Está en la página 1de 255
ACOUSTICS ACOUSTICS 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 02138 Library of Congress Catalog Card Number: 86-70671 International Standard Book Number: 0-88318-494-X 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 MLL-T. 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 above-named 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 present-day 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 first-year 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 concepts PREFACE ‘The high-fidelity 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 thirty-two 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 Direot-radiator Units Lanulspeakers Part NXUL Horns vill 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 well-planned 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. Free-feld 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 free-field 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. ELECTRO-MECHANO-ACOUSTICAT. 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, DIRECT-RADIATOR LOUDSPEAKERS vant XVII Basic Theory of Divect-radtotor Loudspeakers Pant XVIII Design Factors Affecting Direc-radiator Lowdepeater Performance 6 16 a a 62 0 1 a o 109 16 16 128, 14 a 150 18, 183 183 2 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 high-fidelity 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. I-15, Macmillan & Co.,[ih, London, 1806; P.M. Morse, “Vibra tion and Sound," 2d ed, pp. 217-225, MeCiraw-Hill Book Company, Ine., New York, 1948; L. B. Kinsler and A. R. Frey, “Fundamentals of Acoustic” pp. 118-137, John Wiley & Sons, Ine, New York, 1950; R. W. B. Stephens and A. F, Bave, “Wave Motion and Sound,” pp. 32-13, 400-406, 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 f--2 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 Charles-Boyle 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 one-half 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 one-half 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 one-half 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 one-half 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 gram-molecular 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 watt-see per degree centigrade per gram-moleculae 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. 104-114, 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, Ra-F 29) The time derivative of this equation gives Ag-72 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 One-dimensionol Derivation Threedimensional Devnation "The one-dimensional wave equation ig | The three-dimensionsl 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 (2-40), 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 2208 22 THE WAVE EQUATION AND SOLUTIONS — [Chap.2 One-dimensional 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 three-dimensional 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 source-free, homogeneous, isotropie, frictionless gas at rest, The Wave Equation in Spherical Coordinates, ‘The one-dimensional ‘wave equations derived above are for plane-wave 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. (2-4a) 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 One-dimensional Wave Equation, ‘The ‘one-dimensional 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, p-t-an(+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 backward-traveling, wave, fall + 2/0) 2, Regardless of the shape of the outward-going wave (or of the back- ward-traveling 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 backward-traveling 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 Steady-state 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 steady-state wave can be represented by a Jinear summation of sine-wave 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 electsical-cieuit 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 backward-traveling 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 steady-state 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 p-or 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 right-hand 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 outward-raveing 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 the 28 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 constant 30 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 one-fourth 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 with 32 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 one-fourth 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 one-fourth 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 one-fourth 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, vari uM 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 one-twentieth 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 u-2-— 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 backward-traveling 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 pressure-responsive 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 velocity-responsive 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 right-hand 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 velocity-sensitive 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 steady-state relations derived in this chapter are sum. marized in Table 2.1. ‘TABLE 21. General and Steady-state Relations for Small-signal 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 temperature 40 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., watt-seconds 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 sound-level 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 Pree-progressive Wave. From Eqs. (2.57) and (2.59) we have seen that the pressure and particle velocity in a plane free-progressive (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 fre-progressive 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) aaa 42 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 free-progressive 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 10-9 wattseo 2.9. Energy Density in a Spherical Free-progressive Wave. ‘The energy density in a spherical free-progressive 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 free-traveling 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 in-phase 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, 167-168, John Wiley & Sons, Ine, 1980. 1 The average power supplied by an clrtrical generator to a circuit equals the voltage times the in-phase 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 Free-progressive Wave. For a plane free-progressive 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 free-progressive 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 Free-progressive 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 mine 46 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 10-4 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 10-4 nowton/m (re 2 X 10-4 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 ELECTRO-MECHANO-ACOUSTICAL CIRCUITS parr VI Mechanical Circuits SS 3.1, Introduction. ‘The subject of electro-mechano-acoustics (some. times called dynamical analogies) is the application of electrical-cireuit 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. a 48 ELECTRO-MECHANO-ACOUSTICAL 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 clectrical-circuit 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: 249-267 (1988); The Mobility Method of Computing the Vibration’ of Linear’ Mechanical and Acoustical Syatems: Mechanical-clectrical Analogies, J. Appl. Phyr., 8: 373-387 (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: 405-414 (1941). A. Bloch, Electro-mechanical Anslogies and Their Use for the Analysis of Mechani- cal and Bleetrosmechanieal Systems, J. Ina. Elec. Eng., 92: 157-169 (1945), {Among the four circuit elements, the first three are two-poles, This list is exhaus- tive, The transformation element isa four-pole, There are other loses fonr-poes 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 : | Constant-drop | The quantity a is independent of tor. ‘The arrow points to the positive terminal ofthe generator |Constant-ow |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 electrical-cireuit theory because there con- ductance and resistance are often indiscriminately written beside the symbol for a resistance-type 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 CHANO-ACOU 3.3, Mechanical Circuits. Mechanical-circuit 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 Mobility-type Analogy to Impedance-type 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 Prefer rRO-M CHANO-ACOUSTICAL 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 mobility-type and impedance-type 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 nenton-acconda 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 mobility-type 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) Moblity-ype 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 impedance-type 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. Cw 54. RO-MECHANO-ACOUSS 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 mobility-type 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 flus-tusns esaty-toe —Ingetne-tye ‘The analogous quantity here is the time in- () (0) ‘tegral of velocity, which is displacement, Fro. 34. (2) Mobility-type ‘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 impedance-type 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 mobility-type 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. (@) Mobility-typeand 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 impedance-type 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, eonstant-velocity or constant-force. 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 differ- 56 ELECTRO-MECHANO-ACOUS 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 constant-velocity 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 conslant-force generator is represented here by an electromagnetic transducer (e.g, a moving-eoil 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 constant-force 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) Mobility-typo and (®) imped- fance-type 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 zero-velocity (ground) point. This will be illustrated in Example 3.3, (uit) et 13. 3.14. Mobility-type 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 mobility-type analogous circuit for this mechanical deviee, itis frequently useful to raw & mechanical-circuit 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 mechanical-cireuit 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 « mobility-type analogous circuit, This i accomplished simply by replacing the mechanical elements with the analogous mobilty-type elements. The circuit becomes that shown in Fig. 3.17, Remember that, in the moblity-type 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. Mobility-type 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 BLEGTRO-MECHANO-ACOUSTICAL CIRCUITS (Chap-3 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. Three-dement 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 mohility-type 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 moblity-type 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, Mobility-ty = 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) Mobility-type 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 ELECTRO-MECHANO-Acoust AL CIRCUITS (Chap. 3 mobility Solve for the total mobility presentnd to the constant force generator fi Solution, By inspection, the mobility-type 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 probe-tube 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 in-phace 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 elements 64 ELECTRO-MECHANO-ACOUSTICAL 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 cross-sectional 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 cross-sectional plane in the tube. ‘The volume velocity Ud) is equal to the linear velocity u(®) multiplied by the cross-sectional 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) Impedance-type and (2) Pra. 8.25, Bnclosed volume of air V with mobility-type symbole for an acoustic opening for entrance of pressure varia the impedance-type analogous symbol for acoustic mass is shown in Fig, 3.24a, and the mobility-type 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 cross-sectional 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. Umpene-tpe wipe ‘The impedance-type analogous element for” (gy () P 66 ‘TRO-MECHANO-ACOUSTICAL CIRCUITS Chap. 3 ‘The acoustic clement used to represent an acoustic resistance is a fine-mesh 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 newton-seconds U@ = instantaneous volume velocity in eubic meters per second of the gas through the cross-sectional aren of resistance. In the steady state U is an rms quantity. ‘The impedance-type analogous symbol for acoustic resistance is shown in Fig. 3.280 and the mobility-type in Fig. 3.280, Acoustic Generators, Acoustic generators can be of either the constant- volume velocity or the constant-pressure 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 constant-volume 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, newton-m 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 constant-pressure generator. Because the air in the tubes is aot confined, it Pot of (@) (b) own : Fro, 3.80, (a) Impedance-type and (@) Fra. 831. Acoustic device consisting of imobility-type symbols for a constant-vol- four tubes, thre eavities, and two screens time velocity generator. driven by's coustant-pressure 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 impedance-type 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 sum 8 rho-Me HANO-ACOUSTICAL 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, Impeslance-type 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 ie-cineuit 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 10-7 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 constant-foree generator @ produces a serie of tones, among whieh is tance is 500 mks acoustic ohms. If the tuhe haa a cross-sectional 1, = ly = 5.0m, fy = Lem, V = 100 em?, and the cross-sectional 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 constant-force 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) linpedanee-type analogous circuit for the deviee of (a) Solution. By inspection we may draw the impedance-type 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 ep RO-MECHANO-ACOUSTICAL 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 mechano-acoustic 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. Electromagnetic-mechanical 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 variable-reluctance earphone or microphone. ‘The simplest type of moving-coil 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 moving-coil type of transducer are fa Bi (8.26a) e= Blu (8.260) where i = electrical current in amperes J = “open-circuit” force in newtons produced on the mechanical circuit by the current i B = magnetic-flux 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 = “open-circuit” electrical voltage in volts produced by a velocity u ‘The right-hand 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 moving-eil 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 open-circuit 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 Electrostatic-mechanical 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 ¢ = “open-circuit” 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= “open-cireuit" 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 sectrostatic-mochanical 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 electromagnetic-mechanical type. Two possible forms are shown in Fig. 3.37. ‘The mechanical sides are of the impedance-type 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 short-circuited (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 open-cir 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 open-circuit force ‘wanedicer (2) if oa @) Bit fe t Bie % Ba ‘a renin’ Ly & () Fie. 8.48. Combined electrostatie-electromagnetic 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.MECHANO-ACOUSTICAL CIRCUITS [Chap. 3 Fig. 3470, Solving, we get Aw fate =X 10-*p 8280027 X 10-9 = 3078 x 10" fo = jallumn = ~834 X10" f= fot fem 814 x 10% Sa Ly = 10.78 x 10-4 — 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 10-4 lial = 147 X 10° an The voltage drop across the enpacitor is Ied = ATX 10-4 108 = 59 volts The electric stored encegy on the capacitor Ce i B4Crlelt X 10-4 X 8.48 X 108 = 7 10-4 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 one-sixth 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. a 92 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 one-sixth wavelength (that. 2d ed,, pp. 311-826, MeGraw-Hill Book 2P. M. Morse, “Vibration and Sound, Company, Ine., New York, 1948. +1, B. Kinsler and A. Frey, “Pondamentals of Acoustics," pp. 168-173, John Wiley & Sons, Ine., New York, 1950, * Morse, op. rit, pp. 312-313, 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 nie 94 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: 01-09 08, 22: 01-3008 Tae ior dea2 —pIe3008 ba 01-2008 @ a Fic. 43. Directivity patterna for the two simple in-phase 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 the 96 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: 01-08 08 ana v2:01-1508 a Ste ERG KRY [iS $e) Si @-22 01-6008 4-2) 01-6608 @ a iretivity patterns for a linear array of fou simple i-phase 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 from 98 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 in-phase sources, In the present case, we have a maximum pressure at @ = 0° and @ = 180° for b = 2/2. yA: 01-0308 N21 08 Toe od @-a/2 Ol-8108 4-2 Di-6208 @ @ 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. 280-285, Part X} DIRECTIVITY PATTERNS 99 212 (44-2) 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. Near-field and Far-field. ‘The difference between nearsfield and far-field 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 far-field 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) Sound-pressure 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 low-frequency 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 half-space, 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 one-half wave- * Morse, op. ct, pp. 826-846. 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 left-hand 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 far-field. 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: 481-4106 (Feb, 1, 1948) "PUM. Wiene, On the Kelation between the Sound Fields Raviated and Diffracted hay Phan Obstacles, J. dewnat, Soe. Amer., 98: 697-700 (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. 39-40, 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 straight-line and plane- surface arrays. These types of arrays are closely resembled by open- ended organ pipes, direct-radiator 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. Curvod-line 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 more 108 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, Curved-surface Sources (Multicellular Horns). ‘The most common curved-surface sources found in acoustics at the present time are multi- cellular homs. Two typical examples of such curved-surface 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 mean-square 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 o 10 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 mean-square 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 4 ne 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 box-enclosed 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: 987-405 (1048). us RADIATION OF SOUND (Chap. 4 Fic. 4.23, Measured directivity patterns for a typical 12in. direct-radiator 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, diet-aintor loudspeaker mounted in» 27-by20-by 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 well-known physical properties that are simple to assemble into a completed mechanism. With such components (resistors, capacitors, and inductors) the advanced research engineer and the high-school student alike are able to experiment with new circuits. Such complicated devices us electrie-wave 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 side-stepped 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) acoustie-impedance anvlogy; (¢) mechanieal-mobility 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, MeGraw-Ifill 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 newton-seeonds per meter. ‘The naan 9 indieates that the resistive component is a fune- tion of frequency. Xue = mechanical reactance in newton-seconds 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. 187-195, 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 (newton-see/) 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) aeousti-iinpedance analogy ; (2) mechanienl-smobility 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 High-frequency 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 one-half wavelength 4/2, the impedance load 122 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 thenear-feld. At‘a distance, however, the far-field radiation spreads, as we learned in the preceding chapter. 5.3. Plane Circular Piston in End of Long Tube.’ ‘The mechanical impedance (newton-seconds 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)