Xiang Duanqi, Wang Zheng & Chen Jinjing Acoust i c Laboratory, Beijing Institute of Architectural Design and Research, Beijing 100045, People's Republic of China (Received I February 1990: accepted 5 April 1990) ABSTRACT It is of great importance in acoustic design to establish a diffuse sound field in a reverberation chamber. Though various attempts have been made to establish sound diffusion in most existing ret'erberation chambers, the results were not good enough, and consequently extra diffusers have had to be installed. Excellent sound diffusion has been obtained in the reverberation chamber to be discussed in this paper by using an irregular shape and tilted walls with fixed reinforced concrete spherical diffusers. In this paper the authors have emphasized the design of sound diffusion and also discussed the design and determination of the acoustic properties of the reverberation chamber. 1 I N T R O D U C T I O N T h e Ac o u s t i c La b o r a t o r y o f H a n g z h o u El e c t r i s o u n d Fa c t o r y , s i t u a t e d i n H a n g z h o u Ci t y, t he c a p i t a l o f Z h e j i a n g Pr o v i n c e , Ch i n a , c o mp r i s e s t wo pa r t s , a n a n e c h o i c c h a m b e r i n t r o d u c e d i n a n o t h e r p a p e r ~ a n d a r e v e r b e r a t i o n c h a mb e r t o be d i s c u s s e d he r e i n de t ai l . Th i s r e v e r b e r a t i o n c h a mb e r is bui l t t o me a s u r e t h e p e r f o r ma n c e o f l o u d s p e a k e r s a n d mi c r o p h o n e s i n a di f f us e s o u n d f i el d, t he s o u n d p o we r l e ve l o f l o u d s p e a k e r s a n d o t h e r s o u n d s o ur c e s , a nd, a b s o r p t i o n c o e f f i c i e n t o f ma t e r i a l s , f o r wh i c h pe r f e c t s o u n d di f f us i o n, l o n g r e v e r b e r a t i o n t i me , u n i f o r m s o u n d f i el d d i s t r i b u t i o n , a n d l o w b a c k g r o u n d n o i s e l e ve l are r e qui r e d. The r e f o r e , t he d e s i g n s h o u l d me e t t he f o l l o w i n g r e qui r e me nt s . 83 Applied Acoustics 0003-682X/91/$03-50 1991 Elsevier Science Publishers Ltd, England. Printed in Great Britain 84 Xiang Duanqi, Wang Zheng. Chen Jinling - S o u t h w a l l -1-~ . - ' + [ r L l ' - - , , " "~'=:~, ~ I l l ~ I -93oo [ I i v l I 1 Fig. I. i :o-oo 2SO mm Reinforced c o n c r e t e f e l t I OOmm Mi n e r a l w o o l 5 0 0 r n r n B i t u mi n o u s c o n c r e t e (a) (b) 150 rnrn Plain concrete (a) Pl an and (b) cross-sect i on of the r ever ber at i on chamber . - - Re ve r be r a t i on time: at 125Hz, T > 15s, at 500 Hz, T > 12 s, and at 2000 Hz, T > 6 s. - - S o u n d field di st ri but i on: in t he l ow- f r equency r ange AP = Pm~x- Pmi, < + 3"0dB, in t he medi um- and hi gh-frequency range AP = P~x - Pmi, < + 1.5 riB, wher e Pmax and P , , , r epr esent respect i vel y t he ma xi mum and mi ni mum sound pressure level in t he measur i ng area. - - S o u n d diffusion: t he pl anned di rect i onal diffusion d > 0-95 wi t hout ext r a suspended diffusers and scat t eri ng el ement s in t he chamber ; and - - Ba c k g r o u n d noise: less t han 30 dBA. The r ever ber at i on chamber is built wi t h an i r r egul ar shape, wi t h effective capaci t y of 286-6 m 3 and t ot al surface ar ea 414 m 2. Bot h t he pl an and cross- sect i on of t he c ha mbe r are shown in Fig. 1. The acoust i c meas ur ement s conduct ed af t er t he compl et i on of t he r ever ber at i on c ha mbe r have i ndi cat ed t hat all t he desi gn r equi r ement s are met. 2 ACOUSTI C DES I GN 2.1 V o l u m e To est abl i sh a diffuse s ound field in t he chamber , adequat e nor mal modes must be st i mul at ed in t he measur i ng f r equency range, especi al l y in t he low- Acoustic design of a reverberation chamber 85 frequency range. The number of nor mal modes wi t hi n the frequency band Af ne a r t he frequency f can be obt ai ned f r om eqn (1): 2 r 4 n f 2 V r t f s L ] a N = L ~ +~--~o2 +~-~-o_ a f (1) where V= vol ume of the chamber (m3); s = total surface area of the chamber (m2); L = total l engt h of t he edges of t he chamber (m); Co = s ound velocity in air; and AN = number of nor mal modes. Accor di ng to eqn (1), when t he f r e que nc yf i s hi gh the second and t hi rd t erms can be i gnored; t hen t he equat i on may be simplified to: 4rcf z V AN_~ C~ Af (2) In general measur ement s, t he l / 3 oct ave band is oft en used, of whi ch t he ratio bet ween t he frequency band wi dt h Af and its cent ral frequency fm equal s 0"23. In this case eqn (2) can be wri t t en as: AN= 7.4(-ff--V ~ (f---m~ 3 \ Voj t fo ) ( 3 ) where jm = cent ral frequency of I/3 oct ave band; V o = reference vol ume 100m3; and f o = reference frequency = 100 Hz. Fr om eqn (3), t he appr oxi mat e number of nor mal modes wi t hi n a 1/3 oct ave band, of whi ch cent ral frequency is fro, can be obt ai ned. To establish a perfectly diffuse s ound field, at least 15-20 nor mal modes shoul d be excited wi t hi n the 1/3 oct ave band of whi ch t he cent ral frequency is 100Hz. Accor di ng to eqn (3), the effective vol ume V is cal cul at ed as 200-270 m 3. Therefore, t he vol ume of our chamber is desi gned as 268-6 m 3. 2.2 Shape Though havi ng little influence on t he numbe r of nor mal modes, t he shape of the chamber does affect t he di st r i but i on of nor mal frequencies, part i cul arl y in the l ow-frequency range. ISO have r ecommended, 3 therefore, t hat t he ratio bet ween any t wo di mensi ons shoul d not be a r ound number . The r ecommended ratios of wi dt h/ l engt h and hei ght / l engt h are shown in Tabl e 1. In our chamber , t he wi dt h/ l engt h and hei ght / l engt h rat i os are respectively 0-82 and 0.66. To pr event fl ut t er echoes pr oduced bet ween t wo parallel walls and to obt ai n a uni f or m s ound di st r i but i on in t he l ow-frequency range, t wo walls of t he reverberat i on chamber are tilted, wi t h tilting angles 15-- and 5 respectively and t he acut e angl es cut off. 86 Xiang Duanqi, Wang Zheng, Chen Jinjing TABLE 1 Recommended Rat i os of Wi dt h/ Lengt h and Height/' Length Width/length Height~length 0-83 0"65 0-79 0"63 0-68 0"42 0.70 0"59 2.3 Reverberation time The absor pt i on area of each surface in the reverberat i on chamber must be made as small as possible, so t hat the reverberat i on time can be kept l ong enough to measure the absor pt i on propert i es of materials, part i cul arl y mat eri al s with low absor pt i on coefficients. It is specified by ISO 3 t hat in an empt y chamber with a vol ume of 200 m 3, the total absor pt i on area must be less t han the values listed in the first row of Table 2, when the measur ement is conduct ed in 1/3 octave bands. If the vol ume of a reverberation chamber is I7, the al l owed ma xi mum absor pt i on area in t he chamber shoul d be the above- st at ed values mul t i pl i ed by ( V/ 200) 2/3. To reduce t he absor pt i on area, t he walls are overl ai d with cerami c tiles wi t h poi nt i ng of white cement and t he floor and surface of spherical diffusers are paved with waxed terrazzo. The reverberat i on chamber and the spherical diffusers are made up of rei nforced concret e, in order to prevent l ow-frequency r esonant absorpt i on. 2.4 Diffusion It is very i mpor t ant for a reverberat i on chamber to have a sufficiently diffuse s ound field. The vol ume of 268.6 m 3 and low absor pt i on areas were desi gned to give good diffusion but diffusion woul d undoubt edl y be effected, part i cul arl y when measur i ng t he absor pt i on coefficient of materials, when TABLE 2 The Al l owabl e Maxi mum Abs or pt i on Ar ea (Sabine) Volume Frequen O' ( Hz} {m 3 ) 125 250 500 1 000 2 000 4 000 200 6-5 6-5 6-5 7-0 9-5 13'0 268-6 7"9 7.9 7-9 8-5 1 l ' 5 15'8 Acoustic design of a reverberation chamber 87 r : 12~0 r = 6 5 0 . . . . , 1 1 0 0 0 I i , _ ~ r : 9,,~0 r : 5 0 0 0 + ,6oo + e r e t _ _ _ ~ i (a) Sou~h wal l 8610 Go 4 6 6 2 0 i ! 8610 I i i (d) Fig. 2. Dimensions, shape, and arrangements of diffusers. (a) Sizes and shape of diffusers. (b) Arrangement on southern wall. (c) Arrangement on the ceiling. (d) Arrangement on eastern wall. the f l oor is covered wi t h l Om 2 o f absorbent mat eri al . To obt ai n a perfect diffuse sound field, the f ol l owi ng met hods are usual l y adopted. - - Sus pe nde d scat t eri ng el ement s or diffusers in t he chamber , but this will great l y r educe t he r ever ber at i on time. - - I ns t al l i ng a r ot at i ng paddl e in t he cent er of t he chamber , but this will consi der abl y r educe t he measur i ng space. Ther e will also be noi se f r om t he mot or dri vi ng t he paddl e whi ch is difficult to el i mi nat e. To r emedy t he defect s ment i oned above, f our di fferent l y sized spheri cal diffusers made of r ei nf or ced concr et e ar e used in t he chamber , of whi ch t he radii of cur vat ur e are 1.25 m, 0-95m, 0-65 m and 0-5 m, respectively. The diffusers are fixed on t wo nei ghbor i ng walls and t he ceiling in an i r r egul ar ar r angement , so t hat no pai r of s moot h surfaces will face each ot her in t he chamber . The di mensi ons, shape, and a r r a nge me nt of t he diffusers ar e s hown in Fig, 2. Fi gur e 3 is a phot ogr a ph of t he i nsi de of t he r ever ber at i on chamber . 2.5 Background noise The r ever ber at i on c ha mbe r is si t uat ed in an i ndust ri al ar ea near t he ai rport , wher e t he noi se is up to 85 dBA dur i ng t ake- of f and l andi ng of airplanes. To 88 Xiang Duanqi, Wang Zheng, Chen Jinjing Fig. 3. Indoor setting of the reverberation chamber. prevent the i nt rusi on of noise and vi brat i on, a compl et el y separ at e st r uct ur e is used in t he cons t r uct i on of the r ever ber at i on chamber . The out er r oom is made or" bri ck 240 mm thick wi t h a r ei nt br ced concr et e root'. The i nner chamber is a box made of 150-mm thick r ei nl br ced concr et e set on an elastic cushi on made or" 500-mm t hi ck bi t umi nous concr et e and 100-mm t hi ck mi neral wool. The c ha mbe r has a doubl e- i nsul at i on door . 3 ACOUS TI C CHARACT E RI S T I CS OF THE RE VE RBE RAT I ON CHAMBE R Af t er compl et i on of the r ever ber at i on chamber , acoust i c meas ur ement s on r ever ber at i on time, s ound field di st r i but i on, s ound diffusion, and back- gr ound noi se were conduct ed. 3.1 Reverberat i on t i me The meas ur ement of r ever ber at i on t i me was conduct ed at nine l ocal i t i es (see Fig. 4) and at t hree levels (1. 6m, 2 0 m and 2-5m, respectively). The measur i ng f r equency range is f r om 100 to 4000 Hz in the 1/3 oct ave band, and wi t h oct ave band whi t e noi se as the s ound source. The da t a listed in Tabl e 3 show l ong enough r ever ber at i on t i mes and a smal t s t andar d Acoustic design of a reverberation chamber 89 ~7- - "-t~- - - -p'4- ---9"-" - - ~ 1 //.. ' I I I l ~--+--+---+---4 I I I I I ~9-- - - &- - - - ~- -b----~ 3 Reverberation time and directional diffusion ; o , Sound f i el d di s t r i but . i on Fig. 4. Locations of measurements. devi at i on. The abs or pt i on areas cal cul at ed on t he basis of r ever ber at i on t i me are less t han the al l owabl e ma xi mum val ues listed in Tabl e 2. 3.2 Sound field distribution The sound field di st r i but i on was meas ur ed f r om 125 Hz to 4000 Hz at 25 posi t i ons and t wo levels. The s ound sour ce was oct ave band whi t e noi se and t he signals were recei ved in t he 1/3 oct ave band. The difference bet ween t he ma xi mum and mi ni mum s ound pressure levels measur ed at six l ocat i ons and the s t andar d devi at i ons are listed in Tabl e 3. The results obt ai ned have i ndi cat ed t hat t he sound field di st r i but i on is uni f or m and the desi gn r equi r ement s have been satisfied. TABLE 3 The Results of Measurement Content Octave central f requency (H:) 125 250 500 1000 2 000 4 000 Reverberation time a 7"60 21-07 13"67 12"3 9"9 7"8 5-I 6 I'1 0-42 0"26 0"23 0"20 0"16 Sound field distribution AP 5'5 2-5 2'5 2"5 1.5 1"5 6 1"46 0-62 0-56 0"58 0-35 0-35 Absorption area b 2"1 3'2 3-5 4.4 5.5 8.5 Background noise 34.5 23"0 20.7 14-0 12.7 12"2 25"4 dBA Only octave data are listed. b Calculated on the basis of reverberation time. 9 0 Xiang Duanqi, Wang Zheng, Chen Jinjing P . 5 t 1 2 5 H z ~ 2 5 0 H z 5 0 0 H z l O0 0 H z 2 0 0 0 H z 4 0 0 0 H z I . q ~, d.l Aal ,,L= t=,J ~, ~, I. tJ_~ t . . . . f'% H,l. L. ~VV"IM'IL, V'V~lm., t ~ " V ' ~ , ~ \ \ ' v " . P ; p ~ , - s p ,,,,,d 1 0 , - . m l s "If Iv~ ~A., "'~ 1 ~ 1 ,.I .. ~V~A., ' " ~ , ~ ' ~ \ 're , ,, 0 l OdB, . . . . . . . . . . . F i g . 5. R e v e r b e r a t i o n d e c a y c u r v e s . 3. 3 Sound diffusion It has been i ndi cat ed t hat a bet t er diffusion of t he sound field is obt ai ned owi ng to t he non- par al l el walls wi t h r ei nf or ced concr et e spheri cal diffusers, whi ch can be pr ovi ded as follows. (a) The r ever ber at i on decay curves obt ai ned at 27 poi nt s (nine l ocat i ons and t hree levels) and 17 frequenci es ar e s moot h and st rai ght and t he s t andar d devi at i ons are small (see Tabl e 3). Some decay curves are shown in Fig. 5. (b) The sound field di st r i but i on in t he c ha mbe r is qui t e uni f or m (see Tabl e 3). (c) The me a s ur e me nt of di rect i onal di ffusi on pr opos ed by Thi el e "~ is qui t e compl ex and difficult to conduct , so a pl anar di rect i onal diffusion is meas ur ed in t he c ha mbe r by using a st at i onar y s ound sour ce emi t t i ng warbl e t ones of 2000 Hz and a high di rect i onal recei ver compos ed of a sect or sound lens with a condens er mi cr ophone at its focus. As t he recei ver rot at es in a pl ane f r om 0 t o 360 , t he vol t age out put f r om t he mi cr ophone is dr awn cont i nuous l y on pol ar coor di nat e r ecor d paper. The pl anar di rect i onal diffusion d is m d = 1 - - (4) gg/o wher e d = pl anar di r ect i onal diffusion; m = diffusion coefficient in t he meas ur ed chamber , and m o = di ffusi on coeffi ci ent in t he anechoi c chamber . The diffusion coefficient is: AM m -- (5) M Acoustic design o f a reverberation chamber 91 wher e M is t he mean s ound i nt ensi t y f r om all di rect i ons: M = f f ~Adq3 = I - - F I ~ : ~ ( x / - A ) Z d ~ 0 1 ( 6 ) 2 n n L 2 Jo and AM is t he mean devi at i on of s ound intensity: i += IA - Mld~o = L F - f - <7) d o 2x 7: L 2 Jo l i v / AV A M w h e r e ,4 i s t h e s o u n d i n t e n s i t y . The coefficient m o can be obt ai ned by means of t he same met hod as m. The mean val ue of pl anar di r ect i onal diffusion meas ur ed at nine l ocat i ons (the same ones as t hose used f or meas ur i ng t he r ever ber at i on time; see Fig. 4) is 0"96, t hus meet i ng t he desi gn r equi r ement s. 3 . 4 B a c k g r o u n d n o i s e l e v e l The ba c kgr ound noi se level i nsi de t he c ha mbe r was measur ed when t he near by wor ks hop was in oper at i on. The noise s pect r um and dBA are shown in Tabl e 3. It is cl earl y s hown t hat t he meas ur ed noi se levels ar e less t han t he desi gn target. 4 CONCL US I ON The meas ur ement s show t hat all t he acoust i c cri t eri a for t he r ever ber at i on c ha mbe r have been r eached, s ome even been surpassed. The r ever ber at i on c ha mbe r has been shown t o have excel l ent acoust i c pr oper t i es dur i ng t he past 18 mont hs. Thi s pr oj ect has pr oved t hat it is possible to establish sufficient di ffusi on in a r ever ber at i on c ha mbe r wi t hout ext r a suspended diffusers or scat t eri ng el ement s. The use of i r r egul ar shape and tilted walls wi t h fixed r ei nf or ced- concr et e spher i cal diffusers is one of t he bet t er met hods of obt ai ni ng a diffuse s ound field in a r ever ber at i on chamber . RE F E RE NCE S 1. Xiang Duanqi, Wang Zheng &Ch e n Jinjing, AppL Acoust., 29 (1990) 139. 2. Morse, P. M., Vibration and Sound, McGraw-Hill, 1948. 3. ISO 354 1985 E. 4. Thiele, R., Acoustics, 3 (1953) 291.