The acoustic properties of panels with rectangular apertures

T. E. Vigran
Acoustics, Department of Electronics and Telecommunications, Norwegian University of Science and
Technology, NO-7491 Trondheim, Norway

(Received 13 January 2014; revised 24 March 2014; accepted 3 April 2014)

A model for the acoustic properties of a plate perforated with slots of rectangular shape is proposed.
The model is based on known expressions for the complex density and compressibility of a pore of
rectangular shape together with the radiation impedance of a rectangular shaped piston in a baffle.
For the so-called end correction of a rectangular aperture in a plate, an approximate solution is
shown to fit an exact solution for the imaginary part of the radiation impedance, the latter solution
based on the work of Lindemann [J. Acoust. Soc. Am, 55, 708–717 (1974)]. Two different
procedures are tested to calculate the mutual influence of the apertures on the end correction, the
one calculating the mutual impedance of neighboring pistons in the plate, the other by calculating
the end correction of a piston placed in the end of an infinitely long tube. The model is used
calculating the input impedance and absorption coefficient of a Helmholtz resonator with such a
plate, comparing with measurement results. The fit between predicted and measured results, using
plates with narrow slits, is good, but it is believed that the model also cover a wider range of
C 2014 Acoustical Society of America.
dimensions for such a slotted plate. V
[http://dx.doi.org/10.1121/1.4871363]
PACS number(s): 43.55.Ev, 43.40.Rj, 43.20.Rz [FCS] Pages: 2777–2784

I. INTRODUCTION In an analog way as with microperforated panels (with

Perforated plates or panels are widely used in the design
of absorbents for regulating the acoustic conditions in rooms.
Further use of perforated plates is found in various types of
acoustic silencers for air-conditioning systems, for silencing
engines in cars, trains etc. When having some millimeter
characteristic size of the perforations, such plates are nor-
mally combined with a fabric or a porous material to obtain
the necessary resistive component. Making the perforation
size in the sub-millimeter range, creating what is normally
denoted a microperforated panel, no additional resistive
material is needed. One commonly finds the perforations in
the form of holes or slots.
Modeling the acoustic impedance of such plates is based
on studies on the sound propagation in narrow tubes of
different cross-sectional areas containing a viscous, heat-
conducting fluid, normally air. Exact solutions for tubes of
circular cross section have long existed in the literature, see,
e.g., Rayleigh,1 and later interest in these solutions, and vice
versa for other cross-sectional shapes, has mainly been
directed by the research on sound propagation in porous
media, see, e.g., Allard and Atalla.2 A large amount of litera-
ture is also devoted to the application of these solutions to
the design of microperforated panels, which were initiated
by Maa’s work on hole-perforated panels.3,4
The present note is devoted to modeling the acoustics
of perforated panels with rectangular slots, such as
depicted in Fig. 1. We shall denote these apertures as
micro-slits when having a sub-millimeter width. We shall
then look for an exact solution for the sound propagation
in a tube of rectangular shape represented by the complex
density and bulk modulus. This is offered by Stinson,5 see
also Stinson and Champoux,6 where also pores of other
shapes are treated.
circular holes), Maa7 offered an approximate model for the
micro-slit case, however, not based on the exact model for
the rectangular case but for a slit of infinite length. A neces-
sary addition to such a model is an expression for the end
correction of the slits, based on the radiation impedance. In
Maa’s model, this is based on the radiation impedance of an
elliptical piston given by Rayleigh8 not a rectangular one as
will be used in this note.
An exact expression for the radiation impedance of a
rectangular piston in a baffle wall, using the letters b and a
for the dimensions width and length, is given by Mechel.9
An approximate solution is given by Levine,10 also cited by
Mechel9 saying “An approximation for large k0a (>5) and
not too small b/a.” In fact, it is found that another approxi-
mate expression for the end correction, also given by
Mechel,11 is better suited for smaller b/a. In the following
text, we shall compare this expression with an exact one
based on the work by Lindemann,12 who showed that the
radiation impedance of a baffled piston of arbitrary shape is
the Laplace transform of an impulse response directly related
to the geometrical shape of the piston.
For a perforated plate, the influence on the radiation im-
pedance by the neighboring apertures or pistons may become
important, modifying the end correction as calculated in the
manner expressed in the preceding text. A common way of
arriving at a correction factor for the end correction is to
divide the space in front of the plate into separate rectangular
tubes, one for each aperture in the plate, thereby calculating
the end correction of the aperture placed in the end of such
an infinitely long tube. In this way, Ingard13 has calculated
the mass end correction for various types of apertures, such
as circular and rectangular apertures in a rectangular tube.
Alternatively, the influence of neighboring apertures may be

J. Acoust. Soc. Am. 135 (5), May 2014 0001-4966/2014/135(5)/2777/8/$30.00 C 2014 Acoustical Society of America
V 2777

Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 142.68.229.57 On: Mon, 12 May 2014 04:17:13
 2p
ak ¼ ðkþ1=2Þ ;
l
2p
bn ¼ ðnþ1=2Þ : (4)
w

In Eq. (1), the argument qa is equal to q0 (1.2 kg/m3 for

air) to be used in Eq. (3), whereas qa0 in Eq. (2) is equal to
Prq0, where Pr is the Prandtl number. For air we get Pr
0.71.
FIG. 1. Sketch of perforated plate with slots of width w and length l. In the case of an infinitely long slit, i.e., letting l ! 1,
Eqs. (1) and (2) may be evaluated as6
quantified by calculating the mutual radiation impedance of
moving pistons. Here we shall use an integral expression pffi !1
tanhð jks Þ
given by Sha et al.,14 the one calculating the mutual imped- qs ðxÞ ¼ q0 1  pffi or
jks
ance of rigid rectangular pistons in a rigid infinite baffle.
In this note, we shall, after presenting this model for a pffi !1
tanðks = jÞ
plate with rectangular apertures, compare data with some qs ðxÞ ¼ q0 1 pffi ; (5)
ks = j
measurement results of the absorption coefficient deter-
mined in a standing wave tube. This will be using a plate
with narrow slits, which certainly does not verify the full and
potential of this model, but should demonstrate part of its cP0
usefulness. Ks ðxÞ ¼ " pffiffiffiffiffiffi # ; (6)
tanhð jPrks Þ
1 þ ðc  1Þ  pffiffiffiffiffiffi
jPrks
II. THEORY
where
A. Complex density and bulk modulus
rffiffiffiffiffiffiffiffiffi
w xq0
For a single rectangular aperture in a plate of thickness ks ¼ : (7)
d, we shall use the expressions for the complex density and 2 l
compressibility given by Stinson and Champoux6 developed
for a pore in a porous material. The cross-sectional dimen- From the Eqs. (1) to (4) and from Eqs. (5) and (6), respec-
sions in Ref. 6 are denoted 2a and 2b, and the assumption is tively, we may calculate the complex wave number k0 for the
that b/a  1. The notion concerning these dimensions might sound propagation in a tube of rectangular or slit shaped
vary, and they might not be consistent within the same cross section
book.9,11 As indicated on Fig. 1, we have therefore chosen to rffiffiffiffi
use the letters w and l for the width and length of the aper- 0 q
k ¼x ; (8)
ture, i.e., in Ref. 6 w ¼ 2b and l ¼ 2a. For the complex den- K
sity and the bulk modulus, the inverse of the compressibility,
we may write which for the rectangular case may be expressed as
q0
qr ðxÞ ¼ ; (1) x 1=2
Fðqa Þ kr0 ¼ c  ðc  1Þ  Fðq0a Þ=Fðqa Þ : (9)
c0
and
This will give the attenuation in decibel/meter approximately
cP0 equal to 8.68 times the imaginary part of the complex
Kr ðxÞ ¼ ; (2)
c  ðc  1Þ  Fðq0a Þ wave number. This is amply illustrated by Cummings15 for
different cross-sectional shapes including circular and trian-
where c and P0 are the adiabatic constant (1.4 for air) and gular shapes as well.
the barometric pressure, respectively, and the function F Here we shall return to the expressions for the com-
may be written plex density, making a comparison between the rectangu-
0 1 lar and the slit-shaped case to illustrate when Eq. (5) is a
64jxqa X1 X 1
1 useful approximation to Eq. (1). This is illustrated in Fig. 2
Fðqa Þ ¼ 2 2  @  A:
ll w 2 2 2 2 showing the real and imaginary part of the complex density
k¼0 n¼0 a b a þ b þ jxq =l
k n k n a in a slit of width w equal to 0.2 mm compared with rectan-
(3) gular slots having the same width but with lengths l equal
to 0.2, 0.4, and 2.0 mm. In the latter two cases, only the
Here l is the dynamic coefficient of viscosity, for air approx- imaginary part is shown as the differences in the real part
imately 2105 kg/(ms), and the coefficients ak and bn are are quite small, lying between the two results shown. As
given by indicated by the results, a ratio l/w on the order of 10

2778 J. Acoust. Soc. Am., Vol. 135, No. 5, May 2014 T. E. Vigran: Panels with rectangular apertures

Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 142.68.229.57 On: Mon, 12 May 2014 04:17:13
 " #
Dd 1 1  ð1 þ b2 Þ3=2
¼ bþ
w 3p b2
"
qffiffiffiffiffiffiffiffiffiffiffiffiffi
1 1
þ ln b þ 1 þ b2
p b
!#
qffiffiffiffiffiffiffiffiffiffiffiffiffi
1 2
þ ln 1þ 1þb ; (11)
b

where b is the ratio of the sides (¼ w/l).

Lindemann12 has shown that, in his own words: “The
radiation impedance of a baffled piston of arbitrary shape is
the Laplace transform of a time function closely and directly
related to the geometric shape of the piston.” This implies
that the normalized radiation impedance of any piston may
be expressed as
Zr
¼ L½ A  dðrÞ  f ðrÞ ðr ¼ c0  tÞ; (12)
Z0

FIG. 2. Real and imaginary part of the complex density in sound propaga-
tion in a slit-shaped tube of width 0.2 mm as compared with a tube of rectan-
gular cross section, having the same width but of length 0.2, 0.4, and
2.0 mm.
should justify using the slit model as an approximation for
the rectangular one.
B. The end correction for a slotted plate
The mass end correction of an aperture in a plate, also
sometimes denoted the attached length, is calculated from
the radiation impedance of a piston of the same shape. This
radiation impedance may, as mentioned in Sec. I, however,
be strongly influenced by the neighboring apertures, and we
shall give a correction factor to the end correction as a func-
tion of the perforation rate of the plate.
where L indicates the Laplace transformation of a function
comprising the delta-function and a function f(r) called the
characteristic pulse of the piston. The variable A is the area
under the characteristic pulse also being equal to the area of
the given piston.
In Lindeman’s paper, a number of different piston
shapes are treated, giving the expressions for the function
f(r) for these shapes, among them the rectangular one which
we shall use here.
Following Lindemann,12 Fig. 3 shows three examples of
these characteristic pulses on a relative scale for a square
piston as well as for two rectangular pistons with different
side ratio. According to Eq. (12), the corresponding impulse
responses, which can be transformed to the frequency
domain and thus give the complex radiation impedances, are
a d-function at time zero plus the negative value of the

1. Radiation impedance for a rectangular piston

As mentioned in Sec. I, Mechel9 gives an exact expres-
sion for the radiation impedance of a rectangular piston with
short side or width w ¼ b and long side l ¼ a, in a baffle wall.
Furthermore, an approximate solution is given by Levine,10
also cited by Mechel9 saying “An approximation for large
k0a (> 5) and not too small b/a.” The so-called end correc-
tion for a slot in a plate, based on the low frequency approxi-
mation of the imaginary part of the radiation impedance, is,
however, better described by another expression given by
Mechel,11 see following text. We shall compare this formula
for the end correction Dd, expressed by its equivalent imagi-
nary part of the normalized radiation impedance, i.e.,
 
Zr x Dd
Imag ¼ w ; (10)
Z0 c0 w

with an exact one based on the work by Lindemann.12 The

approximate expression for the relative end correction FIG. 3. Characteristic pulses for a square piston and for two rectangular
according to Mechel11 is given by pistons.

J. Acoust. Soc. Am., Vol. 135, No. 5, May 2014 T. E. Vigran: Panels with rectangular apertures 2779

Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 142.68.229.57 On: Mon, 12 May 2014 04:17:13
 pulses shown. Here we shall present the normalized imagi- TABLE I. Plate characteristics (see Fig. 1).
nary part of the radiation impedance, comparing it with the
Plate Slit width Slit length (l) C-C (X) C-C (Y)
approximate solution from Eqs. (10) and (11). This is shown sample (w) (mm) (mm) (mm) (mm)
in Fig. 4 for the three cases illustrated in Fig. 3, where the
straight lines are the approximations. As seen, these approxi- I 0.15 200 - 10
mations are very good in the lower frequency range, II 0.15 16.5 29.5 10
certainly dependent upward in frequency on the ratio
between the sides.
As mentioned in Sec. I, other approximations are also also be included. Approximate expressions based on the
used for the end correction of narrow slits, e.g., Maa,7 who perforation rate exist for panels with infinitely long slots, as
used the radiation impedance of an elliptical piston based on well as for hole-perforated panels, see, e.g., Mechel.11 For
Rayleigh’s derivation.1 It is therefore interesting to compare parallel infinitely long slots of width w, a commonly used
the approximate values for the end correction in this case approximation for the one-sided end correction is
with the one based on a rectangular piston. Maa7 presents a  
Dd 1 pe
series solution for the elliptic integral, the formula (13) in ¼  ln sin ; (13)
his paper, and also a table (his Table I,) of values for the rel- w p 2
ative end correction. This is shown in Fig. 5 together with
the one for a rectangular piston given in Eq. (11). As seen, where e is the perforation rate. Furthermore, for circular
the relative end correction for the elliptic case, referred to holes arranged in a square pattern, the expressions found
the minor axis, is always smaller than for the rectangular here and similar ones; see, e.g., Atalla and Sgard,17 are in
case. The curve for the elliptic case is here calculated by an fact approximations of the so-called Fok’s function, see
approximation for the elliptic integral given by Abramowitz Vigran.18 One of these approximations implies that the end
and Stegun.16 As seen, the table values given by Maa fits correction is multiplied by a factor Q, given by
very well to this approximation, whereas his formula (13) is pffiffi pffiffiffiffi
Qcirc:holes ðeÞ ¼ 1  1:47 e þ 0:47 e3 : (14)
useful only for side ratios near to 1. For a square shape, using
Eq. (11), the relative value will be 0.946, whereas the ellip-
Similarly, Ingard13 presents the one-sided approximate end
tic case will be p/4  0.785. For comparison, the value for a
correction for a plate with square apertures in a regular
circular piston, referred to its radius, is 8/(3p)  0.849.
pattern as
pffiffiffi
2. Mutual impedance. Correction factor for the end Dd  0:48  A  Qsquare ðeÞ
correction pffiffiffi pffiffi pffiffi
¼ 0:48  A  ð1  1:25  eÞð e  0:4Þ; (15)
Finally, using the expressions given in the preceding
text to calculate the impedance of a perforated plate, the mu- where A is the area of the aperture. However, Ingard13 also
tual influence of the apertures on the end correction must gives expressions for the general case, rectangular apertures

FIG. 4. The imaginary part of the normalized radiation impedance, given by

the frequency transform of the impulse responses according to Eq. (12) with FIG. 5. Relative end correction for rectangular and elliptic slits; the rectan-
the functions f(r) as shown in Fig. 3. The straight lines represent the approxi- gular one according to Eq. (11) and the elliptical one based on an approxi-
mations after Mechel (Ref. 11). mation in Ref. 16. Discrete values and curve after Maa (Ref. 7).

2780 J. Acoust. Soc. Am., Vol. 135, No. 5, May 2014 T. E. Vigran: Panels with rectangular apertures

Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 142.68.229.57 On: Mon, 12 May 2014 04:17:13
 in a rectangular tube, expressions we shall give in the fol-
lowing text.
As mentioned in the preceding text, one should also be
able to quantify the influence on the end correction by calcu-
lating the force on a given piston from the neighboring ones,
i.e., calculating the so-called mutual impedance. For the case
of two rigid pistons, see Fig. 6, the mutual impedance
according to Sha et al.14 may be expressed as
ð þw ð þl
Z0 k0 ejkh
Zmut ¼ j  ðl  jxjÞðw  jyjÞ  dxdy; (16)
2p w l h

where the variable h is given by

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
h ¼ ðx þ x1 Þ2 þ ðy þ y1 Þ2 : (17)

An example is shown in Fig. 7, where the normalized imagi-

nary part of the mutual impedance of square pistons, being a
half-width apart, is compared with the normalized radiation
FIG. 7. The imaginary part of the normalized radiation impedance of a
impedance of the piston shown in Fig. 4. square piston taken from Fig. 4 compared with the mutual normalized im-
Applying this procedure to calculate the influence on pedance due to a similar piston, the pistons being a half-width apart.
the end correction of a single piston, we shall, as an approxi-
mation, calculate the mutual impedance caused by four 0  12 0  1 2
neighboring pistons, i.e., summing the impedance due to l w
these four pistons. An example, where we use square pistons, Bsin pm l C Bsin pn w C lw
B 1 C B 1 C
is given in the following text where we compare with the gmn ¼ B C B C  ;
@ l A @ w A l1  w1
results using the model presented in the following text for pm pn
l1 w1
the end correction for such a piston placed in the end of an
infinite tube. 1
 mn ¼ for m ¼ 0 and n 6¼ 0; m 6¼ 0 and n ¼ 0;
In this case, following Ingard,13 a rectangular piston, as 2
shown in Fig. 1, with side’s l and w is placed in a tube of  mn ¼ 1 for m and n 6¼ 0: (19)
cross dimensions l1 and w1, the latter dimensions being equal
to the center-center distances between pistons in the x and y To arrive at a correction factor Q(e) for the end correction
directions. For frequencies well below the lowest cross mode for rectangular apertures, restricting us to the low frequency
in this tube (x < xmn), the one-sided end correction may, range, we define the factor as
based on Appendix I in Ref. 13, be expressed as
DdI
Qrect ðeÞ ¼ ; (20)
1 X
2X 1
  gmn  tmn DdM
Dd ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ; (18)
p m¼0 n¼0 ½m=l1 2 þ ½n=w1 2 where the indexes I and M refer to the end correction accord-
ing to Ingard,13 i.e., Eq. (18), and the equation by Mechel11
where the star symbol on the summation implies that the given by Eq. (11), respectively. Using the procedure to calcu-
term m ¼ 0, n ¼ 0 is excluded. The terms gmn and  mn, late the effect of the mutual impedance on a piston caused by
respectively, are given as four surrounding pistons, the correction factor is calculated as
ImagðZmut Þ
Qrect ðeÞ ¼ 1  ; (21)
ImagðZM Þ

where Zmut is the mutual impedance using Eq. (16) and ZM is

FIG. 6. Sketch illustrating the calculation procedure for mutual impedance
of rectangular pistons.
the impedance corresponding to the end correction given by
Eq. (11).
As a first example, we shall compare the results using
these methods on square apertures in a regular pattern, illus-
trated by the insert sketch in Fig. 8, showing the placement
of the four pistons to calculate the mutual impedance. The
corresponding cross-sectional area of the infinite tube for
using Eq. (18) is indicated by the dotted line. In Fig. 8, three
curves are shown, corresponding to Eqs. (20) and (21) to-
gether with the approximate one given by Eq. (15). It looks
like the latter one is a good approximation even further than

J. Acoust. Soc. Am., Vol. 135, No. 5, May 2014 T. E. Vigran: Panels with rectangular apertures 2781

Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 142.68.229.57 On: Mon, 12 May 2014 04:17:13
 indicated for this formula. The mutual impedance calcula-
tion shows generally a stronger dependence on the perfora-
tion rate, but, as mentioned in the preceding text, using only
four neighboring pistons to represent all pistons can only be
an approximation, probably working well at low perforation
rates.
For the general case of rectangular shaped apertures in a
regular pattern, see Fig. 1, four variables are determining the
perforation rate; the width and length of the apertures to-
gether with the center-to-center distances in two directions.
As a second example, we shall illustrate this case using
Eq. (20) with the ratio w/l as a parameter but with a fixed
spacing s between the apertures in both directions; see the
inset sketch in Fig. 9. In this figure, the result for the square
case, shown in Fig. 8, is plotted together with the result for a
ratio of ten, respectively, 100 for the length to width ratio. It
is interesting to note the stronger influence of the rectangular
shaped neighboring pistons.
Calculating the correction factor based on the mutual
impedance, which is not plotted in Fig. 9, shows an even
stronger dependence on the perforation rate than for the FIG. 9. Correction factor for the end correction of rectangular apertures in a
regular pattern in the special case of equal spacing between apertures in
square case in Fig. 8. both directions.
It should be possible to arrive at an approximate expres-
sion for the factor Q(e) also in this general case, expressed
by the four variables mentioned in the preceding text. from this graph, the approximate formula is useful for the
However, this has presently not been tried, and the presented higher perforation rate even for a length to width ratio (l/w)
calculations comparing with measurement results are there- of 10 only but it fails for low perforation rates such as below
fore based on Eq. (18). Finally, it might be interesting to 10%; i.e. the e1/2  0.3.
compare the end correction using this equation with the one
commonly used for parallel infinitely long slits given in
III. MEASUREMENTS
Eq. (13). The relative end correction comparing these equa-
tions is shown in Fig. 10 using the same arrangement of Measurements were conducted on a square sample;
apertures as shown in the insert sketch of Fig. 9. As seen 200 mm  200 mm, of a 1.0 mm thick aluminum plate with
slits of width 0.15 mm. Samples were fastened to a thin
frame in a standing wave tube of the same square cross sec-
tion, implying an upper working frequency of approximately
850 Hz. The measuring method is based on the ASTM
measuring standard E 2611-09, see Ref. 19, using the
so-called two-load method. The details concerning the set-up
and the measurement equipment are given by Vigran18 and
are not repeated here.
Details concerning the samples used are given in Table I,
referring to the sketch of such a plate in Fig. 1. In fact, the two
samples are based on using the same plate. In sample I, the
slits span the whole plate, whereas in sample II, the length of
the slits are (16.5 6 0.5) mm simply by taping 13 mm wide
and 0.5 mm thick plastic strips across the plate. The basic plate
(sample I) is formerly measured in a similar setup with differ-
ent cavity depths in front of a hard backing plate and com-
pared with predictions based on the infinite slit case. These
results are given by Vigran,20 and even if the fit between
measured and predicted results is reasonably good, there was
a discrepancy as to the frequency of the maximum absorption.
One of these former results, the one with an air cavity of
150 mm, is presented here together with a recent measure-
FIG. 8. Correction factor for the end correction of square apertures in a ment result using the ASTM standard where the cavity depth
regular pattern calculated by two methods, the one based on calculating the
mutual impedance (Ref. 14) caused by four surrounding apertures (pistons),
is a simulated one. As seen from Fig. 11, the two measure-
the other calculating the end correction of a piston in an infinite long tube ment results are surprisingly equal, spanning some 7 yr
(Ref. 13). between and using two different methods. The former

2782 J. Acoust. Soc. Am., Vol. 135, No. 5, May 2014 T. E. Vigran: Panels with rectangular apertures

Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 142.68.229.57 On: Mon, 12 May 2014 04:17:13

FIG. 10. The relative end correction for rectangular apertures of different
width to length ratio, arranged as shown in the insert sketch of Fig. 9, com-
pared to the approximate expression for the infinite length case, the latter
given by Eq. (13).
measurement results exhibits some strong oscillations,
particular in the range 120–170 Hz, presumably due to
resonances due to the “bars” or beams of the plate.
Assuming clamped end conditions for such a beam, the low-
est resonant frequency will be approximately 130 Hz, but the
actual end conditions are difficult to ascertain. The fastening
of the plate was different in these two cases; formerly small
screws were used to fix the plate to the frame, whereas glue
was used now, maybe introducing some damping to the
plate. Finally, predicted results fit the measured results rather
well.
Making shorter slit lengths by taping strips across the
plate, realizing sample II, gives the result shown in Fig. 12.
The cavity depth used in this case is 50 mm, and again the
fit between measured and predicted results is reasonably
good.
FIG. 11. Absorption coefficient for sample I with cavity depth of 150 mm,
measured in a standing wave tube with a physically hard backing vs a simu-
lated one.
J. Acoust. Soc. Am., Vol. 135, No. 5, May 2014
Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 142.68.229.57 On: Mon, 12 May 2014 04:17:13
FIG. 12. Absorption coefficient for sample II with cavity depth of 50 mm.
IV. CONCLUSIONS
A general model for the acoustic properties of a plate
perforated with rectangular apertures (slots or slits) is pro-
posed. The model is based on the literature on the complex
density and compressibility of a pore of rectangular shape
and the radiation impedance of a rectangular shaped piston.
The latter is used to calculate the so-called end correction of
a rectangular aperture, which must be modified to account
for the mutual influence of apertures in the plate. Two differ-
ent correction factors depending on the perforation rate are
tested: The one based on calculating the mutual impedance
between apertures represented by rigid pistons, the other
simulating the effect by calculating the end correction of a
piston placed in the end of an infinitely long tube. These cor-
rection factors are compared for apertures of square shape,
but for the general case, the latter model is kept, partly
because of computation time. However, a next step will be
to develop practical approximation formulas for this correc-
tion factor.
The model is compared with measurements results on
the absorption coefficient of a resonator type absorber, a
plate backed by an air cavity. However, only plates with nar-
row slits are tested, and it is believed that the model has a
much larger range of application, which surely should be
tested.
ACKNOWLEDGMENT
The author wants to thank Professor Peter Svensson for
valuable comments and discussions. Also thanks to Ph.D.
candidate Bjørn Kolbrek for providing a subroutine for the
mutual impedance calculation.
1
J. W. Strutt, Baron Rayleigh, The Theory of Sound, 2nd ed. (Dover, New
York, 1945), Vol. II, Chap. XIX.
2
J. Allard and N. Atalla, Propagation of Sound in Porous Media. Modelling
Sound Absorbing Materials (Wiley, Chichester, U.K., 2009), pp. 1–358.
3
D.-Y. Maa, “Microperforated wideband absorbers,” Noise Control Eng. J.
29, 77–84 (1987).
4
D.-Y. Maa, “Potential of microperforated panel absorber,” J. Acoust. Soc.
Am. 104, 2861–2866 (1998).
5
M. R. Stinson, “The propagation of plane sound waves in narrow and wide
circular tubes, and generalization to uniform tubes of arbitrary cross-
sectional shapes,” J. Acoust. Soc. Am. 89, 550–558 (1991).
T. E. Vigran: Panels with rectangular apertures 2783
 6 14
M. R. Stinson and Y. Champoux, “Propagation of sound and the assign-
ment of shape factors in model porous materials having simple pore geo-
metries,” J. Acoust. Soc. Am. 91, 685–695 (1992).
7 15
D.-Y. Maa, “Theory of microslit absorbers,” Chinese J. Acoust. 20, 1–10
(2001).
8 16
J. W. Strutt, Baron Rayleigh, The Theory of Sound, 2nd ed. (Dover, New
York, 1945), Vol. II, Chap. XVI.
9 17
F. P. Mechel, Formulas of Acoustics, 2nd ed. (Springer Verlag, Berlin,
2008), Chap. F, Sec. F.7.
10 18
H. Levine, “On the radiation impedance of a rectangular piston,” J. Sound
Vib. 89, 447–455 (1983).
11
F. P. Mechel, Formulas of Acoustics, 2nd ed. (Springer Verlag, Berlin,
19
2008), Chap. F, Sec. F.12.
12
O. A. Lindemann, “Transient fluid reaction on a baffled plane piston of
arbitrary shape,” J. Acoust. Soc. Am. 55, 708–717 (1974).
13
U. Ingard, “On the theory and design of acoustic resonators,” J. Acoust.
20
Soc. Am. 25, 1037–1061 (1953).
2784 J. Acoust. Soc. Am., Vol. 135, No. 5, May 2014
Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 142.68.229.57 On: Mon, 12 May 2014 04:17:13
K. Sha, J. Yang and W.-S. Gan, “A simple calculation method for the self-
and mutual-radiation impedance of flexible rectangular patches in a rigid
infinite baffle,” J. Sound Vib. 282, 179–195 (2005).
A. Cummings, “Sound propagation in narrow tubes of arbitrary cross-
section,” J. Sound Vib. 162, 27–42 (1993).
M. Abramowitz and I. A. Stegun, Handbook of Mathematical Functions
(Dover, New York, 1972), Chap. 17.
N. Atalla and F. Sgard, “Modeling of perforated plates and screen using
rigid frame porous models,” J. Sound Vib. 303, 195–208 (2007).
T. E. Vigran, “Normal incidence transmission loss in impedance tube—
measurement and prediction methods using perforated plates,” Appl.
Acoust. 73, 454–459 (2012).
ASTM E 2611-09. Standard Test Method for Measurement of Normal
Incidence Sound Transmission of Acoustical Materials Based on the
Transfer Matrix Method (American Society for Testing and Materials,
West Conshohocken, PA, 2009)
T. E. Vigran, Building Acoustics (Taylor and Francis, London, 2008), Chap. 5.
T. E. Vigran: Panels with rectangular apertures