Journal of Sound and Vibration 409 (2017) 81–93

Contents lists available at ScienceDirect

Journal of Sound and Vibration

journal homepage: www.elsevier.com/locate/jsvi

Acoustic plane wave diffraction from a truncated semi-infinite

cone in axial irradiation
Dozyslav Kuryliak, Victor Lysechko n
Department of Physical Bases of Materials Diagnostics, Karpenko Physico-Mechanical Institute of NAS of Ukraine, Lviv, Ukraine

a r t i c l e i n f o abstract

Article history: The diffraction problem of the plane acoustic wave on the semi-infinite truncated soft and
Received 31 January 2017 rigid cones in the case of axial incidence is solved. The problem is formulated as a
Received in revised form boundary-value problem in terms of Helmholtz equation, with Dirichlet and Neumann
4 July 2017
boundary conditions, for scattered velocity potential. The incident field is taken to be the
Accepted 19 July 2017
total field of semi-infinite cone, the expression of which is obtained by solving the aux-
Handling Editor: A.V. Metrikine
iliary diffraction problem by the use of Kontorovich-Lebedev integral transformation. The
diffracted field is sought via the expansion in series of the eigenfunctions for subdomains
Keywords: of the Helmholtz equation taking into account the edge condition. The corresponding
Semi-infinite truncated cone
diffraction problem is reduced to infinite system of linear algebraic equations (ISLAE)
Plane acoustic wave
making use of mode matching technique and orthogonality properties of the Legendre
Analytical regularization procedure
functions. The method of analytical regularization is applied in order to extract the sin-
gular part in ISLAE, invert it exactly and reduce the problem to ISLAE of the second kind,
which is readily amenable to calculation. The numerical solution of this system relies on
the reduction method; and its accuracy depends on the truncation order. The case of
degeneration of the truncated semi-infinite cone into an aperture in infinite plane is
considered. Characteristic features of diffracted field in near and far fields as functions of
cone's parameters are examined.
& 2017 Elsevier Ltd All rights reserved.

1. Introduction

The semi-infinite truncated cones are related to canonical scatterers of the acoustic field. Presently there is not enough
theoretical data collected regarding its basic scattering features. The solution of diffraction problems from such scatterers
provides a reliable test bed for modelling of various elements of acoustic measuring and operating systems. The aperture,
obtained by cut of the cone apex, may be used to produce the necessary distribution of sound pressure. In addition, such
cones may be applicable as a model of object diagnostics or for controlling the flow of fluid through an aperture.
Problems of diffraction of electromagnetic and acoustic waves on finite cones incorporating the edge have been the
subject of many studies in literature [1–6]. In these publications, hollow and closed cones were examined by different
asymptotic and rigorous approaches such as geometrical theory of diffraction [1], mode matching technique [2–5] and
Wiener-Hopf method [6]. The finite cones in the electromagnetic case are also widely considered numerically [7,8].
For solution of the diffraction problem on finite hollow cones in [9,10] the analytical regularization procedure is developed.

n
Corresponding author.
E-mail address: vtlysechko@gmail.com (V. Lysechko).

http://dx.doi.org/10.1016/j.jsv.2017.07.035
0022-460X/& 2017 Elsevier Ltd All rights reserved.
82 D. Kuryliak, V. Lysechko / Journal of Sound and Vibration 409 (2017) 81–93

The basis of this procedure is the exact inversion of the singular part of the operator of diffraction problem. The effectiveness
and the power of this approach for the solution of diffraction problem on finite acoustic cones is demonstrated in [11]. The
regularization idea [9,10] is very close to the ideas which have been developed in [12–14] for electromagnetic and elastic
problems.
Another class of canonical diffraction structures which has been intensively studied in acoustic and electromagnetic
contexts with impedance and penetrable boundary conditions are the infinite cones [15–17] (and wedges [18,19]). Wave
scattering problems for infinite truncated cones substantially have been considered in the electromagnetic case [20,21] and
in the context of potential theory for the truncated cone such problem was examined in [22]. In Acoustics, a detailed analysis
of the radiation features is carried on only for a limiting case when the truncated cone degenerates into an infinite plane
with the circular aperture [23,24]. So in order to extend a class of problems comprising the edges in Acoustics, this paper
engages with the accurate analysis of diffraction features of semi-infinite truncated cones. Some results regarding these
problems have appeared in [25]. Here the regularization idea [9–11] is used to solve this kind of diffraction problem. As a
result, the analytical treatment of this problem provides the following: an algorithm that is valid for any frequency range
and geometrical parameters of the cone; the reliable benchmark to test out an accuracy of the purely numerical methods
and the high frequency asymptotic approaches; effectiveness in computation of the diffraction characteristics rather than
the computation by purely numerical method.

2. Formulation of the problem

Let us consider the acoustically soft (rigid) semi-infinite truncated cone (see Fig. 1) in the spherical coordinate system
(r, θ , φ) as
Q1: { r ∈ (c1, ∞); θ = γ ; φ ∈ [0, 2π )},

where c1 is the truncation radius and γ is the cone-generating angle. The cone Q1 is irradiated by a plane acoustic wave,
which propagates along the positive direction of the z-axis and is defined as the velocity potential of unit amplitude
U0(r , θ ) = exp(ikr cosθ ). (1)

Here k = ω/c0 = 2π /λ is the wave number, ω is the angular frequency, c0 is the velocity of sound wave, λ is the wavelength.
Time factor exp( − iωt ) is suppressed throughout this paper.
We formulate the corresponding diffraction problem as a mixed boundary value problem for the Helmholtz equation

∇2U (r , θ ) + k 2U (r , θ ) = 0, (2)

with respect to the scattered potential U (r, θ ), which satisfies the Dirichlet (S-case) or Neumann (R-case) boundary con-
ditions on the conical surface in the form
⎡U (r , θ ) + U (i)(r , θ )⎤ = 0 for S−case;
⎣ ⎦(r, θ )∈ Q 1 (3a)

∂θ⎡⎣ U (r , θ ) + U (i)(r , θ )⎤⎦ =0 for R−case.

(r, θ )∈ Q1 (3b)
(i )
Here U (r, θ ) is the incident field potential, which will be determined later as the field excited by the plane wave (1) in semi-
infinite conical region {0 < r < ∞ , γ < θ ≤ π}; ∇2 is the Laplacian defined by

∂2 2 ∂ 1 ∂ ⎛⎜ ∂ ⎞
∇2 = + + 2 sinθ ⎟.
∂r 2
r ∂r r sinθ ∂θ ⎝ ∂θ⎠
We search for the solution of the mixed boundary value problem (2) and (3) in the class of functions that satisfy the
radiation condition, as well as the finiteness of energy in any bounded volume. The second condition in our case is reduced
to the fulfillment of the Meixner condition at the edge of Q1. Due to these two additional conditions, the diffraction problem
(2) and (3) is properly posed.

Fig. 1. Geometry of the problem.

 D. Kuryliak, V. Lysechko / Journal of Sound and Vibration 409 (2017) 81–93 83

3. Representation of the diffracted field

For solution of the diffraction problem (2) and (3), we exploit the mode matching technique. Therefore the space
R 3: { r ∈ (0, ∞) , θ ∈ [0, π ] , φ ∈ [0, 2π )} surrounding the cone is divided into three sub-regions as
D1: {r ∈ (c1, ∞); θ ∈ [0, γ )} ,
D2 : {r ∈ (c1, ∞); θ ∈ (γ , π ]} ,
D3: {r ∈ (0, c1); θ ∈ [0, π ]} (4)
(t ) (i )
in which the total field U (r, θ ) decomposes as a sum of the incident field U (r, θ ) and the diffracted field U (r, θ ) in the form
⎧
⎪ U (r , θ ) , (r , θ ) ∈ D1, D3;
U (t )(r , θ ) = ⎨
⎪ (i )
⎩ U (r , θ ) + U (r , θ ), (r , θ ) ∈ D2 . (5)

It is clear that φ ∈ [0, 2π ) for each region.

In order to satisfy the Helmholtz Eq. (2) for each region (4) together with appropriate boundary conditions (3) on Q1, we
find the scattered field in the form as
⎧ ∞ Kνp(sr )
⎪ ∑ y (2,1) P
νp − 1/2(cosθ ) , (r , θ ) ∈ D1;
⎪ p
K νp(sc1)
⎪ p=1
1 ⎪ ∞ (2,2) K μ (sr )
U (r , θ ) = ⎨∑ y Pμk − 1/2( − cosθ ) k , (r , θ ) ∈ D2 ;
sr ⎪ k
k=1
K μk(sc1)
⎪ ∞
⎪ (1)
Iz n(sr )
⎪ Φ1(sr ) + ∑ x¯ n Pz n − 1/2(cosθ ) I (sc ) , (r , θ ) ∈ D3.
⎩ n= 1 zn 1 (6)

Here yp(2,1) , yk(2,2) , x¯ n(1) are unknown expansion coefficients to be determined; Pη− 1/2(·) is the Legendre function; Iη(·) is the
modified Bessel function; Kη(·) is the Macdonald function; s = − ik ; zn = n − 1/2 with Φ1(sr ) ≡ 0 for S-case and zn = n + 1/2
with Φ1(sr ) ≡ x¯ 0(1)I1/2(sr ) /I1/2(sc1) and unknown x¯ 0(1) for R-case; νp, μk are positive roots of transcendental equations

Pη − 1/2(cosγ )│η = νp = 0, Pη − 1/2( − cosγ )│η = μk = 0 for S−case; (7a)

Pη1− 1/2(cosγ )│η = νp = 0, Pη1− 1/2( − cosγ )│η = μk = 0 for R−case, (7b)
1
where Pη− 1/2(·) is the associated Legendre function defined in [26] as

Pη1− 1/2( ± cosγ ) = ± ∂γPη − 1/2( ± cosγ ).

It should be noted that expression (6) corresponds to the total field in domains D1, D3 and to the diffracted field in D2. For
satisfying the Meixner condition at the aperture's edge we search the unknown expansion coefficients in the class of se-
( )
quences ςn = O(n−1) for n → ∞, where ςn = x¯ n yn(2,1) , yn(2,2) . So, according to the representation (6), we seek the field in conical
regions D1 and D2, formed by cone's lateral sides and aperture, as a superposition of normal waves of semi-infinite cones.
These waves go to infinity and behave as an outgoing spherical waves ∼eikr /r for r → ∞. The aperture of truncated cone
causes the disturbances to expansion coefficients and does not make the change to the spectrum of spatial modes.
In order to use the mode matching technique for the solution of this problem, let us obtain the expression of the plane
acoustic wave in conical region {0 < r < ∞ , γ < θ ≤ π}. For this purpose the auxiliary diffraction problem needs to be
considered.

4. Representation of an acoustic plane wave in the conical region

Let us in the spherical coordinate system (r, θ , φ) consider a perfectly soft (rigid) semi-infinite cone
Q 0: { r ∈ (0, ∞); θ = γ ; φ ∈ [0, 2π )}

irradiated by an acoustic plane wave which represents in terms of Kontorovich-Lebedev integral as

π3 1 Pν − 1/2(cos θ ) dν
U0(r , θ ) = exp(ikr cos θ ) =
2 πi
∫Γ ν cos(πν )
Iν(sr )
sr
.
(8)

Here Γ is an integration contour which is a straight line in the complex plane passing parallel to imaginary axis
Γ ⊂ Π: { |Re(ν )| < 1/2}. For convenience of the analysis of integral (8) let us suppose that the wave number k is an imaginary
84 D. Kuryliak, V. Lysechko / Journal of Sound and Vibration 409 (2017) 81–93

value and then, according to our definition, the parameter s (s = − ik ) becomes real-valued (Im(s ) = 0, Re(s ) > 0). This gives
the convergence plane wave integral representation for θ < π /2. The solution for the real-valued k is obtained at the end of
our analysis.
(∞)
The total field is decomposed as a sum of two terms: diffracted field US,R (r, θ ) and incident field U0(r, θ ) in the form

U t (∞) = US,R
(∞)
(r , θ ) + U0(r , θ ), (9)

where scattered field for conical region {0 < r < ∞ , γ < θ ≤ π}, also expressed in terms of Kontorovich-Lebedev integral, has
the form

1
(∞)
US,R (r , θ ) =
πi
∫Γ νF (S,R)(ν)Pν− 1/2( − cosθ )Iν(sr ) dsrν . (10)

Here F (S,R) is the unknown even function with respect to ν, and it is regular in the strip Π; the even property of F (S,R) ensures
compliance of radiation conditions at infinity.
Enforcement of the boundary conditions (3) on the surface of Q0 together with (8) and (10) leads to

π3 Pνl − 1/2(cosγ )
F (S,R)(ν ) = ∓ .
2 cos(πν )Pνl − 1/2( − cosγ ) (11)

Here upper sign and l ¼ 0 refer to S-case and lower one and l ¼ 1 correspond to R-case; for S-case the function F (S)(ν ) has
the set of simple poles at ν = μk and ν = n − 1/2, where k , n = 1, ∞; for R-case poles of the function F (R)(ν ) are located at
ν = 1/2, ν = μk and ν = n − 1/2, where k , n = 2, ∞. It should be noted that function F (R)(ν ) at the point ν = 1/2 has both a
0
simple pole and indeterminate form of type 0 . Let us now substitute expression (11) into (10). This leads to the integral
representation of the diffracted field, which is convergent if θ > 2γ − π /2. Next we close the contour of integration by a
semicircle of radius R in the right half plane. Then we tend radius to infinity R → ∞; making use of Jordan Curve Theorem
this contour integral can be changed into residue series. In addition to R-case we use L'Hopital's rule. As a result, we obtain
the total field representation of semi-infinite cone in the form of the convergent series for full conical region
{0 < r < ∞ , γ < θ ≤ π} and the complex value s (Re(s ) ≥ 0, Im(s ) ≤ 0) as
(∞)
US,R = Ψ (sr , θ )/ sr ,

where

Ψ (sr , θ ) = 2π 3 ×
⎧ ∞ μk Pμk − 1/2(cosγ )Pμk − 1/2(−cosθ )
⎪ ∑k = 1 cos(πμk )∂μPμ − 1/2(−cosγ ) Iμk(sr ) for S−case;
⎪ k
⎪ 1 + tan2(γ / 2)
×⎨ 2π
I1/2(sr )−
⎪
⎪ ∞
μ P1
k μ − 1/2
k
( cosγ ) Pμ − 1/2(−cosθ )
k
⎪ − ∑k = 2 cos(πμk )∂μP1 Iμk(sr ) for R−case.
(−cosγ )
⎩ μk − 1/2

So, we take the incident field for our diffraction problem as

U (i)(r , θ ) = US,R
(∞)
(r , θ ), (12)

which ensures the fulfilment of boundary conditions in the region D2.

It should be noted that problem considered in this section is well-known and has been examined by other methods (see,
for example [27]).

5. Solution of the problem

To find the unknown expansion coefficients in (6), we use the mode matching of the total scattered field and its normal
derivative on the spherical surface of the radius r = c1 in the form

⎧ U (t )(r , θ )r = c + 0 ;
⎪ 1
θ ∈[0, γ )
(t )
U (r , θ ) r = c1− 0 =⎨
θ ∈[0, π ] ⎪ U (r , θ )r = c1+ 0 ,
(t )
⎩ θ ∈(γ , π ] (13a)
 D. Kuryliak, V. Lysechko / Journal of Sound and Vibration 409 (2017) 81–93 85

⎧ ∂ U (t )(r , θ )r = c + 0 ;
⎪ r 1
θ ∈[0, γ )
(t )
∂rU (r , θ ) r = c1− 0 =⎨
θ ∈[0, π ] ⎪ ∂rU (r , θ )r = c1+ 0 .
(t )
⎩ θ ∈(γ , π ] (13b)

Substitution of relationship (5) together with (6) and (12) into (13) produces the functional (series) equations with
respect to unknown expansion coefficients (the additional term x¯ 0(1) arises for R-case) with Legendre function kernel, defined
on θ ∈ [0, π ]. Since the behavior of the velocity potential in the vicinity of the cone's edge has a singularity of the form
∇U (r , θ ) = O ρ−1/2 (∇is the nabla operator; ρ is the distance to the edge in local coordinate system), we present these
( )
equations by way of limiting passing as
N
Φ1(sc1) + lim ∑ x¯ n(1)Pzn− 1/2(cosθ ) =
N →∞
n= 1
⎧ P
⎪ lim ∑ yp(2,1) Pνp− 1/2(cosθ ), θ ∈ [0, γ );
⎪
⎪ P →∞
p=1
=⎨
⎪ K
⎪ Ψ (sc1, θ ) + lim ∑ yk(2,2) Pμk − 1/2( − cosθ ), θ ∈ (γ , π ];
⎪
⎩ K →∞
k=1 (14a)

N I′z n (sc1)
Φ′1(sc1) + lim ∑ x¯ n(1)Pzn− 1/2(cosθ ) =
N →∞
n= 1
Iz n(sc1)
⎧ P K ′νp (sc1)
⎪ lim ∑ yp(2,1) Pνp− 1/2(cosθ ) , θ ∈ [0, γ );
⎪
⎪ P →∞
p=1
Kνp(sc1)
=⎨
⎪ K K ′μ (sc1)
⎪ Ψ ′(sc1, θ ) + lim ∑ yk(2,2) Pμk − 1/2( − cosθ ) k , θ ∈ (γ , π ].
⎪
⎩ K →∞ K μk(sc1)
k=1 (14b)

In these equations, the prime denotes the derivative with respect to the argument.
∞
By virtue of orthogonal properties of Legendre functions {Pνp− 1/2(cosθ )}∞
p = 1 and {Pμk − 1/2( − cosθ )}k = 1 in angles areas θ ∈ [0, γ )

and θ ∈ (γ , π ] respectively, the next representation may be written

Pz n − 1/2(cosθ ) = q(z n, γ ) ×
⎧ P
ν α +(νp, γ )
⎪ lim ∑ p Pνp− 1/2(cosθ ), θ ∈ [0, γ );
⎪ 2
⎪ P →∞ p = 1 νp − z n
2
×⎨
⎪ K
μ α−(μ , γ )
⎪ lim ∑ k 2 k 2 Pμk − 1/2( − cosθ ), θ ∈ (γ , π ].
⎪
⎩ K →∞ k = 1 μk − z n (15)

Here we employ the notation

⎧⎡ −1
⎧ ∂P ( ± cosγ )⎤⎦ ;
⎪ Pz n − 1/2(cosγ ); ⎪ ⎣ ν ν − 1/2
⎪ ν=η
q(z n, γ ) = ⎨ 1 α (η, γ ) = − 2⎨
±
⎪P
⎩ z n − 1/2 (cosγ ) ; ⎪ ⎡
± ∂P 1
−1
( ± cosγ )⎤⎦ ,
⎪
⎩ ⎣ ν ν − 1/2 ν=η

where the upper and under lines correspond to S-case and R-case respectively; upper signs and η = νp correspond to
θ ∈ [0, γ ) and lower one and η = μk correspond to θ ∈ (γ , π ].
Now, substitute a relationship (15) into the left-hand side of Eqs. (14) for each of regions and equate terms with the same
Legendre functions. Eliminating the two unknown coefficients yp(2,1) , yk(2,2) (also x¯ 0(1) in R-case) from the obtained system and
passing to the limit P, K, N → ∞ (N = P + K − κ , where κ = 0 for S-case and κ = 1 for R-case), we obtain infinite system of
linear algebraic equations (ISLAE) for unknown coefficient in the form of

A11X (1) = F (1). (16)

∞
Here X(1) = {x }(1)
n
n=1
, x n(1) = q(zn, γ )x¯ n(1); A11 is the infinite matrix operator
⎧ sc1W [KξqIz n ]sc1 ⎫∞
⎪ (11) ⎪
A11: ⎨ aqn = ⎬ ,
⎪ ⎡ ⎤
⎣ ξq − z n ⎦Kξq(sc1)Iz n(sc1) ⎪
2 2
⎩ ⎭q, n = 1
86 D. Kuryliak, V. Lysechko / Journal of Sound and Vibration 409 (2017) 81–93

where {ξq}∞
q = 1 is the growing sequence of the roots for soft (7a) and rigid (7b) cones;

⎧ ∞ ∞
⎪ {νp}p = 1⋃{μ }k = 1 for S−case,
k
q=1 = ⎨
{ξq}∞ ⎪ ∞ ∞
⎩ {1/2}⋃{νp}p = 2 ⋃{μk }k = 2 for R−case;

∞
W [·] is the Wronskian, W [αβ ]o = α(o)β′(o) − α′(o)β (o); F (1) = {f }
(1)
q
q=1
is the known vector defined as

⎧ 0, ξq ∉ νp
∞ ⎪
⎪
{ }
f q(1)
q=1
= ⎨
⎪−
π 3 Pξq − 1/2(cosγ )
, ξq ∈ μk
for S−case;
⎪
⎩ 2 cos(πξq)Kξq(sc1)
⎧ π tan(γ /2) 1
⎪− , ξq =
⎪ 2 K1/2(sc1) 2
∞ ⎪
{f } (1)
q
= ⎨ 0, ξq ∉ νp for R−case.
q=1 ⎪
⎪ π 3 Pξ1q − 1/2(cosγ )
⎪− , ξq ∈ μk
⎩ 2 cos(πξq)Kξq(sc1)

Using the analytical regularization procedure [9] for Eq. (16), the diffraction problem thus has been reduced to finding
ISLAE of the second kind

X (1) = A−1(A − A11)X (1) + A−1F (1). (17)

Here A , A−1 are couple of operators defined in [9] as

∞
A: {a qn = (ξq − z n)−1 }
q, n = 1
;

⎧ −1⎫
∞
A−1: ⎨ τkq =
⎩ ({M −1
− (ξq, }
γ ) ′M′− (zk , γ )(zk − ξq) ) ⎬
⎭k, q = 1
,

where split functions M −(ν, γ ), M+(ν, γ ) ( M −(ν, γ ) = M+( − ν, γ )) are regular and do not equal to zero corresponding in both left
Re(ν ) < 1/2 and right Re(ν ) > − 1/2 half planes. These functions are derived from factorization of the even meromorphic
function

⎧ −1
{
cos(πν ) ⎪ Pν − 1/2(cosγ )Pν − 1/2( − cosγ )
⎨
}
for S−case;
M (ν, γ ) =
π ⎪ − 1
1
{ 1
⎩ − Pν − 1/2(cosγ )Pν − 1/2( − cosγ ) }
for R−case,

(
which is regular in the strip Π: { |Re(ν )| < 1/2} and has simple zeroes and poles at ±zk , ±ξq k , q = 1, ∞ respectively; M+(ν, γ ), )
M −(ν, γ ) = O ν ( )
±1/2
if |ν| → ∞ in regularity regions, where the upper sign corresponds to S-case and the lower to R-case.
( )
ISLAE (17) allows obtaining the solution in the class of sequences x¯ n(1) = O n−1 , when n → ∞ and provides the fulfilment
of the Meixner condition at the edge.
For R-case, the unknown x¯ 0(1) is defined as

sinh(sc1)
x¯ 0(1) = .
sc1 (18)

The other unknown expansion coefficients (6) in both S and R cases can be readily obtained through the solution of ISLAE
(17) by way of
 D. Kuryliak, V. Lysechko / Journal of Sound and Vibration 409 (2017) 81–93 87

x n(1)
x¯ n(1) = q(z n, γ )
;
⎧ ∞ x (1 )
⎪ νpα +(νp, γ ) ∑n = 1 2 n 2 for S−case;
⎪ νp − z n
yp(2,1) = ⎨
⎪ x¯ (1)δ 1 + ν α +(ν , γ ) ∑∞ xn(1) for R−case;
⎪
⎩
0 p p p n= 1 ν 2 − z 2
p n

⎧ ∞ x (1 )
⎪ μk α−(μk , γ ) ∑n = 1 2 n 2 −
μk − z n
⎪
⎪ 2π 3 μk Pμ − 1/2(cosγ )I μ (sc1)
⎪ − cos(πμ )∂k P k
for S − case;
⎪ k μ μk − 1/2(−cosγ )
⎪
⎪ (1)
yk(2,2) ( π
( 2γ
= ⎨ x¯ 0 − 2 1 + tan 2 I1/2(sc1) δk +
⎪
) 1
)
⎪ 2π 3 μk P1 (cosγ )I μ (sc1)
⎪ + μk − 1/2 k
(
1 − δk1 + )
⎪ cos(πμk )∂μP1 μk − 1/2
(−cosγ )
⎪
⎪ ∞ x (1 )
⎪ +μk α−(μk , γ ) ∑n = 1 2 n 2 for R−case.
⎩ μk
− zn

Obtained in such a way, the solution of the diffraction problem is valid when cone aperture is not equal to the right angle
(γ ≠ π /2). In case when the semi-infinite truncated cone degenerates into an aperture in infinite plane (γ = π /2), the wave
diffraction problem is reduced to Eq. (17), where indexes zn, ξq (n, q = 1, 2, 3, …) are specified as half-integers in the form of
z n = 2n − 3/2, ξq = 2q − 1/2 for S−case; (19a)

z n = 2n − 1/2, ξq = 2q − 3/2 for R−case (19b)

with the column vector

⎧ 1
⎪ for S−case;
∞ q
( − 1) πΓ (q) ⎪ K2q − 1/2(sc1)
F (1) = {f }(1)
q
q=1
= ⎨
8 Γ (q + 1/2) ⎪ 1 − 2q
⎪K for R−case
⎩ 2q − 3/2(sc1)

and the regularization operator A−1 is given in [11].

The other unknowns are determined through x¯ n(1) by way of
⎧ (8p − 2)Γ (p + 1 / 2) ∞ x n(1)
⎪ ( − 1) p + 1 ∑n = 1 for S−case;
⎪ π Γ (p ) (2p − 1 / 2)2 − z n2
yp(2,1) = ⎨
⎪ x¯ (1)δ 1 + ( − 1) p + 1(4p − 3)Γ (p − 1 / 2) ∑∞ x n(1)
for R−case;
⎪
⎩ 0 p π Γ (p ) n = 1 (2p − 3 / 2)2 − z 2
n

⎧ k + 1(8k − 2)Γ (k + 1 / 2) ∞ x n(1)

⎪ ( − 1) ∑n = 1 −
π Γ (k ) (2k − 1 / 2)2 − z n2
⎪
⎪ − 2π (2k − 1/2)I2k − 1/2(sc1) for S−case;
⎪
⎪
yk(2,2) = ⎨
⎪
(
x¯ 0(1) − 2π I1/2(sc1) δk1− )
⎪ − 2π (2k − 3/2)I2k − 3/2(sc1) 1 − δk1 + ( )
⎪
⎪ (4k − 3)Γ (k − 1 / 2) ∞ x n(1)
⎪ +( − 1)k ∑n = 1 for R−case.
⎩ π Γ (k ) (2k − 3 / 2)2 − z n2

Here Γ(·) is the Gamma function.

6. Numerical calculation

For numerical computation of Eq. (17) we apply the truncation method. For our purposes, the truncation order is chosen
from the condition N = [|sc1|] + ϵ , where ϵ = 4, … , 10.
Fig. 2 shows a normalized pressure distribution U (t )/U (i) of the soft cone in the form of isobars in coordinates (θ , r /λ ). In
the figure the white regions correspond to the maxima while the black ones to the minima. Two values of wave parameter at
c1/λ = 0.1592 and c1/λ = 1.5916 for cone aperture γ = 120° are chosen. These parameters correspond to small and large
88 D. Kuryliak, V. Lysechko / Journal of Sound and Vibration 409 (2017) 81–93

Fig. 2. Normalized near total field distribution of the semi-infinite truncated soft cone (γ = 120° ): (a) c1/λ = 0.1592; (b) c1/λ = 1.5916 .

aperture sizes respectively. It is clear from Fig. 2(a) that isobars have monotone distribution over coordinates (θ , r /λ ). The
bright region is confined by coordinates {0.12 < r / λ ≤ 0.1592; 140° < θ ≤ 180°} and adjoined directly to the aperture. The
total field in that region covers 0.8< U (t ) <0.9. Notice, that practical application follows the condition λ >> a1 (a1 = c1 sin γ is
the radius of circular aperture) to exclude phenomena caused by diffraction. As it is seen from Fig. 2(a) the maximally
illuminated region occupies approximately 2/3 of the angular aperture sector, which allows one to use this structure as a
probe. At the point r/λ = 0.05, which is removed from the aperture approximately to r/λ = 0.1, the maximum of the field
decreases by less than a half. Such slight descends of the field magnitude may be used in diagnostics of bulk defects. A clear
manifestation of diffraction effects is observed for the case when c1/λ = 1.5916 (see Fig. 2(b)). These effects notably hinder
distinguishing the bright illuminated regions and make difficulties in further application in diagnostics.
Analogous pictures of the near field distribution for the rigid cone Q1 are also observed at c1/λ = 0.1592 (Fig. 3 (a)) and
c1/λ = 1.5916 (Fig. 3(b)). It should be noted that the major contribution to these distributions of the rigid cone is caused by
the piston mode, which is determined by (18).
From Fig. 3(a) it can be seen that the near field is formed monotonous for value c1/λ = 0.1592. This dependence is
analogous to the previous one (see Fig. 2(a)). However, by virtue of exciting of the piston mode, the magnitude of ratio
U (t )/U (i) here is greater. In the brightly illuminated region, which directly adjoins to the aperture
{0.1 < r / λ ≤ 0.1592; 140° < θ ≤ 180°}, the total field magnitude exceeds the incident field approximately twice. This effect
may be used for probing. At the same time, as may be observed from Fig. 3(b) the narrow illuminated region occupies
{1.1 < r /λ ≤ 1.5916; 170° < θ ≤ 180°}. This formation is caused by the diffraction effect, when a1/λ ≈ 1. This means that one
can obtain a bright field of pressure even under the condition λ ∼ a1.
The power scattering W − in the region γ ≤ θ ≤ π normalized by the factor (2ρ0 c0)−1πc12 (ρ0 is density) is defined as

Fig. 3. Normalized near total field distribution of the semi-infinite truncated rigid cone (γ = 120° ): (a) c1/λ = 0.1592; c1/λ = 1.5916 .
 D. Kuryliak, V. Lysechko / Journal of Sound and Vibration 409 (2017) 81–93 89

−1 2π π
(
W − = 2πkc12 ) lim
r →∞
∫0 dφ ∫γ iU (r , θ ) ∂rU *(r , θ ) r 2 sinθ dθ .
{ } (20)

Here the asterisk indicates complex conjugate value of ∂rU (r, θ ). Taking into account (6) and the orthogonality properties of
Legendre functions [9], the expression (20) takes a form

2
π sinγ
∞
1 yk(2,2)
W− = ∑ ×
2(kc1)2 k=1
μk K μk(sc1)
⎧ 1
⎪ Pμk − 1/2( − cosγ )∂μPμk − 1/2( − cosγ ) for S−case;
×⎨
⎪ Pμ − 1/2( − cosγ )∂μPμ1 − 1/2( − cosγ ) for R−case.
⎩ k k (21)

In the case of γ = π /2 the formula (21) takes a rather simple form as

2
∞ (2,2)
π 1 yk
W− = ∑ ,
2(kc1)2 k=1
ξk Kξk(sc1)

where ξk is defined in (19).

The frequency dependence of W − for S-case is given in Fig. 4. As it is seen the behavior of curves in Fig. 4 (a) shows the
progressive decrease of their oscillations as parameters kc1 and γ increase. For narrow cones (γ ≤ 20°) the coefficient W −
retains a practically stable value in the range 7 < kc1 < 20. The growth of the parameter γ (see Table 1) leads to formation of
peaks at kc1 = 2.5, where the value of W − is beyond 1.
The case when the plane wave passes out through the aperture, i.e. γ > 90°, is shown in Fig. 4(b). From this figure it can
be seen that the maximum of W − is significantly shifted toward the high-frequency region, and for the two angles
γ = 150°; 160° these maxima appear when the aperture radius corresponds to a half wavelength (a1 ∼ λ /2) (see Table 2).
Besides, we observe the maximum of magnitude of scattering by the edge for γ ≥ 150°.
The frequency dependences of W − in the case of the truncated rigid cone, for γ ≤ 90° and γ > 90° are depicted in Fig. 5
(a) and Fig. 5(b) respectively. From Fig. 5(a) it can be seen that the curves are monotonically shifted to larger values of W −
with growth of γ. As a parameter kc1 increases, W − tends to a constant when the value of the maximum radiated power is
less than 1. The situation changes as the plane wave is passed out through the aperture (Fig. 5(b)). In that case W − >> 1 in the
low-frequency region. This indicates that the scattered field almost does not penetrate into the region D1. Starting from the
parameter kc1 > 5, the curves have oscillations which attenuate in the cone cavity. It means that the wide cavity creates a
lower barrier for penetration of the field into the opening cut.
It should be noted that the same behaviour of diffracted power of soft and rigid semi-infinite cones (see Fig. 4(a) and
Fig. 5(a)) happens and may be observed for rigid and soft finite cones respectively (see publication [11], Fig. 6). For the case
when γ = π /2 the oscillating behavior of W − can be explained in terms of Babinet's principle, when the diffracted features of
soft and rigid aperures coincide with rigid and soft disks respectively.
The far-field pattern D(θ ) defined U (r , θ ) ∼ D(θ )eikr /r as r → ∞ equals

Fig. 4. Dependencies of normalized power of diffracted field on the parameter kc1 for S-case: (a) γ ≤ 90° ; (b) γ > 90°.
90 D. Kuryliak, V. Lysechko / Journal of Sound and Vibration 409 (2017) 81–93

Table 1
Local maxima of magnitude values of W − for γ ≤ 90° .

γ 30° 40° 50° 60° 70° 80° 90°

W− 0.72858 1.11746 1.36969 1.4783 1.47272 1.38617 1.24266

kc1 3.45 2.97 2.68 2.51 2.44 2.45 2.52
ka1 1.76 1.91 2.05 2.17 2.11 2.41 2.52

Table 2
Local maxima of magnitude values of W − for γ > 90° .

γ 110° 120° 130° 140° 150° 160° 170°

W− 0.93844 0.829071 0.81932 0.947344 1.1937 1.50703 2.43491

kc1 2.85 3.15 3.70 4.50 6.43 9.11 16.20
ka1 2.68 2.73 2.83 2.89 3.16 3.12 2.81

Fig. 5. Dependencies of normalized power of diffracted field on the parameter kc1 for R-case: (a) γ ≤ 90°; (b) γ > 90°.

Fig. 6. Far-field patterns of the soft cone: (a) γ = 60°, 90° ; (b) γ = 150° .

⎧
⎪
∞ yp(2,1) Pνp− 1/2(cosθ )
⎪ ∑ ;
Kνp(sc1)
π⎪ p=1
D(θ ) = ⎨
2⎪ ∞ yk(2,2) Pμk − 1/2( − cosθ )
⎪ ∑ .
⎪ K μk(sc1)
⎩ k=1

This expression for its physical meaning defines the diffracted field in the region D2 (γ < θ ≤ π ) and the total field in D1
(0 ≤ θ < γ ).
In analysing of far-field patterns for the soft cone, the main attention is focused on frequencies at which resonance peaks
of scattering power occur. The dependence of D(θ ) is examined for several cone's angles at local maxima of W − (kc1 = 2.51,
γ = 60°; kc1 = 2.52, γ = 90° see Table 1; kc1 = 6.43, γ = 150° see Table 2); the results are displayed in Fig. 6. From Fig. 6(a) we
establish that if the cone is irradiated from the side of the cut apex, the levels of diffracted field in the range 0° ≤ θ ≤ γ and
 D. Kuryliak, V. Lysechko / Journal of Sound and Vibration 409 (2017) 81–93 91

Fig. 7. Influence of the wave parameter kc1 on the far-field patterns of the rigid cone: (a) γ = 60° ; (b) γ = 120° .

total one in γ ≤ θ ≤ 180° are almost the same. Next, let us consider the cavity with γ = 150°, the aperture radius of which
nearly equals a wavelength (see Fig. 6(b)). This corresponds to the case when an acoustic wave passes through the aperture.
In this situation the level of diffracted field in the direction θ = 180° is twice greater than the total field in θ = 0°. This reveals
the effect of the soft conical cavity blocking the field penetration from the region D2 to D1.
Our analysis of truncated semi-infinite rigid cone examines the far-field patterns for kc1 ≤ 10; it is depicted in Fig. 7.
The behavior of the curves in this figure shows the essential growth of the radiation in the direction of the incidence
acoustic wave propagation (θ = 0°) when the wave parameter kc1 is increased. Besides, it is clearly seen that the level of
diffracted field essentially increases in the region of specular reflection from the lateral side of the cone 100° < θ < 140°.
Fig. 7(b) illustrates the far-field patterns, when the cone generating angle equals 120°, i.e. irradiation occurs from the
side of conical cavity. A comparison of Fig. 7(a) and (b) reveals a good congruence of curves upon 0° < θ < 60° (region
D1) and the essential divergence, when θ > 120°. The behavior of curves in Fig. 7(b) shows an essential level of diffracted
field in the direction of cone generator θ = 120° and in the direction opposite to the propagating incident wave
(θ = 180°).
Now let us analyze the influence of the wave parameter kc1 , in the same frequency band as previously, on scat-
tering characteristics of truncated rigid cone with cone-generating angles: γ = 20° and γ = 160°. The corresponding far-
field patterns are plotted in Fig. 8. As follows from Fig. 8(a) in the diffracted field, the formation of sidelobes (the
peaks in off-axis radiation) are not observed all over the angular range. From the behavior of curves in Fig. 8(b), the
appreciable penetration of the acoustic wave it may be seen in the cavity from the side of the cut; this shows the focus
properties of conical cavity. Here the level of the total field (region D1 ) is essentially greater than the diffracted field
(region D2 ).

7. Verification of results

Verification of the validity of numerical calculations can be realized by testing of fulfillment of the mode-matching
conditions (13) along the virtual spherical surface. By way of illustration, the dependence of absolute values of the total field
(13a) on the spherical surface of specified radius c1 as function of the polar angle θ is displayed in Fig. 9. In this figure, the
solid curves depict the field's magnitude from the side of region D3 (0° ≤ θ ≤ 180°); dotted curves depict the field's mag-
nitude from the side of regions D1 (0° ≤ θ < γ ) and D2 (γ < θ ≤ 180°). Fig. 9 demonstrates a good congruence of these curves
for all observation angles θ.

Fig. 8. Influence of the wave parameter kc1 on the far-field patterns of the rigid cone: (a) γ = 20° ; (b) γ = 160° .
92 D. Kuryliak, V. Lysechko / Journal of Sound and Vibration 409 (2017) 81–93

Fig. 9. Verification of the mode-matching condition for cone with γ = 30° , kc1 = 30 : (a) S-case; (b) R-case.

Fig. 10. Comparison of normalized power of diffracted field for circular aperture in an infinite plane: (a) S-case; (b) R-case.

We also verified our calculations by comparing them with those obtained for circular soft and rigid apertures in an
infinite plane; the results are displayed in Fig. 10. As one can see from this figure, when γ = π /2, our results (solid lines) are in
excellent agreement with those of [24] (dotted lines).

8. Conclusions

Using the method of eigenfunction expansion together with mode matching technique and analytical regularization
procedure, the correct solution of the canonical diffraction problem from soft and rigid semi-infinite truncated cones in the
case of axially-symmetric irradiation is obtained. The solution of the diffraction problem has been reduced to ISLAE of the
second kind; this solution satisfies all the necessary conditions, including the edge condition, and the condition of the piston
mode excitation in the case of rigid scatterer. The transition from the cone to the circular aperture in an infinite plane is
considered for both soft and rigid cases.
The influence of geometrical parameters and of the frequency on the radiation features of soft and rigid semi-infinite
truncated cones is found out. The significant contrast in far fields between the cases when the acoustic plane wave passes
through the aperture of the conical cavity from the side of the cut, and vice versa, when it passes out through the aperture,
is shown. The focusing and blocking effects of the field are established for the case when an acoustic wave passes out from
the conical cavity. The effect of low-frequency resonance scattering for the soft truncated cone is revealed and it is quite
similar to that which occurs in the case of the finite rigid cone. The influence of the aperture radius of the truncated soft and
rigid cones on near-field distributions is considered. It is shown that the total field has a monotonous distribution in the
vicinity of the aperture under the condition c1 sin γ << λ .

References

[1] J.B. Keller, Backscattering from a finite cone, IEEE Trans. Antennas Propag. 8 (2) (1960) 175–182, http://dx.doi.org/10.1109/TAP.1960.1144832.
[2] F.H. Northover, The diffraction of electric waves around a finite, perfectly conducting cone. Pt. 1. The mathematical solution, J. Math. Anal. Appl. 10
(1965) 37–49, http://dx.doi.org/10.1016/0022-247X(65)90145-9.
[3] F.H. Northover, The diffraction of electric waves around a finite, perfectly conducting cone. Pt. 2. The field singularities, J. Math. Anal. Appl. 10 (1)
(1965) 50–69, http://dx.doi.org/10.1016/0022-247X(65)90146-0.
[4] A. Syed, The diffraction of arbitrary electromagnetic field by a finite perfectly conducting cone, J. Nat. Sci. Math. 2 (1) (1981) 85–114.
[5] I.V. Vovk, V.T. Hrinchenko, Sound wave radiation from a finite hollow cone, in: Y.A. Mytropolskyi (Ed.), Linear Boundary Problem of Mathematical
Physics, Institute of Mathematics, Academy of Sciences of Ukrainian SSR, Kyiv, 1973, pp. 129–139.
 D. Kuryliak, V. Lysechko / Journal of Sound and Vibration 409 (2017) 81–93 93

[6] D.C. Pridmore-Brown, A Wiener-Hopf solution of a radiation problem in conical geometry, J. Math. Phys. 47 (1–4) (1968)
79–94, http://dx.doi.org/10.1002/sapm196847179.
[7] L. Shi, H. Sun, W. Dong, X. Lv, A dual-band multifunction carborne hybrid antenna for satellite communication relay system, Prog. Electromagn. Res. 95
(2009) 329–340, http://dx.doi.org/10.2528/PIER09070306.
[8] P. Vsetula, Z. Raida, Sierpinski-based conical monopole antenna, Radioenginering 19 (4) (2010) 633–638.
[9] D.B. Kuryliak, Z.T. Nazarchuk, Analytical-numerical methods in the theory of wave diffraction on conical and wedge-shaped surfaces, Nauk. Dumka
Kyiv (2006).
[10] D.B. Kuryliak, Z.T. Nazarchuk, Convolution type operators for wave diffraction by conical structures, Radio Sci. 43 (4). 〈http://dx.doi.org/10.1029/
2007RS003792〉.
[11] D.B. Kuryliak, Z.T. Nazarchuk, V.O. Lysechko, Diffraction of a plane acoustic wave from a finite soft (rigid) cone in axial irradiation, Open J. Acoust. 5 (4)
(2015) 193–206, http://dx.doi.org/10.4236/oja.2015.54015.
[12] V.P. Shestopalov, A.A. Kirilenko, S.A. Masalov, Convolution-type matrix equations in the theory of diffraction, Nauk. Dumka Kyiv (1984).
[13] I.I. Vorovich, V.M. Aleksandrov, V.A. Babeshko, Nonclassical Mixed Problems of the Theory of Elasticity, Nauka, Moscow, 1974.
[14] G.Ya. Popov, Exact Solutions of Certain Boundary-value Problems of Mechanics of a Deformable Solid, Astroprint, Odessa, 2013.
[15] V.P. Smyshlyaev, Diffraction by conical surfaces at high frequencies, Wave motion 12 (4) (1990)
329–339, http://dx.doi.org/10.1016/0165-2125(90)90003-m.
[16] Y.A. Antipov, Diffraction of a plane wave by a circular cone with an impedance boundary condition, SIAM J. Appl. Math. 62 (4) (2002)
1122–1152, http://dx.doi.org/10.1137/s0036139900363324.
[17] M.A. Lyalinov, Scattering of an acoustic axially symmetric surface wave propagating to the vertex of a right-circular impedance cone, Wave motion 47
(4) (2010) 241–252, http://dx.doi.org/10.1016/j.wavemoti.2009.10.006.
[18] A.D. Rawlins, Diffraction by, or diffusion into, a penetrable wedge, P. Roy. Soc. A 455 (1987) (1999)
2655–2686, http://dx.doi.org/10.1098/rspa.1999.0421.
[19] V.M. Babich, M.A. Lyalinov, V.E. Grikurov, Diffraction Theory: the Sommerfeld-Malyuzhinets Technique, Alpha Science International Ltd., Oxford, UK,
2008.
[20] D.B. Kuryliak, Axially-symmetric field of electric dipole over truncated cone. I. Comparison between mode-matching technique and integral trans-
formation method, Radio Phys. Radio Astron 4 (2) (1999) 121–128.
[21] D.B. Kuryliak, Axially-symmetric field of electric dipole over truncated cone. II. Numerical simulation, Radio Phys. Radio Astron 5 (3) (2000) 284–290.
[22] J. Bazer, S.N. Karp, Potential flow through a conical pipe with an application to diffraction theory, Research Report No EM-66, New York University,
Institute of Mathematical Sciences, August 1954.
[23] R.D. Spence, The diffraction of sound by circular disks and apertures, J. Acoust. Soc. Am. 20 (4) (1948) 380–386, http://dx.doi.org/10.1121/1.1906389.
[24] C.J. Bouwkamp, Theoretical and numerical treatment of diffraction through a circular aperture, IEEE Trans. Antennas. Propag. 18 (2) (1970)
152–176, http://dx.doi.org/10.1109/TAP.1970.1139646.
[25] V.O. Lysechko, D.B. Kuryliak, The near acoustic field of a circular aperture obtained by truncation of the semiinfinite cone vertex, Rep. Natl. Acad. Sci.
Ukr. 8 (2016) 63–69, http://dx.doi.org/10.15407/dopovidi2016.08.063.
[26] I.S. Gradshtein, I.M. Ryzhik, Tables of integrals, series, and products, Gosudarstvennoe Izdatelstvo Fiziko-Matematiceskoj Literatury, Moscow, 1963.
[27] J.J. Bowman, The cone, In: J.J. Bowman, T.B.A. Senior and P.L.E. Uslenghi, (Eds.), Electromagnetic and Acoustic Scattering by Simple Shapes, North-
Holland Publishing Co; Amsterdam (1969) 637–701.