Acoustics  Leo Beranek
0%(1)0% encontró este documento útil (1 voto)
273 vistas255 páginasDescripción:
Libro Recomendado!
Fecha en que fue cargado
Apr 15, 2016
Derechos de autor
© © All Rights Reserved
Formatos disponibles
PDF o lea en línea desde Scribd
Compartir este documento
Compartir o incrustar documentos
¿Le pareció útil este documento?
Descripción:
Libro Recomendado!
Copyright:
© All Rights Reserved
Formatos disponibles
Descargue como PDF o lea en línea desde Scribd
0%(1)0% encontró este documento útil (1 voto)
273 vistas255 páginasAcoustics  Leo Beranek
Descripción:
Libro Recomendado!
Copyright:
© All Rights Reserved
Formatos disponibles
Descargue como PDF o lea en línea desde Scribd
ACOUSTICSACOUSTICS
Leo L. Beranek
1984 Edition
Acoustic Laboratory
Massachusetts Institute of Technology
Bolt Beranek and Newman, Inc.
1993 Edition
975 Memorial Drive, Suite 804
Cambridge, MA 02138Library of Congress Catalog Card Number: 8670671
International Standard Book Number: 088318494X
Copyright©1954, 1986, 1990, 1993, 1996,
by the Acoustical Society of America
All rights reserved. No part ofthis publication may be
eproduced, stored in a retrieval system, or transmitted,
in any form or by any means (electronic, mechanical,
photocopying, recording, or otherwise) without the
prior written permission of the publisher.
Published by the Acoustical Society of America
though the American Institute of Physics, Ine
'500 Sunnyside Blvd, Woodbury, New York 11797
Printed in the United States of America
PREFACE
Acoustics isa most fascinating subject. Music, architecture, engineer
ing, science, drama, medicine, psychology, and linguistics all seek from it
answers to basic questions in their fields. In the Acoustics Laboratory
at MLLT. students may be found working on such diversified problems
as auditorium and studio design, loudspeaker design, subjective percop
tion of complex sounds, production of synthetic speech, propagation of
sound in the atmosphere, dispersion of sound in liquids, reduction of noise
from jetaireraft engines, and ultrasonic detection of brain tumors. The
annual meetings of the Acoustical Society of America are veritable five
ring shows, with papers and symposia on subjects in all the abovenamed
fields, Opportunites for employment are abundant today because man
‘agement in industry has recognized the important contributions that
acoustics makes both to the improvement of their products and to the
betterment of employee working conditions.
There is no easy road to an understanding of presentday acoustics.
First the student must acquire the vocabulary that is peculiar to the
subject. Then he must assimilate the Iaws governing sound propagation
and sound radiation, resonance, and the behavior of transducers in an
Acoustic medium, Last, but certainly not of least importance, he must
Jearn to understand the hearing characteristics of people and the reac
tions of listeners to sounds and noises.
‘This book is the outgrowth of course in acoustics that the author
has taught to seniors and to firstyear graduate students in electrical
engineering and communication physies. The basic wave equation and
some of its more interesting solutions are discussed in detail in the first
part of the text. The radiation of sound, components of acoustical sys
tems, microphones, loudspeakers, and horns are treated in sufficient detail
to allow the serious student to enter into electroacoustic design.
‘There is an extensive treatment of such important problems as sound
in enclosures, methods for noise reduction, hearing, speech intelligibility,
and psychoacoustic criteria for comfort, for satisfactory speech intelligi
Dility, and for pleasant listening conditions
The book differs in one important respect from conventional texts on
acoustics in that it emphasizes the practical application of electrical
circuit theory in the solution of a wide variety of problems. Wherever
possible, the background of the electrical engineer and the communica
tion physicists utilized in explaining acoustical conceptsPREFACE
‘The highfidelity expert will find the chapters onloudapesker enclosures,
horus, and rooms particularly interesting because they show how the per
formance of loudspeakers either in baffles oF attached to horns may be
accurately and simply calculated. ‘These chaptere alco ilustrate the
necessity of considering in design the overall system, including the
amplifier, the loudspeaker, the baflle or horn and considering also the
room in which they are to be operated. Numerical examples and sum
rary charts are given to facilitate application ofthis material to music
reproduction systems,
Tn view of the incressed interest in noise control, the author has
pt this subject in mind in waiting Chapters 1, 2, 4, and 10 to 18.
‘These chapters served as the basis of a special summer program on
noise reduction at MILT. in 1958. ‘The material of Chapters 11 and
1 is new, and it is hoped that it will be of value to those interested in
noise and its effect on human beings,
Tn short, the engineer or scientist who wishes to practice in the field of
coustios and who does not intend to confine his efforts to theoretical
tatters must know the material of this text
Problems for each chapter are included at the end ofthe text for use by
thestudent. References to collateral reading in English are given in the
text, although no attempt has been made to give a bibliography of the
primary sources of material. Suggestions to instructors for best use of
the text aro given immodintely after this preface.
‘The author wishes to express his deep appreciation to Francis M.
Wiener and Rudolph H. Nichols, Jr for their assistance in the detailed
review and editing of the text and the preparation of some original
material. Many members of the Acoustics Laboratory at M.L:T. have
read one or more chapters and have given valuable assistance to the
author. Of these, particular mention is made of Mary Anne Summer
field, Walter A. osenblith, Kenneth N. Stevens, Jerome R. Cox,
Jordan J. Baruch, Joanne J. English, and Norman Doeling
‘The illustrations are due to the highly capable and untiring efforts of
Clare Twardsik. ‘The author is deeply indebted to his typist, Eliasbeth
H. Jones, to his secretary, Lydia Bonazzoli, and to his wife, Phylis, who
‘made it possible for him to complete the text within a reasonably short
span of time.
Leo L. Beranex
SUGGESTIONS FOR INSTRUCTORS
‘This text is divided into thirteen chapters, comprising thirtytwo parts.
Bach patt is intended to he approximately 1 week's work, although this
will vary among students owing to differences in their previous training.
If the entire class expects to take a full year of acoustics, the parts
should be taught in sequence, with the exception of Part XXVIII, Meas
urement of Acoustic Levels, which may be referred to in associated labor
ratory experiments and demonstrations throughout the course. If only
‘8 purt of the eluss plans to continue throngh both terms, thefundamental
‘material should be taught in the first term and the more applied material
in the second, One suggested division, in this case, is as follows:
Birst Term Second Term
Part I. Introduetion Part V. Energy Density
‘and Intensity
Part I. ‘Terminology
Part IX. Circuit’ Theorems,
Part HI. ‘The Wave Bqua
ergy and Power
Part XI. Diroctivity Index
Part IV tions of the and Directivity
Wave Equation Factor
Part VI. Mochanieal Cire Patt XIV, General Characters
ae isties of Micro
— phones
Pare VIt. Acoustical Part XV. Pressure
gin) Microphones,
Part VIL “Pransducers Part XVI. Gradient and Com
Part X.Directivity bination Micro
Patterns phones
oe Part XVII. Design Factors Af
Part XL aiaion Impede leas aera
ve ator Loudspeakers
Part XIU. Acoustic Part XX. Bass Reflex Enclo
Hlements cae
Part XVI. Basie Theory of Part XI Hora Driving
Direotradiator Units
Lanulspeakers Part NXUL Hornsvill SUGGESTIONS FOR INSTRUCTORS
First Term Second Term,
Part XIX. Simple Enclosures Part XXV, Sound Transmis
sion through Walls
Part XXII, Sound Fields in between Enclosures
Small Regularly Part XXVI. Noise Control Pro
cedures and Noise
Shaped Brelonures ‘
Part XXIV. Sound Fields in Pitt XXVIT Acmatic Tanemi
Large Irregularly ‘at Seen
Shaped Bnelowures PAC XXIX. Reciprocity
of Transducers
Part XVII. Measurement of Par XXXI. Speech
Acoustic Levele Intelligiblity
Part XXXII. Psychoacoustie
Part XXX, Hearing Criteria
A course in acousties should be accompanied by a set of wellplanned
JIaboratory experiments. For example, the material of the first few chap
ters will be more significant if accompanied by a laboratory experiment,
(on noise measurement. ‘This will familiarize the student with the meas
urement of sound pressure and with the use of a frequency analyzer He
will appreciate more fully the meaning of sound pressure, sound intensity,
decibels, sound energy density, and power level; and he will understand
the accuracy with which noise can be measured,
A suggested minimum of 10 experiments, listed both numerically for a
year's course and by term, is as follows:
First Term Second Term
No. 1. Noise measurement, No. 3. Freefeld calibration of
‘microphones
No. 5. Design and testing of a
loudspeaker baffle
No. 8. Prediction and control of
noise in a ventilating
No.2. Measurement of the cone
stants of an electro
‘mechanical transducer
No.4. Measurement of freefield
response of a loudspeaker
system
No. 6. Study of sound fields ina No, 9, Audiometsic testing of
small rectangular hearing
enclosure No. 10. Application of psycho
acoustic eriteria in the
design of an auditorium,
No.7. Study of sound fields in a
large irregular enclosus
An assignment of two problems per week should provide sufficient
application of the material of the text, ‘The short list of problems for
each chapter should he supplemented by timely problems derived from
the instructor's experience,
CONTENTS
PREFACE
SUGGESTIONS FOR INSTRUCTORS
CHAPTER 1. INTRODUCTION AND TERMINOLOGY
rant I Introduction
pane IL Terminology
CHAPTER 2. THE WAVE EQUATION AND SOLUTIONS
vant IIL The Wore Equation
vant IV Solutions of the Wave Equation
pant V_ Energy Density and Intenaty
CHAPTER 3. ELECTROMECHANOACOUSTICAT. CIRCUITS
Pant VI" Mechanieal Cirewite
pat VIL Acoustical Cireuite
part VIII Transducers
rant IX Circuit Theorems, Bnergy, and Power
CHAPTER 4, RADIATION OF SOUND
vane X Directiity Patterns
vant XI Directivity Index and Directivty Factor
CHAPTER 5. ACOUSTIC COMPONENTS
pax XIT Rodiation Impedances
vant XIII Acoustic Blomenta
CHAPTER 6, MICROPHONES
rant XIV. General Characeritice of Microphones
pane XV Preseure Microphones
vant XVI Gradiont and Combination Microphones
CHAPTER 7, DIRECTRADIATOR LOUDSPEAKERS
vant XVII Basic Theory of Divectradtotor Loudspeakers
Pant XVIII Design Factors Affecting Direcradiator Lowdepeater
Performance
6
16
a
a
62
0
1
a
o
109
16
16
128,
14
a
150
18,
183
1832 CONTENTS
JAPTER 8, LOUDSPEAKER ENCLOSURES
vane XIN Simple Bnolonures
pant XX Baswreler Bnelosures
CHAPTER 9, HORN LOUDSPEAKERS.
pant XXE_ Horm Driving Wits
rawr XXIL Horns
CHAPTER 10. SOUND IN ENCLOSURES
ant XXII Sound Fields in Small Regularly Shaped Enclosures
pant XXIV Sound Fielda in Large Irregularly Shaped Bnelosures
rant XXV_ Sound Tronemiseion throuph Walls between Enclosures
CHAPTER 11, NOISE CONTROL
pant XXVI Procedures and Sources
vant XXVIL Acoustic Pronemission Pathe
CHAPTER 12, ACOUSTIC MEASUREMENTS
pane XXVIT Measurement of Acoustic Lele
vant XXIX Reciprocity Calibration of Transducers
CHAPTER 13. HEARING, SPEECH INTELLIGIBILITY, AND
PSYCHOACOUSTIC CRITERIA
pane XXX Hearing
Pant XXX Speech Inteligibility
rant XXXII Poychoacoustic Criteria
PROBLEMS.
DIX J, DECIBEL CONVERSION ‘TABLES
APPENDIX Il, CONVERSION FACTORS
‘APPENDIX IL
APP!
INDEX
208
208
239
259
250
268
285
285
298
at
392
saz
45
361
361
317
388
388
406
az
431
464
469
479
PREFACE TO THE PAPERBACK EDITION
With the advent ofthe compact dise, with miniature highfidelity systems
ambulating everywhere, and with emphasis on combination voice and
 in'a medium Gee Fig: 2 where the
incremeefromlefciorahtatacpace rate  wound bremure» cheng in apece at 2
a op/or (re Pag 2) pace rato gradp = 128 4 522 4 2
where i,j and ae unit velar in the
and diectins, respectively, and
Pit fe remire ats pone
Assume that the sides of the box are completely frictionless; i.e, any
viscous drag between gas particles inside the box and those outside is
negligible. ‘Thus the only forces acting on the enclosed gas are due to
the pressures at the faces of the box.
The difference between the forces acting on the two sides of our tiny’
box of gas is equal to the rate at which the force changes with distance
times the ineremental length of the box:
Forge seting bo acoerat he box Inthe  Fore ating ta aelrte the box in the
positive x dretion = ~ (32.2) ay a  postive dieton = ~ [1 (222) ay ae
te)  4 (2249) see + (22 2) ae ay]
an)
Note that the positive gradient causes an acceleration of the box in the
negative direction of 2
+ Nonvector derivations of the wave equation are given in Rapleigh, “Theory of
Sound,” Vol. 2,pp. I15, Macmillan & Co.,[ih, London, 1806; P.M. Morse, “Vibra
tion and Sound," 2d ed, pp. 217225, MeCirawHill Book Company, Ine., New York,
1948; L. B. Kinsler and A. R. Frey, “Fundamentals of Acoustic” pp. 118137,
John Wiley & Sons, Ine, New York, 1950; R. W. B. Stephens and A. F, Bave, “Wave
Motion and Sound,” pp. 3213, 400406, Edward Amold & Co,, London, 1950; and
other pes,
A vector derivation of the wave equation ie given in two papers that must be read
fogethor: W.J. Cunningham, Application of Vector Analysis tothe Wave Equation,
J Acoust Soe. Amer, 22: 61 (1950); nd R. Vo Hartley, Note on “Application of
‘Vector Analysis to the Wave Equation,” J. Acous, So. Amer, 22: 511 (1950).18 THE WAVE EQUATION AND SOLUTIONS — (Chap. 2
f2 aw fener a
By Newton's aw, the foree per nit volume (/7) of Ba. (2) must be
equal tothe tine rie of change othe momentum prunit vlan othe
box. We have already assured tat our box i dorm packet
thatthe sas of fhe gas within it lays constants "That is
fo Neyo
% sa
a Dg
Vat ae
send FB an
average velocity of the gus  where qis the average vector velocity of
Jn the "box" in the direction, oi the  the gas in the box,” 9” is the average
space average of the instantaneous den  density of the gas in the box, and
sity of the gas inthe box, and MC = p'V  AF = p'V isthe total mass of the gas in
fi the otal mess ofthe gas in the bor,  the box, D/Dt ia not a simple partial
erivative but represents the total rate
of the change of the velocity of the par
ticular bit of gas in the box regardless of
its position,
Pa 29
ee)
where tg and gar the component
She set paride vest
hanes in density ofthe gas due  "iF te veto prt vlsty nema
nd wove isemal enough, then  enough, the stee! change of womestam
stares danty 9h appro  fe pres n the ean be apa
tnsnly ual tbe average dency,  indy the ate of change of omens
Thee, Samat dned pointy Dq/DP = 9/95 and
the instantaneous" ewity. pea be
throsinated bythe average dealy po
thn,
ae “at p e
2
(2a)
‘The approximations just given are generally acceptable provided the
sound pressure levels being considered are below about 110 db re 0.0002
microbar. Levels above 110 db are so large as to create hearing dis
comfort in many individuals, as we shall see in Chap. 13 at the end of this
book,
The Gas Law. If we assume an ideal gus, the CharlesBoyle gas law
applies to the box. It is
Py =RT es)
where P is the total pressure in the box, V is the volume equal to
Ar By 42, T is the absolute temperature in degrees Kelvin, and R is a
‘constant for the gas whose magnitude is dependent upon the mass of gas,
Pact 11 THE WAVE EQUATION 0
chosen. Using this equation, we can find a relation between the sound
pressure (excess pressure) and an incremental change in V for our box.
Before we can establish this relation, however, we must know how the
temperature 7’ varies with changes in P and V and, in particular, whether
the phenomenon is adiabatic or isothermal
‘At audible frequencies the wavelength of a sound is long compared with
the spacing between air molecules. For example, at 1000 eps, the
wavelength 4 equals 0.34 m, as compared with an intermolecular spacing
of 10° m, Now, whenever a portion of any gas is compressed rapidly,
its temperature rises, and, conversely, when it is expanded rapidly, its
temperature drops. At any one point in an alternating sound field,
therefore, the temperature rises and falls relative to the ambient, tem
perature. This variation oceurs at the same frequency as that of the
sound wave and is in phase with the sound pressure,
‘Let us assume, for the moment, that the sound wave has only one fre
quency. At points separated by onehalf wavelength, the pressure and
the temperature Quctuations will be 180° out of phase with each other.
Now the question arises, Is there sufficient time during onehalf an
alternation in the temperature for an exchange of heat to take place
between these two points of maximally different temperatures?
Tt has been established? that under normal atmospheric conditions the
speed of trave) of a thermal diffusion wave at 1000 eps is about 0.5 1n/se
‘and at 10,000 epsit is about 1.5 m/see. The time for onehalf an alterna
‘ion of 1000 eps is 0.0005 sec. In this time, the thermal wave travels a
distance of only 0.00025 m. This number is very small compared with
onehalf wavelength (0:17 m) at 1000 eps. At 10,000 eps the heat travels,
7.5 X 10*m, which is a small distance compared with a half wavelength
(1.7 X 10? m), It appears safe for us to conclude, therefore, that there
is negligible heat exchange in the wave in the audible frequency range.
Gaseous compressions and expansions of this type are said to be adiabatic,
For adiabatic expansions, the relation between the total pressure and
the volume is known to bet
PV? = constant 26)
where + is the ratio of the specific heat of the gas at constant pressure to
the specifie heat at constant volume for the gas. This equation is
130 4 mass of gas i choren so that ite weight in grams is equal to ite molecular
weight Genowa to chemists as the grammolecular weight, or the mole, then the
Yolume of thia mace st O°C and 0.76 m Hg is the sue for all gates and equals
0.02242 m?. ‘Then R= 8314 wattsee per degree centigrade per grammoleculae
Weight, If the mass of gas chosen ie times its molecular weight, thea e = 8.314
‘Soe LL, Beranck, "Acoustic Measurements,” p. 49, John Wiley & Sous, Tne,
Now York, 1940
"MW. Zemaasky, “leat and "Thermodynamics," 24 ed, pp. 104114, MoGran
Hil Book Company, Inc, New York, 1,EQUATION AND SOLUTIONS — (Chap. 2
20 THE WAVE
‘obtained from the gas law in the form of Eq. (2.5), assuming adiabatic
conditions. For air, hydrogen, nitrogen, and oxygen, é¢., gases with
diatomic molecules,
yal
Expressing Eq, (2.6) in differential form, we have
av an
Let
P=Petp V=Vetr (28)
where Ps and Vs are the undisturbed pressure and volume, respectively,
fand p and r are the incremental pressure and volume, respectively, owing
ry
ay
vagesioe as
ee opiet
Beara oo
_ o
Fis, 22. Change in volume of the box with change in position. From (a) snd (6)
I acen that the inesomentel change im volume of the box equals + = (96,/9¢) S
to the presence of the sound wave. Then, to the same approximation as
that made preceding Eq, (2.4) and because p « Po and 1 Vo,
RaF 29)
The time derivative of this equation gives
Ag72 ax
‘The Continuity Equation. The continuity equation is s mathematical
expression stating that the total mass of gas in a deformable “box” must
remain constant. Because of this law of conservation of mass, we are
able to write a unique relation between the time rate of change of the
ineremental velocities at the surfaces of the box.
aot 11 THE WAVE EQUATION a
Ea
See
a 
et ibae
‘The diflerence of the two qvantitios
above multiplied by the area ay 3z ives 
the increment in volume +
ot,
Warsyar ane  Yeave eam)
ved a1
Differentiating with respect to time  Diflrentiating with respict to. time
veld, yields,
132) 
a
Fa vodivg 2.30)
Where aj the instantincons particle  where it the instantaneous particle
velocity velocity
The Wave Equation in Rectangular Coordinates
Onedimensionol Derivation Threedimensional Devnation
"The onedimensional wave equation ig  The threedimensionsl wave equation
obtained by combining the equation of  ix obtained by combining the equation of
‘mation (24a), the gus las (2.10), and the  motion (240), the gas law (2.10), andthe
continuity equation (2.134), Combing  eontinuty equation (2.125). Combine:
tom of (210) and (2.14) gives tion of (210) and (219) gives
Dierotte (1 with pect o  Dierntate (2.4) wih rape to
wre eet) a
Piderevte a) with apes to  Ta the degen of ech sd of
% ace  2
. ~ di gat ~ mar 220822 THE WAVE EQUATION AND SOLUTIONS — [Chap.2
Onedimensional Derivation ‘hveeimensionat Derivation
Replacing the div (grad p) by Wp, wo get
vip = podiv ¥
tp = pediv 8 (a7)
interchangesbilty of the 2  where vis the operator called the Lapla
and £ derivatives, and combining (2.152)  cian. Combining (2.199) and (2.17), we
tnd (2162), we get ee
~ 52% ei v essay
Let us, by definition set,
we
eat (2.19
oe (2.19)
‘We shall see later that cis the speed of propagation of the sound wave in
the medium,
We obtain the onedimensional wave  We obtain the threedimensional wave
cenwation caution
re) vende aay
Jn rectangle coordinates
vip 2 4 OF 4 OP
pe ME MP any
We could als ave sited p and  We sould ao have eliminated p and
retained in whih ease we would have  rsttnedq, im whith ease we would have
533 (222%)
oy Law where 4g = grad (div q) when there is
a (2220)  Co rotation in the medium,
Equations (2.20) and (2.22) apply to sound waves of “small” magni
‘tude propagating in a sourcefree, homogeneous, isotropie, frictionless gas
at rest,
The Wave Equation in Spherical Coordinates, ‘The onedimensional
‘wave equations derived above are for planewave propagation along one
dimension of a rectangular coordinate system. In an anechoic (echo
free) chamber or in free space, we frequently wish to express mathe
‘matically the radiation of sound from a spherical (nondireetional) source
of sound. In this ease, the sound wave will expand as it travels away.
from the source, and the wave front alvrays will he a spherical surface.
To apply the wave equation to spherical waves, we must replace the
‘operators on the left side of Eqs. (2.20) and (2.22) by operators appro
priate to spherical coordinates,
Assuming equal radiation in all directions, the wave equation in one=
dimensional spherical coordinates is
ap
oe (223)
Pert IV] SOLUTIONS OF THE WAVE RQUATION 23
Simple differentiation will show that (2.23) can also be written
apr) _ 1 HG)
ar Bae Ce)
Itis interesting to note that this equation has exactly the same form as
Eq (2.204). Hence, the same formal solution will apply to either equa
tion except that the dependent variable is p(z,¢) in one case and pr(r,t) in
the other case.
‘Example 21. In the steady stat, that ia, 2u/2¢ = jou, determine matheratially
hhow the sound pressure in a plane progreetive sound wave fone
‘ould be determined fom ineasurement of partsle veloeity alone
Solution. From Bq. (24a) we find in the steady state that
where p and ware now rms
Lively,” Written in diferent
of the sound pressure and particle velocity, respme=
form,
1 the particle velocity is 1 em /e0, wit 1000 radians/ste, and ais 0 em, then
{8p = =30.005 X 1000 1.18 x a0
31059 newton mn?
‘We shall have an opportunity ia Chap, 6 of this test toate @ practical application
‘of these equations to the mesturement of particle vlocty hy a velocity microphone
pant IV Solutions of the Wave Equation
2.3. General Solutions of the Onedimensional Wave Equation, ‘The
‘onedimensional wave equation was derived with either sound pressure or
particle velocity as the dependent variable. Particle displacement, or the
variational density, may also be used as the dependent variable. ‘This
can be seen from Eqs. (2.4a) and (2.18a) and the conservation of mass,
which requires that the product of the density and the volume of a small
box of gas remain constant. That is,
2'V = poVo = constant, (2.25)
and so
pave = —Vae (2.26)
Let
p= pte (2.27)
where pis the incremental change in density. Then, approximately, from
Eqs. (2.8) and (2.26),
per = —Vep (2.28)a THE WAVE EQUATION AND SOLUTIONS —[Chep.2
Differentiating,
ar _ Vode
ee a asetl
so that, from Eq. (2.132),
2 au
et nese (2.29)
Also, we know that the particle velocity is the time rate of change of the
particle displacement.
watt e220)
Inspection of Eqs. (24a), (2.180), (2.29), and (2.30) shows that the
pressure, particle velocity, particle displacement, and variational density
are related to each other by derivatives and integrals in space and time.
‘These operations performed on the wave equation do not change the form
of the solution, as we shall see shortly. Since the form of the solution is
not changed, the same wave equation may be used for determining
density, displacement, or particle velocity a well as sound pressure by
substituting p, or &, or u for p in Eq. (2.20a) o p, &, oF q for p in
(2.200), assuming, of course, that there is no rotation in the medium,
General Solution. With pressure as the dependent vatiable, the wave
‘equuataon 38
ap _ lap
Bet ae
(2a)
‘The gencral solution to this equation is a sum of two terms,
ptan(+3) 82)
where fi and fs are arbitrary functions, We assume only that they have
continuous derivatives of the first and second order. Note that hecause
tand x occur together, the first derivatives with respect to z and t are
exactly the same except for a factor of +e.
‘The ratio z/c must have the dimensions of time, so that e is a speed.
From ¢ = yPe/pe (Eq. (2.19)] we find that
108\"
18,
in airat an ambient pressure of 10° newtons/im? and at 22°C. ‘This quan
lity is nearly the same as the experimentally determined value of the
speed of sound 344.8 [see Eq. (1.8)], 80 that. we recognize c as the speed at
which a sound wave is propagated through the ait,
From the general solution to the wave equation
‘observe two very important facts:
ven in Eq, (2.82) we
Pat IV] SOLUTIONS OF THE WAVE EQUATION 25
1. The sound pressure at any point x in space can be separated into
‘two components: an outgoing wave, f(t — z/c), and a backwardtraveling,
wave, fall + 2/0)
2, Regardless of the shape of the outwardgoing wave (or of the back
wardtraveling wave), it is propagated without change of shape. To
show this, let us assume that, at ¢ = f,, the sound pressure at x — 0 is
f(t). At a time ¢ = ty + t the sound wave will have traveled a dis
tance x equal to tye m. At this new time the sound pressure is equal to
= fills + te ~ bec/e) = full). In other words the sound pressure has
ropagated without change. ‘The same argument can be made for the
backwardtraveling wave which goes in the —z direction.
It must be understood that inherent in Eqs. (2.31) and (2.92) are two
assumptions. First, the wave is a plane wave, i, it does not expand
laterally. ‘Thus the sound pressure is not a function of the y and
coordinates but is a function of distance only along the x coordinate.
Socond, it is assumed that there are no losses of dispersion (scattering of
the wave by turbulence or temperature gradients, et.) in the air, so that.
the wave does not lose energy as it is propagated. Dissipative and dis
persive cases are not treated analytically in this book, but are discussed
briefly in Chaps. 10 and 11
Steadystate Solution. In nearly all the studice that we make in this
‘text we are concerned with the steady state. As is well known from the
theory of Fourier series, a steadystate wave can be represented by a
Jinear summation of sinewave functions, each of which is of the form
VO = V2 orl cos (st + 6 (2.330,
For example, if ys is sound pressure, we write
2 = Spt) = Y V2 [pel 08 (ot + 0.) (2.380
where w, = 2rf,;f, = frequency of vibration of the »th component of the
wave; 9, is the phase angle of it; and /2 ¢ (or p") is the peak
magnitude of the component. Because the wave is propagated without
change of shape, we need consider, in the steady state, only those solutions
to the wave equation for which the time dependence ateach point in space
Js sinusoidal and which have the same angular frequencies w, as the source
Borrowing from electsicalcieuit theory, we represent & sinusoidal
funtion with a frequency w hy the real part of a complex exponential
function, Thus, ata fixed point in space 2, we have the sound pressure,
YG) = VIRe lolze"] (2.340)
PUz,t) = V2 Re [plz )e (2.340)26 INE WAVE EQUATION AND SOLUTIONS — (Chap. 2
where p(x) is a complex function (fe, it hes a real and an imaginary part)
that gives the dependence of p on z. The product of +/2 times the
magnitude of p(z) is the peak value of the sinusoidal sound pressure
function at z, ‘The phuse angle of p(z) is the phase shift measured from
some reference position, Generally we omit writing Re although it
always must be remembered that the real part must be taken when using
the final expression for the sound pressure, In the steady state, there
fore, we may replace fs and fs of Eq. (2.82) by a sum of functions each
having a particular angular driving frequency Ww, s0 that
PD = Y pales) = Y VERE lipyremore + promenjeee] (2.35)
‘The part of Eq. (2.85) within the brackets is the same as that within the
brackets of Eq. (2.34). ‘The factor +/2 is introduced s0 that later p.*
and p_' may represent complex rms functions averaged in the time
dimension. ‘The + and ~ subscripts indicate the forward and backward
traveling waves respectively.
tis apparent that the first term of Hq. (2.85) represents an outward
traveling wave whose rms magnitude p,* does not change with time {
or position z. A similar statement may be made for the second term,
which is the backwardtraveling wave.
It is customary in texts on acousties to define a wave number k,
ee
kat. Blo (2.36)
Also, let us drop Re and the subscript » for convenience. Any one term
of Eq. (2.35), with these changes, becomes
PEED = V2 [Dla)er] = V2 (Dyer + peer) (2.37)
Similarly, the solution to Eq. (2.22a), assuming steadystate conditions,
ula) = VF (get + u_eitatey (2.38)
tis understood that the real part of Eqs. (2.37) and (2.38) will be used in
the final answer. The complex magnitudes of p, and por wy and ware
determined from the boundary conditions
‘The complex rms pressure and particle velocity are found directly from
Bqs. (2.37) and (2.38) by eanceling /Ze™ from the righthand sides,
When the remaining fonetion is converted into magnitude and phase
angle, the magnitude ts the quantity that would be indicated by a rms sound
pressure meter. Note, however, that when we take the real part of
P(e.) 0F w(t), the quantity +/2 e must bein the equation if the proper
values for the instantaneous pressure and particle velocity are to be
obtained
Pat lV] SOLUTIONS OF THE WAVE EQUATION 27
Brample 22. Assume tha for the steady state, at x point x = 0, the sound pres
rare in a onedimensional outwardraveing wave has the recurrent form shown by
the dotted curve in the sketch below. This wave form i ven by the real part of
the equation
P10.) = VF (er 4 2
(e) What are the partite velocity and the particle displacement «function of time
st z= 5m? () What ate the rms valies of these two quantities? (c) Are the
‘ins values dependent upon 2?
° 3005 ‘oor Tas
seconds,
Solution. a. We have for the solution of the wave equation giving both = and ¢
(sco Eq. (2.371
nia = Va
Prom Ba 2s) wee hat
= ch omen
wie) = jh
wat = 3 wed
And from Hq, (280) we have
PB (4. sansa 4 2B grimuerae
Sow a8" ies
Aba = Sm, 2/2 = 5/348 = 00145 see
ata
tat) = 23 (aesoneosian 4 detmaresieny
sod
tt = 2 (igen sto etme)
‘aking the apart ofthe two preci equation
size) = 2 co (ots — ut) +200 cts 2.9
ean = 23 phgnin ou 0.) + 52 sn ns —270)]
Note hat enh rm nthe parte dplcemen i on of in pase withthe
velocity and Use the wave shape is different. As might be expected, differen tistion
‘nphasioes the28 THE WAVE EQUATION AND SOLUTIONS —(Chap.2
‘Thoee equations are plotted below:
q
q
2. The rms magnitude of a sine wave is equal to ite peak value divided by VB.
‘This may be verified by squaring the cine wave and Gnding the average value over one
cyele and then taking the square root of the rat. If two sine waves of diferent
square root of the sums of the squares of the individual rms magoitudes, 99 that
iby VEER ~ 0011 m/tc0
tn de VGR) + Ga) <8 x10
Thoms values w and é, are independent of x fora plane progressive sound wave,
24. Solution of Wave Equation for Air in a Rigidly Closed Tube. For
this example of wave propagation, we shall consider a hollow cylindrical
Devin ston
Holiow yd tube
fig emiston 7
= t
t
tele
re, 23. Rigidly terminated tube with rigid side walls. ‘The velocity at =
value of 72 ny com al 9 /ce
has
tube, closed at one end by a rigid wall and at the other end by a flat
vibrating piston (soe Fig. 2.3). ‘The angular frequency of vibration of
PatiVj SOLUTIONS OF TH
WAVE EQUATION 29
the piston is w, and its rms velocity is ue. We shall assume that the
diameter of the tube is sufficiently small so that the waves travel down
the tube with plane wave fronts. In order that this be true, the ratio of
the wavelength of the sound wave to the diameter of the tube must be
‘greater than about 6.
Fo. 24, Porton of the tuhe showing the diection and magnitude of movement ofthe
‘ie particles sv fnction of LAL position , the particle volocity and displace
ent ere maxituin, At position B, they are zero.
Particle Velocity. ‘The form of solution we shall select is Eq. (2.38).
Set tis equal to the rims velocity of the vibrating piston at x = 0, and set
Lequal to the length of the tube, The boundary conditions are
Atz = 0, u(0,) = V2 we, so that
aye
Ata = thu = 0, s0 that
ef uel = 0 239)
Remember that
sny = Se
Hence
= xahsin Al (2.40)
and
wet ;
wn ean)
whieh gives ws
lad) = VB te LS ny
sin k(t — 2}
un wk 49)
Note that the «/% and the time exponential have been left out of Eq
(243) so that both ve and w are complex rms quantities averaged ove
time,
Refer to Fig, 24. If the length Land the frequency are held constant30 THE WAVE EQUATION AND SOLUTIONS —{Chap. 2
the particle velocity will vary from a value of zero at x = 1 to a maxi
mum at ~ z= 2/4, that is, at J — x equal to onefourth wavelength.
In the entire length of the tube the particle velocity varies according
to a sine function. Between the end of the tube and the d/4 point,
the oscillatory motions are in phase. In other words, there is no pro
gressive phase shift with z. This type of wave is called a standing wave
because, in the equation, z and cf do not occur as a difference or &
sum in the argument of the exponential funetion. Hence the wave is not,
propagated
In the region between 1 — z= /4 and lz =/2, the particle
velocity still has the same phase except that its amplitude decreases
sinusoidally. At I— x =/2, the particle velocity is zero, In the
region between  — x = 2/2 and 1 ~ x = the particle velocity varies
with 2 according to a sine function, but the particles move 180° out of
phase with those between 0 and 2/2. This is seen from Eq. (2.43),
wherein the sines of arguments greater than © are negative.
If we fix our position at some particular value of x and if Lis held con
stant, then, as we vary frequency, both the numerator and denominator
of Eq. (2.43) will vary. When A is some multiple of x, the particle
velocity will become very large, except at x= 0 or at points where
E( ~ x) isa multiple of x, that is, at points where — z equals multiples
of X/2. ‘Then for kl = nz
(2.44)
Equation (2.43) would indicate an infinite rms velocity under this condi
tion. In reality, the presence of some dissipation in the tube, which was
neglected in the derivation of the wave equation, will keep the particle
velocity finite, though large.
‘The rms particle velocity 1 will he zero at those parts of the tube where
k( — 2) = nefandnisan integer or zero. That is,
(2.45)
In other words, there will be planes of zero particle velocity at points
along the length of the tube whenever 1 is greater than X/2.
Some examples of the particle velocity for 1 slightly greater than various
multiples of 4/2 are shown in Fig. 2.5. Two things in particular are
apparent from inspection of these graphs. First, the quantity n deter
mines the approximate number of half wavelengths that exist between
the two ends of the tube. Secondly, for a fixed wu, the maximum velocity
1 For the type of source we have assumed and no diaspation, this ease breaks down
for Hl = ne
Pat IV] SOLUTIONS OF THE WAVE EQUATION 31
of the wave in the tube will depend on which part of the sine wave falls
atx = 0. For example, if 1 — nd/2 = /4, the maximum amplitude in
the tube will be the same as that at the piston. If] — nh/2 is very near
zero, the maximum velocity in the tube will become very large.
Let us choose a frequency such that n = 2 as shown, Two factors
determine the amplitude of the sine function in the tube. First, atz = 0
the sine eurve must pass through the point we. Second, atx = I the sine
curve must pass through zero. It is obvious that one and only one sine
wave meeting these conditions can be drawn so that the amplitude is
determined, Similarly, we could have chosen a frequency such that
a2) —
FH
aut
x0 a
Fic. 25. Variation ofthe particle velocity (x for =0, 482 function ofthe distance along the
tue of Fig 2 for thee frequencies, ce, forthree wavelengths. ALx = 0, therms particle velocity
the particle velocity is zero. The period T= 1/f
n « 2, but where the length of the tube is slightly less than two half wave
lengths, Tf this ease had been asked for, the sine wave would have started
off with a positive instead of a negative slope at x = 0.
Sound Pressure. ‘The sound pressure in the tube may be found from
the velocity with the aid of the equation of motion (Eq. (2.4a)], whieh, in
the steady state, becomes
p= ~iepfude (2.46)
‘The constant of integration in Eq, (2.46), resulting from the integration
of Eq, (24a), must be independent of 2, because we integrated with32 TH
WAVE EQUATION AND SOLUTIONS — [Chep. 2
respect to x. ‘The constant then represents an increment to the ambient.
pressure of the entire medium through which the wave is passing. Such
An increment does not exist in our tube, so that in Eq, (2.46) we have set
the constant of integration equal to zero. Integration of Eq. (2.46),
after we have replaced u by its value from Eq. (2.42), yields
pe) = “Ine VE na HL 2 ean)
P= ~ineuy 2H 2) 4s)
Note that the /% and the time exponential have been left out of Eq.
2.48) so that both p and tz are complex rms quantities averaged in time.
‘The rms pressure p will be zero at those points of the tube where
i(U = 2) = nx + 7/2, where n is an integer or zero,
(2.49)
The pressure will equal zero at, one or more planes in the tube whenever 1
s greater than )/4. Some examples are shown in Fig. 2.6. Here again,
quantity n fs equal to approximate number of half wavelengths in tube
Refer once more to Fig. 2.5 which is drawn for ¢= 0. ‘The instan
aneous particle velocity is at its maximum (as a function of time). By
comparison, in Fig. 2.6.at ¢ = 0, the instantaneous sound pressure is zero.
At a later time = 7/4 = 14f, the instantaneous particle velocity has
vecome zero and the instantancous sound pressure has reached its m:
num. Equations (2.42) and (2.47) say that whenever k(0 ~ 2) is a
small number the sound pressure lage by onefourth period behind the
yarticle velocity. At some other places in the tube, for example when
1 ~ 2) lies between d/4 and 2/2, the sound pressure leads the particle
relocity by onefourth period.
‘To see the relation between p and u more clearly, refer to Figs. 2.5 and
2.6, for the case of n = 2. In Fig. 2.5, the particle motion is to the right
whenever w is positive and to the left when it is negative. Hence, at the
2/2 point the particles on either side are moving toward each other, 80
hat onefourth period later the sound pressure will have built up to a
naximum, as can be seen from Fig. 2.6, At (I — z) = 4/2, the particles
wre moving apart, so that the pressure is dropping to below barometric
ws ean be seen from Fig. 2.6
Figures 2.5 and 2.6 also reveal that, wherever along the tube the magni
ude of the velocity is zero, the magnitude of the pressure is a maximun
und viee versa. Hence, for maximum pressure, Eq. (2.45) applies.
Specific Acoustic Impedance. It still remains for us to solve for the
ipevifie acoustic impedance Z, at any plane x in the tube. Taking the
Part IV] SOLUTIONS OF THE WAVE EQUATION 33
ratio of Eq. (2.48) to Bq. (2.43) yields
= —iowe cot Ml! = 5X, mks rayls (2.50)
where X, is the reactance.
Where we have set
1
(21)
‘That is, U is the astance between any plane x in Fig. 2.3 and the end
———— wa ——___
ry] oT ft aan
ME
rel Y
: omer 
POEUIET,
0 sl
Fa, 2.6, Variation of the sound pressure p(s) as «function of the distance along the
tube for three frequencies, i.e, for three wavelengths, Atz = 0, the rms particle
Velocity is we and at x =f, itis zero, ‘The period 7 equals 1/f
of the tube at L The —j indicates that at low frequencies where
cot kl! & 1/k’ the particle velocity leads the pressure in time by 90° and
the reactance X,is negative. Atall frequencies the impedance is reactive
and either leads or lags the pressure by exaetly 90° depending, respec
tively, on whether X, is negative or positive. ‘The reactance X, variuM THE WAVE EQUATION AND SOLUTIONS — (Chop. 2
us shown in Hig. 2.7, If the value of kl’ is small, we may approximate the
cotangent by the first two terms of a series
1a
cot ht = yp — (2.52)
‘This approximation is valid whenever the product of frequency times the
distance from the rigid end of the tube to the point of measurement is
very small. If the second term is very small, then it may be neglected
with respect to the frst,
2 ft
fo ne
vty i :
"hI ' 1
 ! !
! 1
 ! !
} 
b
I  
+e t
0 i i wl
A 1 I 
fo ry eM 0
Fro, 27. Tho specific acoustic reactance (Prne/tem) along the tube of Fig. 2.3 for a
particular frequency, sc, » particular wavelength where 3(4/2) isa litle less than the
fue length For this ease, the numberof zeros is 3, and the number of poles is
Let us see how small the ratio of the distance to the wavelength \ must
be if the second term of Eq. (2.52) is to be 3 per cent or less of the first
term. That is, let us solve for I'/A from
2al" d 3)
By 5 0.03 (2.53)
which gives us
ve :
5 = 0.05 (2.54)
In other words, if cot ki’ is to be replaced within an accuraey of 3 per cent
by the first term of its series expansion, I’ must be less than onetwentieth
wavelength in magnitude,
Assuming I" < 4/20, Eq. (2.50) becomes
1 1
pot _ re ae
2m ikem —i5h = a0 7 pat = jo, HB rays (2.55)
Heneo, the specific acoustic impedance of a short length of tube can be
Part IV} SOLUTIONS OF THE WAVE EQUATION 35
represented as a ‘capacitance” elle specific acoustic compliance, of mes;
nitude C, = U'/poc*. Note also that C, = '/7Ps, because of Eq. (2.19).
The acoustic impedance is of the same type, except that an area factor
sppears so that
u2— pet
Bu ~ jo(V Joo) ~ jul,
where V = I'S is the volume and S is the area of cross section of the tube.
C, is called the acoustic compliance and equals V/pec?. Note also that
C4 = V/yPo, from Eq. (2.19).
‘mks acoustic ohms (2.56)
Example 2.8. A cylindrical tube isto be used in an acoustic device as an impedance
clement. (a) ‘The impedance desired is thet of a compliance. What length should
it have to yield w reactance of 1.4 X 10% mks ray xt an angular frequency of 1000
radiana/sec? (6) What is the relative magnitude of the fir. and second terms of
Bq. (252) for this case?
Solution. ‘The reactance of auch a tbe is
P
@ en 14x10 he
Henee, = 0.1m,
o e
renee, the second term is about 3 per eent ofthe fist term.
2.6, Freely Traveling Plane Wave. Sound Pressure. If the rigid
termination of Fig. 2.3 is replaced by a perfectly absorbing termination, a
backwardtraveling wave will not occur. Hence, Eq. (2.37) becomes
Plat) = VE pyemern 7)
where p. is the complex rms magnitude of the wave. This equation also
applies to a plane wave traveling in free space.
Particle Velocity. From Eq, (2.4a) in the steady state, we have
1 op
un 12 58)
Japa 3 58)
Hence,
V2 Pe pcre = PEL
u(t) = rr a pot (2.59)
‘The particle velocity and the sound pressure are in phase. This is mathe
matical proof of the statement made in connection with the qualitative
discussion of the wave propagated from a vibrating wall in Chap. 1 and
Fig 11
Specific Acoustic Impedance. The specific acoustic impedance is
2, =F = pe mks rayls (2.60)36 THE WAVE EQUATION AND SOLUTIONS — [Chap.2
‘This equation says that in a plane freely traveling wave the specific
acoustic impedance is purely resistive and is equal to the product of the
average density of the gas and the speed of sound, This particular
‘quantity is generally called the characteristic impedance of the gas because
its magnitude depends on the properties of the gas alone. It is a quantity
that is analogous to the surge impedance of an infinite electrical line.
For air at 22°C and a barometric pressure of 10° newtons/m?, its mag
nitude is 407 mks rayls.
2.6. Freely Traveling Spherical Wave. Sound Pressure. A solution
to the spherical wave equation (2.24) is
vi ec + acm) a 261)
where A, is the magnitude of the rms sound pressure in the outgoing wave
at unit distance from the center of the ephere and A_ is the same for the
reflected wave.
If there are no reffecting surfaces in the medium, only the first term of
this equation is needed, i.e.,
Pert)
Pen) = VEAL pu 0)
Varticle Velocity. With the aid of Eq. (2.48), solve for the particle
velocity in the r direction,
ure)
VE oo( 43h)
VEL (14 be 203)
Specific Acoustic Impedance. ‘The specific acoustic impedance is found
from Eq. (2.62) divided by Eq. (2.63),
kr peckr
BR Sore
7 (90° = tan mks (2.64)
oe ~ i kr rayls (2.64)
Plots of the magnitude and phase angle of the impedance as a function
of kr are given in Figs. 2.8.and 2.9. The real and imaginary parts, R, and
X,, are plotted in Fig. 2.10.
Por large values of kr, that is, for large distances or for high frequencies,
this equation becomes, approximately,
Z.% poe mks rayls (2.65)
The impedance here is nearly purely resistive and approximately equal to
the characteristic impedance for a plane freely traveling wave. In other
words, the specific acoustic impedance a large distance from a spherical
source in free space is nearly equal to that in a tube in which no reflections
pcour from the end opposite the source.
Pat IV} SOLUTIONS OF
va =
1 ee 14
ete
oe
 See
tncretmceste, CHEV 
gt REE a 
£ a
gL 
od 
okt oe  ry t
coe * ‘ood “oa 05 10 5 10 20
(erat feat r/R)
Fo, 28. Plot of the magnitude of the specific acousticsimpedance ratio !Z/ow in a
spherical feeely traveling wave as a funetion of kr, where is the wave number equal
tow/ror 2e/sand ris the distance from the center of the spherical source. [21] isthe
‘magnitude of ratio of pressure to particle velocity in a spherical free traveling weve,
sand pr ie the characteristic impedance ef wit
so
Ao
* LUI iz
7} + itt 4
« a .
t :
5° ani
ol  lt" ercraegeetis NN
£40) particle velocity
3 Bisiecaate
Ey : {ty
 i 
Pa
9 
ih
— + . ee
oo * “ood “ox 08° "10, 5° 10 =
Uirytorfee(200)
Fie, 29. Mot ofthe phate angle, n degre ofthe pee acousticmpedance rt
Zulosc in a spherical wave a3 8 function of kr, where kis the wave number equal to
wfeor 2r/a, und ris the distance from the centr of the spherical source.38 THE WAVE EQUATION AND SOLUTIONS — [Chap.2
Equations (2.62) and (2.63) are significant hecause they reveal the
difference between the responses of a microphone sensitive Lo pressure
and a microphone sensitive to particle velocity as the microphones are
brought close to a small spherical source of sound at low frequencies. AS
ris made smaller, the output of the pressureresponsive microphone will
double for cach halving of the distance between the microphone
29)
1
0s
03]
02
oa
005}
03]
0.02]
Po
0.005
0.008]
002]
Irapedance
con ‘aalogy
0.0005}
0.0008]
0.0002]
00014 = ire
oor * G05 “02 03 ‘10 510 20
be
Fio, 2.10. Teal and imaginary parts of the normalized spociie acoustic impestance
Z.Joue of the air loul ons plating sphere of radius r loeated in free spac. Fe
{quency is plotted on a norinalized seale where kr = 2xjr/e = 2er/d.” Note als that
the ordinate is equal to Zu/owS, where Zw i the mechaninl impedance, nil to
ZaS/no, whore Zs the acoustic impeance. The quantity 8 is the aren for which
the impedance is being determined, and py ig the characteristic impesanc: of the
edie
center of the spherical source. Expressed in deeibols, the output increases
Gdb for cach halving of distance, For the velocityresponsive micro
phone, the output variation is not so simple. Only at sufficiently large
distances (kr* >> 1) does the output increase 6 dh for each halving of
distance. For shorter distances the second term inside the parentheses
on the righthand side of Eq. (2.63) becomes large, and the magnitude of
w increases at a rate exceeding +6 db for each halving of distance. For
Pat VJ SOLUTIONS OF THE WAVE EQUATION
very short distances (Kr? < 1), the rate of increase of w approaches :
limit of +12 db for each halving of distance. It is for this reason tha
the voice of a radio crooner sounds “bassy”” when he sings very neat tot
velocitysensitive microphone which was designed to have ite best
response when located a large distance from the source of sound
‘Another significant thing is to be learned from Eq, (2.64). At low fre
quencies it is very difficult to radiate sound energy from a small loud
speaker. A small loudspeaker may be likened to a pulsating balloon o!
some small radius r. The specific acoustic impedance Z, of the air pre
sented to each square centimeter of the balloon is given by Eq. (2.64) and
Fig. 2.10. At low frequencies, the impedance becomes nearly purely
reactive, and the resistance becomes very, very small, Hence, the powe!
radiated by a small loudspeaker becomes very small. At high fre.
quencies, kr > 2, the impedance Z, becomes nearly purely resistive and
has its maximum value of pc, s0 that the power radiated for a given value
of p, reaches its maximum.
‘The important steadystate relations derived in this chapter are sum.
marized in Table 2.1.
‘TABLE 21. General and Steadystate Relations for Smallsignal Sound Propa.
gation in Gases
ame Genel equation  Steuyatate euatin
Wave equations 718
Joy ya 22)
X= Bar
apr) _ 1 apr)
oF 7a ae
Equation of motion oa
Bm Mfume Jude
grad p= Jao
Diplacement
Incremental density
Incremental temperature40 THE WAVE EQUATION AND SO
TIONS — [Chap. 2
parr V Energy Density and Intensity
2.7, Bnergy Density. Energy density is an important concept in
acoustics because, in dealing with sound in enclosures, it is necessary to
study the flow of energy from a source to all parts of the room. The
energy density, ie., wattseconds per unit volume, is greater near the
source than farther away and is the variable that appears in the equations
describing the acoustical conditions. On the other hand, the ear and
most soundlevel meters respond to rms sound pressure. We need to
ascertain, therefore, the relation between energy density and sound
pressure in sound fields.
‘The energy density associated with the small “box” of gas at any par~
ticular instant is the sum of the kinetic and potential energies per unit
volume of the air particles in the box. ‘The kinetic energy density due to
the excess pressure of the sound wave Drx is
1 Mut _1
Dax = 5 ye = yee (2.66)
where w is the average instantaneous velocity of the air particles in the
ox, pe is the average density, and 3f/V is the mass per unit volume.
‘The potential energy density due to the sound wave Dre may be found
from the gas law. For very small changes in the volume of the box, we
may write [see Eqs. (2.8)]
(2.67)
(2.9) and substitute the resulting expression for
dz, the potential energy density becomes
= pap lw
Pa 25P,
Des (2.08)
When the sound pressure p is equal to zero, the potential energy due to
the sound wave must be zero. The arbitrary constant of integration is
therefore also equal to zero.
‘The total energy density due to the sound wave D = Dax + Des, oF
Dw = 3 (on + 5) (2.69)
‘This equation is true at any instant at a given point in space
2.8, Energy Density in Plane Waves. Energy Density in a Plane
Preeprogressive Wave. From Eqs. (2.57) and (2.59) we have seen that
the pressure and particle velocity in a plane freeprogressive (outgoing)
Part VI ENERGY D
PNSITY AND INTE
41
wave are equal to
(zt) = Re V2 pret = /2 Ips cos [k(ct — 2) + 6] (2.70)
V2 Me psc = V2 Pol
Ee Feel cos h(ct ~ 2) + 6] (2.71)
u(z,t) = Re
where py = pyle.
‘The instantaneous energy density for such a wave in the steady state is,
from Lage. (2.68) and (2.19), equal to
D(z) = psl*
DG
pan) cost Ia(ct — 2) + 8]
= Dpal? 1+ cos 2at ~ 2/e + Of)
mc 2
WAL (1 + cos 2u(t — 2/0 + 6/4) @72)
“This equation says that for a plane freprogressive wave, at all times, the
Kinetic and potential energy densities are equal at a given point in space
but that they vary with position or with time sinusoidally from zero to
twice their average value. ‘The situation here is different from that fora
pendulum where the kinetic energy and the potential energy vary in
opposite phase, &.c, one ie a maximum when the other iz a minimum
Here, energy is being transported away from the source. Conversely,
the pendulum is a conservative system.
‘When averaged over either a length of time equal tol = 1'/2 = 1/2f or
a distance in space 2 ~ /2~ o/2f, we find the average energy density to
be equal to
Ipal? 2 13)
al wattce/m 273)
D,
where ps is the magnitude of the rms value (in time) of the sound pres
sure measured at any point in the sound wave. Note also that pect = 7Ps
as stated before. Inspection of Eq. (2.60) shows that we may let
Pe
pot
where 1, is the rms value (in time) of the velocity at any point in the
wave, Then,
uy (274)
Dara = 1sl¥00 (275)
‘Equations (2.73) and (2.75) give the relations among rms sound pressure,
particle velocity, and energy density
Energy Density in a Plane Standing Wave. From Eqs. (2.42) and
(2.47) we have that
piety = VE need EM cow — 2) aaa42 THE WAVE EQUATION AND SOLUTIONS — (Chap.2
where @ is the phase angle of us.
608 (ot + 0) sin k= 2
UE) = VE ug) SRE Sin BE 2) em
In this case, the kinetic and potential energy are 90° out of time phase
‘The situation is analogous to that for a pendulum becuse in both eases
the systems are conservative.
‘The instantaneous energy density for such a wave in the steady state is,
from Eqs. (2.69) and (2.19), equal to
08 2(ut + @) cos 2(l — 2)
T= cos 2E
D(z) = udlioe 2.78)
‘When averaged over either a length of time f equal to 7/2 or a distance
in space equal to 4/2, we find the average energy densily to be equal to
Dog = ltl, wateane/mt ex
where us is the magnitude of the rms velocity of the piston at 2 = 0.
This equation shows that, for « constant value of jus, the average energy
density varies from uy/?p9/2 to infinity depending on the value of
H = 2el/r
A better way of representing the average energy density is in terms of
the rms pressure. If, by definition, we let the rms value of the pressure
be related to the mms velocity uy at x = O by the formula
pi = one (280)
we have
(zt) = V2 [pil sin (wt + 6) cos k(t — 2) (281)
Then Eq. (2.79) becomes
= iat
Dow = PO (282)
Here, {pi is the magnitude of the rms value (in time) of the mazimum
value (in space) of the sound pressure. If we measure the rms value of
the sound pressure in space by moving a microphone backward and for
ward over a wavelength and averaging the varying output in a rms
rectifier, then [Poul = [psl/V2 and
Dig = Weal’ yattace/mt (283)
where pow is the magnitude of the rms value of the sound pressure
averaged in both space and time, Note that Eq. (2.83) is identical to
By. (273)
Part V1 ENERGY DENSITY AND INTENSI 3
Example 24. Calculate the average energy density in a plane freeprogressive
Sinusoidal sound wave with a maximum particle displacement of 0.01 em at a
quency of 100 eps.
Soluion, From Eq, (2.80) we find that the ras yrtile velovity 1 tga wine
80
2e x 100 x OL >
one = 00 X 0.01 9.645 ma /sce
VP x10
‘The average energy density is given by Ka. (275),
Dag = (0.0485)* X LB = 234 X 109 wattseo
2.9. Energy Density in a Spherical Freeprogressive Wave. ‘The
energy density in a spherical freeprogressive wave can be shown to be
equal to
81)
where [p, is the magnitude of the rms value (in time) of the sound pres
sure at a point a distance r from the center of the spherical souree.
If the produet of the distance r and the frequency is large (2k? >> 1),
the average energy density is the same as for plane freetraveling oF
standing wave, as can be seen from Eqs. (2.73) and (2.83). Near the
sonree, however, the eneray density hecomes very large. ‘This occurs
because the impedance {sce Eq. (2.64)] becomes largely reactive and the
stored energy becomes high.
2.10. Sound Energy Flow—Intensity, Later in this text we make
frequent reference to the flow of sound energy through an acoustie system,
Because of the law of the conservation of energy, the total acoustie energy
starting from a source must be completely accounted for in the system.
At any part of an acoustic system, we should be able to state the amount
of energy flowing through that part per unit time, and it should equal the
power emanating from the source minus any intervening losses.
In Part IT we defined intensity as the average time rate at which energy
is flowing through unit area of the acoustie medium. In the mks system,
the units of intensity are watts per square meter. The intensity is
actually the product of the sound pressure times the inphase component
of the particle velocity.
General Equation for Intensity. We can find the average intensity J
1 given direction at a given point in the medium by performing the
operation
T= Reptgcos ¢ (285)
SL. B Kinder and A. TR. Frey, “Fundamentals of Acoustss,” pp, 167168, John
Wiley & Sons, Ine, 1980.
1 The average power supplied by an clrtrical generator to a circuit equals the
voltage times the inphase component of the current. "Tht power ean be shown to
quel Re (E*1), where B and 7 are the complex ema voltage and curtont, respectively.a THE WAVE EQUATION AND SOLUTIONS — [Chap.2
where p* is the complex conjugatef of the rms sound pressure p, qis the
complex rms particle velocity in the direction the wave is traveling, and
@ is the angle between the direction of travel and the direction in which
the intensity is being determined. ‘The symbol Re indicates that the
real part of the product is to he taken.
Intensily in a Plane Freeprogressive Wave. For a plane freeprogressive
sound wave the intensity equals
= Re pot Pt co
T= Re pom Be é 2.80)
Another way of looking at the question of intensity for & plane progres
sive wave is to say that all the energy contained in a column of gas equal
in Jength to em must pass through unit area in 1 see. Hence, the inten
sity is
T= Dang 005 4 287)
So, regardless of whether the intensity is determined from (2.86) or (2.87),
wwe get for & plane freeprogressive wave that
Pal oo g =
1 = BE cose
Intensity im a Plane Standing Wave. In a plane standing wave the
pressure and particle velocity are 90° out of phase in time [see Fqs. (2.76)
and (2.77)] so that the real part of the product ptu is zero. Hence, for
a plane standing wave,
ius love cos (2.88)
T=0 (289)
Physically, tbis means that as much sound energy returns to the source as
travels away from it.
Intensity in a Spherical Freeprogressive Wave. For a spherical progres
cive wave, we got the pressure p from Eq. (2.62). By definition, let
pe a em (2.00)
Then,
Plot) = V2 pact eo)
‘The quantity p, is equal to the complex rms pressure at any point a dis
tance 7 from the center of the source. Hence, the particle velocity
u(r) at any point a distance r is
Em (14 jbo asm
HIE pis represented by [ple then pe is [ple Similarly, if p i represente hy
e+ jy, then 9" i represented by px ~ Je
ur.
Ps
Vi ENERGY DENSITY AND INTENSITY 45
or the complex rms particle vel
we B15]
Substitution of the sound pressure at, p,, and Eq, (2.98) into Bq, (2.85)
vields
) a
I= Rept iF Be cos & eo)
‘owe
here, as before, ¢ is the angle between the direction of travel of the wave
and the direction in which the intensity is being determined.
‘Wo can derive these results in a different way. Equation (2.08) states
that, for kr large, p and u fora spherical wave are nearly in time phase and
P(r) = pucu(r) a8 shown by Eq. (2.05). Hence, for kr large, we see fron.
Eq, (2.88) that in a spherical wave for large distances I = ur?pec 008 6.
‘The total power at any radius r is equal to W = er! = 4r"pl?/pee.
Hence, for a spherical wave,
Ww
T=, forg=0 (2.95)
By the law of conservation of energy, W is independent of r if there are
uio losses in the gas so that the intensity varies inversely as the square of
the distance r.
From Eq. (2.90) we see also that: the square of the rms magnitude of
the sound pressure at any point varies inversely with the square of the
distance r. Hence, because the intensity / at any point varies similarly,
it is directly proportional to the square of the sound pressure at that.
point. ‘This result agrees with that shown in Eq. (2.94)
Example 2.5. A spherical sound source is radiating sinusoidally into free space
 watt of acoustic power at 1000 eps. Caleulate (a) intensity in the direction the
wave js traveling; (6) sound pressure; (e) particle velocity; (d) phase angle between
(0) and (¢); (2) energy density; and (f) sound preseure level at @ point 30 em from the
‘eonter of the source, (Assume 22°F and 0.751 m Hg.)
Solution. a. ‘The intensity maybe found fo
2.95).
>, The rins sound pressure eomes from Ig. (2:94).
Ind = Vase = VORB XAT = 18.97 nowtone/m*
‘ The rms particle volocity is given by Bg. (2.98),
be = (Or x 1000/344.8)(08) = 5.46
be VTE A 1897 VT + HR
tel = Fe VUES 1897 VT BB aaz4 mine46 THE WAVE EQUATION AND SOLUTIONS — (Chap.
4. The phase angle @ between pe and ue may be found from Faq. (264)
0 = 90" — tant kr = 90" — 796" = 10.4"
The energy density is given by Hq. (2.84)
BEC + ah) = pata (
Pow pact AI * aunt) ~ TEx TO +
+2162 X 104 wattsee/n*
J. The sound pressure love is found from Hq. (1.18).
oe 18.97
SPL = 20 log 5 OTT
= 119.5 db re 2 X 104 nowton/m (re 2 X 104 mierobar)
‘This sound pressure level is about 15 db higher than the highest level tht is meas
tured at 25 1t above a full aymphiony orchestra. In other words, 1 watl of acoustic
power ervatcs a very high sound pressure level atl ft from the souree
CHAPTER 3
ELECTROMECHANOACOUSTICAL CIRCUITS
parr VI Mechanical Circuits
SS
3.1, Introduction. ‘The subject of electromechanoacoustics (some.
times called dynamical analogies) is the application of electricalcireuit
theory to the solution of mechanical and acoustical problems. In clas
sical mechanics, vibrational phenomena are represented entirely by
differential equations. ‘This situation existed also early in the history of
telephony and sadiv. As telephone and radio communication developed,
it became obvious that a schematic representation of the elements and
their interconnections was valuable. These schematic diagrams made it
Possible for engineers to visualize the performance of a circuit without
laboriously solving its equations. The performance of radio and tele
vision systems can be studied from a single sheet of paper when sich
schematic diagrams are used. Such a study would have been hopelessly
4ifficult if only the equations of the system were available.
There is another important advantage of a schematic diagram besides
its usefulness in visualizing the system. Often one has a piece of equip
ment for which he desires the differential equations. ‘The schematic
iagram may then be drawn from visual inspection of the equipment.
Following this, the differential equations may be formed directly from
the schematic diagrams. Most engineers are trained to follow this pro
cedure rather than to attempt to formulate the differential equations
directly.
Schematic diagrams have their simplest applications in circuits that
contain Jumped elements, é¢., where the only independent variable is
time. In distributed systems, which are common in acoustics, there may
be as many as three space variables and a time variable. Here, a
schematic diagram becomes more complicated to visualize than the
differential equations, and the classical theory comes into its own again.
There are many problems in acousties, however, in which the elements
are Jumped and the schematic diagram may be used to good advantage.
a48 ELECTROMECHANOACOUSTICAL CIRCUITS (Chap. 3
Four principal requirements are fulfilled by the methods used in this
text to establish schematic representations for acoustic and mechanical
devices. They are’
1. The methods must permit the formation of schematic diagrams from
visual inspection of devices
2. They must be capable of such manipulation as will make possible
the combination of electrical, mechanical, and acoustical elements into
one schematic diagram.
3. They must preserve the identity of each element in combined cir
cuits 60 that one can recognize immediately a force, voltage, mass,
inductance, and so on.
4. They must use the familiar symbols and the rules of manipulation
for electrical circuits
Several methods that have been devised fulfill one or two of the above
four requirements, but not all four. A purpose of this chapter is to
present a new method for handling combined electrical, mechanical, and
acoustic systems. It incorporates the good features of previous theories
‘and also fulfils the above four requirements. ‘The symbols used conform
with those of earlier texts wherever possible.!“*
3.2. Physical and Mathematical Meanings of Circuit Blements. ‘The
circuit elements we shall use in forming a schematie diagram are those of
clectricalcircuit theory. These elements and their mathematical mean
ing are tabulated in Table 3.1 and should be learned at this time. There
are generators of two types. ‘There are four types of circuit elements:
resistance, capacitance, inductance, and transformation. There are three
generic quantities: (a) the drop across the circuit element; (0) the flow
through the circuit element; and (¢) the magnitude of the circuit element.
Attention should be paid to the fact that the quantity ais not restricted
to voltage ¢, nor b to electrical current i. In some problems a will repre
sent force f, or velocity wu, or pressure p, or volume velocity U. In those
cases b will represent, respectively, velocity u, or force f, or volume
°B, Gehlshoj, “lectromechanical and Hectroscoustieal Anslogies,” Academy of
‘Technical Sciences, Copenhagen, 1947
*F. A. Firestone, A Now Anelogy between Mechanieal and Hlectrcal Systems,
J. Acoust. Soe. Amer, 4: 249267 (1988); The Mobility Method of Computing the
Vibration’ of Linear’ Mechanical and Acoustical Syatems: Mechanicalclectrical
Analogies, J. Appl. Phyr., 8: 373387 (1938).
*H. F, Olson, “Dynamical Analogies” D, Van Nostrand Company, Inc, New
York, 1988.
“WP. Mason, Electrical and Mechanical Anslogies, Bell Syslem Tech, J., 20:
405414 (1941).
A. Bloch, Electromechanical Anslogies and Their Use for the Analysis of Mechani
cal and Bleetrosmechanieal Systems, J. Ina. Elec. Eng., 92: 157169 (1945),
{Among the four circuit elements, the first three are twopoles, This list is exhaus
tive, The transformation element isa fourpole, There are other loses fonrpoes
Which one might have chosen in addition, e., the ideal gyrator,
Part. VI) MECHANICAL CIRCUITS 49
velocity U, or pressure p. Similarly, the quantity ¢ might be any
appropriate quantity such as mass, compliance, inductance, resistance,
etc, The physical meaning of the circuit elements ¢ depends on the way
in which the quantities a and b are chosen, with the restriction that ab
thas the dimension of power in all eases, ‘The complete array of alterna
tives is shown in Table 3.2.
‘TABLE 3.1. Mathematical and Physical Significance of Symbols
] T ‘Mooring
somtt  tame :
 Constantdrop  The quantity a is independent of
tor. ‘The arrow points to the
positive terminal ofthe generator
Constantow The quantity b is independent of
 generator what is connected to the gen
erator. ‘The arrow points in the
ireation of positive fow
Resistancetype  a
sement
Sent fend f ba fom
 ee
— ee
clement
gst $ type element, be a
138 aa
‘An important idea to fix in your mind is that the mathematical opera~
tions associated with a given symbol are invariant. If the element is of the
inductance type, for example, the drop @ across it is equal to the time
derivative of the flow b through it multiplied by its sizec. Note that this
rale is not always followed in electricalcireuit theory because there con
ductance and resistance are often indiscriminately written beside the
symbol for a resistancetype element, The invariant operations to be
associated with each symbol are shown in columns 3 and 4 of Table 3.1. Part VI) MECHANICAL CIRCUITS 51
CHANOACOU
3.3, Mechanical Circuits. Mechanicalcircuit elements need not
“kK always be represented by electrical symbols. Since one frequently draws
" a mechanical circuit directly from inspection of the mechanical device,
bis. ‘more obvious forms of mechanical elements are sometimes useful, at least
until the student is thoroughly familiar with the analogous cireuit. We
shall accordingly devise a set of ‘‘mechanieal” elements to be used as an
introduction to the elements of Table 3.1
em Ma
leney
TABLE 3.3. Conversion from Mobilitytype Analogy to Impedancetype
Analogy, or Vice Versa
Acoustical
 ats = [MECHANICAL NULOGES {AOOUSTIAL MALDGIES
Impedance
analogy
P
u
ovis pe [_trowtnce ype  Meity ype  tnpetonn Spe
ca ee (a
i Ec a
ry ToS we PS BS att!
on Ra
Je= Me
‘and Acoustical Circuits
Impedance
‘analogy
f
T
‘Mechanical
Mesh aus (nobiy Boe) ech to was empress)
{us 2 poe
In electrical circuits, a voltage measurement is made by attaching the
Jeads from a voltmeter across the two terminals of the element. Voltage
is a quantity that we can measure without breaking into the circuit, ‘To
measure electric current, however, we must break into the circuit because
this quantity acts Ghrough the clement. In mechanical devices, on the
other hand, we ean measure the velocity (or the displacement) without
disturbing the machine by using « capacitive or inertially operated vibra~
tion pickup to determine the quantity at any point on the machine. It
is not velocity but force that is analogous to electric current. Force can
i not be measured unless one breaks into the device.
Values for a, & and ¢ in Electrical, Mechani
ese
TABLE 3.2
:
ans
yas
oie
° ’
1 PreferrROM
CHANOACOUSTICAL CIRCUITS (Chap.
It becomes apparent then that if a mechanical element is stetly
analogous to an electrical element it must have a velocity difference
appeating between (or acrou) is wo terminals ands free acing Cough
it Analogounly als, the produet of ie rma foree/ in newlane ard the
inphase eompovent of the nn velocity w in meter por second Je the
oF ———r—C— rrr
torreeponds to a voltage and a force to a ourant, the mobilgnine
tmalogy. tis also known as the “averse” asalogy
‘Many text teachin addition a "dieet” analogsy Tt the opposite of
the mobility analogy in that force is made to correspond to voltage and
rh ———F_————
Impedance type analogy To fain the student with bath consepte
‘all examples will be given here both in mobilitytype and impedancetype
analogies.
Mechanical Impedance Zy, and Mechanical Mobily tw. ‘The mechan
Oe —r———i—
tt mechanical device, "We commonly use the symbol Ze or mechani
impedance, where the subscript M stands for “mechanical.” "The unit
tre nentonacconda per meter, or mka mechaneal hina
"The mechanical mcilty isthe inverse of the mechanical impedance,
1Linthe complex ata volrtytatce nea gees eaat a ooh
devieesWeommonly ure the symbol fs for mechanical mating Te
Unite are meters por seco per newton, or mia rascal
14. cal mohms.+
——
Je selon a omelet opto
that force. ‘The units the kilogram, AUN ght ase
LL ———_N
convection ip needed to et it in motion” Howevty, the
Je. Me force acting on t mase and tho renllant acederalon
{ArH reckoned with reapet to ube earth neta rae) co thet
Inreality the second terminal of masse the earth
he mechanival symbol used to represen mast is shower i Fig. 31
‘The upper end of the mase moves with a velocity wth respect fo te
ground. ‘The }shaped configuration represents, the "aesnd” terminal
Of the mass end has aero velocity. "The force cam be measured by a
tuilable device inorted between the point I and the next elomen cr
reverator connecting toi
Maan Mw obeys Newton second aw that
au)
Sy = Mu @y
{The word “mohm” stands for mobility ohm. The vty are meters pee second
Pact VE) MECHANICAL CIRCUITS 53
‘whore /() is the instantaneous force in newtons, Mw is the mass in kilo
grams, and 1(0) is the instantaneous velocity in meters per second,
In the steady state [see Eqs. (2.38) to (2.35)], with an angular frequency
«sequal Lo 2r times the frequeney of vibration, we have the special ease of
Newton's second law,
joM we 82)
where j = \/=i as usual and f and w are rms complex quantities,
The mobilitytype analogous symbol that, ;
we use as a replacement for the mechanical I
symbol in our circuits is a capacitance type. [> re
Wisshown in Fig. 320, Themathematical fu eae Py Say
4
‘operation invariant for this symbol is found
from Table 3.1. In the steady state we have
am or = L658) ty toe tpt
jue oN
(a) 0)
‘This equation is seen to satisfy the physical Fra. 3.2. (a) Moblityype
law given in Bq, (3.2). Note the similarity $4 (impedance type ym
in appearance of the mechanical and analo
coussvmbolsin Figs. 8.1 and 3.2a, Inelectrical eireuits the time integral
of the current through a capacitor is charge, The analogous quantity
here is the time integral of force, which is momentum,
‘The impedancetype analogous symbol for a mass is an inductance.
It is shown in Fig. 3.25, ‘The invariant operation for steady state is
4a = jucb orf = juM yn. Ialso satisfies Bq. (82). Note, however, that
in this analogy one side of the mass element is not neces
1) Ju, sarily grounded; this often leads to confusion. In elec
trical circuits the time integral of the voltage across an
inductance is fluxctums. The analogous quantity here
is momentum,
Mechanical Compliance Cy. A physics} structure is
2ll Yur said to be a mechanical compliance Cw if, when itis acted
ie. 83. Me. on by a force, its displaced in direct proportion to the
chasiealsymbol force. The unit is the meter per newton. Compliant
for a mishan~ elements usually have two apparent terminals
‘The mechanical symbol used to represent a mechanical
compliance isa spring. It is shown in Fig. 3.3. ‘The upper end of the
element moves with a velocity u: and the lower end with a velocity us
“The force required to produce the difference hetween the velocities u; and
tu: may be measured by breaking into the machine at either point } or
point 2. Just as the same vrrent would be measured at either end of an
element in an electrical cirewit, 20 the same force will be found here at
either end of the compliant element.
Cw54. ROMECHANOACOUSS
AL CIRCUITS [Chap.3
Mechanical compliance Cy, obeys the following physical ia,
ant foa or 0 Z fue ea)
where Cy is the mechanical compliance in meters per newton and u(t) is
the instantaneous velocity in meters per second equal to us — u, the
difference in velocity of the two ends,
In the steady state, with an angular frequency w equal to 2x times the
frequency of vibration, we have,
i
jae 5)
where f and u are taken to be rms complex quantities.
‘The mobilitytype analogous symbol used as 2 replacement for the
mechanical symbol in our circuits is an inductance. Tt is shown in Fig,
3.4a. ‘The invariant mathematical operation
f © that this symbol represents is given in Table
3.1. In the steady state we have
w= jeoluf G6)
_
ae gee
Ls Ly In electrical circuits the time integral of the
voltage acrose an inductance is flustusns
esatytoe —Ingetnetye ‘The analogous quantity here is the time in
() (0) ‘tegral of velocity, which is displacement,
Fro. 34. (2) Mobilitytype ‘This equation satisfies the physical law
ni) npotaeetype given in Eq, (QS). Note the wimaiy in
fei fr athe Sompir cance the mechanical and analogous
symbols in Pig 8.8 and 3.
‘The impedancetype analogous symbol for a mechanical compliance is
a capacitance. It is shown in Fig. 3.45. ‘The invariant operation for
steady state isa = b/jue, or f = u/jwCy. It also satisfies Bq. (3.5). In
electrical circuits the time integral of the current through capacitor is
the charge. The analogous quantity here is the displacement,
Mechanical Resistance Ry, and Mechanical Responsiveness tw. A
physical structure is stid to be a mechanical resistance Ry if, when it is
acted on by a force, it moves with a velocity direetly proportional to the
force. ‘The unit is the mks mechanical ohm.
‘We also define here a quantity ry, the mechanical responsiveness, that,
is the reciprocal of tw. ‘The unit of responsiveness is the mks mechanical
mohm,
‘The above representation for mechanical resistance is usually limited
to viscous resistance. Frictional resistance is excluded because, for it,
the ratio of force to velocity is not a constant. Both terminals of resistive
ements can usually be located by visual inspection,
Pact Vi) MECHANICAL CIRCUTTS 55
‘The mechanical element used to represent viscous resistance is the fluid
dashpot shown schematically in Fig. 3.5, ‘The upper end of the element.
moves with a velocity w and the lower with a velocity us. ‘The force
required to produce the difference between the two velocities u; and ts
may be measured by breaking into the machine at either point 1 oF
point 2
Mechanical resistance Ry obeys the following physical law,
foram de
where f is the force in newtons, 1 is the difference between the velocities
tu, and uz of the two ends, Hy is the mechanical resistance in mechanical
ohms, ie, newtons/(meter per second), and ry is the mechanical
responsiveness in mks mechanical mohms, i.e., meters per second per
newton.
‘The mobilitytype analogous symbol used to replace the mechanical
symbol in our circuits is a resistance, It is shown in Fig. 3.6a, ‘The
ody tye gta
al be (a) 0) al be
Fa. 35, Mechanical Fic.3.6. (@) Mobilitytypeand Fra. 3.7. Mechenical
symbol for mechanical @) impedaneetype symbols for eymboi for a constant
(iseous) resistance. mechanical resistance. velocity generator,
invariant mathematical operation that this symbol represents is given in
Table3.1. Incither the steady or transient state we have
we raf G8)
In the steady state w and f are taken to be rms complex quantities. ‘This
equation satisfies the physical law given in Ea. (3.7)
‘The impedancetype analogous symbol for a mechanical resistance is
shown in Fig, 3.00. It also satisfies Bq. (3.7).
‘Mechanical Generators. ‘The mechanical generators considered will be
cone of two types, eonstantvelocity or constantforce. A constant.
relocity generator is represented as a very strong motor attached to a
shuttle mechanism in the manner shown in Fig. 8.7. ‘The opposite ends
of the generator have velocities uy and us. One of these velocities either
1 oF ua, is determined by factors external to the generator. The differ56 ELECTROMECHANOACOUS
L CIRCUITS (Chap. 3
ence between the velocities w; and wy, however, is velocity w that is inde
pendent of the external load connected to the onerator.
‘The symbols that we used in the two analogies to replace the mechanical
symbol for a constantvelocity generator are shown in Fig. 38. The
invariant mathematical operations that these symbols represent are also
given in Table 3.1. ‘The tips of the arrows point to the “positive”
terminals of the generators. ‘The double circles in Fig. 3.8a indicate that
the internal mobility of the generator is zero. ‘The dashed line in Fig.
3.80 indicates that the internal impedance of the generator is infinite.
‘A conslantforce generator is represented here by an electromagnetic
transducer (e.g, a movingeoil loudspeaker) in the primary of which an
electric current of constant amplitude is maintained. Such a gencrator
produces a foree equal to the product of the current é, the flux density B,
and the effective length of the wire I cutting the flux (f = Bli). This
device is shown schematically in Fig. 3.8. ‘The opposite ends of the
aaa, 4 fu
== ee
(a) (b) (a) (8)
generator have velocities u and ws that are determined by factors external
to the generator. ‘The force that the generator produces and that may
be measured by breaking into the device at either point 1 or point 2 is a
constant force, independent, of what is connected to the generator.
‘The symbols used in the two analogies to replace the mechanical
symbol for a constantforce generator are given in Fig. 3.10. ‘The
invariant mathematical operations that these symbols represent are also
given in Table 3.1. ‘The arrows point in the direction of positive flow.
Here, the dashed line indicates infinite mobility, and the double circles
indicate zero impedance.
Levers. simpie vever. It is apparent that the lever is a device
closely analogous to a transformer. ‘The lever in its simplest form con
sists of a weightless bar resting on an immovable fulerum, so arranged
that a downward force on one end causes an upward force on the other
id (see Fig. 8.11). From elementary physics we may write the equation
‘of balance of moments around the fulerum,
Suds = fils
Past VI} MECHANICAL CIRCUITS a7
of, if not balanced, assuming, small displacements,
9)
Also,
.10)
‘The above equations may be represented ¥ the ideal transformers of
ws nl ), rh ipa pe
FLOATING LevER. As an example of a simple floating lever, consider a
weightless bar resting on a fulcrum that yields under force. ‘The bar is,
seta HE) OG}
Aa Mosity ype Impedance. oe
2 (@) ()
Fie, 3.11. Simple lever. Fro, 3.12, (a) Mobilitytypo and (®) imped
fancetype symbols for a simple lever,
so arranged that a downward force on one end tends to produce an upward
force on the other end, An example is shown in Fig. 3.13.
‘To solve this type of problem, we first write the equations of moments.
' 4
et 4
‘Summing the moments about the
center support gives
Lf hfe at
Mobis consti
7, move vp a8 Sonn ery
hfe (B11) Fic. 3.18. Floating lever.
the end support gives
(i + bf
When the forces are not balanced, and if we assume infinitesimal dis
placements, the velocities are related to the forces through the mobilities,
so that
(3.12)ICAL CIRCUITS (Chap. 3
mau QE fora =0
so that
(3.13)
and, finally,
14)
This equation may be represented by the analogous circuit of Fig. 3.14.
The lever loads the generator with two mobilities connected in series, each
of which behaves as a simple lever when the other is equal to zero. It
will be seen that this is a way of obtaining the equivalent of two series
masses without a common zerovelocity (ground) point. This will be
illustrated in Example 3.3,
(uit)
et
13. 3.14. Mobilitytype symbol for a Fic, 3.16, Sixelement mechanical de
floating, lever. view.
Example 8.1. The mechanical dovieo of Fig. 3.15 consists of piston of mass Mwy
iding on an oil surface inside a eylinder of mass Afw;.. This eylinder in urn slides
in an oiled groove cut ina rigid body... ‘The sliding. (viscous) resistances art Iewy and
Re respectively. ‘The cylinder is held by a spring of compliance Cw. ‘The mechan
cal generator maintains a constant sinusoidal velocity of angular frequency wy whose
rms magnitude is xim/see. Solve for the force J produced by the generator.
Solution. Although the force will be determined ultimately from an analysis of the
mobilitytype analogous circuit for this mechanical deviee, itis frequently useful to
raw & mechanicalcircuit diagram. This interim step to the desiel circuit will be
especially helpful to the student who is inexperienced in the use of analogies. Its use
virtually eliminates errors from the final eieuit
‘To draw the mechanical circuit, note fst the junction points of two oF more ele
nents, This locates all element terminals which move with the same velocity. "Phere
ue in this example two velocities, w and us, in addition to “ground,” or zero velocity
These two velocities are represented in the mechanicalcireuit diagram by the veloetica
of two imaginary rigid bars, land 2 of Fig, 3.16, which oscillate in a verlical direction
The cirevit drawing is made by attaching al clement terminals with velocity 1 to tho
iret bar and all terminals with velocity wz tn the second hat. All terminals with zero
Past Vi) MECHANICAL CIRCUITS 59
velocity are drawn to a ground bar. Note that & mass always has one termiaal on
ground.t "Three elements of Fig. 3.15 have one terminal with the velocity a the
generator, the mass Mus, and the viscous resistance wi, These are attached to
bar 1. Four elements have one terminal with the velocity us the viscous resstaacet
‘Ru: and Ra, the mass Mus, and the compliance Cy, These ate attached to bar 2
Five elements have one terminal with zero velocity: the generator, both masecs, the
‘viscous resistance Ry, and the compliance Cx.
Fic. 3.16. Mechanioal cireut for the device of Fig. 3.15,
We are now in a position to transform the mechanical circuit into « mobilitytype
analogous circuit, This i accomplished simply by replacing the mechanical elements
with the analogous mobiltytype elements. The circuit becomes that shown in Fig.
3.17, Remember that, in the moblitytype analogy, force “flows” through the
elements and velocity is the drop across them, The resistors must have lower case
v's written alongside them. As defined shove, rw = 1/Ry, aad the unit is the mks
‘mechanical mob,
‘The equations for this circuit are found in the usual manner, using the rules of
Table 3.1. Let us determine zy = u/f, the mechanical mobility presented to the
8.17. Mobilitytype analogous eireuit for the device of Fig. 8.15.
enerator. ‘The mechanical mobility of the three elements in parallel on the right=
Dband side of the echematie diagram is
a
ho
Theta * tan * Jat
om r
Jour + Bas + se
Including the element ry: the mechanical mobility for that part ofthe circuit through
which Je flows is, then,
oeetaeetesssssenae =
ao™
r
SoMa Rare +
‘Note that the input mechanical mobility ew is given by
“Tih
{ An exception to this rule may oeeur when the mechanical device embodies one or
more floating levers, a8 we just learned.60 BLEGTROMECHANOACOUSTICAL CIRCUITS (Chap3
and
f= apcity, 7 Moe
Substituting /. al fs into the second equation preceding gives us the input mobility
tse)
2.355)
mt —— 1
Jes + Neus + 5h
he res is
fm Zyu — newtons 1)
Example $2. Ava further ©
masses of Zand 4 kg shown in Fi
maple of « mechanical eireuit, let us vonsider the two
3.18, They are assumed to rest on a frietionless
uo
mp +R
OORT ese
‘ist iebonese
‘race
Fa, 3.18. Threedement mechanical deviee.
plane surface and to be conncetod together through a generator of constant velocity
that it also free to slide on the Fretiouloas plane surface, Let ite velocity be
uy = 2.08 1000 ean one
Draw the mohilitytype analogous eieuit, and determine the force f produced by the
generator. Also, determine the mobility presente to the generator,
Solution. ‘The masses do not have the sume velovity with respect to ground. The
difference hetweon the velocities of the two mass is ux. The element represeating
mass is that shiown in Fig. 3:20 with one end grounded and the other moving at the
Velocity of the mas,
— the moblitytype vieeait for this example is shown in
Fig. 2.19. The velority up equals ay + is, where wy i
t Mont the velocity with respect to ground of My), and uz is that
us
Theta) + Weed
Some
Fic. 3.19, Mobilityty = dellandvs
nonin erent for the Maan
device of Figs 3.18, ai 4X10? — go nomtone tain
a+ ve > ny
The j indieates that the time phase of the force is 0° leading with respect to that uf
the Velocity of the generator.
Part Vi) MECHANICAL CIRCUITS 6
CObviowsly, when one mass is large compared with the other, the foree is that neces
sary to mov the sinaler one alone. ‘This example revels the only type of ease
whieh masses can be in verias without the introduction of floating levers. At most,
‘only two masses ean Ie in series boeansr a common grou is necessary
‘The mobility presented to the generator ie
ee (0), = Ma tan
“ FF ane Gol wiM ves
= ja seg = —FES X 1 oe a8)
Example 3, An example of « mechanical devi embodying a floating lever ie
shown in Fig. 3:20.” The maine atached pointe 2 and 3 may be sansa to be
GAs
fe:
o Mee *
te :
(2) Mie
@) (oy
Fic, 8.20, (a) Mechanical device em Fis. 8.21, (a) Mobilitytype analogous cr
hdying a floating lever. (@) Mechan cuit for the deviee of Fig 8.20. (0) Same
‘eal diagram of (a). The eompliances as (a) but with transformers removed,
ff the springs are very large so that all
Sf feand fz go to move M ys and Ms
resting on very cormpliant springs, The driving force fi will be assurned to have a
Srequeney well above the resonance frequencies of the masses and their spring supports
so that
Algo, assure that anne is attached 0 the weightless lever bar at point 1, with #62 ELECTROMECHANOAcoust
AL CIRCUITS (Chap. 3
mobility
Solve for the total mobility presentnd to the constant force generator fi
Solution, By inspection, the mobilitytype analogous eireit is drawn as shown in
Fig. §.21@ and 0, Solving for ew = ui/fy we get
i apes @.19)
Fat Fa + Me]
Note that if f» 0, the mobility is simply that of the mass Man. Also, if +0, the
mobility is that of Muy and Mya, that is,
hie F Mand
8.20)
t is possible with one or more floating levers to have one of more masses with no
sound terminal
pant VII Acoustical Circuils
3.4. Acoustical Elements. Acoustical circuits are frequently more
lifficult to draw than mechanical ones because the elements are less easy
o identify. As was the case for mechanical cireuits, the more obvious
orms of the elements will be useful as an intermediate step toward draw
ng the analogous circuit diagram. When the student is more familiar
vith acoustical circuits, he will be able to pass directly from the acoustie
levice to the final form of the equivalent circuit.
In acoustic devices, the quantity we are able to measure most easily
vithout modification of the device is sound pressure. Such a measure
nent is made by inserting a small hollow probe tube into the sound field
the desired point. This probe tube leads to one side of a microphone
iaphragm. ‘The other side of the diaphragm is exposed to atmospheric
ressure. A movement of the diaphragm takes place when there is a
ifference in pressure across it. ‘This difference between atmospheric
ressure and the pressure with the sound field is the sound pressure p.
Because we can measure sound pressure by such a probetube arrange
nent without disturbing the device, it seems that sound pressure is
nalogous to voltage in electrical circuits. Such a choice requires us to
onsider current as being analogous to some quantity which is propor
ional to velocity. As we shall show shortly, good choice is to make
urrent analogous to volume velocity, the volume of gas displaced pe
econd
Part VIE ACOUSTICAL CIRCUITS 63
A strong argument can be made for this choice of analogy when one
considers the relations governing the flow of air inside such acoustio
devices as loudspeakers, microphones, and noise filters. Inside a certain
type of microphone, for example, there is an ar eavity that connects to
the outside air through a small tube (eee Fig. 3.22). Assume, now, that
the outer end of this tube is placed in a sound wave. The wave will
cause a movement of the air particles in the tube. Obviously, there i a
junction between the tube and the cavity at the inner end of the tube at
point A. Let us atk ourselves the question, What physical quantities are
Continuous at this junction point?
int the sound pressure just inside the tube at 4 is the same as that in
the cavity just outside A. That is to say, we have continuity of sound
Pressure. Second, the quantity of air leaving the inner end of the small
tube in a given interval of tm isthe quantity that enters the cavity
thestme interval of time. Thatis, the mass per second of gas leaving the
small tube equala the iass per second of gas entering the volume,
Because the pressure is the same at both ee
places, the density of the gas must also be are,
the same, and it follows that there is eon.
Einuity of volume velocity (cubie meters per
secuud) at this Junction. Analogously, in ete
the case of eesti, there sconinaty of ad
electric current at a junction, Continuity F ed cavity con
of volume velocity must exist even if there 3eag, W tbe outer
are several tubes or eavities joining nearone tonal area The jncton
point. A violation of the law of conserva Plane between the tube and the
tion of mass otherwise would occur. ee
We conclude that the quantity that flows through our acoustical ele
ments must be the volume velocity U in cubie meters per second and the
drop across our coustieal elements must be the pressure pin newton per
square meter. This conclusion indicates that the impedance type of
analogy is the preferred analogy for acoustical circuits. The product of
the effective sound pressure p times the inphace component of the effec:
tive volume velocity U gives the acoustie power in watts.
Tn this part, we shall discuss the more general aspects of acoustical
circuits. In Chap. 5 of this book, we explain fully the approximations
involved and the rules for using the concepts enunciated here in practical
problems.
Acoustic Mass M4, Acoustic mass is a quantity proportional to mass
but having the dimensions of kilograms per meter‘. It is associated with
44 mass of air accelerated by a net force which acts to displace the gas
without appreciably compressing it. ‘The concept of acceleration without
compression is an important one to remember. "It will assist you in dis
tinguishing acoustic masses from other elements64 ELECTROMECHANOACOUSTICAL CIRCUITS (Chap.3
‘The acoustical element that is used to represent an ucoustic mass is a
tube filled with the gas as shown in Pig. 3.23
‘The physical law governing the motion of a mass that is acted on by a
force is Newton's second law, (Q) = Mw du(t)/dt, ‘This law may be
txpromed in acoustical term as follows,
Med
pw = Bree
alo '3
0 = Ma 21)
here p() = instantancous diference between presures in netons per
square meter existing at each end of a mass of gas of Mu kg
undergoing acceleration
Mu/S? = acoustic mass in kilograms per meter‘ of the gas
undergoing acceleration. ‘This quantity is nearly equal to
the mass of the gas inside the containing tube divided by
the square of the crosssectional area, To be more exact
we must note that the gas in the immediate vicinity of the
snds of the tube also adds to the mass, Hence, there are
‘cud curvecliuns” which must be considered. ‘These cor
rections are discussed in Chap. 5 (pages 132 to 139)
U(® = instantaneous volume velocity, of the gas in cubic meters
per second across any crosssectional plane in the tube.
‘The volume velocity Ud) is equal to the linear velocity
u(®) multiplied by the crosssectional area S,
In the steady state, with an angular frequency w, we have
M.
p= joMU (8.22)
where p and U are taken to be rms complex quantities.
aye at e
‘rai
vet
Impedance type Moby type
(a) )
Fic. 3.24. (a) Impedancetype and (2) Pra. 8.25, Bnclosed volume of air V with
mobilitytype symbole for an acoustic opening for entrance of pressure varia
the impedancetype analogous symbol for acoustic mass is shown in
Fig, 3.24a, and the mobilitytype is given in Fig. 3.24). In the steady
state, for either, we get Eq. (3.22). ‘The arrows point in the direction of
positive flow or positive drop,
Poot VU ACOUSTICAL GIREUETS 6
Acoustic Compliance Cy. Acoustic compliance is a constant quantity
having the dimensions of meter* per newton, It is associated with a
volume of air that is compressed by a net force without an appreciable
average displacement of the center of gravity of air in the volume. In
other words, compression without acceleration identifies an acoustic
complianee.
‘The acoustical clement that is used to represent an acoustic eompliance
is a volume of air drawn as shown in Fig. 3.25.
‘The physical law governing the compression of a volume of air being
acted on by a net force was given as f(0) = (1/Cu)fu(t) dt. Converting
from mechanical to acoustical terms,
Oye
SOAS
fooge wae
or
me) (3.23)
where p(@) = instantaneous pressure in newtons per square meter acting.
to compress the volume V of the air,
CyS* = acoustic compliance in meters* per newton of the
volume of the air undergoing compression. ‘The acoustic
compliance is nearly equal to the volume of air divided by
‘yPo, as we shall see in Chap. 5 (pages 128 to 131).
U(® = instantaneous volume velocity in cubie meters per second
of the air flowing into the volume that is undergoing com
pression. ‘The volume velocity U(C) is equal to the linear
velocity u(0) multiplied by the crosssectional area S.
In the steady state with an angular fre
quency a, we have
u 
* iat 3.28)
C,
uv >
a
te ta fe Be
ee ee ae
quantities. Umpenetpe wipe
‘The impedancetype analogous element for” (gy ()
P66
‘TROMECHANOACOUSTICAL CIRCUITS Chap. 3
‘The acoustic clement used to represent an acoustic
resistance is a finemesh sezeen drawn as shown in
Fig. 8.27
‘The reciprocal of acoustic resistance is the acouse
tic responsiveness rs. The unit is the mks acoustic
mohm with dimensions meter per second per newton.
1.327, Finemesh _ The physical law governing dissipative effects ina
Sc wblover mechanic! system vas given by 0) = Raw, oF ia
Ssanacousticalsym terms of acoustical quantities,
bol for eceetiolse  teal quantities,
Pe) = RU = LUD (8.25)
where p(®) = difference between instantaneous pressures in newtons per
square meter across the dissipative element. In the steady
state p is an rms complex quantity.
Ra = Ru/S* = acoustic resistance in acoustic ohms, ie, newton
seconds per meter
14 = rwS? = acoustic responsiveness in acoustic mohm:
meter* per newtonseconds
U@ = instantaneous volume velocity in eubic meters per second of
the gas through the crosssectional aren of resistance. In
the steady state U is an rms quantity.
‘The impedancetype analogous symbol for acoustic resistance is shown
in Fig. 3.280 and the mobilitytype in Fig. 3.280,
Acoustic Generators, Acoustic generators can be of either the constant
volume velocity or the constantpressure type. The prime movers in our
uv »
m7
trite th roe
Ly ie of} Lot
(@) ir) () 6)
Fic. 3.28 (a) Impedancetype symbol Fra, 9.29. a) Impedancetype an
fo acoustic estanen aod @S'mabity=moitetgye apes or eos
type symbol for acoustic reponsivenesa. pressure generator,
acoustical circuits will be exactly like those shown in Figs. 3.7 and 8.9
exeept that us often will be zero and 1, will be the velocity of a small piston
of area S. Remembering that w= a — us, we see that the generator of
ig. 8.7 has a constantvolume velocity U'= uS and that of Fig. 3.9 «
constant pressure of p = f/S.
‘The two types of analogous symbols for acoustic generators are given
in Figs, 3.29 and 3.°% ‘The arrows point in the direction of the positive
Past VIS ACOUSTICAL CIRCUITS 7
terminal or the positive flow. As before, the double eircles indicate zero
impedance or mobility and a dashed line infinite impedance or mobility.
Mechanical Rotational Systems. Mechanical rotational systems are
handled in the same manner as mechanical rectilineal systems. ‘The
following quantities are analogous in the two systems,
Rectilineal systems Rotational systems
J = force, newtons 1 = torque, newtonm
1 = velocity, m/see 9 = angular velocity, radians ace
= displacement, m 4 = angular displacement, radians
Zu = fie = mechanical impedance, mks Zp = 7/2 = rotational impedance, mks
rmechanieal ohms rotational obme
su =e/f= mechanical mobility, mks zg = 6/7 = rotational mobility, mks
mechanieal mohme rotational mobs
Ry = mechanical resistance, mks me Re = rotational resistance, mks rota
chanical ohms tional ohms
ry = mechanical responsiveness, mks rx = rotational responsiveness, mks 0
mechanical mohms tational mohms
Mog = mass, kg In = moment of inertin, kgm?
Cur = mechanieal compliance, m/aewlon Cy = rotstionel compliance, radians/
Wy = mechanical power, watts Wa = rotational power, watts
Bxample 84, ‘The acoustic device of Fig. 3.31 consists of three cavities Vs, Vs and
Vs, two finesmesh sereens Its» and Ru» four short lengths of tube 7, Ts, Ts, and Ta,
and a constantpressure generator. Because the air in the tubes is aot confined, it
Pot of
(@) (b) own :
Fro, 3.80, (a) Impedancetype and (@) Fra. 831. Acoustic device consisting of
imobilitytype symbols for a constantvol four tubes, thre eavities, and two screens
time velocity generator. driven by's coustantpressure generator,
experiences negligible eampression. Because the ai in each ofthe cavities is confined,
it experiences little average movement. Let the force of the generator be
‘f= 10 cos 1000 newtons
the radius of the tube @ = 0.5 om; the longth ofeach of the four tubes t = 5 em; the
‘volume of each of the three eavities V = 10/em#;and the magnitude ofthe two acous
tic resistances Ra = 10 mks acoustic ohms. Neglecting end corrections, solve for the
volume velocity Usat the end of the tube 7.
‘Solution. Remembering thst there is continuity of volume velocity and pressure
at the junctions, we ean draw the impedancetype analogous eireuit from inspection
It is shown in Fig. 3.32. The bottom line of the schematic diagram represents
laimospheric pressure, which means that here the variationel pressure p is equal to
ero, At each of the junctions of the elements 1 to 4, «different variational pressure
fan be observed. ‘The end of the fourth tube (7) opens to the atmosphere, which
requires that M aa be connected directly tothe bottom line of Fig. 8.32
Note that the volume velocity of the gas leaving the tube 1, is equal to the sum8 rhoMe
HANOACOUSTICAL CI
CULTS. (Chap a
of the volume veloities of the gas entering ¥; and 7 ‘The volume veloity of the
tess Teaving 7. isthe same as that flowing through the sereen Mea and is equ to the
sum of the volume velocitin of the gas entering Vs and 7.
‘One test of the validity of an analogous circuit is its beliavior for direct current, If
one removes the piston and blows into the end ofthe tube 7, (Fig. 3.31), a steady Row
of air from Tis observed. Some resistance to this fow will he offered by the two
seteens Mas and Las. Similarly in the sehematie diagram of Fig, 3.32, steady pees.
sure p will produce a steady tow U through Ma resisted only by Rea: and Ita.
to, 3.32, Impeslancetype analogous eireut for the acoustie deviee of Fig. 3.31
Asan aside, let us note that an acoustic compliance ean occur in a circuit without
one of the terminals being at ground potential only if it iy produced by an elastir
diaphregm. For example, if the resistance Ra, in Fig. 3.31 were replaced by an
impervious but clastic diaphragm, the element Ra; in Fig. 3.32 would be replaced
by a compliancestype element with both terminals shove ground potential. In this
‘ease a steady flow of air could not he maintained through the device of Fig. 3.31, a=
«an alao be seen from the eirewit of ig. 3.82, with Ray replaced by a compliance
Determine the element sizes of Fig. 3.32
J 2 10*cos 1000
pag = perm ae reset
Mas = Max = Mas = Mac = 22 = 138X998 «750 kame
= 7.15 X 10 m*/newton
v1
Ca = Cn = Ca = =e
Ray = Ray = 10 mks ncoustic ohms
iecineuit theory, we solve for Uy indircetly, First, arbi
tarily let Uy = 1 mut/se, an the ratio p/Us.
Pe = juM alls = 7.5 X 10 newtons/m*
Uy = feCaps = “5.36 X 10° m*/a0e
WUT Us = 006
i = (Ras + jalfac + py = 9.46 + 514.6 X 108
Us = jaune = “0.1088 + 530.77 10°
Uy Us m 01842 + 50.77 x 107
Ba = (ay + joM as) + Bs = 17.37 + 32.001 x 108
Oy = jeCasps = 0.1496 + 51.242 X 10°
U 2 Us $0, = 0.602 4 51.919 x 10"
jeMaW + ps = 1898 +7261 X10 = F fore = 1
‘The desired value of Uo is
0.1273 cos 10008
15.95 + 92.01 x 1
4.88 10° cos (10001 ~ 90°)
4.88 X 10* sin 10000
Jn other words, the impedance is principally that of the four aoouste masses in series
0 that Ue Ings p by nearly 90°.
a
Ueno
Pact VII) Acous:
AL crneurrs co)
Example 8.5. A Iichnbolts resonator is frequently sed asa means foro
‘an undesired froquency component From an acoustic aystern. An example is given it
Fig. 3.392, A constantforee generator @ produces a serie of tones, among whieh is
tance is 500 mks acoustic ohms. If the tuhe haa a crosssectional
1, = ly = 5.0m, fy = Lem, V = 100 em?, and the crosssectional aren of Isis 2 em’,
‘what frequeney is eliminated from the #93
@
M, M
Mas
a Dire
cn
@)
Pia. 8.83. (a) Acoustic device consisting of a constantforce generator G, piston P,
tube T with length f+ 1, microphone M, and Helmolta resonator 8 with volume V
and connecting tube as shown, (B) linpedaneetype analogous circuit for the deviee
of (a)
Solution. By inspection we may draw the impedancetype analogous circuit of
‘ig. 4.386, The element sizes are
18 x 0.05
3x10
May = Ma B= = 118 kg/mé
Ply 3IBX 001 59 bg ns
May = Bb = BOM 50 ke
10
eo 7 * m/newton
Co = 5B,” Ta cage ~ TS x TOA me
Re) = 500 mks neoustic ohms
tis obvious that the volume velocity Us of the transducer Af willbe zero when the
shunt branch is at resonance, Hence,
eh = I — 1540 satin one
Vitals ABB
f= 285 epROMECHANOACOUSTICAL ELRCULTS: (Chap. 3
parr VIL Transducers
A transducer is defined as a device for converting energy from one form.
to another. Of importance in this text is the electromechanical trans
ducer for converting electrical energy into mechanical energy, and vice
versa. ‘There are many types of such transducers. In aeousties we are
concerned with microphones, earphones, loudspeakers, and vibration
pickups and vibration producers which are generally linear passive
roversible networks.
‘The type of electromechanical transducer chosen for each of these
instruments depends upon such factors as the desired electrical and
mechanical impedances, durability, and cost. It will not be possible here
to discuss all means for electromechanical transduction. Instead we
shall Timit the discussion to electromagnetic and electrostatic types.
Also, we shall deal with mechanoacoustic transducers for converting
‘mechanical energy into acoustic energy.
3.6. Blectromechanical Transducers. Two types of electromechanical
transducers, electromagnetic and electrostatic, are commonly employed
in loudspeakers and microphones. Both may be represented by trans
formers with propertico that permit the joining of mechanical aud elec
‘rieal cireuits into one schematie diagram.
Electromagneticmechanical Transducer. This type of transducer can
be characterized by four terminals. Two have voltage and curr
associated with them. ‘The other two have velocity and force
the measurable properties. Familiar examples are the moving
loudspeaker or microphone and the variablereluctance earphone or
microphone.
‘The simplest type of movingcoil transducer is a single length of wire in
uniform magnetic field as shown in Fig. 3.34. When a wire is moved
upward with a velocity was shown in Fig, 3.34a, a potential difference e
will be produced in the wire such that terminal 2 is positive. If, on the
other hand, the wire is fixed in the magnetic field (Fig. 3.34) and a cur
rent iis caused to flow into terminal 2 (therefore, 2 is positive), a force will
be produced that acts on the wire upward in the same direction as that
indicated previously for the velocity.
‘The basic equations applicable to the movingcoil type of transducer are
fa Bi (8.26a)
e= Blu (8.260)
where i = electrical current in amperes
J = “opencircuit” force in newtons produced on the mechanical
circuit by the current i
B = magneticflux density in webers per square meter
Pact VIN TRANSDI
RS nm
I= effective Jength in meters of the electrical conductor that
moves at right angles across the magnetic lines of force of flux
density B
1 = velocity in meters per second
= “opencircuit” electrical voltage in volts produced by a
velocity u
‘The righthand sides of Eqs. (3.26) have the same sign because when
w and f are in the same direction the electrical terminals have the same
sign.
+ i
2 J
x (lle HAle
if
Veloty x wth postive The current produces a
‘recon upward fove f ating upward
@ @)
Fro, 84, Simplified form of movingeil transducer consisting of single length of
wire cutting a magnetic field of flux density J. (a) The conductor is moving ver
tically at constant velocity so as to generate an opencircuit voltage across terminals
Vand 2. (b) A constant current i entering terminal 2 to produce force on the eon
ductor in & vertical direation.
‘The analogous symbol for this type of transducer is the “ideal” trans
former given in Fig. 3.35. The “windings” on this ideal transformer
have infinite impedance, and the transformer obeys
Eqs. (3.26) at all frequencies, including steady flow. _{ B!1_{
The mechanical side of this symbol necescarily is of
the mobility type if current flows in the primary. te
‘The invariant mathematical operations which this
symbol represents are given in Table 3.1, They lead Fre. 3.35. Analo
directly to Eqs. (3.26). The arrows point in the ‘gous symbol for ‘
directions of postive flow or positive potent. eepromagnetic
Electrostaticmechanical Transducer. This type of ca ae Fig. x aa
transducer may also be characterized by four termi 'e mechanical side
nals. Av two of them vatagesand eurentacan be yeh the MODIY
mmeasured. "At the other te, forts and velostes
can be measured. ‘This transducer is satisfactorily described by the fol~
lowing mathematica! relation,
= (8270)
fan (27)72 ELECT RO.MEGHANO ACOUSTICAL CIREULTS [Chap 3
where ¢ = “opencircuit” electrical voltage in volts produced by
displacement §,
displacement in meters of @ dimension of the piezoclectrie
device
= clectrical charge in coulombs stored in the diclectrie of the
piezoelectric deviee.
J= “opencireuit" force in newtons produced by an electrical
charge 4,
upling coefficient with dimensions of newtons per coulomb
or volts per meter. It is a real number when the network is
Tinear, passive, and reversible.
‘An example is a piezoelectric erystal microphone such as is shown in
Fig. 8.36. A force applied uniformly over the face of the crystal causes
an inward displacement of magnitude & As. result of this displacement,
4 voltage © appears across the electrical terminals 1 and 2. Let us
assume that a positive displacement
ware (nWard) of the erystal causes termi
wiveyaai nal 1 to become postive. For small
€
‘Sceneried displacements, the induced vollage is
stituted proportional to displacement, ‘The
Tue” inverse of thi effect occurs when no
$e!" external force acts on the erystal face
a but an electrical generator is con
© crete nected to the terminals 1 and 2. If
O ws the external generator is connected s0
that terminal 1 is positive, an internal
duces ented ee ei areatal 8 Force f is produced which acts to ex
pand the crystal. For small displace
ments, the developed force f is proportional to the electric charge g stored
in the dielectric of the crystal.
Equations (8.27) are often inconvenient to use because they contain
charge and displacement. One prefers to deal with current and velocity,
which appear directly in the equation for power. Conversion to current
and velocity may be made by the relations
ua B= jot (3.280)
a
dt
jg (3.288)
1 The coupling coetMcient is frequently defined differently in advanced texts on
sectrostaticmochanical transilucres, For example, in some teats it ib defined as
the squure root of CyCar'/8, where Ciy and Cy are defined after By, (7). Phe
muthor does not intend that the definition for coupling coeficient in this text should
pe adopted ag standard; rather, the teri ie used simply for convenience in the
Jiseussion,
Part VIN TRANSDUCERS 3
so that Eys. (8.27) become
(8.29a)
(3.290)
Unfortunately, the usual analogous symbol for this type of transformer
is not as simple as that for the electromagneticmechanical type. Two
possible forms are shown in Fig. 3.37. ‘The mechanical sides are of the
impedancetype analogy. Let us discuss Fig. 3.37a first,
The element (%, is the mechanical compliance of the transducer. In
order to measure C',, a sinusoidal driving force f is applied to the trans
ducer terminals 3 and 4, and the ; a
resulting, sinusoidal displacement g—tos
ismeasured. Duringthismeasure , &
ment the electrical terminals are ce 7
shortcircuited (e= 0). Averylow 2 “
driving frequency is used so that (a)
the mass reactance and mechan 4 Grit y
ical resistance can be neglected, 1H °
The clement C’; is the electrical f We sew
capacitance of the crystal mea, .
sured at low frequencies with the (o)
mechanical terminals opencir pio, 3.37, Two forma of analogous apm
cuited (u = 0). Application of a bols for pisoslctse anedutor, ‘The
Current it. the primary will ectaned ‘ale te of the pens
produce a voltage across the con ‘¥?*
denser (%, of ¢/jwC',. This in turn produces an opencircuit force
‘wanedicer
(2)
if
oa
@)
Bit fe t Bie
%
Ba
‘a renin’
Ly
&
()
Fie. 8.48. Combined electrostatieelectromagnetic transducers.
ial diagram of the device. (2) Analogous circuit with mobilities on mechanical sie,
‘The @ operator ia used for the piezoelectric transduce? (c) Same as (3), except that
24 replaces the three parallel mobilities as shown by ().
D POWER 89
ie Ys
Loud Im
Ce Rex?
ote
(ny
(@) Block mechan
(e) Dual of. )
Because the circuit of (2 has infinite mobility, 6) simplifies to thi form. (@) and (A)
Solution of U by superponition,90 ELECTRO.MECHANOACOUSTICAL CIRCUITS [Chap. 3
Fig. 3470, Solving, we get
Aw fate =X 10*p 8280027 X 109
= 3078 x 10"
fo = jallumn = ~834 X10"
f= fot fem 814 x 10%
Sa Ly = 10.78 x 104 — 9 x 10
In = ESOT X10" 2 185 newton/mt
= 189.8 db re 0.0002 mierobar
st = ato
Example 310. Two transducers, ones piezoelectric erystal and the other a moving
coil in a magnetic fel aro connected to.a mane Ms f 1.543 kg as shown in Fig. 3.480
Determine the stored electrical energy inthe condenser Cy at 100 eps for the following
constants:
= volt
Re = 10 ohms
Bom i weber/m?
130m
Cy = 4X 107 Farad
Maas = LOkg.
Cua; = 10" m/newton
‘r= 1.28 X10"
= 628 radians/eco
Solution, The transducers are shown schematically in () of Fig. 3.48, A further
simplification of this diagram is shown in (e). Let us determine the value of zu first.
We note that the dual of (dis given by (e),
fhe t= jth 4) hy
= s1620 + 968) ~ j1503 ~ 0
In other words, the mobility ie infinite at 100 eps, Hence, eireit (¢) simplifies to
that shown in (), By superposition, #; can be broken into two parts and 4) given
by the two circuits (9) and (A), 0 that f= sf — ¥.
28 x07
Ho ~ 1X30 XO
680
f= it! ~ it = ja X 104
lial = 147 X 10° an
The voltage drop across the enpacitor is
Ied = ATX 104 108 = 59 volts
The electric stored encegy on the capacitor Ce i
B4Crlelt
X 104 X 8.48 X 108 = 7 104 wathnee
CHAPTER 4
RADIATION OF SOUND
In order fully to specify a source of sound, we need to know, in addition
to other properties, its directivity characteristics at all frequencies of
interest. Some sources are nondirective, that is to say, they radiate
sound equally in all directions and as such are called spherical radiators.
Others may be highly directional, either because their size is naturally
large compared to a wavelength or because of special design.
‘The most elementary radiator of sound is a spherical source whose
radius is small compared to onesixth of a wavelength. Such a radiator
is called a simple source oF a point source. Ita propertion aro eposified
by the magnitude of the velocity of its surface and by its phase relative
to some reference. More complicated sources such as plane or curved
radiators may be treated analytically as a combination of simple sources,
each with its own surface velocity and phase.
‘A particularly important consideration in the design of loudspeakers
and horns is their directivity characteristics. This chapter serves as an
important basis for Inter chapters dealing with loudspeakers, baffies,
horns, and noise sourees.
The basic concepts governing radiation of sound must be grasped
thoroughly at the outset. It is then possible to reason from those con
cepts in deducing the performance of any particular equipment or in
planning new systems, Examples of measured radiation patterns for
common loudspeakers are given here as evidence of the applicability of
the basic concepts,
pant X Directivity Patterns
"The directivity pattern of a transducer used for the emission or for recep
tion of sound is a description, usually presented graphically, of the
response of the transducer as a function of the direction of the transmitted
ot incident sound waves in a specified plane and at a specified frequency.
a92 RADIATION OF SOUND (Chop. 4
The beam width of a directivity pattern is used in this text as the
angular distance between the two points on either side of the principal
axis where the sound pressure level is down 6 db from its value at @ = 0.
4.1. Spherical Sources." A spherical source is the simplest to eon
sider because it radiates sound uniformly in all directions. As we saw
from Eq, (2.62), the sound pressure at a point a distance rin any direction
from the center of a spherical source of any radius in free space is equal to
V2 AS jury
pin) = ¥ (4)
where A, is the magnitude of rms sound pressure at unit distance from
the center of the sphere,
_ os ro
or
as
ner = so
Dr=0DR
cy
Fro, 4.1, Directivity pattern for a nondirectional source, Such a pattern is drawn
on a particular plane intersecting the center of the source. ‘The directivity index DI
defined in Pare XI) equals 0 db at all angles,
Directivity Pattern. On a polar diagram, the directivity pattern on any
plane surface intersecting the center of such a spherical source is given in
Fig. 4.1. It is obviously a nondirectional source.
Simple Source (Point Source). For the special case of a very small
source, whose radius a is small compared with onesixth wavelength (that.
2d ed,, pp. 311826, MeGrawHill Book
2P. M. Morse, “Vibration and Sound,
Company, Ine., New York, 1948.
+1, B. Kinsler and A. Frey, “Pondamentals of Acoustics," pp. 168173, John
Wiley & Sons, Ine., New York, 1950,
* Morse, op. rit, pp. 312313,
Pact Xi pine
STIVETY PAY
Ns 93
is, tw & 1), the velocity at the surface of the sphere is [see Eq, (2.63)]
V2A, ¢
wat) Ln
pea joxfa
puts 42)
A sourve for which (his formula is valid is called a simple sonrce.
Substitution of Bq. (4.2) into Bq. (4.1) yields
Pan) (43)
Us = rms volume velocity in eubie meters per sevond of the very
small source and ix equal to (40°),
rms ound pressure in newLons per square meter ata distance
1 from the simple source
Strength of a Simple Source.* ‘The rms magnitude of the total sir flow
at the surfare of a simple source in cubie meters per second (or eubie
centimeters per sevond in the egs system) is given hy Us and iy called the
strength of a simple source.t
Intensily al Distance r. Ata distance r from the center of a simple
sonree the intensity is given by
P
walts/m? ay
When the dimensions of a source are much smaller than a wavelength,
the radiation from it will be much the same no matter what shape the
radiator has, as long as all parts of the radiator vibrate substantially in
phase. The intensity at any distance is directly proportional to the
square of the volume velocity and the frequency.
4.2. Combination of Simple Sources.‘ ‘The basie principles governing
the directivity patterns from loudspeakers can be learned by studying
combinations of simple sources. ‘This approach is very similar to the
consideration of Huygens wavelets in opties. Basically, our problem is
to add, vectorially, at the desired point in space, the sound pressures
arriving at. that point from all the simple sources. Let us see how this
method of analysis is applied.
Two Simple Sources in Phase, ‘The geometric situation is shown in
Fig. 4.2. Itis assumed that the distance r from the two point sources to
the point 4 at which the pressure p is being measured is large compared
‘with the separation 6 between the two sources.
“The spherical sound wave arriving at the point p from source 1 will have
fm sonar (ext the peak mnagitude of the otal ae Hos i
taude is used. tn the
‘IL F. Olam, "Elements of Aconstical Engineering,”
Nonteand Company, Ine, New York, 1947,
tea of the ris wag
V2 reine
ed ppt, 1D. Van
text, the xlrength of wimp nie94 RADIATION OF SOUND Chap. 4
traveled a distance r — (b/2) sin 6, and the sound pressure will be
pir) = YEAS pep tnmenan i
‘The wave from source 2 will have traveled a distance r + (b/2) sin 8, 50.
that
Vids
patra) =
ele Oa HD, (4.50)
‘The sum of pi + ps, assuming r>> b, gives
ViAy
pir) cig HAA (ileb ah fe gion) (46)
Multiplication of the numerator and the denominator of Eq. (4.6) by
ia, 42. Two simple (point) sources vibrating in phase located a distance b apart and
at distance r and angle 8 with respect to the point of measurement A.
exp (jrb sin 6/2) ~ exp (~Jjeb sin 6/A) and replacement of the expo
nentials by sines, yields
— VIAL joie staeeny Sit. [(2xb/d) sin 6]
ae “sin [(#b/d) sin @] ed)
‘The equation for the magnitude of the rms sound pressure p is
ip) = 24s  in [C2xb/) sin 6) en
7  Bsin (Gab /A) sin 8]
‘The portion of this equation within the straight lines yiels the directivity
pattern
Referring to Fig. 4.2, we see that if bis very amall compared with a
wavelength, the two sources essentially coalesce and the pressure at a
Part Xi] DIRECTIVITY PATTERNS 9
distance r at any angle @ is double that for one source acting alone, Th:
directivity pattern will be that of Fig. 4.1
As b gets larger, however, the pressures arriving from the two source
will be different in phase and the directivity pattern will not be a circle
0/4: 0109 08,
22: 013008
Tae ior
dea2 —pIe3008 ba 012008
@ a
Fic. 43. Directivity patterna for the two simple inphase sources of Fig. 4.2. Sym:
metry of the directivity patterns occurs about the xis Passing through the two
sources. Hence, only a single plane ie necessary to deeeribe the directivity character.
istis at any particular frequeney. The boxes give the directivity index at. 9 = 0"
One angle of zero directivity index is also indicated, (The directivity index is
‘cussed in Part XI.)
In other words, the sources will radiate sound in some directions better
than in others. Asa specific example, let b = /2. For @ = 0 or 180° it
is clear that the pressure atriving at a point 4 will be double that from
either source, However, for 6 = 490° the time of travel between the96 RADIATION OF SOUND
pt
pletely cancels the radiata from the other. Hee the presure at all
points along the +90° axis is zero. Remember, we have limited our dis
tumion tors
Directivity patterns expressed in devibels relative to the presuire at
= 0, are given in Fig, 43 for the two inephave sources with & = 2/4
\/25¥5 3/3; and 20,
very important observation can be made from the directivity p
tomns for thissimpl typo of radiator that applies to all types of rai
niger the exlont of Uno radiator (here, the ge
ill be tho prinipal lube along the @'= 0 axis at any give
quency and the greater the number of side lobes. As we shall see in the
next paragraph, iis posible tosup
pres tho side lobes, thats Lo sey,
those other than the prineipal lobes
atOand 180%, by simply increasing
the number of elements.
Linear Array of Simple Sources.*
‘The goometite situation for this
type of radiating array is shown in
_ Big 14. The rns sound pressure
ico vibrating input Mets produced at a point a by m identi
Tauern gpa SE eee ee
a straight Tine, the sources diy
and with the extent
tance
d= (x ~ Hb small compared with the dist
nity  in [(n 2) sin
livin (Gxb/A) sin
49)
As a special ease,
very large and th
before,
us assume that the number of points becomes
separation b becomes very small. ‘Then, as
: d= (n= 1b & nb (4.10)
a sin [xd
a aay
P= pe
where ps is the ide of the rms sound pressure at a distance r from
the array at an angle @ = 0. As before, it is assumed that the extent of
the array d is small compared with the distance r
Plots of Eq. (1.9) for w= 4 and d= A/M, 4/2, A, 3n/'
shown in Fig. 4.5. lar plots for n —+ & and b> 0, that is
4.5, and 4.6 for a given ratio
of array length Lo wavelength is in the suppression of the “side lobes.”
Part XI DIRECTIVITY PATTERNS 7
‘That is, sound is radiated well in the @ = 0° and @ = 180° directions for
all three arrays. However, as the array becomes longer and the number
of elements becomes greater, the radiation becomes les in other directions
‘than at @ = 0° and @ = 180°.
4/4: 0108 08
ana v2:011508
a
Ste
ERG
KRY
[iS
$e)
Si
@22 016008 42) 016608
@ a
iretivity patterns for a linear array of fou simple iphase sourees evenly
spaced overs length The boxes give the directivity index at @'= 0% One angle
Of zero dretivity index is also indested by the azom.
Doublet Sound Source. A doublet, sound source is pair of simple
sound sources, separated a very small distance b apart and vibrating in
opposing phase. The geometric situation is shown in Fig. 4.7. The dis
tance r to the point A is assumed to be large compared with the separation
b between the two sources
Tt ean be clearly seen that the sound pressure at, @ = 90° and @ = 270°
will be zero, because the contribution at those points will be equal from98 RADIATION OF SOUND (Chap. 4
the two sources and 180° out of phase. ‘The pressures at @ = 0° and
9 = 180° will depend upon the ratio of 6 to the wavelength ». For
example, if b = 2, we shall have zero sound pressure at those angles just
as we did for b = 4/2 in the case of two inphase sources, In the present
case, we have a maximum pressure at @ = 0° and @ = 180° for b = 2/2.
yA: 010308
N21 08
Toe od
@a/2 Ol8108 42 Di6208
@ @
Fro. 46, Directivity patterns for a linear line array radiating uniformly along ite
length d. ‘The boxes give the directivity index at @'= 0". One angle of sero direc
tivity index is also indieated by the arrow.
‘The usual case of interest, however, is the one for
ben (4.12)
In this case, the complex rms pressure py at a point A can be shown to
equal’
* Kinsler and Prey, op. cit, pp. 280285,
Part X} DIRECTIVITY PATTERNS 99
212 (442) anvem 1)
‘msstrength in cubic meters per second of each simple source.
ratio of the complex rms sound pressure py produced by the
oublet to the complex rms sound pressure p, produced by a simple source
is found by dividing Eq. (4.13) by Bq. (4.8). This division yields
pelty@) =
Ba 4 si) ost aay
‘When the square of the distance r from the acoustic doublet is large
Fra. 4.7. Doublet sound souree, This type of source consists of two simple (point)
sources vibrating 180° out of phase. They are located a distance b apart and are at
fan angle # and a distance r with respect 10 the point of measurement A. The lower
half of the graph shows by the aren ofthe circles the magnitude of the sound preagure
‘sa function of angle #. ‘The upper half of the graph showa the variation ofthe radial
‘and tangential componenta ofthe particle velocity as a function of angle @
compared with 2/36 (k'+* >> 1), Eq. (4.18) reduces to
pe = OUD cos em (4.15)
For this case the pressure varies with @ as shown in Figs. 4.7 and 4.8. It
changes inversely with distance r in exactly the same manner as for the
simple source.
Near the acoustic doublet, for r? < \*/36, Eq. (4.18) reduces to
pr = 29UB cos pen 10)100 WADIA
YN OF SOUND. Jechap.
Vor th eos @ as ws in Big 18, bt
iLchangesinversely with Use square of the distanver, We arestill assume
ing that r > 6.
Nearfield and Farfield. ‘The difference between nearsfield and farfield
behaviors of sources must always he borne in mind, When the diveetivity
pattern of a loudspeaker or some other sound source is presented in
technical publication, itis always understood that the data were taken 2
a distance r sulliiently large so that the sound prosure was decreasing.
nearly with distance along a radial line connecting with the source, a8
was the ease for Eq. (4.15). ‘This is the farfield ease. For this to be
true, two conditions usually have to he met. First, the extent 6 of the
the pressure also varies wi
‘DI=88 DB
W/ decty 20 0 (9/0)
(a) ()
Fis, 48. Dinectivity pattern for & doublet sound sonree. (a) Soundpressure ratio
Pipes. &. (b) 2 logis p/pe vs. 8. The boxes ive the dirvetivity index xt 9 = 0
ne angle of ser0 directivity index ie aleo indicated by the arrow,
radiating array must be small compared with r, and 7? must he large com
pared with 2/36. In acousties the size factor indicated is usually taken
to be larger than 3 to 10,
One more item is of interest. in connection with the acoustic doublet,
‘The particle velocity is composed of two compouents, one radially
directed, and the other perpendicular to that direction. At @ = 0 and
180° the particle velocity is directed radially entirely (soe Rig. 4.7). At
4 = 90 and 270° the particle velocity is entirely perpendicular to the
raglial line. In between, the radial component varies ws the eos @ and the
perpendicular component as the sin 0.
‘An interesting fact is that at @ = 90° and 270° a doublet sound source
appears to propagate a transversely polarized sound wave, ‘To dem
strate this, take two unbaflled small loudspeakers into an anechoie
chamber. “Unbaffled loudspeakers (trunsducers) are cyuivalent to
doublets herause the pressure increases on one side of the diaphragm
Part X} DIRECTIVITY PATTERNS 101
whenever iG decreases on the other Hold the two transducers about
0.5 m apart with both diaphragms facing the floor (wot facing each other).
Tot one transiuver radiate a lowfrequency sound and the other act as &
microphone connected to the input of an audio amplifier. As we see
from Fig. 4.7, no sound pressure will he produced at the diaphragm of the
microphone, but there will be transverse particle velocity. A particle
is always the result of a pressure gradient in the direction of the
velocity. ‘Therefore, the diaphragm of the microphone will be eaused
to move when the two transducers are held as described above. When
fone of the transducers is rotated through 90° about the axis joining the
units, the diaphragm of the microphone will not move because the
pressure gradient will be in the plane of the diaphragm, Hence, the
sound wave appears to be plane polarized.
‘You have now learned the elementary principles governing the diree
tional characteristies of sound sources. We shall be able to use these
Fia, 49, Rigid cireular piston in. rigid baffle. The point of measurement A ie
located at distance ran angle @ with respect ta the center of the piston
principles in understanding the measured or calculated behavior of some
‘of the more complicated sound sources found in acoustics,
43. Plane Piston Sources. Rigid Circular Piston in Infinite Bate
Many radiating sources can be represented by the simple concept of
vibrating piston located in an infinitely large rigid wall. ‘The piston ir
assumed to be rigid so that all parts of its surface vibrate in phase and ite
velocity amplitude is independent of the mechanical or acoustic loading
on its radiating surface. ‘The rigid wall surrounding the piston is usually
called a baffle, which, by definition, is a shielding structure or partition
used to inerease the effective length of the external transmission path
between the front and back of the radiating surface.
‘The geometry of the problem is shown in Fig. 4.9. We wish to know
the sound pressure ata point A located at a distance r and an angle @ from
the center of the piston. To do this, we divide the surface of the piston
into a number of small eloments, each of which isa simple source vibrating
in phase with all the other elements. ‘The pressure at 4 is, then, the sum
in magnitude and phase of the pressures from these elementary elements,102 RADIATION OF SOUND (Chap. 4
‘This summation appears in many texts' and, for the case of r large com
pared with the radius of the piston a, leads to the equation
V2 ifestuera® [@ (ka sin 6)
r asin
108 Jan
Ja(_) = Bessel function of the first order for cylindrical coordinates*
Fro, 4.10, Directivity patterns for a rigid cieular piston in an inGnite bale ws
function of ka = 2ra/hy mhere a is the radius of the piston. The boxes give the
directivity index at @= O°. One angle of zero directivity index sls indieated. The
DI never hecomes less than 8 db because the piston radiates only into halfspace,
The portion of Eq. (4.17) within the square brackets yields the direc
tivity pattern and is plotted in decibels as a function of @ in Fig. 4.10 for
six values of ka = 2ra/a, that is, for six values of the ratio of the cireum
ference of the piston to the wavelength.
‘When the circumference of the piston (2na) is less than onehalf wave
* Morse, op. ct, pp. 826846. A table of Besocl functions is given
age 444
Part XI DIREGTIVETY PATTERNS 103,
length, thats, ka < 0.5, the piston behaves essentially likea point souree.
When ka hecomes greater than 8, the piston is highly directional. We
see rom Fig. 4.24 that an ordinary loudspeaker aiso becomes quite direc
tive at higher frequencies in much the same manner as does the vibrating
piston,
Rigid Circular Piston in Bnd of a Long Tube? In many instances,
sound is radiated from a diaphragm whose rear side is shielded from the
front side by a box or a tube. If the box does not extend appreciably
beyond the edges of te diaphragm, its performance may be estimated by
comparison with that of a rigid piston placed in the end of a long tube
‘Phe geometrical situation is shown in Fig. 4.11. ‘The pressure at point
A is aguin found by suunming the pressures from a number of small ele
ments on the surface of the piston, each acting as a simple source. ‘The
A
Long estar ston
‘of slong tube, ‘The point of measurement
"espeet tothe eenter of the piston.
solution of this problem is complex, however, becwuse radiation ean take
place in all directions and the sound must diffract around the edge of the
tube to get to the lefthand part of space (Fig. 4.11). Hence, a theory
that includes the effects of diffraction must be used in solving the problem
analytically.
‘The results from such a theory are shown in Fig, 4.12 for six values of
ka. It is assumed here also that the distance r is large compared with a,
so that the directivity pattern applies to the farfield.
Rigid Circular Piston without Baffe. To complete the eases wherein
pistons are commonly used, we present the results of theoretical studies on
the directivity pattern of a rigid piston of radius a without any baffle,
radiating into free space. ‘These results are shown graphically in Fig
4.13 for four values of ka. It is interesting to note the resemblance
hetween these curves and those for an acoustic doublet. In fact, to 8
lirst approximation, an unbaflted thin piston is simply a doublet, because
an axial movement in one direction compresses the air on one side of it
and causes a rarefaction of the air on the other side.
Levine and J. Schwinger, On the Radiation ofS
le Pine, Phy. fen, T8: 4814106 (Feb, 1, 1948)
"PUM. Wiene, On the Kelation between the Sound Fields Raviated and Diffracted
hay Phan Obstacles, J. dewnat, Soe. Amer., 98: 697700 (1951),
from an Unflanged Cire:RADIATLON OF SOUND
Part X} DIRECTIVITY PATTERNS 105
Square or Rectangular Piston Sources.» natn SQUARE PISTON IN INFI
NITE marruE. This type of radiating souree is not very common in
acoustics. Tt suffices to say here that the directivity pattern for such a
piston in a plane perpendicular to the piston and parallel to a side is,
identical to that fora linear array of simple sources as given by Eq. (4.11).
‘The pattern in a plane that is parallel to either of the two diagonals differs
very little from that in a plane parallel to a side.
i PISTON IN INFINITE BAFFLE. The directivity pat
radiating source with dimensions d, and ds are given
,] sin ((rds/A) sin Oa]
(ed2/2) sin By (418)
* Olson, op. it, pp. 3940,106 RADIATION OF SOUND (Chap. +
where # = angle between the normal to the surface of the piston and the
projection of the line joining the middle of the surface and the
observation point on the plane normal to the surface and
parallel to ds
4, = same as 8, with ds substituted for dy
Note that the directivity pattern is equal to the product of the dir
fivity pattems for two line arrays at right angles to each other [s
Bq. (4.11)].
44. Curved Sources."® Tn the preceding paragraphs of this part
we have dealt, with the radiation of sound from straightline and plane
surface arrays. These types of arrays are closely resembled by open
ended organ pipes, directradiator loudspeakers, simple horns, and other
devives. One characteristic common to all these arrays, except the
doublet array, is that they become more directional as the ratio of their
length to the wavelength becomes greater. ‘This is usually an undesir
able trait for a loudspeaker to exhibit beeause it meaus that the spectrum
athe
F10.4.14. Parabolic megaphone suitable fur use by a cheesleader ina football stadium,
of music or speech ax reproduced will vary from one position to another
around the loudspeaker.
‘To overcome, in part, this increase in directivity with inereasing fre
queney, curved surfaces are commonly employed as sound radiators
Thi aces can be made up of a number of small loudspeakers or small
homs or as a megaphone with a eurved front.
Curvodline Source (Parabolic Megaphone). An example of a eurved
line source is the parabolic megaphone of Fig. 4.14. ‘The megaphone
‘opening is thin enough (1 in.) to be roughly equivalent to a simple line
source for frequencies below 4000 cps. ‘The horn is parabolie because the
sectional area is proportional to Uhe distance from the apex.
‘The sound pressure at a point A a distance r from the apex is found by
summing the pressures in amplitude and phase originating from an
assumed curved line coimposed of simple sources. When the distance ris
"Ibi, pp 40,
Part XI DIRECTIVITY PATTERNS 107
Fro, 4.15. Directivity patterns for the parabolic mogaphone of Fig 4.14 in the plane
containing the ate of the opening
(6)
Fie, 4.16, Multicclular horas with curved radiating fronts, (@) 3X 5 = 16 cells
0) 2X4 = Bealls,
large compared with the radius 2 of the horn, we obtain the directivity
patterns shown in Fig, 4.15.
‘This simple case illustrates basic principles applicable to all curved
radiating sources. At low frequencies the curved source is largely non
directional. As the frequency is increased, the souree becomes more108 RADIATION OF SOUND (Chap. 4
directional, achieving its minimum angle of spread when the chord of the
curved source is approximately equal to one wavelength. At high fre
quencies, the directivity pattern becomes broader, reaching its maximus.
width when it becomes equal to the width of the are expressed in degre
for example, 90° in Fig. 4.14,
Curvedsurface Sources (Multicellular Horns). ‘The most common
curvedsurface sources found in acoustics at the present time are multi
cellular homs. Two typical examples of such curvedsurface horus are
shown in Fig. 4.16,
‘The beam width of the directivity pattern was defined as the angular
distance between the two points on either side of the principal axis where
160
pean} i
i
e
Rarer oes
&
E
8
OT gg tg FR
Frequency in cycles per second
Fic. 4.17, Beam widths of multicellular horns constructed as shown in the inscst
und as sketehed in Fig 10,
the sound pressure level isdown 6 dbf from its valueat@ = 0. ‘The beam
widths of the directivity patterns for two, three, four, and five cell widths
of multicellular horns were measured on commercial units and are shown
n Fig. 4.17. ‘These data are useful in the design of sound systems,
It should be noted that the minimum beam width occurs when the are
the multicellular horn about equals X. Also, at high frequencies the
beam reaches a width of about n  25° — 15° for the size of cell shown in
Fig. 4.17. ‘The theoretical directivity indexes for these maximum widths
of beams are found from the nomogram in Fig. 4.22, which is based on the
sketch of Fig. 4.21
{No standard value has boon chosen for dhe nuanber of devibels down from the
4 = 0 value of the sonnal pressire level in determining beam width, Values nf 3s
12d 10 dy are often encountered in the literature
Pact XH DIRECTIVITY INDEX 109
pare XI Directir
ily Inder and Directivity Factor
Charts of the directivity patterns of sound sources are sufficient in
many cases, such as when the source is located outdoors at a distance
from reftecting surfaces. Indoors, it is necessary in addition to know
something about the total power radiated in order to calculate the
reinforcing effect of the reverberation in the room on the output of the
sound source. A number is calculated at each frequency that: tells the
degree of directivity without the necessity for showing the entire diree
tivity pattern. This number is the directivity factor or, when expressed
in decibels, the directivity index.
4.5. Directivity Factor [Q(/)]. ‘The directivity factor is the ratio of
the intensityt on a designated axis of a sound radiator at astated distance
r to the intensity that would be produced at the same position by a point,
source if it were radiating the same total acoustic power as the radiator.
Free space is assumed for the measurements. Usually, the designated
axis is taken as the axis of maximum radiation, in which case Q({) always:
exceeds unity. In some cases, the directivity fuctor is desired for other
axes where Q(/) may assume any value equal to or greater than zero.
4.6. Directivity Index (DI(f)}, ‘The directivity index is 10 times the
logarithm to the buse 10 of the divectivity factor.
DI(A) = 10 log QC) (4.19)
4.7. Calculation of Q(/) and DI(f). ‘The intensity J ata point removed
distance r from the acoustical center of a source of sound located in free
space is determined by first measuring the effective sound pressure p and
Jetting / = pl/a. If the source is a point source so that I is not a
function of @and is located in free space, the total acoustic power radiated
W, = tert
If the source is not @ point source, the total acoustic power radiated is
determined by summing the intensities over the surface of a sphere of
radius r. ‘That is, the total radiated power is
ar er)
where the coordinate of any point in space is given by the angles @ and
and the radius r (see Fig. 4.18) and p%(6,9,7) equals the meansquare
sound pressure at the point designated by 6, ¢, and r.
{See the definition for intensity om page 11, The intensity equals the sound pres
sure squared, divided by aye fora plane wave in Free space o10 RADIATION OF SOUND Chap. 4
Usually an analytical expression for p(0,6) does not exist. In practice,
therefore, data are taken at the centers of a number of areas, approxi
mately equal in magnitude, on the surface of a sphere of radius r sur
rounding the source. As an example, we show in Fig, 4.19a a spherical
Fic. 4.18. Coordinate system defining the angle @ and ¢ and the length r of a line
connecting a point A to the center of a sphere. ‘The area of the incremental surface
aS =r? sin Ode ds.
erton view
()
i
2 KN 9
Ly .
View trom top of sphere Pian view
(a) (e)
io, 4.19, (a) Division of a spherical surface into 20 equal areas of identical shape,
(@) and (0) Division of hemisphere into 8 parte of equal aree but unequal shape.
surface divided into 20 equal parts of the same shape. The measured
intensities on each of these paris may be called I,, Is, I, ete. ‘The total
power radiated W is
w
+ TSu0 (4.21)
Part XI] DIRECTIVITY INDEX m
where Si, Sx. . . Sie are the areas of the 20 parts of the epherical sur
face. If, as in Fig. 4.19a, the surface is divided into 20 equal parts, thea
81 = & = S Ste. For less critical cases, it is possible to
divide the spherical surface into 16 parts of equal area but of different
shape, as we can see from Fig. 4.19 and c.
By definition, the directivity factor Q(f) is
Part a —
oe WY Front sin oaoas
where [pad is the magnitude of the meansquare sound pressure on the
designated axis of the sound source at a certain distance r (see Fig, 4.23,
0° axis, as an example)
For the special case where, for any particular value of @, the sound
pressure produced by the sound source is independent of the value of 4,
that is to say, there is an axis of symmetry, Eq. (4.22) simplifies to
on (4.22)
spat
2 [96 sina
‘The magnitude signs are left. of for convenience.
Many sources, such as loudspeakers, are fairly symmetrical about the
principal axes so that Eq. (4.23) is valid. In this case, data are generally.
taken at a number of points with the angles @ in a horizontal plane
around the source so that.
a = (4.23)
AN = 
ae SNe sin 28
(424)
where Ad = separation in degroes of the successive points around the
sound source at which measurement of p(@,) was made (see
4.23 as an example).
180°/A9 = number of measurements that were made in passing from a
point directly in front of the souree to one directly behind the
source (0 to 180"). The sound source is assumed to be
symmetrical so that the variation between 360 and 180° is
the same as that between 0 and 180°,
If the source is mounted in an infinite baffle, measurement. is possible
only in a hemisphere. Hence, the value of n in Eq. (4.23) varies from
1 to 90°/a8. If the source in an infinite baffle is nondirectional in the
hemisphere, which is usually the case for ka < 0.5, then the directivity
factor Q = Q = 2, that is, DI = 3 db
If the directivity pattern is not quite symmetrical, then the factor of 4ne RADIATION OF SOUND (Chop. 4
in the numerator of Eq. (4.23) becomes 8 and the value of m varies from
1 to 360°/A8. This, in effect, averages the two sides of the directivity
pattern.
‘The directivity index at @ = 0° for each directivity pattern shown in
Part X and in this part is written alongside each directivity pattern.
» 00
0
i e
as 8
i S
gu 20
bu :
$19) 10S
ee re
z ints tae ea
ES
BiE 5
5
7
a f
gh pete
tan(uofe
so 4.20, Diecttey indexes or the elt sun (2) one ae ty of pt in an
‘Minite place bales (2) pst inthe end of long tube, and @) a piston i ee
Space without any bale
‘The reference axis is the principal axis at @ = Oin every case. An angle
@ at which the directivity index equals 0 db is also marked on these
graphs, Hence, the directivity index at any other angle @ can be found
by subtracting the decibel value for that axis from the decibel value at the
axis where DI = 0.
Fra, 4.21. Radiation into a slid cone of space defined by the angles a and 0.
For easy reference, the directivity indexes for a piston in (1) an infinite
plane baffle, (2) along tube, and (3) free space are plotted as a function of
ka in Fig. 4.20.
‘Many horn loudspeakers at high frequencies (above 1500 ops) radiate
sound uniformly into a solid rectangular cone of space as shown in Fig.
4.21. These horns are of the type discussed in Par. 4.4 (page 106). The
directivity indexes in the frequency range above 1500 cps as a function of
Part XI} DIRECTIVITY INDEX us
a and 8 may be estimated with the aid of the nomogram given in Fig.
4.22.01
Detailed calculations are shown in Table 4.1 for a boxenclosed loud
speaker having the directivity pattern at a frequency of 1500 eps shown
Bieta
index DB
30:
oe
50g
60.
70°.
105
ds
135
wot £
‘0° 0
Fra. 4.22. Nomogram for determining the directivity indexes of a source of eound
fadiating uniformly into a solid cone of apace of the type shown by Fig. 4.21. [After
‘Molloy, Calculation of the Divectivity Index Jor Various Types of Radiators, J. Acoust
Soe. Amer, 20: 387408 (1948).]
in Fig. 428, The left (L) and right (2) sides of the directivity character
isties are not alike, so that the averaging process of the previous para
graph is used.
= C.T. Motlay, Caleulation of the Directvity Index for Various Types of Radiators,
J. Acoust, Soe, Amer, 20: 987405 (1048).us RADIATION OF SOUND (Chap. 4
Fic. 4.23, Measured directivity patterns for a typical 12in. directradiator loud
speaker in a 27 by 20 by I2in, rectangular box. "The squares give the directivity
index at # = 0", One angle of zero directivity index is nls indicated.
sL 
iE 1
or  7
_ 
Rae ]
:
olen 1 oobi. 1
Legal att the ah
Freqeney in ys per scond
Fic. 424, Dinetivity indexes for O° axe ofthe directivity pattern of Fig. 4.23 com
puted an though the souree were symmmeriiealsboot the O" ans. 'Phe dats ppl to &
{apical 12m, dietaintor loudspeaker mounted in» 27by20by 12, rectangular
Part XI DIRECTIVITY INDEX 15
After a directivity factor has been calculated at each frequency, a plot
of directivity index DI({) in decibels is made with the aid of Eq. (4.19)
For the loudspeaker with the directivity patterns of Fig. 4.23, the direc
tivity index as a function of frequency is shown in Fig. 4.24,
TABLE 4.1, Calculation of Directivity Index DIG)
Directivity
355  5  0.087 1.00  0.95  0.09
343  15  0.250 1.00  0.79  0.96
a5  25  0.423 om  050  030
325  35  0.574 0.45  0.25  0.26
ais  45  0.707 oz  o10  0.19
aos  55  0.819 14  oot  on
as  65  0.900, o.o7  0.01  0.06
285  75  0.968, 0.03  o.o1  0.03
275  5  0.996 0.01  oor  0.01
265  95  0.990 o.01  oor  oor
255  105  0.960 oor  oor  o.08
245  15  0.906 oor  0.01  0.01
23s  sas  oso o.o1  oor  0.01
0.01
0.01
0.00
0.00
0.00
Be X 573"
a)
DICH = 10 log 104 = 10.2 db
{ALY = 1500 eps for a commercially available loudspeaker having the directivity,
patterns shown in Fig. 4.23. ‘The quantity a@ = 10° = »/18 radians,CHAPTER
ACOUSTIC COMPONENTS
pant XIT Radiation Impedances
rece eee
The fields of radio and television have advanced rapidly beeause of the
bility of electrical components with wellknown physical properties
that are simple to assemble into a completed mechanism. With such
components (resistors, capacitors, and inductors) the advanced research
engineer and the highschool student alike are able to experiment with
new circuits. Such complicated devices us electriewave filters often can
he designed hy selecting from among readily available parts until a
desired performance characteristic is achieved: a feat that otherwise
might require a lengthy mathematical aualysis,
No such satisfactory situation exists in the field of acoustics. Acous
tieal elements have not been available commercially. Advanced text~
books have often sidestepped the theoretical treatment. of their per.
formance, Even those texts which deal primarily with acoustic devices
sive limited information on how to predict the performance of eavities,
holes, tubes, screens, slots, and diaphragms—the elements of acoustical
circuits. ‘This text does not. pretend to advance the science of aroustieal=
circuit theory to anything approaching a state of completion. Much
basic research remains to be done. It does attempt to interpret the
{atest theories in such a way that the reader ean construct and understand
the performance of the usual types of acoustie devices,
Loudspeakers, microphones, and acoustic filters are the most common
devices composed of mechanical and acoustical elements. One obvious
acoustical element.is the air into which the sound is radiated, Others are
ir cavities, tubes, slots, and porous sereens hoth behind and in front of
actively vibrating diaphragms. ‘These various elements have acoustic
impedances associated with them, which can, in some frequeney ranges,
he represented ax simple lumped elements. ‘In other frequeney ranges,
‘listributed elements, analogous to electric lines, must be used in explain.
ing the performance wf the devices.
U6
Part XIN) RADIATION IMPEDANCES wt
The frst acoustical element that we shall deal with is the radiation
impedance of the air itself. Radiation impedance is a quantitative state
ment of the manner in which the medium reacts against the motion of a
vibrating surface.
Sound is produced by vibrating surfaces such as the diaphragm of a
loudspeaker. In addition to the energy required to move the vibrating
surface itself, energy is radiated into the air by the diaphragm, Part of
this radiated energy is useful and represents the power output of the loud
speaker. The remainder is stored (reactive) energy that is retumed to
the generator. Consequently, the acoustic impedance has a real part,
accounting for the radiated power, and an imaginary part, accounting for
the reactive power.
a ao #410
© @
Fic, 5.1. Exact radiation impedances and mobilities for all values of ka for a sphere
with “a surface that vibrates radially, (a) Mechanicalimpedence analogy (0)
acoustieimpedance anvlogy; (¢) mechaniealmobility analogy? (@) scoustie mobility
aanslogy. The quantity aia the radins of the sphe
‘The four simplest types of vibrating surface treated here are (1) a
pulsating sphere, (2) a plane circular piston mounted in an infinite surface
(baffle), (3) a plane circular piston in the end of a long tube, and (4) a
plane circular piston without baffle. We have already derived the
radiation impedance for a pulsating sphere. ‘The mathematical solution
of the radiation from a circular piston mounted in an infinite baffle appears
in many advanced texts so that only the results will be presented here.
‘More complicated problems are to solve analytically for the radiation
impedances and directivity patterns of a long tube and a vibrating piston
without baffle. Those solutions are now available, and the results are
given in this part. Most other types of vibrating surfaces are exceed
ingly difficult in mathematical treatment, and the results will not be
resented here,
1PM. Morse, “Vibration aod Sound,” 2d ed., pp. 326 :46, MeGrawIfill Book
Company, Ine,, New York, 1948,ir ACOUSTIC COMPONENTS. (Chop. 5
Pore rela yee tote Guat and ie 2
eee een eee
eee i eaaeN cine releieeee
eae eluate aces
siiiccomparel wits he once eon
son lernerlaeag ay thet
Aeneas ceer aeeurary
Sacco ucemsa aera ates
Ee oni cane eee
gaits ene nae aes
Seca eteeeiegts
igucrger —— uiigntieutval Fie bt The de
sion nina te mobi and impelavee are g
2 Plane Chesac Piston in Yntite Beas "tho mechan! impede
Zu
Wa + iN = rate
where Za =
= rallius of piston in
density of gus in kilograms per eubie meter,
speed of sound in meters p
mechanical resistance in newtonseeonds per meter. ‘The
naan 9 indieates that the resistive component is a fune
tion of frequency.
Xue = mechanical reactance in newtonseconds per meter,
k= w/e = 2s/X = wave number.
Ju, Ky = bwo types of Bessel fanetion
be
second,
Tw
ven by the series.
Kinder ar ALR. Frey,
Wiley & Sons, Hue, New York, 1050,
"Mors, op rity pp. 392, 383. Morse gives in
(ka) that eqaly K ken) 2a
#G,N, Watson, "Theory of Bowl
Lomlon, 1922
lamentabs of Acoustien,” pp. 187195, Jobe
ible VIIL on page 447 a function
Puetions," Cambridge University Pres
Pert XU) ne
UW) (62)
KW) (63)
where W = 2ka,
0.0005 =
0.0003
0.0002
coo! stu shu shu
aor “oo” 1 03," 19 30" 10
alu
Fia. 53. Real and imaginary parts of the
(Zu /ratpr) of the air oad on one side of « plan
Infinite fat baile. Frequeney is plotted on a normalized seale, where kn = 2efa/e =
2ea/d. Nota also that the ordinate is equal to Zera%/oye, where Z is the beoustie
impedance.
Graphs of the real and imaginary parts of
Za _ Wut
waipe ~~ ray
64)
‘are shown in Fig. 5.3 as a function of ka. The German §t indicates that
the quantity varies with frequency.
Similar graphs of the real and imaginary parts of the mechanical
mobility,
aunty = vatpye(tu + jz) = ratpae (att a tea) 65)120 ACOUSTIC COMPONENTS [Chap. 5
are shown in Fig. 5.4. ‘The mechanical mobility is in meters per newton
second, ie., mks mechanical mohms.
‘The data of Fig, 5.3 are used in dealing with impedance analogies and
the data of Fig, 5.4 in dealing with mobility analogies
We see from Fig. 5.3 that, for ka < 0.5, the reactance varies as the first
power of frequency while the resistance varies as the second power of fre
quency. At high frequencies, for ka > 5, the reactance becomes small
200)
100}
1g]
50
330}
os
og
02
oy
05 =
09
x2}
Webi analogy
00 ty ot
Oo "093° 01 03 19 30° 10
‘ha
Fic, $4. Real end imaginary parts of the normalised mechanical mobility (rato)
Of the air load upon one side of a plane piston of endius @ mounted in an infinite flat
balfle. ‘Frequency ig plotted on & normalized stale, where ba ~ 2sfa/e = 2ra/>.
[Note also that the ordinate is equal (o z4py/ra, where 24 is the acoustic. mobility.
compared with the resistance, and the resistance approaches a constant
value,
‘The mobility, on the other hand, is better behaved. ‘The responsive
ness is constant for ka < 0.5, and it is also constant for ka > 5 although
its value is larger:
Approximate Analogous Circuits. ‘The behavior just noted suggests
that, except for the wiggles in the curves for ka between 1 and 5, the
impedance and the mobility for a piston in an infinite baffle can be
approximated over the whole frequency range by the analogous circuits
Part X11] RADIATION IMPEDANCES 12a
of Fig. 5.5. ‘hose circuits give the mechanical and acousticimpedances
‘and mobilities, where
Rus = x0%p9 mks mechanical ohms (newtonsee/) 68)
Ru = Rus + Rw 128a%pec/9e
= 4.53a%pee mks mechanical ohms (5.7)
Reus = 1.386a"p¢ mks mechanical ohms 68)
Cus = 0.6/apee* m/newton 69)
Man = Sa'pa/3 = 2.67a%p9 ke (6.10)
ras = I/na'pe ~ 0:318/a%pee mks mechanical mohms 6.1)
rus = 0.7210"p9 mks mechanical mohms (12)
Rea = pec/ra® = 0318 pu¢/a? mks acoustic ohms (613)
Ra = Raat Ras = 128pe/
0.158.e/a? mks acoustic ohms (5.14)
Rey = 0.1404p.e/0% mks acoustic ohms (6.15)
Cay = 5.940" / pyc? m*/newton (5.16)
May = 8po/30'a = 0.27 p0/a kg/m* (5.17)
Tax = 7a*/pec mks acoustic mohms (6.18)
Pas = 7.128"/pye mks acoustie mohms 6.19)
ot y
ao Im ten
am Mh Mn Za Ie iM,
Rue Raa
() @)
te) @
Fic, 5.5. Approximate radiation impedances nnd mobilities for a piston inn jofi
ic
halfe of for a piston in the rad of along tube for all values of fa. (e) Mechanieal
Impndanee analogy, (0) aeoustiiinpedance analogy ; (2) mechanienlsmobility analogy
{Ud} acousticemobility analogy.
All constants are dimensionless and were chosen to give the best
average fit to the functions of Figs. 5.3 and 5.4
Low and Highfrequency Approximations. At low and at high fre
quencies these circuits may be approximated by the simpler cireuits given
in the last column of Table 5.1
It is apparent that when ka < 0.5, that is, when the circumference of
the piston 2ra is less than onehalf wavelength 4/2, the impedance load122 Aco
STIG COMPONENTS [Chap. 5
presented by the air on the vibrating piston is that of a mass shunted by a
very large resistance. In other words J? = (Jt, + 2)* is large com
pared with w*M,*. In fact, this loading mass may be imagined to be a
layer of air equal in area to the area
of the piston and equal in thickness to
about 0.85 times the radius, because
Piston of adus @
(ra*)(0.85a)p9 = 2.67a"ps = Mus
tong wte At high frequencies, ka > 5, the air
Fic. 56. Piston vibrating in the load behaves exactly as though it were
‘end of long tube connected to one end of a tube of the
same diameter as the piston, with the
other end of the tube perfectly absorbing. As we saw in Eq, (2.60), the
input mechanical resistance for such a tube is xa"pee. Hence, intuitively
one might expect that at high frequencies the vibrating rigid piston beams
the sound outward in lines perpendicular to the face of the piston. This
207 + a
el rT
o2}+

003
fe
a
IE oor
0.005,
0003}
0.002
0001]
0.0005}
00003} :
00002} —
000: V1 s Iu rh
oo “00a” or 03,” 10 30
,, 0
Fic. $:7, Real and imaginary parts of the normalized mechanival impedance Za
roo) of the air lou spon one ide of « plane piston of rai # monsted sn thi end
Of a Tong tube, Frequrncy is plotted on ssl eal, where k= 2xfa/e =
Braid Note also thr anlinate is equal to Zena/p, where Zain the acoustic
‘impedance,
Part XI) RADIATION IMPEDANCES 123,
inactually the case for thenearfeld. At‘a distance, however, the farfield
radiation spreads, as we learned in the preceding chapter.
5.3. Plane Circular Piston in End of Long Tube.’ ‘The mechanical
impedance (newtonseconds per meter) of the air load on one side of a
plane piston mounted in the end of a long tube (see Fig. 5.6) and vibrating
sinusoidally is given by a complicated mathematical expression that we
shall not reproduce here.
209
100}
xo
EY
ES
1]
5
F330]
B29
Pe
os
a

oa
005
003
ce]
oo
oo “ors* o1 "03," 10 30° 10
Fio, 58. Real and imaginary parts of the normalized meshanieal mobility (wolycey)
of the aic load upon one side of a plane piston of radius a mounted in the end of
long tube. "Frequency i plotied on a normalized seal, where ka = 2efa/e ~ 2ya/h
Note leo that the ordinate is equal to zapu/xa?, where ca isthe acoustic mobility
Graphs of the real and imaginary parts of the normalized mechanical
impedance 2 /xa"pi as a function of ka for a piston so mounted are shown,
in Fig. 5.7. Similar graphs of the real and imaginary parts of the normal
ized mechanical mobility are shown in Fig, 58. The data of Fig. 5.7 are
used in dealing with impedance analogies and those of Fig. 5.8 in dealing
with mobility analogies.
To a fair approximation, the radiation impedance for a piston in the
end of a long tube may be represented over the entire frequency range by
1, Levine and J. Schwinger, On the Radiation of Sound from an Unflanged Cite
lar Pipe, Phy, Re, TB: 385 06 (148)
Documentos similares a Acoustics (Leo L. Beranek)
Más de Augusto Bonelli Toro
Mucho más que documentos.
Descubra todo lo que Scribd tiene para ofrecer, incluyendo libros y audiolibros de importantes editoriales.
Cancele en cualquier momento.