Manual de Ruido Industrial

INDUSTRIAL NOISE CONTROL MANUAL
Vincent Salmon
James S. Mills
Andrew C. Petersen
Industrial Noise Services, Inc.

Palo Alto, California
Contract No. HSM 99-73-82
U. S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE

Public Health Service
Center for Disease Control
National Institute for Occupational Safety and Health
Division of Laboratories and Criteria Development
Cincinnati, Ohio 45202
June 1975
The contents of this report are reproduced herein as
received from the contractor except for minor changes.
The opinions, findings, and conclusions expressed are
those of the authors and not necessarily those of
NIOSH. Hention of company or product names is not
to be considered as an endorsement by NIOSH.
DHEW Publication No. (NIOSH) 75-183
NIOSH Project Officer: Roy M. Fleming
ii
BIBLIOGRAPHIC DATA 11. Keporl No. NIOSH-75-183
SHEET
~. Tide and Sublide S. Report Dale
June 1975
-
INDUSTRIAL NOISE CONTROL MANUAL 6.
' •. Authorlsl • 8. Performing Organization Repl.

Vincent Salmon, James S. M~lls, and Andrew C. Petersen No.
9. Performing Organizalion Name and Address 10. Project/Task/Work Unit No.
Industrial Noise Services, Inc. 11. ContraCt/GraD I No.

Palo Alto, California RSM-99~73-82
12. SpoDsoring Organization Name and Addless 13. Type of Report & Period
Covered
National Institute for OCcupational Safety and Health
4676 Columbia Parkway 14.
1S. Supplemenla ry NOle s
This manual contains fundamental information to aid the user in understanding,

16. Abstracts
measuring, and controlling noise. It was written for persons having little or no exper-
ience in solving noise control problems, realir.ing that a large number of businesses are
not prepared to deal with their noise problems eff~tiuely. There are seven chapters in
the manual covering the following subjects: fundamental principals of sound, noise measu-
rement, noise control techniques, noise control materials, case histories of successful
applications of noise control methods in actual industrial situations, how to choose a
qualified consultant, and references to additional pertinent literature. The manual
should be used as a guide to help the reader develop solutions to his particUlar noise
problems using proven methods.
17. Key WOlds and Documenl Analysis. 170. Descriplors
Industrial hygiene
Acoustic measurement
Noise reduction-
Industrial plants
iNoise analyzers
Noise generators
~ound level meters
~cdustic insulation
~coustic screening
~ontrol equipment
17b. Identifiers/Open-Ended Terms
17e. COSATI Field 'Group 06/J

18. Availability Statemenl 19. Sccurit)'Class (This es
R,,~~~t)1 .a.C::C::IFIr.n
Release unlimited 20. Security C1"ss (This
Page
IIN'-1. A~<;I rrl' n
"OR'" NT'$-" .REV. 10 ' " ENllORSlD IIY ANSI AND UNESCO. THIS FORM MAY III:: KHJI{OIlUCI::U
ABSTRACT
This manual contains fundamental information to aid the user

in understanding, measuring, and controlling noise. It was
written for persons having little or no experience in solving
noise control problems, realizing that a large number of
businesses are not prepared to deal with their noise problems
effectively. There are seven chapters in the manual covering
the following subjects: fundamental principles of sound,
noise measurement, noise control techniques, noise control
materials, case histories of successful applications of noise
control methods in actual industrial situations, how to choose
a qualified consultant, and references to additional pertinent
literature. The manual should be used as a guide to help the
reader develop solutions to his particular noise problems
using proven methods.
iii
FOREWORD
Following passage of the Occupational Safety and Health Act

of 1970, the United States Department of Labor established a
workplace noise exposure regulation which affects the majority
of businesses having excessively noisy environments. Many
companies may not be prepared to deal with their noise prob-
lems effectively but are trying to implement noise reduction
programs nonetheless. Company personnel can be faced with
uncertainty when attempting to choose a noise control method
or device, noise control materials, noise measuring instru-
mentation, or a qualified consultant. Industries need practical
noise control advice which draws upo~ the successes and failures
of previously tested methods in order to reduce the time and
expense involved in achieving an acceptable noise environment.
Accordingly, a contract was awarded for the development of a

manual which would provide essential information for under-
standing noise control techniques, including a large number
of noise control case histories fro~ actual industrial situa-
tions. Several excellent reference books have been available
for many years, but none have stressed case histories suffi-
ciently. (However, during the term of this contract, at least
two noise control books were published which do contain useful
case histories; these are included in the list of references.)
It was decided that this manual, in particular, should present
sufficient practical information to provide the user with a
solid foundation for understanding, measuring, and controlling
noise. For readers desiring additional information, the
manual would give references which cover areas of interest
in greater detail.
It was recognized that noise control is not a new discipline,

and that many effective noise control programs have been
instituted by industry, both on an in-house basis and by hiring
qualified noise control consultants. A wide range of noise.
problems and control solutions needed to be compiled from many
industrial programs in order for this manual to benefit a
large number of companies. Each case history necessarily would
contain sufficient detail to guide the reader in using the
method to solve the same or a similar noise problem. A mini-
mum of thirty varied case histories was specified to the contractor.
Although the coverage provided by the case histories documented
in the manual is not as broad as was anticipated, it is hoped
that subsequent publications of additional case histories can
~reced;ng page blank!

v - --~
be made available to eXfand this important chapter of the
manual. A free ezchanye of mutually be~eficial information can
ai~ many co~panies in solving their noise problems effectively.
In planning~or future supplemental sections to the case

history chapter of this manual, we invite all interested
IJill"ties to subF1i tease> histories of solutions to industrial
noise p~ohlems. Mo case tlistories will be published without
giving acknowledgement to each contributor who so desires;
and case histories will be revised, if necessary, to fit the
general format listed below. The time of publ~cation for any
supplement would, of course, be dependent upon the amount of
response to this invitation.
Case History Format:
A. Description of the process, machine, and noise problem.
B. Noise measurements made and discussion of findings.
C. Control approaches -- advantages and disadvantages.
D. Results in terms of noise reduction achieved and cost.
E. Pitfalls to avoid in implementing the control methods.
F. Figures -- noise data (e.g., octave band levels).
G. Sketches of area layout, machine-operator relationship,

construction details of noise control devices.
H. Photographs of machines before and after modifications

8 x 10 glossy preferred.
Case Histories can be sent to:
Noise Section/Physical Agents Branch

Division of Laboratories and Criteria Development
National Institute for Occupational Safety and Health
532 u.S. Post Office and Courthouse
cc
c >c
vi
CONTENTS
LIST OF ILLUSTRATIONS xi
LIST OF TABLES xv
CHAPTER 1 FUNDAMENTALS OF SOUND 1
Basic Principles 1
Definitions and Relationships 2
Hearing and Noise Level Weighted Scales 5
Combining Noise Levels 8
Noise Generation and Noise Fields 10
Sound Power 13
Summary of Symbols and Definitions 14
CHAPTER 2 MEASUREt~NT OF NOISE 15
Compliance Measurements 15
Basic Instruments, Use and Calibration 15
Dosimeters 19
Diagnostic Measurements 19
Sound Level Meters 19
Octave Band Filters 24
Fractional Octave-Band Analyzers 24
Photographic Equipment 24
Tape Recorders 25
Other Equipment 26
Making Readings 27
Relating to A-levels 27
Vibration 28
Ears and Mind 28
General Tips 29
Combining Decibels 29
Data Requirements 30
Follow-Up Measurements' 32
CHAPTER 3 NOISE CONTROL TECHNIQUES 33
Source 33
Distance or Relocation 34
Vibration Control or Isolation 34
Damping 36
Lagging 37
Air and Gas Flow Noise Reduction 38
Air Jet Flow 39
vii
Hydraulic Systems 3'9
Motor Air Noise 42
Enclosed Drives 42
Balancing 42
Noise Source Enclosures or Barriers 43
Path of Sound 43
Acoustical Shields and Barriers 43
Total Enclosure 46
Room Acoustics 57
Receiver 60
Administrative Controls 60
Personal Protective Equipment 61
Audiometry 62
CHAPTER 4 NOISE CONTROL MATERIALS 63
Absorption Materials 63
Transmission Loss Materials 65
Damping Materials 66
Vibration Isolators 67
Material Selection 67
Acoustic Considerations 67
Vibration Isolators 71
Silencers 72
Nonacoustic ~onsiderations of Materials 72
CHAPTER 5 CASE HISTORIES 75
Case History Data 75

Outline of Complete Procedure for Development
of Noise Control 75
1 - Steel Wire Fabric Machine 81
2 - BOO-Ton Verson Blanking Press - Vibration
Isolation 88
3 -' Blood Plasma Centrifuge 95
4 Blanking Press Ram III
5 - Spinning Frame Air Noise 116
6 Barley Mill 119
7 Sheeter for Boxboard 129
8 Air Scrap Handling From Bobst Cutting
presses, Folding Carton Manufacturing 133
9 - Jordans for Paper Mill for Boxboard 136
10 Folding Carton Packing Stations, Air
Hammer Noise Source 142
11 - In-Line Gravure Printing and Cutting
Press Foldinq Cartons 146
12 Air Scrap Handling Ducts for Corrugated
Container Industry 149
13 - Paper Machine, Wet End 152
viii
14 - Punch Press Noise 157
15 - Straight and Cut Machines 170
16 - Cut-Punch Press in Metal Fabricating
. Plant 177
17 - Parts Conveying Chute 184
18 - Nail Making Machine 187
19 - Wood Planer 191
20 - Punch Press 196
21 - ~1aterials Handling, Air Motors 205
22 - Textile Braiding Machines 207
23 - Metal Cut-Off Saw 210
24 - Planer-Wood 213
25 - Punch Press Total Enclosure 215
26 - Dewatering Vacuum Pump, Paper Mill 216
27 - Steam Line Regulators 218
28 - Plastics Scrap Grinder 220
29 - Newspaper Printing Press 223
30 - Chemical Process Plants 224
CHAPTER 6 SELECTING AND USING A CONSULTANT 231
Knowing When a Consultant is Needed 231

Selection of a Consultant 232
Guideline Questions 233
The Proposal 236
Other Services 238
CHAPTER 7 ANNOTATED REFERENCES 239
ix
ILLUSTRATIONS
1.1 Frequency-wavelength chart 3

1.2 Response characteristics of weighting
scales and of ear at threshold 6
1.3 Method for combining decibel levels 9
1.4 Noise field around a single machine 12
3.1 Effect of reduced pump speed on noise level 40
3.2 Side branch muffler 41
3.3 Dissipative muffler 41
3.4 Source-barrier-receiver geometry 45
3.5 Split panels for services 47
3.6 Welded angle iron frame 48
3.7 Enclosure panels secured to frame by
quarter-turn fasteners 48
3.8 Vibration isolation and toe coving 49
3.9 Enclosure panel interior treatment 50
3.10 Door and hatch detail 51
3.11 Window detail 51
3.12 Cooling and ventilation system 52
3.13 Examples of partial and complete enclosures 53
3.14 Effect of enclosure sound leaks on potential
noise reduction 54
3.15 Example of isolating wall 56
3.16 Sound field behavior in a reverberant room 58
4.1 Transmission loss plateaus 69
4.2 The mass law relation between average sound
transmission loss and mass per unit area of
partition 70
5.1.1 Noise levels for 8-foot fabric machine,
1 meter from main drive gear 83
5.1. 2 Noise levels for 8-foot fabric machine
operator station 84
5.1. 3 Noise levels for 8-foot fabric machine
1 meter behind operator station in wire
spool area 85
5.1. 4 Steel plate barrier with window 86
5.1. 5 Lexan barrier in two sections 87
5.2.1 Vertical acceleration on pier, before and
after isolation 90
5.2.2 Horizontal acceleration on pier, before and
after isolation 91
5.2.3 Attenuation of vertical and horizontal
acceleration at pier by adding isolation 92
5.2.4 Quasi-peak levels 4 feet, from press foot,
before and after isolation 93
5.2.5 A-weighted quasi-peak levels 4 feet from
press, before and after isolation 95
xi
l Preceding page blank
5.3.1 Sketch of one centrifuge 98
5.3.2 Operator exposure in -5°C box room 99
5.3.3 Sketch of centrifuge showing locations
for close-in diagnostic readings 100
5.3.4 Centrifuge noise measured parallel-to
motor shaft 101
5.3.5 Centrifuge noise measured 15 cm directly
below pulley 102
5.3.6 Centrifuge noise measured perpendicular
to spout 103
5.3.7 Centrifuge noise measured between centrifuge
motor and pUlley belt 104
5.3.8 Centrifuge noise measured 2 cm from centrifuge
motor exhaust 105
5.3.9 Centrifuge noise measured at tub height 106
5.3.10 Operator's exposure in reconstituting room 107
5.3.11 Noise in reconstituting room, measured
2 cm from bottom edge of exhaust face of
refrigeration unit 108
5.3.12 Noise in reconstituting room, measured 2 cm
below rear intake face of refrigeration unit 109
5.3.13 Motor exhaust muffler '110
5.4.1 Method used to cover slots in blanking press
ram 113
5.4.2 Quasi-peak readings of blanking press before
noise control 114
5.4.3 Quasi-peak readings of blanking press after
ram ringing was contained 115
5.5.1 Air exhaust vent modification for spinning
machine noise control 118
5.6.1 Barley mill noise levels, measured betw~en
Moorspeed and Ross mills 124
5.6.2 Barley mill noise levels, measured at guard
rail of hay shredder 125
5.6.3 Barley mill noise levels, measured at control
operator chair 126
5.6.4 Floor plan of barley mill 127
5.6.5 Sectional view of barley mill 128
5.7.1 F1oorp1an of sheeter for boxboard 131
5.7.2 Sheeter drive box enclosure 131
5.7.3 Layout of sheeter and operators 132
5.8.1 Scrap handling system for cutting press 135
5.9.1 Noise levels 1 m from Jordan refiner 139
5.9.2 Noise levels inside enclosure of control room
for Jordans 140
5.9.3 Noise reduction lagging for Jordan shell 141
xii
5.10.1 Air hammer stripper and packer line 144
5.10.2 Barrier wall theory 145
5.11.1 Top view of in-line gravure-cut press
with sound barrier curtain 148
5.11.2 Side view of in-line gravure-cut press
with sound barrier curtain 148
5.12.1 Noise levels in scrap duct for corrugated
box industry, before and after covering 151
5.13.1 Noi~e levels at wet end aisle of paper
machine 154
5.13.2 Paper mill--wet end 155
5.13.3 Noise levels inside control booth at wet
end of paper machine 156
5.14.1 Layout of punch press room 160
5.14.2 Ambient noise level with furnace and punch
presses off 161
5.14.3 Ambient noise level with furnace on, punch
presses off 162
5.14.4 Ambient noise level with furnace on, two
Summit, two Bliss, and one Benchmaster
presses in operation 163
5.14.5 operator's exposure at punch I: stamping
stock 164
5.14.6 Operator's exposure at punch I: stamping
without stock 165
5.14.7 Operator's exposure at punch VII 166
5.14.8 Operator's exposure at punch VII,
operating without work piece 167
5.14.9 Operator's exposure at punch VII,
operating without stock and without
air exhaust 168
5.14.10 Design of nozzle 169
5.14.11 Sketch of Plexiglas barrier 169
5.15.1 Straight and cut machine: close-in
measure near west side of clutch cutter
mechanism 172
5.15.2 Straight and cut machine: close-in
measure near east side of clutch cutter
mechanism 173
5.15.3 Straight and cut machine: operator's
near field exposure 174
5.15.4 Straight and cut machine: operator's
far field exposure 175
5.15.5 Barrier wall for straight and cut machine 176
5.16.1 Cut-punch press, operator's exposure 179
5.16.2 Cut-punch press, close-in diagnostic data,
14 crn from gears 180
5 cm from the dog-flywheel interface 181
xiii
5 cm from piston-collar impact on air
cylinder 182
5.16.5 Sketch of hanging barrier for cut-punch
press 183
5.17.1 Chute for conveying cartridge cases 185
5.17:2 Noise spectra measured 3 feet from chute 186
5.18.1 Nail making machine: operator position
noise levels 189
5.18.2 Nail making machine: operator position
noise levels after vibration isolators
installed 190
5.19.1 Noise levels for wood planer -
straight cutter 193
5.19.2 Noise levels for wood planer with helical
cutter 194
5.19.3 Effect of board length on noise from
wood planer 195
5.20.1 Comparison of punch press noise levels
with air ejection and with mechanical
ejection 200
5.20.2 Comparison of punch press noise levels
with and without a shield between operator
and air ejection noise 201
5.20.3 Comparison of punch press noise levels:
standard die versus slanted die 202
5.20.4 Noise levels 30 feet from bench press area,
before and after sound absorption treatment 203
5.20.5 Noise 30 inches from punch press, before
and after test cardboard enclosure 204
5.21.1 Effect of muffler on air exhaust from hoist 206
5.22.1 Textile braiding machine: comparison of
noise levels from steel carriers and from
polyurethane carriers 209
5.23.1 Metal cut-off saw: operator's exposure,
before and after enclosure of saw 212
5.26.1 Octave-band analysis of paper mill
vacuum pump noise 217
5.27.1 Main valve plug with throttling vanes to
reduce noise in steam line regulator 219
5.28.1 Plastics grinder; range of noise levels, dBA 221
5.28.2 Plastics grinder; range of noise levels
after soundproofing, dBA 222
5.30.1 Noise reduction achieved by reducing fan
speed, using increased blade pitch to
offset decrease in speed 225
5.30.2 Compressor discharge noise reduction achieved
by lagging expansion joint 226
5.30.3 Noise reduction achieved by redesigning pump
by-pass loop 227
5.30.4 Noise reduction achieved by adding silencer
to air blower intake 228
xiv
TABLES
1.1 Corrections from flat levels to A-levels 7
1.2 Method for approximating decibel combinations 9
1.3 Example of level combining 10
2.1 Sound level meters and accessories available 21
3.1 Required static deflection for common

industrial speeds or forcing frequencies 35
3.2 A-level noise reduction by use of heavier pipe 42
4.1 Sound absorption coefficients of common

acoustic materials 64
4.2 Sound absorption coefficients of common

construction materials 64
4.3 Transmission loss of common materials 68
5.30.1 Sources of noise and methods of noise

reduction 229
xv
Chapter 1
FUNDAMENTALS OF SOUND
The need for noise control is certainly not newi the ancient
Romans complained about chariot noise. But as the industrial revo-
lution progressed, noise has become more intense and more persist-
ent. Machinery to do specific jobs was designed, manufactured and
sold to its users without much, if any, concern for noise.
Recently, both State and Federal Governments have become active in
dealing with environmental pollution of all kinds, of which noise
is a part. Thus, there is a new emphasis on reduction of noise at
its source.
When faced with the problem of high or unacceptable noise
levels, you must be able to make a choice of the many different
ways of achieving reduction of these levels. In some cases you
will be able to apply more or less standard solutions that can be
found in the literature about your particular industry or machine.
However, the general problem of noise reduction is not so simple.
Rather than applying a standard recipe, you must learn and apply
the first principles of sound and vibration to the problem and
design the noise reduction means yourself. This chapter is designed
to introduce you to these basic principles so that you can apply
them.
Basic Principles
Sound is transmitted through any elastic medium (gas, solid,

or liquid) and propagates as wave motion. A convenient way to
visualize wave motion is to use a long spring, such as a "Slinky"
toy. If such a spring is stretched vertically and the top is
quickly lowered, a motion down the length of the spring will result.
The bunching together and spreading apart of the individual spring
elements is a wavelike motion. These motions help us to visualize
and thus understand wave motion.
The visual wave motion of the spring corresponds to the

invisible wave motion of sound waves. Actually, sound propagates
by elastic interactions among atomic or molecular components of
the medium through which it travels. This means that the speed
of sound depends on the mass of the molecules (density) and on
their elastic reactions (pressure).
To discuss the properties of wave motion, let us return to the
coil spring example. Again, stretch the spring vertically and then
suddenly push down on the top coils, but this time continue to
repeat this motion. You will see groupings of compressed coils
traveling down the spring. Between these compressed coil groupings
will be groupings whose coils are expanded farther apart than usual.
Of course, the coils do not move down; they stay in one average
position and oscillate about it. What moves in the wave motion is
the grouping of spring elements. This motion of .the grouping of
elements travels with a definite speed. If you change the rate of
up and down motion of your hand holding the top of the spring, the
speed of the groups will stay the same, but the spacing between
groups will change. This spacing is called the wavelength.
The wave speed in the spring remains constant while the wave-
length of the groupings changes. The same situation occurs with
sound in the air. Airborne sound travels at a speed (c) of 344 meters/
second at 20°C (which is 1127 feet/second at 6SoF). The wave
motion in air involves particles (atoms and molecules) that vibrate
because of the sound energy present. The number of complete cycles
per second executed by these particles is called the frequency of
the wave. In our spring analogy, we noticed that we could change
the frequency or wavelength, the sound speed remaining unchanged.
Definitions and Relationships
We are basically interested in four parameters descriptive of

sound waves in air: sound speed, frequency, wavelength, and ampli-
tude. The sound speed depends on the temperature of the medium,
while amplitude depends on both temperature and density.
In many noise control problems you will be concerned with both

frequency and wavelength. In modern terminology, the unit of fre-
quency is cycles per second or hertz, abbreviated Hz. The symbol
for frequency is f.
Wavelength is the distance between similar configurations or

groupings of molecules in the sound wave. The Greek letter lambda
( t.) is the symbol for wavelength. From the coil spring experiment
it can be noted that the spacing between groups decreased as the
frequency of starting motion was increased. In mo~e precise form,
we relate sound speed, frequency, and wavelength by the formula
c = ft.. (1.1)
Figure 1.1 shows th~s relation for sound in air at 6SoF. At 1000 Hz
the wavelength is about 1.1 ft, 13.5 in., or 34 cm. In Chapter 3
we shall discuss how these concepts are used in nois~ contr~l tech-
nology.
2
, 00. 0 r- ---r--....,.........,.....,...,.....,...."I"""I"'.,......-~...,..-_r_...,.""""'T"""'T"_r"I""T"',._--_r_-.,......"""'T".....,.....,....,....,_n
50.0 ~--+--:'---+----+---f---3IliI~-+---+----t----+-----t
20.0 1----.3Il.::-'-----+---3\E---t----+-----t-.....3Ill~+_---+----__t_--_t
, O. 0 1-----+----.:~-f----_+___'IliIr_-+_-----+---+_--'''"-_+----_+~-__t
en
a:
ill.
~ 5. J 1-----+-----f---..llIlr--+----+_~----+---+_--__+-.....3Illr_-_+_--__t
ll.'
:2'
a:
o
en
ill
I 2.0 I - - - - - + - - - - - f - - - - _ + _ - - " " " ' -
U
Z
~
ill '.0 l----+-------II-----+----t-----"IiiL---i----+-~-__+----_+_--____t
ill
ll.
Z
>,= clf =wavelength
I
~
.5 f =frequency
"
Z c = velocity of sound
W
-l
W c at 20° C, 68 0 F
>
« .2 =344 meters per sec = "29 feet per sec
:'!: =13,540 inches per sec = 769.8 miles per hour
I
. 1 ....._ _......_ ......_"_......._"_.......l....l..l....-_ _......._ ....... ---'~ ....................._ _......._ _"__ _""_........lo..oI.........
10 20 50 100 200 500 1000 2000 5000 10,000

FREQUENCY - Hz
FIGURE 1.1 - Frequency - wavelength chart
Airborne sound is typically what we are involved with in

industrial noise problems, and the two parameters of most concern
are frequency and amplitude. The amplitude, or magnitude, of a ,,"
noise cannot be described easily without an explanation of some of

the basic physics of sound. The perception of amplitude, called
loudness, and the measurement of amplitude are related to this
understanding.
Sound has traditionally been measured by observing its effect on

something. Early methods used diaphragms and direct recorders
that were displaced by sound waves alone, much as is the eardrum.
Scientists were able to calculate the amount of sound pressure
3
required to produce displacements. However, simple and reliable
quantitative measurements were not possible until stable micro-
phones and amplifiers were developed. The term "sound pressure"
describes the alternating pressure above and below atmospheric
pressure. The magnitude of the pressure fluctuations can be
sensed by noting that a fluctuation of only 0.1% of atmospheric
pressure represents an intolerably loud sound. Mathematical
descriptions of these pressure fluctuations are not as important
here as are the concepts of the physical motion and the definitions
of sound pressure, sound power, and sound pressure level. The
relationships between these terms is somewhat complicated, but
the following discussion will provide an introduction.
The equation relating sound power to sound pressure (for

plane and spherical waves) is
W = p2 sip c (1. 2)
where W is the sound power, p is the sound pressure, S is the

surface area associated with the measured sound pressure, p is
the greek letter "rho" used for density of the medium, and c is
the speed of sound. For noise control in which spherical waves
are encountered, S = 41T'r 2 , whereTT= 3.1416 and r is the distance
from source to measurement point.
The sound pressure, p, could be measured to indicate the
magnitude of a sound, but the range of values between a high and
low sound pressure could easily have a ratio of a million-to-one.
To eliminate the problem of this wide numerical range, "sound
pressure level" (Lp) is used. The logarithmic relationship
between sound pressure level and sound pressure is defined by the
equation.
(1. 3)
where Po is a reference pressure. The reference pressure has

been chosen to be a value which is approximately the threshold of
hearing at 1.000 cycles per second (0.0002 dyne/cm 2 , 20 micrometer/
m2 , or 20 microPa). The unit of sound pressure level is the
"decibel" (dB), so that if p = Po' then Lp = 0 dB. From the
definition of reference pressure, it follows that a person with
excellent hearing could hear sounds levels at least as low as 0 dB.
Experiments have developed the approximate relation between

loudness and sound level. For levels around that of conversational
speech, the sensation change that is described as "twice as loud"
corresponds to a change in level. of ~ to 10 dB. Thus, sound at
60 dB would feel about twice as loud as sound at 50 dB; and at
90 dB, half as loud as 100 dB.
The sensation of pitch is well-correlated with frequency.

The musical interval called the "octave" is recognized by most people.
It corresponds to a frequency ratio of two to one. Thus,middle C,
at 262 Hz, is an octave above 131 Hz and an octave below 524 Hz.
Often you will need to estimate the frequency of such machinery
4
sounds as those from fan blades and gear teeth. Remember that you
need the number of events per second, ~hereas rpm is usually given.
For example, if a large air-moving centrifugal fan has 24 blades
and operates at 1165 rpm, the fan blade passage frequency is (1165)
(24)/60 = 466 Hz. This tone should be audible to the ear. With
practice, you can learn to listen for such definite tones in the
presence of other components of the noise.
Quality refers to that aspect of sound that enables us to
distinguish between two similar sounds. The physical equivalent
of quality is the character of the spectrum. The spectrum is an
objective and precise numerical statement of how much of the total
sound exists in all frequency regions, from low to high. Spectra
are measured with electrical filters that pick out and pass only
the frequency regions for which they are designed. The one in most
common use by the noise control engineer is the octave band filter,
which is discussed in Chapter 2. With this, the resulting spec-
trum gives the sound levels in each octave band, usually 63 through
8000 Hz.
Hearing and Noise Level Weighted Scales
The whole human hearing system, which we refer to as the ear,

can receive and interpret a wide variety of sounds. When sound of
only one frequency is present, and the sound pressure variation
with time is sinusoidal, the sound is said to be a pure tone. Com-
plex sounds, such as the usual noise spectrum, may be considered
to be made up of a multiplicity of random phase pure tones.
We have seen that the ear is capable of sensing very small

amounts of sound energy. Figure 1.2 shows the relationship between
excellent hearing and the response of the weighting circuits in
standard sound level meters.
Th~ combination of noise levels and the exposure time to such

levels is called noise exposure. "(Noise exposure calculation is
explained in Chapter 2.) The spectrum of noise is also important,
as can be seen from examination of Figure 1.2. The ear is much
more sensitive to components with frequencies above 1000 HZ, and
is much more easily damaged by noise of high intensity at, say,
3000 Hz than at 300 Hz. Instead of dealing with the many frequen-
cies in the hearing range, it would be convenient to be able to use
a single number descriptive of the noise spectrum character, to sim-
plify the measurement and evaluation of noise.
Botsford, Young, and others have successfully demonstrated the

value of this single number approach. It turns out that A-scale
readings of noise show reasonably acceptable correlation with the
potential for causing hearing loss. At the same time, some simple
indicator of spectrum character is needed. Botsford takes the
difference between the A-weighted or A-scale noise and the C-weight-
ed or C-scale noise levels measured on a standard sound level meter.
Reexamination of Figure 1.2 shows that the C-weighted noise in-
cludes more of the lower frequencies than does the A-weighted.
Thus, if we consider the following three possibilities, we can see
what is meant about quality of noise.
5
co
"'0 10 -10 >-
I I-
w
(/) 0 0 >co
_"'0
Z !::::I
0
c.. ·10 10 (/) a::
(/)
w sound level ~<t
a:: ·20 20 (/) w
meter scales WW
W
> >:I:
30 - l -
I- ·30 I-u..
<t ear "response" at threshold <to
....J
W ·40 40 ~
a:: a::
·50 50
100 1000 10,000
FREQUENCY - Hz
FIGURE 1.2 - Response characteristics of weighting scales and of ear at threshold
(L C - LA) > 0 Low-frequency quality
(L C - LA) = 0 Medium-frequency quality

(L
C
- LA) < 0 High-frequency quality
For a fixed C-level, noise in the last two cases will induce more
hearing loss than low-frequency quality noise. It must be noted
that consideration of the actual level is left out of this spectrum
quality parameter and must be considered first before quality judge-
ments are made. As you get deeper into noise level measurements
and analyses, you will find that this kind of single number descrip-
tion aids you when you need a quick, or on the spot, analysis of
noise.
The evolution of the use of A-scale noise levels, LA' has

basically been the result of experimental evidence on the relative
hearing loss from constant-level sounds of different frequencies.
'High-frequency sounds cause much more hearing loss than low. The
similarity to hearing sensitivity was shown earlier in Figure 1.2.
Thus, A-weighted sound has become the common legal description for
6
,
sound lev~ls, because it incorporates a good single approximation
to the weightings that show good correlation with such effects as
loudness, speech interference and annoyance, as well as hearing
loss. The American National Standard for sound level meters states
that, unless otherwise indicated, sound level means LA' in units of
dBA. To convert flat response readings to A levels, apply the weight-
ing factors at octave band center frequencies shown in Table 1.1.
TABLE 1.1 - Corrections from flat levels to A-levels
Octave Band
Center
dB
Frequency
(Change)
(Hz)
31.5 -39.5
63 -26
125 -16
250 -8.5
500 -3.0
1000 0
2000 +1.0
4000 +1.0
8000 -1.0
16000 -6.5
So far we have talked about noise as complex sounds or noise

spectra and characterized the quality of noise, but we have not
actually defined noise. The standard usage or accepted defini-
tion is that noise is unwanted sound. This is more a psychologi-
cal than a physical definition. The sound of a racing car engine
running may not be noise to you but to your neighbor it may be.
7
In a legal sense, noise is whatever sound may be subjected to
regulation because of undesirable effects. Usually, sound level
and duration are the predominant criteria. Effects of sounds or
noise include not only hearing loss, but also annoyance, as well
as speech and sleep interference.
Combining Noise Levels
A frequent calculation in noise control engineering is combining

levels. Suppose we have two levels and we are interested in find-
ing out what our soun~ level meter would read if the two sounds
were simultaneous. We make the basic assumption that the noises
are random; that is, only weak pure tones are present. The for-
mula for calculating the combined level, L ' given L l and L 2 , is
c
L = L + 10 log (10(L 2 - Ll)/10 + 1) (1.4)
c l
A practical example would be to have two closely spaced noise
sources and to stand, say, about halfway between them. Measure
the levels with a sound level meter, first with one on and the
other off, then vice versa. Combination of the levels should give
you the level if they were both running. This combination of sound
levels of random noise requires adding the power or energy of each
source and using the decibel or logarithmic transformation to get
what is read on a sound level meter.
Some peculiarities in the measurement of sound and the use of

decibels are worth mentioning. Doubling of the energy raises the
level by 3 dB. For example, two equal random noises of 40 dB each
combine to a 43 dB total. Two equal 90 dB random noises combine
to 93 dB. The combination process uses the physical fact that the
square of the sound pressure is usually closely proportional to the
sound energy.
The chart of Figure 1.3 enables this combination to be per-

formed without the formula.
For rough work, the scheme of Table 1.2 is easy to remember

and to use; its results are usually good to ~ 1 dB. To remember
the break points at which 3, 2, or 1 dB is to be added to the larg-
er level, note that the break point differences of 1, 4, and 9 turn
out to be the squares of the first three integers. The exact val-
ues of these break points are 1.1, 3.8, and 9.1 dB.
As an example of the use of Table 1.2 and Figure 1.3, consider

the problem of calculating the noise from several identical ma-
chines that are at different distances from the point at which the
combined level is needed. Suppose that the noise from each machine
8
-J o
ct
5
~
3
Z NUMERICAL 01 FFERENCE BETWEEN

wm
W-c
TWO LEVELS BEING ADDED - dB
:!:,
~-J
WW2J.----H.I
m>
WW
U-J
Za:
Ww
a: c)'
Wa:
u.
ct
!:!:-J
00
-JZ
ct ct ,
U
a:
W
::2 Ol.-_"'----l'--....._....L.._......._..-._.l.-_~___L .;.J
~ 3 4 5 6 7 8 9 10 11 12 13 14
NUMERICAL DIFFERENCE
BETWEEN TOTAL AND SMALLER LEVELS - dB
FIGURE 7.3 - Method for combining decibel levels·
* From Handbook of Noise Measurement, General Radio Co., Concord, MA.
TABLE 1.2 - Method for approximating decibel combinations
Add to Larger Scale

Difference
(dB)
0 3.0
1 2.6
2 2.1
3 1.8
i
4 1.4
5 1.2
6 1.0
7 0.8
8 0.6
9 0.5
10 0.4
11 0.3
12 0.2
9
operating alone is given by the levels in the first line of Table
1.3. In the second line We have arranged these, for convenience
only, in order of increasing level. Successive lines below show
the steps of combining levels in pairs by the rules of Table 1.2.
The result is 98 dB. By rigorous calculation, the correct value
is 97.5 dB.
TABLE 1.3 - Example of levef combining
Original levels 85 92 90 84 93 87
Rearranged 84 85 87 90 92 93
84 & 85 =88
Lr-I 88
I .
88 & 87 =91 91
I .
91 & 90 = 94 94
92 & 93 :.=96
•
96
I
94 & 96 =98 98
Noise Generation and Noise Fields
When investigating noise it is often convenient to consider

separately the source, path, and receiver components of the prob-
lem. At the source we are concerned with the forces that cause
surfaces to vibrate or air to generate turbulence, or both, and
cause sound to be radiated. Along the path the sound spreads out
and may be absorbed. The receiver is (for our problems) the human
ear, with attention to the noise exposure experienced.
It is important to be acquainted with how the noise is gener-

ated and radiated. Such knowledge can sometimes permit the intro-
duction of simple modifications to machines to reduce the amount
of sound that is radiated. The two chief items about which infor-
mation is usually sought are the exciting force and the size,
10
nature, and orientation of radiating surfaces. The usual excita-
tion forces are from unbalanced rotational and reciprocating
sources, impact aerodynamic sources, friction, magnetic sources
(AC transformers), hydrodynamic sources (water hammer), and liquid
cavitation.
Sound is radiated into the air by such surfaces as machine

frames, panels, and guards; unrestrained stock and workpieces;
such stressed parts as saws and diesiand friction-excited sur-
faces such as brakes. Fan blades leave a wake of alternating
eddies behind. Steam and air venting makes noise by shear against
the surrounding relatively still air; the sound power is typically
proportional to the gas velocity to about the fifth power.
When sound spreads out from a small source with no major re-
flecting surfaces nearby, the sound level along the path decreases
with distance. If the level at distance r is L , then at r the
eve 1 · 0 0
1 1S
(1. 5)
The term "inverse square law field" is given to the region in which
equation (l.5) holds. Conversely, if we are reasonably certain
that inverse square law conditions pertain, and we know the level at
one position, then by (l.5) we can calculate the level at any other
position in this region, along the same radius from the source at
which the given level was measured. However, if the source is di-
rectional, then it will be necessary to measure at many angular
values to get a complete description of the sound field.
As we have seen, the magnitude of the loudness sensation de-

pends on the sound level at the ear. However, a complication devel-
ops when sound in a room, such as in a factory space, is to be cal-
culated. At a sufficient distance from the source of noise, we
will hear not only the direct sound, but also all the sound that is
reflected from the room surfaces. The behavior of sound level with
distance from a noisy machine is sketched in Figure 1.4.
There are three relatively distinct regions to which we shall

refer. Close to the machine, or in the near field, levels will
depend very much on near-by noise spots, such that levels
may remain constant or may decrease. In the second region we have
inverse square law behavior, where the noise from the machine is
much greater than that reflected from the walls and other room
surfaces. In the third region, the reflected sound predominates,
and the level is relatively constant. Typically, because we are
not dealing with ideal point noise sources, the falloff of noise
level with distance does not behave in a smooth way. Differences
in the way the noise falls off with distance away from the source
will indicate to some degree the kind of noise control to be used.
11
The critical distance (r c ) concept is useful. Critical distance
is ,that at which the noise level changes from inverse square be-
havior and becomes relativ~ly constant (reverberant level) as shown
in Figure 1.4. (For applications see Chapter 3.)
70
60
2 inverse
square
50
ID
"0
3 reverberant
1 near
...J 40
ill
>
ill
...J
0 3C
z
::J
0
___________l__
CJ)
20
10
3-dB change point
rc r4
LOG DISTANCE - r
FIGURE 7.4 - Noise field around a single machine. Critical distance is r c'
The junction between regions 2 and 3 (Figure 1.4) is at the

critical distance, r c . Beyond that distance, increased absorption
on the room surfaces will reduce the reverberant level but not the
level in the inverse square law level of region 2. The importance
of this for noise control engineering is that if a worker is in
region 2 (and most are), acoustical treatment of the room surfaces
will not reduce the noise he receives from his own machine. The
chief exception occurs when there are many machines, close to-
gether in a low-ceiling room. If a factory room has average ab-
sorption coefficient a and has n identical floor mounted machines
that emit noise equally in all directions, the critical distance is
rc = 0.20 / S a In (1. 6)
12
where S is the total area of the walls, floor, and ceiling and is
measured in the same units as r c . Note that additional absorption
(increasing a) will· reduce only those levels in zone 3, beyond
the critical distance, r c . Thus, the ambient room background levels
can be reduced by increasing the acoustical absorption at room sur-
faces.
Sound Power
For calculating the reverberant sound field in region 3, it is

necessary to know the total sound power emitted by the machines or
a standard source in a room. Sound power cannot be measured direct-
ly but must be calculated from measurements made 'under closely con-
trolled conditions, usually in the laboratory. One procedure uses
a nonreflecting (anechoic) environment: another uses a highly reflec-
ting room for the machine in question. It is usual to state, not
the sound power, but the sound power level. This is defined by
(1. 7)
where W is the sound power in watt, and Wo is the reference power

of 10- 12 watt, and corresponds to Po
or 20 micronewton/m 2 reference
sound pressure.
Except for absorption in the air, the radiated sound power is

independent of distance from the machine. However, the intensity
(power per unit area) does change inversely as the square of the
distance away from a point source. The area over which intensity is
measured is easily related to the shape of the measurement surface
and the distance away from the source.
For calculating noise levels in the open, away from all reflec-
ting surfaces, the sound pressure levels at various distances can
be calculated by equation (1.5). However, in a room, and in factory
workshops 'in particular, it is first necessary to start with sound
power. Also, sound power characterizes the total noise from the
machine and is useful in comparing different machines on a noise
basis. In terms of hearing damage, however, it is the A-weighted
sound pressure level that is needed and to which other measures
must eventually be converted.
At our reference power Wo = 10- 12 watt, the sound pressure

level measured 0.28 meter away from this ideal point source would
be zero with no reflecting surfaces.
The general relationship for a point source well away from

any reflecting surfaces, emitting sound power level Lw is:
Lp = Lw - 10 log S = Lw - 20 log r - 11 (1. 8)
13
where r is in meters and S is in square meters: if r is in feet,
replace the numeric 11 by 0.7. The constants are the result of
the units used. Now for point sources this can be an extremely
useful relationship. If the source is on a large flat surface
such as a floor in middle of large room with no reflectors other.
than the surface, the relation is
Lp = Lw - 20 log r - 8 (1. 9)
where r is in meters.
In the future, possibly the sound power (W) of a noise source
will be the single more important parameter to know when analyzing
the sound level in a room, especially with noise-rated machines.
After some additional reading in references such as those listed
in Chapter 7, you will be able to use equation (1.9) to aid in
writing purchase specifications. Combination of the level you
read on your sound level meter before a new machine is installed
with the one you predict using equation (1.9), plus information on
the acoustics of the room, will tell you whether the new noise
level will be acceptable. If several machines are present, decibels
may be combined by the rules discussed earlier to give the total
level.
Summary of Symbols and Definitions

Some of the definitions used in sound measurements are-
summarized below.
W = Sound power
p = Sound pressure
p = Density of medium in which sound is being propagated
c = Speed of sound in medium
S = Area over which sound pressure is measured
r = Distance away from source
W
o
= Reference sound power, 10 -12 watt
= Reference sound pressure, 20 micronewton/meter 2
= Radius of reference sphere = 1 meter
Pc = 410 rayls, where p is the sea level density of air and c is
corresponding sound speed at 20°C.
14
Chapter 2
MEASUREMENT OF NOISE
In the usual industrial noise situation, there will be two

types of measurements:
(1) Compliance measurements, which are made in

accordance with some relatively precise set
of instructions, usually based on laws or
regulations.
(2) Diagnostic measurements, which aid in engineering

control of noise by locating specific noise
sources and determining their magnitudes, and
by indicating the types of sontrols needed,
their locations, and the amount of reduction
sought.
The two types will be explored in detail in this chapter.
Compliance Measurements
Compliance measurements are made with some relatively precise

set of instructions, usually based on laws and regulations. The
purpose is usually to determine the extent of compliance with the
limits set forth in the laws or regulations. Thus, in a n6ise ex-
posure compliance survey for industrial noise, the data will be the
slow A-weighted levels measured at the ear location of the workers,
together with the times spent at the levels encountered. From these
data, the daily noise dose is calculated by means specified in the
regulations.
Basic Instruments, Use and Calibration. The chief instrument

for noise measurements is the sound level meter (SLM) , which should
be a Type 1 or 2, made in accordance with American National Standard
S1.4 (1971), "Specification for Sound Level Meters." The Type 2
instrument has broader tolerances on performance than the Type 1 and
is acceptable under the OSHA Occupational Noise Exposure regu-
lations. It is usually less bulky, lighter, and less expensive than
the Type 1 SLM. The SLM consists of an omnidirectional microphone,
a calibrated attenuator, a stabilized amplifier, an indicating
meter, and weighting networks. The networks provide the three com-
mon sound level meter filters: A, Band C, and can be set to
either Fast or Slow response. More details on types of meters and
their manufacturers are given in the discussion of diagnostic meas-
urements later in this chapter.
15
When the noise levels are known to change very little through-
out the working day, a simple SLM reading suffices. However, the
reading must be taken properly. The standard procedure is to lo-
cate the microphone at the ear position concerned, but with the
worker at least three feet away. This is the so-called free field
measurement that is specified as preferred in American National
Standard Sl.13-1971, "Methods for the Measurement of Sound Pressure
Levels." For a general standing position, the preferred microphone
height is 1.5 m (about 5 ft); for a seated worker, 1.1 m (about
3.5 ft).
When it is necessary to make sound measurements that will

withstand scrutiny in the courts, several criteria are important:
(1) The data should be obtained by a qualified and dis-
interested individual, to avoid charges of bias. .
An acoustical consultant of known accomplishment,
preferably also a registered professional engineer
is appropriate.
(2) The instruments and measurement procedures used

should conform fully with the applicable American
National Standards.
(3) Instruments should be calibrated before and after

each significant set of readings. If the cali-
bration is out of tolerance, readings back to
the previous calibration must be repeated.
(4) The calibration should be traceable to the National

Bureau of Standards.
Obtaining reliable data depends on periodic calibration of

the instruments. The preferred calibrators deliver an acoustical
signal of known frequency and sound pressure level. Some calibra-
tors provide a variety of signals of different frequencies and
levels. To ensure that the calibrators are correct*, it is best
to buy two and reserve one (as a comparison instrument) which have
been certified as to sound level by means traceable to the
National Bureau of Standards.
The manufacturer's instructions for holding the SLM should be

followed. The practice in the United States is to calibrate the
microphone for a correct reading ~~0n the sound grazes the diaphragm
* For moderate use, annual calibration of the calibrator is often

satisfactory.
16
of the microphone. Thus, the axis of the microphone is at a
90-degree angle to the direction the sound is travelling. Instru-
ments made in Europe (such as the B&K) are calibrated to be correct
when the microphone is aimed at the source. However, these instru-
ments can be used with a random incidence corrector fitting, which
effectively converts them to the u.s. type of use.
For both types the observer, microphone, and sound source must
be at right angles to the direction of the sound. To be absolutely
sure there is no i~terference from the body of the observer, posi-
tion the microphone at least a meter away from the observer. With
steady noise conditions, possible interference can be determined
easily: with the microphone at a fixed position, decrease the dis-
tance between the observer and microphone to less than 1 m. The
position at which a significant change in reading occurs indicates
(for that specific situation only) how far away the observer need
be.
For most industrial situations, a reading on the slow A-scale

is specified for compliance measurements. Despite the averaging
properties of the "Slow" setting, industrial noise is often vari-
able enough that reading the meter becomes a problem.
As a general matter, it is best to first explore A-levels in

the region of interest before obtaining the final sound level for
compliance measurements. Sometimes directional effects can change
the reading a few dB in a few feet. One example is a noise source
that is partially shielded by a machine structure, with the oper-
ator in and out of the acoustical shadow. Several readings may be
needed to completely delineate the noise in the range of positions
used by the worker in question.
In sampling noise for obtaining statistical data using the

"Fast" setting on the meter, you should allow no less than one-half
second between observations; in using the "Slow" setting on the
meter, an interval of at least two seconds is recommended. Machin-
ery that operates at varying speeds or cycles of operation produces
noise that fluctuates ln level. Cyclical noises are those that
change repetitively in excess of 5 dB. A suggested sampling method
is to take readings, with the SLM set to slow response, every 15
seconds for a period of 3 to 5 minutes from which an average value
can be calculated.
It is often necessary to record a single reading that is the
average over a short period. The following rules may be used.
(1) If the difference between average minima and

average maxima is less than 6 dB, use the average
of these two extremes.
17
(2) If the difference is greater than 6 dB, use
the reading 3 dB below the average maxima.
(3) It is ~lways preferable to record the range

of readings if over 6 dB, plus comments on
probable cause. Typical causes include machine
cycling and very low frequency pulsation from
air handling equipment.
When an operator performs various duties during the working day,

he can be exposed to varied permissible noise exposures specified
by the Occupational Safety and Health Administration (OSHA). These
can be used to determine the total daily noise dose, as follows:
When the daily noise exposure is composed of two or more noise
exposures of different levels, their combined effect is the sum
of the fractions, at all levels, of the actual time (hr) divided by
permissible time. The sum of these fractions must be equal to or
less than 1.0 for a permitted daily noise dose.
Permissible noise exposures are:
Hours/Day dBA Hours/Day dBA

8 90 l~ 102
6 92 1 105
4 95 ~ 110
3 97 ~ or less 115 maximum
2 100
Exposure to 'steady noise of 115 dBA is the maximum allowed. Expo-

sure to impulsive or impact noise should not exceed 140 dB peak.
This cannot be measured on the basic sound level meter and requires
more expensive special meters.
Some general precautions should be observed when using the

sound level meter. Wind or air currents can distort readings (a
wind screen attachment will minimize this problem and also protect
the microphone in dirty industrial environments). Vibration of the
meter, supported or on a tripod, can distort readings. Excess room
humidity or temperature can also be a problem, as well as magnetic
distortion of the meter from adjacent power equipment. Barriers or
walls can obstruct sound and reduce measurements of a SLM, or, by
reflection, increase measurements. Background noise should be at
least 10 dB below the total measured level. (The instruction book-
let with the meter gives corrections to readings for level differ-
ences below 10 dB.)
To avoid dropping the meter when it is hand held, keep the

safety cord that comes with most meters wrapped around your wrist.
18
Many other nonacoustical data are required to support the
noise dose information. Such data include plant location and
product, pertinent personnel and their positions in the organiza-
tion (especially important if they are to be witnesses), persons
present during measurements, time span of measurements, room lay-
out and dimensions, sketches of machines, descriptions of machines,
production data, the daily time machines are producing, worker and
measurement locations and photographs.
Dosimeters. If the employee (such as a maintenance or repair

person) moves through differing noise situations, a dosimeter is a
useful instrument for measuring noise. At the present time, there
is no completed American National Standard covering the perform-
ance of dosimeters, although these instruments have been available
for at least two years. Calibration is essential to dosimeters,
and the manufacturer's instructions should be followed until stand-
ards for use are available. Any newly received dosimeter should
be left on continuously for at least eight hours and calibrated
periodically to determine if there are any shifts in reading.
Some observation of the employee wearing the dosimeter may be

needed to attest to normal activity. By deliberately favoring high
or low noise level positions, the employee can influence the indicated
dose upward or downward. Readout should'not be available to the
employee directly from the instrument, lest the employee be encour-
aged to play with the instrument. The dosimeter should contain
simple provisions (as circuit board replacement) for changes antic-
pated in OSHA regulations.
Diagnostic Measurements
Diagnostic measurements are made to aid engineering control

of noise by locating specific noise sources and their magnitudes
and by indicating the types of controls needed, their locations,
and the amount of reduction sought. Such diagnostic measurements
are not always straightforward, and their success depends mostly
on the capabilities and experience of the person making the meas-
urements. In a very real sense, diagnostic measurements of noise
constitute a sort of detective activity, in which clues must be
properly read, hunches followed up, and data cross-checked.
No single detailed method can be specified, for each situa-

tion is unique. However, workable general procedures and the
equipment used are described below.
Sound Level Meters. Diagnostic measurements can be made with

the Type 1 or Type 2 sound level meters that are limited to A, B,
and C scales (B is not normally used) and "Fast" or "Slow" set-
tings. The "Slow" setting is the basis for OSHA compliance
regulations.
19
To the basic sound l~vel meter, special attachments are
available, including:
Windscreen
Earphone coupler
Vibration pick-up and adaptor
Tripod
Extension cable for microphone.
This equipment package can be used to solve many noise source

problems for which specific data based on an octave band analysis
are not required. For example, absorption of a material is
frequency dependent, but using the absorption data at 500 Hz
will generally result in a satisfactory result because the reduced
absorption at lower frequencies is offset by the lesser importance
of these frequencies in an A-scale reading. Although these readings
may result in using slightly better materials than the absolute
minimum, in most cases this cost difference on a single installation
is small. So the limitations of measurements with only A- and C-
scale are of degree only. Many acceptable solutions have been
developed without full octave band data. Many of the case histories
included in Chapter 5 used only these basic data.
Table 2.1 lists the"suppliers and model numbers of several

sound level meters. The table also provides data on presence andl
or availability of A, B, and 0 scales, "Fast" or "Slow" meter
movements, calibrator, windscreen, and microphone extension cord;
decibel range (for Type 2); and availability of an octave band
analyzer, accelerometer, and amplifier.
Following Table 2.1 is a list of suppliers who state explic-
itly that their sound level meters meet at least Type 2 specifica-
tions under American National Standard Sl.4-l97l. These specifi-
cations" are somewhat more rigorous than those of IEC 123, which are
sometimes acceptable. If you are not sure that an instrument
thus qualifies, any reputable manufacturer can supply the backup
data.
20
TABLE 2.1 - Sound level mete rs and accessories available
Scales Speed Capabilities Accessories

Name, Model A B C F S Peak dB Ranqe Cal WS EXI OBA Acc AnlDl
ACOS SLM 3 (1) X X X 50-130 X X X X X
Advanced Acoustical Research 101A X X X X X 40-140 X X X
Amstro SL-1 (1) X X X X X 40-125 X X X
Bausch & Lomb X X X X X 40-140 X
Bendi x 300 X X X X X 55-140 X
B & K 2205 X X X X X 32-140 X X
B & K 2209 (2)
Cast Ie CS 16A
X
X
X X X X
X
X 6-140
60-130
X
X (3)
X X X X
-
Castle CS17A X X X X 20-140 X
Columbia SPL-103 X X X X X 40-140 X
Dahlberg GS-9510 X X X X X 20-130 X X X
--
Dawe 1400G (1) X X X X X 24-140 X (3) X X X X
Digital Acoustics DA100 (2) X X X X X X 14-154 (4)
Edmont 60-510 X X 60-120 X
Electro-Sonic 2010 X X X X X 40-140
General Radio 1565B X X X X X 40-140 X X X X
General Radio 1933 (2) X X X X X X X X X (4) X
Mamash S716 X X X X X 18-140 X X
Mechanalysi s 308 X X X X X 45-140 X X X
Metroson i cs dB-201 X X X X X 40-140 X X
M-S-A X X X X X 32-140 X X
Peekel GRB (1) X X X X X 20-150 (4) X
Pulsar X X X
Quest X X 60-120 X
RAC SPL-103 X X X X X 40-140 X
Rhode & Schwartz EL T (2) X X X X 55-120 X X X
Rhode & Schwartz E ZGA (2) X X X X X X 20-160 X X X X X
Scott 452 X X X X X 40-140 X X
Tracor SPL-103 X X X X X 40-140 X
Tri pi ell 375 X X X X X 40-140 X X
Unico SLP-21 (1) X X X ? 25-130 X X
F ~ last speed
S = slow speed
Cal = Cal i brat or Notes:
WS = wind screen
Ext = extension cord for microphone (1) Conforms to standard IEC 123, slightly less rigorous than ANSI
Peak = true peak reading capabil ity S1.4-1971, Type 2.
OBA = octave band analy zer (2) Conforms to standard ANSI S1.4-1971, Type 1, precision.
Acc = accelerometer (3) Calibrator is of falling ball type.
Ampl = amplifier for extending low level range (4) Octave band analyzer Integral with instrument.
21
METER AND MANUFACTURER/SUPPLIER ADDRESSES
ACOS: Cowl Industries Ltd., Milwaukee, WI
Advanced Acoustical Research Corp., Sawmill River Road, Elmsford,

'NY 10523
Amstro Corp., 120 Clinton Road, Dept. D, Fairfield, NJ 07006
Bausch & Lomb, Rochester, NY 14602 (as Columbia)
Bendix: National Environmental Instruments, Inc., P.O. Box 590,

Warwick, RI 02888
B&K Instruments, Inc., 5111 W 164 Stree~ Cleveland, OH 44142
Castle Associates: R-D Eck, Inc., 12 Dale Street, Waltham,

MA 02154
Columbia Research Laboratories, Inc., MacDade Blvd. & Bullens

Lane, Woodlyn, PA 19094
Dahlberg Electronics, Inc., P.O. Box 549, Minneapolis, MN 55440
Dawe Instruments Ltd., Concord Road, Western Avenue, London W3,

England
Digital Acoustics, Inc., 1415 E McFadden, 'Santa Ana, CA 92705
Edmont-Wilson, Coshocton, OH 43812
Electro-Sonic: Monotherm Insulation Systems, Inc., 551 So.

Yosemite Ave., Oakland, CA 95361
General Radio Co., 300 Baker Avenue, Concord, MA 01742
Mamash Applied Science Laboratories Ltd., Ramat Aviv, Tel Aviv,

Israel
Mechanalysis: IRD-Mechanalysis, Inc., 6150 Huntley Road,

Columbus, OH 43229
Metrosonics, Inc., Box 18090, Rochester, NY 14618
Mine Safety Appliances Co., 201 N Braddock Ave., Pittsburgh,

PA 15208 (as B&K 2205)
Peekel Laboratorium voor Electronica N.V., Alblasstraat 1,

Rotterdam, The Netherlands
22
Pulsar Instrument Co., 650 So. Vaqueros Street, Sunnyvale,
CA 94086
Quest Electronics, 510 Worthington Street, Oconomowoc,

WI 53066
RAC: Research Appliance Co., Rt. 8, Allison Park, PA 15101

(as Columbia)-
Rhode & Schwartz, Muehldorfstr. 15, 8000 Muenchen 80, West

Germany
Scott Instrument Laboratories, 533 Main Street, Acton,

MA 01720
Simpson Electric Co., 5200 W. Kinzie Street, Chicago, IL 60644

(as General Radio)
Tracor Medical Instrument Division, 6500 Tracor Lane, Austin,

TX 78721 (as Columbia)
Triplett Corp., Bluffton, OR 45817
Unico: National Environmental Instruments, Inc., Fall River,

MA 02720
23
Octave Band Filters. This equipment is normally an attach-
ment to a sound level meter. It measures the sound pressure
level in each octave band, normally from 63 Hz to 8000 Hz. Sepa-
rate octave band analyzers often include weighted band readings
of A and C (slow and fast) plus peak readings. This instrument
can spot the band in which a maximum occurs. This aids in deter-
mining the noise source, as for example in verifying that the
blade-pass frequency of a fan is in the band with maximum level.
With the data from the octave band analysis, barriers and enclo-
sures may be designed in detail, using the prediction of the re-
quired attenuation as a function of band frequency .. By investi-
gating the source in the machine having the same frequency as in
the band maxima found at the operator position, you can determine
the exact reason for the noise level and possibly change it by
redesign, local barrier or enclosure, damping, or other methods,
which will be discussed in Chapter 3.
Fractional Octave-Band (Proportional-Band) Analyzers. De-

pending upon the degree of exactness required, proportional-band
analyzers can be obtained for 1/3 octave or 1/10 octave. This
fine resolution would find critical use where machine redesign is
contemplated and all specific noise sources need to be identified
to lead to corrective design to reduce noise. This is normally
not the problem on existing equipment, for which noise can
usually be controlled by barriers, enclosures, isolators, damping,
and so on. Exceptions occur when sources of single frequency
noise may be close together in frequency, and fine resolution may
be required . . The usual example is that of isolating the contri-
butions from gearing systems, where several gear tooth passage
frequencies (and harmonies) may be close together.
Photographic Equipment. The role of photography in documen-

ting the situation cannot be over-estimated. You will need a
single lens reflex camera, a 28 or 35 mm close-focusing wide angle
lens, and an electronic flash, preferably with automatic control
and remote sensor. The flash should be \vide angle. Be careful
with the flash gun if explosive atmospheres may be present; you
may not be allowed to use it. If at all possible, include a scale
of length in all photos. Use Tri-X or similar high speed film.
If you cannot use flash, a tripod may be needed for time exposure
photographs. An essential part of the photographic effort is doc-
umenting the acoustical and production ronditions for each shot.
If you have a plan qr sketch of the workroom, establish a dimen-
sion grid and refer to that in your notes on each shot. The first
exposure in each roll should be a close-up of a sheet of paper on
which you indicate the necessary identification information.
Start with establishing shots that show the entire area and the
relations of the machines to each other. Take pictures from sev-
eral angles. Then corne in closer to show the whole machine, again
at several angles. For every microphone position at which sound
24
level data are obtained, photograph the setup. Use close-ups for
added clarity. When noise sources are identified or even suspected,
photograph their locations from several angles, showing the rela-
tions to other parts of the machine.
The photographs will be useful for documenting the measure-

ments, for revealing machine features that must be considered when
noise control means are designed, for showing worker positions, for
identifying competing noise sources, and for possible evidence in
any legal procedures.
Tape Recorders. A suitable battery-operated magnetic tape

recorder can be useq to obtain a noise sample for repeated analysis
in the laboratory. A major difficulty is that tape recorders are
adjusted for optimum operation with speech or music signals. Indus-
trial noise signals may be different enough in peak-to-average
level characteristics and in spectrum shape that the full perform-
ance of the tape recorder cannot be realized. Competent advice
should be sought before purchasing a tape recorder for noise. Some
instrument manufacturers are now supplying such tape recorders.
A convenient combination is a precision (Type 1) SLM and a

two-channel (stereo) tape recorder. The SLM serves as the iriput
device to the tape recorder. By proper setting of the controls,
and with a knowledge of the peak levels present, it is possible to
make tape recordings from which most of the original information
may be recovered. There is one exception: for complex noise sig-
nals, the original and playback signal waveforms will usually not
be the same. Hence, do not expect a tape recorder (unless spe-
cially adjusted or equalized) to provide the same peak-to-average
level information that was in the original noise signal.
To recover actual levels from the tape record, you must record
an acoustic calibration signal of known sound level. This signal
is conveniently supplied by the acoustic calibrator used on the
microphone of the SLM. Use it to calibrate the SLM, and connect
the latter to the signal channel of the tape recorder. The tape
recorder should have a meter that indicates recording level. By
a series of qualifying measurements on the tape recorder, you
should have established for each octave band the level above the
top end of the meter scale at which distortion becomes excessive,
and the level below th~bottom end of the meter scale at which in-
ternal noise limits the signal that can be recorded. The distor-
tion limit is often considered to be the level at which the play-
back is 1 dB below that expected on a linear basis.
While the calibration signal is being recorded, use the other

stereo channel for voice recording of notebook information: data,
time, job number, location, description of setup, description of
calibration signal, and most importantly, settings of all controls
25
on SLM and tape recorder, together with the time, and a brief
description of the content of the run.
On playback, the scheme for recovering the absolute levels,

depends on that fact that the output voltage from the SLM depends
only on the reading of its meter and not on the settings of its
attenuator(s). For example, suppose that the calibration siqnal
sound pressure level is 114 dB and that the meter scale is 0 to
10 dB. Then the SLM range switch will be set at 110 dB, and the
meter movement will read 4 dB. The output voltage from the SLM
due to the calibration signal is then recorded. Next, suppose
you tape a noise that fluctuates between 85 and 95 dEC (not dBA)
Then you set the SLM range at 90 and note this setting on the
voice channel. On playback into a suitable system (which may be
the same SU1) with a decibel output scale, you set the controls
so that the calibrate signal reads 4 dB on the output meter.
Then when the unknown noise signal is played back, zero dB

on the output meter corresponds to the range setting on the SLM
when the signal was recorded "flat", that is, with the dBC or
preferably "linear" setting. Then suppose on playback you wish
to insert an A-weighting filter; if the SLM is being used as the
playback indicator, set it to A-scale. If the output meter reads
2, then the level was 92 dBA (original recording with the SLM
range at 90).
If the original signal has large low-frequency content, such

as noise from process air blowers, it may all too easily overload
the tape recorder. For such signals, recording can be done with
the A (or B) scales to remove some of the excessive low-frequency
noise. Thus, on playback using a "flat" system, the total level
will be in dBA (or dBB). Octave band levels can be then approxi-
mated by adding the weighting for band center. However, if the
spectrum slopes rapidly, additional and more uncertain corrections
are needed.
Another use for a tape recorder is as an electronic notebook.

In this use, all data, descriptions, comments, and suggestions for
noise control can be placed directly on tape. Use a cassette
recorder, preferably one taking standard cassettes. It should also
have an automatic volume control and an automatic voice switch to
initiate recording simply by talking. Set recording level low so
that you must speak loudly when the microphone is directly in front
of your mouth. This procedure will increase the signal to noise
ratio so that you can make understandable tapes even when the noise
is in excess of 105 dBA.
Other Equipment. For diagnostic measurements you will need,

as described above, a standard Type 1 SLM,·preferably with peak-
reading capability; octave band analyzer; calibrator; heavy-duty
26
tripod strong enough to hold the meter; extension cord; lightweight
tripod to hold microphone on extension cord end and also to serve
as a wand for probe measurements; accelerometer and integrator net-
work; tape recorder; voice microphone for tape recorder; and a
camera, flash, and tripod.
Other equipment that you will find useful in diagnostic meas-

urements includes a'time and motion study stop watch calibrated in
hundredths of,a minute (not seconds) for measuring exposure time; a
steel tape calibrated in both inches and centimeters; an optical
range finder (to 150 feet) for getting dimensions (es~ecially ceil-
ing heights) when building or workroom plans are unobtainable; an
inclinometer for getting heights of stacks beyond the range of the
rangefinder; a hand tally counter for counting repeated operations,
such as a punch press; and a vibrating reed tachometer for obtain-
ing rotational rates of fans, etc~
Making readings. At the worker's ear, obtain the peak sound

level, the A and C levels, and the octave band sound pressure
levels from (at the most) 31.5 Hz to 16,000 Hz (usually 63 through
8000 Hz suffices). Be sure the same machines are operating through-
out the measurement period. The A-level weights each octave band
level to make the total consistent with the judgement of the human
ear. Conversion of flat response to A-level was discussed in
detail in Chapter 1 (Table 1.1).
Relating to A-levels. A useful technique is to identify the

A-weighted band that contributes most to the A-level and set the
octave band analyzer for the frequency of this maximum band. Then
with the probe microphone, locate the area on the machine respon-
sible for the maximum contribution in this band.
If a peak reading SLM is available, it can be used to advan-

tage when you suspect that impacts or high frequency noise bursts
are important components of the noise. Maximum peak level per-
mitted by current regulations is 140 dB. For many industrial
noises, the peak sound level minus the A-level is in the range of
14 to 16 dB. For sounds like chains rattling against metal panels,
this difference may exceed 30 dB. Candidate sources of impact
noise include metal-to-metal contact, rapid linear acceleration of
parts, short duration air noise bursts, and metal parts traveling
down metal chutes. The peak-to-dBA difference should increase as
the impact noise source is approached.
A useful field technique in diagnostic measurements is to

observe the difference between the A-level and the maximum A-
weighted octave band. The greater the difference, the more likely
it is that several A-weighted octave bands of approximately the
same level are contributing to the total A-level. If the differ-
ence is small, it is likely that pure tones are causing the level
in one A-weighted octave band to be much greater than others.
27 .
When the C-level is greater than the A-level, there are
strong low frequency components, but when A is greater than C,
the energy is mainly in the high frequency bands.
For making close to the source readings, many sound level

meters come with an extension microphone cable that can be used
to probe into those portions of the machine at which candidate
noise sources are located. Sources include surfaces excited, for
example, by rotary unbalance, vibrating panels, metal-to-metal im-
pact, vibration of unrestrained workpieces, rapid acceleration of
machine parts, friction, worn gears, poor bearings, hydraulic
systems, air and stearn discharge, and materials handling noise
(chutes, conveyors). The action needed to reduce the A-level of
noise can bp selected in part by a knowledge of the frequency
region that contributes most to the A-level. You will rarely
find that the bands centered at 250 Hz and below contribute sig-
nificantly to the A-levels of noise (see Table 1.1 in Chapter 1)
For most industrial noises, the maximum A-weighted contribution
is in the 500 to 2000 Hz rpgion.
Vibration. Vibration measurements can often be used to

verify the connection between received sound and its source. Can-
didate locations are thin machine structural panels and belt
guards. However, there is no reliable connection between amount
of vibration and A-level. Nevertheless, in more difficult diag-
nostic problems, vibration measurements can prove useful. If the
spectrum of the average alternating acceleration of a surface is
similar to that of the noise around it, it is likely that most of
the noise below about 400 Hz carne from that surface. Above 400 HZ,
average surface velocity is a good indicator. Some sound level
meters have vibration attachments based on accelerator input. With
some meters, a resistor/capacitor integrating network is available
so that acceleration, velocity, or displacement may be read from
the SLf1. There is as yet no standard means of calculating noise
levels from vibration measurements, especially at higher frequen-
cies. Use the mean square reading over a surface, rather than a
single reading. .
Occasionally vibration requires analysis with more sophisti-

cated instruments, such as fixed narrow band analyzers, or 1/2,
1/3, or 1/10 octave analyzers. Such instruments are usually used
in the laboratory, but field use is possible.
Ears and Mind. Trained ears play an important role in diag-

nostic measurements. They serve as spectrum analyzers and direc-
tional indicators. Masking of the observers own speech can also
provide rough indications of sound level, once a personal cali-
bration has been developed. The observer, must be listening criti-
cally during all measurements to detect when an unusual noise
event may have occurred. This event is often a change in operation
of a machine and may require a new set of data on that condition.
28
The high-frequency directionality of the ears can be greatly
enhanced by cupping the hands behind the ears. On turning the
head, it will be easy to localize the noise sources in the plane
of rotation. By doing this at a couple of positions around the
machine, you can often home in on a suspected source location.
Because most of the A-level is ordinarily due to higher frequency
noise, this localization by ear is very helpful.
Another clue to source location is the shielding of noise by

a part of the machine. If the ear or a probe microphone is swept
past the suspected region, there will be a sudden drop in level
(particularly in high frequency octave bands) as the source is
eclipsed.
General Tips. Diagnostic measurements are directed toward

noise from specific machines, not the noise from many sources that
is received by a worker in a compliance measurement situation.
Ideally, only the machine in question should be operating to avoid
contamination of the measurements by noise from other sources.
The machine may then be put through its several operating modes:
idling, full production, or with some functions selectively dis-
abled. These modes of operation aid in identification of noise
sources that may be close together or have noise levels nearly
identical.
This ideal arrangement is usually impossible, and measurements

must ordinarily be made during normal production. However, if the
noise is relatively continuous and does not contain audible pure
tones, it is possible to calculate for each octave band the level
due to the machine alone. The only requirement is that at some
time during the measurements, the machine in question must be off.
This often occurs, owing to jams, maintenancei rest breaks, or
change of work. Measure the noise from all other sources, as well
as the noise with all machines operating. If there are small
changes between the two noise level measurements, it is possible
to infer the noise level (in each octave band) from the machine as
if it alone were operating. The principle is simple: the total
noise is the sum of the two contributions: machine and everything
else. However, since decibel levels do not add directly, cal-
culation is needed. The scheme works best when the difference of
the two levels is over 2 dB.
Combining Decibels. Decibel levels cannot be added directly.

If there are two equal noise sources, the combined noise level is
3 dB greater. If the two noise sources are unequal, a smaller num-
ber of decibels on a graduated scale of difference is added, as
discussed in Chapter 1 (see Table 1.2 and Figure 1.3).
This same method can be used in combining octave band values

two at a time to get overall spectrum (flat) values, or if octave
29
bands are A-weighted, to obtain overall A-values (approximately
what the ear hears) .
Thus, when data are available for levels from several ma-
chines, as explained above, and the decibel decrease is recorded
when the machine under study is shut down, you can determine the
decibel decrease to be subtracted from the reading with all ma-
chines on to give the contribution from the machine under study.
In the following list, a is decibel decrease when the machine under
study is shut down, and-b is the number of decibels to be subtrac-
ted from the reading with all machines on to give the contribution
from the machine under study.
a b a b a· b
0-:-5 9-:6 4-:-5 1-:9 8-:5 0-:-7
1.0 6.9 5.0 1.7 9.0 0.6
1.5 5.3 5.5 1.5 9.5 0.5
2.0 4.3 6.0 1.3 10.0 0.5
2.5 3.6 6.5 1.1 11.0 0.4
3.0 3.0 7.0 1.0 11. 5 0.4
3.5 2.6 7.5 0.9 11. 5 0.3
4.0 2.2 8.0 0.7 12.0 0.3
Final levels should be rounded off to the nearest decibel.
Decibels can also be combined mathematically by using the

formula in constructing the above table:
L2 = L 10 log (lOB - 1)
l + (2.1)
where B = (L
c
- Ll)/lO
L = Combined level, background plus machine level

c
L = Background level (all other machines)
l
L = Machine level
2
Data Requirements. Depending on the noise source and the
possible methods of noise reduction contemplated, some or all of
the following data may be required:
(1) Sound pressure levels, A and C, at location of interest
(operator position) for compliance measurements plus
peak.
(2) Octave band analysis made I ft from the source, or at
any distance for which noise from that source
predominates.
30
(3) Octave band analysis made I meter from a surface
that envelopes the source. Use in sound power
calculations.
(4) A and C at critical distance: the point where

there is a 3 dB increase when approaching the
noise source, above the generally constant level
remote from the source (only that source operating) .
(5) Background noise level: A and C, and octave band.
(6) Noise source probe readings in near field, comparing

fast and slow readings related to machine operation
if possible. Use for diagnostic measurements.
(7) Sketch of area, machine, noise source, and operator

location. Augment with photographs if required for
later analysis.
(8) Vibration measurements to determine acceleration

levels of vibrating surfaces (in some cases velocity
and displacement) plus description of vibrating region.
For isolation requirements, obtain weight on footings
and lowest forcing frequency (rpm of motor, drive, and
other vibration contributors) .
(9) General data, such as type and description of machine,

type of stock and product, production rate, require-
ments for monitoring and access, use of automation,
locations with respect to other machines, dimensions
of enveloping surface, times of measurements. Also
information on all personnel concerned.
(10) Plot data on octave band form (Codex graph paper

No. 31460 or equivalent) .
The basic aim of this documentation is to record as clear an under-

standing of the noise sources as possible for detailed analysis
later. The documentation should be such as to be clearly under~
stood by others and to serve as reminders of the problem details
for final corrective recommendations.
Several things about the noise can be learned from the basic
sound level readings, A and C, and the octave band data (Items
(1) through (3) above):
C greater than A indicates low frequency components predom-

inate
A greater than C indicates high frequency components
predominate
31
Peak minus A indicates impact when greater than 14-16 dB
A minus maximum A-weighted band of 6 dB indicates possibility
of 4 high A-weighted octave bands

of 3 high octave bands

of 2 high octave bands
A minus maximum A-weighted band of close to zero indicates

strong pure tones
The data about an enveloping surface can be used later if they are
found to be an advantage in estimating sound power. Sound power is
estimated by placing an imaginary surface one meter away from a
surface that envelopes the machine and touches it at all essential
areas. Measure the noise in each octave band one meter away from
this imaginary surface where the noise (A-weighted, for example)
changes by about 5 dBA or more. The enveloping surface is quickly
defined by drawing a box around the machine with a chalk mark on
the floor. By moving the microphone over a surface one meter away
from the enveloping surface, you can locate each-measurement spot.
If this is done carefully, the value of the sound power can be ob-
tained. The sound power level associated with a given surface is
, L = L + 10 log S (2.2)
w P
where L is the sound pressure level,
p
S is the area in m2 at each position, and
L is the sound power level in dB re 10 -12 watt.

W
Since equation (2.2) must be used for the l~vel of each octave band
for the total surface area, you must use the combination rules de-
veloped in Chapter 1 to obtain the total power for each octave band
and then the total A- or C-weighted power. Sound power information
will be used to determine expected sound pressure levels if room
acoustics are changed, which will be discussed in Chapter 3.
Follow-Up Measurements. After the changes have been made to

reduce the noise level (to be discussed in Chapter 3) I a follow-
up series of mpasurements should be mdor to record the attenuation
achieved. These measurements arp usu~l ly made at the operator
position, and revised noise dose is calculated to determine if the
results are now in compliance ~ith current regulations. This gives
a permanent file record covering the original noise level and noise
dose calculations, method of noise reduction or insertion loss* of
noise attenuator, and final noise level and noise dose calculations.
Such data are useful, should a similar problem arise in the future.
* Insertion loss is defined as the reduction in level at a measure-

ment point after noise control means have been introduced.
32
Chapter 3
NOISE CONTROL TECHNIQUES
The objectionable noise sources have been identified and meas-

ured as outlined in Chapter 2, Measurement of Noise. In noise con-
trol analysis the next approach is to consider where the most effec-
tive and economical control of the noise can be obtained: at the
source, the path, or the receiver, and in what order, if more than
one treatment is needed.
Source
A full knowledge of the operation or process is needed to an-

swer t~e basic questions: Can the machine or operation be elimi-
nated or replaced by a quieter operation with equal or better effi-
ciency (weld for rivet, etc.)? Can the present noisy machine be
replaced economically by newer equipment designed with lower noise
levels (perhaps as well for increased production)? Can the specif-
ic noise source be corrected by minor design changes (avoiding .
metal-to-metal contact by use of plastic bumpers, reolacing noisy
drives by quieter types, use of improved gears, improved mainte-
nancei etc.)?
Can the specific machine elements causing noise be corrected

by a local source approach rather than considering the entire ma-
chine as a noise source? (Later data covering the specific designs
for barriers, enclosures, damping vibrations, vibration isolation,
lagging of vibrating surfaces, mufflers and silencers for air and
gas flows, reduced air velocities of free jets, etc., may be con-
sidered for the individual noise producing elements of the total
operation.) Can the noisy machine elements be moved (such as
pumps, fans, air compressors that service the basic machine but
need not be an integral part)? Can the machine parts be vibration
isolated to reduce airborne noise from vibrating panels or guards?
The major constraints in changes of noise sources will be
stated including interference with the productivity of the machine
or the operator, and the relative costs of alternative choices
developed. Some of the specific noise reduction methods that can
be applied at the noise source are:
1. Distance or relocation
2. Vibration control or isolation
3. Damping of vibrating surfaces
4. Lagging of vibrating surfaces
33
5. Air (and gas) flow and jet noise reduction by
design, mufflers
6. Hydraulic systems, acoustic filters, surge tanks
7. Motor air cooling mufflers'
8. Mechanical drive enclosures
9. Balancing of rotating equipment
10. Noise source enclosures
Distance or Relocation. Can a noise source be moved further

away from the operator? Distance reduces noise level from a point
source by 6 dB per each doubling of distance in a free field. In
a closed space such as a factory work room, reverberant sound will
eventually determine when no further reduction by distance may be
obtained. The distance at which this occurs is called the critical
distance (see Chapter 1). Relocation may apply to machine service
units such as pumps, fans, drives, hydraulic systems, and air and
stearn flows that may be relatively easily moved and do not need
constant attention.
Vibration Control or Isolation. Vibration control or isola-

tion can be investigated on two bases: isolate a vibration source,
such as a motor-pump assembly that is part of a machine, or vibra-
tion isolate the entire machine. A separate noise level test with
the component on and off would help to determine if the component
is a major contributor to the total noise.
Vibration control is necessary for quiet operation, and equip-

ment mountings should be reviewed and changed from solid mountings
to vibration isolating mountings of springs, rubber, cork, felt, or
fiberglass as the situati~n requires. Vibration causes flat panel
surfaces of machines, guards, and building structures to radiate
noise. The excitation can come from rotating equipment, vibrators,
tumblers, or shakers, and the introduction of isolation between the
machine and such surfaces is often the complete solution to some
noise problems.
Vibration isolators are commercially available. They are se-

lected by specifying the weight supported, the deflection required,
and the lowest vibratory frequency of the unit to be isolated.
They are made from, elastomers (rubber in compression and shear,
ribbed rubber) ; other compressible materials (cork); fibrous mats
(felt, fiberglass); and steel springs.
Basic {solation requires a knowledge of the lowest forcing

frequency (f) of the machine to be isolated, as related to the
34
natural frequency or the isolator (f n ) plus machine, and the
weight on the footing to be isolated. The transmissibility of
vibratory energy is greatest (and is to be avoided) when the ratio
of f/f n = 1. Isolation begins above f/f n = ~ The isolator plus
machine resonance frequency f n is determined from
f n = 3.13 ~ where d is the static deflection of isolator
under load in inches. This relation holds only when the deflection
is strictly proportional to the load (linear systems).
The steps for selecting spring isolators are (Bell 1973), B.4:
(1) Establish that part of the total weight that is on the

footing in question.
(2) Determine lowest forcing deflection required for degree

of transmission percent required (see Tabl~ 3.1); 5% is
normally adequate.
(4) Choose a suitable isolator that will sustain the load

and have the proper deflection. Isolator manufacturers
often list spring constants (lb/in. deflection).
(5) Ensure that deflection is uniform for each footing.
(6) For more complex isolation problems, submit the data

from vibration tests to isolator manufacturers for more
exact specifications.
TABLE 3.1 - Required static deflection (inches) for common industrial speeds or forcing
frequencies (base is assumed i mmovabl e)
Speed Vibration Transmission (percent)

rpm freq. 0.5% 1.0% 5.0% 10% 25%
3600 60 0.55 0.27 0.06 0.03 0.01
2400 40 1.2 0.62 0.13 0.07 0.03
1800 30 2.2 1: 1 0.27 0.12 0.05
1600 27 2.8 1.4 0.29 0.15 0.07
1400 23 3.6 1.8 0.38 0.20 0.09
1200 20 4.9 2.5 0.52 0.27 0.12
1000 17 7.1 3.6 0.74 0.39 0.18
800 13 5.6 1.2 0.61 0.28
600 10 2.1 1.1 0.49
400 7 4.6 2.4 1.1
35
The selection of isolator pads follows the same general
method, using the data from the suppliers as to the recommended
grade, material, and thickness. However, many pads are highly
nonlinear and cannot be selected directly on the above basis.
A pump-motor set mounted on a common platform furnishes a

typical isolation problem, and also one that has a simple solu-
tion: use four properly selected vibration isolators.
Other vibration problems can be more complex, and the sup-

pliers of isolating devices should be consulted. Give them the
machine weight, operating frequencies, weight distribution on
footings, and test measurements covering acceleration, velocity,
and displacement at machine footings and at other points on ma-
chine to aid in determining the isolator requirements. For exam-
ple, machines with low forcing frequencies may require a heavy
concrete inertia block, generally 1.5 to 2.0 times the weight of
3upported equipment. In addition the inertia block and entire
support structure could rest on spring isolators.
Another problem is a machine on a rather limber floor. Such

designs and specifications call for special expertise. Complex
vibration in more than one plane also requires specialized assist-
ance from the suppliers or a qualified consultant. Under optimum
conditions, the reduction in noise level (in dBA) should range
from 2 dB for a machine with no vibrating panels, mounted on a
very heavy inertia block, to perhaps as much as 14 dB for a heavy
machine on a second story and limber floor.
Damping. Damping is the reduction of vibration and resultant,

noise by adding a layer or layers of vibration-absorbing material
to either side of the vibrating surface. For example, metal parts
striking a sheet metal pan cause noise that can be damped with a
wide variety of materials including roofing paper, sheet lead,
damping compounds, special tapes, acoustic lagging, and other com-
mercially available materials to fit the problem.
Surface acceleration readings made in the diagnostic measure-

ments can point out problem surface areas that are efficient noise
radiators. Power radiated as low-frequency sound from a vibrating
surface can be estimated from acceleration readings:
W (watt) = 5.3 x 10- 2 g2 s2 (3.1)
g = Average (root mean square over surface)

acceleration in g - units
2
S = Area of surface, m
36
The free field sound pressure level from a point source can
be determined from sound power level from the formula:
L = L - 20log r - 0.7 ( 3 • 2)
p W (r = distance from source in ft) .
This relation does not hold close to a machine with many noise
sources, or at a sufficient distance in a workroom where reverber-
ant sound predominates.
The dominant frequency, f , can be estimated from acceleration

D
and velocity:
2
f
D
= a/27Tv a = acceleration (m/sec ) ( 3. 3)
v = velocity (m/sec)
f = frequency (Hz)
Lagging. Lagging of pipes or other surfaces radiating noise
is a combination of a resilient member, noise absorption, and
shielding. It usually consists of 1 to 4 inches of resilient acous-
tic sound absorber covered with a layer or layers of heavy shield-
ing such as sheet lead, lead filled vinyl, or other metal protec-
tive coverings.
Some typical constructions and attenuations are given below

(in dB) :
500 Hz 1000 Hz 2000 Hz
I-in. Molded fiberglass and

aluminum foil covering 1.5 4.8 13.8
I-in. Molded fiberglass and

lead impregnated vinyl 5.0 12.0 24.0
2-in. Molded fiberglass and

lead impregnated vinyl 4.5 13.5 26.0
Notes: 1. Fiberglass is 4 lb/cu ft. Lead vinyl is 0.87 lb/sq ft.

2. Note low attenuation at 500 Hz, less at lower frequen-
cies.
3. Good seal at all joints is critical.
4. In special cases two layers of 2-in. fiberglass plus
lead impregnated vinyl between layers plus a cover
layer of lead impregnated vinyl would increase atten-
uation.
5. Sheet lead of same weight/area would also satisfy.
6. Source: Dear, 1972, P.15.
37
Air and Gas Flow Noise Reduction. To reduce the noise of air
dU':gas flow and discharge, silencers are used: absorptive silenc-
ers contain porous materials to absorb and reduce noise; reactive
silencers depend on reflection (impedance mismatch) of the sound
waves as the basic noise reduction method. A combination of both
types is used in some designs, with various configurations.
Commercially available silencers are specified by the insertion

loss (by octave bands) and by other specifications such as velocity
of flow, temperature, and allowable pressure drop. Large industrial
silencers are also known as snubbers and are sometimes combined with
spark arrestors. There is a great deal of art in silencer design.
One unsolved problem with dissipative silencers is fouling of the
absorbent by particulate matter. In general, under such conditions,
do not use· a dissipative silencer.
Fans and blowers when near or part of an operation can be a

major noise source. Fan types used are propeller, axial, and cen-
trifugal. Blades on centrifugal fans may be radial, forward curved,
or backward curved, but backward curved blades are the quietest.
The resulting air noise is a combination of b~ade-pass frequency and
harmonic peaks plus broadband aerodynamic noise and turbulence.
Reduced speeds will reduce noise, and replacement with lower

noise level fans such as backward curved blade types can be consid-
ered. If this is not practical or economical, the air flow noise
can be reduced by commercial or custom-made silencers. Custom-made
silencers that can be constructed in maintenance shops include
acoustical labyrinths, parallel baffle silencers, acoustic lined
plenums, acoustic lined ducts, and acoustic lined bends ..
If duct walls are lined with an absorbent with absorption

coefficient, a, the dB reduction obtained is given by
6 L = 12. 6 paL 4/ 8 (3.4)
6L = Change in sound pressure level

P = Perimeter of duct, inches
8 = Area of duct, square inches
a = Coefficient of absorption (note that this is
frequency dependent, so octave band data will
be used to determine required insertion loss
and length of duct to be treated.
Plenum chambers can also be lined with sound absorbing mate-

rial. An approximate relation for the reduction in level is:
( 3 • 5)
38
a = Coefficient of absorption of liner
S
p
= Area treated on plenum walls
Sb = Discharge area of blower

An absorbent lined bend should add about 5 dB attenuation,
with length of treatment at about five times duct width. Commer-
cial silencers are available for greate~ attenuation to fit any
fan or duct size, and suppliers can give insertion loss at each
octave band under varied conditions of flow. Note that noise
travels both upstream and downstream, and silencers may thus be
needed on both intake and delivery sides of the fan.
Air Jet Flow. In industrial applications where high velocity

air is discharged, shear-induced boundary layer turbulence can be
a serious noise source. Normally, maximum levels from this aero-
dynamic noise occur in the 2000, 4000, and 8000 octave bands. If
velocity can be reduced, the noise level is reduced. Since air
noise power (from several mechanisms) varies approximately as the
5th power of velocity, noise level reduction can be estimated from:
6L ( 3 • 6)
V
l
= Present high velocity
V
2
= Reduced velocity
A free jet impinging on a surface increases the noise level up to
7 dB.
Low noise level jets are available commercially that use a

divider at the exit to change a single jet stream into multiple
smaller jet streams. These devices should be tested .to see if they
can do the job required. Some hand air guns use a version of by-
pass jet engines, in which a mantle of slower moving air is aspi-
rated over the jet. Jet noise reductions of as much as 10 dB can
be realized.
If the jet in question is from a vent, it could be moved to

be vented outside the plant or it could be muffled.
Air operated tools can be equipped with commercially available

exhaust air mufflers.
Hydraulic Systems (Miller, 1973), P.13. Hydraulic systems, as
part of a machine or operation, can have specific noise sources.
Motor-pump assembly contributes most to the overall noise pattern,
being vibrated both by overall vibration and by the blade tooth or
piston pass frequency plus harmonics. Noise levels depend on type
of pump: gear pumps are more noisy than screw pumps, piston pumps
are in between. Noise levels of most pumps increase with speed.
Some screw pumps are available that, with a quieted electric motor
produce about 81 dBA at 3 ft.
39
Fluid line noise sources are often sharp bends, flow restric-
tions, and undersize sections. These configurations cause cavita-
tion and turbulence, causing vibration and noise. The major fac-
tors that affect fluid flow and possible noise are pipe diameter, 0,
fluid velocity, V, fluid density, p, and coefficient of viscosity,~.
These parameters are correlated by the Reynolds number, N :
R
N
R
= OVP/~ ( 3. 7)
A Reynolds number of less than 1000 will usually guarantee quiet

laminar flow. In general a pipe size large enough to ensure a fluid
velocity of less than 10 fps should be used for water-like liquids .
.Vibration isolation of the pump-motor unit will attenuate noise

from the vibration of the floor. Other noise reduction means in-
clude substitution of a larger pump at reduced speed (Figure 3.1) i
addition of hydraulic mufflers, which may be tuned reactive, un-
tuned reactive, or dissipative. Tuned or side branch mufflers may
be used for limited frequency band attenuation for constant speed
100
.----
-
"," I
" constant power
CD
.."".--
..... ..,..
"0 90
....J " , --'constant di spl acement
w
>
W
....J
W
(f)
"
0
z 80
1800 2700 3600

PU~C:> SPEED - RPM
FIGURE 3.1 - Effect of reduced pump speed on noise level
40
pumps (Figure 3.2). Dissipative mufflers, shown in Figure 3.3,
may be used where the diameter d is given by
d = 0.014 Q ( 3 • 8)
where d = Inner pipe diameter, in.

Q = Flow rate, gal/min
f
length
----T#--1
;\/4,3'\/4, 5;\/4
FIGURE 3.2 - Side branch muffler
11 ------
,o
~----' l_
FIGURE 3.3- Dissipative muffler
Chamber diameter, D, and length, I, are determined from:
f
c = 0.282 cd/IV (3 . 9)
where c = Velocity of sound, fps

f = Cut-off frequency, Hz
c
2
V - 0.78S-D 1
Both types of mufflers could be used in the same system for effec-
tiveness over a wider frequency range. Less attenuation is achieved
if each is used independently for the specific frequency.
41
If the pipe size for avoiding turbulent flow is not econom-
ical, other means of attenuation must be used, such as storing
energy elastically with a flexible hose or surge chamber, use of
elastic spacers, wrapping and supporting of the line, use of iso-
lation type hangers, or increase in pipe wall thickness. Noise
reductions for several pipe wall thicknesses are shown in Table
3.2.
TABLE 3.2 - A-level noise reduction by use of heavier pipe
Noise reduction, dB
Pipe
Diameter Pipe Grades
(inches)
XS80' 120 160' • XXS
-
2 6 8 12 16
4 7 10 13 16
6 8 12 15 18
8 9 14 18 19
10 9 15 19
12 10 16 20
• Extra heavy
•• Double extra heavy
Fluid lines can also be lagged (see discussion of lagging above) .
Motor Air Noise. Motor air noise, if found to be a problem,

can be reduced by acoustic line air flm..' c:hambers which are
available for some motors. Totally enclosed fan cooled (TEFC)
motors with integral quieting are also being manufactured. Less
than 80 dBA at 5 ft is claimed for motors up to 5000 HP.
Enclosed Drives. Vibrating surfaces of a drive enclosure are

noise sources. Damping these surfaces reduces radiated noise. If
the drive enclosure is a steel box structure, the inside can be
lined with a combination absorption layer (such as open cell poly-
urethane foam) with a damping layer backing. Such materials must
be oil-resistant. If there are air vent openings for cooling,
noise traps can be added. (For further details see Case History 7
in Chapter 5.) Quieter drives can also be considered, such as sub-
stituting belts for gears, "silent" chains for roller chains, and
so on. Precision made and carefully aligned gears are always
quieter than other types.
Balancing. It is always preferable to reduce noise at the

source by avoiding the generation of noise. Unbalanced rotating
42
parts cause surfaces to vibrate and generate noise. All signifi-
cant rotating parts should be balanced in situ, that is, in the
machine at its final position. This is-Particularly true for large
fans, where there may be unbalanced aerodynamic forces on the
blades. Rotating mechanisms that drive reciprocating devices can
use counterbalance weights.
Noise Source Enclosures or Barriers. These will be discussed

in greater detail in treating the noise path in the next section
and are noted here to call attention to the fact that the technique
can be used close to localized noise sources. Examples of such
sources are drives, pumps, gears, stearn and air vents, punch oper-
ations (die area), or other localized metal-to-metal contact areas
of a machine.
Path of Sound
After as much noise as 'possible has been reduced at each

source, the next step is to reduce noise along the path from the
source to the receiver (operator). The sound may be blocked by
~coustical shields, barrier walls, partial enclosures, or total
en~los0res. All these techniques depend on interposing in the path
a material, called an isolator, whose wave transmitting properties
are as different as possible from those of the path. In air paths,
such materials are ideally solid, nonporous, and limp. For liquid
paths, a stream of gas bubbles is a good isolator. Along a solid,
path, resilient vibration isolators perform the same function,
weakening the path. Our discussion will be in terms of the most
important path, in air.
Acoustical Shields and Barriers. An acoustical shield is an

approximately square piece of material, usually (for visual access)
safety glass or clear plastic (polycarbonate or polymethylmethacry-
late), placed between worker and noise source. Such a shield is
effective only if its smaller dimension is at least three times the
wavelength contributing most to the noise exposure received. Thus,
shields of reasonable size are effective only against high frequency
sound, such as from concentrated air or gas jets. Examples include
air ejection systems in punch presses, plasma guns, air guns, and
metal spray guns. Improperly designed shields can interfere with
access. However, they can be combined with automatic safety mecha-
nisms to serve a mechanical safety function as well. A shield must
be carefully engiueered to the job if maximum noise reduction is
to be obtained with minimum interference with manual access. For
high frequency noise, reductions up to 8 dB can be obtained.
Shields must be vibration-isolated from the machine.
An acoustical barrier is a much larger shield, typically free

standing on the floor and usually much wider than high. Visual
43
access can be provided by viewing ports of safety glass or clear
plastic. For both barriers and shields, the sound arrives at the
operator around the edges as well as through the material of the
barrier. Because the former path most often limits the attenua-
tion that can be obtained, the transmission loss of the barrier
material need not be large. Indeed, for the best reasonable bar-
rier, a reduction of 15 to 20 dB in A-level is about the maximum
possible. (For a method of barrier attenuation estimating, see
Case History 6 in Chapter 5.) Thus, at best the barrier (well-
damped) need be no heavier than 0.5 Ib per sq ft, if the A-weighted
spectrum peaks at 1000 Hz; less may suffice if the maximum occurs
at higher frequencies.
For best results, at least the machine side of a useful bar-

rier should be covered with an acoustical absorbent, preferably
oil resistant and cleanable. The barrier material can be plywood;
it ordinarily needs no added damping, and 1/2 inch will usually be
thick enough for a 10-dB reduction. If steel is used, damping
must be added to control resonances. Handles and sometimes casters
can be provided on the barrier for ease of moving. If the barrier
is hinged (for stability), the joints can be severe noise leaks.
These' leaks can be controlled by a resilient strip of 1/8-inch-thick
Neoprene over the gap. Use a width of at least three inches, placed
on the concave side of the bend. Fasten by one long edge, with the
other free to slide as the hinge bends.
Machine guards for mechanical protection can often be replaced

by acoustical guards, covered on the inside by absorbent materials.
These guards constitute a type of partial enclosure over the noise
source. If visual access is needed, safety glass or clear plastic
panels can be used. However, the acoustical guard will have to be
more complete than the mechanical, and must be vibration-isolated
from the machine. Much engineering must go into fitting the guard
so as to cover all the noise leaks. The importance of the leaks Cdn
be seen from an example. If the guard is designed for a reduction
of 15 dB and has a total area of 6 sq ft, then a crack around it
that is about 1/4-inch wide will pass as much sound as the guard
material. In most acoustical guards, the leaks will limit the noise
reduction possible; the chief task of the noise control engineer is
to design for minimum leaks.
When a barrier is wrapped around a machine, with its top more

or less open, it becomes a partial enclosure. Such an enclosure can
be effective in reducing noise to workers nearby. How~ver, the noise
escapes through the top and contributes to the reverberant sound in
the workroom. In addition, specular (mirrorlike) reflection from
the ceiling can contribute reflected-path levels that can become ob-
vious when the direct path is reduced by the enclosure, as shown in
Figure 3.4.
44
7'" ceiling
/ '" '" ray grazing top of barrier
/ /
/
'"
/
/
'" '" '" ray reflected from ceiling
/
/ "'-
/
/
'" '" Jl
/
/ recei ver
~
~----.
barri er
source
machine
floor
FIGURE 3.4- Source-barrier-receiver geometry. The angle into acoustic shadow should be greater
than 30° lor at least 70-dB attenuat ion. Cei ling reflection can offset barrier attenuat ion if ceiling
height is less than 7.5 times distance from source to receiver.
These spill-over noise effects can be reduced by covering the

inside of the enclosure with absorbent. Also," suspended absorbents
may be placed over the openings to reduce the escaping noise. If
all other machines in the workroom are quieted, then the ceiling
reflection may become apparent. Such reflections are usually specu-
lar, and the patch of ceiling at which the reflection takes place
can be located geometrically on building plans. Absorbent placed
on the ceiling at this location will effectively reduce the reflected
sound received.
Partial and total enclosures will usually need access for in-
coming material, product, scrap removal, operator, maintenance per-
son, and vision. Doors, windows, and hatches will handle most
access problems, but the usual precautions about avoiding leaks hold
strongly at these openings. Hinged or sliding doors can use a gas-
ket for a seal. A convenient material is the closed-cell foamed
elastomer weather stripping sold with a pressure-sensitive adhesive.
Special acoustical gaskets, designed specifically for sealing leaks,
are also available. For less stringent sealing, the magnetic strip
gaskets used on refrigerator doors supply both seal and positive
closure. Hatches can be dogged down by quarterturn latches.
45
Windows for visual access may need added internal illumination to
make visual monitoring easy and positive. Heat buildup should be
no problem with an open top in a partial enclosure. Noise reduc-
tion also removes acoustic signals that some workers use in evalu-
ating the performance of a machine. Hence, if the reduction is
great enough, acoustic cues may have to be separately supplied.
This is easily done by a rugged microphone (at the site where the
essential information is generated), feeding a small loudspeaker
at the worker position.
Openings for workpiece, product, and scrap flow can permit

noise to escape. Such openings should be in the form of tunnels
lined with absorbent material. The length of the tunnel determines
the amount of noise attenuation obtainable. In the design, theab-
sorbent can be selected for maximum effect on the noise spectrum
at that opening. Use of lined tunnels should be accompanied by
some degree of automation.
Total Enclosure. The next step of noise control, a total

enclosure, has the same problems (and solutions) as the partial
enclosure. The chief added problem is that of heat buildup. This
problem is easily handled by adding a ventilating blower, together
with silencers for both supply and exhaust air. Some internal duc-
ting may be needed if there are heat-sensitive components in the
machine, but these ducts can also selectively supply cooling air
and remove hot air. The minimum flow rate of cooling air, Q (in
cfm) depends on W, the watts of heat generated, and on ~T, the
temperature rise permitted (degree F). For air cooling at sea
level, Q = 1.76 W/~T. More flow is needed at higher altitudes.
There is no doubt that the total enclosure will require a

change in work habits, and as such will usually be resisted by all
concerned. The shock of the change can be eased if the people
most involved--the workers and the foremen--are provided an oppor-
tunity to enter into the design discussions. Enclosures can also
force consideration of modernizing equipment, say, for automatic
feed by conveyor, so that less personal attention to the machine
is required. Such automation may also offset the difficulties that
arise from less free access to the machine. In most instances the
noi~e control engineer will have little difficulty with the acousti-
cal aspects of enclosure design. The chief job is to ensure an
industrially viable design, taking account of the requirements for
access, minimum change in productivity, and minimum installed cost.
To meet these requirements, the noise control engineer must work
closely with the industrial, plant, and process engineer; with
foremen and workers; with maintenance crews; and with management.
As a general matter, enclosures must not touch any part of the

machine and should be vibration-isolated from the floor. Neverthe-
less, the enclosure must be pierced for such services as electricity,
46
air, steam, water, oil, or hydraulic power. These services can be
regrouped, together with mechanical controls, to a convenient loca-
tion where the enclosure panel can be split. A resilient acousti-
cal seal can then be made from two ring-shaped pieces of liB-inch
(or heavier) Neoprene. See Figure 3.5. Slot each piece at the
pipe or conduit and overlap the two pieces with the slots facing
away from each other. Seal the straight edges with strips of Neo-
prene or similar oil-resistant, heavy, and resilient material .
./
/~ -
./ ./~ sealing strip; glue line

split paneis ./ 1/8 X 3 inches Neoprene
"~
semicircular cut-outs
for services split washers of 1/8 inch Neoprene,
3 inches wide on radius
FIGURE 3.5 - Split panels for services
For mechanical controls operating through an arc-shaped hole

in a panel, the seal can be of abutted, multiple strips of Neoprene.
The control lever should be as thin as possible. Better ·yet, re-
place it by a servo control operated from the outside.
Many of the features of a convenient enclosure design are

illustrated in Figures 3.6 through 3.13. The general design is
based on panels secured (by quarter-turn captive screws) to an angle
iron frame (Figures 3.6 and 3.7). Thus, rapid access is provided
for all types of servicing of the machine. This type of enclosure
is to be as close as possible to the machine. Up to 20 dB of noise
reduction is usually easily obtained. The angle iron frame can be
of bolted sections, to permit quick and complete disassembly and
removal.
47
FIGURE 3.6 - Welded angle iron frame. This frame can be welded in segr:;ents
that are bolted together.
enclosure panel
gasket of closed cell toamed elastomer
edge of angle iron frame
.1-4-::------- quarter-turn captive screw fastener: space

about every 6 to 9 inches. Wing screw
fasteners can be used if rapid access is
needed wi thout tools.
FIGURE 3.7 - Enclosure panels secured to frame by quarter-turn fasteners.
48
Machine vibration may still create a problem by vibrating
the floor, which then acts as a resonant sounding board to vi-
brate the enclosure. This problem is easily handled by vibra-
tion-isolating mounts, using steel springs, or elastomers in shear
(Figure 3.8). Special care in design is needed if the exciting
force is of short duration but is repeated, as in a punch press.
Not all vibration-isolator suppliers recognize the need for care-
ful selection of isolators in this special repeated impact situ-
ation. Be sure that you have enough data on the machines and the
isolators to ensure an effective design. You will need data on
three time scales: (1) duration of the impact, (2) time between
impacts: and (3) the minimum period of oscillation of machine on
a suggested isolator.
.. enclosure panel
toe caving, bent 1/8 X 9 inches cold-rolled

steel; 5 inches deep, 4 inches high
vibration isolating mount, elastomer in shear;

static deflection at least 1/4 inch
strip seal to floor, 118 X 3 inches Neoprene
clearance about 1/2 inch
FIGURE 3.8 - Vibration isolation and toe coving
In any machine, the time comes when major repairs are due:
additions or changes may also be called for. The enclosure design
suggested in Figures 3.5 through 3.13 affordS some flexibility in
this regard. The panels can be made separately and fastened in
place with a gasket material (such as weatherstripping) to close
off chance leaks. If the panel material is a metal, its resonances
can be distributed more uniformly in frequency if the panel is
49
enclosure panel; usual I y steel
damping sheet; can also be spray, trowel on
absorbent: seal raw edges with paint.

main surface to be oil-resistant and
cleanable
FIGURE 3.9 - Enclosure panel interior treatment
50
angle iron stiffener for door frame ...
hinge spacer _..
enclosure panel _
hinge .....
door panel -.
gasket of closed cell foamed elastomer

(weatherstri ppi ng)
door panel stiffener of 1/8 X 1 inch _ _- .

cold-rolled steel
FIGURE 3.10 - Door and hatch detail. Interior of doors and hatches have same acoustical treatment as
enclosure panel. Secure doors by vibration resistant latches or by quarter-turn fasteners. Doors and
hatches must make airtight seal to enclosure panel.
enclosure panel
hold down clamp; use quarter-turn screws
I--....L.~ bead of silicone rubber to make resilient

gasket. use silicone oil on window to
get quick release for servicing window
panel frame/stiffener of 1/8 X 1 inch

cold-rolled steel
glazing: Plexiglas, Lucite, Lexan or safety glass
FIGURE 3.11 - Window detail
51
reinforced by bolted-on angle iron (bolting adds more damping than
welding). The stiffeners should be placed so as to divide the
panel into smaller areas, no two of which should be the same size
and shape. Frames for doors, windows, and hatches can also be
used as stiffeners .
. . . . - - - - - - 3 foot inlet silencer 3 foot exhaust si Iencer
j ....- - - - - - a t least 4 feet
flexible collar connector_--.--;----r
vibration isolating
mounts
collection duct
""",---centrifugal blower
FIGURE 3.12 - Cooling and ventilation system. Three-foot silencers are satisfactory
up to 15-dB isolation. .
Windows pose a special problem because they are a weak spot

acoustically. Generally, if more than 20-dB .reduction is needed,
double glazing must be used. The inside layer should be safety
glass, because it must withstand cleaning to remove oil, grease,
and dirt. All panes should be set into soft elastomer gaskets.
Room temperature-setting silicone rubbers are useful here. The
visual access that windows provide should be carefully thought out
in terms of the information the operator needs. Glareless light-
ing of the components to be monitored is helpful. In extreme
cases, closed circuit video monitoring can be used.
A special adaptation of the total enclosure for the machine

is a total enclosure for the operator when this is the more prac-
tical or economical approach. Such enclosures may require air
52
parti al enclosure parti al enc losure
complete enclosure
FIGURE 3.13 - Examples of partial and complete enclosures
53
intake and exhaust fans, with noise traps, lighting, heating, or
in some cases, air conditioning. As in machine enclosures, some
inside absorpt.ion is recommended, such as an acoustic tile ceiling,
and special care must be taken in window and door design to avoid
leaks. See Figure 3.14 for effect of leaks.
90
% opening
70
0.0
50
40
0 30 0.1
w
N
...J
« 20 1.0
w 2.0
a:
m
"0
5.0
Z
0 10 10.0
f=
«
:::l 7 20.0
z
w
~
~ 5 30.0
«.
w 4 40.0
(/)
0 3 50.0
z
o 10 20 30 40 50 60
TRANSMISSION LOSS POTENTIAL
FIGURE 3.14 - Effect of enclosure sound leaks on potential noise reduction
54
The most important information required for the acoustical
design of enclosures is the transmission loss, Lt. This quantity
represents, in decibels, the reduction of sound power in going
through an isolating wall. It is measured by a well-defined and
accepted standard. The transmission loss varies considerably with
frequency, and the term sound transmission class (5TC) is sometimes
used. This is a single number indication of average transmission
loss. However, it applies specifically to speech sounds and acous-
tical privacy requirements. It should be used with caution in in-
dustrial situations, where sound spectra differ from those of
speech and where reduction of A-level is the requirement. If a
single number is needed, use the transmission loss at the band an
octave below that for which the maximum A-weighted octave band
level occurs at the operator position.
The noise reduction, Rn , is the difference in level at opera-

tor and in enclosure after installation of the enclosure. It de-
pends on several factors: the average transmission loss of the
enclosure surfaces, the amount of absorption in the enclosure, and
to some extent, the acoustics of the work room. The worker is
usually close enough to the machine that room acoustics effects are
not important. If the average absorption coefficient of the inte-
rior of the enclosure is at least 25% at the band an octave below
that for the maximum A-weighted octave band level, then approxi-
mately Rn = Lt. A more useful measure of the effect of the enclo-
sure is the insertion loss. This is the reduction of level at the
operator position before and after introduction of the enclosure.
Its calculation requires knowledge of total noise power from the
machine, its directional characteristics, the total absorption with-
in the enclosure, the transmission loss of the enclosure, the sha22
of the enclosure facing the worker, and the acoustics of the work-
room.
A problem occurs in calculating the net transmission loss when

the enclosure has panels, doors, hatches, windows, silencers, and
leaks, each with its own area 5i and associated transmission loss Li.
However, a formula which can be used is
-L /10
L == 1010g 5 - 1010g L: 5 10 i (3.10)
t i
This amounts to adding up all the sound power that escapes and
dividing by the total area. As an example consider a machine con-
trol room in a corner that has ceiling-high walls that separate it
completely from the rest of the shop, where the level is 100 dBA.
The design of the wall is shown in Figure 3.15. It is desired to
compare the performance of single and double glazed windows at a
midrange octave band. We assume that there is negligible leakage
through the roof and that all leaks have been well sealed.
55
24 feet wide
concrete block areas

3 feet high
door
glass 3 feet glass
wide 5 feet high
glass area 105 sq 1t

door and frame 24 sq ft
concrete block 135 sq ft
total 264 sq ft
FIGURE 3.15 - Example of isolating wall
The calculations can conveniently be put in the form shown

below.
Values of S·J. lO-Li/lO
S. L. 10-Li/lO Single DJuble
l l Glazing Glazing
Single Glazing 105 31 7.94 x 10- 4 0.0834
IX:>uble Glazing 105 45 3.16 x 10- 5 0.00332
IX:>or 24 31 7.94 x 10- 4 0.0191 0.0191
Concrete Block 135 45 3.16 x 10- 5 0.00427 0.00427
Sums 264 0.1068 0.0267
10 log 264 = 24.2
Lt (single)= 24.2 - 10 log (0.1068) = 34 dB

L (double) = 24.2 - 10 log (0.0267) = 40 dB
t
56
For the particular octave band for which the above was calcu-
lated, the control room levels would be 66 dB for single glazing
and 60 dB for double. Note how the single glazed window is the
weak element, while with double glazing the door is the chief source
of sound entering. Such calculations reveal where improvement may
be needed, and how much.
Room Acoustics
In addition to the direct path from the noise source to the

receiver (operator), there is also the sound reflected repeatedly
from walls, ceiling, and other surfaces, such as equipment. The
study of the sound fields generated by such reflections is termed
room acoustics.
The sound level from a point source, with rio reflecting sur-
faces nearby, changes at the rate of 6 dB drop per doubling of dis-
tance from the source. In a normal factory space with acoustically
hard walls, floors, ceilings, partitions, and machines, sound is
multiply reflected, and an operator may receive this reflected
sound from essentially all directions. The result is a reverberant
buildup of noise. This means that as the distance from the machine
increases, the noiBe level will diminish for awhile but a point
will be reached where little or no reduction is achieved. Beyond
this distance is the reverberant field, and the distance at which
the change-over occurs is called the critical distance, r c ' The
reverberant field is the only portion of the total noise that is
affected by changes in room acoustics, For most industrial situa-
tions the operator is inside the critical distance. Thus, the noise
that he receives is affected but little by adding absorption to the
walls or ceiling of the work room. This absorption will affect the
sound level experienced by those who walk through, at some distance
from the machines.
The total absorption of a room is the sum of the products of

each room area, S, and its associated absorption coefficient, a.
The room constant, R, is a value based on the summation of all
areas and is:
R = IS a/ (1 - Ci) = Sal (l - a ) (3.11)
Here Q is the average absorption coefficient (I Sa/ S). As the

value of R is increased in a room, the reverberant field sound level
is reduced. Figure 3.16 shows relative sound level versus distance
from the acoustic center of a noise source, for several values of R.
If we had a single source of noise in a room, which factories

normally do not have, the relationship of Lp (sound pressure level),
Lw (sound power level), and room constant, R (distance from the
57
o.----~r__--- -----..,...._---__r---_r-----.. . .
...... ---~
- 5 1 - - - - - - - -.......~+_-----_t_----+_---+_-----_+----__1
Sa
,R = -:--=--
1 - 6:
...J
W
>
~ -1 0r---------t-~~,,~r_--=~'!!!!IIt_-lliiiiiiiiiiiiiii*-;;;;;;;;;;;;;;;;;;;;;;;;;;~~R~:=~5~0~;;;;;;;;~
w
a:
=>
en 100
en
W
a: -15l-----------+------"'WlIl~--"""""-=+-----+--------I--------i
a..
o 200
z
=>
oen
~ -201--------I-----T----.:"CT---+:::::!!!~-lIiiiiiiiiiiiiiii~~5:.:0~0.;;;;;;;l
f-
<X:
...J
W I
a: I 1000
free field
-251----------+-------+-----t--"~--"'l~-----_+_----j
I
2000
I
-301----------+-----+----t------i-'~------=="""""'o;;;;:::_-l--5000
I
I
-35 .......----:--=----~:___-----":""'"""---~:___--~---...&.---'------'
0.5 1.0 2.0 4.0 8.0 32 64 128
DISTANCE FROM ACOUSTIC CENTER - FT
FIGURE 3.16,- Sound field behavior in a reverberant room
58
noise source, r, is in feet and area, S, is in square feet) is:
L
p
= LW + 10 log ((Q/4~r2)+(4/R))+10.3 (3.12)
Here Q is the directivity of the source. For rand S in m and m2 ,

omit the 10.3.
From this relation, it is seen that the inverse square law and
reverberant fields are equal when
2
Q/4~r
c
= 4/R (3.13 )
where the critical distance is
(3.14)
where ais the average absorption coefficient. For the ordinary fac-
tory workroom, ain middle frequencies is usually between .05 and 0.1.
For machines near a floor, the directivity factor Q = 2. Then approx-
imately
r == 0,2 I Sa (3.15 )
c
If the room has N identical machines,
r == 0.2 /Sa/N (3.16)

c
Note that because ais usually an increasing function of frequency,
the critical distance increases somewhat with frequency.
If a rough estimate of an overall r c is all that is needed, this

can be accomplished by using nothing moxe than your ears. Focus
your attention on a dominant, easily identified noise source. Walk
away from the source machine until the level appears relatively con-
stant to your ears. Then walk rapidly toward the source. Stop
when you are first aware of an increase in level. Repeat until you
have localized the point at which the direct sound begins to predom-
inate, at the critical distance position. Measure from there to the
source.
As an example, consider a workroom 80 ft x 120 ft with a 25-ft

ceiling, having 10 identical automatic screw machines. Suppose that
the average absorption coefficient, at the maximum A-weighted octave
band level, is 0.15. Then with a total surface area of 29,200 sq ft,
r
c == 0.2
[29,200(0.15/l0)J ~ = 4.2 ft (3.17)
Workers on screw machines will be within 4 ft, which is just at this
critical distance. Hence they will receive about equal amounts of
direct and reverberant sound.
59
If the noise level at a distance from the source requires
attenuation, absorption must be added to the room in the form of
absorptive materials on ceiling and walls, or by hanging absorbent
baffles, which are commercially available. The amount of attenua-
tion needed, ~ L, can be related to room: constant R by
(3.18)
where Rl and R2 are before and after values, respectively.

We again repeat the caution: adding room absorption will normally
note reduce the direct noise from the source, which is what the oper-
ator receives. Indeed, adding absorption increases the critical
distance, so the operator can be still further from his machine and
still receive the major amount of noise from it. In the example
above, adding 100% absorption would decrease the received level
only 3 dB.
Receiver
If noise source corrections and noise path corrections do not

reduce the noise to acceptable limits or before these corrections
have been applied, attention should be directed toward the receiver.
This may be in the form of administrative controls or personal pro-
tective equipment or both.
Administrative Controls. Administrative controls can in some

cases be developed to limit the duration of exposure to noise levels
above 90 dBA to the permissible noise exposure time specified in the
Section 1910.95 of OSHA (Occupational Safety and Health Administra-
tion) regulations:
Hours/Day dBA Hours/Day dBA

- - - -
8 90 l~ 102
6 92 1 105
4 95 ~ 110
3 97 l:I or
2 100 less 115 max.
For combinations of exposures, the sum of the actual daily exposure

time divided by the allowed exposure time for each time period must
equal or be less than 1.0 to be in compliance regarding noise levels
and should be equal or less than 0.5 to reduce the need for wearing
ear protectors. The sum of these ratios is called the daily noise
dose. Administrative controls include the following approaches:
(1) Workers at 90 dBA are not exposed to higher levels.
(2) Workers at higher levels are removed from the noise after
the above time limits and spend the balance of the day in
lower than 85 dBA areas.
(3) Work time is divided between more operators, that is, split
shifts are arranged between high noise and low noise areas.
60
(4) When less than full-time operation of a machine is
required, time is split into partial days instead
of an occasional full day operation.
(5) Exposure is reduced by shift scheduling to reduce

number of exposed employees and number requiring
protective equipment.
Personal Protective, Equipment. Present (October 1974) Depart-

ment of Labor "Guidelines, to Occupational Noise Standards"
(Bulletin 334), say: "The use of personal protective equipment is
considered by the Department to be an interim measure while engi-
neering and administrative controls are being perfected. There·
will be very few cases in which the use of this equipment will be
acceptable as a permanent solution to noise problems."
The Departrnentof Labor Bulletin 334 Guidelines cover the

following major points on protective equipment:
(1) Only approved ear protectors that have been tested in

accordance with ANSI Standard Z24.22-l957, should be
used.
(2) Five dB less than stated attenuation of equipment

should be allowed because test data were obtained under
ideal conditions that are not normal in day-to-day oper-
ation.
(3) Ear muffs and ear plugs should be fitted and supplied
through a properly trained person who can educate the
workers in the use and maintenance of the muffs and
plugs.
(4) Wax impregnated cotton and fine glass wool are accept-
able, but cotton stuffed in the ears has very little
value and is not acceptable.
In addition, it is our experience that ear plugs should be

inspected periodically to ensure that they have not deteriorated
and that the workers have not cut off the ends for greater comfort
(but no protection) .
Selling the use of protective equipment to employees and super-

visors will usually require that an educational and promotional pro-
gram precede the required use of ear protection. There should be
continual follow-up by supervisors to see that the program is ac-
cepted and that ear protection is worn where needed. This program
can be aided by signs in areas where protective equipment is manda-
tory. Supervisors should be aware that if a plug or muff is uncom-
fortable, it may not be worn. Other difficulties are with long
61
hair, which can hide a plug that has deliberately been loosely in-
serted; plugs may be cut off and ear muff bands weakened to reduce
pressure. Tight .muffs over broad-temple glasses can restrict blood
flow. Devices are available to ease this last problem.
Audiometry. Audiometric tests should be made of all indivi-

duals working regularly in areas in which the noise level is above
85 dBA. This is an excellent way to determine how successful engi-
neering or other controls have been. Several recommendations for
an audiometric program include:
(1) Test facilities and procedures are to meet established

standards of ANSI S3.1-1960 and 3.6-1969 (or latest)
(2) Tests are to be made as described in ANSI standard

S3.1-1960 and 3.6~l969, and by a person certified
as an aUQ10metrlc technician by a course of training in
accordance with Intersociety guidelines.
(3) Audiometers used are to meet specifications ln ANSI

standard S3.6-1969 or latest and shall have a certi-
ficate of at least yearly calibration. They are to
be subjected to a biological check at least once a
month; a log of checks and calibrations is to be
maintained.
62
Chapter 4
NOISE CONTROL MATERIALS
This chapter describes the four types of materials most often

used in noise control: absorbers and isolators for airborne sound,
and vibration isolators and damping materials for controlling vi-
bration solid-borne sound. Guidelines for selecting materials are
also given, based on acoustic and nonacoustic considerations.
By far the best location for noise control is as close to the

source as possible. Once the sound is airborne, it is difficult
and expensive to control. Control at the receiver by ear plugs,
ear muffs, or control booths is not usually included in the meaning
of engineering control, which commonly refers to machine-oriented
efforts at the source or along the early path.
Absorption Materials
With absorption, sound in air is changed into heat. Suitable

materials are usually fibrous, lightweight, and porous. The fibers
should be relatively rigid. If a cellular material is used, the
cells must intercommunicate. Foams should be reticulated to the
proper degree.
Examples of absorbents include acoustical ceiling tile, cloth-

ing, loss-type mufflers, and foamed elastomers. Physically the flow
resistance of fibrous materials is the most important characteristic.
For optimum results, the flow resistance must usually be increased
as the thickness of the absorbent decreases, to maintain peak absorp-
tion.
The flow resistance can be sensed rather crudely by attempting

to blow through the material. Comparison with an accepted material
of the same thickness provides a personal calibration. The effec-
tiveness of an absorbent is measured by the absorption coefficient.
Ideally, this is the fraction of the sound energy flowing toward the
absorbent that enters the material and is not reflected; thus" a
perfect absorbent would soak up all the sound incident on it.
Industrially useful absorbents have coefficients above 60% in the
frequency range from 500 Hz and up.
Absorbents on room surfaces reduce the amount of reverberant

sound. The relative reduction of such sound levels by equal amounts
of any two absorbents is given approximately by 10 times the loga-
rithm of the ratio of their absorption coefficients. Table 4.1 shows
average absorption coefficients of various absorbents, and for com-
parison Table 4.2 gives absorption coefficients of relatively nOnab-
sorbent construction materials plus those for some special materials.
63
TABLE 4.1 - Sound absorption coefficients of common acoustic materials
Frequency Hz
Materials· 125 250 500 1000 2000 4000
Fibrous glass
(typically 4 Ib / cu ft) hard backing
1 inch thick o.oi 0.23 0.48 0.83 0.88 0.80
2 inches thick 0.20 0.55 0.89 0.97 0.83 0.79
4 inches thick 0.39 0.91 0.99 0.97 0.94 0.89
Polyurethane foam (open cell)
1/4-inch thick 0.05 0.07 0.10 0.20 0.45 0.81
1/2-inch thick 0.05 0.12 0.25 0.57 0.89 0.98
1 inch thick 0.14 0.30 0.63 0.91 0.98 0.91
2 inches thick 0.35 0.51 0.82 0.98 0.97 0.95
Hairfelt
1/2-inch thick 0.05 0.07 0.29 0.63 0.83 0.87
1 inch thick 0.06 0.31 0.80 0.88 0.87 0.87
• For specific grades see manufacturer's data;note that the term NCR, when used, is a single
term rating that is the arithmetic average of the absorption coefficients at 250, 500, 1000,
and 2000 Hz.
TABLE 4.2 - Sound absorption coefficients of common construction materials
Frequency Hz
Material 125 250 500 100C 2000 4000
Brick
Unglazed 0.03 0.03 0.03 0.04 0.04 0.05
Painted 0.01 0.01 0.02 0.02 0.02 0.02
Concrete block, pai nted 0.10 0.05 0.06 0.07 0.09 0.08
Concrete 0.01 0.01 0.015 0.02 0.02 0.02
Wood 0.15 0.11 0.10 0.07 0.06 0.07
Glass 0.35 0.25 0.18 0.12 0.08 0.04
Gypsum board 0.29 0.10 0.05 0.04 0.07 0.09
Plywood 0.28 0.22 0.17 0.09 0.10 0.11
Soundblox concrete block
Type A (slotted), £3 inch 0.62 0.84 0.36 0.43 0.27 0.50
Type 8, 6 inch 0.31 0.97 0.56 0.47 0.51 0.53
Carpet 0.02 0.06 0.14 0.37 0.60 0.65
64
The absorption coefficient depends not only on the material
but also on what is in front and back of it. Most coefficients
are stated for an unobscured front, but with a rigid impervious
backing spaced various distances away from the material. Noise
control engineers use designations of the Acoustical and Insulating
Materials Association to describe the material mountings:
(1) Cemented to backing with about 1/8 in. air

space
(2) Spaced 3/4 in. away by furring strips
(4) Laid directly on surface - very little air

space
(7) Suspended 16 in. from the backing
When the mounting is not specified, usually it is No. (1) or (4).
Materials may have special facings. For resistance to grease

and water that would clog pores, a thin plastic film covering is
often used. Such films, as well as perforated hardboard facings,
tend to produce a maximum in the mid-frequency absorption coefficient.
Some thin construction materials, notably plywood, can show increased
low frequency absorption by panel resonance, if they are not securely
fastened down.
The standard reverberation room method of measurement of absorp-

tion coefficient (ASTM C423-66, or latest version) essentially sub-
jects the absorbent to sound from all angles. Data on absorption co-
efficient cannot be regarded as useful and meaningful unless they
have been obtained in this standard fashion.
Often a single-number statement of average absorption is needed.

Not always useful to the noise control engineer, but at least stand-
ardized in presentation is the noise reduction coefficient (NRC).
This is the average, to the nearest, 0.05, of the coefficients at 250,
500, 1000, and 2000 Hz.
Transmission Loss Materials
The noise isolation properties of materials are stated in terms

of transmission loss. As with absorption, the concept of energy
flow is used; here it is the energy transmitted through the material,
relative to that flowing toward it. Transmission loss is 10 log
(incident)/(transmitted energy), and usually increases with frequency
at the rate of 6 dB per doubling of frequency. Measurement is diffi-
cult, and only a handful of laboratories in this country are quali-
fied to make the standard measurement (ASTM E90-61T, or latest revi-
sion). Data on the transmission loss of materials appearing in
advertising literature cannot be regarded as meaningful unless
they have been determined in this standard manner.
65
As a result of the search for a single number to indicate the
average full transmission 1(35, the concept of sound transmission
class (STC) was developed. It is useful specifically in assessing
the degree to which intelligible speech is prevented from being
transmitted through a wall. Because the intelligibility of speech
depends chiefly on the higher frequency sounds, the STC is arrived
at by requiring greater loss at higher frequencies than at low.
The STC should be used with caution in industrial work, however,
where the noise spectrum can be much different from that of speech.
The noise control engineer will need the transmission loss in each
octave band for the proper application of isolating materials.
Isolating materials are effective because the material is as

different as possible from air, on a wave-transmitting basis. In
many situations where a large transmission loss is needed, the mate-
rials can be installed in several layers, rather than one thick one.
The chief requirement is that there be minimum mechanical connection
between layers, other than the air space. Large increases in trans-
mission loss are possible by subdividing the total mass.
Damping Materials
Damping materials are used to reduce resonance effects in sol-

ids. Essentially, damping materials are absorbents for solid-borne
sound and convert it into heat.
DampiLg materials are used in many applications. If a machine

panel (such as a belt guard) is subjected to vibration, it will
radiate sound strongly at its resonant frequencies. Damping the
panels or guards can thus reduce this radiated sound. In another
application, parts that fall into (and are carried along) metal
chutes can excite the chute panels by repeated impact. Damping
materials along the chute surfaces will reduce the noise; these
materials must also be selected with heat resistance and mechanical
integrity in mind. Damped stock tubes are available for quieting
screw machine operation. Panels for isolating enclosures can trans-
mit large amounts of sound in the critical frequency region. Damp-
ing can greatly reduce the severity of the decrease of transmission
loss in that region.
There are two types of damping materials: homogeneous and

constrained layer. A homogeneous material is applied in a rela-
tively thick cast, sprayed or troweled layer, depending on the thick-
ness and type of metal to be damped. A constrained layer material
consists of a thin layer of the actual damping (lossy) material,
with a backing of thin metal or stiff plastic. The mechanical action
is one of making the damping layer much more effective than if it
were homogeneous. Constrained layer damping materials can be pur-
chased as an adhesive/metal foil tape combination, where the adhesive
is selected for its energy loss properties as well as its adhesion.
These damping tapes are especially useful on thin panels (1/16-in.
steel or less).
66
Vibration Isolators
Vibration isolators act on the same principle as isolators for

airborne sound: introducing into the transmission path a material
whose wave-transmitting properties are as different as possible from
the medium carrying the wave. For vibration in solids, such mate-
rials are spring-like. Examples include resilient elastomer and
metal springs, elastomer pads, and in extreme cases, air springs.
The weaker the spring, usually the greater the isolation.
Solid rubber or rubber-fabric pads are not too effective, be-

cause the displacement is small and is not proportional to the load.
This latter effect, called nonlinearity, is best described by a
curve showing load on the isolator to produce a given deflection.
Some manufacturers provide these data.
If an isolator is too weak vertically, it may not be laterally

stable. There are available side-restrained metal spring isolators
to avoid this difficulty. In extreme cases it may be necessary to
use many isolators, all acting along lines that pass through the
center of gravity of the machine. This use of mat~rials is most
effective when the vibration situation is reversed, i.e. when a
delicate mechanism is to be protected from external shock and vibra-
tion. '
The proper amount of damping is needed with vibration isola-

tion in many applications. Steel springs alone are highly undamped;
if they rest on elastomer pads, there is much improvement.
Material Selection
The most commonly used materials for control of noise in indus-

try are isolators and absorbers for airborne sound and vibration
isolators and dampers for solid-borne sound. Selection of materials
is governed by many factors other than acoustical, and installed
cost becomes less important than possible interference with produc-
tion. Here we first discuss the acoustic factors and then consider
the others.
Acoustic Considerations. Isolators for airborne sound have

many configurations, ranglng from a simple shield to a barrier to a
total enclosure. The most important acoustical parameter is the
weight per unit area: the heavier the barrier, the more isolation
is obtained. A characteristic of most isolating materials, such as
sheet steel, glass, or solid wood panels, is that they are resonant.
At the frequencies for which the various types of resonances occur
(and particularly in the coincidence region), the isolation afforded
becomes much smaller. At frequencies above the resonance, the iso-
lation again increases.
67
An idealized and smoothed way of describing the above behavior
is given by Beranek in Noise and Vibration Control (B.5). Isola-
tion, expressed (in dB of transmission loss) as the ratio of energy
transmitted to energy incident, first increases with frequency at
the rate of 6 dB per octave. At the coincidence region, a plateau
is reached for which the average transmission loss is practically
constant (see Table 4.3). The frequency extent of this plateau de-
pends on the damping, or deadness, of the panel material. Thus,
lead has a short plateau and steel a long one. Above the plateau,
the isolation again increases. Damping can be added to reduce the
plateau length.
TABLE 4.3 - Transmission loss of common materials
Frequency Hz
Material I ty'sq It 125 250 500 1000 2000 4000 8000
Lead
1/32-inch thick 2 22 24 29 33 40 43 49
1/64-inch thick 1 19 20 24 27 33 39 43
Plywood
3/4-inch thick 2 24 22 27 28 25 27 35
1/4-inch thick 0.7 17 15 20 24 28 27 25
Lead vinyl 0.5 11 12 15 20 26 32 37
Lead vinyl 1.0 15 17 21 28 33 37 43
Steel
18-gauge 2.0 15 19 31 32 35 48 53
16-gauge 2.5 21 30 34 37 40 47 52
Sheet metal (viscoelastic laminate-core) 2 15 25 28 32 39 42 47
Plexiglas
1/4-inch thick 1.45 16 17 22 28 33 35 35
1/2-inch thick 2.9 21 23 26 32 32 37 37
1-inch thick 5.8 25 2B 32 32 34 46 46
Glass
1/8-inch thick 1.5 11 17 23 25 26 27 28
1/4-inch thick 3 17 23 25 27 28 29 30
Double glass
1/4 x 1/2 x 1/4-inch 23 24 24 27 2B 30 36
1/4 x 6 x 1/4-inch 25 28 31 37 40 43 47
5/B-inch Gypsum
On 2 x 2-i nch stud 23 28 33 43 50 49 50
On staggered stud 26 35 42 52 57 55 57
Concrete, 'I-inch thick 48 ,29 35 37 43 44 50 55
Concrete block, 6-inch 36 33 34 35 38 46 52 55
Panels of 16-gauge steel, 'I-inch
absorbent, 20-gauge stee I 25 35 43 48 52 55 56
68
Figure 4.1 shows the plateau average height and width for some
common single panel isolating materials. Note that the variable is
the~product of frequency and weight/area (abbreviated as fw and
given in Hz-lb/ft 2 ). To select material and weight/area, first
decide how much noise reduction is needed. A good approximation is
obtained by measuring the octave band spectrum and A~level at the
ear position of the worker closest to the machine. If this position
is indefinite, measure at 1 meter (3.28 feet) from the major noise-
producing surface of the machine. Then the minimum noise reduction,
NR, needed is given by:
NR = LA - Lc ( 4. 1)
where LA is the equivalent 8-hour level, and L c is the criterion

level of 90 or 85 dBA, as appropriate. It is wise to plan for
85 dBA. The transmission loss (TL) for industrial situations is
related to the noise reduction by TL = NR - 6. To afford design
margin, it is wise to use TL = NR.
I I I I I I I I I I I I I 1IIIfII" I I I I I I I
60 ~~ ~
~
/ lead (59)
50
~~
,o~c7
m
""0
Cf)
Cf)
0
40 ~' ~
. ~~
;/
,0
\0<::-
steel (40)
concrete (12)
-l
"0~
<::-~
brick (11l ,
Z
0
Cf) 30 . oeP'u e ; / plaster (9) I
aluminum (14)
Cf) . <::-u' glass (13)
~
::2:
Cf)
Z
« 20
II
'"
plywood (3)
f-
lO
V
/
,
10 3 104 lOb
FREQUENCY-WEIGHT/UNIT AREA (fw) [~z-Ib/ft 2 J

F I GU RE 4,1 - Transmission loss pIa teaus. Ends of plateaus occur at the folIo wing fw values:
plywood 500, 2600; glass 1200, 13,000; aluminum 1400, 25,400; plaster 1400, 11,200; brick 3500,
12,250; concrete 4000, 18,000; steel 4500,49,500; lead 30,000, 120,000. Numbers in ( ) are
weights in Ib/~ ftPe;-;nchoTthickness.·:~'--'~J
.. '. ,:~'. '. ' . ·,-jj·;j~~t!
69
The next question is: In what octave band is the maximum noise
reduction needed? To estimate this, modify the measured octave band
levels by subtracting from each its midband A-weighting. Locate the
octave band in which the maximum A-weighted level occurs. Use the
designating frequency of the octave band below this in calculating
the weight.
Now use Figure 4.1. Choose a material whose plateau height is

greater than TL required. If the plateau is too low, added damping
will be needed to avoid extreme coincidence effects. From the curve,
find the value of the fw (frequency X weight/area) product corres-
ponding to the TL needed. Divide this fw product by the designating
frequency of the octave band below the maximum in the A-weighted oc-
tave band spectrum. This gives the minimum weight per unit area for
the panel material. See Table 4.3 for the TL of some common mate-
rials and Figure 4.2 for a comparison of various materials and wall
designs (from Harris 1957, p. 20-6), B.7.
WEIGHT IN KG PER SQUARE METER

S 10 20 50 100 200
60~-----r---...,..--F'---,-""",-"--,--,r-T~-----"~~-"",---,:",~==:--,,,--""':;":"',,,
651--------r-------1--
50
.D 451----
o
Z
40 - -- -..
en
en
o-l ~5 - ' - -.----------
z
o
en
en
:2
en
Z
«
a:
......
5 ----
II -- ----
I
O~I------:-----:~-~-~--:~~7-!S~9......\ 0 : - - - - - -......
20:----J.~0--4..l.0--5.i..0-6J.0-7J...0-S";'0-9J.O....lIOO
WEIGHT IN POUNDS PER SQUARE FOOT

Source: C.M. Harris, ed., Handbook 01 Noise Control, McGraw-Hili, New York, 1957.
FIGURE 4.2 - The mass law relation between average sound transmission loss and mass per unit
area of partition
10 r----
Reproduced frompy
l~_~~-_ ,~~_::allable co _. )
70
If the panel material is resonant (long plateau), it must be
damped to reduce plateau length. For irregular panels, the damping
material can be sprayed on or troweled on. As a convenient guide,
several popular materials are satisfactory when the dry thickness_ of
the damping treatment is about twice that of base metal, usually
steel. Manufacturers will furnish recommended treatment thickness.
Design recommendations based on this material selection are

covered in more detail in Chapter 3.
Many damping materials are quite temperature-dependent. Be

sure that the material selected has optimum performance for the
temperature range in which it will be used. Most manufacturers can
supply the necessary engineering information for making an optimum
selection. Damping materials in sheet form for gluing to a vibrat-
ing surface can be supplied with a pressure-sensitive adhesive.
The interior of enclosures must be sound absorbent to prevent

reflective buildup of sound within the enclosure. Often the func-
. tions of absorption and damping can be combined in a single composite
material. The absorbent is usually a foamed polyurethane; the mini-
mum thickness of the foam should be 1 inch. Special oil and water-
impervious film coverings can be specified. This covering protects
the material from becoming liquid soaked. However, deposits on the
film must be removed periodically to prevent loss of high frequency.
absorption. Use a mild warm detergent solution, applied with a
cloth. Wipe carefully to prevent damage to the film.
Absorbents protected by a film still have exposed edges. These

may be sealed by a latex paint that anchors itself to the pores of
the absorbent and closes the edges. An alternative is a thin film
that is adhesively fastened over the edges. By choice of thickness
of air space between the absorbent and the mounting surface, the
shape of the absorption-frequency curve can be controlled. This
can be matched to the need for maximum noise reduction in frequency
regions determined by the noise measurements.
Vibration Isolators. The selection of vibration isolators is

based on placement of the resonance frequency of both machine mass
and isolator spring constant. All of these should be well below the
operating frequency. As a general rule, a machine on a heavy rigid
foundation is well isolated when the resonance frequency is less
than 1/5 the lowest operating frequency. The latter is usually that
of rotary unbalance in the slowest rotating part. If the machine is
on a lightweight floor or is hung from a springy roof, the ratio
should preferably be less than 1/5. Vibrating pipes or suspended
fans can be in this category. Steel spring pipe hangers, supplied
with damping elastomer seat, are the usual pipe hanger materials
selected. For many machines on a solid foundation, a crossribbed
rubber pad often suffices. Manufacturer's literature can be help-
ful in making an optimum selection, provided that the spring con-
stant and maximum deflection is supplied.
71
A special problem arises with punch presses or other sources of
periodic impulse noise. Here the problem is to reduce the trans-
mission of both the shock and the ensuing vibration. The optimum
choice of vibration isolator is governed by the relation among three
time intervals. First is tl' the effective duration of the exciting
shock. Second is t2' the period (time for one cycle) of motion re-
sulting from resonance between mass of the machine and the effective
stiffness of the vibration/shock isolator. Third is t3' the inter-
val between repetitions of the impulse. The isolator should be
chosen so that tl < t2 < t3 to obtain near-optimum results. The value
of tl for a punch press or shear is approximately the time between
contact of the tool with the workpiece and the completion of the
cutting action. This can be determined from machine parameters; di-
rect measurement with an oscilloscope gives more reliable data.
Vibration isolation can reduce A-levels from 2 to 15 dB. The

lesser amounts are obtained when the base is solid and heavy, as for
a foundation set directly on the eart~. See Chapter 3 for more de-
tailed discussion of vibration isolation.
Silencers. Silencers or mufflers can be considered as sec-

tions of a duct or pipe that have been acoustically designed and
treated or shaped specifically to reduce the noise transmission.
The noise may originate from a machine, or be flow-generated by
gas or liquid flow.
Silencers for airborne sound are best selected by reference to

manufacturers' literature. The usual silencers are lined with spe-
cial absorbents. The reductions available are tabulated for the
octave bands for several silencer lengths. Knowing the reduction
desired, you can choose readily. Silencers are often available in
several pressure drop categories. Final selection requires some
expertise in air-handling system design, and a mechanical engineer
should be consulted in critical situations. The degree of silenc-
ing also depends on the amount of steady flow and on whether the
sound is traveling upstream or down. Some manufacturers supply
such data.
Nonacoustic Considerations of Materials. Selection may also

require consideratiori of nonacoustic factors, which often control
the final choice. These factors may be broadly classified as en-
vironmental and regulatory. Environmental factors include:
- Moisture, water spray, water immersion
- Oil, grease, dirt
Vibration
- Temperature
- Erosion by fluid flow.
72
Regulatory factors include:
- Lead-bearing material forbidden near food processinq

lines.
- Restrictions on materials that may be in contact with

.foods being processed - glass, monel or stainless steel
permitted.
- Requirements for material not to be damaged by disin-

fecting.
Firebreak requirements on ducts, pipe runs, shafts.
- Flamespread rate limits on absorbents.
- Fire-endurance limits on absorbents.
Restrictions on shedding of fibers in air by

absorbents.
Elimination of uninspectable spaces in which vermin

may hide.
- Requirements for secure anchoring of heavy equipment.
Restrictions on hole sizes in machine guards (holes

can reduce radiated noise of vibrating sheets).
A good example of the influence of these factors is seen in

the selection of absorbents for use inside machine enclosures.
It is typical of ordinary maintenance practice to over-lubricate
rather than to install or service oil or grease seals. Hence it
is common to· find oil and grease deposits on machines, often with
dirt, metal chips, and other debris. Such deposits greatly degrade
the perforMance of absorptive coatings, which are porous materials
that easily wick oil and water. However, absorbent materials are
now available with a thin imperforate skin or film covering of
Mylar, Saran or Tedlar, which prevents fluid wicking. Nevertheless,
the sheer weight of grease deposits will degrade higher frequency
performance even without wicking; fire hazard will also be increased.
Hence the film must be strong enough that the deposits can be
cleaned off with a cloth wet with warm detergent, plus mild rubbing.
Such maintenance will be necessary with machine enclosures lined
with absorbent. The "time between cleanings can be greatly length-
ened if oil and grease seals are installed or if deflecting shields
are used on severe oil spray, such as comes from impacting parts in
a punch press.
73
Curtain types of isolating materials are convenient for con-
structing an enclosure rapidly. The usual material is lead-loaded
vinyl plastic. Where leaded materials cannot be used, as in some
stages of food processing, a barium-loaded type is available. Monel
and stainless steel are the only common metals usually permitted in
contact with food. Fibrous absorbents in shop-made silencers and
mufflers can be eroded by high speed gas flow, say, above 30 mls
(100 fps). The fibers may pose a health hazard and also can inter-
fere with machine operations. The situation is worsened if vibra-
tion is present, as this tends to break and shake out small fibers.
The material used should have some bonding agent to hold fibers
securely in place. In addition, the absorbent can be covered with
wire screen or perforated metal. If the latter is used, the ratio
of open to total area should be over 0.5. The effective absorption
will be decreased if lesser open areas are used. Foamed absorbents
shed much less than fibrous types, but all need sealing of raw
edges by a film-making paint or by a thin plastic cover.
Since most damping materials are very temperature sensitive,

the operating temperature must be known before making an informed
selection. For unusual conditions, consult the manufacturer.
Often if a large lot of material is to be purchased, it can be
specially formulated for the conditions specified.
Fire resistance is often required by building codes. Absorb-

ents are available for several degrees of resistance. With suit-
able materials, fire breaks are sometimes unnecessary in isolating
walls that are filled with absorbent. Since local building codes
may not be applicable to structures that can be described as a
part of the machine, prudent language must be used in describing
the function of the enclosure.
A most important nonacoustical factor in the selection of

noise control materials is net cost. The noise control engineer
must always be aware of this and should design so that labor plus
materials cost is minimized. A part of the net cost is also the
loss in production while a machine is being treated, so time to
restore production must be considered. Ease of maintenance must
also guide the selection. Achieving a viable design means that
material selection cannot be accomplished on a purely acoustical
basis.
74
Chapter 5
CASE HISTORIES
The case histories presented here should be of some use to

production and safety engineers, health personnel and other factory
personnel who are not specialists in noise control. The case histo-
ries are examples of engineering tasks that have been completed by
professional noise control engineers. Collected here are actual
cases on various industrial devices. These devices were typically
machines utilized in a production process and in some cases had been
cited by safety officials for unsafe high noise levels.
The case histories presented here were chosen from some of

those where the amount of noise reduction actually achieved was
measured. Such engineering results, even if not directly applicable
to your situation, may, illustrate general principles that will point
the way to a successful result in your problems.
Each case history was selected because the work stations asso-
ciated with it were in violation of the recommended noise exposure
levels in all but two or three examples, where uncomfortable noise
levels affected productivity even though they were lower than the
limits in the OSHA noise regulations.
Case, History Data
Ideally, case histories should be complete in all details.

However, the prime duty of a practicing noise control engineer is
to solve his client's problems at minimum net long term cost. In
this the experience of the noise control engineer is invaluable in
knowing when information is unnecessary or has been handled verbally.
Indeed, the client may already have done part of the work.
In the following outline, there is described the whole process

of accomplishing noise control that is viable in both engineering
and economic senses. The outline will also serve as a check list
to guide you in learning and applying the principles of noise con-
trol engineering that have been discussed earlier. The case his-
tories that follow contain the essential data for the simpler prob-
lems and somewhat more for the complicated ones.
Outline of Complete Procedure for Development of Noise Control
1. Plant data
- SIC classification of industry

- Location, address; division
- Product or process,
75
2. Problem definition
- Compliance plan
Compliance measurements, daily noise dose
- Diagnostic measurements and source locations
- Design of experimental noise control
- Design of final noise control
- Supervision of construction, installations
- Post-installation checkout, performance evaluation
- Oral briefings
- Preparation of technical paper
3. Machine data
3.1. Identification
- Make, model, serial number, factory number

- Appearance (drawing or photo); identification
of significant parts, functions
- Layout drawing of workroom, all machines shown
- Location of aisles, vertical clearances; serv-
ice lines; conveyors; hazard-posted areas
3.2. Operating data
- Functions of machine; relation to others

- Type of input: gage, size, shape of stock
- Type of output: shape, size
- Type of scrap: how collected
- General product flow with respect to other
machines
- Use of automation: conveyors, robots
- Services and ratings: electrical, air, water,
fuel, steam, hydraulic, internal combustion
engine, vibrator
- Production rate (maximum)
- Down time: jams, breakdowns; repair, mainte-
nance, set-up; reload, idling; operator at
rest room, meals
- Constraints on operation: access, both physi-
cal and visual; for worker, input stock, out-
put product and scrap; access for repair,
maintenance, set-up, reload; safety, union reg-
ulations, sanitation, special materials for
food industries, rodent control; operator need
for aural cues; limits on capital and operat-
ing expenses
76
Special machine features: noise control
features already installed; use of vibra-
tion isolators; use of air; evidence of
over-lubrication
4. Noise situation data
4.1. General observations (ear)
- Noise high or low pitched

Directional location by cupping hands behind
ears
- Presence of pure tones
- Level constant, varying slowly or with much
impact noise
- Feeling of vibration in floor
- Workers communicating byword or sign
- Use by workers of aural cues in detecting
and evaluating machine performance, jams
4.2. Measuring equipment
- Name, make, model, SiN

- Calibration data: when; traceable to NBS
- Check list for diagnostic acoustical meas-
urements: SLM, octave band analyzer, 1/2-in.
and l-in. microphones, tripod, extension
cord for microphone, windscreens, calibrator
and adaptors; accelerometer; control box for
acceleration, velocity, displacement; strobo-
scope; vibrating reed tachometer
- Check list for optional equipment for diag-
nostic acoustical measurements: two channel
tape recorder, connecting cords, microphone
for voice channel, blank reel, AC cord,
charger; range finder, measuring roller,
steel tape (inches and centimeters); flash-
light; pressure-sensitive labels; camera
with wide-angle lens, flash; spare batteries
for all equipment (alkaline only); ear muffs,
safety glasses, safety shoes, hard hat, paper
towels, handsoap; pliers, diagonal cutters,
screwdrivers; circuit tester
4.3. Acoustical measurements
- A, C, peak and octave band readings

- Measure at ear positions of worker, worker
absent, with all machines going, then with
machine in question off
77
- Run machine at different speeds to
locate resonances
- Run with portions of machine selectively
disabled
- Measure rpm's with stroboscope, vibrat-
ing reed tachometer
- Measure at suspected noise sources on
machine; photograph the set-ups; locate
microphone precisely
- Set to octave band for which A-weighted
spectrum at ear of worker maximizes.
Probe around machine to locate sources
- Locate around machine an imaginary box
that touches all major surfaces; record
the dimensions; at I m away from box, ob-
tain levels for calculating total sound
power, using procedures described by Diehl,
in Machinery Acoustics, reference B.G
- On slow A-scale obtain contours of equal
level around machine, others off; locate
paths of workers among contours. Repeat
with all machines on
4.4. Vibration measurements
- A, C, peak and octave band readings

- Probe over the surface (pickup coupled so
it is not rattling) for acceleration and
velocity
- Calculate root mean square values over
surface
- Observe if maximum A-weighted band for rms
acceleration or velocity coincides with
maximum A-weighted band for airborne noise
- Run machine at different speeds to locate
resonant excitation of vibration
- Selectively disable parts of machine to
locate exciting sources
4.5. Auxiliary data
- Data per (3.2)

Unusual conditions: breakdowns; machine
with bad bearing, gears, loose parts
- Tape recordings of noise situations that
are short-lived or non-repetitive, together
with calibration signal; also useful for
later narrow-band analysis, judging rpm,
pure tones .
78
- photographs of all pertinent parts of
machine, from establishing shots to close-
ups of name plate
- Names, position and possibly addresses of
operating, supervisorial and management
personnel concerned
- Time of entry to plant, time spent at
each machine, time left plant
5. Development of noise control
5.1. Preliminary report
- Data, raw and reduced; evaluation, inter-

pretation
- Preliminary noise control recommendations,
taking full account of const~aints in (3.2)
above
- Preliminary estimate of noise reduction
expected
- Preliminary estimate of capitalized installed
cost
- Preliminary estimate of possible change in
productivity and change in piece part cost
- Recommendations on use of automation
- Conference to discuss implications of report
5.2. Development of revised recommendations
- Remeasure as needed
- Re-estimate noise reduction, costs
- Prepare recommended experimental program if
problem sufficiently unusual
- Prepare sketches showing acoustically essential
features of the noise control devices; if re-
quired, prepare drawings
- Recommend special materials; provide alternate
suppliers
- Estimate construction, installation costs
5.3. Installation, use
- Monitor construction and installation for

adherence to acoustical specifications
- Introduce corrective measures for improperly
installed devices
- Evaluate emergency alternate materials
- Measure installed performance; correct
deficiencies
79
- Measure daily noise dose to applicable
workers
6. New work
- Recommend improvements if similar noise control is

to be applied to other machines of the same class
- Recommend action on problems remaining
- Provide briefings on results to technical and
management people
- Prepare paper for publication
- Help prepare formal compliance reports
80
CASE HISTORY #1 - STEEL WIRE FABRIC MACHINE
Description of the Problem
This 8-foot fabric machine man~factures wire netting spaced at

6-inch spacings, starting with individual wires from large spools
that run through the length of the machine. A perpendicular wire,
known as staywire, is fed across at 6-inch intervals and spot welded
at each intersection. This staywire is then cut off at the left-
hand side of the machine. The long wires are then moved through the
machine another 6 inches, and the staywire operation is repeated.
This machine produces 6 x 6 inch #8 or #10 wire netting, which is
used as concrete reinforcement in the home building industry. Ma-
chine is made by Keystone Steel and Wire Company.
Measurement and Analysis
At the operator position, the noise level was found to be 99

dBA and 102 dBC, indicating low frequency components. This kind of
noise (L C - LA = 3) is very unpleasant.
Daily noise dose was found to be 2.5, where the acceptable level
is 1.0.
Criteria were established to reduce the noise exposure to 1.0 or

less, which is equivalent level to 90 dBA or less.
The octave band measurements made at the main drive gear (Figure
5.1.1), at the operator station (Figure 5.1.2), at the wire spool
area (Figure 5.1.3) showed that noise sources included (1) general
mechanical noises due to needed maintenance, (2) the wire wrapper, a
ratchet-action machine operated from main drive gears and found to
cause 10aa-Hz peak noise, and (3) mechanical sources within the ma-
chines, which could lend themselves to isolation.
Control Approaches Considered
In addition to the direct noise corrections implied above, an-

other solution could have been to construct a noise shelter for the
operator. This solution was dropped in favor of working on speci-
fic noise sources. A program was established to:
(1) Overhaul the machine: replace bearings, reduce metal-to-

metal banging, replace worn gears, and so on.
(2) Replace ratchet type drive on wire wrapper with chain

drive. (This device pulled the long wires through the
fabric machine.)
81
(3) Add steel plates (10 1b/ft 2 ) to the frame of the machine.
These plates were welded to the frame to block direct air
path noise to the operator from gears. The machine frame
casting had many openings, which were covered by these
steel plates, as shown in Figure 5.1.4.
Noise Reduction Achieved
Noise at the operator station was reduced from 99 dBA to 93 dBA

(93 dBC). With this reduction, an additional source was noted and
determined to be the staywire lifter arms. These were covered with
a 3/8-inch-thick piece of Lexan (see Figure 5.1.5) for the full
length of the operator position, hinged so that it could be easily
removed for maintenance.
Operator station levels were reduced to 89 dBA and were there-

fore in compliance with current standards.
A shelter could have solved the problem, but where possible,

attack on direct noise is recommended. When major noise sources are
reduced, the contribution of other noise sources can be better de-
termined and corrected. By redesigning the wrapper device from a
ratchet to a chain drive, the production rate was increased by 50%.
Costs
Costs were mainly internal plant labor for machine overhaul,

plus the cost of the steel barrier plates welded to the frame (esti-
mated at less than $100), plus the cost of the piece of Lexan at
$5.00/sq ft, or about $50 plus installation labor, in plant.
Pitfalls
The major pitfall in this kind of approach is moving too fast.

Testing each technique under actual conditions is far better than
moving rapidly into failure. This solution required approximately
2-years-Iapsed time from beginning to end.
82
OCTAVE PASS BANDS IN HERTZ
90 180 355 710 1..00 - 2800 - S600 - 1'200
120
110 _dBC
_dBA
(l)
"0
...J 100
w
>
W
...J
W
a:
::>
U) 90
U)
w
a: /-
0-
0
z /
::>
0U) 80
A weighted
0
Z
«
(l)
I
w I
>
«
I-
0
()
70
...
60
63 125 250 500 1000 2000 4000 8000

50
5
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.1.1 - Noise levels for 8-foot fabric machine, 1 meter from main drive gear
83
OCTAVE PASS BANDS IN HERlZ
180 355 710 UOO - 2800 - 5600 - 11200
120
110
aJ
"0
_dBC
...J 100 _dBA

w
>
W
...J
W
a:
:J 90
en
en
W
a:
a-
0
z
:J
0 80
en
0
/
Z
<C / A weighted
aJI
W I
>
<C 70
I-
0
0
/
60
I
I
63 125 250 500 1000 2000 8000
50
5 5 5
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.1.2 - Noise levels for a-foot fabric machine operator station
84
OCTAVE PASS BANDS IN HER12
90 180 35S 710 l«lO - 2800 - 5600 - 11200
120
110
CD
", 100
...J
W
>
W
...J -dBC
___ dBA
W
~ 90
if)
if)
W
CI:
a..
o
z
o~
if)
80
o
z<{
CD A weighted
W
> 70
<{
f-
()
o
60
=1=
I 63 121 250 500 I c~o
-+-
2CCO .000 8000
50
5 5
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.1.3 - Noise levels for 8-foot fabric machine, 1 meter behind operator station
in wire spool area
85
FIGURE 5.1.4 - Steel plate barrier with window (stop sign hung on it).
86
FIGURE 5.1.5 - Lexan barrier in two sections; slides up 'or access.
87
CASE HISTORY #2 - aOO-TON VERSON BLANKING PRESS -
VIBRATION ISOLATION
The SOO-ton Verson press is a massive unit weighing about

275,000 pounds, mounted on four footings set on heavy concrete
piers. Production on this press was automobile chassis steel sec-
tions of 1/4-inch steel about 10 inches wide and a to 10 feet long.
Normal operating speed was 30/min. Steel stock was fed to the
press from a reel. Noise levels were about 120 dB impact, 105 dB
quasi-peak, and 94.5 dBA at operator location, which was about four
feet in front of the press.
Control Approach Chosen
As a starting point to the total solution of the noise problem,

it was decided to vibration isolate the press and determine the
attenuation gained before working on other noise sources, which are
not part of this case history.
The press was operated in a single shot mode. Hence quasi-peak

readings for each octave band were more meaningful for ear effect
than rms readings (slow A-scale). The peak value is the maximum
level reached by the noise, whereas quasi-peak is a continuously in-
dicating measure of the average (over 600 msec) of the high levels
reached just before the time of indication and is thus lower than
the actual peak but is greater than slow A-scale values.
Vibration data were recorded for the support foundation, floor

near press, adjacent building column, and press structure at the
press feet, before and after installation of the isolators.
Isolators
From the data supplied on strokes per minute and press weight,
the isolators were specified to be Vibration Dynamics Corporation
(of Lagrange, Illinois) series BFM micro/level isolators, under the
press feet. No price lists are available because each isolation
problem is specifically engineered and quoted. Cost was about
$2,000 for the isolators, and installation by in-plant labor was
probably about $1,000.
Noise Reduction and Vibration Reduction Achieved
Adding isolating pads reduced the vertical acceleration at the

pier by 9.5 dB, as shown in Figure 5.2.1. Most of the reduction
88
occurred in the 2, 4, and 8 kHz bands. The vertical foot-to-pier
acceleration reduction was 30 dB.
In Figure 5.2.2 is shown the horizontal acceleration at the

pier. Adding isolation effected a 12 dB reduction in acceleration.
The horizontal foot-to-pier acceleration was reduced 36 dB by the
isolating pads.
The attenuation afforded by the pads is shown in Figure 5.2.3,

for both vertical and horizontal motion of the pier. Note that it
is the vertical motion that is responsible for most of the sound
radiated by the floor.
Figures 5.2.4 and 5.2.5 are comparisons of the sound level

readings at 4 ft before and after isolation (quasi-peak readings,
single shot operation). The calculated dBA level in Figure 5.2.5
shows a reduction of 6.5 dBA in airborne noise.
Isolators reduced vibration in support foundation, floor, build-

ing, column, and press structure. It has been found that a primary
cause of background, or ambient, noise is the vibration in the build-
ing structure, which is presumed to be caused by the anchor bolt
after-shock.
Calculation here shows that there was a 105-dBA quasi-peak

sound level before isolation and a 98.5-dBA level after isolation.
With a relationship of about 10 dB quasi-peak to rms, a reduction in
level from 94.5 dBA to 88 dBA at operator location has been made.
Additional presses will add their own noise and will increase levels
to above 90 dBA. Other operational noise sources in the press must
be controlled separately.
Pitfalls and Rewards
The major pitfall of this approach is that airborne sound level

reduction due to vibration isolation is almost impossible to predict.
However, a serious noise control program in such operations must in-
clude isolation devices for all presses.
A reward is that the.die life and maintenance of such machines

is significantly increased for presses that are vibration isolated.
Isolators improve operation and maintenance by reducing failures of
anchor bolts, foundation failure, or breaking of press feet.
89
90 180 355 710 1<00 - 2800 - S600 - 11200
o 1 - - -.......----+----+---+----+----1------4-- total
-10 1----+-----+----+----+----+
Cl
...
Q)
I-
z I
o
I- -30 1-_ _-+-_---lI~~---+_I_-_+_
<l::
a:
W
-l
W
()
()
l
-.
<l:: -40
-
-+-
t
--
!:+ 1 -r
--
---j--
~t
-j.
-t-
+ I
---L
63 ~so
'1
'" -J ."
i
"Cit:']
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.2.1 - Vertical acceleration on pier, before and after isolation
90
ISO 355 710 1.0(10 - 'Soo - 5600 - 11200
o
total
-10
0)
~
-20
CD
"0
Z
0
I-
~
a: -30
w,
...J
W
()
()
~
-50 I - - - - - l - -__I----+-----i---+--±
5
63 125 250
I
.:00
T
-t -
-,
Ie ;0
1f BOOO
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.2.2 - Horizontal acceleration on pier, before and after isolation
91
90 180 355 710 ,.00 - 2800 - 5600 - 11200
40 1---4----1----+---+-...,......-1---1---+----+----'1
OJ 30
-0
Z horizontal
0
I--
e::(
a: 20
w
...J
W
()
()
e::(
LL
0 10
z
0
I--
e::(
~
Z
W
I-- 0
I--
i
e::(
_l_
-+-
63 12~
2
250 :;00
I
lC00
i
20CO 4000 8000
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.2.3 - Attenuation of vertical and horizontal accelerations at pier by adding isolation
92
180 355 710 1400 - 2800 - 5l>OO - 11200
100
..J 90
w
>
W
..J
W
a:
~
rn
80
w
a:
0...
o
Z
::>
~ 70
o
z
«
III
'.
W
> 60
«
I-
U
o
I
-r
=+=
-1-
--t-
$
63
1 :25 750 500
T
- -
'--,--
1:-':"11)
I
2C)C
f
-
dOOO
. --+.
8000
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.2.4 - Quasi-peak levels 4 feet from press foot before and after isolation
93
180 355 710 10600 - 2800 - S600 - 11200
105
100
98.5
-l 90
w
> ,
W
-l
W
a:
after isolation,
,,
~ 80
en
W
a:
0..
o
Z
:J
g 70
o
z
.ex:
[IJ
ill
:;;: 60
I-
1
a
o
- l_
50
-
I
I t
6J 125 250 500 10'JO 4000 8000
40
100 1000 10000

FREQUENCY IN HERTZ
F IQURE 5.2.5 - A-weighted quas i-peak levels 4 feet from press foot before and after isolation
94
CASE HISTORY #3 - BLOOD PLASMA CENTRIFUGE
This plasma production room has two parallel banks of centri-

fuges, 15 to a bank, plus refrigeration units. A sketch of one
centrifuge is shown in Figure 5.3.1. Centrifuge spinning frequency
is 13,000 rpm (217 Hz). Though centrifuges appeared to be the major
noise source, refrigeration units were also considered. The same
refrigeration units without centrifuges are used in a separate re-
constituting room.
Measurements and Analysis
Operator level was 97 dBA with one bank in operation; 100 dBA
was predicted with both banks operating. Octave band readings of
operator levels are shown in Figure 5.3.2.
Often close-in measurements of a complex noise source will aid

in determining specific areas requiring correction. With this
thought in mind, we made close-in diagnostic readings at the loca-
tions shown in Figure 5.3.3. Octave band readings for each of these
locations are shown in Figures 5.3.4. through 5.3.9.
See Figure 5.3.3 for sketch of centrifuge showing locations of

close diagnostic readings and octave band charts, from 31.5 to
8000 Hz. Note from these the following facts:
(1) From Figure 5.3.2, operator position, note that the

major noise contribution to the A-reading is in the
500-Hz and 1000-Hz bands.
(2) All other locations show similar maximum bands when

close readings are made of the various suspect items
of this assembly, Figures 5.3.4 through 5.3.9.
(3) Highest levels with similar bands are noted for Figure
5.3.8, close to motor exhaust, showing that the motor
exhaust is the major noise source.
(4) From these octave band readings, the spectral content

of the overall noise can be gleaned. Of course, it
is the A-weighted level at the operator's work station
(dotted line Figure 5.3.2) that is OSHA regulated and
needs to be reduced, since the A-level measurement
characterizes the noise much as the human ear responds
to it. Figure 5.3.2 shows that the largest contribution
to the A-readings are from the 500- and 1000-Hz bands,
95
so our noise control diagnosis and correction
efforts should concentrate on these bands.
(5) The centrifuge spinning frequency of 217 Hz from

the original data adds to the weak 250-Hz level,
but as can be seen from Figure 5.3.2, is not a
large contributor to existing A-levels and can be
safely ignored.
Data from the reconstituting room are shown in Figures 5.3.10

through 5.3.12. This room has no centrifuge, so the noise source is
predominantly the refrigeration unit, which has a maximum at 500 Hz
(though only at 82 dBA) , but does confirm that this is a possible
noise source for the SOD-Hz band in the centrifuge plus compressor
room.
We conclude from the above data that the maximum band at 1000 Hz
is a major contributor. The same maximum occurs on Figure 5.3.8
(close to the motor exhaust) and in Figure 5.3.4, where the noise is
presumed to be reflected from the highly reflective metal pulley
guard surfaces. These data confirm that the major noise source to
be attenuated is the motor exhaust.
The results of our measurements indicated an unacceptable expo-

sure when the operator was exposed for 4 hours to both centrifuge
banks (Figure 5.3.2 predicted level at 100 dBA). Under these condi-
tions, the operator's daily dose was 2.0, and thus exceeded the
acceptable exposure of 1.0 as specified by OSHA noise regulations.
There are three general locations for controlling noise: at

the source, along the transmission path, or at the receiver. Two
could be used in this case. A properly designed and constructed
muffler or partial enclosure for the centrifuge motor exhaust would
provide the necessary source control, using materials and geometric
configuration that have effective attenuation in the 1000 to 4000 Hz
bands. Such a control measure would provide an expected 3 to 7 dBA
attenuation.
Noise control measures can also be used along the path of trans-
mission. The paths of airborne noise transmission were from direct
and reverberant fields such as walls, floor, and ceiling, which
supply very little noise absorption.
For direct field reduction, barriers with the proper trans-

mission loss, dimensions, and orientation may be used. The reverber-
ant field can be controlled by the addition of absorbent materials.
These combined measures would reduce operator exposure 5 to 15 dBA.
96
Although the above solution would satisfy the noise control

requirements, other solution possibilities should be discussed with
company representatives as they are able to specify important opera-
ting, maintenance, and production constraints that may limit the
type of noise control means.
In this case, after discussion, two of the possible control

methods were eliminated. The hard and impervious walls had to re-
main because of the need for daily sterilization by high pressure
steam and water hosing, and hand scrubbing. A porous absorbent sur-
face would provide areas for bacterial growth and would not with-
stand the rigorous daily cleaning. A barrier to shield the opera-
tor from the centrifuge motor noise was eliminated from considera-
tion because it would block the proper flow of refrigerated air.
A small warming of the plasma would produce an unusable product.
It was decided to try the motor exhaust muffler and see how much
noise reduction could be achieved.
The muffler was designed with a stainless steel outer skin,

lined with acoustical absorbent spaced 2 inches from the inside of
the steel shell, using small blocks. The 2-inch air space allows
absorption since it reflects from the inner steel surface and back
through the absorbent. The muffler has one 90-degree bend, as
shown in the sketch in Figure 5.3.13. In-house shop cost was esti-
mated at $300.
Noise at the operator's position was reduced from 97 dBA to

92 dBA, satisfactory for a four-hour employee. The motors of both
centrifuges must be treated if they operate together.
Pitfalls
An airtight seal at the junction of muffler and motor is very

important; we used resilient caulking compound as sealants. An
air leak of area no more than ten percent of the muffler cross sec-
tion would pass as much noise as if the muffler were not there at
all. An intake muffler would probably also help, but noise from
other sources would then become prominent.
Effectiveness of noise control is reduced if path of vibration

transmission is not held to minimum. Accordingly, it was very im-
portant to use as few absorbent spacers as possible. In so doing,
the steel skin vibration was kept small. The spacers can be made
of damping material to reduce resonant vibration of the skin.
97
-5° box
centrifuge
'-- ~ product tub

- - - -- - - - - -- - --- - - - -- - - -- - -------- - --
FIGURE 5.3.1 - Sketch of one centrifuge (front view)
98
90 180 JS5 710 1400 - '800 - 5600 - 11:200
predicted level
two centri fuge
100 1----+---+---+-----1----I-----!-+----!.... banks on: dBA
-j...,t--_ _ =l=l- one c~ ri fuge
- bank on: dBA
-l 90
W
>
W
-l
W
a:
:>
en 80
en
W
a:
a..
o
z
:>
oen 70
o
Z
<t: I
CD
w / A - wei ghted
>
<t: 60
l-
t)
o
50
40
I
I
63 125 250 '00 1C'10
I
=t=
2000 4000 8000
100 1000 10,000

FREQUENCY IN HERTZ
FIGURE 5.3.2 - Operator exposure in -5° C box room (below clock, 3cm from clock wall)
99
100
90
.... ~
A
, .- .... 1-- , - ..... ~
..;,;, -
If ./ --. II' . . r--.
80 ~ - l- t- 1"- - \-
I
)
70 ,• ~ J
I IJ
60
I
i iI'
I I
FIG. 5.3.4 FIG. 5.3.2 FIG. 5.3.7
operator exposure
~
" l-
PUlley~
, , mOIO'Q ~
~ ~ ""'"
V- 1/
,
II
~!""
•
FIG · 5.3.5
belt/1'
6 l-
~t.::
centri fuge
"\. l..:-'"
eXhaus~ .....
I
•
FIG 5.3.8
I spout
~ ..-
"\. ~
product tub 7-
-
i
, \J .4 ~
~'
,-l-- I -
~
...
, l' ,... II I
~
I
~.
•
FIG . 5.3.6
•
FIG. 5.3.9
FIGURE 5.3.3 - Sketch of centrifuge showing locations for close - in diagnostic readings
100
lBO 355 710
100
90
lD
"0
...J
W
>
W
...J 80
w
a:
::l """1-
CI)
CI)
W I
a: I
1· - I
a.. 70
o
1 :.-1:
+
Z
::l
o
CI) -:~- ±
-_./ A-weighted t·
o
t~
~ 60
lD
l.U
>
~
I-
()
o I
50
/
/ .:..L:.
J
1-J _·=
.....
-I
',' ,
40 i I
100 1000 10,000

FREQUENCY IN HERTZ
FIGURE 5.3.4 .. Centrifugenoisemeasu:edparallef to motor shaft (2 cm from pulley belt guard; between
#25 guard and motor exhaust of adjacent centrifuge)
~-------
C Reproduced trom
[ ,\ best available COpy· ::
G - ~-- --.,-;;
101
90 180 355 710 1~OO - 2800 - 5600 - 11200
100
[l) 90
"0
...J
W
>
W
...J
W 80
a:
::>
en
en
W
a:
e..
0 70
I-
Z
::>
0
=1_. .
en
0
Z
-l-
--L
,r:... - wei ghted
± 1
<l: -~.-
[l) 60
w
>
<l: ~
I-
() ~
0
50
*-
I
63 '", , ~OOG sr.oo
40 " i
,, I'
I
100 1000 10,000

FREQUENCY IN HERTZ
FIGURE 5.3.5 - Centrifuge noise measured 15cm directly below pulley
102
180 355 710 1'00 - 2800 - 5600 - 11200
100
ro 90
"0
..J
W
>
W
..J
80
W
a:
::>
en
en
W
a:
0-
o 70 -
z
::> I -
oen
o
z
03 60
I 'i A - weighted
W
>
<l::
l-
t)
o -'-
50
--1
j
=t~
63 175 750 ~OO 1\"')0 :-r,,~I,; 4000 BJ00
40
100 1000 10,000

FREQUENCY IN HERTZ
FIGURE 5.3.6 - Centrifuge noise measured perpendicular to spout (3 cm away, facing clock)
103
--t-
'd;J{.. -- 78C'1 S600 - 1',00
--,----,-----.,
l- I
100 1-----+----+----+----- - - - 'j

-1
--1---
- -
t ,---+----1---1
+
%J 90 1------ ---- ----I---)~-1------

-r /1. ----
1]
---"1liII.7---t--------- f---4
-.J
I' '
lT/ -r- -I -~!-
W i
>
~-t ----(---.
'1 - --------t---l---.-t--
W
-.J
W 80
a: -r 1---
1 -I
:::>
U)
U)
w
a:
-~
j
i
j /
,
,I'
-+-
-j
L
i
-I-
1.
-
--
-- I
- -
-
I
--+-
~
T
--
-
D..
o
Z
:::>
o
U)
70
~-,I{- Iwe;ghl-t, -I---T-,j--~--,.--
o
z
« I
- :.
~
l
i I '
j 1 I
-
ro 60 -- --1- ----- t ----J_-----+------+-----1f-----t---.--i
::_ +T 1 l r
, I _ ~ _
UJ
>
«
t-
O
1 1 t
o
50
J ~rl-~fl- -l---l---
f- -- -: -r
40
100 1000 10,000

FREQUENCY IN HERTZ
FIGURE 5.3.7 - Centrifuge noise measured between centrifuge motor and pulley belt (normal
to shaft 10 cm away)
104
180 355 710 l~oo - 2800 - 5600 - 11200
100
-l 90
w ~ - ..-
>
W ~
-l
~
W
a:
~ 80
en
W
a: I
a.. I
o I
z
5en 70
o I
Z
<{
m A - weighted
w
> 60
<{
I-
o
o
I
I ,
i
50
40
I
~J 250 500
t dOOO 8000
100 1000 10,000

FREQUENCY IN HERTZ
FIGURE 5.3.8 - Centrifuge noise measured 2 cm from centrifuge motor exhaust
105
180 355 710 1400 - 2800 - 5600 - 11200
100
..J
W 90
>
W
..J
W
a:
::J
en
lZ
a:
80
CL
o
Z
::J
oen
o 70
Z
ex: I
CD
w I
>
ex: I A - weighted
I- 60
u
o
I
I
!-----1:
50
t:
-r-
-l" - r- --
I --+- -
6) I-
40
100 1000 10,000

FREQUENCY IN HERTZ
FIGURE 5.3.9 - Centrifuge noise measured at tub height (2 em from front tub edge)
106
., - 90 180 355 710 1400 - 2800 - 5600 - 11200
100
...J
W
> 90
w
...J
W dBA
a:
:::>
(/)
(/)
w 80
a:
0..
o
Z
:::>
o
(/)
o 70
z
«
CD
w I
> I
« A - weighted
t; 60
o I
I
I
50 -+-
40
5
63 125 150 JOO
*
--
I
-~r~
IC":O iO::~ _000 800C
100 1000 10,000

FREQUENCY IN HERTZ
FIGURE 5.3.10 - Operator's exposure in reconstituting room (center of room, ear level of a standing
worker to the right of tank # 24)
107
90 180 3SS 710 TAoo - 2800 - 5600 - 11200
100
CD
"0
...J 90
w
>
W
...J
W
a:
:> 80
en
en
W ~
a:
C-
O
z
:>
"
0 70
en
0 I
z I
«
CD
w
I A - weighted
> 60 I
«
f-
a
0
I
50 -r
---L.
I
40
I
,so
--.--
,JC
r-t--
\r'10
J
](·,"0 '000 BOOO
100 1000 10,000

FREQUENCY IN HERTZ
FIGURE 5.3.11 - Noise in reconstituting room, measured 2 cm from bottom edge of exhaust face
of refrigeration unit
108
180 J~5 710 1.00 7800 - '600 - 11700
100
.J 90
W
>
W
.J
W
a:
~
en
w
a:
a..
80
I P~
- -t'~ ...
t1
o
z
1----1 ,I
:J
oen 70
o
z t- I -"'_
~t·;9hI1 -t
«
[D
w
«>
I-
U
60 I
i c
ti--_.
=1=J: i I
.
i
o
50
1-
~--+-+--t.·
{ t1_ 1-.
~~ .u;-_.
-j
=t ~ f
j __--~-:-----I
I ~J= ::t i~
40
I T 1- -1.--
'---r-~-"-"""""-·rf" - - - ' - - - . . -......

.-
I
.. (
-,-
+
-l- t
,.
'--r-...,----+--.-....,.-,,-II----+----,--+--..---,--+,--,-,,---
l
100 1000 10,000
FREQUENCY IN HERTZ
FIGURE 5.3.12 - Noise in reconstituting room, measured 2 em below rear intake face of
refrigeration unit
109
~I
18 inches
I
I
I
I 6 inches
r,--------(."""-
I " l---------.. .~
I Exterior shape
. .~_ _ 12 inches
I and dimensions
I 12 inches
I
I Cross section of muffler
I showing absorbent, bend,
and dimensions. Absorbent
I
is 2 in. from surfaces.
---(
'"
FIGURE 5.3.13 - Motor exhaust muffler
110
CASE HISTORY #4 - BLANKING PRESS RAM
In forming operations, large blanking presses are used. The

ram, which is like a connecting rod in a reciprocating engine, is
hollow. The forming die runs in grooves on the side of the press,
like a piston in the cylinder of a reciprocating engine, and com-
pletely closes off the end of the hollow ram. There are slots in
the ram that are used normally when the press is used in blanking
operation to extricate the work from the die. This is similar to
removal of a cookie from a cookiecutter. These slots are in the
side of the ram (see Figure 5.4.1). When the press is being used
in the forming mode, these slots are not required, and when the
die "snaps through", it cuts off the work. 'This gives rise to
high noise levels.
In a vibration isolated forming press, operator position noise

levels were LA = 94 dBA, LC = 100 dBC in the "slow" reading position.
An octave band measurement disclosed that levels in the 250-, 500-,
1000-, and 2000-Hz bands were much higher than in other bands. This
ringing noise, which had a maximum near 2 kHz, was easily discern-
ible by ear. By careful listening, it was determined that the
source was radiation from the slots in the ram.
The technical conclusion was that the ram hollow interior plus
slots was essentially behaving as a shock-excited Helmholtz resona-
tor. A Helmholtz resonator is a closed volume of air connected by
a tube, which will resonate at various frequencies (like blowing
across a glass jug opening).
The one approach that would obviously work would be to fill the
cavity in the ram with rubber-like material. Another approach would
be to simply plug the slots and keep the noise inside the ram. The
second approach was chosen because it was easy to try and certainly
not expensive to test. It would also allow the machine to be easily
reconverted to a blanking operation.
The ram slots were each covered with a plywood plate sealed
with a Neoprene gasket, as shown in Figure 5.4.1. Weatherstripping
(nonhardening sealant) was used to prevent small ·leaks. These con-
trol measures were easily installed.
111
The first attempt was satisfactory and achieved a 6-dBA reduc-

tion of quasi-peak A-weighted level from 99 to 93 dBA OP. See '
Figures 5.4.2 and 5.4.3. Applying this to the observed slow A
reading of 94 dBA yields the observed 88 dBA.
This case history demonstrates both the simplicity (the solu-

tion) and the complexity (the resonator) of noise control. It also
demonstrates a more subtle feature: simple solutions are worth
trying if there is a good physical reason for them.
Pitfalls
The obvious pitfall here would be to apply this solution to a

press that had not first been vibration isolated. If the press were
on other than piers isolated from the building, or had sheet metal
guards, one would probably not have been able to measure any improve-
ment. Filling the ram cavity would have been another pitfall. It
would have accomplished the noise reduction, but would have prevented
easy reconversion of the press to blanking operation.
112
weather stri pp i ng
,cement
"-
ram plate (typical)
gasket
through bol t
PRESS FRAME
ram SLOT
NOTE:
use Neoprene
gasket both
sides; seal
washers and joints
with weatherstripping
cement
FIGURE 5.4.1 - Method used to cover slots in blanking press ram
113
90 1eo 355 710 '.400 7800 - 5600 - 11200
120
110
=r::
CD
"C
-I 100
W
>
W
4 -dBA~ QP
-I
W
a:
~ 90
en
W
a:
a..
o
z
3 80
il f + +
en
o
z
«
CD
t
w
~_ _-+-...I__--1A_w_e_i_gh_t_e-td- - - - e - - - - I r
1£.
>
« 70
I-
o
o
ft + I :+ , -
60 .--t-f~r--- ---:----1----+--.
+ t
1t
-~ --l ::,t
.- :: ~ T ~. j-
t
. 1
-<- -r -r-
..., '!
JI 'f
50 -+-----"-.,.--.,--,-r----r-----
100 1000 10.000

FREQUENCY IN HERTZ
FIGURE 5.4.2 - Quasi-peak readings of blanking press before nOise control
114
90 180 355 710 1400 - 2800 - 5600 - 11200
120
110
rn
"0 100
..J
W - _ Quasi-peak
>
W
..J dBA QP
W
a: 90
:J
en
en
W
a:
0-
o
z
:J
80
oen
o
z
«
rn
! --I
A weighted
.
I
w 70
«>f-
£
f !
a
o
60
,, -'-
:j=
I----#~f-----t---_+---_+---~-;~-.-- --:~t:----=t---±----l
_.~ .1- =f 1
50
I
t·.l
f ,j _.sr.
L&.-r--.-....,...'-T-"T"'"1--r---'----r-- ....
5
+
"ll:"
-1
~,'ol'
i
' "-T"-....,....--II-.--r--r-T"""Ti-+!------t-t - -
5 I 2
T
J:', lr~
·:f
3'~'O')
, I
100 1000 10,000

FREQUENCY IN HERTZ
FIGURE 5.4.3 - Quasi-peak readings Of blanking press after ram ringing was contained
115
CASE HISTORY #5 - SPINNING FRAME AIR NOISE
Cord manufacturers use a machine called a spinning frame to

convert yarn to cord.. In the process of spinning this yarn into
cord or thread, lint or small pieces of yarn fall away. At each
spinning station along the frame, air suction removes this lint
by a system that works essentially like a vacuum cleaner. This
system requires a rather large air moving system for each spinning
frame, and the noise created from these air moving systems causes
the ambient noise levels to range from 88 to 93 dBA at the work
stations throughout this system. This unit was a Whitins Model
M-2.
Measurements were made with a Type 2 sound level meter. At

about 2 cm from the air exhaust of the lint scavenger system, the
levels were: LA = 100 dBA, LC = 100 dBC. The. major noise source
was unquestionably the air escaping from the lint removal system,
as was verified by the fact that LA = L e . This problem is cornmon
in high velocity air systems.
The obvious solution to a problem of this nature is to use a

muffler or an acoustical isolator. However, a more fundamental
approach considered was to slow the escaping air at the scavenger
exhaust. This could be accomplished by simply giving the exhaust
vent a bigger open area, as shown in Figure 5.5.1. The velocity
of the escaping air was estimated to be 115 ft/sec (the fan moved
1800 cfm through an area of about 37 in. 2 ). To simply open the
fan cover was not practical, since the air must be directed up-
ward.
The reason this control approach is a good one to consider can

be best summarized in the following relationship:
5
X = 10 10910 (Vo/V n )
where X is the reduction in decibels
V is the original air velocity

o
V is t~e new air velocity.
n
This equation is widely used by noise control engineers to estimate
the relative noise reduction if air stream slowdown is possible.
116
Our design increased the area through which the air exhausted
by a factor of 10: from 0.26 ft 2 to 2.6 ft 2 . Because the flow is
practically incompressible, Vo/V = 10- 1 and X = 50 dB reduction.
However, the net noise reductlonnwill ordinarily be less because
other noise sources are still present. A useful rule of thumb is
to expect a useful reduction of at most 10 dB if a major source is
completely removed. The chief exception to this is the intense
and often high frequency pure-tone single source, such as a whistle,
stearn vent, or automatic 'control valve.
The noise was measured, with all but one fan cover unchanged;
it was LA = 93 dBA and LC = 94 dBC, a reduction of 7 dB in A-level.
It is thought that this reduction fairly well represents the back-
ground level without this fan running.
Pitfalls
The most common pitfall in a treatment of this kind is to at-

tempt to do a makeshift or sloppy final job. Care must be taken
for the final result to be effective. A professional metal shop can
fabricate the device shown in Figure 5.5.1 much more easily and,
certainly in quantity~ much less expensively than it could be fabri-
cated in your plant. The rubber gaskets and sealant are both impor-
tant to the overall effectiveness of the job.
This kind of noise control is certainly straightforward and is

not expensive. The cursory results reported here make it well worth-
while. The plant for which this was designed will no longer find it
necessary to require hearing protection for its workers.
117
NOTE:
1 - for final assembly, seal all gaskets with
nonharden i ng weatherstri ppi ng cement
2 - make certain existing fan cover is strong
enough to support modification
vent - muffler, to be designed and

attached only if required
new air exhaust stack, directs

air upward to vent - muffler- --1~--_ ...
exi sting fan cover, modified to

accept new air exhaust stack
FIGURE 5.5.1 - Air exhaust vent modification for spinning machine noise control
118
CASE HISTORY #6 - BARLEY MILL
Excessive noise levels existed around Moorspeed and Ross barley

mills (rolls S-inch-diameter, 15-inch-long), a hay shredder, and a
control operator's chair in a cattle feed grinding mill. The ob-
jective was to reduce the noise level at the operator's position
for OSHA compliance.
A and C level and octave band measurements were made between

the Moorspeed and the Ross mills (Figure 5.6.1) and at the hay
shredder with both mills in normal continuous operation (Figure
5.6.2). With LC - LA = 9 dB, excessive low frequency noise levels
were predicted. These were confirmed by octave band measurements.
Octave band measurements at the control operator's chair are shown
in Figure 5.6.3. A sketch of the room showing the relative loca-
tions of the equipment is shown in Figure 5.6.4.
Roller crushing actions produced high noise levels, and cor-

rection by machine redesign was believed to be too costly a method
for solving this problem. When the source is too difficult or
uneconomical to attempt to correct, working on the noise path will
often result in a more economical solution. Therefore, a partial
enclosure, open at the top, was chosen.
Although walls can be of solid construction with a minimum of

access doors, in this case access was needed for adjustment,
maintenance, repair, and roll replacement. For roll replacement, a
forklift truck entry was required. For ease of quick access, a
fixed barrier wall was discarded in favor of a lead-vinyl curtain
wall extending, if required, up to the 17-ft height of the roof
-support beams. All three noise sources could be enclosed by two
curtain walls enc19sing the corner of the building as shown in
Figure 5.6.4. The curtains run on rails for easy sliding back and
are held together by Velcro closures.
Figure 5.6.1 shows that, if the sound levels from 250 Hz up

are reduced by at least 14 dB, the resulting A-level readings would
be less than 90 dBA for compliance outside the curtain walls.
119
Barrier wall attenuation is limited in three ways: (1) direct
transmission loss in each octave band, (2) noise over the wall, and
(3) room absorption, noise-source side.
(1) Direct transmission loss (TL): The manufacturer of lead-

vinyl fiberglas curtains, 0.75 Ib!sq ft, was chosen. Manufacturer's
literature gave the transmission loss in each octave band as fol-
lows:
125 Hz 250 Hz 500 Hz 1000 Hz 2000 Hz
TL = 11 dB 16 dB 20 dB 26 dB 31 dB
It is seen that the transmission loss is not a limiting factor.
(2) Noise over wall: Barrier wall attenuation can be esti-

mated from data in Noise and Vibration Control, Leo L. Beranek,
page 178, using the dimensions from Figure 5.6.4 and from the sec-
tional view in Figure 5.6.5.
N = ? (A + B - D) = f (16.6 + 16.6 - 18)
N = -A-
30.4 (Fresnel number)
125 Hz 250 Hz 500 Hz 1000 Hz 2000 Hz
Ix = 9.6 ft 4.8 2.4 1.2 0.6
N = 3. 2 6. 3 12.6 25.2 50.4

Attenua-
tion (dB) 14 16 18 20 20
(from graph page 178, Beranek) . In practical situations, the attenu-

ation is limited to about 20 dB.
By a rough first approximation procedure we can obtain an esti-

mate of the reduction afforded by the curtain walls. In the listing
below we start with the worst-case octave band levels of Figure
5.6.2 and then list the transmission loss and barrier effects just
calculated. Subtracting the minimum of these two reduction mecha-
nisms yields a tentative spectrum of the resulting sound in the room.
After A-weighting and combining levels, the predicted reduced room
level is 84 dBA.
120
Octave bands 125 250 500 1000 2000
Noi se sour-c-e-----:;-l-;:;0:-::6:-----:;-l-;:::0~1------;:;:9-;:::8:----9=7-----:9"""0-
Direct TL 11 16 20 26 31
Over wall 14 16 18 20 20
Reduced levels 95 85 80 79 70
A-weighting -16 -9 -3 0 1
A-weighted 79 76 77 77 71
A-level 84 dBA
For visual access the enclosure can have 10 x 20-inch plastic

windows placed to order; use only minimum number. The curtains
should be long enough to drag a bit on the floor to reduce leaks.
Some rerouting of power, steam, and air lines may be required.
The approximate 1973 costs were: $4.00/ft 2 for curtains made

to order with grommets; Velcro fasteners, $3.00/ft; track, $1.50/ft;
rollers (one per grommet), about $2.50 each; windows, $25.00 each;
total cost about $4,000.
The preceding simplified treatment neglects an important fact:

we have not gotten rid of the noise, but have merely redistributed
it. Thus, the total sound power from the machines escapes from the
topless enclosure and spreads throughout the room. Close to the
curtains, there should be some reduction, but very little farther
away. Absorption is required to actually reduce .the sound power.
This was next considered.
(3) Absorption, noise-source side of wall: When noise sources

are confined to a space with less absorption than before, they may
build up higher levels due to reverberation. The sound barrier
curtain material can be obtained with sections of sound absorbent on
the inside, to counteract this effect. In the barley mill, however,
this was not recommended as the porous open material could easily
become dust clogged. Shortly after this noise control job was com-
pleted, absorbents covered with a plastic film became available.
At the time the recommendation was for an easily installed and main-
tained material, Owens-Corning Fiberglas Noise Stop Baffles. These
are 23 x 48 x 1.5 inches and comprise an absorbent board wr~pped in
a washable, noncombustible plastic film; each baffle is supp~ed
with two wires through the 23-inch dimension. These wires terrn~nate
in hooks; to install, stretch wires ~ three feet on center, parallel~
to the line joining the two mills and about flush with the top of
the enclosure rails. The baffles are available through local acous-
tical material suppliers. The cost is about $lO.OO/baffle.
The enclosure developed by the curtain walls is in effect a

separate small room, and the noise reduction can be estimated from
the relationship of total absorption before and after adding the
sound absorption panels. This relationship is
121
dB Attenuation = 10 log a /a
2 1
where: a 2 is new total absorption
a l is original absorption
(from reference: Harris Handbook of Noise Control, pages 18-19).

Original absorption, a:
Area x Coeff. = Abs. (sabins)
Long wall 32' x 17' x 2 x 0.02 22
End wall 17' X 17' x 2 x 0.02 11
Roof 17' X 20' x 1 x 0.02 7
Absorption by adding 100 panels 2 x 4 feet:
100 X 2' x 4' x 2 sides x 0.8 (average A-weighted

absorp. coeff. of
panel) = 1280
Original absorption 40
New total absorption 1320
dB attenuation = 10 log :~ = 10 log 1~~0 = 15.2 dB
Resultant level = measured level - reduction = 86 dBA.
The measured final noise level was 87 dBA, giving a 7 dBA reduc-
tion. This was 3 dBA lower than the maximum desired level, and was
the result of paying careful attention to elimination of leaks.
The room formed by the curtain did not realize such.a reduction, but
since these machines required no attention while running, the noise
exposure of personnel was significantly reduced below D = 1. The
major remaining path is reflection from the ceiling.
~itfalls
Barrier walls of various heights can often be used between a

noise source and a machine operator. A major pitfall is that, in a
122
room with a high level of reverberant noise, the partial barrier
will be short-circuited by the reflected noises from walls, ceilings,
and other surfaces. In such cases, attenuation based on the partial
wall theory will not be obtained, often resulting in no attenuation
at all in highly reverberant rooms. Curtain walls must be kept
closed to get attenuation. Sound absorbing units must be kept clean
to be efficient.
Even in a semireverberant room, a reduced barrier height can be

used. In this case a 7-ft barrier should ideally reduce the level
to 89 dBA at the receiving location. However, since the semi rever-
berant conditions will introduce more reflected sound with the lower
barrier, the high wall used in this case history is recommended
because the added absorption within the barrier area has, in effect,
made a separate small room and created the condition on which the
barrier wall theory was based.
123
90 ISO 355 710 1400 - 2BOO - 5600 - 11200
120
_ peak
110
_ dBC
m
"0 100
-l _dBA
w
>
W
-l
W
a: 90
I 1-
:J
U?
U?
w
a:
I
a. - I
0
Z 80
- =I A weighted
:J
0
U?
0
z /
«
m ~
w 70
>
«
I-
0
0
60 r-
50
I
150
I
500
i
=r::
IC')()
I
I
±
"*
=1=
2000
I
2
4000
I I
5
I
SOOO
~ I
100 1000 10,000

FREQUENCY IN HERTZ
FIGURE 5.6.1 - Barley mill noise levels, measured between Moorspeed and Ross mills (4m out,
1 m up, both mi lis on)
124
180 355 710 1./00 - 2BOO - 5600 - 11200
120
_peak
m
:' IQG
...J
110
-r
_dBA
UJ
>
UJ
...J
UJ
a:
:l
90
C/)
C/)
UJ
a:
0..
o A weighted
~ 80
o
C/)
o
I
z
« I
m I
UJ 70
>
«
~
()
o
60
I
I
63 125 250
I
1000 2WO 8000
50 5 5
100 1000 10,000

FREQUENCY IN HERTZ
FIGURE 5.6.2 - Barley mill noise levels, measured at guard rail of hay shredder (2 m opposite
blower, both mills on)
125
180 3SS 710 1.400 - 2800 - 5600 - 11200
120
110
_ peak
r
_dBC
CD
"0
100
-i
w
>
W
..J _dBA
w
a: 90
::>
(J)
(J)
.~-
W _J _
a:
a..
0
z
::> 80
0(J)
-j-
0
Z ~ A weighted
<{
CD
w /
/ ±
> 70
<{
I-
()
0 I
I
60
I
I
-i:
63 125 250 500 lenD 20CQ 4000 8000
50
S
100 1000 10,000

FREQUENCY IN HERTZ
FIGURE 5.6.3 - Barley mill noise levels, measured at control operator chair (ear level, operator absent)
126
--,,.....--_-----------------r----------or--------,
~. ~\ '
• I Moorspeed Ross
\ lead - vinyl curtain~ test 7
2,,\:3'
g' \
;'
_ _ _
. __
17'l~ _°r:~ h~:'~:'_ -f 32' _ door
I operator
42' I
1// ~~-- 70'

roof 17' at walls, 26' at
door
rt.
~
... .1
FIGURE 5.6.4 - Floor plan of barley mill
127
FIGURE 5.6.5 - Sectional view of barley mill
128
CASE HISTORY #7 - SHEETER FOR BOXBOARD
The sheeter, starting from large rolls of boxboard about 6 feet

in diameter, cuts the web to length using a rotary knife that can be
adjusted to rotary speed, and therefore sheet length, by means of
variable speed drive (Reeves Drive). The cut sheets are delivered
to pallet. The speed is about 700 ft/min.
At the operator control station near the sheeter (see Figure

5.7.1), the noise level was found to be 93 dBA. Close-in probe read-
ings at the variable speed drive were high, indicating that the
drive is a major noise source. Readings were as follows:
96 dBA close to front drive guard, in aisle.

98 dBA close to front drive guard, in aisle 3.
105-107 dBA close to front drive vent openings.
The drive box enclosure was a steel shell 6 ft high, 3.5 ft wide,
and 3.5 ft deep, having two vent openings in the side for natural
air cooling (see Figure 5.7.2).
Other operator locations that were far from the drive were
checked:
90 dBA operator at delivery

88 dBA operator at rollstand in feed (see Figure 5.7.3
for general layout.
From the close-in readings, the drive was determined to be the

major noise source and not the roll unwind stands, rotary cutter, or
delivery belts to finished pallet of boxboard.
To reduce the drive noise within the steel box enclosure, it

was decided to line the interior walls with an acoustic absorbing
polyurethane foam with a layer of 0.017-inch-thick sheet lead to
provide damping of the steel surface panels. To reduce the noise
corning out of the air vents, an acoustic traps was designed to ab-
sorb the noise at vent but allow full normal air circulation. This
acoustic trap is shown in Figure 5.7.2.
129
The level at the operator control panel near the drive unit
was found to be 89 dBA, reduced from 93 dBA. In addition, some
reduction was obtained in other operator positions:
86-87 dBA from 90 dBA operator at delivery.

86 dBA from 88 dBA operator at roll stand.
Close readings at vents were reduced to 94 dBA from 105 dBA; this
is not an operator position.
Sound absorbing polyurethane foam with a lead septum designed

for combined damping and absorption is available from various sup-
pliers at less than $4/ft 2 ; material cost was about $400, and in-
house labor to glue in place and fabricate a holder for the sound
trap was about another $400; total cost was about $800.
The polyurethane sound absorbing material shows a 0.90 absorp-

tion coefficient at 1000 Hz and above. Although the absorption
coefficient at lower frequencies is less, the effect of increasing
A-weighting with decreasing frequency makes the loss of absorption
at lower frequencies less important.
Pitfalls
Without close-in reading to locate the drive unit as the

major noise source, we could have jumped to the conclusion that
the entire sheeter, including the drive unit, must be installed
in an acoustic enclosure, thereby spending a great deal more money
for the solution.
This kind of noise reduction is typically not as satisfactory

as one would like. The major problem that can arise is the exist-
ence of other direct sound paths from the knives to the operator.
Another pitfall for sheeters is the knife design. Some of

the older models have straight knives instead of an angular striking
or cutting edge. Straight knife sheeters will probably require an
acoustic absorbing lined metal or wood hood over the knife assembly
and perhaps under the knife assembly.
130
.....l~==::;~-'-.:o.!:r:k~f=- _
sheet web
(cut)
~
I vents (see FIG. C 7.2)
variable speed drive
I •
I I 6 it high X 3.5 ft X 3.5 ft
L ...J in steel cabinet
FIGURE 5.7.1 - Floorplan of sheeter·for boxboard
II I I
I
vent
I
noise
acoustic
foam rubber /barriers
6'
vent
I I
,~
3.5' .~~I
-w " ,_4 --~~-f,
FIGURE 5.7.2 - Sheeter drive box enclosure
131
sheeter --_~r----, rotary knife
roll web
roll operator l ope,a,,,,!

rr
delive,y
drive
~ ~';
I ,
/
1 ........- 1
L ..J
FIGURE 5.7.3 - Layout of sheeler and operators
132
CASE HISTORY #8 - AIR SCRAP HANDLING FROM BOBST CUTTING PRESSES,
FOLDING CARTON MANUFACTURING
In the folding carton industry, printed sheets are cut on Bobst

and similar cutting presses, equipped with automatic strippers for
removal of waste material between cartons. When the press is oper-
ated and is in good mechanical adjustment, there is no serious noise
problem. Often, however, noise from the scrap disposal system re-
sults in levels above 90 dBA on the pressman platform.
This popular scrap disposal system (see Figure 5.8.1) used a

horizontal air vane conveyor to move the scrap from under the strip-
ping station to the intake of a centrifugal fan that pushes the scrap
to a baler or to bins at a baler in a remote location.
The noise problem arises from the pieces of paper scrap striking
the sides of the intake conveyor under the press stripper, the sides
of the intake hood to the fan, and the fan and outlet ducts. These
all contributed noise that resulted in noise levels of over 90 dBA at
the pressman station~ Depending on amount of scrap and size of
pieces, the noise level reached 95 dBA each stroke of the press,
normally making the noise almost continuous.
In this type of problem, it was not considered necessary to

make octave band measurements when simple direct sound level read-
ings would tell the story of the obvious problem before and the
results after damping. Octave band levels aid in determination of
the noise source, but in this case the noise source was known and
before and after levels could be expressed in dBA.
The sheet metal of the stripper intake, fan intake from hori-
zontal air vane, the fan, and outlet ducts were all damped (and
transmission loss improved) by gluing a layer of lead sheeting to
the outside surfaces, using a resin glue recommended by supplier of
sheeting. Sheeting used was 1/32-inch thick, 2 Ib/sq ft.
Other sheet damping materials that are on the market could have
been used as effectively, as discussed below.
The damping of the sheet metal reduced the noise level at the
pressman platform to 88 to 90 dBA.
The concept of using sheet lead to damp the sheet metal ducts
came from supplier literature citing successful sheet metal damping
on ducts and fans and other surfaces. (Cost is about $0.90/ft 2 .)
For less damping, a 1 Ib/ft 2 material may be used·at $0.46/ft 2 .
For minimum damping, stiff roofing felt may do. For even greater
damping, there are many products on the market in sheet form and
tape form. Suppliers can be consulted on specific problems~ prices
range from $1.50 to $3.50/ft 2 ..
For very high vibration- and noise levels, a further duct treat-
ment step would be lagging, which is a spring-absorber-mass combina-
tion of 1 to 3 inches of resilient acoustic absorbing material
(Fiberglas or polyurethane) with a heavy cover sound barrier of
sheet lead or lead-filled vinyl sheeting over the entire surface.
See Case History #9 for an example of this method.
134
5 hp motor ___
fan
...~
~ ,.
./8 vertical discharge duct
inch diameter
damped to 10 feet from floor
"1' . . ) !J
intake duct
7C~ I I
I I
: I
- - I I - - - -
I I
out load stripper I I press feed in load
I I
: I
top
view
III I I II
pressman platform
side
view
-stripper-- -------
= out load- press feed - in load
pressman platform
floor ~ _ _.......tL--.l._--I_ _-,""~-r--_-.J...LL.. -.l. ...L.._~"""_"""
air - veyor
FIGURE 5.8.1 - Scrap handling system for cutting press
135
CASE HISTORY #9 - JORDANS FOR PAPER MILL FOR BOXBOARD
In a boxboard waste paper mill, Jordans are used to refine the

waterborne pulp. A Jordan consists of a conical shell and rotating
conical plug, both with steel blades on their mating surfaces. The
refining action can be adjusted by decreasing the spacing of the re-
volving cone to the stationary cone within the limits of power
available from the drive motor. In this case, drive motor power is
500 H.P.
The action of the Jordan refiner results in high noise levels,

as shown in Figure 5.9.1. With maximum refining power the noise
level was found to be 97 dBA at the aisle (102 dBA one foot from
Jordan surface) during start-up with only one Jordan in operation.
As normal operation required six of the eight Jordans in operation
with varied power settings, depending on degree of refining required,
the operating noise level in the aisle was often at about 100 dBA.
Although the paper maker did not have to be in the area continuously
during the shift, he was required to make periodic adjustments, which
took more time than the maximum exposure_ of 2 hours at 100 dBA per-
mitted by current regulations. Some noise reduction was therefore
required.
It was determined from vibration measurements that the major

vibration surfaces were the conical shell, end cover plate, and out-
let box. Shell acceleration measurements were made with a hand-held
pickup attachment to the sound level meter to determine the major
vibration surfaces as summarized below:
Areas selected for treatment Vibration

(g)
Shell surface, large end 15.9

Shell surface, rib 20
Flange face 10
Shell surface, center 15.9
Shell surface, rib 11.2
Shell surface, small end 5
Outlet box 7.9
Areas not treated
Bearing support 2.5
Overhead piping 1
Drive motor and bell (1) 0.9
Drive motor and bell (2) 1.4
136
A-level noise readings only were obtained because octave band
data would not have added any information for these purposes.
Two possible solutions were considered: (1) complete enclo-

sure and (2) partial lagging of Jordan shell. Total enclosure of
each Jordan was possible but was not considered as practical as
lagging because there were eight Jordans near each other, because
adjustment and maintenance would create additional problems, and
because the customary sound absorption material inside the enclo-
sures would not be compatible with the wet conditions of this paper
mill. Partial lagging was chosen since it could be installed on
the major vibrating surfaces producing the noise. Separation of
lagging could be designed to allow for maintenance dismantling of
the front face and plug. The noise absorbing material of the lag-
ging could be covered with layers of sheet lead, taped to form a
water-resistant surface that would withstand cleaning normal to
paper mills.
The lagging consisted of a resilient (and absorbing) layer of

3 inches of TIW fiberglass, covered by an impervious and heavy
layer of lead sheeting to serve as the mass element and to contain
the noise. In theory, a I-lb/sq ft lead sheeting, perfectly covering
the area, would give an attenuation of 24 dB at 500 Hz and 28 dB at
1000 Hz. From a practical standpoint of making a more impervious
continuous barrier fqr both noise and water from cleaning hoses,
two layers of lead were used, each layer taped for tightest joints
under the conditions of use.
Alternate materials could be considered: the absorbing layer

could be polyurethane foam, and the barrier layers could be lead-
filled vinyl. However, the fiberglass is much more resilient, and
thus is preferred.
In another paper mill, a third method was chosen: enclosure

for the operators. As the time actually spent in making the ad-
justments during the shift was small, the daily noise exposure in-
dex would be at or below the required 1.0 if l~ to 2 hours expo-
sure at 100 dBA were balanced with the remainder of the shift in
a protected area below 85-dBA levels.
Since this reduction would require only a minimum shelter for

15 dBA attenuation, a simple construction could be used consisting
of ~-inch plywood, two sides on a 2 x 4 inch framing, with one door
and a viewing window plus lighting. Due to paper mill conditions,
the enclosure was heated and air conditioned for worker comfort.
Record keeping could be done inside the enclosure.
137
Figure 5.9.2 shows the attenuation obtained inside the control
room; the noise level was reduced from 97 dBA outside (Figure 5.9.1)
to 77 dBA inside, well within compliance levels. Reducing the noise
levels further would require an enclosure with improved transmission
loss construction (such as concrete block).
A sketch was made following the lagging design parameters dis-

cussed above, requiring 3-inch Fiberglas grade TIW, plus two top
layers of 1/64-inch lead sheeting, all held to the surface by "stic-
klips" glued to the surface (normal method of acoustic application
in buildings). Each layer of the lead was waterproof taped to
protect the absorbing layer from water cleanup. Two layers were
used to get best waterproofing.
Shafts~ bearing boxes, and the drive motor were not lagged
because vibration readings did not justify the action, as noted
above.
The sketch shown in Figure 5.9.3 was used by an acoustic con-

tractor for estimating and installation purposes. Cost was $600
per Jordan, averaged over all six. Total enclosure cost would have
been at least $4,000 for each Jordan, based on a quotation from an
enclosure supplier; thus lagging achieved a net savings of $3,400
per machine.
For a single Jordan paper-pulp refiner, the noise level at the

operator station was reduced by 11 dBA, from 97 dBA to 86 dBA. With
all eight Jordan refiners similarly lagged and running at normal
production, the aisle noise level ranged from 88 to 91 dBA. Even
for unusually high production in the fastest mode, these same levels
ranged from 92 to 94 dBA. This lagging produced acceptable daily
noise exposures of less than unity.
Pitfalls
Paper mill conditions should be studied carefully because the

noise absorbing materials cannot be exposed to the wet conditions
of a paper mill. Maintenance methods were reviewed before design,
leading to a separation of the lagging at the front flange so
that it could be removed for maintenance.
Cover plates requiring occasional removal were not lagged, as

noted in Figure 5~9.3. A disadvantage is the lack of smoothness
of this surface--it makes lagging look like upholstering!
138
___I

180 355 710 1400 - 2800 - S600 - 11200
CD
"l:l
I 100
..J
W _ 9 7 dBA
>
W
..J
W
a:
::J 90
en
en
W
,,"
a:
Q.
0 ~
-"
Z
::J 80
0
en ,-" A weighted
I
0
z
«
CDI I
W
> 70
« I
~
()
0
-j-
60
I
I I
63 125 250 500

I
IC:JO 20aO 4000 8000
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.9.1 - Noise levels 1 meter from Jordan refiner
139
90 180 355 710 1<100 - 2800 - 5600 - 11200
CD
';' 100
--l
ill
>
ill
--l
ill
0: 90
:l
(f)
(f)
ill
0:
0..
o
~ 80
o
(f) _ 7 7 dBA
o
z
«
CD
W 70
>
«
f-
o ~
o
// A weighted
60
63 12~ 250 ~~o 2000 _000 8000
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.9.2 - Noise levels inside enclosure of control room for Jordans
140
separate front flange
c. o. clear
cover shaded area with 3-inch fiberglass TIW plus 2 plies of 1/64-inch
sheet lead. Tape for water resistant joint. Fasten stik-klips and caps.
studs clear
FIGURE 5.9.3 - Noise reduction lagging for Jordan shell
141
CASE HISTORY #10 - FOLDING CARTON PACKING STATIONS,
AIR HAMMER NOISE SOURCE
In the manufacture of folding cartons, the individual cartons

are cut, and the cut sheets are stacked by the cutting press on a
pallet. To deliver the multiple sheets from the press, the cartons
are held together with a nick or uncut portion. When stacked, the
individual cartons are separated by stripping with an air-driven
chisel which breaks the nicks and frees an entire stack. When no
additional operations are needed, these stacks are packed in cases
for shipment.
Air hammers/chisels produce noise that has not yet been elimi-
nated by equipment manufacturers. Currently available air hammer
mufflers do not reduce the noise to an acceptable level. The air
hammer operator therefore must wear ear protection. The problem in
this case was to protect other workers (packers) from the air
hammer noise. A typical production air hammer stripping and pack-
ing set-up is shown in Figure 5.10.1.
The production sequence for this operation is for the strip-

per to air hammer a stack of cartons (precounted by the cutting
press) and place them on the conveyor at Point C. The packer, at
the end of the conveyor E, prepares the case, packs the stacks of
cartons, seals, labels, and stacks the finished pack on a delivery
skid. Two packers are required to handle the output from one strip-
per. The stripper is actually using the air hammer about 50% of
his time, with the balance of the time used in stacking or preparing
the load. Thus he can get some relief from continuous use of his
ear muffs by hanging them around his neck while not actually using
the hammer. It is easier to sell the use of ear muffs when needed
if the operator can get some relief when muffs are not needed.
Measurements
As the specific wavelength requirement is minimum in this prob-

lem; no octave band readings were made; all dat~ were based on A and
C scale readings from an acceptable Type 2 sound level meter.
It was decided to protect the packers from the air hammer strip-
ping noise by using a barrier wall. The previous chapters described
the theory of barrier walls in detail, but a convenient rule of
thumb is that useful protection is afforded by the barrier wall
beyond 30 degrees into the acoustical shadow. Note that in Figure
142
5.10.1, the packers behind a wall 10-feet long and 6-feet high are
within this protected zone in both top view and side view of the
operation.
The barrier wall need be no better acoustically than the attenu-

ation afforded around the sides and top of the wall. The wall was
therefore fabricated using a 2 x 4-inch frame faced on both sides by
~-inch plywood for a simple sturdy barrier wall.
If there had been any reason to reduce noise reflections from

the noise source side, this side could have been faced with sound
absorbing acoustic materials, but it was not needed in this case.
We used the rule of thumb of aiming for the packer to be well

within the 30 degree line from the acoustic shadow line. Other means
of estimating the attenuation of barrier walls are covered by Beranek
in Noise and Vibration Control, page 178, and illustrated in Figure
5.10.2. The attenuation calculated for this barrier wall ranges from
10 to 15 dB, depending on the wavelength. This agreed with the meas-
ured attenuation of 7 to 12 dB and the noise reduction from the 92
to 97 dBA range to about the 85 dBA average measured at the packer's
ear level.
Cost
The barrier costs were:
~-inch plywood, 2 sides, 5 sheets, 4 x 8 160 sq ft $30.00
2 x 4 inch framing 60 ft 10.00
In-plant labor 60.00
Approximate total $100.00
Pitfalls
In this installation, we were fortunate not to have low ceilings,

which would have established a serious sound reflection problem and
defeated the barrier wall. Barrier walls will not give good results
in a highly reverberant, low-ceilinged room. If there had been a low
ceiling, useful level reduction would still have been possible by
adding sound absorbing material at the reflecting portion on the
ceiling (about 12 feet over the barrier wall and the noise source).
The amount of attenuation gained is easily estimated by using the
r~tio of absorption of new material to that of the existing ceiling.
Ceiling reflection is a major pitfall of the use of barrier walls
indoors. The design of the wall alone is based on free field
conditions.
143
stripper
c
/
--conveyor _ - --E
B
r~
0 barrier wall
packers
ceiling
C conveyor- - - E
B
packer ear level
stripper
-~
barrier wall
/
FIGURE 5.10.1 - Air hammer stripper and packer line
144
---- e-10feet
-------- - - - - -
4A ,ee\ __ - - -
... - ...
-- -- ---
O· .:.--
---
barrier wall receiver
source
2
N=X(A+ B-D) 250 Hz 500 Hz 1000 Hz 2000 Hz
N::; ~(5.5+ 10-14) A 4.8 2.4 1.2 0.6
N=~ N 0.625 1.25 2.5 5

dB 10 11 13 15
Source: Beranek, Noise and Vibration Control, page 178.

FIGURE 5.10.2 - Barrier wall theory
145
CASE HISTORY #11 - IN-LINE GRAVURE PRINTING AND CUTTING
PRESS :CLDING CARTONS
In the manufacture of folding cartons, one method is to print

the cartons in a web, using multiple gravure color stations and
feeding the printed web into a reciprocation cutting press.
The reciprocation cutting press, using. a rule die, cuts the

cartons and delivers cut cartons to a delivery belt. The rotary
printing operation was not noisy, but the cutting press noise from
the cutting head was in the range of 93 to 95 dBA at the normal
operator position. The take-off operators were far enough from the
noise source, so noise at their station was below 90 dBA.
Figure 5.11.1 shows the operator location, control station,

cutting head, and carton delivery. To reduce the noise of the
cutter head at the operator position, we decided on a barrier wall.
As access to the unit for job changes and maintenance was important,
the barrier wall was specified to be the lead-filled vinyl sound
stopper curtain material available on a made-to-order basis and
designed to be portable.
The curtain unit ordered was 7 feet high and 8 feet long with
a viewlng port 10 x 20 inches, since the attenuation required for
OSHA compliance was only about 5 dBA minimum.
The operator location noise at the control console was reduced

from the 93-95 dB range to an 86-87 range. The operator performed
inspection and adjustment at the cutter head for a few hours daily,
as required, but was still within the time exposure limits.
Total cost, using a lead-filled vinyl curtain at about $4/ft 2 ,

was about $300, including hanging fasteners, viewing window in cur-
tain, and pipe supports.
Pitfalls
To get any attenuation from barrier walls, the receiver must

be located with respect to the noise source so as to be beyond 30
degrees into the acoustical shadow line, as a rule of thumb. Note
that in the top view, Figure 5.11.1, the pressman operator is just
within this line. In Figure 5.11.2, showing over-the-wall vertical
plane limitations of this same rule of thumb, the pressman operator
146
is well within this limiting area. The curtain met the objective
since only a small attenuation of about 5-6 dBA was required and
the actual real attenuation was 7-8 dBA. More attenuation would
require a larger curtain.
A design pitfall in barrier walls is that if room conditions

are too reverberant and the ceiling is too low, the barrier wall is
bypassed. Low ceiling reflections can be overcome by adding an ab-
sorbent to the reflecting area of the ceiling over the barrier
wall.
In Case History #10, a relatively permanent wood construction

wall was used. This case required a different treatment because
regular access was required to the cutter head between the console
and the press. The free standing, easily movable curtain wall
provided both protection during operation and easy access to the
press for set up.
147
take-off \ drive
I
skids
Do delivery
cut head gravure printing - web
---------i
D° -------ef ----- >30 0
'\ ?/
§t;.. . ._--
..console
noise source
J
1- sound barrier curtain, 7 feet high, 8 feet long

pressman
FIGURE 5.11.1 - Top view of in-line gravure-cut press with sound barrier curtain
barrier - _.......;......--
____-=r ~ _-_-_-~.,....~ressman ear level
noi se source
FIGURE 5.11.2 - Side view of in-line gravure-cut press with sound barrier curtain
148
CASE HISTORY #12 - AIR SCPAP HANDLING DUCTS FOR
CORRUGATED CONTAINER INDUSTRY
In a corrugated box factory, slit side-trim is removed from

the conveyor by air. Trim blower fans with extra heavy blades cut
the trim while conveying it via ducts to bins and balers. The 12-
inch ducts are suspended from the ceiling about 10 feet from the
floor, crossing a 40-foot-long work room en route to the bins and
baler room.
Ear level noise levels were 93 dBA.
The trim is carried along in the ducts by the air, which nor-
mally moves about 6000 ft/min. The trim often strikes the duct
walls (mainly at bends), causing noise. Noise reduction was de-
sired to improve worker communication for operations under ducts
and to meet the requirements for minimal OSHA compliance.
Octave band data were collected because this case required

not only a reduction of a few decibels to comply. with minimum regu-
lations, but noise reduction for safety reasons, to control speech
interference. Octave band data are a truer measure of speech inter-
ference noise levels than a single number dBA reading. For the
minimum compliance data, the dBA reading would have been adequate.
The solution chosen was to wrap the problem duct locations

with 2 inches of mineral wool building insulation to furnish a
resilient and absorbing layer. Over this insulation were placed
two impervious layers of heavy tar paper, spirally wrapped with
50% overlap.
Noise was reduced considerably in the problem area: levels

changed from 93 dBA to 72 dBA, a reduction of 21 dB. To the ear,
the noise could hardly be heard above other noise. The octave band
comparison is shown in Figure 5.12.1. Although the standard mate-
rials used were very economical, special acoustic absorber pipe
coverings with lead-filled vinyl sheeting could also have been
used and may have given even more attenuation. This was not needed
here.
149
We do not have detailed costs for this case history. How-
pver, since the materials are inexpensive, the major cost must
have been labor. The job was probably done for less than $200.
If less attenuation had been required for OSHA compliance

only, the sheet metal ducts could have been damped and transmission
loss improved by gluing 1/32-inch sheet lead to the duct outer sur-
face. Comparison with the experience at other installations indi-
cated that a 5-dB attenuation would probably have been attained.
(See Case History #8 for other methods.)
Pitfalls
Use the most economical methods to attain the attenuation re-

quired. The building insulation plus roofing paper used here is a
very economical solution.
Note that a large overlap was used; lack of overlap on any

wrapping will cause leaks and redu~e attenuation.
150
90 180 355 710 1.-00 - 2800 - 5600 - "200
rn
""0
...J 100
w
>
W
.....J
w 93 dBA
a:
~ 90
if)
-
if)
w
a:
0..
0
z
~
before -
- - -
"
0if) 80
0
z
<:
rn
w
I
72 dBA
> 70
<:
f-
a
0
60 I----+---+----+---+----+---+_-
63 115 150 500

I
leGO 1020 ~OOO 8000
100 1000 10000
FREQUENCY IN HERTZ
FIGURE 5.12.1 - Noise levels in scrap duct for corrugated box industry. before and after covering
151
CASE HISTORY #13 - PAPER MACHINE, WET END
Th~ major noise sources of the wet end of this paper machine
were the couch roll suction air movement, the pumps, and the
whipper roll. The whipper roll supplies a beating action on the
felt of the paper machine to provide continual web felt cleaning.
Noise in the aisle at the wet end is shown in the octave band
analysis of Figure 5.13.1. The level is 92 to 94 dBA in the oper-
ator aisle. Higher readings of more than 100 dBA were obtained
close to the couch roll. See Figure 5.13.2 for a sketch of the
area.
Paper machine manufacturers have developed a quieter couch

roll in which the suction holes in the drum are in a staggered
rather than a regular pattern. However, the replacement cost of a
couch roll is high, and it will probably be used only on mill expan-
sion projects or new mill construction.
An alternative method to reduce the operator noise exposure

was by constructing a personnel booth to house the operator, plus
the operating controls, during most of the operating shift. The
wet end paper machine operator spent an hour or less making couch
roll adjustments during a typical operating day. If the balance of
each day were exposure in the mill operating aisle at 92 to 94 dBA,
the resulting exposure would be over the dose limits of unity.
However, if the operator spent the 1 hour at 100 dBA (couch roll
adjustments), 2 hours on general observations near machine at 92
dBA, and the balance of the shift in areas under 85 dBA, including
a personnel booth, his daily noise dose would be:
1 hour actual 2 hours actual
2 hours allowed +
6 hours allowed = 5/6 = 0.83
(100 dBA) (92 dBA)
Since this dose is less than 1.0, it is within normal exposure

compliance.
The recommendation for the wet end of the machine (couch roll
and whipper noise exposure) was to provide an operator enclosure
with operating controls and instruments, and with viewing windows
to observe machine operation.
152
Calculations indicated that the required 15 dB attenuation
could be attained with a simple structure consisting of 2 x 4-inch
framing with ~-inch plywood walls inside and out, plus one solid
door and two windows 3 x 5 foot each, double glazed. Ceiling and
upper half of walls were covered with acoustic tile to reduce rever-
berant noise. The room was provided with light, heat, and air condi-
tioning for worker comfort. In-plant construction cost was $2,500.
Results achieved by the enclosure are shown in Figure 5.13.3.

Inside noise level was reduced to 75 dBA from 92 to 94 dBA outside.
Greater attenuation can be obtained by purchasing special

acoustic shelters or by using more elaborate (from acoustic stand-
point) construction such as concrete block walls, double windows,
or interior sound absorption.
Pitfalls
Most of the difficulties to be avoided are non-acoustical.

It is essential that the operator has no interference with visual
monitoring of machine operation. This consid"eration fixes the booth
location and window placement.
153
ISO 355 710 1400 - 2800 - 5600 - 11200
m
"0
..J 100
w
>
W
..J
94 dBA
w
a:
::>
if) 90
if)
w ~
a:
a. /
0 ~
z
::>
0 80
if) ~
0 ~
Z
<t:
m
W ,- A weighted
>
<t: 70
l-
u I
0 I
60
63 125 250 500 I coo 20~O

I
4000 8000
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.13.1 - Noise levels at wet end aisle of paper machine
154
control area
major noi se source enclosed 10' X 14'
,,/ 100 dBA )94 dBA 92 dB
i couch roll 7
: WET END
,..- ,
~ OOOOOOOO:control :
~ DDDGcBo~DDL----' ~
top hydropu I per
bottom hydropulper
dock
Ii Iler hydropulper
FIGURE 5.13.2 - Paper mill - wet end
155
90 ISO 355 710 I~OO - 2800 - 5600 - 11200
en
"0
I 100
..J
W
>
W
..J
W
a: 90
::J
C/)
C/)
W
a:
0-
0
Z
::J 80
0
C/)
0
Z 75 dBA
«
en
I
W 70
>
«
I-
0
0
60
A weighted ±--
63
I125
2
250 ,00
5
1COO
I
2000 ~OOO SOOO
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.13.3 - Noise levels inside control booth at wet end of paper machine
156
CASE HISTORY #14 - PUNCH PRESS NOISE
Punch presses in use in this shop were Summit, Bliss Diamond,

and Benchmaster. Within the room were four large presses and four
small punches. One of the Summit presses was chosen as represen-
tative of the large press group, and the Benchmaster was chosen as
representative of the small press group. The general room layout
is shown in Figure 5.14.1.
Octave band measurements were made of the ambient noise levels

when all the presses and the nearby furnaces were shut down. Read-
ings were taken near the central supervisor's desk. The overall
A-weighted level was a very low 58 dBA, as shown in Figure 5.14.2.
This indicated that there were no other serious noise sources. Also
noted was the difference in noise level with and without the furnaces.
Figure 5.14.3 shows that, with the furnace on, the level increased
to 69 dBA, still quite low for most industrial noise levels. Thus,
the furnace was also eliminated as an irritant noise source.
Figure 5.14.4 is an octave band analysis taken from the center

desk with two Summit presses, two Bliss punches, and one Benchmaster
in operation. The level at the desk is 97 dBA, a definite over-
exposure condition.
The Summit punch, Location I in Figure 5.14.1, was chosen as

typical of a large press. Operator levels, shown in Figure 5.14.5
and 5.14.6, were 106 dBA during operating cycle and 90 dBA during
preparation (punch I off, reaching for a new sheet, etc.). At 106
dBA, the permitted exposure time is 0.87 hr. The octave band analy-
sis showed that major noise contributions to the A-level came from
the 1000-Hz and higher bands.
Figure 5.14.6 shows the spectrum of noise from operation with

nothing in the die. Although a reduction was noted in the 500-Hz
band and a small reduction in the 250-Hz and 1000-Hz bands, the
2000-8000-Hz bands, which were main contributors to A-level noise,
remained the same as with the full operation. The 1000-8000-Hz
bands that were the chief contributors to the A-level were entirely
due to the effect of air exhaust noise from jets for removing parts
and pushing them into the collection chute. For these higher fre-
quencies and short wavelengths, barriers are efficient. Close7in
diagnostic measurement~ were made behind the press, but no new
noise sources were noted except the directionality of some of the
air ejection noise.
157
The reduction sought was from 103 dBA to 86 dBA, with 90 dBA
acceptable. This required reductions of:
13 dB in 500-Hz band
For a separate study of a typical small press, the Benchrnaster

(punch VII) was chosen. The operator's position octave band analy-
sis in Figure 5.14.7 shows somewhat less noise than the large press;
it has the same general configuration and air jet noise source.
Figure 5.14.8 shows the noise levels with no stock in the press, and
Figure 5.14.9, the press in punching operation with no stock and no
air ejection. Again data were very similar to those for the larger
press.
Our recommendations were: reduce air noise along path by

installing a barrier between noise source and operatot; and reduce
noise from air ejection at the source. The latter was considered
first.
Noise caused by high air velocity can be reduced by decreasing

the linear flow velocity by increasing the nozzle opening, for same
air mass flow. If the diameter of the nozzle is doubled, in a con-
stant volume velocity system, flow velocity is reduced to one-fourth,
and noise level is reduced nearly 30 dB (noise of air jet varies
approximately as fifth power of velocity). However, thrust would
also be reduced, to one-fourth of original value. For proper ejec-
tion, the nozzle must be aimed more accurately and more efficiently
toward the target. If the distance from the nozzle to the target
were reduced 50%, a 30% velocity reduction would give the same
thrust. Experiments must be conducted to determine the maximum
thrust required for minimum noise.
A barrier between source and operator can add to the attenuation

obtained. The barrier should be box shaped around the die (with
far side and bottom missing). This barrier replaces the present
guard, and handles both mechanical and acoustical guard functions.
Materials suggested include ~-inch plywood, ~-inch Plexiglas, and
~-inch Lexan, made with airtight corner joints. Noise absorbent
material, Mylar faced for dirt and oil protection, was added inside
the box; it must be kept clean during normal operations.
158
Control Approaches Chosen
Based on suggested possible methods of nozzle construction, a

quiet nozzle cover was made. The design of this nozzle is shown
in Figure 5.14.10. AJr pressure, controlled by a reducing valve,
was reduced to the minimum to do the ejection job. (Low-noise air
jets are also available commercially.)
A sketch of the barrier is shown in Figure 5.14.11. To afford

visual access, the material chosen for the barrier was ~-inch Plexi-
glas. The three-sided barrier was locally designed, aiming to have
minimum leakage at bottom of barrier (toward the operator).
For absorption, I-inch acoustical (fully reticulated) polyure-

thane, with Mylar film covering for ease of cleaning, was glued to
the inside surface, leaving a mInimum uncovered portion for oper-
ator vi~wing of punch action.
Accurate costs were not available for this in-plant effort;

however, the materials were less than $100 and labor was estimated
at $250.
After experiments with reduced jet velocity and with the bar-
riers described, t~e following noise levels were attained:
Large punch press reduced from 106 dBA to 85 dBA.

Small punch press reduced from 99.5 dBA to 82.5 dBA.
Pitfalls
The major pitfall for barriers will be to see that it is used.

Also when used, the bottom opening or noise leak toward the oper-
ator should be kept at a minimum. Another pitfall to continued
efficiency will be allowing the Mylar-covered noise absorbent to
become dirt and grease laden; periodic cleaning is needed.
A pitfall associated with air volume reduction is the tendency

of operators to increase pressure or remove the nozzle.
Attenuation will depend on the success of air velocity reduc-

tion in maintaining the needed thrust for ejection in conjunction
with noise reduction of barrier. Unless these experiments were
done with operator involvement, they may not accept the alterations.
If a mechanical method could be developed to replace air jet

part ejection, this would be the best alternative.
159
V Roussell
r---,
D
V I Benchmaster
D
VII Benchmaster
I I
I I
IL... _ _ -.J
center desk furnace
D IV Bliss
D III Bliss
II Diamond I Summit VIII Summit
FIGURE 5.14.1 " Layout of punch press room
160
180 355 710
80
Ld 70
>
W
....J
W
II
::>
~ 60
w
II 58 dBA
0-
o
Z
::>
o
C/) 50
o
Z
-<
CO I
W I
> I
~ 40
,,
U A weighted
o
,
30 -1..._
+
20
5
63 125 250 500 IC"O
1
]O~O 4000 8(1C()
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.14.2 - Ambient noise level with furnace and punch presses off (microphone 1.5 meter
above floor, directly above desk chair)
161
90 180 355 710 1400 - 2BOO - 5600 - 11200
80
m
"
...J
W
~ 70 69 dBA
...J
W
a:
::>
(j)
(j)
w
a: 60
11.
o
,
Z
::>
o
(j)
I -
~ 50
«
m I
w I A weighted
«> I
f-
()
40
I
o
$
30 --I-
f
-t-
63 125 250 :JO

t-
--+ .
1['0 ?C')O '000 80GO
20
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.14.3 - Ambient noise level with furnace on, punch presses off (microphone 1.5 meter
above floor. directly above desk chair)
162
90 180 355 710 lAOQ - 2800 - 5600 - 11200
97 dBA
90
IT)
"'0
...J
UJ 80
>
UJ
...J
W
a:
::J
CI)
CI) 70
UJ
a: I
Q. I
0
z I
::J
0CI) I
60
0
Z
«
IT)
I'A weighted ::
w
> I
« 50
f-
<-)
0
40
I
I
I
63 125 250 500 20~J AOOQ 8000
30
5
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.14.4 - Ambient noise level with furnace on, two Summit, two Bliss, and one Benchmaster
presses in operation (microphone 1.5 meter above fI~or, directly above desk chair)
163
lBO J55 710 1400 - 2BOO - 5600 - 11200
dB fast, I inear scale

110
CD dBA fast
"0
dBA slow
-.J
w
>
w
-.J 100
W
a:
~
(j)
(j)
w.
a: 90
0..
0
Z
~
0(j)
0
Z 80
« I
CD
w
>
«
I'A weighted'
I-
0
0 70
I
I . I
60 -~
-+
I
I,
50
I
I
125 250 500
5
---.-
IC10
I
2CCO 4000
5
Booo
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.14.5 - Operator's exposure at punch I: stamping stock (air exhaust on, noise level
a t operator's ear)
164
· OCTAVE PASS BANDS IN HERTZ
180 355 710 1400 - 2BOO - 5600 - '1200
110
106 dBA
m
"0
--I 100
w
>
W
--I
W
a: .,
::J
(f) 90 -' 90 dBA
(f)
w punch I off
a:
Q..
0 punch lion
Z
::J
0(f) 60
0
z
«
m
w
> 70
«
f-
0
0
I
60 t----+--.......---t---+----+---t----+----"1f-----+----t
I
I
I
I
63 125
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.14.6 - Operator's exposure at punch I: stamping without stock (air exhaust on, noise
level at operator's ear)
165
90 180 355 710 10100 - 2800 - 5600 - 11200
dBA level from punches I, III, and VII
100
dBA level with VII on only _-!-_ _--I~--_+--_ _+_
--I
W 90
>
W
--I
W
a::
:J dBA level from I and III only
(f)
[3 80
a::
0-
o
Z
:J
G
(f) 70
o
Z
<t:
OJ
W
> IA weighted .:.
~ 60
u
o I
I
I
I
50
-t-
63 , ~5 250 500 lCOO ~OCO 4000 8000
40
2 5
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.14.7 • Operator's exposure at punch VII
166
90 180 355 710 1‫ןס‬OO - 2800 - 5600 - 11200
100
dBA
co
-0
90
-l
UJ
>
UJ
-l
UJ
a: 80
~
en
en
W
a:
0..
o
~ 70
oen
I
o I
Z
<{
CO
I
UJ 60 I - - - - - t - - - - - i - l - ' A-wei ghted -+----+---+----+----if------l
>
,
<{
f-
U
o
I
50
I
I
40 L--r-~__T'_:.,....T""T"'T'"""---12i.:.5--,.......;2:.i.50:.....,..-,.__~50~O-r-..,..,,....,:.:1O+-OO:....-_--...:2-+0C:.:..O_-.--...:.O~OO~__r.,;:.8+00..,O- r-----J
5
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.14.8 - Operator's exposure at punch VII, operating without work piece
167
180 J55 710 1400 '800 ~ - 11'00
100 I---_t_--__t--~+_--_+---+_--_t_--__t---_;_--_;
90
...J
W
>
W
...J
W dBA
a:
~ 80
en
en
W
a:
Q.
o
Z
~
oen 70 ~-~~~=::t:...--_=ji!~-+_-----.:+_----=:J:_----=~~~=---~
o /
Z
<{
ro /
w I
>
<{
60 I----+---HL-.----L----+----+---+-----+---+------i
f-
()
o
I
50 1----+---.j~__+---+---_+---+----+-----4---+---~
I
40
,
L..-..,.....Lr-T'6J......"T'""'1......_ I.... 50
' 5_ _....-_'..
"--;-
-r_.,....-...,501-0..........-.,....,.l-+r~_o 000-rr-_......I
'0+0_0_-r-_4.....0l-00""'"T""'"T--r8.....
oS oS
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.14.9 - Operator's exposure at punch VII, operating without stock and without air exhaust
168
1/2-inch cap 118-inch diamater
r-"\
/1-- ----; I
/
,.--f----
... / I I .J-_
I I
I I I
inlet
I I
I
I
'----f--- - - _
I
_ _ _ .....J
I )
r
~-------f '--
out let
FIGURE 5.14.10 - Design of nozzle
,...
-, die
~'/4-;nCh
0 V 1- i nch
Plexiglas
polyurethane foam and Mylar
I I
close fit
FIGURE 5.14.11 - Sketch of Plexiglas barrier
169
CASE HISTORY #15 - STRAIGHT AND CUT MACHINES
, .
The straight and cut machine straightens heavy gauge wire in

an in-feed to cut-off unit set to cut repeat lengths, resulting in
noise levels of 92 dBA at operator position.
Figures 5.15.1 and 5.15.2 are close-in octave band analyses of

the diagnostic measurements made in front of the clutch mechanism.
In Figure 5.15.1, curve A shows peak cutting levels, and curve B is
the slow response of the same cutting levels (wide separation indi-
cates impact noise). Curve C is idling, noncutting machine noise
level. The differences indicate dominance of the total spectrum by
the cutting noise. In Figure 5.15.2, curve D is the peak response
and curve E is the slow response. Curves D and E exceed curves A
and B, indicating some directionality of the cutting noise.
Figures 5.15.3 and 5.15.4 are octave band analyses made at the
operator position. Most of the operator time is represented by
Figure 5.15.3, with the cutting cycle noise level at 92 dBA (idling
cycle at only 83 dBA), indicating the dominant noise source of the
clutch cutter mechanism is the same form as in the close-in diag-
nostic measurements. A-weighted levels of these readings show that
the required attenuation is 9 dB at 1000 Hz, 7 dB at 2000, and 10
dB at 8000 Hz to reduce noise level at the operator position to
85 dBA.
Control Approaches Chosen
Having determined from management discussions that the machine

should not be changed or minor redesign attempted, we recommended
attenuation by a barrier wall that would block the sound path from
the cutting assembly to the operator.
Barrier materials for obtaining the required attenuation were

1/4-inch plywood, with 1/8- to 1/4-inch Plexiglas for viewing ports
where necessary. The barrier wall was extended 26 inches past the
extremities of the area encompassed by the cutter and was close to
the cutter, about 6 to 8 inches away. The barrier was hung in place,
supported by chain from overhead. In addition, an absorbent layer
was hooked to the barrier on both sides. To prevent clogging of
absorbent, the I-inch polyurethane foam absorbent was supplied with
Mylar facing. See Figure 5.15.5.
Normally, the noise absorbent for barriers is used only on the

machine noise source side. In this case, however, noise absorbent
170
was used on the operator side of the barrier as well to reduce
sound field build-up in the space between barriers. With the bar-
rier close to the cutter, the operator would be within the safe
sound shadow area--the area beyond a line at least 30 degrees from
the edge of the acoustical shadow line.
As the barrier was built in-plant, no actual costs are avail-

able but material costs are estimated at about $100.
The cutting cycle noise levels at the operator location were

reduced from 92 dBA to 85 dBA, a 7-dBA reduction. Idle cyc,le noise
level was reduced from 83 dBA to 76 dBA.
Pitfalls
Barriers are easy to remove by the operator for many reasons,

real and imaginary, and use must be maintained by supervision. Loca-
tion of an effective portable barrier must be standardized so that
the barrier is not by-passed. Barriers can be by-passed by noise
reflections from a low ceiling. If this problem had existed in this
case, a section of the ceiling above and about 4 feet on each side
of the barrier could have been treated with an absorbing material.
171
90 180 J55 710 1400 - 2800 - 5600 - 11200
130
120
A peak cutting levels
B slow response measure of cutting levels
C idling, noncutting level
co
"0 110
...J
w
>
W
...J
~ 100
::J
en
en
W
a:
0-
o
Z 90
::J
oen
B
o
Z
<!
CO
w 80
>
<!
I-
()
o
--+-
70
A-weighted
60
250
i500
I
lC'1O 2000 4000 8000
100 1000 10000
FREQUENCY IN HERTZ
FIGURE 5.15.1 - Straight and cut machine: close-in measure near west side of clutch cutter
mechanism (1.2 meter above floor, 0.5 meter from cutter)
172
90 180 355 710 1.000 - 2800 - ~ - 11200
130
'120
o peak cutting levels

[l) E slow response measure of cutting levels
"'C
,110
...J
w
>
W
...J
W
gj 100
en
en
W
cr:
11.
o
Z
::J 90
oen
o
Z
<t:
[l)
~ 80
<t:
l-
t)
o
-I-
70
60 500 1000
i
200,) 4000
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.15.2 - Straight and cut machine: close-in measure near east side of clutch cutter mechanism
(1.2 meter above floor, 0.5 meter from cutter)
173
180 355 710 1400· - 2800 - S600 - 11200
-+-_-- cutting level dBA

90
l!l
"0
80
...J
W
>
W
...J
W
([
:l 70
rn
rn
W
([ /
a.. /
0 /
Z /
:l 60
0
rn
0
/+
z I A weighted -
«
OJ I
I
w I
> 50
«
I-
0
0
40
63 125 250 500 1000 2000 4000 8000

30
100 1000 10006

FREQUENCY IN HERTZ
FIGURE 5.15.3 - Straight and cut machine: operator's near field exposure
174
180 355 710 1.1()() - 2800 - 5600 - 11200
100
90
cutting level dBA
I t
CD 80
"'0
-l
W
>
W
-l 70
w
a:
~
C/)
C/)
w
a:
0-
o 60
I
Z
~
o
C/)
/
o /
z
« /
CD 50
w
>
«
l-
t)
o
40
30
5
63 125
2
250 500
5
lC~O
1000
I
20CO 4000 8000
10000
100
FREQUENCY IN HERTZ
FIGURE 5.15.4 - Straight and cut machine: operator's far field exposure
175
\\
,if ceiling T T
right side
view
r noise angle~ / chain !
suspended'\j"'--+~--------"~
front
view
/
barri er
U--
IT 5 feet --tHI
operator ear level
1
. ~-
-
floor
FIGURE 5.15.5 - Barrier wall for straight and cut machine
176
CASE HISTORY #16 - CUT-PUNCH PRESS IN METAL
FABRICATING PLANT
This punch press had been modified to produce metal stampings

out to a predetermined size. This machine was the first stage of
a stamping operation in which the metal was sized and roughly shaped.
In two following stages, each part was finished.
Figure 5.16.1 gives the octave band analysis of the operator

exposure, which is 102 dBA while punching and 88 dBA during idling.
Figure 5.16.2 shows close-in octave band data for gear noise, illus-
trating the continuous nonpunching noise source in the gear mechanism.
Figure 5.16.3 shows close-in measurements of the dog and flywheel
noise. Figure 5.16.4 shows similar close-in measurements of noise
from piston-collar impact on the air cylinder.
A reduction of gear mechanism noise 7 dB at 500 HZ, 10 dB at

1000 HZ, and 9 dB at 2000 Hz would reduce the idling, nonpunching
level from 88 dBA to 81 dBA. A reduction of dog-flywheel and piston-
collar impact by 17 dB at 500 Hz, 21 dB at 1000 Hz, and 25 dB at
2000 Hz would reduce the punching level from 102 dBA to about 85 dBA.
As machine change was not practical, changes had to be made in

the noise transmission path to the operator on two of these noise
sources. The other source, piston-collar impact on the air cylinder,
was modified at the source by adding washers made from Unisorb Type D
pad between the piston stop and the collar to reduce metal-to-metal
impact noise.
The gear noise and dog-flywheel impact noises were attenuated

by constructing an extended barrier about these noise sources. To ob-
tain the attenuation required, l-inch plywood was used. The enclo-
sure was attached to the right side of the press (as the operator
looks at press) and extended upward to the top of the press, downward
to operator chest level, and outward several inches past the flywheel
guard. The top, bottom, and right-hand edges had a small 6-inch
extension at the barrier extending 90 degrees away from the operator,
as shown in Figure 5.16.5.
An absorbent was added to both sides of enclosure using Mylar-

covered l-inch acoustical foam absorbent available from several
177
suppliers. The joint between the enclosure and the right-hand side
of the press was sealed to p:_event noise leakage; a 2-inch~wide
strip of closed cell foam weatherstripping was specified.
Normally, absorbing material is used only on the noise source

side of a barrier wall; however, if other noise sources might re-
flect from the barrier wall to the operator, absorbing materials on
the operator side will reduce this noise component.
Sound pressure levels during idling were reduced from 88 dBA

to 81 dBA. Punch operational sound pressure levels were reduced
from 102 dBA to 88 dBA, thus bringing the entire operation into
compliance.
Though not recorded, costs are estimated at less than $200 for
plywood, polyurethane foam, and the labor for attachment.
178
180 355 710 10100 - 2800 - 5600 - 11200
110
punching dBA
100
m
"0
..J 90
w
>
W
.J
W
([
:::l 80
C/)
C/)
w
([
a..
0
z
:::l
0 70
C/)
0
Z
<t:
m
w
> 60
<t:
I-
a I I
0 I
I
50 1------f~---+---+_:_--___4---_+---J----l_--_+---_l
40
03 125
2
250 5ao
!
- --t-
IC~O
I
2000 '000
--+-
8000
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.16.1 - Cut-punch press, operator's exposure. A, idling; B, punching.
179
180 355 710 1400 - 2800 - 5600 - 11200
110
100
dBA
ro
"0
..J
90
w
>
W
..J
W
a:
:J
(/)
80
(/)
w
a:
0-
0
Z
:J 70
0
(/)
I
0
z
«
m A-weighted
w
> 60
« I
l-
()
0
I
I
50
40
63 125 250 500 10ao
I
2000 ~ooo
i
8000
100 1000 10000
FREQUENCY IN HERTZ
FIGURE 5.16.2 - Cut-punch press, close-in diagnostic data, 14 cm from gears
180
• 180
9l) 355 710 UOO - '2800 - 5600 - 11200
140 r--"';"-~-~--r----,---.----r-----r---r----r----,
130
rn
"1:l
...J 120
W
>
W
...J
W
a:
~
en 110
en
ill
a:
a..
0
z
~
0
en 100
0
Z
«
rn
w
> 90
«
f-
a
0
BO
63 125 250 500 lGOO 2000 AOOO 8000

70
2 5
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.16.3 - Cut-punch press, close-in diagnostic data, 5 cm from the dog-flywheel interface
181
., 90 180 JSS 710 l.wo - 2800 - 5600 - 11200
140 _....;;:.--..--;,;.........,-....;.;.:.-~......;.;;";"""....,...-;...":",.--r-......;.;;";,,,,,,...,.-----.:.:..:..:...--,_.:..:..:..._..,...-.......;.,....,
130 1---+----+----1---+----+----+---1---+----1
OJ
"0
120
...J
w
>
W
...J
W
a: 110
::l
C/)
C/)
W
a:
a..
0
z
::l 100
0
C/)
0
z
«
OJ
w
> 90
«
I-
U
0
80
I
-+-
70 L--r-....,........,:.6,;..JT""'T-r-,............
12_S_--,.......;2..... ,......,....,...,.'C-t.'_O
50....,.._..,-..:..501-0_' 20+D_O_-r-_400t-0---r--r-,.8-10r-OOT""'T"_--J
S S
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.16.4 - Cut-punch press, close-in diagnostic data, 5 em from piston-collar impact
on air cylinder
182
noj se sourcp
J
/
top view
/
c:============::::!.II- barri er
o operator
no i se sources
hanging barrier ~
operator ear level
side view
FIGURE 5.16.5 - Sketch 0' hanging barrier for cut-punch press
183
CASE HISTORY #17 - PARTS CONVEYING CHUTE
This case was taken from published data*, because of the im-
portance of illustrating the method for other applications.
Chutes for conveying small parts can radiate much noise from
the impact of parts on the sheet metal of the chute. The noise
(for a given part) can be reduced by keeping to a minimum the dis-
tance the part must fall to the chute. For reducing the remaining
noise, the chute must be stiffened and damped.
Constrained layer damping is used, in which the treatment can

be placed on either the parts side or the underside of the chute.
If placed on the parts side, the metal layer should be wear-resis-
tant to the impacting parts. In this example, 30-caliber cartridge
cases were carried in the chute shown in Figure 5.17.1. The bottom
of the chute was 14-gauge steel, which was lined with 0.035-inch
cardboard and then covered with a wear plate of 20-gauge galvanized
steel. Rubber deflector plates were positioned to funnel parts to
the center of the chute, so that they would not hit the untreated
sides of the chute.
Figure 5.17.2 shows the spectra measured 3 feet to one side of

the chute. The level has been reduced from 88 dBA to 78 dBA, a
decrease of 10 dB. Greater reduction could have been obtained if
multiple layers of thinner cardboard were used (in solid contact
with the cover sheet). Still better would be replacement of the
cardboard by commercially available damping materials specifically
formulated for constrained layer use.
Pitfalls
Much noise still comes out of the top of the conveyor. A cover
over it, lined with absorbent, should reduce the noise an additional
5 to 10 dB. Prior to any noise control effort, the relative amounts
of noise from top and bottom shGuld be determined. Ordinarily damp-
ing is always to be used.
* Cudworth, A. L., Field and Laboratory Example of

Industrial Noise Control. Noise ControlS (1): 39, 1959.
184
14 gauge steel chute
?O gauge galvanized steel
0.035 inch cardboard
FIGURE 5.17.1 - Chute for conveying cartrtdge cases
185
110
CD 100
"0
...J
W
>
w 90
...J
w
a:
:>
en
: /~
en
W
a:
a..
0 70
-.....
untreated __
/
V./ I""""" - f' '\
-..... "" -' (\
Z
:>
0
en
0 sandwich on bottom O f /
z chute - rubber on
«
CD deflection plates
w
>
« 50
l-
t)
0
40
overall 20 - 75 - 150 - 300 - 600 - 1,200 - 2,400 - 4,800-
10,000
FREQUENCY BAND - CYCLES PER SECOND
FIGURE 5.17.2 - Noise spectra measured 3 feet from chute
186
CASE HISTORY #18 - NAIL MAKING MACHINE*
A nail making machine was operating under conditions causing

severe impacts. The vibration was solidly transmitted to a weak
concrete floor, which radiated considerable noise. There were 10
machines, operating at 300 strokes per minute. Operator noise level
was 103.5 dBA.
An octave band analysis was made at the operator position and

is shown in Figure 5.18.1.
It was decided to use vibration isolating mounts to reduce

floor radiated noise. Because of the repeated shock situation, se-
lection at the isolator followed these rules:
(1) The natural period of isolator plus machine should be

much greater than the shock pulse duration (10 msec) .
(2) The natural period of isolator plus machine should be

less than the time between pulses (200 msec).
Elastomer type isolators were used, having a static deflection

of 0.1 inch under machine load. This corresponds to a natural
period of 100 msec, thus fulfilling the design conditions.
Figure 5.18.2 shows octave band spectra at the operator's

position after all machines had been vibration isolated. The A-
levels have been reduced about 8.5 dB to 95 dBA. This is still in
excess of permitted levels, and additional noise control is needed.
Pitfalls
To maintain the isolation, maintenance people should be warned

not to short-circuit the isolators by any solid connection from
machine to floor. This short-circuiting can also occur when dirt
and grease are allowed to build up around the pods.
* From Crocker, M. J., and J. F. Hamilton, Vibration Isolation for

Machine Noise Reduction. Sound and Vibration 5 (11): 30, 1971
187
Additional Noise Reduction
As a reduction to a noise level of 95 dBA is not considered

satisfactory for full day operator exposure, additional noise reduc-
tion could be obtained by the design of a barrier between the major
noise source in the machine and the operator. Depending on the
needs for vision through the barrier, plywood, lead-filled vinyl
curtain, or Plexiglas could be used. Such a barrier should yield a
reduction of 5 to 8 dBA at the operator position. (For calculated
design parameters, see Case History #6 and for rule of thumb para-
meters, see Case History #10.) This noise reduction should result
in compliance with 87 to 90 dBA.
Where there is a series of machines, additional reduction of

several decibels could be obtained by added room absorption, either
in the form of spray-on acoustic absorbent on ceilings and walls or
in the form of hanging absorbent baffles from the ceiling. (See
Chapter 3 on Room Acoustics.)
188
90 180 355 710 lMJO - 2800 - 5600 - 11200
110
[]) 103.5 dBA

"C
...J 100
UJ
>
UJ
...J
W
a:
::J
C/)
90
C/)
w
a:
C-
O
Z
::J
0C/) 80
0 /
z
~
[])
I
UJ
>
~
70
I-
U
0
63 125 250 500 lC00 20CO 4000 8000
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.18.1 - Nail making machine: operator position noise levels
189
oM 90 180 355 710 1400 - 2800 - 5600 - 11~
'----""'T""--""""'-- --,--....;...--r-----r---,-.....;.,,-...,----,-----,
CD
"0
~ 100
ill
>
ill
~
95 dBA
ill
a:
~ 90
en
ill
a:
0-
o
z
6en 80
o
z
«
CD,
, ,,
ill
«> 70
~ A weighted
()
o I
I
I
63 125 250 500 lC'lO 2000 AOOO 8000
5 5 5'
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5,18.2 - Nail making machine: operator position noise levels after vibration isolators instal
190
CASE HISTORY #19 - WOOD PLANER*
Wood planers use a high-speed rotating cutter head to produce

lumber with a finished surface. Noise levels to the operator are
high.
There are apparently many noise sources for investigation:
(1) The board, excited by cutter knife impacts.
(2) The heavy structure under the cutter head, excited

by vibration transmitted through the board.
(3) Modulation of air flow by cutter knife chopping at

the chip collector air stream.
(4) Motor windage, hum.
(5) Dust collector blower, vibration noise.
(6) Machine surfaces excited by impacts.
Measurements and Analysis and Control
The octave band operator levels are shown in Figure 5.19.1.

Analysis resulted in the following possibilities for control of
planer noise:
(1) Restrain the board from vibrating. Feed belts on

both sides can be used with considerable backup
mass and pressure. This would require a radical
machine design change.
(2) Contact the board by means that add damping, to

reduce resonant vibration. If this is done as
an add-on, it must occur beyond the feed and
delivery ports of the planer. Thus it would be
helpful only for long lengths of board.
(3) Use a helical knife cutter head, which will also

reduce idling noise. A helix angle larger than is
commonly available would be desirable.
* Stewart, J. S., and F. D. Hart, Analysis and Control of Wood

Planer Noise. Sound and Vibration 6 (3): 24, 1972.
191
(4) Enclose the planer and board. This is a brute
force method that depends for its success on con-
trolling the amount of sound that escapes from
the feed and delivery areas; most of the sound
contributing to A-level noise is between 500
and 5000 Hz. This increases the enclosure design
problem, because enclosures are more effective
at high frequencies. (See Case History #24.)
Results achieved by the helical knife cutter head are shown in

Figure 5.19.2: reduction from 106 dBA to 93 dBA. Figure 5.19.3
shows the operator noise level related to length of board planed,
comparing the helical knife cutter with the straight knife cutter.
The helical knife is by far the quieter.
Pitfalls
To meet OSHA operator levels for full day operation, the plant
would need to make further noise level reduction, perhaps by the
design of a total enclosure with an acoustic lined tunnel for the
in-feed and out-feed.'This should not be tried until it has indeed
been determined that the openings are the chief sources. In many
mills, however, the planer is not operated on a full-time basis,
thus allowing a higher noise level for the shorter time period that
an operator is present. At 93 dBA, 5.3 hours is permitted.
192
..J
W
>
W
..J
W
a:
::>
C/)
C/)
w
a:
0-
o
z
::>
o
C/)
o
z
«
en
w
>
«
I-
o
o
o
a:
I
I-
FIGURE 5.19.1 - Noise levels for wood planer - straight Gutter
193
co
"0
-J
W
>
W
-J
W
a:
::>
en
en
W
a:
a..
o
z
::>
g
o
z
«
[()
w
>
«
I-
o
o
o
a:
:r:
I-
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.19.2 - Noise levels for wood planer with helical cutter
194
.'
110
straight cutter ~ ~"I""I_ _
« 105
m
"C
UJ
a::
::J
U)
o0..
X 100
UJ
a::
o
I-
«
a::
UJ
0..
o
95
90
2 3 4 5 6
BOARD LENGTH - FEET
FIGURE 5.19.3 - Effect of board length on noise from wood planer
195
CASE HISTORY #20 - PUNCH PRESS*
Punch presses constitute a most troublesome source of indus-

trial noise, both because of their number and because of their high
noise output.
From various papers on the subject of punch presses, the follow-

ing list of noise sources has been gathered. These may not all be
present on anyone press but are listed as a guide to specific press
noise source analysis.
(1) Shock excitation of the workpiece, machine guards, ,

floor and building.
(2) Gears, drive, bearings, and components, such as '

clutch and brake mechanism and drive shaft.
(3) Plunger resonance.
(4) Air ejection, air jet cleaning, and air cylinder

exhausts.
(5) Die design.
(6) Stripper plate design.
* American Industrial Hygiene Association, Industrial Noise Manual,

Detroit, 1966. Examples 11.C" 11.EE.
Allen, C. H., and R. C. Ison, A Practical Approach to Punch Press

Quieting, Noise Control Engg. 3 (1): 18, 1974.
Bruce, R. D., Noise Control of Metal Stamping Operations. Sound

and Vibration ~ (11): 41, 1971.
Shinaishin, O. A" On Punch Press Diagnostics and Noise Control.

Proc. Inter-Noise 72: 243, 1972.
Shinaishin, O. A., Sources and Control of Noise in Punch Presses,

Proceedings Purdue University Conference on Reduction of Machine
Noise, May 1974, page 240.
Stewart, N. D., J. A. Daggerbart, and J. R. Bailey, Identification

and Reduction of Punch Press Noise. Proc. Inter-Noise 74: 225,
1974.
(7) Ejection of parts leaving press on chute or bin.
(8) Vibration of sheet metal being fed to the press.
(9) Start and stop of automatic feed to the press.
(10) Building acoustics.
Shock excitation of surrounding structures--This effect can be

minimized by properly designed vibration mounts for the entire press
to reduce excitation of floors, walls, and other equipment. As an
example of this see Case History #2.
Drives, etc.--Good maintenance can contribute to noise reduc-

tion. The noise of drive gears can be reduced by damping the gear
body, improving gear surface quality and tolerances, precision in-
stallation and bearings, better lubrication, and/or changing gear
material for a better damped material. On existing equipment many
of the above aids cannot be done at reasonable expense, but gear
drives are often enclosed in a box-like structure whose surfaces
radiate noise. These surfaces can be damped with off-the-shelf mate-
rials, or the drive unit, if space is available, can be enclosed,
fully or partially. Heat dissipation must be considered. Solid
metal or plastic guards can be changed to expanded metal or wire
mesh for less noise, or the guard surface can be vibration damped.
The entire guard, if solid, must be vibration isolated from the
vibrating machine.
Plunger resonance--If a hollow plunger or ram is a Helmholtz

resonant type of noise source, its noise radiation can often be
reduced by covering the hole in the plunger. See Case History #4.
Air ejection of punched parts--If possible, substitute mecha-

nical ejection to eliminate a large noise source. One comparison,
shown in Figure 5.20.1, (AIHA 1966) resulted in an 8-dBA noise reduc-
tion. Multiple jet nozzles are also available for reduced noise.
Reduce to a minimum the air velocity used for ejection (since noise
level is related to velocity) by reducing the air pressure avail-
able. Achieve better efficiency of air jet by accurate setting and
aiming where needed.
Shield the area of punch-air ejection from the operator. An

example of the result of this method, in Figure 5.20.2, shows the
noise levels of a press with and without a 24 x 48 inch shield to
protect operator ~rom air ejection noise.
197
Die design--Changes in die design can reduce noise by attempt-
ing to design for spreading the punching action, slanting the
blanking punch or die, or other means of promoting consecutive
shear action instead of instant action. Shinaishin reported the
results of a slanted die as shown in Figure 5.20.3. Changes in die
materials can reduce noise. As presses produce sound energy from
vibration of metal plates upon impact, the velocity of impact can
be reduced by using hard rubber mounts (snubbers). Another possi-
bility is laminated and more massive plate, reducing the size of
plate and radiating area.
A change of work stock material from steel to a lead-steel

composition has also reduced impact noise; Shinaishin reported a
14-dB reduction with this test method. Noise radiation can be
lessened by reducing plate area by cutting out any surface areas
that perform no function.
These comment~ emphasize that the tool engineer must now con-
sider designing for noise reduction as well as for mechanical per-
formance. Within such general framework outlined, any improvements
in noise level will come by experiment and testing results.
Stripper plates--Stripper plates in some dies contribute to

noise levels due to metal-to-metal contact, which could be changed
to plastic or elastomeric contact with better damping and reduced
noise.
Ejection of parts to chute or bin--Noise levels can be reduced

by damping metal chutes, using damping materials on the market or
making a constrained layer design. See Case History #17.
Vibration of sheet metal being fed to press--Noise levels can

be reduced by preventing vibration, such as by adding a hold-down
conveyor. The noise can also be constrained by using an acoustic
tunnel in-feed, or the operator can be shielded by properly designed
barriers.
Start and stop feed mechanisms--Noise can be reduced by rede-

sign: substitute plastic contact areas where possible; enclose the
noise source partially; or add barriers between noise source and
operator.
Building acoustics--In a room with many noise sources, the oper-

ator may be in the reverberant field. Such noise can be reduced by
adding absorption. From Bruce, an example of use of absorption to
reduce noise in a press room is shown in Figure 5.20.4, 30 feet
from presses. Closer to presses, noise reduction would be less
with probably no more than 2 to 3 dB at the operator position.
The press area can also be enclosed or walled off from the rest of
the plant.
198
Use of Enclosures
Allen and Ison (1974), P.18, reported a partial enclosure of ram,

die, in-feed, and ejection on a 50-ton test press. A reduction
of 13 dBA was obtained for an enclosure; see Figure 5.20.5. The
model enclosure was made of cardboard, ~ lb/sq ft, lined with
1 in. of polyurethane foam. Later a steel enclosure was installed,
for durability.
Total enclosures with opening via an acoustic tunnel may be

required and should follow recommended enclosure design as re-
viewed in Chapter 3.
Pitfalls
The remaining radiation came chiefly from the flywheel cover,

which was neither damped nor vibration isolated. Diagnostic meas-
urements should indicate the relative contributions from each
source, so that the residual noise will be known.
199
eo
1 355 710 1400 7S00 - ~ - 11200
100 1---+----+-----1---+---+---+------11----+----1
97 dBA
-l
W 87 dBA
>
W
-l
W
a: 80 1---+-----,,.=--:t...L...----1---+----+---=.~_t_---I_--_+_--_l
::::l
(/)
(/)
W mechanical ejection
a:
D-
O
Z
::> 70 1---+----+------1---+----+---_t_---I_---+----j
o
(/)
o
z
«
cp
w
>
«
I-
o
o
I
-l--
63 175 750 500 lena 7CDO .000 8000
100 1000 10000

FREQUENCY IN HERTZ
Source: AIHA, 1966.
FiGURE 5.20.1 - Comparison of punch press noise levels with air ejection and with mechanical ejection
200
180 J55 710 l.«lO - 2800 - 5600 - "200
101 dBA
100 1----+-----1---+----+---+----+-----11----+----1
90 I---=+=----.:+---+--~--II!!!!!~~---~--+--
89 dBA
...J
W
>
W
...J
W
§ 80 1----+----1--I-J,....-+----+---+---7'-t--...;;;;;:~~
.........__- - - - i
(J)
(J) with shield
w
a:
0-
o
Z
701---+----+-----j~--+----+----+--'---+---+-----j
o:J
(J)
o
z
«(D
I
W
«>
I-
U
o
6J 125 250 500

I
le00 2000 4000 8000
5 2 I
100 1000 10000

FREQUENCY IN HERTZ
Source: AIHA, 1966.
FIGURE 5.20.2 - Comparison of punch press noise levels with and without a shield between operator
and air ejection noise
201
180 355 710 uoo 2SOO - 5600 - 11200
108 dBA
III
"'C
...J
standard die
100 1---_+_......._~-+_~rt_:l~----::IIllrl_-~~~r2"~_+_--~--
I
100 dBA
w
>
W
...J
W
§en 90~---=t:l---\:~-I-~~-~=------i=----~~~~~--+:--_--.j
en
W
cr
a..
o
z
80 1----II'----H~-_+---_t_--_+--~+_~.,......_+-t-~c.._1---_1
o:::l
en
o
z
«
III
W
> 701----4----t---+----+---+----+-----+----+----i
«
t
~
()
o
63 115 150 ~oo

-t1
+-
Ier,O
i
2000 AOOO 8000
100 1000 10000

FREQUENCY IN HERTZ
Source: Shinai shin, 1972.
FIGURE 5.20.3 - Comparison of punch press noise levels: standard die versus slanted die
202
90 180 355 710 10100 - 2800 - S600 - 11200
(])
"C
100 1 - - - 4 - - - - + - - - - - - - 4 - - - + - - - - + - - - - + - - - 1 - - - + - - - - 1
-l
W
>
W
-l
before sound absorption 94 dBA
W
a:
~
en
90 1--i:---=~~.-.~~<,......1~=-=-±~;;:_=t--1t-M~3A-1
W
a:
0-
o
Z
:J 801-----+-----1--...,....-+----+---+-~lk__+_--___1f--~=__-+---__l
o
en
o after sound absorption
z
«
(])
W
>
I 1
701-----+------4---+----+---+-----t-----If----t----t
«
l-
t)
o
63 125 250 500 lena 2000 ~ooo 8000
100 1000 10000

FREQUENCY IN HERTZ
Source: Bruce, 1971.
FIGURE 5.20.4 - Noise levels 30 feet from bench press area before and after sound absorption treatment
203
90 180 710 1~ - 2100 - 5600 - 11200
110 1----t----+---t---t----+----+---;t---+----1
CD
"0
, 100 1-----+-----t---+----+---+----+---___1---+-----1
.....J
ill
> 96 dBA
ill
.....J
ill
§ 901-----+---+----+---+----+---+----:::....~;;;.....--+------1
-
C/)
C/)
ill
a:
0-
o 83 dBA
Z r~~ _ _ . normal operation
::l 801----+-----+---+--#---+---t------+-----+---+-----1
oC/)
o
Z
<t:
CD
W
> 701----+--~.-:::-...:..::=!'+o.:::-----+-----l-----+---t---+------I
<t: with test enclosure
l-
e..:>
o
60 I---+----+----l---+----+----+---l---+---~
S
63 12S
2
250 SOO
S
ICOO
1 -
i
2000 _000
s
8000
100 1000 10000

FREQUENCY IN HERTZ
Source: Allen and Ison, 1974.
FIGURE 5.20.5 - Noise 30 inches from punch press before and after test cardboard enclosure
204
CASE HISTORY #21 - MATERIALS HANDLING, AIR MOTORS
Air operated motor hoists are a noise source in many indus-

tries that make extensive use of materials-handling systems.
Control Approach
As the noise source is the exhaust air, this exhaust can be

muffled by using off-the-shelf mufflers selected for the air
pressure and delivery of the exhaust.
A typical octave band analysis, before and after installation

of an exhaust muffler, is shown in Figure 5.21.1. Note the rising
spectrum that is characteristic of freely-escaping high pressure
gas. Another case showed the following A-weighted sound levels at
the floor for a I-ton air hoist:
No Muffler With Muffler

up - no load 98 85
Up - 600 lb 96 84
Down - no load 102 88
Down - 600 lb 100 86
Air exhaust from other tools can be similarly muffled. Newer

designs include mufflers, which should be specified at purchase.
205
90 180 3S5 710 lAOO - 7800 - 5600 - 11200
115 dBA
11 0 t--..:....-+---+----+---+----+---I----::~--+_--_l
CD
"0
100
-l
w after muff ler
>
w
-l
w
a: 90
~
(/)
(/)
w
a:
a..
0
z
~ 80
0
(/) 81 dBA
0
Z
"
c:{
rn
I
W before muffler
>
c:{
70
l-
t)
0
60
63 125 250 500 1000 2000 8000
S 5 S
100 1000 10000
FREQUENCY IN HERTZ
Source: AFS-Foundry Environmental Control of Noise, 1972.
FIGURE 5.21.1 - Effect of muffler on air exhaust from hoist
206
CASE HISTORY #22 - TEXTILE BRAIDING MACHINES*
In braiding operations, a bobbin of thread is rotated on a

carrier base in a special slotted cam. This cam revolves as it
is rotated around the machine, with several other carriers and
cams. The carriers are thrown from one cam to another. With steel
carriers, the major source of the intense noise present is the re-
sulting metal-to-metal impact. The manufacturer was willing to
consider machine modifications to reduce noise in the case history
reported here.
Measurement and Analysis and Control
In a laboratory study, the metal-to-metal contact was easily

identified as the chief noise source. It was recognized that a
carrier with inherent damping properties should reduce the noise.
Replacement of the carrier by a nonmetallic one was thus considered.
Of the several materials tried, the best combinations of strength,
light weight, and damping was an injection moldable polyurethane.
The carriers were installed in a l3-carrier braider operating

at a handle speed of 340 rpm. with the microphone 10 inches above
the top plate of the braider and 18 inches out, the levels shown
in Figure 5.22.1 were obtained. A reduction of 11 dB was obtained.
The above results were obtained in the laboratory. For an

in-plant test, a row of 84 braiders was converted to plastic car-
riers. The adjacent row was left with steel carriers; other ~ows
of braiders were operating. The microphone was 3 feet from the
centerline between the test rows, and 3 feet above the floor. The
noise levels for various combinations of machines is shown below
(an x indicates on).
Level, dBA 97 97 90 85
Steel test row x x
Plastic test row x x
All other x x x x
Residual noise from the motor cooling system remained and lowered
the noise reduction to the 7 dB achieved in this production test.
* From Cudworth, A. L., and J. E. Stahl, Noise Control in the Textile

Industry. Proc. Inter-Noise 72: 177, 1972.
207
Pitfalls
Since this study, it has been round that the plastic carriers
are not strong enough for some operations requiring heavy yarn (or
wire). This suggests consideration of a composite carrier with a
steel core for strength and a cladding of heavy polyurethane for
damping. To our knowledge, this concept has not yet been tried.
This emphasizps the need for considering non-acoustical parameters
along with the acoustical.
208
180 J55 710 l~OO - 2800 - !600 11200
94 dBA
CD
"'0
...J
w
>
W
...J
W
a:
80
-
:J 70
C/)
C/)
ill
a:
a..
0 polyurethane
Z
=>
0 60
C/)
til'
0
z
«
CDI
",'"
W
>
« 50
I-
()
0
-j-
5
63 125 250 500 le,oo
I
7000 4000 aoao
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.22.1 - Textile braiding machine: comparison of noise levels from steel carriers and
from polyurethane carriers
209
CASE HISTORY #23 - METAL CUT-OFF SAW*
A common problem in industry is that of protecting workers from

noise produced by machines that the worker must guide or manipulate
directly. An example is a cut-off saw used on metal shapes. Noise
comes from two main vibrating sources: the saw blade itself and the
workpiece, which is largely unconstrained. The saw itself is actu-
ated downward and into the work by a lever attached to the hinged
and counter-balanced (or spring-loaded) saw and motor.
The worker must visually monitor the cutting operation. Also,

and perhaps more importantly, the vibration and opposing force
,transmitted to him through the lever arm also furnish useful cues
on the progress Of the cutting operation. The problem is to reduce
the noise he receives, without undue interference with work flow,
with visibility, and with the use of the lever arm.
The solution was an enclosure covering the whole saw. Work-

pieces pass transversely through slots in the enclosure. Flaps of
lead-loaded vinyl close off the opening and reduce to a small
amount the unavoidable leakage area when a workpiece is present.
The front, above saw bed height, is closed by two doors whose'sur-
face is mostly ~-inch clear plastic (polymethylmethacrylate). This
plastic provides very good vision. The doors close with a gap the
width of the control lever. Each door has a flap of lead-loaded
vinyl about 3 inches wide to close the gap. The lever pushes aside
t.he flaps only where it protrudes. Thus the leakage toward the wor-
ker is greatly reduced.
Figure 5.23.1 shows the spectra at the worker position before

and after the enclosure was installed. The decrease in A-level is
13 dB. The standard panels used in the enclosure are very much
better than indicated by the reduction measured, illustrating again
the importance of leaks in determining the performance of enclosures.
Pitfalls
Several features of the design could be improved. The ears of

the workers are very close to the leak at the door flaps. It should
be relatively simple to offset the saw feed lever to the right (for
* Handley,J. M., Noise--The Third Pollution. lAC Bulletin

6.0011.0, 1973.
210
the right-handed worker). This has several advantages: (l)it
places his right hand in a more comfortable position, (2) with
the door gap and flaps moved to the right, his vision is greatly
improved, and (3) the noise leak is moved further from his ears.
A nonacoustical improvement would be to have the doors slide open,
rather than open out, which can be a safety hazard.
211
90 leo J55 710 1.we - 2800 - 5600 - 11200
97 dBA
90 r--~--:=~=~-~~=-+:-----::+-~;;;;;;;;;;;:;;;+:--I
before enclosure
84 dBA
70 1-----4----+----+----+----+---+----1----+-----1
--+-
6J 125 250 500 1000 2000 4000 . eooo
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.23.1 - Metal cut-off saw: operator's exposure before and after enclosure of saw
212
CASE HISTORY #24 -PLANER-WOOD*
Wood planers in the forest products ind~stry produce noise

levels of 102-108 dBA at the operator (feeder) work stations. Noise
levels are 103 dBA at the grading station and trimmer and 95 dBA
elsewhere in the planing mill.
In the area cited in the article, enclosures were installed on

30 large planers. Out of the general program, the following specif-
ic guidelines for viable enclosures were developed by experience.
(1) Walls and roof should be from 8 to 10 feet high, using

staggered studs; thus keeping inside and outside wall independent
with separate sills and headers. Isolate wall structure from floor
with felt or mastic. Fill space between walls with rock wool or
fiberglass. Sheathing used was l~-inch tongue and groove or equiv-
alent plywood. Additional acoustical board was used on upper two-
thirds of walls and ceilings for noise absorption. Removable wall
or roof sections should be installed as needed for major machine
repairs.
(2) Floors are usually adequate as constructed for a normal

planer installation, but if the planer is elevated on piers, the
enclosure walls should be extended to the main floor or acoustical
floor similar to the walls constructed between piers.
(3) Doors should be refrigeration type, with beveled or

stepped edges. They should open out, so that suction from blowers
keeps them closed. Doors or jams should be sealed with weather-
stripping. Use heavy duty hinges. Alternatively, standard acous-
tical doors may be purchased.
(4) Windows should be as small as practical, using double-

glazed shatterproof or screened glass with an air space between.
(5 ) Infeed and out feed openings should be as small as possible.
A tunnel type opening provides room for multiple baffles of old
conveyor belting or lead-filled vinyl to block the noise path. Slit
the belt at intervals to accommodate various board widths, keeping
* From Pease, D. A., Forest Industries, March 1972.
213
the unused portion of the tunnel width blocked. The outfeed tunnel
should be at least as 16ng ai the longest boards fed through the
~laner so that noise caused by the vibrating board is confined
inside. Install funnel-shaped metal facing inside to guide the
stock into the tunnel opening.
(6) Opening for ducts and pipes should be just enough over-
size to permit packing the annular space with insulation.
(7) Make-up air openings, to compensate for air exhausted by

the blower system of the planer, must be constructed as a silencer
to control noise leakage. Construct a chimney several feet high,
with baffles arranged inside so that incoming air must follow a
zig-zag path; baffles should be lined with acoustic material.
Another method is a smooth-wall chimney with a "weather cap" baffle
lined with acoustic material at the top.
The article states that noise levels were reduced to less than
90 dBA, to comply with OSHA regulations.
Pitfalls
At these high levels leaks are critical, and should be kept

to the minimum.
The absorbent must be covered by a plastic film to avoid foul-

ing by the dust. In addition, because lumber is not always fed in
straight, the absorbent must be protected by a heavy galvanized
screen.
The feed tunnels should be long enough to hold the whole board,
or else there must be positive hold-down to prevent board vibration.
214
CASE HISTORY #25 - PUNCH PRESS TOTAL ENCLOSURE
Two Minster model P2~2000, 200-ton straights ide presses were

running over 250 strokes per minute in stamping out laminations
for a particular motor model. The press is located in a metal-
construction building. Dies are changed often.
Noise level at the operator station was 104 dBA~ and the gen-
eral plant level was 92 dBA.
Panels forming a total enclosure were constructed using:
1 layer absorbent polyurethane acoustical foam

1 layer 1/64-inch sheet lead
1 layer 3-inch fiberglass Trw blanket
1 layer fiberglass cloth to withstand industrial solvents.
The enclosure used was circular, 1?6-inch in diameter,- 16-feet

high, with top of domed construction. Access doors allow for main-
tenance, and there is a stock feed opening. Finished parts leave
the enclosure by means of two under-floor part guides. Supply lines
were rerouted under floor, using flexible conduits. A 3500 cfm
heat exhaust system with a silencer was added to each dome.
The operator is outside the enclosure except to change dies,

change feed reels, or make adjustments.
Total enclosures reduced noise level for operator to 83 dBA

and general plant level (with other equipment) to 87 dBA.
215
CASE HISTORY #26 - DEWATERING VACUUM PUMP, PAPER MILL*
In paper manufacturing, a fine-grained mineral slurry is

dewatered by a suction press and a vacuum pump. The mixture
of air and water discharges into a drain; slugs of water make
the discharge pulsate irregularly. Although no workers stay
long in the discharge area, the levels are high even at consider-
able distances from the discharge pipe. In one instance, the
pipe was inadvertently tuned to the pump pulse frequency, thus
compounding the problem.
The solution is a special combination muffler and water/air

separator called a snubber. Besides use of the snubber, the
line length should be changed to de tune the system.
Figure 5.26.1 shows the effect of adding a snubber and the

greater effect of two snubbers. The low frequency resonances
are greatly reduced by this direct scheme. A-levels close to
the discharge pipe were changed from 112 dBA without snubbers to
103 dBA with one, and 93 dBA with two snubbers. Some of the low
frequency reduction is probably due to the detuning effect of
the change in the acoustic length that the snubbers introduce.
* Young, R. L., Practical Examples of Industrial Noise Control.

Noise Control 4 (2): 11, 1958
216
120 ,.....--""'T""----r---r---...,..---"""T'"----,.---r---'"T"""---,
...J
UJ 110
>
UJ
...J
UJ
a:
::>
en
en
UJ
a:
a..
0 90
z
::>
0
en
0
Z 80
-<
III
UJ
>
-<
I- 70
()
0
60 L-_ _.....L. ......_ _...J..._ _.....L. .L...._ _-I-_ _......I - ' -_ _......
20- 75- 150- 300- 600- 1,200- 2,400- 4,800-

75 150 300 600 1,200 2,400 4,800 10,000
FREQUENCY BAND IN Hz
FIGURE 5.26.1 - Octave-band analysis of paper mill vacuum pump noise
217
CASE HISTORY #27 - STEAM LINE REGULATORS*
Steam lines with regulators are used in many industries and

can be a problem noise source if they are in an area occupied by
employees.
The method used here, which can be used to regulate other gas
flows, was to modify the design of the main valve plug. The rede-
signed valve plug has throttling vanes, as shown in Figure 5.27.1,
to reduce the turbulence of the steam flowing through the space
between the regulator's main valve and its valve seat, which is
the noise source.
For a 2~-inch steam line handling 50,000 lb/hr through a

reduction of 555 to 100 psia, the redesigned valve reduced pipe
line noise from 97 dBA to 85 dBA.
* From Electrical World, January 1973.
218
FIGURE 5.27.1 - Main valve plug with throttling vanes to reduce nois':l in steam fine regulator
219
CASE HISTORY #28 - PLASTICS SCRAP GRINDER*
In the molding room, primary noise sources are from scrap

grinders and plastic granulators. The noise has increased during
the past few years because of the growth in number of grinders
and increasing toughness of the newer plastics.
Noise level maxima of 125 dBA in the initial grinding phase

have been. recorded, and 100 dBA are common. Figure 5.28.1 shows
the octave band analysis of an untreated grinder, giving the range
of readings for one load of 4 pounds of polycarbonate.
Control Approach
Although the optimum mechanical conditions of the plastics

scrap grinder, such as sharp blades, proper screen size, blade-
to-screen clearance, and proper feeding procedures, help reduce
grinder noise on existing equipment, this alone could not bring
the unit within acceptable noise limits. Much of the noise came
from resonant excitation of metal panels.
A damping material was applied to all surfaces: hopper,

interiors of pedestals, stands, and covers. In general, a ~-inch
coating has been satisfactory for most grinders from bench models
to 18 x 3D-inch throat grinders.
The results of the soundproofing, shown in Figure 5.28.2

(for one load of 4 pounds of polycarbonate), bring noise levels
down to OSHA criteria, reducing the maximum level from 100 dBA
to a range of 88 dBA to 90 dBA.
* Morse, A. R., Plastic Technology, July 1968.
220
90 180 355 710 1400 - 2800 - 5600 - 11200
11 0 I----+-----+---+----+----I----+---~I----+----I
CD
"0
100 100 dBA, max
...J
W
>
W
...J
W
a:
::>
(/)
90
(/)
w
a:
0-
0
Z
::> 80
0(/)
0
Z
"
<{
CDI
W
> 70
<{
I-
0
0
---t-
63 125 250 500 1000 2000 4000 8000
5 5
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5.28.1 - Plastics grinder,- range of noise levels, dBA
221
ISO 355 710 1400 - 2800 - 5600 - 11200
110 I---+_---+----i---+----+---l----~--+_--_;
CD
"'C
I 100 t - - - - t - - - - f - - - + - - - - + - - - t - - - - t - - - - - f - - - + - - - - f
..J
W
>
W
...J
W
II
:J 9 0 t - - - - t - - - - f - - - + - - - - + - - - t - - - - t - - - - - 1 - - 90 dBA
rn
rn
W
BB dBA
II
a..
o
z
isrn Bo~-4+--~---- ..--~---~----:3t---~--._011!1 - - _ 4 ...
o
z
«
CDI
W
>
-
«
t-
()
o
63 12~ 250 ~oo 1000 2000 8000
100 1000 10000

FREQUENCY IN HERTZ
FIGURE 5. 2B.2 - Plastics grinder; range of noise levels after soundproofing, dBA
222
CASE HISTORY #29 - NEWSPAPER PRINTING PRESS*
This pressroom is equipped with five double 3 to 2 Hoe

folders and four double 2 to 1 Hoe folders with a complement
of 45 colormatic press units.
The following methods were used for noise reduction:
Enclosures for folders reduced noise from III to 101 dBA.
In the reel room, all openings in the floor or deck plates

between the press room level and the reel room were sealed and
isolated. The opening in the arch of the press was closed to the
smallest dimension that would still allow paper to feed through
to the unit. On the basis of dosimeter data, the noise exposure
was reduced to acceptable levels.
On the press room floor, an existing folder enclosure was

retained and improved. A control booth was constructed for noise
isolation. An S-foot wall was added on the press room floor as
noise barrier, plus a 4-foot panel at the top of the wall, angled
upward and toward press. Wall surfaces were lined with 2-inch
absorbent polyurethane. The S-foot wall was constructed of:
26 gauge metal, l/S-inch masonite, 3/4-inch air space, 3/S-inch
plywood, plus 26 gauge metal. The panel was 2-inch polyurethane,
1/2-inch plywood, and 26 gauge metal.
Sound traps were made at the tops of ladders at catwalk

level. No isolation of the stairs, from reel room to press room,
was necessary, ~s they are outside the press enclosures and not
affected by the high noise levels of the press. Wall panels are
easily removed for maintenance.
Pressmen going inside enclosure for adjustments on a short-

time basis wear ear protection.
Materials used for sound absorption were flame resistant

and approved by insurance inspection.
Noise levels were reduced to comply with OSHA standards.
* From Editor & Publisher, November 10, 1973.
223
CASE HISTORY #30 - CHEMICAL PROCESS PLANTS*
Existing chemical process plant noise reduction requires

source analysis to determine the method of noise reduction.
Methods
As a result of this study, a list of noise sources is shown

in Table 5.30.1, with recommended methods of noise reduction.
Some specific examples and the results obtained by each noise con-
trol method are cited in Figures 5.30.1 through 5.30.4. The atten-
uation attained is shown in each figure.
* From Judd, S. H., Noise Abatement in Process Plants, Chemical

Engineering, January 11, 1971.
.224
105
II I I
sound pressure level measured 3 feet
100 above fan, 3 feet out from rim
en
,
"0
--l 95
w
>
w
"
--l 90
W
a:
~ 85
"
-~ '4
~
- ......
V 12,000 fee't per minute tip speed
~<'
en
en
...........
'-
W
a: 80
..........
- -.. -- .. ..~......
11..
I~ ....
0
Z
~ 75
~ ............... ....
0
en
10, 000 feet per minute tip speed
~-
-
70
65
~ .....
20 - 75 - 150 - 300 - 600 - 1,200 - 2,400 - 4,800 -
75 150 300 600 1,200 2,400 4,800 10,000
OCTAVE BAND, Hz
FIGURE 5.30.1 - Noise reduction achieved by reducing fan speed, using increased blade pitch to offset
decrease in speed (measured 3 feet above fan, 3 feet out from rim)
225
120 I I I I I r
sound pressure level measured
'¥r~.. fi rst stage expans ion [!) 115 -
~
:~1; "0 1-foot from expansion joint

":::: joi nt on a two-stage
>
~~~-: centrifugal 9,600-rpm, ~ 110
W
:;.~j~
.~\~
6,000-scfm at 100 psig >
W 105
compressor ~
W J,
a: 100
:::>
en
en
w 95
~
'U~~~
~ before -,
I
I
"~
a:
"~ ".... .......--

0.. 90
0
z
:::> 85 ,.- ~
26 gauge galvanized steel
1/ 16-i nch lead
0
en
80
after -
'-,,
3-inch, ti-pound glass fiber 75
20 - 75- 300 - 1,200 - 4,800 -
75 150 600 2,400 10,000
OCTAVE BAND, Hz
FIGURE 5.30.2 - Compressor discharge noise reduction achieved by lagging expansion joint (measured
1 foot tram expansion joint)
226
new 10-inch loop
• in 115
, ~. I /. II 1. T
soun pressure level measured 10-lnches
110 ~ from #2 pump discharge line -
t CD
u "0105
r ...J
b pump #2 w j~
>
: ' .....
i 100
n I ill
pump #1 ...J _origina: 8-inch
e w 95 "jump over" -
-1Il~
diSCharge1
line 8-inch by
6-i nch taper
c:;::
~90
I
~I I"
-",
out
".~...
(J)
original 8-inch
w
er 85 ~
-IIt.~ 'I
:-"""'III
~
0..
"jump over"
oZ 80
~.
new lO-inch loop· ~
IlPump III 'pump 1/2 => ~
~ 75 I I "'
"
inlet psi 35 580
outlet psi 580 840 70
8 volute impellers 4 2 20 - 75- 300 - 1,200 - 4,800 -
rpm 75 150 600 2,400 10,000
3,600 3,600
gpm 2,100 2,100
OCTAVE BAND, Hz
FIGURE 5.30.3 - Noise reduction achieved by redesigning pump by-pass loop (measured 10 inches
from tl2 pump discharge line)
227
11 Olr--T""'"--,r--.....,---r-""'T"-~-..,...-'t""""'_
sound pressure level measured I
lD 105 ....- + - - + - - - 6 feet from air intake slot,
"0
-.J- 1 00 t-~a__-~-+_-~-+_-+_-,.......--,.......-___'II
ill
>
ill
-.J
ill
a:
115,000 ~ 851~-~r_+-_+...."",.1IIIF_-_+_-4.J'"'"""~~-~
en
scfm to ill
drain a:
(L
80 t---+--+--+-----'+=..,.;;po..,...+--~r___f',...___4
blower
~
!
75
:::>
,~ 701--+--+---4----1
65,....--......._ ........._ ......._ ......._""'--_............-""
20 - 75 - 300 - 1,200 - 2,400 -
section through intake si lencer 75 150 600 2,400 4,800
on a 17-psig centrifugal blower
OCTAVE BAND, Hz
FIGURE 5.30.4 - Noise reduction achieved by adding silencer to air blower intake (measured 6 feet
from air intake slot)
228
TABLE 5.30.1· Sources 01 noise and methods 01 noise reduction
Equipment Source of noi se Method of noi se reduction
Heaters Combustion at burners Acoustic plenum' (10 Bwg. plate)

Seal s around control rods and over
sight holes
Inspiration of premix air at burners Inspirating intake silencer
Draft fans Intake sil encer or acoustic p!enum
Ducts Lagging
Motors TEFC cooling air fan Intake si lencer
Undirectional fan
WP II cooiing system Absorbent duct liners
Mechanical and el ectrical Enclosure
Airfin coolers Fan Decrease rpm (increasing pitch)
Tip and hub seals
Increase number 01 blades"
Decrease static pressure drop"
Add more fin tubes"
Speed changer Belts in place of gears
Motors Quiet motor
Slower motor
F an shroud Streamline airflow
Stiflenlng and damping (reducing vibration)
Compressors Discharge piping and expansion joint Inline silencer and/or lagging
Anti surge bypass Use quiet valves and enlarge and stream-
line piping"
Lag valves and piping
Inline silencers
Intake piping and suction drum Lagging
Air intake Silencer
Di scharge to air Si lencer
Timing gears (aXial) Enclosure (or constrained damping on case)
Silencers on discharge and lagging
Speed changers Enclosure (Or constrained damping on case)
Engines Exhaust Si lencer (mUffler)
Air intake Si lencer
Cooling Ian Enclose intake or di scharge Or both
Use quieter tan
Mi scellaneous Turbine steam di scharqe Si lencer
Air and steam vents Si lencer
Use quiet valve
Eductors Lagging
Piping Limit velocities
Avoid abrupt changes in size and direction
Lagging
Valves Limit pressure drops and velocities
Llmi t mass 1I0w
Use constant velocity or other quiet valve
Divide pressure drop
Size adequately for total flow
Si ze for control range
Pumps Enclosure
'If oil fired, provide for drainage of oil leaks and inspection .
•• Usually limited to replacement or new facilities.
229
Chapter 6
SELECTING AND USING A CONSULTANT
Knowing When a Consultant is Needed
Having read the previous chapters, you will already have a

good feeling for the situations that you can deal with on your own.
If you are still unsure of the solution or if preliminary measures
have proved unsatisfactory, it may be time to consider the use of a
consultant.
Situations where a consultant is needed arise when the machine

to be quieted is complex~ as when there are present many noise
sources of approximately equal strength. Locating the sources and
obtaining their relative noise strengths will need perhaps more so-
phisticated equipment and procedures than you may have. If you
find that the A-level at all points at a constant distance from the-
machine (but within the critical distance) covers a range of 5 dB
or less, this is likely to be the case.
You may also need a consultant if unusual situations exist. A

common one with belt-driven blowers, for example, is a slow but
considerable variation in level. Another is impact noise, as from
a punch press, where several events take place in rapid succession.
A narrow-band analysis of a tape recording is usually called for.
Inadvertent tuning of some part of the machine may lead to pure
tone ringing that is difficult to locate. For such situations,
using a consultant is often the most rapid way of getting results.
If you have installed noise control means that don't work, you
may (albeit reluctantly) have to use a consultant to retrieve the
situation. Although this may be a painful decision, it will usu-
ally occur but once. You should document the situation thoroughly
and use the consultant to supply information on what went wrong.
Sometimes you may be approaching a lawsuit situation, where

data must be obtained and presented (as an expert witness) by a
disinterested third party. Many consultants can provide this com-
plete service.
Now that you have decided to obtain a consultant, how do you

proceed? You should first be warned that there is no legal bar to
anyone offering services as an acoustical consultant. Consequently,
it is up to you to avoid those who are unsuitable because of lack
of training or experience, as well as simple venality or greed.
iPreceding page blank

231
Selection of a Consultant
People billing themselves as consultants can be broadly classi-

fied according to whether or not they have a special interest in
recommending a particular acoustical product or solution. Both
types, properly used, have their special advantages and disadvan-
tages.
In the former group of "special interest consultants", individ-

uals vary in their backgrounds from product salesmen to professionals
who are quite capable in their line of business. This group, which
is most commonly indicated by the degree of their association with
manufacturing or retail sales of acoustical products, should be used
directly only if, by use of the techniques described in previous
chapters, you have satisfied yourself that their solution is appli-
cable to your problem. In this case you have progressed to the point
where the "consulting" aspect consists mainly in soliciting propos-
als for design and installation. The main problem remaining is to
write your contract in such a way that you are guaranteed (to the
extent possible) of actually solving your problem in a cost effec-
tive manner. The advantage of using this group directly is that you
avoid ,consultant costs and pay for the product. In effect you are
acting as your own consultant.
The disadvantage in dealing with a product-oriented consultant

is that a costly mistake, out of proportion to the independent con-
sultant's fees, is rendered more likely. Examples abound of cases
in which thousands of dollars were spent in implementing a particu-
lar solution, only to find that no good was done. (A common mis-
take is to utilize acoustical tile in situations where reverberant
noise is not the problem.) If there are any doubts in your mind as
to the proper method for solving your problem, then an "independent
consultant" (one free from ties to a particular line of products)
should be called in. Since this "independent consultant" is what
is usually meant by the word "consultant", it is this type of pro-
fessional that will be discussed for the remainder of this chapter.
The word independent will be dropped.
In choosing a consultant a first step is to inquire of the two

organizations in the field that are interested in the qualifications
of their members. The most inclusive is the Institute of Noise Con-
trol Engineering (INCE), P. O. Box 1758, Poughkeepsie, New York
12601. This group will have as members only those who have passed
suitable examinations. Applicants must pass both the Engineer-in-
Training examination given for registered professional engineers
and a special examination on noise control engineering that was
first given in 1974. There will be no grandfathering - the term
applied to certification purely on the basis of past activity in
the field. It is expected that this certification of noise control
232
engineers will receive approval by and support from the government-
al agencies needing such engineers directly, or in the contract
work they support.
The second source of information on qualified acoustical con-

sultants is the National Council of Acoustical Consultants (NCAC),
8811 Colesville Road, Suite 225, Silver Spring, Maryland 10910.
This group has a rigorous code of ethics ~equiring (as with all reg-
istered professional engineers) that no member be associated with
the sale of a product. Consequently some consultants, otherwise
well qualified, are not members because of this association. A
membership list is available. Not all the completely classified
consultants are yet members.
At the present writing, INCE certification is just beginning

and the NCAC list is not yet complete. Therefore the most common
recourse is to question the prospective consultant yourself. A
series of questions is given below. These questions are rather
completely presented here, and you may wish to ask only those that
are pertinent to your particular task.
Guideline Questions
Education
(1) What schools did you attend?
(2) What courses did you take bearing on acoustics?
(3) What degrees did you receive? When?
(4) In what special conferences, seminars, symposia, or

graduate courses in acoustics have you been involved,
either as a student or as an instructor?
Experience
(1) For how many years have you been professionally

active in acoustics?
(2) Please supply a list of recent clients that you have

served, preferably in my geographical area, and on
problems similar to those in which I am interested.
(3) What teaching or training have you done in acoustics,

and to what groups--university, industry, trade
associations, civic groups, engineers, symposia?
233
Status
(1) Are you now an independent consultant? For how

many years? Full time or part time?
(2) If part time:
(2.1) Who is your chief employer or in what other

business ventures are you involved?
(2.2) Is your employer aware and does he approve

of your part time activity as an acoustical
consultant?
(2.3) May we contact your employer concerning you?
(2.4) What restrictions does your employer place

on you as a part time acoustical consultant?
(3) Are you associated with the manufacture or sale of

a product that could create a conflict of interest
in your activities as an acoustical consultant?
Professional Affiliations
(1) Of what engineering or scientific societies or

associations are you a member? Representative
ones are the Acoustical Society of America, the
Institute of Noise Control Engineering, and the
National Council of Acoustical Consultants.
(2) What is your present grade of membership and length

of time in that grade, for each association?
(3) Have you been accorded any professional honors ln

these associations such as offices, committee
chairmanships, awards, or prizes?
(4) Are you a registered professional engineer? In

what states? In what disciplines?
(5) Of what professional engineer associations are

you or your firm a member?
(6) Of what trade associations, chambers of commerce,

or similar business groups are you or your firm a
member?
234
Special Capabilities
(1) In what areas of acoustics do you specialize?
Noise measurement and control

Architectural acoustics
Hearing conservation
Shock and vibration measurement and control
Nondestructive testing
Medical ultrasonics
Underwater acoustics
(2) What equipment do you have for conducting

acoustical measurements in the field? In the
laboratory?
(3) With what national standards do you comply in

conducting your acoustical measurements?
(4) Are you listed by any governmental or trade

association body as an acceptable or certified
acoustical test laboratory?
(5) What equipment do you have for the absolute

calibration of test apparatus?
(6) Can you serve as an expert witness, either for

your client or as a friend of the court? What
experience have you had?
Business Practices
(1) Please indicate your fee structure. Do you handle

this by hourly charges, estimates for total job,
retainer charges~ or all of these? (1974 hourly
fees vary from $30 to $60, depending on the experi-
ence and training of the consultant.)
(2) If you use a contract form, please supply a sample.
(3) In your charges, how do you treat such expenses as

travel, subsistence, shipping, report reproduction,
and computer time? Consultants usually charge to
you the time spent during travel for you on Monday
through Friday, 8:00 a.m. to 5:00 p.m. There may
be a charge for use of highly specialized and ex-
pensive equipment.
(4) What insurance and bonding do you have?
235
(5) Are you operating as an individual, partnership,
or corporation?
(6) What statements do you have in your contracts

covering commercial security, liability, patent
rights?
(7) What restriction is there on the use of your name

in our reports, in litigation, in advertisements?
(8) What is the character and extent of reports that

you prepare?
(9) What facilities do you have for producing design

shop drawings on devices that you may develop for
the specific purposes of a consulting task?
(10) Do you have branch offices? Where?
(11) What size is your staff? What are their qualifi-

cations? Who will be working on this project?
The Proposal
Once you have selected a consultant, you can arrange to obtain

his services in several ways. With most professional people, a
verbal commitment is sometimes all that is necessary. However, you
may wish to request a written proposal that spells out the steps to
be taken in the solution of your problem.
Often, in a larger job, proposals from several points of view

are evaluated and used as one of the bases for the final selection
of the consultant. In this case answers to pertinent questions in
the preceding section may be sought in the proposal rather than in
the interview. If so, evaluation of the proposal from this point of
view is self-evident from the above discussion. If the questions
you are interested in are not answered to your satisfaction, don't
hesitate to ask for further clarification. In the discussion below,
we are concerned with the section of the proposal that outlines the
consultant's approach to your problem.
Aside from background qualifications of the consultant, the

proposal should answer the questions:
(1) How much is the service going to cost? Smaller

jobs are often bid on an hourly basis, with a mini-
mum commonly specified of one-half day's work, plus
direct expenses. Larger jobs are usually bid at
a fixed amount, based on the work steps described.
236
( 2) What is the consultant going to do? The answer to
this question may range all the way from a simple
agreement to study the problem to a comprehensive
step-by-step plan to solve it.
(3) What will be the end result? The answer to this

question is all too often not clearly understood;
the result is usually a report that specifies the
consultant's recommendation. If you do not want
to pay for the preparation of a written report, and
a verbal one will do, specify this in advance. Since
recommendations often call for construction to be
carried out by others, whose work is not subject to
the consultant's control, results can usually not be
guaranteed. Rather, an estimate of the noise reduc-
tion to be attained is all that can be expected.
If the consultant is to provide drawings from which
the contractor will work, one must specify sketches
or finished drawings. Generally, sketches are suf-
ficient. If special materials are required, the con-
sultant should agree to specify alternative selec-
tions if possible. If you want a guaranteed result,
experimental work will usually be necessary.
In the case of a proposal to quiet machine noise, the proposal,

if detailed, will probably callout the following steps:
(1) Determine the daily noise dose, so that the amount

of reduction required is known.
(2) From diagnostic measurements, determine the location

and relative strength of the major noise sources on
the machine in question, all other competing noise
sources being more than 10 dB below the intended
noise.
(3) Design preliminary noise control means; discuss

design with production people for pos~ible inter-
ference with access to the machine.
(4) Prepare and submit final recommendations in a report,

with construction data.
(5) In a post-report conference, resolve any questions

or compromises; submit memorandum of conference.
(6) If experimental work is needed, it can be added

between (3) and (4) above.
237
Other Services
If you wish, the consultant can also, as additional services,

provide monitoring of construction to determine compliance with
specifications. The consultant can also make post-installation
measurements to confirm predictions and supply oral briefings as
needed.
By working with the consultant during his measurements, you

can learn a great deal about how"\:o handle the special situation
for which he has been retained. However, he brings to the job an
instrument that is most difficult to reproduce: ears trained to
listen and to guide the use of the physical instruments. It takes
much practice and not a little aptitude to achieve this condition.
This aspect of a consultant's expertise is most difficult to re-
place.
If the consultant is to serve as an expert witness for you,

you will find that he is not automatically on your side. Rather,
he is more like a friend of the court, devoted to bringing out the
facts he has developed, with careful separation of fact from ex-
pert opinion. Complete frankness is needed if you want to avoid
unpleasant surprises. For example, the consultant may be asked by
the opposing attorney for a copy of his report to you. Thus, this
report should be prepared with this eventuality in mind.
If the consultant is retained to develop a quieter machine

for you, there should be a meeting of minds on handling of patent
rights. Ordinarily the patent is assigned to the client, with per-
haps a royalty arrangement for the inventor. .
For many situations the consultant will need photographs and

plans of machines and shop layout to facilitate his evaluation.
Permission to obtain these data can be handled in a manner con-
sistent with your industrial security system. A qualified consul-
tant will not have to be told to regard this materials as private,
not to be divulged to others without your prior consent. If you
regard him as the professionai person he is, your association can
be fruitful to all concerned.
238
Chapter 7
ANNOTATED REFERENCES
B.l American Industrial Hygiene Association. Industrial Noise

Manual, Second Edition. Detroit, MI, American Industrial
Hygiene Association, 1966. 171 p. Although ~he instrument
section is outdated, the described measurement techniques are
still applicable. Much data are given on ear plugs and muffs.
The chapter on engineering control is very practical; it is
copiously illustrated and describes many useful techniques.
A most valuable section on examples presents compact, illus-
trated case histories in which the noise reduction obtained
is given, usually with octave band spectra.
B.2 American Foundrymen's Society. Control of Noise, Third Edi-

tion. Des Plaines, IL, American Foundrymen's Society, 1972.
67 p. The engineering section of this manual was prepared
by an experienced consultant. It is written without equa-
tions, but with many charts, graphs, and tables. Although
the many examples are taken from foundry technology, the con-
trol techniques are quite general in application. There are
many compact case histories, together with data on the noise
reductions obtained. The point of view is very practical.
B.3 American Petroleum Institute. Guidelines to Noise. American

Petroleum Institute. Washington, DC. Medical Research Re-
port EA 7301. 1973. III p. This commissioned report sum-
marizes measurement instruments and procedures, explicit
noise reduction techniques, new plant design for low noise,
and source characteristics. The appendices have detailed
information on noise control materials, levels from machinery,
and addresses of suppliers of noise control materinls.
B.4 Bell, L. H. Fundamentals of Industrial Noise Control.

Trumbull, CT, Harmony Publications, 1973. 250 p. This prac-
tical book is written from the point of view of the practic-
ing noise control engineer/consultant. A minimum amount of
mathematics is used; many examples and exercises are given.
The chapters on enclosures, fans, gears, silencers, and vibra-
tion control are quite useful. A feature of the book is the
compact case histories, for which photographs and drawings
amply describe the techniques used.
B.5 Beranek, L. L. (ed.). Noise and Vibra+ion Control. New York,

NY, McGraw-Hill, 1971. 630 p. This is the major modern ref-
erence source for the noise control engineer. The treatment
is often mathematical, but there are plenty of illustrative
worked-out problems. Especially useful are the treatments of
239
transmission loss of simple and complex panels; mufflers and
silencers; sound in rooms; vibration isolation; and sound
power measurement.
B.6 Diehl, G. M. Machinery Acoustics. New York, NY, John Wiley

& Sons, 1973. 204 p. The chief contribution of this book is
in a detailed description of practical techniques, backed by
analysis, for the in situ measurements required for calculat-
ing sound power. Every professional noise control engineer
should be aware of these techniques. The sections on noise
sources and reduction procedures have a great deal of directly
useful information, especially for enclosure design.
B.7 Harris, C~ M. (ed.). Handbook of Noise Control. New York, NY,

McGraw-Hill, 1957. 985 p. Although old, this is still the
fundamental reference handbook for the noise control engineer.
Of particular interest are these sections: 13, vibration iso-
lation; 14, vibration damping; 21, acoustical filters and
mufflers; 23, gear noise; 24, bearing noise; 25, fan noise;
26, noise in water and stearn systems; 27, heating and venti-
lating system noise; and 30, electric motor and generator noise.
Of course, recent developments in acoustical materials and
measuring equipment are missing, but the fundamentals are un-
changed.
B.8 King, A. J. The Measurement and Suppression of Noise. London,

Chapman and Hall, 1965. 180 p. This British book is devoted
chiefly to noise from electrical machinery. Much attention is
paid to the design of duct silencers.
B.9 Peterson, A.P.G., and E. E. Gross. Handbook of Noise Measure-

ment. General Radio Co., 1972. 317 P. This is an excellent
source of data on measurement of sound pressure and calculation
of sound power levels. Valuable details are given on sound
analysis techniques, characteristics of many types of acousti-
cal instruments, and a summary of noise reduction procedures.
An especially useful section covers precautions to be observed
to ensure that valid data are acquired.
B.IO Noise and Its Control. Pollution Engineering Magazine.
Reprint of articles, 1973. This very readable set of arti-
cles summarizes characteristics of machine noise sources and
noise control techniques. It will provide a general back-
ground to the problems.
B.ll Thumann, A., and R. K. Miller. Secrets of Noise Control.

Atlanta, FairmontPresss, 1974. 256 p. This book presents
much practical noise control information in graphs and tables,
with a minimum of mathematics. Especially useful are data on
cost estimating, a listing of suppliers of noise control prod-
ucts, means of source location, silencers, and check lists for
management of noise control. There are many useful worked-
out problems. A comprehensive list is supplied for all the
standard methods of measurement that a professional noise con-
trol engineer should use.
240
B.12 Yerges, L. F. Sound, Noise and Vibration Control. New York,
NY, 1969. Van Nostrand Reinhold, 1969. 203 p. This practi-
cal book has practically no mathematics and relies almost com-
pletely on tables, charts, and graphs for its data. The
author, an experienced acoustical ·consultant, provided a great
deal of directly useful information on materials selection,
noise characteristics of machinery, design of noise control
means, and translation of subjective reactions to noise ir~~0
causes and solutions.
Periodical References, Arranged by Topic
P.l Absorber, resonant

- Mikeska, E. E., and R. N. Lane. Measured Absorption Charac-
teristics of Resonant Absorbers Employing Perforated Panel
Facings. J. Acoust. Soc. Am. 28: 987, September 1956.
Gives results for many differen~configurations, tuning from
100 to 800 Hz.
P.2 Air moving equipment - general

- American Society of Heating, Air-Conditioning and Refriger-
ating Engineers Handbook. Chapter 33, "Sound and Vibration,"
1970. The definitive treatment of air conditioninq noise.
- Diehl, G. M. Think Quiet. Compressed Air Magazine. Reprint

of set of articles, 1971. This is a forerunner of the
author's book on machinery noise. The emphasis is on air
moving machinery.
- Kodaras, M. J. Suppression of Ventilating Noise. Noise Con-

trol. ~, (2): 42 (1958). Treatment of ducts, mechanical
rooms.
- Sanders, G. Noise Control in Air-Handling Systems. Sound/

Vibration. 1 (2): 8, February 1967. Descriptive. Covers
concepts of noise control in air conditioning, cooling
towers, gas turbines, blowers, internal combustion engines,
jet engines.
P.3 Brewing
- Melling, T. H. Noise in the Brewing Industry--the Sources,
Its Control. Proc. Inter-Noise 72: 313 (1972). Describes
bottling and keg line noise. Few data on control.
P.4 Compressor
Diehl, G. M. Stationary and Portable Air Compressors. Proc.
Inter-Noise 72: 154 (1972). The chief noise sources are
discussed thoroughly, and a few noise control suggestions
are offered.
241
P.5 Cooling tower
- Dyer, I., and L. N. Miller. Cooling Tower Noise. Noise Con-
trol. 5 (3): 44, May 1959. Most of noise from fans. Esti-
mation procedures given. No noise control suggestions.
- S~elbach, H., Jr., and F. M. Oran. What to do About Cooling-

Tower Noise. Sound. 2 (5): 32, September-October 1963.
Describes characteristics of the noise, gives a detailed
estimating procedure, and describes effect of intake and
discharge silencers.
P.6 Damping
- Miles, R. C. Steel-Viscoelastic Composites. Sound/Vibration.
7 (7): 27, July 1973. Describes a panel of two layers of
steel bonded by a lossy adhesive. Excellent damping is ob-
tained.
- Warnaka, G. E., et ale Structural Damping as a Technique for
Industrial Noise Control. Reprint from J. Am. Indus. Hyg.
Assn., January 1972. Comparison of noise radiated from damped
and undamped machine surfaces.
P.7 Diagnosis: locating .sources

- Baade, P. K. Identification of Noise Sources. Proc. Inter-
Noise 72 Tutorials, p. 98 (i972). Valuable and practical in-
formation on techniques for locating noise sources.
- Gilbrech, D. A., and R. C. Binder. Portable Instrument for

Locating Noise Sources in Mechanical Equipment. J. Acous.
Soc. Am. 30: 842, September 1958. Two microphones correlated
by a multiplier allow the direction of a sound source to be
found.
- Hodgson, T. H., and C. E. Ebbing. Diagnostic Tests for Locat-

ing Noise Source. Proc. Noise-Con 73: 266 (1973). Des-
cribes use of critical listening, noise contours, close-in
measurements, sound power estimating, vibration measurements,
time domain event analysis, cross-correlation, and coherence
measurement. Many illustrations, references.
- Miller, T. D. Machine Noise Analysis and Reduction. Sound/_

Vibration. 1 (3): 8, March 1967. Uses rotational perio-
dicity to locate sources in bearings, electrical equipment,
fans. Briefly considers control means.
- Porter, M. A., and J. Q. Delap. Discrete Source Identification

in the Presence of High Ambient Noise Levels. Proc. Inter-
Noise 72: 409 (1972). Uses damped, absorbent-lined tube in
front of sound level meter to reduce pickup of undesired sound
for close-in measurements.
P.B Enclosures, panels

- Nordby, K. S. Attenuation of Non-Metallic Panels. Proc.
Inter-Noise 72: 127 (1972). For moderate amounts of trans-
mission loss, light weight panels can suffice. Paper describes
characteristics of plastics, foams, aluminum, safety glass.
242
P.9 Fans
Carrier. Sound and the Centrifugal Fan. 1969. This engineer-
ingguide provides alignment charts for estimating octave band
levels from centrifugal fans.
- Grahanl, J. B. Acoustical Considerations for the Selection of

Fans for New Facilities. Proc. Noise-Con 73: 343 (1973).
Gives data on effect on noise of fan type, variable inlet
vanes, blade angle, and operating point on fan characteristics.
P.IO Gas turbine

- Tatge, R. B. Noise Control of Gas Turbine Power Plants. Sound/
Vibration. 7 (6): 23, June 1973. Defines NEMA noise limit
curves, noise reduction for regenerator equipment, use of silen-
cers.
P.ll Gears
- Dunlap, T. A., and W. G. Halvorsen. Transmission Noise Reduc-
tion. SAE Paper 720735. September 1972. Discusses measure-
ments of bending and torsional compliance, and noise reduction
by detuning drive train components, decoupling housing areas,
and housing damping.
- Route, W. D. Gear Design for Noise Reduction. SAE Paper 208E·

June 1960. Discusses influence of contact ratio, pre.ssure
angle, angles of recess and approach, dimension control of
mounting, and tooth errors.
P.12 General
- Crocker, M. J. Noise Control Approaches. Proc. Inter-Noise
72 Tutorial, p. 17 (1972). Excellent summary of procedures.
- Cudworth, A. L. Field and Laboratory Examples of Industrial

Noise Control. Noise Control. 5 (1): 39, January 1959.
Case histories on ejector chutes~ tote boxes, chipping hammer,
air nozzles, and air compressors.
- Engstrom, J. R. Noise Reduction by Covers. Noise Control.

1 (2): 19, March 1955. Case histories on multislide punch
press, riveting machine.
- Engstrom, J. R. Riveting Hammer and Paper Shredder. Noise

Control. 2 (2): 18, March 1956. Case histories on riveting
hammer, paper shredder.
- Handley, J. M. Noise--The Third Pollution. lAC Bulletin

6.0011.0, 1973. Case histories from Industrial Acoustics
on cut-off saw, looms, punch press, power transformer, process
steamboiler.
- Ihde, W. M. The Application of Engineering Noise Control
Measures. National Safety News. 107 (6): 85, June 1973.
This article provides an intensely practical summary of charac-
teristics of a wide variety of noise sources and of procedures
for noise control. There are many illustrations and examples.
243
- Yerges, L. F. Methods of Noise Control for Machinery Already
Installed. Proc. Noise-Con 73: 376 (1973). Excellent and
compact summary of procedures, costs. Case histories on
grinders, automatic screw machines.
- Young, R. L. Practical Examples of Industrial Noise Control.

Noise Control. 4 (2): 11, March 1958. Case histories on
chain drives, pinch rollers, tumbling barrels, vacuum pumps,
supercharger test stand, boiler fabricating operations.
P.13 Hydraulic pumps

- Dean, R. C. Origins of Noise in Hydraulic Pumps. Proc.
Inter-Noise 72: 249 (1972.). Discusses chief noise sources in
lobe and lug type pumps, and how to interpret measurements and
inlet conditions.
- Miller, J. E. Silencing the Noisy Hydraulic System. Machine
Design: June 14, 1973, p. 138-143.
- Parker Hannifin Co. Reducing the Operatlng Noise of Industrial
Hydraulic Systems. 1972. Exhaustive and informative treatment,
with excellent list of noise reduction techniques.
P.14 Materials, composite

- Cavanaugh, w. V. Composite Materials for Noise Reduction.
Proc. Inter-Noise 72 Tutorial, p. 59 (1972). Discussion of
many materials, advantages and shortcomings.
- Miller, R. K. Acoustical Materials for the Food Process Indus-

, try. Proc. Noise-Con 73: 519 (1973). GiveB noise levels,
sanitary constraints on absorption materials, and isolator
selection.
P.15 Pipes
- Dear, T. A. Noise Reduction Properties of Selected Pipe
Covering Configurations. Proc. Inter-Noise 72: 138 (1972).
Exhaustive study of many pipe lagging systems for reducing
r~diating noise, under controlled conditions. Excellent
reference.
P.16 Product noise

- Guinter, J. J. Action Plan for Product Noise Reduction.
Sound/Vibration. 6 (12): 18, December 1972. Outlines
procedures.
P.17 Pumps, centrifugal

- Evans, L. M." Control of Vibration and Noise from Centrifugal
Pumps. Noise Control. 4 (1): 28, January 1958. Emphasizes
selection and proper operating point.
P.18 Punch press

- Allen, C. H., and R. C. Ison. A Practical Approach to Punch
Press Quieting. Noise Control Engineering: Vol. 3, No. 18, 1974
244
- Bruce, R. B. Noise Control of Metal Stamping Operations.
Sound/Vibration. 5 (11): 41, November 1971. Absorption,
partial and total enclosures are discussed. Control at the
source is also considered.
- Shinaishin, O. A. Punch Press Diagnostics and Noise Control.

Proc. Inter-Noise 72: 243 (1972). Suggests changes in die,
stock, snubbers, perforated plates, enclosures.
P.19 Refinery noise

- DeBiase, J. L. Criteria and Design Specifications for Plant
Noise Control. Sound/Vibration. 6 (9): 26, September 1972.
Good listing of sources, control procedures. Also where to
measure OSHA levels.
P.20 Screw machine

- Schweitzer, B. V. Silent Stock Tube for Automatic Screw
Machines. Noise Control. 2 (2): 14, March 1956. Describes
development use of stock'tubes with damping.
P.2l Silencers
- Sanders, G. J. Silencers: Their Design and Applications.
Sound/Vibration. 2 (2): 6, February 1968. Describes types,
and attenuation/frequency characteristics. Useful data on
materials resistant to high temperatures, flow velocities.
P.22 Sound power measurement

- Huebner, G. Sound Power Determination of Machines in situ.
Proc. Inter-Noise 72: 381 (1972). Discusses errors from
being in near field, from finite nUmber of measurements, from
measurement error, and from room reflections. Latter unim-
portant if area ratio (total absorption in room)/(area of
measuring surfaces) is greater than 10.
P.23 Stack noise

- Hersh, A. S., et ale Refinery Auxiliary Stack Redesign to
Eliminate Induced Resonant Tone. Proc. Noise-Con 73: 441
(1973). Many small changes, including avoiding turbulence,
are described.
P.24 Valve noise

- Heymann, F. J. Some Experiments Concerning Control Valve
Noise. Proc. Noise-Con 73: 382 (1973). Several plug and
seat configurations are described. Much useful data on get-
ting data from which to predict noise from scale models.
P.25 Vibration and radiated power

- Chan, C.M.P., and D. Anderson. Correlation of Machine Struc-
ture Surface Vibration and Radiated Noise. Proc. Inter-
Noise 72: 261 (1972). Mean square velocity or acceleration
correlates well with sound power.
245
P.26 Vibration and shock isolation
- Miller, H. T., et all. Practical Design of Machinery Founda-
tions for Vibration and Noise Control. Proc. Inter-Noise 72:
185 (1972). Discusses real life parameters that must be
considered.
P.27 Woodworking industry

- Christman, R. P., et al. Sound Pressure Levels in the Wood
Products Industry. Noise Control. 2 (5): 33, September
1956. Octave spectra for many machines.
Additional specific references used in the case histories are

included as footnotes to the specific case.
246 '" u.s. GO'IIRIIIIDlT PRIIlJUlG OfFICE, 1911- 757 -0 57/6 739

Manual de Ruido Industrial

Uploaded by

Copyright:

Available Formats

Manual de Ruido Industrial

Uploaded by

Document Information

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

Manual de Ruido Industrial

Uploaded by

Copyright:

Available Formats

What is the purpose of the manual?

What is the purpose of the manual?

What topics are covered in the manual?

What topics are covered in the manual?

INDUSTRIAL NOISE CONTROL MANUAL

Industrial Noise Services, Inc.

Contract No. HSM 99-73-82

U. S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE

DHEW Publication No. (NIOSH) 75-183

NIOSH Project Officer: Roy M. Fleming

' •. Authorlsl • 8. Performing Organization Repl.

9. Performing Organizalion Name and Address 10. Project/Task/Work Unit No.

Industrial Noise Services, Inc. 11. ContraCt/GraD I No.

This manual contains fundamental information to aid the user in understanding,

17. Key WOlds and Documenl Analysis. 170. Descriplors

17e. COSATI Field 'Group 06/J

This manual contains fundamental information to aid the user

Following passage of the Occupational Safety and Health Act

Accordingly, a contract was awarded for the development of a

It was recognized that noise control is not a new discipline,

~reced;ng page blank!

In planning~or future supplemental sections to the case

Case History Format:

A. Description of the process, machine, and noise problem.

B. Noise measurements made and discussion of findings.

C. Control approaches -- advantages and disadvantages.

D. Results in terms of noise reduction achieved and cost.

E. Pitfalls to avoid in implementing the control methods.

F. Figures -- noise data (e.g., octave band levels).

G. Sketches of area layout, machine-operator relationship,

H. Photographs of machines before and after modifications

Case Histories can be sent to:

Noise Section/Physical Agents Branch

CHAPTER 1 FUNDAMENTALS OF SOUND 1

CHAPTER 2 MEASUREt~NT OF NOISE 15

CHAPTER 3 NOISE CONTROL TECHNIQUES 33

CHAPTER 4 NOISE CONTROL MATERIALS 63

CHAPTER 5 CASE HISTORIES 75

Case History Data 75

CHAPTER 6 SELECTING AND USING A CONSULTANT 231

Knowing When a Consultant is Needed 231

CHAPTER 7 ANNOTATED REFERENCES 239

1.1 Frequency-wavelength chart 3

1.1 Corrections from flat levels to A-levels 7

1.2 Method for approximating decibel combinations 9

1.3 Example of level combining 10

2.1 Sound level meters and accessories available 21

3.1 Required static deflection for common

3.2 A-level noise reduction by use of heavier pipe 42

4.1 Sound absorption coefficients of common

4.2 Sound absorption coefficients of common

4.3 Transmission loss of common materials 68

5.30.1 Sources of noise and methods of noise

Sound is transmitted through any elastic medium (gas, solid,

The visual wave motion of the spring corresponds to the

Definitions and Relationships

We are basically interested in four parameters descriptive of

In many noise control problems you will be concerned with both

Wavelength is the distance between similar configurations or

10 20 50 100 200 500 1000 2000 5000 10,000