force to the
particles is strong and the pressure fluctuations can be transmitted quite
readily. The speed of sound is then high. If the density of the medium is
high, the vibrations or fluctuations are slow and the sound velocity will
be smaller. The speed (m/s) at which a sound wave travels through a
solid or liquid medium is shown in the equation below
where c = sound speed (m/s)

E = Young’s elastic modulus (N/m2)
ρ = density of the medium (kg/m3)
There is a common belief that the progression of the particles through

a medium would be much more restricted in a denser media. However,
usually very dense materials have much higher elasticity than less dense
media, so the particle moves more freely. This equation is important to
the fundamental understanding of sound transmission or isolation of
acoustic media.
For sound waves travelling through air, the speed of sound, c, is related
to the ratio of specific heat, specific gas constant and air temperature as
follows
where
c = sound speed (m/s)

γ = ratio of specific heats (= cp/cv = 1.40)
cp, cv = specific heat at constant pressure and volume respectively (kJ/kgK)
P = atmospheric pressure (N/m2)
ρ = density of air (kg/m3)
Applying an ideal gas equation, it can be shown that the sound speed at
1 atmospheric pressure is only related to the air temperature as follows
where c = sound speed (m/s)

T = air temperature (TK = T(oC) + 273.16)
Example
Calculate the sound speed of a minibus at 20oC and 1 atmosphere.
Assume that the bus is travelling in second gear at 40 km/hr.
Solution
Since the temperature is in degrees Celsius, first convert to the Kelvin

scale. Referring to equation TK = T(oC) + 273.16
TK = 273.16 + 20 = 293.16 K
Activity 7.1
Now, I wish you to familarize yourself with the various equations
learned and to develop your practical knowledge on their applications.
Derive equation starting with equation . List
some of the factors that affect sound speed in air. What will be the
sound speed on top of Victoria Peak in Hong Kong compared with that
at the bottom of the Peak in Central, provided that the atmospheric
pressure and temperature at the top of Victoria Peak are 0.89 atm and
15oC respectively, and at the bottom are 1 atm and 25oC.
Using equation , it can also be shown that sound travels faster
in materials having high elastic modulus and low density. Table 7.1
shows the value for some common materials.
Table 7.1 Speed of sound in various materials (Hearn 1985)
Elastic Approximate
Density
Medium modulus sound speed
(kg/m3)
(GN/m2) (m/s)
Vacuum 0 0 0
Air at 20oC and 1 atmosphere — 1.204 343
Cast iron 90 7640 1200
Water 2.0 1000 1410
Aluminium alloy 69 2770 4991
Steel 208 7800 5164
Unit 7 9
Activity 7.2
What do you notice about the relationship between sound speeds and
different materials as shown in Table 7.1? Discuss the implications of
your observations within your tutorial group.
Decibels
The measure of sound or acoustic pressure level (Lp), sound or

acoustic power level (Lw) and sound or acoustic intensity level (LI)
used in practice, describe the relative ‘strength’ of the sound. They are
expressed in decibels (dB), which relate the quantity to the reference
quantity. For example, Lp relates the amplitude of the sound pressure to
a reference sound pressure. This reference sound pressure corresponds
to the lowest audible pressure by humans of 2 × 10-5 Pa from a 1000 Hz
pure tone. The average normal human ear can perceive pitch sensations
corresponding to sound waves in a range of frequency between 20 and
20,000 Hz. Because the sensitivity of the ear varies with frequency,
a middle-range frequency of 1,000 Hz is taken as reference for the
average normal threshold of hearing.
In practice, the sound pressure of the faintest sound that a normal

healthy individual can hear is about 2 × 10-5 Pa. Noise, commonly
defined as unwanted sound, is an environmental phenomenon to
which we are exposed before birth and throughout life. Noise is any
sound — independent of loudness — that can produce an undesired
physiological and psychological effect in an individual or group. The
pressure associated with the sensation of pain is about 200 Pa. Because
this inconveniently large range in the magnitude of these numbers and
our ears’ response is not directly proportional to pressure, the concept of
the decibel is introduced in this section. The decibel compresses these
numbers into a more meaningful scale from the threshold of hearing at
0 dB to above the threshold of pain at about 140 dB (see Figure 7.1).
The levels corresponding to most noise sources lie between these two
extreme limits. These are typical levels in Figure 7.1; however, actual
noise levels may vary depending on the particular item. You should note
that 0 dB is approximately the softest sound the healthy human ear can
hear, and continued exposure to noise above 75 dB over time will cause
hearing damage. To know if a sound is loud enough to damage your
ears, it is important to know both the loudness level (measured in dB)
and the length of exposure to the sound. In general, the louder the noise,
the less time required before hearing damage will occur. According to
the Factories and Industrial Undertaking (Noise at Work) Regulation
in Hong Kong, the maximum exposure time at 85 dB is eight hours
or when the maximum value of the unweighed instantaneous sound
pressure is likely to be greater than 140 dB. If you must be exposed to
noise, it is recommended that you limit the exposure time and/or wear
ear protectors. Remember that noise levels above 85 dB will harm
hearing over time. Noise levels above 140 dB can cause damage to
hearing after just one exposure.
Figure 7.1 Comparative scaling of sound pressure levels and decibels

(Kiely 1998)
Sound pressure level
As introduced in Section 2.4 of Unit 11 of T234/T237, the sound

pressure level (Lp) at a point is measured in decibels (dB) and is equal
to 20 times the logarithm to the base 10 of the ratio of the root mean
square sound pressure to the reference sound pressure, 2×10-5 Pa. You
will note that this is not an absolute scale but a comparative scale.
where
P = root-mean-square (rms) sound pressure (Pa or N/m2)

P0 = reference root-mean-square sound pressure (i.e. 2×10-5 Pa or N/m2)
Unit 7 11
Example
Find the decibel reading corresponding to a pressure amplitude
P = 0.2 N/m2.
Solution
Using the above definition for Lp we begin with the term in brackets on
the right-hand side of the equation by dividing P by P0.
P/P0 = 0.2/2×10-5 = 10000
Next we take the log of this value (in this unit, log always means log10)
log(10000) = 4
and multiply by 20 to get Lp
Lp = 20 × 4 = 80 dB
Sound intensity
Sound intensity represents the energy-flow through the medium per unit
area per unit of time. Since energy per unit time equals the sound power
W, the power of the point source will spread over the surface of a sphere
of radius r (see Figure 7.2). You should note that a sound source could
be regarded as a point source if its dimensions are small compared
with its distance from the receiver and if it radiates an equal amount of
energy in all directions. Aircraft noise is an example of a point source.
The relationship between the intensity I and power W is
where
I = sound intensity at distance r from the point source of sound (W/m2)

W = sound power, total energy radiated per unit time of a sound source (W)
S = surface area of the spherical wave front (4πr2 m2) and
r = distance from the point source (m)
Figure 7.2 Spherical sound wave propagation generated by a point source

In a sound field consisting of waves coming from a source, it is usual

to refer to the sound intensity. Intensity is the amount of energy
passing through unit area per unit time. From the equation above, you
may observe that the intensity decreases inversely as the square of
the distance from the point source. Another important relationship in
environmental noise pollution involves the sound intensity with the
sound pressure experienced by the air molecules as the sound wave
propagates through the medium. This is equal to the square of the
pressure divided by the product of density times the speed of sound for
the medium. You should remember that in Hong Kong, air density could
be equal to 1.20 kg/m3 at atmospheric pressure and normal temperature
ranges from 10 to 30oC.
You should note that this equation can be applicable to both plane and
spherical waves
where
P = sound pressure (Pa)

ρ = medium density (kg/m3)
c = sound speed (m/s)
ρc = specific acoustic impedance of the medium (kg/m2s).
Example
If the sound power from an electric drill in a construction site near the
classrooms of the Open University of Hong Kong equals 100 W, what
are the intensity and the sound pressure at 10 m from the drill to the
classrooms of the University?
Solution
At a distance 10 m from the drill to the University, the drill is

comparatively small and can be considered as a point source, so that the
sound intensity at this distance from the source is .
From
Unit 7 13
Sound Intensity level
From , the sound intensity at a point, I, is directly proportional to
the square of the acoustic pressure, P.
Therefore, in which
I = sound intensity (W/m2)
I0 = reference sound intensity, generally 10-12 W/m2
However,
Therefore,
where I0 = 10-12 W/m2
Activity 7.3
Now, try to integrate your knowledge about the various equations
learned.
Determine the reference sound intensity, I0 = 10-12 W/m2 for the equation
, given an air density of 1.20 kg/m3 and a sound speed
of 343 m/s.
Sound power level
The sound power level of a source is expressed in decibels (dB) and is

equal to 10 times the logarithm to the base 10 of the ratio of the sound
power of the source to a reference sound power. The reference sound
power in air is normally taken to be 10-12 watt.
where W0 = 10-12 W
Example
Calculate the sound power level of 1 watt generated from a rock drill on
a quarry located just north of Anderson Road, Kowloon.
Solution
Relationship between sound intensity level LI, sound

pressure level Lp and sound power level Lw
where P = sound pressure (Pa), ρ = medium density and
c = velocity of sound in the medium.
Converting LI to Lp:
For standard temperature (0oC) and pressure (1 atmospheric pressure),

ρ = 1.29 kg/m3 and
Therefore,
and
where P0 = 2×10-5 N/m2
Therefore, LI ≈ Lp
You should note that the sound intensity level and pressure level are, for
all practical purposes, equal in magnitude to each other.
Converting Lw to LI:
Since , therefore W = IS
Similarly, W0 = I0S0 where W0 = 10-12 W and I0 = 10-12 W/m2, therefore

S0 = 1 m2
Unit 7 15
Therefore, Lw = LI + 10logS dB
From the above equation, we can estimate the sound power level if the
sound intensity level of a sound (or sound pressure level) and the radius
of the spherical surface is known.
Example
Calculate the sound power level of a sound at a distance of 2 m from a
uniformly radiating source whose sound pressure level is 78 dB.
Solution
Lw = LI + 10logS dB
Lw ≈ Lp + 10logS dB
where S = surface area of a sphere = 4πr2 = 16π
Lw = 78 + 10log16π = 95.0 dB
Addition of sound pressure level
The previous sections illustrate that sound pressure levels cannot be

added arithmetically, because they involve logarithmic scales. Imagine
that there are two sound sources, 1 and 2 (e.g. a minibus and a truck),
producing individual sound pressure levels of L1 and L2 dB. The total
sound pressure level will be:
The subscript 1 refers to the sound pressure level of source 1.

Rearranging terms in the equation, the equation would become
Similarly, for the second sound source

Therefore, total sound pressure level is
The above relation can be extended to several sound sources. For

example, if there are n sources acting simultaneously
or
in which Σ is the summation sign and i takes integer values from 1 to n.
Example
Three bulldozers are producing individual sound pressure levels of
83 dB measured at the pavement. What is the total sound pressure level
when (a) two and (b) three bulldozers operate simultaneously?
Solution
For part (a), Lp = 10log(108.3 + 108.3) = 86 dB
For part (b), Lp = 10log(108.3 + 108.3 + 108.3) = 87.77 dB
Activity 7.4
Now, I would like you to carry out this activity to find out the practical
values of the above equations and calculations.
From the above example, you may find it convenient to calculate two
identical sounds that result in a sound pressure level increase of 3 dB,
i.e. if L1 = L2, then Lp = =
= = 3 + L1. If n identical sounds are added, what
is the expression to calculate the total sound pressure levels? Noise
from the construction of an International Theme Park in Penny’s Bay
of North Lantau is mainly caused by ten breakers. If the total sound
pressure level is 94 dB, calculate the sound pressure level of each
breaker.
You should know the time distribution and variation of noise level
are important in the evaluation of source-and-effect relationships. For
example, random or sudden noise is more noticeable and unsettling than
constant sound levels are. Likewise, unusual sounds, cycles of periodic
Unit 7 17
sounds, and background noise levels need to be related in the source-

and-effect analysis.
Sound propagation and attenuation
Every sound source has a characteristic sound power, but sound

pressure level may vary on many external factors such as orientation
and distance of the receiver from the source. Based on the distance,
sound source is regarded as point, line and plane sources.
Spherical propagation
Section 4.1 of Unit 11 of T234/237 explains how sound waves propagate
through the media. In this section, we consider the case of a source
radiating uniformly in all directions, suspended in the centre of a large
space. Practical examples of sound waves spreading spherically from a
point source (see Figure 7.3) are noise generated from the aircraft flying
overhead, exhaust of industrial chimney or a ventilation system located
on the top of a building.

Figure 7.3 Sound waves spreading spherically from a point source

From equations, and in which S = 4πr2
Total sound power,
We also have equations, and
Therefore, and
By substituting the above equations into , we can relate
sound pressure level to the sound power level, by
Lp = Lw – 20logr – 11
where
Lp = sound pressure level (dB)

Lw = sound power level of the source re 10-12 watts (dB)
r = distance of the receiver from the sound source (m).
Self-test 7.1
Sound equations in action
Explain how you can derive the equation Lp = Lw – 20logr – 11.
With equations, and
Putting the above equation into , we get
Taking the logarithm on both sides,
Lw = 10log400 + Lp + 10(log4π) + 10(logr2) – 10logρc
Lw = 10log400 + Lp + 10(log4π) + 20(logr) – 10logρc
Lp = Lw – 20(logr) – 10log400 – 10(log4π) + 10logρc
Assume air density, ρ = 1.2 kg/m3 and c = 340 m/s
Lp = Lw – 20(logr) – 10log400 – 10(log4π) + 10log(1.2×340)
Lp = Lw – 20logr – 11
Unit 7 19
Example
Suppose an aircraft is flying with sound power of 100 W to Hong Kong
International Airport. Assuming this energy is radiated equally in all
directions, calculate the sound pressure level 300 m from the aircraft on
the ground.
Solution
Hemispherical propagation
In the previous section, we understand that spherical propagation

represents a noise source located in free space and radiating in all
directions. However, in many situations, the noise source is near the
ground and we have hemispherical propagation. Similar to spherical
propagation, and where S = 2πr2
Total sound power,
With equations, and
Puting the above equation into , we get
Lw = 10log400 + Lp + 10(log2π) + 10(logr2) – 10logρc
Lw = 10log400 + Lp + 10(log2π) + 20(logr) – 10logρc

Lp = Lw – 20(logr) – 10log400 – 10(log2π) + 10logρc
Assume, air density, ρ = 1.2 kg/m3 and c = 340 m/s
Lp = Lw – 20(logr) – 10log400 – 10(log2π) + 10log(1.2×340)
Lp = Lw – 20(logr) – 8 dB.
Example
The noise radiated from the air compressor in a construction site has the
sound power level of 109 dB(A). Calculate the sound pressure level to
be expected at a distance of 100 m from the compressor, assuming the
radiation is unidirectional.
Solution
Lp = Lw – 20(logr) – 8 dB
Lp = 109 – 20log100 – 8 = 61 dB(A)
Propagation from a line source
Another kind of noise source, called a line source, is one from which
the sound waves spread cylindrically (see Figure 7.4). Examples are a
noisy pipeline carrying a turbulent liquid, a continuous stream of road
traffic, and a railway.
Figure 7.4 Sound waves spreading cylindrically from a line source

(Unit 11 of T234/T237, OUHK)
Unit 7 21
Again, we can make use of the important equations, and
in which . We can consider the power radiated by unit length
of the line, l = 1 and S = πr. The relationship between sound power per
unit length of a line source at ground level and sound pressure with
distance r from the line is
total sound power,
With equations, and
Putting the above equation into , we get
Lw = 10log400 + Lp + 10(logπ) + 10(logr) – 10logρc
Lp = Lw – 10(logr) – 10log400 – 10(logπ) + 10logρc
Assume, air density, ρ = 1.2 kg/m3 and c = 340 m/s
Lp = Lw – 10(logr) – 10log400 – 10(logπ) + 10log(1.2×340)
Lp = Lw – 10(logr) – 5 dB.
Air absorption
Attenuation is the loss of power suffered by radiation as it passes

through matter. Sound waves, propagating through air or any other
medium, experience attenuation because air has viscosity, and some
energy has to be used to overcome intermolecular friction. Therefore,
energy is always dissipated by transmission through air due to heat
conduction, viscosity, etc. The rate of air attenuation with distance
depends on the frequency of the noise source, which also varies with the
temperature and humidity, but pressure variations have little effect. Air
attenuation becomes important over long distances because a significant
amount of energy can be removed from the sound waves. The
approximate air attenuation at a temperature of 20oC may be calculated
by the expression
where
Aa = air attenuation (dB)

f = frequency (Hz)
r = distance between source and receiver (m)
ϕ = relative humidity (%).
Example
1,000 Hz of noise emitted from an electricity substation on Austin
Road, is transferred 305 m to a residential building in Tsimshatsui at
20oC. The relative humidity is 60%. Calculate the reduction value of air
absorption.
Solution
From air attenuation, we now move on to consider both steady noise

and variable noise (or noise that varies over time).
Measurement of steady and time-varying

noise
The time distribution and variation of noise level are important in the
evaluation of source-and-effect relationships. For example, random
or sudden noise is more noticeable and unsettling than constant sound
levels are. Likewise, unusual sounds, cycles of periodic sounds, and
background noise levels need to be related in the source-and-effect
analysis. This section introduces sound measurements by describing
sound pressure, sound level, units for sound level measurements and
the basic features of sound level meters. Frequency of sound is also
described.
Unit 7 23
Noise that varies over time
If the level of a noise varies over time, it is said to be time varying or

unsteady. Types of time variation in the level of a noise may include
non-periodic fluctuation (i.e. road traffic noise), periodic fluctuation (i.e.
some factory machines), or impulse fluctuation (i.e. repeated hammer
blows). A time-varying noise can be described fully only by reporting
the entire noise level distribution. However, for practical purposes,
we can use a single number statistic of the distribution, or sometimes
several statistics, to summarize some of the information about the
distribution, e.g. the maximum (Lmax) or (Lmin) minimum noise level
that occurred. Therefore, a varying noise pattern may be described by a
statistical description of the noise as follows.
Percentile sound levels
These are the levels exceeded for N percent of the observing time.
Examples are:
L1 = noise level exceeded for 1% of the observation time.
L10 = noise level exceeded for 10% of the observation time, and is
widely used for measurement of road traffic noise.
L90 = noise level exceeded for 90% of the observation time, and can
generally be regarded as a background level.
Equivalent continuous sound level
As mentioned, the sound pressure levels measured in a noise

environment usually vary with time. The equivalent continuous sound
level Leq is used to describe the noise environment with a single number.
It is the level which, produced constantly over the same observation
time, would result in the receiver experiencing the same sound energy
as from the fluctuating sound.
By definition, it is considered as the mean energy level because it

corresponds to the total sound energy being uniformly spread over the
observation time. The equation is
where ti = measurement time, Li = measured sound pressure level (dBA)

over a period of ti, T = total measurement time.
For calculation, this equation becomes
in which we have n discrete measurement of Lp
Example
The noise exposure pattern of a person is as follows:
105 dB for 2 hours
88 dB for 2.5 hours
103 dB for 3 hours
93 dB for 0.5 hours
Calculate the average level for the period of 8 hours.
Solution
Octave band analysis
It is mentioned in Section 3.1, Unit 11 of T234/T237 that we are

frequently interested in knowing how sound energy is distributed as a
function of the sound frequency, because noises consist of not just one
frequency but a whole range of frequencies at different levels occurring
simultaneously. To accomplish this, the measured acoustic energy can
be electronically filtered into contiguous frequency bands. For each
frequency band we measure the overall sound pressure level of the
filtered sound within the band. You should be aware that most sound
level metres designed for noise measurements are limited to octave
bands. Therefore, the octave band is a common contiguous band used in
the study of environmental noise pollution. Here, the upper frequency
limit of a band is twice that of the lower frequency limit. Thus, if fl
and fu are the lower and upper frequency limits of an octave band,
respectively, then
fu = 2 fl
Unit 7 25
Since it is conventional to use a logarithmic scale for plotting

frequencies, the average or centre frequency, fc, between the lower and
upper frequency limits of an octave band is
Therefore,
Our ears, which are capable of hearing over a frequency range of

between 20 Hz and 20 kHz, respond differently to different frequencies,
being most sensitive in the 2 to 5 kHz range. Consequently, the
measurement of the frequency spectrum of a steady noise is important
in assessing the possible effects of a noise. The current set of octave
bands used with modern acoustic instrumentation is shown in Table 7.2.
Table 7.2 Sets of octave bands (Irwin and Graf 1979)
Lower frequency Center frequency Upper frequency

fl (Hz) fc (Hz) fu (Hz)
11 16 22
22 31.5 44
44 63 88
88 125 177
177 250 355
355 500 710
710 1,000 1,420
1,420 2,000 2,840
2,840 4,000 5,680
5,680 8,000 11,360
11,360 16,000 22,720
The octave band sound pressure level in decibels represents a measure

of the mean-squared pressure of the sounds within the particular
frequency band. Figure 7.5 shows a worked example of measured
octave band data for a fan heater (i.e. the range of frequencies and
associated levels). The frequency on the abscissa axis is represented
with a logarithmic scale. Also, as a matter of convention, the octave
band decibel levels are plotted at the corresponding centre frequency.
Finally, straight lines join the decibel levels to display the spectrum
data.
70
65
60
Sound pressure level

55
(dB)
50
45
40
35
30
63 125 250 500 1000 2000 4000 8000
Frequency (Hz)
Figure 7.5 Octane band sound spectrum for a fan heater

You should note that the octave bands described above are called 1/1
octave, and it is usually sufficient for general sound measurement. For
a more detailed analysis of a sound wave, finer sets of contiguous bands
are also available such as 1/2 and 1/3 octave bands.
A-weighting and noise levels
The weighting networks in sound-level metres attempt to alter the

measured signal in a similar way to the human hearing mechanism.
The A, B and C weighting networks approximate human response for
low, moderate and high sounds. From the measured equal loudness
level contours as shown in Figure 7.6, the curves of equal loudness are
developed for the following conditions:
A-weighting: human response for low sound levels, i.e. it follows

the 40 phon curve.
B-weighting: human response for moderate sound levels, i.e. it

follows 70 phon curve.
C-weighting: human response for high sound levels, i.e. it follows

100 phon curve.
As you have noted, the loudness level of a sound shown in Figure 7.6 is
determined by the subjective comparison of the loudness of the sound
to that of a 1000 Hz pure tone. The unit of loudness of sound is phon,
which measures the intensity of a sound relative to a reference tone
of defined intensity and frequency. The reference tone usually used
has a frequency of 1 kilohertz and a root-mean square sound pressure
of 2 × 10-5 Pascal, as introduced in the previous section. The observer
listens with both human ears to the reference tone and the sound to
be measured alternately. The reference tone is then increased until the
observer judges it to be of equal intensity to the sound to be measured.
Unit 7 27
If the intensity of the reference tone has been increased by n* decibels

to achieve this, the sound being measured is said to have an intensity of
n phons. The decibel and phon scales are not identical, as the phon scale
is subjective and relies on the sensitivity of the human ear to detect
changes of intensity with frequency. As seen from the 0 phon curve of
Figure 7.6, human hearing is most sensitive in the range 2000 to 5000
Hz. You should note that the response in this range is very close to 0 dB
or of 2 × 10-5 Pa.
Figure 7.6 Curves of equal loudness (Unit 11 of T234/T237, OUHK)
As seen from the curves of Figure 7.6, the threshold of hearing is a

function of frequency. For example, with normal hearing, you would
just be able to hear a 5000-Hz tone at a 0-dB level, whereas you would
require a pressure level of about 25 dB to be able to just hear a 100-Hz
tone. Standardized values of the weighting function are given in
Table 7.3.
Table 7.3 Octave band A, B and C weighting values (Irwin and Graf 1979)
Frequency A-weighting B-weighting C-weighting

(Hz) (dB) (dB) (dB)
16 -56.7 -28.5 -8.5
31.5 -39.4 -17.1 -3.0
63 -26.2 -9.3 -0.8
125 -16.1 -4.2 -0.2
250 -8.9 -1.3 0
500 -3.2 -0.3 0
1,000 0 0 0
2,000 1.2 -0.1 -0.2
4,000 1.0 -0.7 -0.8
8,000 -1.1 -2.9 -3.0
16,000 -6.6 -8.4 -8.5
In Hong Kong, the A-weighting network has been adopted in most

environmental legislation and ordinances such as the Noise Control
Ordinance and Factories and Industrial Undertakings (Noise at Work)
Regulation, to compensate for human hearing characteristics regardless
of the sound intensity level. The A-weighting network is built into the
sound level metre so that the metre may approximate the response of
the human ear. By contrast, D weighting is occasionally used for certain
aircraft noise measurements.
Example
Determine the overall A-weighted sound pressure level of a noise for a
loading task from the helicopter of the Government Flying Services in
Hong Kong International Airport with the following frequency analysis:
Frequency (Hz) 250 500 1000 2000 4000 8000

Sound pressure level 97 101 97 90 89 94
(dB)
Unit 7 29
Solution
Frequency (Hz) 250 500 1000 2000 4000 8000

Band pressure level 97 101 97 90 89 94
(dB)
A-weighting (dB) -8.9 -3.2 0 1.2 1.0 -1.1
Sound pressure level 88.1 97.8 97 91.2 90 92.9
(dB(A))
Lp = 10log(108.8 + 109.8 + 109.7 + 109.1 + 109.0 + 109.3)
Lp = 102.1 dB(A)
Basic theory of sound insulation and acoustics
Introduction
As with thermal insulation, introduced in Unit 2, sound insulation means

providing a barrier to the flow of acoustic energy. The most obvious
way of doing this is to provide a barrier in the path of propagation.
Therefore, the knowledge of the parameters of the barrier must be
understood in order to determine how much insulation it will provide.
There are three general paths for noise transmission.
Path 1
From within the buildings to the outside, for example, noise passes
through the plant room envelope to the outside environment.
Figure 7.7a High-frequency noise generated during the process of metal drilling
(Y H Yau & Associates) The workshop for installation of refrigeration
system is located at Lam Tei, Tuen Mun, September 22, 2001
From outside to inside, for example, transportation noise to the noise

sensitive area nearby.
Figure 7.7b Tram and road traffic noise (Y H Yau & Associates) Johnston Road,
December 24, 2003
From one room to other parts of the building, for example, noise is
transmitted via the enclosure for the noisy machine to other parts of the
building.
Figure 7.7c Enclosure for the abrasive wheels (Y H Yau & Associates) Wheel
and Brake Shop located at Fanling, January 19, 2004
Sound reduction between a source room and a receiving room will

depend on the nature of the dividing walls, ceiling or floor. It will also
depend on the acoustic properties of the rooms themselves. Before
discussing sound insulation theory, we begin by looking at the most
important determinant of room acoustic, reverberation time. You
studied reverberation in Section 4.4, Unit 11 of T234/T237 and will be
discussing it in more detail in this section.
Reverberation
Acoustical reverberation is encountered when the sound emitted from

a noise source is reflected many times by the room surfaces. In a real
situation, if a noise source is located in a reverberant room, then the
sound energy will be reflected many times before it is absorbed. Since
all surfaces such as walls, floor and ceiling in a reverberant room are
perfectly reflecting, the sound pressure will theoretically increase to a
large value. In practice, the sound pressure level does grow to a high
level in a small hard room used to enclosure a noisy machine (see
Figure 7.7c).
Unit 7 31
By considering direct and reflected sound, the following equation

describes the sound pressure level Lp at any point inside a room:
dB where
Lw = sound power level (dB)

Q = directivity factor
r = distance from noise source (m)
R = room constant (m2) and is defined by where
α = average sound absorption coefficient
S = total surface of the room (m2)
We can observe that as α approaches zero, R also approaches zero.
Then the equation becomes

Lp = LW + log(∞) or Lp = ∞ dB under the condition of zero absorption.
This means in a small room with hard surfaces, as shown in Figure 7.4c,
the sound pressure level can become surprisingly excessive.
In a real situation, a pronounced reverberant field exists. It is

characterized by a rather uniform or slowly varying sound pressure level
throughout large regions of the reverberant field. Under reverberant
field conditions,
then the equation of becomes
or
Activity 7.5
Now, let us think about the practical applications of the various terms
and equations learned.
From the equation , discuss within your tutorial

group the relationship between sound pressure level and sound power
level. Describe how you could determine the sound power level by
using this equation.
Example
A diaphragm (pneumatic) pump with a directivity of unity and a sound
power level of 130 dB in the 8,000 Hz octave band operates in a
pump room with a room constant of 10.32 m2 in this frequency range.
Calculate the sound pressure level in the reverberant room.
Solution
The sound pressure level due to the reverberant field is given by equation
Lp = 130 –10 log(10.32) +6 = 125.9 dB
Self-test 7.2
Reverberation exercise
Repeating the above example, can you determine the minimum distance
from the pump for the calculated sound pressure level to increase by
1 dB?
From the equation
Re-arrange the above equation to the following to make r the subject:
From the equation, the value of Lp = 125.9 + 1 = 126.9 dB
After substation,
r = 0.88 m
Therefore, r should be less than 0.88 m.

Unit 7 33
Reverberation time
This is the time required for the energy density in the acoustical field to
reduce to a level 60 dB below its steady-state value. This time period is
important for effective use of the space. The following expression is for
the time-dependent energy density:
where
δ' = instantaneous space-average energy density (J/m3)

α = average sound absorption coefficient
S = total surface of the room (m2)
c = velocity of propagation (m/s)
V = volume of the room (m3)
w = power of the sound source (W)
Assuming the noise source is turned off after the steady-state energy
density is reached, then
=0
At time t = 0, , the steady-state density
Solving for δ', the following equation is obtained
which is the expression for the transformation of the

energy with respect to time. The reverberation time, T, can be obtained
from the above equation and the definition of reverberation time:
The reason for the above expression is that in a time period, T, the
steady-state density decays to 1/106 of its original value. That means it
is reduced by 60 dB. This equation can be further simplified to
which is equal to
or
Since different surfaces of a room will have different absorption

coefficients, the average sound absorption coefficient, α, of the room as
a whole must be determined. We define αS in this equation by
or
where
αi = absorption coefficient of the ith surface

Si = area of the ith surface of the room (m2)
Substitute into
, then
We understand that c = 343 m/s at 20oC as introduced in the previous

section, sound speed, and substitute c into the equation, then
which is the familiar form of the Sabine formula for
reverberation time.
Example
Try to calculate the reverberation time of classroom A0415 in the Open
University of Hong Kong. Assume the room dimensions are 10m long ×
6m wide × 3m high, and the average absorption coefficient is 0.25.
Solution
By equation,
The volume of the room is V = 10×6×3 = 180 m3 and its surface is
S = 10×6×2 + 10×3×2 + 6×3×2 = 216 m2
= 0.25×216 = 54 m2
= 0.54 s
Unit 7 35
The assumption of equation is that the reverberant sound

energy in a room builds up or decays continuously. In reality, the
process is a discrete one and the energy changes at each reflection. Many
more complicated formulae are available for the calculation. However,
these complicated equations are beyond the scope of this course. The
absorption of air in the room has to be taken into consideration at high
frequencies. Therefore, at high frequencies, the Sabine formula as
described previously should be modified as shown below:
where
A = total absorption of room boundaries

X = absorption factor that depends on frequency and humidity, with
values according to Table 7.4
XV = the total air absorption in volume V.
You should note that, according to the above equation, the effect of air
attenuation increases the absorption from A to A+XV.
Table 7.4 Absorption factors, X, for various frequencies and relative humidities,
at a temperature of 20oC
(Section 1.5, Noise Block, ENVR S335, OUHK)
Frequency (Hz) Relative humidity (%)

30 40 50 60 70 80
×10–3 ×10–3 ×10–3 ×10–3 ×10–3 ×10–3
1000 3.28 3.28 3.28 3.28 3.28 3.28
2000 11.48 8.20 8.20 6.56 6.56 6.56
4000 39.36 29.52 22.96 19.68 16.40 16.40
Self-test 7.3
Air absorption
Note that up to this point, you have considered air absorption as a factor
in the determination of reverberation time. Can you explain why the
effect of air attenuation is more pronounced for large rooms than for
small rooms?
In a large room, air absorption can be important because large rooms

have larger volumes than small rooms.
Example
Determine the reverberation time in a lecture theatre of the Open
University of Hong Kong if the effect of air attenuation should be
considered. The theatre has a volume of 5000 m3. The air in the theatre
is 21oC and the relative humidity is 70%. If the reverberation time
empty at 4000 Hz is 2 seconds, determine the number of square metres
of absorbent in the structure.
Solution
From Table 7.4, X = 16.4×10–3, XV = 16.4×10–3×5000 m3 = 82 m2
Using modified Sabine’s formula:
A = 320.5 m2
Optimum reverberation time
In general, shorter reverberation times are needed in lecture theatres

intended for listening to speech than in concert halls for listening to
music. Using the empirical method for optimum reverberation time is
that suggested by the following expression:
where
T = reverberation time (s)

V = Volume of the room (m3)
r = 4 for speech, 5 for orchestra and 6 for choir.
Unit 7 37
Example
Determine the reverberation time for a the Great Hall in Hong Kong
to be used mainly for orchestral music and to hold 400 people. The
optimum volume per person (m3) for this type of hall is 7.1.
Solution
You should realize that the volume per person depends on the purpose
for which the building is to be used. A volume of 7.1 m3 per person is
given in this question.
The total volume of the hall = 400×7.1 m3 =2840 m3
where r = 5 for orchestra
= 1.38
Measurement of reverberation time
Reverberation time is measured by first creating a sound of suitable

amplitude and then observing the decay at different frequencies
immediately after the source is switched off. The main difficulty is the
accurate measurement of very small times, for example 0.5 seconds or
less. Narrow-band sound, centered on the frequency of interest, can be
produced through loudspeakers, and the decay after source-off can then
be recorded on a chart recorder or tape recorder. Sometimes it is more
convenient to use a pistol as a source of impulsive broadband noise and
then filtre the decaying sound electronically before recording, as shown
in Figure 7.8. Reverberation times in a typical living room might vary
from 1.5 sec at 100 Hz to 0.8 sec at around 2,000 Hz.
Figure 7.8 Measurement of reverberation using a pistol as sound source
If the reverberation time is measured in a special reverberant room

with a long reverberation time (sometimes known as a reverberation
chamber), before and after the introduction of sound-absorbing material,
then Sabine’s formula can be used to calculate the absorption coefficient
of the material at a certain frequency. This method of measuring
absorption coefficient produces a value that corresponds to sound
arriving at the absorbing surface from all possible angles.
Another method, which measures absorption coefficient for sound

arriving only at right angles to the material surface, is known as the
impedance tube or standing wave tube method. Usually, absorption
coefficients measured by the standing-wave tube method are lower
than those obtained by the reverberation chamber method for the same
material.
Limited by the small sample used, only certain kinds of absorber can
be tested in an impedance tube, porous materials being particularly
suitable. Other kinds of absorber, for example panel absorbers, can be
tested in a reverberation chamber. The panel absorber is a resonating
device.
The formula for the calculation of resonant frequency of a panel

absorber is
where
M = mass per unit area of the panel (kg/m2)

D = depth of the air gap in meter (m).
This formula also applies to panels in the form of suspended ceilings or

closed double windows.
Example
A residential building in Mong Kok found that sealing existing windows
is inadequate to avoid the noise effects from road traffic. The windows
should be double-glazed with internal glass of mass 5 kgm–2. Each
double-glazed window has an air gap of 100 mm and is lined with
acoustic absorbent. Determine the expected resonant frequency.
Solution
M = 5 kgm–2
D = 0.1 m
By
= 85 Hz
Unit 7 39
Sound reduction index
Using panels or walls for reducing acoustic energy from a noise

source is an obvious way to control a noise problem (see Figure 7.9).
With proper design, noise reduction of up to 30 dB can be achieved.
The sound reduction index, R, of a wall or a panel is a measure of its
ability to prevent airborne noise passing through it, which is defined as
where t = transmission coefficient, defined as the ratio
of the sound power transmitted through the panel to the sound power
incident on it.
Figure 7.9 Sound reduction index measurement
Therefore,
in which W1 is the sound power on the source side and W2 is the sound
power on the observed side. Therefore, the sound reduction index is
also known as the sound transmission loss. When one is calculating the
value of t for a panel comprising n elements (e.g. a concrete wall with a
wooden door), the overall value is given by
in which Ai and ti represent the individual areas and transmission

coefficients of the elements of the panel, Atotal and ttotal represent the
total area and coefficient for the complete panel.
The effective value of sound reduction index, Reff for the panel is defined
as
dB
Example
A partition of total area 15 m2 consists of a 250 mm concrete wall and
contains a wooden door of area 3 m2. Assume the mean sound reduction
of the concrete wall is 52 dB and the door 19 dB. Determine the sound
reduction of the composite panel.
Solution
By
Concrete wall:
tconcrete = 10–5.2
Wooden door:
twood = 10–1.9
By equation, with n = 2.
ttotal = 0.0025229
Therefore,
= 26 dB (to the nearest 1 dB)

Unit 7 41
Measurement of sound reduction index
As shown in Figure 7.9, two reverberant rooms are needed for

the measurement of the sound reduction index of a panel. In field
measurement, such reverberant rooms have highly reflective surfaces.
One of the rooms is the source room, and the other is the receiving
room. The field measurement consists of producing a suitable sound on
one side of the test panel, such as an array of loudspeakers in the source
room, and measuring the reduction in sound pressure level at the other
side (the receiving room).
If we call Lsource = average sound pressure level in the source room, and
Lreceiving = average sound pressure level in the receiving room, then the
sound reduction index is given by
in which
S = area of the test panel (m2)

A = absorption in the receiving room (m2).
Requirements for good sound insulation
Mass
The unit of density is mass per volume. In order to compare how dense
the materials are, the same volume should be used. From the table
below, we can see that lead is the densest. This may imply that lead
is the best for insulation. However, the above comment is made from
property of mass only. To achieve complete insulation effectiveness,
more tests should be carried out. For example, the insulation under
different levels of frequency should be examined for better comparison
of different materials as sound insulators.
Density of materials
Materials Density (kg/m3)

Concrete 2300
Glass 2500
Plywood 580
Hardboard 810
Plasterboard 750
Brick 1900
Steel 8100
Lead 11200
For a single partition, the mass law may be obtained approximately from
RAV = 10 + 14.5log10m where

RAV = average sound reduction (dB)
m = mass per unit area (kg/m2).
From the above equation, you should observe that the greater the mass
per unit area, the larger the insulation provided by a partition. The
reason is that the packed molecules on the surface of a material act like
a wall to insulate or reduce the sound energy transmitting to the other
side. There is a limitation of using mass as the insulation strategy. The
reason is that structural considerations may prevent the use of heavy
material as an insulator.
Example
Determine the sound insulation of a concrete wall (100 mm thick) with
mass of 415 kg/m2.
Solution
RAV = 10 + 14.5log10m
RAV = 10 + 14.5log10(415) = 48 dB
This gives an average insulation of about 48 dB.
Completeness
In a real situation, the presence of weak links has a catastrophic effect

on sound insulation; for example, 50 dB insulation of a single-brick-
wide wall would be reduced to 20 dB by a hole that is only 1/100th of
its total area. This suggests obvious ways to try to raise the insulation of
the poorest parts. The first consideration is to ensure that air gaps around
doors and windows are eliminated. However, care should be taken to
ensure that the partition is taken right up to the ceiling; otherwise, a
reduction in the insulation of the partition by 10 dB or more may result.
Multiple leaf construction
Multiple leaf construction involves walls made of two or more layers

of material with separating cavities. This kind of construction is used
for sound insulation in which the individual panels are light and offer
relatively poor insulation individually. For example, a 280 mm cavity
brick wall will offer little acoustic advantage over a 230 mm solid brick
wall for sound insulation.
Unit 7 43
For a lightweight panel, a double partition may be an improvement if

the following requirements are achieved:
• A gap should be large enough, at least 50 mm.

• The two panels should have different superficial weights.
• No air paths should go through the panels.
• The gap is placed with sound-absorbent material.
• The panels are not coupled by the method of construction.
Windows
Double-glazing is common for windows in which if the gap is
appropriate, considerable improvement in sound reduction can be
obtained. In general, the performance of double-glazing should be
better than that of single glazing. Typical insulation curves are shown in
Figures 7.10 and 7.11.
Figure 7.10 Sound insulation curve for single glazing

(Smith, Peters and Owen 1996)
Figure 7.11 Sound insulation curve for double-glazing

(Smith, Peters and Owen 1996)
Floors
Airborne sound transmission is always a design consideration in floor
construction. Conventional wood joist floor construction provides very
little acoustical resistance to impact sound such as from footsteps. The
design of a floor in a room is therefore very important in the sound
insulation. The floor can be made massive with a soft floor finish, or the
floor should be isolated as much as possible from the rest of the building
structure.
A standard method of isolation is the floating floor. This means that

the joists of a wooden floor are isolated from the structure beneath by
a resilient layer such as dense fiberglass. Figure 7.12 shows several
construction details for floors that have popular construction methods
used for controlling impact sound.
a b
c d
Unit 7 45
Figure 7.12 Construction details for floor systems that control impact sound
(Section 1.5, Noise Block, ENVR S335, OUHK)
The characteristics of each design in Figure 7.9 are given below.
Figure 7.12(a). This gives the required insulation for impact and
airborne sound by the following combination of materials: concrete,
plaster and screed together weigh more than 365 kg m-2, and a soft floor
finish is used. If a hard floor finish is used, airborne insulation is still
adequate, but the floor fails for impact sound.
Figure 7.12(b). This gives the required insulation for airborne and impact
sound if concrete and plaster together weigh more than 220 kg m-2.
Figure 7.12(c). This is a floating floor construction giving required

insulation for airborne and impact sound if concrete and plaster weigh
more than 220 kg m-2.
Figure 7.12(d). This gives the required insulation for airborne and impact
sound if concrete and screed together weigh more than 220 kg m-2 and a
soft floor finish is used. If a medium or hard floor finish is used, airborne
insulation is still adequate, but the floor fails for impact sound.
Figure 7.12(e). In this floating floor construction, adequate sound

insulation is possible for both airborne and impact sound, as long as one
of the following sets of conditions is met: heavy lath and plaster with
80 kg m-2 of pugging, or heavy lath and plaster with 15 kg m-2 of
pugging on heavy walls, or heavy lath and plaster with no pugging but
with very heavy walls or plasterboard, and one coat plaster with
15 kg m-2 pugging and very heavy walls.
Noise prediction models
Transportation noise
Although measurement of noise levels will nearly always be the

most desirable method of ascertaining a noise environment, in many
situations, often the most important ones, noise prediction is essential.
Examples are:
• a new railway or roadway is to be constructed through an existing

urban area;
• residential developments and railway or roadway are being planned

concurrently;
• there is an existing transportation noise problem and we wish

to evaluate various strategies such as using barriers, enclosures,
screening structures or special building designs to mitigate road
traffic or rail noise affects sensitive receivers;
• noise levels need to be evaluated for future traffic conditions.
Road traffic noise
A stream of vehicles on a roadway can be regarded acoustically as a

distributed line source. The noise level at a point beside the stream
is then dependent on the strength of the source, the path length from
source to receiver, and any excess attenuation during propagation.
‘Calculation of Road Traffic Noise’ (CRTN), published by the
Department of Transport, Welsh Office, UK, is a memorandum
describing the procedures for calculating road traffic noise. The CRTN
methodologies are widely used in many countries and are accepted by
EPD as the basis of road traffic noise prediction. The variables likely to
affect source strength and propagation are listed below.
Noise source strength variables:
• traffic flow
• vehicle speed
• % of heavy vehicles
• roadway gradient
• roadway surface (e.g. bituminous or concrete surfaces)
• traffic conditions (e.g. free flow, acceleration etc.).
Unit 7 47
Propagation variables:
• distance
• ground cover (e.g. soft or hard)
• height of noise emissions from vehicles
• height of propagation above ground surface
• screenings (e.g. buildings, walls or purpose-built noise barriers)
• reflections (e.g. façade effect or opposite facade effect)
• meteorological conditions.
The CRTN model requires that all source roadways likely to contribute
to noise levels at the prediction site be divided into straight segments so
that noise generation characteristics and propagation conditions from
source to receiver are constant for all points along the segment. When
the source and propagation variables are known for each segment, the
noise level can be calculated at the prediction point for an ‘infinite
roadway’ passing through the segment. This ‘infinite roadway’ level
is then corrected depending on the horizontal angle subtended at the
receiver by the segment. The sound energy contribution from each
segment is then summed to predict the total A-weighted L10 at the
prediction site. The following is a precision of the prediction model.
a Segment the roadway as described above.
b For each segment, calculate the basic noise level for the given
total traffic flow. Use the following equations to calculate L10 (1h) or
L10 (18h) where L10 (1h) and L10 (18h) mean the noise level exceeded for
10% of the one-hour period and 18-hour period at peak traffic flow
respectively:
L10 (1h) = 43.2 + 10 log10 q dB(A)
L10 (18h) = 28.1 + 10 log10 Q dB(A)
where q = vehicles/hour and Q = vehicles/18 hour.
c Calculate the correction for the mean traffic speed and the
percentage of heavy vehicles in the traffic stream. You should add
this correction to the basic noise level. You can either use
Figure 7.13 or the equation of the curves below:
where V = mean traffic speed and p = % of heavy vehicles.

Figure 7.13 Correction for mean traffic speed and % heavy vehicles (‘Calculation
of Road Traffic Noise’ 1988)
d Calculate the correction for roadway gradient, G (%) from the

equation below, and add this to the basic noise level.
Correction = 0.3×G dB(A) where G = gradient (%)
e Calculate the correction for distance and add this to the basic noise
level. The distance is the perpendicular distance from the point to
the ‘infinite roadway’ passing through the segment. For greater than
50% ‘hard ground’, use Figure 7.14. For greater than 50% ‘soft
ground’, use Figure 7.15.
Unit 7 49
Figures 7.14 and 7.15 Correction for distance over hard ground (Figure 7.14)
and over grassland (Figure 7.15) as a function of the
horizontal distance from the edge of the nearside
carriageway and height above the ground — valid for
distances greater than 4 m (Unit 12, T234/T237, OUHK)
f Calculate the correction for angle of view to the roadway from the
equation below and add this to the basic noise level.
dB(A)
where θ = angle of view (degree).
g To ascertain the reduction in noise level afforded by a barrier

beside an ‘infinite’ segment, a cross-section is drawn along the
perpendicular from the road to the prediction point (see
Figure 7.16). The path length distance between the first diffracted
ray (a + b) and the direct ray — assuming the barrier was not
present — (c) is calculated. Using the appropriate part of the

diagram (depending on whether the prediction point is in the shadow
zone or in the illuminated zone), the potential barrier correction for
the segment can be estimated. Then you can add this correction to
the basic noise level. You should note that if shielding is present,
then the distance attenuation for ‘hard’ ground must always be used
no matter what the ground type.
Figure 7.16 Potential barrier correction as a function of path difference

(‘Calculation of Road Traffic Noise’ 1988)
h Combine the sound levels from any two segments. This total can
then be combined with a third segment, and so on, to obtain the total
contribution of all segments.
i This method predicts levels in the free field. If levels are to be

predicted in the reverberant field, at 1 m from the façade of a
building, an additional 2.5 dB(A) is added to the predicted level.
If buildings are present along the side of the roadway opposite the
prediction point, an additional 1 dB(A) is added to the predicted
level.
Rail traffic noise
Two major sources of the noise generated by a rail system are from the
passenger cars and the locomotive. Figure 7.17 shows the A-weighted
sound pressure level against the time duration of a passing train. It
shows that the locomotive noise is not contribute from speed, whereas
car noise is directly related to both the train speed (the faster the
Unit 7 51
speed, the higher the noise) and the train bypass time (the longer the
bypass time, the higher the noise exposure). Since the train has lower
bypass times at high speeds, the two factors related to car noise tend
to compensate each other. Consequently, the train speed is omitted in
predicting rail traffic noise.
Figure 7.17 Typical noise generated by a passing train showing the engine peak
level followed by the mean car noise level (Cunniff 1977)
The distance from the track centreline to the point, the number of
daytime operations, Nd, and the number of nighttime operations, Nn, are
required parameters for predicting the day-night average sound level,
Ldn from the operations of a rail system. The numbers, Nd and Nn, would
be based on the mean daily level of activity over a normal working
week. Seasonal variations can be included by dividing the total yearly
activity by 365 days in arriving at Nd and Nn.
Figure 7.18 Relationship between Ldn, distance from track, and equivalent
number of daily operations (Cunniff 1977)
Having established Nd and Nn, the equivalent number of operations, N,

can be calculated as follows.
N = Nd + 10Nn
Having N, enter Figure 7.18 at the prescribed distance and proceed

vertically to the appropriate N contour. Move across horizontally to
read the predicted Ldn. If necessary, refer to Table 7.5 to account for the
variables listed. For multiple occurrences of these variables, only use
the larger of the adjustment values.
Table 7.5 Adjustments to Ldn Noise Contours (Cunniff 1977)
Corrections to desired
Variables affecting noise output
Ldn value (dB)
Passenger trains only (if combination of passenger and freight — -1
assume all freight)
Presence of helper engines:
Level grade or descending grade 0
Ascending grade +2
Mainline welded or jointed track 0
Low-speed classified jointed track +4
Presence of switching frogs or grade crossings +4
Tight radius curve
Radius less than 600 ft +4
Radius 600 to 900 ft +0.5
Radius greater than 900ft 0
Presence of bridgework
Light steel trestle +14
Heavy steel trestle +5
Concrete structure 0
Construction noise
Construction noise was a major problem when the EPD was set up
in 1986. Pile drivers used to operate 12 hours a day in urban areas,
affecting one in 12 people. The Noise Control Ordinance came into
effect in 1989 and included controls on construction noise that have
been progressively tightened. Nowadays, piling is limited to three to
five hours a day in built-up areas; quieter piling equipment must be
used, and other forms of noisy activities are controlled.
Construction noise remains a problem in Hong Kong. The EPD is trying

to solve the problem and is working to encourage greater compliance
among builders, who tend to treat fines for noise offences as a cost of
doing business.
Noise from construction mainly comes from the equipment used, for
example electric tools for drilling and dressing stone and concrete,
Unit 7 53
compressors, pneumatic tools, pump, rollers, concrete mixers. Noise

coming from construction is restricted by regulation. The law designates
some times of the day for construction work to take place.
Every construction site in Hong Kong has to get a construction noise

permit before starting operation. The details of Noise Control Ordinance
can be found in the Bilingual Laws Information System (http://www.
blis.gov.hk/eng/home.htm) or a concise guide to the noise control
ordinance provided by the EPD (http://www.epd.gov.hk/epd/sites/
default/files/epd/english/environmentinhk/noise/guide_ref/files/CG_
E-06n.pdf)
‘A Concise guide to the Noise Control Ordinance’ (Noise from

construction site) (http://www.epd.gov.hk/epd/sites/default/files/epd/
english/environmentinhk/noise/guide_ref/files/CG_E-06n.pdf)
Under the Ordinance, construction activities are grouped into two

categories: general construction work and percussive piling (for
example, piling by a hydraulic hammer or a drop hammer). Each of
these categories of work is controlled by a system of Construction Noise
Permits, as shown below.
On 17 August 1989, the carrying out of general construction work using

powered mechanical equipment during the restricted hours, that is
between 7 pm and 7 am or at any time on a general holiday (including
Sunday), is prohibited under the Ordinance unless a valid Construction
Noise Permit is in force.
From 1 November 1996 during the restricted hours in Designated

Areas, the use of specified powered mechanical equipment (for
example, hand-held breakers and dump trucks) and/or the carrying
out of the prescribed construction activities (for example, erection or
dismantling of formwork and hammering) is subject to more stringent
control. The same system of Construction Noise Permits for controlling
powered mechanical equipment is used. The Designated Areas, that is
densely populated built-up areas, are defined under the Noise Control
(Construction Work Designated Areas) Notice. The Notice came into
operation on 1 May 1996.
Carrying out percussive piling is prohibited between 7 pm and 7 am

and on holidays, unless specifically exempted by an order made by
the Chief Executive in Council. As of 17 November 1989, percussive
piling during the daytime may only be carried out in accordance with
a Construction Noise Permit. To tighten the control, the use of noisy
diesel, pneumatic and steam hammers for percussive piling is banned in
built-up areas surrounded by noise-sensitive uses from 1 October 1999.
An application for a Construction Noise Permit for the two categories

of work must be made to the Noise Control Authority in a respective
prescribed form and accompanied by a cheque for the prescribed fee. To
assess whether an application for carrying out general construction work
and percussive piling is permissible, the Authority will estimate the
effect of the noise generated by the equipment at any Noise Sensitive
Receiver (e.g. domestic premises) in the vicinity in accordance with

the assessment procedures in the Technical Memoranda of the Noise
Pollution Control Ordinance (see Section 4 of ‘A concise guide to the
noise control ordinance’ of the EPD website).
The conditions of a Construction Noise Permit for carrying out

percussive piling are subject to appeal (see Section 9 of A Concise
Guide to the Noise Control Ordinance of the EPD website).
In the previous section on transportation noise, sound propagation and

attenuation, we understand that if the sound power level of the source
(Lw dB) is known, then the sound pressure level, Lp at a distance r from
the source may be predicted by the formula:
Lp = Lw – 20logr – 11
Assumption for the above equation is that the source is in free space
or that it is above a perfect sound absorber. If the source is near hard
reflecting ground, the amount of energy received is doubled (because
half of the sound which was directed downwards by the source is now,
after reflection, directed upwards and adds to that directed upwards in
the first place). Therefore, sound pressure levels are increased by 3 dB,
and 8 replaces the number 11 in the equation above.
In addition to the above equation, the following is useful in calculating

the sound pressure level (L2) at any other distance (r2), provided a sound
pressure level (L1) at any one distance (r1) is known:
The factor of 20 in the above equation is used for a point source only.
You should realize that the factor is equal to 20 only when the drop-
off rate is 6 dB per doubling of distance. The value of the factor gives
the number of dBs reduction per decade (i.e. a factor of 10) change of
distance. This equation still is valid if there is a power law relationship
between the intensity and distance, even when the power is not 2 (that
is the inverse square law). However, the replacement of 20 by 10 is for
a line source only, which is equivalent to a drop-off rate of 3 dB per
doubling distance.
Example
Noise measurement has shown that noise from a water pump drops off
with distance at a constant rate of 8 dB per doubling of distance when
travelling over an area of grassland in the Hong Kong Stadium. The
sound pressure level at 20 m from the pump is 60 dB. Determine the
sound pressure level at a distance of 30 m from the pump.
Unit 7 55
Solution
From equation , let n = 20, the equation will become
then, L1 – L2 = 8 and =2
therefore, 8 = n log(2)
n = 11.54 dB per decade of distance
Hence,
L1 = 60
r1 = 20
r2 = 30
L2 = 58.0
Therefore, the level at 30 m is 60 dB.
Example
The noise level from a construction site in Causeway Bay arises from
the operation of the powered mechanical equipment: a power rammer, a
rock drill and a scraper.
Calculate the Leq over 1 hour at a point on the construction site

boundary, from the data given.
As the construction site is located in Causeway Bay, we can assume

‘hard ground’ propagation condition.
Key:
Lw = A-weighted sound power level

r = distance of noise source from reception point
T = total measurement time.
Equipment Lw (dB) r (m) ‘on time’ (minutes per hour)

Power rammer 105 81 40
Rock drill 112 52 22
Scraper 119 60 28
Solution
By equation
Lp = Lw – 20logr – 8
Power rammer: Lp = 105 – 20log81 – 8 = 58.8 dB
Rock drill: Lp = 112 – 20log52 – 8 = 69.7 dB
Scraper: Lp = 119 – 20log60 – 8 = 75.4 dB
By equation
= 73.0 dB
Industrial noise
In a factory or a workplace, it is very common to move large quantities
of air or convey products by fans and blowers. To meet the air- and
product-handling requirements, the various kinds of fans and blowers as
shown in Figure 7.19 are most often used. Fans are used to move large
volumes of air for ventilation, for example drying operations, which
are low-speed low-static-pressure units. Blowers are high-pressure
rotary positive displacement type that would better be described as
compressors.
Fans
Centrifugal fans
There are three basic types of centrifugal fan — backward, forward

curved, and radial. Noise generated from centrifugal fans mainly comes
from superposition of discrete tones at the impeller or blade passing
frequency and broadband aerodynamic noise.
Unit 7 57
Backward curve Forward curve Radial blade
Figure 7.19 Examples of centrifugal fans (Bell 1982)
The frequency (octave band centre frequency) of discrete tones is

estimated by
where
n = harmonics, n=1 for fundamental, n=2 for second harmonic and so on

B = number of blades
N = fan rotational speed (rpm).
The occurrence of discrete tones is from two main sources. First, when
a blade in a centrifugal fan passes a point in space, a pressure is created
because of the displacement of air. Second, when the blade passes through
the cut-off point in the scroll, abrupt pressure changes also occur at the
blade passing frequency and higher integer-ordered harmonics.
There are some guidelines to predict the amplitude of the discrete noise
coming from the fan:
1 The cut-off clearance is very crucial (the clearance between the

blade and scroll at the cut-off point). A clearance of 5% to 10% of
the wheel diameter is suggested as an optimum.
2 Backward-inclined blades are always quieter than forward-inclined

blades.
Figure 7.20 Noise generation at the scroll cut-off point (Bell 1982)
The equation for the prediction of noise levels for average sound power
level in the range of 500 to 4000 Hz (octave frequency) is
L = 10logQ + 20logPstatic + K where
L = power sound level (dB)

Q = volumetric flow rate (cfm)
Pstatic = static pressure (in. H2O)
K = a constant depends on fan type, e.g. forward and backward curved
blades K = 35
radial types = 43.
Example
In order to maintain a sufficient fresh air supply rate for an air-
conditioned workplace located in Kwun Tong Industrial Estate, a
centrifugal fan (forward curved) with a static pressure of 2 in. H2O is
installed, which produces an airflow rate of 6,000 cfm. Calculate the
average sound power level for the fan.
Solution
K = 35
L = 10log(6000) + 20log(2) + 35
= 78.8 dB.
Axial fan
In the axial fan, the airflow is along the axis of the fan. In order to
avoid a circular flow pattern and increase performance, guide vanes are
installed downstream of the rotor. The name of the axial fan with exit
guide vanes is vane axial; those without vanes are called tub axial.
Unit 7 59
Figure 7.21 A belt-driven vane axial fan
In general, axial fans operate at higher pressures than centrifugal fans

do and usually are considerably noisier. The common uses of axial fans
include heating, ventilating and air-conditioning systems.
Due to the number of blades and relatively high rotational speeds,

the noise from axial fans is better estimated with the consideration of
discrete blade passing tones. The equation of estimation is same as the
equation used for centrifugal fans, except that the K value is now 48 and
the equation is valid for four octave bands: 500, 1000, 2000, 4000 Hz.
L = 10logQ + 20logPstatic + 48
Gas jet noise
Another most common noise source in industrial environments is the

gas jet. The basic noise mechanism is referred to as aerodynamic noise.
Some examples include blow-off nozzles, steam valves, pneumatic
control discharge vents, and oil burners. The noise source from gas jets
is often in the range of 100 to 110 dB(A).
One common example of a gas jet is the high-velocity airflow emanating

from a reservoir through a nozzle. The gas accelerates from near close
to zero velocity in the reservoir to the peak velocity in the core at the
exit of the nozzle. The peak velocity of a gas jet is strongly dependent
on the pressure difference between the reservoir pressure Pr and the
external ambient pressure Pa. For example, when the pressure ratio
Pr/Pa is increased, the velocity of the gas at the discharge nozzle
increases. However, when the pressure of the ratio hits about 1.9, the
flow velocity through the nozzle becomes sonic, that is, it reaches
the speed of sound. Further increase in reservoir pressure does not
significantly increase the flow velocity anymore. When the critical
pressure ratio of 1.9 is reached, we say that the nozzle is choked.
In reality, the noise generation from gas jets results from the creation
of fluctuating pressures due to turbulence and shearing stresses as the
high velocity interacts with the ambient gas. Complex radiating sources
called eddies are formed from the interaction with the high-frequency
noise being generated near the nozzle in the mixing shearing region and
the low-frequency noise being generated downstream in the region of
large-scale turbulence. With only turbulence flow present, the pressure
fluctuations are random functions of space and time, and only statistical
methods can be used to describe the character of the noise. So, the
spectral character of gas-jet noise is generally broadband.
Figure 7.22 Octave band spectral analysis of a part ejection nozzle at

approximately 1m (Lewis H Bell and Associates 1982)
Figure 7.22 is an octave band spectral analysis of an air jet from a

0.6 cm diameter copper tube (nozzle) crimped to blow parts from the
die area of a power press. Since the reservoir pressure is usually in the
range of 40 to 95 psia, the gas nozzle can be considered choked and
the peak velocity in the core of the jet near Mach = 1. The character of
noise is a broadband hiss with peak levels near 101 dB in the octave
whose centre is 2000 Hz.
The magnitude and spectral character of the noise from jets cannot be
accurately predicted to random pressure fluctuations. In addition are
the complex nature of the jet itself and uncertainties associated with
factors like temperatures, turbulence, nozzle types, etc. For the sets
of empirical data generated dominantly by the aerospace industry,
first-order estimates of the acoustical power spectral character can be
obtained. The overall sound power W from a subsonic or sonic jet can
be calculated by the equation:
Unit 7 61
where
W = overall sound power (W)

K = a constant of proportionality (with no unit)
ρ = density of ambient air (kg/m3)
A = area of the jet nozzle (m2)
V = jet flow velocity (m/s)
C = speed of sound.
Because of turbulence flow of the gas from the gas nozzle, the V in
the above equation is a fluctuating velocity and varies throughout the
jet stream. It is quite difficult to measure or calculate the value of V.
It is simpler to use average velocity and assume the energy carrying
eddies are of the same size as the jet diameter. Also, the total radiated
acoustical power W is proportional to the kinetic energy of the jet
stream. With the above assumption, the total radiated power is then a
fraction of the total power discharged from the nozzle. Therefore, the
radiated power W is
(W)
where
V = average flow velocity through the nozzle (m/s)

M = mach number of flow (V/c) (with no unit)
ρ = density of ambient air (kg/m3)
A= nozzle area A (m2)
e = constant of proportionality of the order of 10–4.
The above expression is first-order approximation and is applicable

to many industrial situations in which average velocity of the jet is in
the range of 0.15c < V ≤ c, where c is speed of sound in the jet exhaust
medium. The factor eM 5 can be considered an efficiency factor and
found from empirical data (see Figure 7.20).
Figure 7.23 Gas-jet radiation efficiency factor and range of Mach numbers
(Lewis H Bell and Associates 1982)
From Figure 7.23, the efficiency factor is given as a range, which

reflects the uncertainty associated with turbulence, temperature, etc.
Example
Estimate the overall acoustical power of W of a choked 0.6 cm diameter
air jet exhausting to the atmosphere. Calculate also sound power level
L w.
Solution
Since the nozzle is choked, it is wise to take the average velocity of the
nozzle to be the speed of sound, that is 344 m/s. The density of air is
approximately equal to 1.18 kg/m3.
The area of the nozzle A =
= 3.16×10–5m2
From Figure 7.23, the radiation efficiency factor (eM 5) for Mach 1 is
eM 5 = 2×10–5 (use the centre of the range).
Unit 7 63
By equation
= 0.0152 W
The equation linking sound power level and acoustical power is given by
= 101.8 dB.
When the gas is hot and highly turbulent, such as from a gas burner,
another first-order correction can be applied:
Temperature correction =
where
Tg = absolute temperature of the gas jet (unit in oR)

Ta = absolute temperature of ambient air (unit in oR).
Example
What is the total acoustical power level Lw for the choked 1/4 in.
diameter air jet of the above example if the temperature of the jet is
raised from 60oF to 400oF?
Solution
The power level Lw for 60oF was calculated to be 101.8 dB.
Tg = 460+400 = 860oR
Ta = 460 + 60 = 520oR
The correction for a temperature is given by

temperature correction =
= 4.4 dB
The temperature-corrected power level is then
L = 101.8 + 4.4 = 106.2 dB.
Probably, the air jet is considered a point source with typical

directionality as shown in Figure 7.24 below.
Figure 7.24 Typical directivity pattern for a small subsonic gas jet
(Lewis H Bell and Associates 1982)
From Figure 7.24, the peak level occurs in the angular range of 15o to
45o from the axis of the jet. The relative sound pressure level as given in
the figure can be considered the directivity index DI term.
For better understanding of the following example, some equations are

quoted for reference.
The equation linking sound pressure level and sound power level
Lp = Lw – 20log(r) – 11
where
Lp = sound pressure level

Lw = sound power level of point source.
Unit 7 65
To account for the directionality, a term DI must be added to the basic

equation above to its generality.
Lp = Lw + DI – 20log(r) – 11
Example
Calculate the radiated overall sound pressure level at a radial distance
of 8m from the nozzle of the above example at angular positions of 0o,
45o, 90o and 150o.
Solution
From the previous example,
the overall sound power level = 101.8 dB.
The sound pressure level Lp at 8 m is
Lp = 101.8 + DI – 20log(8) – 11
= 72.7+ DI.
Then, applying the directivity correction from figure at 0o, we found that
DI = 0 dB, so the sound pressure level Lp is
Lp = 72.7 + 0 = 72.7 dB
At 45o, DI = 4 dB
Lp = 72.7 + 4 = 76.7 dB
At 90o, DI = -5 dB
Lp = 72.7 – 5 = 67.7 dB
At 150o, DI = -9 dB
Lp = 72.7 – 9 = 63.7 dB.
Hydraulic pumps
The four basic types of hydraulic pump that are commonly found in
industrial applications are centrifugal, screw, reciprocating and gear.
The theory of pumping is that discrete parcels of fluid are taken in at
the inlet, compressed and recombined at discharge. Ideally, the flow
would be steady and have no fluid or pressure pulsations. In reality,
the fluid flow and discharge pressures are not steady but contain
periodic components because of the pump compression mechanism.
The compression includes pistons, gears, screws, etc. Most noise comes
from these pulsations. Additional broadband noise is often present due
to mechanisms such as cavitations, turbulence, etc. The broadband
sources are usually secondary in magnitude and related to poor pump

design.
For periodic pulsation, the frequencies of the discrete tones fn can be

calculated from the equation:
fn = nPrN
n = harmonic number, 1, 2, 3
Pr = number of compressions or pumping events per revolution
N = pump rotational speed (r/s).
The above equation can apply for fans, gears, internal combustion
engines, etc. One point to note is the determination of the number
of compressions or pumping events. Some pump manufactures will
combine the fluid delivery with several compression units in order to
minimize the flow pulsation.
For example, the flow delivery of pairs of cylinders may be combined,

yielding a fundamental tone (n = 1) at half the frequency calculated by
just counting the cylinders.
Many different kinds of pump are available in the market. Again, the
first approximation given by Irwin and Graf shows that the overall
acoustical power Lw in the four octave bands 500, 1000, 2000 and 4000
Hz can be estimated as follows:
Lw = 10log hp + Kp
Kp is the pump constant, 95 dB for centrifugal, 100 dB for screw and

105 dB for reciprocating pumps (below the speeds of 1600 rpm, the Kp
value should subtract 5 dB). Moreover, under the assumption of equal
energy in each band, the individual octave band levels can be taken as
6 dB less than the overall power level.
Example
A particular 120-hp screw pump operates at 2500 rpm. Estimate (a) the
acoustical sound power level in four octave bands from 500 to 4000 Hz,
and (b) the acoustical sound power level for 1000-Hz octave band.
Solution
a By equation, Kp = 100 dB for screw pump
Lw = 10log hp + Kp
= 10log(120) + 100
= 120.8 dB This is the total for the four octave band.
Unit 7 67
b Because individual octave band levels can be taken as 6 dB less

than the overall power level, the acoustical sound power level for
1000-Hz octave band
= 120.8 dB – 6 dB
= 114.8 dB.
Air compressor noise
Compressors are used extensively in industry. Their role is to raise the

pressure of a gas, and they are usually driven by a motor or turbine.
The noise emission of a compressor is a function of the type of unit.
Diehl points out that the driving engine in portable air compressors is
the major source of noise, not the compressor. Furthermore, a cooling
fan is a noise source. For example, a reciprocating compressor typically
generates a strong low-frequency pulsating noise, which in turn depends
on rotational speed and the number of cylinders. Diehl proves that
the noise generated by a centrifugal compressor is a function of many
more parameters like the house power input, inlet design, turbulence,
rotational speed, the interaction of rotating and stationary vanes,
the radial distance between impeller blades and diffuser vanes, gas
molecular weight and mass flow.
Similar to fans, blade passage frequency is an important frequency

component in certain types of compressor. In diffuser type machines,
blade-rate component (BRC) is of primary importance. The frequency
of this component is determined by equation:
where
fBRC = blade-rate component frequency

Nr = number of rotating blades
Ns = number of stationary vanes
KBRC = greatest common factor of Nr and Ns.
Example
Estimate the frequency of the blade-type compressor of a diffuser
type compressor with Nr = 12 and Ns = 8 which operates at a speed of
6000 rpm.
Solution
KBRC = 4 for 12 and 10
fBRC =
= 2400 Hz.
In this example, fBRC falls outside the frequency range of interest. If fBRC
falls within the range of interest, an increment of several decibels in the
octave band sound power level in which it occurs would be expected.
The total sound power level in the four octave bands 500, 1000, 2000,
4000 Hz may be estimated for centrifugal and reciprocating compressors
by the following equation:
Lw = 10log hp + Kc
where Kc = air compressor constant, 86 dB for hp from 1 to 100.
Figure 7.25 Sound power level estimate for both centrifugal and reciprocating air
compressors (the sound power level for the four octave bands from
500 to 4000 Hz) (Irwin and Graf 1979)
Again, the estimation of sound power level is equally divided among the
four octave bands. Each band level is 6 dB below the total determined
by the equation (Lw = 10log hp + Kc).
Unit 7 69
Example
A centrifugal air compressor is rated at 60 hp. Estimate the sound power
level in the four octave bands from 500 Hz to 4000 Hz and in each of
the four bands individually.
Solution
Using the equation,
Lw = 10log hp + Kc
Lw = 10log 60 + 86
Lw = 103.8 dB, for four octave bands.
Then in each of the four octave bands,
Lw = 103.8 – 6
= 97.8 dB.
Basic principles of vibration
Introduction
Vibration is structure-borne noise. That means vibration is transmitted
through structure rather than through air. Sound, in contrast, is
transmitted through air (medium with atoms). This also explains why
sound can’t transmit through a vacuum. Sound transmission requires a
medium. Sometimes, sound can also be produced by vibration.
Vibration is a repetitive, periodic, or oscillation of a mechanical system.

The rate of the vibration cycles is termed frequency. Frequency is
the number of times a complete motion cycle takes place during the
period of one second (the unit of frequency is Hz). Repetitive motions
that are somewhat clean and regular, and that occur at relatively low
frequencies, are commonly called oscillations. Any repetitive motion,
even at high frequencies, with low amplitudes, and having irregular and
random behaviour falls into the general class of vibration.
The mechanical system in which motion takes place can consist of

a single component occurring at a single frequency, as with a tuning
fork, or of several components occurring at different frequencies
simultaneously, for example the piston motion of an internal combustion
engine of a car. In reality, the vibration signals usually consist of many
frequencies occurring simultaneously so that we cannot immediately
see just by looking at the amplitude-time pattern graph how many
components there are and even at what frequencies they occur.
The components can be observed by plotting a graph of vibration

amplitude against frequency. The breaking down of vibration signals
in individual frequency components is called frequency analysis.
Normally, frequency analysis is a cornerstone of diagnostic vibration
measurements. Vibrations coming from different equipment or set-
ups can have different properties. For example, vibration generated by
a pump (constant liquid flow with pure liquid) is a kind of periodic,
since the pump is rotating in the same cycle repeatedly. If the liquid
containing solid as impurities is transmitted through a pump, the
resulting vibration from the pump will not be periodic. The solid in
the liquid alters the regular cycling of the pump and this in turn affects
the vibration. The plot of vibration amplitude against frequency is the
frequency spectrogram. Using the above frequency analysis, we are able
to find the source of undesirable vibration of a particular instrument.
Unit 7 71
Where does the vibration come from?
Actually, vibration occurs everywhere at levels that are normally

undetectable. To be more specific about the source of vibration, it
usually occurs due to the dynamic effects of manufacturing tolerance,
clearances, rolling and rubbing contact between machine parts and
the imbalanced force in rotating and reciprocating members. Often,
small insignificant vibrations can excite the resonant frequencies of
some other structural vibration and noise sources. These so-called
insignificant vibrations can therefore become important.
Every coin has two sides. Sometimes, mechanical vibration performs a

very important role. For example, we generate vibration intentionally
in component feeders, concrete compactors, ultrasonic cleaning baths
for jewellery or glass, rock drills and pile drivers. Vibration testing
machines are used extensively to impart a controlled level of vibration
energy to products and sub-assemblies that require examining their
physical or functional response and ascertaining their resistibility to
vibration environments. The use of this vibration testing is for the
design of building structures especially in countries like Japan, where
there is the constant danger of an earthquake.
Only through accurate measurement can the use and control of vibration
be done smoothly.
How to quantify the vibration level
The vibration amplitude, which is the characteristic that describes the

severity of the vibration, can be quantified in several ways. Details are
shown in Figures 2(a) and 2(b), Section 2.2, Unit 11 of T234/T237.
In order to compare two frequency graphs from a machine, some

parameters can be used. The relationship between peak-to-peak level,
the peak level, the root mean square (RMS) level and average level of a
sine wave is summarized below:
1 Peak-to-peak value indicates the maximum excursion of the wave.

It is a very useful quantity; for example, the vibratory displacement
of a machine part is critical for maximum stress or mechanical
clearance considerations.
2 Peak value indicates the level of short-duration shock and indicates

only what maximum level has occurred. This value doesn’t give any
information about the time history of the wave.
3 Average value takes into consideration the time history of the

wave. However, this value is considered of limited practical interest
because it has no direct relationship with any useful physical
quantity.
4 Root mean square, RMS value, is the most relevant measure of

amplitude because it both takes the time history of the wave into
account and gives an amplitude value which is directly related to the
energy content and so the destructive abilities of the vibration.
5 Period is the time required to complete one cycle. That is, the
amplitude of the wave reaches a positive peak from the time axis,
then goes back through the rest position (the time axis), reaches a
negative peak, and returns to the rest position.
Measuring amplitude
After going through the basic terminology used to describe vibration

level, it is useful to know how these parameters were measured.
We can also measure amplitude from the time waveform as shown in

2(a), Section 2.2, Unit 11 of T234/T237.
Unit 7 73
Refer to the time axis, at time equals 0. That is the ‘at rest’ reference
position. The period (T is the symbol) of the waveform is measured in
seconds.
In general, the severity of a vibration problem can be described by

amplitude. The three ways of measuring amplitude are shown below:
1 Displacement measures the distance the shaft moves in relation to a

reference point. It is measured by peak-to-peak measurement.
2 Velocity measures the displacement of the shaft in relation to time.

The maximum speed is the peak measurement.
3 Acceleration measures the total force required to move the vibrating

element in the opposite direction. The total force is measured from
the waveform by the amount of energy under the curve or the RMS
value.
Measuring frequency
We can measure frequency, which is defined as how often the wave

moves. The waveform shown in Figure 2(b), Section 2.2, Unit 11 of
T234/T237 is one cycle. Frequency (f) generally can be calculated by
the time required for one cycle. The relationship between frequency and
period is frequency is the inverse of the time period.
Example
From the following figure, please find the amplitude, period and
frequency.
x against time
1.5
0.5
0
x
0 1 2 3 4 5 6 7 8 9 10 11
-0.5
-1
-1.5
time (second)
Solution
Amplitude = 1 unit
Period = 6.3 sec
Frequency =
Vibration analysis overview
Theoretical vibration profiles
A harmonic function is the simplest type of periodic motion. It is shown

in Figure 7.21, which is the harmonic function for small oscillation of a
simple pendulum. For example
X = X0sin (ω * t)
where
X = Vibration displacement (thousandths of inch or mils)

X0 = Maximum displacement or amplitude (mils)
ω = Circular frequency (radian per second)
t = Time.
Figure 7.26 Small oscillation of a simple pendulum, harmonic function
where T = period
A real example of simple harmonic motion is the motion of a mass on

a string and that of the shadow of a revolving peg. Both motions repeat
in a time, T, period. The frequency f again equals 1/T. The position,
velocity and acceleration of the shadow are
X = A’ cos (2 π f t)
V = - 2 π f×A’ sin (2 π f t)
A = - 2 π f(2 π f)2 A’ cos (2 π f t)
Unit 7 75
where t = time, A’ = amplitude
X and A are related by

A = –(2 π f)2 X
From the above equation, acceleration is proportional to the negative of

the displacement.
For spring, 2 π f =
where
k = spring constant
m = mass.
Example
An 8 kg mass attached to the end of a spring has a maximum
displacement of 0.1 m at t = 0. The spring constant is 100 Nm–1. Find
the position, velocity and acceleration of the mass as a function of time.
Solution
The amplitude A = 0.1 m
By equation
2πf =
=
= 10 s–1
X = A’ cos (2 π f t) = (0.1 m) cos (10 s–1)t
V = -2 π f×A’ sin (2 π f t) = -10 s–1(0.1 m) sin (10 s–1)t

= – (1 ms–1) sin (10 s–1)t
A = -2 π f(2 π f)2 A’ cos (2 π f t) = - (10 s–1)2 (0.1) cos (10 s–1) t

= – (10 ms–1) cos (10 s–1) t
Activity 7.6
Now, let us carry out the following desk-top activities to test your
knowledge about this section.
1 Determine the frequency if the motion of a mass on a spring repeats

itself each 3 seconds.
2 What is the acceleration of a mass on a spring when the position is

0.1 m? The frequency is 4 Hz.
3 A 5 kg mass on a string constant 20 Nm–1 undergoes simple

harmonic motion. (a) What is the frequency? (b) What is the period?
4 What mass on a spring of spring constant 10 Nm–1 will result in

simple harmonic motion with a frequency of 0.5 Hz?
I’m sure from doing the above calculations you are able to integrate
your knowledge and to find out their applications.
Actual vibration profile
Figure 7.23 is quite smooth and easily described by periodic function.

But in reality, the profile looks very complex in shape.
The process of vibration analysis requires the gathering of complex

machine data, which must then be deciphered. This is understandable,
since there are usually many sources of vibration. Each source generates
its own curve of waveform, and these waveforms of different sources
are added, and displayed as a composite profile. Basically, these profiles
can be displayed in two formats, ‘time domain’ and ‘frequency domain’.
Time domain
Vibration data, plotted as amplitude versus time, are referred to as a

time-domain data profile. The following figure shows the complexity of
these types of data for a piece of industrial machinery.
Figure 7.27 Example of a typical time-domain vibration profile for a piece of

machinery (Mobley 1999)
Unit 7 77
Figure 7.27 shows the actual time traces or time plots. That is the
change in amplitude of frequency of the machine as time passes. Time-
domain profiles are the sum of all vibration components (frequencies,
impact etc.) that are present in the machine-train and its installation.
Time traces usually include all frequency components, but the
individual components are more difficult to isolate than with frequency-
domain data.
The use of the time-domain plot is for all linear and reciprocating
motion machinery. They are very useful in the overall analysis of
machine-trains to study changes in operating conditions. However, time-
domain is difficult to use. The problem of the graph is that it shows all
the vibration data to represent the total displacement at any given time;
it is difficult in determining the contribution of any particular vibration
source.
There is another disadvantage for this type of graph. Since the above
data were taken during the lifetime of a machine, the graph plotted was
used to compare the result of historical data at exactly the same running
speed and loads. However, this is not practical because of vibrations
in day-to-day plant operations environment and changes in operating
parameters. This significantly affects the profile and makes it impossible
to compare historical data.
Frequency domain
In practice, simple harmonic vibration functions are related to the

circular frequencies of the rotating or moving components. Frequencies
are the production of the running speed of the machine, which is
expressed in revolutions per minute (rpm) or cycles per minute
(cpm). Therefore, for this domain, the basic information needed is the
calculations of these frequencies and then the analysis of the operating
condition of the machine.
Frequencies-domain data are obtained by simply converting time-

domain data using a mathematical technique referred to as a fast Fourier
transfer (FFT). FFT allows each vibration component of a complex
machine-train spectrum to be shown as a discrete frequency peak. The
frequency-domain amplitude can be the displacement per unit time
related to a particular frequency, which is plotted as the y-axis against
frequency as x-axis. This is opposed to the time-domain spectrum,
which sums the velocities of all frequencies and plots the sum as the
y-axis against time as the x-axis. An example of a frequency-domain
plot or vibration signature is shown below.
Figure 7.28 Typical frequency-domain vibration signature (Mobley 1999)
Frequency-domain data are required for equipment operating at more

than one running speed and all rotating applications. The x-axis of the
spectrum is frequency normalized to the running speed, so a change
in running speed will not affect the plot. A vibration component that is
present at one running speed will still be found in the same location on
the plot for another running speed after the normalization, although the
amplitude may be different.
How to interpret the vibration data
The main idea in using vibration signature analysis for predictive

maintenance, diagnostic and other application is the ability to
differentiate between normal and abnormal vibration profiles. Many
vibrations are normal for a piece of rotating or moving machinery.
Examples of these are normal rotations of shafts and other rotors,
contact with bearings or gear-mesh and so on. However, specific
problems with machinery generate abnormal but identifiable vibrations.
Examples of these are loose bolts, misaligned shafts, worn bearings,
leaks and incipient metal fatigue.
Predictive maintenance using vibration signature analysis is based

on the following facts, which form the basis for the methods used to
identify and quantify the root causes of failure:
a All common machinery problems and failure modes have distinct

vibration frequency components that can be isolated and identified.
b A frequency-domain vibration signature is generally used for the

analysis because it is comprised of discrete peaks, each representing
a specific vibration source.
Unit 7 79
c There is a cause for every frequency component in a machine-train’s

vibration signature.
d When the signature of a machine is compared over time, it will

repeat until some event changes the vibration pattern.
An increase or decrease in amplitude may indicate degradation of the

machine-train; this may not always be held. Variation in load, operating
practices and a variety of other normal changes also generate a change
in the amplitude of one or more frequency components within the
vibration signature. It is also important to note that lower amplitude
does not necessarily indicate an improvement in mechanical condition
of a machine-train. Therefore, it is better to understand the sources of
all amplitude variations.
Vibration measurement equipment
In practice, vibration data are obtained by the following procedure:
1 Mount a transducer onto the machinery at various locations,

typically machine housing and bearing caps.
2 Use a portable data-gathering device, referred to as a vibration

monitor or analyser, to connect to the transducer to obtain vibration
readings.
Transducer
The transducer is most commonly used to obtain vibration measurement

as an accelerometer. It incorporates piezoelectric films to convert
mechanical energy into electrical signals. The device generally
incorporates a weight suspended between two piezoelectric films. The
weight moves in response to vibration and squeezes the piezoelectric
films, which sends an electrical signal each time the weight squeezes it.
Portable vibration analyser
The portable vibration analyser incorporates a microprocessor that

allows it to convert the electrical signal mathematically to acceleration
per unit time, perform an FFT and store the data. It can be programmed
to generate alarms and displays of the data. The data stored by the
analyser can be downloaded to a personal or a more powerful computer
to perform more sophisticated analyses, data storage and retrieval, and
report generation.
Typically, vibration levels of 0.1 inch per second or less (measured in

velocity) are acceptable for many types of rotating machinery. That’s
about 1 mil (measured in Displacement) for machinery operating at
2000 rpm or slower. The Balmac Vibration Severity Chart can help
determine acceptable vibration levels.
Figure 7.29 Portable vibration analyser (http://www.castlegroup.co.uk/Shop/

catalog/product_info.php?products_id=219)
This kit includes a vibration meter, incorporating the latest technological

developments to allow complex vibration measurement and monitoring
to be undertaken. The measurement falls in 10 octave bands. The
continuous monitoring of Lav (linear average value) and interval data
logging is ideal for environmental measurements. The results can be
printed out to a portable printer or downloaded to a computer running
Castle dBdata4W vibration analysis software.
Figure 7.30 Balmac Vibration Severity Chart (http://www.balmac.thomasregister.

com/olc/06685705/faq.htm)
Vibration source
All machinery with moving parts generates mechanical forces during
normal operation. As the mechanical condition of the machine changes
due to wear, there will be changes in the operating environment or load
variation. Understanding machinery dynamics and how forces create
unique vibration frequency components is the key to understanding
vibration sources.
Unit 7 81
Vibration does not just happen. There is a physical cause, referred to

as a forcing function, and each component of a vibration signature has
its own forcing function. The components that make up a signature are
reflected as discrete peaks in the FFT or frequency-domain plot.
The vibration profile that results from motion is the result of a force
imbalance. By definition, balance occurs in moving systems when
all forces generated by or acting on the machine are in a state of
equilibrium. In real applications, however, there is always some level
of imbalance and all machines vibrate to some extent. The following
is a discussion of rotating machinery and machinery undergoing
reciprocating and/or linear motion.
Rotating machinery
A rotating machine has one or more machine elements that turn with
a shaft, for instance, rolling-element bearings, impellers, agitators in
reactor and other rotors. In a perfectly balanced machine, all rotors turn
true on their centerline and all forces are equal. However, in industrial
machinery, it is common for an imbalance of these forces to occur. In
addition to imbalance generated by a rotating element, vibration may be
caused by instability in gases, liquids or even solids flowing through the
rotating machine.
Rotor imbalance
Although mechanical imbalance generates a unique vibration profile,

it is not the only form of imbalance that affects rotating elements.
Mechanical imbalance is the condition in which more weight is on
one side of a centreline of a rotor than the other. In many cases, rotor
imbalance is the result of an imbalance between centripetal forces
generated by the rotation. The source of rotor vibration also can be an
imbalance between the lifting force generated by the rotor and gravity.
Rotating machines are designed to generate vertical lift of the rotating

element when operating within normal parameters. This vertical lift
must overcome gravity to properly centre the rotating element in its
bearing-support structure. However, because gravity and atmospheric
pressure vary with altitude and atmospheric pressure, actual lift may not
compensate for the downward forces of gravity in some environments.
When the deviation of actual lift from designed lift is significant, a rotor
might not rotate on its true centreline. This offset rotation creates an
imbalance and a measurable level of vibration.
Flow instability and operating conditions
Rotating machines subject to imbalance caused by turbulent or

unbalanced media flow include pumps, fans, and compressors. A good
machine design for these units incorporates the dynamic forces of the
gas or liquid in stabilizing the rotating element. The combination of
these forces and the stiffness of the rotor-support system (i.e. bearing
and bearing pedestals) determine the vibration level. Rotor-support
stiffness is important because unbalanced forces resulting from flow

instability can deflect rotating elements from their true centreline, and
the stiffness resists the deflections.
Deviations from a machine’s designed operating envelope can affect

flow stability, which directly affects the vibration profile. For example,
the vibration level of a centrifugal compressor is typically low when
operating at 100% load with laminar airflow through the compressor.
However, a radical change in vibration level can result from decreased
load. Vibration resulting from operation at 50% load may increase by
as much as 400% with no change in the mechanical condition of the
compressor. In addition, a radical change in vibration level can result
from turbulent flow caused by restrictions in either the inlet or discharge
piping.
Turbulent or unbalanced media flow (i.e. aerodynamic or hydraulic

instability) does not have the same quadratic effects on the vibration
profile as that of load change, but it increases the overall vibration
energy. This generates a unique profile that can be used to quantify
the level of instability present in the machine. The profile generated
by unbalanced flow is visible at the vane or blade-pass frequency of
the rotating element. In addition, the profile shows a marked increase
in the random noise generated by the flow of gas or liquid through the
machine.
Mechanical motion and forces
A clear understanding of the mechanical movement of machines

and their components is an essential part of vibration analysis. This
understanding, coupled with the forces applied by the process, are the
foundation for diagnostic accuracy.
Almost every unique frequency contained in the vibration signature of a

machine-train can be directly attributed to a corresponding mechanical
motion within the machine. For example, the constant endplay or axial
movement of the rotating element in a motor-generator set generates
elevated amplitude at the fundamental (1×), second harmonic (2×),
and third harmonic (3×) of the shaft’s true running speed. In addition,
this movement increases the axial amplitude of the fundamental (1×)
frequency.
Forces resulting from air or liquid movement through a machine also

generate unique frequency components within the machine’s signature.
In relatively stable or laminar-flow applications, the movement of
product through the machine slightly increases the amplitude at the vane
or blade-pass frequency. In more severe, turbulent-flow applications, the
flow of product generates a broadband, white noise profile that can be
directly attributed to the movement of product through the machine.
Other forces, such as the side load created by V-belt drives, also
generate unique frequencies or modify existing component frequencies.
For example, excessive belt tension increases the side load on the
machine-train’s shafts. This increase in side load changes the load zone
Unit 7 83
in the machine’s bearings. The result of this change is a marked increase

in the amplitude at the outer-race rotational frequency of the bearings.
Applied force or induced loads can also displace the shafts in a

machine-train. As a result, the machine’s shaft will rotate off-centre,
which dramatically increases the amplitude at the fundamental (1×)
frequency of the machine.
Reciprocating and/or linear-motion machinery
This section describes machinery that exhibits reciprocating and/or

linear motion(s) and discusses typical vibration behaviour for these
types of machine.
Machine descriptions
Reciprocating linear-motion machines incorporate components that

move linearly in a reciprocating fashion to perform work. Such
reciprocating machines are bi-directional in that the linear movement
reverses, returning to the initial position with each completed cycle
of operation. No reciprocating linear-motion machines incorporate
components that also generate work in a straight line but do not reverse
direction within one complete cycle of operation.
Few machines involve linear reciprocating motion exclusively. Most

incorporate a combination of rotating and reciprocating linear motions
to produce work. One example of such a machine is a reciprocating
compressor. This unit contains a rotating crankshaft that transmits
power to one or more reciprocating pistons, which move linearly in
performing the work required to compress the media.
Train noise
There are four basic sources of noise in train operations: wheels or
rail noise, engine noise, noise from the auxiliary equipment, and
aerodynamic noise.
In the field of vibration noise (structure-borne), wheel and rail noise is

the most common and dominant source of noise produced by railways.
The wheel and rail are set into vibration by the action of rolling one
over the other, making wheel/rail noise. The spectrum analysis of
measurements reveals some response peaks emerging from a substantial
content of broadband noise, suggesting that both resonant-modes and
broadband response are being excited.
Roughness of the wheel tread and of the railhead is considered the

most relevant source of wheel/rail noise and vibration. However, other
sources have been identified which can be attributed to parametric
excitation, including wheel imbalance, and rail stiffness between wheel
and rail. These sources may be relevant for broadband excitation.
In the last 20 years, because train builders are pushed faster to enhance
their competitive position, the maximum train speeds have vastly
increased. Whereas wheel/rail noise increases roughly at the cube of
speed, aerodynamic noise increases at around the sixth power of speed.
For speeds up to 250 km/h aerodynamic noise is noticeable even in the
presence of wheel/rail noise.
In designing the structure of KCR west rail, the re-radiated structure-

borne noise is the component of greatest importance. As the train runs
along the tracks, vibrations at the wheel-rail interface are transmitted
via floating slab track (FST), which is used to support the rails into the
viaduct deck. The excitation of the deck that causes an audible rumble
to be radiated is generally that related to the transverse vibration of
the panels forming the webs and flanges of the deck cross-section. The
mass of concrete viaducts is big and less flexible cross-sections. These
features are inherently well suited to minimize the re-radiated low
frequency rumble characteristic of elevated railways.
In this section, the vibration problem is discussed. The control measure

for vibration generated from the train is in Unit 8.
An interesting situation is that we can hear the sound of train moving

along the rails. This is an example of structure-borne vibration.
Figure 7.31 Vibrations from the train move along the rails (http://www.nohsc.gov.
au/OHSInformation/OHSSolutions/noise/SOUNDA.HTM)
Industrial noise
Industrial noise includes the noise from industrial plants. Depending
on the type of industry, noise levels generated can be very different.
For example, in a tough glass factory, the water-pumping system for
sanding purposes generates a certain level of vibration noise. That is,
the pump vibration results in the generation of noise. This noise greatly
affects the workers around the pump. In the worst situation, hearing
loss and psychological effects will be caused to the workers. With better
Unit 7 85
understanding the principle of vibration generation, the control method

can be more easily understood in Unit 8.
The following are examples of industrial noise coming from vibration.

Only the problem of vibration is discussed here. For the control part, it
is suggested you read through Unit 8. The vibration noises are coming
from two sources: the pumping system, piping system or rotating
machine.
Pumping system
Circulation pump
A circulation pump produces pressure variations in the water in a

heating system. The sound waves are transmitted through the pipes
to the radiators, whose large metal surfaces transmit airborne sound
afterward.
Figure 7.32 A circulation pump (http://www.nohsc.gov.au/

OHSInformation/OHSSolutions/noise/SOUNDA.HTM)
Structure-borne noise from pump sets
Vibration from an operating pump set may be transmitted to the interior

of the building through the building structure when the pump set is
directly mounted on a supporting structure without proper isolation. The
vibration transmitted may activate the building structure to generate
noise, which causes noise disturbance to the workers inside the plant.
Figure 7.33 Structure-borne noise from pumps sets (http://www.epd.gov.hk/

epd/english/environmentinhk/noise/guide_ref/pump_sys_7.html)
Piping system or rotating machines
Piping system
Vibration from the water flow (turbulent flow) in pipes has a high
chance of being transmitted from the pipe runs to the interior of the
building through the building structure where the pipes are mounted.
It becomes more severe when the pipes are in direct contact with large
planes such as walls or slabs. The vibration transmitted may activate the
building structure to generate noise, which causes noise disturbance to
residents inside the building.
Figure 7.34 Noise from vibration pipe (http://www.epd.gov.hk/epd/

english/environmentinhk/noise/guide_ref/pump_sys_5.html)
Unit 7 87
Structure-borne noise from air-cooled chillers

Vibration from an air-cooled chiller’s operation may be transmitted
indoors through the building structure at points where the chiller is
rigidly fixed to the structure without proper isolation. The vibration
transmitted may activate the building structure to generate noise, which
causes noise disturbance to residents inside the building. The cooling
system may be the major source of noise as a result of intense pressure
shocks in the liquid from compressors.
Figure 7.35 Structure-borne noise from air-cooled chillers (http://www.epd.gov.

hk/epd/english/environmentinhk/noise/guide_ref/vent_sys_4_2.html)
Noise from construction site
Concrete cutter
When a concrete cutter is cutting concrete, high-frequency sound is

generated. The sound comes from the cutting process (tough concrete
and the cutter). We can observe that the whole machine vibrates.
Figure 7.36 Concrete cutter (http://www.laku.com.my/concrete-cutter.html)

Plate compactor
From the design specification of a plate compactor, the exciter is found
to operate up to 5000–6000 rpm. At this high speed, the vibration
occurring inside the machine generates a certain amount of noise.
Figure 7.37 A plate compactor (http://www.laku.com.my/plate-compactor.html
Pedestrian vibrating roller
The operating theory is using the vibration of a big metal plate to press
the sand down evenly. This running of the roller generates some noise
after the machine is turned on.
Figure 7.38 Pedestrian vibrating roller
There are many more instruments used in construction sites. You may
find them in some of the manuals or references on noise and vibration
control. You should also note that for most of the instruments, they
have specified times and noise limits for their optimal operation. This
is the main reason an engineer or technician should have a thorough
knowledge of noise and vibration principles.
Unit 7 89
References
Bell, L H (1982) Industrial Noise Control — Fundamental and
Applications, New York: Marcel Dekker, Inc.
‘Calculation of Road Traffic Noise’ (1988) Department of Transport,

Welsh Office, HMSO.
‘Community Noise’, A Symposium Sponsored by ASTM Committee

E33 on Environmental Acoustics, ASTM, Kansas City, MO, 24–26 May
1978, ASTM Special Technical Publication 692, ASTM, 1979.
Cunniff, P F (1977) Environmental Noise Pollution, New York: John

Wiley & Sons, Inc.
Hearn, E J (1985) Mechanics of Materials, 2nd edn, Oxford: Pergamon.
Irwin, J D and Graf, E R (1979) Industrial Noise and Vibration Control,

Englewood Cliffs, NJ: Prentice-Hall.
Kiely, G (1997) Environmental Engineering, New York: McGraw-Hill.
Liu, H F and Liptak, B G (1997) Environmental Engineers’ Handbook,

2nd edn, New York: Lewis Publishers.
Mobley, R K (1999) Vibration Fundamentals, Boston: Newnes.
Perry, R H, Green, D W and Maloney, J O (1997) Perry’s Chemical

Engineers’ Handbook, 7th edn, New York: McGraw-Hill.
Sharland, I (1986) Woods Practical Guide to Noise Control, Colchester:

Woods of Colchester.
Smith, B J, Peters, R J and Owen, S (1996) Acoustics and Noise

Control, 2nd edn, New York: Longman.
Key environmental concepts

Absorption — the reduction of sound energy as it passes though an
acoustic medium and its conversion into heat.
Absorption coefficient — the measurement of the ability of a medium

to absorb sound. It is given by the ratio of the absorbed energy to that
of the incident energy (i.e. = 0 for total reflection and = 1 for total
absorption). Reverberation absorption coefficient is the value abstained
when the sound field is diffuse.
Airborne noise — sound that arrives at the point of interest, such as one
side of a partition, by propagation through air.
Ambient noise — the total noise level in a specified environment, i.e. no

particular sound is singled out for interest.
Anechoic chamber — a room for noise testing that is constructed so that

the boundary surfaces absorb all the incident sound.
Audibility threshold — the sound pressure level at a specific frequency,

which persons with normal hearing can just, detect. It is conventionally
quoted as 2 × 10-5 Pa at 1000 Hz.
Background noise — the prevailing noise level in a specified

environment measured in the absence of the noise source being studied,
which includes airborne and structural borne noise.
Compliance — the displacement per unit of applied force; the reciprocal

of stiffness.
Damping — method of dissipating vibration energy.
Damping ratio — the ratio of the actual damping in a system to the

critical damping, at resonant frequency.
Directivity factor — the ratio of the sound intensity at a specified

distance from the source of the sound in a specified direction to the
sound intensity at the same distance, but averaged over all directions.
Directivity index — ten times the logarithm of the directivity factor.
Free field — an environment in which there is no reflective surface.
Far field — that part of the sound field where the sound wave is
spreading spherically, i.e. the sound decays at 6dB for a doubling of the
distance from the source.
Hearing loss — an increase in the audibility threshold caused by

disease, injury, age or exposure to a high level of noise. It may be
temporary or permanent.
Unit 7 91
Impedance — ratio of applied force to resulting velocity under harmonic

excitation. Driving point impedance is when force and velocity are
measured at the same point; transfer impedance if force and velocity are
at different points.
Infrasonic — frequencies below the audible range, below 16 Hz.
Noise — unwanted sound.
Reverberation — the persistence of sound in an enclosure after a sound

source has been removed.
Reverberant chamber — a room designed for the study of acoustic

properties of materials and for other purposes in which a diffuse field
is required. It has highly reflective walls, so that the average energy
density is the same throughout the chamber: the converse of an anechoic
chamber
Reverberation time — the time for the sound pressure, at a specific

frequency, to decay by 60dB.
Sound power level — of airborne sound, ten times the logarithm to

the base 10 of the ratio of the sound power under consideration to the
standard reference power of 10-12 W quoted in decibels.
Sound pressure level — of airborne sound, ten times the logarithm

to the base 10 of the ratio of the square of the sound pressure under
consideration to the square of the standard reference pressure of 2x10-5
Pa quoted in decibels.
Sound reduction index — the ratio of sound energy incident on a

partition to that which is transmitted through the partition.
Structural borne noise — sound that arrives at the point of interest, such
as the edge of a partition, by propagation through a solid structure.
Threshold shift — see hearing loss.
Transmission coefficient — the ratio of the transmitted energy to the

incident energy of a panel of structure.
Ultrasonic — frequencies above the audible range, above 20 kHz.
Weighting network — a series of filtres in a sound level meter or

signal processor, which approximates under defined conditions to the
frequency response of the human ear. A variety of weighting networks
has been devised, the most common one being A-weighting.
Feedback on activities
Activity 7.1
From ideal gas equation PV = mRT, where
P = gas pressure (N/m2)

V = gas volume (m3)
m = mass of gas (kg)
R = specific gas constant (0.287 kJ/kgK)
T = gas temperature (K)
Rearrange the above equation, then where m/V = ρ
Therefore, P = ρRT and substitute into
Then,
Substitute γ = 1.40 and R = 287 J/kgK into the above equation
Therefore,
From , the sound speed in air depends on pressure and
temperature because ρ = m/V and V is associated with temperature.

Therefore, at the top of Victoria Peak, the sound speed will be lower as
result of both the lower pressure and the lower temperature, compared
with that at the bottom area in Central. The speed of sound in gases
does not change for regular changes in pressure. As the pressure
increases, the density of the air increases, too. This means the value of
is practically constant. By contrast, the temperature has a bigger
influence on the speed of sound in gases. The value of varies
with the temperature according to the ideal gas equation, .

Therefore, sometimes it may be easier to use the following equation for
the approximation
where t = temperature of air

co = speed of sound in air at 273K
Unit 7 93
Activity 7.2
This means that sounds are capable of travelling at different speeds
through all kinds of media, except a vacuum. The actual speed depends
not only on the elastic modulus of the medium but also the medium
density. In particular, the speed in air varies with the square root of
temperature, so that the sound speed will be lower in winter than in
summer.
Elastic modulus is defined as the force you need to provide to elongate

your material. The elastic modulus is not provided because air doesn’t
have a fixed shape, and there is no way to define air as elongated or not.
The elastic modulus is measured by pulling a sample of a material in

a tensile testing machine, an instrument that measures force. Stress,
denoted by the Greek letter σ (sigma), as the force (F) normalized
by the cross-sectional area (A) of the material (http://math.ucdenver.
edu/~jmandel/classes/5663f04/handouts/Elastic_Modulus.pdf)
Strain, denoted by the Greek letter ε, as the change in length of the fibre
normalized by the initial length.
Elastic modulus =
From Table 7.1, it is observed that the speed of sound increases as the
ratio of elastic modulus to density increases. This means both elastic
modulus and density affect the speed of sound. If the material used
for sound transmission is fixed, the temperature is another property
affecting the speed of sound. In general, the higher the temperature, the
higher the speed of sound. For example, one can tolerate higher sound
levels in wintertime than in summer time.
For the vacuum, there is no single molecule for sound transmission;

therefore, the speed of sound is the lowest.
Activity 7.3
If we assume a sound wave travelling as a free progressive wave, i.e.
the sound wave was not obstructed as it propagated through the air, the
equation is valid where ρ = 1.20 kg/m3 and c = 343 m/s. Since

the lowest audible pressure from a 1000 Hz pure tone is equal to
2 × 10-5 Pa, therefore, .
Activity 7.4
If there are n sound sources,
Therefore, total sound pressure levels, Lp = 10logn + L1
94 = 10log10 + L1
L1 = 84 dB
Activity 7.5
You note that Lp under reverberant conditions does not depend on the
distance r from the noise source. Since there is no ‘acoustic wattmeter’
available for the direct measurement of sound power, the sound power,
Lw may be calculated from the measurements of sound pressure, Lp and
the room surface area, S.
Activity 7.6
1 T = 3 sec , f = 1/T = 0.333 Hz
2 A = –[2π(4s–1)]2(0.1 m) = –63.2ms–2
3 a f = 0.318 Hz
b Period = 3.14 sec
4 since, 2 π f = , m = k/(2πf)2 = 1.01 kg

ENVR-S335 - U7 Basic Principles of Noise Concepts

Uploaded by

Copyright:

Available Formats

ENVR-S335 - U7 Basic Principles of Noise Concepts

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

ENVR-S335 - U7 Basic Principles of Noise Concepts

Uploaded by

Copyright:

Available Formats

ENVR S335

External Course Assessor

Copyright © The Open University of Hong Kong, 2003, 2009, 2011,

All rights reserved.

The Open University of Hong Kong

This course material is printed on environmentally friendly paper.

Basic concepts and the physical assessment of noise 4

Basic principles of vibration 70

Key environmental concepts 90

We are also exposed to different levels of pounding, roaring or shrill

Home Workplace Recreation

(Y H Yau & Associates)

Noise in Hong Kong, or in any other area, is an unavoidable by-product

Specifically, this unit:

• describes fundamental characteristics of transportation, construction

• explains how to perform calculations by hand involving sound

• describes measurement procedures and terminology relating to

• describes techniques to diagnose the relative importance of different

• describes factors affecting transportation, construction and industrial

Basic concepts and the physical

Noise also means ‘unwanted sound’. Usually, the sound of a piano is

One person’s music is often another’s noise.

Although we may enjoy listening to our own music, others may

Noise is an unwanted audible sensation, sound, or an intermittent

Basic definitions and symbol conventions

The pressure fluctuation is periodic, and a complete fluctuation in

= 1/(5 Hz) = 0.20 s.

If we imagine a body moves with a uniform speed v, it will travel a

For a sound wave travelling at the speed of propagation, c, the period

Since wavelength, λ, is the distance of a sound wave in one complete

TK = 273.16 + 21.1 = 294.26 K

In fact, speed of sound relates to the elasticity and density of the

where c = sound speed (m/s)

There is a common belief that the progression of the particles through

c = sound speed (m/s)

where c = sound speed (m/s)

Since the temperature is in degrees Celsius, first convert to the Kelvin

Derive equation starting with equation . List

Using equation , it can also be shown that sound travels faster

Table 7.1 Speed of sound in various materials (Hearn 1985)

The measure of sound or acoustic pressure level (Lp), sound or

In practice, the sound pressure of the faintest sound that a normal

Figure 7.1 Comparative scaling of sound pressure levels and decibels

Sound pressure level

As introduced in Section 2.4 of Unit 11 of T234/T237, the sound

P = root-mean-square (rms) sound pressure (Pa or N/m2)

P/P0 = 0.2/2×10-5 = 10000

and multiply by 20 to get Lp

I = sound intensity at distance r from the point source of sound (W/m2)

Figure 7.2 Spherical sound wave propagation generated by a point source

In a sound field consisting of waves coming from a source, it is usual

P = sound pressure (Pa)

At a distance 10 m from the drill to the University, the drill is

Sound Intensity level

From , the sound intensity at a point, I, is directly proportional to

the square of the acoustic pressure, P.

I = sound intensity (W/m2)

I0 = reference sound intensity, generally 10-12 W/m2

where I0 = 10-12 W/m2