Documente Academic
Documente Profesional
Documente Cultură
Applied Acoustics
journal homepage: www.elsevier.com/locate/apacoust
a r t i c l e i n f o a b s t r a c t
Article history: Panel-type sound absorbers are commonly used to absorb low-frequency sounds. Recently, a new type of
Received 14 January 2011 panel/membrane absorbers has been proposed as a next-generation sound absorber free from environ-
Received in revised form 28 April 2011 mental problems. On the other hand, it is known that placing a honeycomb structure behind a porous
Accepted 11 May 2011
layer can improve sound absorption performance and a similar effect can be obtained for microperforat-
Available online 2 July 2011
ed-panel absorbers. Herein, the sound absorption characteristics of a panel sound absorber with a hon-
eycomb in its back cavity are theoretically analyzed. The numerical results are used to discuss the
Keywords:
variations in the sound absorption characteristics due to the honeycomb as well as the mechanism for
Panel-type absorber
Honeycomb
sound absorption.
Sound absorption ! 2011 Elsevier Ltd. All rights reserved.
Sound transmission
0003-682X/$ - see front matter ! 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.apacoust.2011.05.017
944 M. Toyoda et al. / Applied Acoustics 72 (2011) 943–948
are designed and installed for a specific purpose. Thus, the interior cos h ! þ ik0 z cos h "
v j ðx; zÞ ¼ P e ! P!j e!ik0 z cos h eik0 x sin h ; ð4Þ
surfaces must be reinforced and supported by a structural element q0 c 0 j
such as a frame and a joist. Honeycombs can also serve as rein-
forcement. However, when the panel is intended to absorb sound, where q0 is the density of air, c0 is the speed of sound, k0 = x/c0 is
the vibration of the panel can be disturbed and sound absorption the wavenumber of air, and Pþ1 ¼ 1. In the case of a layer with a hon-
may deteriorate if the panel is tightly bonded to the honeycomb. eycomb, h = 0 can be substituted into cosines in Eqs. (3) and (4). The
Furthermore, honeycombs can be used for acoustical purposes if sines in the equations remain even in the honeycomb layer. This
the surface or structure is sufficiently strong, which can improve condition means the cell width of the honeycomb is infinitely small
the sound absorption performance. and the surface of a cell wall is perfectly smooth. The former
In this study the sound absorption characteristics of a panel assumption is valid if the cell width is less than one-half wave-
sound absorber with a honeycomb in its back cavity is theoretically length of air [13–15]. Although the latter assumption would be also
analyzed. The numerical results are used to discuss the effect of the ideal, the treatment of a honeycomb layer is on the safe side; when
honeycomb on the sound absorption characteristics and the hon- the surface is not smooth, additional sound absorption due to
eycomb mechanism. The theoretical analysis assumes the panel energy loss at the surface can occur especially at high frequencies.
extends infinitely. Both the vibration of the panel and the sound Allowing for admittances of the panel and the back wall, the
field are expressed by the standard form of a standing wave. The boundary conditions can be written as
discussion herein focuses on the influence of the honeycomb on
the sound absorption characteristics. v 1 ðx; 0Þ ¼ !ixwðxÞ þ A1 p1 ðx; 0Þ; ð5Þ
v 2 ðx; 0Þ ¼ !ixwðxÞ ! A2 p2 ðx; 0Þ; ð6Þ
2. Theoretical consideration p2 ðx; d1 Þ ¼ p3 ðx; d1 Þ; ð7Þ
v 2 ðx; d1 Þ ¼ v 3 ðx; d1 Þ; ð8Þ
Fig. 1 shows a model of a panel sound absorber with an air- v 3 ðx; d1 þ d2 Þ ¼ Ab p3 ðx; d1 þ d2 Þ: ð9Þ
backed cavity. The panel is a thin plate that extends infinitely
and is located at z = 0 with density q, thickness h, Young’s modulus By substituting Eqs. (2)–(4) into Eqs. (5)–(9), the solutions for un-
E, loss factor g, Poisson’s ratio m, and complex flexural rigidity known quantities W and P% j can be obtained.
# #2 Then the oblique inci-
D = E(1 ! ig)h3/{12(1 ! m2)}. The specific acoustic admittance of dence absorption coefficient ah ¼ 1 ! #P! # can be calculated, and
1
its illuminated surface is A1, and that of the back side (transmitted the sound-field-averaged absorption coefficient is determined by
side) is A2. The back cavity is divided into two layers with depths d1 taking the average from 0" to 78" for the angle of incidence h over
and d2. A honeycomb can be inserted into one or both of these lay- the half space.
ers. However, this study assumes the panel and the honeycomb are
not in mechanical contact (i.e., they do not have a force transmis-
sion). The back wall located at z = d1 + d2 is rigid with surface spe- 3. Numerical examples and discussion
cific acoustic admittance Ab. The time factor e!ixt is suppressed
throughout, and i is an imaginary unit, x is the angular frequency, 3.1. Effect of honeycomb: normal and oblique incidences
and t is time.
In this case, the equation of motion for the panel displacement Fig. 2 shows numerical examples of the normal and oblique
w(x) can be written as (h = 30", 60", 75") absorption coefficients relative to the field-inci-
dence-averaged absorption coefficient. Herein the back cavity is
@ 4 wðxÞ assumed to be completely filled with the honeycomb by setting
D ! qhx2 wðxÞ ¼ p1 ðx; 0Þ ! p2 ðx; 0Þ; ð1Þ
@x4 d1 = 0.045 m and d2 = 0 m (therefore, the thickness of the honey-
where pj(x, z) (j = 1, 2, 3) is the sound pressures of jth layer. Assum- comb is also d1). The specific acoustic admittance Ab is set to zero,
ing an incident plane wave with a unit pressure amplitude upon the and the specific acoustic admittances A1 and A2 are set to 0.026.
panel with an angle of incidence h, w(x), pj(x, z), and the particle The admittance of 0.026 means absorption coefficient of 0.1 with-
velocities vj(x, z) can be expressed in terms of unknown quantities out phase shift, which is considered herein as the mean value of
W and P% sound absorption coefficients due to the surface roughness of ply-
j as
wood. As a matter of course, the sound absorption characteristics
wðxÞ ¼ Weik0 x sin h ; ð2Þ of plywood surface depend on frequency. We however give the
! " constant value of 0.026 at all frequencies in order to abstract the
pj ðx; zÞ ¼ P þj eik0 z cos h þ P!j e!ik0 z cos h eik0 x sin h ð3Þ effect of surface absorption. Fig. 2a and b shows those without
and with a honeycomb, respectively.
For the normal and oblique incidence cases, the peak frequen-
cies drastically change with the angle of incidence as well as the
x presence of a honeycomb (Fig. 2a). Without a honeycomb, the peak
frequency increases as the angle of incidence increases. Addition-
A1 A2 Ab ally, the peak value varies with the angle. In this example, the peak
is highest for h = 60". In contrast, the peak frequency does not
p1 p2 p3 change with the angle of incidence upon applying a honeycomb
(Fig. 2b). (It should be noted that the peak frequency changes more
v1 v2 v3
than two octaves without a honeycomb.) Therefore, in the case of a
honeycomb-inserted absorber, the peak frequency is independent
of the angle of incidence. Attaching a honeycomb causes the peak
0 z frequency characteristics to become similar to those in the case of
d1 d2 normal incidence. However, the peak value for the honeycomb
case is similar to the case without a honeycomb, except where
Fig. 1. Geometry of the analytical model for a panel absorber backed by two air h = 75" and a significant peak is not present (compare Fig. 2a and
layers and a rigid wall. b).
M. Toyoda et al. / Applied Acoustics 72 (2011) 943–948 945
Absorption coefficient
d1 = 0.05 m
Absorption coefficient
0.6 0.6
0.4 0.4
0.2 0.2
0.0 0.0
31.25 62.5 125 250 500 1k 2k 4k 31.25 62.5 125 250 500 1k 2k 4k
Frequency [Hz] Frequency [Hz]
the honeycomb thickness is half the total cavity depth, and then
0.6 the peak diffuses. Although the previous section states that the
honeycomb increases the peak value, the aforementioned value
is not the maximum.
0.4 For a larger cavity depth, this tendency appears to be more dras-
tic. Fig. 4 shows the results for the total cavity depth of
d1 + d2 = 0.2 m. In this case the peak frequency decreases as the
0.2 honeycomb thickness increases, but the peak does not become
sharper. Actually, the peak becomes more diffuse. Hence, inserting
a honeycomb in the cavity is ineffective. Thus, inserting a honey-
0.0 comb into the back cavity does not necessarily improve the acous-
31.25 62.5 125 250 500 1k 2k 4k
tic performance of a panel absorber.
Frequency [Hz]
Fig. 5 shows the influence of the angle of incidence on the obli-
Fig. 2. Effect of sound incident conditions on the absorption characteristics: (a) que incidence absorption coefficient of the absorber with the same
without a honeycomb, (b) with a honeycomb completely filled in the back cavity. parameters. In the case without a honeycomb (Fig. 5a), the peak
Parameters of the panel are as follows: E = 6 ' 109 Pa, q = 610 kg/m3, h = 0.002 m,
absorption coefficient increases with the angle. However in the
g = 0.01, m = 0.3, d1 = 0.045 m, d2 = 0 m, A1 = A2 = 0.026, Ab = 0.
case with a honeycomb (Fig. 5b–d), the absorption coefficient de-
creases for a large angle of incidence. Hence, if these characteristics
In the field-incidence-averaged case, the characteristics are the are averaged, the overall field-incidence-averaged absorption coef-
results of the average of the characteristics for each angle of inci- ficient with a honeycomb decreases.
dence. Thus, the peak becomes sharper and higher with a honey-
comb (Fig. 2b). The peak frequency remains nearly constant in
1.0
the normal incidence case, but is lower than the peak frequency d1 = 0 m
of the field-incidence-averaged case without a honeycomb. Thus, d1 = 0.05 m
the honeycomb shifts the peak frequency to a lower frequency d1 = 0.1 m
and increases the peak value, effectively improving the sound 0.8
d1 = 0.15 m
absorption performance of panel absorbers. Although the peak be- d1 = 0.2 m
Absorption coefficient
Absorption coefficient
Absorption coefficient
0.6 0.6
0.4 0.4
0.2 0.2
0.0 0.0
31.25 62.5 125 250 500 1k 2k 4k 31.25 62.5 125 250 500 1k 2k 4k
Frequency [Hz] Frequency [Hz]
(c) 1.0
0 degree (d) 1.0 0 degree
30 degree 30 degree
60 degree 60 degree
0.8 0.8
75 degree 75 degree
Field incidence Field incidence
Absorption coefficient
Absorption coefficient
0.6 0.6
0.4 0.4
0.2 0.2
0.0 0.0
31.25 62.5 125 250 500 1k 2k 4k 31.25 62.5 125 250 500 1k 2k 4k
Frequency [Hz] Frequency [Hz]
Fig. 5. Angle dependence of the absorptivity of the honeycomb-backed panel absorbers. d1 + d2 = 0.2 m. d1 = (a) 0, (b) 0.05, (c) 0.1, (d) 0.2 m. Other parameters are the same as
in Fig. 2.
4. Mechanism ratio under a normal incidence is naturally zero. However for obli-
que incidences, the peak reflection ratio increases when the angle
The phenomenon where peak absorption decreases with honey- is large, but the reflection ratio for all angles reaches zero around
comb thickness, which is described in the previous section, can be
interpreted from the previous results by the authors [15]. In stud-
ies on the sound transmission through a single wall, transmission
1.0
loss increases when honeycombs divide the back side of the wall. 30 degree
In the case of a panel sound absorber, acoustical damping in a cav- 60 degree
ity that absorbs the sound energy transmitted through the panel 75 degree
0.8
via the vibration by the incidence sound is large. Therefore, when 85 degree
the transmitted energy through the panel decreases due to the
honeycomb, the energy in the cavity itself becomes smaller. Conse-
Reflection ratio
0.6
quently, the sound absorption efficiency is reduced. To clarify the
mechanism for the variation of the absorption coefficient by a hon-
eycomb, herein we discuss sound reflection from the viewpoints of
the honeycomb itself and sound transmission through a single pa- 0.4
nel attached to a honeycomb.
Eqs. (3), (4), (7), and (8) can model a honeycomb located in free
air where P! 0.2
3 ¼ 0 with the boundary conditions p1(x,0) = p2(x,0) and
v1(x, 0) = v2(x, 0). The second layer of the honeycomb is modeled by
substituting h = 0 into cosines in Eqs.
# #2(3) and (4). Then the reflec-
tion ratio of the honeycomb Rh ¼ #P! # can be obtained by solving 0.0
1 31.25 62.5 125 250 500 1k 2k 4k
the simultaneous equations. Fig. 6 shows the effect of the angle of
Frequency [Hz]
incidence on the reflection ratio when the honeycomb is 0.1 m
thick. Considering the treatment of a honeycomb, the reflection Fig. 6. Angle dependence of the reflection ratio of a honeycomb itself.
M. Toyoda et al. / Applied Acoustics 72 (2011) 943–948 947
1.7 kHz and 3.4 kHz due to z-directional resonances inside the hon- 35
eycomb cells. Fig. 7 shows the effect of honeycomb thickness for d1 = 0 m
h = 75". The reflection ratio at low frequencies increases when d1 = 0.015 m
30
the honeycomb thickness is large, except at resonance frequencies. d1 = 0.035 m
d1 = 0.1 m
Hence, the honeycomb layer reflects sound according to its thick-
(b) 6.2
1.0
d1 = 0.015 m
d1 = 0.035 m 6.0
d1 = 0.1 m
0.8
Transmission loss [dB]
d1 = 0.2 m
d1 = 0 m
5.8 d1 = 0.015 m
Reflection ratio
0.6 d1 = 0.025 m
d1 = 0.035 m
5.6
d1 = 0.05 m
d1 = 0.1 m
0.4
5.4
0.2
5.2
0.0 249 249.5 250 250.5 251
31.25 62.5 125 250 500 1k 2k 4k Frequency [Hz]
Frequency [Hz]
Fig. 9. Effect of honeycomb thickness d1 on the field-incidence-averaged transmis-
Fig. 7. Effect of honeycomb thickness d1 on the reflection ratio under an oblique sion loss of the panel with a honeycomb: (a) around 125 Hz, (b) around 250 Hz.
incidence of h = 75". Parameters of the panel are the same as in Fig. 2.
948 M. Toyoda et al. / Applied Acoustics 72 (2011) 943–948
2. As the honeycomb thickness increases, the peak sharpens and Culture, Sports, Science and Technology (MEXT) of Japan, Grant-in-
the peak value increases. Aid for Scientific Research (C), 20560550, 2010.
3. However, beyond a certain point, increasing the thickness
decreases the peak value because the transmitted energy starts References
to decrease.
4. If the total back cavity is sufficiently deep, the absorption peak [1] Fuchs HV. Alternative fiberless absorbers – new tools and materials for noise
control and acoustic comfort. Acust/Acta Acust 2001;87:414–22.
vanishes as the honeycomb becomes thicker because surface [2] Bruel PV. Sound insulation and room acoustics. London: Chapman and Hall;
absorption on the illuminated side becomes predominant com- 1951 [chapter 4].
pared to absorption due to multireflections in the back cavity. [3] Kimura S. Experimental studies on the sound absorbing characteristics of
acoustic materials. Rep Inst Ind Sci Univ Tokyo 1961;10(5):1–48 (in Japanese).
[4] Ford RD, McCormick MA. Panel sound absorbers. J Sound Vib 1969;10:411–23.
[5] Hiraizumi M, Takahashi K, Sone T, Nimura T. Study on the sound absorption by
5. Conclusion panel vibration – analysis in the model of circular panel with clamped
boundary. J Acoust Soc Jpn 1974;30:276–84.
In this study, the sound absorption characteristics of a panel [6] Sakagami K, Takahashi D, Gen H, Morimoto M. Acoustic properties of an
infinite elastic plate with a back cavity. Acustica 1993;78:288–95.
sound absorber with a honeycomb structure in its back cavity [7] Sakagami K, Gen H, Morimoto M, Takahashi D. Acoustic properties of an
are theoretically analyzed, particularly the effect of the honeycomb infinite elastic plate backed by multiple layers. Acust/Acta Acust
on the sound absorption characteristics and its mechanism. Insert- 1996;82:45–53.
[8] Sakagami K, Gen H, Morimoto M, Takahashi D. Acoustic reflection of an elastic
ing a honeycomb generally decreases the absorption peak value
plate with periodic ribs and a back cavity. Acust/Acta Acust 1999;85:1–11.
due to multireflections in the back cavity because the honeycomb [9] Bosmans I, Lauriks W, Lombaert G, Mermans J, Vermeir G. Sound absorption of
reflects the transmitted sound through the panel, reducing the stretched ceilings with an impervious synthetic membrane. J Acoust Soc Am
sound energy absorbed in the back cavity. However, the peak value 1999;106:233–9.
[10] Sakagami K, Kiyama M, Morimoto M, Takahashi D. Sound absorption of a
can increase, and the peak frequency decreases with a thin honey- cavity-backed membrane: a step towards design method for membrane-type
comb under certain limited configurations. Thus, the honeycomb absorbers. Appl Acoust 1996;49:237–47.
must be carefully selected to improve the absorption characteris- [11] Attenborough K. Sound absorbing materials and sound absorbers. In: Ver IL,
Beranek LL, editors. Noise and vibration control engineering. New York: Wiley;
tics of a panel-type absorber. 2005.
[12] Sakagami K, Morimoto M, Yairi M. Application of microperforated panel
absorbers to room interior surfaces. Int J Acoust Vib 2008;13:120–4.
Acknowledgments [13] Sakagami K, Yamashita I, Yairi M, Morimoto M. Sound absorption
characteristics of a honeycomb-backed microperforated panel absorber:
revised theory and experimental validation. Noise Control Eng J
This study was partially supported by the Program for the 2010;58:157–62.
Improvement of Research Environment for Young Researchers [14] Toyoda M, Takahashi D. Sound transmission through a microperforated-panel
from the Special Coordination Funds for Promoting Science and structure with subdivided air cavities. J Acoust Soc Am 2008;124:3594–603.
[15] Toyoda M, Kugo H, Shimizu T, Takahashi D. Effects of an air-layer-subdivision
Technology (SCF) commissioned by the Ministry of Education,
technique on the sound transmission through a single plate. J Acoust Soc Am
Culture, Sports, Science and Technology (MEXT) of Japan. Addition- 2008;123:825–31.
ally, this study was partly supported by the Ministry of Education,