ArchitecturalScienceReviewVolume511pp21-30

The Acoustical Characteristics of the Kocatepe Mosque in Ankara Turkey

Zuumlhre Suuml and Semiha Yilmazerdagger

PrograminArchitecturalAcousticsSchoolofArchitectureRensselaerPolytechnicInstituteTroyNY12180USADepartmentofInteriorArchitectureandEnvironmentalDesignBilkentUniversity06800BilkentAnkaraTurkey

daggerCorrespondingauthorTel9003122902592Fax9003122664136E-mailsemihabilkentedutr

Received5October2006accepted5September2007

Abstract ThisresearchinvestigatestheacousticalcharacteristicsofmosquesoftheclassicalOttomanperiodandthecontemporaryperiodKocatepeMosquethemajormosqueinAnkaraTurkeyasacasefromthecontemporaryeraisauniquecombinationof16thcenturyOttomanaestheticsand20thcenturytechnologyAlthoughpreviousandrecentmosqueshavebeeninspiredbythiscombinationtheuseofareinforcedconcretedomewithoutcavityresonatorscouldcauselongreverberationtimesespeciallyinlowfrequencies ThepurposeofthisstudywastoinvestigatethisconditionbystudyingKocatepeMosqueTheacousticalcharacteristicsofthemosquewereanalyzedbycomputersimulationObjectiveroom-acousticindicatorsincludingreverberationtime(RT)earlydecaytime(EDT)clarity(C80)definition(D50)lateralfraction(LF)speechtransmissionindex(STI)andstrength(G)arepresentedTheresultsshowthattheacousticalqualityofKocatepeMosqueisnotoptimalwhenitisemptyandclosesttooptimalconditionswhenfullyoccupiedThenewtechniquestoovercomeexcessivelowfrequencyattenuationcausedbyhugeconcreteshellsoftodayrsquosmosquesarediscussedinconclusionKeywords AcousticalparametersMosqueacousticsODEONReverberationtimeRoomacousticsSpeechintelligibility

IntroductionThe mosque as an important building type of Muslim

architecture has evolved to meet Islamic needs A varietyof different worship activities happen within these multi-functional public spaces these different uses have differentacoustical requirements As in many other religionsworshippers sometimes need solitude while at othertimes they want to feel in absolute unity with the otherspresent Acoustics are one of the basic means of creatingdifferent effects (Karabiber 1999) Although mosques areuniquely importantbuildings in everyMuslimcommunitytheir acoustical quality has not received adequate attentionespeciallyinrecentexamples

The architect Sinanrsquos mosques have been studied in aEuropean Commission Fifth Framework INCO-MEDProgram called ldquoConservation of the Acoustical HeritagebyRevival and IdentificationofSinanrsquosMosquesrsquoAcoustics(CAHRISMA)rdquothisstudygenerallyprovedthemtobeveryefficient(YukselBinanampUnver2003)TheCAHRISMAproject completed in 2003 studied the acoustics of oldByzantine churches and mosques built by the 16th centuryTurkish Architect SinanAnotherprioritywastointroducetheconceptofHybridArchitecturalHeritagetheideaisthatthearchitecturalheritageconceptconsideredinconservationand restoration projects should be upgraded to cover

acousticalaswellasvisualaspectsforspaceshavingacousticimportance(CAHRISMA2007Karabiber2000)

Studies have compared the acoustical quality of variousspaces One study by Fausti Pompoli and Prodi (2003)comparedtheacousticsofmosquesandchurchesTheyfoundthatreverberationtimeintheunoccupiedconditionisverylong giving them a unique feeling of majesty In anotherstudyKarabiber andErdogan (2002) compared theancientmosqueofKadırgaSokulluMehmetPasawitharecentoneSisliMerkezTheyconcludedthatalthoughtherewasnogreatdifferencebetweenthetotal soundabsorptionof thespacestheacousticsintheancientmosquewasbetter

Sinanrsquos important mosques have been analyzed in otherstudies besides the CAHRISMA project (Kayili 2002Topaktas2003)Kayili(2000)arguedthattheacousticsystemsSinanappliedespeciallyusingHelmholtz(cavity)resonatorstechnology are among the most successful applications ofacoustic science It isbelieved that this applicationbecameatradition inOttomanmosques However ithasnotbeenpossibletoinvestigatethestateofresonatorsineverymosquedomeassuchanopportunityonlyoccursduringrestorationwork however that opportunity did arise in the SultanAhmet(Blue)MosqueNaturallywhentheopeningsoftheresonatorsarecompletelyclosedtheyareunabletofunction

wwwearthscancoukjournalsasre doi103763asre20085104

ArchitecturalScienceReview Volume51Number1March200822

InlargespacesandatlowfrequenciesthiscausesundesirablereverberationcurvesThiswasconfirmedbymeasurementsofreverberationtimeintheSuleymaniyeandSelimiyeMosqueswhere a long reverberation time was measured especiallyin the low frequencies (Kayili 2005) However the recentmosqueswhichhavedomesbuiltupwithreinforcedconcretewithoutcavityresonatorshavenotbeenstudied

InthisstudytheKocatepeMosquewaschosenasanexampleoftheOttomanstyleconcretedomedmosquesto investigatethisspecificconditionanditsoutcomeinmosqueacousticsalsotocontributetotheongoingresearchonhybridarchitecturalheritageTheKocatepeMosquewithitsarchitecturalgrandeurandcongregationcapacityhasaconsiderablesocialrolewithinthecommunityitservesItisalsooneoftheleadingexamplesofthecontemporaryeraofreinforcedconcretedometypologyin mosque acoustics and architecture The main differencebetween the recentmosqueswith theirnew technology andtheOttomaneraSinanmosquesarethecavityresonatorsthatwerefrequentlyusedinancientdomestobalancelowfrequencyreverberationTheeffectsofthismissingarchitecturalelementinrecentmosquessuchastheKocatepeMosquearediscussedinacousticalterms

Theacousticalqualityofthemosqueshasalsobeendiscussedintheliterature(Abdou2003Karabiber1999Kayili2002ProdiampMarsilio2003Utami2005)Mosque design is mainlyMosquedesignismainlyinfluencedbyworshipconsiderations threedistinctactivitiesarecarriedoutinamosqueeitherseparatelyortogetherOneisprayerseitherindividuallyorinagroupledbyaleadertheImamThesecondisattendanceatasermonbeingdeliveredon its own or within the Friday noon prayers The third islisteningtoorrecitingsomeversesfromtheHolyQuranAllthese activities require a high level of speech audibility andintelligibility Toensuregood listeningconditionsacousticalneedsmustbeconsideredinthedesignphase(Abdou2003)

Insummarytherearethreedistinctacousticalrequirementsformosquesbull Audibilityofthenamaz(prayer)ordersoftheImam(prayer

leader)bull Recognizablesermonofthepreacher

bull ListeningtoorjoiningintherecitalofthemusicalversionsoftheHolyQuran

Thus intelligibility of both speech and other sounds isextremelyimportantespeciallyimportantforholytonesthatmustbebothspaciousandeffective(HamadahampHamouda1998Karabiber1999)Severalacousticalparametersgovernspeech audibility intelligibility and spaciousness of soundthe parameters usually employed in the acoustical analysisof mosques are reverberation time sound pressure leveldistribution and sound transmission index (Suarez SendraSuarez SendraNavarro amp Leon 2004) In order to clarify of the mannerthe purpose of this study is also to evaluate and predict theacoustical quality of the Kocatepe Mosque by assesing theseparametersasmajorvariables

Kocatepe MosqueHistory

Theideaofbuilding theKocatepeMosquedatesback toaprojectcompetitionin1940ThewinneratthisstagewasajointprojectbyVedatDalokayandNejatTekelioğlu Thecomplexwastoconsistofamosquefortwothousandpeoplethemainoffices of the Presidency of Religious Affairs a high IslamicinstitutealibraryanauditoriumamuseumaparkinglotfortwohundredcarsretailshopsalargekitchenandapolyclinicAsecondcompetitionwasheldin1957andthepartoftheprojectcontainingthemainofficesofthePresidencyofReligiousAffairswasfinishedin1964Athirdcompetitionin1967waswonbyHuumlsrevTaylaandFatihUluenginTheKocatepeMosquewasinauguratedin1987withacapacityforacongregationof15000(Tuumlrkiye Diyanet Vakfı [Presidency of Religious Affairs] 2007)TuumlrkiyeDiyanetVakfı[PresidencyofReligiousAffairs]2007)

Architectural FeaturesCommandingthehillofKocatepethemosqueisamodern

city landmark Its architecture combines the traditionalOttomanstyleofthe16thcenturywith20thcenturytechnology(seeFigure1)WithitsfourminaretsthemosqueresemblesSelimiyeitscentraldomeandhalfdomesareinspiredbytheSultanAhmetMosqueofMehmetAga from the school ofSinantheArchitect

FIGURES

Figure 1 Interior and exterior views from Kocatepe Mosque Ankara

Figure 2 3D display of the mosque

Figure 3 Plan and elevation view of the source and receiver locations

Figure 1 Exterior and interior views of the Kocatepe Mosque Ankara (Tuumlrkiye Diyanet Vakfi 2007)

23ZuumlhreSuumlandSemihaYilmazer AcousticalCharacteristicsofMosques

Themainmosquehasanestimatedvolumeof68696m68696mandafloor area of 4288msup2 It is 64 x 67m with a heightfloorareaof4288msup2 It is 64 x 67m with a height It is 64 x 67m with a height It is64x67mwithaheightof485mthedomehasadiameterof255mTherearefoursmallerhalfdomessurroundingthecentraldomeThesehalfdomes are enlarged with 12 domes The main dome restson four largepillars (elephant feet)eachwithadiameterof3metersUnlikeearliermosquestheelephantfeetareonly3 meters in diameter because of modern technology TheinteriorofthemosqueisthusmorespaciousandthereislessvisualandauditorydistractionthanwithlargercolumnsThemainpartofthemosque(harem)issurroundedbygalleriesinaldquoUrdquoformThegallerieshavetwostoriesquiteunliketheclassicalIstanbulmosquesAsaresultmorespaceisaddedtotheharem

The mosque as a whole is built on a reinforced concreteframeworksystemandtheoutsidesarecoveredwithartificialstoneThedomethehalfdomesandtheminaretsareconcreteAllthedomesarecoveredwithtraditionalleadThepointsofthecupolassurroundingthecourtyardandtheonesover thesidedoorsareofmarble Thepointsof themaindome theonessurroundingitandtheminaretsaremadeofgoldplatedcopperIntheinteriorclassicalOttomandecorationdominateswithmarbleandgoldleafstained-glasswindowsanddecorativetilesThewritingsonthemaindomethelionchestwritingsandthewritingscirclingthedomelikeabeltareofgoldplatedbrassOnemaindoorandfoursidedoorsaremadeofnailedwood without nails The sides of the staircases are coveredwithmarbleSmallbalconiesarebuiltatthesamelevelasthelandingsallowingmorespaceforlandingsThenichewithaheightof10metersandthepulpitwithaheightof870metersarebothmarbledecoratedwithspecialornaments

Thestainedglasspanesthatcreateasignificanteffectintheinteriordecorationof themosque aremadeof special glassand represent a transition from classical Ottoman style tomodernstyleTheinnerareaofthemosqueislitbyonemainchandelier32satellitechandeliersand4cornerchandeliersThemainchandelieris55metersindiameteritschainsaremadeofcrystalballsandtheframesaregoldplatedThereisacentralheatingsystemprovidingheatfromthefloorTheHaremhasahand-wovencarpetofaparticulartypewiththesamepattern seen in theAfyonUluMosque Thegalleriesalsohavecarpetsthataremachine-wovenwithspecialpatterns(Tuumlrkiye Diyanet Vakfı [Presidency of Religious Affairs]Tuumlrkiye Diyanet Vakfı [Presidency of Religious Affairs]2007)

Research MethodTechnological developments provide new opportunities for

acousticaldesign Theacoustical assessmentof a space in thedesignphaseusedtobedonewithscalemodelsthataretime-consumingandunpracticalItisnowpossibletousecomputersimulation to analyze acoustical properties prior to the actualconstruction of a building and acoustical design can becomeanintegralpartofthearchitecturaldesignprocessComputersimulation has proven to be a viable tool in designing musicbuildingssuchasconcerthallsoperahousesandmulti-purposeauditoria These not only minimize unexpected acousticproblemsbuttheyalsoallowmanyalternativestobetestedinashorttimespan(SchmidtampKirkegaard2004)

ComputersimulationsoftheKocatepeMosquewerecarriedout using ODEON Room Acoustics Program 605 softwarewhich was released by the Technical University of Denmark(Naylor1993)Thecalculationmethodofthissoftwareisbasedonpredictionalgorithmsincludingtheimage-sourcemethodandraytracingTheODEONRoomAcousticsProgramalsotakesintoaccountthestatisticalpropertiesoftheroomrsquosgeometryandabsorption(Rindel2000)

In2001asapartoftheCAHRISMAprojectthreeByzantinechurchesandthreemosquesweremodeledintheOdeonroomacousticprogram(WeitzeChristensenRindelampGade2001)These geometries are dominated by spherical and cylindrical(concaveconvex) shapes with large domes dominating theceilingAnotherfeatureisnumerouscolumnsandgalleriesthatobstructthedirectsoundInthemodelschangescanbemadeincalculationparameterssuchasthetransitionorder(TO)thenumberofraysandthenumberofsubdivisionintoplanesurfacesofconcaveconvexsurfacesThisisdonetoinvestigatetheeffectoftheacousticparametersindifferentrooms(accordingtoISO3382) This study shows that the ray tracing method yieldsthe best results (TO=0 for complex rooms with many curvedsurfacesmodeledwithrelativelysmallplanesurfaces

Inthisstudya3DModeloftheKocatepeMosqueisobtainedusingCADSoftwareThemodelisimportedtotheODEONRoomAcousticsProgramFigure2showsthe3Ddisplayofthemodel

Two omni-directional point sources and a receiver arespecifiedforeachactivitypatternFigures3and4illustratethelocationsofthereceiverandthesourcesSource1indicatesthe mihrab and Source 2 indicates the minbar Which isactivateddependsontheactivity

FIGURES

Figure 1 Interior and exterior views from Kocatepe Mosque Ankara

Figure 2 3D display of the mosque

Figure 3 Plan and elevation view of the source and receiver locations

Figure 2 3D display of the mosque

ArchitecturalScienceReview Volume51Number1March200824

Different materials at different sections of the mosquewith different sound absorption coefficients are assigned(Table 1) After fixing the calculation parameters theselected receiver surfaces are divided into grids assigning096msup2 for eachpersonpraying (Abdou2003) Inorderto get the results of different acoustical parameters andtheir distribution throughout the mosque the maps andcumulativedistributiongraphsforcalculatedparametersareobtainedforthesesurfaces

The process is repeated for different activities takingplace in the Kocatepe Mosque First reverberation time ismeasuredintheemptymosqueThefirstactivitymodeistheprayermodeforthisthesourceistheimamatthemihrabfacing towards the mihrab set at a height of 080m Fordaily prayers the mosque is assumed one third full ThesecondmodeisthepreachingmodeandrecitalofthemusicalversionsoftheHolyQuranThesourceistheimamattheminbarfacingtheworshippersThemosqueisassumedfullfor Friday prayers for listening to the Friday Speech forreligious days and festivals The receiver surfaces for bothactivitymodes are set at aheightof080m approximatelyseating level For the two different activity modes musicand speech related parameters are analyzed these includereverberation time (RT) early decay time (EDT) clarity(C50) definition (D80) speech transmission index (STI)lateralfraction(LF)andstrength(G)

Acoustical Analysis

Reverberation Time (RT) Reverberation time is the time required for a sound ina

roomtodecayby60dBafterthesoundsourcehasstoppedRT is the basic parameter that should be calculated in theassessment of room acoustics RT characterizes the decayofsoundforspecifyingtheefficiencyofaroominfulfillingthe acoustical requirements depending upon the activity orfunctionthattakesplaceForliturgicalpurposes(orchestrachorusororgan)theoptimumrangeforreverberationtimeis30to35sforspaceslargerthan10000m(Egan1994)

InthemosqueRTunoccis748sinthemiddlefrequenciesThisismuchlongerthantheoptimumlevelevenforliturgicalpurposes Long sounddecayat low frequencies can reducetheintelligibilityofspeechForgoodintelligibilityRTvaluesat low octave-band frequencies should remain flat down to100 Hz At low frequencies an increase in RT values ofaround 10-20 would still yield a natural sound but inthecaseofKocatepetheincreaseisabout50intheemptymosque Thebass ratiowhich is theaverageRTat low tomid frequencies is 154 in the Kocatepe Mosque whereasthe optimum range for music performances is 12 to 125(Kuttruff 1991) This value implies an over-warm spacerichinbasssoundthisiseffectiveforspatialimpressionbutdetrimentaltospeechintelligibility

FIGURES

Figure 1 Interior and exterior views from Kocatepe Mosque Ankara

Figure 2 3D display of the mosque

Figure 3 Plan and elevation view of the source and receiver locationsFigure 3 Plan and elevation view of the source and receiver locations

Table 1 SoundabsorptioncoefficientsofdifferentmaterialsusedintheKocatepeMosque

4

TABLES

Table 1 Sound absorption coefficients of different materials used in Kocatepe Mosque

Material63

Hz

125

Hz

250

Hz

500

Hz

1

kHz

2

kHz

4 k

Hz

8

kHz

Scattering

Factor

Smooth concrete 001 001 001 002 002 002 005 005 01

Marble 001 001 001 001 001 002 002 002 02

Solid wood panel 015 015 010 006 008 010 005 005 01

Double glazing 002 002 006 003 003 002 002 002 01

Solid wooden door 014 014 010 006 008 010 010 010 01

Prayers-11 per msup2 055 055 086 083 087 090 087 087 05

Prayers-13 per msup2 010 010 021 041 065 075 071 071 05

Ceramic tiles 001 001 001 001 002 002 002 002 03

Carpet on concrete 002 002 006 014 037 060 065 065 01

Chandelier 035 035 025 018 012 007 004 004 02

Lime plaster 004 004 005 004 008 004 006 006 01

Table 2 Relation between scores of speech transmission quality and STI (RASTI)

Quality Score STI (RASTI) value

Bad 0 to 032

Poor 032 ndash 045

Fair 045 ndash 060

Good 060 ndash 075

Excellent 075 to 10

25ZuumlhreSuumlandSemihaYilmazer AcousticalCharacteristicsofMosques

For the first activity mode the prayer mode in a onethird full mosque the global estimation results show thatthemosquehasanRT13occof435satthemidfrequenciesForthesecondactivitymodethepreacher-recitalmodeinafullmosqueRT33occis319sat themid frequencies (Figure5) Distribution maps illustrate that 90 of the receiversurfaceshasanRT13occof62sandRT33occ281to389s(seeAppendixM1)

RT13occismuchbetterthanRTunoccbutitisstillhigherthantheoptimumrangeAnyvaluegreaterthan10sisespeciallydetrimental to the intelligibilityof speech (Kuttruff1991)The values obtained are unsatisfactory for both music andspeechApositivepointworthmentioningisthatthesoundfield isnearlyeven Therearea fewechospotsunderneaththedomebutthisisacceptablegiventhedimensionsofthedomeRT33occisintheoptimumrangeduringtherecitalsofthemusicalversionoftheHolyQuranbutinthepreachingmodeitreducestheintelligibilityofspeech

ThenumberofworshippersattendingFridayprayergreatlyaffectsthetotalsoundabsorptioninsidethemosqueandtheduration reverberation time becomes more controlled Inthese conditions RT33occ is satisfactory for liturgical musicunderneath the main and secondary domes and there arealmostnoechospotsdespitethehugeconcavesurfaceaboveThis results from the dimensions of the domes and theirheightsabovethefloorWhenthemaindomeiscompletedtoasphereitisfarabovethereceiversurfacesThereforenoneof the receiver surfaces is within this problematic focusingvolume

Early Decay Time (EDT)EDTisroughlydefinedasthesensationofRTItisabetter

indicatorofthesubjectivefeelingofreverberationtimeandiscriticalindefiningtheacousticalqualityofamusichallThe

2

Figure 4 3D view of the source and receiver locationsFigure 4 3D view of the source and receiver locations

Figure 5 Calculated values for RTunocc and RT13occ and RT33occ for frequencies from 125 to 4000 Hz shaded area optimum RT

(Egan 1994)

3

125 Hz 250 Hz 500 Hz 1 kHz 2 kHz 4 kHz

RTunocc 171 124 943 554 378 250

RT13occ 1199 765 311 360 299 221

RT33occ 545 338 335 304 274 203

Figure 5 Calculated values for RTunocc and RT13occ and RT33occ for frequencies from 125 to

4000 Hz shaded area optimum RT (Egan 1994)

3

125 Hz 250 Hz 500 Hz 1 kHz 2 kHz 4 kHz

RTunocc 171 124 943 554 378 250

RT13occ 1199 765 311 360 299 221

RT33occ 545 338 335 304 274 203

Figure 5 Calculated values for RTunocc and RT13occ and RT33occ for frequencies from 125 to

4000 Hz shaded area optimum RT (Egan 1994)

ArchitecturalScienceReview Volume51Number1March200826

EDT should not be higher than plusmn10 for good acousticsForliturgicalpurposesEDTshouldbebetween27and385s(Templeton1993)

For the first activity (ie the 13 full prayer mode)EDT13occ is 63 s and for the second activity (ie the fullpreacher-recital mode) EDT33occ is 44 s (see Appendix BM2) Neitherof these is intheoptimumrange HowevertheyareconsistentwithRT13occandRT33occimplyingagooddistributionofsoundthroughoutthemosqueEventhoughnotashighscatteringcoefficientsastobeexpectedareassignedtosurfaceslikecolumnheadsstalactitesandornamentationaroundmihrabminbarandbalconieswithinthesimulationconsidering their relatively small modeled size there is stillaproperamountofdiffusioninthemosquethateliminateshotanddeadspotsandprovidesevendistributionofsoundespecially in high frequencies Using higher scatteringcoefficientsforthosesmallsurfaceswouldprobablybemoreaccurateandcausemuchmoredrasticresultsandmuchofhomogeneityofsoundfield

Clarity (C80) Clarity is defined technically as the ratio of early sound

energy (arriving within 80 ms of direct sound) to late orreverberant sound energy (arriving more than 80 ms afterthe direct sound) This quality characterizes the separationin time of sounds from individual instruments or groups ofinstruments Thisparametermustbewithincertainlimits ifthemusicaldetailsaretobeheard Ingeneralasatisfactoryacousticalspaceshouldhaveclaritybetween-2and+2dBtosatisfybothmusicandspeechcriteriaandbetween-1and+3dBforchoralmusic(Kuttruff1991)

InthefirstactivitymodeC8013occisbetween-4and+4dBinthesecondactivitymodeC8033occisat+18dBfor90ofthereceiverpoints(seeAppendixM3)C8013occvaluesarewithin the acceptable range atmost of the receiver locationsclosertothemihrabandaroundthemuezzinplatform ThevaluesobtainedaresuitableforbothmusicandspeechThisindicatesanadequatedistributionofearlyreflectionsinthesereceiverlocationsValuesbegintodecreaseunderthepartsofthemaindomethatareclosertothesidewalls

Definition (D50)Definitionisthemeasurederivedfromtheearrsquosresponseto

consecutiveimpulsesitcharacterizestheratioofeffectiveenergytothetotalenergyinanimpulseresponseupto50msThereisagoodrelationshipbetweendefinitionandspeechintelligibilityThisparametershouldbegreaterthan20tosatisfybothmusicandspeechperformances(Templeton1993)

In thefirst activitymodeD5013occ is above30 in thesecondactivitymodeD5033occvariesbetween22and50(see Appendix M4) D5013occ and D5033occ are acceptableunderneath thehalf of themaindome that is closer to themihrabEvensurfacesunderneaththesecondarydomesexhibitgoodvaluesThisisprobablycausedbythefirstreflectionsoffthedomesanddiffusedsurfacesofthestalactitesbetweenthecolumnsanddomes Thedomesdefineanacoustical spaceunderneath

Lateral Fraction (LF)Lateralfractiondefinestherelationshipbetweenasenseof

spatialimpressionandthearrivalofreflectedsoundfromwallstothesideofthelistener(Templeton1993)Formusictheparametershouldbebetween01-035andgreaterthan035forallotherpurposes(Barron1993)

BothforthefirstandsecondactivitymodestheLF13occandLF33occ values are above030 (seeAppendixBM5) Bettervaluesareexpectedatpointsclosertothewall surfaces Theworst place in terms of LF13occand LF33occ is underneath themaindomeLF33occisbetternearthesidewallsandaroundthemuezzinplatform

Speech Transmission Index (STI) STI is directly related to speech To ensure good speech

intelligibility theenvelopeof the signal shouldbepreservedallowingthevariousfrequencybandstocontributetospeechqualityThescaleofoptimumvaluesforthesoundtransmissionindexisshowninTable2

ForthefirstactivitymodeSTI13occisaround052whichisconsideredfairforthesecondactivitymodeSTI33occisaround060 which is considered a good class (see Appendix M6)TheresultsshowthatthereisanevendistributionofSTI13occthroughoutthemosque At locations justaroundthesourceor imambettervalues are seen Enclosedbuildingsof largevolumearenotdesignedforoptimal speechcommunicationsotheseresultsaresatisfactoryconsideringthevolumeofthemosque STI33occ results are even better indicating that theintelligibilityof the imamat theminbar is satisfactorywhenthemosqueisfull

Theoptimumvolumeperpersonfortheintelligibilityofthespeech isbetween4-6m(MaekawaampLord1994) WhentheKocatepeMosqueisconsideredfullthevolumeperpersoncorrespondsto45mThusthevolumeofthemosquemeetsacousticalrequirements

Strength (G)Soundpressurelevel(SPL)mapsindicatetheGresultswhen

anomni-directionalsourcetypeandapowerof31dBOctavebands are selected from the appropriate Point Source EditorofODEONsoftwareForsymphonicandliturgicalmusicGvaluesshouldbegreaterthan3dB(Lynge2001)

Thesimulationresultsforthefirstandsecondactivitymodesshow that sound distribution is satisfactory underneath themaindomeandevenunderthesmallerdomes(seeAppendixM7) These smaller domes are effective in distributing thesoundfromtheimamtofurtherspotsG33occvaluesarehigherclosertotheminbarthanintherestofthemosquethisisgiventhattheimamisthesource

Table 2 RelationbetweenscoresofspeechtransmissionqualityandSTI(RASTI)

4

TABLES

Table 1 Sound absorption coefficients of different materials used in Kocatepe Mosque

Material63

Hz

125

Hz

250

Hz

500

Hz

1

kHz

2

kHz

4 k

Hz

8

kHz

Scattering

Factor

Smooth concrete 001 001 001 002 002 002 005 005 01

Marble 001 001 001 001 001 002 002 002 02

Solid wood panel 015 015 010 006 008 010 005 005 01

Double glazing 002 002 006 003 003 002 002 002 01

Solid wooden door 014 014 010 006 008 010 010 010 01

Prayers-11 per msup2 055 055 086 083 087 090 087 087 05

Prayers-13 per msup2 010 010 021 041 065 075 071 071 05

Ceramic tiles 001 001 001 001 002 002 002 002 03

Carpet on concrete 002 002 006 014 037 060 065 065 01

Chandelier 035 035 025 018 012 007 004 004 02

Lime plaster 004 004 005 004 008 004 006 006 01

Table 2 Relation between scores of speech transmission quality and STI (RASTI)

Quality Score STI (RASTI) value

Bad 0 to 032

Poor 032 ndash 045

Fair 045 ndash 060

Good 060 ndash 075

Excellent 075 to 10

27ZuumlhreSuumlandSemihaYilmazer AcousticalCharacteristicsofMosques

ConclusionTwo important criteria in mosque acoustics are an even

distributionofsoundandoptimumreverberationtimesWhenalltheparametersstudiedareconsideredtheresultsshowthattheKocatepeMosqueisagoodplaceforrecitingthemusicalversionof theHolyQuranas itcreatesaritualandtranquilacoustical atmosphere However for the prayer mode themosquehasinadequateintelligibilityofspeechThisisverifiedbytheabruptjumpsinreverberationtimeatlowfrequenciesintheactivitymodesandwhenthemosqueisempty Thereis excessive reverberation and a highbass ratio in the emptymosque However a flat distribution of reverberation onfrequency is important for intelligibilityof speech regardlessofvolumesoundabsorptionshouldbebalancedfordifferentfrequencies Ancient mosques achieved this bass balance byspecialacousticaltreatmentssuchascavityresonators(Kayili2005)

Excessive reverberation may arise when the walls and theceiling of an auditorium are covered with highly reflectivematerialthatcausesadegradationofspeechsoundsespeciallywhenthesurfaceisconcaveThesoundreinforcementsystemwouldincreasethedirectsoundlevelatlisteningpositionsandyieldbetterintelligibility

In the second activity mode the number of worshippersattendingFridayprayergreatlyaffects total soundabsorptioninsidethemosqueandtheparametersfallwithinthedesirableranges Themosqueissuccessful intermsofsoundfocusingconsideringalloftheactivitypatternsInmanynewmosquesldquofocusingrdquoisaseriousproblembutinKocatepethisacousticaldefect is overcome by the proportions and locations of thedomes(ProdiampMarsilio2003Utami2005)Asthemaindome is above the receiver surfacesnoneof theworshippersstayintheprobablefocusingvolumeSecondarydomeshavenofocusingeffectonthecontrarytheyprovidesoundenergytodistantpartsofthemosqueandtheyprovideadiffusedfieldbeneaththem

TheancientMosquesofSinanarestilluniqueandmaybeareferenceandmodelformosquesyettobebuilt(CAHRISMA2007 Kayili 2005 Topaktas 2003) Except for cavityresonators the basic architectural features of Kocatepe areinspiredbysomeoftheSinanMosquesTheinteriortreatmentofthemosqueisimportantinprovidingadivineenvironmentThestalactitesasa transitionalarchitecturalelementbetweendome and columns the ornamentation on the mihrabthe minbar the columns and balconies-- all have not justaestheticappealbutalsoacousticalpurposesespeciallyinhighfrequencies

Itisclearlyobservedfromtheresultsthatneweramosqueswith concrete domes have a greater tendency to keep andenhance the low frequency soundswithin the audience areaThis low frequency interferes with the speech intelligibilityandputsconstraintsontheclarityofthelectureortheordersconveyedbytheImamThematerialsusedinancientmosquesandthecavityresonatortechniquewithvarioussizedglasspotswithin the dome were wise and efficient in providing goodacousticalqualityRecentlydevelopedacousticaldevicessuchasHelmholtzresonatorsprovidenewdesignopportunitiesthesecouldduplicatetheprincipledesigncriteriaofSinanandcouldimprovetheacousticsof futuremosquesaswellasKocatepe

However considering the feasibility of Helmholtz resonatorswithin big spaces such as mosques the authors would liketo introduce micro-perforated panel absorbers to be studiedmore comprehensively in future applications and studies ofmosqueacousticsThisnewtechnologyusesthesamephysicsprincipals asHelmholtz resonators in substantial attenuationof lowfrequencieswhilekeepingthehighfrequencies inthedesiredrangesThearchitecturalapproachandfeelingofthespaces could then be combined with the new technologicalimprovements of this era and architectural acoustics as asciencecouldbegreatlyfurthered

InthisstudythedrawbacksofnewconstructiontechniquesontheacousticsofcontemporarymosqueshavebeenanalyzedTheKocatepeMosqueisoneoftheleadingexamplesoftheOttomanstyle reinforced concrete dome typology in mosque acousticsand architecture The acoustic simulationsdemonstrated thatthismosquetypologyhasstrongaestheticfeaturesbutmoderateacoustical effects in comparison to the authentic OttomanmosquesTheexcessivereverberationinthelowfrequencyrangeoftheenergydecayisfoundtobethebasicproblemofthatnewtypologyThekeytoimprovetheproblemsintodayrsquosmosquesistoadjustandcombinethematerialtechnologyoftheeratotheaestheticsofancientmosquessuchthatitfitstotheconceptandeventuallytothefunctionalityofthemosquesacoustics

References Abdou AA (2003) Measurement of acoustical characteristics

of mosques in Saudi Arabia The Journal of Acoustical Society of America 113(3)1505-1517

Barron M (1993) Auditorium Acoustics and Architectural DesignLondonEampFNSpon

CAHRISMA(2007)Orsted ndashDTUAvailableathttpserveroersteddtudkwwwoldatcahrismaretrieved4September2007

EganMD(1994)Architectural AcousticsNewYorkMcGraw-HillFaustiPPompoliRampProdiN(2003)Comparingtheacoustics

ofmosquesandByzantinechurches19th International Symposium CIPA (The International Committee for Architectural Photogrammetry) AntalyaTurkeyTechnicalUniversityofIstanbul

HamadahHAampHamoudaHM(1998)AssessmentofspeechintelligibilityinlargeauditoriacasestudyKuwaitStateMosqueApplied Acoustics54(4)273-289

KarabiberZ(1999)AcousticalproblemsinmosquesAcasestudyonthethreemosquesinIstanbulForum Acusticum BerlinCD-ROM4pgs

Karabiber Z (2000) A new approach to an ancient subjectCAHRISMAprojectInProceedings of the 7th International Congress on Sound and Vibration (1661-1668) Garmisch-PartekirchenGermanyInternationalInstituteofAcousticsandVibration

KarabiberZampErdoganS(2002)Comparisonoftheacousticalpropertiesofanancientandarecentmosque Forum AcusticumSeville Spain Spanish Acoustical Society (SEA) Available athttpwwwcicaes~forum2002retrieved4September2007

KayiliM(2000)UseofcavityresonatorsinAnatoliasinceVitruviusInternational Institute of Acoustics and Vibration 31621-1628

KayiliM(2002)AnadolursquodagelenekselakustiksistemlerveMimarSinan uygulamaları (Anatolian traditional acoustic works andworksdonebySinantheArchitect)[Turkish]InProceeding of 6th National Acoustical Congress TAKDER 26-28 October (233-238)AntalyaTurkeyTurkishAcousticalSociety

23ZuumlhreSuumlandSemihaYilmazer AcousticalCharacteristicsofMosques

Themainmosquehasanestimatedvolumeof68696m68696mandafloor area of 4288msup2 It is 64 x 67m with a heightfloorareaof4288msup2 It is 64 x 67m with a height It is 64 x 67m with a height It is64x67mwithaheightof485mthedomehasadiameterof255mTherearefoursmallerhalfdomessurroundingthecentraldomeThesehalfdomes are enlarged with 12 domes The main dome restson four largepillars (elephant feet)eachwithadiameterof3metersUnlikeearliermosquestheelephantfeetareonly3 meters in diameter because of modern technology TheinteriorofthemosqueisthusmorespaciousandthereislessvisualandauditorydistractionthanwithlargercolumnsThemainpartofthemosque(harem)issurroundedbygalleriesinaldquoUrdquoformThegallerieshavetwostoriesquiteunliketheclassicalIstanbulmosquesAsaresultmorespaceisaddedtotheharem

The mosque as a whole is built on a reinforced concreteframeworksystemandtheoutsidesarecoveredwithartificialstoneThedomethehalfdomesandtheminaretsareconcreteAllthedomesarecoveredwithtraditionalleadThepointsofthecupolassurroundingthecourtyardandtheonesover thesidedoorsareofmarble Thepointsof themaindome theonessurroundingitandtheminaretsaremadeofgoldplatedcopperIntheinteriorclassicalOttomandecorationdominateswithmarbleandgoldleafstained-glasswindowsanddecorativetilesThewritingsonthemaindomethelionchestwritingsandthewritingscirclingthedomelikeabeltareofgoldplatedbrassOnemaindoorandfoursidedoorsaremadeofnailedwood without nails The sides of the staircases are coveredwithmarbleSmallbalconiesarebuiltatthesamelevelasthelandingsallowingmorespaceforlandingsThenichewithaheightof10metersandthepulpitwithaheightof870metersarebothmarbledecoratedwithspecialornaments

Thestainedglasspanesthatcreateasignificanteffectintheinteriordecorationof themosque aremadeof special glassand represent a transition from classical Ottoman style tomodernstyleTheinnerareaofthemosqueislitbyonemainchandelier32satellitechandeliersand4cornerchandeliersThemainchandelieris55metersindiameteritschainsaremadeofcrystalballsandtheframesaregoldplatedThereisacentralheatingsystemprovidingheatfromthefloorTheHaremhasahand-wovencarpetofaparticulartypewiththesamepattern seen in theAfyonUluMosque Thegalleriesalsohavecarpetsthataremachine-wovenwithspecialpatterns(Tuumlrkiye Diyanet Vakfı [Presidency of Religious Affairs]Tuumlrkiye Diyanet Vakfı [Presidency of Religious Affairs]2007)

Research MethodTechnological developments provide new opportunities for

acousticaldesign Theacoustical assessmentof a space in thedesignphaseusedtobedonewithscalemodelsthataretime-consumingandunpracticalItisnowpossibletousecomputersimulation to analyze acoustical properties prior to the actualconstruction of a building and acoustical design can becomeanintegralpartofthearchitecturaldesignprocessComputersimulation has proven to be a viable tool in designing musicbuildingssuchasconcerthallsoperahousesandmulti-purposeauditoria These not only minimize unexpected acousticproblemsbuttheyalsoallowmanyalternativestobetestedinashorttimespan(SchmidtampKirkegaard2004)

ComputersimulationsoftheKocatepeMosquewerecarriedout using ODEON Room Acoustics Program 605 softwarewhich was released by the Technical University of Denmark(Naylor1993)Thecalculationmethodofthissoftwareisbasedonpredictionalgorithmsincludingtheimage-sourcemethodandraytracingTheODEONRoomAcousticsProgramalsotakesintoaccountthestatisticalpropertiesoftheroomrsquosgeometryandabsorption(Rindel2000)

In2001asapartoftheCAHRISMAprojectthreeByzantinechurchesandthreemosquesweremodeledintheOdeonroomacousticprogram(WeitzeChristensenRindelampGade2001)These geometries are dominated by spherical and cylindrical(concaveconvex) shapes with large domes dominating theceilingAnotherfeatureisnumerouscolumnsandgalleriesthatobstructthedirectsoundInthemodelschangescanbemadeincalculationparameterssuchasthetransitionorder(TO)thenumberofraysandthenumberofsubdivisionintoplanesurfacesofconcaveconvexsurfacesThisisdonetoinvestigatetheeffectoftheacousticparametersindifferentrooms(accordingtoISO3382) This study shows that the ray tracing method yieldsthe best results (TO=0 for complex rooms with many curvedsurfacesmodeledwithrelativelysmallplanesurfaces

Inthisstudya3DModeloftheKocatepeMosqueisobtainedusingCADSoftwareThemodelisimportedtotheODEONRoomAcousticsProgramFigure2showsthe3Ddisplayofthemodel

Two omni-directional point sources and a receiver arespecifiedforeachactivitypatternFigures3and4illustratethelocationsofthereceiverandthesourcesSource1indicatesthe mihrab and Source 2 indicates the minbar Which isactivateddependsontheactivity

FIGURES

Figure 1 Interior and exterior views from Kocatepe Mosque Ankara

Figure 2 3D display of the mosque

Figure 3 Plan and elevation view of the source and receiver locations

Figure 2 3D display of the mosque

ArchitecturalScienceReview Volume51Number1March200824

Different materials at different sections of the mosquewith different sound absorption coefficients are assigned(Table 1) After fixing the calculation parameters theselected receiver surfaces are divided into grids assigning096msup2 for eachpersonpraying (Abdou2003) Inorderto get the results of different acoustical parameters andtheir distribution throughout the mosque the maps andcumulativedistributiongraphsforcalculatedparametersareobtainedforthesesurfaces

The process is repeated for different activities takingplace in the Kocatepe Mosque First reverberation time ismeasuredintheemptymosqueThefirstactivitymodeistheprayermodeforthisthesourceistheimamatthemihrabfacing towards the mihrab set at a height of 080m Fordaily prayers the mosque is assumed one third full ThesecondmodeisthepreachingmodeandrecitalofthemusicalversionsoftheHolyQuranThesourceistheimamattheminbarfacingtheworshippersThemosqueisassumedfullfor Friday prayers for listening to the Friday Speech forreligious days and festivals The receiver surfaces for bothactivitymodes are set at aheightof080m approximatelyseating level For the two different activity modes musicand speech related parameters are analyzed these includereverberation time (RT) early decay time (EDT) clarity(C50) definition (D80) speech transmission index (STI)lateralfraction(LF)andstrength(G)

Acoustical Analysis

Reverberation Time (RT) Reverberation time is the time required for a sound ina

roomtodecayby60dBafterthesoundsourcehasstoppedRT is the basic parameter that should be calculated in theassessment of room acoustics RT characterizes the decayofsoundforspecifyingtheefficiencyofaroominfulfillingthe acoustical requirements depending upon the activity orfunctionthattakesplaceForliturgicalpurposes(orchestrachorusororgan)theoptimumrangeforreverberationtimeis30to35sforspaceslargerthan10000m(Egan1994)

InthemosqueRTunoccis748sinthemiddlefrequenciesThisismuchlongerthantheoptimumlevelevenforliturgicalpurposes Long sounddecayat low frequencies can reducetheintelligibilityofspeechForgoodintelligibilityRTvaluesat low octave-band frequencies should remain flat down to100 Hz At low frequencies an increase in RT values ofaround 10-20 would still yield a natural sound but inthecaseofKocatepetheincreaseisabout50intheemptymosque Thebass ratiowhich is theaverageRTat low tomid frequencies is 154 in the Kocatepe Mosque whereasthe optimum range for music performances is 12 to 125(Kuttruff 1991) This value implies an over-warm spacerichinbasssoundthisiseffectiveforspatialimpressionbutdetrimentaltospeechintelligibility

FIGURES

Figure 1 Interior and exterior views from Kocatepe Mosque Ankara

Figure 2 3D display of the mosque

Figure 3 Plan and elevation view of the source and receiver locationsFigure 3 Plan and elevation view of the source and receiver locations

Table 1 SoundabsorptioncoefficientsofdifferentmaterialsusedintheKocatepeMosque

4

TABLES

Table 1 Sound absorption coefficients of different materials used in Kocatepe Mosque

Material63

Hz

125

Hz

250

Hz

500

Hz

1

kHz

2

kHz

4 k

Hz

8

kHz

Scattering

Factor

Smooth concrete 001 001 001 002 002 002 005 005 01

Marble 001 001 001 001 001 002 002 002 02

Solid wood panel 015 015 010 006 008 010 005 005 01

Double glazing 002 002 006 003 003 002 002 002 01

Solid wooden door 014 014 010 006 008 010 010 010 01

Prayers-11 per msup2 055 055 086 083 087 090 087 087 05

Prayers-13 per msup2 010 010 021 041 065 075 071 071 05

Ceramic tiles 001 001 001 001 002 002 002 002 03

Carpet on concrete 002 002 006 014 037 060 065 065 01

Chandelier 035 035 025 018 012 007 004 004 02

Lime plaster 004 004 005 004 008 004 006 006 01

Table 2 Relation between scores of speech transmission quality and STI (RASTI)

Quality Score STI (RASTI) value

Bad 0 to 032

Poor 032 ndash 045

Fair 045 ndash 060

Good 060 ndash 075

Excellent 075 to 10

25ZuumlhreSuumlandSemihaYilmazer AcousticalCharacteristicsofMosques

For the first activity mode the prayer mode in a onethird full mosque the global estimation results show thatthemosquehasanRT13occof435satthemidfrequenciesForthesecondactivitymodethepreacher-recitalmodeinafullmosqueRT33occis319sat themid frequencies (Figure5) Distribution maps illustrate that 90 of the receiversurfaceshasanRT13occof62sandRT33occ281to389s(seeAppendixM1)

RT13occismuchbetterthanRTunoccbutitisstillhigherthantheoptimumrangeAnyvaluegreaterthan10sisespeciallydetrimental to the intelligibilityof speech (Kuttruff1991)The values obtained are unsatisfactory for both music andspeechApositivepointworthmentioningisthatthesoundfield isnearlyeven Therearea fewechospotsunderneaththedomebutthisisacceptablegiventhedimensionsofthedomeRT33occisintheoptimumrangeduringtherecitalsofthemusicalversionoftheHolyQuranbutinthepreachingmodeitreducestheintelligibilityofspeech

ThenumberofworshippersattendingFridayprayergreatlyaffectsthetotalsoundabsorptioninsidethemosqueandtheduration reverberation time becomes more controlled Inthese conditions RT33occ is satisfactory for liturgical musicunderneath the main and secondary domes and there arealmostnoechospotsdespitethehugeconcavesurfaceaboveThis results from the dimensions of the domes and theirheightsabovethefloorWhenthemaindomeiscompletedtoasphereitisfarabovethereceiversurfacesThereforenoneof the receiver surfaces is within this problematic focusingvolume

Early Decay Time (EDT)EDTisroughlydefinedasthesensationofRTItisabetter

indicatorofthesubjectivefeelingofreverberationtimeandiscriticalindefiningtheacousticalqualityofamusichallThe

2

Figure 4 3D view of the source and receiver locationsFigure 4 3D view of the source and receiver locations

Figure 5 Calculated values for RTunocc and RT13occ and RT33occ for frequencies from 125 to 4000 Hz shaded area optimum RT

(Egan 1994)

3

125 Hz 250 Hz 500 Hz 1 kHz 2 kHz 4 kHz

RTunocc 171 124 943 554 378 250

RT13occ 1199 765 311 360 299 221

RT33occ 545 338 335 304 274 203

Figure 5 Calculated values for RTunocc and RT13occ and RT33occ for frequencies from 125 to

4000 Hz shaded area optimum RT (Egan 1994)

3

125 Hz 250 Hz 500 Hz 1 kHz 2 kHz 4 kHz

RTunocc 171 124 943 554 378 250

RT13occ 1199 765 311 360 299 221

RT33occ 545 338 335 304 274 203

Figure 5 Calculated values for RTunocc and RT13occ and RT33occ for frequencies from 125 to

4000 Hz shaded area optimum RT (Egan 1994)

ArchitecturalScienceReview Volume51Number1March200826

EDT should not be higher than plusmn10 for good acousticsForliturgicalpurposesEDTshouldbebetween27and385s(Templeton1993)

For the first activity (ie the 13 full prayer mode)EDT13occ is 63 s and for the second activity (ie the fullpreacher-recital mode) EDT33occ is 44 s (see Appendix BM2) Neitherof these is intheoptimumrange HowevertheyareconsistentwithRT13occandRT33occimplyingagooddistributionofsoundthroughoutthemosqueEventhoughnotashighscatteringcoefficientsastobeexpectedareassignedtosurfaceslikecolumnheadsstalactitesandornamentationaroundmihrabminbarandbalconieswithinthesimulationconsidering their relatively small modeled size there is stillaproperamountofdiffusioninthemosquethateliminateshotanddeadspotsandprovidesevendistributionofsoundespecially in high frequencies Using higher scatteringcoefficientsforthosesmallsurfaceswouldprobablybemoreaccurateandcausemuchmoredrasticresultsandmuchofhomogeneityofsoundfield

Clarity (C80) Clarity is defined technically as the ratio of early sound

energy (arriving within 80 ms of direct sound) to late orreverberant sound energy (arriving more than 80 ms afterthe direct sound) This quality characterizes the separationin time of sounds from individual instruments or groups ofinstruments Thisparametermustbewithincertainlimits ifthemusicaldetailsaretobeheard Ingeneralasatisfactoryacousticalspaceshouldhaveclaritybetween-2and+2dBtosatisfybothmusicandspeechcriteriaandbetween-1and+3dBforchoralmusic(Kuttruff1991)

InthefirstactivitymodeC8013occisbetween-4and+4dBinthesecondactivitymodeC8033occisat+18dBfor90ofthereceiverpoints(seeAppendixM3)C8013occvaluesarewithin the acceptable range atmost of the receiver locationsclosertothemihrabandaroundthemuezzinplatform ThevaluesobtainedaresuitableforbothmusicandspeechThisindicatesanadequatedistributionofearlyreflectionsinthesereceiverlocationsValuesbegintodecreaseunderthepartsofthemaindomethatareclosertothesidewalls

Definition (D50)Definitionisthemeasurederivedfromtheearrsquosresponseto

consecutiveimpulsesitcharacterizestheratioofeffectiveenergytothetotalenergyinanimpulseresponseupto50msThereisagoodrelationshipbetweendefinitionandspeechintelligibilityThisparametershouldbegreaterthan20tosatisfybothmusicandspeechperformances(Templeton1993)

In thefirst activitymodeD5013occ is above30 in thesecondactivitymodeD5033occvariesbetween22and50(see Appendix M4) D5013occ and D5033occ are acceptableunderneath thehalf of themaindome that is closer to themihrabEvensurfacesunderneaththesecondarydomesexhibitgoodvaluesThisisprobablycausedbythefirstreflectionsoffthedomesanddiffusedsurfacesofthestalactitesbetweenthecolumnsanddomes Thedomesdefineanacoustical spaceunderneath

Lateral Fraction (LF)Lateralfractiondefinestherelationshipbetweenasenseof

spatialimpressionandthearrivalofreflectedsoundfromwallstothesideofthelistener(Templeton1993)Formusictheparametershouldbebetween01-035andgreaterthan035forallotherpurposes(Barron1993)

BothforthefirstandsecondactivitymodestheLF13occandLF33occ values are above030 (seeAppendixBM5) Bettervaluesareexpectedatpointsclosertothewall surfaces Theworst place in terms of LF13occand LF33occ is underneath themaindomeLF33occisbetternearthesidewallsandaroundthemuezzinplatform

Speech Transmission Index (STI) STI is directly related to speech To ensure good speech

intelligibility theenvelopeof the signal shouldbepreservedallowingthevariousfrequencybandstocontributetospeechqualityThescaleofoptimumvaluesforthesoundtransmissionindexisshowninTable2

ForthefirstactivitymodeSTI13occisaround052whichisconsideredfairforthesecondactivitymodeSTI33occisaround060 which is considered a good class (see Appendix M6)TheresultsshowthatthereisanevendistributionofSTI13occthroughoutthemosque At locations justaroundthesourceor imambettervalues are seen Enclosedbuildingsof largevolumearenotdesignedforoptimal speechcommunicationsotheseresultsaresatisfactoryconsideringthevolumeofthemosque STI33occ results are even better indicating that theintelligibilityof the imamat theminbar is satisfactorywhenthemosqueisfull

Theoptimumvolumeperpersonfortheintelligibilityofthespeech isbetween4-6m(MaekawaampLord1994) WhentheKocatepeMosqueisconsideredfullthevolumeperpersoncorrespondsto45mThusthevolumeofthemosquemeetsacousticalrequirements

Strength (G)Soundpressurelevel(SPL)mapsindicatetheGresultswhen

anomni-directionalsourcetypeandapowerof31dBOctavebands are selected from the appropriate Point Source EditorofODEONsoftwareForsymphonicandliturgicalmusicGvaluesshouldbegreaterthan3dB(Lynge2001)

Thesimulationresultsforthefirstandsecondactivitymodesshow that sound distribution is satisfactory underneath themaindomeandevenunderthesmallerdomes(seeAppendixM7) These smaller domes are effective in distributing thesoundfromtheimamtofurtherspotsG33occvaluesarehigherclosertotheminbarthanintherestofthemosquethisisgiventhattheimamisthesource

Table 2 RelationbetweenscoresofspeechtransmissionqualityandSTI(RASTI)

4

TABLES

Table 1 Sound absorption coefficients of different materials used in Kocatepe Mosque

Material63

Hz

125

Hz

250

Hz

500

Hz

1

kHz

2

kHz

4 k

Hz

8

kHz

Scattering

Factor

Smooth concrete 001 001 001 002 002 002 005 005 01

Marble 001 001 001 001 001 002 002 002 02

Solid wood panel 015 015 010 006 008 010 005 005 01

Double glazing 002 002 006 003 003 002 002 002 01

Solid wooden door 014 014 010 006 008 010 010 010 01

Prayers-11 per msup2 055 055 086 083 087 090 087 087 05

Prayers-13 per msup2 010 010 021 041 065 075 071 071 05

Ceramic tiles 001 001 001 001 002 002 002 002 03

Carpet on concrete 002 002 006 014 037 060 065 065 01

Chandelier 035 035 025 018 012 007 004 004 02

Lime plaster 004 004 005 004 008 004 006 006 01

Table 2 Relation between scores of speech transmission quality and STI (RASTI)

Quality Score STI (RASTI) value

Bad 0 to 032

Poor 032 ndash 045

Fair 045 ndash 060

Good 060 ndash 075

Excellent 075 to 10

27ZuumlhreSuumlandSemihaYilmazer AcousticalCharacteristicsofMosques

ConclusionTwo important criteria in mosque acoustics are an even

distributionofsoundandoptimumreverberationtimesWhenalltheparametersstudiedareconsideredtheresultsshowthattheKocatepeMosqueisagoodplaceforrecitingthemusicalversionof theHolyQuranas itcreatesaritualandtranquilacoustical atmosphere However for the prayer mode themosquehasinadequateintelligibilityofspeechThisisverifiedbytheabruptjumpsinreverberationtimeatlowfrequenciesintheactivitymodesandwhenthemosqueisempty Thereis excessive reverberation and a highbass ratio in the emptymosque However a flat distribution of reverberation onfrequency is important for intelligibilityof speech regardlessofvolumesoundabsorptionshouldbebalancedfordifferentfrequencies Ancient mosques achieved this bass balance byspecialacousticaltreatmentssuchascavityresonators(Kayili2005)

Excessive reverberation may arise when the walls and theceiling of an auditorium are covered with highly reflectivematerialthatcausesadegradationofspeechsoundsespeciallywhenthesurfaceisconcaveThesoundreinforcementsystemwouldincreasethedirectsoundlevelatlisteningpositionsandyieldbetterintelligibility

In the second activity mode the number of worshippersattendingFridayprayergreatlyaffects total soundabsorptioninsidethemosqueandtheparametersfallwithinthedesirableranges Themosqueissuccessful intermsofsoundfocusingconsideringalloftheactivitypatternsInmanynewmosquesldquofocusingrdquoisaseriousproblembutinKocatepethisacousticaldefect is overcome by the proportions and locations of thedomes(ProdiampMarsilio2003Utami2005)Asthemaindome is above the receiver surfacesnoneof theworshippersstayintheprobablefocusingvolumeSecondarydomeshavenofocusingeffectonthecontrarytheyprovidesoundenergytodistantpartsofthemosqueandtheyprovideadiffusedfieldbeneaththem

TheancientMosquesofSinanarestilluniqueandmaybeareferenceandmodelformosquesyettobebuilt(CAHRISMA2007 Kayili 2005 Topaktas 2003) Except for cavityresonators the basic architectural features of Kocatepe areinspiredbysomeoftheSinanMosquesTheinteriortreatmentofthemosqueisimportantinprovidingadivineenvironmentThestalactitesasa transitionalarchitecturalelementbetweendome and columns the ornamentation on the mihrabthe minbar the columns and balconies-- all have not justaestheticappealbutalsoacousticalpurposesespeciallyinhighfrequencies

Itisclearlyobservedfromtheresultsthatneweramosqueswith concrete domes have a greater tendency to keep andenhance the low frequency soundswithin the audience areaThis low frequency interferes with the speech intelligibilityandputsconstraintsontheclarityofthelectureortheordersconveyedbytheImamThematerialsusedinancientmosquesandthecavityresonatortechniquewithvarioussizedglasspotswithin the dome were wise and efficient in providing goodacousticalqualityRecentlydevelopedacousticaldevicessuchasHelmholtzresonatorsprovidenewdesignopportunitiesthesecouldduplicatetheprincipledesigncriteriaofSinanandcouldimprovetheacousticsof futuremosquesaswellasKocatepe

However considering the feasibility of Helmholtz resonatorswithin big spaces such as mosques the authors would liketo introduce micro-perforated panel absorbers to be studiedmore comprehensively in future applications and studies ofmosqueacousticsThisnewtechnologyusesthesamephysicsprincipals asHelmholtz resonators in substantial attenuationof lowfrequencieswhilekeepingthehighfrequencies inthedesiredrangesThearchitecturalapproachandfeelingofthespaces could then be combined with the new technologicalimprovements of this era and architectural acoustics as asciencecouldbegreatlyfurthered

InthisstudythedrawbacksofnewconstructiontechniquesontheacousticsofcontemporarymosqueshavebeenanalyzedTheKocatepeMosqueisoneoftheleadingexamplesoftheOttomanstyle reinforced concrete dome typology in mosque acousticsand architecture The acoustic simulationsdemonstrated thatthismosquetypologyhasstrongaestheticfeaturesbutmoderateacoustical effects in comparison to the authentic OttomanmosquesTheexcessivereverberationinthelowfrequencyrangeoftheenergydecayisfoundtobethebasicproblemofthatnewtypologyThekeytoimprovetheproblemsintodayrsquosmosquesistoadjustandcombinethematerialtechnologyoftheeratotheaestheticsofancientmosquessuchthatitfitstotheconceptandeventuallytothefunctionalityofthemosquesacoustics

References Abdou AA (2003) Measurement of acoustical characteristics

of mosques in Saudi Arabia The Journal of Acoustical Society of America 113(3)1505-1517

Barron M (1993) Auditorium Acoustics and Architectural DesignLondonEampFNSpon

CAHRISMA(2007)Orsted ndashDTUAvailableathttpserveroersteddtudkwwwoldatcahrismaretrieved4September2007

EganMD(1994)Architectural AcousticsNewYorkMcGraw-HillFaustiPPompoliRampProdiN(2003)Comparingtheacoustics

ofmosquesandByzantinechurches19th International Symposium CIPA (The International Committee for Architectural Photogrammetry) AntalyaTurkeyTechnicalUniversityofIstanbul

HamadahHAampHamoudaHM(1998)AssessmentofspeechintelligibilityinlargeauditoriacasestudyKuwaitStateMosqueApplied Acoustics54(4)273-289

KarabiberZ(1999)AcousticalproblemsinmosquesAcasestudyonthethreemosquesinIstanbulForum Acusticum BerlinCD-ROM4pgs

Karabiber Z (2000) A new approach to an ancient subjectCAHRISMAprojectInProceedings of the 7th International Congress on Sound and Vibration (1661-1668) Garmisch-PartekirchenGermanyInternationalInstituteofAcousticsandVibration

KarabiberZampErdoganS(2002)Comparisonoftheacousticalpropertiesofanancientandarecentmosque Forum AcusticumSeville Spain Spanish Acoustical Society (SEA) Available athttpwwwcicaes~forum2002retrieved4September2007

KayiliM(2000)UseofcavityresonatorsinAnatoliasinceVitruviusInternational Institute of Acoustics and Vibration 31621-1628

KayiliM(2002)AnadolursquodagelenekselakustiksistemlerveMimarSinan uygulamaları (Anatolian traditional acoustic works andworksdonebySinantheArchitect)[Turkish]InProceeding of 6th National Acoustical Congress TAKDER 26-28 October (233-238)AntalyaTurkeyTurkishAcousticalSociety

ArchitecturalScienceReview Volume51Number1March200824

Different materials at different sections of the mosquewith different sound absorption coefficients are assigned(Table 1) After fixing the calculation parameters theselected receiver surfaces are divided into grids assigning096msup2 for eachpersonpraying (Abdou2003) Inorderto get the results of different acoustical parameters andtheir distribution throughout the mosque the maps andcumulativedistributiongraphsforcalculatedparametersareobtainedforthesesurfaces

The process is repeated for different activities takingplace in the Kocatepe Mosque First reverberation time ismeasuredintheemptymosqueThefirstactivitymodeistheprayermodeforthisthesourceistheimamatthemihrabfacing towards the mihrab set at a height of 080m Fordaily prayers the mosque is assumed one third full ThesecondmodeisthepreachingmodeandrecitalofthemusicalversionsoftheHolyQuranThesourceistheimamattheminbarfacingtheworshippersThemosqueisassumedfullfor Friday prayers for listening to the Friday Speech forreligious days and festivals The receiver surfaces for bothactivitymodes are set at aheightof080m approximatelyseating level For the two different activity modes musicand speech related parameters are analyzed these includereverberation time (RT) early decay time (EDT) clarity(C50) definition (D80) speech transmission index (STI)lateralfraction(LF)andstrength(G)

Acoustical Analysis

Reverberation Time (RT) Reverberation time is the time required for a sound ina

roomtodecayby60dBafterthesoundsourcehasstoppedRT is the basic parameter that should be calculated in theassessment of room acoustics RT characterizes the decayofsoundforspecifyingtheefficiencyofaroominfulfillingthe acoustical requirements depending upon the activity orfunctionthattakesplaceForliturgicalpurposes(orchestrachorusororgan)theoptimumrangeforreverberationtimeis30to35sforspaceslargerthan10000m(Egan1994)

InthemosqueRTunoccis748sinthemiddlefrequenciesThisismuchlongerthantheoptimumlevelevenforliturgicalpurposes Long sounddecayat low frequencies can reducetheintelligibilityofspeechForgoodintelligibilityRTvaluesat low octave-band frequencies should remain flat down to100 Hz At low frequencies an increase in RT values ofaround 10-20 would still yield a natural sound but inthecaseofKocatepetheincreaseisabout50intheemptymosque Thebass ratiowhich is theaverageRTat low tomid frequencies is 154 in the Kocatepe Mosque whereasthe optimum range for music performances is 12 to 125(Kuttruff 1991) This value implies an over-warm spacerichinbasssoundthisiseffectiveforspatialimpressionbutdetrimentaltospeechintelligibility

FIGURES

Figure 1 Interior and exterior views from Kocatepe Mosque Ankara

Figure 2 3D display of the mosque

Figure 3 Plan and elevation view of the source and receiver locationsFigure 3 Plan and elevation view of the source and receiver locations

Table 1 SoundabsorptioncoefficientsofdifferentmaterialsusedintheKocatepeMosque

4

TABLES

Table 1 Sound absorption coefficients of different materials used in Kocatepe Mosque

Material63

Hz

125

Hz

250

Hz

500

Hz

1

kHz

2

kHz

4 k

Hz

8

kHz

Scattering

Factor

Smooth concrete 001 001 001 002 002 002 005 005 01

Marble 001 001 001 001 001 002 002 002 02

Solid wood panel 015 015 010 006 008 010 005 005 01

Double glazing 002 002 006 003 003 002 002 002 01

Solid wooden door 014 014 010 006 008 010 010 010 01

Prayers-11 per msup2 055 055 086 083 087 090 087 087 05

Prayers-13 per msup2 010 010 021 041 065 075 071 071 05

Ceramic tiles 001 001 001 001 002 002 002 002 03

Carpet on concrete 002 002 006 014 037 060 065 065 01

Chandelier 035 035 025 018 012 007 004 004 02

Lime plaster 004 004 005 004 008 004 006 006 01

Table 2 Relation between scores of speech transmission quality and STI (RASTI)

Quality Score STI (RASTI) value

Bad 0 to 032

Poor 032 ndash 045

Fair 045 ndash 060

Good 060 ndash 075

Excellent 075 to 10

25ZuumlhreSuumlandSemihaYilmazer AcousticalCharacteristicsofMosques

For the first activity mode the prayer mode in a onethird full mosque the global estimation results show thatthemosquehasanRT13occof435satthemidfrequenciesForthesecondactivitymodethepreacher-recitalmodeinafullmosqueRT33occis319sat themid frequencies (Figure5) Distribution maps illustrate that 90 of the receiversurfaceshasanRT13occof62sandRT33occ281to389s(seeAppendixM1)

RT13occismuchbetterthanRTunoccbutitisstillhigherthantheoptimumrangeAnyvaluegreaterthan10sisespeciallydetrimental to the intelligibilityof speech (Kuttruff1991)The values obtained are unsatisfactory for both music andspeechApositivepointworthmentioningisthatthesoundfield isnearlyeven Therearea fewechospotsunderneaththedomebutthisisacceptablegiventhedimensionsofthedomeRT33occisintheoptimumrangeduringtherecitalsofthemusicalversionoftheHolyQuranbutinthepreachingmodeitreducestheintelligibilityofspeech

ThenumberofworshippersattendingFridayprayergreatlyaffectsthetotalsoundabsorptioninsidethemosqueandtheduration reverberation time becomes more controlled Inthese conditions RT33occ is satisfactory for liturgical musicunderneath the main and secondary domes and there arealmostnoechospotsdespitethehugeconcavesurfaceaboveThis results from the dimensions of the domes and theirheightsabovethefloorWhenthemaindomeiscompletedtoasphereitisfarabovethereceiversurfacesThereforenoneof the receiver surfaces is within this problematic focusingvolume

Early Decay Time (EDT)EDTisroughlydefinedasthesensationofRTItisabetter

indicatorofthesubjectivefeelingofreverberationtimeandiscriticalindefiningtheacousticalqualityofamusichallThe

2

Figure 4 3D view of the source and receiver locationsFigure 4 3D view of the source and receiver locations

Figure 5 Calculated values for RTunocc and RT13occ and RT33occ for frequencies from 125 to 4000 Hz shaded area optimum RT

(Egan 1994)

3

125 Hz 250 Hz 500 Hz 1 kHz 2 kHz 4 kHz

RTunocc 171 124 943 554 378 250

RT13occ 1199 765 311 360 299 221

RT33occ 545 338 335 304 274 203

Figure 5 Calculated values for RTunocc and RT13occ and RT33occ for frequencies from 125 to

4000 Hz shaded area optimum RT (Egan 1994)

3

125 Hz 250 Hz 500 Hz 1 kHz 2 kHz 4 kHz

RTunocc 171 124 943 554 378 250

RT13occ 1199 765 311 360 299 221

RT33occ 545 338 335 304 274 203

Figure 5 Calculated values for RTunocc and RT13occ and RT33occ for frequencies from 125 to

4000 Hz shaded area optimum RT (Egan 1994)

ArchitecturalScienceReview Volume51Number1March200826

EDT should not be higher than plusmn10 for good acousticsForliturgicalpurposesEDTshouldbebetween27and385s(Templeton1993)

For the first activity (ie the 13 full prayer mode)EDT13occ is 63 s and for the second activity (ie the fullpreacher-recital mode) EDT33occ is 44 s (see Appendix BM2) Neitherof these is intheoptimumrange HowevertheyareconsistentwithRT13occandRT33occimplyingagooddistributionofsoundthroughoutthemosqueEventhoughnotashighscatteringcoefficientsastobeexpectedareassignedtosurfaceslikecolumnheadsstalactitesandornamentationaroundmihrabminbarandbalconieswithinthesimulationconsidering their relatively small modeled size there is stillaproperamountofdiffusioninthemosquethateliminateshotanddeadspotsandprovidesevendistributionofsoundespecially in high frequencies Using higher scatteringcoefficientsforthosesmallsurfaceswouldprobablybemoreaccurateandcausemuchmoredrasticresultsandmuchofhomogeneityofsoundfield

Clarity (C80) Clarity is defined technically as the ratio of early sound

energy (arriving within 80 ms of direct sound) to late orreverberant sound energy (arriving more than 80 ms afterthe direct sound) This quality characterizes the separationin time of sounds from individual instruments or groups ofinstruments Thisparametermustbewithincertainlimits ifthemusicaldetailsaretobeheard Ingeneralasatisfactoryacousticalspaceshouldhaveclaritybetween-2and+2dBtosatisfybothmusicandspeechcriteriaandbetween-1and+3dBforchoralmusic(Kuttruff1991)

InthefirstactivitymodeC8013occisbetween-4and+4dBinthesecondactivitymodeC8033occisat+18dBfor90ofthereceiverpoints(seeAppendixM3)C8013occvaluesarewithin the acceptable range atmost of the receiver locationsclosertothemihrabandaroundthemuezzinplatform ThevaluesobtainedaresuitableforbothmusicandspeechThisindicatesanadequatedistributionofearlyreflectionsinthesereceiverlocationsValuesbegintodecreaseunderthepartsofthemaindomethatareclosertothesidewalls

Definition (D50)Definitionisthemeasurederivedfromtheearrsquosresponseto

consecutiveimpulsesitcharacterizestheratioofeffectiveenergytothetotalenergyinanimpulseresponseupto50msThereisagoodrelationshipbetweendefinitionandspeechintelligibilityThisparametershouldbegreaterthan20tosatisfybothmusicandspeechperformances(Templeton1993)

In thefirst activitymodeD5013occ is above30 in thesecondactivitymodeD5033occvariesbetween22and50(see Appendix M4) D5013occ and D5033occ are acceptableunderneath thehalf of themaindome that is closer to themihrabEvensurfacesunderneaththesecondarydomesexhibitgoodvaluesThisisprobablycausedbythefirstreflectionsoffthedomesanddiffusedsurfacesofthestalactitesbetweenthecolumnsanddomes Thedomesdefineanacoustical spaceunderneath

Lateral Fraction (LF)Lateralfractiondefinestherelationshipbetweenasenseof

spatialimpressionandthearrivalofreflectedsoundfromwallstothesideofthelistener(Templeton1993)Formusictheparametershouldbebetween01-035andgreaterthan035forallotherpurposes(Barron1993)

BothforthefirstandsecondactivitymodestheLF13occandLF33occ values are above030 (seeAppendixBM5) Bettervaluesareexpectedatpointsclosertothewall surfaces Theworst place in terms of LF13occand LF33occ is underneath themaindomeLF33occisbetternearthesidewallsandaroundthemuezzinplatform

Speech Transmission Index (STI) STI is directly related to speech To ensure good speech

intelligibility theenvelopeof the signal shouldbepreservedallowingthevariousfrequencybandstocontributetospeechqualityThescaleofoptimumvaluesforthesoundtransmissionindexisshowninTable2

ForthefirstactivitymodeSTI13occisaround052whichisconsideredfairforthesecondactivitymodeSTI33occisaround060 which is considered a good class (see Appendix M6)TheresultsshowthatthereisanevendistributionofSTI13occthroughoutthemosque At locations justaroundthesourceor imambettervalues are seen Enclosedbuildingsof largevolumearenotdesignedforoptimal speechcommunicationsotheseresultsaresatisfactoryconsideringthevolumeofthemosque STI33occ results are even better indicating that theintelligibilityof the imamat theminbar is satisfactorywhenthemosqueisfull

Theoptimumvolumeperpersonfortheintelligibilityofthespeech isbetween4-6m(MaekawaampLord1994) WhentheKocatepeMosqueisconsideredfullthevolumeperpersoncorrespondsto45mThusthevolumeofthemosquemeetsacousticalrequirements

Strength (G)Soundpressurelevel(SPL)mapsindicatetheGresultswhen

anomni-directionalsourcetypeandapowerof31dBOctavebands are selected from the appropriate Point Source EditorofODEONsoftwareForsymphonicandliturgicalmusicGvaluesshouldbegreaterthan3dB(Lynge2001)

Thesimulationresultsforthefirstandsecondactivitymodesshow that sound distribution is satisfactory underneath themaindomeandevenunderthesmallerdomes(seeAppendixM7) These smaller domes are effective in distributing thesoundfromtheimamtofurtherspotsG33occvaluesarehigherclosertotheminbarthanintherestofthemosquethisisgiventhattheimamisthesource

Table 2 RelationbetweenscoresofspeechtransmissionqualityandSTI(RASTI)

4

TABLES

Table 1 Sound absorption coefficients of different materials used in Kocatepe Mosque

Material63

Hz

125

Hz

250

Hz

500

Hz

1

kHz

2

kHz

4 k

Hz

8

kHz

Scattering

Factor

Smooth concrete 001 001 001 002 002 002 005 005 01

Marble 001 001 001 001 001 002 002 002 02

Solid wood panel 015 015 010 006 008 010 005 005 01

Double glazing 002 002 006 003 003 002 002 002 01

Solid wooden door 014 014 010 006 008 010 010 010 01

Prayers-11 per msup2 055 055 086 083 087 090 087 087 05

Prayers-13 per msup2 010 010 021 041 065 075 071 071 05

Ceramic tiles 001 001 001 001 002 002 002 002 03

Carpet on concrete 002 002 006 014 037 060 065 065 01

Chandelier 035 035 025 018 012 007 004 004 02

Lime plaster 004 004 005 004 008 004 006 006 01

Table 2 Relation between scores of speech transmission quality and STI (RASTI)

Quality Score STI (RASTI) value

Bad 0 to 032

Poor 032 ndash 045

Fair 045 ndash 060

Good 060 ndash 075

Excellent 075 to 10

27ZuumlhreSuumlandSemihaYilmazer AcousticalCharacteristicsofMosques

ConclusionTwo important criteria in mosque acoustics are an even

distributionofsoundandoptimumreverberationtimesWhenalltheparametersstudiedareconsideredtheresultsshowthattheKocatepeMosqueisagoodplaceforrecitingthemusicalversionof theHolyQuranas itcreatesaritualandtranquilacoustical atmosphere However for the prayer mode themosquehasinadequateintelligibilityofspeechThisisverifiedbytheabruptjumpsinreverberationtimeatlowfrequenciesintheactivitymodesandwhenthemosqueisempty Thereis excessive reverberation and a highbass ratio in the emptymosque However a flat distribution of reverberation onfrequency is important for intelligibilityof speech regardlessofvolumesoundabsorptionshouldbebalancedfordifferentfrequencies Ancient mosques achieved this bass balance byspecialacousticaltreatmentssuchascavityresonators(Kayili2005)

Excessive reverberation may arise when the walls and theceiling of an auditorium are covered with highly reflectivematerialthatcausesadegradationofspeechsoundsespeciallywhenthesurfaceisconcaveThesoundreinforcementsystemwouldincreasethedirectsoundlevelatlisteningpositionsandyieldbetterintelligibility

In the second activity mode the number of worshippersattendingFridayprayergreatlyaffects total soundabsorptioninsidethemosqueandtheparametersfallwithinthedesirableranges Themosqueissuccessful intermsofsoundfocusingconsideringalloftheactivitypatternsInmanynewmosquesldquofocusingrdquoisaseriousproblembutinKocatepethisacousticaldefect is overcome by the proportions and locations of thedomes(ProdiampMarsilio2003Utami2005)Asthemaindome is above the receiver surfacesnoneof theworshippersstayintheprobablefocusingvolumeSecondarydomeshavenofocusingeffectonthecontrarytheyprovidesoundenergytodistantpartsofthemosqueandtheyprovideadiffusedfieldbeneaththem

TheancientMosquesofSinanarestilluniqueandmaybeareferenceandmodelformosquesyettobebuilt(CAHRISMA2007 Kayili 2005 Topaktas 2003) Except for cavityresonators the basic architectural features of Kocatepe areinspiredbysomeoftheSinanMosquesTheinteriortreatmentofthemosqueisimportantinprovidingadivineenvironmentThestalactitesasa transitionalarchitecturalelementbetweendome and columns the ornamentation on the mihrabthe minbar the columns and balconies-- all have not justaestheticappealbutalsoacousticalpurposesespeciallyinhighfrequencies

Itisclearlyobservedfromtheresultsthatneweramosqueswith concrete domes have a greater tendency to keep andenhance the low frequency soundswithin the audience areaThis low frequency interferes with the speech intelligibilityandputsconstraintsontheclarityofthelectureortheordersconveyedbytheImamThematerialsusedinancientmosquesandthecavityresonatortechniquewithvarioussizedglasspotswithin the dome were wise and efficient in providing goodacousticalqualityRecentlydevelopedacousticaldevicessuchasHelmholtzresonatorsprovidenewdesignopportunitiesthesecouldduplicatetheprincipledesigncriteriaofSinanandcouldimprovetheacousticsof futuremosquesaswellasKocatepe

However considering the feasibility of Helmholtz resonatorswithin big spaces such as mosques the authors would liketo introduce micro-perforated panel absorbers to be studiedmore comprehensively in future applications and studies ofmosqueacousticsThisnewtechnologyusesthesamephysicsprincipals asHelmholtz resonators in substantial attenuationof lowfrequencieswhilekeepingthehighfrequencies inthedesiredrangesThearchitecturalapproachandfeelingofthespaces could then be combined with the new technologicalimprovements of this era and architectural acoustics as asciencecouldbegreatlyfurthered

InthisstudythedrawbacksofnewconstructiontechniquesontheacousticsofcontemporarymosqueshavebeenanalyzedTheKocatepeMosqueisoneoftheleadingexamplesoftheOttomanstyle reinforced concrete dome typology in mosque acousticsand architecture The acoustic simulationsdemonstrated thatthismosquetypologyhasstrongaestheticfeaturesbutmoderateacoustical effects in comparison to the authentic OttomanmosquesTheexcessivereverberationinthelowfrequencyrangeoftheenergydecayisfoundtobethebasicproblemofthatnewtypologyThekeytoimprovetheproblemsintodayrsquosmosquesistoadjustandcombinethematerialtechnologyoftheeratotheaestheticsofancientmosquessuchthatitfitstotheconceptandeventuallytothefunctionalityofthemosquesacoustics

References Abdou AA (2003) Measurement of acoustical characteristics

of mosques in Saudi Arabia The Journal of Acoustical Society of America 113(3)1505-1517

Barron M (1993) Auditorium Acoustics and Architectural DesignLondonEampFNSpon

CAHRISMA(2007)Orsted ndashDTUAvailableathttpserveroersteddtudkwwwoldatcahrismaretrieved4September2007

EganMD(1994)Architectural AcousticsNewYorkMcGraw-HillFaustiPPompoliRampProdiN(2003)Comparingtheacoustics

ofmosquesandByzantinechurches19th International Symposium CIPA (The International Committee for Architectural Photogrammetry) AntalyaTurkeyTechnicalUniversityofIstanbul

HamadahHAampHamoudaHM(1998)AssessmentofspeechintelligibilityinlargeauditoriacasestudyKuwaitStateMosqueApplied Acoustics54(4)273-289

KarabiberZ(1999)AcousticalproblemsinmosquesAcasestudyonthethreemosquesinIstanbulForum Acusticum BerlinCD-ROM4pgs

Karabiber Z (2000) A new approach to an ancient subjectCAHRISMAprojectInProceedings of the 7th International Congress on Sound and Vibration (1661-1668) Garmisch-PartekirchenGermanyInternationalInstituteofAcousticsandVibration

KarabiberZampErdoganS(2002)Comparisonoftheacousticalpropertiesofanancientandarecentmosque Forum AcusticumSeville Spain Spanish Acoustical Society (SEA) Available athttpwwwcicaes~forum2002retrieved4September2007

KayiliM(2000)UseofcavityresonatorsinAnatoliasinceVitruviusInternational Institute of Acoustics and Vibration 31621-1628

KayiliM(2002)AnadolursquodagelenekselakustiksistemlerveMimarSinan uygulamaları (Anatolian traditional acoustic works andworksdonebySinantheArchitect)[Turkish]InProceeding of 6th National Acoustical Congress TAKDER 26-28 October (233-238)AntalyaTurkeyTurkishAcousticalSociety

25ZuumlhreSuumlandSemihaYilmazer AcousticalCharacteristicsofMosques

For the first activity mode the prayer mode in a onethird full mosque the global estimation results show thatthemosquehasanRT13occof435satthemidfrequenciesForthesecondactivitymodethepreacher-recitalmodeinafullmosqueRT33occis319sat themid frequencies (Figure5) Distribution maps illustrate that 90 of the receiversurfaceshasanRT13occof62sandRT33occ281to389s(seeAppendixM1)

RT13occismuchbetterthanRTunoccbutitisstillhigherthantheoptimumrangeAnyvaluegreaterthan10sisespeciallydetrimental to the intelligibilityof speech (Kuttruff1991)The values obtained are unsatisfactory for both music andspeechApositivepointworthmentioningisthatthesoundfield isnearlyeven Therearea fewechospotsunderneaththedomebutthisisacceptablegiventhedimensionsofthedomeRT33occisintheoptimumrangeduringtherecitalsofthemusicalversionoftheHolyQuranbutinthepreachingmodeitreducestheintelligibilityofspeech

ThenumberofworshippersattendingFridayprayergreatlyaffectsthetotalsoundabsorptioninsidethemosqueandtheduration reverberation time becomes more controlled Inthese conditions RT33occ is satisfactory for liturgical musicunderneath the main and secondary domes and there arealmostnoechospotsdespitethehugeconcavesurfaceaboveThis results from the dimensions of the domes and theirheightsabovethefloorWhenthemaindomeiscompletedtoasphereitisfarabovethereceiversurfacesThereforenoneof the receiver surfaces is within this problematic focusingvolume

Early Decay Time (EDT)EDTisroughlydefinedasthesensationofRTItisabetter

indicatorofthesubjectivefeelingofreverberationtimeandiscriticalindefiningtheacousticalqualityofamusichallThe

2

Figure 4 3D view of the source and receiver locationsFigure 4 3D view of the source and receiver locations

Figure 5 Calculated values for RTunocc and RT13occ and RT33occ for frequencies from 125 to 4000 Hz shaded area optimum RT

(Egan 1994)

3

125 Hz 250 Hz 500 Hz 1 kHz 2 kHz 4 kHz

RTunocc 171 124 943 554 378 250

RT13occ 1199 765 311 360 299 221

RT33occ 545 338 335 304 274 203

Figure 5 Calculated values for RTunocc and RT13occ and RT33occ for frequencies from 125 to

4000 Hz shaded area optimum RT (Egan 1994)

3

125 Hz 250 Hz 500 Hz 1 kHz 2 kHz 4 kHz

RTunocc 171 124 943 554 378 250

RT13occ 1199 765 311 360 299 221

RT33occ 545 338 335 304 274 203

Figure 5 Calculated values for RTunocc and RT13occ and RT33occ for frequencies from 125 to

4000 Hz shaded area optimum RT (Egan 1994)

ArchitecturalScienceReview Volume51Number1March200826

EDT should not be higher than plusmn10 for good acousticsForliturgicalpurposesEDTshouldbebetween27and385s(Templeton1993)

For the first activity (ie the 13 full prayer mode)EDT13occ is 63 s and for the second activity (ie the fullpreacher-recital mode) EDT33occ is 44 s (see Appendix BM2) Neitherof these is intheoptimumrange HowevertheyareconsistentwithRT13occandRT33occimplyingagooddistributionofsoundthroughoutthemosqueEventhoughnotashighscatteringcoefficientsastobeexpectedareassignedtosurfaceslikecolumnheadsstalactitesandornamentationaroundmihrabminbarandbalconieswithinthesimulationconsidering their relatively small modeled size there is stillaproperamountofdiffusioninthemosquethateliminateshotanddeadspotsandprovidesevendistributionofsoundespecially in high frequencies Using higher scatteringcoefficientsforthosesmallsurfaceswouldprobablybemoreaccurateandcausemuchmoredrasticresultsandmuchofhomogeneityofsoundfield

Clarity (C80) Clarity is defined technically as the ratio of early sound

energy (arriving within 80 ms of direct sound) to late orreverberant sound energy (arriving more than 80 ms afterthe direct sound) This quality characterizes the separationin time of sounds from individual instruments or groups ofinstruments Thisparametermustbewithincertainlimits ifthemusicaldetailsaretobeheard Ingeneralasatisfactoryacousticalspaceshouldhaveclaritybetween-2and+2dBtosatisfybothmusicandspeechcriteriaandbetween-1and+3dBforchoralmusic(Kuttruff1991)

InthefirstactivitymodeC8013occisbetween-4and+4dBinthesecondactivitymodeC8033occisat+18dBfor90ofthereceiverpoints(seeAppendixM3)C8013occvaluesarewithin the acceptable range atmost of the receiver locationsclosertothemihrabandaroundthemuezzinplatform ThevaluesobtainedaresuitableforbothmusicandspeechThisindicatesanadequatedistributionofearlyreflectionsinthesereceiverlocationsValuesbegintodecreaseunderthepartsofthemaindomethatareclosertothesidewalls

Definition (D50)Definitionisthemeasurederivedfromtheearrsquosresponseto

consecutiveimpulsesitcharacterizestheratioofeffectiveenergytothetotalenergyinanimpulseresponseupto50msThereisagoodrelationshipbetweendefinitionandspeechintelligibilityThisparametershouldbegreaterthan20tosatisfybothmusicandspeechperformances(Templeton1993)

In thefirst activitymodeD5013occ is above30 in thesecondactivitymodeD5033occvariesbetween22and50(see Appendix M4) D5013occ and D5033occ are acceptableunderneath thehalf of themaindome that is closer to themihrabEvensurfacesunderneaththesecondarydomesexhibitgoodvaluesThisisprobablycausedbythefirstreflectionsoffthedomesanddiffusedsurfacesofthestalactitesbetweenthecolumnsanddomes Thedomesdefineanacoustical spaceunderneath

Lateral Fraction (LF)Lateralfractiondefinestherelationshipbetweenasenseof

spatialimpressionandthearrivalofreflectedsoundfromwallstothesideofthelistener(Templeton1993)Formusictheparametershouldbebetween01-035andgreaterthan035forallotherpurposes(Barron1993)

BothforthefirstandsecondactivitymodestheLF13occandLF33occ values are above030 (seeAppendixBM5) Bettervaluesareexpectedatpointsclosertothewall surfaces Theworst place in terms of LF13occand LF33occ is underneath themaindomeLF33occisbetternearthesidewallsandaroundthemuezzinplatform

Speech Transmission Index (STI) STI is directly related to speech To ensure good speech

intelligibility theenvelopeof the signal shouldbepreservedallowingthevariousfrequencybandstocontributetospeechqualityThescaleofoptimumvaluesforthesoundtransmissionindexisshowninTable2

ForthefirstactivitymodeSTI13occisaround052whichisconsideredfairforthesecondactivitymodeSTI33occisaround060 which is considered a good class (see Appendix M6)TheresultsshowthatthereisanevendistributionofSTI13occthroughoutthemosque At locations justaroundthesourceor imambettervalues are seen Enclosedbuildingsof largevolumearenotdesignedforoptimal speechcommunicationsotheseresultsaresatisfactoryconsideringthevolumeofthemosque STI33occ results are even better indicating that theintelligibilityof the imamat theminbar is satisfactorywhenthemosqueisfull

Theoptimumvolumeperpersonfortheintelligibilityofthespeech isbetween4-6m(MaekawaampLord1994) WhentheKocatepeMosqueisconsideredfullthevolumeperpersoncorrespondsto45mThusthevolumeofthemosquemeetsacousticalrequirements

Strength (G)Soundpressurelevel(SPL)mapsindicatetheGresultswhen

anomni-directionalsourcetypeandapowerof31dBOctavebands are selected from the appropriate Point Source EditorofODEONsoftwareForsymphonicandliturgicalmusicGvaluesshouldbegreaterthan3dB(Lynge2001)

Thesimulationresultsforthefirstandsecondactivitymodesshow that sound distribution is satisfactory underneath themaindomeandevenunderthesmallerdomes(seeAppendixM7) These smaller domes are effective in distributing thesoundfromtheimamtofurtherspotsG33occvaluesarehigherclosertotheminbarthanintherestofthemosquethisisgiventhattheimamisthesource

Table 2 RelationbetweenscoresofspeechtransmissionqualityandSTI(RASTI)

4

TABLES

Table 1 Sound absorption coefficients of different materials used in Kocatepe Mosque

Material63

Hz

125

Hz

250

Hz

500

Hz

1

kHz

2

kHz

4 k

Hz

8

kHz

Scattering

Factor

Smooth concrete 001 001 001 002 002 002 005 005 01

Marble 001 001 001 001 001 002 002 002 02

Solid wood panel 015 015 010 006 008 010 005 005 01

Double glazing 002 002 006 003 003 002 002 002 01

Solid wooden door 014 014 010 006 008 010 010 010 01

Prayers-11 per msup2 055 055 086 083 087 090 087 087 05

Prayers-13 per msup2 010 010 021 041 065 075 071 071 05

Ceramic tiles 001 001 001 001 002 002 002 002 03

Carpet on concrete 002 002 006 014 037 060 065 065 01

Chandelier 035 035 025 018 012 007 004 004 02

Lime plaster 004 004 005 004 008 004 006 006 01

Table 2 Relation between scores of speech transmission quality and STI (RASTI)

Quality Score STI (RASTI) value

Bad 0 to 032

Poor 032 ndash 045

Fair 045 ndash 060

Good 060 ndash 075

Excellent 075 to 10

27ZuumlhreSuumlandSemihaYilmazer AcousticalCharacteristicsofMosques

ConclusionTwo important criteria in mosque acoustics are an even

distributionofsoundandoptimumreverberationtimesWhenalltheparametersstudiedareconsideredtheresultsshowthattheKocatepeMosqueisagoodplaceforrecitingthemusicalversionof theHolyQuranas itcreatesaritualandtranquilacoustical atmosphere However for the prayer mode themosquehasinadequateintelligibilityofspeechThisisverifiedbytheabruptjumpsinreverberationtimeatlowfrequenciesintheactivitymodesandwhenthemosqueisempty Thereis excessive reverberation and a highbass ratio in the emptymosque However a flat distribution of reverberation onfrequency is important for intelligibilityof speech regardlessofvolumesoundabsorptionshouldbebalancedfordifferentfrequencies Ancient mosques achieved this bass balance byspecialacousticaltreatmentssuchascavityresonators(Kayili2005)

Excessive reverberation may arise when the walls and theceiling of an auditorium are covered with highly reflectivematerialthatcausesadegradationofspeechsoundsespeciallywhenthesurfaceisconcaveThesoundreinforcementsystemwouldincreasethedirectsoundlevelatlisteningpositionsandyieldbetterintelligibility

In the second activity mode the number of worshippersattendingFridayprayergreatlyaffects total soundabsorptioninsidethemosqueandtheparametersfallwithinthedesirableranges Themosqueissuccessful intermsofsoundfocusingconsideringalloftheactivitypatternsInmanynewmosquesldquofocusingrdquoisaseriousproblembutinKocatepethisacousticaldefect is overcome by the proportions and locations of thedomes(ProdiampMarsilio2003Utami2005)Asthemaindome is above the receiver surfacesnoneof theworshippersstayintheprobablefocusingvolumeSecondarydomeshavenofocusingeffectonthecontrarytheyprovidesoundenergytodistantpartsofthemosqueandtheyprovideadiffusedfieldbeneaththem

TheancientMosquesofSinanarestilluniqueandmaybeareferenceandmodelformosquesyettobebuilt(CAHRISMA2007 Kayili 2005 Topaktas 2003) Except for cavityresonators the basic architectural features of Kocatepe areinspiredbysomeoftheSinanMosquesTheinteriortreatmentofthemosqueisimportantinprovidingadivineenvironmentThestalactitesasa transitionalarchitecturalelementbetweendome and columns the ornamentation on the mihrabthe minbar the columns and balconies-- all have not justaestheticappealbutalsoacousticalpurposesespeciallyinhighfrequencies

Itisclearlyobservedfromtheresultsthatneweramosqueswith concrete domes have a greater tendency to keep andenhance the low frequency soundswithin the audience areaThis low frequency interferes with the speech intelligibilityandputsconstraintsontheclarityofthelectureortheordersconveyedbytheImamThematerialsusedinancientmosquesandthecavityresonatortechniquewithvarioussizedglasspotswithin the dome were wise and efficient in providing goodacousticalqualityRecentlydevelopedacousticaldevicessuchasHelmholtzresonatorsprovidenewdesignopportunitiesthesecouldduplicatetheprincipledesigncriteriaofSinanandcouldimprovetheacousticsof futuremosquesaswellasKocatepe

However considering the feasibility of Helmholtz resonatorswithin big spaces such as mosques the authors would liketo introduce micro-perforated panel absorbers to be studiedmore comprehensively in future applications and studies ofmosqueacousticsThisnewtechnologyusesthesamephysicsprincipals asHelmholtz resonators in substantial attenuationof lowfrequencieswhilekeepingthehighfrequencies inthedesiredrangesThearchitecturalapproachandfeelingofthespaces could then be combined with the new technologicalimprovements of this era and architectural acoustics as asciencecouldbegreatlyfurthered

InthisstudythedrawbacksofnewconstructiontechniquesontheacousticsofcontemporarymosqueshavebeenanalyzedTheKocatepeMosqueisoneoftheleadingexamplesoftheOttomanstyle reinforced concrete dome typology in mosque acousticsand architecture The acoustic simulationsdemonstrated thatthismosquetypologyhasstrongaestheticfeaturesbutmoderateacoustical effects in comparison to the authentic OttomanmosquesTheexcessivereverberationinthelowfrequencyrangeoftheenergydecayisfoundtobethebasicproblemofthatnewtypologyThekeytoimprovetheproblemsintodayrsquosmosquesistoadjustandcombinethematerialtechnologyoftheeratotheaestheticsofancientmosquessuchthatitfitstotheconceptandeventuallytothefunctionalityofthemosquesacoustics

References Abdou AA (2003) Measurement of acoustical characteristics

of mosques in Saudi Arabia The Journal of Acoustical Society of America 113(3)1505-1517

Barron M (1993) Auditorium Acoustics and Architectural DesignLondonEampFNSpon

CAHRISMA(2007)Orsted ndashDTUAvailableathttpserveroersteddtudkwwwoldatcahrismaretrieved4September2007

EganMD(1994)Architectural AcousticsNewYorkMcGraw-HillFaustiPPompoliRampProdiN(2003)Comparingtheacoustics

ofmosquesandByzantinechurches19th International Symposium CIPA (The International Committee for Architectural Photogrammetry) AntalyaTurkeyTechnicalUniversityofIstanbul

HamadahHAampHamoudaHM(1998)AssessmentofspeechintelligibilityinlargeauditoriacasestudyKuwaitStateMosqueApplied Acoustics54(4)273-289

KarabiberZ(1999)AcousticalproblemsinmosquesAcasestudyonthethreemosquesinIstanbulForum Acusticum BerlinCD-ROM4pgs

Karabiber Z (2000) A new approach to an ancient subjectCAHRISMAprojectInProceedings of the 7th International Congress on Sound and Vibration (1661-1668) Garmisch-PartekirchenGermanyInternationalInstituteofAcousticsandVibration

KarabiberZampErdoganS(2002)Comparisonoftheacousticalpropertiesofanancientandarecentmosque Forum AcusticumSeville Spain Spanish Acoustical Society (SEA) Available athttpwwwcicaes~forum2002retrieved4September2007

KayiliM(2000)UseofcavityresonatorsinAnatoliasinceVitruviusInternational Institute of Acoustics and Vibration 31621-1628

KayiliM(2002)AnadolursquodagelenekselakustiksistemlerveMimarSinan uygulamaları (Anatolian traditional acoustic works andworksdonebySinantheArchitect)[Turkish]InProceeding of 6th National Acoustical Congress TAKDER 26-28 October (233-238)AntalyaTurkeyTurkishAcousticalSociety

ArchitecturalScienceReview Volume51Number1March200826

EDT should not be higher than plusmn10 for good acousticsForliturgicalpurposesEDTshouldbebetween27and385s(Templeton1993)

For the first activity (ie the 13 full prayer mode)EDT13occ is 63 s and for the second activity (ie the fullpreacher-recital mode) EDT33occ is 44 s (see Appendix BM2) Neitherof these is intheoptimumrange HowevertheyareconsistentwithRT13occandRT33occimplyingagooddistributionofsoundthroughoutthemosqueEventhoughnotashighscatteringcoefficientsastobeexpectedareassignedtosurfaceslikecolumnheadsstalactitesandornamentationaroundmihrabminbarandbalconieswithinthesimulationconsidering their relatively small modeled size there is stillaproperamountofdiffusioninthemosquethateliminateshotanddeadspotsandprovidesevendistributionofsoundespecially in high frequencies Using higher scatteringcoefficientsforthosesmallsurfaceswouldprobablybemoreaccurateandcausemuchmoredrasticresultsandmuchofhomogeneityofsoundfield

Clarity (C80) Clarity is defined technically as the ratio of early sound

energy (arriving within 80 ms of direct sound) to late orreverberant sound energy (arriving more than 80 ms afterthe direct sound) This quality characterizes the separationin time of sounds from individual instruments or groups ofinstruments Thisparametermustbewithincertainlimits ifthemusicaldetailsaretobeheard Ingeneralasatisfactoryacousticalspaceshouldhaveclaritybetween-2and+2dBtosatisfybothmusicandspeechcriteriaandbetween-1and+3dBforchoralmusic(Kuttruff1991)

InthefirstactivitymodeC8013occisbetween-4and+4dBinthesecondactivitymodeC8033occisat+18dBfor90ofthereceiverpoints(seeAppendixM3)C8013occvaluesarewithin the acceptable range atmost of the receiver locationsclosertothemihrabandaroundthemuezzinplatform ThevaluesobtainedaresuitableforbothmusicandspeechThisindicatesanadequatedistributionofearlyreflectionsinthesereceiverlocationsValuesbegintodecreaseunderthepartsofthemaindomethatareclosertothesidewalls

Definition (D50)Definitionisthemeasurederivedfromtheearrsquosresponseto

consecutiveimpulsesitcharacterizestheratioofeffectiveenergytothetotalenergyinanimpulseresponseupto50msThereisagoodrelationshipbetweendefinitionandspeechintelligibilityThisparametershouldbegreaterthan20tosatisfybothmusicandspeechperformances(Templeton1993)

In thefirst activitymodeD5013occ is above30 in thesecondactivitymodeD5033occvariesbetween22and50(see Appendix M4) D5013occ and D5033occ are acceptableunderneath thehalf of themaindome that is closer to themihrabEvensurfacesunderneaththesecondarydomesexhibitgoodvaluesThisisprobablycausedbythefirstreflectionsoffthedomesanddiffusedsurfacesofthestalactitesbetweenthecolumnsanddomes Thedomesdefineanacoustical spaceunderneath

Lateral Fraction (LF)Lateralfractiondefinestherelationshipbetweenasenseof

spatialimpressionandthearrivalofreflectedsoundfromwallstothesideofthelistener(Templeton1993)Formusictheparametershouldbebetween01-035andgreaterthan035forallotherpurposes(Barron1993)

BothforthefirstandsecondactivitymodestheLF13occandLF33occ values are above030 (seeAppendixBM5) Bettervaluesareexpectedatpointsclosertothewall surfaces Theworst place in terms of LF13occand LF33occ is underneath themaindomeLF33occisbetternearthesidewallsandaroundthemuezzinplatform

Speech Transmission Index (STI) STI is directly related to speech To ensure good speech

intelligibility theenvelopeof the signal shouldbepreservedallowingthevariousfrequencybandstocontributetospeechqualityThescaleofoptimumvaluesforthesoundtransmissionindexisshowninTable2

ForthefirstactivitymodeSTI13occisaround052whichisconsideredfairforthesecondactivitymodeSTI33occisaround060 which is considered a good class (see Appendix M6)TheresultsshowthatthereisanevendistributionofSTI13occthroughoutthemosque At locations justaroundthesourceor imambettervalues are seen Enclosedbuildingsof largevolumearenotdesignedforoptimal speechcommunicationsotheseresultsaresatisfactoryconsideringthevolumeofthemosque STI33occ results are even better indicating that theintelligibilityof the imamat theminbar is satisfactorywhenthemosqueisfull

Theoptimumvolumeperpersonfortheintelligibilityofthespeech isbetween4-6m(MaekawaampLord1994) WhentheKocatepeMosqueisconsideredfullthevolumeperpersoncorrespondsto45mThusthevolumeofthemosquemeetsacousticalrequirements

Strength (G)Soundpressurelevel(SPL)mapsindicatetheGresultswhen

anomni-directionalsourcetypeandapowerof31dBOctavebands are selected from the appropriate Point Source EditorofODEONsoftwareForsymphonicandliturgicalmusicGvaluesshouldbegreaterthan3dB(Lynge2001)

Thesimulationresultsforthefirstandsecondactivitymodesshow that sound distribution is satisfactory underneath themaindomeandevenunderthesmallerdomes(seeAppendixM7) These smaller domes are effective in distributing thesoundfromtheimamtofurtherspotsG33occvaluesarehigherclosertotheminbarthanintherestofthemosquethisisgiventhattheimamisthesource

Table 2 RelationbetweenscoresofspeechtransmissionqualityandSTI(RASTI)

4

TABLES

Table 1 Sound absorption coefficients of different materials used in Kocatepe Mosque

Material63

Hz

125

Hz

250

Hz

500

Hz

1

kHz

2

kHz

4 k

Hz

8

kHz

Scattering

Factor

Smooth concrete 001 001 001 002 002 002 005 005 01

Marble 001 001 001 001 001 002 002 002 02

Solid wood panel 015 015 010 006 008 010 005 005 01

Double glazing 002 002 006 003 003 002 002 002 01

Solid wooden door 014 014 010 006 008 010 010 010 01

Prayers-11 per msup2 055 055 086 083 087 090 087 087 05

Prayers-13 per msup2 010 010 021 041 065 075 071 071 05

Ceramic tiles 001 001 001 001 002 002 002 002 03

Carpet on concrete 002 002 006 014 037 060 065 065 01

Chandelier 035 035 025 018 012 007 004 004 02

Lime plaster 004 004 005 004 008 004 006 006 01

Table 2 Relation between scores of speech transmission quality and STI (RASTI)

Quality Score STI (RASTI) value

Bad 0 to 032

Poor 032 ndash 045

Fair 045 ndash 060

Good 060 ndash 075

Excellent 075 to 10

27ZuumlhreSuumlandSemihaYilmazer AcousticalCharacteristicsofMosques

ConclusionTwo important criteria in mosque acoustics are an even

distributionofsoundandoptimumreverberationtimesWhenalltheparametersstudiedareconsideredtheresultsshowthattheKocatepeMosqueisagoodplaceforrecitingthemusicalversionof theHolyQuranas itcreatesaritualandtranquilacoustical atmosphere However for the prayer mode themosquehasinadequateintelligibilityofspeechThisisverifiedbytheabruptjumpsinreverberationtimeatlowfrequenciesintheactivitymodesandwhenthemosqueisempty Thereis excessive reverberation and a highbass ratio in the emptymosque However a flat distribution of reverberation onfrequency is important for intelligibilityof speech regardlessofvolumesoundabsorptionshouldbebalancedfordifferentfrequencies Ancient mosques achieved this bass balance byspecialacousticaltreatmentssuchascavityresonators(Kayili2005)

Excessive reverberation may arise when the walls and theceiling of an auditorium are covered with highly reflectivematerialthatcausesadegradationofspeechsoundsespeciallywhenthesurfaceisconcaveThesoundreinforcementsystemwouldincreasethedirectsoundlevelatlisteningpositionsandyieldbetterintelligibility

In the second activity mode the number of worshippersattendingFridayprayergreatlyaffects total soundabsorptioninsidethemosqueandtheparametersfallwithinthedesirableranges Themosqueissuccessful intermsofsoundfocusingconsideringalloftheactivitypatternsInmanynewmosquesldquofocusingrdquoisaseriousproblembutinKocatepethisacousticaldefect is overcome by the proportions and locations of thedomes(ProdiampMarsilio2003Utami2005)Asthemaindome is above the receiver surfacesnoneof theworshippersstayintheprobablefocusingvolumeSecondarydomeshavenofocusingeffectonthecontrarytheyprovidesoundenergytodistantpartsofthemosqueandtheyprovideadiffusedfieldbeneaththem

TheancientMosquesofSinanarestilluniqueandmaybeareferenceandmodelformosquesyettobebuilt(CAHRISMA2007 Kayili 2005 Topaktas 2003) Except for cavityresonators the basic architectural features of Kocatepe areinspiredbysomeoftheSinanMosquesTheinteriortreatmentofthemosqueisimportantinprovidingadivineenvironmentThestalactitesasa transitionalarchitecturalelementbetweendome and columns the ornamentation on the mihrabthe minbar the columns and balconies-- all have not justaestheticappealbutalsoacousticalpurposesespeciallyinhighfrequencies

Itisclearlyobservedfromtheresultsthatneweramosqueswith concrete domes have a greater tendency to keep andenhance the low frequency soundswithin the audience areaThis low frequency interferes with the speech intelligibilityandputsconstraintsontheclarityofthelectureortheordersconveyedbytheImamThematerialsusedinancientmosquesandthecavityresonatortechniquewithvarioussizedglasspotswithin the dome were wise and efficient in providing goodacousticalqualityRecentlydevelopedacousticaldevicessuchasHelmholtzresonatorsprovidenewdesignopportunitiesthesecouldduplicatetheprincipledesigncriteriaofSinanandcouldimprovetheacousticsof futuremosquesaswellasKocatepe

However considering the feasibility of Helmholtz resonatorswithin big spaces such as mosques the authors would liketo introduce micro-perforated panel absorbers to be studiedmore comprehensively in future applications and studies ofmosqueacousticsThisnewtechnologyusesthesamephysicsprincipals asHelmholtz resonators in substantial attenuationof lowfrequencieswhilekeepingthehighfrequencies inthedesiredrangesThearchitecturalapproachandfeelingofthespaces could then be combined with the new technologicalimprovements of this era and architectural acoustics as asciencecouldbegreatlyfurthered

InthisstudythedrawbacksofnewconstructiontechniquesontheacousticsofcontemporarymosqueshavebeenanalyzedTheKocatepeMosqueisoneoftheleadingexamplesoftheOttomanstyle reinforced concrete dome typology in mosque acousticsand architecture The acoustic simulationsdemonstrated thatthismosquetypologyhasstrongaestheticfeaturesbutmoderateacoustical effects in comparison to the authentic OttomanmosquesTheexcessivereverberationinthelowfrequencyrangeoftheenergydecayisfoundtobethebasicproblemofthatnewtypologyThekeytoimprovetheproblemsintodayrsquosmosquesistoadjustandcombinethematerialtechnologyoftheeratotheaestheticsofancientmosquessuchthatitfitstotheconceptandeventuallytothefunctionalityofthemosquesacoustics

References Abdou AA (2003) Measurement of acoustical characteristics

of mosques in Saudi Arabia The Journal of Acoustical Society of America 113(3)1505-1517

Barron M (1993) Auditorium Acoustics and Architectural DesignLondonEampFNSpon

CAHRISMA(2007)Orsted ndashDTUAvailableathttpserveroersteddtudkwwwoldatcahrismaretrieved4September2007

EganMD(1994)Architectural AcousticsNewYorkMcGraw-HillFaustiPPompoliRampProdiN(2003)Comparingtheacoustics

ofmosquesandByzantinechurches19th International Symposium CIPA (The International Committee for Architectural Photogrammetry) AntalyaTurkeyTechnicalUniversityofIstanbul

HamadahHAampHamoudaHM(1998)AssessmentofspeechintelligibilityinlargeauditoriacasestudyKuwaitStateMosqueApplied Acoustics54(4)273-289

KarabiberZ(1999)AcousticalproblemsinmosquesAcasestudyonthethreemosquesinIstanbulForum Acusticum BerlinCD-ROM4pgs

Karabiber Z (2000) A new approach to an ancient subjectCAHRISMAprojectInProceedings of the 7th International Congress on Sound and Vibration (1661-1668) Garmisch-PartekirchenGermanyInternationalInstituteofAcousticsandVibration

KarabiberZampErdoganS(2002)Comparisonoftheacousticalpropertiesofanancientandarecentmosque Forum AcusticumSeville Spain Spanish Acoustical Society (SEA) Available athttpwwwcicaes~forum2002retrieved4September2007

KayiliM(2000)UseofcavityresonatorsinAnatoliasinceVitruviusInternational Institute of Acoustics and Vibration 31621-1628

KayiliM(2002)AnadolursquodagelenekselakustiksistemlerveMimarSinan uygulamaları (Anatolian traditional acoustic works andworksdonebySinantheArchitect)[Turkish]InProceeding of 6th National Acoustical Congress TAKDER 26-28 October (233-238)AntalyaTurkeyTurkishAcousticalSociety

27ZuumlhreSuumlandSemihaYilmazer AcousticalCharacteristicsofMosques

ConclusionTwo important criteria in mosque acoustics are an even

distributionofsoundandoptimumreverberationtimesWhenalltheparametersstudiedareconsideredtheresultsshowthattheKocatepeMosqueisagoodplaceforrecitingthemusicalversionof theHolyQuranas itcreatesaritualandtranquilacoustical atmosphere However for the prayer mode themosquehasinadequateintelligibilityofspeechThisisverifiedbytheabruptjumpsinreverberationtimeatlowfrequenciesintheactivitymodesandwhenthemosqueisempty Thereis excessive reverberation and a highbass ratio in the emptymosque However a flat distribution of reverberation onfrequency is important for intelligibilityof speech regardlessofvolumesoundabsorptionshouldbebalancedfordifferentfrequencies Ancient mosques achieved this bass balance byspecialacousticaltreatmentssuchascavityresonators(Kayili2005)

Excessive reverberation may arise when the walls and theceiling of an auditorium are covered with highly reflectivematerialthatcausesadegradationofspeechsoundsespeciallywhenthesurfaceisconcaveThesoundreinforcementsystemwouldincreasethedirectsoundlevelatlisteningpositionsandyieldbetterintelligibility

In the second activity mode the number of worshippersattendingFridayprayergreatlyaffects total soundabsorptioninsidethemosqueandtheparametersfallwithinthedesirableranges Themosqueissuccessful intermsofsoundfocusingconsideringalloftheactivitypatternsInmanynewmosquesldquofocusingrdquoisaseriousproblembutinKocatepethisacousticaldefect is overcome by the proportions and locations of thedomes(ProdiampMarsilio2003Utami2005)Asthemaindome is above the receiver surfacesnoneof theworshippersstayintheprobablefocusingvolumeSecondarydomeshavenofocusingeffectonthecontrarytheyprovidesoundenergytodistantpartsofthemosqueandtheyprovideadiffusedfieldbeneaththem

TheancientMosquesofSinanarestilluniqueandmaybeareferenceandmodelformosquesyettobebuilt(CAHRISMA2007 Kayili 2005 Topaktas 2003) Except for cavityresonators the basic architectural features of Kocatepe areinspiredbysomeoftheSinanMosquesTheinteriortreatmentofthemosqueisimportantinprovidingadivineenvironmentThestalactitesasa transitionalarchitecturalelementbetweendome and columns the ornamentation on the mihrabthe minbar the columns and balconies-- all have not justaestheticappealbutalsoacousticalpurposesespeciallyinhighfrequencies

Itisclearlyobservedfromtheresultsthatneweramosqueswith concrete domes have a greater tendency to keep andenhance the low frequency soundswithin the audience areaThis low frequency interferes with the speech intelligibilityandputsconstraintsontheclarityofthelectureortheordersconveyedbytheImamThematerialsusedinancientmosquesandthecavityresonatortechniquewithvarioussizedglasspotswithin the dome were wise and efficient in providing goodacousticalqualityRecentlydevelopedacousticaldevicessuchasHelmholtzresonatorsprovidenewdesignopportunitiesthesecouldduplicatetheprincipledesigncriteriaofSinanandcouldimprovetheacousticsof futuremosquesaswellasKocatepe

However considering the feasibility of Helmholtz resonatorswithin big spaces such as mosques the authors would liketo introduce micro-perforated panel absorbers to be studiedmore comprehensively in future applications and studies ofmosqueacousticsThisnewtechnologyusesthesamephysicsprincipals asHelmholtz resonators in substantial attenuationof lowfrequencieswhilekeepingthehighfrequencies inthedesiredrangesThearchitecturalapproachandfeelingofthespaces could then be combined with the new technologicalimprovements of this era and architectural acoustics as asciencecouldbegreatlyfurthered

InthisstudythedrawbacksofnewconstructiontechniquesontheacousticsofcontemporarymosqueshavebeenanalyzedTheKocatepeMosqueisoneoftheleadingexamplesoftheOttomanstyle reinforced concrete dome typology in mosque acousticsand architecture The acoustic simulationsdemonstrated thatthismosquetypologyhasstrongaestheticfeaturesbutmoderateacoustical effects in comparison to the authentic OttomanmosquesTheexcessivereverberationinthelowfrequencyrangeoftheenergydecayisfoundtobethebasicproblemofthatnewtypologyThekeytoimprovetheproblemsintodayrsquosmosquesistoadjustandcombinethematerialtechnologyoftheeratotheaestheticsofancientmosquessuchthatitfitstotheconceptandeventuallytothefunctionalityofthemosquesacoustics

References Abdou AA (2003) Measurement of acoustical characteristics

of mosques in Saudi Arabia The Journal of Acoustical Society of America 113(3)1505-1517

Barron M (1993) Auditorium Acoustics and Architectural DesignLondonEampFNSpon

CAHRISMA(2007)Orsted ndashDTUAvailableathttpserveroersteddtudkwwwoldatcahrismaretrieved4September2007

EganMD(1994)Architectural AcousticsNewYorkMcGraw-HillFaustiPPompoliRampProdiN(2003)Comparingtheacoustics

ofmosquesandByzantinechurches19th International Symposium CIPA (The International Committee for Architectural Photogrammetry) AntalyaTurkeyTechnicalUniversityofIstanbul

HamadahHAampHamoudaHM(1998)AssessmentofspeechintelligibilityinlargeauditoriacasestudyKuwaitStateMosqueApplied Acoustics54(4)273-289

KarabiberZ(1999)AcousticalproblemsinmosquesAcasestudyonthethreemosquesinIstanbulForum Acusticum BerlinCD-ROM4pgs

Karabiber Z (2000) A new approach to an ancient subjectCAHRISMAprojectInProceedings of the 7th International Congress on Sound and Vibration (1661-1668) Garmisch-PartekirchenGermanyInternationalInstituteofAcousticsandVibration

KarabiberZampErdoganS(2002)Comparisonoftheacousticalpropertiesofanancientandarecentmosque Forum AcusticumSeville Spain Spanish Acoustical Society (SEA) Available athttpwwwcicaes~forum2002retrieved4September2007

KayiliM(2000)UseofcavityresonatorsinAnatoliasinceVitruviusInternational Institute of Acoustics and Vibration 31621-1628

KayiliM(2002)AnadolursquodagelenekselakustiksistemlerveMimarSinan uygulamaları (Anatolian traditional acoustic works andworksdonebySinantheArchitect)[Turkish]InProceeding of 6th National Acoustical Congress TAKDER 26-28 October (233-238)AntalyaTurkeyTurkishAcousticalSociety

ArchitecturalScienceReview Volume51Number1March200828

KayiliM(2005)Acoustic Solutions in Classic Ottoman Architecture Manchester England Foundation for Science Technology andCivilization also available at httpwwwfstccouk retrieved 4September2007

KuttruffH(1991)Room AcousticsLondonElsevierLyngeC(2001)ODEON Room Acoustics Program Version 50 User

Manual Industrial Auditorium and Combined EditionsLyngbyDenmarkOrstedsPlads

Maekawa Z amp Lord P (1994) Environmental and Architectural AcousticsLondonEampFNSpon

Naylor GM (1993) ODEON Another hybrid room acousticalmodelApplied Acoustics38131-143

ProdiNampMarsilioM (2003) On theeffectofdomedceilingin worship spaces A scale model study of a mosque Building Acoustics10(2)117-133

RindelJH(2000)TheuseofcomputermodelinginroomacousticsJournal of Vibroengineering 3219-224

Schmidt AMD amp Kirkegaard PL (2004) On architecturalacoustics design using computer simulation In Eleventh International Congress on Sound and Vibration St PetersburgRussia International Institute of Acoustics and Vibration(IIAV)

SteenekenHJMampHoutgastT (1980) Aphysicalmethodformeasuring speech-transmissionquality Journal of the Acoustical Society of America67(1)318-326

Suarez R Sendra JJ Navarro J amp Leon AL (2004) Theacoustics of the cathedral mosque of Cordoba Proposals forarchitectural intervention Acta Acustica United with Acustica 90362-375

TempletonD(1993)Acoustics in the Built Environment Advice for the Design TeamOxfordButterworth

Topaktas IL (2003) Acoustical properties of classical Ottomanmosques Simulation and measurements Unpublished PhDthesis Department of Architecture Middle East TechnicalUniversityAnkara

Tuumlrkiye Diyanet Vakfi (Presidency of Religious Affairs) (2007)Kocatepe Mosque httpwwwdiyanetvakfiorgtreserlerkocatepecamiikocatepe_camiipdfretrieved4September2007

Utami SS (2005) An acoustical analysis of domes coupled torooms with special application to the Darusshol MosqueUnpublishedMSthesis BrighamYoungUniversitySaltLakeCityUSA

WeitzeCAChristensenCLRindelJHampGadeAC(2001)ComputersimulationsoftheacousticsofmosquesandByzantinechurches In 17th International Congress on Acoustics RomeItalianAcousticalSociety

YukselZKBinanCampUnverR(2003)Aresearchprojectinthe intersectionofarchitecturalconservationandvirtual realityCAHRISMA In 19th International Symposium CIPA (The International Committee for Architectural Photogrammetry AntalyaTurkeyTechnicalUniversityofIstanbul

5

APPENDIX

Calculated parameters distribution maps for different activity patterns

M1 Reverberation time distribution maps for 500 Hz source at mihrab and minbar

M2 Early decay time distribution maps for 500 Hz source at mihrab and minbar

AppendixCalculatedparameterdistributionmapsfordifferentactivity

patterns

M1 a) Reverberation time distribution maps for 500 Hz source at mihrab b) Reverberation time distribution maps for 500 Hz source at minbar

29ZuumlhreSuumlandSemihaYilmazer AcousticalCharacteristicsofMosques

6

M3 Clarity distribution map for 500 Hz source at mihrab and minbar

M4 Definition distribution map for 500 Hz source at mihrab and minbar

5

APPENDIX

Calculated parameters distribution maps for different activity patterns

M1 Reverberation time distribution maps for 500 Hz source at mihrab and minbar

M2 Early decay time distribution maps for 500 Hz source at mihrab and minbar

6

M3 Clarity distribution map for 500 Hz source at mihrab and minbar

M4 Definition distribution map for 500 Hz source at mihrab and minbar

M2 a) Early decay time distribution maps for 500 Hz source at mihrabb) Early decay time distribution maps for 500 Hz source at minbar

M3 a) Clarity distribution map for 500 Hz source at mihrab b) Clarity distribution map for 500 Hz source at minbar

M4 a) Definition distribution map for 500 Hz source at mihrab b) Definition distribution map for 500 Hz source at minbar

ArchitecturalScienceReview Volume51Number1March200830

7

M5 Lateral fraction distribution map for 500 Hz source at mihrab and minbar

M6 Speech transmission index distribution map source at mihrab and minbar

7

M5 Lateral fraction distribution map for 500 Hz source at mihrab and minbar

M6 Speech transmission index distribution map source at mihrab and minbar

8

M7 Strength distribution map source at mihrab and minbar

M5 a) Lateral fraction distribution map for 500 Hz source at mihrab b) Lateral fraction distribution map for 500 Hz source at minbar

M6 a) Speech transmission index distribution map source at mihrab b) Speech transmission index distribution map source at minbar

M7 a) Strength distribution map source at mihrab b) Strength distribution map source at mihrab

ArchitecturalScienceReview Volume51Number1March200830

7

M5 Lateral fraction distribution map for 500 Hz source at mihrab and minbar

M6 Speech transmission index distribution map source at mihrab and minbar

7

M5 Lateral fraction distribution map for 500 Hz source at mihrab and minbar

M6 Speech transmission index distribution map source at mihrab and minbar

8

M7 Strength distribution map source at mihrab and minbar

M5 a) Lateral fraction distribution map for 500 Hz source at mihrab b) Lateral fraction distribution map for 500 Hz source at minbar

M6 a) Speech transmission index distribution map source at mihrab b) Speech transmission index distribution map source at minbar

M7 a) Strength distribution map source at mihrab b) Strength distribution map source at mihrab