BBC Research and Development Report Number 1995-07

16 Pages·1995·0.55 MB·English

BBC RD 1995/7
Research and
Development
Report
FURTHER INVESTIGATIONS OF
PARTITIONS HAVING
BUILT-IN ACOUSTIC TREATMENT
G.D. Plumb, M.A.(Cantab)
Research and Development Department
Technical Resources Division
THE BRITISH BROADCASTING CORPORATION
BBC RD 1995/7
FURTHER INVESTIGATIONS OF PARTITIONS
HAVING BUILT-IN ACOUSTIC TREATMENT
G.D. Plumb, M.A. (Cantab).
Summary
Laboratory measurements were made of the sound insulations and absorption coefficients of a
range of metal-framed partitions having built-in acoustic treatment. The partitions were made
using glass wool and gypsum boards. The sound insulations of the single leaf partitions having
built-in acoustic treatment were considerably lower than those of comparable partitions made
using mineral wool, gypsum boards and fibreboard. For the double leaf partition having
built-in acoustic treatment, the sound insulations were, on average, 8dB higher than those of
thecomparable partitions containingmineralwool.
The use of the new double and triple leaf partitions containing built-in acoustic treatment
should result in similar savings of floor areas to those obtained by using the partitions
containingmineralwoolasbuilt-in acoustic treatment. However,itshouldbe possibletoobtain
a flatter reverberation time curve using the new partitions and the sound insulations should be
higher,providedthat double leaf partitionsare used.
Issued undertheAuthorityof
Research & Development Department,
Technical Resources Division GeneralManager
BRITISH BROADCASTING CORPORATION Research &Development Department
(R013) 1995
(cid:211) BritishBroadcastingCorporation
Nopartofthispublicationmaybereproduced,storedina
retrieval system, or transmitted in any form or by any
means, electronic,mechanical, photocopying, recording,
orotherwise,withoutpriorpermission.
(R013)
FURTHER INVESTIGATIONS OF PARTITIONS
HAVING BUILT-IN ACOUSTIC TREATMENT
G.D. Plumb, M.A. (Cantab).
1. INTRODUCTION ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... .. 1
2. THEDESIGN STRATEGY .. ... ... ... ... ... ... ... ... ... ... ... ... ... .. 1
3. ABSORPTIONCOEFFICIENTMEASUREMENTS . ... ... ... ... ... ... ... .. 2
4. SOUNDINSULATIONMEASUREMENTS . ... ... ... ... ... ... ... ... ... .. 5
4.1. Singleleafpartitions ...................................................5
4.2. Doubleleafpartitions ..................................................9
5. CONCLUSIONS . ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... . 10
6. REFERENCES .. ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... . 11
(R013)
FURTHER INVESTIGATIONS OF PARTITIONS
HAVING BUILT-IN ACOUSTIC TREATMENT
G.D. Plumb, M.A. (Cantab).
1. INTRODUCTION glass wool were used in the partitions tested; 75mm
and150mm.
Partitions containing built-in acoustic treatment have
already been designed and tested1. The primary For sections of the partitions having wideband
purpose of building-in acoustic treatment is to reduce absorption, the faces were covered with 24%*
the overall thicknesses of the partitions and therefore perforated, 0.7mm thick, galvanised steel sheet. The
increase the available floor areas in studios. The perforated steel sheet is smooth, robust and should
expected outcome of building-in acoustic treatment have virtually no effect on the absorption. Other
was a reduction in sound insulation. The partitions materials, such as perforated hardboard or aluminium
alreadytested were made using mineral wool, gypsum sheet, wererejected ongroundsof insufficientstrength
boardandfibreboard. orhighercost.
For thesepartitions,there wastoomuchoverlapin the For relatively unabsorbent sections of the partitions,
performancesofthelowandhighfrequencyabsorbent thefaceswerecoveredwithunperforated0.7mmthick
sections of the partitions. This overlap was a steel sheet. For these relatively unabsorbent sections,
consequence of the materials used, and resulted in an the cavity was still completely filled with glass wool
excess of absorption at mid frequencies. At the time, for three reasons: to damp the motion of the steel, to
this was not considered to be a difficulty. This was provide adequate sound insulation and to prevent
because the partitions were only intended for use in flanking transmission of sound at the boundaries
certain very low cost studio areas where the acoustic between absorbent and relatively reflective sections.
performances were of secondary importance to the The relatively reflective sections were expected to
spacesavingrequirements. providea certain amount of low frequencyabsorption,
because of damped resonances in the steel facing and
As a result of more recent work2,3, the combination of in the plasterboard/coreboardbackingto the absorbent
glass wool and gypsum boards alone in partitions leaves.
having built-in acoustic treatment, was expected to
result in better overall acoustic performances. The For single leaf partition tests, either one or two layers
range of low cost studios in which partitions having of 70mm thick metal studs were used, the choice
built-in acoustic treatment could be used should then depending upon the depth of glass wool required. In
beincreasedbytheuseofthenewpartitions. practical studio constructions containing 150mm
depth of glass wool, it would be better to use 150mm
Measurementsweremadeofthesoundinsulationsand studs rather than two layers of 70mm studs screwed
absorption coefficients of a number of partitions that together. In the tests, two layers of 70 mm studs
usedgypsumboardsandglasswool4as built-inacoustic screwed together were used to simplify the test
treatment.It was expectedthat the soundinsulationsof procedure. For absorbent Shaftwall leaves (special
the new partitions would be comparable with those of types of leaves that can be constructed from one side
theolder designsofpartitions; also thattherewouldbe only), 90mm thick I studs (containing 70mm
lessoverlapintheabsorption coefficient curves for the thicknessofglasswool)wereused.
low and high frequency absorbent sections of the
partitions. Inmulti-leafpartitions,onlysomeoftheleavesneedto
be absorbent. The remainder can be constructed as
conventional Camdens. In a typical studio con-
2. THE DESIGN STRATEGY struction, the single leaf partitions may be either
absorbent partitions or conventional metal-framed
Most of the absorption of the new partitions having single Camdens5. For a double leaf partition, the
built-inacoustictreatmentwasprovidedbyglasswool Shaftwall leaf would typicallyfaceinto thestudio and
in the cavities of the absorbent leaves. Moderate would be absorbent. The other leaf would usuallyface
depths of this glasswool have beenshown2to provide into a corridor and could therefore be a conventional
reasonablewidebandabsorptionwhenthematerialwas metal-framed single Camden. If both leaves of a
laiddirectlyonareflectivesurface.Twothicknessesof double leaf partition were absorbent, the sound
*
Intheselectionsofsteelsheetthatwereperforatedtheratiooftheareaoftheperforationstotheareaofsteelremainingwas24:76.
(R013)
performance which extended lower in frequency than
thepartitioncontainingmineralwool.
Fig.4 shows the absorption coefficient of an
unperforated partition containing a 70mm depth of
glass wool. The curve is typical of that observed for a
damped membrane absorber. The peak of absorption,
apparentat100–125 Hz,islinked with a resonancein
the metal front panel, which is damped by the glass
wool. The peak of absorption for the partition
containing glass wool is much higher than that for the
glassormineralwool partition containing mineral wool. This may be
because glass wool provides a more suitable level of
perforatedsteelsheet damping to the steel sheet than the mineral wool,or it
(N.B.Inthesectionsofsteelsheetthatwere perforated, may be a result of better sealing at the edges of the
theratiooftheareaoftheperforationstotheareaof
steelstuds.
steelremainingwas24:76)
Fig.1 -Cross-sectionthrough the framecontainingan Fig.5(seepage4)showstheabsorptioncoefficientsof
absorbentpartition.
insulation would probably be inadequate. In a triple
leafpartition,bothouterleaveswouldusuallyface into
studio areas and would be absorbent. The inner leaf
wouldbenon-absorbing,socouldbeasingleCamden.
glasswool mineralwool
3. ABSORPTION COEFFICIENT
MEASUREMENTS
ISO-Standard absorption coefficient measurements
were made on a number of single leaf absorbent
partitions, suspended horizontally above the floor of 12.5 9.5 12.5 15.0 19
mm mm mm mm mm
the receive room in the BBC Transmission Suite.
fibreboard coreboard/
They were built in to a frame of size 3.6m wide by plasterboard gypsumplank
3.0 m long, made from 25mm thick chipboard. Fig. 1
shows a cross-section through the frame, containing
one of the partitions tested. The frame was used for
tworeasons:toenclosetheedgesofthepartitionswith
a reflective boundary and to supportthe partitionsat a
known distance above the floor. The width of the
cavity between the partition and the floorwas50mm. perforated unperforated
Fig.2 shows the meanings of the symbols used in the steelsheet steelsheet
keys to the subsequent graphs of measurements.
Figs. 3– 6 show ISO-Standard measurements of the
50mm Cstud
absorption coefficients of the partitions built into the
frame.
70mm Cstud
Fig. 3shows themeasured absorptioncoefficients ofa
perforated partition containing a 70 mm depthofglass 150mm Cstud
wool. The infill pieces were 3000 × 600 × 19 mm
thick coreboard, and the studs were 90 mm thick
100mm CHstud
Shaftwall I studs. The absorption curve is typical of
that usually obtained for a 70mm thickness of a
90mm I stud
porous absorber6. The measurement is compared with
that of a partition containing mineral wool. The two
curves are very similar to each other,apart from some 150mm I stud
variation in the roll-off in the absorption at lower
frequencies. The partition containingglass wool hada Fig.2-Symbolsusedin thekeysto thegraphs.
(R013) -2 -
1.4 1.4
1.2 1.2
nt nt
e 1.0 e 1.0
ci ci
effi effi
o 0.8 o 0.8
c c
n n
o o
pti 0.6 pti 0.6
or or
bs 0.4 bs 0.4
a a
0.2 0.2
0 0
63 125 250 500 1000 2000 4000 8000 63 125 250 500 1000 2000 4000 8000
1/3octavebandcentrefrequency, Hz 1/3octavebandcentrefrequency, Hz
(a)partitioncontainingglass wool (a)partitioncontainingglasswool
(b)partitioncontainingmineral wool (b)partitioncontainingmineral wool
Fig.3-Theabsorptioncoefficientsof perforatedpartitions Fig.4 -Theabsorption coefficients ofunperforated
containing70mmdepth of insulationmaterial. partitions containing70mm depthof insulationmaterial.
a perforated partition containing 150mm depth of those for the unperforated absorber containing 70mm
glass wool. The result is compared with that for a thick glass wool (Fig.4), shows that increasing the
similar partition containing mineral wool. At higher depthofglass wool lowers the frequenciesover which
frequencies, the two curves show quite close thepartitionabsorbs.
agreement with each other. The partition containing
glasswool is averyeffectivewidebandabsorber,more In a studio, a proportion of the walls will require
so than the partition containing mineral wool. The perforated panels to control the reverberation times at
reasonfor this difference inperformance is notknown. mid and high frequencies. In addition, a proportion of
It is improbable that this large difference in the walls will be unperforated to provide low
performance is linked with the properties of the glass frequency absorption. The remainder of the walls can
wool and mineral wool because they are both porous be conventional metal-framed Camdens which have
absorbers having similar densities. It may be that the relativelysmalllevelsofabsorption.
coreboard, damped by the glass wool, is a more
effective lowfrequencyabsorberthanplasterboardand Fig. 7 (see page 5) shows the overall absorption
fibreboarddampedbymineralwool. coefficients of partitions, having perforated sheets for
50%ofthefrontalareaandunperforatedsheetsforthe
Fig.6 (overleaf) shows the measured absorption remainder, containing either 70mm thick glass wool
coefficients of an unperforated partition containing or 150mm thick glass wool. The absorption coeffi-
150mmdepthofglasswool. Theresults arecompared cients were predicted from the absorption coefficients
with those for a similar partition containing mineral of the partition having perforated sheets for 100% of
wool. The peak of absorption at 100Hz islinked with the frontal area, and unperforated sheets for 100% of
a resonance in the metal front panel, which is damped the frontal area respectively. For this prediction, the
by the glass wool. The peak of absorption for the mean absorption coefficient was calculated by
partition containing glass wool is much higher than summing the individual absorption areas, while
that for the partition containing mineral wool. This assumingthattheperforatedandunperforatedabsorber
may be because glass wool provides a more suitable regions did not interact. In previous work1, this had
level of damping to the steel sheet than the mineral beenshowntobeareasonableapproximation.
wool,oritmaybearesultofbettersealingattheedges
of the steel studs. A comparison of the results with Results are compared with predictions for absorbent
(R013) -3 -
1.4 1.4
1.2 1.2
nt nt
e 1.0 e 1.0
ci ci
effi effi
o 0.8 o 0.8
c c
n n
o o
pti 0.6 pti 0.6
or or
bs 0.4 bs 0.4
a a
0.2 0.2
0 0
63 125 250 500 1000 2000 4000 8000 63 125 250 500 1000 2000 4000 8000
1/3octavebandcentrefrequency, Hz 1/3octavebandcentrefrequency, Hz
(a)partitioncontainingglasswool (a)partitioncontainingglasswool
(b)partitioncontainingmineralwool (b)partitioncontainingmineral wool
Fig.5 -Theabsorption coefficients ofperforatedpartitions Fig.6 -Theabsorption coefficients ofunperforated
containing150mmdepth of insulationmaterial. partitions containing150 mmdepth ofinsulationmaterial.
partitions containing mineral wool. For the partitions irregular at lower frequencies and has low levels of
with mineral wool, perforated steel sheets were used absorptionbelow80Hz.
for 33% of the frontal area, the remainder using
unperforated sheets. The different percentages of Fig.8 shows the predicted reverberation times in a
unperforatedfacesforthe glasswoolandmineralwool typical studio built from absorbent partitions
partitions were chosen because the unperforated containing either 70mm or 150mm depths of glass
partitions with glass wool are much more effective as wool or mineral wool. For thecalculations,the sizeof
low frequency sound absorbers than the unperforated the studio was taken as 5m × 4m × 3m high. The
partitionswithmineralwool. floorwas coveredwith carpettiles andtheceilingwas
treatedwithsuspendedceilingtiles.
At higher frequencies, the partitions containing glass
wool have higher absorption coefficients than those For the room design, the area of perforated partition
containing mineral wool because a greater proportion was selected to give a mid band reverberation time of
of the faces of the partitions are perforated. At lower approximately 0.2 seconds. The area of unperforated
frequencies, the partitions containing glass wool have partition was selected to give a reasonably flat
considerably higher absorption coefficients than those absorption coefficient curve. For the partitions
containing mineral wool because the unperforated containing mineral wool, there was insufficient
glass wool partitions are more effective at absorbing available wall area for the desired level of low
low frequency sound than the partitions containing frequency absorption. This accounts for the bass rises
mineralwool.Thepartitionscontainingglasswoolare in the reverberation times for the studios constructed
much more suited to the acoustic treatment of rooms from partitions containing mineral wool.For the glass
containing other high frequency absorbers such as wool partitions, not all the partition area needed to be
carpet tiles or ceiling tiles. In comparison to the absorbent.Relativelyuntreatedareas wereprovidedby
partition containing 150mm glass wool,the measured usingconventionalmetal-framedCamdens.
absorption coefficient curve for the partition
containing 70mm depth of glass wool is more Above 250 Hz, the reverberation time curves show
(R013) -4 -
1.4 very close agreement with each other. Below 250Hz,
the reverberation time characteristics are somewhat
1.2
irregular.With oneexception,therise in reverberation
ent 1.0 timeatlowerfrequencieswas excessive andadditional
effici low frequency absorption would be required. For
o 0.8 150 mm depth of glass wool, the rise in reverberation
c
n time at lower frequencies was well controlled and the
o
pti 0.6 predicted reverberation time curve was acceptably
or smooth.Ifabsorptionweretheonlycriteriononwhich
bs 0.4
a the partitions were selected, the partition containing a
0.2 150mm depth of glass wool would be the preferred
choice.
0
63 125 250 500 1000 2000 4000 8000
1/3octavebandcentrefrequency, Hz
4. SOUND INSULATION
(a) 70mmglasswool MEASUREMENTS
50%perforated, 50%unperforated
4.1. Singleleafpartitions
(b) 150mmglasswool
50%perforated, 50%unperforated The sound reduction indices were measured of a
(c)70mmmineral wool variety of single leaf partitions built into the aperture
between the source and receive rooms in the BBC
33%perforated, 67%unperforated
Transmission Suite. The results are shown in Figs.9,
(d) 150mmmineral wool
10, 11 and 12. Either 70mm or 150mm thick metal
33%perforated, 67%unperforated studs were used, completely filled with either one or
two layers of compressed 75 mm thick glass wool
Fig.7 -Thepredictedabsorptioncoefficientsof partitions batts,as appropriate.On onesideof thestuds,asingle
having partially perforatedandpartially layer of 19mm thick gypsum plank was fitted (with
unperforated fronts. the boards installed horizontally), followed by one
layer of 12.5 mm thick plasterboard. The joints were
filled with acoustic sealant. Either unperforated or
2.0
24% perforated steel sheets (0.7mm thick) were
screwedtotheothersideofthestuds.
1.0 Fig. 9 (overleaf) shows the sound insulations of a
perforated partition containing a 70 mm thickness of
s 0.8
e, glasswool. The soundinsulationsabove200Hz were
m
ti 0.6 considerably lower than those of the partitions
n
o containing mineral wool. The reason for this is not
ati
er 0.4 known although considerable leakage of sound at the
erb perimeter ofthe partitionwasobservedwhenlistening
ev inthereceiveroom.
r
0.2
Nospecific reason fortheweaknesswasdiscoveredon
dismantling the partition. It may be that it is more
difficult to reliably seal the joint at the edges of the
0.1 gypsum plank and plasterboardlayers thanthe joint at
63 125 250 500 1000 2000 4000 8000
the edges of the plasterboard and fibreboard layers. It
1/3octavebandcentrefrequency, Hz
isalsopossiblethattheglasswoolislesseffectivethan
mineral wool at preventing the transmission of the
(a) using70mmglasswool
soundthathasleakedthroughanygap.
(b) using150mmglasswool
(c) using70mmmineralwool Below 100Hz, the new partition had higher sound
(d) using150mmmineralwool insulationsthantheolderpartitions.Thisisdueinpart
to the additional masses of the boards.It isalso linked
with the differentfundamental resonant frequenciesof
Fig.8 -Aprediction ofthe reverberation times
in roomsconstructedfrompartitionscontaining the partitions which are related to the masses and
mineralorglasswool. stiffnessesoftheboards.
(R013) -5 -
100 100
B B
x, d 80 x, d 80
e e
d d
n n
nI 60 nI 60
o o
cti cti
u u
d d
e 40 e 40
R R
d d
n n
u u
o 20 o 20
S S
0 0
63 125 250 500 1000 2000 4000 8000 63 125 250 500 1000 2000 4000 8000
1/3octavebandcentrefrequency, Hz 1/3octavebandcentrefrequency, Hz
(a)(34)partitionwithglasswool, (a)(39)partitionwithglasswool,
coreboardandplasterboard coreboardandplasterboard
(b)(38)partitionwithmineralwool, (b)(44)partitionwithmineralwool,
fibreboardandplasterboard fibreboardandplasterboard
(c)(42)partitionwithmineralwool, (c)(48)partitionwithmineralwool,
fibreboardanddoubleplasterboard fibreboardanddoubleplasterboard
(Rw valuesareshownin brackets) (Rwvaluesareshownin brackets)
Fig.9-Thesoundinsulationsof perforatedpartitions Fig.10-Thesoundinsulationsof unperforatedpartitions
containing70mmdepth of insulationmaterial. containing70 mmdepth ofinsulationmaterial.
Fig. 10showsthesoundinsulationsofanunperforated containing 70mm thickness of glass wool, the
partition containing 70mm thickness of glass wool. partition containing 150mm thickness of glass wool
As for the perforated partition containing 70mm has appreciably higher sound insulations above
thickness of glass wool, the sound insulations above 100 Hz. This is because the greater thickness of glass
200Hz were considerably lower than those of the wool absorbs more of the sound that has leaked past
partitions containing mineral wool, so similar com- the plasterboard and gypsum plank layers. Below
ments apply. The sound insulations below 200Hz 125 Hz, the two partitions had similar sound
were higher than for the partitions containing mineral insulations because the masses of the two partitions
wool. As expected, the sound insulations, above weresimilar.
100 Hz, of the unperforated partition were consider-
ably higher than those of the perforated partition. Fig.12showsthesound insulationsofanunperforated
Below 125Hz, the converse was true because of the partitioncontaining a150mmthicknessofglasswool.
differentfundamentalresonantfrequencies. In this case, the overall performanceofthe partitionis
comparable to that ofthepartitionscontainingmineral
Fig.11 shows the sound insulations of a perforated wool. The sound insulations between 315Hz and
partitioncontaininga150mmthicknessofglasswool. 2 kHz were lower than those of the partitions
As before, the sound insulations above 200Hz were containing mineral wool and the sound insulations
considerably lower than those of the partitions below 160Hz were higher than for the partitions with
containing mineral wool, and the sound insulations mineralwool.
below 100Hz were higher than for the partitions with
mineral wool. Compared with the perforated partition Compared with the unperforated partition containing
(R013) -6 -