";{;I';; ," ',',' t"" ~:.a car · 'e .:
"':measureinents'~
BrUel& Kjrerl+1 BrOel& KjaerAustralia Ply. Ltd.
HEIl D OFFlCE: 33 MaprsBoy R<>ad, Cooco
Advertising/Administration: Sandy Eastman REPORT Tel.: (02)5273173 Fax:(02)5232163 • An Anechoic Chamber at the Australian Defence Force Academy. SubscriplionRates J. C. S. Lai . Surface Mail: 1 yearA$33; 2 years A$59.50; 3 years A$86.50. Airmail: 1 year A$45; 2 years A$71.50; 3 years A$98.50.
Address all correspondence to: Information for Contributors . The Chief Editor C/-School of Physics The University of New SoulhWales P.O. Box 1. Kensington, N.S.w. 2033
Acoustics Australia is published by the Australian Acoustical Society (IncorporatedinN.S.w.) 35-43 Clarence Street, Sydney, N.S.W.2000,Australia. Responsibility for the contents of articles rests upon the authors not the Australian Acoustical Society. Articles may be reproduced in full provided adequate reference is quoted. Printed by Advertiser Index CronuliaPrinting Co. Ply. Ltd., 16 Cronulla Street, Cronulla 2230. (02)5235954. .. insidebackcover ISSN 0814-6039 The cartoon on p.19 was drawn by Doug Cato. (Cf}{]~[Q)~WfPJ
1P@U!J~lElDl0001Pl1l.~(6lE
~WIPl$l!IJffilij
=- ~@@IF1ID1E For more infor mation phone THECHADWICK GROUP Sydney (02) 428 1388 Melbo urne (03) 560 2422 Canberra (062)806333 R. Acoassfics AUDIOIViIETRIC rESY /BOOTHS Noise lsol. t;,," Class 39 ACOUSTIC DOORS STC36,38,40 ,43 ,dB SOUND FllE4.DROOIViS • Inslal/alion Service available o Certified by Commonwealth Experimental Building Station , Ryde. N.sW • Basic Con structio n - Composile Panel Tim be r Hnisb o 100% Australian Content Manufactured by Kell & Rigby "",,,,..,,,,.COm'''' 8 DunlopS!reel, Enlield, N.S.W. Telephone: (02) 642 5999 ·Vol. 15 No. 1 Acoustics Aualraha NEWS------A.C.T. The final meeting for 1986 was held Mini-Editorial at the Playground Theatre Restaurant First Meeting and Cabaret. Approximately 14 mem In this issue there is evidenceofa On Tuesday, 24th February, 1987 bers and friends partook of dinner and little editorial tidying-up. The previous sections Australian News, International Louis Challis spoke to almost 20 ~~t:ro~~."Rise and Fall of the Roman people who attended the first meeting News and People have been amalga of the Society to be held in the A.C.T. mated into one section simply called under the auspices of the N.S.W. News. In these days of satellites, desk Division. His talk was on the Acoustic Notes from the Gen. Sec. ~~c~~~t~~~s~~d;~eS,;,:OA~at~~c~i~:rS~ Considerations of a Major Architec 1988 Annual Conference ... the South tural Project with particular reference Australian Division Committee has ac :~;te~et~;nO~iJ~~.ates- it gets around to the New Parliament House Project. ceptedresponsibilityfororganisingand He outlined the generai responsi running the 1988 Annual Conference. Speaking of FAX machines, authors bilities of the acoustic consultant and No dates or themes have been finalised and suppliers of any written material, the need for interaction wifh the others at this stage but it is expected that especially late material (late in the sense involved with the project. The new it will probably be organised for the that it is long past the editor's deadline parliament house is a particularly com latter half of the year. but the printer is still poised with his plex project because of its largescale finger on the GO button) may now send and the innovative design. Louisex National Committee on Physics ... in their contributions directly using FAX plained that fhe time restrictions led to recent years the National Committee has No. (02)5232163. This will deposit your his development of new computer aided taken a special interest in acoustics exemplary prose in the office of Cronulla techniques for rapid assessment of the and in so doing has initiated and Secretarial Services to be processed .acoustic environment within the large sponsored a scientific meeting on immediately, if not sooner, by Sandy spaces for the House of Represen Underwater Sound and supported our Eastman who is now in charge of all tatives and the Senate. The use of successful nomination for the Commis our administrative and advertising busi speech, which had been recorded in sion on Acoustics. There have been ness. For all business matters please the models of the spaces, was of great initiatives in other areas including optics call Sandy on (02)S273173. benefit in demonstrating to the clients The initial AAS representative on the the need for more sound absorption and National Committee on Physics was The editorial committee is grateful to Jane Raines for her work as Advertising ~~rf;~~;~UIplacement ot the reflecting Paul Dubout and subsequently Ray Piesse. The representation has con Manager prior to her recent retirement from that position. The room acoustic considerations tinued for about 10 years. The Council are not the only ones which are of the AAS has now nominated Dr. Howard Poffard important in such a large project. Louis Neville Fletcher to serve on this com Chie/Editor explained some of the other acoustic mittee until May 1989. Neville Fletcher problems which had to be solved. has a long-time interest in acoustics His talk was followed by a lively ques and is at present Director of CSIRO tion time before a dinner which was ~:sr~~uteof Physical Sciences in Can- Australian Acoustical attended by half the audience. Society Records Neville Fletcher was the Chairman for Engineering Award ... noted in Con the meeting. sulting Engineer for September 1986 Council at its 37th meeting decided While there are only a few members among the annual ACEA Engineering to implement a centralised computer of the Society currently resident in Awards, an award of High Commen ~~::~erfhc~r~ff~~~iVeevne;;sn;,~~~ta~~i~= the A.C.T. there does seem to be a dation for Louis Challis & Associates substantial interest in acoustics. Ilis for their project 'Observation Booth for stration and reduce the work load of some of its voluntary office bearers. therefore hoped that, if more activities i~~:~gFA18A Hornet Aircraft - Vic- are held in Canberra, more can be en The records which will include details couraged to join the Society and maybe of membership, subscription payments, tor~::'rJ.Groupof the Society can be FASTS Request members interests and addresses will be maintained by the Science Centre (FASTS is the Federation of Austra Foundation in Sydney. They will be lianScientificandTechnologicalSoci suitable for keeping the Society's N.S.W. eties of which the Australian Acoustical Registers of members up to date, Society is a member.) February Technical Meeting compiling the Society's Directory and "FASTS is entering into an arrange for mailing purposes. The first meeting for 1987 was a joint ment with the airlines to promote meeting with the Institution of Engi FASTS Member Society 1988 Confer neers, on 18 February. ences overseas. Where people can link The topic for the meeting was the a few Conferences together they are "Reduction of Noise from fans at more likely to travel overseas. Eraring Power Station". The two speak If you would like your 1988 Confer ers were A. Hanna, a senior engineer enceto be oromoted in this way please from the Electricity Commission and send in whatever details you can. B. A. Russel, the technical director We will be looking at some initial from James Howden, Aust. The meas promotion in April 1987. Dates and urement procedures and reduction venues would be appreciated, draft techniques for dealing with the noise programs if you have them." WANTED TO BUY ~i~~u~~~gdelow frequency sources were G.P.D. Box 2181, Canberra, A.C.T. B & K 2209 Sound Level Meter. 2601. Te/.:(062)473554. B&K4426 Statistical Analyser. Basic units or with Bruel & Kjaer Move accessories. From Queensland Details to: Both Warren Middlefon and Peter The telephone number given in our Koorockin have independently spent last issue was not correct. Bruel & Caleb Smith Kjaer's new address is 24 Tepko Road, P.O. BOl<306 time recently in U.K. Warren has now Toronto,N.S.W.2283 returned to resume teaching duties at Terrey Hills 2084; postal address is Phone: (0 O.I.T. while Peter is staying on to tour P.O. Box 177, Terrey Hills 2084; tele 49)S93370 Europe. phone (02)4502066; telex AA26246. NEWS. The prize mon ey for th. awa rd comes .nd thao ppOtlunltytoc h.t ...llh lrl.nd s Iro m the Depa rlme nt ollndus lry, Tec h and coil . agu al Irom the Icou atic. 'ie ld. no lol1l .nd Resource. The new ' ddr• ., lor B & K Is 24 Letter to the Editor hpko Ro.d , Tartey HIIII , N.S,W. 2084. (02)450~ . C., rS i•• DIDEIrIIS (The Clgital Inte •• ctlwe Follo...in; a workshOll H ul on , t tM Cie$81 Engine Monitor ing Sysloml is 1986 A.A,S. Con le"nc . In Toowoo mb . the firsl comp rehen sive diese l engin e di.gno stic s)'St. m In lhe world. The Standards N, il e L..... l. lo r = ~ .\'S le m ~ ol -..l'ldanclllsw (ll c hoiee . It nOl co....1dar troelag a l .. lab.jl)' a 11'... . 1 arnounto llirne lor an II'''''' don (21 f' equene y , p.e tr.l eo nl.m lb :=~~:snc:...~=, 121 ';:~~:~ onc:a annoyanw Is r_ r· "'1981~ (31 Tonea. hluea,acr aac ..... and '-· T .. raionla ... titled IrMl.. ne yru mI>IeL "RaCOl'lmanded Oat i; n Sou nd l _ l. (. 1 lda mlea /ionwt lhlt-.. nollasourea a nd R_berat ionTlrnaalor8uoilding [oc c upa tion.I • • 1C.). pr.... nt .. w. r. the fade ral Me mbe r, Inlarlo ..., Is In.. nded 10 uslat (5) Sl'na Irom O1her c.. s n linsom Jim Cal1ton and lire Sta le Member Jim desig loptowlde.c.rtai naeouatic ni.. gl...... dornestleu p.. lSl. Longley . ... "lfon I.,..;lhinoccupiad .p.eaa in lls also ...... • ..., • • isUnogbuild'~ l In his introd ucto ry ta lk. C~1tWinl... In..... dedlor the . pp1Ie, t1on In t 1rIeo ~ ~==:::=I.':.,~:;g"::.,~~ia. • • pla ined ho ... lhoe building was d. (81 AcINltyd i.tu 'b ance l'pae<:h, le la. lion . nd ...... ntolbuildin g ter . slgnedtofi lin wilh the ... .,our>ding ia ls In d ae rvieaa used in thoese spaces. ~::::nlr~on.""'w~~~~~~,h~~ ~:'da~::C1·~m:,,~~.:r.,.~~;t~,:~ .I. epl. (9) Combination 01 wl...,.1 and aural ci. tion. TIIa IWlWpremises in<:orpo r. t. Recent Activities 01 djl turtlanee ('haedl lghta / traltic. I. rger Slo rege lac imi&S. e.palld ed sa , laclO1ynigh tshi lt• • IC.I. wice ca pa bilities . a lec ture room lor Acoustics/Vibratio n (101 SOCiologic al sla tu, Committ ees (I l l Dasirab l. ao und 10 one (music l, , n'lhe ma 10 , nothar [no,"), • On. o! In. mllor ev... " In tha (121 In,bl llly lOal fec l .ourc e o' noisa u. a 01 Acou .tl c. a nd Ylbralion lor Ina (13) Hea ring c hafac ler ist ics 01 Indi y... r jU$t pas Md ..... lhe .m . lgam .lion widua l. o!tha two .... s 01 work under . com· (U) A; e a nd ; endar0 1 lislene rs mon bo ard (AV/ -l, SAA Council e n A IUlihar l uggu lion w, s lh, t , nnoy· dotle d lhecon stllutlono ll he Aco u. Uc. 1 anee mey be ,lf ect ed by whelh ar the VlbraUon$ fand. roh80l rd lale In 1985 lll l.ner wlS orona Or up rlgh l bOt lhls Th. Council l uppo rtod the propo s. ' IS . ee med 10 be . co nj.c lu r. only sl; nifiCllnt a nd Inle ".la led gro wth be lwee n 11M two . r. .. WII 10f. .. e n, J .A . RO' e (P rlnc lp a 'J e spoc l"lIy In th e e nvironmenla l "fee 01 J. A. ROSE ACOUS TICS ~~~I~I)Iand "lb .lI lon meas Ufing and • One or lhO re cen t ml jOf public .· non s wn the rev i. lon ot A520 21, Tha ne w ha. dquan ers building 01 ACOUl lics . Alrc, .ft Noi• • Inlru. lo n_ Brue/& Kja er Austra/ia . Bui!ding Slting . nd Conl fruction. Th. rawlI.d s tl nd ard Inco , po r. lea .c tu.l VIPAC Award noise me asu ramenl . 'or landing s a nd ~r,~~m~~c~~fe";.UpT~:~~in~." a l:,e ll t~:l. i A complete range of noise control materials: ABSORPTION- BARRIER- oeCIBAIl AF - plain ac oustic foa m DECIBAIl LV - ioo oeo vinyl DECIBAIl VAF - ce nc rcreo vinyl faced DECIBAIl LVF -cicc oe o vinyl/foam ac oustic foa m composite ba rrier & absorber DECIBAIl MAF - metallized po lyester rocec oc oosuc foa m DECIBAIl LVM -IOClded vinyl ma tting DeCIBAR FAF - toug h PUfilm faced ac oustic foa m DAMPiNG - DECIBAIl TAF - fedile faced acoustic foa m DECIBAIl DC - vibration damping compo und Suppl iers to Major OEM's & to industry DECIBAIl DS - vibration damp ing sheet INDUSlRIAL 7 Leva nswe ll Road. NOISE Moora bbin. Vic . CONlROL Australia 3189. pry.limired Ph: (03) 555 9455 I Desig ners, manu fac turers a nd suppl iefs of noise contlol prod uc ts a nd ma terials Machine Condition Monitoring Using Vibration Analyses J. Mathew Department of Mechanical Engineering Monash University Clayton Vic. 3168 ABSTRACT: Machine condition monitoring is concerned with obtaining data which would aid in evaluating machinery mechanical integrity. Various techniques have been developed in recent times and these can be broadly classified into six categories: aural visual operational variables, temperature, wear debris and vibration monitoring. This article is mainly concerned with the last of these, vibration monitoring. The various forms of vibration analysis techniques applicable to machine condition monitoring techniques are described. It is shown that relying entirely on anyone technique may be unwise. It would be preferable to process the basic time domain signal in several ways to highlight different aspects of the signal which would in turn relate to different failuremodes. The evidence obtained from vibration monitoring can finallybe integrated with data obtained using other techniques to confidentlY identify machinery failure. KEYWORDS: Condition monitoring, vibration monitoring, vibration analysis, time domain, frequency domain, quefrencydomain. 1. INTRODUCTION It is generally accepted that preventative maintenance can adopt consists of placing a spanner or rod against the machine and one of three different approaches. Breakdown maintenance is their earmuffs. An extension to this method is the stethoscope. practised when machines are operated until they fail before These range from simple rod cum tube to earpiece devices, maintenance techniques arescheduled. Scheduled maintenance to ttiose that pick up the rod or tube vibrations using a micro is concerned with interrupting the machines at regular intervals phone. The signal obtained from the microphone can be either formaintenance,inordertoreducethenumberofunplanned amplified or filtered before being fed toa headset [3]. stoppagesthatcanariseasaresultofabreakdownmaintenance strategy. However. the selection of an optimum maintenance interval is a difficult task in many situations. Too frequent maintenance can bea waste of resources as well as valuable production time. It may also increase the risk of damage to machinery that may arjse from human errors during reassembly. On the other hand,too long an interval would result in more failures than desired. A more acceptable maintenance policy would be to determine the maintenance interval by monitoring the actual condition of machines. This latter approach would certainly result in increased plant availability which would in OPERATIONAL turn result in greater return from the capital invested. Other VARIABLES CONDITION benefits include reduced maintenance costs and improved safety, particularly in machinery where failure constitutes a MONITORING health or physical hazard. TECHNIQUES The clear advantages offered by the application of the "on condition" approach has in recent times led to the developmentofa vast number of techniques for monitoring machinery condition [1, 21.These can be broadly classified under the categories described in Figure 1. Aural and visual monitoring techniques are considered to be basic forms of monitoring machine condition. If is commonly accepted that skilled personnel who have intimate knowledge of machines, are capable of identifying failure occurring by simply listening to the sounds of distress emitted by nearby machines. There are those who indeed have enhanced their skills by being able to listen to machine vibrations. The method Figure1: Conditionmonitoringtechniques VillUal monito,ing can sometim ..... Iso provide a di,1>CI. indica lion of !h. machine condWon without!h • ..-d fo' fUf'l!'leranalysis.Techniques'angefromu . ings impl.magn ifying ~ 11__ Of low power mia O$COpClS10 lilllrt .ssisled d.vicea ...L - 1. "' I auchas boroacopesaoclstr'Oboscopn l. t Other IOf.... ot visual ".j- mon~()finginelude .... ng dye peno:Ilfanta 10 prov;oe. clearer definition of Clack. occurring on the machine $Urf_ aocl the ~ lIStI of heat sensl tive paints also known . s thermog ,aphic .:.. peinta. U""allylh esepai ntsnee<:f lobe .. lact ed lo mat ch the lampe ,at" l1110 be meas u'ed wtl ich would involve ...... Pfior .!. I ..L knowle e",mentsNl'lObe_ lI.dvancedbelorea~ llbIechllnge i. produced in itI operational variabl.... Tem perature mon itoring conti !11i of measuremenl1i o. the Operllliona! temperature and the temperillu re of com ponent ..... ces l61. The monitori"llof opera tional tem per.lUra C¥Ibe considered ...... bset of ltMloperationelvariablaadi llCuu.ed 11",e (ms) Pfeviously. The monitoring of comPQnenlItri&cetBmperalure& ~ _ ha.beenfou ndt orai ete tO_ OCCUffing in ll'\lChina e!ements fit/4lf 5· T.... ~-'umr ol. particularly in journal be.rings where lubrication wei either 'nadaqu. te or ablll nt l7l.Techn iquesf or monitoring lem par8lure M.ch ine vitJrations arise from cyd ic axdta tion forces witllin of mach ine com ponanq inclucl. the UH of op\k;lll pyrometers . the m.chine. These lorces are somet ime. buill into the design t:hermocouplaa, thatmograpl>y and ,",st . nce thermometers. of the machine or c. n be dce to real che nges in dynamic Wewrdllb< i$ islJ8rll!fllled .l. elativenlOY ing tu1 llCesof jlfopert;. of individualmachineelement.d .... towea rorf. it....e. load bear ing machine elements. Hence it is posti b'- IO.s:sess ThaMexcitation !Of'tol;are transmitted to edjecent componen b the condition of these surtaces Hthe weer dabr is were coIecteer . ....lyai. l. uau. lJycarriadouI so thaI some char.eteri llica not readily obse rved vi.ually, are higi'olighttld, Saver.1 techniques ha... e;ltwr b-. p.opoMd or UHd in mac:hine monitoring end they are Ihown in Figure • . 2..1 Waveform Anal yai. Thi. techni que consiall of recording the time history of tha . vent on e sloreg e oacilloscopeor . r.. lti m• • nal...... r. Apan !rom.n obvious funda mente l eppreclet ion of the signal, lUch •• jj the ligna! w.' Iirlusoida l or random , it is particularly ul8flJ ln the study of non~ dyconditionl. nd ahOfttr-*'t impuI$M. Disaet. ~ occumnoin o-s and bearings lWChil broken teethon the fOfr1'l(lfandCf8Cks in1heinnef . nd outer races of th. I.tler, can be identifoedr. latively eal ily I12!. I An .. . mple is Ihown in Figura 5, wtlet . the waveform of the Q UH UN O caling vibration acceler. tionofa aingle .t8\ltl gearbox with a brok. ntooth onthepin ion is pretllnted . Thep inioninlhi s exemp li WIl l !MectIy coupl ed to • 5.6 kW, 2865 rpm AC electric motor. Under nominal load,ashllftspeed of approxi mately3000 rpm was obtained. Hence a single discrete fault on the pinion such as the broken tooth would produce pulses in the time domain with a period ofoccurrance of 20 msorso. This feature is clearly evident in Figure 5. Waveform analysis ~~~_:~~c~r~~;~~u~i~hidsehn:~fy~~~:de:t~na~a~~~~:ti~~:s~h~~:~! ~ r1n;J~ysiS,waveforms can indicate the occurrence of resonances ]- ~ 2.2 Indices ThePeaklevel,RMSlevelandtheirratioCrestfactor,areoften used to quantify the time signal. The Peak level is not a statistical quantity and hence may not be reliable in detecting XOisplacemendmils) damage in continously operating systems. The RMS value, however, would be more satisfactory for steady state Rgurs 6: Orbitmonitoring; applications. Even so, the value of either of these parameters -Bassliflllorbit---OrbitwithwornbeBrings tend to be governed by the amplitude of large components occurring in the time signal. Hence,unreliabledatawouldbe Mechanical glitch was minimised by accurate machining of the obtained when monitoring machinery where the time signal target area of the probes. Electrical glitch was minimised by contains information pertaining to more than one element, pushing an aluminium sleeve onto the target area. In this say a multistage gearbox where the time signal would contain particular test, the bearing inner diameter was deliberately information from the high and low speed gears as well as the enlarged by boring out bearing material. The results clearly bearings. showed that the orbit diameter increased particularly in the The Crest factor, defined as the ratio of the Peak to RMS vertical direction, indicating that this rotor bearing system was levels, has been proposed asa trending parameteras it includes stiffer in the horizontal direction. Hence the added bearing both parameters [14J. However, investigations by the author clearance due to the simulated wear condition resulted in have shown that this parameter usually increases marginally larger vertical relative displacements. with incipient failure and subsequently decreases due to the gradually increasing RMS value typical of progressive failure. 2.5 Statistical Analysis Ouite often, the trend recorded by this parameter has been found to be similar to another time domain parameter, the 2.5.1 PrObBbi/ityDensity Kurtosis factor. Evidence of this similarity will be produced The probability density is the probability of finding instant later when the Kurtosis factor is discussed. aneous amplitude values within a certain amplitude interval, divided by the size of the interval. All signals will have a 2.3 TimeSynchronousAverage characteristic probability density curve shape. These curves Synchronous averaging is the time signal averaged over a large if derived from machinery vibration signals can subsequently number of cycles, synchronous with the running speed of the be used in monitoring machine condition. Examples of the machine. This technique not only removes background noise application of this parameter to rolling element bearing but also periodic events not exactly synchronous with the monitoringarealreadyavailable[12,14,17].Afurtherexample machine being monitored. It is especially useful in gear of damage detection in a high speed rolling element bearing is vibration diagnosis where multiple shafts are present. All shown in Agure 7. Damagewas induced by producing a small components not synchronous with the shaft of interest can be radial groove on the outer race of the bearing. This action deleted. Atypical measurement setup would consist of the resulted in the vibration acceleration waveform looking not transducer usually an accelerometer,a tachometer which would unlike Figure 5. The waveform when the bearing was in good produce the reference pulse and a signal averager. Signals of condition was characteristic of a random waveform with a shafts not synchronous with the reference shaft can also be probability density curve that was similar to the normal averaged if the repetitive rate of the reference pulses are distribution or bell shaped curve. Note that the expression of altered with the aid ofa pulse frequency multiplier [151. the vertical axis in the logarithmic scale tends to produce a distorted shape when compared with probability curves 2.4 Shaft Orbits expressed in linear scales. This is deliberately carried out to Lissajous patterns ara'obtained by displaying time waveforms obtained from two transducerswhose outputs are phaseshifted by 90 degrees, on an oscilloscope where the time base is substituted with the signal from one of the probes. When shaft relative displacement probes are used, the pattern so obtained is the shaft orbit and can be used to indicate journal bearing wear, shaft misalignment, shaft unbalance, lubricationinstabilitiesinhydrodynamicbearingsandshaftrub [13J.The techniques of proximity analysis are well established, particularly with applications to turbomachinery. Eddy current tranducersare commonly used in addition to appropriate signal conditioners. However, care needs to be exercised when mounting these eddy current devices, so that the effects of mechanicaland electrical runoutare minimised [161.Otherwise, spurious signals will be obtained during measurements. An example of orbit analysis used in monitoring journal bearing NormollsedAxlS wear is presented in Figure 6. Two eddy current proximity Rgurs 7: Theprobebilitydistributionofbeering vibretion probes were placed at 90 degrees radially along the shaft. ecce/eretionBmplitude.-baseline dara,--- damagedbearing m~M lowp TheptObebilitydel'lsity 1lOfft'IeIiMd101h8~~ . ~ in N ti.... domIin wavef<:lrm. When . pplied 10 the e..-1Qr"~ t>N-na..tleneomparedwiththe....".jlfeWlul .umple of • t-ring wher e the out . ,_ Wei torN(IOOd~iII_IO'-~ difhw_in denwll ed, lhe trerld ofthe KurtoIieloc1Ol ielhowtl inFil! ... ll ~ .Thehigh""'oI~denlityMIheme ~ theempljtudeofthilperlfTlll.d~lill nifi· ~ lUllllelti"lltNtthewavetormo1tl-e cantlv.f1It11"'initill· time llomei ...oill" bel;e iflCl.. si"llly more random .. failufl p,O\lreued . Thi ". 'l urlVllhowe<:lthattheKutt Olis pa.ameter .• llhoull h ueeful in.om e!or m.of fa' lll.tt. o;ouldnot be ,~"d on .. . IflHlditlll pa rem. " ,lt for the pllrpo&ll of P helarll<'ltdilClePl ncvoccunedinthe oond~ltd,l .... mp" shown in Figl.O"9. In aome U$et. the lfWKb teCOfdad by both .n. ~ par_w lfI.lmoItidentJcel .. 1haI simple OIltIec101met • • whIdlOOSlsillnilicantlv ... ~ K...... mer-. .ceninl1Mdbtuoed in plac:eofv.elllleriu" rnorOtorinliIroIinll""'-biIIIW'lllcondilion. 3. THE fREQUE NCY DOMAIN Fir/vIe"A ~ -"' _ -,- DillitIlI8ltFo..rilr"*VMofthe1irne-..bm""~ 1he_popuIIr~ofdlrMnll_lreq.8leydomlin 2.5.2 PrrJbMJtiitvol &-* aisP"IIln.lil;rwtInllC*lfU"lOob&lIir.cll13 lcenpn;Mde Theprobebiityof~hM..,~uMdin~ vMIIblI inlOlfftltlonworh regIfd tr:I fNIdline 00f'0ti0II. beP"iJ~ (17).I1i11he..... ofNprot.biIity EnveIoIlinllorciltnodullcinllNtime..-tormprioflO d«'oMv_ lf"dllftoes1heprobebikylhM ....--.. P«forminllthelIltFo..rilrtr....,.,illIIo~popuIIrity ~""""""-=eedII....,~ pMude .AgM\. endisnclucModinlh8~6IIeuIIiononftequency It-. C8Il be moriIOred 10 pIO'OOde.. inclie._ 01 domIin~FNIy .lJ)IIChlinformltionean""'be beftIlIblUe. de"ftledUllng~fi""_l\nMItoPlllinllm. EKhof~~ ~ intotmIOonOlWvinbendiofinllrIll. 2.U I+ob«JilifyDensirr cen berellted_lho:M'n in FIQl.O'.10 ,",-"'-of1heprcbabTitydeneify_eanbellwenbed by . -'es of q le no.mbelindiceL n- .. . rnornenbollhe C&ro'lIlndanelogo:xdlDlTIIChIr'icelmorr.ntalbOUllltoeQl1lroid 3.1 The S!llnatu•• SpKIfum • • ill of. plene. The tnt.., d IIeClOI'dof u- momenb'" The vitirational charKtltllii ticaol anv mld1inl •• 101OIftI ...... krowrl, 1tllI mean and ~ ~ • . Odd momen t. ,elet• .. tant uniquol, duelo the _ iouItr _ teo c:hlrea lriltic:l of thl intormetion. bouttl-e poeition olthe pRolcoenli ly .. llnN.1O the me<:lilIn "",Iue. Even momenb Indiclole the IptNd in dl.tri bution. UlUallv moment s llr. alll th.n two , r, no.-melised bVr. movI"ll the mea n and dividi"" bV thelUlnda rddaviation ".ed to the order of the moment. Th. third fllOfTlIflt is Sk' WIllIS. and the fourth i. KurtOlli. For prk6c .1 lIiynell the odd momen ta ere usualiVcloH to _0...... m.hilihal' .ven momenb ire --..itive 10 impulIiyer-. in the oillnel. ]:=("/.,--..,,;...;! ...\/, ,., Kunoaill hal been" l8cted .. . oompromi.. ..-... bit-. 'VP1~ the --.rti ... ~ momenll If"d tIM _----..itNe hilII* ~ .ThisPill...,«tlllb-njlfopooedlObe..,.;u...1D ...... inrollonllelement~[17).~ ..... irOIperdenI: • _ ,>I ~oIthiit echn;que118J'-Ihowo'ltNt.highKun:.il v -*ionlybe~i1m.oriIlinllm.~ __ ~ .. n. ~ _ l , _lIa(houe-rIKMrr!) /oI _of..,~..nn .ThiI~don""in lOIftI~of""ein~lUCh.aecked_ _ _...... ,..ZIhocn01_ ~ The mwlhod of .-nbIy , rnounUrlg end Nt*tion of .... lNChine, .. pIrjo.pMin~.eOon .~ . ~ .whIn. ~ been~ .... • • ~ ... spec1JUrnlhouldbeobtloined .... normel l\ftling ThiI~wil prO¥ide. bail.b'" conc:Sitionto. ,.,. ,Z:SpecnI"""'" coonper;eoninOfdertoloo::tte~ fNquwlcieeinwhlch ~n:._invb'M icftlewlkeYe~ .Agu.l1 ~ lO~spKtt .dwlwdftom~tr..-duc.r1olOO , .... .-mple ofhow.,..etive llignatw.IIpettr .. providldthllPlCtrldo notcontllntoomuctolnfo ~ lndic:l1l Increil_tl'lfouoghoulth.lrequllOcy,ange.The dutttInId . ndt hlref Of. of limited uw . the poeitlona of thoM h'equ. ncy <:omponenlJl r. loIled to the ge M mesllCOfTlponenlllld two hlr monb. Indee d thie t.ch nlque 3.I.21r1dice, ~ ~ .widelvueed.OftllOlI"1op11"to< ofsp«:tr.I The emount ol lnfol'matioo 11'_ in . plgI~ ear> be reo:orchOflI.ittl .. l:I Of1.tllperecud .. ...:l~thlmln '1du:ldif. chapectrllel\lngelh8toccunoed ...... tt>elebofetoty Of otllc» UIing.d igitll ... tln.-IadFat. . in lliingle runber. VlriouIlndicel hIVlI bien proposed [19~ .....,.. . MrouIl OllmPlrilOnl•• lhenmede ,Mor.recently . By end 1ervI.lt-. $pICtral ~ ...... beer> found 10 be the pr_ '- been IUtomIted son-"-I by UIing pCftMliI mono -.itNe INn .... domein nkII IO.och...... M1d ~ ~with..-yend_ in1lwmonitorWlQof~ RM$ ...... beImgs [18L ~ mlnd._--.dy ,o-s inMliglncl . n-devicM Unplify ltle dWI a:quisi!lon jlgniII l2OI. k11OfT11 procaaendthe_vl~ ~Id.1'IiIbII -.tt.~ intr...o.'lOOfdedby1lwfTequwocy ~IXIfTlIlIUIIn_ inI.,.....w uto ...... ~~caotbolQUite"" .Fv-13 lO~jlgniII~ condition. 3.U~ whlnepplild ~ ~ Li. ~ tlonCll'lbeUled lXlI'Iditionhlcloct....cl.Tl'iatypeol' rnonitorlnOjounwI"'ingcondition,l...:lbeen~ Pf~[7 ~ ~!, r f i;f;!ld0~ ~~ J : _~ .'=-. ---, :;,----;_=- ----=H.. , Freque"Cy (Hz) 11.. (1-0.15) ,...rr :~--.. __ T_ ....-..- ~ ,..,."'"'-l. ...- ...... '.1" ~ IrwI ~ ~fJI~~ _ ,fWS · _ ~ _ _ _ IWS,."...... _ ~ R.. fI1MS--.t "" ~ IIiPflI¥. W"..-.II~ .a.J.-.d""""""' ... -.IIMd rJD IgIrf,-/Md __ /d Iltinl /wmolW fJI"...... __ ICl_'-' 3 ~ '. Mtuk$ 4. THE Qt,lEFRENCY DOMA IN ~ illhe~ bthllClltPl1"Jmwhodlisdtfioed. An..,. ... livemeltlodoftryinll lO...... IIt~oc:a,ning ~ .lI:ilUMd m in tNlIillNllunl lpel;\J\Im ~ Io fof.1j:leCIr1ltl"... which .. thIItlCJKlfumof theloltill"ithmofthllt __ ~ tohighlightperiooflCiIieI1:hIIoco..o'inthe~ inthe Iromlhlltbaselioe~ ~ llill"ol!Urlo ..... IIowIItbIeIOlow._ 1imi1121L ~ i&..-..d __ ol_ speclr.withlh .. n-'< I8memrn.- .. the IOhig highl:periocIil; ~ ~occurrinll inlhllttimeciornlltin.;gnaThto.r..tlon will indicM. if. .. ocamng. SomMi..-. ""' .... IOW c~ ilnoII.IrMlllmaner lndc:..lf-.t ~ d thllt iI~ IO.OClI'IItWtI t.od'wIdl:h~ ~ .. ~~ to~ .lorilOMld ex-.;iMd ""'*' doing eo I26L Cepmum IItnllIyW ... e.o.d"..., ._l.-d jf .. ~~ --..:l. ~""uNd inthll....,...~(INfbolIvibr"" SornIIrty,'*"OWmMb _ uMdil'odtrminar~in 127}.~~C*l ott.nbedftectlldinlhllt~ ~~ /)f1liclllbllndl ~_IItntio:ip81ed_II" IO"notedlhfl.~'" duelOlhell'..-.oIoI tN~-..p of IItml)UnIItII.xpoll'ience"noquir.clin ___ in IN specfnm. An • ...,p. iI in f9n 14. whid'I ~ pr-.d ...,p;NdoIIIt ;l~ Cllllifl9 1TlIlnI...... ,. 1imi'I1NSk•.• ..otntionapeclJ .. "-1hIIt~of . Iinglll.clltgllt' ~ III207S __ dw9und11tr.....-yingduty .m. tIIlhe lT\lldMoe. ~IItion IignM.Tht \lltr'gll _ corraponds to thll st.ll fundarnenU1lmodlAlllioncomponenf ~lIOlw mcricin andlN1l1141.6_ il,rIIthmonic~ spedr ...... T...... tolnlOllnitudlltof ~ ...... ,roc:r... speo;nI~of"- 12 ThIoEtweIopotd s,.cvum ~ -M1hincr-.ing _ inv-s. n.. "-ve been...... -.:- wt-. T...... "..lop.cl~..... bMn ~ ..-.IO .. ~ q.eII'''''O''lwvelecorded~ in . .-tul inmoniloring.....,...... -m-f .. b¥1lfOducinll bIItrIdwdIh~of An _ ~ il Pf..-rl lld in F1gure15. ThIIt· ..-uwlylhorllto.nliol'limpu .... I22L. t..". .lYJlicaIof _ . b!oIdIrlIIdI01:hIIfilllt"'Ntude thllQllpe;lr"~ ~ tO'mlltglltinrolIltl ~"'"""' "lltbeBrif,gl.Ohenincipient v-wrtio nl could be fIItIItd,1y C1-.-.:1 . The If.,," _ ~ inrollinllI~"ll.elf'lnOl ...... detel:ted usinll obt.n.d lJIing ealing aculeIlittion _ 11",,1$ trom IIw w.glllt ~~.IpIItCtJ.lc:ornp.-_... tllllt _ llY conl'ibutlOnaby stag.ell_bc xwhich w...... bjlltCledtolll1O'1l1tr1oMlcondilion. impulses eee -.lilly SW.mpMl bV thIIt vibr. 'oon com T"- iniIillI lna llllH in gamNlllde o! Ihi fuoCIIImentl l componlf'Of ~ofll'lOnldominlomlltlementlor.xt'~lnf\ullnces and ill I. t-n>ooic wal dllllt to pini"" in the p itch Ii"" 0 1 IIw of .-bylnlC hinery. Tht wchniQueinvo:llva r tly, l high Pl!tN "' spu r II-I . Conlinued Operllliol'le--CI further da~ 1Itauch .. tlltering operat ion 10 ,emove domi ... ~"lllow f,eq uency scorl""o!thtladdendum.nddlodCIenCIllmo l t"'tHtl'l. The eompon.,tl in ItMt spectrum . Tht resulting . ig... 1 is then tell_.linI llydiloontinlleddlllltoe xcessivlltvibo'lI00n .rd '.clif,..jplltrtialivorfllllv.AnomMIlf -..oelledthllt~ ~ .ee.mg~ in ~O'NdW>ery""'bMndlltte<;tedwith... I«:tw'oiquIIt123L Cautiorl_be u .ciMdwt-.~roling""""" .~ """ moCIe ~ingswilto.ITIOI" . • __ high ~ ~ _ """"' noI Pf fI' __ in"" colIectIIId fl'omluci>bearftlla.~ .Ihe~ .....,noII" ~ Incs-.:tinllWOM ~in beaings. 13 ,.... s-d AtoeIyM ~ oldlltllt""'" YIll:.-tKtoroiqo»tII~IN ~in . ~inIo~pn:>I)Of'I_.lO triv o...rrency<..s) mofttor.lMtIdof~lItOIf-.brOIdOf_ .inwhich ~ I2.LTht dIItfett"~tII~ _ Aponl. -A __ -....-...... __ ~ -..otJ-.as monitorirlllof.I'>I cN.-aetarinil:f<~moMOIinll·M...,d lheemsti"ll ~ ~ lItl*:bga ttWonoorpor-lII·1JetIcfinll vibrllttion ...... e!w-.;teoilbr: lMquon;:y moni~ toring. 1ilInce, to' . qmpIa, thIIt ~1fId01 ItMt il* ...... or beMng deI_ftIItQuIItrlcle COllklblltmonilo"«l.Ao- . IhiI .PPfO"Ch if ..-..d IOlelv __ ~probIoImI. F lhl~ .Hence. bvmenvorher"equerqrc:omponenl.in 'rend i"" , par. ""' 1f tflal dIIlCfibes.tl comPCl'*' II in IOmIIt WllV,wOllldpr(lY\de. be tter meaSU'eo f lailll'. OCCtlrring T'" I hoc k PIli.. INllhod (251 ml y blltconli doo,ed to be ll IPllciIItliNd .ppliclllion olcN rleteriltic lr"ll lllll'oCV monitoring flit... of higtl lpHd lolling eIIItmeot bMf,"llI IMII'" in ener ll'l' being....medlllurtr.onic fl'lIQUIItfIcies.l'hiItee hniqllllUlllS ... 1ItCCeIer_I\M'Ied'l'llltChllriclllylltfld llllctricallv llO.freq...... -.cy of32kHzlO dllteetu-distr ...... III1 • .ellltWeIv ~ inindultrv'ornonilor 1irnJM-.. __ .ndilwidlltlv highspeed~ / ~ n. f~ ~ lICOUStiI;noise . Signat ure 8j)8CVlIl oo mparilont .no-! Ihet r. IIA , Mect..- _ Condition the base sp ect ral noise level in.c,eased $ignlficant!y with Monlloring. ChopmenendH:ol.London. 19n, O>eptsr l Q. ~ IOIhoCondition Monilanntlcf ed vance d damag e end te nde d 10 o verwh elm the sheft II N_ ,M~endAMoQltes.A. modula tion components. HllI1ce. tna e mp ha til of the modul ation Machinery, Her Mojelty'. s..t~ Office, London. 1m. components In the specttum wes red uced , resultin g In thlI .,.... 1 MlIIoew. J.. .-.d Alredaon. R.J.. ..".", Cond~ion Monitoring of Ioweirog of thlI gomnitudes of anociloted quel ..-.cy l'rocesdingaol!hsJoo.n>oI~Inu.""",,,Vibr-'ionsndSli AA __ CT.on/fIrencaonConditionmJlB'.lUraArwlytil". Othflr fofm.C8P8 of trt1_ lyosislnducla~ ot Monitoring.Ilrighton,~.21·23MsyI9llll.Ilp.21-033. ---'''' natu. a ce pstra end 8ing le number Indicea derived ffom 8 Co1lKon. RA , Mec:hInicsl Feul! o;.g..... end CondiIion euccenlve c epstt a ln mucht hesa me w ayllllllg nal1.lrespectral Mol'litoring . ChepInen.nd Hal.1.ondon.1971. o.ac-. 8.-.d 9. techn iq.- a nd ep ectrll l lnd ices are used. However. one m ust 9KuMe1I.8 .T" anclStedd , J.S.• '"CcwraIMIonofYilrIllon, W_ beP ~ . _. 13 Collecon. RA . Vlbronon Monitoring ..... Wiley. IMw d lfl.. en t disclpl lnee .r a e mploytid. for eumpl e. spectrometf ic york. 1m.Chapter " oiI.nalyl is can only ptOYide infOfmet ion on the concentret iona , . ar. un.SG .• "Ths$iognlrtut. Aoolyall of Sonic IlurinCl Vllt.Iiona ". oI_metalelnanoilumpl e.Parnclecou ntingontheomer IEEET .. nuctionoonSonlcoandUItr~ .YoI.Sl).27,8. henclonIyp ~ RA .VbationMonilOrInu_O iag_ .WDsy.New oI_debril.Vibf'ationrnonitoringilwellsuitedtowardathe 15 monit oring ol '(lL1llng maehineryend ClIn p 8_ .J.,endAlfredeon.R.J~·'l1IeCondibonMorOloringol monitoring rolling element bearinge but is unsuitllble fof 1 monitoring jouma l bearing s. Similarly. eeps ttum enelYlie Ie I I\ok>g Elernom Bs.ingI IlIWItiYibnltion """""" ", ASME useful too l in dleg noalng gear fajhJrBl bu t 10 far ha s not bMn Tren"C1Ion1.JoutNtlolViltstion.~S~nReliobiily In Oaaign, Yol. 106. J uly 11lM. pp 7..<163. em ployed in det ect ing .0011ng eleme nt bea ring f.~ures. nd ." 19AJft~ .R.J.,endM-'t- .J.,·'froqusncyOomalntTMllhadlkr ~ y. with good reBSOf'\. it woul d appel ' that the monitoring Rolling E...... t B_ln g. " . T•• naadlo nl of Ih. deleCtion,diag~erdprognotisol_orfallureocculfing lnstiIutIor1of~YoLM £10.2.JuIy19B15.pp.10B-1t2. in lTI8Chioerycan be g.eatIy enhenc:ed if a qndroma spproach R.J~ 2ll~ .J . n All, ''l1Ie MonIlcMg of GM,t>oa ~ for Iaedopled, in wllichacombinetionoll'\'fllplOfTl&iclentifoed by VUstion_Sling StMdyCondillonl .., a pprop " nInflfor!hl$oufldlofB_n gTn>UbII". 2"'_ .M .. _~ A~IO!hsCondition_ingol 25a-d'lack.RC~U· " ~OfIIign, ~."'Mal-!Y ·.SUbonsryOlfc..lDn _ ,Tod~lmemslion"."""lIItl981.pp.»J3, EIac Rev\Iow.No. 3, 1981• • "'''''Ie. M., . nd Associeler., AGuld. 10 the Condition MonitoriiIg 71 Rsndl •. R.B~··A.dvwiesainIhoApplicatlonofe-cru-nMslyailIO~ of Mschlnety . ... Majesty', SWllor>ery Office. London. 1979, Gsart>oaOillgnooil".Secondlnt.. nolionefCooo""ence",, ~ . pp.U~. In Rotiling Mscninery. I.Mech.E.. camb< pp. 189-11. An Anechoic Chamber at the Australian Defence Force Academy J . C. S. lai a ~=rs~e~~~ie~':.dl~~~ef:I~II:::Ty8;;O~?~:~~ :ge i D. p. JtIMn I 01 Mee::M nical f ngl _ ing squarect ' rom the lJOU,ce. Flgure 2:dl. ptay. the . ver. ge ~ ., U..... r'li ty Col Unl....ni l, 01 N_ South WI I• • maximum devi. llons from thtl ill... rse lq U8r. 1_ 01 Au. " .. . n o.f erlC.F ore. ...u dem y the trav erses . Iong two dll1arenl microp hone palh. A.c.r. 26OO,...... Iia for the frequ ency rang8 100 Hz. 10 12.5 kHz . The lell~th ". of tr averse for each pa ltl it;a l le... 2,6m. The req UIre The A",st .,l ian Oeleroc:e FDlce Academy, incorpor at l'I'>ems of ISO 3745 h..... 81so been plolled III Figure 2 ing • UI'l,....,,, ity College of th e Unlve" ity of New South for compa rison. "' Ilow ing 0.5 d B varl .tion 10 . ccount ~ nberr. ~ Wa le.. I, loe aled in Id jaeen t 10 the Royal Ior t l'leo nsll'VtlHlll3ancf ~ .. fthtlrne.lUrf ngl MIIII. ry College, Dul'llroon . Th. fir,l inl. 01 slu de nts some irregu l.ri ti.. of the movem.enl 01 the mic rophone, occ ul red at the beg.nt'l Il'lll 01 1986 . F.c,lil.., na.... bee n It-e chamber adequ8 lely utis fies the ISO 3745 requ ire ~ built 10 cater 101' the neeQs 01 1xI11l teach ,ng a m8I'lt, tor .!ow'e, cut -olll requenc:yof 150 Hz. ,esearch. In Itle fiel d 01 aco ustIC" an anec::hOlC It se ne, of tests h8s been concIucte ~~O'Z~'--' viny1idenechlor idej . The " lU llin g free-lield dimens ions , r- r-: are 3.5 III x 3.5 III • 3.5 III belw ee", the nps of the dented s1ab8 and the lowe r cul-o fl lrequency is 150 Hz, ~ \-~------'::- The lmer shell h., • wlre-mes h floor ma de of 3.2 III - ltl ick lIal v. nized stainlesa Ileel caIeS, ~ t!'le mesh aper- ~ --,,:::' , V .._ ' ''0'1.' T~ ':?i:r~":.=C~ plr': =r ~~~i:::. Vjcledfor$4l~o fmicrop/'lOfles.nd sm allitem. ~ Z~ ~ ~. amaler ,. z. •• •• ~:~J::-:P~ ~~r::;~~:.dl are available 'n The d"II' " 10 which the anechoic chambe r approK hes 8'1 idea l free field has eeen lasted by measuri ng the ...."et ion of sound pr essure With . d S tence along the r. d•• 1 ~ fechon ' rom en omn i-d lrec Ii onal sound .cu rc. 8ccordlflg to ISO 3745 standard /" - --,------/ ' IftQul~[M[NT ....//' /~ ._'\"-- _ ,,/ ' 't'AOU CiroOU'[ .OQOII ~ -- 'O Q Z'lO SQO I• •• •• 'of , ,_ . .. , 1"' 1 Applications of Underwater Acoustics to Australia's Maritime Defence Marshall Hall Defence Science & Technology Organisation Weapons Systems Research Laboratory Maritime Systems Division Royal Australian Navy Research Laboratory P.O. Box 706, Darlinghurst, N.S.W. 2010 ABSTRACT: Over the past three decades, Australian Defence Departments have sponsored significant research effort on aspects of underwater acoustics that have an effect on the performance of sonar systems. The reasons for this effort, and its achievements to date are discussed and related to changes in the requirements of the Maritime defence forces. INTRODUCTION ledge of our maritime environment is crucial to military operations in our own defence" [Dlbb, 1986, p.64]. There has been a great "technological change" in Oceanographic parameters are complicated functions maritime warfare over the past few decades, stimulated of space; many of them fluctuate significantly with time of course by the Second World War and the subse (particularly in the upper layers of the water column); quent "Cold War", and made possible by the advances and some of them can have critical effects on the per in electronics that have been a feature of the 20th formance of maritime defence equipment. century. A feature of early sonars was that, in the absence of Prior to this century, the navy's main form of remote a perceived requirement to do otherwise, the dimen sensing was the telescope, which assisted the human sions of the transducers (projectors and receivers) observer to detect, localise and classify significant were comparable with the wavelength of the carrier objects within the field of vision. Advances in the signal (which was generally some tens of centimetres). science of electromagnetic propagation brought about As a result, the width of the transmitted beam was radio and radar, while progress in acoustics enabled large (say 30°). The phenomenon of background noise; the introduction of underwater sonar (initially stimu either reverberation (the reception of large numbers of latedbytheneedtodetecticebergs,decadesbefore small random echoes from inhomogeneities in the radar was developed). Sound is used for detecting medium), or ambient noise, was therefore significant underwater targets because, at useful wavelengths and but could be studied in terms of averages over com frequencies, it is absorbed by sea-water far less than paratively large regions. Initial research into the is any electromagnetic transmission. The average properties of the ocean therefore emphasised a coarse speed of sound in water is around 1500 mls (compared scale approach. To reduce the background noise, and with 2 x 108 mls for light), so to generate a carrier hence improve the performance of these sensors, the sound signal with a wavelength of 10 em, for example, beamwidthwasdecreasedbyincreasingthetransducer a frequency of 15 kHz (toward the upper limit of sizes. The effect of this has been that, in order to be average human audibility) would be required. able to predict the larger sensor's background noise level,thenoiseorreverberation producing properties Radar and sonar have both played important parts of the ocean must be re-examined in finer detail. The during maritime warfare over the past few decades, noise problem was not "solved" by reducing the beam and are expected to remain important for the foresee widths, since with increasing capabilities by platforms able future. Their importance to Australia has recently and weapons to act at great distances, it has become been reported in the following terms: "Australia necessary to detect targets at longer ranges (that is, requires a manifest ability to detect, identify, and track still at low signal-to-noise ratios). Thus, although potentially hostile forces within our area of direct decades have elapsed, the need for environmental military interest ... modern technology in the form of information has not (yet) diminished. over-the-horizon radar (OTHR) offers the prospect ... In addition to the design of better performing broad-area real-time surveillance of our air and sea sensors, another important facet of naval operations approaches out to 1500 nautical miles. Another promis is the routing of ships and submarines in order to ing new surveillance technology is the towed acoustic (a) minimise time spent at sea; (b) minimise the time array, which is especially useful against submarines in spent in enduring rough weather (for equipment per favourablewaterconditions ... " [Dibb, 1986, p.6]. formance); and (c) avoid regions where costly above Over the past few decades there has been a steady surface or undersea hostilities would be liable to occur. emphasis on predicting the performance of a sensor For the first two of these objectives, it is important to within a given situation, and also on improving per monitor large-scale features such as the ocean's formance, by design changes, wherever possible. A "weather patterns" and their associated currents (as conclusion drawn from these endeavours has been that well as of the meteorological conditions). For the third the marine environment needs to be understood in con objective, it is necessary to develop an atlas of the siderable detail. In the Australian context, for example, (slowly varying) ocean properties that affect sonar this conclusion has been recently re-phrased as: performance, as well as models of the rapidly varying "Australian hydrographic and oceanographic know- properties in terms of observable parameters. In this article then, we will first discuss sonar as an (c) a submarine can often increase the transmis application of military oceanography, and describe the sion loss by remaining in the acoustic relevant environmental parameters. Secondly, we will "shadow zone" in the thermocline beneath summarise the progress that has been made by the surface mixed-layer. research into understanding these parameters, with (d) the strength of the echo is decreased by twice emphasis on the ocean areas around Australia. the transmission loss that applies to the dis tance between the sonar and target. SONAR B. Passive Sonar The word "sonar" is applied to either a passive (listening) system or an active (echo-ranging) system; With the realisation that the detection ranges of and both types have been in use since before World active sonars appeared to be limited, there has been War I!. an emphasis (especially since about 1970) on making the best possible use of passive sonar, which is basic A. Active Sonar ally an "array" of hydrophones (underwater micro Early active sonars operated at carrier frequencies of phones). A passive sonar simply listens for noises around 25 kHz, and the average detection range of radiated by vessels (usually from their propulsion submarines was only about 1 kilometre. The average systems), and can be either mounted on the sea-floor, detection range was subsequently increased in the affixed to a submarine hull, towed by a vessel, or post-war period, partly by reducing the carrier fre dropped into the sea by an aircraft (the "sonobuoy", quency (to improve the signal to noise ratio), and of which BARRA is a modern Australian example). increasing the size (and hence mass and cost) of the The signal-to-noise ratio at a passive sonar is deter systems commensurately. The Australian MULLOKA is mined by parameters such as: an example of a modern active sonar, developed by (1) The "loudness" of the target; DSTO over the past decade and currently fitted to (2) The transmission loss that occurs from the target several RAN vessels. to the sonar; and The signal-to-noise ratio at an active sonar is deter (3) The background-noise level at the sonar. mined by: For stationary sonars,the noise is generally the ambient noise due to the effects of wind-driven (1) the loudness (source strength) of the pings emitted surface-roughness, biological organisms or distant byaprojector; shipping. For moving sonars, flow noise is also a (2) the transmission loss that occurs from the projector contributing factor, especially at the lower frequencies. to the target. Transmission loss is the ratio of the Ambient noise can be reduced by designing a (large) power output of the projector to the intensity at a directional system that responds only to sounds from given point and is a function of range from the a particular (but controllable) direction. transmitter; Frequencies low in the audio band are often used, (3) the target strength (scattering cross-section) of the for the following reasons: target; (1) sound at these frequencies travels long distances (4) the transmission loss from the target back to the in the ocean since the frequency is low enough projector (which is generally the same as item (2) that the effects of absorption, scattering, and unless the vessels are over a steeply sloping bottom reflectivity on transmission loss are small; bottom); (2) at these wavelengths, it is feasible to build an (5) the self-noise of the sonar; and array of hydrophones that is sulliciently large to (6) reverberation,adecayingnoise-likesignalheard be highly directional; and by a receiver after every ping. Reverberation is (3) ships and submarines radiate substantial amounts caused by backscattering from inhomogeneities. of noise at these frequencies. With the advent of long-range guided torpedoes and ENVIRONMENTAL PARAMETERS submarine-launched missiles, it has become desirable to detect submarines at long ranges. The limited detec RELEVANT TO SONAR PERFORMANCE tion ranges that are achieved by modern active sonars From the preceding chapter we can compile the in some circumstances are dueto a combination of the following list of oceanographic parameters that can following factors: allectsonar performance: (1) Attypicalsonarfrequencies,beam-widthsofrealis Sea-surface roughness' and backscatter. tically-sizedtransducersaresuchthatsignificant Sound-speed profiles in the ocean. reverberation is generated by biological scatterers Acoustic properties of the sea-floor. such as deep-sea fishes with gas-filled swim Acoustic propagation in deep water. bladders (in deep-water), or by the roughness of Acoustic propagation in shallow water. the sea-floor and sea-surface (in shallow water). Ambient sea noise. An important consequence of reverberation being Acoustic scattering. the effective background noise is that there is We shall briefly review current progress in each of nothing to be gained by increasing the output from these fields, with emphasis on the Australian context. the projector (the reverberation also increases and Several of these parameters (together with related the signal-to-noise ratio remains unchanged). oceanographic parameters) are being charted in Aus (2) Since the transducer is affixed to the key of the tralia's maritime "area of direct military interest" ship (it is very expensive to deploy otherwise). (ADMI). the followinq problems occur: A. Sea-surface Roughness and Backscatter (a) the receiver's self-noise is increased by the proximity of the ship with its moving parts; Backscatter at short wavelengths (approximately 10 (b) propagation in the horizontal direction is liable em) is associated with the variation in the slope of the to degradation by (i) scattering and attenua surface and is therefore dependent on the instantane tion by gas bubbles generated by the moving ous wind-speed. Surface backscattering is anisotropic, ship and/or the wind; and (ii) downward in that the return from the downwind direction is less refraction, when the water near the surface is heated and a negative sound-speed gradient 'Footnote: is generated ("afternoon effect"). The effect This parameter is often referred to as a meteorologi ive horizon of the sonar can come in to a cal, rather than an oceanographic, parameter; but is range of less than 1 km. included here for completeness. than that lro m upwind. Damped ca pilla ry waves and C. Aco ..stlc Prope rties 01 Ihe Sea- Floo r short gravity waves tend to ride on the leeward. rather Ihan the wind ~ward.side 01 longe r gravity waves. The sea-uocrs are two-phase media in which the solid parll cles have variable properties such as chemiCal result Is that th~ leeward side 01 a 1000 wave Is steeper than its Wlnd-ward slde and tneretore relurns s composition. grain-size distrib ulion and degr ee 01co m stronger rettecuon . pacllo n (related to " porosity") . One appr oac h to pre Back scatt er at long-wa velengths. 10 a tirsl order dictingtheiracous tic properties has been to treat tlle m eprcxtmancn, is associated with the larger-scale rough· as a one-phase solid and to direct ly measure In . itu lne speeds and alt enuation rates of tlle compressiol'lat ness that is at the "Bragg-fesonaoce" wsvelength . and.shear waves. and thedenslly. Wit.htheae quatll ities Since. the surface waves are moving, the beckecatter avail able, eom me acoustic rel lectlvlty (at any Ire exhibits Doppler shills . A sonar wavelength (X,) of 20 m lor example is in resonance with surface waves quencyan d incide nt angle) and the behaviou r al lhal have 8 wavelength ()..l 01 10m . (The phase speed energy that penetrates the eee-nccr. may be calc u lated. Measurem ents have been made 01 the acoustic ot these grav!ty waves is V ~ " {OM 2" ) - 4 mI a, and the trequenctes 01 tho Bragg peaka in the Doppler interaction in the deep ocean using sou.ces.a nd recervere placed nearth e sea-Iloor, so as to provIde a spectrum are l:>f = ± 2V/ A., - ± 0.4 Hz IIbove and below the came r trequ ency.) wide range 01 interaction angles, a stable predict able water environment above lhe sca-ucor , and an absence B. Sound-S peed Pr olil ea In the Oc ea n 01 any enec t due 10 the sea eurtec e. These measure ~entshave been modelled using computer programs , A typical sound-speed crerue in the ocean exhibits In order to determine Ihe venenen 01 sound speed with the follow ing teatur ea: depth into the sea floor (Lawrence, 1985). This typeo ! (1) From lhe surface to a typica l dep th 01 several experiment deals wilh the top lew hundred metres of tens 01 metres (allhoug h oecasionally zero).t hore re an sedi ment, in cont rast to the deeper depths thet are isothe rmal tayer In which the $Cund-speed slowly examined by seismic profiling . Also, charts showing a increases (or the renec nve index slowty decreases) variety 01 lealures including bathymetry , sediment wit h depth. This eurlace mixed·tayer rel racts sound type. ane sediment thickne as are being prod uced. Work rays upwards (With a radil.lSor curvature 01 around 90 has also been performed on surveying the geographi krn] 10 reflect off lhe surface . and tberetore forms a cal variation 01 bott om interac1lon propert ies. using the surrace-euct. simpie r proc edure 01 near'1lllrlace explosive charges (2) Beneathth e mixed·layar. lhere ia a '·tharmocline" and hydrophones. These measurements determine the in which the temperature. and hence sound-speed, loss of aco uatic onergy on reflection as a funct ion 01 both have negative gradients. At medium and low angle and 01 frequency. latitudes. the thermocilne extenda to seve-at hundred metres in depth. D. Deep-Water Aco uslic Pr opagati on (3) At around 1 km depth. tho temperature gradient In geM ral terms. deep water propagation is weI! becomes vmy small, and the sound-speed gradient understood . especially lor cases where scatler lng near increases to zero and ecounuestc Increase (to a value the see-surfac e is unimportant, where variation in the of 0,018 ml s per m at abyssal deplhs) . At great depth horizontal plane i\l very gradual [eo tha t mode-coupling ~ (4-S km) the eound-apeed can Incre ase to its value at is neglig ible). and where bot.tom roughness i not tne sortece so that. regardless of the parti cu lar depth important. Under these co ndItions, well-establis hed 01 the sea floor, lhe eou-e ocean layer isanacoust ic approxim atians lo the acoustic w8ve equation (such as duct. adiabatic mode theory or the parabolic equ ation theory) Target Data Channels SGO"'"1ll HullMountedCL7Sonar can be used to study the acoustic effects of oceano Surface-Roughness graphic features such as mesoscale eddies. These Acoustic surface-duct transmission is degraded as studies have shown that such features have important the surface-roughness increases. Since MLT and effects on long-distance propagation, such as altering roughness are both wind-generated (although MLT the range at which zones of high intensity occur refiectsthe history of the wind over a period of several [Lawrence, 1983] or refracting the sound horizontally days) it is plausible that these two variables could be to produce a "bearing error". correlated. If the correlation were significant, there (1) Surface-duct propagation could be interesting effects on the probabilitydistribu tionoftransmission loss. For cases considered to date The acoustic ocean surface-duct is analogous to the however [Hall and Sandy, 1985], the correlation radar atmospheric humidity duct, although the shape between roughness and thickness is only about 10%, of the refractive index profile in the duct differs (quasi and the consequent effect on the probabilitydistribu linear for the acoustic case, but quasi-exponential for tion of transmission loss is insignificant. radar). The acoustic surface duct has a maximum, or cut-off, (2) Convergence-Zone Propagation wavelength, which for the typical profile can be At low frequencies (for which the surface-duct has expressed as no effect), long range propagation occurs via upward Xm,,~8x10-JhJ/2 refraction at depths ot a to s km. Fora shallow sound where X"' is the maximum acoustic wavelength source, this phenomenon gives rise to "convergence (metres), and h is the mixed-layer thickness (metres). zone" propagation in which the sound rays (in the For a typical layer thickness of 50 m, the maximum vertical plane) return to the surface, and partially trapped wavelength is therefore 2.8 m which corre re-focus there, at ranges of 60 km and multiples sponds to a cut-off frequency of around 540 Hz. thereof. Inthehorizontal plane, convergence zones are annuli centred on the sound source. This phenomenon For surface-duct propagation at high frequencies can also be compared with its skywave (Over the (above around 15 kHz), it is known that the transmis Horizon) radar analogue by calculating the ratios of sion loss tends to increase with wind-speed even at horizontal skip distance to vertical "reflection"dis short ranges where a direct ray can reach the receiver tance. The results, which are surprisingly close, are as without reflecting off the sea surface. The reason for follows: this effect is not well understood, but since it is mainly noticeable at high frequencies, it is believed to be due (a) for convergence zone sonar, convergence range/ to scattering in the neighbourhood of the sea-surface refractiondepth=60/5=12;and [Hall,1980]. (b) for skywave radar, skip distance/reflection height The main environmental parameters that determine =1500/100=15. transmission loss in a surface-duct are surface-rough This approximate agreement is only a coincidence, ness (and wind-speed), the mixed-layer thickness and, however, since the sonar ratio is determined by the when it occurs, near-surface solar heating. variation in the acoustic refractive index of the ocean, Near-surface Solar Heating whereas the radar ratio is determined by the angles at which radar beams are strongly reflected by the iono Transient thermoclines, or small temperature rises, sphere. develop near the ocean surface when solar heating is high and the wind-speed is low. This "afternoon effect" E. Shallow-water Acoustic Propagation (as the Navy terms it) causes rays from a shallow sonar The term "shallow-water" usually refers to regions to pass through the surface-duct rather than to be over the continental shelf, where the depth of the sea trapped by it. A one-dimensional thermal and mechani floor increases to around 200 m. In general, shallow cal diffusion model has recently been developed [Hill, water propagation is not as predictable as for the 1983]. deep-water case, mainly for the reason that the sound The one-dimensional model has been applied to speed profile in the seafloor, which is now the domi regions in the deep ocean where horizontal variation nant factor, is difficult to predict to the required is unlikely (in practical terms, this means far from precision. boundary currents). The variable calculated was the Other factors that lead to a high variability in probability that "afternoon effect" would occur for a shallow-water propagation are: period of some hours during an afternoon. It was found [Hill. 1983] that afternoon effect is unlikely to occur at (a) the importance of the shape of the sea-floor. Large latitudes higher than 400S (due to the high average scale roughness leads to horizontal refraction of wind speed); whereas it is reasonably likely to occur the sound paths [Kamenyitsky, 1971], while small in regions and seasons where the average wind-speed scale roughness yields bottom scattering and loss is low. of coherence; The assumption of one-dimensionality is not always (b) the importance of the shape of the sound-speed justified, however, since it is liable to be inapplicable profile in the water column. A negative-gradient in frontal regions where there is a gradient in the sea (as occurs in summer) leads to short skip distances surface temperature. In the Tasman Sea, where large between successive bottom reflections; whereas a scale eddies are continually formed, "steep" gradients zero-gradient (typical of winter) leads to long skips of sea surface temperature are commonly found, and, between reflections. Thus the reflectivity of the on the cold side of a front, the vertical temperature sea-floor is a strong determinant under summer profile can exhibit a near-surface temperature rise in conditions, but is less of a limitation under winter addition to any solar heating effect. conditions. The complement of this situation is that sea-surface roughness can be significant Mixed-Layer Thickness (MLT) under winter conditions, but less important when Acoustic transmission improves as the MLT in the sound-speed gradient becomes negative. creases. High winds (storms) generate turbulence that Ingeneral,andforthereasonslistedabove,acoustic increases MLT, whereas the cumulative effect of near propagation losses in shallow water are higher (over surface solar heating and mixing is to generate new the same horizontal range), and detection ranges of (thin) mixed-layers. Work on modelling these processes ~~~::.sare correspondingly less, than occur in deep is currently underway. F. Ambient Noise G. Acoustic Scattering Ambient noise levels in the ocean are consistently In addition to propagation loss, the other environ high because the low attenuation of sound allows mental parameter that determines the performance of sources at large distances to contribute. For example, activesonarsisacousticbackscattering. ships distributed over an ocean basin produce a low For shallow projectors in deep water, the reverbera frequency background noise referred to as traffic noise. tion that occurs is often dueto inhomogeneities within Regions of high shipping such as the Tasman Sea the ocean medium ("volume backscattering"). Back show underwater sound pressure levels that are com scattering by the rough sea surlace is usually unim parabletothoseofabusycitystreet. In regions of portant because the grazing angle of the sound rays to low shipping densities and poor propagation on the the surface is very small (less than 2°), for both other hand, such as the Timor Sea, traffic noise levels surface-duct and convergence-zone propagation (the are 20-30 dB lower [Cato, 1976J. vertical distance travelled by convergence-zone rays is The interaction of the wind and waves at the sea due to refraction by the sound-speed profile rather surface produces high noise levels which persist to than by the angle-of-Iaunch that is important to OTHR). the depths of the ocean, exceeding traffic noise levels A further factor that contributes to sea-surface rough in most areas once wind speeds rise above about 15 knots. Although this is the most important source of ness being less important to sonar than to radar is underwater noise, it is difficult to predict accurately, that wave crests have higher slopes than do wave because a number of complex mechanisms of noise troughs. generation are involved and these are not yet well understood. Some of our recent theoretical work, how Reverberation from a deep sea-floor is often neglig ever, has been successful in accurately predicting low ible since, for a near-horizontal beam, it is due to frequency (less than 10 Hz) surface generated noise backscatteringatverylongranges(dependingonthe and is currently being applied to higher frequencies exact bottom-depth). Bottom reverberations received [Cato, 1983]. To test these theories and to improve our (from a very rough bottom) by sidelobes can be impor understanding of the noise generation mechanisms we tant, however, for some systems in certain range are engaged in sophisticated experiments in laboratory brackets. tanks, reservoirs and at sea. The major scatterers in the ocean medium include Marine animals produce a wide variety of sounds bubbles, gas-filled swim bladders in fishes, and (for the which although intermittent can at times dominate the higher frequencies) the various species of zooplank background noise. In particular, there are regular ton. Bubbles and swimbladders have a resonance choruses at sunset, and to a lesser extent sunrise, when the sounds of countless fish and invertebrates frequency that varies inversely with their average cause the ambient noise to rise by 20 to 30 dB [Cato, radius. 1978b]. The occurrence of these choruses is sufficiently regular to be reliably predicted once the sources and Swimbladders their behaviour and distribution have been established. We are currently working with University biologists in For bladders, the resonance disappears at the small this area. sizes due to friction within the bladder tissues [Hall, 1981]. Fish with swimbladders are distributed through The sounds of large whales are of sufficient intensity out the world oceans, although their population densi to affect sonar performance even though relatively ties tend to depend on re-supply of nutrients. Most small numbers of individuals are involved. The intense species form a "Deep Scattering Layer" (DSL) at clicking sounds of sperm whales, which are plentiful in depths of several hundred metres by day, and rise to deep water, result in choruses of similar level to those the upper layers by night (tofeedonthenear-surlace of fish and invertebrates. However, they are far less plankton). In the Australasian area the lowest levels of predictable because of the nomadic behaviour of these biological (volume) backscattering occur in the Coral whales. Herds of humpback whales produce a variety Sea Basin [Hall, 1973a], while the highest values are of intense sounds during their annual migrations along found near the Chatham Rise south-east of New Zea the Australian coastline [Cato, 1984]. Sonar operators land (an important commercial fishery area). An inter need to be aware of the characteristics of these sounds estinq DSL has been observed near the Equator, over so that when detected they can be recognised lor what the Ceylon Abyssal Plain south of the Bay of Bengal. they are. One of the problems about quantifying these This DSL was 1.7 km deep [Hall, 1971] and from its characteristics is that the sounds change from one year resonance frequency (of 3.5 kHz) the mass of the tolhenext. individual fish is estimated to be around 150 grams. P~)) ~ ~ Bubbles "Cato, D. H. (1978a) "Review of ambient noise in the ocean: non biologicai sources". Bull. Austral. Acoust, Soc. 6, Bubbles scatter sound strongly, especially at their 31-36. resonance frequency, and therefore scatter and attenu ate sound significantly when prevalent. Scientists from Cato, D. H. (1978b) "Marine biological choruses observed In Sydney University and RANRL have obtained photo tropical waters near Australia". J. Acous!. Soc. Am. 64. graphically a reasonable number of measurements of wind-generated bubble size spectra under various wind 'Cato, D. H. (1980) "Some unusual sounds of apparent bio logical origin responsible for sustained background noise conditions. The main results are [Walsh, 1986]: that in the Timor Sea". J. Acoust, Soc. Am. 68, 1056-1060. very few bubbles are observed at wind speeds less than 6 ms-': that above that speed the number of Cato, D.H. (1983)"Theoretical and measured noise from water bubbles increases as W3.4whilethe volume fraction of wave orbital motion" Proc.11th International Congress on air to water increases as W4.' (W = wind-speed); and Acoustics,Paris,July1983,VoI.2,pp449-452. 4 that the bubble-size spectrum has an r slope (r = Cato, D. H. (1984) "Recording humpback whale sounds of bubble radius). Walsh's work was for deep ocean con Stradbrokelsland".lnFocusonSlradbroke,R.J.Cole ditions beyond the N.S.W. continental shelf. These man, J. Covacevich and P. Davie (Editors), Boolarong, results are in good agreement with analogous results Brisbane,pp283-290. already obtained overseas, which also showed that the bubble populations decrease quasi-exponentially with Oibb, P. (1986) "Review of Australia's defence capabilities". depth. Mulhearn [1981] developed mathematical models Report to the Minister for Defence. Aust. Govt. Pub\. Ser to explain the results of acoustic measurements of vice, Canberra. bubbles in coastal waters (off California) at low wind Hall, M. (1971) "Volume backscaltering in the South China speeds as being due to one population of bubbles ~!~.and the Indian Ocean". J. Acous!. Soc. Am: '50,940- attached to small particles in the water and stabilised by surfactant skins, and a second population arising Hall. M. (t973a) "Volume backscaltering in the Tasman Sea, from decaying matter on orin the sea-floor. the Coral Sea and the Indian Ocean". J. Acoust. Soc. Am. 54,473-477. DISCUSSION a As can be seen from the previous chapters, there Hal~~~~~1;;~~~~~~~~~i:~,~.s~~t~~~~sinS~c~i: 51:eF3cO~ are many parameters that are relevant to the perform ance of the various types of sonar systems. For each of these parameters, it is desirable to know its spectrum Hall, M. (1980) "Surface-duct propagation: an evaluation of models of the effects of surface roughness". J. Acous!. (if applicable) and how it varies with position in space, Soc. Am. 67, 803-811- and time (suchasseason,ortimeofday, depending on the parameter). Some parameters are determined by, Hall. M. (1981) "Measurements of acoustic volume back or affected by, other environmental conditions that are scattering in the Indian and Southern Oceans". Aust, J. easily monitored (such as surface roughness, or the Mar. & Freshwater Res. 32, 855-876. sea-surface temperature). Many parameters are affected 'Hall,M. and Quill,A. F. (1983) "Biological sound scaltering in by conditions that are not easily monitored (such as ~~3_~~~aneddy". Aust. J. Mar. & Freshwater Res. 34, size-and depth-distributions of deep-sea fishes, or the detailed structure of the deep-sea floor); and in these cases it is easier to measure the acoustic parameters 'Hall, M. (1984) (Editor) "International Conference, Develop directly (namely, volume backscattering strength and ments in Marine Acoustics December 4-6,1984". Austra acoustic bottom .reflectivity, for the above examples) lian Acoustical Society, Sydney. than to measure the causative parameter and then Hall, M. and Sandy, R. T. (1985) "An investigation into the deduce the acoustic effect. It is fortunate that the effect of the correlation between sea-surface roughness significant parameters that fluctuate in an apparently and duct-thickness on the distribution of acoustic pro random manner (e.g., mixed-layer thickness) generally pagation loss", Technical Memorandum No. 5/85, RAN do so at depths that are not too great. Research Laboratory, Sydney. Like any branch of "mission-oriented" scientific Hill. J. W. (1983) "The Structure and Dynamics of Oceanic research, the future paths that underwater acoustics Transient Thermoclines". Eighth Australasian Fluid Mech will take are particularly difficult to foresee. As has ~~~csConference, Uni. of Newcastle, N.S.W.,pp 4A.13-4A. happened in the past, advances in the science will stimulate technological innovations that will, in turn, Kamenyitsky, A. J. A. (1971) "Some aspects of shallow water generate new and unexpected questions about the soundpropagation".Ph.D.thesis,SchoolofPhysics,Uni marine environment. versity of N.S.w., Sydney. ACKNOWLEDGEMENTS Lawrence, M. W. (198.3)"Modelling of Acoustic Propagation ~i~oSSWarmcore Eddies". J. Acoust, Soc. Am. 73, 474- Dr. D. H. Cato contributed the section on Ambient Noise. Dr. P. J. Mulhearn contributed to the sub section on Bubbles. Lawrence. M. W. (1985) "Ray Theory Modelling Applied to Low Frequency Acoustic Interaction with Horizontally Stratified (Received 23 February, 1987) Ocean Bottoms". J. Acoust. Soc. Am. 78,649-658. Mulhearn, P.J. (1981)"Distribution of Microbubbles in Coastal Waters". J.Geophys.Res.86,No.C7,6429-6434. 'Burgess, A. S. and Kewley, D.J. (1983)"Wind-generated sur face noise source levels in deep water east of Australia". Walsh, A. (1986) "Photographic Measurements of Bubble J.Acous!. Soc. Am. 73, 201-210. Populations from Oceanic Whitecaps". M.Sc.thesis, Dept. of Geology & Geophysics, Univ. of Sydney. Cato, D. H. (1976) "Ambient sea noise in waters near Aus tralia". J. Acoust. Soc. Am. 60, 320-328. 'AdditionaJreferencesnotcitedinfhetext. * * * Predicting the Reactances of Irregularly Shaped Helmholtz Resonators by the Finite Element Method K.P. Byrne and D.W. Kelly School of Mechanical and Industrial Engineering The University of New South Wales P.O. Box 1 Kensington N.S.W. 2033 ABSTRACT: A procedure for using a standard structural finite element program, which can be run on microcomputers, for determining the reactive component of the specific acoustic impedance at the throat inlet of irregUlarly shaped Helmholtz resonators is described. The procedure is applied to three irregularly shaped Helmholtz resonators and the predicted results are compared with experimental measurements. It is found thet there is good agreement. The significance of this work is that experimental measurements have been made to compere with the numerical solutions for a range of irregular cavities. The excellent agreement indicates that the numerical procedures can be used with confidence to predict important properties of Helmholtz resonators. INTRODUCTION Helmholtz resonators frequently occur in acoustic systems and controlpuretonenoise.ltisnoteworthythataccuratepredietion they are often used in noise control equipment. A Helmholtz of the naturalfrequencyofa Helmholtz resonator is difficult resonator essentially consistsofa cavity and a throat as shown when the throat length is small compared with the other throat in Figure 1. The gas in the cavity is compliant and acts as a dimensions because the effective throat length is dominated spring. The gas in and near the throat acts as a mass. At the by end corrections associated with the "attached masses" at natural frequency of the spring-mass system the "slug" of gas the ends of the throat. associatedwiththethroatcanbeexcited,byexternalacoustic The finite element technique described in this paper was pressure fluctuations of this frequency, into large amplitude used to predict the reaetance at the inlets to the throats of oscillations. The frictional losses associated with these three irregUlarly shaped Helmholtz resonators. The reactance oscillations allow significant dissipation of acoustic energy at of these resonators was measured by attaching them to the this frequency to occur. end of an impedance tube. The predicted end measured Helmholtz resonators are sometimes incorporated into reactances are then compared. mufflers to enhance low frequency performance. Another common application is associated with perforated facings which Throat are spaced from hard backing surfaces. Each hole of the facing can be considered to be the throat of a Helmholtz resonator. Acousticians have been interested in Helmholtz resonators for many years and a substantial body of literature relating to these devices has been generated. A notable contribution to this literature was made bylngard [1Jwhoconsideredcavities of regular shape. However, Helmholtz resonators must often be of irregular shape so that they will fit into an available space. The performance of such irregularly shaped resonators cannot be reliably predicted by use of results such as those given by Ingard. The purpose of this paper is to show that a standard structural finite element program, which can be run on a FigureI: GeneralHelmholtzResonator desktop computer, can be modified to determine, as a function of frequency, the reaetive component of the acoustic impedance at the inlet to the throat of the resonator. Once the frequency FORMULATION OF THE PROBLEM dependent reactance is known, useful quantities such as the Consider the arbitrary Helmholtz resonator shown in Figure 1. natural frequency and the bandwidth of the resonator when The required quantity, the specific acoustic reactance at the coupled toa specified system can be determined. An accurate inlet to the throat of the resonator, is given by the ratio of the knowledge of the natural frequencyofa Helmholtz resonator magnitude of the acoustic pressure at the inlet to the throat to is of importance as Helmholtz resonators are often used to the magnitude of the particle velocity at, and normal to, the throat inlet when the gas in the system is oscillating sinu Hence, in summary, a structural finite element program can soidally. The acoustic pressure, p, in the gas must satisfy the be used for acoustical purposes by taking the following steps: acoustic wave equation and the boundary conditions. The 1. At all nodes all degrees of freedom except that associated acoustic wave equation, intermsoftheacousticpressure.p, withuxl-p)aresettozero. is given by Equation (1). 2. At nodes where the acoustic pressure is zero, Ux(apl is set V'p·= (1/c'Ha'p/ilt') to zero. w~~:~~:a~~see~e~~~~~e~u:i~~~'EqUation12),can be used to 3'"2: :Ia~~:I~:e:~r~~~~::~~:~io;~6~a~~defined by specify the effect of a rigid boundary. It is also useful in 4. ThestructuraldensitYQsisreplacedbyl/c'. ~~~~~~ningthe particle velocity once the pressure field is 5. At nodes on rigid boundaries no action is taken. Once the proceeding steps have been taken the eigenvalues (2) and eigenvectors of the acoustical system can be determined using standard algorithms in the finite element package. The Q is the gas density and uis·the particle velocity, a vector acoustic pressures p, (au x) can be found and these can be quantity. used to determine quantities such as the particle accelerations and velocities by use of the numerical approximation to THE STRUCTURAL-ACOUSTICAL ANALOGY Equation (2), the linearised Euler equation. Equation (1), subject to boundary conditions which can be The analogy which has just been described is not new. It is. defined by Equation (2), can be solved by a structural finite for example. given in [21and [41.The finite element method element program which incorporates regular linear elastic ele- has. however, been developed predominantly for structural ments if the following structural-acoustical analogy is invoked. applications and while structural engineers are familiar with its Equation (1) can be written in cartesian coordinates as use, the same appears not to be true for designers of acoustical systems. (alaxHaplaxl+ (alayHap/ayl+ lalazHaplaz) = 11/c'lla'p/ilt') We also note that the finite element method can be used to (3) model the acoustic Equation (1) directly. This would leadtoa specialised acoustic code which would avoid the need for The x direction equation of motion in an elastic solid of implementing the analogy described in this section. densityQ scanbewrittenasEquationI4): Using the Finite Element Method to Determine laaxxlax)+ (aTxylayl + (aTxzlaz) = esla'ux/ilt') 14) The Reactance of Helmholtz Resonators The Helmholtz resonators considered had the shapes shown a , Txy and.T are the stress components and is the xx xz u, in Figure 2. The dimensions shown are internal dimensions. displacement in the x direction. It is evident that Equations (3I The throat width for all resonators was 20 mm and the internal and (4)are analogous if dividerwas3mmthick. They are reasonably regular because theYCOUldbereadilymanufacturedthis~;t axx " aplax, Txya aplay. Txza ap/az, esa 1/c'anduxa p, (5) The structural-acoustical analogy can be completed by the following steps. If the structural displacement components in the yand zdirections, uyand uzare made zero. the stress strain relationships are given by Equation 16) and the strain displacement relationships are given by Equation (7) ?1l'~ axx } _ [C"'" Cxy,cxz] t~xx Txy - Cyy Cyz 'Yxy} (6) t Txz SYM Czz • 'Yxz ~xx= auxlax, 'Yxy= auxlay, 'Yxz= auxlaz Thus if Cxx = Cyy = Czz = 1 and Cxy = Cxz = Cyz = 0 and u, = P then the structural-acoustical analogy relationships of Equation (51will be satisfied. ~[::D'o; At rigid boundaries. the particle velocity normal to the boundary is zero and so by Equation (2) nxlaplax) + nylaplay)+ nzlap/az) = 0 pl~SO nx,nyandnzarethedirectioncosinesofthelocalnormaltot~~ boundary. Thus by the structural-acoustical analogy Equation 7 19)becomes ~/229 110) /~ This equation requires that the structural analogy of an ts acoustically rigid boundary is a traction free boundary. Rgure2:Resonetorerrengements lt is posslble, becauseof the prismatic shapeof the resonators A modal analysis procedure [2] was then used to determine shown in Figure 2,to obtain the required results by analysis of the response of the systemtoa driving sinusoidal acceleration a two dimensional finite element model such as that shown in at the open end of the long pipe. Ught modal damping ensured Figure 3. However, there would be no problem in implementing that numerical instabilities did not occur at the natural a three dimensional finite element model if this was necessary. frequencies of the system. At each frequency the amplitude of It.can be seen with reference to Figure 3 that the cavity is the pressure and velocity were determined as averagesacross divided into rectangular subregions ("finite elements") the inlet to the throat and the ratio taken and plotted on connecting "nodes"defined at the corners of each element. Figures4,5and6. The acoustic pressure is assumed to vary linearly on each The finite element approximation is known to converge with element between values at the nodes. The values of the refinement of the model by reducing the sizeof the "elements" acoustic presssure at each node then become the unknown used. As a final check a refined model, in which each element parameters which must be determined. The elements can have shown in Figure 3 was divided uniformly into four elements to an arbitrary quadrilateral shape but the use of standard mesh 1000,------,--~---.------rr.,-----,------,------, generators leads to the rectangular shape conforming to the geometry of the cavity shown in the figure. The long pipe attached toa resonator provides the driving acoustical system with a broad range of frequencies both above and below the natural resonancefrequencies of the resonatorwhen connected to an infinite half space. The finite element mesh shown in Figure 3 was typical of that used to model the cavity and the throat. The mesh was refined near the entrances to the throat to give an accurate representation of the added mass which extends the "slug" of air shown in Figure 1 outside the throat. The mesh contained 522 nodes defining 455 elements. The finite element program used was the STRAND finite element system [3] implemented on an NEC/APC III personal computer. This program is also available on equivalent IBM and SPERRYmicrocomputers.The subspacealgorithm provided E with the software was used to determine the lowest 15 resonant frequencies of the combined system. The lowest and highestofthosefrequenciesfortheresonatorB,forexample, were 40 and 950 Hertz compared to the lowest natural frequency of the resonatorof about 200 Hertz. 100 200 300 400 500600 700 FrequencylHzl Rgura4:Predictadl-)anameasuredf o)throatreacIlInCeS ~ for resonstor A MeasuredresistancefKJ. ~ == - I---- ~ I---- ...... E E I--- I---- - RJlJl '··es L1Jll::r""'"eteme nts I Open end __ ~BO~mm~_ 100 200 300 400 500 600700 Frequency(Hz) Figure 5: Predicted 1-) end measured fo) throatreacrences Rgure3:Typicalfinitaelementmodel for resonetor B. MeasuradresisrencefKJ. F' ...... cy CHZ) Rfl thaOllginal,~ ,indielltingthatanaocuratllaolvtionhad"""" ' oo tainad on the '0lI'-' mash and tkat, in fa". a aimpl er mesh " 'uld probablyh awlb ""n"",,,d Thellgr ~.The aig"'fi~ nc made. ~ .natur al frequency of the resonetol's !Where the (l iculation of each point plotted o n Figu,es 4,5 end 6,equ ires rlllctlmceis_ollndthe ' elCUlnCe CUrve8CfOt$lIle horizoo1ta1 lin inIIxpensiltl luper-pos;tion end the time requ" ed lldi lectly axislis predietel'l"IIIVoocural.ly bythen umerice l "",thod, prQPOrtionaI to the num bers of poInll ploned . Th. eXllnded tOllltlt" with the llIopa of thll reaetsnce line at th il PQint. ~rnet05Ol""fortheeill'lfWel"""end.nvllC1Ollilofrlttle These two ite...... 'e thll critiCllldata in the des illn of etficiant COlllequence becaUM llOI'IIl of the aignific ~udeof theimpedencebecornes"IIIVlarlle. witho R~ '2Augll.fr'986J CONCLUDIN G REM ARKS I tn this PlIper we haY componenl 01 the acoustic imptollance of irT~ ,1v"""ped JASA, Val. 26. No. 6. 1953. pp. 1037·1061 . ~ and Helmholtz tIIoQI'loa1()Q. A procedI.a"lI lor the experimeotal 2 Coo".. R.D. "Conce Appr;celion. 01 Rnile EI""""" detllmlina lionof lIle Sllme charllCtoriaticsfor reaona tOlflM Ana/yM- Second Edrtion··. Wl..,. 1981. al50 been d elCtibad lind lIpplied to three specifi c d eaigN with 3 $T1NISTRANO/STOUT U- Guide ~ .MacNeeI aMS CINASTRM Theorlllic:alM.>uoI lIipect ratio parallel to thll t!lI"a.t varyinogffom 0.6 end 2.0. Cor~ltion.LoaAngelel. APPEN DIX A EXPERIMENTA L DETERMINA TION OFT HE REACTANCE AT A RESONATOR THROAT INlET ThIolP"C'fc-...ic~ ~theinlel:ol._ throM _..-..dwfth . -.ding _lUbe. The~ .. ~oltlll ..... --... in Flgur. A..l. A ling" ' eIOnMOf. 229 wide._~to _lI'ldol.plIaic pipotwiUl .. m.- cr_of 229rrwnl"d.'-vlhol 6OOOmm. ThIo..-«lf 1hroel_1oclItId 10 thel:lt_ tymmwic; with rap«I 10 till pipe t;rOM MCOon. A hom ~_1ocIled~theot'*.-ldofthepipl . Aprobe m icrop/lonl . whoM poMion wfth 'MP'Ct to the f_ of the .MOtWtOIc:or,jdbe -...... l.-IlO __ .. _ pr_.in tllltubl At~....nlch ...... ltM hwt Iheory ._~ . TNCOf'I"(IloP;....-.;a..oI ... inc:it*1l-.cl""""'=tll(l _ intt.pipe ... --. inF9nA.1 The ...... -...:I ... ~withtt. incidInt.... ~_M::CII:;U'C d, tortt.hoc:l:lIwlh_o.c.yin"'(fnctionol~ FigoIwA. _,.~_tI A/tI-ouoh.OI«~oltt.lIl"dlrofa.pipe...-.gth . d _ ,to~ the ... « 1101twl.~... .. '.i1 SinceP;...:I P,;-. P,.. P""".r.~can be .tt.o;q d"'1Oh...." ofh_ . rn.apecificKOUllil:~~thelnl"'~ - ~ ltM...:IoIhplpelRd ...... -t_ .i';.anbe....,;m" P,Jp;. ReIl> lA.2I • • h 'lti2oft--comple>:~oftt."""'1I Ihilpoint P1t. P,l!ltheOllfTllllh repr_iDnollhflwlocilV " .. P.IP 6 .. - .. lA-3I .th ilpaint (P.- P,.II"e . T...... Eq,uMionlA-lIt1nbe writt.... 1.-.:I .. ~ ..::o.miclmpelHt'q.i;lfIC.tIn ThuI ..... fIOffT\riIed ;: "' "e ICP; + P;1IlP;- P,.II bewrinenas i'::_· i ~ -- ~,,,.,,.,...'''-'''''' "" - - The normilliMd ...-r;.".tnd .-:tMl~ comjlOfl ,. I" e _ !1 _ R"lIIl + R'_ 2Ree. 81 lA.61 :$~ Fir/ The prececl i"" QUBfl{i1'" ,... .. to the tM:e of tt. ' _ lIl". n-",y can btl ludl iy .N1ed to lhe In""of the .1IlOO8tOI throat by multiplyi"" I,,*m by the ,Itio 01 the .. .. of tht pipe aoss-MCtion to the ... of I'" 'MOI\lfIOI throet. It is ""id_t lvt R " Ill 6 mUIt b. found . Thao qUlrltltleo can be found from thI ...... ,... 11 mIde with the probe microphonltullhownlnlhe loilowWtg The com plu ,,,,,,... ntatlon 01 the 1OOIl. llc prll' '''' II • pointat.dllltanc:edl'oml"'lnlttr1ae-~lha P~"'"• (P;.... f"'leiOd... p,.,...- .-foII";": lA-71 p.c:anbe,~ onthe~plarlII.lIhownin Fig",. A.:l It e.l be .-. ~du.t u d _ _ p...... -V . A repr-.tative plot 01 P~ _II-.In f"'lglQA. 4. It can be S8Ilnfrom F'llure A.3l hiltlheminimurn v.lue of P~ 5;nce l'l _ P,JP~d,_ dl + IJ2ln d ad« 1.thU lfllC8ding which is closast to the face of the ,esoroator occurs when eqUll~ can be written in the k1110wingform bV use of the mele is iI djffe..mee of " betw8WI the arllumena of the ,null"' '' 1 ... 9 9'(2 ...... whid> for 9 < < 1. can be complex ,ep,e sema tOn of the iocidenl and "'f1I1C1ec1WIVes. approximaled as 1 /I Supposethatthis valueofd iBdenotedd,.Thu a l ... 6 - Sincek .. ~"f).,lhiseQUationClnbeWfitlen8$ SubJlitution 01 thil 'esult ;n Equation (A.13) Vlelds th" fol!owinll eQUlllion for R. Thus bV me.asu,ing d" the di\ltanc" of the li,st minimum f,om the tece of the resc eeter rt ia possible to detarmine 6 R _ 11 - {til ... ~d,('f- ~:rln.. l IJl :i~ ~ needed in Equlti ons !A.51and !A.Gl. It now remains 1I"'!of, ...~d,I.J-,- ~2m.11lA.IG) Suppose Ihal tha 'alio6 of the maximum p'e nu ' e to th" p'lffl.Sures at the fifSIend seeond minima 8,errl8llau,ad. II can pr...... be soon from Figu,e A.3 that the maximum ' e will be .J-l and ~2 cen be conVfln;"nttv lound bV m""su'. in~the approximatel" PI ... P" The fi,st minimum. which occurs al dec ibel dilfarellCfl belWeoo the maximum and thoem,n,ma. a~d, d" willbe P, - P,"-' wMe the second minimum, which This value of R, along wilh ltw1value of 6 found from Equation -P ,e~.Supposethatlhe,etiO$ occursatd2,wil'beP,ead, (A,101 can be used in Equations !A.5). nd !A.G) to find the of the maximum to these minima arew,itten as reQuired ra, ultt 0 1I.J-, - IP, ... P,II(Pj""'" - Pr"""""1 (A,11l PEACE ENGINEERING " as speci alised excluaivelyinll'.elield of""ll inee'ed ""iseconlrolp.roduc"'o,ovet'5v ea, . ~ avnt Inlhaltim e we ".... ranged produclalOeatefro...-ltypud flOisel\flXJlllffla. W.... 'aadVlO$l\ar.tlWknowJadll'l""""VOU·A~ ealllOourotlIOltCf_douraoanltl .... putVOUintouehwilh aomeonewho .... dlSWSlIyo<.-probOtrnand~naoossilrvPro _ an oblilJalionf_ ._ en QUOIa1lon CuSlom buill pren endoa u'IlI. ,.~ .... ~ ..... ~~ti:i~\!~:~ '.~; :~:ola~OC::'b: _ ai' ..- . aoouat>elouvfea - ro maner ...... , yo<.-r8ql.li.. manta. we have the p'od ut l. _ tt(:Oma,lOl1OiHc:onlrol TECHNICAL NOTES-- A doctor's sound dictionary by relating to seismic data results obtained for wave propagation and scattering at a welded quarter space. Ever wondered what your doctor's listening to when AII.usersof ultrasonics have an interest in transducers, heplacesthatcold,shinystethoscopeonyourchest? A University of Queensland researcher has the answers. ~~:~~r~t~d~fa~u~~::jS~~~~j~rE~~~I~~en~r~e~~s~:~~~st~~ He has just developed the world's first dictionary of the near field of an ultrasonic transducer particularly the different chest sounds a doctor hears. appropriate to a meeting of this kind. P. J. King Professor Charles Mitchell compiled the audio-visual (University of Nottingham) reported low temperature tape after he found difficulty teaching his students the ultrasonic attenuation evidence of three distinct loss multitude of sounds in patients' chests. phenomena associated with vanadium centres in doped "It's extremely hard to teach students and even GaAs. A. McKie (University of Hull) described laser doctors the different sounds. You obviously can't listen generation and detection of ultrasound and a system in at the same time," said Professor Mitchell, from which IS now being used for making ultrasonic the university's department of medicine. measurements of soiids at high temperatures. The But now Professor Mitchell, with the aid of computer potential of this system was demonstrated by detailed graphics, film footage and real-life sounds, can explain measurements of the temperature dependence of the what it's like listening to a patient with asthma, bron velocity of sound in mild steel at temperatures in excess of BDDoC. ~~~~s~/neumonia,fluid accumulation and even lung .The meeting was attended by about 3D people and The more basic sounds, from a normal pair of lungs Will be remembered for both the high quality of the to crackles, wheezes and gurglings, have also been contributions and the lively constructive discussions immortalised on tape. that followed. The Group A.G.M. was also held during "It has always been easier in training to see what the day. something looks like, to show them a picture or let (D. P. Almond in Physics BUlletin, Feb. 19B7) them feel alumpin breast cancer or whatever," Profes sor Mitchell said. "Now they will not only know what Sounding out fractures particular diseases and conditions sound like but what A fractured bone has a lower resonant frequency they look like, for example when the bronchial tubes lhananunfracturedbone,becausetheareaaround vibrate together to produce wheezing." the fracture is less rigid. This principie has been used The video tape, which will assist medical students by a team in the Medical Electronics Laboratory at the University of Kent to develop a technique for bone ~~~ctl~~~efsa~n~ft~~~a~r~S:~~i~~li~~~n:~ga~O~xr:~~~a~fracture assessment. of Professor Mitchell's work in developing a system of The method introduces into the bone a random noise classifying lung sounds to assist doctors in diagnosing at a very low level and the resulting movement is medical conditions and establishing their severity. detected with an accelerometer. The output from the (Chris Thomas in the Sydney Morning Herald, 14 Feb. 19B7) accelerometer is then analysed to give the frequency response of the bone, in both amplitude and phase. Current research in ultrasonics and Breaks in the amplitude response occur at the resonant frequencies. physical acoustics As weil as detecting fractures, the method gives an Thisnow traditional annual meeting of the Physical electronic measurement of bone healing-something Acoustics Group of the Institute of Physics was held at that conventional fracture detection methods cannot do. University College, London. It has been the custom for This is because the low resonant frequency of the one of the centres active in various areas of physical broken bone slowly rises as the fracture heals, until it acoustics to host the meeting and to provide a major gives a frequency recording similar to that of an part of the programme; hence half of the contributions untracturad bone. For instance, an unbroken shin bone were from workers at University College. has a resonant frequency of about 73D Hz. Several papers were on various aspects of ultrasonic The technique, which has been tested on patients wave scattering which are of importance in the develop from the Kent and Canterbury Hospital and the London ment of nondestructive evaluation techniques. P. Smith Hospital, has been patented by the university's indus (U.C.L.) discussed the problem of using ultrasonic trial liaison company. scattering data to deduce the nature of the scattering A simpler version of the technique has also been object in a medium which is viscoelastic. L. J. Bond developed to measure the rigidity of the fracture. This (U.C.L.) reported a recent advance in overcoming the is similar to the manual sensing method currently used, zero-of-time problem which has led to some controversy in which the limb is considered un-united if the con over the application of 1-0 Born inversion techniques sultant can flex the fractured bone. However this relies for sizing defects. M. Punjani and L. J. Bond (U.C.L.) on the consultant's sensitivity to bone movement inside discussed the phenomena which occur when an ultra both skin and muscle; the new technique gives elec sonic wave is scattered by a partially closed crack. tronicobjectivity. The natural occurrence of these cracks in solids poses A small vibration is introduced to the bone, then a major obstacle to reliance on ultrasonic techniques detected by an accelerometer, amplified and again for crack sizing and detection. M. Plant (U.C.L.) repre introduced to the bone. Such a system oscillates at a sented the college's close involvement in the develop frequency determined by the phase shift round the ment of the acoustic microscope, with a new analysis of system and, in particular, by the phase shift through the propagation of surface waves at the interface the fracture. A non-rigid fracture has a iarge phase between the coupling fiuid and the sample being shift and asa result the oscillation frequency drops. investigated. Again, this measurement can be used to monitor bone E. Aristodemou (British Geological Survey, Edin healing,astheoscillationfrequencyshouldreturnto burgh) demonstrated that seismologists and those that of the surrounding unfractured bone. engaged in ultrasonic testing have much in common (Physics BUlletin, Feb. 1987) TECHNlCAL NOTES. A computerised Bosendorfer? The famous Viennese piano manufacturers Bosen Time-warped speech dorfer have come up with the ultimate way in which you In the 1960s most researchers assumed that speech can have, say, Vladimir Ashkenazy or Andre Previn recognition was simply a matter of distinguishing the playing in your own living room. Like recordings, they "shape" of each "phoneme" (syllable or consonant would not actually be there. But the sound you would group) and translating that into words. But that hear would be every bit as "live" as the moment the aproach has proved unrewarding because it under performers originally played the notes. Or any pianist estimates the variability and ambiguity of speech. could be heard "playing" at a concert when they were Compare "This new display can recognise speech" on the other side of the world. with "The nudist play can wreck a nice beach". The piano company is trumpeting the arrival of its Today a different mood prevails. I.B.M.'s Dr. Fred new computer piano system - the Bosendorfer 290SE Jelinek jokes that his system improves every time he - as a breakthrough in piano reproduction that will gets rid of an "expert". What he means is that given revolutionise the business. The basis of the system is lots of data, computers are better at deducing what to an optical device that scans the keyboard action 800 measure so as to distinguish words than humans are. times a second as the piano is being played and stores At its simplest this means measuring the statistical the information digitally on an audio cassette. By similarity between a stored template (of a word usually) playing back the cassette, the keys are activated in and the sound that has been heard. exactiy the manner of the pianist with all the weight, But it is never as easy as that. For a start, words vary speed and style used in the original performance. in length according to the speed at which they are spoken and according to their context. They have to be A pianist playing in one concert hall could be heard "time-warped" to a standard length. But it does not simultaneously elsewhere as the signals could be help to time-warp them by a set amount. Say the word passed from one Bosendorfer 290SE to another. "This "three" slowly and it is the "ee" that gets lengthened, is vastly in advance of the old piano-roll system which not the "thr". The answer is dynamic time-warping, a reached a peak of popularity in the 1920s before the mathematical trick that matches two spectrograms of advent of records," saidsone of the Bosendorfer uneven length. experts who has been engaged on the project for the But if you try hard enough you can dynamically past 14 years. "But this piano is unlikely to be bought time-warp one word into almost any other. The time by many people for enjoyment in their front rooms. It warping has to be constrained. The cleverest way of costs about $140,000 and is a serious tool likely to be doing this leads to a whole new approach to speech of most use to music colleges and recording studios." recognition called "hidden Markov modelling", after a Students could use it to compare one performance Russian mathematician who analysed the opera, with another and it would be possible to listen to per Eugene Onegin.1t was first applied to speech recogni formances of the great maestros of the day if they tion by Mr. Jim Baker, now the chief executive officer could be persuaded to play programs for the tape of Dragon Systems. system. "Tapes could be reproduced and distributed to It gets away from the idea of comparing word tem all purchasers of the computer piano system," the plates, comparing instead tiny fragments of speech spokesman said. with stored patterns and, in particular, the probability that one fragment will be preceded and followed by The equipment can easily be used to make correc another. It is "hidden" because the answer it gives for tions to imperfections that always crop up in perform each sound is itself statistical and based on the com ance, so Bosendorfer expect to sell 290SEs to top-line puter's own ability to learn from examples. record companies. "All it takes it a computer terminal The statistical approach stumbles over short words, and it would bean enormous saving in time and labour not long ones, which include more distinctive features. to make corrections and alterations before the record ing engineers start work on a tape," the piano-man ::~~~e:~~b!!~:'~~~ti~rii~n~~~;ei~hee~~~~~i~~i~nr~;~~'c~~~said. Eventually, Bosendorfer hope to bring the equip in. "In America" is a more likely phrase than "it ment to Australia. America", but "if America" and "is America" are both (The Australian, 11 Nov. 1986) plausible. No single approach, acoustic or linguistic, is as good as their combined efforts. (The Australian, 18 Nov. 1986) Bangkok also has problems Information for Contributors Discotheques violating city noise and light levels face Articles for publication normally occupy 4-5 printed closure from next month, Environmental Health Division pages (approximately 4 pp. double-spaced typing per Director Voravit Lebnak said yesterday. A Bangkok page). Authors may be asked to pay additional type Metropolitan Administration ordinance issued in 1986 setting charges for pages in excess of 5. Frequent set the maximum noise level at 90 decibels and banned headings and sub-headings are desirable and an laser beams. abstract of approximately 200 words should be The ordinance was issued after the B.M.A. organised inoluded. Reprints may be ordered, preferably prior to a meeting between 50 discotheque operators and printing (they are then cheaper). officials from the division and Public Health Ministry. So far, 39 discotheques have been allowed to operate Diagrams will normally be reduced to single column because they comply with the regulations. width: authors are requested to plan diagram propor From next month, said Dr. Voravit, B.M.A. officials tions and letter size accordingly. Full stand-alone will inspect the discotheques. Violators will get a warn captions should be provided for each diagram (these. ing for a first offence and subsequently face 15-dayor will be typeset). indefinite closure. Dr. Voravit said people who are Types of articles accepted include technical, tutorial subject to excessive noise once or twice a week are and review. Short reports (1 page printed) on current prone to hard hearing. He said laser beams for illumi research or a group's activities are welcome, as are nation can cause cell coaqulation or inflammation, shorter notes for inclusion under Technical Notes. which may lead to blindness. (Bangkok Post, 17 Oct. 1986) ....Ih o"' ''l c 1<:0" lor aequenti. , oct ....e IyIII• ....,."tornati c .. "IIiC'lg It • •• ..., ~lIbt"l!d,. nd lhere II. NEW PRODUCTS CIlOIce of I...... tNerophollI -.itMbn .nd 2OIlV. 28V."" O V",iCroP_ po!a rl... ,_YOl". CIRR US P\lte< . nd pIoIlld. or the uni t c... be A eou fld inalioon can be l/Wdo d,.p1ay. tor.o;on. nllnelO fldl dOnI l,\e I" ln ...._ _ in .... 1ie1d S ELF CAUBAATIHG SUI tillUOUl li octlWllPICI ... Pri n- .. CImIS fine _ II _ ~ted and l.. rQ ::::~;::;:~:.h;:c~el~."D~:~~.~~".~~: ~::~S~o:'hn:Of!I~: ~~i •• worlll who hlMl diMe..1l . COUlllc. 1 e nvlro..me..ts likely Finally, lhe CAL 2.21C II deli vered 10 llfI ... cou..I.. " d I.. Ihe lield com plete ...ilh . tuooe ~ . u lll l. II .... IBIo.Iand AwIe TheCl.lion.ndj)lo~fo.tne ~lIoool"PI01DOO"' "Ilas unl'-- In ...... ,.,d lor .... 60 not • • coIlen t Df>31 int erfl c. ll joltl tty ao· _ b& " _id ..Idlonolc: ..... lC1IrialiaCMI ' the ...... -""lice"'" in l,"" ",," II tt:lO t'Wee'nCiIf\lI Re,.a""h ,ttoeonglnIIOf . .... ACO RN; CirNa In.... _ Me ,. 1uI1·~'lJllIllol'O.l Soeu,-Anne SA (f'._1 Inc Ouest 1$0 kHz. I.. pattie ...... 6t05 is qulf'lOdIg Inle EEE .88 ~ .allel l e... ." 1,, limed by lhe comP" '''. Th" • . . nd h... .. I ..,ert ace ... Ie'II 'l'Ieynolnnge r have ln bl by 1011>0.... In~ pe'lod L&q 10"11" tha n . ..do!l&l< ee .I. l inlerl ac:•. TI>a Type "' _ " lhe " Clnry u.. b."Ie," "",n tc. nbe ".I " "l. 'o d «33 io p. rtlcula.ly .n y to Opa' ". AFTER Ihe lYe nl. All Ih. ull. hu 10 do ilac: q" lred.tl bli..dlv l..to memory. The Ci. ... . plOll' . "'''' e l ".n ''' (I"i.. d.l l MA TCHEO MICROPHONE PAIR ei""" 1'0"" co",p uler Sou" d l . vel B.... I &. Kja,, · S"'eo Mic.o- Mol'... lor .... "'pl . Iht CRl 2,36. pho "" Set s 3529 d 3530 con !lisl 01 CRl 2.39A, SA 1G-1l O. "' ic:.o 15 oe two e• •el" li)' "'otc "'" p.l rs 01 o.....i ' ro'" I n ..... 100... ll\n lor ", p, cld .pa,, " (A-8) as lhe CRL 2.22. CRl 2.37AJ 3 o ny •• " 1'00 .ec:o'diflQ. The .ha.p . cle.r OP 15 equ,PPI'CI " ..,l .....1 ai, p.o- lIereo Im' lIII .. 1'11<:11is gained by Ihls 1I' "' ..... Id 'lell on d iff.. entc:ocn- .ac:o' d,"9 Ihod depencts e.. ,,10M P\lt anc:ld',.. t in.t ...... ntsbu t"'" "'aICl\ ing ol t indM d... ICfOllI>or>n PfOll· ·CO · . lhe _ t"' ll to'IMf'I"I'¥IIy. j;me .. "'I.~ ,", Tile CAL 230 A Rail T_ "".tysoor por>o.e . Brue l &. I(). ", '. matc:1Ie .ias;~~~~~~o':a~a~c ~~~~ d ~~~r~:;~ port~ to therma l printers, ca sse lle BookReviews The best pon ibl. s1e,ao Irnalle h.as memory units, <;<>mpule' sySle m. elC. bee n .nsu ,e d by matChlno Ula miC,O 10' gen e,a llon 01 hard copy ,e po.ts In BUILDINGS FOR MUSIC phone ' In lhoe 3529 and 3530 ~il, to a wide vafiery ot tormals (e.g . time .. ilhin ld B in amp lil...... ,esponu OVer hlslo. ies: excee cence level. listings, Michael Forsylh Ine e ntlrel, eq uencyra1l98 20HZI O elc). Optlon,ln ciud e Inllxpe nslva appli_ 20 kHz ar'ld .. ithln 10 d8{l,ee , In phase cat lonsofl .. ar"pac kage ',I ',equlred ove' the ' ange !;(1Hz Ie 2OktU. ma nulacl"r ad by L• •son Davl. Labo,a_ ~~~:~!%jrorJ~~3~r: ~;.~slii,Ind~, e 'n add itioo to , la ndard acces , o,ie. to,les U,S.A. and e.clu s lvely repre ::n~Y aueh as cabl e d ips, .. indsc ,ee n•• and b~~' Yipac Ply , lt d. - Instrumen t Thl. boo~ filis an Impo.la nl ga p In fl>eSO~~J~o~:i.IgI~~ud,;o;:~i~d~n ~i lypes01 p,otectlon grid .. hlch allow thoe Iha Archltactu ,al ecou sncs liler~!t.. e. II I'&quancy ra,ponse o. the microphone I, nol a te' I boo k on room acou st,cs :e t;; . t. iIOre ~ ~s~armonlc :~nc COmponenl, sea'ch Thi, book ls .,. . a nli. lf ord asi gnars an Furllwr Jnlorm aflr:m: AN/TECI/ , 1-5 ~~~a;,,~r~'·'sf~I~lt ~~~~;"'~,ea~v~~~~ ~ "'Oa~a:;'ca~ r'~::'~u'ci:= o~.nf~i: Carr", SIreel, Udcombe, N.S,W, 2141, from ova.... u bUI co, t much mO" alon• .. . llke . n . c' obal .. hecloses (02J 84IJI111. Th. muffs .. III b" "Iljlec l. lly use 'ul 10. his eyes a nd Gllngs to tho 'wes 01 a n ascend ing balloo n". Tnea dage lhat . VIPAC " G ood acoust ics ha,.s much 10 do .. lth ~:~~:.'~ ~:~I:i~~ar~\:~!oi:% ~:ir ho.... " 11you see as willi ho.... ell you =~~SE~DOSIMETER J SOUND LEVER mull. .. ou ld also be usetut In In heappllcation~olmedical ...hlchoov.r<'ll data gel Inlo lhe COmpule r. Dr. C. A.. HIli Bod IIi' collng lill , I ult" so und. ncc metna aw e a llh 01 use Ihe ln. lilute oI Canc e rR es..l,ch , Royl l lui .nd rele vanl Inlormation supple R. W.Harri, l,Ia rsden Hospil . I .' . ...,,1I knO"'n lp r menl ed by an e.t etlslv(t list 01 ref..r· Ih.l r plo",",'i ng C ~~nal'. cusses the bioe lfects whlche.nbaob h." . the .c lua lv alue lndOd lsp layed \~~ are no'" a"aila ble in boun d la ined by ultraso~nd. The biophysics The dl. pla y we. .... lldon " andt h. c ha plerde a lsw ith lhe phe nome na ...hich p romp ll a t tha bollO", which gave key. Further Informalion: Audio Engineer. may OCCur during Ihe propagl tlon 01 " 'o kes 10 c hl nge Ihe dis play we re ing SocM)', 60 Eas t 42nd Su e er. New ultra souod in biologica l ""'d ie. Tha ll ea sy to unders ta nd . On. option wa s Yor~. NYlOf 65·0075, Priee$: Memb&M Include thermal as w.11 as no n·thlf mal I UPpel ad 10 . llow the display to be U.S,$2S, non-m&mbe re U.S.$35. mec han isms. ..flect s due 10 "dlal lon du mp.d to prlnle r bul this did not NEW PUBLICATIONS ... tasks is provided. The information upon characteristics of both thevibra shows how effects are related to char tion environment and the task. Data acteristlcsofthevibration, the display from published experimental studies are The following publications have been and other aspects of the environment. used as the basis of a series of design received by the Society and are held, Published experimental studies are used recommendations which may be used temporarily, in the Acoustics Labora asthebasisofaseriesofdesignre ~onr;;,i~~~~~eta~hk~.influence of vibration tory, School of Physics, University of commendations which may be used to N.S.W.They are available for inspection minimise the influence of vibration on Technical Report No. 136,57pp. or loan by members. Photocopies (not visual tasks. S.J. ElliollandP.A. Nelson In contravention of copyright condi Part II: Manual tasks (Report No. 134), The application of adaptive filtering to tions) may be ordered by contacting 35pp. ::;;'n.activecontrolofsoundandvibra- Cronulla'Secretarial Services on R. W. McLeod and M. J. Griffin (03)527-3173. A charge will be made SUMMARY: Design guidance relevant for photocopying and postage. to the effects of vibration on manual Technical Report No. 137,124pp. L. C. Chow and R. J. Pinnington Acta Acustlca activities is provided. The information describes the mechanisms by which On the prediction of oil layer damping Vol. 11 Nos. 1, 2, 3 (1986) on plates. Archives of Acoustics vibration may affect task performance Vol. 10 No.2 (1985) and shows how effects are dependent I. INCE Newslellers 42, 43, 44 (1986) Applied Acoustics Vol. 19 Nos. 5, 6 (1986) Vol. 20 No.1 (1987) Gibson, A. V. 19 Contents include: A. L. Brown and INDEX Hariharan,P.51 K. C. Lam, Urban noise surveys; T. D. JOhnston,R.67 ~~~:~ngand R. Perrin, Vibrations of Volume 14 1986 Martin, J. 43 Marty, S. 51 Vol. 20 No. 2 (1987) Format: Author, issue, pages, title Mitani, Y. 75 Contents: A. L. Brown and K. C. Lam, Ohta, M. 75 Levels of ambient noise in Hong Kong; A. ARTICLES Oreb, B. F. 51 A. Preis, Intrusive sounds; J. S. Schroeder, M. R. 9 Mirza, Learning English basic sounds Stoneman, S. A. T. 14 BAGOT, H., No.2, 35·41 Troup, G. J. 67 through syllable ulterances; P. R. Kes· Bells, their design and tuning wick and M. P. Norton, A comparison Welsh, M. C. 14 ot modal density measurement tech CALDERSMITH, G., No.2, 47-51 niques. Vibration geometry and radiation fields in acoustic guitars D. INTERVIEW ChlneseJ. of Acousllcs (in English) Vol. 5 No.1 (1986) FLETCHER, N. H., No.3, 71-74 PAUL DUBOUT, No.1, 18 Nonlinearities in musical acoustics AC~~~:rc"ts~~~~C;;iss,%~ng~~~~~ha~rog~GIBBINGS, D. L. H., and GIBSON, A. V., E. LETTERS caused by random surface; Wu shuox- No. 1,19-22 ~~[OPhOnesensitivity calibration at BULTEAU, v., No.1, 5 :~~,e~e~~~~~t~o~~d;~h:~feP~~~~ti~~ Noise: Problems and Remedies buildings adjacent to urban streets; Ll Changli et ai, A research on speech JOHNSTON, R., CLINCH, P. G. and ~~eO~Gr~'~J1~ih~;'~clause quality of channel vocoder; GongXiufen etal,Ultrasonlcinvestigationofnon ~~~~~IeG~/~O~~I3ir:~;6~esonancein FRICKE,F.,No.2,41 linear effect in biological medium; Sha clarinet playing Types of articles published Jiazhengetal,Frequencyandefficiency MARTIN,J., No.2, 43-46 of sound column at resonance. The acoustics of the recorder Vol. 5 No.2 (1986) OHTA, M. and MITANI, Y., No.3, 75·78 Contents include: MaDayou, Contri A practical evaluation method for the ADVERTISER INDEX butions to acoustics in China; Zhu stochastic noise reproduction effect of Weiqing, Evolutionary spectra of non sound absorbing materials Bruel & Kjaer .. statlonaryunderwatersoundreverbera tlon process; Rao Yu'an et al, Normal SCHROEDER, M. R., No.1, 9·12 Chadwick 2 values of human tympanograms; Jiang Computers in acoustics Jlnchang, Structure and function of the STONEMAN, S. A. T. and WELSH, M. C., Industrial Noise Control 6 sounding apparatus in cicada. No. 1,14-17 Vol. 5 No.3 (1986) Experimental studies of acoustic phen Kell & Rigby 2 omena in turbomachinery Contenfs Include: Wang Hongzhang Peace Engineering 26 et ai, Measurement and calculation of electromechanical parameters of piezo B. REPORTS Readers are asked to mention this ceramic thin cylindrical tube transcen publication when replying to adver ders; Tao Duchun, A study on ship BROWN,W., No.3, 66 radiated noise rhythms (II); Dai Genhua Acoustics group - Public Works De tisements. andWangHongyu,Experimentalinvesti partment gatlonon the characteristics of dia DUNLOP, J.I., No.1, 13 meter-expanded silencers. ~~~teu:~~;ontrolof the acoustic impe- Telecom Australia Research Labora New Company faries, P.O. Box 249,Clayton North 3168 MARTY, S., OREB, B. F., HARIHARAN, Ron Carr Associates Pty. LId. has P., No. 2,51-52 been formed to continue and expand ~::~~~:~t~~~~~sa~~8tl:'~~lonAnalysis at guitar top-plate modes the business of Ron Carr and Company Pty. Ltd. The new company is a part University of Southampton nership between Prof. Harold Marshall Technical Reports Nos. 133 and 134 C. AUTHOR INDEX and Christopher Day of Marshall Day Bagot,H.35 ~;:~;~f~~~in New Zealand and Peter ~a~~~P~ai~~di~~~~r~:f~;1e~~~~~~e~~s~Brown. W. 66 Part I: Visual displays (Report No. 133), Caldersmith, G. 47 Marlin Beech-Jones, who worked for 33pp. Clinch, P. G. 67 Ron Carr for three years, will join the M. J. Moseley and M. J. Griffin Dunlop,J.1. 13 new practice, which is located at 22 SUMMARY: Design guidance relevant Fletcher,N.H.71 Tratalgar Road, Camberwell, Victoria to the effects of vibration on visual Gibbings, D. L. H. 19 3124. Phone 8827268. Fax 8826277. 8ENTS FUTUREEI1' ~ • Indicates an Australian Conference . June 23-25, LISBON October 5-9, CZECHOSLOVAKIA 5th FASE CONGRESS NOISE AND ENVIRONMENT 1987 ~~;f~/S:C~;;t- :~.SEB~:~i~abi~~ct;be~~fa~i::1~6,DXfust;~;e~n~bnf~::~~le,S~C;~: Codex, Portugal. May 11-15, INDIANAPOLIS ~~{!v~~ia~1I, 832 27 Bratislava, czecno- MEETING OF ACOUSTICAL SOCIETY July 6-9, LONDON OF AMERICA ULTRASONICS INTERNATIONAL CON· • November 12-13, HOBART FERENCE AND EXHIBITION ACOUSTICS IN THE EIGHTIES. Details: Mrs. B. Goodlriend, A.S.A., 335 Detells: M. vukovoiec, Butterworth East 45th St., New York, NY 10017, fJ~%~~4faff:ne~~adf'n:'~i~:: U.S.A. ~~i7J::::;,L~~;r:yOGegx5~~,Btl Street, ~;i::~~x May 13-15, POLAND • July 8-10, BATHURST November 16-20, MIAMI IV SYMPOSIUM ON HYDROACOUSTICS ANZASCA CONFERENCE Details: Dr. R. Salamon, Institute 01 "Controlling fhe Environment." MEETING OF ACOUSTICAL SOCIETY Telecommunication, Gdansk Technical Details: Ann Heydon, Dept. of Architec OF AMERICA University, ul Majakowskiego 11/12, ture,UniversityotSydney2060. Details: Mrs. B. Goodfrend, A.S.A., 335 80-952 Gdansk,Poland. East 45th St., New York, NY 10017, July, ANTWERP, BELGIUM U.S.A. May 19-21, POLAND 15-25, SUMMER SCHOOL ON INTER • December 1-4, LAUNCESTON INTERNATIONAL CONFERENCE. NAL FRICTION PROCESSES. "Prospects in Modern Acoustics, Edu 27-30, CONFERENCE ON INTERNAL 8th AUSTRALASIAN CONFERENCE ON cation and Development." FRICTION AND ULTRASONIC ATTEN COASTAL AND OCEAN ENGINEERING. Details: Prof. Dr. Sliwinski, UATION IN SOLIDS. A. University ~rt~~~;n~~~~~r~7c~a~~~:Te~ir~:ittU%~r~ of Gdansk, Institute of Experimental Details: R. de Batist, S.C.K. - C.E.N., Boeretang 200, 2400 MOL, Belgium. ton,ACT,2600. ~~:Sics,80 952 Gdansk, Wita Stwosza August 24-28, U.S.S.R. • May 20-27, MELBOURNE 11th INTER. SYMPOSIUM ON NON 1988 MAINTENANCE ENGINEERING CON LINEAR ACOUSTICS FERENCE 1987 • January 25-29, SYDNEY "Effective Maintenance: the road to ~~!~!it~;5,7%~gi;~3~Ji~;bt;~~:~;~:~BICENTENARY CONGRESS OF profit" PHYSICISTS Details: Institution of Engineers, 11 September 8-11, POLAND Details: Dr. Collocott, CSIRO Division National Circuit, Barton, A.C.T. 2600. XXXIV OPEN SEMINAR ON ACOUSTICS of Applied Physics, P.O. Box 218, Lind field 2070. June 1-4, YUGOSLAVIA Details: Inst. of Telecommunication & March 15-17, GERMANY XXXI ETAN CONFERENCE ~~~~~i~YS;~~~~i:9:~~~n50_~;~'wr~~: Details: Prof. P. Pravica, Electrotech DAGA '88 nical Faculty, Bulevar Revoluci]e 73, Details: R. Martin, Abt.l-Mechanik and Belgrade, Yugoslavia 11000. September 13-16, BIRMINGHAM Akustik,BundesaJle 100, D-3300Braun CONFERENCE OF BRITISH SOCIETY schweig. June 8-10, PENNSYLVANIA OF AUDIOLOGY May 16-20, SEATTLE NOISE-CON 87 "High Technology for Noise Control". ;~~~~~~n~n~hr~~~~~~~iO~fe.,~fHearing MEETING OF ACOUSTICAL SOCIETY Details: Conference Secretariat, NOISE- Details: Mr. N. Bland, Centre for the OF AMERICA Details: Mrs. B. Goodfriend, A.S.A., 335 ~7;;~fm:m/,~r7g'o,wJ~.ernRoad, Bir- East 45th St., New York, NY 10017, ~~r}:,~fr~T2ft~;:Y;6r~~gr~~~:nn~U.S.A. September 15-17, CHINA June 6-10, YUGOSLAVIA June 9-11, UMEA, SWEDEN INTER-NOISE 87 XXXII ETAN CONFERENCE 4th INTERNATIONAL MEETING ON LOW Details: Prof. P. Pravica, Electrotechni FREQUENCY NOISE AND VIBRATION. ~~!~t.~!~~r-~7~~~I~tjr:I~~~1~;~ncuncal Faculty, Bulevar Revotuci je 73, YU-ll000 Belgrade. :~~~~~£~~~i{4m!if:.H;~~~:~~~~~~September 22-25, WARWICK August 21-25, STOCKHOLM NEW MATERIALS AND THEIR • June 17-19, BRISBANE APPLICATIONS 5th INTER. CONGRESS ON NOISE AS A PUBLIC HEALTH PROBLEM ?~8~PUTINGSYSTEMS CONFERENCE ~~;~1g;,~~7~~~~~~~~~~;~ti~~:;d~~Details: Noise '88, C/- Reso Congress SWIX 80X, U.K. Service,S-11392Stockholm. £:;f~~~/67r~~~:i~~rtg~,;.~~ei~~o.11 June 19, MADRID ACOUSTICS AND OCEAN BOTTOM ADVERTISING RATES ~:~~iZt~:'T:f/~~~n~7,Calle Serrano, $360 Spec/a/sizes $410 June 22-26, SEATTLE ndll~dY""oracx ano wrute $235 11 cm by 11 em (111page) $175 INTERNATIONAL SYMPOSIUM ON FISHERIES ACOUSTICS m~~~~~th~;~t~.~~d~h,h~glhem high $;~~ $150 Discount rates available for three con- $8 ':::,;~~~~s.ads and for sustaining ~f;'~J;~iJ,?~lf,¥!k~il~ RU5TRRL/R/~RCOUST/CRLSOCIETY> DIVISIONAL COMMITTEES ('Indicates Councillor) New South Wales Victoria South Australla WesternAustra/la Queensland Chairman Serge Hlistunov Stephen Samuels' Peter Swift' Giafranco Rasile' Robert Hooker' Vice-Chairman Ray Piesse' Graeme Harding' Robert Boyce' Peter Wilkins Ronald Rumble Robert Monteith Robert Williamson Ian Bailey ~~:::~~Z;~~~~~::I~~t'Geoff Barnes Ken Martin Graeme Yates ~~I~I~0~1sd~~~ton Graeme Atkins Sin Chan David Bies Brian Bickford Lex Brown Bruce Cremin Marie Kormendy Russ. Brown Peter Knowland ~~:~ie~~~~:I~;=~~:~~~~enserRobert Langford Fritz Kamst Jack Rose Timothy Klar Michael Norton Warren Middleton Colin Tlckell ~~~~d~;~~insJohn Lambert Warren Renew' Andrew Zelnik Max Lane ;il~:~~:~:'Ron Windebank Greg Wild ADDRESSES: All correspondence on NATIONAL matters should be addressed to the General Secretary. All correspondence on REGIONAL matters should be addressed to the appropriate Division Secretary. GenerafSecretary NewSouthWa/es Victoria A.A.S.-GeneraISecretary A.A.S.-N.S.w.Division A.A.S.-VictoriaDivision National Science Centre ~gl~r:r:~~nUt~eet~gi6ra~:~~nU~eet191 Royal Parade Sydney.N.S.W.2000 Sydney,N.S.W.2000 Parkville, Vic. 3052 South Austra/la Western Australia Queens/and I.H.Balley A.A.S.-Qld. Division ~.:a~~~~a~t~on ~X~W.OfPhysics C/- Div. Noise Abatement Hyde Park, 64-70 Mary Street S.A.5061 Kent St. Bentley, W.A. 6102 Brisbane,Qld.4000 SUSTAINING MEMBERS AQ·VIB DIVN. OF AQUA·COOL CMA FOAM GROUP NAP SILENTFLO PTY. LTD. TOWERS PTY. LTD. 1 HEATHCOTE ROAD P.O. BOX 173 UNIT 7, 2 STANTON ROAD LIVERPOOL, N.S.W. 2170 CLAYTON, VIC. 3168 SEVEN HILLS 2147 THE CHADWICK GROUP PEACE ENGINEERING PTY. LTD. ASSOCIATION OF AUSTRALIAN 292 BURNS BAY ROAD UNIT 2/20 MARIGOLD STREET ACOUSTICAL CONSULTANTS LANE COVE 2066 REVESBY, N.S.W. 2212 C/- SCIENCE CENTRE PITSTOCK PTY. LTD. 35 CLARENCE STREET FOAMLITE AUSTRALIA PTY. LTD. ACOUSTICS DIVISION SYDNEY. N.S.W. 2000 P.O. BOX 331 5-15 COTIAM AVENUE SUNNYBANK. OLD. 4109 BANKSTOWN, N.S.W.2200 AUSTRALIAN GYPSUM LTD. INDUSTRIAL NOISE CONTROL G.P.O. BOX 4325 SOUND ATTENUATORS PTY. LTD. MELBOURNE 3001 AUSTRALIA 7 LEVANSWELL ROAD 39 KOORNANG ROAD MOORABBIN, VIC. 3189 SCORESBY 3179 BRADFORD INSULATION, C.S.R. LTD. JAMES HARDIE INDUSTRIES LTD. SOUNDGUARD PTY. LTD. P.O. BOX 1754 G.P.O. BOX 3935 P.O. BOX 246 NORTH SYDNEY 2060 SYDNEY 2001 RYDALMERE, N.S.W. 2116 BRUEL & KJAER (AUST.) PTY. LTD. JOHN LYSAGHT (AUST.) LTD. K. H. STRAMIT LTD. 24 TEPKO ROAD P.O. BOX 77 52 MANDOON ROAD TERREY HILLS. N.S.W. 2084 PORT KEMBLA, N.S.W. 2505 GIRRAWE~N2145gth e besl way 10 eSla t> " om wo rking scie ntists. A limited num ber of ple nary leclures will be p resen ted Iish a n alm' naHve body (Extracte rom dl b~ eminent . Cie nliat• . inclUding Nobel In March 1986 the Noise Council was 1_INCfN ewslelle, #0 . 43) l. unched at a rece pHonl n lhe House ::~luop:inners.eerectee by fhe spec ialist 01 Lords hoaled by Lord Nathan. The aims a nd obj"" tives 01 the Noise Coun The "u slra llen Acousllca l Society is I' ll a ", to pro mote and respon d to issues suppo" ing Ihll Congress lind witl be ", tating lo nolse a"" vib' aflon and to 1987 Annual Subscriptions ~~ °L~er~: ~ei~nlc ~~W6s" : make Inde pe ndent t""hnl ea l and sCie n ::;g~:::o;: d se Me e~pe't iava ilable to Inte rnalional ThO lollowing molion s were pass ed Fu"h er enquirln : Dr. S, J. Collco tl. s"" nallonal age ncies. ce ntral and local el t he 16th Annual Ge nerlll Meellng C.S.I.fl.O, D,vision of Applied Physics . men~ gove,n CQmmerc e and ind ustry. or the SocielY held o n 2 Octo be' . 1986 P.O. Bo~ 216. Lindtield 2070. 'ret.: The membe rs of the Council include p) Tha l as from. and including. 1s t (02) ~672 611 Lo,d Ellioll 01 Morpettl as Cha irman. Ap,it. 1987 Annual SUbsc ription Roy Eme,son an d GI>OIl lean vl~allas fa tes will be - Membe rs $50. Appo intment Deputy Chairme n and Ia n Acton. Peter Mfillat" $40. SUb' Cfibe.. $40 . Ande rson. Mike Anke rs. Richa,d " the, Slllde nls $3O. 0 •. Andrew Hede has r", ently bee n Ion. Oon Ba,nell . G'aham Ju l