Accepted Manuscript

Acoustic performance of periodic steel cylinders embedded in a viscoelastic medium

Gyani Shankar Sharma, Alex Skvortsov, Ian MacGillivray, Nicole Kessissoglou

PII: S0022-460X(18)30830-7
DOI: https://doi.org/10.1016/j.jsv.2018.12.013
Reference: YJSVI 14549

To appear in: Journal of Sound and Vibration

Received Date: 21 December 2017

Revised Date: 22 November 2018
Accepted Date: 8 December 2018

Please cite this article as: G.S. Sharma, A. Skvortsov, I. MacGillivray, N. Kessissoglou, Acoustic
performance of periodic steel cylinders embedded in a viscoelastic medium, Journal of Sound and
Vibration (2019), doi: https://doi.org/10.1016/j.jsv.2018.12.013.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT

Acoustic performance of periodic steel cylinders

embedded in a viscoelastic medium

PT
Gyani Shankar Sharmaa,∗, Alex Skvortsovb , Ian MacGillivrayb , Nicole
Kessissogloua
a School

RI
of Mechanical and Manufacturing Engineering, The University of New South
Wales, Sydney, Australia
b Maritime Division, Defence Science and Technology, Melbourne, Australia

SC
Abstract

U
The acoustic performance of a locally resonant phononic crystal comprising steel
cylindrical scatterers arranged periodically in a viscoelastic medium is analyt-
AN
ically and numerically studied. Analytical expressions for the damping and
dipole resonance frequency of the phononic array of solid cylinders embedded
in a viscoelastic medium are derived. Based on effective medium approxima-
M
tion, effective geometric and material properties of the composite layer with
embedded cylindrical scatterers are also derived. Results from the analytical
D

model are compared with those from a numerical model based on the finite el-
ement method. The phononic crystal is immersed in water and the effects of
TE

water backing and steel-air backing on its acoustic performance are investigated.
Dipole resonance of the cylinders as well as constructive interference between
EP

waves scattered by the cylinders and reflected from the steel backing plate are
shown to lead to high sound absorption, attributed to conversion of longitudi-
nal to shear waves which are subsequently dissipated in the rubber medium.
C

Different geometric and material parameters are examined to provide further

insight into the physical mechanisms associated with high sound absorption of
AC

the phononic crystal.

Keywords: acoustic coating, phononic crystals, soft elastic material, dipole

∗ Correspondingauthor
Email address: [email protected] (Gyani Shankar Sharma)

Preprint submitted to Journal of Sound and Vibration December 10, 2018

 ACCEPTED MANUSCRIPT

resonance, effective medium approximation

PT
1. Introduction

It is well known that composite media comprising periodic inclusions in a

RI
viscoelastic matrix can exhibit remarkable acoustic properties. These compos-
ite media are commonly referred to phononic crystals and there is a vast body
5 of literature on this subject, for example, see reviews [1–7] and the references

SC
therein. In particular, it has been found that materials with a soft rubber-like
matrix are favourable candidates for viscoelastic coatings in maritime applica-
tions [8–12]. This is due to the ability of these materials to form scattering

U
resonances at frequencies corresponding to acoustic wavelengths much larger
10 AN
than the dimension of the scatterers or even the spacing between the scatter-
ers. Such resonances are often called ‘sub-wavelength’ resonances and lead to
strong conversion of sound waves into shear waves, which are subsequently at-
M
tenuated due to the relatively large value of the imaginary component of the
shear modulus for rubber-like materials [9, 12–14].
D

15 Changing the shape of inclusions in a soft matrix is an approach to control

the resonance frequency and hence absorption [15]. Cylindrical voids have desir-
TE

able low frequency acoustic performance as, in theory, the resonance frequency
can be arbitrarily small [16–21]. The scattering of sound waves by cylinders
periodically arranged in a host matrix has traditionally been investigated for
EP

20 exploration of new wave phenomena, including dynamic density [22], homogeni-

sation [23], multiple scattering effects [24], acoustic cloaking [25], and as a model
C

of fibre-reinforced composites [26, 27]. Soft materials with solid inclusions also
have specific practical importance due to their consistent acoustic performance
AC

under static pressure which is often not the case for voided or porous compos-
25 ites [28, 29]. It has been demonstrated that soft materials with solid inclusions
can be described by an effective frequency-dependent dynamic density, while
the effective stiffness can be estimated from a frequency-dependent quasi-static
approximation [30]. To estimate dynamic density, a number of analytical frame-

2
 ACCEPTED MANUSCRIPT

works have been proposed and validated ranging from simple phenomenological
30 models [11, 12, 31, 32] to high-fidelity multi-scattering corrections [33–35].

PT
There are a number of interesting studies on resonant scattering by gratings
of heavy cylinders embedded in a light porous matrix [15, 36–41]. These stud-
ies provide important insight into sound interaction with the gratings, including

RI
quarter-wavelength and Biot resonance absorption, effect of inclusion shape and
35 multi-scattering phenomena. The porous matrix has been modelled either us-

SC
ing Biot theory or as an equivalent fluid. The results of these studies provide
practical guidance for design of effective vibroacoustic coatings for aerospace
applications. For rubber-like coatings used in underwater applications, the low

U
value of shear modulus leads to monopole and dipole resonances associated with
the inclusions which do not exist in porous media. These resonances often be-
40

AN
come the core performance characteristic of viscoelastic coatings for maritime
applications.
M
The present paper investigates sound absorption by a rubber medium with
periodic hard inclusions for use as an acoustic coating externally applied to ma-
45 rine vessels. An analytical model of acoustic wave scattering by a grating of hard
D

cylindrical inclusions embedded in a soft matrix is developed. Using homogeni-

sation theory, the matrix with inclusions is replaced by a medium with effective
TE

acoustic properties. This reduces the original two-dimensional problem of scat-

tering by a grating to a one-dimensional problem that can be easily treated by
EP

50 the transfer matrix method, which intrinsically incorporates all multi-scattering

phenomena. The proposed framework is applied for realistic cases of materi-
als in water or with steel and air backing. The analytical model is validated
C

with finite element simulations using the commercial software COMSOL and
demonstrates good agreement.
AC

55 2. Physical model

The phononic crystal model considered here consists of steel cylinders ar-
ranged periodically in a soft viscoelastic medium. The ‘softness’ of the medium

3
 ACCEPTED MANUSCRIPT

PT
RI
Incident

SC
plane
wave
Water h 2a Water
y

U
Steel Rubber
x
AN t

(a)
M
Incident
plane
D

wave
Water h 2a Air
y
TE

Steel Rubber Steel

x
t s
EP

(b)

Figure 1: Schematic diagram of a unit cell of a phononic crystal comprising a steel cylinder
C

embedded in rubber with (a) water backing and (b) steel-air backing.
AC

4
 ACCEPTED MANUSCRIPT

is characterised by the very low value of the ratio given by

|µ/λ|  1, (1)

PT
where λ and µ are the first and second complex Lamé coefficients, respectively,
√
with µ = µ0 + iµ00 , λ = λ0 + iλ00 , and i = −1 the imaginary unit. The coeffi-

RI
cient µ is also the complex shear modulus, where the real part µ0 is important
for stiffness and the imaginary component µ00 is important for damping. The
cylinders are assumed to be of infinite height, thereby reducing the model to

SC
60

two dimensions. Figure 1 schematically shows a layer of steel cylinders of radius

a uniformly separated by a distance h in a host rubber medium of thickness

U
t. The length of the phononic crystal extends to infinity in the y-direction and
only a periodic unit cell in that direction, comprising a single steel scatterer
65 AN
embedded in rubber, is shown in Fig. 1. The phononic crystal is subjected to
plane wave excitation and the fluid medium on the incidence side is water. The
acoustic performance of the phononic crystal is investigated for two types of
M
backing on the transmission side, namely, water (Fig. 1(a)) or a steel plate of
thickness s and air (Fig. 1(b)).
D

70 3. Analytical model of effective acoustic properties

TE

The analytical model to obtain the effective acoustic properties of an array

of hard inclusions in an elastic medium can be derived from the motion of the
EP

inclusions induced by an incident sound wave, for which the governing equation
is given in a number of publications [14, 32, 42–44]. It has been shown that
75 pulsating oscillations associated with monopole resonance results in frequency
C

dependent stiffness [20, 45], while translational oscillations related to dipole res-
onance lead to frequency dependent density [34, 35, 44]. When both resonances
AC

are present, the effective properties of the material include both frequency de-
pendent effective stiffness and density. For the case of hard inclusions in a soft
80 elastic medium considered in this work, monopole resonance is relatively weak
[46]. Only the effective density is frequency dependent while the effective stiff-
ness is determined by the conventional “mixture law”. Analytical expressions

5
 ACCEPTED MANUSCRIPT

describing dipole resonance of solid cylinders in an elastic medium, vibrational

damping of the array with inclusions, and the effective geometric and material

PT
85 properties of a layer of steel cylinders in the host elastic medium, are derived in
what follows.

RI
3.1. Motion of a solid inclusion in a soft elastic medium
Consider a lattice of scatterers comprising solid inclusions embedded in a soft
elastic medium. The equation of motion governing the vibration of an inclusion

SC
is given by [14, 32]
ρi vi üi (t) = Fi , (2)

U
where vi is the volume of the inclusions, Fi is the force acting on the inclusion
given by
AN
Fi = ma (ü(t) − üi (t)) + ξ(u̇(t) − u̇i (t)) + k(u(t) − ui (t)) + ρvi ü(t), (3)

and u(t), ui (t) are respectively the longitudinal displacement of the host medium
M
and solid inclusions embedded in it. Parameters ρ and ρi are the density of the
90 host elastic medium and steel inclusions, respectively. According to [32], the
D

first term on the right hand side of Eq. (3) corresponds to the added mass effect
whereby ma is the effective mass of the elastic medium in motion. The second
TE

term on the right hand side of Eq. (3) corresponds to viscous drag where ξ is the
viscous damping coefficient. The third term describes the elastic restoring force
EP

95 exerted by the matrix on the inclusion when it is displaced from equilibrium,

where k is the effective stiffness coefficient. The fourth term is the buoyancy
induced by the accelerating matrix. For the two-dimensional model considered
C

here, all four parameters, ma , ξ, k, and vi are defined per unit length of the
inclusion.
AC

Equations (2) and (3) can be written in a more compact form as

1+β
üi (t) − ü(t) + ψ(u̇i (t) − u̇(t)) + ω02 (ui (t) − u(t)) = 0, (4)
δ+β
where
ξ
ψ= , (5)
ρvi (δ + β)

6
 ACCEPTED MANUSCRIPT

s
k
ω0 = , (6)
ρvi (δ + β)

PT
100 and ψ is the effective damping coefficient, δ = ρi /ρ is the ratio of the density
of the scatterers to that of the host medium, and β = ma /ρvi is the ratio

RI
of the effective mass of the host medium to the mass of the steel inclusions.
Parameter ω0 corresponds to the frequency of the dipole resonance of the solid
inclusion in the soft medium caused by the restoring force from the medium.

SC
105 The effective stiffness and damping coefficients of the viscoelastic matrix with
periodic inclusions are determined in what follows.

U
3.2. Effective mass and effective stiffness

AN
A conventional way to account for the effects of a periodic lattice of scatterers
in a host matrix is to impose periodic boundary conditions at the edges of an
elementary cell comprising a single inclusion in the host medium [47–49]. An
M
alternative analytical approach corresponds to changing the shape of the cell.
For a two-dimensional hexagonal lattice, the unit cell can be approximately
modelled as a circular domain [50, 51]. Other cell shapes can also be modelled
D

as a circle with appropriate adjustment of filling fraction [51]. The elementary

TE

cell then reduces to the motion of a rigid cylindrical inclusion within a larger
concentric cylinder filled with the soft elastic medium [32, 51]. The mass ratio
β, effective mass ma and effective stiffness coefficient k can be obtained as [32]
EP

1+α
β= , (7)
1−α
C

ρvi (1 + α)
ma = , (8)
1−α
AC

8πµ0 (1 + α2 )
k= , (9)
(1 + α2 ) ln(1/α) + (α2 − 1)
where µ0 is the real part of the shear modulus of the host elastic medium as
defined earlier and α is the filling fraction. To convert the case of a cylindrical
110 inclusion of radius a inside a concentric cylindrical host medium of radius b to

7
 ACCEPTED MANUSCRIPT

the case of a square lattice of inclusions in a host medium with filling fraction
α = πa2 /h2 , the radius of the cylindrical host medium b should be related to
√

PT
the spacing between the inclusions h according to b = h/ π.
Substituting Eqs. (7) and (9) in Eq. (6) leads to the following expression for
the dipole resonance frequency

RI
cs
ω0 = γ , (10)
a

SC
where 1/2
8(1 + α2 )

γ= , (11)
(δ + β)((1 + α2 ) ln(1/α) + (α2 − 1))
p
is the non-dimensional dipole resonance frequency and cs = µ0 /ρ is the shear

U
115 wave speed in the host medium. It is noteworthy that in a soft elastic medium,

AN
the dipole resonance frequency ω0 is determined by the shear modulus of the
host material and the filling fraction of the scatterers.
As the filling fraction reduces, the dipole resonance frequency of an array
M
of hard cylinders in an elastic medium tends to zero, since γ 2 ∝ 1/ ln(1/α)
120 as α → 0. This result is in striking contrast to the case of an array of hard
spheres in an elastic medium where the dipole resonance frequency tends to
D

p
a non-zero limit as the filling fraction reduces, with γ = 9ρ/(2ρi + ρ) [31].
TE

Further, as filling fraction α → 0, the dipole resonance frequency is unaffected

by the arrangement of the scatterers as previously reported [34]. Equation (10)
125 can therefore be used as the first approximation for any lattice type.
EP

3.3. Effective damping

An analytical expression for damping in the soft elastic medium embedded
C

with a solid cylinder is derived using a well-known analogy between a viscous

fluid and a viscoelastic medium, whereby the governing equation and boundary
AC

conditions for the displacement vector of an elastic solid has the same form as
the respective equation and boundary conditions governing the velocity vector
in a viscous fluid [42, 52, 53]. Similar to the previous section, the oscillatory
motion of a cylindrical inclusion of radius a inside a cylindrical shell of radius
b filled with a viscous fluid is used to estimate the damping coefficient for the

8
 ACCEPTED MANUSCRIPT

rectangular lattice of cylindrical inclusions. The frequency dependent viscous

damping coefficient ξ is given by [54, 55]

PT
ξ = ρvi ωIm(H), (12)

where ω is the frequency of oscillations,

RI
N
H=1− , (13)
D

SC
N = 2τ 2 [Q00 (τ, χ) − Q00 (χ, τ )] − 4τ [Q10 (τ, χ) + Q01 (χ, τ )]
√ √ (14)
+4τ α[Q01 (τ, χ) + Q10 (χ, τ )] − 8 α[Q11 (τ, χ) − Q11 (χ, τ )],

U
√
AN
D = τ 2 (1 − α)[Q00 (τ, χ) − Q00 (χ, τ )]
(15)
+2τ α[Q01 (τ, χ) − Q10 (χ, χ) + Q10 (χ, τ ) − Q00 (χ, χ)]
M
+2τ α[Q01 (χ, τ ) − Q01 (τ, τ ) + Q10 (τ, χ) − Q10 (τ, τ )],
√ p
where τ = iσ, χ = iσ/α, σ = ωa2 /ν, α = (a/b)2 is the filling fraction, ν is
D

the kinematic viscosity, Qmn (τ, χ) = Im (τ )Kn (χ), and In , Kn are the modified
Bessel functions of the first and second kind, respectively, of the order n. For a
TE

soft elastic medium, the kinematic viscosity is replaced with [42, 43]

µ
ν= . (16)
iρω
EP

The effective damping coefficient for solid cylindrical inclusions embedded in

a soft material can now be obtained by substituting Eq. (12) into Eq. (5) which
C

yields
ω
ψ= Im(H). (17)
δ+β
AC

Note that if the shear modulus µ is real, then H is real and the viscous damping
coefficient given by Eq. (12) is zero.
The asymptotic behaviour of the effective damping coefficient ψ as a function
of frequency can be deduced as follows. The viscous damping coefficient ξ and
dipole resonance frequency ω0 are both regarded as real. If the shear modulus

9
 ACCEPTED MANUSCRIPT

is complex, then from Eqs. (4) and (10), the imaginary component of the dipole
resonance frequency leads to the non-dimensional effective damping coefficient

PT
of the form
ψ ω0 µ00
≈ . (18)
ω0 ω µ0
This mechanism describes viscous dissipation in the near field of an oscillating

RI
130 scatterer and is dominant as ω → 0, where multiple scattering can be ignored.
Hence, at low frequencies, the main contribution to ψ arises from the imaginary

SC
component of the complex shear modulus µ00 .
At high frequencies, another mechanism enhancing sound dissipation be-
comes more important, namely, the conversion of longitudinal waves into shear

U
waves. The effective damping coefficient ψ, having dimension of frequency, cor-

AN
responds to the inverse time scale of the conversion of longitudinal to shear
waves, that is, ψ ' 1/τ0 . The time scale τ0 can be evaluated from the esti-
mate that dE/dt ∼ E/τ0 ∼ P , where E is the kinetic energy of the oscillat-
M
ing inclusion and P is the total power of the generated shear waves. Using
E ' (m + ma )u̇2 /2, where m is the mass per unit length of the cylindrical
scatterers and P ' 2πaρcs u̇2 , it follows that as ω → ∞
D

ψ 1 4
≈ . (19)
ω0 γ 1+δ
TE

Hence at high frequencies, the non-dimensional effective damping coefficient

approaches a saturation frequency independent limit. The asymptotes given by
EP

135 Eqs. (18) and (19) will be validated numerically in the results section.

3.4. Homogenised properties of a unit cell

The layer of scatterers can be modelled as a homogenized layer with effective
C

geometric and acoustic properties using a similar homogenization approach to

AC

that proposed in [19]. Figure 2 schematically shows the layer of scatterers

approximated by the homogenized layer of effective thickness le . For an array
of cylindrical inclusions of radius a separated by spacing h, the effective thickness
of the homogenized layer can be computed using [56]
2h  πa 
le = ln sec . (20)
π h

10
 ACCEPTED MANUSCRIPT

2a

PT
𝑙𝑒

RI
h

SC
t s t s

U
Figure 2: Schematic diagram illustrating the layer of scatterers approximated by a homoge-
nized layer with effective geometric and material properties.

AN
The effective stiffness of the cylinders in the elastic medium averaged over
the unit cell, κg , can be obtained from the ‘inverse rule of mixtures’ as [57, 58]
M
κκi
κg = , (21)
ακ + κi − ακi
where κi is the longitudinal modulus of the steel inclusion and κ = λ + 2µ is the
D

longitudinal modulus of the host elastic medium. For more accurate modelling,
the bulk modulus of the inclusion should be used instead of its longitudinal
TE

140 modulus as the inclusions undergo uniform compression. However, for a soft
host medium, the effect of the near-rigid inclusion compressibility on the total
EP

compressibility is negligible. Further, the bulk modulus of the steel inclusion

has a similar magnitude to the longitudinal modulus.
The effective global density ρg was derived from Eq. (4) and is given by [32]
C

ρg = ρ0g − iρ00g , (22)

AC

where
ρ0g ((ω0 /ω)2 − (1 + β)/(δ + β))((ω0 /ω)2 − 1) + (ψ/ω)2
= 1 + α(δ − 1) , (23)
ρ ((ω0 /ω)2 − 1)2 + (ψ/ω)2

ρ00g αψ(δ − 1)2

= , (24)
ρ ω(δ + β)(((ω0 /ω)2 − 1)2 + (ψ/ω)2 )

11
 ACCEPTED MANUSCRIPT

and ω0 , ψ are given by Eqs. (10) and (17), respectively. In the limit of no
145 inclusions (α = 0), or when the density of the inclusions is equal to the density

PT
of the host medium (δ = 1), Eq. (22) reduces to ρg = ρ.
The effective density ρe and effective longitudinal modulus κe of the homog-
enized layer are then obtained by ‘weighted averaging’ over a periodic cell which

RI
is assumed to be square with length h, as follows [19, 59]

ρg + ρ(αe − 1)
ρe = , (25)

SC
αe

κκg αe
κe = , (26)
κ − κg + αe κg

U
where αe = le /h and ρg , κg are given by Eqs. (21) and (22), respectively.

3.5. Sound absorption coefficient

AN
The effective impedance of the homogenized layer is computed from the
√
M
150 effective material properties as Ze = ρe κe . The impedance of the remainder
√
of the matrix is simply that of the host medium, Z = ρκ. The transmitted
and reflected pressures are then obtained using the transfer matrix method in
D

terms of the impedance of the host medium, impedance and thickness of the
TE

homogenized layer, width and thickness of the host rubber medium and steel
155 plate, and impedance of the fluids on the incidence and transmission sides of
the phononic crystal. The absorption coefficient is then calculated in terms of
EP

the incident, reflected and transmitted pressures, as described in Ref. [20].

3.6. Conditions of model validity

C

The analytical model developed in this work is suitable for large sound wave-
AC

lengths compared with the inclusion radius and spacing. It can be shown that
the sound wavelength is greater than the scatterer radius near the dipole reso-
p
nance. For longitudinal speed in the rubber medium given by cl = κ/ρ, then
from Eq. (10), ω0 a/cl = γcs /cl , which is always small due to Eq. (1). Further-
more, for validity of the effective properties obtained using the homogenisation

12
 ACCEPTED MANUSCRIPT

approach employed here, the longitudinal wavelength should be greater than

the spacing between the scatterers. This leads to the condition

PT
h h cs
ω0 =γ  1, (27)
cl a cl
which can be violated only for very low filling fraction. The homogenisation

RI
160 approach is also likely to fail at high filling fractions. Additionally, the density
of inclusions should be much greater that the density of the host medium in order
to justify the expression for the effective thickness given by Eq. (20) that was

SC
derived for rigid scatterers. The geometrical and material parameters chosen
for the results in the numerical simulations satisfy all these criteria.

U
165 4. Numerical model

AN
The numerical model was implemented using the commercial software COM-
SOL Multiphysics (v5.2a) which employs the finite element method. The host
rubber medium, scatterers and steel backing were modelled as solid domains;
M
water and air were modelled as fluid domains. A physical boundary condition
170 was applied at the interface between the solid and fluid domains to simulate the
D

structural-acoustic interaction. For computation efficiency, only a periodic cell

comprising a single scatterer as shown in Fig. 1 was modelled. The infinite do-
TE

main in the direction normal to sound propagation was simulated using periodic
boundary conditions. In the fluid domain on the incidence side, a plane wave
of unity pressure amplitude was applied to simulate the incident acoustic pres-
EP

175

sure. Perfectly matched layer boundary conditions were used to model anechoic
termination of the outgoing waves at the boundaries of the fluid domains. The
C

reflected and transmitted pressures were recorded at the interface between the
fluid and solid domains on the incidence and transmission sides of the phononic
AC

180 crystal, respectively.

5. Results and discussion

The material properties of the solid and fluid media are listed in Table 1. The
real values of the bulk and shear moduli of the host rubber medium represent

13
 ACCEPTED MANUSCRIPT

the elastic component of the moduli whereas the imaginary values represent the
185 loss factors. The shear modulus of the elastic medium is low compared to the

PT
bulk modulus, implying that the rubber medium is soft. The loss factors for
steel are very small compared to rubber and as such assumed to be zero, (i.e.
so the bulk and shear moduli of steel were assumed to be real). Further, the

RI
material properties of rubber and steel were assumed to be independent of the
190 forcing frequency. The bulk modulus of an isotropic material is K = λ + 2µ/3

SC
and hence the longitudinal modulus of the host rubber medium is κ = K +4µ/3.

Table 1: Material properties.

3
Solids Density (kg/m ) Bulk modulus (GPa) Shear modulus (GPa)

U
Rubber 1000 1+0.01i 0.006+0.0018i
Steel
Fluids
7890
Density (kg/m3 )
AN 175 80.77
Longitudinal speed of sound (m/s)
Water 1000 1500
M
Air 1.2 340
D

Figure 3 presents the non-dimensional dipole resonance frequency of the pe-

riodic array of steel cylindrical scatterers in rubber for a range of filling fraction
TE

values, computed analytically using Eq. (11) and numerically using the finite
195 element method. In the numerical results, the resonance frequency corresponds
to the frequency of the first absorption peak after observation of the deformation
EP

plots at the peak absorption frequencies, and is discussed in more detail later
in the paper. The results in Fig. 3 show that the resonance frequency predicted
analytically and numerically are in very good agreement. The non-dimensional
C

200 dipole resonance frequency of the steel cylinders embedded in rubber increases
AC

with an increase in cylinder radius or a reduction in lattice spacing between the

cylinders.
Figure 4 shows the non-dimensional damping coefficient ψ/ω0 as a function
of non-dimensional frequency ω/ω0 for different filling fractions α. On this
205 logarithmic plot, the inclined parallel lines at low frequencies obtained using

14
 ACCEPTED MANUSCRIPT

101

Non-dimensional dipole resonance

PT
100

RI
SC
10-1 -3
10 10-2 10-1 100

U
Filling fraction

AN
Figure 3: Non-dimensional dipole resonance frequency γ as a function of filling fraction
α = πa2 /h2 of a periodic array of steel cylindrical scatterers in an elastic medium obtained
analytically using Eq. (11) (solid line) and numerically (circles).
M
Eqs. (10) and (17) approximately correspond to the inverse frequency asymptote
given by Eq. (18), whereby the latter is independent of α. At high frequencies,
D

ψ/ω0 scales inversely with the non-dimensional dipole resonance frequency γ and
corresponds to the asymptote of Eq. (17) given by Eq. (19). The high frequency
TE

210 limit also follows inverse scaling with the density ratio (δ = ρi /ρ) as predicted by
Eq. (19). The agreement between the non-dimensional damping coefficient and
EP

its asymptotes improves as the filling fraction is reduced. A minimum value

in the non-dimensional damping coefficient always exists between the inverse
frequency asymptote given by Eq. (18) and the frequency independent limit
C

215 given by Eq. (19).

The density of the host rubber medium, steel scatterers and homogenized
AC

layer as a function of frequency are presented in Fig. 5. The scatterer radius,

lattice spacing and thickness of the rubber are a = 1 cm, h = 5 cm and t = 10
cm, respectively. Unlike the density of the pure materials, the effective density
220 of the homogenized layer is complex. At low frequencies, the effective density
of the homogenized layer is similar to the density of steel. Around the dipole

15
 ACCEPTED MANUSCRIPT

PT
RI
101

SC
Non-dimensional damping

U
100
AN
M
10-1 -1
10 100 101 102
D

Non-dimensional frequency
TE

Figure 4: (Colour online) Non-dimensional damping coefficient ψ/ω0 as a function of non-

dimensional frequency ω/ω0 for filling fractions α of 0.03 (solid black line), 0.1 (solid red line)
and 0.3 (solid blue line), obtained using Eqs. (10) and (17). The dashed lines (overlapped)
and dotted lines represent the corresponding low frequency and high frequency asymptotes
EP

of ψ/ω0 given by Eqs. (18) and (19), respectively. The values of non-dimensional dipole
resonance frequency γ corresponding to α of 0.03, 0.1 and 0.3 are 0.597, 0.815 and 1.49,
respectively.
C
AC

16
 ACCEPTED MANUSCRIPT

×104
1

PT
0.5
Density (kg/m3)

RI
0

-0.5

SC
-1
0 2 4 6 8 10

U
Frequency (kHz)

AN
Figure 5: (Colour online) Real (black lines) and imaginary (red lines) components of the den-
sity of the host rubber medium (dashed lines), steel scatterers (dotted lines) and homogenized
layer (solid lines) for the system geometry discussed in the text. The imaginary component
M
of density of the rubber and steel are zero.

resonance frequency, the real part of the effective density becomes greater than
D

the density of steel and then decreases to a very low value. The imaginary part
of the density of the homogenized layer is negative and tends to zero as frequency
TE

225 tends to zero. It becomes largely negative around the resonance frequency and
becomes close to zero at frequencies greater than the resonance frequency.
Figure 6 presents the effective phase speed and attenuation in the host rubber
EP

medium and homogenized layer for the same geometry as Fig. 5. At all frequen-
cies, the phase speed in the homogenized layer is lower compared to the phase
C

230 speed in the host rubber medium due to scattering by the presence of the steel
cylinders. The phase speed in the homogenized layer increases with frequency
AC

around dipole resonance, stabilises, then decreases slightly to a relatively con-

stant value. Unlike the phase speed, the attenuation in the homogenized layer
is significantly higher than the attenuation in the host rubber medium, partic-
235 ularly at frequencies around and above the dipole resonance frequency due to
the strong conversion of longitudinal waves into highly-damped shear waves.

17
 ACCEPTED MANUSCRIPT

104

Speed (m/s) or Attenuation (dB/m)

PT
102

RI
100

SC
10-2
0 2 4 6 8 10

U
Frequency (kHz)

AN
Figure 6: (Colour online) Phase speed (black lines) and attenuation (red lines) in the host
rubber medium (dashed lines) and the homogenized layer (solid lines).
M
The acoustic performance of a layer of periodic steel cylinders embedded in a
viscoelastic medium with water backing and steel-air backing is now examined.
The scatterer radius, lattice spacing, thickness of the rubber and thickness of the
D

240 steel plate are a = 1 cm, h = 5 cm, t = 10 cm, s = 2 cm, respectively. In Fig. 7,
TE

the absorption coefficients of the phononic crystal with water or steel-air backing
are compared. Results from a corresponding uniform rubber medium in the
absence of periodic scatterers are also presented, showing excellent agreement
EP

between analytical and numerical results for both the water and steel-air backing
245 cases. In the absence of the scatterers, the absorption is very low for the water
backing case since the impedance of water is similar to the impedance of rubber,
C

thereby resulting in high transmission and low reflection of sound waves. For the
AC

case of steel-air backing, the high impedance mismatch between rubber and steel
causes low transmission and high reflection of waves, also resulting in low sound
250 absorption. When scatterers are present, dipole resonance occurs, producing the
first peak in the absorption coefficient at around 1190 Hz for the water backing
case and 1180 Hz for the steel-air backing case. The dipole resonance results in

18
 ACCEPTED MANUSCRIPT

Absorption coefficient
0.8

PT
0.6

RI
0.4

SC
0.2

0
0 2 4 6 8 10

U
Frequency (kHz)

AN
Figure 7: (Colour online) The effect of backing materials on the absorption coefficient of a
layer of steel cylindrical inclusions in rubber. Results for the phononic crystal with water
backing (red lines) and steel-air backing (black lines) obtained analytically (solid lines) and
M
numerically (dashed lines) are compared with results for uniform rubber with water backing
(green lines) and steel-air backing (blue lines).
D

high sound absorption due to increased damping associated with the conversion
of longitudinal to shear waves, as shown earlier in Fig. 6. The second peak in the
TE

255 absorption coefficient for the water backing case, occurring at around 3850 Hz, is
due to interference between the waves scattered by the cylinders. The frequency
of this peak is subsequently shown to be independent of the thickness of the host
EP

rubber medium. The second peak in the absorption coefficient for the steel-air
backing case, occurring at around 4730 Hz, is due to interference between the
waves scattered by the cylinders and waves reflected from the steel plate, and is
C

260

subsequently shown to depend on the distance between the scatterers and the
AC

steel backing plate.

To assist in the interpretation of the physical mechanisms governing the
acoustic performance, deformation plots at the first and second absorption peak
265 frequencies of the phononic crystal for both the water backing and steel-air back-
ing cases are presented in Fig. 8. These plots are obtained numerically using

19
 ACCEPTED MANUSCRIPT

PT
RI
U SC
(a)
AN (b)
M
D
TE
EP

(c) (d)

Figure 8: (Colour online) Deformation amplitude (in metres) of a layer of periodic steel
C

cylinders in rubber with water backing at (a) 1190 Hz and (b) 3850 Hz, and with steel-air
backing at (c) 1180 Hz and (d) 4730 Hz.
AC

20
 ACCEPTED MANUSCRIPT

the finite element method. For both the water and steel-air backing cases, the
first absorption peak corresponds to dipole resonance at which the scatterers

PT
oscillate in the host rubber medium in the direction of sound propagation. At
270 frequencies away from the dipole resonance frequency, the cylinders act like rigid
scatterers. The second peak in the absorption coefficient for water backing is due

RI
to constructive interference between waves scattered by the cylinders, leading
to high deformation of the rubber medium between the cylinders. For steel-air

SC
backing, the second peak in the absorption coefficient is due to constructive in-
275 terference between waves scattered by the cylinders and reflected from the steel
plate due to the high impedance mismatch between rubber and steel, also lead-

U
ing to high deformation of the rubber medium between the cylinders as shown
later in the paper. The physical mechanisms associated with both absorption

280
AN
peaks are attributed to conversion of longitudinal to shear waves, which are sub-
sequently dissipated in the elastic medium which has very high shear damping,
M
as previously reported [60, 61].
The effects of different geometric and material parameters on the acous-
tic performance of the phononic crystal are now examined. Figure 9 presents
D

the absorption coefficients for the layer of steel cylinders in rubber of different
thickness in the direction of sound propagation with water backing (Fig. 9(a))
TE

285

or steel-air backing (Fig. 9(b)). The first peak corresponding to the dipole reso-
nance frequency is not significantly affected by variation in the thickness of the
EP

rubber medium for the two types of backing. For the water backing case, the
frequency of the second absorption peak is also mostly unaffected as the rubber
290 thickness is varied. However for the steel-air backing case, the frequency of the
C

second absorption peak increases as the thickness is reduced, attributed to the

fact that the reflected waves from the steel plate travel a shorter distance in the
AC

host medium before interference with the waves scattered by the cylinders.
The effect of cylinder radius on the absorption coefficient is presented in
295 Fig. 10. For both backing cases, the dipole resonance frequency exhibits non-
monotonic behaviour with variation in cylinder radius a. This is attributed to
the fact that the dipole resonance ω0 given by Eq. (10) is a function of the

21
 ACCEPTED MANUSCRIPT

PT
Absorption coefficient
0.8

RI
0.6

0.4

SC
0.2

U
0
0 2 4 6 8 10
AN
Frequency (kHz)
(a)
1
M
Absorption coefficient

0.8
D

0.6
TE

0.4

0.2
EP

0
0 2 4 6 8 10
Frequency (kHz)
C

(b)
AC

Figure 9: (Colour online) The effect of rubber thickness on the absorption coefficient of a
layer of steel cylindrical inclusions in rubber with (a) water backing and (b) steel-air backing.
Results are obtained analytically (solid lines) and numerically (dashed lines). The rubber
medium has thickness of 8 cm (blue lines), 10 cm (black lines) and 12 cm (red lines).

22
 ACCEPTED MANUSCRIPT

PT
Absorption coefficient
0.8

RI
0.6

0.4

SC
0.2

U
0
0 2 4 6 8 10
AN
Frequency (kHz)
(a)
1
M
Absorption coefficient

0.8
D

0.6
TE

0.4

0.2
EP

0
0 2 4 6 8 10
Frequency (kHz)
C

(b)
AC

Figure 10: (Colour online) The effect of cylinder radius on the absorption coefficient of a
layer of steel cylindrical inclusions in rubber with (a) water backing and (b) steel-air backing.
Results are obtained analytically (solid lines) and numerically (dashed lines). The cylinders
have radius of 0.5 cm (red lines) 1 cm (black lines) and 1.5 cm (blue lines).

23
 ACCEPTED MANUSCRIPT

PT
Absorption coefficient 0.8

RI
0.6

0.4

SC
0.2

U
0
0 2 4 6 8 10
Frequency (kHz)
AN (a)
1
M
Absorption coefficient

0.8
D

0.6
TE

0.4

0.2
EP

0
0 2 4 6 8 10
Frequency (kHz)
C

(b)
AC

Figure 11: (Colour online) The effect of shear modulus of the host rubber medium on the
absorption coefficient of a layer of steel cylindrical inclusions in rubber with (a) water backing
and (b) steel-air backing. Results are obtained analytically (solid lines) and numerically
(dashed lines). The shear modulus of the host rubber medium is 10% (blue lines) and 50%
(red lines) of the original shear modulus (black lines).

24
 ACCEPTED MANUSCRIPT

ratio γ/a, whereby ω0 decreases with increase in a whilst the non-dimensional

dipole resonance γ increases with a (see Fig. 3). However, the absorption peak

PT
300 significantly increases for both backing cases with an increase in cylinder radius,
attributed to greater conversion of longitudinal waves into shear waves due to
oscillation of the cylinders. Further, the absorption at the dipole resonance

RI
frequency for the water backing case is slightly higher compared to the steel-
air backing case, as the presence of the steel plate restricts the motion of the

SC
305 cylinders. For the water backing case, the frequency of the second peak moves
to a higher frequency as the cylinder radius increases, attributed to the fact
that the scattered waves travel a shorter distance before interference due to

U
reduced spacing between the cylinders. In contrast, for the steel-air backing
case, the frequency of the second absorption peak shifts to a lower frequency
310
AN
with an increase in cylinder radius. However, the absorption at the second
peak increases with an increase in the cylinder radius, attributed to increased
M
interference between the scattered and reflected waves as a result of reduced
spacing between the cylinders.
Figure 11 shows the effects of the shear modulus of the host rubber medium
D

315 on the sound absorption. The complex shear modulus of the host medium is
varied to be 10% and 50% of its original value while keeping all other material
TE

and geometric parameters constant. For both backing cases, the first absorption
peak corresponding to the dipole resonance frequency of the scatterers shifts to
EP

a lower frequency as the shear modulus is reduced. For the water backing case,
320 at higher frequencies, the amplitude of sound absorption reduces with reduction
in the shear modulus due to the reduction in shear damping of the host rubber.
C

For the steel-air backing case, the frequency of the second peak is unchanged
as it depends mainly on the longitudinal speed of sound in the medium and
AC

the distance between the scatterers and the steel backing. Similar to the water
325 backing case, the amplitude of sound absorption around the second peak also
reduces with reduction in the shear modulus due to the reduction in shear
damping of the host rubber.

25
 ACCEPTED MANUSCRIPT

6. Conclusions

The acoustic performance of a phononic crystal comprising periodic steel

PT
330 cylinders in a viscoelastic rubber medium with water backing or steel-air back-
ing has been analytically and numerically investigated. In the analytical model,

RI
the equation of motion describes oscillating motion of a solid inclusion in an
elastic medium arising from an incident sound wave, and incorporates effective
mass, stiffness and damping properties of the host elastic medium. The effective

SC
335 density, longitudinal modulus and thickness of the layer of scatterers were de-
rived using effective medium approximation. Using the effective impedance and
thickness of the homogenized layer as well as the impedances of the host elas-

U
tic medium and fluids on the incidence and transmission sides of the phononic

340
AN
crystal, the reflected and transmitted pressures were then calculated from which
the absorption coefficient was subsequently obtained. Conditions under which
the analytical model is valid have been described. Analytical and numerical
M
results showed that dipole resonance arising from oscillatory motion in the di-
rection of sound propagation of the solid cylinders in an elastic medium leads to
D

high sound absorption. A second absorption peak was also observed associated
345 with interference between waves scattered by the cylinders and reflected from
TE

the steel backing plate. The effects of different geometric and material param-
eters on the sound absorption were discussed. The approach presented here
may be beneficial for tailored designs of acoustic metamaterials for underwater
EP

applications, including rapid prototyping and trend analysis.

References
C

350

[1] A. Khelif, A. Adibi, Phononic Crystals: Fundamentals and Applications,

AC

Springer, 2015.

[2] M. R. Haberman, A. N. Norris, Acoustic metamaterials, Acoustics Today

12 (3) (2016) 31–39.

26
 ACCEPTED MANUSCRIPT

355 [3] G. Ma, P. Sheng, Acoustic metamaterials: From local resonances to broad
horizons, Science Advances 2 (2) (2016) e1501595.

PT
[4] S. H. Lee, O. B. Wright, Origin of negative density and modulus in acoustic
metamaterials, Physical Review B 93 (2) (2016) 024302.

RI
[5] P. A. Deymier, Acoustic metamaterials and phononic crystals, Vol. 173,
360 Springer Science & Business Media, 2013.

SC
[6] X. Zhou, X. Liu, G. Hu, Elastic metamaterials with local resonances: an
overview, Theoretical and Applied Mechanics Letters 2 (4) (2012) 041001.

[7] N. Aravantinos-Zafiris, M. Sigalas, M. Kafesaki, E. Economou, Phononic

U
crystals and elastodynamics: Some relevant points, AIP Advances 4 (12)
365 (2014) 124203. AN
[8] M. Hinders, B. Rhodes, T. Fang, Particle-loaded composites for acoustic
M
anechoic coatings, Journal of Sound and Vibration 185 (2) (1995) 219–246.

[9] S. M. Ivansson, Sound absorption by viscoelastic coatings with periodi-

D

cally distributed cavities, The Journal of the Acoustical Society of America

370 119 (6) (2006) 3558–3567.
TE

[10] S. M. Ivansson, Numerical design of Alberich anechoic coatings with su-

perellipsoidal cavities of mixed sizes, The Journal of the Acoustical Society
of America 124 (4) (2008) 1974–1984.
EP

[11] V. Leroy, A. Bretagne, M. Fink, H. Willaime, P. Tabeling, A. Tourin,

375 Design and characterization of bubble phononic crystals, Applied Physics
C

Letters 95 (17) (2009) 171904.

AC

[12] V. Leroy, A. Strybulevych, M. Lanoy, F. Lemoult, A. Tourin, J. H. Page,

Superabsorption of acoustic waves with bubble metascreens, Physical Re-
view B 91 (2) (2015) 020301.

380 [13] Z. Liu, X. Zhang, Y. Mao, Y. Zhu, Z. Yang, C. T. Chan, P. Sheng, Locally
resonant sonic materials, Science 289 (5485) (2000) 1734–1736.

27
 ACCEPTED MANUSCRIPT

[14] M. Duranteau, T. Valier-Brasier, J.-M. Conoir, R. Wunenburger, Random

acoustic metamaterial with a subwavelength dipolar resonance, The Jour-

PT
nal of the Acoustical Society of America 139 (6) (2016) 3341–3352.

385 [15] J.-P. Groby, C. Lagarrigue, B. Brouard, O. Dazel, V. Tournat, B. Nennig,

Using simple shape three-dimensional rigid inclusions to enhance porous

RI
layer absorption, The Journal of the Acoustical Society of America 136 (3)
(2014) 1139–1148.

SC
[16] G. Gaunaurd, One-dimensional model for acoustic absorption in a vis-
390 coelastic medium containing short cylindrical cavities, The Journal of the

U
Acoustical Society of America 62 (2) (1977) 298–307.

AN
[17] S.-C. S. Lin, T. J. Huang, Tunable phononic crystals with anisotropic in-
clusions, Physical Review B 83 (17) (2011) 174303.

[18] T. Meng, H. Hong-Xing, Improved low-frequency performance of a com-

M
395 posite sound absorption coating, Journal of Vibration and Control 18 (1)
(2012) 48–57.
D

[19] G. S. Sharma, A. Skvortsov, I. MacGillivray, N. Kessissoglou, Sound trans-

TE

mission through a periodically voided soft elastic medium submerged in

water, Wave Motion 70 (2017) 101–112.

[20] G. S. Sharma, A. Skvortsov, I. MacGillivray, N. Kessissoglou, Acoustic

EP

400

performance of gratings of cylindrical voids in a soft elastic medium with

a steel backing, The Journal of the Acoustical Society of America 141 (6)
C

(2017) 4694–4704.
AC

[21] V. Romero-Garcı́a, A. Krynkin, L. Garcia-Raffi, O. Umnova, J. Sánchez-

405 Pérez, Multi-resonant scatterers in sonic crystals: Locally multi-resonant
acoustic metamaterial, Journal of Sound and Vibration 332 (1) (2013) 184–
198.

28
 ACCEPTED MANUSCRIPT

[22] P. Martin, A. Maurel, W. Parnell, Estimating the dynamic effective mass

density of random composites, The Journal of the Acoustical Society of

PT
410 America 128 (2) (2010) 571–577.

[23] A. N. Norris, J.-M. Conoir, Multiple scattering by cylinders immersed in

fluid: High order approximations for the effective wavenumbers, The Jour-

RI
nal of the Acoustical Society of America 129 (1) (2011) 104–113.

SC
[24] J.-M. Conoir, A. N. Norris, Effective wavenumbers and reflection coeffi-
415 cients for an elastic medium containing random configurations of cylindrical
scatterers, Wave Motion 47 (3) (2010) 183–197.

U
[25] A. N. Norris, Acoustic cloaking, Acoustics Today 11 (1) (2015) 38–46.

AN
[26] V. Varadan, Y. Ma, V. Varadan, Multiple scattering of compressional and
shear waves by fiber-reinforced composite materials, The Journal of the
420 Acoustical Society of America 80 (1) (1986) 333–339.
M
[27] V. K. Varadan, V. V. Varadan, Y.-H. Pao, Multiple scattering of elastic
waves by cylinders of arbitrary cross section. I. SH waves, The Journal of
D

the Acoustical Society of America 63 (5) (1978) 1310–1319.

TE

[28] A. C. Spero, C. M. Godoy, A. Harari, J. M. Teague, Pressure resistant

425 anechoic coating for undersea platforms, US Patent 7,205,043 (Apr. 17
2007).
EP

[29] P. Méresse, C. Audoly, C. Croënne, A.-C. Hladky-Hennion, Acoustic coat-

ings for maritime systems applications using resonant phenomena, Comptes
C

Rendus Mecanique 343 (12) (2015) 645–655.

AC

430 [30] M. Caleap, B. W. Drinkwater, Metamaterials: supra-classical dynamic ho-

mogenization, New Journal of Physics 17 (12) (2015) 123022.

[31] I. Chaban, Calculation of the effective parameters of microinhomogeneous

media by the self-consistent field method, Soviet Physics - Acoustics 11
(1965) 81.

29
 ACCEPTED MANUSCRIPT

435 [32] V. Fedotovskii, T. Vereschagina, Resonance dispersion of the longitudinal

waves in disperse composites, Acoustical Physics 56 (4) (2010) 506–513.

PT
[33] R.-B. Yang, A. K. Mal, Multiple scattering of elastic waves in a fiber-
reinforced composite, Journal of the Mechanics and Physics of Solids 42 (12)
(1994) 1945–1968.

RI
440 [34] J. Mei, Z. Liu, W. Wen, P. Sheng, Effective mass density of fluid-solid

SC
composites, Physical Review Letters 96 (2) (2006) 024301.

[35] J. Mei, Z. Liu, W. Wen, P. Sheng, Effective dynamic mass density of com-
posites, Physical Review B 76 (13) (2007) 134205.

U
[36] V. Tournat, V. Pagneux, D. Lafarge, L. Jaouen, Multiple scattering of
445

(2004) 026609.
AN
acoustic waves and porous absorbing media, Physical Review E 70 (2)
M
[37] T. Weisser, J.-P. Groby, O. Dazel, F. Gaultier, E. Deckers, S. Futatsugi,
L. Monteiro, Acoustic behavior of a rigidly backed poroelastic layer with
periodic resonant inclusions by a multiple scattering approach, The Journal
D

450 of the Acoustical Society of America 139 (2) (2016) 617–629.

TE

[38] C. Lagarrigue, J.-P. Groby, O. Dazel, V. Tournat, Design of metaporous

supercells by genetic algorithm for absorption optimization on a wide fre-
quency band, Applied Acoustics 102 (2016) 49–54.
EP

[39] J.-P. Groby, C. Lagarrigue, B. Brouard, O. Dazel, V. Tournat, B. Nen-

455 nig, Enhancing the absorption properties of acoustic porous plates by pe-
C

riodically embedding Helmholtz resonators, The Journal of the Acoustical

AC

Society of America 137 (1) (2015) 273–280.

[40] C. Lagarrigue, J. Groby, V. Tournat, O. Dazel, O. Umnova, Absorption

of sound by porous layers with embedded periodic arrays of resonant in-
460 clusions, The Journal of the Acoustical Society of America 134 (6) (2013)
4670–4680.

30
 ACCEPTED MANUSCRIPT

[41] J.-P. Groby, B. Brouard, O. Dazel, B. Nennig, L. Kelders, Enhancing rigid

frame porous layer absorption with three-dimensional periodic irregular-

PT
ities, The Journal of the Acoustical Society of America 133 (2) (2013)
465 821–831.

[42] Y. A. Kobelev, Analogy between linear sound scattering processes in a

RI
viscous liquid and in an isotropic elastic medium, Acoustical Physics 54 (6)
(2008) 769–773.

SC
[43] S. R. Aglyamov, A. B. Karpiouk, Y. A. Ilinskii, E. A. Zabolotskaya, S. Y.
470 Emelianov, Motion of a solid sphere in a viscoelastic medium in response

U
to applied acoustic radiation force: Theoretical analysis and experimen-
tal verification, The Journal of the Acoustical Society of America 122 (4)
(2007) 1927–1936. AN
[44] G. Bonnet, V. Monchiet, Low frequency locally resonant metamaterials
M
475 containing composite inclusions, The Journal of the Acoustical Society of
America 137 (6) (2015) 3263–3271.
D

[45] G. Gaunaurd, H. Überall, Resonance theory of the effective properties of

perforated solids, The Journal of the Acoustical Society of America 71 (2)
TE

(1982) 282–295.

480 [46] H. Zhao, Y. Liu, J. Wen, D. Yu, X. Wen, Tri-component phononic crystals
EP

for underwater anechoic coatings, Physics Letters A 367 (3) (2007) 224–232.

[47] Y. Wu, Y. Lai, Z.-Q. Zhang, Effective medium theory for elastic metama-
C

terials in two dimensions, Physical Review B 76 (20) (2007) 205313.

AC

[48] H. Meng, J. Wen, H. Zhao, X. Wen, Optimization of locally resonant acous-

485 tic metamaterials on underwater sound absorption characteristics, Journal
of Sound and Vibration 331 (20) (2012) 4406–4416.

[49] I. V. Andrianov, V. I. Bolshakov, V. V. Danishevs’ kyy, D. Weichert, Higher

order asymptotic homogenization and wave propagation in periodic com-

31
 ACCEPTED MANUSCRIPT

posite materials, Proceedings of the Royal Society A: Mathematical, Phys-

490 ical and Engineering Sciences 464 (2093) (2008) 1181–1201.

PT
[50] B. Eftekharinia, P. Parvin, H. Habibiyan, H. Hoseinzadeh, Analysis of am-
plification properties of a photonic crystal fiber, Optik-International Jour-
nal for Light and Electron Optics 125 (4) (2014) 1565–1571.

RI
[51] V. V. Tyutekin, Propagation of elastic waves in a medium with cylindrical

SC
495 channels, Soviet Physics. Acoustics 2 (3) (1956) 307–318.

[52] R. M. Christensen, Mechanics of composite materials, Courier Corporation,

2012.

U
[53] A. N. Norris, Impedance of a sphere oscillating in an elastic medium with

500 (2006) 2062–2066.

AN
and without slip, The Journal of the Acoustical Society of America 119 (4)
M
[54] S. S. Chen, M. W. Wambsganss, J. A. Jendrzejczyk, Added mass and
damping of a vibrating rod in confined viscous fluids, Journal of Applied
Mechanics 43 (2) (1976) 325–329.
D

[55] H. Chung, S. Chen, Hydrodynamic mass, Tech. rep., Argonne National

TE

505 Lab., IL (USA) (1984).

[56] P. A. Martin, R. A. Dalrymple, Scattering of long waves by cylindrical

EP

obstacles and gratings using matched asymptotic expansions, Journal of

Fluid Mechanics 188 (1988) 465–490.

[57] R. F. Landel, L. E. Nielsen, Mechanical properties of polymers and com-

C

510 posites, CRC Press, 1993.

AC

[58] N. S. Bakhvalov, G. Panasenko, Homogenisation: averaging processes in

periodic media: Mathematical problems in the mechanics of composite
materials, Springer Science & Business Media, 2012.

32
 ACCEPTED MANUSCRIPT

[59] J. Sánchez-Dehesa, A. Krokhin, Introduction to acoustics of phononic crys-

515 tals. homogenization at low frequencies, in: Phononic Crystals, Springer,

PT
2016, pp. 1–21.

[60] H. Zhao, Y. Liu, D. Yu, G. Wang, J. Wen, X. Wen, Absorptive properties of

three-dimensional phononic crystal, Journal of Sound and Vibration 303 (1)

RI
(2007) 185–194.

SC
520 [61] J. Wen, H. Zhao, L. Lv, B. Yuan, G. Wang, X. Wen, Effects of locally
resonant modes on underwater sound absorption in viscoelastic materials,
The Journal of the Acoustical Society of America 130 (3) (2011) 1201–1208.

U
AN
M
D
TE
C EP
AC

33