13.1 INTRODUCTION Sound waves or acoustic pluses are vibrations that consist of a succession of rapid variations in air Pressure, usually of a small magnitude. These are transmitted through the fluid (liquid or gaseous medium) and produce the sensation of hearing in the hearing meachanism. Sound forms an impor tant part of man’s environment. In some ways, it is quite useful to us. It makes speech communica- tion possible, warns a person of danger, enables a person to enjoy music, etc. However, much of the Sound one encounters unfortunately is unwanted sound termed noise. Some of the detrimental effects of noise are that excessive noise causes irritation, distraction or annoyance i.e, it hurts people psychologically and physiologically and tends to reduce man’s efficiency. Further, it inter- feres with the communication between people and tends to mask audible warning signals thereby reducing the safety in workshops. In addition, prolonged exposure to noise can cause noise- induced hearing loss, Acoustical measurements find wide application in our day-to-day life. For example, in noise control studies, i.e. the development of less noisy machinery and equipment, these days form an important area of the environmental pollution or noise pollution control. In addition, the other applications include the design and testing of sound recording and reproducing equipment, study f water-borne sound (sonar) for underwater direction and range finding, study of aero-dynamic Aoise and its interaction with structures, etc 13.2 CHARACTERISTICS OF SOUND Sound is characterised by the following parameters 1. its intensity in W/m? or its pressure in bar, 2. its frequency in cycles per second or Hz,s and do not fon the subjective agen based on measured results, extends from 0.009) Tepresents an increase of ss the magnitude of soung defined in decibels as: Teference pressure (pj. =Acoustics Measurement 2x 10° 180- | I 2x 10 160-}- Rocket engine, supersonic boom (Immediate hearing damage resuits) 2x 108 +} = 140 Hydraulic press (1 m), (Threshold of pain) | 2 A 2x 10? 3 120 ++ Four engined jet aircraft overhead (150 m) 8 (very loud sound) | ie 7 20 > | 3 120-1 Airport noise, train whistle (100 m), noisy factory is fesird | t og | : 2 r SS 80 > Heavy duty automobiles (6 m, light automobiles at 6 m z gs | (oud sound) a 52 ] £ 2x10 + & 60+ Typewriter (1 m), normal conversational speech (1m) Z 5 (moderate sound) g € | ZB exw > 40 7 Rusting of leaves at 10 m (avorage threshold of hearing) 8 | } BS a3 20 | Faint sound of whisper 0 Acute threshold of hearing for youngmen (1000 to 4000 Hz) Typical values of sound pressure levels of common sounds with reference sound pressure of 0.002 yibar, the threshold of hearing at 1 kHz 13.3.2 Sound Power Level Sound power is the total sound energy radiated by a sound source per unit time. The power in the {ound emitted from a source can have an extremely large range. The power can be as low as 10° Wata faint whisper or 10°° W at normal conversational speech. Further, it could be as large as 10 for saturn rocket or 50 kW for a jet airliner. Because of this wide range of values of power in the various power sources, the sound power level is conveniently defined in the units of decibels, Power level is often abbreviated as PWL or L,, and is defined as: PWL = 10 logy = B (13.3) et here Wis acoustic power of the source and Wr, isthe reference acoustic power (Wry = 10°! W), the power ratio 10"! means -120 dB, the above equation may be written as: PWL = (10 logy) W + 120) dB (13.4) Where Wis in watts. : “igure 13.2 Rives the typical values of sound power levels of some common sounds.ysis wnt and Analy son, Measuremer Instrumentation, w) (100000000 5 Space rocket a { 100 Jot aircraft (50000 Ww) 160 +} | Turboprop (500 W) wo 4 { Light aircraft (5 W) 120 ++ Full symphony orchestra (1 W) Fan (10000 cfm) (0.05 W) ae 100 60 40 Faint sound whisper (10-° W) 20 ° Sound power levels of common sounds with reference po PL, 0B (Relative to 10-7? w) | Typical values of sour "Ww X's interesting to-note that small sound power levels can have a Significant effect on the ear. I is estimated that the acousti “coustic Power emitted by the roars of the football crowd during a match is Pie canes bo best 8 cap ot coflee! Burthe, the power of ¢ fat symphony orchestra is only one acoustic watt or so! Sound pressure level (SPL) of a machine, that is adiating noi e machine as e 18 Noise, depends on the machine a well as on the surroundings. The same machine hae chow different SPLs when placed in different Surroundings. It is for this reason that sound powes is ind power level (PWL) h, i BP rssnce be directly defined but is found by the measurement of sound Ee unten defined, PWL cann sources, acou: ‘els or sound intensity levels tic Power W of all sources can be sic Si aIt sound intensity level Z, is defined 1,= 10 logy) dp 13.5) he value Of ey is taken as 10°”? Wim? in Eq. (13.5) “fis the time averaged product of sound pressure ’p* and particle velocity ‘u’ or Tepeu 13.6) Pi tna froe field. I aan pc is the root mean square sound pre ‘whet Pron’ essure, p the density of air, c the speed of sound in the medium and Pc is called characteristics acoustic impedance of the medium, 13.3.4 Addition of Sound Pressure Levels If there are two or more sources radiating noise, the total sound pressure level may be determined from the individual sound pressure levels of the sources. It may be done by addition based on energy or power levels which are dependent on mean square pressures. Thus, the SPL of each Z| values, by dividing the decibels by 10 (Pe J bused may be converted from number of decibels to and calculating antilogs. After adding the relative (p’/p2,) values of various sources, the conver- sion to dB levels may be made for the combined sound pressure levels, using Eq. (13.1). Let us consider two such sound pressure levels of X dB and ¥ dB x) Y ) = 10 log | alog(*)+ atog (2. Resultant SPL = 10 log [a (Xx) be (3) Similarly, if there are several sources, with SPL’s X, ¥, Z, .. . etc. then x y Ay = 10 log | alog{*.) + atog(¥) + alog (2 38 Tol PL. = 101g [stg (+ «(75) loe( % + | 3.8) Now if in Eq. (13.7) X = ¥, i.e. there are two sound sources of equal SPL, then, x si log {2alog (2 Total SPL = ove | al «(3h x a 301+ =10 {o 01 (x) =X+3dB (39) lend %€ have sounds of SPL of 100 dB each, then the resultant sound pressure level would. be 100+ 3) — 103 apThe background noice en hhaving individual soung ty y, determine the combiney soingst hyil | 13.3.5 Variation of Intensity of Sound with Distance 6 Ma. 'eVelg has been observed that as the distance of me asurement from the sound source inert Sure {intensity of sound decreases. The simplest source of sound is the point source, where the sound is radiated equally in all direction from the apparent centre. For a point source, the relationship feween the distance d and intensity 1 is given by Fiat (13.10) 4nd his applies to spherical wave fonts « Te Wd? (13.11) Equation (13.12) is commonly termed as inverse square In practice, the industrial sound/noise sources have finite dimensions, i.e. they may be rectangu- lar, cylindrical or other composite shapes. But they are rarely point sources. However, as a general rule of thumb, if the distance from the noise source is greater than five times the maximum dimension of the source, then the inverse square law is applicable Itcan also be seen that px Fi (13.12) d Problem 13.4 noisy ‘The sound pressure level measured at 10 m from an automobile horn is 110 dB. Determine the sound pressure level at distances of (a) 20 m and (b) 80 m. Assume that the inverse square law holds good between intensity and distance. Solution | Using Eq. (13.12) we get, | Ie 1/0? | Therefore, if dis doubled than / becomes one-fourth. 4 (a) At 10m, SPL = 10 log a = 110 dB lo 1/4 At20m, SPL = 10 log *) lo 4 1 = 10 log | 7-| + 10109 | lo =110-6 = 104 dB This shows that doubling the distance reduces the sound pressure level by 6 dB. 4 1 | 10 log {+ 10log Gy | =110- 18.1 i 91.9 dBmeasurements. They ae bry a be near-prefect sound absorbers Re © law, i.e. there will be a decrs, source. Usually, the obstruction st are also termed free-field cong, sure level and sound power al we a in metres) from the source, 89 that are near-perfect reflec . OTS of » ths the SPL is everywhere ney yt gic conditions, i.e. conditions so as distance is increased but the dectag, se testing of machinery and equipmea, ge room (minimum volume 300 mye external noises approximates io ig re level is different for different freer 500 and 5000 Hz. It is less sensiive ncies. Therefore the SPL. necessy @ listener is different. A higher SPL to produce the same loudness sensi the ear has been evaluated and eye! The unit of loudness is phon. I's at the point where the curve ox® hhave the same loudness expres Phons when heard at 1000 Hz. Fis®* indicating sound pressure d pressure level at 1000 Hzis Tevel at 100 Hz, the average m™dB door Nder ind me tis but the 120 Phons 100 ae e : 3 10 A 100 re 1000 10000 mas A typical representation of equal loudness contours indicating SPL at different frequencies with no obstruction from the sound source and the reading is taken. The block diagram of the imsmument is shown in Fig. 13.4. This includes a microphone, an electronic amplifier with fre- quency weighting network and a meter or recorder calibrated in decibels A rectifier circuit is incorporated to produce a signal Proportional to root mean square value. ‘The input signal generated by the sound pressure alternates from Positive to negative levels. This Signal is squared, which has the effect of producing all positive signals varying in amplitude. By taking the square root of this signal, it is possible to obtain the output of the meter in the form of averaged root mean square value of the signal. The frequency response of the unweighted system is generally flat or linear over the entire ‘ailable range from 20 Hz to 20 kHz. Further, some commercially available instruments have the Per ftequency measuring range of approximately 30 kHz. The results of the measurements using the linear response are quoted in dB or dB linear. A frequency weighting network is used in the circuit to Provide a response similar to that of the far: With the network, there is a different amount of amplification for each frequency so fiat the overall measurement made gives greater emphasis to some frequencies than to others, It is ‘en from the equal loudness contours that the ear is more responsive to frequencies between $00 £4 5000 Hz than to those above or below these frequencies. The weighting networks, therefore Breater amplification in this range, than at higher or lower frequencies. ‘Three. standaed ighting networks A, B and C are used to approximate the equal loudness curves, The response hulbe A scale approximates to 40 phon equal loudness curve and is used for SPL. below $0 ae Bs of measurements made with this scale are expressed as dB A or 4B ‘A'. The B aie Tein the range 55-95 dB and the response approximates to the 70 phon equal loudness cate Scale, intended for use above 85 dB has approximately a flat response, except at the two ' Figure 13.5 shows the linear and the various weighted responses of the sound level meters,RectiierAcoustics Measurement 13.5.2 Frequency Analysis of Noise Signal she frequency content or sound spectrum of noise signal is determined by frequency analsysis. It is Tefal to the noise control people/analysts for determining the probable sources of noise and it also felps them to suggest means to controV/confine them. Further, noise is a symptom of malfunction: ing in the machinery. Frequency analysis can quite often describe various frequency spectrum itsted malfunction. In addition, spectral characteristics are important in describing the transmission Srsound, say through a wall or its absorption by some materials. Frequency spectrum analysis of a sound signal consist of distribution of sound pressure level at sarious frequencies versus the frequencies ranging from low to high frequencies. This is achieved ith the help of filters, each filter allowing only that part of the spectrum which lies inside the fandwidth of the filter. Frequency analysis can be performed by using either 1/1 octave band or ¥ tetave band or 1/3 octave band. A band of 1/1 octave represents a range of frequencies in which the upper frequency and the lower frequency is in the ratio of 2: 1. In other words, if the upper frequency is f, and the lower frequency is f, then, fu=2f (3.14) The procedure of measuring pressure levels using the I/I octave band analysis consists of setting the meter at the geometric mean frequency and take the reading. The geometric mean is the average of the logarithm of the upper and lower frequencies and is termed as the centre frequency fe Now, log f, (geometric mean) = + (log fi + log f) 2, =log Ji ha o fe= Vf fe (3.15) The centre frequencies of 1/1 octave band have been standardised internationally ranging from 3155 0 16000 H, and are given in Table 13.1. Many portable sound level meters (SLM) have such builtin filters and it is possible to measure sound pressure level for each band by changing the centre frequency. ‘Tobe 13.1 Standard Octave Band Frequencies in the Audio Frequency Range Low band edge Centre frequency High band edge 2 315 a “4 6 88 88 12s 176 176 250 353 353 500 706 706 1,000 144 1414 2,000 2,828 2.928 4,000 5656 5,656 8,000 11,312 16,000 22,624 hu3i2pectrum, the octave band ny, nt Re itcins 1/2 octave p, ang) iy a band analysis. In these mu aang ‘of these smaller bands ig wedUbdi rents ding sform yrticle set noise Commonly, the two microphones used are of 12 mm indiameter with « spacer 12 mm in length between the im pmicrophones. Velocity wis estimated from the pree Iwi gradient between the two microphones and the bree sie pressure of the microphones ‘p’ is used to find the By Gables Wich is averaged over tine. aon thely, FET analysis is used. In the method, sound inten tity in the frequency domain is taken from ihe imaginary m of the cross-spectrum between two signals: fro Freely spaced microphones, using a dual channel RFT analysis. It is seen that intensity /, in ‘r* direction is t(G;) a (13.16) where ‘ais the angular frequenc; tary part of cross-spectrum betw ae signals, p the medium density and Ar in the spacing be- tween the microphones. A computer with software is nor mally used for calculations and plotting. The measurements are carried out with distance ‘r’ from the noise radiating surface being 40 ~ 80 mm. For measure- ments, the noise radiating surfaces are divided into suit- ably spaced grades and intensity measurements are done ateach point of the grid by direct use of Eq. (13.6), from Which intensity maps can be obtained indicating the sions of high sound radiation. In another method called sweeping or scanning method, the probe is slowly swept (wsually, 0.2 m/s) over the surface while the analyser is averaging the measurements. Problem 13.6 Noise radiating surface Fig. 13.6 Sound intensity probe ‘n @ certain sound level meter, the output voltage was directly proportional to the sound Pressure generated by a sound source. The instrument was calibrated with a sound pres- Sure excitation of 1 dyn/cm? and the meter scale was set at -60 dB referenced to 1 V. For a given noise source, the output voltage developed by the instrument is 6 mV. * Determine the SPL of the noise source. ff Solution The output voltage of the sound level meter (SLM) for an excitation of dyn/cm? is given by: -60 dB = 20 log E/E. 20 log E10 ( ®. E=10°V Now. 6 my output voltage of SLM = 6 x 10/10 = 6 dyn/cm? E.=1V)Prot = 2 10°* dynes ) was processed in an a, filters. The results cbtainen ™ 1000 2000 aap R co ean Is of each centre frequency inthe my + alog (8.4) + alog (8.1) + alog (72) 9 (5.2)) ‘variations into the analogous et ‘as sensitivity, frequency = ploy a thin diaphragm to convert p= electrical output by employing s*e octave 2) ‘lectrical esponse, pressure second- or grooves 10 Polarising voltage Provide damping (100 ~ 300 v de) Back plate y Ssretched dlaphragm of thickness = 0.002 om > To amphtior Protective cover 7) qe * Insulator AA gap = 0.002 em Microphone housing Air leak capillary for static pressure equalisation FRRBF 4 Schemetic diagram of a condenser type microphone where E= output voltage of the capacitor, Q = charge provided by the polarising voltage (nearly constant), d= separation of the plates. The microphone is provided with a capillary air-leak for equalisation of pressure on both sides of the diaphragm to prevent the diaphragm from bursting. Further, the back plate of the condenser microphone is provided with damping holes. The motion of the diaphragm causes air flow through the holes which results in energy dissipation because of fluid friction. This damping effect is uiilised in the control of resonant peaks of diaphragm response. These microphones have good frequency response even at high frequencies and also have good acoustic sensitivity. However, they are susceptible to errors caused by high humidity, which pro- duces excessive background noise due to the leakage of the charge. 13.6.2 Electret Microphone Electret condenser microphone is an improved version of ordinary condenser microphone. In the condenser microphone we need to supply a de polarising voltage, whereas in electret microphone, the polarized element, namely the thin coating of the electret material supplies the bonded charge to the back plate of the electret microphone. In other words, it is a self-polarising type of con- denser micorphone. The principle of operation of the electret microphone, shown in Fig. 13.8 is precisely the same 4 that of the conventional condenser microphone. They have excellent frequency response and are ‘elatively free from noise in humid environments. However, the aging of the electret material can ‘ause errors of the order of 0.2 dB per year. 13.6.3 Piezo-electric Crystal Type Microphone Certain materials possess the ability to generate an electrical potential when subjected to mechani- cal strain. Such materials are generally crystalline in nature and are known as piezo electric i2stals. The commonly employed materials are quartz, Rochelle’s salt (potassium-sodium tartarate), ad zinconate titanate, barium titanate and ammonium dihydrogen phosphate,and is widely used as microphone the air and cannot be used above 59 “Chg. In practice, lead zirconate titanae gp force produced by the acoustic pres» which in turn produces voltge wm, element is mechanically coupled nts IS employ direct contact bevex: (ie. element placed in bending mr nent placed in bending mode) or & Piezo-electric material ines contact ie. element is subjected to direct compression). Piezo-electric microphones are very corse and can Measure accurately sound pressure levels below 24 JB. Further, their response at tater frequencies is also very good. However, their response at higher frequencies isnot as good tev eondenser microphones. In addition, they require impedance matching and are also sensitive to vibrations. 13.6.4 Electrodynamic Type of Microphone ‘This type of microphone is based on the principle of generation of emf when a moving conductor jt placed in a magnetic field. Figure 13.10 skows the typical construction of this type of micro: phones. The sensing diaphragm is attached to a coil or ribbon placed within the poles of a Fermanent magnet. The movement of the diaphragm due to acoustic pressure generates the analo- gous induced voltage in the coil Microphone housing Permanent magnet transducer ~ C. Further, Thin iapragm © (PZT) is e Sure on the le To electronic processing ‘output in Sof iron core | circuitry -d_ with the i tween the x ode) or by \ r by direct Arleak capillary for static pressure equalisation The output voltage V of the electrodynamic microphone is given by V= Blu x 10° volts Where B= density of magnetic flux (gauss) length of the conductor and velocity (m/s) Such microohones have the advantage of being self-generating but their frequency response is Poor due to the high inertia of the moving coil. Review Questions _—_————— 13.1 Indicate which of the following is correct: (@ Decibel is a unit of (@) sound pressure level only