This action might not be possible to undo. Are you sure you want to continue?
https://www.scribd.com/doc/48554217/TheoryofVortexSoundCambridgeTextsinAppliedMathematicsbyMSHowe
10/06/2014
text
original
Theory of Vortex Sound
Theory of Vortex Sound is an introduction to the theory of sound generated by
hydrodynamic ﬂows. Starting with a reviewof elementary theoretical acoustics,
the book proceeds to a uniﬁed treatment of low Mach number vortexsurface
interaction noise in terms of the compact Green’s function. Problems are pro
vided at the end of each chapter, many of which can be used for extended student
projects, and a whole chapter is devoted to worked examples.
It is designed for a onesemester introductory course at the advanced un
dergraduate or graduate levels. Great care is taken to explain underlying ﬂuid
mechanical and acoustic concepts, and to describe as fully as possible the steps
in a complicated derivation.
M.S. Howe has been Professor in the Department of Aerospace and Mechanical
Engineering at Boston University since 1992. He is a Fellow of the Institute of
Acoustics (U.K.) and of the Acoustical Society of America.
Cambridge Texts in Applied Mathematics
Maximum and Minimum Principles
M. J. Sewell
Solitons
P. G. Drazin and R. S. Johnson
The Kinematics of Mixing
J. M. Ottino
Introduction to Numerical Linear Algebra and Optimisation
Philippe G. Ciarlet
Integral Equations
David Porter and David S. G. Stirling
Perturbation Methods
E. J. Hinch
The Thermomechanics of Plasticity and Fracture
Gerard A. Maugin
Boundary Integral and Singularity Methods for Linearized Viscous Flow
C. Pozrikidis
Nonlinear Wave Processes in Acoustics
K. Naugolnykh and L. Ostrovsky
Nonlinear Systems
P. G. Drazin
Stability, Instability, and Chaos
Paul Glendinning
Applied Analysis of the Navier–Stokes Equations
C. R. Doering and J. D. Gibbon
Viscous Flow
H. Ockendon and J. R. Ockendon
Scaling, SelfSimilarity, and Intermediate Asymptotics
G. I. Barenblatt
A First Course in the Numerical Analysis of Differential Equations
Arieh Iserles
Complex Variables: Introduction and Applications
Mark J. Ablowitz and Athanassios S. Fokas
Mathematical Models in the Applied Sciences
A. C. Fowler
Thinking About Ordinary Differential Equations
Robert E. O’Malley
A Modern Introduction to the Mathematical Theory of Water Waves
R. S. Johnson
Rareﬁed Gas Dynamics
Carlo Cercignani
Symmetry Methods for Differential Equations
Peter E. Hydon
High Speed Flow
C. J. Chapman
Wave Motion
J. Billingham and A. C. King
An Introduction to Magnetohydrodynamics
P. A. Davidson
Linear Elastic Waves
John G. Harris
Vorticity and Incompressible Flow
Andrew J. Majda and Andrea L. Bertozzi
Inﬁnite Dimensional Dynamical Systems
James C. Robinson
An Introducion to Symmetry Analysis
Brian J. Cantwell
Backlund and Darboux Transformations
C. Rogers and W. K. Schief
FiniteVolume Methods for Hyperbolic Problems
Randall J. LeVeque
Introduction to Hydrodynamic Stability
P. G. Drazin
Theory of Vortex Sound
M. S. Howe
Theory of Vortex Sound
M. S. HOWE
Boston University
caxniioci uxiviisiry iiiss
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo
Cambridge University Press
The Edinburgh Building, Cambridge cn: :iu, United Kingdom
First published in print format
isnx:+ ,¬ïcs::ï::ï:ï
isnx:+ ,¬ïcs::c:::+:
isnx:+ ,¬ïcs::c¬ï:¡,
© Cambridge University Press 2003
2002
Information on this title: www.cambridge.org/9780521812818
This book is in copyright. Subject to statutory exception and to the provision of
relevant collective licensing agreements, no reproduction of any part may take place
without the written permission of Cambridge University Press.
isnx:c cs::c¬ï:¡s
isnx:c cs::ï::ï:x
isnx:c cs::c:::+o
Cambridge University Press has no responsibility for the persistence or accuracy of
uiis for external or thirdparty internet websites referred to in this book, and does not
guarantee that any content on such websites is, or will remain, accurate or appropriate.
Published in the United States of America by Cambridge University Press, New York
www.cambridge.org
hardback
paperback
paperback
eBook (NetLibrary)
eBook (NetLibrary)
hardback
To Shˆ on Ffowcs Williams
Contents
Preface page xiii
1 Introduction 1
1.1 What is Vortex Sound? 1
1.2 Equations of Motion of a Fluid 2
1.3 Equation of Linear Acoustics 4
1.4 The Special Case of an Incompressible Fluid 7
1.5 Sound Produced by an Impulsive Point Source 10
1.6 FreeSpace Green’s Function 12
1.7 Monopoles, Dipoles, and Quadrupoles 13
1.8 Acoustic Energy Flux 18
1.9 Calculation of the Acoustic Far Field 20
Problems 1 23
2 Lighthill’s Theory 25
2.1 The Acoustic Analogy 25
2.2 Lighthill’s :
8
Law 29
2.3 Curle’s Theory 32
2.4 Sound Produced by Turbulence Near a Compact Rigid Body 36
2.5 Radiation from a Noncompact Surface 37
Problems 2 39
3 The Compact Green’s Function 41
3.1 The Inﬂuence of Solid Boundaries 41
3.2 The Helmholtz Equation 44
3.3 The Reciprocal Theorem 46
3.4 TimeHarmonic Compact Green’s Function 49
ix
x Contents
3.5 Compact Green’s Function for a Rigid Sphere 53
3.6 Compact Green’s Function for Cylindrical Bodies 58
3.7 Symmetric Compact Green’s Function 63
3.8 LowFrequency Radiation from a Vibrating Body 65
3.9 Compact Green’s Function Summary and Special Cases 70
Problems 3 79
4 Vorticity 82
4.1 Vorticity and the Kinetic Energy of Incompressible Flow 82
4.2 The Vorticity Equation 84
4.3 The Biot–Savart Law 88
4.4 Surface Force in Incompressible Flow Expressed in Terms
of Vorticity 93
4.5 The Complex Potential 100
4.6 Motion of a Line Vortex 106
Problems 4 112
5 Vortex Sound 114
5.1 The Role of Vorticity in Lighthill’s Theory 114
5.2 The Equation of Vortex Sound 116
5.3 Vortex–Surface Interaction Noise 124
5.4 Radiation from an Acoustically Compact Body 128
5.5 Radiation from Cylindrical Bodies of Compact Cross Section 130
5.6 Impulse Theory of Vortex Sound 131
Problems 5 132
6 Vortex–Surface Interaction Noise in Two Dimensions 136
6.1 Compact Green’s Function in Two Dimensions 136
6.2 Sound Generated by a Line Vortex Interacting with
a Cylindrical Body 139
6.3 Inﬂuence of Vortex Shedding 145
6.4 Blade–Vortex Interaction Noise in Two Dimensions 150
Problems 6 154
7 Problems in Three Dimensions 156
7.1 Linear Theory of Vortex–Airfoil Interaction Noise 156
7.2 Blade–Vortex Interactions in Three Dimensions 158
7.3 Sound Produced by Vortex Motion near a Sphere 162
7.4 Compression Wave Generated When a Train Enters a Tunnel 166
Problems 7 172
Contents xi
8 Further Worked Examples 175
8.1 Blade–Vortex Interactions in Two Dimensions 175
8.2 Parallel Blade–Vortex Interactions in Three Dimensions 186
8.3 Vortex Passing over a Spoiler 191
8.4 Bluff Body Interactions: The Circular Cylinder 194
8.5 Vortex Ring and Sphere 199
8.6 Vortex Pair Incident on a Wall Aperture 204
Bibliography 209
Index 213
Preface
Vortex sound is the branch of ﬂuid mechanics concerned with the conversion
of hydrodynamic (rotational) kinetic energy into the longitudinal disturbances
we call sound. The subject is itself a subsection of the theory of aerodynamic
sound, which encompasses a much wider range of problems also involving, for
example, combustion and ‘entropy’ sources of sound. The book is based on an
introductory onesemester graduate level course given on several occasions at
Boston University. Most students at this level possess an insufﬁcient grasp of
basic principles to appreciate the subtle coupling of the hydrodynamic and
acoustic ﬁelds, and many are illequipped to deal with the novel analytical
techniques that have been developed to investigate the coupling. Great care
has therefore been taken to discuss underlying ﬂuid mechanical and acoustic
concepts, andtoexplainas fullyas possible the steps ina complicatedderivation.
A considerable number of practical problems occur at low Mach numbers
(say, less than about 0.4). It seems reasonable, therefore, to conﬁne an intro
ductory discussion speciﬁcally to low Mach number ﬂows. It is then possible
to investigate a number of idealized hydrodynamic ﬂows involving elementary
distributions of vorticity adjacent to solid boundaries, and to analyze in detail
the sound produced by these vortex–surface interactions. For a broad range of
such problems, and a corresponding broad range of noise problems encountered
in industrial applications, the effective acoustic sources turn out to be localized
to one or more regions that are small compared to the acoustic wavelength. This
permits the development of a uniﬁed theory of sound production by vortex–
surface interactions in terms of the compact Green’s function, culminating in a
routine procedure for estimating the sound, and providing, at the same time, an
easy identiﬁcation of those parts of a structure that are likely to be important
sources of sound. Many examples of this type are discussed, and they are simple
enough for the student to acquire an intuitive understanding of the method of
xiii
xiv Preface
solution and the underlying physics. By these means the reader is encouraged to
investigate both the hydrodynamics and the sound generated by a simple ﬂow.
Experience has shown that the successful completion of this kind of project,
involving the implementation of a widely applicable yet standard procedure for
the prediction of sound generation at low Mach numbers, motivates a student
to understand the ostensibly difﬁcult parts of the theory. One or more of the
problems appended to some of the later chapters can formthe basis of a project.
The ﬁnal chapter contains a set of worked examples that have been investi
gated by students at Boston University. I wish to thank my former students
H. AbouHussein, A. DeBenedictis, N. Harrison, M. Kim, M. A. Rodrigues,
and F. Zagadou for their considerable help in preparing that chapter.
The mathematical ability assumed of the reader is roughly equivalent to
that taught in an advanced undergraduate course on Engineering Mathematics.
In particular, the reader should be familiar with basic vector differential and in
tegral calculus and with the repeated sufﬁx summation convention of Cartesian
tensors (but a detailed knowledge of tensor calculus is not required). An el
ementary understanding of the properties of the Dirac δ function is desir
able (Lighthill, 1958), including its interpretation as the formal limit of an
csequence, such as
δ(x) =
c
π(x
2
÷c
2
)
. c →÷0.
Much use is made of the formula
δ( f (x)) =
¸
n
δ(x − x
n
)
[ f
/
(x
n
)[
.
where the summation is over real simple roots of f (x) =0.
M. S. Howe
1
Introduction
1.1 What is Vortex Sound?
Vortex sound is the sound produced as a byproduct of unsteady ﬂuid mo
tions (Fig. 1.1.1). It is part of the more general subject of aerodynamic sound.
The modern theory of aerodynamic sound was pioneered by James Lighthill
in the early 1950s. Lighthill (1952) wanted to understand the mechanisms of
noise generation by the jet engines of new passenger jet aircraft that were
then about to enter service. However, it is now widely recognized that any
mechanism that produces sound can actually be formulated as a problem of
aerodynamic sound. Thus, apart from the high speed turbulent jet – which may
be regarded as a distribution of intense turbulence velocity ﬂuctuations that gen
erate sound by converting a tiny fraction of the jet rotational kinetic energy into
the longitudinal waves that constitute sound – colliding solid bodies, aeroengine
rotor blades, vibrating surfaces, complex ﬂuid–structure interactions in the lar
ynx (responsible for speech), musical instruments, conventional loudspeakers,
crackling paper, explosions, combustion and combustion instabilities in rock
ets, and so forth all fall within the theory of aerodynamic sound in its broadest
sense.
In this book we shall consider principally the production of sound by un
steady motions of a ﬂuid. Any ﬂuid that possesses intrinsic kinetic energy, that
is, energy not directly attributable to a moving boundary (which is largely with
drawn fromthe ﬂuid when the boundary motion ceases), must possess vorticity.
We shall see that in a certain sense and for a vast number of ﬂows vorticity may
be regarded as the ultimate source of the sound generated by the ﬂow. Our
objective, therefore, is to simplify the general aerodynamic sound problem to
obtain a thorough understanding of how this happens, and of how the sound
can be estimated quantitatively.
1
2 1 Introduction
Fig. 1.1.1. Typical vortex sound problems.
1.2 Equations of Motion of a Fluid
At time t and position x = (x
1
. x
2
. x
3
), the state of a ﬂuid is deﬁned when the
velocity v and any two thermodynamic variables are speciﬁed. Five scalar
equations are therefore required to determine the motion. These equations are
statements of the conservation of mass, momentum, and energy.
1.2.1 Equation of Continuity
Conservation of mass requires the rate of increase of the ﬂuid mass within a
ﬁxed region of space V to be equal to the net inﬂux due to convection across the
boundaries of V. The velocity v and the ﬂuid density ρ must therefore satisfy
1.2 Equations of Motion of a Fluid 3
the equation of continuity, which has the following equivalent forms
∂ρ
∂t
÷div(ρv) = 0.
1
ρ
Dρ
Dt
÷div v = 0.
div v = ρ
D
Dt
1
ρ
¸
¸
¸
¸
¸
¸
¸
¸
¸
¸
¸
¸
¸
¸
¸
¸
¸
. (1.2.1)
where
D
Dt
=
∂
∂t
÷v · ∇ ≡
∂
∂t
÷:
j
∂
∂x
j
(1.2.2)
is the material derivative; the repeated sufﬁx j implies summation over j =
1. 2. 3. The last of Equations (1.2.1) states that div v is equal to the rate of
change of ﬂuid volume per unit volume following the motion of the ﬂuid. For
an incompressible ﬂuid this is zero, i.e., div v = 0.
1.2.2 Momentum Equation
The momentumequation is also called the Navier–Stokes equation; it expresses
the rate of change of momentum of a ﬂuid particle in terms of the pressure p,
the viscous or frictional force, and body forces F per unit volume. We consider
only Stokesian ﬂuids (most liquids and monatomic gases, but also a good ap
proximation in air for calculating the frictional drag at a solid boundary) for
which the principal frictional forces are expressed in terms of the shear coefﬁ
cient of viscosity η, which we shall invariably assume to be constant. Then the
momentum equation is
ρ
Dv
Dt
= −∇p ÷η
∇
2
v ÷
1
3
∇(div v)
÷F. (1.2.3)
Values of ρ. η and ν = η¡ρ (the ‘kinematic’ viscosity) for air and water at
10
◦
C and one atmosphere pressure are given in the Table 1.2.1:
Table 1.2.1. Density and viscosity
ρ, kg/m
3
η, kg/ms ν. m
2
/s
Air 1.23 1.764 10
−5
1.433 10
−5
Water 1000 1.284 10
−3
1.284 10
−6
4 1 Introduction
1.2.3 Energy Equation
This equation must be used in its full generality in problems where energy is
transferred by heat conduction, where frictional dissipation of sound is occur
ring, when shock waves are formed by highly nonlinear events, or when sound
is being generated by combustion and other heat sources. For our purposes it
will usually be sufﬁcient to suppose the ﬂow to be homentropic; namely, the
speciﬁc entropy s of the ﬂuid is uniform and constant throughout the ﬂuid, so
that the energy equation becomes
s = constant. (1.2.4)
We may then assume that the pressure and density are related by an equation
of the form
p = p(ρ. s). s = constant. (1.2.5)
This equation will be satisﬁed by both the mean (undisturbed) and unsteady
components of the ﬂow. Thus, for an ideal gas
p = constant ρ
γ
. γ = ratio of speciﬁc heats. (1.2.6)
1.3 Equation of Linear Acoustics
The intensity of a sound pressure p in air is usually measured on a decibel scale
by the quantity
20 log
10
[ p[
p
ref
.
where the reference pressure p
ref
=2 10
−5
N/m
2
. Thus, p = p
0
≡1 atmo
sphere ( =10
5
N/m
2
) is equivalent to 194 dB. A very loud sound ∼120 dB
corresponds to
p
p
0
≈
2 10
−5
10
5
10
(
120
20
)
= 2 10
−4
<1.
Similarly, for a ‘deafening’ sound of 160 dB, p¡p
0
∼ 0.02. This corresponds
to a pressure of about 0.3 lbs/in
2
and is loud enough for nonlinear effects to
begin to be important.
The passage of a sound wave in the form of a pressure ﬂuctuation is, of
course, accompanied by a backandforth motion of the ﬂuid at the acoustic
1.3 Equation of Linear Acoustics 5
particle velocity :, say. We shall see later that
acoustic particle velocity ≈
acoustic pressure
mean density speed of sound
.
In air the speed of sound is about 340 m/sec. Thus, at 120 dB : ∼ 5 cm/sec; at
160 dB : ∼ 5 m/sec.
In most applications the acoustic amplitude is very small relative to the mean
pressure p
0
, and sound propagation may be studied by linearizing the equations.
To do this we shall ﬁrst consider sound propagating in a stationary inviscid ﬂuid
of mean pressure p
0
and density ρ
0
; let the departures of the pressure and density
from these mean values be denoted by p
/
, ρ
/
, where p
/
¡p
0
<1, ρ
/
¡ρ
0
<1. The
linearized momentum equation (1.2.3) becomes
ρ
0
∂v
∂t
÷∇p
/
= F. (1.3.1)
Before linearizing the continuity equation (1.2.1), we introduce an artiﬁcial
generalization by inserting a volume source distribution q(x. t ) on the right
hand side
1
ρ
Dρ
Dt
÷div v = q. (1.3.2)
where q is the rate of increase of ﬂuid volume per unit volume of the ﬂuid, and
might represent, for example, the effect of volume pulsations of a small body
in the ﬂuid. The linearized equation is then
1
ρ
0
∂ρ
/
∂t
÷div v = q. (1.3.3)
Now eliminate v between (1.3.1) and (1.3.3):
∂
2
ρ
/
∂t
2
−∇
2
p
/
= ρ
0
∂q
∂t
−div F. (1.3.4)
An equation determining the pressure p
/
alone in terms of q and Fis obtained
by invoking the homentropic relation (1.2.5). In the undisturbed and disturbed
states we have
p
0
= p(ρ
0
. s). p
0
÷ p
/
= p(ρ
0
÷ρ
/
. s) ≈ p(ρ
0
. s) ÷
∂p
∂ρ
(ρ. s)
0
ρ
/
.
s = constant. (1.3.5)
6 1 Introduction
The derivative is evaluated at the undisturbed values of the pressure and density
( p
0
. ρ
0
). It has the dimensions of velocity
2
, and its square root deﬁnes the speed
of sound
c
0
=
∂p
∂ρ
s
. (1.3.6)
where the derivative is taken with the entropy s held ﬁxed at its value in the
undisturbed ﬂuid. The implication is that losses due to heat transfer between
neighboring ﬂuid particles by viscous and thermal diffusion are neglected
during the passage of a sound wave (i.e., that the motion of a ﬂuid particle
is adiabatic).
From (1.3.5): ρ
/
= p
/
¡c
2
0
. Hence, substituting for ρ
/
in (1.3.4), we obtain
1
c
2
0
∂
2
∂t
2
−∇
2
p = ρ
0
∂q
∂t
−div F. (1.3.7)
where the prime (
/
) on the acoustic pressure has been discarded. This equation
governs the production of sound waves by the volume source q and the force
F. When these terms are absent the equation describes sound propagation
from sources on the boundaries of the ﬂuid, such as the vibrating cone of a
loudspeaker.
The volume source q and the body force F would never appear in a complete
description of sound generation within a ﬂuid. They are introduced only when
we think we understand how to model the real sources of sound in terms of
volume sources and forces. In general this can be a dangerous procedure be
cause, as we shall see, small errors in specifying the sources of sound in a ﬂuid
can lead to very large errors in the predicted sound. This is because only a tiny
fraction of the available energy of a vibrating ﬂuid or structure actually radiates
away as sound.
When F = 0, Equation (1.3.1) implies the existence of a velocity potential
ϕ such that v = ∇ϕ, in terms of which the perturbation pressure is given by
p = −ρ
0
∂ϕ
∂t
. (1.3.8)
It follows from this and (1.3.7) (with F = 0) that the velocity potential is the
solution of
1
c
2
0
∂
2
∂t
2
−∇
2
ϕ = −q(x. t ). (1.3.9)
This is the wave equation of classical acoustics.
1.4 The Special Case of an Incompressible Fluid 7
Table 1.3.1. Speed of sound and acoustic wavelength
c
0
λ at 1 kHz
m/s ft/s km/h mi/h m ft
Air 340 1100 1225 750 0.3 1
Water 1500 5000 5400 3400 1.5 5
For future reference, Table 1.3.1 lists the approximate speeds of sound in air
and in water, and the corresponding acoustic wavelength λ at a frequency of
1 kHz (sound of frequency f has wavelength λ = c
0
¡f ).
1.4 The Special Case of an Incompressible Fluid
Small (adiabatic) pressure and density perturbations δp and δρ satisfy
δp
δρ
≈ c
2
0
.
In an incompressible ﬂuid the pressure can change by the action of external
forces (moving boundaries, etc.), but the density must remain ﬁxed. Thus,
c
0
= ∞, and Equation (1.3.9) reduces to
∇
2
ϕ = q(x. t ). (1.4.1)
1.4.1 Pulsating Sphere
Consider the motion produced by small amplitude radial pulsations of a sphere
of mean radius a. Let the center of the sphere be at the origin, and let its normal
velocity be :
n
(t ). There are no sources within the ﬂuid, so that q ≡ 0. Therefore,
∇
2
ϕ = 0. r >a.
∂ϕ¡∂r = :
n
(t ). r =a
¸
where r = [x[.
The motion is obviously radially symmetric, so that
∇
2
ϕ =
1
r
2
∂
∂r
r
2
∂
∂r
ϕ = 0. r >a.
Hence,
ϕ =
A
r
÷ B.
8 1 Introduction
where A ≡ A(t ) and B ≡ B(t ) are functions of t . B(t ) can be discarded be
cause the pressure ﬂuctuations (∼ −ρ
0
∂ϕ¡∂t ) must vanish as r →∞. Apply
ing the condition ∂ϕ¡∂r = :
n
at r = a, we then ﬁnd
ϕ = −
a
2
:
n
(t )
r
. r >a. (1.4.2)
Thus, the pressure
p = −ρ
0
∂ϕ
∂t
= ρ
0
a
2
r
d:
n
dt
(t )
decays as 1¡r with distance from the sphere, and exhibits the unphysical char
acteristic of changing instantaneously everywhere when d:
n
¡dt changes its
value. For any time t , the volume ﬂux q(t ) of ﬂuid is the same across any closed
surface enclosing the sphere. Evaluating it for any sphere S of radius r >a, as
shown in Fig. 1.4.1, we ﬁnd
q(t ) =
S
∇ϕ · dS = 4πa
2
:
n
(t ).
and we may also write
ϕ =
−q(t )
4πr
. r >a. (1.4.3)
Fig. 1.4.1.
1.4 The Special Case of an Incompressible Fluid 9
1.4.2 Point Source
The incompressible motion generated by a volume point source of strength q(t )
at the origin is the solution of
∇
2
ϕ = q(t )δ(x). where δ(x) = δ(x
1
)δ(x
2
)δ(x
3
). (1.4.4)
The solution must be radially symmetric and given by
ϕ =
A
r
for r >0. (1.4.5)
To ﬁnd A, we integrate (1.4.4) over the interior of a sphere of radius r = R >0,
and use the divergence theorem
rR
∇
2
ϕ d
3
x =
S
∇ϕ · dS, where S is the
surface of the sphere. Then
S
∇ϕ · dS ≡
−A
R
2
(4π R
2
) = q(t ).
Hence, A = −q(t )¡4π and ϕ = −q(t )¡4πr, which agrees with the solution
(1.4.3) for the sphere with the same volume outﬂowin the regionr >a = radius
of the sphere. This indicates that when we are interested in modelling the effect
of a pulsating sphere at large distances r ·a, it is permissible to replace the
sphere by a point source (a monopole) of the same strength q(t ) = rate of
change of the volume of the sphere. This conclusion is valid for any pulsating
body, not just a sphere. However, it is not necessarily a good model (especially
when we come to examine the production of sound by a pulsating body) in the
presence of a mean ﬂuid ﬂow past the sphere.
The Solution (1.4.5) for the point source is strictly valid only for r >0,
where it satisﬁes ∇
2
ϕ =0. What happens as r →0, where its value is actually
undeﬁned? To answer this question, we write the solution in the form
ϕ = lim
c→0
−q(t )
4π(r
2
÷c
2
)
1
2
. c >0. in which case ∇
2
ϕ = lim
c→0
3c
2
q(t )
4π(r
2
÷c
2
)
5
2
.
The last limit is just equal to q(t )δ(x). Indeed when c is small 3c
2
¡4π(r
2
÷c
2
)
5
2
is also small except close to r = 0, where it attains a large maximum∼3¡4πc
3
.
Therefore, for any smoothly varying test function f (x) and any volume V
10 1 Introduction
enclosing the origin
lim
c→0
V
3c
2
f (x) d
3
x
4π(r
2
÷c
2
)
5
2
= f (0) lim
c→0
∞
−∞
3c
2
d
3
x
4π(r
2
÷c
2
)
5
2
= f (0)
∞
0
3c
2
r
2
dr
(r
2
÷c
2
)
5
2
= f (0).
where the value of the last integral is independent of c. This is the deﬁning
property of the threedimensional δ function.
Thus, the correct interpretation of the solution
ϕ =
−1
4πr
of ∇
2
ϕ = δ(x) (1.4.6)
for a unit point source (q = 1) is
−1
4πr
= lim
c→0
−1
4π(r
2
÷c
2
)
1
2
. r ≥ 0. (1.4.7)
where
∇
2
−1
4πr
= lim
c→0
∇
2
−1
4π(r
2
÷c
2
)
1
2
= lim
c→0
3c
2
4π(r
2
÷c
2
)
5
2
= δ(x). (1.4.8)
1.5 Sound Produced by an Impulsive Point Source
The sound generated by the unit, impulsive point source δ(x)δ(t ) is the solution
of
1
c
2
0
∂
2
∂t
2
−∇
2
ϕ = δ(x)δ(t ). (1.5.1)
The source exists only for an inﬁnitesimal instant of time at t = 0; therefore at
earlier times ϕ(x. t ) = 0 everywhere.
It is evident that the solution is radially symmetric, and that for r = [x[ >0
we have to solve
1
c
2
0
∂
2
ϕ
∂t
2
−
1
r
2
∂
∂r
r
2
∂
∂r
ϕ = 0. r >0. (1.5.2)
The identity
1
r
2
∂
∂r
r
2
∂
∂r
ϕ ≡
1
r
∂
2
∂r
2
(rϕ) (1.5.3)
1.5 Sound Produced by an Impulsive Point Source 11
permits us to write Equation (1.5.2) in the form of the onedimensional wave
equation for rϕ:
1
c
2
0
∂
2
∂t
2
(rϕ) −
∂
2
∂r
2
(rϕ) = 0. r >0. (1.5.4)
This has the general solution rϕ = +(t −r¡c
0
) ÷ +(t ÷r¡c
0
), where + and
+ are arbitrary functions. Hence, the general solution of (1.5.2) is
ϕ =
+
t −
r
c
0
r
÷
+
t ÷
r
c
0
r
. r >0. (1.5.5)
The ﬁrst termon the right represents a spherically symmetric disturbance that
propagates in the direction of increasing values of r at the speed of sound c
0
as t
increases, whereas the second represents an incoming wave converging toward
x = 0. We must therefore set + = 0, since it represents sound waves generated
at r = ∞that approach the source rather than sound waves generated by the
source and radiating away from the source. This is a causality or radiation
condition, that (in the absence of boundaries) sound produced by a source
must radiate away from the source. It is also consistent with the Second Law
of Thermodynamics, which requires natural systems to change in the more
probable direction. An event in which sound waves converge on a point from
all directions at inﬁnity is so unlikely as to be impossible in practice; it would be
the acoustic analogue of the farscattered pieces of a broken cup spontaneously
reassembling.
To complete the solution it remains to determine the function +. We do this
by extending the solution down to the source at r = 0 by writing (c.f., (1.4.7))
ϕ =
+
t −
r
c
0
r
= lim
c→0
+
t −
r
c
0
(r
2
÷c
2
)
1
2
. r ≥ 0. (1.5.6)
Let us substitute this into Equation (1.5.1) and examine what happens as c →0.
By direct calculation we ﬁnd
∇
2
ϕ =
1
r
∂
2
∂r
2
(rϕ) = −
3c
2
+(t −r¡c
0
)
(r
2
÷c
2
)
5
2
−
2c
2
+
/
(t −r¡c
0
)
c
0
r(r
2
÷c
2
)
3
2
÷
+
//
(t −r¡c
0
)
c
2
0
(r
2
÷c
2
)
1
2
1
c
2
0
∂
2
ϕ
∂t
2
=
+
//
(t −r¡c
0
)
c
2
0
(r
2
÷c
2
)
1
2
12 1 Introduction
Therefore,
1
c
2
0
∂
2
ϕ
∂t
2
−∇
2
ϕ =
3c
2
+(t −r¡c
0
)
(r
2
÷c
2
)
5
2
÷
2c
2
+
/
(t −r¡c
0
)
c
0
r(r
2
÷c
2
)
3
2
→4π+(t )δ(x) ÷0 as c →0. (1.5.7)
where the δ function in the last line follows from (1.4.8), and the ‘÷ 0’ is
obtained by noting that for any smoothly varying test function f (x) and any
volume V enclosing the origin
V
2c
2
f (x) d
3
x
r(r
2
÷c
2
)
3
2
≈ f (0)
∞
−∞
2c
2
d
3
x
r(r
2
÷c
2
)
3
2
= f (0)
∞
0
8πc
2
r dr
(r
2
÷c
2
)
3
2
= 8πc f (0) →0 as c →0.
Hence, comparing (1.5.7) with the inhomogeneous wave equation (1.5.1),
we ﬁnd
+(t ) =
1
4π
δ(t ).
and the Solution (1.5.6) becomes
ϕ(x. t ) =
1
4πr
δ
t −
r
c
0
≡
1
4π[x[
δ
t −
[x[
c
0
. (1.5.8)
This represents a spherical pulse that is nonzero only on the surface of the
sphere r = c
0
t >0, whose radius increases at the speed of sound c
0
; it vanishes
everywhere for t 0, before the impulsive source is triggered.
1.6 FreeSpace Green’s Function
The freespace Green’s function G(x. y. t −τ) is the causal solution of the wave
equation generated by the impulsive point source δ(x − y)δ(t − τ), located at
the point x = y at time t = τ. The formula for G is obtained from the solution
(1.5.8) for a source at x = 0 at t = 0 simply by replacing x by x −y and t by
t −τ. In other words, if
1
c
2
0
∂
2
∂t
2
−∇
2
G =δ(x −y)δ(t −τ). where G =0 for t τ. (1.6.1)
then
G(x. y. t −τ) =
1
4π[x −y[
δ
t −τ −
[x −y[
c
0
. (1.6.2)
1.7 Monopoles, Dipoles, and Quadrupoles 13
This represents an impulsive, spherically symmetric wave expanding from the
source at y at the speed of sound. The wave amplitude decreases inversely with
distance [x −y[ from the source point y.
Green’s function is the fundamental building block for forming solutions of
the inhomogeneous wave equation (1.3.7) of linear acoustics. Let us write this
equation in the form
1
c
2
0
∂
2
∂t
2
−∇
2
p = F(x. t ). (1.6.3)
where the generalized source F(x. t ) is assumed to be generating waves that
propagate away from the source region, in accordance with the radiation
condition.
This source distribution can be regarded as a distribution of impulsive point
sources of the type on the right of Equation (1.6.1), because
F(x. t ) =
∞
−∞
F(y. τ) δ(x −y) δ(t −τ) d
3
y dτ.
The outgoing wave solution for each constituent source of strength
F(y. τ)δ(x −y)δ(t −τ) d
3
y dτ is F(y. τ)G(x. y. t −τ) d
3
y dτ.
so that by adding up these individual contributions we obtain
p(x. t ) =
∞
−∞
F(y. τ)G(x. y. t −τ) d
3
y dτ (1.6.4)
=
1
4π
∞
−∞
F(y. τ)
[x −y[
δ
t −τ −
[x −y[
c
0
d
3
y dτ (1.6.5)
i.e., p(x. t ) =
1
4π
∞
−∞
F
y. t −
[x−y[
c
0
[x −y[
d
3
y. (1.6.6)
The integral formula (1.6.6) is called a retarded potential; it represents the
pressure at position x and time t as a linear superposition of contributions from
sources at positions y, which radiated at the earlier times t −[x−y[¡c
0
, [x−y[¡c
0
being the time of travel of sound waves from y to x.
1.7 Monopoles, Dipoles, and Quadrupoles
Avolume point source q(t )δ(x) of the type considered in Section 1.4 as a model
for a pulsating sphere is also called a monopole point source. For a compressible
14 1 Introduction
medium the corresponding velocity potential it produces is the solution of the
equation
1
c
2
0
∂
2
∂t
2
−∇
2
ϕ = −q(t )δ(x). (1.7.1)
The solution can be written down by analogy with the Solution (1.6.6) of
Equation (1.6.3) for the acoustic pressure. Replace p by ϕ in (1.6.6) and set
F(y. τ) = −q(τ)δ(y). Then,
ϕ(x. t ) =
−q
t −
[x[
c
0
4π[x[
≡
−q
t −
r
c
0
4πr
. (1.7.2)
This differs from the corresponding solution (1.4.3) for a pulsating sphere
or volume point source in an incompressible ﬂuid by the dependence on the
retarded time t −
r
c
0
. This is physically more realistic; any effects associated
with changes in the motion of the sphere (i.e., in the value of the volume out
ﬂow rate q(t )) are now communicated to a ﬂuid element at distance r after
an appropriate time delay r¡c
0
required for sound to travel outward from the
source.
1.7.1 The Point Dipole
Let f = f(t ) be a timedependent vector. Then a source on the right of the
acoustic pressure equation (1.6.3) of the form
F(x. t ) = div(f(t )δ(x)) ≡
∂
∂x
j
( f
j
(t )δ(x)) (1.7.3)
is called a point dipole (located at the origin). As explained in the Preface, a
repeated italic subscript, such as j in this equation, implies a summation over
j = 1. 2. 3. Equation (1.3.7) shows that the point dipole is equivalent to a force
distribution F(t ) = −f(t )δ(x) per unit volume applied to the ﬂuid at the origin.
The sound produced by the dipole can be calculated from (1.6.6), but it is
easier to use (1.6.5):
p(x. t ) =
1
4π
∞
−∞
∂
∂y
j
( f
j
(τ)δ(y))
δ
t −τ −
[x−y[
c
0
[x −y[
d
3
y dτ.
Integrate by parts with respect to each y
j
(recalling that δ(y) = 0 at y
j
= ±∞),
and note that
∂
∂y
j
δ
t −τ −
[x−y[
c
0
[x −y[
= −
∂
∂x
j
δ
t −τ −
[x−y[
c
0
[x −y[
.
1.7 Monopoles, Dipoles, and Quadrupoles 15
Then,
p(x. t ) =
1
4π
∞
−∞
f
j
(τ)δ(y)
∂
∂x
j
δ
t −τ −
[x−y[
c
0
[x −y[
d
3
y dτ
=
1
4π
∂
∂x
j
∞
−∞
f
j
(τ)δ(y)
δ
t −τ −
[x−y[
c
0
[x −y[
d
3
y dτ.
Thus,
p(x. t ) =
∂
∂x
j
f
j
t −
[x[
c
0
4π[x[
. (1.7.4)
The same procedure shows that for a distributed dipole source of the type
F(x. t ) = div f(x. t ) on the right of Equation (1.6.3), the acoustic pressure
becomes
p(x. t ) =
1
4π
∂
∂x
j
∞
−∞
f
j
y. t −
[x−y[
c
0
[x −y[
d
3
y. (1.7.5)
A point dipole at the origin orientated in the direction of a unit vector n
is entirely equivalent to two point monopoles of equal but opposite strengths
placed a short distance apart (much smaller than the acoustic wavelength) on
opposite sides of the origin on a line through the origin parallel to n. For
example, if n is parallel to the x axis, and the sources are distance c apart, the
two monopoles would be
q(t )δ
x −
c
2
δ(y)δ(z) −q(t )δ
x ÷
c
2
δ(y)δ(z)
≈−cq(t )δ
/
(x)δ(y)δ(z) ≡ −
∂
∂x
(cq(t )δ(x)).
This is a ﬂuid volume dipole. The relation p = −ρ
0
∂ϕ¡∂t implies that the
equivalent dipole source in the pressure equation (1.3.7) or (1.6.3) is
−ρ
0
∂
∂x
(c ˙ q(t )δ(x)).
where the dot denotes differentiation with respect to time.
1.7.2 Quadrupoles
A source distribution involving two space derivatives is equivalent to a combi
nation of four monopole sources (whose net volume source strength is zero),
16 1 Introduction
and is called a quadrupole. A general quadrupole is a source of the form
F(x. t ) =
∂
2
T
i j
∂x
i
∂x
j
(x. t ) (1.7.6)
in Equation (1.6.3). The argument above leading to Expression (1.7.5) can be
applied twice to show that the corresponding acoustic pressure is given by
p(x. t ) =
1
4π
∂
2
∂x
i
∂x
j
∞
−∞
T
i j
(y. t −[x −y[¡c
0
)
[x −y[
d
3
y. (1.7.7)
1.7.3 Vibrating Sphere
Let a rigid sphere of radius a execute small amplitude oscillations at speed U(t )
in the x
1
direction (Fig. 1.7.1a). Take the coordinate origin at the mean position
of the center. In Section 3.5, we shall prove that the motion induced in an ideal
ﬂuid when the sphere is small is equivalent to that produced by a point volume
dipole of strength 2πa
3
U(t ) at its center directed along the x
1
axis, determined
Fig. 1.7.1.
1.7 Monopoles, Dipoles, and Quadrupoles 17
by the solution of
1
c
2
0
∂
2
∂t
2
−∇
2
ϕ =
∂
∂x
1
(2πa
3
U(t )δ(x)). (1.7.8)
By analogy with (1.7.3) and (1.7.4), we have
ϕ(x. t ) =
∂
∂x
1
2πa
3
U
t −
[x[
c
0
4π[x[
. (1.7.9)
Now,
∂
∂x
j
[x[ =
x
j
[x[
. (1.7.10)
Applying this formula for j = 1, we ﬁnd (putting r = [x[ and x
1
= r cos θ)
ϕ =−
a
3
cos θ
2r
2
U
t −
r
c
0
−
a
3
cos θ
2c
0
r
∂U
∂t
t −
r
c
0
.
near ﬁeld far ﬁeld
The nearﬁeld term is dominant at sufﬁciently small distances r from the origin
such that
1
r
·
1
c
0
1
U
∂U
∂t
∼
f
c
0
.
where f is the characteristic frequency of the oscillations of the sphere. But,
sound of frequency f travels a distance
c
0
¡f =λ ≡one acoustic wavelength
in one period of oscillation 1¡f . Hence, the nearﬁeld term is dominant when
r <λ.
The motion becomes incompressible when c
0
→∞. In this limit the solution
reduces entirely to the nearﬁeld term, which is also called the hydrodynamic
near ﬁeld; it decreases in amplitude like 1¡r
2
as r →∞.
The far ﬁeld is the acoustic region that only exists when the ﬂuid is com
pressible. It consists of propagating sound waves, carrying energy away from
the sphere, and takes over from the near ﬁeld when r ·λ. There is an interme
diate zone where r ∼ λ in which the solution is in a state of transition from the
near to the far ﬁeld. The analytical model (1.7.8), in which the sphere is replaced
by a point dipole at its center, involves the implicit assumption that the motion
18 1 Introduction
close to the sphere is essentially the same as if the ﬂuid is incompressible. It fol
lows from what we have just said that a <λ, that is, the diameter of the sphere
is much smaller than the acoustic wavelength. In general, a body is said to be
acoustically compact when its characteristic dimension is small compared to
the wavelengths of the sound waves it is producing or with which it interacts.
The intensity of the sound generated by the sphere in the far ﬁeld is propor
tional to ϕ
2
:
ϕ
2
→
a
6
4c
2
0
r
2
∂U
∂t
t −
r
c
0
2
cos
2
θ.
The dependence on θ determines the directivity of the sound. For the dipole it
has the ﬁgure of eight pattern illustrated in Fig. 1.7.1b, with peaks in directions
parallel to the dipole axis (θ = 0. π); there are radiation nulls at θ =
π
2
(the
curve should be imagined to be rotated about the x
1
axis).
1.8 Acoustic Energy Flux
At large distances r from a source region we generally have
p(x. t ) ∼
ρ
0
+
θ. φ. t −
r
c
0
r
. r →∞. (1.8.1)
where the function + depends on the nature of the source distribution, and θ
and φ are polar angles determining the directivity of the sound. From the radial
component of the linearized momentum equation
∂:
r
∂t
= −
1
ρ
0
∂p
∂r
≡
1
r
2
+
θ. φ. t −
r
c
0
÷
1
c
0
r
∂+
∂t
θ. φ. t −
r
c
0
. (1.8.2)
The ﬁrst term in the second line can be neglected when r →∞, and therefore
:
r
∼
1
c
0
r
+
θ. φ. t −
r
c
0
≡
p
ρ
0
c
0
. (1.8.3)
By considering the θ and φ components of the momentumequation we can show
that the corresponding velocity components :
θ
. :
φ
, say, decrease faster than 1¡r
as r →∞. We therefore conclude fromthis and (1.8.3) that the acoustic particle
velocity is normal to the acoustic wavefronts (the spherical surfaces r = c
0
t ).
In other words, sound consists of longitudinal waves in which the ﬂuid particles
oscillate backwards and forwards along the local direction of propagation of
the sound.
1.8 Acoustic Energy Flux 19
The acoustic power H radiated by a source distribution can be computed
from the formula
H =
S
p:
r
dS =
S
p
2
ρ
0
c
0
dS. (1.8.4)
where the integration is over the surface S of a large sphere of radius r centered
on the source region. Because the surface area = 4πr
2
, we only need to know
the pressure and velocity correct to order 1¡r on S in order to evaluate the
integral. Smaller contributions (such as that determined by the ﬁrst term in
the second line of (1.8.2)) decrease too fast as r increases to supply a ﬁnite
contribution to the integral as r →∞.
In acoustic problems we are therefore usually satisﬁed if we can calculate
the pressure and velocity in the acoustic far ﬁeld correct to order 1¡r; this will
always enable us to determine the radiated sound power. The formula :
r
=
p¡ρ
0
c
0
is applicable at large distances from the sources, where the wavefronts
can be regarded as locally plane, but it is true identically for plane sound waves.
In the latter case, and for spherical waves on the surface of the large sphere of
Fig. 1.8.1, the quantity
I = p:
r
=
p
2
ρ
0
c
0
(1.8.5)
is called the acoustic intensity. It is the rate of transmission of acoustic energy
per unit area of wavefront.
Fig. 1.8.1.
20 1 Introduction
Fig. 1.9.1.
1.9 Calculation of the Acoustic Far Field
We now discuss the approximations necessary to evaluate the sound in the far
ﬁeld from the retarded potential representation:
p(x. t ) =
1
4π
∞
−∞
F
y. t −
[x−y[
c
0
[x −y[
d
3
y. (1.9.1)
We assume that F(x. t ) ,= 0 only within a ﬁnite source region (Fig. 1.9.1), and
take the coordinate origin O within the region.
When [x[ →∞and y lies within the source region (so that [x[ ·[y[)
[x −y[ ≡ ([x[
2
−2x · y ÷[y[
2
)
1
2
= [x[
¸
1 −
2x · y
[x[
2
÷
[y[
2
[x[
2
¸
1
2
≈ [x[
¸
1 −
x · y
[x[
2
÷ O
[y[
2
[x[
2
¸
Then,
[x −y[ ≈ [x[ −
x · y
[x[
when
[y[
[x[
<1. (1.9.2)
Also,
1
[x −y[
≈
1
[x[ −
x · y
[x[
≈
1
[x[
1 ÷
x · y
[x[
2
Therefore,
1
[x −y[
≈
1
[x[
÷
x · y
[x[
3
when
[y[
[x[
<1. (1.9.3)
The Approximation (1.9.3) shows that, in order to obtain the farﬁeld approx
imation of the Solution (1.9.1) that behaves like 1¡r = 1¡[x[ as [x[ →∞, it is
sufﬁcient to replace [x−y[ in the denominator of the integrand by [x[. However,
in the argument of the source strength F it is important to retain possible phase
1.9 Calculation of the Acoustic Far Field 21
differences between the sound waves generated by components of the source
distribution at different locations y; we therefore replace [x −y[ in the retarded
time by the righthand side of (1.9.2). Hence,
p(x. t ) ≈
1
4π[x[
∞
−∞
F
y. t −
[x[
c
0
÷
x · y
c
0
[x[
d
3
y. [x[ →∞. (1.9.4)
This is called the Fraunhofer approximation.
The source region may extend over many characteristic acoustic wavelengths
of the sound. By retaining the contribution x · y¡c
0
[x[ to the retarded time we
ensure that any interference between waves generated at different positions
within the source region is correctly described by the farﬁeld approximation.
In Fig. 1.9.1 the acoustic travel time from a source point y to the far ﬁeld point
x is equal to that from the point labelled A to x when [x[ →∞. The travel time
over the distance OA is just x · y¡c
0
[x[, so that [x[¡c
0
− x · y¡c
0
[x[ gives the
correct value of the retarded time when [x[ →∞.
1.9.1 Dipole Source Distributions
By applying the farﬁeld formula (1.9.4) to a dipole source F(x. t ) = div f(x. t )
we obtain (from (1.7.5))
p(x. t ) ≈
1
4π
∂
∂x
j
¸
1
[x[
∞
−∞
f
j
y. t −
[x[
c
0
÷
x · y
c
0
[x[
d
3
y
¸
≈
1
4π[x[
∂
∂x
j
∞
−∞
f
j
y. t −
[x[
c
0
÷
x · y
c
0
[x[
d
3
y. [x[ →∞. (1.9.5)
because the differential operator ∂¡∂x
j
need not be applied to 1¡[x[ as this would
give a contribution decreasing like 1¡r
2
at large distances from the dipole.
However, it is useful to make a further transformation that replaces ∂¡∂x
j
by
the time derivative ∂¡∂t , which is usually more easily estimated in applications.
To do this, we observe that
∂ f
j
∂x
j
y. t −
[x[
c
0
÷
x · y
c
0
[x[
=
∂ f
j
∂t
y. t −
[x[
c
0
÷
x · y
c
0
[x[
∂
∂x
j
t −
[x[
c
0
÷
x · y
c
0
[x[
=
∂ f
j
∂t
y. t −
[x[
c
0
÷
x · y
c
0
[x[
−
x
j
c
0
[x[
÷
y
j
c
0
[x[
−
(x · y)x
j
c
0
[x[
3
≈ −
x
j
c
0
[x[
∂ f
j
∂t
y. t −
[x[
c
0
÷
x · y
c
0
[x[
as [x[ →∞.
22 1 Introduction
Hence, the far ﬁeldof a distributionof dipoles F(x. t ) = div f(x. t ) is givenby
p(x. t ) =
−x
j
4πc
0
[x[
2
∂
∂t
∞
−∞
f
j
y. t −
[x[
c
0
÷
x · y
c
0
[x[
d
3
y. (1.9.6)
Note that
x
j
[x[
2
=
x
j
[x[
1
[x[
.
where x
j
¡[x[ is the j th component of the unit vector x¡[x[. Thus, the additional
factor of x
j
¡[x[ in (1.9.6) does not change the rate of decay of the sound with
distance from the source (which is still like 1¡r), but it does have an inﬂuence
on the acoustic directivity.
A comparison of (1.9.5) and (1.9.6) leads to the following rule for inter
changing space and time derivatives in the acoustic far ﬁeld:
∂
∂x
j
←→−
1
c
0
x
j
[x[
∂
∂t
. (1.9.7)
1.9.2 Quadrupole Source Distributions
For the Quadrupole (1.7.6)
F(x. t ) =
∂
2
T
i j
∂x
i
∂x
j
(x. t ).
and
p(x. t ) =
1
4π
∂
2
∂x
i
∂x
j
∞
−∞
T
i j
(y. t −[x −y[¡c
0
)
[x −y[
d
3
y.
By applying (1.9.4) and the rule (1.9.7), we ﬁnd that the acoustic far ﬁeld is
given by
p(x. t ) ≈
x
i
x
j
4πc
2
0
[x[
3
∂
2
∂t
2
∞
−∞
T
i j
y. t −
[x[
c
0
÷
x · y
c
0
[x[
d
3
y. [x[ →∞.
(1.9.8)
1.9.3 Example
For the (1, 2) point quadrupole
F(x. t ) =
∂
2
∂x
1
∂x
2
(T(t )δ(x))
Problems 1 23
Fig. 1.9.2.
Equation (1.9.8) shows that in the acoustic far ﬁeld
p(x. t ) ≈
x
1
x
2
4πc
2
0
[x[
3
∂
2
T
∂t
2
t −
[x[
c
0
. [x[ →∞.
If we use spherical polar coordinates, such that
x
1
= r cos θ. x
2
= r sin θ cos φ. x
3
= r sin θ sin φ.
we can write the pressure in the form
p(x. t ) ≈
sin 2θ cos φ
8πc
2
0
[x[
∂
2
T
∂t
2
t −
[x[
c
0
. [x[ →∞.
The directivity of the sound (∝ p
2
) is therefore represented by sin
2
2θ cos
2
φ.
Its shape is plotted in Fig. 1.9.2 for radiation in the x
1
. x
2
plane (φ = 0. π). The
fourlobe cloverleaf pattern is characteristic of a quadrupole T
i j
for whichi ,= j .
Problems 1
1. A plane sound wave propagating parallel to the x axis satisﬁes the equation
1
c
2
0
∂
2
∂t
2
−
∂
2
∂x
2
ϕ = 0.
24 1 Introduction
with general solution
ϕ = +
t −
x
c
0
÷+
t ÷
x
c
0
.
where +and + are arbitrary functions respectively representing waves pro
pagating in the positive and negative x directions.
Showthat for a wave propagating in the positive x direction in an ideal gas
: =
p
ρ
0
c
0
. ρ =
p
c
2
0
. T =
p
ρ
0
c
p
.
where : is the acoustic particle velocity; p, ρ, and T are respectively the
acoustic pressure, density, and temperature variations, and c
p
is the speciﬁc
heat at constant pressure.
2. Calculate the acoustic power (1.8.4) radiated by an acoustically compact
sphere of radius R executing small amplitude translational oscillations of
frequency ω and velocity U(t ) = U
0
cos(ωt ), where U
0
= constant.
3. As for Problem 2, when the sphere executes small amplitude radial oscilla
tions at normal velocity :
n
= U
0
cos(ωt ), U
0
= constant.
4. A volume point source of strength q
0
(t ) translates at constant, subsonic ve
locity U. The velocity potential ϕ(x. t ) of the radiated sound is determined
by the solution of
1
c
2
0
∂
2
∂t
2
−∇
2
ϕ = −q
0
(t )δ(x −Ut ).
Show that
ϕ(x. t ) =
−q
0
(t − R¡c
0
)
4π R(1 − M cos O)
. M =
U
c
0
.
where R is the distance of the reception point x from the source position at
the time of emission of the sound received at x at time t , and O is the angle
between U and the direction of propagation of this sound.
2
Lighthill’s Theory
2.1 The Acoustic Analogy
The sound generated by turbulence in an unbounded ﬂuid is usually called
aerodynamic sound. Most unsteady ﬂows of technological interest are of high
Reynolds number and turbulent, and the acoustic radiation is a very small by
product of the motion. The turbulence is usually produced by ﬂuid motion
over a solid boundary or by ﬂow instability. Lighthill (1952) transformed the
Navier–Stokes and continuity equations to forman exact, inhomogeneous wave
equation whose source terms are important only within the turbulent region.
He argued that sound is a very small component of the whole motion and that,
once generated, its backreaction on the main ﬂow can usually be ignored. The
properties of the unsteady ﬂow in the source region may then be determined
by neglecting the production and propagation of the sound, a reasonable ap
proximation if the Mach number M is small, and there are many important
ﬂows where the hypothesis is obviously correct, and where the theory leads to
unambiguous predictions of the sound.
Lighthill was initially interested in solving the problem, illustrated in
Fig. 2.1.1a, of the sound produced by a turbulent nozzle ﬂow. However, his
original theory actually applies to the simpler situation shown in Fig. 2.1.1b, in
which the sound is imagined to be generated by a ﬁnite region of rotational ﬂow
in an unbounded ﬂuid. This avoids complications caused by the presence of the
nozzle. The ﬂuid is assumed to be at rest at inﬁnity, where the mean pressure,
density, and sound speed are respectively equal to p
0
. ρ
0
. c
0
. Lighthill com
pared the equations for the production of acoustic density ﬂuctuations in the
real ﬂow with those in an ideal linear acoustic medium that coincides with the
real ﬂuid at large distances from the sources.
25
26 2 Lighthill’s Theory
Fig. 2.1.1.
Todothis, bodyforces are neglected, andthe i thcomponent of the momentum
equation (1.2.3) is cast in the form
ρ
∂:
i
∂t
÷ρ:
j
∂:
i
∂x
j
= −
∂p
∂x
i
÷
∂σ
i j
∂x
j
≡ −
∂
∂x
j
( pδ
i j
−σ
i j
). (2.1.1)
δ
i j
is the Kronecker delta (= 1 for i = j. and 0 for i ,= j ), and σ
i j
is the
viscous stress tensor deﬁned (for a Stokesian ﬂuid) by
σ
i j
= 2η
e
i j
−
1
3
e
kk
δ
i j
. (2.1.2)
where
e
i j
=
1
2
∂:
i
∂x
j
÷
∂:
j
∂x
i
(2.1.3)
2.1 The Acoustic Analogy 27
is the rate of strain tensor. Next multiply the continuity equation (1.2.1) by :
i
:
:
i
∂ρ
∂t
÷:
i
∂(ρ:
j
)
∂x
j
= 0.
By adding this to Equation (2.1.1), we obtain the Reynolds form of the momen
tum equation
∂(ρ:
i
)
∂t
= −
∂π
i j
∂x
j
. (2.1.4)
where
π
i j
= ρ:
i
:
j
÷( p − p
0
)δ
i j
−σ
i j
. (2.1.5)
is called the momentum ﬂux tensor, and the constant pressure p
0
is inserted
for convenience.
In an ideal, linear acoustic medium, the momentum ﬂux tensor contains only
the pressure
π
i j
→π
0
i j
= ( p − p
0
)δ
i j
≡ c
2
0
(ρ −ρ
0
)δ
i j
. (2.1.6)
and the momentum equation then reduces to
∂(ρ:
i
)
∂t
÷
∂
∂x
i
¸
c
2
0
(ρ −ρ
0
)
¸
= 0. (2.1.7)
If the continuity equation (1.2.1) is written in the slightly modiﬁed form
∂
∂t
(ρ −ρ
0
) ÷
∂(ρ:
i
)
∂x
i
= 0. (2.1.8)
we can eliminate the momentum density ρ:
i
between (2.1.7) and (2.1.8) to
obtain the equation of linear acoustics satisﬁed by the perturbation density
ρ −ρ
0
1
c
2
0
∂
2
∂t
2
−∇
2
¸
c
2
0
(ρ −ρ
0
)
¸
= 0. (2.1.9)
Because the turbulence is neglected in this approximation, and there are no
externally applied forces or moving boundaries, the unique solution of this
equation that satisﬁes the radiation condition of outgoing wave behavior is
simply ρ −ρ
0
= 0.
It can now be asserted that the sound generated by the turbulence in the
real ﬂuid is exactly equivalent to that produced in the ideal, stationary acoustic
28 2 Lighthill’s Theory
medium (which is governed by (2.1.9) in turbulencefree regions) forced by the
stress distribution
T
i j
= π
i j
−π
0
i j
= ρ:
i
:
j
÷
( p − p
0
) −c
2
0
(ρ −ρ
0
)
δ
i j
−σ
i j
. (2.1.10)
where T
i j
is called the Lighthill stress tensor. This is Lighthill’s acoustic
analogy.
Indeed, we can rewrite (2.1.4) as the momentum equation for an ideal, sta
tionary acoustic medium of mean density ρ
0
and sound speed c
0
subject to the
externally applied stress T
i j
∂(ρ:
i
)
∂t
÷
∂π
0
i j
∂x
j
= −
∂
∂x
j
π
i j
−π
0
i j
.
or
∂(ρ:
i
)
∂t
÷
∂
∂x
i
¸
c
2
0
(ρ −ρ
0
)
¸
= −
∂T
i j
∂x
j
. (2.1.11)
By eliminating the momentum density ρ:
i
between this and the continuity
equation (2.1.8) (the same procedure used above for the linear problem), we
obtain Lighthill’s equation
1
c
2
0
∂
2
∂t
2
−∇
2
¸
c
2
0
(ρ −ρ
0
)
¸
=
∂
2
T
i j
∂x
i
∂x
j
. (2.1.12)
This is the exact, nonlinear counterpart of (2.1.9). The problem of calculating
the turbulence generated sound is therefore equivalent to solving this equation
for the radiation into a stationary, ideal ﬂuid produced by a distribution of
quadrupole sources whose strength per unit volume is the Lighthill stress tensor
T
i j
. The quadrupole character of the turbulence sources is one of the most
important conclusions of Lighthill’s theory; it implies (see Section 2.2) that
freeﬁeld turbulence is an extremely weak sound source, and that in a typical
low Mach number ﬂow only a tiny fraction of the available ﬂow energy is
converted into sound.
In the deﬁnition (2.1.10) of T
i j
, the term ρ:
i
:
j
is called the Reynolds stress.
For the simpliﬁed problem of Fig. 2.1.1b it is a nonlinear quantity that can be
neglected except where the motion is turbulent. The second term represents
the excess of momentum transfer by the pressure over that in the ideal (linear)
ﬂuid of density ρ
0
and sound speed c
0
. This is produced by wave amplitude
nonlinearity, and by mean density variations in the source ﬂow. The viscous
2.2 Lighthill’s :
8
Law 29
stress tensor σ
i j
is linear in the perturbation quantities, and properly accounts
for the attenuation of the sound; in most applications the Reynolds number
in the source region is very large, and σ
i j
can be neglected, and the viscous
attenuation of the radiating sound is usually ignored.
2.2 Lighthill’s v
8
Law
The formal solution of Lighthill’s equation (2.1.12) with outgoing wave be
havior is given by (1.7.7) with p(x. t ) replaced by c
2
0
(ρ −ρ
0
)
c
2
0
(ρ −ρ
0
)(x. t ) =
1
4π
∂
2
∂x
i
∂x
j
∞
−∞
T
i j
(y. t −[x −y[¡c
0
)
[x −y[
d
3
y. (2.2.1)
This is strictly an alternative, integral equation representation of Equation
(2.1.12); it provides a useful prediction of the sound only when T
i j
is known
or has been determined by some other means. This is because the terms in the
deﬁnition (2.1.10) of T
i j
not only account for the generation of sound, but also
govern acoustic selfmodulation caused by acoustic nonlinearity, the convec
tion of sound waves by the turbulent velocity, refraction caused by sound speed
variations, and attenuation due to thermal and viscous actions. The inﬂuence
of acoustic nonlinearity and of thermoviscous dissipation is usually sufﬁciently
weak to be neglected within the source region, although they may affect prop
agation to a distant observer. Convection and refraction of sound within and
near the source ﬂow can be important, for example in the presence of a mean
shear layer (when the Reynolds stress will include terms like ρU
i
u
j
, where U
and u respectively denote the mean and ﬂuctuating components of v), or when
there are large variations in the mean thermodynamic properties of the medium
within the source region; such effects are described by the presence of unsteady
linear terms in T
i j
(Ffowcs Williams, 1974).
Thus, to predict the radiated sound from Lighthill’s equation (2.2.1) it is
usually necessary to suppose that all of these acoustic effects in the source ﬂow
(which really depend on ﬂuid compressibility) are in some sense negligible.
This means that in practice it must be possible to derive a good approximation
for T
i j
by taking the source ﬂow to be effectively incompressible. This is often
possible when the characteristic Mach number M ∼ :¡c
0
is small (speciﬁcally,
when M
2
<1), and when the wavelength of the sound is much larger than the
size of the source region.
Consider the particular but important case in which the mean density and
sound speed are uniform throughout the ﬂuid. The variations in the density
ρ within a low Mach number, high Reynolds number source ﬂow are then of
30 2 Lighthill’s Theory
order ρ
0
M
2
(Batchelor, 1967). Thus, ρ:
i
:
j
= ρ
0
(1 ÷ O(M
2
)):
i
:
j
≈ ρ
0
:
i
:
j
.
Similarly, if c(x. t ) is the local speed of sound in the source region, it may also
be shown that c
2
0
¡c
2
= 1 ÷O(M
2
), so that
p − p
0
−c
2
0
(ρ −ρ
0
) ≈ ( p − p
0
)
1 −c
2
0
c
2
∼ O(ρ
0
:
2
M
2
).
Hence, if viscous dissipation is neglected we make the approximation
T
i j
≈ ρ
0
:
i
:
j
. provided that M
2
<1. (2.2.2)
In the acoustic region outside the source ﬂow c
2
0
(ρ − ρ
0
) = p − p
0
. If the
irrelevant constant pressure p
0
is suppressed, the Solution (2.2.1) of Lighthill’s
equation therefore becomes
p(x. t ) ≈
∂
2
∂x
i
∂x
j
ρ
0
:
i
:
j
(y. t −[x −y[¡c
0
)
4π[x −y[
d
3
y
≈
x
i
x
j
4πc
2
0
[x[
3
∂
2
∂t
2
ρ
0
:
i
:
j
y. t −
[x[
c
0
÷
x · y
c
0
[x[
d
3
y. [x[ →∞.
(2.2.3)
where in the second line we have used the formula (1.9.8) for the far ﬁeld
of a quadrupole distribution. Quantitative predictions can be made from this
formula provided the behavior of the Reynolds stress ρ
0
:
i
:
j
is known.
To determine the order of magnitude of p, we introduce a characteristic
velocity : and length scale ¹ (of the energycontaining eddies) of the turbulence
sources. The value of ¹ depends on the mechanism responsible for turbulence
production, suchas the widthof a jet mixinglayer. Fluctuations in:
i
:
j
occurring
in different regions of the turbulent ﬂowseparated by distances >O(¹) will tend
to be statistically independent, and the sound may be considered to be generated
by a collection of V
0
¡¹
3
independent eddies, where V
0
is the volume occupied
by the turbulence (Fig. 2.2.1). The characteristic frequency of the turbulent
ﬂuctuations f ∼ :¡¹, so that the wavelength (c
0
¡f ) of the sound ∼¹¡M · ¹
(because M = :¡c
0
< 1). Hence, we arrive at the important conclusion that
the turbulence eddies are each acoustically compact. This means that when
the integral in (2.2.3) is conﬁned to a single eddy, the retarded time variations
x · y¡c
0
[x[ across that eddy can be neglected; that is, if the coordinate origin
is temporarily placed at O within the eddy, we can set x · y¡c
0
[x[ = 0 in the
integration over that eddy. The value of the integral over the eddy then may be
estimated to be of order ρ
0
:
2
¹
3
.
2.2 Lighthill’s :
8
Law 31
Fig. 2.2.1.
The order of magnitude of the time derivative for changes in the source region
is
∂
∂t
∼
:
¹
.
Therefore, it follows from (2.2.3) that, for one eddy, the farﬁeld acoustic pres
sure satisﬁes
p ∼
¹
[x[
ρ
0
:
4
c
2
0
=
¹
[x[
ρ
0
:
2
M
2
. (2.2.4)
The acoustic power radiated by the eddy is determined by the surface inte
gral (1.8.4) taken over a large sphere centered on the eddy. Thus, in order of
magnitude,
acoustic power radiated by one eddy ∼4π[x[
2
p
2
ρ
0
c
0
∼
¹
2
ρ
0
:
8
c
5
0
= ¹
2
ρ
0
:
3
M
5
.
(2.2.5)
This is Lighthill’s ‘eighth power’ law.
The total power radiated from the whole of the turbulent region of volume
V
0
, containing V
0
¡¹
3
independent eddies, is
H
q
≈
V
0
¹
3
(¹
2
ρ
0
:
3
M
5
) =
:
¹
ρ
0
:
2
M
5
V
0
.
32 2 Lighthill’s Theory
Dimensional arguments and experiment indicate that the rate H
0
, say, at which
energy must be supplied by the action of external forces to maintain the kinetic
energy of a statistically steady turbulent ﬁeld occupying a volume V
0
is given
in order of magnitude by
H
0
∼
:
¹
ρ
0
:
2
V
0
.
Therefore, the mechanical efﬁciency with which turbulence kinetic energy is
converted into sound is
H
q
H
0
∼ M
5
. (2.2.6)
This is smaller than about 0.01 for Mach numbers M  0.4, conﬁrming
Lighthill’s hypothesis that the ﬂow generated sound is an inﬁnitesimal by
product of the turbulent motion.
2.3 Curle’s Theory
In most applications of Lighthill’s theory it is necessary to generalize the so
lution (2.2.1) to account for the presence of solid bodies in the ﬂow. Indeed,
turbulence is frequently generated in the boundary layers and wakes of ﬂow
past such bodies (airfoils, ﬂow control surfaces, etc.), and the unsteady surface
forces (dipoles) that arise are likely to make a signiﬁcant contribution to the
production of sound. The procedure in such cases is to introduce a system of
mathematical control surfaces that can be deformed to coincide with the sur
faces of the different moving or stationary bodies, although for the moment
we shall discuss only cases involving stationary bodies. Before doing this we
establish an integral transformation formula that is used repeatedly in problems
of this kind.
2.3.1 Volume and Surface Integrals
Let V be the ﬂuid outside a closed control surface S (Fig. 2.3.1) deﬁned by the
equation
f (x) = 0. where
¸
f (x) > 0 for x in V.
f (x)  0 for x within S.
(2.3.1)
and consider the Heaviside unit function
H( f ) =
¸
1 for x in V.
0 for x within S.
2.3 Curle’s Theory 33
Fig. 2.3.1.
Then, for an arbitrary function +(x) deﬁned in V and on S,
∞
−∞
+(x)∇H d
3
x =
S
+(x) ndS ≡
S
+(x) dS. (2.3.2)
or
∞
−∞
+(x)
∂ H
∂x
j
d
3
x =
S
+(x) n
j
dS ≡
S
+(x) dS
j
. (2.3.3)
where H ≡ H( f ) and n is the unit normal on S directed into V.
Proof
∇H( f ) ≡ δ( f ) ∇ f (2.3.4)
is nonzero only on S, where ∇ f is in the direction of n. The volume integral
is therefore conﬁned to the region between the inner and outer faces of a shell
of inﬁnitesimal thickness (between the broken line surfaces in Fig. 2.3.1) that
just encloses S, and in which the volume element is
d
3
x = ds
⊥
dS.
where s
⊥
= 0 on S and s
⊥
is measured parallel to n. Because f = 0 on S we
can write, for small values of s
⊥
,
f =
∂ f
∂s
⊥
S
s
⊥
.
where (∂ f ¡∂s
⊥
)
S
≡ [∇ f [ > 0 is evaluated on S.
34 2 Lighthill’s Theory
Therefore,
δ( f ) = δ([∇ f [s
⊥
) ≡
δ(s
⊥
)
[∇ f [
.
Hence,
∞
−∞
+(x)∇H d
3
x ≡
∞
−∞
+(x)∇ f δ( f ) d
3
x =
∞
−∞
+(x)
∇ f
[∇ f [
δ(s
⊥
) ds
⊥
dS
=
S
+(x)ndS. because n =
∇ f
[∇ f [
.
2.3.2 Curle’s Equation
Curle (1955) has derived a formal solution (called Curle’s equation) of
Lighthill’s equation (2.1.12) for the sound produced by turbulence in the vicin
ity of an arbitrary, ﬁxed surface S, deﬁned as above by an equation f (x) = 0
(Fig. 2.3.2). This surface may either enclose a solid body, or merely constitute
a control surface used to isolate a ﬁxed region of space containing both solid
bodies and ﬂuid or just ﬂuid.
To derive Curle’s equation, multiply the momentum equation (2.1.11) by
H ≡ H( f ), and use the deﬁnition (2.1.10) of T
i j
to obtain
∂
∂t
(ρ:
i
H) ÷
∂
∂x
i
Hc
2
0
(ρ −ρ
0
)
= −
∂
∂x
j
(HT
i j
) ÷(ρ:
i
:
j
÷ p
/
i j
)
∂ H
∂x
j
.
(2.3.5)
Fig. 2.3.2.
2.3 Curle’s Theory 35
where
p
/
i j
= ( p − p
0
)δ
i j
−σ
i j
(2.3.6)
is the compressive stress tensor. Repeat this operationfor the continuityequation
(2.1.8):
∂
∂t
(H(ρ −ρ
0
)) ÷
∂
∂x
i
(Hρ:
i
) = (ρ:
i
)
∂ H
∂x
i
. (2.3.7)
The Formula (2.3.4), ∇H = ∇ f δ( f ), shows that Equations (2.3.5) and (2.3.7)
formally determine the momentum density ρ:
i
and the density ﬂuctuation
(ρ − ρ
0
) in the exterior region V (where H( f ) ≡ 1) in terms of the Lighthill
stresses T
i j
in V and sources distributed over the control surface.
An analog of Lighthill’s equation (2.1.12) can now be obtained by eliminat
ing Hρ:
i
between (2.3.5) and (2.3.7). This is the differential form of Curle’s
equation
1
c
2
0
∂
2
∂t
2
−∇
2
¸
Hc
2
0
(ρ −ρ
0
)
¸
=
∂
2
(HT
i j
)
∂x
i
∂x
j
−
∂
∂x
i
(ρ:
i
:
j
÷ p
/
i j
)
∂ H
∂x
j
÷
∂
∂t
ρ:
j
∂ H
∂x
j
. (2.3.8)
The equation is valid throughout all space, including the region enclosed by
S where H( f ) vanishes. The second and third terms on the righthand side
respectively represent dipole and monopole sources distributed over S. They
have the following interpretations:
1. If S is merely an artiﬁcial control surface it will enclose ﬂuid, possibly also
solid bodies, and may or may not contain turbulence; the surface dipole and
monopole sources then represent the inﬂuence of this region on the sound
radiated in V; in other words the aggregate effect of the dipole and monopole
sources accounts for the presence of solid bodies and turbulence within S
(when T
i j
,= 0 in S) and also for the interaction of sound generated outside
S with the ﬂuid and solid bodies in S.
2. If S is the boundary of a solid body, the surface dipole represents the pro
duction of sound by the unsteady surface force that the body exerts on the
exterior ﬂuid, whereas the monopole is responsible for the sound produced
by volume pulsations (if any) of the body.
Because Curle’s formof Lighthill’s equation is valid throughout all space, the
outgoing wave solution is found from the general solution (1.6.6) of the wave
36 2 Lighthill’s Theory
equation (1.6.3), as before, using the special form (1.7.5) for dipole sources.
When account is taken of the transformation formula (2.3.3) this yields Curle’s
equation
Hc
2
0
(ρ −ρ
0
) =
∂
2
∂x
i
∂x
j
V
[T
i j
]
d
3
y
4π[x −y[
−
∂
∂x
i
S
[ρ:
i
:
j
÷ p
/
i j
]
dS
j
(y)
4π[x −y[
÷
∂
∂t
S
[ρ:
j
]
dS
j
(y)
4π[x −y[
. (2.3.9)
where the square bracket notation such as [T
i j
] ≡ T
i j
(y. t −[x −y[¡c
0
) implies
evaluation at the retarded time. Note that, because H( f ) ≡ 0 inside S, the sum
of the three integrals on the righthand side must also vanish when the ﬁeld
point x is within S.
2.4 Sound Produced by Turbulence Near a Compact Rigid Body
When the surface S (in Fig. 2.3.2) is rigid, Curle’s equation (2.3.9) reduces to
Hc
2
0
(ρ −ρ
0
) =
∂
2
∂x
i
∂x
j
V
[T
i j
]
d
3
y
4π[x −y[
−
∂
∂x
i
S
[ p
/
i j
]
dS
j
(y)
4π[x −y[
. (2.4.1)
We now use this solution to determine the order of magnitude of the sound gen
erated by an acoustically compact body within a turbulent ﬂow. Compactness
usually requires the Mach number M ∼ :¡c
0
<1, and we shall assume this to
be the case in the following.
The contribution from the quadrupole integral in (2.4.1) is estimated as in
Section 2.2. To deal with the surface dipole, note ﬁrst that for turbulence of
velocity : and correlation scale ¹, the orders of magnitude of the pressure and
viscous components of the compressive stress tensor
p
/
i j
= ( p − p
0
)δ
i j
−σ
i j
are
( p − p
0
) ∼ ρ
0
:
2
. σ ∼ η
:
¹
;
that is,
( p − p
0
)
σ
∼
ρ
0
:¹
η
=
:¹
ν
.
where ν = η¡ρ
0
is the kinematic viscosity. The dimensionless ratio Re = :¹¡ν
is the Reynolds number and is always very large (∼10
4
or more) in turbulent
ﬂow. This means that viscous contributions to the surface force can be neglected.
2.5 Radiation from a Noncompact Surface 37
In the far ﬁeld the pressure p(x. t ) = c
2
0
(ρ −ρ
0
)(x. t ), and H( f ) = 1. Thus,
applying the farﬁeld dipole approximation (1.9.6), and neglecting retarded time
variations x · y¡c
0
[x[ because S is compact, the dipole sound pressure p
d
, say,
can be written
p
d
≈
x
i
4πc
0
[x[
2
∂
∂t
S
( p − p
0
)
y. t −
[x[
c
0
dS
i
=
x
i
4πc
0
[x[
2
dF
i
dt
t −
[x[
c
0
.
[x[ →∞. (2.4.2)
where F(t ) is the unsteady force exerted on the ﬂuid by the body. The contri
bution to p
d
from a surface element of diameter ¹ within which the turbulence
surface pressure ﬂuctuations are correlated is evidently of order
1
c
0
[x[
:
¹
(ρ
0
:
2
¹
2
) =
¹
[x[
ρ
0
:
2
M.
which exceeds by an order of magnitude (1¡M ·1) the sound pressure (2.2.4)
produced by a quadrupole in V of length scale ¹. If A is the total surface area
wetted by the turbulent ﬂow, there are A¡¹
2
independently radiating surface
elements, and the total power radiated by the dipoles is
H
d
∼ 4π[x[
2
p
2
d
ρ
0
c
0
∼ Aρ
0
:
3
M
3
. (2.4.3)
The direct power radiated by quadrupoles occupying a volume V
0
is H
q
∼
(V
0
¡¹)ρ
0
:
3
M
5
, the same as in the absence of the body (see Section 2.2). The
sound produced by the turbulence near S is therefore dominated by the dipole
when M is small, and as M →0 the acoustic power exceeds the quadrupole
power by a factor ∼1¡M
2
· 1. Precisely how small M should be for this to
be true depends on the details of the ﬂow, which determine the appropriate
values of A and V
0
¡¹.
This increase in acoustic efﬁciency brought about by surface dipoles on an
acoustically compact body occurs also for arbitrary, noncompact bodies when
turbulence interacts with compact structural elements, such as edges, corners,
and protuberances.
2.5 Radiation from a Noncompact Surface
Consider a compact turbulent eddy in x
2
> 0 adjacent to an inﬁnite, plane rigid
wall at x
2
= 0 (Fig. 2.5.1). Let us apply the rigid surface form(2.4.1) of Curle’s
equation to calculate the radiation. At high Reynolds numbers and at x in the
38 2 Lighthill’s Theory
Fig. 2.5.1.
acoustic far ﬁeld (where p(x. t ) ≡ c
2
0
(ρ −ρ
0
) and H( f ) ≡ H(x
2
) = 1) we ﬁnd
p(x. t ) ≈
x
i
x
j
4πc
2
0
[x[
3
∂
2
∂t
2
T
i j
y. t −
[x[
c
0
d
3
y
÷
x
2
4πc
0
[x[
2
∂
∂t
y
2
=0
( p − p
0
)
y. t −
[x[
c
0
÷
x
1
y
1
÷ x
3
y
3
c
0
[x[
dy
1
dy
3
.
[x[ →∞. (2.5.1)
Retarded time variations have been neglected in the integral over the volume
of the compact turbulent eddy. We have not done this in the surface pressure
integral, because this tends to extend over a larger region than the Reynolds
stress ﬂuctuations responsible for it (indeed, the acoustic component of the
pressure, as opposed to the near ﬁeld hydrodynamic pressure, extends out to
inﬁnity on the wall, decaying only very slowly like 1¡[x[).
The value of the surface integral cannot be estimated by a naive orderof
magnitude calculation of the kind performed in Section 2.4 for a compact body,
because for an inﬁnite plane wall the domain of integration includes the acoustic
region, and therefore involves an unknown and possibly important contribution
from the acoustic pressure that we are trying to calculate! The difﬁculty was
resolved by Powell (1960) by the ingenious device of applying Curle’s solution
Problems 2 39
(2.4.1) at the image ¯ x = (x
1
. −x
2
. x
3
) in the wall of the far ﬁeld observation
point x. At the image point H( f ) = H(x
2
) ≡ 0, and therefore
0 ≈
¯ x
i
¯ x
j
4πc
2
0
[x[
3
∂
2
∂t
2
T
i j
y. t −
[x[
c
0
d
3
y
−
x
2
4πc
0
[x[
2
∂
∂t
y
2
=0
( p − p
0
)
y. t −
[x[
c
0
÷
x
1
y
1
÷ x
3
y
3
c
0
[x[
dy
1
dy
3
.
[x[ →∞. (2.5.2)
The surface integral term in this formula is equal but opposite in sign to that
in the original solution (2.5.1), which is now seen to exactly represent the
quadrupole sound generated by a system of image quadrupoles in the wall!
Adding (2.5.1) and (2.5.2), we ﬁnd
p(x. t ) ≈
(x
i
x
j
÷ ¯ x
i
¯ x
j
)
4πc
2
0
[x[
3
∂
2
∂t
2
T
i j
y. t −
[x[
c
0
d
3
y
≈
(x
i
x
j
÷ ¯ x
i
¯ x
j
)
4πc
2
0
[x[
3
∂
2
∂t
2
ρ
0
:
i
:
j
y. t −
[x[
c
0
d
3
y. [x[ →∞. (2.5.3)
Therefore, the apparently strong contribution from the surface pressure di
poles actually integrates to a termof quadrupole strength. This is a consequence
of the Kraichnan–Phillips theorem (see Howe, 1998a), according to which the
net unsteady component of the normal force between an inﬁnite plane wall and
an incompressible ﬂuid must vanish identically
y
2
=0
( p − p
0
)(y. t ) dy
1
dy
3
≡ 0. (2.5.4)
Thus, extreme care must be exercised when using Curle’s equation to estimate
the sound produced by turbulence interacting with large surfaces. As a general
rule, the surface contribution will be comparable to that from the turbulence
quadrupoles whenever the characteristic wavelength of the sound is smaller
than the radius of curvature of the surface.
Problems 2
1. Show that the acoustic efﬁciency of a compact sphere of radius R executing
small amplitude translational oscillations at velocity U = U
0
sin(ωt ) is
H
a
H
0
∼
ωR
c
0
3
.
40 2 Lighthill’s Theory
where
H
a
=
πωR
3
6
ρ
0
U
2
0
ωR
c
0
3
. H
0
=
2ωR
3
3
ρ
0
U
2
0
are respectively the average acoustic and hydrodynamic powers fed into the
ﬂuid during the quarter cycle 0  ωt  π¡2.
Explain the signiﬁcance of averaging only over 0  ωt  π¡2.
2. What is the efﬁciency in Problem 1 when the sphere pulsates with small
amplitude normal velocity :
n
= U
0
sin(ωt )?
3. The wake behind a bluff body ﬁxed in a nominally steady, lowMach number
ﬂow at speed U produces a drag force equal to C
D
A
1
2
ρ
0
U
2
, where C
D
is the drag coefﬁcient (which may be regarded as constant), and A is the
projected crosssectional area of the body in the ﬂow direction. Derive an
approximate formula for the farﬁeld acoustic pressure radiated by the body
when U contains a small amplitude, timeharmonic component such that
U = U
0
÷u cos(ωt ), where U
0
and u are constant and u <U
0
.
4. Show that Powell’s solution (2.5.3) for the sound generated by turbulence
adjacent to a rigid plane wall is identical with the solution of Curle’s differ
ential equation (2.3.8) determined by the modiﬁed Green’s function
G(x. y. t −τ) =
1
4π[x −y[
δ
t −τ −
[x −y[
c
0
÷
1
4π[¯ x −y[
δ
t −τ −
[¯ x −y[
c
0
.
where ¯ x = (x
1
. −x
2
. x
3
). G(x. y. t − τ) is the solution of what problem of
linear acoustics?
3
The Compact Green’s Function
3.1 The Inﬂuence of Solid Boundaries
Let us return to the general problem of linear acoustics. To ﬁx ideas, we shall
frame the present discussion in terms of Equation (1.3.9)
1
c
2
0
∂
2
∂t
2
−∇
2
ϕ = −q(x. t ) (3.1.1)
for the velocity potential, but our conclusions will be applicable quite generally.
This equation determines ϕ in terms of a speciﬁed source distribution q(x. t ). In
the absence of solid boundaries (in free space) the results of Section 1.6 enable
us to represent ϕ in the form
ϕ(x. t ) =
∞
−∞
−q(y. τ)G(x. y. t −τ) d
3
y dτ. (3.1.2)
where G(x. y. t −τ) is the free space Green’s function
G(x. y. t −τ) =
1
4π[x −y[
δ
t −τ −
[x −y[
c
0
. (3.1.3)
that is, G(x. y. t −τ) is the outgoing wave solution of
1
c
2
0
∂
2
∂t
2
−∇
2
G = δ(x−y)δ(t −τ). where G = 0 for t  τ. (3.1.4)
In our discussion of Curle’s extension of Lighthill’s theory in Chapter 2, it
was found that the presence of a solid boundary S in the vicinity of the tur
bulence quadrupole sources T
i j
resulted in the appearance of additional dipole
and monopole sources distributed over S. represented by the second and third
terms on the righthand side of Equation (2.3.8). Curle’s solution (2.3.9) of this
41
42 3 The Compact Green’s Function
Fig. 3.1.1.
equation was derived by using the retarded potential formula (1.6.6), obtained
by use of the free space Green’s function (3.1.3). A similar representation in
volving surface distributions of dipoles and monopoles is obtained for ϕ when
we attempt to solve Equation (3.1.1) using Green’s function (3.1.3) in the sit
uation illustrated in Fig. 3.1.1, where the source distribution q(x. t ) is adjacent
to a rigid boundary S on which the normal derivative ∂ϕ¡∂x
n
= 0.
It would be very convenient if we could modify the functional form of
G(x. y. t − τ) so that it automatically takes account of the contributions from
the dipole and monopole sources on S, inasmuch that no surface integrals occur
in the ﬁnal solution. To do this, we must ﬁnd a solution of Green’s function
equation (3.1.4) that satisﬁes appropriate boundary conditions on S. The so
lution ϕ of (3.1.1) is then once again given by Formula (3.1.2) in terms of
the modiﬁed Green’s function, there being no additional surface integrals to
evaluate.
The main practical difﬁculty is the calculation of the modiﬁed Green’s func
tion. Although it is always possible in principle, exact analytical representations
are known only for solid bodies of very simple shapes (such as spheres, circular
cylinders, and halfplanes). However, it turns out that a relatively simple and
general approximate formula can be found for the modiﬁed Green’s function
for those problems where it is known that the typical wavelength of the sound
produced by the source distribution q(x. t ) is large compared to one or more
principal dimensions of the solid body S. This is called the compact Green’s
function.
3.1 The Inﬂuence of Solid Boundaries 43
To simplify the calculation of the compact Green’s function, we use the
Fourier integral formula
δ(t −τ) =
1
2π
∞
−∞
e
−i ω(t −τ)
dω. (3.1.5)
which expresses the δ function as a linear combination of timeharmonic os
cillations of frequency ω. The formula is proved by observing that no real
system can oscillate at inﬁnitely large frequencies, and therefore in all practical
problems e
−i ω(t −τ)
can be replaced by e
−i ω(t −τ)−c[ω[
for arbitrarily small c >0.
Then,
1
2π
∞
−∞
e
−i ω(t −τ)
dω ≡ lim
c→÷0
1
2π
∞
−∞
e
−i ω(t −τ)−c[ω[
dω
= lim
c→÷0
c
π[c
2
÷(t −τ)
2
]
.
The ﬁnal term on the right is the usual deﬁnition of δ(t −τ) as the limit of an
‘csequence’ (Lighthill, 1958).
If we now put
G(x. y. t −τ) =
−1
2π
∞
−∞
ˆ
G(x. y. ω)e
−i ω(t −τ)
dω. (3.1.6)
then the substitution of this and (3.1.5) into the Green’s function equation (3.1.4)
shows that, for each frequency ω,
ˆ
G(x. y. ω) is the solution of
∇
2
÷κ
2
0
ˆ
G(x. y. ω) = δ(x −y). (3.1.7)
where κ
0
= ω¡c
0
is called the acoustic wave number. Sound of frequency ω
has wavelength
λ =
2π
κ
0
.
Thus, a solid body of characteristic dimension ¹ is compact for waves of
frequency ω provided that
¹
λ
=
κ
0
¹
2π
<1. (3.1.8)
This condition will be used below in Section 3.4 to calculate the compact
Green’s function.
44 3 The Compact Green’s Function
3.2 The Helmholtz Equation
The equations
∇
2
÷κ
2
0
ˆ ϕ = 0.
∇
2
÷κ
2
0
ˆ ϕ = ˆ q(x. ω) (3.2.1)
are known respectively as the Helmholtz equation and the inhomogeneous
Helmholtz equation. The source term ˆ q(x. ω) represents one frequency com
ponent of the source q(x. t ) of Equation (3.1.1), so that
q(x. t ) =
∞
−∞
ˆ q(x. ω)e
−i ωt
dω. (3.2.2)
Therefore, because (differentiating under the integral sign)
1
c
2
0
∂
2
∂t
2
∞
−∞
ˆ ϕ(x. ω)e
−i ωt
dω=
1
c
2
0
∞
−∞
−ω
2
ˆ ϕ(x. ω)e
−i ωt
dω
≡−
∞
−∞
κ
2
0
ˆ ϕ(x. ω)e
−i ωt
dω.
the solution ˆ ϕ(x. ω) of the inhomogeneous equation is related to the solution of
(3.1.1) by
ϕ(x. t ) =
∞
−∞
ˆ ϕ(x. ω)e
−i ωt
dω. (3.2.3)
3.2.1 The Point Source
Equation (3.1.7) determines Green’s function
ˆ
G(x. y. ω) for the inhomoge
neous Helmholtz equation. The free space Green’s function can be found by
the method used in Section 1.5 for the wave equation. If we temporarily set
y = 0, we have to ﬁnd the radially symmetric solution of
∇
2
÷κ
2
0
ˆ
G = δ(x). (3.2.4)
In the usual way (see Section 1.5), we have
∂
2
∂r
2
(r
ˆ
G) ÷κ
2
0
(r
ˆ
G) = 0 for r = [x[ > 0.
and therefore
ˆ
G =
Ae
i κ
0
r
r
÷
Be
−i κ
0
r
r
. (3.2.5)
where A. B remain to be determined.
3.2 The Helmholtz Equation 45
To do this recall that our solution represents one component of a time
dependent acoustic problem of frequency ω. Since the time factor is e
−i ωt
,
the two terms on the righthand side of (3.2.5) correspond to propagating sound
waves of the form
Ae
−i ω
t −
r
c
0
r
÷
Be
−i ω(t ÷
r
c
0
)
r
.
the second of which represents waves converging on the source from inﬁnity,
and must therefore be rejectedbecause of the radiationcondition. Hence, B = 0.
The value of the remaining constant A is found by extending the solution to
include the region occupied by the source at r = 0 by writing
ˆ
G = lim
c→0
Ae
i κ
0
r
(r
2
÷c
2
)
1
2
.
By substituting the solution into (3.2.4) and using the Formula (1.4.8) we ﬁnd
A = −1¡4π.
The free space Green’s function for the inhomogeneous Helmholtz equation
(the solution of (3.1.7)) is now obtained by replacing r = [x[ by [x −y[
ˆ
G(x. y. ω) =
−e
i κ
0
[x−y[
4π[x −y[
. (3.2.6)
Because the source ˆ q(x. ω) in the second of Equations (3.2.1) can be ex
pressed as a superposition of point sources by means of
ˆ q(x. ω) =
∞
−∞
ˆ q(y. ω)δ(x −y) d
3
y.
the solution of the inhomogeneous Helmholtz equation in an unbounded me
dium can be written
ϕ(x. ω) =
∞
−∞
ˆ
G(x. y. ω) ˆ q(y. ω) d
3
y ≡
−1
4π
∞
−∞
ˆ q(y. ω)e
i κ
0
[x−y[
[x −y[
d
3
y.
(3.2.7)
3.2.2 Dipole and Quadrupole Sources
The method of integration by parts described in Section 1.7 can be used to show
that the corresponding solutions for the dipole and quadrupole sources
ˆ q(x. ω) =
∂ f
j
∂x
j
(x. ω) and ˆ q(x. ω) =
∂
2
T
i j
∂x
i
∂x
j
(x. ω) (3.2.8)
46 3 The Compact Green’s Function
are respectively
ˆ ϕ(x. ω) =
−1
4π
∂
∂x
j
∞
−∞
f
j
(y. ω)e
i κ
0
[x−y[
[x −y[
d
3
y.
and
ˆ ϕ(x. ω) =
−1
4π
∂
2
∂x
i
∂x
j
∞
−∞
T
i j
(y. ω)e
i κ
0
[x−y[
[x −y[
d
3
y. (3.2.9)
Example For a point dipole at the origin orientated in the x
1
direction
ˆ q(x. ω) =
∂
∂x
1
( f
1
δ(x)).
Therefore,
ˆ ϕ(x. ω) =
−1
4π
∂
∂x
1
∞
−∞
f
1
δ(y)e
i κ
0
[x−y[
[x −y[
d
3
y ≈
−i κ
0
x
1
f
1
e
i κ
0
[x[
4π[x[
2
. [x[ →∞.
(3.2.10)
3.2.3 Green’s Function for the Wave Equation
Let us verify the general relation (3.1.6) between the Green’s functions G(x. y.
t −τ) and
ˆ
G(x. y. ω), respectively, for the wave equationandthe inhomogeneous
Helmholtz equation in the special case in which there are no solid boundaries.
According to this formula we ﬁnd, using the expression (3.2.6) for the free
space Green’s function
ˆ
G,
G(x. y. t −τ) =
−1
2π
∞
−∞
ˆ
G(x. y. ω)e
−i ω(t −τ)
dω
=
1
8π
2
[x −y[
∞
−∞
e
−i ω(t −τ−
[x−y[
c
0
)
dω
=
1
4π[x −y[
δ
t −τ −
[x −y[
c
0
.
which is precisely Equation (3.1.3).
3.3 The Reciprocal Theorem
The calculation of the compact Green’s function is greatly simpliﬁed by appli
cation of the reciprocal theorem. We need to consider only a special case of
3.3 The Reciprocal Theorem 47
Fig. 3.3.1.
this very general theorem of mechanics, which was ﬁrst used with great effect
in acoustics by Lord Rayleigh (1945).
Consider the two acoustic problems indicated in Fig. 3.3.1, in which sound
of frequency ω is generated by two unit point sources at x = x
A
and x = x
B
in
the presence of a solid body S. We denote the functional forms of the respective
velocity potentials generated by these sources by
ˆ
G(x. x
A
. ω) and
ˆ
G(x. x
B
. ω),
where
∇
2
÷κ
2
0
ˆ
G(x. x
A
. ω) = δ(x −x
A
). (3.3.1)
∇
2
÷κ
2
0
ˆ
G(x. x
B
. ω) = δ(x −x
B
). (3.3.2)
In addition
ˆ
G(x. x
A
. ω) and
ˆ
G(x. x
B
. ω) must satisfy appropriate mechani
cal boundary conditions on S. We take these to have the same general linear
form
∂
ˆ
G
∂x
n
(x. x
A
. ω) =
ˆ
G(x. x
A
. ω)
Z(x. ω)
.
∂
ˆ
G
∂x
n
(x. x
B
. ω) =
ˆ
G(x. x
B
. ω)
Z(x. ω)
.
for x on S. (3.3.3)
48 3 The Compact Green’s Function
where x
n
is measured in the normal direction from S into the ﬂuid and Z(x. ω)
is the surface impedance. For a rigid surface, Z(x. ω) = ∞.
At large distances from S, in the acoustic far ﬁeld, both solutions are assumed
to exhibit the characteristics of outgoing sound waves, such that (with implicit
time dependence e
−i ωt
)
ˆ
G(x. x
A
. ω) ∼
f
A
(θ. φ)e
i κ
0
r
r
.
ˆ
G(x. x
B
. ω) ∼
f
B
(θ. φ)e
i κ
0
r
r
.
r ≡ [x[ →∞. (3.3.4)
where it may be supposed that the coordinate origin is in the neighborhood of
S. The angular dependencies of the farﬁeld radiations from the two sources
are determined by the factors f
A
(θ. φ) and f
B
(θ. φ), which are functions of the
polar angles θ. φ deﬁning the orientation of the far ﬁeld point x, and generally
depend strongly on the details of the interaction of the volume ﬂows from each
source with S.
The reciprocal theorem states that
ˆ
G(x
A
. x
B
. ω) =
ˆ
G(x
B
. x
A
. ω). (3.3.5)
That is, the potential at x
A
produced by the point source at x
B
is equal to the
potential at x
B
produced by an equal point source at x
A
.
Proof. Multiply Equation (3.3.1) by
ˆ
G(x. x
B
. ω) and Equation (3.3.2) by
ˆ
G(x. x
A
. ω), subtract the resulting equations and integrate over the volume
bounded by the surface S and by a large surface Y in the acoustic far ﬁeld.
Green’s identity
ˆ
G(x. x
B
. ω)∇
2
ˆ
G(x. x
A
. ω) −
ˆ
G(x. x
A
. ω)∇
2
ˆ
G(x. x
B
. ω)
= div(
ˆ
G(x. x
B
. ω)∇
ˆ
G(x. x
A
. ω) −
ˆ
G(x. x
A
. ω)∇
ˆ
G(x. x
B
. ω)).
and the divergence theorem permit the volume integral of the term obtained
from the lefthand sides to be expressed as surface integrals over S and Y,
whereas the integrals involving the δ functions can be evaluated explicitly. This
procedure gives
S÷Y
ˆ
G(x. x
A
. ω)
∂
ˆ
G
∂x
n
(x. x
B
. ω) −
ˆ
G(x. x
B
. ω)
∂
ˆ
G
∂x
n
(x. x
A
. ω)
dS
=
ˆ
G(x
B
. x
A
. ω) −
ˆ
G(x
A
. x
B
. ω).
3.4 TimeHarmonic Compact Green’s Function 49
The surface integral over S vanishes because of the impedance conditions
(3.3.3). The surface integral over Y vanishes because of conditions (3.3.4) and
because ∂θ¡∂x
n
and ∂φ¡∂x
n
are each of order 1¡r as r →∞, and therefore
∂
ˆ
G
∂x
n
(x. x
A
. ω) ∼ f
A
(θ. φ)
i κ
0
e
i κ
0
r
r
∂r
∂x
n
.
∂
ˆ
G
∂x
n
(x. x
B
. ω) ∼ f
B
(θ. φ)
i κ
0
e
i κ
0
r
r
∂r
∂x
n
. as r →∞.
This proves the theorem.
The result is usually expressed as the simple reciprocal relation
ˆ
G(x. y. ω) =
ˆ
G(y. x. ω). (3.3.6)
3.4 TimeHarmonic Compact Green’s Function
We are now ready to derive the compact Green’s function
ˆ
G(x. y. ω) for the
problem depicted in Fig. 3.4.1. We have to solve
∇
2
÷κ
2
0
ˆ
G(x. y. ω) = δ(x −y).
∂
ˆ
G
∂x
n
= 0 on S. (3.4.1)
where the rigid body S is assumed to be acoustically compact. The inﬂuence of
Fig. 3.4.1.
50 3 The Compact Green’s Function
a solid body on the production of sound by neighboring sources is equivalent to
an additional distribution of monopoles and dipoles on S. The compact Green’s
function includes a ﬁrst approximation for the net effect of these monopole and
dipole distributions, obviating the need to evaluate surface integrals.
In practice, we are interested primarily in calculating the sound in the far
ﬁeld of the body. Let ¹ denote the characteristic diameter of the body, and take
the coordinate origin at O within S. The source point y is assumed to be close
to S (so that [y[ ∼ ¹) and the observer at x is taken to be in the acoustic far
ﬁeld. The compactness condition (3.1.8) therefore implies that
κ
0
¹ <1 and κ
0
[y[ <1.
In these circumstances the compact approximation for
ˆ
G(x. y. ω) can be
found very easily from the solution of the reciprocal problem:
∂
2
∂y
2
1
÷
∂
2
∂y
2
2
÷
∂
2
∂y
2
3
÷κ
2
0
ˆ
G(y. x. ω) = δ(y −x).
∂
ˆ
G
∂y
n
= 0 on S.
(3.4.2)
where the source is at the farﬁeld point x, and
ˆ
G(y. x. ω) is determined as a
function of y close to S. The solution of (3.4.1) is then given by the reciprocal
theorem (Section 3.3)
ˆ
G(x. y. ω) =
ˆ
G(y. x. ω) (the potential
ˆ
G(x. y. ω) at the
farﬁeld point x produced by the point source at y is exactly equal to the potential
ˆ
G(y. x. ω) produced at the nearﬁeld point y by an equal point source at the
farﬁeld point x).
To solve (3.4.2), we put
ˆ
G(y. x. ω) =
ˆ
G
0
(y. x. ω) ÷
ˆ
G
/
(y. x. ω)
≡
−e
i κ
0
[x−y[
4π[x −y[
÷
ˆ
G
/
(y. x. ω)
where
ˆ
G
0
(y. x. ω) is the spherically spreading wave generated by the point
source at x when the presence of the solid is ignored. The term
ˆ
G
/
(y. x. ω)
is the velocity potential of the motion produced in the ﬂuid when this wave
impinges on S.
When [x[ →∞, the approximations
[x −y[ ≈ [x[ −
x · y
[x[
≡ [x[ −
x
j
y
j
[x[
and
1
[x −y[
≈
1
[x[
÷
x · y
[x[
3
≈
1
[x[
3.4 TimeHarmonic Compact Green’s Function 51
and the condition κ
0
y
j
∼ κ
0
¹ <1 imply that
ˆ
G
0
(y. x. ω) ≡
−e
i κ
0
[x−y[
4π[x −y[
≈
−e
i κ
0
[x[
4π[x[
e
−
i κ
0
x
j
y
j
[x[
≈
−e
i κ
0
[x[
4π[x[
1 −
i κ
0
x
j
y
j
[x[
÷O(κ
0
¹)
2
(3.4.3)
The linear dependence on y
j
in the second line of this formula represents the ﬁrst
approximation (of order κ
0
¹) in a power series expansion of rapidly decreasing
terms that describes the variation of the incident spherical wave close to the
body. Thus, regarded as a function of y, the terms shown explicitly in
ˆ
G
0
(y. x. ω) =
−e
i κ
0
[x[
4π[x[
÷
e
i κ
0
[x[
4π[x[
i κ
0
x
j
y
j
[x[
÷· · · ≡ constant ÷U
j
y
j
÷· · · .
where U
j
=
e
i κ
0
[x[
4π[x[
i κ
0
x
j
[x[
. (3.4.4)
can be regarded as the velocity potential of a uniform ﬂow at velocity U
j
im
pinging on the solid.
At distances [y[ ·¹ from S, the distortion of this ﬂow produced by the body
must be small. Let it be represented by the velocity potential
ˆ
G
/
(y. x. ω) = −U
j
ϕ
∗
j
(y) ÷O(κ
0
¹)
2
. where ϕ
∗
j
(y) →0 when [y[ ·¹.
The function ϕ
∗
j
has the dimensions of length and ∼¹ in order of magnitude
(Batchelor 1967). Then,
ˆ
G(y. x. ω) =
ˆ
G
0
(y. x. ω) ÷
ˆ
G
/
(y. x. ω) =
−e
i κ
0
[x[
4π[x[
÷U
j
(y
j
−ϕ
∗
j
(y)) ÷· · · .
(3.4.5)
where the terms shown explicitly represent a potential ﬂow past the body.
Near the body
ˆ
G(y. x. ω) satisﬁes (3.4.2) with the righthand side replaced
by zero (because the source is in the far ﬁeld). Hence,
U
j
∇
2
(y
j
−ϕ
∗
j
(y)) ÷O(κ
0
¹)
2
= 0.
But U
j
(y
j
− ϕ
∗
j
(y)) = O(κ
0
¹), and therefore, correct to the neglect of small
terms of order O(κ
0
¹)
2
,
∇
2
(y
j
−ϕ
∗
j
(y)) = 0. i.e. ∇
2
ϕ
∗
j
(y) = 0.
52 3 The Compact Green’s Function
where the rigid surface condition requires
∂
∂y
n
(y
j
−ϕ
∗
j
(y)) = 0 on S. (3.4.6)
Summarizing our conclusions from Equations (3.4.4)–(3.4.6): When x is in
the acoustic far ﬁeld, and y is close to the body
ˆ
G(x. y. ω) =
−e
i κ
0
[x[
4π[x[
1 −
i κ
0
x
j
[x[
(y
j
−ϕ
∗
j
(y)) ÷O(κ
0
¹)
2
.
y ∼ O(¹). [x[ →∞. (3.4.7)
The ﬁrst termin the large brackets represents the contribution fromthe spherical
wave
ˆ
G
0
(x. y. ω) evaluatedat y =0. The next termis O(κ
0
¹) andincludes a com
ponent −i κ
0
x
j
y
j
¡[x[ from the incident wave plus a correction i κ
0
x
j
ϕ
∗
j
(y)¡[x[
produced by S.
The vector ﬁeld
Y(y) ≡ y −ϕ
∗
(y) (3.4.8)
is called the Kirchhoff vector for the body; the j th component
Y
j
(y) ≡ y
j
−ϕ
∗
j
(y)
satisﬁes Laplace’s equation ∇
2
Y
j
= 0 with ∂Y
j
¡∂y
n
= 0 on S, and can be
interpreted as the velocity potential of an incompressible ﬂow past S that has
unit speed in the j direction at large distances from S. The function ϕ
∗
j
(y) decays
with distance from S, and satisﬁes
∂ϕ
∗
j
∂y
n
(y) = n
j
on S. (3.4.9)
because ∂y
j
¡∂y
n
≡ n
i
∂y
j
¡∂y
i
= n
i
δ
i j
= n
j
. Hence, ϕ
∗
j
(y) is just the instanta
neous velocity potential of the motion that would be produced by translational
motion of S as a rigid body at unit speed in the j direction.
Deﬁnition
ˆ
G(x. y. ω) =
−e
i κ
0
[x[
4π[x[
1 −
i κ
0
x
j
[x[
(y
j
−ϕ
∗
j
(y))
. y ∼ O(¹). [x[ →∞.
(3.4.10)
is called the compact Green’s function for source points y near the body and
observer positions x in the acoustic far ﬁeld.
3.5 Compact Green’s Function for a Rigid Sphere 53
In Section 3.7, we shall introduce a very much more elegant representation
of the compact Green’s function that greatly expands its utility.
3.5 Compact Green’s Function for a Rigid Sphere
Let the sphere have radius a and take the coordinate origin O at its center, as
illustrated in Fig. 3.5.1. We have to determine the Kirchhoff vector whose j th
component
Y
j
(y) = y
j
−ϕ
∗
j
(y) for j = 1. 2. 3
is equal to the velocity potential of incompressible ﬂow past the sphere having
unit speed in the j direction at large distances from the sphere.
Consider the case j = 1 shown in the ﬁgure. The ﬂowis evidently symmetric
about the y
1
axis. Take spherical polar coordinates (r. ϑ. φ) with ϑ measured
from the positive y
1
axis. Then, y
1
= r cos ϑ and the condition (3.4.9) to be
satisﬁed on the sphere is
∂ϕ
∗
1
∂r
= cos ϑ at r = a. (3.5.1)
The axisymmetry of the problem suggests that we look for a solution of
Laplace’s equation in the form
ϕ
∗
1
= +(r) cos ϑ.
which satisﬁes the axisymmetric form of Laplace equation
1
r
2
∂
∂r
r
2
∂
∂r
÷
1
r
2
sin ϑ
∂
∂ϑ
sin ϑ
∂
∂ϑ
+(r) cos ϑ = 0.
Fig. 3.5.1.
54 3 The Compact Green’s Function
provided that
r
2
d
2
+
dr
2
÷2r
d+
dr
−2+ = 0.
The solutions of this equation are proportional to r
n
where n is a root of the
quadratic equation
n
2
÷n −2 = 0. i.e.. n = −2. 1.
Hence,
Y
1
≡ y
1
−ϕ
∗
1
= r cos ϑ −
Ar ÷
B
r
2
cos ϑ.
where A and B are constants. The condition that ϕ
∗
1
→ 0 as r → ∞implies
that A = 0, and condition (3.5.1) supplies B = −a
3
¡2. Therefore,
Y
1
= r cos ϑ ÷
a
3
2r
2
cos ϑ ≡ y
1
1 ÷
a
3
2r
3
.
Because of the symmetry of the sphere it is clear that we also have
Y
2
= y
2
1 ÷
a
3
2r
3
. Y
3
= y
3
1 ÷
a
3
2r
3
. r = [y[.
Thus, the compact Green’s function (3.4.10) for the sphere is
ˆ
G(x. y. ω) =
−e
i κ
0
[x[
4π[x[
¸
1 −
i κ
0
x
j
y
j
[x[
1 ÷
a
3
2[y[
3
¸
. y ∼ O(a).
[x[ →∞. (3.5.2)
This represents the farﬁeld acoustic potential produced by a point source at y
close to the sphere. Because κ
0
[y[ is small the second term in the brace brackets
is always small compared to 1. This appears to suggest that, after all, the sphere
has a relatively small effect on the production of sound! This is certainly true
for monopole sources, but most sources of interest in applications are dipoles or
quadrupoles, and in these circumstances we shall see that it is the small, second
term that determines the leading order approximation for the farﬁeld sound.
3.5.1 Radiation from a Dipole Adjacent to a Compact Sphere
Let us apply the compact Green’s function (3.5.2) to determine the farﬁeld
sound generated by a dipole source close to a sphere of radius a <λ =acoustic
3.5 Compact Green’s Function for a Rigid Sphere 55
Fig. 3.5.2.
wavelength. With the origin at the center of the sphere, we consider the outgoing
wave solution of
∇
2
÷κ
2
0
ˆ ϕ = f
1
∂
∂x
1
{δ(x
1
− L)δ(x
2
)δ(x
3
)}. where
∂ ˆ ϕ
∂x
n
= 0 on [x[ = a.
The dipole is orientated in the x
1
direction and lies on the x
1
axis at (L. 0. 0),
as in Fig. 3.5.2. The solution is given by
ˆ ϕ(x. ω) =
f
1
∂
∂y
1
{δ(y
1
− L)δ(y
2
)δ(y
3
)}
ˆ
G(x. y. ω) d
3
y.
where the integration is over the ﬂuid, and ∂
ˆ
G¡∂x
n
= 0 on the sphere. The
source term is zero everywhere except at (L. 0. 0). To evaluate the integral we
write
ˆ ϕ(x. ω) = f
1
∂
∂y
1
{
ˆ
G(x. y. ω)δ(y
1
− L)δ(y
2
)δ(y
3
)} d
3
y
−f
1
δ(y
1
− L)δ(y
2
)δ(y
3
)
∂
ˆ
G
∂y
1
(x. y. ω) d
3
y.
The ﬁrst integral is zero because δ(y
1
−L) = 0 on the boundaries of the region
of integration, so that
ˆ ϕ(x. ω) = −f
1
∂
ˆ
G
∂y
1
(x. y. ω)
y=(L.0.0)
. (3.5.3)
Thus far the calculation is exact. To determine the solution in the far ﬁeld
given that the sphere is acoustically compact we use the compact approxima
tion (3.5.2) for
ˆ
G(x. y. ω). We see immediately that the differentiation with
respect to y
1
will be applied only to the small second term in the braces
56 3 The Compact Green’s Function
of (3.5.2), giving
ˆ ϕ(x. ω) =
−i κ
0
f
1
x
j
e
i κ
0
[x[
4π[x[
2
¸
∂
∂y
1
¸
y
j
1 ÷
a
3
2[y[
3
¸¸
y=(L.0.0)
=
−i κ
0
f
1
x
1
e
i κ
0
[x[
4π[x[
2
1 −
a
3
L
3
=
−i κ
0
f
1
cos θ e
i κ
0
[x[
4π[x[
1 −
a
3
L
3
. [x[ →∞. κ
0
a <1.
where θ is the angle between the x
1
axis and the x direction (so that x
1
=
[x[ cos θ).
By setting a = 0 in this formula, we recover the far ﬁeld (3.2.10) of a dipole
source in the absence of the sphere. The presence of the sphere accordingly
reduces the amplitude of the sound relative to that produced by a freeﬁeld
dipole. The amplitude is zero when L → a, because in this limit the surface
of the sphere is effectively plane in the vicinity of the dipole and an equal and
opposite image dipole is formed in the sphere. The net radiation is therefore
equivalent to that produced by a quadrupole source, and to calculate the sound
in this case it would be necessary to use a more accurate approximation to
ˆ
G(x. y. ω). This conclusion applies only to dipoles orientated radially with
respect to the sphere (see Problem 1), but it is also true for any compact rigid
surface when a dipole orientated in the direction of the local surface normal
approaches the surface.
3.5.2 Sound Produced by a Vibrating Sphere
Let the surface S of a ﬁxed body execute small amplitude vibrations with normal
velocity :
n
(x. ω) (Fig. 3.5.3). The ﬂuid motion is the same as that generated by
a distribution of monopoles of strength :
n
(x. ω) per unit area of S when S is
Fig. 3.5.3.
3.5 Compact Green’s Function for a Rigid Sphere 57
assumed to be stationary (rigid). The corresponding source strength ˆ q(x. ω) in
the inhomogeneous Helmholtz equation (3.2.1) is
ˆ q(x. ω) = :
n
(x. ω)δ(s
⊥
−c). (c →÷0).
where s
⊥
is distance measured in the normal direction from S into the ﬂuid,
and c > 0 places the sources just within the ﬂuid adjacent to S. The velocity
potential ˆ ϕ(x. ω) is therefore
ˆ ϕ(x. ω) =
ﬂuid
:
n
(y. ω)δ(s
⊥
−c)
ˆ
G(x. y. ω) d
3
y (c →÷0)
=
S
:
n
(y. ω)
ˆ
G(x. y. ω) dS(y) where
∂
ˆ
G
∂x
n
(x. y. ω) = 0 on S.
(3.5.4)
Consider the sound produced when the sphere of Fig. 3.5.1 vibrates with
small amplitude about its undisturbed position centred at the origin with velocity
ˆ
U(ω)e
−i ωt
along the x
1
axis. Then,
:
n
(y. ω) =
ˆ
U(ω) cos ϑ.
If the vibrations are at sufﬁciently low frequency, the sphere will be compact,
and when the observer at x is in the acoustic far ﬁeld the integral in (3.5.4) can
be evaluated using the compact approximation (3.5.2) for
ˆ
G(x. y. ω):
ˆ ϕ(x. ω)
≈
−e
i κ
0
[x[
4π[x[
¸
S
:
n
(y. ω) dS(y) −
i κ
0
x
j
[x[
S
y
j
1 ÷
a
3
2[y[
3
:
n
(y. ω) dS(y)
¸
.
The ﬁrst integral represents the net volume ﬂux through S and vanishes identi
cally for rigid body translational motion. The second integral is nonzero only
for j = 1, when y
1
= a cos ϑ and [y[ = a on S, and we can take dS = 2πa
2
sin ϑ dϑ (so that the surface integral becomes 3πa
3
ˆ
U(ω)
π
0
cos
2
ϑ sin ϑ dϑ =
2πa
3
ˆ
U(ω)). Hence,
ˆ ϕ(x. ω) ≈
i κ
0
ˆ
U(ω)a
3
x
1
e
i κ
0
[x[
2[x[
2
≡
i ω
ˆ
U(ω)a
3
cos θe
i κ
0
[x[
2c
0
[x[
. [x[ →∞.
where θ is the angle between the x
1
axis and the radiation direction x (see
Fig. 1.7.1).
The solution for a sphere oscillating at an arbitrary time dependent velocity
U(t ) can be derived from this result provided the sphere remains compact. This
58 3 The Compact Green’s Function
means that if we write
U(t ) =
∞
−∞
ˆ
U(ω)e
−i ωt
dω.
then
ˆ
U(ω) ,= 0 only for κ
0
a < 1. If this condition is satisﬁed we can use the
Formulae (3.2.2) and (3.2.3) to obtain the timedependent velocity potential in
the form
ϕ(x. t ) ≈
a
3
cos θ
2c
0
[x[
∞
−∞
i ω
ˆ
U(ω)e
−i ω(t −[x[¡c
0
)
dω
=
−a
3
cos θ
2c
0
[x[
∂
∂t
∞
−∞
ˆ
U(ω)e
−i ω(t −[x[¡c
0
)
dω
=
−a
3
cos θ
2c
0
[x[
∂U
∂t
(t −[x[¡c
0
). [x[ →∞.
This agrees with the farﬁeld result obtained in Section 1.7, and conﬁrms the
model used there in which the vibrating sphere was replaced by a point dipole
of strength 2πa
3
U(t ) at its center.
3.6 Compact Green’s Function for Cylindrical Bodies
The reciprocal calculation of the Green’s function described in Section 3.4 for
the compact body in Fig. 3.4.1 can be immediately extended to the case of a
cylindrical body of compact cross section.
Figure 3.6.1 illustrates the situation for an inﬁnite circular cylinder of radius
a whose axis lies along the y
3
axis, and whose diameter 2a ∼ ¹ is acoustically
compact. The source point y is adjacent to the cylinder and for the moment
(see Section 3.7) is assumed to be within an axial distance [y
3
[ < λ from
the coordinate origin O. In this region the Expansion (3.4.7) remains valid
with ϕ
∗
3
(y) ≡0, because the impinging ﬂow described by the velocity potential
(3.4.4) for j = 3 is unaffected by the cylinder. Hence, we can take
ˆ
G(x. y. ω) =
−e
i κ
0
[x[
4π[x[
1 −
i κ
0
x
j
Y
j
[x[
. y ∼ O(¹). [x[ →∞. (3.6.1)
where the Kirchhoff vector Y has the components
Y
1
= y
1
−ϕ
∗
1
(y). Y
2
= y
2
−ϕ
∗
2
(y). Y
3
= y
3
. (3.6.2)
3.6.1 Circular Cylinder
The potentials ϕ
∗
1
(y). ϕ
∗
2
(y) for the circular cylinder of radius a can be found
by the method of Section 3.5.
3.6 Compact Green’s Function for Cylindrical Bodies 59
Fig. 3.6.1.
For j =1 the ﬂow is symmetric about the y
1
axis and is independent of
the spanwise coordinate y
3
(Fig. 3.6.2). Using polar coordinates (y
1
. y
2
) =
r(cos ϑ. sin ϑ), the condition (3.4.9) to be satisﬁed on the cylinder is
∂ϕ
∗
1
∂r
= cos ϑ at r = a. (3.6.3)
Fig. 3.6.2.
60 3 The Compact Green’s Function
As in the case of the sphere, we try a solution of the form
ϕ
∗
1
= +(r) cos ϑ.
which satisﬁes the polar form of Laplace’s equation
1
r
∂
∂r
r
∂
∂r
÷
1
r
2
∂
2
∂ϑ
2
+(r) cos ϑ = 0.
provided that
r
2
d
2
+
dr
2
÷r
d+
dr
−+ = 0.
The general solution is + = Ar ÷ B¡r. The component Ar must be rejected
because it does not decay as r →∞. Therefore,
Y
1
≡ y
1
−ϕ
∗
1
= r cos ϑ −
B
r
cos ϑ.
and condition (3.6.3) yields B = −a
2
. Therefore,
Y
1
= r cos ϑ ÷
a
2
r
cos ϑ ≡ y
1
1 ÷
a
2
r
2
.
Similarly,
Y
2
= y
2
1 ÷
a
2
r
2
.
Hence, the compact Green’s function for a circular cylinder, with source
near the origin, is
ˆ
G(x. y. ω) =
−e
i κ
0
[x[
4π[x[
1 −
i κ
0
x
j
Y
j
[x[
. y ∼ O(¹). [x[ →∞. (3.6.4)
where
Y
j
= y
j
1 ÷
a
2
y
2
1
÷ y
2
2
. j = 1. 2; Y
3
= y
3
. (3.6.5)
3.6.2 Rigid Strip
The rigid strip of chord 2a and inﬁnite span provides a simple model of a
sharpedged airfoil. In Fig. 3.6.3 the airfoil occupies −a  y
1
 a. y
2
= 0.
−∞ y
3
 ∞. The airfoil has no inﬂuence on a uniform mean ﬂow in the y
1

direction, nor on one in the y
3
direction, so that potential functions ϕ
∗
1
(y) ≡ 0
and ϕ
∗
3
(y) ≡ 0.
3.6 Compact Green’s Function for Cylindrical Bodies 61
Fig. 3.6.3.
The potential ϕ
∗
2
(y) ≡ ϕ
∗
2
(y
1
. y
2
) can be determined by the method of con
formal transformation. (Readers unfamiliar with this procedure should consult
Section 4.5.) If z = y
1
÷ i y
2
, the cross section of the airfoil in the z plane is
mapped onto the circular cylinder [Z[ = a in the Z plane by the transformation
Z = z ÷
z
2
−a
2
.
Because Z ∼ 2z as [z[ → ∞a uniform ﬂow at unit speed in the y
2
direction
in the z plane at large distances from the airfoil corresponds to a uniform
ﬂow at speed
1
2
in the direction of the imaginary Z axis at large distances
from the cylinder. This ﬂow can be found by the method discussed above
for the circular cylinder (or see Example 3 of Section 4.5), and determines
Y
2
= y
2
−ϕ
∗
2
(y
1
. y
2
) = Re[n(z)], where n is the complex potential
n(z) = −
i
2
Z −
a
2
Z
= −
i
2
z ÷
z
2
−a
2
−
a
2
z ÷
√
z
2
−a
2
= −i
z
2
−a
2
.
Thus, the compact Green’s function for a strip, with source near the origin,
is given by
ˆ
G(x. y. ω) =
−e
i κ
0
[x[
4π[x[
1 −
i κ
0
x
j
Y
j
[x[
. y ∼ O(a). [x[ →∞. (3.6.6)
where the components of the Kirchhoff vector are
Y
1
= y
1
. Y
2
= Re(−i
z
2
−a
2
). Y
3
= y
3
. z = y
1
÷i y
2
. (3.6.7)
62 3 The Compact Green’s Function
Fig. 3.6.4.
Figure 3.6.4 depicts the streamline pattern of the ﬂow past the strip deﬁned
by the velocity potential Y
2
(y). The streamlines crowd together and change very
rapidly near the sharp edges. This is an indication that edges can be important
sources of noise when located in the near ﬁeld of a dipole or quadrupole (or
any higher order multipole) source, because ∇Y
2
becomes very large there.
Example Calculate the farﬁeld velocity potential when
∇
2
÷κ
2
0
ˆ ϕ = f
2
∂
∂x
2
[δ(x
1
− L)δ(x
2
)δ(x
3
)] . L > a. κ
0
L <1.
where
∂ ˆ ϕ
∂x
2
= 0 on the airfoil −a  x
1
 a. x
2
= 0. −∞ x
3
 ∞.
The dipole source is orientated in the x
2
direction and is positioned just to the
right of the edge at y
1
= a in Fig. 3.6.3. The solution is given by the following
form of Equation (3.5.3)
ˆ ϕ(x. ω) = −f
2
∂
ˆ
G
∂y
2
(x. y. ω)
y=(L.0.0)
≈ −
i f
2
κ
0
x
2
e
i κ
0
[x[
4π[x[
2
∂Y
2
∂y
2
y=(L.0.0)
where, from (3.6.7),
∂Y
2
∂y
2
= Re
−i
∂
∂y
2
z
2
−a
2
. z = y
1
÷i y
2
= Re
z
√
z
2
−a
2
.
3.7 Symmetric Compact Green’s Function 63
Therefore,
ˆ ϕ(x. ω) ≈ −
i f
2
κ
0
x
2
e
i κ
0
[x[
4π[x[
2
L
√
L
2
−a
2
= −
i f
2
κ
0
L cos Oe
i κ
0
[x[
4π[x[
√
L
2
−a
2
. [x[ →∞.
where O = cos
−1
(x
2
¡[x[) is the angle between the normal to the strip and the
radiation direction (x¡[x[) indicated in Fig. 3.6.3.
The amplitude of the sound is increased by a factor L¡
√
L
2
−a
2
relative to
that produced by the same dipole in free space, and is unbounded as L →a,
when the dipole approaches the edge.
3.7 Symmetric Compact Green’s Function
The deﬁnition (3.4.10) of the compact Green’s function can be recast to exhibit
the reciprocal nature of the source and observer positions y and x. To do this
we ﬁrst observe that, for a body of characteristic diameter ¹,
Y
j
(y) = y
j
−ϕ
∗
j
(y) ∼ O(¹).
and, therefore, that κ
0
Y
j
<1. Hence,
ˆ
G(x. y. ω) ≈
−e
i κ
0
[x[
4π[x[
1 −
i κ
0
x
j
Y
j
[x[
=
−e
i κ
0
[x[
4π[x[
1 −
i κ
0
x · Y
[x[
≈
−1
4π[x[
e
i κ
0
[x[−
i κ
0
x·Y
[x[
.
≈
−e
i κ
0
[x−Y[
4π[x[
. Y ∼ O(¹). [x[ →∞. (3.7.1)
where on the last line we have used the usual farﬁeld approximation (1.9.2)
[x −Y[ ≈ [x[ −
x · Y
[x[
. [x[ →∞.
Now let X(x) denote the Kirchhoff vector for the body expressed in terms of x,
i.e., let
X
j
(x) = x
j
−ϕ
∗
j
(x). (3.7.2)
64 3 The Compact Green’s Function
Then, because ϕ
∗
j
(x) → 0 as [x[ → ∞, we also have [X[ ∼ [x[ as [x[ → ∞,
and, therefore, from (1.9.2) and (1.9.3)
[X −Y[ ≈ [x[ −
x · Y
[x[
1
[X −Y[
≈
1
[x[
÷
x · Y
[x[
3
≈
1
[x[
¸
¸
¸
¸
¸
¸
¸
when [x[ →∞. (3.7.3)
Thus, to the same approximation, (3.7.1) can be written
ˆ
G(x. y. ω) ≈
−e
i κ
0
[X−Y[
4π[X −Y[
. Y ∼ O(¹). [x[ →∞.
This result is the basis of our revised deﬁnition of the
Compact Green’s Function for the Inhomogeneous Helmholtz Equation
ˆ
G(x. y. ω) =
−e
i κ
0
[X−Y[
4π[X −Y[
. (3.7.4)
where X = x −ϕ
∗
(x). Y = y −ϕ
∗
(y) are the Kirchhoff vectors for the body
expressed respectively in terms of x and y. The components X
j
and Y
j
are the
velocity potential of incompressible ﬂow past the body having unit speed in
the j direction at large distances from the body; ϕ
∗
j
is the velocity potential of
the incompressible ﬂow that would be produced by rigid body motion of S at
unit speed in the j direction.
Our generalized deﬁnition clearly satisﬁes the reciprocal theorem. Also, be
cause of the symmetrical way in which x and y enter this formula we may now
remove any restriction on the position of the coordinate origin. The approxima
tion is valid for arbitrary source and observer locations provided that at least
one of them lies in the far ﬁeld of the body. When both x and y are in the far
ﬁeld (so that X ∼ x and Y ∼ y) predictions made with the compact Green’s
function will be the same as when the body is absent. This is because for dis
tant sources the amplitude of the sound scattered by a compact rigid object is
O((κ
0
¹)
2
) smaller than the incident sound, that is, is of quadrupole intensity
(Lighthill 1978; Howe 1998a). When x is close to the body the source must be
in the far ﬁeld;
ˆ
G(x. y. ω) then determines the modiﬁcation by the body of low
frequency sound received by an observer near the body.
The deﬁnition (3.7.4) is easily recalled because it is an obvious generaliza
tion of the free space Green’s function (3.2.6). In applications it is necessary
to remember also that it is valid for determining only the leading order ap
proximation to the surface monopole and dipole sources induced on the body
3.8 LowFrequency Radiation from a Vibrating Body 65
by neighboring sources in the ﬂuid. In practice this means that when used in
calculations
ˆ
G(x. y. ω) will normally be expanded only to ﬁrst order in the
Kirchhoff source vector Y(y).
We can go further and use the formula (3.1.6) relating Green’s functions for
the wave equation and the Helmholtz equation to derive the compact approxi
mation for Green’s function of the wave equation:
G(x. y. t −τ) =
−1
2π
∞
−∞
ˆ
G(x. y. ω)e
−i ω(t −τ)
dω
≈
1
8π
2
[X −Y[
∞
−∞
e
−i ω(t −τ−
[X−Y[
c
0
)
dω
=
1
4π[X −Y[
δ
t −τ −
[X −Y[
c
0
.
This remarkable result is formally identical with the classical free space Green’s
function (1.6.2) with X. Y substituted for x. y. However, its use is subject
to the same restrictions as (3.7.4), and it will be valid only when applied to
timedependent source terms producing sound whose wavelength is large com
pared to the characteristic body dimension ¹. With this understanding we can
deﬁne the
Compact Green’s Function for the Wave Equation
G(x. y. t −τ) =
1
4π[X −Y[
δ
t −τ −
[X −Y[
c
0
. (3.7.5)
where X = x −ϕ
∗
(x). Y = y −ϕ
∗
(y) are Kirchhoff vectors for the body. The
components X
j
and Y
j
are the velocity potential of incompressible ﬂow past
the body having unit speed in the j direction at large distances from the body;
ϕ
∗
j
is the velocity potential of the incompressible ﬂow that would be produced
by rigid body motion of S at unit speed in the j direction.
3.8 LowFrequency Radiation from a Vibrating Body
The compact Green’s function (3.7.5) for the wave equation will now be used
to give a complete theory of the lowfrequency sound produced by a vibrating
body. The maximum frequency of the vibrations must be small enough to
ensure that the body (or its cross section, in the case of vibrating a cylinder)
is acoustically compact. The argument follows closely the discussion of the
vibrating sphere in Section 3.5, except that we now work directly with time
dependent quantities.
66 3 The Compact Green’s Function
Fig. 3.8.1.
Let the closed surface S (Fig. 3.8.1) vibrate with normal velocity :
n
(x. t ). As
before, the velocity potential in the ﬂuid (governed by Equation (3.1.1)) is the
same as that generated by the distribution of volume sources
q(x. t ) = :
n
(x. t )δ(s
⊥
−c) (c →÷0) (3.8.1)
distributed over S regarded as a rigid, stationary surface, where s
⊥
is distance
measured in the normal direction from S into the ﬂuid. The velocity potential
ϕ(x. t ) is given exactly by
ϕ(x. t ) = −
∞
−∞
ﬂuid
:
n
(y. τ)δ(s
⊥
−c)G(x. y. t −τ) d
3
y dτ (c →÷0)
= −
∞
−∞
S
:
n
(y. τ)G(x. y. t −τ) dS(y) dτ.
where
∂G
∂x
n
(x. y. t −τ) = 0 on S. (3.8.2)
At low frequencies the ﬁrst approximation to the farﬁeld sound is obtained
by replacing G(x. y. t − τ) in (3.8.2) by its compact approximation (3.7.5).
The details are given below; they illustrate the general procedure that should be
adopted when using the compact Green’s function (in particular, the technique
of expanding to ﬁrst order in Y):
ϕ(x. t ) ≈ −
∞
−∞
ﬂuid
:
n
(y. τ)δ(s
⊥
−c)
4π[X −Y[
δ
t −τ −
[X −Y[
c
0
d
3
y dτ
(c →÷0). [x[ →∞
= −
∞
−∞
S
:
n
(y. τ)
4π[X −Y[
δ
t −τ −
[X −Y[
c
0
dS(y) dτ.
= −
1
4π[x[
∞
−∞
S
:
n
(y. τ)δ
t −τ −
[x[
c
0
÷
x · Y
c
0
[x[
dS(y) dτ
(X ∼ x as [x[ →∞)
3.8 LowFrequency Radiation from a Vibrating Body 67
= −
1
4π[x[
∞
−∞
S
:
n
(y. τ)
¸
δ
t −τ −
[x[
c
0
÷δ
/
t −τ −
[x[
c
0
x
j
Y
j
c
0
[x[
¸
dS(y) dτ.
where the prime denotes differentiation with respect to time. Performing the
integration with respect to τ:
ϕ(x. t ) ≈ −
1
4π[x[
S
:
n
y. t −
[x[
c
0
dS(y)
−
x
j
4πc
0
[x[
2
∂
∂t
S
:
n
y. t −
[x[
c
0
Y
j
(y) dS(y).
The ﬁrst integral represents an omnidirectional monopole sound wave, and is
nonzero only if the volume enclosed by S changes with time (i.e., only for a
pulsating body). It is then the most important component of the far ﬁeld sound –
the second integral is smaller by a factor ∼O(ω¹¡c
0
) < 1 (because ∂¡∂t ∼ ω
and Y
j
∼ ¹).
The monopole term vanishes for a rigid body executing small amplitude
translational oscillations at velocity U(t ), say. Then,
:
n
(y. τ) = n(y) · U(τ) = n
i
(y)U
i
(τ).
where n(y) is the surface normal directed into the ﬂuid. Making the substitution
Y
j
= y
j
− ϕ
∗
j
(y) in the second integral we obtain an acoustic ﬁeld of dipole
type, given by
ϕ(x. t ) ≈ −
x
j
4πc
0
[x[
2
∂U
i
∂t
t −
[x[
c
0
S
n
i
(y)Y
j
(y) dS(y) (3.8.3)
= −
x
j
4πc
0
[x[
2
∂U
i
∂t
t −
[x[
c
0
S
[n
i
y
j
−n
i
ϕ
∗
j
] dS. [x[ →∞.
(3.8.4)
Example: The vibrating sphere Consider a rigid sphere of radius a centred at
the origin and oscillating in the x
1
direction at velocity U(t ) (Fig. 3.8.2). Then
U = (U. 0. 0), and it is only necessary to take i = 1 in (3.8.3) or (3.8.4).
In terms of spherical polar coordinates (r. ϑ. φ) we have
y = r(cos ϑ. sin ϑ cos φ. sin ϑ sin φ)
68 3 The Compact Green’s Function
Fig. 3.8.2.
Therefore,
Y ≡ y
1 ÷
a
3
2[y[
3
=
3a
2
(cos ϑ. sin ϑ cos φ. sin ϑ sin φ)
on the sphere and n
1
= cos ϑ
Hence,
S
n
1
Y
j
dS =
3a
3
2
S
(cos ϑ. sin ϑ cos φ. sin ϑ sin φ) cos ϑ sin ϑ dϑ dφ
=
¸
2πa
3
. j = 1.
0. j = 2. 3
and, therefore, (3.8.3) becomes
ϕ(x. t ) ≈
−a
3
cos θ
2c
0
[x[
∂U
∂t
(t −[x[¡c
0
). [x[ →∞. and
x
1
[x[
= cos θ.
(3.8.5)
which is the result already obtained in Section 3.5 using the solution derived
from the Helmholtz equation.
3.8.1 Far Field Pressure Produced by a Vibrating Body
A more general and illuminating discussion of the lowfrequency sound pro
duced by a vibrating rigid body can be given in terms of the added mass tensor
M
i j
(Batchelor 1967), deﬁned by the surface integral
M
i j
= −ρ
0
S
n
i
ϕ
∗
j
dS.
3.8 LowFrequency Radiation from a Vibrating Body 69
The Condition (3.4.9) satisﬁed by ϕ
∗
j
on S implies that M
i j
= M
j i
, because n
i
can be replaced in the integrand by ∂ϕ
∗
i
¡∂y
n
, and
M
i j
= −ρ
0
S
n
i
ϕ
∗
j
dS = −ρ
0
S
∂ϕ
∗
i
∂y
n
ϕ
∗
j
dS ≡ −ρ
0
S
ϕ
∗
i
∂ϕ
∗
j
∂y
n
dS = M
j i
.
(3.8.6)
The ﬁnal integral is deduced fromthe second by referring to Fig. 3.3.1, recalling
that ∇
2
ϕ
∗
j
= ∇
2
ϕ
∗
i
= 0, and applying the divergence theorem as follows:
S÷Y
∂ϕ
∗
i
∂y
n
ϕ
∗
j
−ϕ
∗
i
∂ϕ
∗
j
∂y
n
dS =
ﬂuid
(ϕ
∗
i
∇
2
ϕ
∗
j
−ϕ
∗
j
∇
2
ϕ
∗
i
) d
3
y ≡ 0.
The integration over Y vanishes as the surface recedes to inﬁnity (Batchelor
1967) because
ϕ
∗
i. j
(y) ∼ O
1
[y[
2
as [y[ →∞.
By evaluating the net force on S produced by the unsteady surface pressure
(or by the method described below in Section 4.4) it can be veriﬁed that when
the body translates at velocity U(t ) without rotation in an ideal, incompressible
ﬂuid, it exerts a force on the ﬂuid in the i direction given by
F
i
(t ) = M
i j
dU
j
dt
. (3.8.7)
For a body of mass m, this means that when an external force F
i
acts through
its centre of mass, the equation of motion of the body can be written
(mδ
i j
÷ M
i j
)
dU
j
dt
= F
i
.
The added mass tensor determines the effective mass of ﬂuid dragged along by
the body in its accelerated motion. The inertia of this ﬂuid, in addition to that
of the body, must also be overcome by the force F when the body accelerates.
In general, however, a couple must also be applied to the translating body to
counter a rotational torque also exerted on the body by the ﬂuid (see Batchelor
(1967) for further discussion).
Let us now apply these concepts to determine from (3.8.4) the sound pres
sure produced by a rigid compact body executing small amplitude translational
oscillations at velocity U(t ). The acoustic pressure is given in the far ﬁeld by
p = −ρ
0
∂ϕ¡∂t (see Section 1.3), and therefore
p(x. t ) =
x
j
4πc
0
[x[
2
∂
2
U
i
∂t
2
t −
[x[
c
0
¸
ρ
0
S
n
i
y
j
dS −ρ
0
S
n
i
ϕ
∗
j
dS
¸
.
[x[ →∞. (3.8.8)
70 3 The Compact Green’s Function
The ﬁrst integral is evaluated by applying the divergence theorem, which trans
forms it into an integral over the volume V
s
of the body:
ρ
0
S
n
i
y
j
dS = ρ
0
V
s
∂y
j
∂y
i
d
3
y = ρ
0
V
s
δ
i j
≡ m
0
δ
i j
. (3.8.9)
where m
0
is the mass of the ﬂuid displaced by the body. The second term in the
brace brackets of (3.8.8) is just the added mass tensor M
i j
.
Thus, the acoustic pressure can be expressed in either of the forms
p(x. t ) ≈
x
i
4πc
0
[x[
2
(m
0
δ
i j
÷ M
i j
)
∂
2
U
j
∂t
2
t −
[x[
c
0
=
x
i
4πc
0
[x[
2
m
0
∂
2
U
i
∂t
2
÷
∂ F
i
∂t
t −
[x[
c
0
. [x[ →∞. (3.8.10)
where the second line follows from (3.8.7), where F
i
is the force exerted by the
body on the ﬂuid in the i direction.
For a sphere of radius a oscillating at speed U(t ) in the x
1
direction
m
0
=
4
3
πa
3
ρ
0
and M
i j
=
1
2
m
0
δ
i j
Therefore,
p(x. t ) ≈
x
i
4πc
0
[x[
2
m
0
δ
i j
÷
1
2
m
0
δ
i j
∂
2
U
j
∂t
2
t −
[x[
c
0
. [x[ →∞.
=
ρ
0
a
3
cos θ
2c
0
[x[
∂
2
U
∂t
2
t −
[x[
c
0
.
which is equivalent to (3.8.5).
3.9 Compact Green’s Function Summary and Special Cases
3.9.1 Compact Bodies and Cylindrical Bodies of Compact Cross Section
General Form
G(x. y. t −τ) =
1
4π[X −Y[
δ
t −τ −
[X −Y[
c
0
(3.9.1)
X = x −ϕ
∗
(x)
¸
Kirchhoff vectors for the body.
Y = y −ϕ
∗
(y)
The vector components X
j
(x) and Y
j
(y) are the velocity potentials of
3.9 Compact Green’s Function Summary and Special Cases 71
Table 3.9.1. Standard Special Cases
Body X
1
X
2
X
3
Sphere of radius a, with
centre at origin
x
1
1 ÷
a
3
2[x[
3
x
2
1 ÷
a
3
2[x[
3
x
3
1 ÷
a
3
2[x[
3
Circular cylinder of radius a
coaxial with the x
3
axis
x
1
1 ÷
a
2
x
2
1
÷x
2
2
x
2
1 ÷
a
2
x
2
1
÷x
2
2
x
3
Strip airfoil
−a  x
1
 a. x
2
= 0.
−∞ x
3
 ∞
x
1
x
3
Re(−i
√
z
2
−a
2
)
z = x
1
÷i x
2
incompressible ﬂow past the body having unit speed in the j direction at large
distances fromthe body (special cases are listed in Table 3.9.1); ϕ
∗
j
is the veloc
ity potential of the incompressible ﬂow that would be produced by rigid body
motion of S at unit speed in the j direction. For a cylindrical body of compact
cross section parallel to the x
3
direction, we take
X
3
= x
3
. Y
3
= y
3
.
3.9.2 Airfoil of Variable Chord
The compact Green’s function deﬁned by (3.6.6) and (3.6.7) for a rigid strip
can be generalized to include the ﬁnite span, variable chord airfoil illustrated in
Fig. 3.9.1. The coordinate axes are orientated as in Fig. 3.6.3 for the strip airfoil,
with y
2
normal to the plane of the airfoil and y
3
in the spanwise direction. The
airfoil span is assumed to be large, and the chord 2a ≡ 2a(y
3
) is a slowly varying
function of y
3
. The potential Y
2
of ﬂow past the airfoil in the y
2
direction may
then be approximated locally by the formula for an airfoil of uniform chord
2a(y
3
).
Fig. 3.9.1.
72 3 The Compact Green’s Function
Therefore, a ﬁrst approximation to the compact Green’s function (3.9.1) for an
airfoil of span L occupying the interval −
1
2
L  y
3

1
2
L is obtained by taking
Y
1
= y
1
. Y
2
=
¸
Re
−i
z
2
−a(y
3
)
2
. [y
3
[ 
1
2
L
y
2
. [y
3
[ >
1
2
L
.
Y
3
= y
3
. z = y
1
÷i y
2
. (3.9.2)
This model has been found to give predictions within a few percent of those
based on the exact value of Y
2
(y) in the case of an airfoil of elliptic planform
whose
aspect ratio =
airfoil span
midchord
> 5.
3.9.3 Projection or Cavity on a Plane Wall
Let the plane wall be rigid and coincide with x
2
= 0 (Fig. 3.9.2). When the pro
jection or cavity is absent the Green’s function with vanishing normal derivative
on the wall is
G
0
(x. y. t −τ) =
1
4π[x −y[
δ
t −τ −
[x −y[
c
0
÷
1
4π[¯ x −y[
δ
t −τ −
[¯ x −y[
c
0
.
Fig. 3.9.2.
3.9 Compact Green’s Function Summary and Special Cases 73
where ¯ x = (x
1
. −x
2
. x
3
) is the image of the observer position x in the plane
wall.
The ﬁgure illustrates the case for a projection, but the following discussion
applies without change to compact (but nonresonant) wall cavities. Assume ﬁrst
that the origin is close to the projection. Let [x[ →∞(noting that [¯ x[ = [x[) and
expand G
0
near the projection to ﬁrst order in y (i.e., correct to dipole order)
G
0
(x. y. t −τ)
≈
1
4π[x[
¸
δ
t −τ −
[x[
c
0
÷
x · y
c
0
[x[
÷δ
t −τ −
[x[
c
0
÷
¯ x · y
c
0
[x[
¸
≈
1
4π[x[
¸
2δ
t −τ −
[x[
c
0
÷
2(x
1
y
1
÷ x
3
y
3
)
c
0
[x[
δ
/
t −τ −
[x[
c
0
¸
.
We require a corrected expression that has vanishing normal derivative (as a
function of y) on the wall and on the projection. By inspection, this is obtained
simply by replacing the factor
2(x
1
y
1
÷ x
3
y
3
)
c
0
[x[
by
2(x
1
Y
1
÷ x
3
Y
3
)
c
0
[x[
.
where Y
1
= y
1
−ϕ
∗
1
(y). Y
3
= y
3
−ϕ
∗
3
(y) are the velocity potentials of hori
zontal ﬂows past the projection that are parallel to the wall and have unit speeds
respectively in the y
1
and y
3
directions as [y[ →∞.
It may now be veriﬁed that (in the usual notation) the required compact
Green’s function is
G(x. y. t −τ) =
1
4π[X −Y[
δ
t −τ −
[X −Y[
c
0
÷
1
4π[
¯
X −Y[
δ
t −τ −
[
¯
X −Y[
c
0
. (3.9.3)
where
Y
1
=y
1
−ϕ
∗
1
(y). Y
2
=y
2
. Y
3
=y
3
−ϕ
∗
3
(y)
X
1
=x
1
−ϕ
∗
1
(x). X
2
=x
2
. X
3
=x
3
−ϕ
∗
3
(x)
¸
. (3.9.4)
and
¯
X = (X
1
. −X
2
. X
3
).
These formulae can be used also for a twodimensional projection or cavity
that is uniform, say, in the x
3
direction simply by setting Y
3
= y
3
. X
3
= x
3
.
To complete this discussion of compact Green’s function, we now give with
out proofs a selection of useful examples.
74 3 The Compact Green’s Function
Fig. 3.9.3.
3.9.4 Green’s Function for a HalfPlane (Howe, 1975a)
Analytical representations of the exact Green’s function
ˆ
G(x. y. ω) are known
for a rigid halfplane x
1
0. x
2
= 0 (which is inﬁnite in the x
3
direction,
Fig. 3.9.3) but are of limited use in applications. However, we can deﬁne a
compact Green’s function for a source at y whose distance from the edge is
small compared to the acoustic wavelength, that is, for κ
0
(y
2
1
÷ y
2
2
)
1
2
< 1. To
do this, we introduce cylindrical polar coordinates
x = (r cos θ. r sin θ. x
3
). y = (r
0
cos θ
0
. r
0
sin θ
0
. y
3
).
Then, if i
3
is a unit vector in the x
3
direction (parallel to the edge),
ˆ
G(x. y. ω) =
ˆ
G
0
(x. y. ω) ÷
ˆ
G
1
(x. y. ω) ÷· · · . (3.9.5)
where, for [x − y
3
i
3
[ →∞and κ
0
y
2
1
÷ y
2
2
<1,
ˆ
G
0
(x. y. ω) =
−1
4π[x − y
3
i
3
[
e
i κ
0
[x−y
3
i
3
[
.
(3.9.6)
ˆ
G
1
(x. y. ω) =
−1
π
√
2πi
√
κ
0
ϕ
∗
(x)ϕ
∗
(y)
[x − y
3
i
3
[
3¡2
e
i κ
0
[x−y
3
i
3
[
.
and
ϕ
∗
(x) =
√
r sin(θ¡2). ϕ
∗
(y) =
√
r
0
sin(θ
0
¡2). (3.9.7)
3.9 Compact Green’s Function Summary and Special Cases 75
ϕ
∗
(x) is a velocity potential of incompressible ﬂow around the edge of the half
plane expressed in terms of polar coordinates (x
1
. x
2
) = r(cos θ. sin θ) (and
similarly for ϕ
∗
(y)). The component
ˆ
G
0
of
ˆ
G represents the radiation from a
point source at y when scattering is neglected;
ˆ
G
1
is the ﬁrst correction due to
presence of the halfplane.
3.9.5 TwoDimensional Green’s Function for a HalfPlane
(Howe, 1975a)
The Green’s function for the wave equation in two dimensions, where conditions
are uniform in the x
3
direction, satisﬁes
1
c
2
0
∂
2
∂t
2
−∇
2
G = δ(x
1
− y
1
)δ(x
2
− y
2
)δ(t −τ).
where G = 0 for t  τ. (3.9.8)
When a line source at y = (y
1
. y
2
) is close to the edge of the halfplane in
Fig. 3.9.3 the corresponding compact Green’s function is obtained by integrat
ing (3.9.6) over −∞  y
3
 ∞, using the method of stationary phase for
κ
0
x
2
1
÷ x
2
2
→∞(see Example 2 of Section 5.2), and then using the Formula
(3.1.6) to calculate G(x. y. t − τ) ≈ G
0
(x. t − τ) ÷ G
1
(x. y. t − τ) ÷ · · ·. In
particular G
1
(x. y. t −τ) is the ﬁrst term in the expansion that involves y, and
is found to be
G
1
(x. y. t −τ) ≈
ϕ
∗
(x)ϕ
∗
(y)
π[x[
δ(t −τ −[x[¡c
0
). [x[ →∞. (3.9.9)
where x = (x
1
. x
2
). y = (y
1
. y
2
) and ϕ
∗
is deﬁned as in (3.9.7).
3.9.6 TwoDimensional Green’s Function for a Plane with an Aperture
A rigid plane x
1
= 0 is pierced by a twodimensional aperture occupying
−a  x
2
 a (Fig. 3.9.4). The twodimensional compact Green’s function (the
solution of (3.9.8)) is applicable for a source at y = (y
1
. y
2
) well within an
acoustic wavelength of the aperture on either side of the plane. The observer at
x = (x
1
. x
2
) is in the acoustic far ﬁeld. The ydependent part of the compact
Green’s function is
G(x. y. t −τ) ≈−
√
c
0
sgn(x
1
)
π
√
2π[x[
χ(t −τ −[x[¡c
0
)
√
t −τ −[x[¡c
0
Re
¸
ln
Z
a
÷
Z
2
a
2
−1
¸
.
Z = y
2
÷i y
1
. (3.9.10)
76 3 The Compact Green’s Function
Fig. 3.9.4.
where
χ(t ) = H(t )
∞
0
ln(maξ
2
¡4c
0
t )e
−ξ
2
dξ
[ln(maξ
2
¡4c
0
t )]
2
÷π
2
.
and m = 1.781072. . . . Note the deﬁnition Z = y
2
÷i y
1
.
3.9.7 Green’s Function for Long Waves in a Rigid Walled Duct
(Howe, 1975b)
Only plane waves can propagate in a cylindrical duct of crosssectional area
A when the characteristic wavelength of sound is large compared with the
diameter ∼
√
A, even if the source region is highly three dimensional. When this
condition is satisﬁed the corresponding compact Green’s function satisﬁes the
onedimensional wave equation, provided the crosssectional area is uniform.
Taking the x
1
direction along the axis of the duct (Fig. 3.9.5a), we have
G(x. y. t −τ) =
c
0
2A
H
t −τ −
[x
1
− y
1
[
c
0
. [x −y[ ·
√
A. (3.9.11)
where H is the Heaviside step function. For a uniformduct with an acoustically
3.9 Compact Green’s Function Summary and Special Cases 77
Fig. 3.9.5.
compact section of variable cross section, such as the neck in Fig. 3.9.5b, the
compact Green’s function becomes
G(x. y. t −τ) =
c
0
2A
H
t −τ −
[X
1
−Y
1
[
c
0
. [x −y[ ·
√
A. (3.9.12)
where X
1
(x) and Y
1
(y) are the velocity potential of incompressible ﬂow in the
duct having unit speed at large distances from the neck.
3.9.8 Compact Green’s Function for a Duct Entrance (Howe, 1998a,b)
The typical geometry is illustrated in Fig. 3.9.6a. Within the duct, several
diameters from the entrance, the crosssectional area is uniform and equal to
A. However, the geometry of the entrance can be arbitrary, and not necessarily
that of the uniform cylinder shown in the ﬁgure. Take the coordinate origin in
the entrance plane of the duct, with the negative x
1
axis lying along the axis of
the duct. Then, there are two cases:
(i) Propagation within the Duct
This is applicable for the case shown in Fig. 3.9.6a, involving a source at y near
the duct entrance and an observer at x within the duct (or vice versa), when the
78 3 The Compact Green’s Function
Fig. 3.9.6.
characteristic acoustic wavelength is large compared to the duct diameter.
G(x. y. t −τ) ≈
c
0
2A
¸
H
¸
t −τ −
[ϕ
∗
(x) −ϕ
∗
(y)[
c
0
¸
−H
¸
t −τ ÷
ϕ
∗
(x) ÷ϕ
∗
(y)
c
0
¸¸
. (3.9.13)
where the velocity potential ϕ
∗
(y) describes incompressible ﬂowfromthe duct,
and satisﬁes
ϕ
∗
(x) ≈x
1
−¹
/
when [x
1
[ ·
√
A within the duct.
≈−A¡4π[x[ when [x[ ·
√
A outside the duct. (3.9.14)
in which ¹
/
is the end correction (Rayleigh, 1945) of the duct opening (≈0.61R
for an unﬂanged circular cylinder of radius R =
√
A¡π).
(ii) Propagation in Free Space (Fig. 3.9.6b)
When either the source or observer is located at a large distance from the duct
entrance in free space
G(x. y. t −τ) =
1
4π[X −Y[
δ
t −τ −
([X(x) −Y(y)[ −[ϕ
∗
(x) ÷ϕ
∗
(y)])
c
0
.
(3.9.15)
where ϕ
∗
is deﬁned as in (3.9.14), and X(x). Y(y) denote the Kirchhoff vector
whose i component is the velocity potential of ﬂow past the stationary surface
formed by the duct entrance having unit speed in the i direction at large dis
tances from the entrance outside the duct (they become exponentially small
Problems 3 79
with distance [x
1
[ or [y
1
[ into the duct). The terms ϕ
∗
(x). ϕ
∗
(y) account for the
additional, weak monopole sound generated by a source near the duct entrance;
the source compresses the ﬂuid in the duct mouth producing a sound wave in the
duct whose reaction on the mouth causes a volume ﬂux equal to the monopole
source strength. The amplitude of this monopole is of the same order as the
usual dipole sound determined by the compact Green’s function.
For a uniform, thinwalled cylindrical duct we can take
X(x) ≡ (x
1
−ϕ
∗
(x). X
2
(x). X
3
(x)). Y(y) ≡ (y
1
−ϕ
∗
(y). Y
2
(y). Y
3
(y)).
If the source coordinate y
1
→−∞within the duct,
G(x. y. t −τ) =
1
4π[x[
δ(t −τ −([x[ − y
1
)¡c
0
). [x[ →∞. (3.9.16)
This represents a monopole wave centered on the duct entrance. This limiting
form of G can be used to calculate the lowfrequency free space radiation
generated by internal sources far from the entrance.
Problems 3
1. Use the compact Green’s function to solve
∇
2
÷κ
2
0
ˆ ϕ = f
2
∂
∂x
2
[δ(x
1
− L)δ(x
2
)δ(x
3
)] .
where
∂ϕ
∂x
n
= 0 on [x[ = a.
for the sound radiated by an azimuthally orientated dipole adjacent to a
compact, rigid sphere.
2. Use the compact Green’s function to solve
∇
2
÷κ
2
0
ˆ ϕ = f
1
∂
∂x
1
[δ(x
1
− L)δ(x
2
)δ(x
3
)].
where
∂ϕ
∂x
n
= 0 on
x
2
1
÷ x
2
2
1
2
= a.
for the sound radiated by a radially orientated dipole adjacent to a rigid
circular cylinder of compact cross section.
80 3 The Compact Green’s Function
3. Repeat Question 2 for the dipoles
f
2
∂
∂x
2
[δ(x
1
− L)δ(x
2
)δ(x
3
)] . f
3
∂
∂x
3
[δ(x
1
− L)δ(x
2
)δ(x
3
)].
4. The constant strength dipole
q(x. t ) = f
2
∂
∂x
2
(δ(x
1
−Ut )δ(x
2
−h)δ(x
3
)). f
2
= constant
translates at constant velocity U past a ﬁxed rigid cylinder of radius a  h
whose axis coincides with the x
3
axis. Show that when M = U¡c
0
<1, the
farﬁeld acoustic potential determined by Equation (3.1.1) is given by
ϕ ≈
f
2
Ma
2
2π[x[(h
2
÷U
2
[t ]
2
)
3
¸
x
1
h
[x[
(3U
2
[t ]
2
−h
2
) ÷
x
2
U[t ]
[x[
(3h
2
−U
2
[t ]
2
)
¸
.
where
[t ] = t −
[x[
c
0
.
5. The volume source
q(x. t ) = q
0
δ(x
1
−Ut )δ(x
2
−h)δ(x
3
). q
0
= constant
translates at constant velocity U past a ﬁxed rigid sphere of radius a  h
whose center is at the origin. Determine from Equation (3.1.1) the far ﬁeld
acoustic pressure p = −ρ
0
∂ϕ¡∂t given that M = U¡c
0
<1.
6. The point source q(x. t ) = q
0
δ(x
1
−Ut )δ(x
2
)δ(x
3
). (q
0
= constant) convects
along the axis of symmetry of the necked duct shown in Fig. 3.9.5 at constant,
low Mach number speed U. If the crosssectional area of the duct is denoted
by S(x
1
), where S(x
1
) → A. x
1
→±∞, use the approximations
X
1
= A
x
1
0
dξ
S(ξ)
. Y
1
= A
y
1
0
dξ
S(ξ)
.
to calculate the acoustic pressure radiated from the neck during the passage
of the source.
7. In incompressible ﬂow the velocity potential generated by a distribution of
sources q(x. t ) near a rigid body is determined by the solution of
∇
2
ϕ = q(x. t ).
Problems 3 81
Show that the monopole and dipole components of the solution at large
distances from the body (in the hydrodynamic far ﬁeld) can be calculated
using the following incompressible limit of the compact Green’s function
G(x. y. t −τ) =
−δ(t −τ)
4π[X −Y[
.
8. A rigid body translates without rotation in the j direction at velocity U
j
(t )
in an ideal, incompressible ﬂuid at rest at inﬁnity. Show that the velocity
potential of the ﬂuid motion is
+ = U
j
ϕ
∗
j
.
In a ﬁxed reference frame, the pressure can be calculated from Bernoulli’s
equation:
∂+
∂t
÷
p
ρ
0
÷
1
2
(∇+)
2
= 0.
Use these results to prove formula (3.8.7) for the force exerted on the ﬂuid
by the body.
9. A compact rigid disc of radius a executes small amplitude vibrations at
velocity U(t ) normal to itself. In the undisturbed state it lies in the plane
x
1
=0 with its center at the origin. If ϕ
∗
1
(x) = ∓(2¡π)
a
2
− x
2
2
− x
2
3
on the
faces x
1
= ±0.
x
2
2
÷ x
2
3
 a of the disc, show that the acoustic pressure
generated by the motion is given by
p(x. t ) ≈
2ρ
0
a
3
cos θ
3πc
0
[x[
∂
2
U
∂t
2
(t −[x[¡c
0
). [x[ →∞. cos θ =
x
1
[x[
.
4
Vorticity
4.1 Vorticity and the Kinetic Energy of Incompressible Flow
4.1.1 Kelvin’s (1867) Deﬁnition
Kelvin was responsible for much of the pioneering work on the mechanics of
incompressible ﬂow. He gave the following deﬁnition of a vortex in a homo
geneous incompressible ﬂuid,
. . . a portion of ﬂuid having any motion that it could not acquire by ﬂuid pressure
transmitted from its boundary.
To understand this consider the ideal (i.e., inviscid) incompressible ﬂow
produced by arbitrary motion of a solid body with surface S (Fig. 4.1.1). The
motion generated from rest by ‘ﬂuid pressure transmitted from its boundary’
can be described by a velocity potential ϕ such that
v(x. t ) = ∇ϕ.
∂ϕ
∂x
n
= U
n
on S.
where U
n
is the normal component of velocity on S.
There are no sources within the instantaneous region V occupied by the ﬂuid,
where ∇
2
ϕ = 0. Therefore, the kinetic energy T
0
of the ﬂow is
T
0
=
1
2
ρ
0
V
(∇ϕ)
2
d
3
x =
1
2
ρ
0
V
(div(ϕ∇ϕ) −ϕ∇
2
ϕ) d
3
x
= −
1
2
ρ
0
S
ϕ
∂ϕ
∂x
n
dS ≡ −
1
2
ρ
0
S
ϕU
n
dS. (4.1.1)
where the divergence theorem has been used to obtain the second line (there
is no contribution from the surface Y at inﬁnity in Fig. 4.1.1, where ϕ ∼
O(1¡[x[
2
)). This formula implies that if S is suddenly brought to rest (U
n
→0)
the motion everywhere in the ﬂuid ceases instantaneously, because
V
(∇ϕ)
2
d
3
x
82
4.1 Vorticity and the Kinetic Energy of Incompressible Flow 83
Fig. 4.1.1.
can vanish only if ∇ϕ ≡ 0. This unphysical behavior is never observed in a
real ﬂuid because (i) no ﬂuid is perfectly incompressible, and signals generated
by changes in the boundary conditions propagate at the ﬁnite speed of sound,
and (ii) diffusion of vorticity from the boundary supplies irrecoverable kinetic
energy to the ﬂuid.
For an incompressible, real ﬂuid we write
v = ∇ϕ ÷u.
and deﬁne the vorticity ω by
ω = curl u ≡ curl v.
If ϕ is taken to be deﬁned as above (for ideal ﬂow) then, because div u = 0 and
the normal component u
n
= n· u = 0 on S, the kinetic energy becomes
T =
1
2
ρ
0
V
(∇ϕ ÷u)
2
d
3
x =
1
2
ρ
0
V
((∇ϕ)
2
÷2∇ϕ · u ÷u
2
) d
3
x
=
1
2
ρ
0
V
((∇ϕ)
2
÷u
2
) d
3
x ÷ρ
0
V
div(ϕu) d
3
x
=
1
2
ρ
0
V
((∇ϕ)
2
÷u
2
) d
3
x −ρ
0
S
ϕu
n
dS
= −
1
2
ρ
0
S
ϕU
n
dS ÷
1
2
ρ
0
V
u
2
d
3
x ≡ T
0
÷
1
2
ρ
0
V
u
2
d
3
x. (4.1.2)
84 4 Vorticity
When the surface motion is arrested, the ﬂow described by the velocity poten
tial ϕ stops instantaneously, but that associated with the rotational velocity u
persists. The crucial difference between rotational and irrotational ﬂows is that,
once established, vortical motions proceed irrespective of whether or not the
ﬂuid continues to be driven by moving boundaries or other external agencies.
Equation (4.1.2) also establishes Kelvin’s theorem that T ≥ T
0
: the kinetic
energy of the real ﬂow(for which u ,= 0) always exceeds that of the correspond
ing ideal, irrotational ﬂow. In other words, the irrotational motion represents
the least possible disturbance that can be produced in the ﬂuid by the moving
boundary.
4.2 The Vorticity Equation
Let v
A
denote the ﬂuid velocity at a point A at x. The velocity v
B
at a neigh
bouring point B at x ÷δx can then be written (Goldstein, 1960)
v
B
≈ v
A
÷(δx · ∇)v
= v
A
÷
1
2
ω ∧ δx ÷
1
2
∇(e
i j
δx
i
δx
j
).
where e
i j
is the rateofstrain tensor (2.1.3) and the gradient in the second line
is taken with respect to δx. The ﬁrst two terms on the second line represent
motion of A and B as a rigid body, consisting of a translation at velocity v
A
together with a rotation at angular velocity
1
2
ω; the last term, being a gradient,
represents an irrotational distortion of the ﬂuid in the neighborhood of A.
If we consider a small spherical ﬂuid particle with center at A, the distortion
corresponds to a deformation into an ellipsoid whose principal axes correspond
to the principal axes of e
i j
. If a spherical ﬂuid particle is suddenly solidiﬁed
without change of angular momentum, it will rotate at angular velocity ω¡2, so
that ω may be deﬁned as twice the initial angular velocity of the solid sphere
when an inﬁnitesimally small sphere of ﬂuid with center at A is suddenly solid
iﬁed without change of angular momentum (but this is not true for arbitrarily
shaped volume elements). The vorticity may therefore be regarded as a measure
of the angular momentum of a ﬂuid particle. This is consistent with our conclu
sion above regarding kinetic energy, inasmuch as the conservation of angular
momentum suggests that vorticity is associated the intrinsic kinetic energy of
the ﬂow, and determines the motion that persists in an incompressible ﬂuid
when the boundaries are brought to rest.
Avortex line is tangential to the vorticity vector at all points along its length.
Vortex lines that pass through every point of a simple closed curve deﬁne the
boundary of a vortex tube. For a tube of small crosssectional area δS the product
4.2 The Vorticity Equation 85
ωδS is called the tube strength, and is constant because
div ω = div(curl v) ≡ 0.
and the divergence theorem therefore implies that
ω· dS = 0 for any closed
surface, and in particular for the surface formed by two cross sections of the
tube and the tube wall separating them, on the latter of which ω· dS = 0. It
follows that vortex tubes and lines cannot begin or end within the ﬂuid. The
noslip condition (Batchelor 1967) requires the velocity at a boundary to be the
same as that of the boundary. A vortex line must therefore form a closed loop,
or end on a rotating surface S at which
n· ω = 2n· Ω. (4.2.1)
where Ω is the angular velocity of S.
We shall show that vorticity is transported by convection and molecular
diffusion. Therefore an initially conﬁned region of vortex loops can frequently
be assumed to remain within a bounded region. In the absence of body forces F,
we ﬁrst use the identity curl curl A = grad div A−∇
2
Ato write the momentum
equation (1.2.3) for homentropic ﬂow in the form
∂v
∂t
÷(v · ∇)v ÷∇
dp
ρ
= −ν
curl ω −
4
3
∇(div v)
.
By using the vector identity
(v · ∇)v = ω ∧ v ÷∇
1
2
:
2
. (4.2.2)
the momentum equation can be cast into Crocco’s form
∂v
∂t
÷ω ∧ v ÷∇B = −ν
curl ω −
4
3
∇(div v)
. (4.2.3)
where
B =
dp
ρ
÷
1
2
:
2
(4.2.4)
is the total enthalpy in homentropic ﬂow. The vector ω ∧ v is sometimes
called the Lamb vector. When the ﬂuid is incompressible (or when the term
in div v representing the small effect of compressibility on viscous dissipation
is neglected) Crocco’s equation reduces to
∂v
∂t
÷ω ∧ v ÷∇B = −ν curl ω. (4.2.5)
in which case dissipation occurs only where ω ,= 0.
86 4 Vorticity
The curl of (4.2.3) and the relation curl curl ω ≡ −∇
2
ω yield the vorticity
equation for a Stokesian ﬂuid of constant shear viscosity:
∂ω
∂t
÷curl(ω ∧ v) = ν∇
2
ω. (4.2.6)
For an incompressible ﬂuid div ω = div v = 0, and
curl(ω∧v) ≡ (v · ∇)ω ÷ωdiv v −(ω· ∇)v −v div ω = (v · ∇)ω −(ω· ∇)v.
therefore the vorticity equation can also be written
Dω
Dt
= (ω· ∇)v ÷ν∇
2
ω. (4.2.7)
The terms on the right represent the mechanisms that change the vorticity of a
moving ﬂuid particle in incompressible ﬂow:
(i) (ω· ∇)v.
Consider a ﬂuid particle at A in Fig. 4.2.1 with velocity v at time t . Let the
vorticity at A be ω = ωn, where n is a unit vector, and consider a neighboring
particle at B a small distance s from Ain the direction n; that is, snis the position
of B relative to A. At time t the points A and B lie on the vortex line through
A, and the velocity at B is v ÷s(n· ∇)v. After a short time δt , A has moved a
vector distance v δt to A
/
and B has moved to B
/
whose position relative to A
/
is s(n÷(n· ∇)v δt ). During this time, the term(ω· ∇)v in the vorticity equation
causes the vorticity of the ﬂuid particle initially at A to change fromωn at A to
ω(n ÷(n· ∇)v δt ) at A
/
. Thus, the vortex line through A
/
lies along the relative
vector s(n ÷ (n· ∇)v δt ) from A
/
to B
/
. Therefore, the ﬂuid particles and the
vortex line through A and B have deformed and convected in the ﬂow in the
Fig. 4.2.1.
4.2 The Vorticity Equation 87
same way; in their new positions A and B continue to lie on the same vortex
line. In the absence of viscosity (when ν∇
2
ω does not appear on the right of
(4.2.7)) vortex lines are therefore said to move with the ﬂuid; they are rotated
and stretched in a manner determined entirely by the relative motions of A and
B. The magnitude of ω increases in direct proportion to the stretching of vortex
lines. When a vortex tube is stretched, the crosssectional area δS decreases and
therefore ω must increase to preserve the strength of the tube.
(ii) ν∇
2
ω: Molecular Diffusion of Vorticity
This term is important only in regions of high shear, in particular near solid
boundaries. Very close to a stationary wall the velocity becomes small and non
linear terms in the vorticity equation (4.2.7) can be neglected. The equation then
reduces to the classical diffusion equation
∂ω
∂t
= ν∇
2
ω.
Vorticity is generated at solid boundaries, and viscosity is responsible for its
diffusion into the body of the ﬂuid, where it can subsequently be convected by
the ﬂow.
It should be understood that viscosity merely serves to diffuse the vorticity
into the ﬂuid from the surface, and does not generate the vorticity. In an ideal
ﬂuid the slipping of the ﬂowover the surface creates a singular layer of vorticity
at the surface called a vortex sheet whose strength is determined by the tan
gential velocity difference between the surface and the ideal exterior ﬂow. This
vorticity stays on the surface; it would start to diffuse into the ﬂuid if the ﬂuid
were suddenly endowed with viscosity. The rate of diffusion would depend on
the value of ν, but the amount of the vorticity available for diffusion from the
surface is independent of ν.
The circulation I with respect to a closed material contour C is deﬁned by
I =
C
v · dx =
S
/
curl v · dS ≡
S
/
ω· dS.
where S
/
is any twosided surface bounded by C. When ν = 0 the motion in
homentropic ﬂow evolves in such a way that the circulation around the moving
contour remains constant:
DI
Dt
=
D
Dt
C
v · dx =
C
∇
−
dp
ρ
÷
1
2
:
2
· dx ≡ 0. (4.2.8)
This is Kelvin’s circulation theorem. It follows that vorticity can neither be
created nor destroyed in a body of inviscid and homentropic ﬂuid.
88 4 Vorticity
Fig. 4.2.2.
4.2.1 Vortex Sheets
A vortex sheet is a useful model of a thin layer of vorticity when viscous diffu
sion can be neglected. Imagine a thin shear layer (Fig. 4.2.2) across which the
velocity changes rapidly from v
−
to v
÷
. We approximate the layer by a surface
f (x. t ) = 0 with unit normal n across which the normal components of veloc
ity are equal (n· v
−
= n· v
÷
), but the tangential components are discontinuous
(n ∧ v
−
,= n ∧ v
÷
). Let f
>

0 respectively on the ± sides of the surface. Near
the sheet, on either side, it can be assumed that curl v
±
= 0, and we can set
v = H( f )v
÷
÷ H(−f )v
−
Hence,
ω = ∇H ∧ (v
÷
−v
−
) = n ∧ (v
÷
−v
−
)δ(s
⊥
). (4.2.9)
where ∇H ≡∇H( f ) =−∇H(−f ) =nδ(s
⊥
), and s
⊥
is distance measured in
the normal direction from the sheet.
In a real ﬂuid the vorticity would diffuse out from the sheet and it could
not therefore persist indeﬁnitely. In an ideal ﬂuid the sheet is subject only to
convection and stretching by the ﬂow at the local mean velocity, which is
v =
1
2
(v
÷
÷v
−
). (4.2.10)
where v
±
are evaluated just above and below the sheet.
4.3 The Biot–Savart Law
In an unbounded ﬂuid the velocity v can always be expressed in terms of scalar
and vector potentials ϕ and A such that
v = ∇ϕ ÷curl A. where div A = 0.
The equations determining ϕ and A are found by taking in turn the divergence
4.3 The Biot–Savart Law 89
and curl (using the formula curl curl A = grad div A −∇
2
A):
∇
2
ϕ = div v. ∇
2
A = −curl v ≡ −ω.
We can take ϕ =0 for incompressible, unbounded ﬂow which is at rest at in
ﬁnity. To ﬁnd A we use the Green’s function for Laplace’s equation determined
by (1.4.6) (i.e., by the incompressible limit of (3.2.6), when κ
0
= 0) to obtain
A =
ω(y. t ) d
3
y
4π[x −y[
.
The velocity is then given by the Biot–Savart formula
v(x. t ) = curl
ω(y. t ) d
3
y
4π[x −y[
. (4.3.1)
This is a purely kinematic relation between a vector v that vanishes at inﬁnity
and ω = curl v.
Because vorticity is transported by convection and diffusion, an initially
conﬁned region of vorticity will tend to remain within a bounded domain, so
that it may be assumed that ω →0 as [x[ →∞. The divergence theorem then
shows that
ω
i
(y. t ) d
3
y = −
Y
y
i
ω
j
(y. t )n
j
dS(y) ≡ 0.
where the surface Y (with inward normal n) is large enough to contain all the
vorticity. By using this result and the expansion (1.9.3) for [x[ →∞we derive
from (4.3.1) the following approximation in the hydrodynamic far ﬁeld:
v(x. t ) ≈ curl
x
j
4π[x[
3
y
j
ω(y. t ) d
3
y
∼O
1
[x[
3
. [x[ →∞. (4.3.2)
Furthermore, the divergence theorem also implies that
div(y
i
y
j
ω(y. t )) d
3
y = 0.
and therefore that
(y
i
ω
j
(y. t ) ÷ y
j
ω
i
(y. t )) d
3
y = 0.
This can be used to express (4.3.2) in either of the following equivalent forms
v(x. t ) ≈ curl curl
I
4π[x[
= grad div
I
4π[x[
. [x[ →∞.
where I =
1
2
y ∧ ω(y. t ) d
3
y. (4.3.3)
90 4 Vorticity
(see Question 2 of Problems 4). The vector I is called the impulse of the vortex
system, and is an absolute constant in an unbounded ﬂow (see Section 4.4).
These formulae supply alternative representations of v in the hydrodynamic far
ﬁeld (where the motion is entirely irrotational) in terms of either the vector po
tential A = curl (I¡4π[x[) or the scalar potential ϕ = div (I¡4π[x[). (Batchelor
(1967) denotes I by P¡ρ
0
; Lighthill (1978, 1986) uses G.)
4.3.1 Kinetic Energy
Using the Biot–Savart formula it can be veriﬁed that the kinetic energy of an
unbounded (threedimensional) incompressible ﬂow is given in terms of the
vorticity by
T =
ρ
0
8π
ω(x. t ) · ω(y. t )
[x −y[
d
3
x d
3
y. (4.3.4)
The following representation can also be derived (using (4.2.2))
T = ρ
0
x · (ω ∧ v)(x. t ) d
3
x. (4.3.5)
4.3.2 Incompressible Flow with an Internal Boundary
Let the rigid body in Fig. 4.1.1 have volume L and move in an incompressible
ﬂuid at rest at inﬁnity with velocity
U = U
0
÷Ω∧ (x −x
0
(t )). (4.3.6)
where U
0
= dx
0
¡dt is the velocity of its center of volume x
0
(t ), and Ω(t ) is its
angular velocity.
In the usual waylet f (x. t ) vanishon S, with f >0inthe ﬂuid. Then H( f )v ÷
H(−f )Uis the velocity everywhere, in both the ﬂuid and solid (where it equals
U(x. t )). But
H( f )v ÷ H(−f )U = ∇ϕ ÷curl A. (div A = 0).
The body has constant volume (div U = 0), but curl U = 2Ω. Now, the noslip
condition on S implies that
div(H( f )v ÷ H(−f )U) = ∇H( f ) · (v −U) ≡ 0
curl(H( f )v ÷ H(−f )U) = H( f )ω ÷ H(−f )2Ω.
4.3 The Biot–Savart Law 91
Hence, ϕ ≡0, and the velocity everywhere is given by the following modiﬁca
tion of the Biot–Savart formula (4.3.1):
v(x. t ) = curl
V
ω(y. t ) d
3
y
4π[x −y[
÷curl
L
2Ω(t ) d
3
y
4π[x −y[
. (4.3.7)
where V is the volume occupied by the ﬂuid. This formula predicts that v =
U when x lies in the region L occupied by the body. Vortex lines may be
imagined to continue into the solid. As for an unbounded ﬂow, the identity
curl (H( f )v ÷ H(−f )U) d
3
x =0 implies that v ∼O(1¡[x[
3
) as [x[ →∞.
Similarly, the asymptotic representations (4.3.3) remain valid provided the inte
gration includes the region occupied by the body (where ω = 2Ω). The formula
is also applicable in inviscid ﬂow, but the contribution from the bound vorticity
in the vortex sheet on the surface of the body must be included in the integrals.
4.3.3 Blowing Out a Candle (Lighthill 1963)
An amusing illustration of the signiﬁcance of vorticity is depicted in Fig. 4.3.1,
where a puff of air is ejected from the tube and directed at the ﬂame of a candle.
Suppose the tube has radius R, and that the air is forced out at constant speed
V by impulsive movement of the piston over an axial distance L. In an ideal
ﬂuid the motion outside is irrotational and resembles at large distances from
the exit a radially symmetric source ﬂow. This ﬂowpersists only while the piston
is moving, during which time the velocity potential at a large distance r fromthe
Fig. 4.3.1.
92 4 Vorticity
exit resembles that produced by a monopole of strength q = π R
2
V:
ϕ ∼ −
V R
2
4r
.
so that the air blows against the ﬂame at distance ¹ at speed
V
ϕ
∼
V
4
R
2
¹
2
.
In reality, vorticity is generated at the tube wall. The air leaves the tube in the
form of a jet, and the exiting ﬂuid is initially contained within a cylindrical slug
of air of length L, whose displacement from the tube forces the potential ﬂow
ϕ. In addition, however, vorticity leaves the tube within a circular cylindrical
vortex sheet at the periphery of the slug, across which the axial velocity changes
from V within the jet to 0 outside. The sheet may be pictured as a succession
of vortex rings of radius R and inﬁnitesimal core radii. The circulation of these
rings per unit length of the jet is V, and they translate at the local mean air
velocity on the sheet equal to
1
2
V. The total circulation ejected from the tube
during the time L¡V in which the piston moves is therefore I =
1
2
LV.
Shortly after leaving the tube the cylindrical vortex rolls up to form a vortex
ring of circulation I which translates by selfinduction (as determined by the
Biot–Savart law (4.3.1)) at speed estimated by Kelvin to equal
V
t
∼
I
4π R
0
¸
ln
8R
0
σ
−
1
4
¸
≈
V L
8π R
0
¸
ln
8R
0
σ
−
1
4
¸
.
where R
0
∼ 1.2R is the radius of the vortex ring and σ ∼ 0.2R
0
is the radius of
its core (assumed to be of circular cross section). The ring arrives at the ﬂame
after a time t
¹
∼ ¹¡V
t
.
The air on the axis of the vortex ring at its center forms a localized jet with
velocity on the centerline equal to
V
j
∼
V L
4R
0
.
If the ﬂame is extinguished it is because the vortex jet blows away the hot
combusting gases from newly vaporized wax.
According to this sequence of events, the candle is only blown out because of
the presence of the vortex. In its absence the ﬂame would barely ﬂicker under the
inﬂuence of blowing by the potential velocity ﬁeld V
ϕ
. The following numeri
cal estimates conﬁrmthis conclusion. Take V = 10m/s, L = 1cm, R = 0.5cm,
and¹ =0.3m. Then, R
0
≈0.6 cm. σ ≈0.12 cm. V
ϕ
≈0.0007 m/s. V
t
≈2.4 m/s.
V
j
≈4.2 m/s. and t
¹
≈0.13 s.
4.4 Surface Force in Incompressible Flow 93
4.4 Surface Force in Incompressible Flow Expressed
in Terms of Vorticity
We now derive the following formula for the force F exerted on an incompress
ible ﬂuid by a rigid body with surface S whose centre of volume has velocity
U
0
:
F ÷m
0
dU
0
dt
= ρ
0
dI
dt
≡
ρ
0
2
d
dt
x ∧ ω(x. t ) d
3
x. (4.4.1)
where I is the impulse deﬁned by the integral in (4.3.3), including any contri
butions from bound vorticity within S, and m
0
is the mass of ﬂuid displaced by
the body.
With reference to Fig. 4.1.1, let
F = external force applied to the body to maintain its motion,
m = mass of the body.
The ﬂuid is assumed to be at rest at inﬁnity. Let V denote the ﬂuid between a
large closed surface Y containing all of the vorticity and the surface S of the
body, and let V
÷
denote the interior of Y including the volume L of the body.
The center of volume of the body is assumed to be in motion at velocity U
0
(t ),
but in general the body may also be rotating at some time dependent angular
velocity Ω. The global equations of motion are
m
dU
dt
÷ρ
0
d
dt
V
v(x. t ) d
3
x = F ÷
Y
p(x. t ) dS
m
dU
dt
= F −F
where U is the velocity of the centre of mass of the body. Subtracting these
equations and extending the volume integral to include the volume Loccupied
by the rigid body (where
L
v d
3
x = LU
0
), we can also write
F ÷m
0
dU
0
dt
= ρ
0
d
dt
V
÷
v(x. t ) d
3
x −
Y
p(x. t ) dS. (4.4.2)
Now let v(x. t ) = curl A, where the vector potential A(x. t ) is deﬁned by the
Biot–Savart integral (4.3.1) taken over V
÷
, which includes the region occupied
by the body. Then, (4.3.3) implies that A = curl(I(t )¡4π[x[) on Y, so that
V
÷
v(x. t ) d
3
x =
V
÷
curl Ad
3
x =−
Y
n∧AdS →−
Y
n∧curl
I(t )
4π[x[
dS.
94 4 Vorticity
Similarly, p →−ρ
0
∂ϕ¡∂t onY, where ϕ =div(I(t )¡4π[x[) (see (4.3.3)). Hence,
−
Y
p(x. t ) dS →ρ
0
d
dt
Y
ϕndS = ρ
0
d
dt
Y
ndiv
I(t )
4π[x[
dS.
The righthand side of (4.4.2) can therefore be written
ρ
0
d
dt
Y
¸
−n ∧ curl
I(t )
4π[x[
÷ndiv
I(t )
4π[x[
¸
dS.
By the divergence theorem, the integral in this expression over the large, but
arbitrary surface Y can be replaced by an integration over the surface of a large
sphere [x[ = R, because
{curl curl −∇div}(I(t )¡4π[x[) = −∇
2
(I(t )¡4π[x[) ≡ 0 for [x[ >0.
On the sphere n = −x¡[x[, and the integrand equals I(t )¡4π R
2
; the integral is
therefore just equal to I(t ). Thus, (4.4.2) reduces to the desired representation
(4.4.1).
4.4.1 Bound Vorticity and the Added Mass
Consider the particular case of a rigid body accelerating without rotation at
velocity U(t ) in an otherwise unbounded, ideal incompressible ﬂuid in the
absence of vorticity (Fig. 4.4.1). We have seen previously (Section 3.8) that
force exerted on the ﬂuid can be written
F
i
= M
i j
dU
j
dt
.
Fig. 4.4.1.
4.4 Surface Force in Incompressible Flow 95
where M
i j
is the added mass tensor (3.8.6). This result will now be derived
from the integral formula (4.4.1).
There is no vorticity in the ﬂuid, but the slipping of the ideal ﬂow over S
generates a singular distribution (a vortex sheet) of bound vorticity that must
be used to evaluate the integral. To calculate the bound vorticity we need an
expression for the velocity everywhere in space, including the region occupied
by the body, where v ≡ U.
In the ﬂuid, we can take
v = U
j
∇ϕ
∗
j
.
Therefore, by introducing a control surface f (x. t ) = 0 that coincides with the
surface S of the body, with f >0 in the ﬂuid and f 0 within S, the required
formula for the velocity is
v = U
j
H( f )∇ϕ
∗
j
÷U
j
H(−f )∇x
j
.
and the vorticity is
ω = curl{U
j
H( f )∇ϕ
∗
j
÷U
j
H(−f )∇x
j
}
= −U
j
∇H ∧ ∇(x
j
−ϕ
∗
j
) (where ∇H = ∇H( f ) = −∇H(−f ))
≡ curl {U
j
(x
j
−ϕ
∗
j
)∇H}.
Equation (4.4.1) accordingly gives the force in the form
F = −m
0
dU
dt
÷
ρ
0
2
d
dt
x ∧ curl{U
j
(x
j
−ϕ
∗
j
)∇H} d
3
x.
Now the vector A=U
j
(x
j
−ϕ
∗
j
)∇H vanishes except on the surface S of the
body. The identities
x ∧ curl A = 2A ÷∇(x · A) −
∂
∂x
j
(x
j
A).
(·)∇H d
3
x =
S
(·) dS
(4.4.3)
therefore imply that
F
i
= −m
0
dU
i
dt
÷ρ
0
d
dt
U
j
(x
j
−ϕ
∗
j
)
∂ H
∂x
i
d
3
x
= −m
0
dU
i
dt
÷ρ
0
d
dt
S
U
j
(x
j
−ϕ
∗
j
)n
i
dS.
But
ρ
0
S
x
j
n
i
dS = ρ
0
Lδ
i j
≡ m
0
δ
i j
.
96 4 Vorticity
where L = volume contained by S, and
ρ
0
S
−ϕ
∗
j
n
i
dS = M
i j
.
where M
i j
is the added mass tensor (3.8.8). Therefore, force on ﬂuid in irrota
tional ﬂow ≡ F
i
= M
i j
dU
j
dt
.
4.4.2 Force Exerted on an Incompressible Fluid by a Moving Body
The integral in (4.4.1) deﬁning the value of dI¡dt can be transformed to remove
the strongdependence of the integrandonthe boundvorticityon S. This vorticity
is produced both by motion of S and by relative motion between S and the ﬂuid
induced by free vorticity in the ﬂow. Thus, any attempt to recast dI¡dt must
be strongly inﬂuenced by both the shape and motion of S. We consider only
the important special case of a body in translational motion without rotation at
velocity U(t ), and show that the i th component of the force F exerted on the
ﬂuid can also be written
F
i
=M
i j
dU
j
dt
−ρ
0
V
∇X
i
· ω∧v
rel
d
3
x −η
S
∇X
i
· ω∧dS. v
rel
= v −U.
(4.4.4)
where v
rel
is the ﬂuid velocity relative to the translational velocity of S, X
i
=
x
i
−ϕ
∗
i
(x. t ) is the Kirchhoff vector already encountered in the deﬁnition of the
compact Green’s function (Section 3.4), and M
i j
= M
j i
= −ρ
0
S
n
j
ϕ
∗
i
dS is
the added mass tensor. X
i
represents the velocity potential of an ideal ﬂow past
S that has unit speed in the i direction at large distances from S (it depends on
t because a ﬁxed coordinate system is being used).
The ﬁrst term on the right of (4.4.4) represents the inviscid component of the
force, associated with the added mass. The contribution from free vorticity is
furnished by the volume integral; the ﬁnal term arises from frictional effects on
S, which are relatively small at large Reynolds numbers. Now v
rel
= 0 on S,
and therefore the contribution to the volume integral from vorticity close to and
on S is negligible; indeed, even in the inviscid limit there is no contribution to
the integral from the surface vortex sheet forming the bound vorticity, because
∇X
i
and the relative Lamb vector ω ∧ v
rel
are orthogonal on S.
To derive this formula from (4.4.1) we introduce the usual control surface
f (x. t ) = 0 enclosing S, with f >0 in the outer ﬂuid region, multiply Crocco’s
homentropic momentum equation (4.2.5) by H ≡ H( f ), and take the curl of
the resulting equation. Using the formula
DH
Dt
≡
∂ H
∂t
÷v · ∇H = 0.
4.4 Surface Force in Incompressible Flow 97
and the noslip condition on S, we ﬁnd
∂
∂t
(Hω) = −
∂
∂t
(∇H ∧ U) −curl((∇H · U)U) −∇H ∧ ∇B
−curl(Hω ∧ v) −ν curl(H curl ω).
Then, because ω = 0 within S,
d
dt
x ∧ ω(x. t ) d
3
x =
d
dt
x ∧ (Hω) d
3
x =
x ∧
∂
∂t
(Hω) d
3
x
= −
x ∧
∂
∂t
(∇H ∧ U) d
3
x −
x ∧ curl ((∇H · U)U) d
3
x
−
x ∧ (∇H ∧ ∇B) d
3
x −
x ∧ curl (Hω ∧ v) d
3
x
−ν
x ∧ curl (H curl ω) d
3
x
= 2L
dU
dt
÷0 ÷2
S
B dS −2
V
ω ∧ v d
3
x −2ν
S
ω ∧ dS.
where the last line follows by use of the identities (4.4.3). Thus, adopting sufﬁx
notation,
dI
i
dt
≡
1
2
d
dt
(x ∧ ω)
i
d
3
x = L
dU
i
dt
÷
S
Bn
i
dS −
V
∇x
i
· (ω ∧ v) d
3
x
−ν
S
∇x
i
· ω∧dS. (4.4.5)
The surface integral
S
Bn
i
dS can be eliminated by recalling that
∇
2
ϕ
∗
i
= 0.
∂ϕ
∗
i
∂x
n
≡ n
j
∂ϕ
∗
i
∂x
j
= n
i
on S.
Then, because ∇ϕ
∗
i
∼ O(1¡[x[
3
) as [x[ → ∞, the divergence theorem shows
that
S
Bn
i
dS = −
V
div(∇ϕ
∗
i
B) d
3
x ≡ −
V
∇ϕ
∗
i
· ∇B d
3
x. Hence, using
Crocco’s equation (4.2.5)
S
Bn
i
dS =
V
div
ϕ
∗
i
∂v
∂t
d
3
x ÷
V
∇ϕ
∗
i
· ω ∧ v d
3
x
−ν
V
div(∇ϕ
∗
i
∧ ω) d
3
x.
The ﬁrst and last integrals on the right are transformed further by the divergence
98 4 Vorticity
theorem, for the ﬁrst
V
div
ϕ
∗
i
∂v
∂t
d
3
x =
M
i j
ρ
0
dU
j
dt
.
where M
i j
= M
j i
= −ρ
0
S
n
j
ϕ
∗
i
dS is the added mass coefﬁcient of (3.8.6).
For the last
−ν
V
div(∇ϕ
∗
i
∧ ω) d
3
x = ν
S
∇ϕ
∗
i
· ω ∧ dS.
Thus, substituting for
S
Bn
i
dS in (4.4.5), we ﬁnd
dI
i
dt
= L
dU
i
dt
÷
M
i j
ρ
0
dU
j
dt
−
V
∇X
i
· ω∧v d
3
x−ν
S
∇X
i
· ω∧dS. (4.4.6)
But the identity
∇X
i
· ω ∧ U = div(U(v · ∇X
i
) −v(U· ∇X
i
) −(v · U)∇X
i
)
implies that
V
∇X
i
· ω ∧ Ud
3
x = 0, so that (4.4.6) can also be written
dI
i
dt
= L
dU
i
dt
÷
M
i j
ρ
0
dU
j
dt
−
V
∇X
i
· ω∧v
rel
d
3
x−ν
S
∇X
i
· ω∧dS (4.4.7)
where v
rel
= v −U.
Equation (4.4.4) is now obtained by substituting from (4.4.7) into (4.4.1)
(recalling that m
0
= ρ
0
L).
4.4.3 Stokes Drag on a Sphere
The ﬁrst termon the righthand side of (4.4.4) is the force necessary to accelerate
the added mass of the body. The i th component of the viscous skin friction is
−η
S
(ω ∧ dS)
i
≡ −η
S
∇x
i
· ω ∧ dS. Thus, (because X
i
= x
i
−ϕ
∗
i
) the net
contribution of the normal pressure forces on S is represented in (4.4.4) by the
terms
−ρ
0
V
∇X
i
· ω ∧ v
rel
d
3
x ÷η
S
∇ϕ
∗
i
· ω ∧ dS.
The second, viscous component is comparable in magnitude to the skin friction,
and is produced by the pressure ﬁeld established by the surface shear stress.
The necessity for such a term is vividly illustrated by the Stokes drag on a
sphere. Let the sphere have radius a and translate at constant velocity U=(U. 0.
0). U >0, along the x
1
axis. At very small Reynolds numbers Re = aU¡ν <1
4.4 Surface Force in Incompressible Flow 99
the inertial terms ω ∧ v and ∇(
1
2
:
2
) can be discarded from Crocco’s equation
(4.2.5), which (for incompressible ﬂow) reduces to the creeping ﬂow equation
∇p = −η curl ω (4.4.8)
in a reference frame moving with the sphere. Both the pressure and the vorticity
therefore satisfy Laplace’s equation ∇
2
p = 0. ∇
2
ω = 0. By symmetry p must
vary linearly with η and U· x, and the condition that p should vanish at large
distances from the sphere supplies the dipole solution
p = Cη
U· x
[x[
3
≡ −Cη div
U
[x[
. [x[ >a.
where C is a constant.
Similarly, ω must be a linear function of U∧x: the identity curl curl(U¡
[x[) = grad div(U¡[x[) and Equation (4.4.8) imply that ω = C curl (U¡[x[) ≡
C(U ∧ x)¡[x[
3
([x[ >a). The value of C is most easily found by substituting
this expression for ω into the Biot–Savart formula (4.3.1) and evaluating the
righthand side at the centre x = 0 of the sphere, where v ≡ U. This yields
C = 3a¡2, and therefore
ω = curl
3aU
2[x[
. [x[ >a. (4.4.9)
The net force F
1
on the ﬂuid is in the x
1
direction, and is given by the ﬁnal
integral on the right of (4.4.4). It is equal in magnitude to D
s
÷ D
p
, where
D
s
and D
p
are the respective components of the Stokes drag on the sphere
produced by the skin friction and the viscous surface pressure. For the sphere
ϕ
∗
1
= −a
3
x
1
¡2[x[
3
, and we readily calculate F
1
= 6πηUa, and
D
s
=η
S
(ω ∧ dS)
1
=−4πηUa. D
p
=−η
S
∇ϕ
∗
1
· ω∧dS=−2πηUa.
The pressure drag is therefore equal to half the skinfriction drag.
This interpretation of D
p
as the component of drag attributable to the normal
pressure forces on the sphere can be conﬁrmed directly using the creeping ﬂow
approximation (4.4.8). Because n
1
= n· ∇ϕ
∗
1
on S we ﬁnd, using the divergence
theorem,
D
p
= −
S
pn
1
dS ≡ −
S
p∇ϕ
∗
1
· dS =
V
∇p · ∇ϕ
∗
1
d
3
x
= −η
V
curl ω· ∇ϕ
∗
1
d
3
x = −η
S
∇ϕ
∗
1
· ω ∧ dS.
100 4 Vorticity
where (4.4.8) and the identity div(A ∧ B) = curl A· B − A· curl B have been
used on the second line.
4.5 The Complex Potential
The remainder of this chapter is devoted to a brief outline of the complex
potential representation of twodimensional, incompressible ﬂows and its ap
plication to determine the equation of motion of a line vortex in such ﬂows. The
results will be applied in later chapters to investigate simple models of sound
production by vortices interacting with surfaces.
4.5.1 Laplace’s Equation in Two Dimensions
Suppose that
n(z) = ϕ(x. y) ÷i ψ(x. y). z = x ÷i y
is regular (analytic) in a region D of the zplane. The real and imaginary parts
ϕ(x. y) and ψ(x. y) are solutions of Laplace’s equation.
∂
2
ϕ
∂x
2
÷
∂
2
ϕ
∂y
2
= 0.
∂
2
ψ
∂x
2
÷
∂
2
ψ
∂y
2
= 0 in D.
Let f (z) be regular in D, and deﬁne a (conformal) transformation Z = f (z)
of D into a region D
/
in the plane of Z = X ÷i Y. Let W(Z) be regular in D
/
with real and imaginary parts +(X. Y). +(X. Y). Then,
∂
2
+
∂ X
2
÷
∂
2
+
∂Y
2
= 0.
∂
2
+
∂ X
2
÷
∂
2
+
∂Y
2
= 0 in D
/
.
The transformation Z = f (z) permits us to deﬁne a corresponding function
n(z) ≡ ϕ(x. y) ÷i ψ(x. y) = W( f (z)), which is regular in D, with derivative
n
/
(z) = f
/
(z)W
/
( f (z)). For corresponding points in D and D
/
we have
ϕ(x. y) = +(X(x. y). Y(x. y)). ψ(x. y) = +(X(x. y). Y(x. y)).
In other words, the solutions + and + of Laplace’s equation in D
/
are also
solutions of Laplace’s equation in D.
These results have the following signiﬁcance. The solution of Laplace’s equa
tion within a given twodimensional bounded region D is equivalent to the so
lution of Laplace’s equation within the transformed region D
/
. If it is possible to
solve the latter problem, the solution to the original problem in D can be found
4.5 The Complex Potential 101
by transforming back to the z plane. Difﬁculties may arise at isolated points
where f
/
(z) =0 and at points where f (z) ceases to be regular, but these can usu
ally be dealt with by careful examination of the behavior of the transformation
near such points.
4.5.2 Hydrodynamics in Two Dimensions
Irrotational motion of an ideal, incompressible ﬂuid in planes parallel to the xy
plane can be investigated by introducing the complex potential n(z) =ϕ(x. y) ÷
i ψ(x. y). The velocity
v = ∇ϕ = (∂ϕ¡∂x. ∂ϕ¡∂y).
The function ψ is called the stream function. For steady motion the velocity at
(x. y) does not change with time, and the ﬂuid particles travel along a ﬁxed sys
temof streamlines each of which is a member of the family of curves ψ(x. y) =
constant.
Both ϕ(x. y) and ψ(x. y) are solutions of Laplace’s equation and satisfy the
Cauchy–Riemann equations:
∂ϕ
∂x
=
∂ψ
∂y
.
∂ϕ
∂y
= −
∂ψ
∂x
.
which imply that ∇ϕ · ∇ψ = 0, i.e., that the streamlines intersect the equipo
tentials ϕ = constant at right angles. If v = (u. :), then the complex velocity
n
/
(z) =
∂ϕ
∂x
−i
∂ϕ
∂y
≡ u −i :
is also regular.
The fact that n(z) is a regular function of z can greatly simplify the solution
of many problems. This will be illustrated by consideration of two methods
based on the theory of complex variables.
Method 1 The real and imaginary parts of every regular function n(z) deter
mine the velocity potential and stream function of a possible ﬂow. A catalog
of ﬂows can therefore be constructed by studying the properties of arbitrarily
selected n(z).
Example 1 n = Uz. U = real constant:
ϕ = Ux. ψ = Uy. Thus, v = (U. 0).
102 4 Vorticity
The motion is uniform at speed U along streamlines parallel to the x direction.
Example 2
n = U
z ÷
a
2
z
. U = real constant, a >0. [z[ >a. (4.5.1)
At large distances from the origin n →Uz, and the motion becomes uniform
at speed U parallel to the x axis. In terms of the polar form z = re
i θ
,
n = U
re
i θ
÷
a
2
r
e
−i θ
. Thus, ϕ = U cos θ
r ÷
a
2
r
.
The radial component of velocity
∂ϕ
∂r
= U cos θ
1 −
a
2
r
2
vanishes at r = a. The motion therefore represents steady ﬂow in the x direc
tion past a rigid cylinder of radius a with centre at the origin (Fig. 4.5.1; c.f.,
Section 3.6).
Example 3
n = −iU
z −
a
2
z
. U = real constant. [z[ >a >0.
ϕ = U sin θ
r ÷
a
2
r
. (4.5.2)
describes potential ﬂow in the y direction past a rigid cylinder of radius a with
center at the origin.
Fig. 4.5.1.
4.5 The Complex Potential 103
Example 4 The function
n =
1
2π
ln z.
ϕ =
1
2π
lnr. ψ =
θ
2π
. z = re
i θ
.
is regular except at z = 0. The ﬂow is radially outward from the origin along
streamlines θ =constant, at speed ∂ϕ¡∂r = 1¡2πr. The origin is a singularity
of the ﬂow where ﬂuid is created at a rate equal to
C
∇ϕ · nds, where C is any
simple closed curve enclosing the origin with outward normal n, and ds is the
element of arc length on C. In particular, taking C to be a circle of radius r,
C
∇ϕ · nds =
2π
0
∂ϕ
∂r
r dθ = 1.
The origin is therefore a simple source of unit strength. When the source is
situated at z
0
= x
0
÷i y
0
n =
1
2π
ln(z−z
0
).
ϕ =
1
2π
ln [z − z
0
[ =
1
2π
ln
(x −x
0
)
2
÷(y −y
0
)
2
.
Example 5 The function
n =
−i I
2π
ln z.
ϕ =
Iθ
2π
. ψ = −
I
2π
lnr. z = re
i θ
.
is regular except at z =0, and describes the irrotational ﬂow outside a line
vortex of strength I concentrated at z =0. The streamlines are circles centered
at z =0, and the ﬂow speed is ∂ϕ¡r∂θ =I¡2πr in the anticlockwise direction
(for I >0). The circulation
C
v · dx =I, where C is any contour encircling
the vortex once, and the contour is traversed in the positive direction (with the
interior on the left). When the vortex is at z
0
= x
0
÷i y
0
n =
−i I
2π
ln(z − z
0
).
Example 6 The function
n =
1
2π
(ln(z − z
0
) ÷ln(z − z
∗
0
)).
ϕ =
1
2π
(lnr
1
÷lnr
2
)
.
represents the ﬂow produced by two unit point sources located at z
0
= x
0
÷i y
0
and z
∗
0
= x
0
−i y
0
(Fig. 4.5.2). The motion is symmetric with respect to the x
axis, and ∂ϕ¡∂y = 0 on y = 0. Therefore, in the region y >0 the potential also
describes the ﬂow produced by a point source at z
0
adjacent to a rigid wall at
y = 0 (the presence of the wall is said to be accounted for by an image source).
104 4 Vorticity
Fig. 4.5.2.
Example 7 The function
n =
−i I
2π
ln(z − z
0
) ÷
i I
2π
ln(z − z
∗
0
).
represents the ﬂowproducedbytwoline vortices of circulations ±Irespectively
at z
0
= x
0
÷i y
0
. z
∗
0
= x
0
−i y
0
(Fig. 4.5.3). The streamfunction ψ = Imn = 0
onthe x axis, whichis therefore a streamline of the ﬂow, onwhich∂ϕ¡∂y = 0. In
the region y >0 the potential describes the ﬂowproduced by a vortex of strength
I at z
0
adjacent to a rigid wall at y = 0 (which is accounted for by an equal and
opposite image vortex). Each vortex translates parallel to the wall at speed u =
I¡4πy
0
determined by the velocity potential of its image. The mean value of the
local rotational ﬂow produced by the selfpotential of each vortex (Example 5)
vanishes on the vortex axis, and cannot therefore cause it to translate.
Method 2 The ﬂow past a system of rigid boundaries in the z plane is repre
sented by means of a conformal transformation Z = f (z) by an equivalent ﬂow
in the Z plane. The transformation is usually chosen to simplify the boundary
Fig. 4.5.3.
4.5 The Complex Potential 105
conditions, thereby permitting the solution in the Z plane to be found in a rela
tively straightforward manner. Point source and vortex singularities of the ﬂow
are preserved under the transformation. Indeed, if Z = Z
0
is the image of a
vortex of strength I at z = z
0
, the complex potential in the neighborhood of
Z
0
(where Z − Z
0
≈ f
/
(z
0
)(z − z
0
)) is determined by
W(Z) = n(z) =
−i I
2π
ln(z − z
0
) ÷terms ﬁnite at z
0
=
−i I
2π
ln
Z − Z
0
f
/
(z
0
)
÷terms ﬁnite at Z
0
=
−i I
2π
ln(Z − Z
0
) ÷terms ﬁnite at Z
0
.
The vortex in the z plane therefore maps into an equal vortex at the image point
in the Z plane.
Example 8 Derive the following formula for the velocity potential of irrota
tional ﬂow around the edge of the rigid halfplane x 0. y = 0 in terms of
polar coordinates (r. θ):
ϕ = α
√
r sin
θ
2
. α =a real constant.
and plot the streamlines.
The transformation Z =i
√
z maps the z plane cut along the negative real
axis (so that −π arg z π) onto the upper half of the Z plane. The complex
potential of ﬂowin the positive X direction parallel to the boundary Y =0 in the
Z plane corresponds to ﬂow around the edge of the halfplane in the clockwise
sense, and has the general representation W = UZ, where U is real. In the z
plane this becomes
n = iU
√
z ≡ −U
√
r sin
θ
2
÷iU
√
r cos
θ
2
. −π θ π.
The polar representation of the velocity is therefore
v = (:
r
. :
θ
) =
∂ϕ
∂r
.
1
r
∂ϕ
∂θ
=
−U
2
√
r
sin
θ
2
. cos
θ
2
.
This satisﬁes the rigid wall condition on the halfplane because the component
of velocity normal to the wall is :
θ
, which vanishes at θ = ±π. The streamlines
of the ﬂow are the parabolas
√
r cos
θ
2
= constant. i.e., y = ±2β
1 −
x
β
.
106 4 Vorticity
Fig. 4.5.4.
where x β. β being a positive constant, as shown in Fig. 4.5.4. When U >0
ﬂuid particles travel along the parabolic streamlines around the edge in the
clockwise direction. The streamline for β =0 corresponds to the upper and
lower surfaces of the halfplane, which maps into the streamline Y =0 on the
surface of the wall in the Z plane. The ﬂowvelocity becomes inﬁnite like 1¡
√
r
as r →0 at the sharp edge.
4.6 Motion of a Line Vortex
In twodimensional incompressible, inviscid ﬂow in planes parallel to x
3
= 0
the vortex lines are all parallel to the x
3
direction, and the vorticity equation
(4.2.7) reduces to
Dω
3
Dt
= 0.
A line vortex is therefore convected without change at the local velocity at its
4.6 Motion of a Line Vortex 107
core. For a vortex of strength I at z = z
0
(t ) in the plane of z = x
1
÷ i x
2
,
the velocity becomes inﬁnite as the core is approached because of the singular
velocity induced by its selfpotential
−
i I
2π
ln(z − z
0
).
But the rotational ﬂow around the core induced by the vortex cannot induce
motion in itself, and this potential must be removed from the complex potential
n(z) before calculating the convection velocity of the vortex.
In applications the complex potential n(z) usually arises in the form
n(z) = −
i I
2π
ln(ζ (z) −ζ (z
0
)) ÷ F(z). (4.6.1)
where ζ (z). F(z) are regular functions of z in the neighborhood of the vortex
core at z = z
0
. In particular, when [z − z
0
[ is small we have
ζ (z) = ζ (z
0
) ÷(z − z
0
)ζ
/
(z
0
) ÷
(z − z
0
)
2
2
ζ
//
(z
0
) ÷· · · .
where the primes denote differentiation with respect to z. Thus, subtracting the
selfpotential from n(z) we ﬁnd, near the vortex,
W(z) = n(z) ÷
i I
2π
ln(z − z
0
)
= −
i I
2π
ln(ζ (z) −ζ (z
0
)) ÷
i I
2π
ln(z − z
0
) ÷ F(z)
≈ −
i I
2π
ln
¸
ζ
/
(z
0
) ÷
1
2
ζ
//
(z
0
)(z − z
0
)
¸
÷ F(z). (4.6.2)
The complex velocity of the vortex is W
/
(z
0
) ≡ {W
/
(z)}
z=z
0
, i.e.,
dz
∗
0
dt
≡
dx
01
dt
−i
dx
02
dt
= W
/
(z
0
).
Using (4.6.2) this becomes
dx
01
dt
−i
dx
02
dt
= −
i Iζ
//
(z
0
)
4πζ
/
(z
0
)
÷ F
/
(z
0
). (4.6.3)
The real and imaginary parts of this equation supply two nonlinear ﬁrstorder
ordinary differential equations for the position (x
01
(t ). x
02
(t )) of the vortex at
time t .
108 4 Vorticity
4.6.1 Numerical Integration of the Vortex Path Equation
In most cases it is necessary to integrate equation (4.6.3) numerically. The
time and space variables should ﬁrst be nondimensionalized with respect to
convenient time and length scales deﬁned by the problem(several examples are
discussed in Chapter 8). The integration is started from a prescribed point on
the trajectory through which the vortex is required to pass.
Let us consider integration by means of a fourthorder Runge–Kutta algo
rithm. Write the equation of motion (4.6.3) in the form
dz
0
dt
= f
∗
(z
0
). where f (z
0
) = −
i Iζ
//
(z
0
)
4πζ
/
(z
0
)
÷ F
/
(z
0
).
and let h be a suitably small integration time step (which need not be constant).
Assume that at time t
n
the vortex is at z
0
(t
n
) = z
n
0
. To determine the complex
position z
n÷1
0
at time t
n÷1
= t
n
÷h, we evaluate
k
1
= h f
∗
z
n
0
. k
2
= h f
∗
z
n
0
÷
1
2
k
1
. k
3
= h f
∗
z
n
0
÷
1
2
k
2
.
k
4
= h f
∗
z
n
0
÷k
3
.
and then ﬁnd
z
n÷1
0
= z
n
0
÷
1
6
(k
1
÷2k
2
÷2k
3
÷k
4
).
Example 1 Calculate the trajectory of a line vortex of strength I >0 adjacent to
a rigid halfplane lying along the negative real axis (x
1
0. x
2
= 0; Fig. 4.6.1a).
The transformation
ζ = i
√
z. z = x
1
÷i x
2
. −π arg z π. (4.6.4)
maps the ﬂuid region −π arg z π into the upper half Imζ >0 of the ζ plane
(Fig. 4.6.1b). Let the vortex at z
0
(t ) map into a vortex at ζ = ζ
0
(t ). The velocity
potential n(ζ ) of the motion in the ζ plane is found by introducing an image
vortex of strength −I at ζ = ζ
∗
0
(t ), as described in Example 7 of Section 4.5,
in which case
n =
−i I
2π
ln(ζ −ζ
0
) ÷
i I
2π
ln(ζ −ζ
∗
0
).
In the z plane this becomes
n(z) = −
i I
2π
ln(ζ (z) −ζ (z
0
)) ÷
i I
2π
ln(ζ (z) −ζ
∗
(z
0
)).
4.6 Motion of a Line Vortex 109
Fig. 4.6.1.
which is of the form (4.6.1). Hence, the equation of motion (4.6.3) becomes
dx
01
dt
−i
dx
02
dt
=
i I
8πz
0
÷
i I
4π
√
z
0
[
√
z
0
÷(
√
z
0
)
∗
]
.
This can be integrated in closed form. Let z
0
=re
i θ
. Then the real and imag
inary parts of the equation are
dx
01
dt
≡ cos θ
dr
dt
−r sin θ
dθ
dt
=
I
8πr
sin θ ÷tan
θ
2
.
dx
02
dt
≡ sin θ
dr
dt
÷r cos θ
dθ
dt
= −
I
8πr
(cos θ ÷1).
Therefore,
dr
dt
= −
I
8πr
tan
θ
2
.
dθ
dt
= −
I
4πr
2
(4.6.5)
that is,
r
dθ
dr
= 2 cot
1
2
θ
110 4 Vorticity
Fig. 4.6.2.
Thus,
r = ¹ sec
1
2
θ. ¹ = constant. (4.6.6)
This is the polar equation of the trajectory plotted in Fig. 4.6.2. The constant
length ¹ is equal to the distance of closest approach of the vortex to the edge
of the halfplane, which occurs at θ =0. Substituting for r in the second of
equations (4.6.5), we ﬁnd
sec
2
1
2
θ
dθ
dt
= −
I
4π¹
2
. Thus, θ = 2 tan
−1
−
It
8π¹
2
.
where time is measured from the instant at which θ = 0. The dependence of r
on t is now obtained by substituting into (4.6.6).
Collecting together these results we have
r = ¹
1 ÷
Ut
¹
2
. θ =2 tan
−1
−
Ut
¹
;
x
01
¹
=
1 −(Ut ¡¹)
2
1 ÷(Ut ¡¹)
2
.
x
02
¹
=
−2Ut ¡¹
1 ÷(Ut ¡¹)
2
. (4.6.7)
where U =
I
8π¹
. Thus, (for I >0) the vortex starts above the halfplane at
t =−∞ at x
01
=−∞, x
02
=2¹ and translates towards the edge, initially at
speed U parallel to the plane. It crosses the x
1
axis at t =0 at x
01
=¹, and
proceeds along a symmetrical path below the halfplane.
4.6 Motion of a Line Vortex 111
Fig. 4.6.3.
Example 2: Vortexmotionoutside acylinder AvortexIis locatedat z
0
=re
i θ
outside a rigid cylinder of radius a(r) with center at the origin (Fig. 4.6.3).
There is no net circulation around the cylinder. The complex potential is ob
tained by placing an image vortex −I at the inverse point z = a
2
¡z
∗
0
together
with a vortex ÷I at the center of the cylinder. The two interior vortices ensure
that the total circulation around the cylinder vanishes. Then
n(z) = −
i I
2π
ln(z − z
0
) ÷
i I
2π
ln
z −
a
2
z
∗
0
−
i I
2π
ln z.
The ﬁrst term on the right is the selfpotential of the vortex (in the notation of
(4.6.1) ζ ≡ z), so that the equation of motion of the vortex is
dz
∗
0
dt
=
i Iz
∗
0
2π(r
2
−a
2
)
−
i I
2πz
0
≡
i Ia
2
2πz
0
(r
2
−a
2
)
.
By multiplying by z
0
and adding the complex conjugate equation we see that
r = constant, and
r
d
dt
e
−i θ
≡ −ire
−i θ
dθ
dt
=
i Ia
2
e
−i θ
2πr(r
2
−a
2
)
.
Therefore,
dθ
dt
=
−Ia
2
2πr
2
(r
2
−a
2
)
.
and (for I >0) the vortex trajectory is a circle traversed in the clockwise direc
tion at speed
:
0
=
Ia
2
2πr(r
2
−a
2
)
. (4.6.8)
112 4 Vorticity
Problems 4
1. Show that in inviscid, homentropic ﬂow (where div v ,= 0) the vorticity
equation (4.2.7) takes the form
D
Dt
ω
ρ
=
ω
ρ
· ∇
v.
2. Use the relation
(y
i
ω
j
(y. t ) ÷ y
j
ω
i
(y. t )) d
3
y = 0 to show that
1
4π[x[
3
(x · y)ω(y. t ) d
3
y = ∇
1
4π[x[
∧
¸
1
2
y ∧ ω(y. t ) d
3
y
¸
.
Deduce the formulae (4.3.3).
3. Calculate the added mass coefﬁcients M
i j
for an inﬁnite, rigid strip of width
2a.
4. Calculate the added mass coefﬁcients M
i j
for an inﬁnite, rigid cylinder of
radius a.
5. Calculate the unsteady lift and drag exerted on a rigid circular cylinder of
radius a produced by a parallel line vortex of circulation I in the presence
of a uniform mean ﬂow normal to the cylinder. Assume the motion is ideal
and that the net circulation around the cylinder vanishes.
6. Repeat Question 5 under the assumption that the vortex is convected solely
by the mean ﬂow (i.e., when the induced component of the motion of I
produced by image vortices in the cylinder is neglected).
7. A rigid sphere of radius a translates at constant velocity U=(U. 0. 0).
U >0, along the x
1
axis. Use the creeping ﬂow approximation ω=curl
(3aU¡2[x[), where the coordinate origin is taken at the center of the sphere,
to deduce the Stokes drag formula D = 6πηUa.
8. A gas bubble in water is set into translational motion at velocity U(t ) by
sound whose wavelength greatly exceeds the bubble radius. If the acoustic
particle velocity near the bubble would equal V(t ) in the absence of the
bubble, show that U=3V when the mass of the air within the bubble is
neglected.
9. Calculate the path of a line vortex of strength I that is parallel to a rigid strip
occupying −a x
1
a. x
2
= 0. −∞x
3
∞. Assume the ﬂuid is at rest
at inﬁnity and that there is no net circulation around the strip. Determine
the unsteady force on the strip.
10. Calculate the path of a line vortex of strength I that is parallel to a rigid
elliptic cylinder of semimajor and minor axes respectively equal to a and
b. Assume the ﬂuid is at rest at inﬁnity and that there is no net circulation
around the cylinder.
Problems 4 113
11. A line vortex of strength I is adjacent to a rigid rightangle corner whose
sides lie along the positive x
1
and x
2
axes, the vortex being parallel to
the edge of the corner. Show that the vortex traverses a path with polar
representation r sin 2θ = constant.
12. Calculate the trajectories of a vortex pair consisting of two parallel line
vortices of strengths ±I moving under their mutual induction towards a
rigid plane parallel to the line of centers of the vortices.
13. Calculate the trajectory of a line vortex of strength I adjacent to the rigid
halfplane x
1
0. x
2
= 0 in the presence of a uniform mean ﬂow at speed
U in the positive x
1
direction.
14. Showthat the transformationζ =
z
2
¡a
2
÷1. a >0maps the upper z plane
cut bya thinrigidbarrier alongthe imaginaryaxis between z = 0and z = i a
onto the upper ζ plane. Deduce that a line vortex I at z = z
0
(t ) traverses a
path determined by the equation
dz
∗
0
dt
= −
i I
4π
¸
a
2
z
0
z
2
0
÷a
2
−
2z
0
¸
z
2
0
÷a
2
−
z
2
0
÷a
2
¸
¸
.
provided the ﬂuid is at rest at inﬁnity.
5
Vortex Sound
5.1 The Role of Vorticity in Lighthill’s Theory
At low Mach numbers in unbounded, homentropic ﬂow the value of Lighthill’s
quadrupole source (2.2.2) can be approximated by means of the Biot–Savart
induction formula (4.3.1):
T
i j
≈ ρ
0
u
i
u
j
. u(x. t ) = curl
ω(y. t ) d
3
y
4π[x −y[
. (5.1.1)
To examine this in more detail, consider the acoustically compact eddy of
Fig. 2.2.1 consisting of vorticity of characteristic length ¹, and take the coordi
nate origin within the eddy. Put
v = u ÷∇ϕ;
u involves the whole incompressible component of velocity, and u ∼ O(1¡[x[
3
)
as [x[ →∞(see Section 4.3). Because div u=0, the scalar potential ϕ describes
compressible motions, and the continuity equation becomes
∇
2
ϕ ÷
1
ρ
Dρ
Dt
= 0.
But p−p
0
∼ ρ
0
u
2
in the eddy, where the characteristic frequency ∼u¡¹. Thus,
Dp
Dt
∼
ρ
0
u
3
¹
. Hence,
1
ρ
Dρ
Dt
=
1
ρc
2
Dp
Dt
∼
u
¹
M
2
. M =
u
c
0
.
Hence, in order of magnitude
∇ϕ = O(uM
2
) within the eddy, where [x[ ∼ ¹. (5.1.2)
114
5.1 The Role of Vorticity in Lighthill’s Theory 115
Now write
∂
2
(u
i
u
j
)
∂x
i
∂x
j
= div(ω ∧ u) ÷∇
2
1
2
u
2
(5.1.3)
and express the solution p(x. t ) = c
2
0
(ρ −ρ
0
) of Lighthill’s equation given by
(2.2.1) in the form
p(x. t ) = p
1
(x. t ) ÷ p
2
(x. t ).
where, using (1.9.6) and (1.9.8) as [x[ →∞,
p
1
(x. t ) =
−ρ
0
x
i
4πc
0
[x[
2
∂
∂t
(ω ∧ u)
i
y. t −
[x[
c
0
÷
x · y
c
0
[x[
d
3
y. (5.1.4)
p
2
(x. t ) =
ρ
0
4πc
2
0
[x[
∂
2
∂t
2
1
2
u
2
y. t −
[x[
c
0
÷
x · y
c
0
[x[
d
3
y. (5.1.5)
When retarded time variations x · y¡c
0
[x[ within the eddy are neglected the
identity(5.1.3) andthe divergence theoremimplythat
ω∧ud
3
y ≡ 0, because
u ∼ O(1¡[y[
3
) as [y[ →∞. To estimate the value of the integral in (5.1.4) it is
therefore necessary to expand the integrand to the next higher approximation
in the retarded time:
(ω ∧ u)
y. t −
[x[
c
0
÷
x · y
c
0
[x[
= (ω ∧ u)
y. t −
[x[
c
0
÷
x · y
c
0
[x[
∂
∂t
¸
(ω ∧ u)
y. t −
[x[
c
0
¸
÷· · · .
We now ﬁnd
p
1
(x. t ) ≈
−ρ
0
x
i
x
j
4πc
2
0
[x[
3
∂
2
∂t
2
y
i
(ω ∧ u)
j
y. t −
[x[
c
0
d
3
y ∼
¹
[x[
ρ
0
u
2
M
2
.
[x[ →∞. (5.1.6)
The order of magnitude of p
2
(x. t ) is estimated by using the momentum
equation (4.2.3). Because div v ∼ O(M
2
) within the source region, we can write
∂u
∂t
÷∇
dp
ρ
÷
1
2
:
2
÷
∂ϕ
∂t
= −ω ∧ u −ω ∧ ∇ϕ −ν curl ω.
Take the scalar product with u
∂
∂t
1
2
u
2
÷div
¸
u
dp
ρ
÷
1
2
:
2
÷
∂ϕ
∂t
¸
=−u· ω ∧ ∇ϕ −νu· curl ω
=−u· ω ∧ ∇ϕ ÷ν(div (u ∧ ω) −ω
2
)
116 5 Vortex Sound
and integrate over the whole of space. The contributions from the divergence
terms vanish because u(
dp¡ρ ÷
1
2
:
2
÷ ∂ϕ¡∂t ) tends to zero at least as fast
as 1¡[y[
3
as [y[ →∞, where also ω = 0. Hence, using the estimate (5.1.2)
∂
∂t
1
2
u
2
(y. t ) d
3
y = −
(u· ω ∧ ∇ϕ ÷νω
2
)(y. t ) d
3
y ∼ ¹
2
u
3
M
2
÷
¹
2
u
3
Re
.
(5.1.7)
where Re =u¹¡ν typically exceeds 10
4
in turbulent ﬂow. The two terms on
the righthand side nominally represent the dissipation of the turbulent motions
respectively by acoustic radiation and by viscous damping. (We have already
seen, however, in Chapter 2, Equation (2.2.5) that a more accurate estimate of
the radiation damping is ¹
2
u
3
M
5
.)
Thus, when retarded time variations are neglected in (5.1.5), we ﬁnd
p
2
(x. t ) ∼
¹
[x[
ρ
0
u
2
M
4
÷
¹
[x[
ρ
0
u
2
M
2
Re
.
and therefore that p
2
< p
1
in turbulent ﬂow where M <1 and Re ·1.
We conclude that the component
div(ρ
0
ω ∧ v) of the Lighthill quadrupole
∂
2
(ρ
0
:
i
:
j
)
∂x
i
∂x
j
is principal source of sound at low Mach numbers.
5.2 The Equation of Vortex Sound
Lighthill’s equation (2.1.12) can be recast in a form that emphasizes the promi
nent rˆ ole of vorticity in the production of sound by taking the total enthalpy
B =
dp
ρ
÷
1
2
:
2
as the independent acoustic variable, in place of Lighthill’s c
2
0
(ρ−ρ
0
). The total
enthalpy occurs naturally in Crocco’s form (4.2.3) of the momentum equation.
Inthe followingwe shall actuallyuse the Approximation(4.2.5) of this equation,
in which the viscous term
4
3
ν∇(div v) is neglected. Indeed, the principal effect of
this term is to attenuate the sound once it has been generated and is propagating
to a distant observer in the sourcefree region of the ﬂow. This attenuation can be
signiﬁcant in applications, but is of no particular interest when studying sound
generation mechanisms. All viscous stresses can be ignored in a high Reynolds
number source ﬂow except possibly within surface boundary layers on bodies
5.2 The Equation of Vortex Sound 117
immersed in the ﬂow. But surface friction is dominated by the vorticity term
−ν curl ω, which is retained in (4.2.5).
In irrotational ﬂow Crocco’s equation (4.2.5) reduces to
∂v
∂t
= −∇B.
In other words,
B = −
∂ϕ
∂t
in regions where ω = 0. (5.2.1)
where ϕ(x. t ) is the velocity potential that determines the whole motion in the ir
rotational regions of the ﬂuid. B is therefore constant in steady irrotational ﬂow,
and at large distances from the acoustic sources perturbations in B represent
acoustic waves.
If the mean ﬂow is at rest in the far ﬁeld, the acoustic pressure is given by
p = ρ
0
B ≡ −ρ
0
∂ϕ
∂t
. (5.2.2)
To calculate the pressure in terms of B elsewhere in the ﬂow, we use the
deﬁnition
dp
ρ
= B −
1
2
:
2
.
Differentiatingwithrespect totime andusingCrocco’s equation(4.2.5), we have
1
ρ
∂p
∂t
=
∂ B
∂t
−v ·
∂v
∂t
=
∂ B
∂t
−v · (−∇B −ω ∧ v −ν curl ω)
=
DB
Dt
÷νv · curl ω.
The small viscous correction can be ignored in high Reynolds number source
ﬂows, where p and B can be taken to be related by
1
ρ
∂p
∂t
=
DB
Dt
. (5.2.3)
5.2.1 Reformulation of Lighthill’s Equation
Multiply Crocco’s equation (4.2.5) by the density ρ and take the divergence
div
ρ
∂v
∂t
÷∇ · (ρ∇B) = −div(ρω ∧ v). (5.2.4)
118 5 Vortex Sound
The ﬁrst term on the left is expressed in terms of B by using the continuity
equation in the form
div v = −
1
ρ
Dρ
Dt
.
and writing
div
ρ
∂v
∂t
= ∇ρ ·
∂v
∂t
÷ρ
∂
∂t
div v
= ∇ρ ·
∂v
∂t
−ρ
∂
∂t
1
ρ
Dρ
Dt
= ∇ρ ·
∂v
∂t
−ρ
∂
∂t
1
ρ
∂ρ
∂t
−
∂v
∂t
· ∇ρ −ρv · ∇
1
ρ
∂ρ
∂t
= −ρ
D
Dt
1
ρ
∂ρ
∂t
= −ρ
D
Dt
1
ρc
2
∂p
∂t
= −ρ
D
Dt
1
c
2
DB
Dt
.
where Equation (5.2.3) has been used on the last line. Substituting into (5.2.4)
and dividing by ρ, we obtain the desired vortex sound equation for homen
tropic ﬂow
D
Dt
1
c
2
D
Dt
−
1
ρ
∇ · (ρ∇)
B =
1
ρ
div(ρω ∧ v). (5.2.5)
The vortex source on the righthand side vanishes in irrotational regions; if
ω = 0 everywhere, and if there are no moving boundaries, the total enthalpy B
is constant, and there are no sound waves propagating in the ﬂuid. If acoustic
waves cannot enter from inﬁnity, it follows that the (homentropic) ﬂow can
generate sound only if moving vorticity is present, and the righthand side
of (5.2.5) may be identiﬁed as the analytical representation of the acoustic
sources. The differential operator on the left describes propagation of the sound
through the nonuniform ﬂow; as in the case of Lighthill’s equation, when the
source region is very extensive it will not normally be permissible to neglect
the interaction of the sound with the vorticity through which it propagates.
5.2.2 Sound Waves in Irrotational Mean Flow
Let an irrotational mean ﬂow be deﬁned by the velocity potential ϕ
0
(x), with
mean velocity U = ∇ϕ
0
. In an unbounded ﬂuid U =constant; the mean velocity
5.2 The Equation of Vortex Sound 119
can vary with position only if the ﬂuid is bounded, either internally by an airfoil,
say, or externally by the walls of a duct of variable cross section (ϕ
0
(x) can be
multiplevalued if the boundaries are multiply connected, but the mean velocity
is always singlevalued).
Consider an irrotational disturbance ϕ
/
(x. t ), and set
ϕ(x. t ) = ϕ
0
(x) ÷ϕ
/
(x. t ).
It can be shown that the general, nonlinear equation satisﬁed by ϕ is
1
c
2
∂
2
ϕ
∂t
2
÷
1
c
2
D
Dt
1
2
(∇ϕ)
2
÷
1
c
2
∂
∂t
1
2
(∇ϕ)
2
−∇
2
ϕ = 0.
The linearized version of this equation describes the propagation of small ampli
tude sound waves determined by ϕ
/
(x. t ). However, when ω = 0 the linearized
equation for B = −∂ϕ
/
¡∂t ≡ −˙ ϕ is more easily derived from (5.2.5), which
becomes
D
Dt
1
c
2
D
Dt
−
1
ρ
∇ · (ρ∇)
˙ ϕ = 0. (5.2.6)
The coefﬁcients of the differential operators in this equation are functions of
both mean and perturbation quantities, but the linearized equation is obtained
merely by replacing these coefﬁcients by their values in the absence of the
sound. In homentropic ﬂowthe mean density and sound speed can be expressed
in terms of the variable mean velocity U(x), and
D
Dt
≈
∂
∂t
÷U· ∇ .
Furthermore, because the mean ﬂow does not depend on time, we can take the
perturbation potential ϕ
/
, rather than ˙ ϕ, as the acoustic variable. The linearized
equation then becomes
¸
∂
∂t
÷U· ∇
¸
1
c
2
∂
∂t
÷U· ∇
¸
−
1
ρ
∇ · (ρ∇)
¸
ϕ
/
= 0.
where c ≡c(x) and ρ ≡ρ(x) are the local sound speed and density in the steady
ﬂow.
5.2.3 Vortex Sound at Low Mach Numbers
When the characteristic Mach number M is small the local mean values of the
density and sound speed are related to their uniform respective values ρ
0
and
120 5 Vortex Sound
c
0
at inﬁnity by relations of the form
c
c
0
∼ 1 ÷O(M
2
).
ρ
ρ
0
∼ 1 ÷O(M
2
).
The vortex sound equation (5.2.5) can therefore be simpliﬁed by (a) taking
c =c
0
, and ρ =ρ
0
, and (b) by neglecting nonlinear effects of propagation and
the scatteringof soundbythe vorticity. The productionof soundis thengoverned
by the simpler equation
1
c
2
0
∂
2
∂t
2
−∇
2
B = div(ω ∧ v). (5.2.7)
and in the far ﬁeld the acoustic pressure is given by the linearized approximation
p(x. t ) ≈ ρ
0
B(x. t ). (5.2.8)
5.2.4 Example 1 (Powell 1963): Sound Generation
by a Spinning Vortex Pair
Two parallel vortex ﬁlaments each of circulation I and distance 2¹ apart rotate
about the x
3
axis midway between them (Fig. 5.2.1) at angular velocity O =
I¡4π¹
2
, provided the Mach number is small enough for the motion to be
regarded as incompressible. Their positions at time t are
¯ x = (x
1
. x
2
) = ±s ≡ ±(s
1
(t ). s
2
(t )) = ±¹(cos Ot. sin Ot ).
Fig. 5.2.1.
5.2 The Equation of Vortex Sound 121
The vorticity distribution is
ω = Ik(δ(¯ x −s) ÷δ(¯ x ÷s)).
where k is a unit vector in the x
3
direction, parallel to the vortices. The vortex
convection velocities are
v = ±Ok ∧ s(t ) at ¯ x = ±s(t ). where U = O¹ <c
0
.
Hence, because k ∧ (k ∧ s) = −s,
ω ∧ v = −IOs(t )[δ(¯ x −s) −δ(¯ x ÷s)].
If this is expanded in powers of the radius ¹ of the circular orbit (and it can be
veriﬁed that this is equivalent to expanding the acoustic pressure in powers of
M =U¡c
0
<1) the i th component of the ﬁrst nonzero term is
(ω ∧ v)
i
≈
∂
∂ ¯ x
j
(2IOs
i
(t )s
j
(t )δ(¯ x)).
so that the vortex sound source is equivalent to the quadrupole
div(ω ∧ v) ≈
∂
2
∂ ¯ x
i
∂ ¯ x
j
(2IOs
i
(t )s
j
(t )δ(¯ x)).
The solution of the vortex sound equation (5.2.7) for this quadrupole source
is (c.f., (2.2.1))
B =
1
4π
∂
2
∂ ¯ x
i
∂ ¯ x
j
2IO(s
i
s
j
)
t −
[x −y[
c
0
δ(¯ y) d
3
y
[x −y[
. ¯ y = (y
1
. y
2
).
In the acoustic far ﬁeld, we use (1.9.7):
∂
∂ ¯ x
j
≈
−¯ x
j
c
0
(r
2
÷(x
3
− y
3
)
2
)
1
2
∂
∂t
.
where r =(x
2
1
÷ x
2
2
)
1
2
is the perpendicular distance from the centroid of the
vortices (the x
3
axis). Thus, by setting ξ = y
3
−x
3
, we can write in the acoustic
far ﬁeld, where B ≈ p¡ρ
0
,
p ≈
ρ
0
IO¯ x
i
¯ x
j
2πc
2
0
∂
2
∂t
2
∞
−∞
(s
i
s
j
)
t −
(r
2
÷ξ
2
)
1
2
c
0
dξ
(r
2
÷ξ
2
)
3
2
. r →∞.
(5.2.9)
122 5 Vortex Sound
In this formula
(s
i
s
j
)(t ) =
¹
2
2
1 ÷cos 2Ot sin 2Ot
sin 2Ot 1 −cos 2Ot
.
but the constant terms in the matrix can be omitted because of the time deriva
tives in (5.2.9). The integration in (5.2.9) can now be performed by the approx
imate method described below in Example 2 in the limit that Or¡c
0
→ ∞,
i.e., in the limit in which the radial distance r greatly exceeds the acoustic
wavelength
∞
−∞
(s
i
s
j
)
t −
(r
2
÷ξ
2
)
1
2
c
0
dξ
(r
2
÷ξ
2
)
3
2
≈
¹
2
2r
2
πc
0
Or
1
2
cos
2O[t ] −
π
4
sin
2O[t ] −
π
4
sin
2O[t ] −
π
4
−cos
2O[t ] −
π
4
.
where [t ] = t −r¡c
0
. Hence, introducing polar coordinates ¯ x = r(cos θ. sin θ)
we ﬁnd
¯ x
i
¯ x
j
¹
2
2r
2
πc
0
Or
1
2
cos
2O[t ] −
π
4
sin
2O[t ] −
π
4
sin
2O[t ] −
π
4
−cos
2O[t ] −
π
4
i j
=
¹
2
2
πc
0
Or
1
2
(cos θ. sin θ)
cos
2O[t ] −
π
4
sin
2O[t ] −
π
4
sin
2O[t ] −
π
4
−cos
2O[t ] −
π
4
cos θ
sin θ
=
¹
2
2
πc
0
Or
1
2
cos
2θ −2O[t ] ÷
π
4
.
and, therefore, (5.2.9) becomes
p ≈
−ρ
0
IO
3
¹
2
πc
2
0
πc
0
Or
1
2
cos
¸
2θ −2O
t −
r
c
0
÷
π
4
¸
= −4
π¹
r
ρ
0
U
2
M
3¡2
cos
¸
2θ −2O
t −
r
c
0
÷
π
4
¸
.
Or
c
0
→∞.
(5.2.10)
The amplitude of the sound decreases like 1¡
√
r (instead of 1¡r) because the
waves are spreading cylindrically in two dimensions. The sound power must
now be calculated by considering the integral
Y
p
2
ρ
0
c
0
dS
over the surface of a large circular cylinder r =constant. Taking the time
5.2 The Equation of Vortex Sound 123
average, we can show that the acoustic power per unit length of the vortices
∼¹ρ
0
U
3
M
4
. This differs by a factor of the Mach number M from the power
radiated by a compact body of threedimensional turbulence, and is character
istic of the acoustic power produced by twodimensional regions of turbulence
in an unbounded ﬂuid.
5.2.5 Example 2
Show that
∞
−∞
f
ξ
r
e
i κ
0
√
r
2
÷ξ
2
dξ ≈r f (0)
2π
κ
0
r
1
2
e
i
(
κ
0
r÷
π
4
)
. κ
0
r →∞. (5.2.11)
Put ξ =jr, then
I ≡
∞
−∞
f
ξ
r
e
i κ
0
√
r
2
÷ξ
2
dξ = r
∞
−∞
f (j)e
i κ
0
r
√
1÷j
2
dj.
As κ
0
r →∞the exponential factor oscillates increasingly rapidly, and the main
contribution to the integral is from the neighborhood of that value of j where
the oscillations are stationary. This occurs at j = 0. The integrand is therefore
expanded about this point. In the ﬁrst approximation, f (j) can be replaced by
f (0), and
e
i κ
0
r
√
1÷j
2
≈ e
i κ
0
r÷i κ
0
rj
2
¡2
.
Thus,
I = r
∞
−∞
f (j)e
i κ
0
r
√
1÷j
2
dj ≈ r f (0)e
i κ
0
r
∞
−∞
e
i κ
0
rj
2
¡2
dj
= r f (0)e
i κ
0
r
2π
κ
0
r
1
2
e
i π
4
.
which yields (5.2.11).
In particular,
∞
−∞
¸
cos 2O
t −
(r
2
÷ξ
2
)
1
2
c
0
−i sin 2O
t −
(r
2
÷ξ
2
)
1
2
c
0
¸
dξ
(r
2
÷ξ
2
)
3
2
= e
−2i Ot
∞
−∞
e
i
2O
c
0
√
r
2
÷ξ
2 dξ
(r
2
÷ξ
2
)
3
2
≈
1
r
2
πc
0
Or
1
2
e
−i [2O(t −
r
c
0
) −
π
4
]
.
Or
c
0
→∞.
124 5 Vortex Sound
Hence,
∞
−∞
cos 2O
t −
(r
2
÷ξ
2
)
1
2
c
0
dξ
(r
2
÷ξ
2
)
3
2
≈
1
r
2
πc
0
Or
1
2
cos
¸
2O
t −
r
c
0
−
π
4
¸
∞
−∞
sin 2O
t −
(r
2
÷ξ
2
)
1
2
c
0
dξ
(r
2
÷ξ
2
)
3
2
≈
1
r
2
πc
0
Or
1
2
sin
¸
2O
t −
r
c
0
−
π
4
¸
.
5.3 Vortex–Surface Interaction Noise
The small Mach number vortex sound equation (5.2.7) is now applied to deter
mine the sound generated by vorticity in the neighborhood of a ﬁxed body whose
surface S may be vibrating at small amplitude (Fig. 5.3.1). The development
here is analogous to the derivation in Section 2.3 of Curle’s equation.
Introduce a stationary, closed control surface S
÷
on which f (x) = 0, such
that f (x)
>

0 according as x lies without or within S
÷
. The body is assumed
to be within S
÷
, and S
÷
will subsequently be allowed to shrink down to co
incide with the body surface S. Multiply equation (5.2.7) by H ≡ H( f ) and
Fig. 5.3.1.
5.3 Vortex–Surface Interaction Noise 125
form the inhomogeneous wave equation for the new variable HB. We use the
transformations
H∇
2
B ≡ H div (∇B) = div(H∇B) −∇H · ∇B
= ∇
2
(HB) −div(B∇H) −∇H · ∇B. (5.3.1)
and H div(ω ∧ v) = div(Hω ∧ v) −∇H · ω ∧ v. (5.3.2)
Then, (5.2.7) becomes
1
c
2
0
∂
2
∂t
2
−∇
2
(HB) = −div(B ∇ H) −∇H · (∇B ÷ω∧v) ÷div(Hω∧v).
This equation is formally valid everywhere, including the region within S
÷
where HB ≡0. The source terms involving ∇H are concentrated on the control
surface. When x lies in the exterior region these surface terms take account of
the presence of the solid body inside S
÷
; if the body is absent (so that S
÷
is
ﬁlled with ﬂuid), the surface sources constitute a representation ‘to the outside
world’ in f >0 of the various hydrodynamic or acoustic processes that may be
occurring within S
÷
.
Using Crocco’s equation (4.2.5), we can make the substitution
∇H · (∇B ÷ω ∧ v) =−∇H ·
∂v
∂t
÷ν curl ω
≡−∇H ·
∂v
∂t
÷ν div(∇H ∧ω).
Hence the vortex sound equation becomes
1
c
2
0
∂
2
∂t
2
−∇
2
(HB) =−div(B∇H) ÷∇H ·
∂v
∂t
÷div(Hω ∧ v) −ν div(∇H ∧ ω). (5.3.3)
The sources on the right of this equation are either concentrated on the control
surface S
÷
or lie in the ﬂuid outside S
÷
; they completely determine B outside
this control surface. The solution in this region can therefore be found by using
any Green’s function G(x. y. t −τ) that satisﬁes
1
c
2
0
∂
2
∂t
2
−∇
2
G = δ(x −y)δ(t −τ). where G = 0 for t τ
for x and y anywhere within the ﬂuid. In Fig. 5.3.1 the ﬂuid occupies the region
V outside the surface S of the solid body; the control surface S
÷
( f (x) = 0)
therefore lies within V.
126 5 Vortex Sound
Thus, for points x within the ﬂuid the solution of (5.3.3) is
HB(x. t ) =
∞
−∞
V
G(x. y. t −τ)
¸
−div(B∇H) ÷∇H ·
∂v
∂τ
÷div(Hω∧v)
−ν div(∇H ∧ ω)
¸
d
3
y dτ.
where all of the source terms within the brace brackets are functions of y and τ.
Those involving ∇H vanish except on the control surface S
÷
. The divergence
terms are removed by application of the divergence theorem, and then using the
formula
V
(·)∇H d
3
y =
S
÷
(·) dS
(see Section 2.3). Let us illustrate the procedure for the ﬁrst term in the brace
brackets of the integrand
V
G{−div(B∇H)} d
3
y = −
V
{div(GB∇H) − B∇G · ∇H} d
3
y
=
S÷Y
GB∇H · dS ÷
V
B∇G · ∇H d
3
y
= 0 ÷
S
÷
B∇G · dS
≡
S
÷
B(y. τ)
∂G
∂y
n
(x. y. t −τ) dS(y).
where all vector operators are with respect to the y dependence, and Y is a
large, closed ‘surface at inﬁnity’ where ω = 0. There are no contributions from
S and Y because ∇H = 0 everywhere except on S
÷
.
The general solution in the region f >0 outside S
÷
accordingly becomes
B(x. t ) =
S
÷
B(y. τ)
∂G
∂y
n
(x. y. t −τ) ÷ G(x. y. t −τ)
∂:
n
∂τ
(y. τ)
dS(y) dτ
−
V
H( f (y))(ω ∧ v)(y. τ) ·
∂G
∂y
(x. y. t −τ) d
3
y dτ
÷ν
S
÷
ω(y. τ) ∧
∂G
∂y
(x. y. t −τ) · dS(y) dτ. (5.3.4)
where for brevity we have omitted the integration sign for τ, which is understood
to vary over the range −∞τ ∞.
5.3 Vortex–Surface Interaction Noise 127
We now choose G to have vanishing normal derivative on the surface S of
the body. When this is done the control surface S
÷
is allowed to shrink down
onto S (whereupon the ﬁrst term in the ﬁrst integral of (5.3.4) vanishes), and
the general solution in the ﬂuid becomes
B(x. t ) = −
V
(ω ∧ v)(y. τ) ·
∂G
∂y
(x. y. t −τ) d
3
y dτ ÷ν
S
ω(y. τ)
∧
∂G
∂y
(x. y. t −τ) · dS(y) dτ ÷
S
G(x. y. t −τ)
∂:
n
∂τ
(y. τ) dS(y) dτ.
where
∂G
∂y
n
(x. y. t −τ) = 0 on S. (5.3.5)
In the acoustic far ﬁeld ([x[ →∞) we can replace B(x. t ) by p(x. t )¡ρ
0
.
The ﬁrst integral represents the production of sound by vortex sources dis
tributed within the ﬂuid. Green’s function takes full account of the inﬂuence
of the body on the efﬁciency with which these sources generate sound. The
second, surface integral involving the surface vorticity is the contribution from
frictional forces on S. To interpret the ﬁnal term, recall that the control surface
S
÷
was taken to be ﬁxed in space. This means that ﬂuid can ﬂow through the
surface. When S
÷
shrinks down to S the implication is that S is also ﬁxed in
space. However, the normal velocity :
n
can still be nonzero if the surface of the
body is vibrating at small amplitude, and this term in the solution is actually
identical with that given previously in (3.8.2) for a vibrating body in the absence
of vortex sources.
In connection with this, it should also be noted that the reciprocal theorem
implies that the normal derivative conditions
∂G
∂y
n
(x. y. t −τ) = 0.
∂G
∂x
n
(x. y. t −τ) = 0 respectively for y. x on S.
are always satisﬁed simultaneously.
The contribution to the sound from surface friction (the ﬁrst surface integral
on the right of (5.3.5)) is nominally of order
1
Re
<1. Re =
:¹
ν
.
relative to the contribution from the volume vorticity (the ﬁrst integral), where
¹ is the characteristic length scale of the turbulence or body and : is a typical
velocity. At high Reynolds numbers the surface termcan therefore be discarded,
and in the important case in which the body does not vibrate the acoustic far
128 5 Vortex Sound
ﬁeld is then given by
p
ρ
0
(x. t ) = −
V
(ω ∧ v)(y. τ) ·
∂G
∂y
(x. y. t −τ) d
3
y dτ. (5.3.6)
5.4 Radiation from an Acoustically Compact Body
Whenthe surface S is acousticallycompact the compact Green’s function(3.9.1)
can be used to evaluate the general solution (5.3.5) in the far ﬁeld. When
[x[ →∞and the origin is within or close to S, we proceed as already described
in Section 3.8 by expanding Green’s function to ﬁrst order in the retarded time
across S:
G(x. y. t −τ) =
1
4π[X −Y[
δ
t −τ −
[X −Y[
c
0
≈
1
4π[x[
δ
t −τ −
[x[
c
0
÷
x
j
Y
j
4πc
0
[x[
2
δ
/
t −τ −
[x[
c
0
.
[x[ →∞.
where the prime denotes differentiation with respect to t . The ﬁrst term in this
approximation, which is independent of y, clearly makes no contribution to the
ﬁrst two integrals in (5.3.5). It makes a contribution to the ﬁnal surface integral
only if the volume of the body is pulsating. When this happens the resulting
monopole radiation fromthe body is usually large compared to all other sources.
We shall, therefore, ignore this possibility, and consider only surface vibrations
for which the volume of S is constant; in particular we shall assume that S
vibrates as a rigid body. In this case, therefore, the ﬁrst approximation in the
Green’s function expansion can again be discarded.
Substituting into (5.3.5) and performing the integrations with respect to τ,
we ﬁnd in the acoustic far ﬁeld (where B = p¡ρ
0
)
p(x. t ) ≈
−ρ
0
x
j
4πc
0
[x[
2
∂
∂t
¸
(ω ∧ v) · ∇Y
j
d
3
y −ν
S
ω ∧ ∇Y
j
· dS(y)
−
S
∂U
n
∂t
Y
j
dS(y)
¸
. (5.4.1)
where the large square brackets ([ ]) denote that the enclosed quantity is to be
evaluated at the retarded time t − [x[¡c
0
, and U
n
is the normal component of
velocity of vibration of S.
5.4 Radiation from an Acoustically Compact Body 129
When the body executes translational oscillations at velocity U(t ) we have
U
n
= U
i
n
i
.
where n
i
is the i th component of the surface normal directed into the ﬂuid.
Then
ρ
0
S
∂U
n
∂t
Y
j
dS = ρ
0
dU
i
dt
S
(n
i
y
j
−n
i
ϕ
∗
j
) dS
= (m
0
δ
i j
÷ M
i j
)
dU
i
dt
.
where if the body has volume L, then m
0
= ρ
0
L is the mass of ﬂuid displaced
by the body, and M
i j
is the added mass tensor of the body (see (3.8.6)).
The solution (5.4.1) can therefore be written
p(x. t ) ≈
−x
j
4πc
0
[x[
2
∂
∂t
¸
ρ
0
(ω ∧ v) · ∇Y
j
d
3
y −η
S
ω ∧ ∇Y
j
· dS(y)
−(m
0
δ
i j
÷ M
i j
)
dU
i
dt
¸
. (5.4.2)
Reference to Equation (4.4.4) shows that this can also be written
p(x. t ) ≈
x
j
4πc
0
[x[
2
∂ F
j
∂t
t −
[x[
c
0
÷
m
0
x
j
4πc
0
[x[
2
∂
2
U
j
∂t
2
t −
[x[
c
0
. [x[ →∞.
(5.4.3)
where F(t ) is the unsteady force exerted on the ﬂuid by the body. This is just
our earlier conclusion (2.4.2) derived from Curle’s equation, with the addition
of the ﬂuiddisplacement effect of the vibrating body, and is equivalent to the
solution (3.8.10) obtained in the absence of vorticity.
The relative contributions from the volume and surface distributions of vor
ticity in (5.4.2) (respectively the ﬁrst and second integrals) for turbulence of
length scale ¹ and velocity : are estimated respectively by
ρ
0
:
2
M
¹
[x[
and ρ
0
:
2
M
¹
[x[
1
Re
. Re =
:¹
ν
.
Thus, in high Reynolds number turbulent ﬂows the surface frictional contribu
tion to the dipole force F can usually be neglected. For a nonvibrating compact
body the principal component of the acoustic pressure in the far ﬁeld is therefore
130 5 Vortex Sound
given by
p(x. t ) ≈
−ρ
0
x
j
4πc
0
[x[
2
∂
∂t
(ω ∧ v)
y. t −
[x[
c
0
· ∇Y
j
(y) d
3
y. [x[ →∞.
(5.4.4)
5.5 Radiation from Cylindrical Bodies of Compact Cross Section
An important special case occurs when vorticity interacts with a cylindrical (or
approximately cylindrical) surface S of compact cross section (such as the strip
airfoil of Fig. 3.6.3). If the vorticity extends over an extensive spanwise section
of the body it may be important to account for differences in the retarded times
of the sound produced at different spanwise positions.
To do this we ﬁrst write the Kirchhoff vector in the form
Y = Y
⊥
÷ky
3
. Y
⊥
= (Y
1
(y). Y
2
(y). 0). (5.5.1)
where k is a unit vector in the x
3
direction. Then, because variations in the
spanwise source position are not necessarily small compared to the acoustic
wavelength, the compact Green’s function in (5.3.6) is expanded as follows:
G(x. y. t −τ) =
1
4π[X −Y[
δ
t −τ −
[X −Y[
c
0
≈
1
4π[x −ky
3
[
δ
t −τ −
[x −ky
3
[
c
0
÷
x
j
Y
⊥j
4πc
0
[x −ky
3
[
2
δ
/
t −τ −
[x −ky
3
[
c
0
≈
1
4π[x[
δ
t −τ −
[x −ky
3
[
c
0
÷
x
j
Y
⊥j
4πc
0
[x[
2
δ
/
t −τ −
[x −ky
3
[
c
0
. [x[ →∞. (5.5.2)
For example, when a stationary cylindrical body interacts with high Reynolds
number ﬂowat lowMach number, the monopole termin (5.5.2) can be discarded
as before, and the acoustic pressure given by (5.3.6) becomes
p(x. t ) ≈
−ρ
0
x
j
4πc
0
[x[
2
∂
∂t
(ω∧v)
y. t −
[x −ky
3
[
c
0
· ∇Y
⊥j
(y) d
3
y. [x[ →∞.
(5.5.3)
The sound is produced by dipole sources orientated in the lift and drag directions
5.6 Impulse Theory of Vortex Sound 131
only ( j = 1. 2). This approximation is applicable also to a thin airfoil of large
but ﬁnite span, and in cases where the chord of the airfoil is a slowly varying
function of y
3
(such as the elliptic airfoil of Fig. 3.9.1).
5.6 Impulse Theory of Vortex Sound
An interesting formula for the sound generated by vorticity near a compact body
can be derived directly from the representation (4.3.3) of the velocity v(x. t ) in
the hydrodynamic far ﬁeld in terms of the impulse I(t ). Indeed,
v(x. t ) ≈ ∇ϕ when [x[ ·¹.
where ¹ is the length scale of the interaction region (the body), and
ϕ(x. t ) = div
I(t )
4π[x[
where I(t ) =
1
2
y ∧ ω(y. t ) d
3
y. (5.6.1)
This expression for ϕ(x. t ) deﬁnes the incompressible motion in the irrota
tional region far from the body. It is the velocity potential of a hydrodynamic
dipole that will be recognized as the acoustic near ﬁeld of an outgoing acoustic
dipole representing sound production by the ﬂow (see Equations (1.7.4) and
(1.7.9)). The acoustic dipole is found simply by replacing I(t ) in (5.6.1) by
I(t −[x[¡c
0
). At large distances from the body (where the undisturbed ﬂuid is
stationary) the pressure p(x. t ) =−ρ
0
∂ϕ¡∂t , and this procedure therefore leads
to the following formula for the sound in terms of the vorticity:
p(x. t ) ≈ −ρ
0
∂
∂x
j
1
4π[x[
∂ I
j
∂t
(t −[x[¡c
0
)
≈
ρ
0
x
j
4πc
0
[x[
2
∂
2
I
j
∂t
2
t −
[x[
c
0
. [x[ →∞
=
ρ
0
x
j
8πc
0
[x[
2
∂
2
∂t
2
(y ∧ ω)
j
y. t −
[x[
c
0
d
3
y. (5.6.2)
Equation (4.4.1) shows that this is equivalent to the compact approximation
(5.4.3). Note, however, that ω in (5.6.2) is the generalized vorticity, including
the bound vorticity on S. This should be contrasted with the representation
(5.4.2) involving the Kirchhoff vector, where bound vorticity occurs only in the
surface integral of the frictional contribution to the sound. But, (5.4.2) is valid
only for a body in translational motion whereas (5.6.2) is applicable for a body
executing any combination of translations and rotations (Fig. 4.1.1).
132 5 Vortex Sound
The impulse I is constant for vorticity in an unbounded ﬂuid (when com
pressibility is ignored). We knowthat the sound is nowgenerated by quadrupole
sources and that it can be represented in terms of the vorticity as in (5.1.6).
M¨ ohring (1978) has shown that it is also possible to express the quadrupole
sound as a thirdorder time derivative of a secondorder moment of the vor
ticity, analogous to the ﬁrst order moment in (5.6.2) (see Problems 5),
namely
p(x. t ) ≈
ρ
0
x
i
x
j
12πc
2
0
[x[
3
∂
3
∂t
3
y
i
(y ∧ω)
j
(y. t −[x[¡c
0
) d
3
y. [x[ →∞. (5.6.3)
Problems 5
1. Kirchhoff’s spinning vortex: A columnar vortex parallel to the x
3
axis
has elliptic cross section deﬁned by the polar equation r = a{1÷c cos(2θ −
Ot ¡2)}, where c <1 and O is the uniform vorticity in the core. The ellipse
rotates at angular velocity
1
4
O, and the velocity distribution within the core
is given by
v = (:
1
. :
2
) = −
1
2
Or(sin θ ÷c sin(θ −Ot ¡2). −cos θ ÷c cos(θ −Ot ¡2)).
Show that to ﬁrst order in c the vortex is equivalent to the twodimensional
quadrupole
div(ω ∧ v) ≈
∂
2
∂x
i
∂x
j
(T
i j
δ(x
1
)δ(x
2
)). i. j = 1. 2
Problems 5 133
where
T
i j
=
cπO
2
a
4
8
cos(Ot ¡2) sin(Ot ¡2)
sin(Ot ¡2) −cos(Ot ¡2)
and that the acoustic pressure is
p ≈ −
c
8
2πa
r
ρ
0
U
2
M
3¡2
cos
¸
2θ −
O
2
t −
r
c
0
÷
π
4
¸
.
Or
c
0
→∞.
where U =
1
2
aO and M =U¡c
0
.
2. Coaxial vortex rings: Use Equation (5.2.7) to calculate the sound produced
by the unsteady motions of an acoustically compact system of N vortex
rings coaxial with the x
1
axis. Take the vorticity of the nth vortex to be
ω
n
= I
n
δ(x
1
− X
n
(t ))δ(r − R
n
(t ))i
θ
, where (r. θ. x
1
) are cylindrical polar
coordinates, R
n
(t ) being the vortex ring radius, X
n
(t ) its location in the x
1
direction, and i
θ
is a unit vector in the azimuthal direction. Show that
p ≈
ρ
0
8c
2
0
[x[
(3 cos
2
O−1)
∂
2
∂t
2
¸
¸
n
I
n
X
n
d R
2
n
dt
. [x[ →∞.
where Ois the angle between the observer direction and the positive x
1
axis
and the term in square braces is evaluated at τ = t −[x[¡c
0
.
In an ideal, incompressible ﬂuid conservation of energy and momentum
implies that
¸
n
I
n
R
n
R
n
d X
n
dt
− X
n
d R
n
dt
= constant.
¸
n
I
n
R
2
n
= constant.
Use these equations to show that
p ≈
ρ
0
12c
2
0
[x[
(3 cos
2
O−1)
¸
d
3
S
dt
3
¸
. [x[ →∞.
where S(t ) =
¸
n
I
n
R
2
n
X
n
.
134 5 Vortex Sound
3. Calculate the sound produced by a vortex ring of total circulation I, coaxial
with the x
1
axis, whose core has elliptic cross section of major and minor
axes 2a. 2b < R, where R is the mean radius of the ring. Assume that the
x
1
coordinate X(t ) of the vorticity centroid satisﬁes
d X
dt
=
I
4π R
¸
ln
16R
a ÷b
−
1
4
÷
3(a −b)
2(a ÷b)
cos 2Ot
¸
. O =
I
π(a ÷b)
2
.
In the notation of Question 2, show that
p ≈
ρ
0
IR
2
(3 cos
2
O−1)
12c
2
0
[x[
¸
d
3
X
dt
3
¸
t −
[x[
c
0
. [x[ →∞
=
ρ
0
U
2
M
2
8π
3
R(3 cos
2
O−1)
[x[
a −b
a ÷b
cos
2I
π(a ÷b)
2
¸
t −
[x[
c
0
¸
.
where U =
I
a ÷b
. M =
U
c
0
.
4. Derive M¨ ohring’s formula (5.6.3) for the acoustic pressure generated by a
compact region of vorticity in an unbounded ﬂuid. Take the cross product of y
with the high Reynolds number vorticity equation ∂ω¡∂t ÷curl (ω∧v) = 0
(expressed in terms of y and t as independent variables), multiply by y
i
, and
use the identity
y ∧ curl A = 2A ÷∇(y · A) −
∂
∂y
j
(y
j
A)
to deduce that the integral in (5.1.6) can be written
y
i
(ω ∧ v)
j
d
3
y = −
1
3
∂
∂t
y
i
(y ∧ ω)
j
d
3
y ÷
1
3
δ
i j
1
2
:
2
d
3
y.
The result nowfollows by noting that the estimate (5.1.7) permits the second
integral on the right to be discarded.
5. The free space Green’s function in two dimensions – the solution of
1
c
2
0
∂
2
∂t
2
−∇
2
G =δ(x
1
−y
1
)δ(x
2
−y
2
)δ(t −τ). where G =0 for t τ.
– can be derived by integrating the threedimensional Green’s function
1
4π[x −y[
δ
t −τ −
[x −y[
c
0
over −∞ y
3
∞.
Problems 5 135
Deduce that
G(x. y. t −τ) =
H(t −τ −[x −y[¡c
0
)
2π
(t −τ)
2
−(x −y)
2
c
2
0
. x = (x
1
. x
2
). y =(y
1
. y
2
).
and that near the wavefront, where [x −y[ ≈ c
0
(t −τ)
G(x. y. t −τ) ≈
H(t −τ −[x −y[¡c
0
)
2π
√
2[x −y[¡c
0
√
(t −τ) −[x −y[¡c
0
.
6. Consider sound production by a compact distribution of vorticity in an un
bounded twodimensional ﬂow (independent of x
3
), where the vorticity ω
is parallel to the k direction (the x
3
axis). Show that
y
i
(ω ∧ v)
j
dy
1
dy
2
≈−
1
2
∂
∂t
y
i
(y ∧ ω)
j
dy
1
dy
2
. where i. j = 1. 2.
Deduce M¨ ohring’s (1980) twodimensional representation
p(x. t ) ≈
ρ
0
x
i
x
j
4πc
2
0
[x[
2
∂
3
∂t
3
dy
1
dy
2
t −[x[¡c
0
−∞
y
i
(y ∧ ω)
j
(y. τ) dτ
(t −τ)
2
−[x[
2
c
2
0
≈
ρ
0
x
i
x
j
4πc
2
0
[x[
2
c
0
2[x[
∂
3
∂t
3
dy
1
dy
2
t −[x[¡c
0
−∞
y
i
(y ∧ ω)
j
(y. τ) dτ
√
t −τ −[x[¡c
0
.
[x[ →∞.
7. Use the result of Problem 6 to derive the Solution (5.2.10) for the sound
produced by a spinning vortex pair.
8. Use the result of Problem 6 to solve Problem 1.
6
Vortex–Surface Interaction Noise
in Two Dimensions
6.1 Compact Green’s Function in Two Dimensions
In this chapter, we apply the general high Reynolds number solution (5.3.6)
to determine sound produced by twodimensional interactions of rectilinear
vortices with a stationary solid boundary. Conditions are assumed to be uniform
in the x
3
direction, parallel to the vorticity. We shall derive the twodimensional
analogue of the general solution
p
ρ
0
(x. t ) = −
V
(ω ∧ v)(y. τ) ·
∂G
∂y
(x. y. t −τ) d
3
y dτ (6.1.1)
by ﬁrst determining a suitable representation of G in two dimensions. Both the
vorticity convection velocity v and the Lamb vector ω∧v are parallel to the x
1
x
2
plane, so that only the y
1
and y
2
components of the gradient ∂G¡∂y contribute
to the integral. Also, because (ω ∧ v)(y. τ) depends only on y
1
. y
2
and τ, the
integration with respect to the spanwise coordinate y
3
involves only the Green’s
function, and may be performed prior to any further calculations of the sound.
Let
G
2
=
∞
−∞
G(x. y. t −τ) dy
3
. (6.1.2)
For twodimensional problems G is a function of y
3
−x
3
, and the function G
2
will therefore satisfy the Green’s function equation
1
c
2
0
∂
2
∂t
2
−∇
2
G
2
=δ(x
1
− y
1
)δ(x
2
− y
2
)δ(t −τ).
where G
2
=0 for t τ. (6.1.3)
obtained by integrating the threedimensional Equation (3.1.4) over −∞ y
3

∞. Intwodimensions, G
2
represents the ﬁeldgeneratedbya uniformline source
parallel to the x
3
axis extending along the whole of the line x
1
= y
1
. x
2
= y
2
.
136
6.1 Compact Green’s Function in Two Dimensions 137
Similarly, the compact Green’s function for a cylindrical body (with gener
ators parallel to x
3
)
G(x. y. t −τ) =
1
4π[X −Y[
δ
t −τ −
[X −Y[
c
0
.
X
1.2
= x
1.2
−ϕ
∗
1.2
(x). Y
1.2
= y
1.2
−ϕ
∗
1.2
(y). X
3
= x
3
. Y
3
= y
3
.
is a functionof y
3
– x
3
, andthe correspondingtwodimensional compact Green’s
function can be found by integrating over −∞ y
3
∞.
Set ξ =y
3
−x
3
. ¯ x =(x
1
. x
2
) and let [¯ x[ =(x
2
1
÷x
2
2
)
1
2
→∞. Taking the origin
of coordinates within the cylindrical body, we have
G
2
≈
1
4π
∞
−∞
δ
t −τ −
[¯ x[
2
÷ξ
2
c
0
÷
¯ x · Y
c
0
[¯ x[
2
÷ξ
2
dξ
[¯ x[
2
÷ξ
2
as [¯ x[ →∞.
where ¯ x · Y = x
1
Y
1
÷ x
2
Y
2
is independent of ξ. To use this to evaluate (6.1.1)
the δ function must be expanded to ﬁrst order in Y
G
2
≈
1
4π
∞
−∞
δ
t −τ −
[¯ x[
2
÷ξ
2
c
0
dξ
[¯ x[
2
÷ξ
2
÷
¯ x · Y
4πc
0
∞
−∞
δ
/
t −τ −
[¯ x[
2
÷ξ
2
c
0
dξ
([¯ x[
2
÷ξ
2
)
.
Only the second term on the right depends on y and therefore contributes to the
radiation integral (6.1.1). We can therefore take
G
2
≈
¯ x · Y
4πc
0
∞
−∞
δ
/
t −τ −
[¯ x[
2
÷ξ
2
c
0
dξ
([¯ x[
2
÷ξ
2
)
=
¯ x · Y
2πc
0
∂
∂t
∞
0
δ
t −τ −
[¯ x[
2
÷ξ
2
c
0
dξ
([¯ x[
2
÷ξ
2
)
=
¯ x · Y
2πc
0
∂
∂t
¸
¸
¸
¸
H(t −τ −[¯ x[¡c
0
)
([¯ x[
2
÷ξ
2
)
∂
∂ξ
√
[¯ x[
2
÷ξ
2
c
0
¸
¸
¸
¸
¸
ξ=
√
c
2
0
(t −τ)
2
−[¯ x[
2
=
¯ x · Y
2πc
0
∂
∂t
¸
¸
H(t −τ −[¯ x[¡c
0
)
(t −τ)
c
2
0
(t −τ)
2
−[¯ x[
2
¸
¸
¸
.
138 6 Vortex–Surface Interaction Noise in Two Dimensions
We shall henceforth in this chapter regard all space vectors as twodimen
sional, such as x = (x
1
. x
2
). y = (y
1
. y
2
), and drop the overbar on ¯ x and the
subscript 2 on G
2
and soforth. The dipole component of the twodimensional
compact Green’s function then becomes
G(x. y. t −τ) ≈
x · Y
2πc
0
∂
∂t
¸
¸
H(t −τ −[x[¡c
0
)
(t −τ)
c
2
0
(t −τ)
2
−[x[
2
¸
¸
¸
. [x[ →∞.
(6.1.4)
In contrast to the threedimensional Green’s function, which is nonzero only
on a spherically expanding wavefront, G has an inﬁnite peak at the wavefront
where t − τ − [x[¡c
0
=0 followed by a slowly decaying tail. At any point x
in the far ﬁeld the ﬁrst sound arrives from the nearest point on the line source
of (6.1.3), after propagating along a ray perpendicular to the source over a
distance equal to [x[, and therefore after a time delay [x[¡c
0
; but the observer at x
receives sound continuously after the passage of this wavefront, generated at
more distant sections of the inﬁnitely long line source; the wavefront arrives at
time τ ÷[x[¡c
0
, and at a later time t sound is received from those source points
whose distance from x is equal to c
0
(t −τ).
The farﬁeld representation (6.1.4) can be approximated further by expanding
about the wavefront (where t −τ = [x[¡c
0
), which contains most of the acoustic
energy. Just to the rear of the wavefront
(t −τ)
c
2
0
(t −τ)
2
−[x[
2
≡ (t −τ)
c
0
(t −τ) ÷[x[
c
0
(t −τ) −[x[
≈
[x[
c
0
2[x[
c
0
(t −τ) −[x[. for t −τ ∼
[x[
c
0
.
Therefore, (6.1.4) becomes
G(x. y. t −τ) ≈
x · Y
2π
√
2c
0
[x[
3
2
∂
∂t
¸
H(t −τ −[x[¡c
0
)
√
t −τ −[x[¡c
0
¸
. [x[ →∞. (6.1.5)
(see Question 1 of Problems 6).
The special case of a cylindrical body adjacent to a plane, rigid wall at
x
2
= 0, or of a cylindrical wall cavity or projection from a wall (see Fig. 3.9.2)
is handled by the twodimensional version of the compact Green’s function
(3.9.3). The procedure described above yields the following expression for the
dipole component of the twodimensional compact Green’s function
G(x. y. t −τ) ≈
x
1
Y
1
π
√
2c
0
[x[
3
2
∂
∂t
¸
H(t −τ −[x[¡c
0
)
√
t −τ −[x[¡c
0
¸
. [x[ →∞. (6.1.6)
6.2 Sound Generated by a Line Vortex 139
where Y
1
≡ Y
1
(y
1
. y
2
) is the velocity potential of incompressible ﬂow past the
cylinder in a direction parallel to the wall (with unit speed at large distances
from the body). G represents the ﬁeld of a dipole orientated parallel to the
wall and perpendicular to the cylinder axis, and the effect of the wall is to
generate an equal image dipole that just doubles the magnitude of the sound
relative to the corresponding dipole (6.1.5) of the cylinder in the absence of the
wall.
6.2 Sound Generated by a Line Vortex Interacting
with a Cylindrical Body
The sound produced by twodimensional vortex motion at low Mach number
M ∼ :¡c
0
near a stationary, rigid cylindrical body of diameter ¹ has charac
teristic wavelength ∼¹¡M · ¹. The acoustic pressure is determined by the
twodimensional version of (6.1.1) using the compact Green’s function (6.1.5),
viz,
p(x. t ) ≈
−ρ
0
x
j
2π
√
2c
0
[x[
3
2
∂
∂t
t −[x[¡c
0
−∞
dτ
√
t −τ −[x[¡c
0
(ω ∧ v · ∇Y
j
)(y. τ) dy
1
dy
2
. [x[ →∞. (6.2.1)
According to the inviscid form of the Formula (4.4.4) applied to a stationary
body, this can also be written
p(x. t ) ≈
x
j
2π
√
2c
0
[x[
3
2
∂
∂t
t −[x[¡c
0
−∞
F
j
(τ) dτ
√
t −τ −[x[¡c
0
. [x[ →∞. (6.2.2)
where F
j
is the force per unit length of the cylinder exerted on the ﬂuid in the
j direction.
In these twodimensional problems the vorticity ω is directed along the x
3
axis, out of the plane of the paper in Fig. 6.2.1 and parallel to the generators of
the cylinder. Let k be a unit vector in this direction and consider a line vortex
of strength I whose position and translational velocity are
x = x
0
(t ). v =
dx
0
dt
(t ).
If the motion elsewhere is irrotational, we have
ω = Ikδ(x −x
0
(t ))
140 6 Vortex–Surface Interaction Noise in Two Dimensions
Fig. 6.2.1.
so that
ω ∧ v = Ik ∧ vδ(x −x
0
(t )) ≡ Ik ∧
dx
0
dt
(t )δ(x −x
0
(t )). (6.2.3)
and (6.2.1) yields the following general formula for the acoustic pressure as
[x[ →∞:
p(x. t )
≈
−ρ
0
Ix
j
2π
√
2c
0
[x[
3
2
∂
∂t
t −[x[¡c
0
−∞
k ∧
dx
0
dτ
(τ) · ∇Y
j
(x
0
(τ))
dτ
√
t −τ −[x[¡c
0
=
−ρ
0
Ix
j
2π
√
2c
0
[x[
3
2
∂
∂t
t −[x[¡c
0
−∞
¸
dx
01
dτ
∂Y
j
∂y
2
−
dx
02
dτ
∂Y
j
∂y
1
¸
x
0
(τ)
dτ
√
t −τ −[x[¡c
0
.
(6.2.4)
6.2.1 Example 1: Sound Produced by Vortex Motion near
a Circular Cylinder
Assume there is no mean ﬂow, and let the vortex strength I be sufﬁciently
small for the local motion to be considered incompressible. Then, the vortex
will traverse the circular orbit discussed in Section 4.6 (Example 2).
Let the cylinder have radius a, the vortex path have radius r
0
, and take the
coordinate origin at the centre of the cylinder. If I >0 the vortex moves in the
6.2 Sound Generated by a Line Vortex 141
Fig. 6.2.2.
clockwise direction in Fig. 6.2.2 at the speed given by (4.6.8) when r is replaced
by r
0
. Then,
v =
dx
0
dt
(t ) = Ok ∧ x
0
(t ).
O =
−Ia
2
2πr
2
0
r
2
0
−a
2
.
x
0
= r
0
(cos Ot. sin Ot ).
and
k ∧
dx
0
dt
= Ok ∧ (k ∧ x
0
) = −Ox
0
.
Therefore, (6.2.4) becomes
p(x. t ) ≈
ρ
0
IOx
j
r
0
2π
√
2c
0
[x[
3
2
∂
∂t
t −[x[¡c
0
−∞
∂Y
j
∂r
(x
0
(τ))
dτ
√
t −τ −[x[¡c
0
. (6.2.5)
where r =
√
y
2
1
÷ y
2
2
is the radial distance from the cylinder axis.
Table 3.9.1 supplies the components of the twodimensional Kirchhoff vector
for the cylinder:
Y
1
= cos ϑ
r ÷
a
2
r
. Y
2
= sin ϑ
r ÷
a
2
r
.
142 6 Vortex–Surface Interaction Noise in Two Dimensions
where (y
1
. y
2
) = r(cos ϑ. sin ϑ). Thus, by introducing polar coordinates for x
x = [x[(cos θ. sin θ)
we ﬁnd x
j
∂Y
j
∂r
= [x[
1 −
a
2
r
2
cos(θ −ϑ)
so that
x
j
∂Y
j
∂r
(x
0
(τ)) = [x[
1 −
a
2
r
2
0
cos(θ −Oτ).
and (6.2.5) becomes
p(x. t ) ≈
ρ
0
IOr
0
2π(2c
0
[x[)
1
2
1 −
a
2
r
2
0
∂
∂t
t −[x[¡c
0
−∞
cos(θ −Oτ) dτ
√
t −τ −[x[¡c
0
.
Hence, using the formula
t −[x[¡c
0
−∞
cos(θ −Oτ) dτ
√
t −τ −[x[¡c
0
=
π
[O[
1
2
cos
¸
θ −O
¸
t −
[x[
c
0
¸
−
π
4
¸
the pressure becomes
p(x. t ) ≈
ρ
0
I[O[
3
2
r
0
2(2πc
0
[x[)
1
2
1 −
a
2
r
2
0
sin
¸
θ −O
¸
t −
[x[
c
0
¸
−
π
4
¸
= ρ
0
U
2
√
M
πr
0
2[x[
r
0
a
2
1 −
a
2
r
2
0
2
sin
¸
θ −O
¸
t −
[x[
c
0
¸
−
π
4
¸
.
[x[ →∞. (6.2.6)
where U =[O[r
0
is the vortex speed.
The acoustic waves decay like 1¡
√
[x[ with distance, which is appropriate
for energy spreading two dimensionally in cylindrically diverging waves, and
have the characteristic dipole amplitude proportional to ρ
0
U
2
√
M. In three
dimensions the pressure would be proportional to ρ
0
U
2
M, which is smaller by
a factor
√
M when M <1. The increased amplitude in two dimensions is a
consequence of the inﬁnite extent of the vortex source parallel to the cylinder.
At any particular retarded time t −[x[¡c
0
, the acoustic amplitude has the double
lobed directivity pattern illustrated in Fig. 1.7.1b for a dipole. Because O0,
the peaks of these lobes at a ﬁxed distance [x[ from the cylinder rotate in the
clockwise direction at angular velocity [O[ following the orbiting vortex, but
with a phase lag of π¡4 radians. The reader can conﬁrm that the instantaneous
force exerted on the ﬂuid by the cylinder lies in the direction of the vector x
0
(t )
6.2 Sound Generated by a Line Vortex 143
joining the center of the cylinder to the vortex. The radiation peak therefore
also lags by π¡4 the peak in the retarded surface force.
6.2.2 Example 2: Sound Produced by Vortex Motion near a HalfPlane
(Crighton 1972)
The trajectory of a line vortex of strength I interacting with a rigid halfplane is
shown in Fig. 4.6.2 for ideal motion at low Mach number. The sound generated
by the vortex is calculated using the twodimensional compact Green’s function
(3.9.9), which can be written
G
1
(x. y. t −τ) ≈
sin(θ¡2)ϕ
∗
(y)
π
√
[x[
δ
t −τ −
[x[
c
0
. [x[ →∞.
ϕ
∗
(y) =
√
r
0
sin(θ
0
¡2). (6.2.7)
where x =[x[(cos θ. sin θ), the coordinates being deﬁned as in Fig. 4.6.2. This
is applicable when the distance r
0
from the edge of the source at (y
1
. y
2
) =
r
0
(cos θ
0
. sin θ
0
) is much smaller than the acoustic wavelength. For a line vortex
at (r
0
. θ
0
) the characteristic frequency ∼:¡r
0
, where : is the vortex translational
velocity. The wavelength is therefore of order
r
0
:
c
0
=
r
0
M
·r
0
for M =
:
c
0
<1.
so that low Mach number motion is sufﬁcient to ensure that the wavelength of
the sound is much larger than the vortex distance from the edge.
Thus, adopting the notation of (6.2.3) and applying (6.1.1) in two dimensions
(Green’s function being given by (6.2.7)), we ﬁnd
p(x. t ) ≈
−ρ
0
I sin(θ¡2)
π
√
[x[
k ∧
dx
0
dτ
(τ) · ∇ϕ
∗
(y)δ(y −x
0
(τ))
δ
t −τ −
[x[
c
0
dy
1
dy
2
dτ
=
−ρ
0
I sin(θ¡2)
π
√
[x[
¸
k ∧
dx
0
dt
· ∇ϕ
∗
¸
. [x[ →∞. (6.2.8)
where on the second line the term in the square brackets is evaluated at the
retarded position x
0
(t −[x[¡c
0
) of the vortex.
Now ϕ
∗
is the velocity potential of an ideal ﬂow around the edge of the
halfplane in the anticlockwise sense (with streamlines as in Fig. 6.2.3). It is
144 6 Vortex–Surface Interaction Noise in Two Dimensions
Fig. 6.2.3.
the real part of the complex potential
n = ϕ
∗
÷i ψ
∗
= −i
√
z. z = y
1
÷i y
2
.
where ϕ
∗
and the stream function ψ
∗
satisfy the Cauchy–Riemann equations
(see Section 4.5)
∂ϕ
∗
∂y
1
=
∂ψ
∗
∂y
2
.
∂ϕ
∗
∂y
2
= −
∂ψ
∗
∂y
1
.
A simple calculation shows that
k ∧ ∇ϕ
∗
= ∇ψ
∗
.
and therefore that
k ∧
dx
0
dt
· ∇ϕ
∗
= −
dx
0
dt
∧ k· ∇ϕ
∗
= −
dx
0
dt
· k ∧ ∇ϕ
∗
= −
dx
0
dt
· ∇ψ
∗
.
The acoustic pressure can therefore be put in the form
p(x. t ) ≈
ρ
0
I sin(θ¡2)
π
√
[x[
¸
dx
0
dt
· ∇ψ
∗
¸
≡
ρ
0
I sin(θ¡2)
π
√
[x[
¸
Dψ
∗
Dt
¸
. [x[ →∞.
(6.2.9)
where [Dψ
∗
¡Dt ] is evaluated at the retarded position of the vortex.
6.3 Inﬂuence of Vortex Shedding 145
Fig. 6.2.4.
The stream function ψ
∗
is constant on each of the parabolic streamlines of
the ideal ﬂow around the edge deﬁned by ϕ
∗
(Fig. 6.2.3). A vortex that trans
lated along one of these streamlines would be silent; the actual edgegenerated
sound depends on the rate at which the trajectory of the vortex cuts across the
streamlines of this ideal edge ﬂow. This is found as follows
ψ
∗
= −
√
r
0
cos
θ
0
2
.
Therefore,
Dψ
∗
Dt
= −
1
2
√
r
0
dr
0
dt
cos
θ
0
2
÷
√
r
0
2
sin
θ
0
2
dθ
0
dt
.
where r
0
. θ
0
are given by (4.6.7) (in which r is replaced by r
0
and θ by θ
0
).
Performing the calculations we ﬁnd (see Fig. 6.2.4)
p(x. t ) ≈
ρ
0
I
2
(4π¹)
2
¹
[x[
1
2
sin
θ
2
¸
It ¡8π¹
2
[1 ÷(It ¡8π¹
2
)
2
]
5¡4
¸
t −[x[¡c
0
. [x[ →∞.
where ¹ is the distance of closest approach of the vortex to the edge (where it
crosses the x
1
axis at time t = 0 in Fig. 4.6.2).
6.3 Inﬂuence of Vortex Shedding
We can use Equation (6.2.9) to form a qualitative picture of the inﬂuence of
vortex shedding on sound generation. Let the circulation of the vortex I in
146 6 Vortex–Surface Interaction Noise in Two Dimensions
Fig. 6.2.3 be in the indicated anticlockwise sense, so that ﬂuid near the edge
is induced to ﬂow in a clockwise direction around the edge, as implied by
the dashed curve in the ﬁgure. When the Reynolds number is large, inertial
forces actually cause the ﬂow to separate from the edge, resulting in the release
of vorticity of opposite sign from the edge into the wake. Let us assume for
simplicity that this shed vorticity rolls up into a concentrated core of strength
I
s
. Equation (6.2.9) then supplies the following prediction for the net acoustic
pressure
p(x. t ) ≈
ρ
0
sin(θ¡2)
π
√
[x[
I
¸
Dψ
∗
Dt
¸
I
÷I
s
¸
Dψ
∗
Dt
¸
I
s
. [x[ →∞.
where the derivatives are evaluated at the retarded positions of I and I
s
respec
tively. Both vortices translate across the curves ψ
∗
= constant in the direction
of decreasing ψ
∗
, and the derivatives therefore have the same sign. Hence, be
cause I and I
s
have opposite signs, sound produced by the shed vortex will
tend to cancel the edgegenerated sound attributable to the incident vortex I
alone.
This conclusionis applicable toa wide range of ﬂuid–structure interactions. A
typical interaction involves a bounded region of vorticity, called a ‘gust,’ swept
along in a nominally steady mean ﬂow. The localized velocity ﬁeld of the gust
is determined by the Biot–Savart formula (4.3.1). At high mean ﬂow speeds
it is sometimes permissible to neglect changes in the relative conﬁguration
of the vorticity distribution during its convection past an observation point,
the vorticity is then said to be frozen (at least temporarily) and the induced
velocity is steady in a frame translating with the gust. If the mean ﬂow carries
the gust past the surface S of a stationary body, the free ﬁeld induced velocity
determined by the Biot–Savart integral is said to produce an upwash velocity on
S; the actual velocity consists of the upwash velocity augmented by the velocity
ﬁeld required to satisfy the noslip condition on S. (In an ideal ﬂuid only the
normal component of the upwash velocity is cancelled on S.)
When a gust convects past a stationary airfoil the high Reynolds number
surface force (responsible for the sound) is given by (see (4.4.4))
F
i
= −ρ
0
V
∇Y
i
(y) · (ω ∧ v)(y. t ) d
3
y. Y
i
= y
i
−ϕ
∗
i
(y). (6.3.1)
The vector ∇Y
i
is the velocity of an ideal ﬂow past the airfoil that has unit
speed in the i direction at large distances from the airfoil. It is singular (or very
large) at the edges of the airfoil. These singularities have the following signiﬁ
cance, when the vorticity length scale is small compared to the airfoil chord the
6.3 Inﬂuence of Vortex Shedding 147
principal contribution to the integral is from vorticity in the neighborhoods of
the singularities. For example, for the strip airfoil of Fig. 3.6.3
Y
2
= Re(−i
z
2
−a
2
). z = y
1
÷i y
2
.
and ∇Y
2
becomes inﬁnite at the leading and trailing edges z = ∓a. An incident,
smallscale gust convecting in the y
1
direction in a mean ﬂow at speed U would
in practice induce shedding fromthe trailing edge at z = a. When this shedding
is ignored the force calculated from (6.3.1) has two principal components,
respectively fromgust elements near the leading and trailing edges. To calculate
the overall force, however, it is necessary to include the contribution from the
shed vorticity, which affects the motion only near the trailing edge when the
length scale of the wake vorticity is small. In the linearized treatment of this
case (discussed in more detail below), when both the gust and wake vorticity
are taken to convect at the same mean velocity U, it is known from unsteady
aerodynamics that the force component produced by the wake is equal and
opposite to that generated by the gust at the trailing edge (Sears 1941).
The effect of this cancellation can be approximated without calculating
any details of the shed vorticity. This is accomplished by formally deleting
the trailing edge singularity from∇Y
2
, and then ignoring the contribution to the
integral (6.3.1) fromthe shed vorticity. To understand this observe that, because
the value of the integral is dominated by vorticity near the edges, it is only the
behaviors of Y
2
near these edges that must be retained in the integrand, and Y
2
can therefore be replaced by the leading order terms in its expansions about the
edges. For the strip airfoil of Fig. 3.6.3, we would write
Y
2
= Re(−i
√
z −a
√
z ÷a) ∼ Re(
√
2a
√
z ÷a) ÷Re(−i
√
2a
√
z −a).
(6.3.2)
The last term is singular at the trailing edge; it is deleted and the following
approximation is used in (6.3.1) with the wake vorticity ignored:
Y
2
∼ Re(
√
2a
√
z ÷a). (6.3.3)
where the branch cut for
√
z ÷a is taken along the real axis from z = −a to
z = ÷∞.
6.3.1 Example: Surface Force Produced by a Periodic Gust
To illustrate the procedure consider incompressible ﬂow parallel to the airfoil
of Fig. 6.3.1 at speed U in the x
1
direction, in which a time harmonic vortex
148 6 Vortex–Surface Interaction Noise in Two Dimensions
Fig. 6.3.1.
sheet of vorticity
ω
I
= γ kδ(x
2
−h)e
−i ω(t −x
1
¡U)
. h >0. ω>0.
is convected past the airfoil at perpendicular distance h, where γ is the cir
culation per unit length of the sheet. The vortex sheet can be regarded as an
elementary model of a periodic wake behind a small diameter circular cylin
der upstream of the airfoil. The gust upwash velocity induces the shedding of
vorticity ω
S
from the trailing edge of the airfoil. When the reduced frequency
ωa¡U is large the hydrodynamic wavelength 2πU¡ω of the gust and wake is
much smaller than the airfoil chord, and the surface force is produced primarily
by the gust interaction with the leading edge at x
1
= −a.
The net force F
2
(per unit span) on the ﬂuid in the x
2
direction can therefore
be calculated from (6.3.1) by setting ω = ω
I
, where
ω
I
∧ v = γUjδ(x
2
−h)e
−i ω(t −x
1
¡U)
.
(j being a unit vector in the x
2
direction) and by replacing Y
2
by the righthand
side of (6.3.3):
F
2
= −ρ
0
γU
∞
−∞
∂Y
2
∂y
2
(y)δ(y
2
−h)e
−i ω(t −y
1
¡U)
dy
1
dy
2
≈ −ρ
0
γU
√
2a
∞
−∞
Re
i
2
√
y
1
÷i h ÷a
e
−i ω(t −y
1
¡U)
dy
1
= −ρ
0
γUa
πU
2i ωa
1
2
e
−ωh¡U−i ω(t ÷a¡U)
. (6.3.4)
The force can also be calculated exactly from linearized thin airfoil theory
with proper account taken of vortex shedding. This is the gust loading problem
6.3 Inﬂuence of Vortex Shedding 149
of classical aerodynamics (Sears 1941). The linear theory wake is treated as
a vortex sheet downstream of the edge, whose elements convect at the mean
stream velocity U. The strength of the vortex sheet is determined by imposing
the Kutta condition that the pressure (and velocity) should be ﬁnite at the edge
(Crighton 1985). For arbitrary values of the reduced frequency ωa¡U it is found
that
F
2
= πiρ
0
γUaS
ωa
U
e
−ωh¡U−i ωt
. (6.3.5)
where S(x) is the Sears function, which is deﬁned in terms of the Hankel
functions H
(1)
0
and H
(1)
1
by
S(x) =
2
πx
¸
H
(1)
0
(x) ÷i H
(1)
1
(x)
¸
. (6.3.6)
In the limit of high reduced frequency
S
ωa
U
∼
iU
2πωa
1
2
e
−i ωa¡U
.
and the substitution of this into (6.3.5) yields the prediction (6.3.4) determined
by the leading edge singularity of Y
2
. The plots in Fig. 6.3.2 of the real and
imaginary parts of S(ωa¡U) and its asymptotic limit show that the approx
imation (6.3.4) and the exact value (6.3.5) of the surface force agree when
ωa¡U >1.
Fig. 6.3.2.
150 6 Vortex–Surface Interaction Noise in Two Dimensions
Linear theory does not permit the corresponding force component at the
leading edge to be removed by vorticity production at the edge, because it re
quires vorticity shed there to be swept over the rigid surface of the airfoil on
which it cannot inﬂuence the force because ω ∧ v · ∇Y
2
≡ 0.
6.4 Blade–Vortex Interaction Noise in Two Dimensions
A rigid, twodimensional airfoil of chord 2a occupies −a x
1
a. x
2
= 0 in
the presence of a uniform mean stream at speed U in the positive x
1
direction.
A line vortex of strength I parallel to the airfoil span convects at the mean
ﬂow velocity at a constant distance h above the airfoil (Fig. 6.4.1). This is the
approximation of linearized thin airfoil theory, which is applicable when
I
h
<U.
that is, when the inﬂuence on the vortex trajectory of the induced velocity
(∼I¡h) of image vortices in the body of the airfoil is negligible. As the vortex
passes the airfoil new vorticity is shed from the trailing edge into the airfoil
wake, which is assumed to consist of a vortex sheet stretching along the x
1
axis
from x
1
= a to x
1
= ÷∞.
Let us ﬁrst consider the potential ﬂow interaction of the vortex and airfoil,
when no account is taken of vortex shedding. Suppose the vortex passes above
the midchord of the airfoil at time t = 0, then
ω = Ikδ(x
1
−Ut )δ(x
2
−h). v = Ui.
Hence,
ω ∧ v = IUjδ(x
1
−Ut )δ(x
2
−h). (6.4.1)
where i and j are unit vectors in the x
1
and x
2
directions. The acoustic pressure
Fig. 6.4.1.
6.4 Blade–Vortex Interaction Noise in Two Dimensions 151
generated when the wake is ignored is given by (6.2.1) with
Y
1
= y
1
. Y
2
= Re (−i
z
2
−a
2
). z = y
1
÷i y
2
.
Thus, ∇Y
1
= i and ω ∧ v · ∇Y
1
≡ 0, and (6.2.1) reduces to
p(x. t ) ≈
−ρ
0
IUx
2
2π
√
2c
0
[x[
3
2
∂
∂t
t −[x[¡c
0
−∞
∂Y
2
∂y
2
(Uτ. h)
dτ
√
t −τ −[x[¡c
0
.
=
−ρ
0
IU cos O
2π
√
2c
0
[x[
1
2
∂
∂t
t −[x[¡c
0
−∞
∂Y
2
∂y
2
(Uτ. h)
dτ
√
t −τ −[x[¡c
0
.
[x[ →∞. (6.4.2)
where O = cos
−1
(x
2
¡[x[) is the angle between the radiation direction x and
the normal to the airfoil (the x
2
axis). The sound can be attributed to a dipole
source orientated in the x
2
direction.
The radiation produced when a vortex passes very close to the airfoil (so that
h <a) is likely to be particularly intense. The dominant interactions occur as
the vortex passes the edges, where the time scales of the motions ∼h¡U. Thus,
the characteristic frequency
ω ∼
U
h
and the reduced frequency
ωa
U
∼
a
h
·1.
We may therefore regard the leading and trailing edges as independent sources
of sound, and calculate their individual contributions by using the local approx
imation (6.3.2). For the acoustic pressure p
LE
, say, produced at the leading edge
(x
1
= −a) we take
Y
2
∼ Re (
√
2a
√
z ÷a).
so that (6.4.2) becomes
p
LE
≈
−ρ
0
IU cos O
2π
√
2c
0
[x[
1
2
∂
∂t
t −[x[¡c
0
−∞
Re
i
√
2a
2(Uτ ÷i h ÷a)
1
2
dτ
√
t −τ −[x[¡c
0
. [x[ →∞.
To evaluate the integral, make the substitution j=
√
t −τ −[x[¡c
0
and
152 6 Vortex–Surface Interaction Noise in Two Dimensions
perform the differentiation with respect to time. Then,
p
LE
≈
−ρ
0
IU cos O
2π
√
c
0
a
[x[
1
2
∂
∂t
∞
0
Re
¸
i
(U[t ] ÷i h ÷a −Uj
2
)
1
2
dj
=
ρ
0
IU
2
cos O
4π
√
c
0
a
[x[
1
2
∞
0
Re
¸
i
(U[t ] ÷i h ÷a −Uj
2
)
3
2
dj.
where [t ] = t −
[x[
c
0
.
The additional substitution j=1¡ξ transforms the integrand into an exact dif
ferential, leading ﬁnally to
p
LE
≈
ρ
0
IU
2
cos O
4π
√
c
0
a
[x[
1
2
∞
0
Re
¸
i ξ
[(U[t ] ÷i h ÷a)ξ
2
−U]
3
2
dξ
=
ρ
0
IU
√
M cos O
4πa
a
[x[
1
2
U[t ]
a
÷1
U[t ]
a
÷1
2
÷
h
a
2
. [x[ →∞. (6.4.3)
where M = U¡c
0
.
The corresponding nondimensional pressure signature
p
LE
ρ
0
IU
√
M cos O
4πa
a
[x[
1
2
is plotted in Fig. 6.4.2 as the solid curve when h¡a =0.2. The pressure ﬁeld
Fig. 6.4.2.
6.4 Blade–Vortex Interaction Noise in Two Dimensions 153
is generated predominantly as the vortex passes the leading edge of the air
foil at the retarded time [t ] =−a¡U, with characteristic frequency ω∼U¡h.
According to Section 6.3, at high reduced frequencies the sound pressure p
TE
,
say, generated by the potential ﬂow interaction of the vortex with the trailing
edge is cancelled by that produced by the wake vorticity; the solid curve in
Fig. 6.4.2 is therefore representative of the whole of the radiation produced by
the blade–vortex interaction.
To determine p
TE
the calculation described above for p
LE
is repeated after
setting
Y
2
= Re(−i
√
2a
√
z −a)
in (6.4.2), leading to
p
TE
≈
−ρ
0
IU cos O
4π
√
c
0
a
[x[
1
2
∂
∂t
t −[x[¡c
0
−∞
Re
1
(Uτ ÷i h −a)
1
2
dτ
√
t −τ −[x[¡c
0
=
−ρ
0
IU
√
M cos O
4πa
a
[x[
1
2
h
a
U[t ]
a
−1
2
÷
h
a
2
. [x[ →∞. (6.4.4)
This is large at the retarded times during which the vortex is close to the trailing
edge. Thus, when the contribution from the wake is ignored (which is equal
and opposite to p
TE
) the overall acoustic pressure signature is given nondimen
sionally by
( p
LE
÷ p
TE
)
ρ
0
IU
√
M cos O
4πa
a
[x[
1
2
.
which is plotted as the broken line curve in Fig. 6.4.2.
The leading and trailing edge generated components p
LE
and p
TE
have dif
ferent waveforms, even though they are produced by the vortex interacting with
geometrically identical airfoil edges. This is because the integral in (6.4.2) de
termines the acoustic pressure at the retarded time [t ] in terms of interactions
between the vortex and the airfoil at all earlier retarded times; it is a further
consequence of the twodimensional character of the acoustic sources, accord
ing to which, after the ﬁrst arrival of sound fromthe nearest point on the source,
additional contributions to the sound continue to be received indeﬁnitely in time
from more distant parts of the source.
154 6 Vortex–Surface Interaction Noise in Two Dimensions
Problems 6
1. Starting from the formula
G ≈
¯ x · Y
2πc
0
∂
∂t
∞
0
δ
t −τ −
[¯ x[
2
÷ξ
2
c
0
dξ
([¯ x[
2
÷ξ
2
)
.
derive the farﬁeld approximation (6.1.5) for the dipole component of the
twodimensional compact Green’s function by writing
δ
t −τ −
[¯ x[
2
÷ξ
2
c
0
=
1
2π
∞
−∞
e
−i ω(t −τ−
√
[¯ x[
2
÷ξ
2
c
0
)
dω.
and applying Formula (5.2.11).
2. Investigate the production of sound by the low Mach number motion of a
line vortex of strength I that is parallel to a rigid circular cylinder of radius a
whose axis coincides with the x
3
coordinate axis (c.f., Section 4.6, Example
2). Assume that there is no net circulation around the cylinder, and that there
is a mean ﬂow past the cylinder which has speed U in the x
1
direction when
[x
1
[ · a. If the vortex is initially far upstream of the cylinder at a distance
h from the x
1
axis, examine the production of sound for different values of
the nondimensional parameter I¡Uh.
3. Aline vortexof strengthIis parallel toa rigidairfoil occupying−a x
1
a.
x
2
= 0. −∞x
3
∞, in the presence of a mean ﬂow at speed U in the
x
1
direction. The vortex is initially far upstream of the airfoil at a vertical
standoff distance h above the plane of the airfoil. There is no net circulation
around the airfoil. If the motion occurs at very small Mach number, calculate
the sound produced when the vortex passes the airfoil for different values
of the nondimensional velocity ratio I¡Uh. When U ·I¡h, estimate the
inﬂuence on the sound of vortex shedding fromthe trailing edge of the airfoil.
4. A line vortex of strength I is parallel to a rigid airfoil occupying
−a x
1
a. x
2
= 0. −∞x
3
∞
in ﬂuid at rest at inﬁnity. The vortex is in periodic motion around the airfoil
under the inﬂuence of image vortices in the absence of a mean circulation
around the airfoil. Calculate the sound produced when the motion occurs at
a very small Mach number, and show that it can be attributed to two dipole
sources orientated in the x
1
and x
2
directions. Explain the signiﬁcance of
these sources in terms of the corresponding components of the unsteady
force between the ﬂuid and airfoil.
Problems 6 155
5. A line vortex of strength I traverses a path of the kind illustrated in the
ﬁgure past a twodimensional, thin rigid barrier of length d at right angles
to a plane wall at x
2
= 0. There is a low Mach number mean potential ﬂow
over the barrier that has speed U parallel to the wall at large distances from
the barrier. If the distance of the vortex from the wall is h when the vortex
is far upstream of the barrier, calculate the sound produced as the vortex
passes the barrier for different values of I¡Uh. Explain what happens when
I¡h < U. Discuss the forces exerted on the barrier by the ﬂow, and how
they contribute to the radiation.
7
Problems in Three Dimensions
7.1 Linear Theory of Vortex–Airfoil Interaction Noise
Consider aninhomogeneous ﬁeldof vorticity, a gust, convectedinhighReynolds
number, homentropic ﬂow past a stationary rigid airfoil (Fig. 7.1.1). The undis
turbed ﬂow has speed U in the x
1
direction, where the origin is at a convenient
point within the airfoil, with x
3
along the span and x
2
vertically upward. The
Mach number M = U¡c
0
is sufﬁciently small that convection of sound by the
ﬂow can be ignored, and the airfoil chord can be assumed to be acoustically
compact.
The vortex sound source div(ω∧v) includes vorticity in the gust together with
any vorticity shed from the airfoil, either in response to excitation by the gust,
or as tip vortices responsible for the mean lift. The problemcan be linearized by
assuming that u <U, where curl u = ω, that is, by requiring the gustinduced
velocity, and the perturbation velocities caused by airfoil thickness, twist, cam
ber, and the angle of attack, to be small. When div(ω ∧ v) is expanded about
the undisturbed mean ﬂow, only the gust vorticity and additional vorticity shed
when the gust encounters the airfoil contribute to the acoustic radiation to ﬁrst
order. In other words, thickness, twist, camber, and angle of attack may ignored,
and the airfoil regarded as a rigid lamina in the plane x
2
= 0. In this approx
imation, quadrupoles are neglected and vorticity convects as a frozen pattern
of vortex ﬁlaments at the undisturbed mean stream velocity U=(U. 0. 0). In
particular, the wake vorticity is conﬁned to a vortex sheet downstream of the
trailing edge.
When convection of sound by the ﬂow is neglected, the linearized form of
the vortex sound equation (5.2.5) becomes
1
c
2
0
∂
2
∂t
2
−∇
2
B = div(ω ∧ U). (7.1.1)
156
7.1 Linear Theory of Vortex–Airfoil Interaction Noise 157
Fig. 7.1.1.
with solution
B(x. t ) =
p(x. t )
ρ
0
≈ −
(ω ∧ U)(y. τ) ·
∂G
∂y
(x. y. t −τ) d
3
y dτ. [x[ →∞.
(7.1.2)
where ∂G¡∂y
2
=0 on both sides (y
2
=±0) of the projection of the airfoil plan
form onto the y
1
. y
3
plane. At sufﬁciently small Mach numbers G may be
approximated by Green’s function for an airfoil of compact chord.
The sound produced when a localized, high Reynolds number frozen gust
ω(x−Ut ) ≡ω(x
1
−Ut. x
2
. x
3
) is swept past the airfoil of Fig. 7.1.1 is therefore
given by (5.4.4) with the convection velocity v replaced by U, and the Kirchhoff
vector Y by
Y
1
= y
1
. Y
2
= y
2
−ϕ
∗
2
(y). Y
3
= y
3
.
Thus, because
ωd
3
y ≡ 0, Equation (5.4.4) reduces to
p(x. t ) ≈
−ρ
0
U cos O
4πc
0
[x[
∂
∂t
¸
ω
3
∂Y
2
∂y
2
−ω
2
∂Y
2
∂y
3
¸
t −[x[¡c
0
d
3
y. [x[ →∞.
(7.1.3)
where O = cos
−1
(x
2
¡[x[) is the angle between the radiation direction and the
normal to the airfoil, and the origin is taken in the airfoil within the interaction
region.
Let the interaction occur at an inboard location where the chord may be
regarded as constant, with both the leading and trailing edges at right angles to
the mean ﬂow (so that ∂Y
2
¡∂y
3
< ∂Y
2
¡∂y
2
). The planform in the interaction
158 7 Problems in Three Dimensions
Fig. 7.2.1.
regionis thenlocallythe same as that of the twodimensional airfoil of Fig. 7.2.1,
and Y
2
can be approximated as in (3.9.2) for constant a ≡a(y
3
), such that 2a
is equal to the local chord of the airfoil. Then, (7.1.3) becomes
p(x. t ) ≈
−ρ
0
U cos O
4πc
0
[x[
∂
∂t
¸
ω
3
∂Y
2
∂y
2
¸
t −[x[¡c
0
d
3
y. [x[ →∞. (7.1.4)
which reveals that only the spanwise component of vorticity contributes to the
production of sound.
According to Equation (5.4.3) (in which dU
j
¡dt = 0 for a stationary airfoil),
this result can also be expressed in the form
p(x. t ) ≈
cos O
4πc
0
[x[
∂ F
2
∂t
t −
[x[
c
0
. [x[ →∞.
F
2
(t ) = −ρ
0
U
ω
3
(y. t )
∂Y
2
∂y
2
(y) d
3
y. (7.2.5)
where −F
2
is the unsteady airfoil lift during the interaction when the motion
is regarded as incompressible, and the vorticity ω
3
includes contributions from
the impinging gust together with any shed into the vortex sheet wake.
7.2 Blade–Vortex Interactions in Three Dimensions
The calculations canbe performedexplicitlyfor a gust inthe formof a rectilinear
line vortex. Let the vortex have circulation I and be orientated with its axis in
the direction of the unit vector n, as indicated in Fig. 7.2.1. The mean ﬂowspeed
is sufﬁciently large that the vortex can be assumed to maintain its rectilinear
form after being cut by the leading edge of the airfoil. Choose the origin on the
airfoil midchord such that the axis of the vortex passes through the origin at
time t = 0. Any point x on the vortex can then be represented in the parametric
7.2 Blade–Vortex Interactions in Three Dimensions 159
form
x = (Ut. 0. 0) ÷sn. −∞s ∞. (7.2.1)
where s is distance measured along the vortex from its point of intersection
with the plane of the airfoil. Then, if s
⊥
denotes vector distance measured in
the normal direction from the vortex axis,
ω = Inδ(s
⊥
).
where the polar angles θ. φ in Figure 7.2.1 deﬁne the orientation of the unit
vector
n = (sin θ cos φ. sin θ sin φ. cos θ). 0 θ π. 0 φ 2π.
Because of vortex shedding from the trailing edge, most of the sound is
generated when the vortex is cut by the leading edge. The inﬂuence of the shed
vorticity can be formally included by the procedure described in Sections 6.3
and 6.4 by expanding Y
2
about its singularity at the leading edge, that is, by
setting
Y
2
∼ Re(
√
2a
√
z ÷a). z = y
1
÷i y
2
.
with the branch cut for the square root taken along the z axis from z = −a to
z = ÷∞. Thus, by recalling the Relation (7.2.1), Equation (7.1.4) becomes
p(x. t ) ≈
−n
3
ρ
0
IU cos O
4πc
0
[x[
∂
∂t
Re
¸
δ(s
⊥
)
i
√
2a
2
√
y
1
÷i y
2
÷a
t −[x[¡c
0
d
2
s
⊥
ds
=
−n
3
ρ
0
IU cos O
4
√
2πc
0
[x[
∂
∂t
Re
∞
−∞
i
√
a
√
U[t ] ÷s(n
1
÷i n
2
) ÷a
ds
=
n
3
ρ
0
IU
2
cos O
8
√
2πc
0
[x[
Re
∞
−∞
i
√
a
(U[t ] ÷s(n
1
÷i n
2
) ÷a)
3
2
ds
=
−n
3
ρ
0
IU
2
cos O
4
√
2πc
0
[x[
Re
¸
i
√
a
(U[t ] ÷s(n
1
÷i n
2
) ÷a)
1
2
(n
1
÷i n
2
)
¸
∞
−∞
.
(7.2.2)
where [t ] = t −[x[¡c
0
is the retarded time.
By referring to Fig. 7.2.2 it will be seen that the last line of (7.2.2) is zero
when
U[t ] ÷a 0.
160 7 Problems in Three Dimensions
Fig. 7.2.2.
before the vortex is cut by the leading edge of the airfoil. At later times the
integration along the vortex axis over the inﬁnite range −∞s ∞must be
split, as indicated in the ﬁgure, into the two parts −∞s −0. ÷0 s ∞,
because the square root in the last line of (7.2.2) is discontinuous across the
airfoil; for example, when n
2
>0 the square root is real and positive on the
upper surface (s = ÷0) and real and negative at s = −0. Hence, the acoustic
pressure becomes
p(x. t ) ≈
ρ
0
IUM cos O
2
3
2
π[x[
[sinφ[
tan θ
H
U[t ]
a
÷1
U[t ]
a
÷1
. (7.2.3)
The pressure pulse accordingly begins with a singular peak at the instant at
which the vortex is severed by the leading edge of the airfoil at the retarded time
U[t ]¡a = −1. The waveform is illustrated by the dotted curve in Fig. 7.2.3,
which is a plot of
p(x. t )
(ρ
0
IUM cos O¡[x[)
=
[sinφ[
2
3
2
π tan θ
H
U[t ]
a
÷1
U[t ]
a
÷1
for θ = 85
◦
. φ = 90
◦
.
The inﬁnite singularity in the pressure is absent for a vortex of nonzero
core radius R, say. If, for example, the vorticity is assumed to be distributed
according to the Gaussian formula
ω(x) =
Ine
−(s
⊥
¡R)
2
π R
2
.
as a function of distance s
⊥
from the vortex axis, the acoustic pressure is found
7.2 Blade–Vortex Interactions in Three Dimensions 161
Fig. 7.2.3.
to be given by (Howe 1998a)
p(x. t )
(ρ
0
IUM cos O¡[x[)
≈
[ sin φ[
3
2
8 tan θ
a
π R
1
2
`(α). [x[ →∞. for R <a.
(7.2.4)
where
`(α) = [α[
1
2
¸
I
−
1
4
α
2
8
÷sgn(α)I 1
4
α
2
8
¸
e
−α
2
¡8
. α =
2a[ sin φ[
R
U[t ]
a
÷1
.
and I
±
1
4
are modiﬁed Bessel functions of the ﬁrst kind. The pressure signature
predicted by (7.2.4) for
R = 0.1a. θ = 85
◦
. φ = 90
◦
is plotted as the solid curve in Fig. 7.2.3, and differs negligibly from the line
vortex prediction when U[t ]¡a >−1.
The brokenline curve inFig. 7.2.3represents the pressure signature produced
by the potential ﬂow interaction between the ﬁnite core vortex and airfoil (i.e.,
when vortex shedding is ignored). It is an odd function of the retarded time
[t ]. The large negative peak produced as the vortex crosses the trailing edge (at
U[t ]¡a =1) is cancelled by an equal and opposite contribution generated by the
wake. The reader can easily show that, when the ﬁnite size of the vortex core is
162 7 Problems in Three Dimensions
ignored, the potential ﬂow, trailing edge generated pressure pulse is given by
p(x. t ) ≈ −
ρ
0
IUM cos O
2
3
2
π[x[
[sinφ[
tan θ
H
1 −
U[t ]
a
1 −
U[t ]
a
.
7.3 Sound Produced by Vortex Motion near a Sphere
The sound generated when a nominally rectilinear vortex is swept past a com
pact rigid body can also be treated in a linearized fashion, by assuming that
each element of the vortex core is convected along a streamline of the steady
undisturbed mean ﬂow at the local mean velocity. It is not generally possible,
however, to include the inﬂuence of vortex shedding in a satisfactory manner,
except perhaps for streamlined body shapes that are amenable to treatment by
the strip theory of unsteady aerodynamics.
To illustrate the procedure, consider a rigid sphere of radius a with center at
the coordinate origin in the presence of a low Mach number irrotational mean
ﬂow which is in the x
1
direction at speed U for [x[ ·a. The mean velocity at
x can therefore be written
U = U∇X
1
(x). (7.3.1)
where X
1
(x) is the x
1
component of the Kirchhoff vector for the sphere, which
has the general form (Table 3.9.1)
X
i
= x
i
1 ÷
a
3
2[x[
3
. (7.3.2)
Suppose a line vortex of strength I is initially far upstream of the sphere
and parallel to the x
3
axis at a distance h above the plane x
2
=0. The vortex is
convected toward the sphere by the mean ﬂow. The part of the vortex that passes
close to the sphere must evidently be deformed to pass around the sphere; more
distant parts of the vortex (at [x
3
[ ·a) are unaffected and remain parallel to the
x
3
direction during the whole of the interaction. The shape of the distorted vortex
will be symmetric with respect to the midplane x
3
=0; the vortex element
initially on x
3
=0 remains on this plane of symmetry as it convects past the
sphere along a mean streamline, as illustrated in Fig. 7.3.1, which shows the
motion in the plane x
3
= 0.
The shape of the vortex at time t is determined by the solution of the equations
dx
1
dt
= U
∂ X
1
∂x
1
(x).
dx
2
dt
= U
∂ X
1
∂x
2
(x).
dx
3
dt
= U
∂ X
1
∂x
3
(x).
7.3 Sound Produced by Vortex Motion near a Sphere 163
Fig. 7.3.1.
for each element of the vortex. If the undistorted parts of the vortex (at [x
3
[ ·a)
are assumed to convect across the plane x
1
= 0 at time t = 0, these equations
are to be integrated subject to the initial conditions
x
1
= Ut . x
2
= h. x
3
= x
0
3
t →−∞.
where x
0
3
is the initial spanwise location of the vortex element.
In terms of the nondimensional variables
T =
Ut
a
. ¯ x =
x
a
.
the equations of motion of a point on the vortex are found to be
d ¯ x
1
dT
= 1 ÷
¯ x
2
2
÷ ¯ x
2
3
−2¯ x
2
1
2
¯ x
2
1
÷ ¯ x
2
2
÷ ¯ x
2
3
5
2
.
d ¯ x
2
dT
=
−3¯ x
1
¯ x
2
2
¯ x
2
1
÷ ¯ x
2
2
÷ ¯ x
2
3
5
2
.
d ¯ x
3
dT
=
−3¯ x
1
¯ x
3
2
¯ x
2
1
÷ ¯ x
2
2
÷ ¯ x
2
3
5
2
.
These are solved (for example, by the Runge–Kutta method described in
Section 4.6) by starting the integration at T =−10, say. It can be safely assumed
that the sphere has no perceptible inﬂuence on vortex elements initially located
at [ ¯ x
3
[ >10. Because the motion is symmetric about ¯ x
3
=0 the solutions are
required only for the N ÷1 vortex elements with respective the initial positions
¯ x
1
= T. ¯ x
2
=
h
a
. ¯ x
3
= ¯ x
n
3
at T = −10.
where ¯ x
n
3
= 10n¡N. 0 ≤n ≤ N, and N is a suitably large integer.
164 7 Problems in Three Dimensions
Fig. 7.3.2.
Figure 7.3.2 illustrates successive calculated positions of the vortex with in
creasing values of the time T = Ut ¡a for the case h¡a = 0.2. The distortion of
the vortex ﬁrst becomes evident at about T = −2. The hairpin loop is formed
because the translation velocities of vortex elements close to the sphere are small
inthe neighborhoodof the stagnationpoints just infront andjust tothe rear of the
sphere; the accelerated motion over the upper surface of the sphere is insufﬁcient
to counteract the formation of the loop. In reality, of course, the motion would
be strongly inﬂuenced by large selfinduced velocities, by image vorticity in
the sphere, and by viscous diffusion of vorticity from the vortex and from the
surface of the sphere, none of which is accounted for in the present calculation.
The sound generated during this potential ﬂow interaction can be calculated
using the Formula (5.4.4) when the sphere is compact:
p(x. t ) ≈
−ρ
0
x
j
4πc
0
[x[
2
∂
∂t
(ω ∧ v)
y. t −
[x[
c
0
· ∇Y
j
(y) d
3
y. [x[ →∞.
(7.3.3)
where
v = U∇Y
1
(y).
7.3 Sound Produced by Vortex Motion near a Sphere 165
There is no unsteady drag contribution to (7.3.3) from j = 1 because ω∧∇Y
1
·
∇Y
1
≡ 0. Similarly, there can be no net sideforce on the sphere because of
the symmetric form of the vortex, and therefore there will be no contribution
from j =3. The sound is accordingly produced by a dipole source orientated
in the x
2
direction; that is, the interaction produces an unsteady lift force in this
direction which is responsible for the sound, which has the representation
p(x. t ) ≈
−ρ
0
U cos O
4πc
0
[x[
∂
∂t
(ω · ∇Y
1
∧ ∇Y
2
)
y. t −
[x[
c
0
d
3
y. [x[ →∞.
(7.3.4)
where O = cos
−1
(x
2
¡[x[) is the angle between the observer direction x and the
x
2
axis.
To evaluate the integral, write
ω = Iδ(s
⊥
)ˆ s. d
3
y = d
2
s
⊥
ds.
where ˆ s is a unit vector locally parallel to ω, s
⊥
is the vector distance measured
in the normal direction from the local axis of the vortex, and s is distance
measured along the vortex in the direction of ω. The integral (7.3.4) can then
be cast in the following nondimensional form, suitable for numerical evaluation,
p(x. t )
ρ
0
IUM cos O¡4π[x[
= −
∂
∂T
∞
−∞
[ˆ s · ∇Y
1
∧∇Y
2
] d ¯ s. ¯ s =
s
a
. M =
U
c
0
.
(7.3.5)
where the integrand is evaluated at the retarded position of the distorted vortex.
Now the integral in (7.3.5) is divergent, because ˆ s · ∇Y
1
∧ ∇Y
2
→ 1 as
¯ s →±∞. The divergence is not real, however, but a consequence of the formal
operations used in the application of the compact Green’s function. The inﬁnite
contributions to the integral from large values of ¯ s are equal at successive
retarded locations of the vortex, and disappear on differentiation with respect
to T. The integral can therefore be evaluated numerically by restricting the range
of integration to a ﬁnite interval, say, −10  ¯ s 10, because the contributions
at larger values of ¯ s are the same for all retarded times, and give no contribution
to the sound when differentiated.
Typical plots of the calculated nondimensional pressure (7.3.5) are shown in
Fig. 7.3.3 for two values of h¡a; they illustrate how the sound level decreases
as the initial standoff distance h of the vortex increases relative to the radius a
of the sphere.
166 7 Problems in Three Dimensions
Fig. 7.3.3.
7.4 Compression Wave Generated When a Train Enters a Tunnel
A train entering a tunnel pushes aside the stationary air, most of which ﬂows
over the train and out of the tunnel portal, but the buildup of pressure just ahead
of the train propagates into the tunnel as a compression wave at the speed of
sound. In a long tunnel the compression wavefront can experience nonlinear
steepening that is ultimately manifested as a loud, impulsive bang or ‘crack’
(calleda micropressure wave) radiatingout of the distant tunnel exit. Inaddition
inaudible lowfrequency pressure ﬂuctuations called infrasound (at frequencies
∼10–20 Hz) are radiated from the tunnel portal into the open air when the train
enters and leaves the tunnel. All of these waves are indicated schematically in
Fig. 7.4.1. Their effects become pronounced when the train speed U exceeds
Fig. 7.4.1.
7.4 Compression Wave Generated When a Train Enters a Tunnel 167
Fig. 7.4.2.
about 200 km/h (125 mi/h), when in particular the micropressure wave and the
infrasound can cause vibrations and annoying structural rattles in neighboring
buildings.
The formation of the compression wave can be studied in terms of the simpler
problem involving an axisymmetric train entering axisymmetrically a semi
inﬁnite, circular cylindrical duct of radius R and crosssectional area A = π R
2
(Fig. 7.4.2). Let the train be travelling at constant speed U in the negative x
1
direction, where the origin O is at the center of the tunnel entrance plane, so that
the x
1
axis coincides with the axis of the tunnel. Denote the pressure, density,
and speed of sound in the air respectively by ¯ p. ρ and c. They vary with position
and time within the tunnel, and their corresponding undisturbed values are
p
0
. ρ
0
. c
0
.
The cross section of the train is assumed to become uniform with constant
area A
0
=πh
2
at a distance L fromthe nose of the train, where h is the uniform
maximumtrain radius. The aspect ratio h¡L of the nose is takentobe sufﬁciently
small, andthe trainproﬁle sufﬁcientlystreamlined, toensure that ﬂowseparation
does not occur. In practice the Mach number M =U¡c
0
does not exceed 0.4,
and the blockage A
0
¡A≤0.2. If heat transfer and frictional losses are neglected
during the initial stages of wave formation, the air ﬂow may be regarded as
homentropic, and the compression wave can be calculated using the vortex
sound equation (5.2.5)
D
Dt
1
c
2
D
Dt
−
1
ρ
∇· (ρ∇)
B =
1
ρ
div(ρω ∧ v). (7.4.1)
The air in the compression wave region ahead of the train may be regarded as
linearly perturbed from its mean state, with B ≈ p¡ρ
0
≡ ( ¯ p − p
0
)¡ρ
0
. In this
168 7 Problems in Three Dimensions
simple model the vorticity ω vanishes everywhere except within the outer shear
layer of the exit ﬂow of the air displaced when the train enters the tunnel (see
Fig. 7.4.2).
Let f ≡ f (x
1
÷Ut. x
2
. x
3
) =0 be a control surface S contained within the
ﬂuid that just encloses the moving train, with f 0 inside S (in the region oc
cupied by the train) and f >0 outside. The surface is ﬁxed relative to the train,
and the inﬂuence of the train on its surroundings can be represented in terms
of monopole and dipole sources on S. In the usual way, multiply (7.4.1) by
H ≡ H( f ) and rearrange (noting that DH¡Dt =0) to obtain
D
Dt
1
c
2
D
Dt
−
1
ρ
∇ · (ρ∇)
(HB)
=
1
ρ
div(Hρω ∧ v) −(∇B ÷ω ∧ v) · ∇H −
1
ρ
div(ρB∇H). (7.4.2)
This is a generalization of Equation (5.3.3). The two terms on the righthand side
involving ∇H respectively represent monopole and dipole sources distributed
over the moving surface f (x
1
÷Ut. x
2
. x
3
) = 0.
When frictional losses are neglected Crocco’s equation (4.2.5) reduces to
∂v¡∂t = −∇B −ω ∧ v, so that the source terms can be written
∂
∂t
(U · ∇H) −v · ∇
∂ H
∂t
−
1
ρ
div (ρB∇H) ÷
1
ρ
div(Hρω ∧ v).
where U = (−U. 0. 0). The compressibility of the air adjacent to S and within
the very lowMach number exterior ﬂowfromthe tunnel portal can be neglected
when M(A
0
¡A)
2
<1 (Howe et al. 2000), and the source approximated further
by
∂
∂t
(U · ∇H) ÷div (v U · ∇H) −div
¸
p
ρ
0
÷
1
2
:
2
∇H
¸
÷div(Hω ∧ v).
where the relation ∂ H( f )¡∂t = −U · ∇H( f ) has been used.
Thus, if the nonlinear terms onthe left of (7.4.2) (whichaffect the propagation
of the compression wave) are also ignored, the equation ﬁnally reduces to
1
c
2
0
∂
2
∂t
2
−∇
2
(HB) =
∂
∂t
(U · ∇H) ÷div (v U · ∇H)
−div
¸
p
ρ
0
÷
1
2
:
2
∇H
¸
÷div(Hω ∧ v). (7.4.3)
7.4 Compression Wave Generated When a Train Enters a Tunnel 169
Fig. 7.4.3.
7.4.1 Linear Theory
When the blockage A
0
¡Ais small it is sufﬁcient to retain only the ﬁrst monopole
source on the righthand side of Equation (7.4.3). This source can be simpliﬁed
when the aspect ratio h¡L <1 by introducing a slender body approximation.
To do this, consider a system of cylindrical coordinates in which r =
√
x
2
2
÷ x
2
3
is the perpendicular distance from the axis of the train. The control surface
equation
f (x
1
÷Ut. x
2
. x
3
) = 0 can be written r =
¯
f (x
1
÷Ut ).
where A
T
(s) = π
¯
f
2
(s) is the crosssectional area of the train at distance s from
the nose, and the nose is assumed to cross the tunnel entrance plane (x
1
=0)
at t =0. As the train moves (to the left in Fig. 7.4.3) the rate at which air is
displaced by a section of the train of length dx
1
is
U2π
¯
f (x
1
÷Ut ) d
¯
f (x
1
÷Ut ) = U2π
¯
f (x
1
÷Ut )
∂
¯
f
∂x
1
(x
1
÷Ut ) dx
1
≡ U
∂A
T
∂x
1
(x
1
÷Ut ) dx
1
.
Thus,
U
∂A
T
∂x
1
(x
1
÷Ut ) = monopole source strength per unit length of the train.
When h¡L <1, we can collapse the monopole source distribution over the
surface of the train into a line source concentrated on its axis, and approximate
the monopole on the righthand side of Equation (7.4.3) by
∂
∂t
(U · ∇H) (x. t ) ≈
∂
∂t
U
∂A
T
∂x
1
(x
1
÷Ut )δ(x
2
)δ(x
3
)
.
170 7 Problems in Three Dimensions
The source strength is proportional to the rate at which the train cross section
changes with distance along the train, and is nonzero only in the vicinity of the
train nose (and also the tail).
The corresponding approximation of Equation (7.4.3) is therefore
1
c
2
0
∂
2
∂t
2
−∇
2
(HB) =
∂
∂t
U
∂A
T
∂x
1
(x
1
÷Ut )δ(x
2
)δ(x
3
)
. (7.4.4)
where B → p¡ρ
0
in the linear acoustic region ahead of the train.
The monopole in this equation does not depend on time when viewed in a
reference frame moving at the uniformsubsonic speedU of the train. The source
therefore creates only a nonacoustic near ﬁeld when travelling within the tunnel
or in free space far fromthe tunnel entrance (c.f., Question 4 of Problems 1 when
q
0
(t ) =constant). The compression wave is produced when the near ﬁeld of the
source interacts with the tunnel portal, as the train nose enters the tunnel. This
occurs over a time ∼R¡U, so that the characteristic thickness of the wavefront
∼R¡M ·R. Equation (7.4.4) can therefore be solved by using the compact
Green’s function (3.9.13) for a duct entrance
G(x. y; t −τ) ≈
c
0
2A
¸
H
t −τ −
[ϕ
∗
(x) −ϕ
∗
(y)[
c
0
− H
t −τ ÷
ϕ
∗
(x) ÷ϕ
∗
(y)
c
0
¸
.
where ϕ
∗
(x) is the velocity potential of a uniform incompressible ﬂow out of
the tunnel portal that has unit speed far inside the tunnel (see (3.9.14)).
Thus, at x within the tunnel, ahead of the train where B = p¡ρ
0
, we have
p ≡ p
t ÷
x
1
c
0
≈ρ
0
∂
∂t
∞
−∞
U
∂A
T
∂y
1
(y
1
÷Uτ)G(x. y
1
. 0. 0; t −τ) dy
1
dτ
=
ρ
0
Uc
0
2A
∞
−∞
{A
/
T
(y
1
− Mϕ
∗
(y
1
. 0. 0) ÷U[t ])
−A
/
T
(y
1
÷ Mϕ
∗
(y
1
. 0. 0) ÷U[t ])} dy
1
. (7.4.5)
where the prime on A
T
denotes differentiation with respect to the argument,
and [t ] =t ÷ (x
1
− ¹
/
)¡c
0
is the effective retarded time. Because nonlinear
propagation terms have been ignored, this approximation determines the initial
form of the compression wave proﬁle, before the onset of nonlinear steepening.
It is therefore applicable within the region several tunnel diameters ahead of
the train, during and just after tunnel entry.
7.4 Compression Wave Generated When a Train Enters a Tunnel 171
The main contributions to the Integral (7.4.5) are from the vicinities of the
nose and tail of the train, where the crosssectional area A
T
is changing. The
compression wave is generated as the nose enters the tunnel, and may be cal
culated by temporarily considering a train of semiinﬁnite length. During the
formation of the wave, and in the particular case in which the Mach number is
small enough that terms ∼O(M
2
) are negligible, the term Mϕ
∗
in the arguments
of A
/
T
in (7.4.5) is small, and we then ﬁnd, by expanding to ﬁrst order in Mϕ
∗
and integrating by parts, that
p ≈
ρ
0
U
2
A
∞
−∞
∂A
T
∂y
1
(y
1
÷U[t ])
∂ϕ
∗
∂y
1
(y
1
. 0. 0) dy
1
. M
2
<1. (7.4.6)
After the nose has passed into the tunnel, ∂ϕ
∗
¡∂y
1
=1 in the region occupied
by the nose, and (7.4.6) predicts the overall (linear theory) pressure rise to be
Lp ≈ ρ
0
U
2
A
0
¡A. But the linear theory, asymptotic pressure rise can also be
calculated exactly, with no restriction on Mach number, to be
Lp =
ρ
0
U
2
A
0
A(1 − M
2
)
.
because this is attained when ϕ
∗
(y
1
. 0. 0) ≈ y
1
− ¹
/
in (7.4.5) (see Equation
(3.9.14)). This implies that the Approximation (7.4.6) can be extrapolated to
ﬁnite Mach numbers by writing
p ≈
ρ
0
U
2
A(1 − M
2
)
∞
−∞
∂A
T
∂y
1
(y
1
÷U[t ])
∂ϕ
∗
∂y
1
(y
1
. 0. 0) dy
1
. (7.4.7)
This extrapolation of the linear theory to ﬁnite values of M turns out to be
applicable for M 0.4 (Howe et al. 2000).
Figure 7.4.4 illustrates schematically an experimental arrangement used by
Maeda et al. (1993) to investigate the compression wave. Wireguided, axisym
metric model trains are projected into and along the axis of a tunnel consisting of
a 7m long circular cylinder of internal diameter 0.147 m. The nose aspect ratio
h¡L = 0.2, the blockage A
0
¡A = 0.116, and the projection speed U ≈ 230
km/h (M ≈ 0.188). The train nose proﬁles include the cone, and the paraboloid
and ellipsoid of revolution, with respective crosssectional areas given by
A
T
(s)
A
0
=
¸
s
2
L
2
.
s
L
.
s
L
2 −
s
L
. 0 s L.
1. s ≥ L.
The data points in the ﬁgure are measurements (made at the point labelled T)
of the pressure gradient dp¡dt 1 m from the entrance for these three different
172 7 Problems in Three Dimensions
Fig. 7.4.4.
nose proﬁles. The solid curves are predictions of Equation (7.4.7), evaluated
using the following formulae for ∂ϕ
∗
¡∂y
1
(Howe 1998b):
∂ϕ
∗
∂y
1
(y) =
1
2
−
1
2π
∞
0
I
0
ξr
R
2K
1
(ξ)
I
1
(ξ)
1
2
sin
¸
ξ
y
1
R
÷Z(ξ)
¸
dξ.
Z(ξ) =
1
π
∞
0
ln
K
1
(j)I
1
(j)
K
1
(ξ)I
1
(ξ)
dj
j
2
−ξ
2
.
where r =
y
2
2
÷ y
2
3
 R and I
0
, I
1
, and K
1
are modiﬁed Bessel functions.
The linear theory underpredicts the maximum observed pressure gradients
by about 8%. The agreement with experiment can be greatly improved by inc
luding contributions from the surface dipoles in Equation (7.4.3) (which in a
ﬁrst approximation are determined by the drag force exerted on the nose of the
train by the linear theory pressure rise) and, to a lesser extent, by including the
vortex sound generated by the tunnel exitﬂow vorticity (the ﬁnal source term
on the right of (7.4.3)).
Problems 7
1. The termω
2
∂Y
2
¡∂y
3
in the Representation (7.1.3) of the sound produced by
a gust interacting with a thin airfoil accounts for the inﬂuence of changes in
Problems 7 173
the airfoil chord 2a(y
3
) over the interaction region. Show that when
da
dy
3
(y
3
) <1
the Formula (7.2.3) for the sound produced by a line vortex is given in a ﬁrst
approximation by
p(x. t ) ≈
ρ
0
IUM cos O
2
3
2
π[x[
[sinφ[
sin θ
¸
cos θ −sin θ cos φ
da
dy
3
0
¸
H
U[t ]
a
÷1
U[t ]
a
÷1
.
where da¡dy
3
is evaluated at y
3
= 0, where the vortex is cut by the airfoil.
Show that this result is identical with that given by (7.1.4) provided that
in (7.1.4) ω
3
is interpreted as the component of the vorticity parallel to the
local leading edge of the airfoil and the convection velocity U is replaced by
its component normal to the local leading edge.
2. A vortex ring orientated with its axis parallel to the ÷x
1
axis is convected in
a low Mach number mean ﬂow at speed U in the x
1
direction past the edge
of the rigid halfplane x
1
0. x
2
= 0. −∞x
3
∞. Use the compact
Green’s function
G(x. y. t −τ) =
1
2π
2
√
2πi c
0
ϕ
∗
(x)ϕ
∗
(y)
[x − y
3
i
3
[
3¡2
∞
−∞
√
ωe
−i ω(t −[x−y
3
i
3
[¡c
0
)
dω.
where ϕ
∗
is deﬁned as in (3.9.6), to calculate the sound produced as the
vortex passes the edge when the inﬂuence of the halfplane on the motion
of the ring is ignored.
3. Calculate the sound produced within and outside a semiﬁnite circular cylin
drical rigid pipe when a vortex ring exhausts axisymmetrically fromthe open
end. Neglect the inﬂuence of the pipe walls on the motion of the vortex and
ignore any change in the vortex radius at the exit.
4. Determine the (quadrupole) sound produced by the headon collision of
two equal ring vortices. Estimate the sound generated when a ring vortex is
incident normally on a plane wall.
174 7 Problems in Three Dimensions
5. Use the Green’s function (3.9.15) and Equation (7.4.4) to determine the
infrasound generated by a train entering a tunnel modeled by the unﬂanged,
circular cylindrical duct in Fig. 3.9.6b. Assume that the train travels along
the axis of the duct and show that the acoustic pressure at the far ﬁeld point
x outside the tunnel is given approximately by
p(x. t ) ≈
ρ
0
U
2
M
4π[x[
1 −
x
1
[x[
∞
−∞
∂A
T
∂y
1
(y
1
÷U[t ])
∂
2
ϕ
∗
∂y
2
1
(y
1
. 0. 0) dy
1
. [x[ →∞.
where [t ] = t −[x[¡c
0
.
8
Further Worked Examples
8.1 Blade–Vortex Interactions in Two Dimensions
The linear theory of the lowMach number, twodimensional interaction of a line
vortex with an airfoil was discussed in Section 6.4. The interaction will now be
examined in more detail, including also the inﬂuence of image vortices on the
motion. The general problem to be considered is depicted in Fig. 8.1.1, which
shows a vortex of strength I moving in the neighbourhood of a rigid airfoil of
chord 2a occupying −a  x
1
a. x
2
=0. There is no mean circulation about
the airfoil. We shall consider cases with and without a mean ﬂow in the x
1
direction and examine the inﬂuence of vortex shedding from the trailing edge.
8.1.1 Equation of Motion of the Vortex
At time t let the vortex be at
x ≡ (x
1
. x
2
) = x
0
(t ). and translate at velocity v
0
=
dx
0
dt
(t ).
If we set z = x
1
÷i x
2
. z
0
= x
01
÷i x
02
, the transformation
ζ =
z
a
÷
z
2
a
2
−1 (8.1.1)
maps the ﬂuid region in the z plane of the airfoil into the region [ζ [ >1 in
the ζ plane. The upper and lower faces of the airfoil (x
2
=±0) respectively
transform into the upper and lower halves of the unit circular cylinder [ζ [ = 1,
and the vortex maps into an equal vortex at ζ = ζ
0
(Fig. 8.1.2). In the absence
of mean ﬂow (U =0), and when there is no mean circulation about the cylin
der (and therefore about the airfoil), the complex potential of the motion is ob
tainedbyplacinganimage vortex−Iat the inverse point ζ =1¡ζ
∗
0
together with
175
176 8 Further Worked Examples
Fig. 8.1.1.
a vortex ÷I at the center of the cylinder. The two interior vortices ensure that
the total circulation around the cylinder vanishes. Then,
n(ζ ) = −
i I
2π
ln(ζ −ζ
0
) ÷
i I
2π
ln
ζ −
1
ζ
∗
0
−
i I
2π
ln ζ.
The velocitypotential of the motioninthe z plane is givenbysettingζ = ζ (z).
Because a mean ﬂow in the x
1
direction is unaffected by the airfoil, we can
include its contribution by adding the complex potential Uz. Then,
n(z) = −
i I
2π
ln(ζ (z) −ζ (z
0
)) ÷ F(z).
where F(z) =
i I
2π
ln
ζ (z) −
1
ζ (z
0
)
∗
−
i I
2π
ln ζ (z) ÷Uz.
This is of the formgiven in (4.6.1), so that the corresponding equation of motion
Fig. 8.1.2.
8.1 Blade–Vortex Interactions in Two Dimensions 177
of the vortex at z
0
is found from (4.6.3) to be
dz
∗
0
dt
≡
dx
01
dt
−i
dx
02
dt
= −
i Iζ
//
(z
0
)
4πζ
/
(z
0
)
÷ F
/
(z
0
).
that is
dz
∗
0
dt
=
i I
4πa
√
Z
2
−1
¸
Z
√
Z
2
−1
−1 ÷
2
[ζ
0
[
2
−1
¸
÷U. (8.1.2)
where Z =
z
0
a
. ζ
0
= Z ÷
Z
2
−1.
Equation (8.1.2) takes no account of the inﬂuence of vortex shedding. In a
linearized calculation (in which image effects in the airfoil are neglected) this
could be done by assuming shed vorticity to lie in a thin vortex sheet downstream
of the trailing edge at x
1
= a, as in Section 6.3. This would lead to a solution in
terms of the Sears function (6.3.6), but we shall not do this, because it limits the
discussion to linearized motions. Instead, we shall apply the method discussed
in Section 6.4, where the effects of vortex shedding are modelled by deleting
singularities from the compact Green’s function.
8.1.2 Formula for the Acoustic Pressure
The sound produced by the lowMach number motion of the vortex is calculated
from (6.2.4):
p(x. t ) ≈
−ρ
0
Ix
j
2π
√
2c
0
[x[
3
2
∂
∂t
t −[x[¡c
0
−∞
¸
dx
01
dτ
∂Y
j
∂y
2
−
dx
02
dτ
∂Y
j
∂y
1
¸
x
0
(τ)
dτ
√
t −τ −[x[¡c
0
. (8.1.3)
where the Kirchhoff vector for the strip airfoil has the components (Table 3.9.1)
Y
1
= y
1
. Y
2
= Re(−i
z
2
−a
2
). z = y
1
÷i y
2
. (8.1.4)
By deﬁning the radiation angle O for an observer at x in the far ﬁeld as in
Fig. 8.1.1, we can write
p(x. t ) ≈
ρ
0
I sin O
2π
√
2c
0
[x[
∂
∂t
t −[x[¡c
0
−∞
dx
02
dτ
(τ)
dτ
√
t −τ −[x[¡c
0
−
ρ
0
I cos O
2π
√
2c
0
[x[
∂
∂t
t −[x[¡c
0
−∞
dx
01
dτ
∂Y
2
∂y
2
−
dx
02
dτ
∂Y
2
∂y
1
x
0
(τ)
dτ
√
t −τ −[x[¡c
0
.
The two integrals in this formula represent the acoustic ﬁelds of dipole sources.
According to Curle’s theory (Section 2.3) and the Formula (6.2.2), the strengths
178 8 Further Worked Examples
of these dipoles are determined by the unsteady force (F
1
. F
2
) exerted on the
ﬂuid (per unit span) by the airfoil. The ﬁrst is aligned with the airfoil chord
(the mean ﬂow direction) and represents the inﬂuence of suction forces at the
leading and trailing edges (Batchelor 1967); the second component F
2
is equal
and opposite tothe unsteadylift experiencedbythe airfoil duringthe interaction.
In general the integrals must be evaluated numerically using the solution of
the equation of motion (8.1.2). Introduce the shorthand notation
W =
d
dz
(−i
z
2
−a
2
) =
−i Z
√
Z
2
−1
(8.1.5)
evaluated at the vortex. Then,
dx
01
dτ
∂Y
2
∂y
2
−
dx
02
dτ
∂Y
2
∂y
1
x
0
(τ)
≡ −a Im
W(Z)
dZ
dτ
.
and the acoustic pressure becomes
p(x. t ) ≈
ρ
0
Ia
2π
√
2c
0
[x[
∂
∂t
¸
sin O
[t ]
−∞
Im
dZ
dτ
dτ
√
[t ] −τ
÷cos O
[t ]
−∞
Im
W(Z)
dZ
dτ
dτ
√
[t ] −τ
¸
. (8.1.6)
where [t ] = t −[x[¡c
0
is the retarded time, and it is understood that Z = Z(τ).
Equations (8.1.2) and (8.1.6) for the vortex motion and the acoustic pressure
will now be applied to several different special cases.
8.1.3 Linear Theory
In the linearized approximation the vortex is swept past the airfoil along a
trajectory parallel to the x
1
direction at precisely the uniform mean ﬂow speed
U. This is the case illustrated in Fig. 6.4.1. When the standoff distance h <a it
was argued in Section 6.4 that the inﬂuence of vortex shedding fromthe trailing
edge could be estimated by deleting the singularity that occurs at the edge from
the Green’s function and ignoring the shed vorticity. Only the second integral
in (8.1.6) contributes to the sound (because dZ¡dτ = U¡a is real and F
1
≡ 0),
and the trailing edge singularity corresponds to the singularity of W(Z) at
Z =1. W(Z) is singular at both the leading and trailing edges (Z =±1), which,
other things being equal, are therefore the most signiﬁcant sources of sound
at high frequencies, because the second integral in (8.1.6) is dominated by
contributions from the neighbourhoods of the singularities. By deleting the
8.1 Blade–Vortex Interactions in Two Dimensions 179
contribution from the trailing edge we are asserting that all of the sound is
produced by the interaction of the vortex with the leading edge. Near this edge
W(Z) ≈
1
√
2
√
Z ÷1
. (8.1.7)
where the branch cut for
√
Z ÷1 runs along the real axis from Z = −1 to Z =
÷∞. Makingthis substitutionin(8.1.6), measuringtime fromthe instant that the
vortex crosses the midchord x
1
=0 of the airfoil, so that Z =Uτ¡a ÷i h¡a, and
changing the integration variable to ξ =1¡
√
t −τ −[x[¡c
0
, we then recover
the result (6.4.3), which can be written,
p(x. t )
ρ
0
IU
√
M cos O(a¡[x[)
1
2
4πa
≈
U[t ]
a
÷1
U[t ]
a
÷1
2
÷
h
a
2
. [x[ →∞. (8.1.8)
The nondimensional acoustic pressure signature (the righthand side of
(8.1.8)) is plotted as the solid curves in Fig. 8.1.3 for h¡a = 0.2. 0.5. 1.0.
Fig. 8.1.3.
180 8 Further Worked Examples
It was pointedout inSection6.3that the linearizedproblemof determiningthe
unsteady force F
2
exerted on the ﬂuid when an incompressible, sinusoidal gust
convects past the airfoil can be solved exactly with full account taken of vortex
shedding, in terms of the Sears function (6.3.6). This force also determines
the low Mach number acoustic radiation by Equation (6.2.2) (because F
1
≡0),
which in the present case can be shown to predict that
p(x. t )
ρ
0
IU
√
M cos O(a¡[x[)
1
2
4πa
≈
√
2πRe
∞
0
(i λ)
1
2
S(λ)e
−λ{h¡a÷iU[t ]¡a}
dλ.
[x[ →∞. (8.1.9)
The corresponding pressure signatures are plotted as the dotted curves in
Fig. 8.1.3. The agreement with the approximate theory of Equation (8.1.8) is
remarkably good even when h¡a is as large as unity, when the characteristic
reduced frequency λ =ωa¡U of the motion is relatively small, and might be
expected to lie outside the range for which (8.1.8) is valid.
8.1.4 Nonlinear Theory
When account is taken of image vortices in the airfoil the trajectory of the
vortex in the neighbourhood of the airfoil is no longer parallel to the mean ﬂow
direction, and must be determined by numerical integration of Equation (8.1.2).
To do this it is convenient to introduce a dimensionless velocity ratio c and time
T deﬁned by
c =
I
4πaU
. T =
Ut
a
.
in terms of which (8.1.2) becomes
dZ
∗
dT
=
i c
√
Z
2
−1
¸
Z
√
Z
2
−1
−1 ÷
2
[ζ
0
[
2
−1
¸
÷1. (8.1.10)
which can be solved for Z by Runge–Kutta integration (Section 4.6).
If the initial standoff distance is h at x
1
= −∞, the integration is started at a
large distance L upstream of the airfoil midchord, say L =10a, by prescribing
the initial position of the vortex to be Z =−L¡a ÷ i h¡a. The upper part of
Fig. 8.1.4 shows a calculated trajectory for
c = 0.2.
h
a
= 0.2.
8.1 Blade–Vortex Interactions in Two Dimensions 181
Fig. 8.1.4.
where time is measured from the instant that the vortex passes the midchord
of the airfoil. The nonlinear inﬂuence of the image vorticity is to shift the
initially rectilinear trajectory of the vortex away fromthe airfoil in the direction
of the vortex force ω ∧ U (U=Ui). The vortex is closest to the airfoil at
Ut ¡a = 0, where x
02
∼ 0.28a, and where convection by the images increases
the translation speed of the vortex from U to approximately
U ÷
I
4πx
02
= U ÷
cU
x
02
¡a
∼ U
1 ÷
0.2
0.28
= 1.71U.
The sound generated as the vortex passes the airfoil is given by (8.1.6). The
inﬂuence of vortex shedding into the wake is included by using the approxi
mation (8.1.7) for W(Z). The integrals must be evaluated numerically, and this
is done by deﬁning a dimensionless vortex convection velocity (u(
ˆ
T). :(
ˆ
T))
by
dZ
d
ˆ
T
= u(
ˆ
T) ÷i :(
ˆ
T). where
ˆ
T =
Uτ
a
.
182 8 Further Worked Examples
Then,
p(x. t )
ρ
0
IU
√
M(a¡[x[)
1
2
4πa
≈ 2
1
2
d
dT
¸
sin O
[T]
−∞
:(
ˆ
T) d
ˆ
T
[T] −
ˆ
T
÷cos O
[T]
−∞
Im(W(Z)(u ÷i :))(
ˆ
T) d
ˆ
T
[T] −
ˆ
T
¸
= 2
3
2
sin O
d
dT
∞
0
:([T] −λ
2
) dλ
÷2
3
2
cos O
d
dT
∞
0
Im(W(Z)(u ÷i :))([T] −λ
2
) dλ
≈
p
1
(x. t )
ρ
0
IU
√
M(a¡[x[)
1
2
4πa
÷
p
2
(x. t )
ρ
0
IU
√
M(a¡[x[)
1
2
4πa
. (8.1.11)
where T =Ut ¡a, [T] =U[t ]¡a, and the integration variable
ˆ
T has been re
placed by λ =
[T] −
ˆ
T. The ﬁnal integrals are easily evaluated numerically
when the path of the vortex has been determined. The upper limit of integra
tion is actually ﬁnite, because the source terms must be set to equal zero as
soon as [T] −λ
2
reduces to the nondimensional time at which the computation
of the vortex path begins (where the vortex is sufﬁciently far upstream that it
effectively produces no sound by interaction with the airfoil).
The components p
1
(x. t ). p
2
(x. t ) of (8.1.11) correspond respectively to the
dipole sound produced by the unsteady suction and lift forces; their nondimen
sional forms
p
1
(x. t )
ρ
0
IU
√
M sin O(a¡[x[)
1
2
4πa
and
p
2
(x. t )
ρ
0
IU
√
M cos O(a¡[x[)
1
2
4πa
are plotted in Fig. 8.1.4. Vortex shedding should smooth out the pressure signa
tures at the retarded times when the vortex is close to the trailing edge. But the
calculated pressures exhibit blips shown as dotted curves in the ﬁgure. These
arise because, although our calculation has accounted for vortex shedding in
evaluating the dipole source strengths (by means of the approximation (8.1.7)),
the effect of shedding was not included in the calculation of the vortex tra
jectory. However, the smoothing inﬂuence of shedding at a sharp edge acts
to remove the blips, and the pressure signatures have proﬁles similar to those
depicted by the solid curves in the ﬁgure, obtained by interpolating smoothly
between the calculated pressures on either side of the blips.
An interesting nonlinear interaction occurs when the initial standoff distance
of the vortex h =0 (Fig. 8.1.5). In the linearized approximation, the vortex
would strike the leading edge of the airfoil at U[t ]¡a = −1, at which time the
8.1 Blade–Vortex Interactions in Two Dimensions 183
Fig. 8.1.5.
linear theory acoustic pressure (8.1.8) is inﬁnite. This singular event does not
occur because the vortex trajectory is deﬂected around the airfoil by the image
vorticity (for a rounded nose the possibility of additional vortex shedding from
the leading edge may be ignored). The upper part of Fig. 8.1.5 illustrates this
for the same value of the velocity ratio c = I¡4πaU =0.2 considered above.
The maximum convection velocity of the vortex (at Ut ¡a = 0) is now more
than twice the mean stream velocity:
U ÷
I
4π(0.15a)
= U
1 ÷
0.2
0.15
= 2.3U.
The corresponding suction and liftdipole acoustic pressures p
1
and p
2
shown
in the ﬁgure are also greatly increased.
8.1.5 Periodic Vortex Motion
When there is no mean ﬂow (U =0) the characteristic velocity and dimension
less time become
V =
I
4πa
. T =
Vt
a
.
184 8 Further Worked Examples
Fig. 8.1.6.
and the vortex equation of motion (8.1.2) reduces to
dZ
∗
dT
=
i
√
Z
2
−1
¸
Z
√
Z
2
−1
−1 ÷
2
[ζ
0
[
2
−1
¸
.
The solutions are closed trajectories orbiting the airfoil periodically. A typical
orbit is plotted in the upper half of Fig. 8.1.6 for I > 0, for the case where the
trajectory passes through the point labelled 0, where x
01
= −2a. x
02
= 0. The
calculated period is T
0
≡ Vt
0
¡a ≈ 35.84.
An orbiting vortex motion of this kind cannot be realized in practice (because
of diffusion from the vortex core and the continual shedding of additional
vorticityfromthe airfoil), but it is still instructive tocalculate the soundproduced
by the motion. By writing
ˆ
T =
Vτ
a
and
dZ
d
ˆ
T
= u(
ˆ
T) ÷i :(
ˆ
T)
in the general formula (8.1.6) for the acoustic pressure, the nondimensional
8.1 Blade–Vortex Interactions in Two Dimensions 185
suction and lift acoustic pressures are found to be given by the following mod
iﬁed form of (8.1.11)
p
1
(x. t )
ρ
0
V
2
√
M(a¡[x[)
1
2
÷
p
2
(x. t )
ρ
0
V
2
√
M(a¡[x[)
1
2
≈ 2
1
2
d
dT
¸
sin O
[T]
−∞
:(
ˆ
T) d
ˆ
T
[T] −
ˆ
T
÷cos O
[T]
−∞
Im(W(Z)(u ÷i :))(
ˆ
T) d
ˆ
T
[T] −
ˆ
T
¸
.
(8.1.12)
where M = V¡c
0
.
The function W(Z) in the integrand is given by (8.1.5) in the absence of
vortex shedding. It follows by inspection and from the numerical solution, that
when T is measured from Z = −2, as indicated in Fig. 8.1.6, the suction and
dipole source strengths have period T
0
, and possess Fourier series expansions
of the form
:(T) =
∞
¸
n=1
a
n
cos
2πnT
T
0
. Im(W(Z)(u÷i :))(T) =
∞
¸
n=1
b
n
sin
2πnT
T
0
.
where the coefﬁcients a
n
. b
n
can be calculated by using the numerical solution
for the orbit to evaluate
a
n
=
2
T
0
T
0
0
:(T) cos
2πnT
T
0
dT.
b
n
=
2
T
0
T
0
0
Im(W(Z)(u ÷i :))(T) sin
2πnT
T
0
dT.
By making the change of integration variable λ =
[T] −
ˆ
T¡
√
T
0
, the right
hand side of (8.1.12) now becomes
4
√
2π
√
T
0
∞
¸
n=1
¸
−a
n
n sin O
∞
0
sin
¸
2πn
[T]
T
0
−λ
2
¸
dλ
÷b
n
n cos O
∞
0
cos
¸
2πn
[T]
T
0
−λ
2
¸
dλ
¸
.
The integrals are evaluated from the real and imaginary parts of
∞
0
e
2πi n{[T]¡T
0
−λ
2
}
dλ =
1
2
√
2n
e
{2n[T]¡T
0
−
1
4
}πi
.
186 8 Further Worked Examples
Hence, the suction and lift force dipole ﬁelds are given respectively by
p
1
(x. t )
ρ
0
V
2
√
M sin O(a¡[x[)
1
2
≈ −
2π
√
T
0
∞
¸
n=1
a
n
√
n sin
¸
2nπt
t
0
−
π
4
¸
p
2
(x. t )
ρ
0
V
2
√
M cos O(a¡[x[)
1
2
≈
2π
√
T
0
∞
¸
n=1
b
n
√
n cos
¸
2nπt
t
0
−
π
4
¸
. [x[ →∞.
where [ ] denotes evaluation at the retarded time t −[x[¡c
0
. The corresponding
nondimensional pressures are plotted in Fig. 8.1.6 (taking the ﬁrst 26 terms
in the series); both have similar orders of magnitude, and exhibit rapid varia
tions at the retarded times at which the vortex is directly above and below the
airfoil.
8.2 Parallel Blade–Vortex Interactions in Three Dimensions
We now examine to what extent the simple twodimensional methods of the
previous section can be adapted to wings of ﬁnite span and variable chord for
problems of the kind shown in Fig. 8.2.1. The general representation of the
sound produced by vortex–airfoil interactions is discussed in Section 7.1, when
Fig. 8.2.1.
8.2 Parallel Blade–Vortex Interactions in Three Dimensions 187
the airfoil chord can be regarded as compact. The general solution is applicable
to airfoils of arbitrary span, but we shall consider only the case where the span
is compact; predictions for a noncompact span will be intermediate between
those discussed here and those in Section 8.1.
Consider a planar airfoil of either rectangular or elliptic planform, orientated
as illustrated in Fig. 8.2.1 at zero angle of attack to a mean ﬂow at speed U in
the x
1
direction. A spanwise line vortex of strength I is swept past the airfoil
at an initial standoff distance h above the airfoil, as indicated in the side view
of Fig. 8.2.1b. When h =0 it will be necessary to take account of nonlinear
interactions with the airfoil.
For an airfoil of compact chord and span the acoustic pressure produced by
the interaction is given by Equation (5.4.4):
p(x. t ) ≈
−ρ
0
x
j
4πc
0
[x[
2
∂
∂t
(ω ∧ v)
y. t −
[x[
c
0
· ∇Y
j
(y) d
3
y. [x[ →∞.
(8.2.1)
It is assumed that the section of the line vortex that interacts with the airfoil
remains rectilinear, with the representation
ω = Ikδ(x
1
− x
01
(t ))δ(x
2
− x
02
(t )). where x
0
= (x
01
. x
02
. 0).
For an elliptic airfoil of span L (between −
1
2
L x
3

1
2
L), the Kirchhoff vector
Y has the components
Y
1
= y
1
.
Y
2
=
¸
Re(−i
z
2
− ˆ a(y
3
)
2
). [y
3
[ 
1
2
L
y
2
. [y
3
[ >
1
2
L
. Y
3
= y
3
. z = y
1
÷i y
2
.
where 2ˆ a(y
3
) is the airfoil chord at the spanwise location y
3
. For the rectangular
airfoil ˆ a(y
3
) ≡ a = constant; for the elliptic airfoil ˆ a(y
3
) assumes a maximum
value of a at y
3
= 0, and we shall write
ˆ a(y
3
)
a
=
1 −
4y
2
3
L
2
. [y
3
[ 
1
2
L. (8.2.2)
Vorticity is shed into the wake of the airfoil in accordance with the Kutta
condition of unsteady aerodynamics. This smooths out conditions at the trail
ing edge, so that sound is generated primarily as the vortex passes over the
leading edge of the airfoil. As before, this can be dealt with in a ﬁrst approxima
tion by ignoring the shed vorticity and deleting the trailing edge singularity of
188 8 Further Worked Examples
Green’s function, by using the following modiﬁcation of the x
2
component
of Y:
Y
2
= Re(
2ˆ a(y
3
)
z ÷ ˆ a(y
3
)). [y
3
[ 
1
2
L.
Then, (8.2.1) becomes
p(x. t ) ≡ p
1
(x. t ) ÷ p
2
(x. t )
≈
ρ
0
I cos +
4πc
0
[x[
∂
∂t
L
2
−
L
2
¸
dx
02
dt
¸
dy
3
÷
ρ
0
I cos O
4
√
2πc
0
[x[
∂
∂t
L
2
−
L
2
¸
Im
dz
0
dt
√
ˆ a(y
3
)
√
z
0
(t ) ÷ ˆ a(y
3
)
¸
dy
3
[x[ →∞. (8.2.3)
where +. Oare respectively the angles shown in Fig. 8.2.1a between the x
1
and
x
2
directions and the radiation direction, z
0
(t ) = x
01
(t ) ÷i x
02
(t ), and quantities
in square braces are evaluated at the retarded time [t ] = t −[x[¡c
0
.
The ﬁrst term on the right is the suction force dipole, aligned with the airfoil
chord, whose strength is determined by the x
2
component of the vortex convec
tion velocity. It is assumed to be nonzero only over the section −
1
2
L  y
3

1
2
L
of the vortex, where dx
02
¡dt ,= 0 because of nonlinear interactions with the air
foil. The second termis the conventional lift dipole radiation. Note that ‘inﬁnite’
contributions to the integrals from [y
3
[ > L¡2 are constant because ω ∧ v is
constant for [y
3
[ > L¡2, and have been discarded (c.f., Section 7.3).
8.2.1 Linear Theory
When there is no backreaction of the airfoil on the vortex the convection
velocity of the vortex is equal to the mean stream velocity
dx
01
dt
= U.
dx
02
dt
= 0.
The radiation is produced entirely by the lift dipole, and if the vortex crosses
the midchord of the airfoil at time t = 0 (8.2.3) reduces to
p
2
(x. t )
ρ
0
IUM cos O(L¡[x[)¡4πa
≈ −
1
2
3
2
1
2
−
1
2
Im
¸
ˆ a
a
U[t ]
a
÷
ˆ a
a
÷i
h
a
3
2
d ˆ y
3
.
(8.2.4)
8.2 Parallel Blade–Vortex Interactions in Three Dimensions 189
Fig. 8.2.2.
where ˆ y
3
= y
3
¡L; ˆ a¡a = 1 for the rectangular airfoil, and is given by (8.2.2)
for the elliptic airfoil. For a rectangular airfoil the integral evaluates to
Im
¸
U[t ]
a
÷1 ÷i
h
a
−
3
2
¸
.
The acoustic pressure signatures (the lefthand side of (8.2.4)) for the rectan
gular and elliptic airfoils are plotted in Fig. 8.2.2 for a vortex standoff distance
h =0.2a. The proﬁles are qualitatively similar to the corresponding plot in
Fig. 8.1.3 for an airfoil of inﬁnite span, although in three dimensions the am
plitude decreases much more rapidly with increasing retarded distance of the
vortex from the leading edge. The peak amplitude is larger and the acoustic
pulse is of smaller duration ∼h¡U for the rectangular airfoil because different
sections of the vortex interact with the leading edge of the elliptic airfoil at
different times during a total interaction time ∼a¡U > h¡U.
8.2.2 Nonlinear Theory
When the standoff distance h =0. image vorticity in the airfoil prevents the
vortex from impinging on the leading edge, and causes the trajectory to be
locally deﬂected above the airfoil (for I > 0). If the leading edge of the airfoil
is suitably rounded (so that no additional vortex shedding occurs) this case
can be treated for a rectangular airfoil by assuming that only the section of the
vortex within the span −
1
2
L  x
3

1
2
L of the airfoil is affected in this way, and
that the distorted path can be approximated by that for locally twodimensional
ﬂow.
190 8 Further Worked Examples
Let
c =
I
4πaU
. T =
Ut
a
. Z =
z
0
a
.
dZ
dT
= u(T) ÷i :(T).
W(Z) =
1
√
2
√
Z ÷1
.
Then, the motion of the section of the vortex within the airfoil span (−
1
2
L 
x
3

1
2
L) is governed by Equation (8.1.10) (where ζ
0
is deﬁned in terms of z
0
as in (8.1.1)), and the suction and lift dipole radiation pressures are given by
p
1
(x. t )
ρ
0
IUM cos +(L¡[x[)¡4πa
≈
¸
d:
dT
¸
p
2
(x. t )
ρ
0
IUM cos O(L¡[x[)¡4πa
≈
∂
∂T
[Im(W(Z)(u ÷i :))]
¸
¸
¸
¸
¸
¸
¸
¸
¸
[x[ →∞.
where [ ] denotes evaluation at the retarded time t −[x[¡c
0
.
These nondimensional pressures are plotted in Fig. 8.2.3 for a velocity ratio
c =0.2 when the vortex is released upstream with h =0. The upper part of the
Fig. 8.2.3.
8.3 Vortex Passing over a Spoiler 191
ﬁgure shows the path followed by those sections of the vortex inboard of the
airfoil tips; it is the same as that depicted in Fig. 8.1.5 for the inﬁnite span
airfoil. As in that case, small acoustic blips spuriously predicted during the
passage of the vortex past the trailing edge have been removed (c.f., Fig. 8.1.4).
The threedimensional acoustic pulses are narrower than those predicted in two
dimensions.
8.3 Vortex Passing over a Spoiler
The sound produced when vorticity interacts at low Mach number with surface
irregularities on a nominally plane, rigid wall is produced by dipoles orientated
in the plane of the wall, that is, by the unsteady wall drag. A simple canonical
interaction (Kasoev, 1976) involving a line vortex near a thin vertical spoiler is
illustrated in Fig. 8.3.1. The wall coincides with the plane x
2
= 0, and the spoiler
extends along the x
2
axis from x
2
= 0 to x
2
= a > 0 for −∞ x
3
 ∞. The
vortex
ω = Ikδ(x −x
0
(t )). where x
0
= (x
01
. x
02
. 0).
is parallel to the spoiler, and is assumed to convect over it in a lowMach number,
irrotational mean stream having uniform speed U in the x
1
direction.
Deﬁne z = x
1
÷i x
2
. z
0
= x
01
÷i x
02
. The transformation
ζ =
z
2
a
2
÷1 (8.3.1)
maps the ﬂuid region onto the upper half Imζ > 0 of the ζ plane. The left and
right faces of the spoiler (x
1
=∓0) transform respectively into the intervals
−1 ζ 0 and 0 ζ 1 of the real ζ axis, and the vortex maps into an equal
vortex at ζ = ζ
0
. The mean ﬂow is parallel to the real axis in the ζ plane,
Fig. 8.3.1.
192 8 Further Worked Examples
with complex potential Uaζ . The complex potential of the whole ﬂow in the
ζ plane is obtained by introducing an image vortex of strength −I at the
complex conjugate point ζ = ζ
∗
0
, and is given by
n(ζ ) = −
i I
2π
ln(ζ −ζ
0
) ÷
i I
2π
ln(ζ −ζ
∗
0
) ÷Uaζ.
Hence, setting ζ = ζ (z) the motion in the z plane is deﬁned by
n(z) = −
i I
2π
ln(ζ (z) −ζ (z
0
)) ÷ F(z).
where F(z) =
i I
2π
ln(ζ (z) −ζ (z
0
)
∗
) ÷Uaζ (z).
which is in the form (4.6.1). The equation of motion of the vortex at z
0
is
therefore (see (4.6.3))
dz
∗
0
dt
≡
dx
01
dt
−i
dx
02
dt
= −
i Iζ
//
(z
0
)
4πζ
/
(z
0
)
÷ F
/
(z
0
).
that is,
dZ
∗
dT
=−i
¸
1
Z(Z
2
÷1)
−
2Z
Z
2
÷1 −[Z
2
÷1[
¸
÷
cZ
√
Z
2
÷1
. (8.3.2)
where Z =
z
0
a
. T =
Vt
a
. V =
I
4πa
. c =
U
V
.
The compact Green’s function for this problem(applicable when the acoustic
wavelength · ∂) is given by (6.1.6), so that the analogue of Equation (6.2.4)
for the farﬁeld acoustic pressure becomes
p(x. t )
≈
−ρ
0
Ix
1
π
√
2c
0
[x[
3
2
∂
∂t
t −[x[¡c
0
−∞
k ∧
dx
0
dτ
(τ) · ∇Y
1
(x
0
(τ))
dτ
√
t −τ −[x[¡c
0
=
−ρ
0
Ix
1
π
√
2c
0
[x[
3
2
∂
∂t
t −[x[¡c
0
−∞
¸
dx
01
dτ
∂Y
1
∂y
2
−
dx
02
dτ
∂Y
1
∂y
1
¸
x
0
(τ)
dτ
√
t −τ −[x[¡c
0
.
(8.3.3)
where the Kirchhoff vector
Y
1
= Re(aζ ) = aRe(
Z
2
÷1) at z = z
0
. (8.3.4)
8.3 Vortex Passing over a Spoiler 193
The radiation is produced by the unsteady drag force F
1
exerted on the ﬂuid
by the spoiler, given (per unit span) by
F
1
= −ρ
0
ω ∧ v · ∇Y
1
dy
1
dy
2
= −ρ
0
Ik ∧
dx
0
dt
· ∇Y
1
(x
0
).
This force vanishes, and therefore no sound is generated, in the linearized
approximation in which the vortex is assumed to translate at the local mean
stream velocity, because in that case
dx
0
dt
= U∇Y
1
(x
0
) and k ∧ ∇Y
1
· ∇Y
1
≡ 0.
Following the procedure of Section 8.1, introduce the notations
dZ
dT
= u(T) ÷i :(T). W =
d
dz
(
z
2
÷a
2
) =
Z
√
Z
2
÷1
(8.3.5)
evaluated at the vortex, and make the substitution
ˆ
T = Vτ¡a in (8.3.3) to obtain
the acoustic pressure in the form
p(x. t ) ≈
ρ
0
IV
√
M sin O
aπ
√
2
a
[x[
1
2
∂
∂T
[T]
−∞
Im
W(Z)
dZ
d
ˆ
T
d
ˆ
T
[T] −
ˆ
T
.
that is,
p(x. t )
ρ
0
V
2
√
M sin O(a¡[x[)
1
2
≈ 2
5
2
∂
∂T
∞
0
Im(W(Z)(u ÷i :))([T] −λ
2
) dλ.
(8.3.6)
where [T] =V[t ]¡a is the nondimensional retarded time and M =V¡c
0
.
The vortex path equation (8.3.2) and the acoustic pressure integral (8.3.6)
are evaluated numerically, taking the initial position of the vortex to be several
spoiler heights a upstream, where its motion is unaffected by the spoiler. The
upper part of Fig. 8.3.2 shows the vortex trajectories when the initial distance of
the vortexfromthe wall is h =0.75a for the twocases (i) of nomeanﬂow, U =0,
and (ii) U =V ≡I¡4πa; the corresponding nondimensional acoustic pressures
(8.3.6) are plotted in the lower part of the ﬁgure, where time is measured from
the instant that the vortex passes the spoiler. The effect of mean ﬂow is to draw
the trajectory marginally closer to the spoiler as it passes the tip of the spoiler
where the interaction is strongest. The convection velocity at this point is also
increased from about 1.98V when U =0 to 3.95V when U =V, and this is
responsible for more than doubling the amplitude and the effective frequency
of the sound.
194 8 Further Worked Examples
Fig. 8.3.2.
8.4 Bluff Body Interactions: The Circular Cylinder
LowMach number, twodimensional interactions of a line vortex with a circular
cylinder provide an interesting contrast to the sharpedge problems discussed
above. Let the cylinder have radius a and be coaxial with the x
3
axis, and let
there be an irrotational mean ﬂowat speedU past the cylinder in the x
1
direction,
with no mean circulation about the cylinder.
Set z = x
1
÷i x
2
and let the vortex of strength I have the complex position
z
0
= x
01
÷ i x
02
at time t . The complex potential n(z) is found by placing an
image vortex −I at the inverse point z = a¡z
∗
0
within the cylinder, a vortex
÷I at the centre to make the circulation vanish, and by adding the potential for
the uniform mean ﬂow past the cylinder (c.f., Section 8.1):
n(z) = −
i I
2π
ln(z−z
0
)÷
i I
2π
ln
z −
a
2
z
∗
0
−
i I
2π
ln z ÷U
z ÷
a
2
z
. (8.4.1)
The velocity potential governing the motion of the vortex at z
0
is obtained
by deleting the selfpotential
−
i I
2π
ln(z − z
0
).
8.4 Bluff Body Interactions: The Circular Cylinder 195
Hence we arrive at the equation of motion
dZ
∗
dT
=
i
Z([Z[
2
−1)
÷c
1 −
1
Z
2
. (8.4.2)
where Z =
z
0
a
. V =
I
2πa
. T =
Vt
a
. c =
U
V
.
and
dz
0
dt
= V
dZ
dT
≡ V(u ÷i :).
8.4.1 The Acoustic Pressure
The farﬁeld sound produced by the vortex is calculated from (6.2.4):
p(x. t ) ≈
−ρ
0
Ix
j
2π
√
2c
0
[x[
3
2
∂
∂t
t −[x[¡c
0
−∞
¸
dx
01
dτ
∂Y
j
∂y
2
−
dx
02
dτ
∂Y
j
∂y
1
¸
x
0
(τ)
dτ
√
t −τ −[x[¡c
0
. (8.4.3)
where the components of the Kirchhoff vector can be written (see Section 4.5)
Y
1
= Re
z ÷
a
2
z
. Y
2
= Re
¸
−i
z −
a
2
z
¸
. z = y
1
÷i y
2
. (8.4.4)
By deﬁning
W
1
=
d
dz
z ÷
a
2
z
≡1 −
1
Z
2
. W
2
=
d
dz
¸
−i
z −
a
2
z
¸
≡ −i
1 ÷
1
Z
2
evaluated at z
0
, and making the change of integration variable
ˆ
T =Vτ¡a,
Equation (8.4.3) can be written
p(x. t ) ≈
ρ
0
IV
√
Mx
j
2π
√
2a[x[
3
2
∂
∂T
[T]
−∞
Im(W
j
(u ÷i :))(
ˆ
T)
d
ˆ
T
[T] −
ˆ
T
.
where M =
V
c
0
. [T] =
V
a
t −
[x[
c
0
.
The subscripts j = 1. 2 in this formula respectively correspond to the acous
tic pressures p
1
. p
2
, say, produced by drag and lift dipoles, whose strengths
are determined by the force (F
1
. F
2
) exerted on the ﬂuid (per unit span) by
the cylinder. The integrals must be evaluated numerically using the numerical
196 8 Further Worked Examples
Fig. 8.4.1.
solution of Equation (8.4.2) for the vortex path. This is done by making the
further change of integration variable λ =
[T] −
ˆ
T, in which case
p
1
(x. t )
ρ
0
V
2
√
M sin O(a¡[x[)
1
2
≈ 2
1
2
∂
∂T
∞
0
Im(W
1
(u ÷i :))([T] −λ
2
) dλ.
p
2
(x. t )
ρ
0
V
2
√
M cos O(a¡[x[)
1
2
≈ 2
1
2
∂
∂T
∞
0
Im(W
2
(u ÷i :))([T] −λ
2
) dλ.
The calculation begins at time T
/
, say, by taking the initial position of the vortex
to be far upstream of the cylinder at z
0
= −L ÷i h, where L ·a is sufﬁciently
large that the source strengths are negligible for T T
/
(Fig. 8.4.1). The upper
limits of integration are then ﬁnite because the source terms vanish as soon as
[T] −λ
2
 T
/
.
Figure 8.4.1illustrates the typical nondimensional waveforms producedwhen
V ≡ I¡2πa = 2U and for h¡a = ±0.7, time being measured from the instant
that the vortex crosses x
1
=0. The amplitude of the sound decreases rapidly
with increasing distance of closest approach of the vortex to the cylinder; near
the cylinder the translational velocity of the vortex is increased because the
8.4 Bluff Body Interactions: The Circular Cylinder 197
mean ﬂow velocity is larger, and also because of the increased inﬂuence of the
image vorticity. In cases where U ·V, the lift dipole will tend to predomi
nate because convection by the image vortices can then be neglected in a ﬁrst
approximation, and the drag ∼(ω ∧U∇Y
1
) · ∇Y
1
≡ 0.
8.4.2 WallMounted Cylinder
The case of ideal motion of a vortex translating past a cylindrical, semicircular
projection on a rigid wall (Fig. 8.4.2) can be treated by the method used for
the spoiler in Section 8.3. The problem is equivalent to that in which a vortex
pair, consisting of a vortex I at z
0
accompanied by an image of strength −I at
z
∗
0
, is incident symmetrically on a circular cylinder. In this case, the lift dipole
vanishes identically.
The velocity potential of the unsteady motion is given by augmenting the
complex potential (8.4.1) by the terms
i I
2π
ln(z − z
∗
0
) −
i I
2π
ln
z −
a
2
z
0
÷
i I
2π
ln z.
Fig. 8.4.2.
198 8 Further Worked Examples
which correspond to the net potential produced by the image. Then, the equation
of motion becomes
dZ
∗
dT
= i
¸
1
Z − Z
∗
÷
Z − Z
∗
([Z[
2
−1)(Z
2
−1)
¸
÷c
1 −
1
Z
2
. (8.4.5)
where Z =
z
0
a
. V =
I
2πa
. T =
Vt
a
. c =
U
V
.
The compact Green’s function is given by (6.1.6), and the farﬁeld acoustic
pressure by
p(x. t )
≈
−ρ
0
Ix
1
π
√
2c
0
[x[
3
2
∂
∂t
t −[x[¡c
0
−∞
k ∧
dx
0
dτ
(τ) · ∇Y
1
(x
0
(τ))
dτ
√
t −τ −[x[¡c
0
=
−ρ
0
Ix
1
π
√
2c
0
[x[
3
2
∂
∂t
t −[x[¡c
0
−∞
¸
dx
01
dτ
∂Y
1
∂y
2
−
dx
02
dτ
∂Y
1
∂y
1
¸
x
0
(τ)
dτ
√
t −τ −[x[¡c
0
.
(8.4.6)
where the Kirchhoff vector
Y
1
= Re
z ÷
a
2
z
. (8.4.7)
The radiation is produced by the unsteady drag force exerted on the ﬂuid by
the cylinder, which vanishes in the linearized approximation, when the vortex
is assumed to convect passively at the local velocity of the undisturbed mean
stream. As before, set
dZ
dT
= u(T) ÷i :(T). W
1
=
d
dz
z ÷
a
2
z
= 1 −
1
Z
2
(8.4.8)
evaluated at the vortex. Then,
p(x. t )
ρ
0
V
2
√
M sin O(a¡[x[)
1
2
≈ 2
3
2
∂
∂T
∞
0
Im(W
1
(u÷i :))([T]−λ
2
) dλ. (8.4.9)
where the angle Ois deﬁnedas inFig. 8.4.1, [T] =V[t ]¡a is the nondimensional
retarded time (T =0 when the vortex is at x
1
=0), and M =V¡c
0
.
The vortex path equation (8.4.5) and the acoustic pressure integral (8.4.9)
must be evaluated numerically, taking the initial position of the vortex to
be several cylinder radii a upstream where its motion is unaffected by the
cylinder. The upper part of Fig. 8.4.2 shows the vortex trajectories when the
8.5 Vortex Ring and Sphere 199
initial standoff distance of the vortex from the wall h =0.5a for the two cases
(i) of no mean ﬂow, U =0, and (ii) U =V. The vortex convection velocity
at y
1
=0 is increased from about 1.23V when U = 0 to 3.07V when U =V;
this is responsible for the increased acoustic amplitude and for more than dou
bling the effective frequency. The waveforms and these general conclusions
are qualitatively similar to those discussed in Section 8.3 for the sharpedged
spoiler.
8.5 Vortex Ring and Sphere
Perhaps the simplest low Mach number, inviscid, threedimensional vortex–
surface interaction amenable to analysis is the axisymmetric motion of a ring
vortex over a sphere. Let a sphere of radius a be placed with its centre at the
origin in the presence of a uniform mean ﬂow at speed U in the x
1
direction.
A vortex ring of radius r
0
(t ) and circulation I is coaxial with the x
1
axis and
translates in the positive x
1
direction under the inﬂuence of the mean ﬂow,
selfinduction and image vorticity in the sphere (Fig. 8.5.1). We shall assume
the vortex core is circular (and remains circular throughout the interaction)
with radius σ(t ) <r
0
. The selfinduced velocity of the ring (in inviscid ﬂow)
is parallel to the x
1
axis at speed u
I
given approximately by Kelvin’s formula
(Batchelor, 1967)
u
I
(r
0
. σ) =
I
4πr
0
¸
ln
8r
0
σ
−
1
4
¸
. (8.5.1)
Fig. 8.5.1.
200 8 Further Worked Examples
The image vorticity consists of a coaxial ring vortex whose circulation I
/
,
radius r
/
0
. and axial location x
/
01
are given by
I
/
= −
I
r
2
0
÷ x
2
01
1
2
a
. r
/
0
=
a
2
r
0
r
2
0
÷ x
2
01
. x
/
01
=
a
2
x
01
r
2
0
÷ x
2
01
. (8.5.2)
where the planes of symmetry of the ring vortex and its image cut the x
1
axis respectively at x
01
(t ). x
/
01
(t ). The motion of the vortex produced by the
combined induction by the image and the mean ﬂow can be expressed in terms
of the Stokes stream function ψ(
·
r. x
1
) (Batchelor, 1967; Ting and Klein, 1991)
ψ(r. x
1
) =
Ur
2
2
¸
1 −
a
3
r
2
÷ x
2
1
3
2
÷
I
/
2π
(+
÷
÷+
−
){K(A) − E(A)}.
(8.5.3)
where +
±
=
(r ∓r
/
0
)
2
÷(x
1
− x
/
01
)
2
. A =
+
−
−+
÷
+
−
÷+
÷
.
K(A) =
π
2
0
dj
1 −A
2
sin
2
j
. E(A) =
π
2
0
1 −A
2
sin
2
jdj.
where r denotes perpendicular distance from the x
1
axis, and K(A). E(A) are
respectively complete elliptic integrals of the ﬁrst and second kinds.
The radius r
0
(t ) andaxial position x
01
(t ) of the ringvortexare thendetermined
by the equations of motion
dr
0
dt
= −
1
r
0
∂ψ
∂x
1
(r
0
. x
01
).
dx
01
dt
= u
I
(r
0
. σ) ÷
1
r
0
∂ψ
∂r
(r
0
. x
01
). (8.5.4)
The core radius σ decreases when r
0
increases, because the vortex lines move
with the ﬂuid particles. If r
0
=h and σ = σ
0
are the initial values when the
vortex ring is far from the sphere, then at any time t
(2πr
0
)πσ
2
= (2πh)πσ
2
0
. i.e.. σ(t ) = σ
0
h
r
0
(t )
so that the selfinduced velocity (8.5.1) becomes
u
I
=
I
4πr
0
¸
ln
8h
σ
0
¸
r
0
(t )
h
¸3
2
−
1
4
¸
. (8.5.5)
8.5 Vortex Ring and Sphere 201
The equations of motion of the vortex are cast in nondimensional terms by
deﬁning
X =
x
01
a
. R =
r
0
a
. V =
I
2πa
. T =
Vt
a
. c =
U
V
.
Then,
d R
dT
= −
1
R
∂+
∂ X
(R. X).
d X
dT
=
1
R
∂+
∂ R
(R. X) ÷
1
2R
¸
ln
8h
σ
0
¸
aR
h
¸3
2
−
1
4
¸
.
(8.5.6)
where + =
c R
2
2
1 −
1
(R
2
÷ X
2
)
3
2
−(R
2
÷ X
2
)
1
2
(
ˆ
+
÷
÷
ˆ
+
−
){K(A) − E(A)}.
ˆ
+
±
=
(R ∓ R
/
)
2
÷(X − X
/
)
2
. A =
ˆ
+
−
−
ˆ
+
÷
ˆ
+
−
÷
ˆ
+
÷
.
R
/
=
R
R
2
÷ X
2
. X
/
=
X
R
2
÷ X
2
.
Figure 8.5.2 illustrates the sections in the vertical plane of symmetry of
the sphere of two typical vortex trajectories predicted by Equations (8.5.6). In
both cases the integration is started ﬁve sphere diameters upstream with the
following initial values for the vortex ring radius and core radius,
h = 0.8a. σ
0
= 0.05h. (8.5.7)
The solid and brokenline curves in the ﬁgure correspond respectively to c = 0
Fig. 8.5.2.
202 8 Further Worked Examples
(no mean ﬂow) and c = 3. The latter value is chosen to make the mean stream
velocity U approximately the same as the selfinduced velocity u
I
at large
distances from the sphere. This has a relatively small effect on the trajectory,
although the convection speed of the ring past the sphere is greatly increased.
8.5.1 Acoustic Pressure
When the sphere is acoustically compact, equation (5.4.4) gives
p(x. t ) ≈
−ρ
0
x
j
4πc
0
[x[
2
∂
∂t
(ω ∧ v)
y. t −
[x[
c
0
· ∇Y
j
(y) d
3
y. [x[ →∞.
(8.5.8)
For the purpose of evaluating the integral we may neglect the ﬁnite core size of
the vortex, and set
ω = I
ˆ
θδ(r −r
0
(t ))δ(x
1
− x
01
(t )). (8.5.9)
where
ˆ
θ is a unit azimuthal vector, locally tangential to the vorticity ω and ori
entated in the clockwise direction when the vortex ring is viewed fromupstream,
as indicated in Fig. 8.5.1.
The force on the sphere is in the mean ﬂow direction – the effective acoustic
source is the unsteady drag – and only the component
Y
1
= y
1
1 ÷
a
3
2[y[
3
≡ y
1
¸
1 ÷
a
3
2
r
2
÷ y
2
1
3
2
(from Table 3.9.1)
of the Kirchhoff vector makes a nontrivial contribution to (8.5.8). The produc
tion of sound is therefore a nonlinear event – the source explicitly involves
only the selfinduced velocity and the velocity induced by the image vortex;
the mean ﬂow velocity U = U∇Y
1
is absent because
(ω ∧U∇Y
1
)
y. t −
[x[
c
0
· ∇Y
1
(y) ≡ 0.
However, the amplitude and characteristic frequency of the sound both increase
with U.
Substituting (8.5.9) into (8.5.8) and evaluating the integral, we ﬁnd
p(x. t ) ≈
−ρ
0
I cos O
2c
0
[x[
∂
∂t
¸
r
0
dx
01
dt
∂Y
1
∂r
−
dr
0
dt
∂Y
1
∂y
1
(r
0
.x
01
)
. [x[ →∞.
8.5 Vortex Ring and Sphere 203
Fig. 8.5.3.
where the quantity in the square braces is evaluated at the retarded position
of the vortex ring, and O is the angle between the radiation direction and the
x
1
axis illustrated in the upper part of Fig. 8.5.3. Expressing this result in
nondimensional form, we have
p(x. t )
ρ
0
V
2
M cos O(a¡[x[)
≈ π
∂
∂T
¸
3R
2
X
2(R
2
÷ X
2
)
5
2
d X
dT
÷
1 ÷
R
2
−2X
2
2(R
2
÷ X
2
)
5
2
R
d R
dT
¸
.
where M =V¡c
0
, and R and X are the solutions of the vortex equations of
motion (8.5.6).
The nondimensional acoustic pressure signatures plotted in Fig. 8.5.3 are
for the same the initial conditions (8.5.7) considered above for the vortex ring
trajectories in Fig. 8.5.2. The thick solid curve is the pressure proﬁle in the
204 8 Further Worked Examples
absence of mean ﬂow (U =0). The positions of the vortex ring in this case at
several different retarded times V[t ]¡a are marked on the thick curve in the
upper part of Fig. 8.5.3 (time being measured from the instant that the ring
crosses the centre of the sphere). Similarly, the thinline curves in the ﬁgure
give the pressure signature and retarded positions for U =3V, when the self
induction velocity u
I
≈ U at large distances from sphere. Both the amplitude
and frequency of the sound are increased because of the increased convection
velocity of the vortex past the sphere.
8.6 Vortex Pair Incident on a Wall Aperture
Hydrodynamic motion in the vicinity of an aperture in a large thin wall generally
produces an unsteady volume ﬂux through the aperture, which is acoustically
equivalent to a monopole source when the aperture is compact. The upper part
of Fig. 8.6.1 depicts a simple model of such a source. The rigid wall coincides
with the plane x
1
=0, and is pierced by a twodimensional slit aperture of
width 2a whose centerline extends along the x
3
axis. A vortex pair aligned with
the x
3
axis, consisting of vortices of strengths ±I at the respective complex
positions
z
0
= x
01
÷i x
02
and z
∗
0
= x
01
−i x
02
is incident on the aperture from the left (x
1
= −∞).
The motion is evidently symmetric with respect to the x
1
axis, and the trans
formation
ζ =
z
√
z
2
÷a
2
(z = x
1
÷i x
2
)
maps the region Imz > 0 cut along the upper section x
2
> a of the wall onto
the upper half of the ζ plane. By the usual method, we accordingly obtain the
equation of motion of the vortex pair in the form (Karweit, 1975)
dZ
∗
dT
=
3i Z
Z
2
÷1
÷
2i
(Z
2
÷1)
3
2
{Z¡
√
Z
2
÷1 −(Z¡
√
Z
2
÷1)
∗
}
. (8.6.1)
where Z =
z
0
a
. T =
Vt
a
. V =
I
4πa
.
and
dz
0
dt
= V
dZ
dT
≡ V(u ÷i :).
Let the initial separation of the vortices at x
1
= − ∞ be 2h. To integrate the
equation, we can set z
0
= − L ÷i h at a convenient initial (but arbitrary) time
8.6 Vortex Pair Incident on a Wall Aperture 205
Fig. 8.6.1.
T = T
/
, where L ·a. When h¡a is smaller than 2¡3
3
2
≈ 0.385 the vortex pair
passes through the aperture in the manner indicated in Fig. 8.6.1 for h¡a = 0.35.
For larger values of h¡a the trajectories of the two vortices separate; the vortices
travel along symmetric paths parallel to the wall on either side of the aperture,
as illustrated for h¡a = 0.6.
The acoustic pressure in the far ﬁeld is given by
p(x. t ) ≈ −ρ
0
ω ∧ v ·
∂G
∂y
d
2
y dτ. (8.6.2)
where G is the compact Green’s function (3.9.10) for the wall aperture
G(x. y. t −τ)≈−
√
c
0
sgn(x
1
)
π
√
2π[x[
χ(t −τ −[x[¡c
0
)
√
t −τ −[x[¡c
0
Re
¸
ln
˜ z
a
÷
˜ z
2
a
2
−1
¸
.
˜ z = y
2
÷i y
1
. (8.6.3)
206 8 Further Worked Examples
and
χ(t ) = H(t )
∞
0
ln(maξ
2
¡4c
0
t )e
−ξ
2
dξ
[ln(maξ
2
¡4c
0
t )]
2
÷π
2
. m = 1.781072.
The dependence on source position y in (8.6.3) is contained entirely in the
logarithmic term, which represents the velocity potential of the ideal ﬂow that
wouldbe producedthroughthe aperture (fromleft toright) bya uniformpressure
drop across the wall.
When vortex shedding from the aperture edges is ignored,
ω = Ikδ(y
1
− x
01
)δ(y
2
− x
02
) −Ikδ(y
1
− x
01
)δ(y
2
÷ x
02
).
where k is a unit vector in the x
3
direction (out of the plane of the paper in
Fig. 8.6.1).
If we deﬁne
˜
W(Z) =
1
˜
Z
2
−1
˜
Z=i Z
∗
. ˆ χ(T) = χ(t ). M =
V
c
0
.
and put
ˆ
T =
Vτ
a
in the integral (8.6.2).
we ﬁnd
p(x. t ) ≈
2
5
2
ρ
0
V
2
sgn(x
1
)
√
πM
a
[x[
1
2
[T]
−∞
Re(
˜
W
∗
(Z)(u ÷i :))(
ˆ
T)
ˆ χ([T] −
ˆ
T) d
ˆ
T
[T] −
ˆ
T
. [T] =
V[t ]
a
.
Therefore, by setting λ =
[T] −
ˆ
T we can write
p(x. t )
ρ
0
V
2
sgn(x
1
)(a¡[x[)
1
2
≈
2
7
2
√
πM
∞
0
Re(
˜
W
∗
(Z)(u ÷i :))([T] −λ
2
) ˆ χ(λ
2
) dλ.
[x[ →∞. (8.6.4)
where
ˆ χ(λ
2
) =
∞
0
ln(mMξ
2
¡4λ
2
)e
−ξ
2
dξ
[ln(mMξ
2
¡4λ
2
)]
2
÷π
2
.
8.6 Vortex Pair Incident on a Wall Aperture 207
As before, the upper limit of integration in (8.6.4) is actually λ =
√
[T] − T
/
,
where T
/
is the nondimensional initial time fromwhich the motion of the vortex
pair is calculated.
The value of the integral depends weakly on the characteristic Mach number
M =
V
c
0
≡
I
4πac
0
.
This is just the selfinduced convection Mach number of the vortex pair when
separated by a distance 2a. We have taken M = 0.03 for the farﬁeld acoustic
pressure signatures plottedinFig. 8.6.1; inair this wouldimplythat V∼10 m/sec.
The ﬂow induced by the vortex pair approaching the wall forms a localized
twodimensional jet between the vortices, directed toward the wall. The resis
tance of the wall to this ﬂow causes the pressure just to the left of the wall
aperture to rise, forcing ﬂuid through the aperture into the region x
1
> 0. The
radiation therefore has the characteristics of an acoustic monopole source for
x
1
> 0 and a sink for x
1
0. Numerical results are illustrated in the ﬁgure for
h¡a = 0.35. 0.6. In each case, the time origin has been adjusted to correspond
approximately with the peak in the radiated acoustic pressure, which occurs
when the vortices pass close to the edges of the aperture. When h¡a =0.6 the
vortices do not penetrate the aperture but are deﬂected by the wall; this pro
duces a relatively larger pressure rise than for h¡a =0.35, where the vortices
pass through the aperture. The maximum acoustic pressure amplitude is found
to occur when h¡a just exceeds the critical value (∼0.385), when the vortex
trajectories pass very close to the aperture edges. Further increases of h¡a be
yond 0.6 result in a gradual reduction in the amplitude of the sound, and a
corresponding increase in the width of the acoustic pulse (i.e., a decrease in the
characteristic frequency of the sound).
Bibliography
Batchelor, G.K. 1967. An Introduction to Fluid Dynamics. Cambridge: University
Press.
Cannell, P. and Ffowcs Williams, J.E. 1973. Radiation from line vortex ﬁlaments
exhausting from a twodimensional semiinﬁnite duct. Journal of Fluid
Mechanics 58: 65–80.
Crighton, D.G. 1972. Radiation from vortex ﬁlament motion near a half plane. Journal
of Fluid Mechanics 51: 357–362.
Crighton, D.G. 1975. Basic principles of aerodynamic noise generation. Progress in
Aerospace Sciences 16: 31–96.
Crighton, D.G. 1985. The Kutta condition in unsteady ﬂow. Annual Reviews of Fluid
Mechanics 17: 411–445.
Crighton, D.G., Dowling, A.P., Ffowcs Williams, J.E., Heckl, M., and Leppington,
F.G. 1992. Modern Methods in Analytical Acoustics (Lecture Notes). London:
SpringerVerlag.
Curle, N. 1955. The inﬂuence of solid boundaries upon aerodynamic sound.
Proceedings of the Royal Society of London A231: 505–514.
Dhanak, M.R. 1981. Interaction between a vortex ﬁlament and an approaching rigid
sphere. Journal of Fluid Mechanics 110: 129–147.
Dowling, A.P. and Ffowcs Williams, J.E. 1983. Sound and Sources of Sound.
Chichester: Ellis Horwood.
Ffowcs Williams, J.E. 1963. The noise from turbulence convected at high speed.
Philosophical Transactions of the Royal Society of London A255: 469–503.
Ffowcs Williams, J.E. and Hawkings, D.L. 1969. Sound generation by turbulence and
surfaces in arbitrary motion. Philosophical Transactions of the Royal Society of
London A264: 321–342.
Ffowcs Williams, J.E. and Hall, L.H. 1970. Aerodynamic sound generation by
turbulent ﬂow in the vicinity of a scattering halfplane. Journal of Fluid
Mechanics 40: 657–670.
Ffowcs Williams, J.E. 1974. Sound production at the edge of a steady ﬂow. Journal of
Fluid Mechanics 66: 791–816.
Goldstein, M.E. 1976. Aeroacoustics. New York: McGrawHill.
Goldstein, S. 1960. Lectures on Fluid Mechanics. New York: Interscience.
Howe, M.S. 1975a. Contributions to the theory of aerodynamic sound with
209
210 Bibliography
application to excess jet noise and the theory of the ﬂute. Journal of Fluid
Mechanics 71: 625–673.
Howe, M.S. 1975b. The generation of sound by aerodynamic sources in an
inhomogeneous steady ﬂow. Journal of Fluid Mechanics 67: 579–610.
Howe, M.S. 1989. On unsteady surface forces, and sound produced by the normal
chopping of a rectilinear vortex. Journal of Fluid Mechanics 206: 131–153.
Howe, M.S. 1998a. Acoustics of Fluid–Structure Interactions. Cambridge: Cambridge
University Press.
Howe, M.S. 1998b. The compression wave produced by a highspeed train entering a
tunnel. Proceedings of the Royal Society A454: 1523–1534.
Howe, M.S. 2001. Vorticity and the theory of aerodynamic sound. Journal of
Engineering Mathematics 41: 367–400.
Howe, M.S., Iida, M., Fukuda, T., and Maeda, T. 2000. Theoretical and experimental
investigation of the compression wave generated by a train entering a tunnel with
a ﬂared portal. Journal of Fluid Mechanics 425: 111–132.
Karweit, M. 1975. Motion of a vortex pair approaching an opening in a boundary.
Physics of Fluids 18: 1604–1606.
Kasoev, S.G. 1976. Sound radiation from a linear vortex over a plane with a projecting
edge. Soviet Physics Acoustics 22: 71–72.
Kelvin, Lord. 1867. On vortex motion. Transactions of the Royal Society of Edinburgh
25: 217–260.
Knio, O.M., Ting, L., and Klein, R. 1998. Interaction of a slender vortex with a rigid
sphere: Dynamics and far ﬁeld sound. Journal of the Acoustical Society of
America 103: 83–98.
Lamb, Horace. 1932. Hydrodynamics. 6th ed. Cambridge: Cambridge University Press.
Landau, L.D. and Lifshitz, E.M. 1987. Fluid Mechanics. 2nd ed. Oxford: Pergamon.
Lighthill, M.J. 1952. On sound generated aerodynamically. Part I: General theory.
Proceedings of the Royal Society of London A211: 564–587.
Lighthill, M.J. 1956. The image system of a vortex element in a rigid sphere.
Proceedings of the Cambridge Philosophical Society 52: 317–321.
Lighthill, M.J. 1958. An Introduction to Fourier Analysis and Generalised Functions.
Cambridge: Cambridge University Press.
Lighthill, M.J. 1963. Laminar Boundary Layers. Edited by L. Rosenhead. Chs. 1, 2.
Oxford: University Press.
Lighthill, James. 1978. Waves in Fluids. Cambridge: Cambridge University Press.
Lighthill, J. 1986. An Informal Introduction to Theoretical Fluid Mechanics. Oxford:
Clarendon Press.
Maeda, T., Matsumura, T., Iida, M., Nakatani, K., and Uchida, K. 1993. Effect of
shape of train nose on compression wave generated by train entering tunnel.
Proceedings of the International Conference on Speedup Technology for Railway
and Maglev Vehicles. Japan Society of Mechanical Engineers (Yokohama, Japan
22–26 November) pp. 315–319.
MilneThomson, L.M. 1968. Theoretical Hydrodynamics. 5th ed. London: Macmillan.
Moore, D.W. 1980. The velocity of a vortex ring with a thin core of elliptical
crosssection. Proceedings of the Royal Society of London A370: 407–415.
M¨ ohring, W. 1978. On vortex sound at low Mach number. Journal of Fluid Mechanics
85: 685–691.
M¨ ohring, W. 1980. Modelling low Mach number noise. Mechanics of Sound
Generation in Flows. Edited by E.A. M¨ uller, pp. 85–96. Berlin: SpringerVerlag.
Powell, A. 1960. Aerodynamic noise and the plane boundary. Journal of the Acoustical
Society of America 32: 962–990.
Bibliography 211
Powell, A. 1963. Mechanisms of Aerodynamic Sound Production. AGARD Report No.
466.
Powell, A. 1964. Theory of vortex sound. Journal of the Acoustical Society of America
36: 177–195.
Rayleigh, Lord. 1945. Theory of Sound. 2 vols. New York: Dover.
Saffman, P.G. 1993. Vortex Dynamics. Cambridge: University Press.
Sears, W.R. 1941. Some aspects of nonstationary airfoil theory and its practical
applications. Journal of the Aeronautical Sciences 8: 104–108.
Ting, L. and Klein, R. 1991. Viscous vortical ﬂows. Lecture Notes in Physics, Vol. 374.
New York: SpringerVerlag.
Index
acoustic wave number, 43
added mass tensor, 68
contribution to surface force, 69, 94
relation to bound vorticity, 94
sphere, 70
aerodynamic sound, deﬁnition, 1
acoustic analogy, 25
airfoil theory, linearized, 148
airfoil, elliptic, 187
rectangular, 187, 190
analogy, acoustic, 25
BiotSavart, application to creeping ﬂow, 99
formula, 88
used to calculate sound source, 114
with bound vorticity, 91
bladevortex interactions, twodimensions
150, 175
linear theory, 178, 188
nonlinear, 180, 189
threedimensions, 158, 186
blockage, 167
bluff body, 194
bound vorticity, 91
boundary, solid, 41
bubble, in sound ﬁeld, 112
candle, blowing out, 91
causality, 11
circulation, deﬁnition, 87
Kelvin’s theorem, 87
per unit length of wake, 148
compact, acoustically, 18
body, 36, 128
turbulence eddies, 30, 38
compact Green’s function, preliminary
deﬁnition, 49, 52
airfoil of variable chord, 71
circular cylinder, 58
cylindrical bodies, 58
deﬁnition for Helmholtz equation, 64
deﬁnition for wave equation, 65
duct (tunnel) entrance, 77, 170
duct with neck, 77, 80
for incompressible ﬂow, 81
general form, 70
halfplane, 74
rigid strip, 60
sphere, 53
symmetric, 63
twodimensional, 136, 154
uniform duct, 76
wall aperture, 75
wall cavity, 72
wall projection, 72
complex potential, 100
circular cylinder, 102
halfplane, 105
source, 103
vortex, 103
complex velocity, 101
compression wave, generated by train, 166
compressive stress, 35
conformal transformation, 100
halfplane, 105
strip 61,
continuity equation, 2
linearized, 5
with source, 5
control surface, deﬁnition, 32
creeping ﬂow, 99
Crighton, 143, 149
Crocco’s equation, 85
Curle, theory, 32
differential equation, 35
integral equation, 36
213
214 Index
decibel, 4
delta function
Fourier integral for, 43
in three dimensions, 9
density, 2
values for air and water, 3
diffusion, of vorticity, 87
dipole, deﬁnition, 14
directivity, 16
drag, 195
far ﬁeld form, 21
ﬂuid volume, 15
in Curle’s equation, 35
in Helmholtz equation, 45
lift, 188, 190, 195
near compact sphere, 54
near edge, 62
drag, 191, 195
efﬁciency, acoustic, 32
of quadrupole radiation, 32
of surface dipoles, 37
end correction, 78
energy, equation, 4
ﬂux, 18
kinetic, 82, 90
enthalpy, see total enthalpy
entropy, 4
far ﬁeld, deﬁnition, 17
acoustic, 19
calculation of, 20
hydrodynamic, 89
in terms of impulse, 89
Ffowcs Williams, 29
Force, impulse formula, 93
added mass contribution, 94
formula for translating body, 96
lift, 182, 185
produced by gust, 147
role of bound vorticity, 94
Sears’ formula, 149
Fraunhofer approximation, 21
Green’s function, see also compact
Green’s function
deﬁnition, 12
for Helmholtz equation, 45
freespace, 12, 46
gust, time harmonic, 148
vortex, 146, 156
Heaviside unit function, deﬁnition, 32
Helmholtz, equation, 44
Green’s function, 45
inhomogeneous equation, 44, 57
homentropic, 4
hydrodynamic, pressure, 38
far ﬁeld, 89
wavelength, 148
ideal acoustic medium, 25, 27
impedance, 48
impulse, deﬁnition, 90
theory of vortex sound, 131
incompressible ﬂuid, 7, 90, 93, 96
infrasound, 166, 174
integrals, transformation using Heaviside
function, 32
intensity, acoustic, 19
irrotational ﬂow, sound waves in, 118
Karweit, 204
Kasoev, 191
Kelvin, circulation theorem, 87
deﬁnition of vorticity, 82
formula for vortex ring speed, 92, 199
theorem on kinetic energy, 84
kinetic energy, 82
in terms of vorticity, 90
Kelvin’s theorem on, 84
Kirchhoff spinning vortex, 132
Kirchhoff vector, deﬁnition, 52
circular cylinder, 60
rigid strip, 61
singularities of, 146, 159, 179
special cases, 71
sphere, 54
KraichnanPhillips theorem, 39
Kutta condition, 149
see also singularities
Lamb vector, 85, 96, 136
Laplace equation
axisymmetric, 53
polar from, 60
two dimensions, 100
leading edge, Kutta condition, 150
suction, 178, 182, 185
lift, 158
lift dipole, 188, 190, 195
Lighthill, 1
acoustic analogy, 25
eighth power law, 29
equation, 28
equation reformulated in terms of vorticity,
117
stress tensor, 28
linear, acoustics, 4
Lighthill’s equation, 27
momentum equation, 5
Mach number, 29
Maeda, 171
Index 215
micropressure wave, 166
M¨ ohring, vortex sound formula, 132, 135
momentum equation, 3
Crocco’s form, 85
linearized, 5
Reynolds’ form, 27
momentum ﬂux tensor, 27
monopole, 9, 13
in Curle’s equation, 35
NavierStokes equation, 3
near ﬁeld, hydrodynamic, 17
no slip condition, 85
nonlinear steepening in tunnel, 166
plane wave, 19, 23
point source, deﬁnition, 9
impulsive, 10, 12
in Helmholtz equation, 44
monopole, 9
potential, velocity, 6
vector, 88
potential ﬂow interaction, 150, 161, 164
Powell, 38
power, acoustic, 19
quadrupole, deﬁnition, 15
far ﬁeld form, 22
image in plane wall, 39
in Helmholtz equation, 45
Lighthill’s, 28
radiation condition, 11, 13
rate of strain tensor, 26, 84
Rayleigh, 47
end correction, 78
reciprocal, theorem, 46, 127
problem for compact Green’s function, 50
reduced frequency, deﬁnition, 148
retarded potential, deﬁnition, 13
retarded time, deﬁnition, 14
Reynolds number, deﬁnition, 36
Reynolds stress, deﬁnition, 28
in Lighthill tensor, 29
linear, 29
RungeKutta integration, 108, 163
Sears, 147
Sears function, 149
self potential, vortex, 107
singularities, deleting from Green’s function,
146, 159, 179, 187
speciﬁc heats, ratio, 4
speed of sound, deﬁnition, 6
values for air and water, 7
sphere, added mass, 68
interacting with a vortex, 162, 199
modelled by point source, 9
pulsating, 7, 67
vibrating, 16, 57, 67
slender body approximation, 169
spinning vortex pair, 120
stationary phase, 75
Stokes, drag, 98
stream function, 200
Stokesian ﬂuid, 3
stress tensor
Lighthill’s, 28
viscous, 26
suction force, 178, 182, 185
dipole, 188, 190
surface, compact, 36
impedance, 48
noncompact, 37
test function, 9
thermodynamics, Second Law, 11
total enthalpy, deﬁnition for homentropic ﬂow,
85
acoustic variable, 116
in terms of velocity potential, 117
relation to pressure, 117
tunnel, 166, 174
turbulent nozzle ﬂow, 25
velocity, acoustic particle, 5, 18
in terms of impulse, 89
potential, 6
vibrating body, 56
low frequency radiation from, 65, 129
vibrating sphere, 16, 57, 67
viscosity, 3
kinematic, 3
values for air and water, 3
volume source, 5
dipole, 15
vortexairfoil interaction, linear theory, 156,
178, 188
nonlinear, 180, 189
periodic, 183
threedimensional, 158, 186
twodimensional, 150
vortex, coaxial rings, 133
complex potential, 103
equation of motion, 106, 175, 180, 192, 195,
198, 204
ﬁnite core, 160
line, 84, 158
motion near halfplane, 108
outside cylinder, 111
ring, 92, 173, 199
self potential, 107
shedding, 145, 159
tube, 84
216 Index
vortex pair, 204
spinning, 120
vortex sheet, 87, 88, 91
wake, 150
vortex sound, at low Mach numbers,
119
for cylindrical bodies, 130, 194
for nonvibrating body, 130
from spinning vortices, 120
from spoiler, 191
from vortex interacting with cylinder,
139
from vortex near halfplane, 143, 173
from vortex near sphere, 162, 199
from wall aperture, 204
in terms of surface force, 129
vortex sound equation, 116, 118
linearized, 156
low Mach number approximation, 120
vortexsurface interaction noise, 124
general formula, 127
high Reynolds number form, 128
inﬂuence of vortex shedding, 143
vortex near sphere, 162
vortex near spoiler, 191
wall mounted cylinder, 197
vorticity, bound, 91
equation, 84
equation for Stokesian ﬂuid, 86
in force formulae, 93, 96
Kelvin’s deﬁnition, 82
molecular diffusion, 87
source of sound, 114
wake, cylinder, 148
vortex sheet, 150
wall drag, 191
wave equation
classical acoustics, 6
for pressure, 6
in irrotational mean ﬂow, 118
inhomogeneous, 13
radially symmetric, 10
wavelength
acoustic, 7, 17, 43
at 1 kHz, 7
hydrodynamic, 148
wave number, 43
This page intentionally left blank
Theory of Vortex Sound
Theory of Vortex Sound is an introduction to the theory of sound generated by hydrodynamic ﬂows. Starting with a review of elementary theoretical acoustics, the book proceeds to a uniﬁed treatment of low Mach number vortexsurface interaction noise in terms of the compact Green’s function. Problems are provided at the end of each chapter, many of which can be used for extended student projects, and a whole chapter is devoted to worked examples. It is designed for a onesemester introductory course at the advanced undergraduate or graduate levels. Great care is taken to explain underlying ﬂuid mechanical and acoustic concepts, and to describe as fully as possible the steps in a complicated derivation. M.S. Howe has been Professor in the Department of Aerospace and Mechanical Engineering at Boston University since 1992. He is a Fellow of the Institute of Acoustics (U.K.) and of the Acoustical Society of America.
S. Drazin Stability. Ockendon and J. Instability. Stirling Perturbation Methods E. Johnson Rareﬁed Gas Dynamics Carlo Cercignani Symmetry Methods for Differential Equations Peter E. O’Malley A Modern Introduction to the Mathematical Theory of Water Waves R. M. S. J. SelfSimilarity. Pozrikidis Nonlinear Wave Processes in Acoustics K. Hinch The Thermomechanics of Plasticity and Fracture Gerard A. J. C. Gibbon Viscous Flow H. Johnson The Kinematics of Mixing J. Chapman Wave Motion J. Ablowitz and Athanassios S. Naugolnykh and L. Drazin and R. Ockendon Scaling. Fokas Mathematical Models in the Applied Sciences A. R.Cambridge Texts in Applied Mathematics Maximum and Minimum Principles M. Maugin Boundary Integral and Singularity Methods for Linearized Viscous Flow C. Sewell Solitons P. Ciarlet Integral Equations David Porter and David S. G. Doering and J. G. Ottino Introduction to Numerical Linear Algebra and Optimisation Philippe G. C. Fowler Thinking About Ordinary Differential Equations Robert E. and Intermediate Asymptotics G. and Chaos Paul Glendinning Applied Analysis of the Navier–Stokes Equations C. Ostrovsky Nonlinear Systems P. G. King . J. I. R. Barenblatt A First Course in the Numerical Analysis of Differential Equations Arieh Iserles Complex Variables: Introduction and Applications Mark J. Hydon High Speed Flow C. D. Billingham and A.
LeVeque Introduction to Hydrodynamic Stability P. Bertozzi Inﬁnite Dimensional Dynamical Systems James C. Robinson An Introducion to Symmetry Analysis Brian J. Cantwell Backlund and Darboux Transformations C. G. S. Howe . Rogers and W. A. K. Drazin Theory of Vortex Sound M. Davidson Linear Elastic Waves John G.An Introduction to Magnetohydrodynamics P. Harris Vorticity and Incompressible Flow Andrew J. Schief FiniteVolume Methods for Hyperbolic Problems Randall J. Majda and Andrea L.
HOWE Boston University . S.Theory of Vortex Sound M.
Singapore. Cambridge. Madrid.org/9780521812818 © Cambridge University Press 2003 This book is in copyright. United Kingdom Published in the United States of America by Cambridge University Press. accurate or appropriate. New York. Cape Town. and does not guarantee that any content on such websites is. or will remain. São Paulo Cambridge University Press The Edinburgh Building. . Subject to statutory exception and to the provision of relevant collective licensing agreements.cambridge. First published in print format 2002        eBook (NetLibrary)  eBook (NetLibrary)  hardback  hardback  paperback  paperback Cambridge University Press has no responsibility for the persistence or accuracy of s for external or thirdparty internet websites referred to in this book.cambridge. no reproduction of any part may take place without the written permission of Cambridge University Press. Melbourne.org Information on this title: www. New York www. Cambridge .
To Shˆ n Ffowcs Williams o .
Contents
Preface 1 Introduction 1.1 What is Vortex Sound? 1.2 Equations of Motion of a Fluid 1.3 Equation of Linear Acoustics 1.4 The Special Case of an Incompressible Fluid 1.5 Sound Produced by an Impulsive Point Source 1.6 FreeSpace Green’s Function 1.7 Monopoles, Dipoles, and Quadrupoles 1.8 Acoustic Energy Flux 1.9 Calculation of the Acoustic Far Field Problems 1 2 Lighthill’s Theory 2.1 The Acoustic Analogy 2.2 Lighthill’s v 8 Law 2.3 Curle’s Theory 2.4 Sound Produced by Turbulence Near a Compact Rigid Body 2.5 Radiation from a Noncompact Surface Problems 2 3 3.1 3.2 3.3 3.4 The Compact Green’s Function The Inﬂuence of Solid Boundaries The Helmholtz Equation The Reciprocal Theorem TimeHarmonic Compact Green’s Function ix
page xiii 1 1 2 4 7 10 12 13 18 20 23 25 25 29 32 36 37 39 41 41 44 46 49
x
Contents 53 58 63 65 70 79 82 82 84 88 93 100 106 112 114 114 116 124 128 130 131 132 136 136 139 145 150 154 156 156 158 162 166 172
3.5 Compact Green’s Function for a Rigid Sphere 3.6 Compact Green’s Function for Cylindrical Bodies 3.7 Symmetric Compact Green’s Function 3.8 LowFrequency Radiation from a Vibrating Body 3.9 Compact Green’s Function Summary and Special Cases Problems 3 4 4.1 4.2 4.3 4.4 Vorticity Vorticity and the Kinetic Energy of Incompressible Flow The Vorticity Equation The Biot–Savart Law Surface Force in Incompressible Flow Expressed in Terms of Vorticity 4.5 The Complex Potential 4.6 Motion of a Line Vortex Problems 4
5 Vortex Sound 5.1 The Role of Vorticity in Lighthill’s Theory 5.2 The Equation of Vortex Sound 5.3 Vortex–Surface Interaction Noise 5.4 Radiation from an Acoustically Compact Body 5.5 Radiation from Cylindrical Bodies of Compact Cross Section 5.6 Impulse Theory of Vortex Sound Problems 5 6 Vortex–Surface Interaction Noise in Two Dimensions 6.1 Compact Green’s Function in Two Dimensions 6.2 Sound Generated by a Line Vortex Interacting with a Cylindrical Body 6.3 Inﬂuence of Vortex Shedding 6.4 Blade–Vortex Interaction Noise in Two Dimensions Problems 6 7 Problems in Three Dimensions 7.1 Linear Theory of Vortex–Airfoil Interaction Noise 7.2 Blade–Vortex Interactions in Three Dimensions 7.3 Sound Produced by Vortex Motion near a Sphere 7.4 Compression Wave Generated When a Train Enters a Tunnel Problems 7
Contents 8 8.1 8.2 8.3 8.4 8.5 8.6 Further Worked Examples Blade–Vortex Interactions in Two Dimensions Parallel Blade–Vortex Interactions in Three Dimensions Vortex Passing over a Spoiler Bluff Body Interactions: The Circular Cylinder Vortex Ring and Sphere Vortex Pair Incident on a Wall Aperture
xi 175 175 186 191 194 199 204 209 213
Bibliography Index
.
Many examples of this type are discussed. the effective acoustic sources turn out to be localized to one or more regions that are small compared to the acoustic wavelength. at the same time. Great care has therefore been taken to discuss underlying ﬂuid mechanical and acoustic concepts. combustion and ‘entropy’ sources of sound. culminating in a routine procedure for estimating the sound. an easy identiﬁcation of those parts of a structure that are likely to be important sources of sound. to conﬁne an introductory discussion speciﬁcally to low Mach number ﬂows. and many are illequipped to deal with the novel analytical techniques that have been developed to investigate the coupling. and they are simple enough for the student to acquire an intuitive understanding of the method of xiii . Most students at this level possess an insufﬁcient grasp of basic principles to appreciate the subtle coupling of the hydrodynamic and acoustic ﬁelds. The subject is itself a subsection of the theory of aerodynamic sound. less than about 0. therefore. and providing. The book is based on an introductory onesemester graduate level course given on several occasions at Boston University. which encompasses a much wider range of problems also involving. A considerable number of practical problems occur at low Mach numbers (say.4). It is then possible to investigate a number of idealized hydrodynamic ﬂows involving elementary distributions of vorticity adjacent to solid boundaries. For a broad range of such problems. and to explain as fully as possible the steps in a complicated derivation. This permits the development of a uniﬁed theory of sound production by vortex– surface interactions in terms of the compact Green’s function. and to analyze in detail the sound produced by these vortex–surface interactions. It seems reasonable. for example. and a corresponding broad range of noise problems encountered in industrial applications.Preface Vortex sound is the branch of ﬂuid mechanics concerned with the conversion of hydrodynamic (rotational) kinetic energy into the longitudinal disturbances we call sound.
In particular. I wish to thank my former students H. involving the implementation of a widely applicable yet standard procedure for the prediction of sound generation at low Mach numbers. such as δ(x) = π(x 2 + 2) . motivates a student to understand the ostensibly difﬁcult parts of the theory. One or more of the problems appended to some of the later chapters can form the basis of a project. A. M. DeBenedictis.  f (xn ) M. M. The mathematical ability assumed of the reader is roughly equivalent to that taught in an advanced undergraduate course on Engineering Mathematics. . A. Howe where the summation is over real simple roots of f (x) = 0. The ﬁnal chapter contains a set of worked examples that have been investigated by students at Boston University. Much use is made of the formula δ( f (x)) = n δ(x − xn ) . AbouHussein. Harrison. Experience has shown that the successful completion of this kind of project. Rodrigues. → +0. An elementary understanding of the properties of the Dirac δ function is desirable (Lighthill. N. the reader should be familiar with basic vector differential and integral calculus and with the repeated sufﬁx summation convention of Cartesian tensors (but a detailed knowledge of tensor calculus is not required). including its interpretation as the formal limit of an sequence. By these means the reader is encouraged to investigate both the hydrodynamics and the sound generated by a simple ﬂow. Kim. 1958). S. Zagadou for their considerable help in preparing that chapter.xiv Preface solution and the underlying physics. and F.
In this book we shall consider principally the production of sound by unsteady motions of a ﬂuid. conventional loudspeakers. must possess vorticity. and so forth all fall within the theory of aerodynamic sound in its broadest sense. aeroengine rotor blades. vibrating surfaces.1 What is Vortex Sound? Vortex sound is the sound produced as a byproduct of unsteady ﬂuid motions (Fig. However. 1 . it is now widely recognized that any mechanism that produces sound can actually be formulated as a problem of aerodynamic sound.1).1 Introduction 1. Lighthill (1952) wanted to understand the mechanisms of noise generation by the jet engines of new passenger jet aircraft that were then about to enter service. apart from the high speed turbulent jet – which may be regarded as a distribution of intense turbulence velocity ﬂuctuations that generate sound by converting a tiny fraction of the jet rotational kinetic energy into the longitudinal waves that constitute sound – colliding solid bodies. Our objective. that is.1. combustion and combustion instabilities in rockets. therefore. 1. crackling paper. musical instruments. Thus. The modern theory of aerodynamic sound was pioneered by James Lighthill in the early 1950s. Any ﬂuid that possesses intrinsic kinetic energy. explosions. is to simplify the general aerodynamic sound problem to obtain a thorough understanding of how this happens. energy not directly attributable to a moving boundary (which is largely withdrawn from the ﬂuid when the boundary motion ceases). complex ﬂuid–structure interactions in the larynx (responsible for speech). It is part of the more general subject of aerodynamic sound. and of how the sound can be estimated quantitatively. We shall see that in a certain sense and for a vast number of ﬂows vorticity may be regarded as the ultimate source of the sound generated by the ﬂow.
the state of a ﬂuid is deﬁned when the velocity v and any two thermodynamic variables are speciﬁed. x2 .1. Five scalar equations are therefore required to determine the motion. Typical vortex sound problems. The velocity v and the ﬂuid density ρ must therefore satisfy . momentum.2 1 Introduction Fig.1. x3 ). These equations are statements of the conservation of mass. 1. and energy. 1.2.1 Equation of Continuity Conservation of mass requires the rate of increase of the ﬂuid mass within a ﬁxed region of space V to be equal to the net inﬂux due to convection across the boundaries of V. 1.2 Equations of Motion of a Fluid At time t and position x = (x1 .
2.2 Momentum Equation The momentum equation is also called the Navier–Stokes equation.2) is the material derivative. but also a good approximation in air for calculating the frictional drag at a solid boundary) for which the principal frictional forces are expressed in terms of the shear coefﬁcient of viscosity η. 1. Density and viscosity ρ.284 × 10−3 ν.1: Table 1.2 Equations of Motion of a Fluid the equation of continuity.1) states that div v is equal to the rate of change of ﬂuid volume per unit volume following the motion of the ﬂuid. the viscous or frictional force. ∂t 1 Dρ + div v = 0. The last of Equations (1. kg/ms 1.2. We consider only Stokesian ﬂuids (most liquids and monatomic gases.2. ρ Dt D 1 div v = ρ Dt ρ where ∂ ∂ ∂ D = + v·∇ ≡ + vj Dt ∂t ∂t ∂x j 3 (1.2. m2 /s 1.3) Values of ρ. and body forces F per unit volume. i.1) (1.1. η and ν = η/ρ (the ‘kinematic’ viscosity) for air and water at 10 ◦ C and one atmosphere pressure are given in the Table 1. kg/m3 Air Water 1. Then the momentum equation is ρ Dv 1 = −∇ p + η ∇ 2 v + ∇(div v) + F.2. which has the following equivalent forms ∂ρ + div(ρv) = 0.2.2. it expresses the rate of change of momentum of a ﬂuid particle in terms of the pressure p.284 × 10−6 . Dt 3 (1.2.764 × 10−5 1. the repeated sufﬁx j implies summation over j = 1.1. .433 × 10−5 1.e. div v = 0.23 1000 η. For an incompressible ﬂuid this is zero. 3.. which we shall invariably assume to be constant.
A very loud sound ∼120 dB corresponds to 2 × 10−5 p 120 ≈ × 10( 20 ) = 2 × 10−4 5 p0 10 1. when shock waves are formed by highly nonlinear events. pref where the reference pressure pref = 2 × 10−5 N/m2 . namely. p/ p0 ∼ 0.2.2.4 1 Introduction 1. Thus.2. (1. or when sound is being generated by combustion and other heat sources. (1. s). Thus. p = p0 ≡ 1 atmosphere ( = 105 N/m2 ) is equivalent to 194 dB.5) This equation will be satisﬁed by both the mean (undisturbed) and unsteady components of the ﬂow. The passage of a sound wave in the form of a pressure ﬂuctuation is.4) We may then assume that the pressure and density are related by an equation of the form p = p(ρ. s = constant. of course. Similarly. so that the energy equation becomes s = constant. (1.02. the speciﬁc entropy s of the ﬂuid is uniform and constant throughout the ﬂuid. γ = ratio of speciﬁc heats. for a ‘deafening’ sound of 160 dB. This corresponds to a pressure of about 0.3 Energy Equation This equation must be used in its full generality in problems where energy is transferred by heat conduction. For our purposes it will usually be sufﬁcient to suppose the ﬂow to be homentropic. for an ideal gas p = constant × ρ γ .3 lbs/in2 and is loud enough for nonlinear effects to begin to be important. where frictional dissipation of sound is occurring.6) 1.3 Equation of Linear Acoustics The intensity of a sound pressure p in air is usually measured on a decibel scale by the quantity 20 × log10  p .2. accompanied by a backandforth motion of the ﬂuid at the acoustic .
3. To do this we shall ﬁrst consider sound propagating in a stationary inviscid ﬂuid of mean pressure p0 and density ρ0 . t) on the righthand side 1 Dρ + div v = q. The linearized momentum equation (1. mean density × speed of sound 5 In air the speed of sound is about 340 m/sec.2. at 160 dB v ∼ 5 m/sec.3. In most applications the acoustic amplitude is very small relative to the mean pressure p0 . s) ≈ p(ρ0 . ρ Dt (1.5) .3): ∂ 2ρ ∂q − div F. s) ∂ρ ρ. In the undisturbed and disturbed states we have p0 = p(ρ0 . where p / p0 1.2. 0 s = constant.3.1.2) where q is the rate of increase of ﬂuid volume per unit volume of the ﬂuid. s) + ∂p (ρ. We shall see later that acoustic particle velocity ≈ acoustic pressure .3. say. s). ρ0 ∂t Now eliminate v between (1. ρ /ρ0 1. Thus.3.1) and (1.1). for example.2. ρ . and might represent. The linearized equation is then 1 ∂ρ + div v = q. ∂t (1.3) An equation determining the pressure p alone in terms of q and F is obtained by invoking the homentropic relation (1. we introduce an artiﬁcial generalization by inserting a volume source distribution q(x. the effect of volume pulsations of a small body in the ﬂuid. and sound propagation may be studied by linearizing the equations.1) Before linearizing the continuity equation (1.3.3 Equation of Linear Acoustics particle velocity v.4) (1.5). (1. let the departures of the pressure and density from these mean values be denoted by p . − ∇ 2 p = ρ0 ∂t 2 ∂t (1. at 120 dB v ∼ 5 cm/sec.3) becomes ρ0 ∂v + ∇ p = F. p0 + p = p(ρ0 + ρ .3.
that the motion of a ﬂuid particle is adiabatic). as we shall see.3. It has the dimensions of velocity2 . we obtain ∂q 1 ∂2 − div F. This equation governs the production of sound waves by the volume source q and the force F.7) where the prime ( ) on the acoustic pressure has been discarded. such as the vibrating cone of a loudspeaker.3. ∂t (1. substituting for ρ in (1.6) where the derivative is taken with the entropy s held ﬁxed at its value in the undisturbed ﬂuid. When these terms are absent the equation describes sound propagation from sources on the boundaries of the ﬂuid..3. 2 c0 ∂t 2 This is the wave equation of classical acoustics. in terms of which the perturbation pressure is given by p = −ρ0 ∂ϕ . The volume source q and the body force F would never appear in a complete description of sound generation within a ﬂuid.7) (with F = 0) that the velocity potential is the solution of 1 ∂2 − ∇ 2 ϕ = −q(x.8) It follows from this and (1. This is because only a tiny fraction of the available energy of a vibrating ﬂuid or structure actually radiates away as sound. s (1. In general this can be a dangerous procedure because.e.9) . − ∇ 2 p = ρ0 2 ∂t c0 ∂t 2 (1.5): ρ = p /c0 . (1.6 1 Introduction The derivative is evaluated at the undisturbed values of the pressure and density ( p0 . Hence. small errors in specifying the sources of sound in a ﬂuid can lead to very large errors in the predicted sound. t).4).1) implies the existence of a velocity potential ϕ such that v = ∇ϕ.3.3.3. They are introduced only when we think we understand how to model the real sources of sound in terms of volume sources and forces. When F = 0.3. The implication is that losses due to heat transfer between neighboring ﬂuid particles by viscous and thermal diffusion are neglected during the passage of a sound wave (i. Equation (1. 2 From (1.3. and its square root deﬁnes the speed of sound c0 = ∂p ∂ρ . ρ0 ).
∂ϕ/∂r = vn (t). There are no sources within the ﬂuid. ∇ 2 ϕ = 0. t). r = a where r = x. and the corresponding acoustic wavelength λ at a frequency of 1 kHz (sound of frequency f has wavelength λ = c0 / f ).4. so that ∇ 2ϕ = Hence. Speed of sound and acoustic wavelength c0 m/s Air Water 340 1500 ft/s 1100 5000 km/h 1225 5400 mi/h 750 3400 λ at 1 kHz m 0.5 ft 1 5 7 For future reference.1. δρ In an incompressible ﬂuid the pressure can change by the action of external forces (moving boundaries. but the density must remain ﬁxed. Thus. Therefore. and let its normal velocity be vn (t).1) 1.3. Let the center of the sphere be at the origin.3.). .3 1. (1. and Equation (1.3. c0 = ∞.1. The motion is obviously radially symmetric.4 The Special Case of an Incompressible Fluid Table 1.1 Pulsating Sphere Consider the motion produced by small amplitude radial pulsations of a sphere of mean radius a.4. r 1 ∂ r 2 ∂r r2 ∂ ∂r ϕ = 0.9) reduces to ∇ 2 ϕ = q(x.1 lists the approximate speeds of sound in air and in water. etc. r > a. r > a. 1.4 The Special Case of an Incompressible Fluid Small (adiabatic) pressure and density perturbations δp and δρ satisfy δp 2 ≈ c0 . ϕ= A + B. Table 1. so that q ≡ 0.
Evaluating it for any sphere S of radius r > a.2) decays as 1/r with distance from the sphere. as shown in Fig. the pressure p = −ρ0 ∂ϕ a 2 dvn = ρ0 (t) ∂t r dt a 2 vn (t) . 1. 1.4.4.4. the volume ﬂux q(t) of ﬂuid is the same across any closed surface enclosing the sphere. For any time t.4. we ﬁnd q(t) = S ∇ϕ · dS = 4πa 2 vn (t). we then ﬁnd ϕ=− Thus.1. and exhibits the unphysical characteristic of changing instantaneously everywhere when dvn /dt changes its value. B(t) can be discarded because the pressure ﬂuctuations (∼ −ρ0 ∂ϕ/∂t) must vanish as r → ∞.1. . r > a.3) Fig. Applying the condition ∂ϕ/∂r = vn at r = a. and we may also write ϕ= −q(t) .8 1 Introduction where A ≡ A(t) and B ≡ B(t) are functions of t. r (1. 4πr (1. r > a.
it is permissible to replace the sphere by a point source (a monopole) of the same strength q(t) = rate of change of the volume of the sphere.1.3) for the sphere with the same volume outﬂow in the region r > a = radius of the sphere. Then ∇ϕ · dS ≡ S −A R2 × (4π R 2 ) = q(t). where it satisﬁes ∇ 2 ϕ = 0. for any smoothly varying test function f (x) and any volume V 5 .5) for the point source is strictly valid only for r > 0. we write the solution in the form −q(t) 4π(r 2 + 2) 1 2 ϕ = lim →0 .2 Point Source 9 The incompressible motion generated by a volume point source of strength q(t) at the origin is the solution of ∇ 2 ϕ = q(t)δ(x). in which case ∇ 2 ϕ = lim 3 2 q(t) 4π(r 2 + 2) 2 5 →0 .4) To ﬁnd A. A = −q(t)/4π and ϕ = −q(t)/4πr .4. This conclusion is valid for any pulsating body. The last limit is just equal to q(t)δ(x). This indicates that when we are interested in modelling the effect of a pulsating sphere at large distances r a. (1.4 The Special Case of an Incompressible Fluid 1. where it attains a large maximum ∼3/4π 3 . not just a sphere. What happens as r → 0.4) over the interior of a sphere of radius r = R > 0. we integrate (1. Therefore. The solution must be radially symmetric and given by ϕ= A r for r > 0. Hence.4. Indeed when is small 3 2 /4π (r 2 + 2 ) 2 is also small except close to r = 0. it is not necessarily a good model (especially when we come to examine the production of sound by a pulsating body) in the presence of a mean ﬂuid ﬂow past the sphere. However.4. and use the divergence theorem r <R ∇ 2 ϕ d 3 x = S ∇ϕ · dS. where S is the surface of the sphere. The Solution (1.4.4. where δ(x) = δ(x1 )δ(x2 )δ(x3 ).5) (1. where its value is actually undeﬁned? To answer this question.4. which agrees with the solution (1. > 0.
1) The source exists only for an inﬁnitesimal instant of time at t = 0. (1. impulsive point source δ(x)δ(t) is the solution of 1 ∂2 − ∇ 2 ϕ = δ(x)δ(t).8) 1. (1. 2 c0 ∂t 2 (1. the correct interpretation of the solution ϕ= −1 4πr of ∇ 2 ϕ = δ(x) (1.5 Sound Produced by an Impulsive Point Source The sound generated by the unit.6) for a unit point source (q = 1) is −1 −1 = lim →0 4π (r 2 + 4πr where ∇2 −1 4πr = lim ∇ 2 →0 2) 2 1 . r ≥ 0. and that for r = x > 0 we have to solve 1 ∂ 1 ∂ 2ϕ − 2 2 ∂t 2 r ∂r c0 The identity 1 ∂ r 2 ∂r r2 ∂ ∂r ϕ≡ 1 ∂2 (r ϕ) r ∂r 2 (1.2) . It is evident that the solution is radially symmetric.5.4.5.4.10 enclosing the origin lim 3 V 2 1 Introduction f (x) d 3 x 2) 5 2 →0 4π (r 2 + = f (0) lim = f (0) 0 ∞ 3 2 d 3x 2) 2 5 →0 −∞ ∞ 4π (r 2 + 2) 2 5 3 2r 2 dr (r 2 + = f (0).7) −1 4π (r 2 + 2) 1 2 = lim 3 2 2) 2 5 →0 4π (r 2 + = δ(x).5.3) r2 ∂ ∂r ϕ = 0. r > 0. t) = 0 everywhere.4. This is the deﬁning property of the threedimensional δ function. (1. therefore at earlier times ϕ(x. Thus. where the value of the last integral is independent of .
which requires natural systems to change in the more probable direction. r > 0. the general solution of (1. It is also consistent with the Second Law of Thermodynamics.5. We must therefore set = 0.5) The ﬁrst term on the right represents a spherically symmetric disturbance that propagates in the direction of increasing values of r at the speed of sound c0 as t increases.7)) ϕ= t− r r c0 = lim t− (r 2 + r c0 2) 2 1 →0 .5.1) and examine what happens as → 0. (1.4) and + t+ r r c0 .2) is ϕ= t− r r c0 (1. This is a causality or radiation condition.5. (1.6) Let us substitute this into Equation (1.. By direct calculation we ﬁnd ∇ 2ϕ = (t − r/c0 ) 3 2 (t − r/c0 ) 2 2 (t − r/c0 ) 1 ∂2 (r ϕ) = − − + 5 3 1 2 2 2 2 + 2) 2 2 + 2) 2 r ∂r (r c0r (r c0 (r + 2 ) 2 1 ∂ 2ϕ (t − r/c0 ) = 1 2 ∂t 2 2 2 c0 c0 (r + 2 ) 2 .4. where are arbitrary functions. r ≥ 0. To complete the solution it remains to determine the function . that (in the absence of boundaries) sound produced by a source must radiate away from the source.5 Sound Produced by an Impulsive Point Source 11 permits us to write Equation (1. (1.1. We do this by extending the solution down to the source at r = 0 by writing (c. An event in which sound waves converge on a point from all directions at inﬁnity is so unlikely as to be impossible in practice. Hence.5.f. r > 0.5. 2 ∂r c0 ∂t 2 This has the general solution r ϕ = (t − r/c0 ) + (t + r/c0 ). it would be the acoustic analogue of the farscattered pieces of a broken cup spontaneously reassembling.5. whereas the second represents an incoming wave converging toward x = 0. since it represents sound waves generated at r = ∞ that approach the source rather than sound waves generated by the source and radiating away from the source.2) in the form of the onedimensional wave equation for r ϕ: ∂2 1 ∂2 (r ϕ) − 2 (r ϕ) = 0.
6. y.8) for a source at x = 0 at t = 0 simply by replacing x by x − y and t by t − τ . (1. 4πx c0 (1.8).1). 1 Introduction 1 ∂ 2ϕ 3 2 (t − r/c0 ) 2 2 (t − r/c0 ) − ∇ 2ϕ = + 5 3 2 c0 ∂t 2 (r 2 + 2 ) 2 c0r (r 2 + 2 ) 2 → 4π (t)δ(x) + 0 as → 0. comparing (1. t −τ ) is the causal solution of the wave equation generated by the impulsive point source δ(x − y)δ(t − τ ). it vanishes everywhere for t < 0.5. (1. before the impulsive source is triggered.5.7) where the δ function in the last line follows from (1. 4π This represents a spherical pulse that is nonzero only on the surface of the sphere r = c0 t > 0. 2 c0 ∂t 2 then G(x.7) with the inhomogeneous wave equation (1.2) where G = 0 for t < τ. located at the point x = y at time t = τ .5. 1. t) = r 1 δ t− 4πr c0 ≡ x 1 δ t− . we ﬁnd (t) = and the Solution (1.6. In other words. δ t −τ − 4πx − y c0 (1.8) 1 δ(t). whose radius increases at the speed of sound c0 . and the ‘+ 0’ is obtained by noting that for any smoothly varying test function f (x) and any volume V enclosing the origin 2 V 2 f (x) d 3 x 2) 2 3 r (r 2 + ≈ f (0) = f (0) ∞ −∞ ∞ 0 2 2 d 3x 2) 2 3 r (r 2 + (r 2 + 8π 2r dr 2) 2 3 = 8π f (0) → 0 as → 0.5. y.5.4.1) . Hence.6) becomes ϕ(x.5. t − τ ) = x − y 1 .6 FreeSpace Green’s Function The freespace Green’s function G(x. The formula for G is obtained from the solution (1.12 Therefore. if 1 ∂2 − ∇ 2 G = δ(x − y)δ(t − τ ).
Dipoles. t) is assumed to be generating waves that propagate away from the source region.7) of linear acoustics.3.4) 1 4π x − y F(y. The wave amplitude decreases inversely with distance x − y from the source point y. which radiated at the earlier times t −x−y/c0 . τ ) δ(x − y) δ(t − τ ) d 3 y dτ. t) = ∞ −∞ F(y. x−y/c0 being the time of travel of sound waves from y to x.6.1.5) δ t −τ − x − y c0 x−y c0 1 i. in accordance with the radiation condition. The outgoing wave solution for each constituent source of strength F(y. and Quadrupoles A volume point source q(t)δ(x) of the type considered in Section 1. t − −∞ x − y d 3 y. y. (1.6) is called a retarded potential.7 Monopoles.7 Monopoles. Dipoles. because F(x. and Quadrupoles 13 This represents an impulsive. so that by adding up these individual contributions we obtain p(x. t − τ ) d 3 y dτ ∞ −∞ ∞ (1.6. (1.1). y. τ ) d 3 y dτ (1. This source distribution can be regarded as a distribution of impulsive point sources of the type on the right of Equation (1.6. t) = 4π F y. 1. τ )G(x.6. For a compressible .4 as a model for a pulsating sphere is also called a monopole point source. t − τ ) d 3 y dτ. spherically symmetric wave expanding from the source at y at the speed of sound.e. t) = = ∞ −∞ F(y.. it represents the pressure at position x and time t as a linear superposition of contributions from sources at positions y. τ )G(x. τ )δ(x − y)δ(t − τ ) d 3 y dτ is F(y.3) where the generalized source F(x. p(x.6. Green’s function is the fundamental building block for forming solutions of the inhomogeneous wave equation (1.6) The integral formula (1.6. t). Let us write this equation in the form 1 ∂2 − ∇2 2 c0 ∂t 2 p = F(x.
such as j in this equation. Replace p by ϕ in (1.6. As explained in the Preface. and note that x−y x−y ∂ δ t − τ − c0 ∂ δ t − τ − c0 =− .7.6) of Equation (1. ( f j (τ )δ(y)) ∂yj x − y Integrate by parts with respect to each y j (recalling that δ(y) = 0 at y j = ±∞).6.6). 3. (1.3) is called a point dipole (located at the origin).3) of the form F(x.6) and set F(y. any effects associated with changes in the motion of the sphere (i. ∂yj x − y ∂x j x − y . implies a summation over j = 1.3) for a pulsating sphere or volume point source in an incompressible ﬂuid by the dependence on the retarded time t − cr0 . 2 c0 ∂t 2 (1. but it is easier to use (1. Equation (1.3) for the acoustic pressure. τ ) = −q(τ )δ(y). The sound produced by the dipole can be calculated from (1. Then a source on the right of the acoustic pressure equation (1.6. in the value of the volume outﬂow rate q(t)) are now communicated to a ﬂuid element at distance r after an appropriate time delay r/c0 required for sound to travel outward from the source.5): p(x. This is physically more realistic. ϕ(x.1) The solution can be written down by analogy with the Solution (1. Then. t) = 1 4π ∞ −∞ δ t − τ − x−y 3 ∂ c0 d y dτ.7. 1.7) shows that the point dipole is equivalent to a force distribution F(t) = −f(t)δ(x) per unit volume applied to the ﬂuid at the origin.6.6..1 The Point Dipole Let f = f(t) be a timedependent vector.7.14 1 Introduction medium the corresponding velocity potential it produces is the solution of the equation 1 ∂2 − ∇ 2 ϕ = −q(t)δ(x). t) = div(f(t)δ(x)) ≡ ∂ ( f j (t)δ(x)) ∂x j (1.4.7.6.3.e.2) This differs from the corresponding solution (1. t) = −q t − x c0 4π x ≡ −q t − 4πr r c0 . a repeated italic subscript. 2.
. (1. (1. and Quadrupoles Then.7. the acoustic pressure becomes p(x. ∂x where the dot denotes differentiation with respect to time. The relation p = −ρ0 ∂ϕ/∂t implies that the equivalent dipole source in the pressure equation (1.2 Quadrupoles A source distribution involving two space derivatives is equivalent to a combination of four monopole sources (whose net volume source strength is zero). and the sources are distance apart.6.7 Monopoles.7) or (1.7. For example. Dipoles. p(x. 1 p(x. t) = div f(x. 1.3. t − x−y c0 x − y d 3 y.1. if n is parallel to the x axis. 4π x .7. t) on the right of Equation (1.3) is −ρ0 ∂ ˙ ( q(t)δ(x)). t) = 4π = Thus. t) = 1 ∂ 4π ∂ x j ∞ −∞ f j y. ∂x This is a ﬂuid volume dipole.6. the two monopoles would be q(t)δ x − 2 δ(y)δ(z) − q(t)δ x + 2 δ(y)δ(z) ≈ − q(t)δ (x)δ(y)δ(z) ≡ − ∂ ( q(t)δ(x)).4) The same procedure shows that for a distributed dipole source of the type F(x.3). t) = ∂ ∂x j fj t − x c0 ∞ −∞ 15 ∂ f j (τ )δ(y) ∂x j ∞ −∞ δ t −τ − x−y c0 x − y δ t −τ − x−y c0 d 3 y dτ 1 ∂ 4π ∂ x j f j (τ )δ(y) x − y d 3 y dτ.5) A point dipole at the origin orientated in the direction of a unit vector n is entirely equivalent to two point monopoles of equal but opposite strengths placed a short distance apart (much smaller than the acoustic wavelength) on opposite sides of the origin on a line through the origin parallel to n.
t) = ∂2 1 4π ∂ xi ∂ x j ∞ −∞ Ti j (y.7.7) 1. The argument above leading to Expression (1.7.3). 1. A general quadrupole is a source of the form F(x. we shall prove that the motion induced in an ideal ﬂuid when the sphere is small is equivalent to that produced by a point volume dipole of strength 2πa 3 U (t) at its center directed along the x1 axis. In Section 3. t − x − y/c0 ) 3 d y. determined Fig. Take the coordinate origin at the mean position of the center.7.5.16 1 Introduction and is called a quadrupole.6.3 Vibrating Sphere Let a rigid sphere of radius a execute small amplitude oscillations at speed U (t) in the x1 direction (Fig. 1. x − y (1.7. .1a).5) can be applied twice to show that the corresponding acoustic pressure is given by p(x.7.1. t) = ∂ 2 Ti j (x.7. t) ∂ xi ∂ x j (1.6) in Equation (1.
in which the sphere is replaced by a point dipole at its center. In this limit the solution reduces entirely to the nearﬁeld term. x = ∂x j x 3 ∂ 2πa U t − ∂ x1 4π x x c0 17 (1. The nearﬁeld term is dominant at sufﬁciently small distances r from the origin such that 1 r 1 1 ∂U f ∼ . 2 ∂ x1 c0 ∂t 2 By analogy with (1.7.10) Applying this formula for j = 1. c0 U ∂t c0 where f is the characteristic frequency of the oscillations of the sphere.7. which is also called the hydrodynamic near ﬁeld.8) . It consists of propagating sound waves. The motion becomes incompressible when c0 → ∞. and takes over from the near ﬁeld when r λ. the nearﬁeld term is dominant when r λ. (1. The far ﬁeld is the acoustic region that only exists when the ﬂuid is compressible.8). we have ϕ(x. But. sound of frequency f travels a distance c0 / f = λ ≡ one acoustic wavelength in one period of oscillation 1/ f . and Quadrupoles by the solution of 1 ∂2 ∂ − ∇2 ϕ = (2πa 3 U (t)δ(x)). we ﬁnd (putting r = x and x1 = r cos θ) ϕ=− a 3 cos θ r U t− 2r 2 c0 near ﬁeld − a 3 cos θ ∂U r t− 2c0r ∂t c0 far ﬁeld . t) = Now. The analytical model (1. xj ∂ .3) and (1. Hence.9) (1. There is an intermediate zone where r ∼ λ in which the solution is in a state of transition from the near to the far ﬁeld.4).7. Dipoles.1. carrying energy away from the sphere. involves the implicit assumption that the motion . it decreases in amplitude like 1/r 2 as r → ∞.7.7.7 Monopoles.7.
1b. t − r c0 ≡ p .8. We therefore conclude from this and (1. From the radial component of the linearized momentum equation 1 ∂p ∂vr =− ∂t ρ0 ∂r ≡ 1 r2 θ. a body is said to be acoustically compact when its characteristic dimension is small compared to the wavelengths of the sound waves it is producing or with which it interacts. For the dipole it has the ﬁgure of eight pattern illustrated in Fig. The intensity of the sound generated by the sphere in the far ﬁeld is proportional to ϕ 2 : ϕ2 → ∂U a6 2 2 ∂t 4c0 r t− r c0 2 cos2 θ. r → ∞. vφ .8 Acoustic Energy Flux At large distances r from a source region we generally have p(x. and therefore vr ∼ 1 c0 r θ. t − r r c0 . The dependence on θ determines the directivity of the sound. decrease faster than 1/r as r → ∞. 1. the diameter of the sphere is much smaller than the acoustic wavelength.8. there are radiation nulls at θ = π (the 2 curve should be imagined to be rotated about the x1 axis). It follows from what we have just said that a λ. t − r c0 + 1 ∂ c0r ∂t θ. . φ. that is. say.8. ρ0 c0 (1.7. In general.3) By considering the θ and φ components of the momentum equation we can show that the corresponding velocity components vθ . π). φ. φ. t − r .2) The ﬁrst term in the second line can be neglected when r → ∞. and θ and φ are polar angles determining the directivity of the sound.8.3) that the acoustic particle velocity is normal to the acoustic wavefronts (the spherical surfaces r = c0 t). (1.1) where the function depends on the nature of the source distribution. In other words. sound consists of longitudinal waves in which the ﬂuid particles oscillate backwards and forwards along the local direction of propagation of the sound. φ. with peaks in directions parallel to the dipole axis (θ = 0. 1.18 1 Introduction close to the sphere is essentially the same as if the ﬂuid is incompressible. c0 (1. t) ∼ ρ0 θ.
1.8 Acoustic Energy Flux The acoustic power from the formula
19
radiated by a source distribution can be computed p2 d S, ρ0 c0
=
S
pvr d S =
S
(1.8.4)
where the integration is over the surface S of a large sphere of radius r centered on the source region. Because the surface area = 4πr 2 , we only need to know the pressure and velocity correct to order 1/r on S in order to evaluate the integral. Smaller contributions (such as that determined by the ﬁrst term in the second line of (1.8.2)) decrease too fast as r increases to supply a ﬁnite contribution to the integral as r → ∞. In acoustic problems we are therefore usually satisﬁed if we can calculate the pressure and velocity in the acoustic far ﬁeld correct to order 1/r ; this will always enable us to determine the radiated sound power. The formula vr = p/ρ0 c0 is applicable at large distances from the sources, where the wavefronts can be regarded as locally plane, but it is true identically for plane sound waves. In the latter case, and for spherical waves on the surface of the large sphere of Fig. 1.8.1, the quantity I = pvr = p2 ρ0 c0 (1.8.5)
is called the acoustic intensity. It is the rate of transmission of acoustic energy per unit area of wavefront.
Fig. 1.8.1.
20
1 Introduction
Fig. 1.9.1.
1.9 Calculation of the Acoustic Far Field We now discuss the approximations necessary to evaluate the sound in the far ﬁeld from the retarded potential representation: p(x, t) = 1 4π
∞ −∞
F y, t −
x−y c0
x − y
d 3 y.
(1.9.1)
We assume that F(x, t) = 0 only within a ﬁnite source region (Fig. 1.9.1), and take the coordinate origin O within the region. When x → ∞ and y lies within the source region (so that x y) 2x · y y2 + 2 x − y ≡ (x − 2x · y + y ) = x 1 − x2 x
2 2
1 2 1 2
≈ x 1 − Then,
x·y +O x2
y2 x2
x − y ≈ x − Also,
x·y x
when
y x
1.
(1.9.2)
1 1 ≈ x − y x − Therefore,
x·y x
≈
1 x·y 1+ 2 x x y x 1. (1.9.3)
1 x·y 1 ≈ + 3 x − y x x
when
The Approximation (1.9.3) shows that, in order to obtain the farﬁeld approximation of the Solution (1.9.1) that behaves like 1/r = 1/x as x → ∞, it is sufﬁcient to replace x − y in the denominator of the integrand by x. However, in the argument of the source strength F it is important to retain possible phase
1.9 Calculation of the Acoustic Far Field
21
differences between the sound waves generated by components of the source distribution at different locations y; we therefore replace x − y in the retarded time by the righthand side of (1.9.2). Hence, p(x, t) ≈ 1 4πx
∞ −∞
F y, t −
x x·y + d 3 y, c0 c0 x
x → ∞.
(1.9.4)
This is called the Fraunhofer approximation. The source region may extend over many characteristic acoustic wavelengths of the sound. By retaining the contribution x · y/c0 x to the retarded time we ensure that any interference between waves generated at different positions within the source region is correctly described by the farﬁeld approximation. In Fig. 1.9.1 the acoustic travel time from a source point y to the far ﬁeld point x is equal to that from the point labelled A to x when x → ∞. The travel time over the distance O A is just x · y/c0 x, so that x/c0 − x · y/c0 x gives the correct value of the retarded time when x → ∞.
1.9.1 Dipole Source Distributions By applying the farﬁeld formula (1.9.4) to a dipole source F(x, t) = div f(x, t) we obtain (from (1.7.5)) p(x, t) ≈ ≈ 1 ∂ 1 4π ∂ x j x ∂ 1 4πx ∂ x j
∞ −∞ ∞ −∞
f j y, t −
x x·y d 3y + c0 c0 x (1.9.5)
f j y, t −
x·y x + d 3 y, x → ∞, c0 c0 x
because the differential operator ∂/∂ x j need not be applied to 1/x as this would give a contribution decreasing like 1/r 2 at large distances from the dipole. However, it is useful to make a further transformation that replaces ∂/∂ x j by the time derivative ∂/∂t, which is usually more easily estimated in applications. To do this, we observe that ∂fj ∂x j = = y, t − x·y x + c0 c0 x ∂ x·y x + t− ∂x j c0 c0 x − yj (x · y)x j xj + − c0 x c0 x c0 x3 as x → ∞.
∂fj x·y x + y, t − ∂t c0 c0 x ∂fj x·y x + y, t − ∂t c0 c0 x
≈−
xj ∂fj x·y x + y, t − c0 x ∂t c0 c0 x
22
1 Introduction Hence, the far ﬁeld of a distribution of dipoles F(x, t) = div f(x, t) is given by p(x, t) = −x j ∂ 4π c0 x2 ∂t
∞ −∞
f j y, t −
x x·y d 3 y. + c0 c0 x
(1.9.6)
Note that xj xj 1 , = x2 x x where x j /x is the jth component of the unit vector x/x. Thus, the additional factor of x j /x in (1.9.6) does not change the rate of decay of the sound with distance from the source (which is still like 1/r ), but it does have an inﬂuence on the acoustic directivity. A comparison of (1.9.5) and (1.9.6) leads to the following rule for interchanging space and time derivatives in the acoustic far ﬁeld: 1 xj ∂ ∂ . ←→ − ∂x j c0 x ∂t 1.9.2 Quadrupole Source Distributions For the Quadrupole (1.7.6) F(x, t) = and p(x, t) = ∂2 1 4π ∂ xi ∂ x j
∞ −∞
(1.9.7)
∂ 2 Ti j (x, t), ∂ xi ∂ x j Ti j (y, t − x − y/c0 ) 3 d y. x − y
By applying (1.9.4) and the rule (1.9.7), we ﬁnd that the acoustic far ﬁeld is given by p(x, t) ≈ xi x j ∂ 2 2 4π c0 x3 ∂t 2
∞ −∞
Ti j y, t −
x·y x + d 3 y, x → ∞. c0 c0 x (1.9.8)
1.9.3 Example For the (1, 2) point quadrupole F(x, t) = ∂2 (T (t)δ(x)) ∂ x1 ∂ x2
2. x → ∞. c0 The directivity of the sound (∝ p 2 ) is therefore represented by sin2 2θ cos2 φ. π).9. A plane sound wave propagating parallel to the x axis satisﬁes the equation 1 ∂2 ∂2 − 2 ϕ = 0. Its shape is plotted in Fig. x2 plane (φ = 0. such that x1 = r cos θ . t) ≈ x1 x2 ∂ 2 T 2 4π c0 x3 ∂t 2 t− x . 2 ∂t 2 ∂x c0 . Problems 1 1. 1. c0 If we use spherical polar coordinates. we can write the pressure in the form p(x. Equation (1.2 for radiation in the x1 .8) shows that in the acoustic far ﬁeld p(x. The fourlobe cloverleaf pattern is characteristic of a quadrupole Ti j for which i = j. x2 = r sin θ cos φ.9. x → ∞.9. 1. x3 = r sin θ sin φ. t) ≈ sin 2θ cos φ ∂ 2 T 2 8π c0 x ∂t 2 t− x .Problems 1 23 Fig.
where U0 = constant. ρ. . 2 c0 T = p . subsonic velocity U.24 with general solution ϕ= 1 Introduction t− x c0 + t+ x . 4π R(1 − M cos ) M= U . c0 where and are arbitrary functions respectively representing waves propagating in the positive and negative x directions. As for Problem 2. U0 = constant. when the sphere executes small amplitude radial oscillations at normal velocity vn = U0 cos(ωt). density. The velocity potential ϕ(x. 3. and T are respectively the acoustic pressure. t) = −q0 (t − R/c0 ) . 2.8. and is the angle between U and the direction of propagation of this sound. A volume point source of strength q0 (t) translates at constant. 2 c0 ∂t 2 Show that ϕ(x. c0 where R is the distance of the reception point x from the source position at the time of emission of the sound received at x at time t.4) radiated by an acoustically compact sphere of radius R executing small amplitude translational oscillations of frequency ω and velocity U (t) = U0 cos(ωt). 4. and cp is the speciﬁc heat at constant pressure. Calculate the acoustic power (1. ρ0 c0 ρ= p . t) of the radiated sound is determined by the solution of 1 ∂2 − ∇ 2 ϕ = −q0 (t)δ(x − Ut). ρ0 cp where v is the acoustic particle velocity. and temperature variations. Show that for a wave propagating in the positive x direction in an ideal gas v= p . p.
in which the sound is imagined to be generated by a ﬁnite region of rotational ﬂow in an unbounded ﬂuid. The turbulence is usually produced by ﬂuid motion over a solid boundary or by ﬂow instability. where the mean pressure. inhomogeneous wave equation whose source terms are important only within the turbulent region. and sound speed are respectively equal to p0 . once generated.1a.1. 2. 2. a reasonable approximation if the Mach number M is small. This avoids complications caused by the presence of the nozzle. The properties of the unsteady ﬂow in the source region may then be determined by neglecting the production and propagation of the sound. and there are many important ﬂows where the hypothesis is obviously correct. of the sound produced by a turbulent nozzle ﬂow.1 The Acoustic Analogy The sound generated by turbulence in an unbounded ﬂuid is usually called aerodynamic sound. its backreaction on the main ﬂow can usually be ignored. The ﬂuid is assumed to be at rest at inﬁnity. density.2 Lighthill’s Theory 2. 25 . illustrated in Fig. Lighthill compared the equations for the production of acoustic density ﬂuctuations in the real ﬂow with those in an ideal linear acoustic medium that coincides with the real ﬂuid at large distances from the sources. He argued that sound is a very small component of the whole motion and that. and the acoustic radiation is a very small byproduct of the motion.1. However. Lighthill was initially interested in solving the problem. and where the theory leads to unambiguous predictions of the sound. ρ0 . c0 .1b. his original theory actually applies to the simpler situation shown in Fig. Most unsteady ﬂows of technological interest are of high Reynolds number and turbulent. Lighthill (1952) transformed the Navier–Stokes and continuity equations to form an exact.
∂t ∂x j ∂ xi ∂x j ∂x j (2. 2.3) is cast in the form ρ ∂σi j ∂vi ∂vi ∂p ∂ + ρv j =− + ≡− ( pδi j − σi j ). and the ith component of the momentum equation (1. and σi j is the viscous stress tensor deﬁned (for a Stokesian ﬂuid) by σi j = 2η ei j − 1 ekk δi j .3) (2. body forces are neglected.26 2 Lighthill’s Theory Fig. and 0 for i = j).1.1. 3 where ei j = ∂v j 1 ∂vi + 2 ∂x j ∂ xi (2. To do this.1.2) .1) δi j is the Kronecker delta (= 1 for i = j.1.2.1.
2.1. linear acoustic medium. Next multiply the continuity equation (1.1. and the constant pressure p0 is inserted for convenience.1) is written in the slightly modiﬁed form ∂ ∂(ρvi ) (ρ − ρ0 ) + = 0.4) is called the momentum ﬂux tensor.6) and the momentum equation then reduces to ∂(ρvi ) ∂ 2 + c (ρ − ρ0 ) = 0. ∂t ∂x j By adding this to Equation (2.1) by vi : vi ∂(ρv j ) ∂ρ + vi = 0.1.8) we can eliminate the momentum density ρvi between (2.1. (2. the momentum ﬂux tensor contains only the pressure 2 πi j → πi0j = ( p − p0 )δi j ≡ c0 (ρ − ρ0 )δi j .1).1. ∂t ∂ xi (2. stationary acoustic . ∂t ∂ xi 0 (2.2.7) If the continuity equation (1.1. In an ideal. ∂t ∂x j where πi j = ρvi v j + ( p − p0 )δi j − σi j . the unique solution of this equation that satisﬁes the radiation condition of outgoing wave behavior is simply ρ − ρ0 = 0.1.7) and (2. and there are no externally applied forces or moving boundaries. we obtain the Reynolds form of the momentum equation ∂πi j ∂(ρvi ) =− . (2. It can now be asserted that the sound generated by the turbulence in the real ﬂuid is exactly equivalent to that produced in the ideal.8) to obtain the equation of linear acoustics satisﬁed by the perturbation density ρ − ρ0 1 ∂2 − ∇2 2 c0 ∂t 2 2 c0 (ρ − ρ0 ) = 0.9) Because the turbulence is neglected in this approximation.2. (2.1.1 The Acoustic Analogy 27 is the rate of strain tensor.5) (2.1.
2.9).10) of Ti j . (2.1.1.12) This is the exact. The problem of calculating the turbulence generated sound is therefore equivalent to solving this equation for the radiation into a stationary. In the deﬁnition (2.1.1.10) where Ti j is called the Lighthill stress tensor. we can rewrite (2.1.1. This is Lighthill’s acoustic analogy. The second term represents the excess of momentum transfer by the pressure over that in the ideal (linear) ﬂuid of density ρ0 and sound speed c0 .8) (the same procedure used above for the linear problem). This is produced by wave amplitude nonlinearity.1. ∂ xi ∂ x j (2. the term ρvi v j is called the Reynolds stress. ∂t ∂ xi 0 ∂x j (2. and that in a typical low Mach number ﬂow only a tiny fraction of the available ﬂow energy is converted into sound.11) By eliminating the momentum density ρvi between this and the continuity equation (2. For the simpliﬁed problem of Fig.1. and by mean density variations in the source ﬂow.2) that freeﬁeld turbulence is an extremely weak sound source. Indeed. The quadrupole character of the turbulence sources is one of the most important conclusions of Lighthill’s theory. it implies (see Section 2.4) as the momentum equation for an ideal.1.9) in turbulencefree regions) forced by the stress distribution Ti j = πi j − πi0j 2 = ρvi v j + ( p − p0 ) − c0 (ρ − ρ0 ) δi j − σi j .1b it is a nonlinear quantity that can be neglected except where the motion is turbulent. The viscous . =− ∂t ∂x j ∂x j or ∂ Ti j ∂ 2 ∂(ρvi ) + c (ρ − ρ0 ) = − . nonlinear counterpart of (2. ideal ﬂuid produced by a distribution of quadrupole sources whose strength per unit volume is the Lighthill stress tensor Ti j . we obtain Lighthill’s equation 1 ∂2 − ∇2 2 c0 ∂t 2 2 c0 (ρ − ρ0 ) = ∂ 2 Ti j . stationary acoustic medium of mean density ρ0 and sound speed c0 subject to the externally applied stress Ti j 0 ∂(ρvi ) ∂πi j ∂ + πi j − πi0j .28 2 Lighthill’s Theory medium (which is governed by (2.
The inﬂuence of acoustic nonlinearity and of thermoviscous dissipation is usually sufﬁciently weak to be neglected within the source region.2. t) = ∂2 1 4π ∂ xi ∂ x j ∞ −∞ Ti j (y. t) replaced by c0 (ρ − ρ0 ) 2 c0 (ρ − ρ0 )(x. This is because the terms in the deﬁnition (2. 1974). the convection of sound waves by the turbulent velocity. This means that in practice it must be possible to derive a good approximation for Ti j by taking the source ﬂow to be effectively incompressible. Convection and refraction of sound within and near the source ﬂow can be important. although they may affect propagation to a distant observer.12) with outgoing wave be2 havior is given by (1. and properly accounts for the attenuation of the sound.2. for example in the presence of a mean shear layer (when the Reynolds stress will include terms like ρUi u j . high Reynolds number source ﬂow are then of . x − y (2.1.7) with p(x.1) it is usually necessary to suppose that all of these acoustic effects in the source ﬂow (which really depend on ﬂuid compressibility) are in some sense negligible.2. to predict the radiated sound from Lighthill’s equation (2.2 Lighthill’s v 8 Law The formal solution of Lighthill’s equation (2.10) of Ti j not only account for the generation of sound. 2. integral equation representation of Equation (2.1) This is strictly an alternative.1. where U and u respectively denote the mean and ﬂuctuating components of v).7. it provides a useful prediction of the sound only when Ti j is known or has been determined by some other means. This is often possible when the characteristic Mach number M ∼ v/c0 is small (speciﬁcally. 1). Thus. and the viscous attenuation of the radiating sound is usually ignored. t − x − y/c0 ) 3 d y. such effects are described by the presence of unsteady linear terms in Ti j (Ffowcs Williams. and attenuation due to thermal and viscous actions. but also govern acoustic selfmodulation caused by acoustic nonlinearity.1.2 Lighthill’s v 8 Law 29 stress tensor σi j is linear in the perturbation quantities. and σi j can be neglected. Consider the particular but important case in which the mean density and sound speed are uniform throughout the ﬂuid. refraction caused by sound speed variations.12). and when the wavelength of the sound is much larger than the when M 2 size of the source region. or when there are large variations in the mean thermodynamic properties of the medium within the source region. The variations in the density ρ within a low Mach number. in most applications the Reynolds number in the source region is very large.
2) 2 In the acoustic region outside the source ﬂow c0 (ρ − ρ0 ) = p − p0 . where V0 is the volume occupied by the turbulence (Fig. t) is the local speed of sound in the source region. If the irrelevant constant pressure p0 is suppressed. 1967). such as the width of a jet mixing layer. t − x − y/c0 ) 3 d y 4π x − y ρ0 vi v j y. This means that when the integral in (2.2. we introduce a characteristic velocity v and length scale (of the energycontaining eddies) of the turbulence sources. it may also 2 be shown that c0 /c2 = 1 + O(M 2 ). t − x x·y d 3 y. Fluctuations in vi v j occurring in different regions of the turbulent ﬂow separated by distances >O( ) will tend to be statistically independent.1) of Lighthill’s equation therefore becomes p(x.3) xi x j ∂ 2 2 4πc0 x3 ∂t 2 where in the second line we have used the formula (1. and the sound may be considered to be generated by a collection of V0 / 3 independent eddies.9. Hence. provided that M 2 1. we arrive at the important conclusion that (because M = v/c0 the turbulence eddies are each acoustically compact. Quantitative predictions can be made from this formula provided the behavior of the Reynolds stress ρ0 vi v j is known.3) is conﬁned to a single eddy. . if the coordinate origin is temporarily placed at O within the eddy.8) for the far ﬁeld of a quadrupole distribution.2. Similarly.30 2 Lighthill’s Theory order ρ0 M 2 (Batchelor. we can set x · y/c0 x = 0 in the integration over that eddy. The value of the integral over the eddy then may be estimated to be of order ρ0 v 2 3 . if c(x.1). so that the wavelength (c0 / f ) of the sound ∼ /M 1). if viscous dissipation is neglected we make the approximation Ti j ≈ ρ0 vi v j . (2. the Solution (2.2. the retarded time variations x · y/c0 x across that eddy can be neglected. t) ≈ ≈ ∂2 ∂ xi ∂ x j ρ0 vi v j (y.2. The value of depends on the mechanism responsible for turbulence production. Hence. The characteristic frequency of the turbulent ﬂuctuations f ∼ v/ . so that 2 2 p − p0 − c0 (ρ − ρ0 ) ≈ ( p − p0 ) 1 − c0 c2 ∼ O(ρ0 v 2 M 2 ). To determine the order of magnitude of p. (2. + c0 c0 x x → ∞. Thus. 2. that is.2. ρvi v j = ρ0 (1 + O(M 2 ))vi v j ≈ ρ0 vi v j .
4) taken over a large sphere centered on the eddy. in order of magnitude. The order of magnitude of the time derivative for changes in the source region is ∂ v ∼ . .2. acoustic power radiated by one eddy ∼4π x2 p2 ∼ ρ0 c0 2 (2.2. containing V0 / 3 independent eddies. for one eddy. is q ρ0 v 8 = 5 c0 2 ρ0 v 3 M 5 .2.1.5) This is Lighthill’s ‘eighth power’ law. = 2 x c0 x (2.2.4) The acoustic power radiated by the eddy is determined by the surface integral (1.3) that. ≈ V0 3 × ( 2 ρ0 v 3 M 5 ) = v ρ0 v 2 M 5 V0 . The total power radiated from the whole of the turbulent region of volume V0 .2 Lighthill’s v 8 Law 31 Fig. Thus. the farﬁeld acoustic pressure satisﬁes p∼ ρ0 v 4 ρ0 v 2 M 2 . ∂t Therefore. it follows from (2.8. 2.2.
2.1) and consider the Heaviside unit function H( f ) = 1 0 for x in V.4.3. The procedure in such cases is to introduce a system of mathematical control surfaces that can be deformed to coincide with the surfaces of the different moving or stationary bodies. ﬂow control surfaces. conﬁrming Lighthill’s hypothesis that the ﬂow generated sound is an inﬁnitesimal byproduct of the turbulent motion.1 Volume and Surface Integrals Let V be the ﬂuid outside a closed control surface S (Fig. although for the moment we shall discuss only cases involving stationary bodies.2.1) deﬁned by the equation f (x) = 0.3 Curle’s Theory In most applications of Lighthill’s theory it is necessary to generalize the solution (2. for x within S. and the unsteady surface forces (dipoles) that arise are likely to make a signiﬁcant contribution to the production of sound. Before doing this we establish an integral transformation formula that is used repeatedly in problems of this kind. for x within S.32 2 Lighthill’s Theory Dimensional arguments and experiment indicate that the rate 0 . Therefore.3.2.). (2. Indeed. say. . at which energy must be supplied by the action of external forces to maintain the kinetic energy of a statistically steady turbulent ﬁeld occupying a volume V0 is given in order of magnitude by 0 ∼ v ρ0 v 2 V0 . turbulence is frequently generated in the boundary layers and wakes of ﬂow past such bodies (airfoils. 2. etc.01 for Mach numbers M < 0.1) to account for the presence of solid bodies in the ﬂow. where f (x) > 0 f (x) < 0 for x in V. the mechanical efﬁciency with which turbulence kinetic energy is converted into sound is q 0 ∼ M 5.6) This is smaller than about 0.3. 2. (2.
(x) d S j . Then. S where (∂ f /∂s⊥ ) S ≡ ∇ f  > 0 is evaluated on S.3.4) is nonzero only on S. 2.1) that just encloses S. (x) n d S ≡ S S (x)∇ H d 3 x = (x) ∂H 3 d x= ∂x j (x) dS.2) (2. Proof ∇ H ( f ) ≡ δ( f ) ∇ f (2.1. for an arbitrary function ∞ −∞ ∞ (x) deﬁned in V and on S.2.3.3. for small values of s⊥ .3) or −∞ (x) n j d S ≡ S where H ≡ H ( f ) and n is the unit normal on S directed into V. where s⊥ = 0 on S and s⊥ is measured parallel to n. The volume integral is therefore conﬁned to the region between the inner and outer faces of a shell of inﬁnitesimal thickness (between the broken line surfaces in Fig. . and in which the volume element is d 3 x = ds⊥ d S. f = ∂f ∂s⊥ s⊥ .3. where ∇ f is in the direction of n. Because f = 0 on S we can write. 2.3. S (2.3 Curle’s Theory 33 Fig.
1. deﬁned as above by an equation f (x) = 0 (Fig. and use the deﬁnition (2.10) of Ti j to obtain ∂ ∂ ∂H ∂ 2 (ρvi H ) + H c0 (ρ − ρ0 ) = − (H Ti j ) + (ρvi v j + pi j ) . 2.12) for the sound produced by turbulence in the vicinity of an arbitrary. ∞ −∞ δ(s⊥ ) . ∇ f  (x)∇ H d 3 x ≡ = ∞ −∞ (x)∇ f δ( f ) d 3 x = ∞ (x) −∞ ∇f δ(s⊥ ) ds⊥ d S ∇ f  (x)n dS. 2 Lighthill’s Theory δ( f ) = δ(∇ f s⊥ ) ≡ Hence.3.5) Fig. .3.2).1. or merely constitute a control surface used to isolate a ﬁxed region of space containing both solid bodies and ﬂuid or just ﬂuid.34 Therefore. ∂t ∂ xi ∂x j ∂x j (2. ﬁxed surface S. multiply the momentum equation (2.2 Curle’s Equation Curle (1955) has derived a formal solution (called Curle’s equation) of Lighthill’s equation (2. because n = S ∇f .1. 2.3.11) by H ≡ H ( f ).2.3. ∇ f  2. This surface may either enclose a solid body. To derive Curle’s equation.
They have the following interpretations: 1. Repeat this operation for the continuity equation (2. Because Curle’s form of Lighthill’s equation is valid throughout all space.6) is the compressive stress tensor.2.3. This is the differential form of Curle’s equation 1 ∂2 − ∇2 2 c0 ∂t 2 = 2 H c0 (ρ − ρ0 ) ∂ 2 (H Ti j ) ∂ ∂H − (ρvi v j + pi j ) ∂ xi ∂ x j ∂ xi ∂x j + ∂ ∂t ρv j ∂H . shows that Equations (2. possibly also solid bodies.3. ∂x j (2. the surface dipole and monopole sources then represent the inﬂuence of this region on the sound radiated in V . ∇ H = ∇ f δ( f ).5) and (2. ∂t ∂ xi ∂ xi (2.1.4). An analog of Lighthill’s equation (2. and may or may not contain turbulence.3. including the region enclosed by S where H ( f ) vanishes. 2. If S is the boundary of a solid body.6. The second and third terms on the righthand side respectively represent dipole and monopole sources distributed over S. whereas the monopole is responsible for the sound produced by volume pulsations (if any) of the body.3 Curle’s Theory where pi j = ( p − p0 )δi j − σi j 35 (2. the surface dipole represents the production of sound by the unsteady surface force that the body exerts on the exterior ﬂuid.7). If S is merely an artiﬁcial control surface it will enclose ﬂuid.3.3.3.7) The Formula (2.12) can now be obtained by eliminating Hρvi between (2.8): ∂ ∂H ∂ (H (ρ − ρ0 )) + (Hρvi ) = (ρvi ) .3.1.5) and (2.7) formally determine the momentum density ρvi and the density ﬂuctuation (ρ − ρ0 ) in the exterior region V (where H ( f ) ≡ 1) in terms of the Lighthill stresses Ti j in V and sources distributed over the control surface. in other words the aggregate effect of the dipole and monopole sources accounts for the presence of solid bodies and turbulence within S (when Ti j = 0 in S) and also for the interaction of sound generated outside S with the ﬂuid and solid bodies in S. the outgoing wave solution is found from the general solution (1.8) The equation is valid throughout all space.3.6) of the wave .
. 2. and we shall assume this to be the case in the following.3) this yields Curle’s equation 2 H c0 (ρ − ρ0 ) = ∂2 ∂ xi ∂ x j ∂ + ∂t [Ti j ] V ∂ d 3y − 4πx − y ∂ xi S [ρvi v j + pi j ] d S j (y) 4π x − y (2.5) for dipole sources.9) d S j (y) .7. σ η ν where ν = η/ρ0 is the kinematic viscosity.1) is estimated as in Section 2.4.1) We now use this solution to determine the order of magnitude of the sound generated by an acoustically compact body within a turbulent ﬂow.3.4. ( p − p0 ) ρ0 v v ∼ = . t − x − y/c0 ) implies evaluation at the retarded time. [ρv j ] 4π x − y S where the square bracket notation such as [Ti j ] ≡ Ti j (y. using the special form (1.3. note ﬁrst that for turbulence of velocity v and correlation scale . When account is taken of the transformation formula (2.2. as before.6. because H ( f ) ≡ 0 inside S. the orders of magnitude of the pressure and viscous components of the compressive stress tensor pi j = ( p − p0 )δi j − σi j are ( p − p0 ) ∼ ρ0 v 2 . This means that viscous contributions to the surface force can be neglected.4 Sound Produced by Turbulence Near a Compact Rigid Body When the surface S (in Fig. 4π x − y (2.9) reduces to 2 H c0 (ρ − ρ0 ) = ∂2 ∂ xi ∂ x j [Ti j ] V ∂ d 3y − 4π x − y ∂ xi S [ pi j ] d S j (y) .3. Note that. 2.36 2 Lighthill’s Theory equation (1. Compactness usually requires the Mach number M ∼ v/c0 1. v σ ∼η .3.3). To deal with the surface dipole. the sum of the three integrals on the righthand side must also vanish when the ﬁeld point x is within S.2) is rigid. The contribution from the quadrupole integral in (2. Curle’s equation (2. that is. The dimensionless ratio Re = v /ν is the Reynolds number and is always very large (∼104 or more) in turbulent ﬂow.
2) where F(t) is the unsteady force exerted on the ﬂuid by the body. and as M → 0 the acoustic power exceeds the quadrupole 1.2). corners.4. 2. t).3) The direct power radiated by quadrupoles occupying a volume V0 is q ∼ (V0 / )ρ0 v 3 M 5 . t − S x d Fi x xi d Si = .1). which determine the appropriate values of A and V0 / . 2. the same as in the absence of the body (see Section 2.9. say.2.6). noncompact bodies when turbulence interacts with compact structural elements. such as edges. Let us apply the rigid surface form (2. ρ0 c0 (2.4) produced by a quadrupole in V of length scale . c0 x x which exceeds by an order of magnitude (1/M 1) the sound pressure (2. and neglecting retarded time variations x · y/c0 x because S is compact. t− c0 4π c0 x2 dt c0 x → ∞. and protuberances.2. At high Reynolds numbers and at x in the . and H ( f ) = 1. Thus.5 Radiation from a Noncompact Surface 37 2 In the far ﬁeld the pressure p(x. and the total power radiated by the dipoles is d ∼ 4π x2 × 2 pd ∼ Aρ0 v 3 M 3 . t) = c0 (ρ − ρ0 )(x. (2. plane rigid wall at x2 = 0 (Fig. Precisely how small M should be for this to power by a factor ∼1/M 2 be true depends on the details of the ﬂow. If A is the total surface area wetted by the turbulent ﬂow.4. there are A/ 2 independently radiating surface elements. The contribution to pd from a surface element of diameter within which the turbulence surface pressure ﬂuctuations are correlated is evidently of order 1 v × (ρ0 v 2 2 ) = ρ0 v 2 M. can be written pd ≈ ∂ xi 4πc0 x2 ∂t ( p − p0 ) y. The sound produced by the turbulence near S is therefore dominated by the dipole when M is small.5 Radiation from a Noncompact Surface Consider a compact turbulent eddy in x2 > 0 adjacent to an inﬁnite.1) of Curle’s equation to calculate the radiation.4. the dipole sound pressure pd .5. applying the farﬁeld dipole approximation (1. This increase in acoustic efﬁciency brought about by surface dipoles on an acoustically compact body occurs also for arbitrary.
t) ≈ xi x j ∂ 2 2 4πc0 x3 ∂t 2 + x2 ∂ 4πc0 x2 ∂t Ti j y. t − x d 3y c0 x x1 y1 + x3 y3 + c0 c0 x x → ∞. and therefore involves an unknown and possibly important contribution from the acoustic pressure that we are trying to calculate! The difﬁculty was resolved by Powell (1960) by the ingenious device of applying Curle’s solution .5. dy1 dy3 . 2 acoustic far ﬁeld (where p(x. (2. 2. the acoustic component of the pressure.1. t − Retarded time variations have been neglected in the integral over the volume of the compact turbulent eddy. as opposed to the near ﬁeld hydrodynamic pressure.38 2 Lighthill’s Theory Fig. The value of the surface integral cannot be estimated by a naive orderofmagnitude calculation of the kind performed in Section 2. decaying only very slowly like 1/x). extends out to inﬁnity on the wall.5.4 for a compact body. t) ≡ c0 (ρ − ρ0 ) and H ( f ) ≡ H (x2 ) = 1) we ﬁnd p(x. We have not done this in the surface pressure integral. because for an inﬁnite plane wall the domain of integration includes the acoustic region.1) y2 =0 ( p − p0 ) y. because this tends to extend over a larger region than the Reynolds stress ﬂuctuations responsible for it (indeed.
2). This is a consequence of the Kraichnan–Phillips theorem (see Howe.4) y2 =0 Thus. the apparently strong contribution from the surface pressure dipoles actually integrates to a term of quadrupole strength. t) dy1 dy3 ≡ 0.3) c0 ρ0 vi v j y. −x2 . Show that the acoustic efﬁciency of a compact sphere of radius R executing small amplitude translational oscillations at velocity U = U0 sin(ωt) is a 0 ∼ ωR c0 3 . t − Therefore. the surface contribution will be comparable to that from the turbulence quadrupoles whenever the characteristic wavelength of the sound is smaller than the radius of curvature of the surface.1) and (2. At the image point H ( f ) = H (x2 ) ≡ 0.4. x3 ) in the wall of the far ﬁeld observation point x. according to which the net unsteady component of the normal force between an inﬁnite plane wall and an incompressible ﬂuid must vanish identically ( p − p0 )(y.2) y2 =0 ( p − p0 ) y.1).5.5. extreme care must be exercised when using Curle’s equation to estimate the sound produced by turbulence interacting with large surfaces. x → ∞. t − x d 3y c0 x x1 y1 + x3 y3 + c0 c0 x dy1 dy3 .5. t − x d 3y c0 x d 3 y. t) ≈ ≈ ¯ ¯ (xi x j + x i x j ) ∂ 2 2 4πc0 x3 ∂t 2 ¯ ¯ (xi x j + x i x j ) ∂ 2 2 4πc0 x3 ∂t 2 Ti j y. . As a general rule.5. (2.1) at the image x = (x1 . 1998a). t − x → ∞.5. which is now seen to exactly represent the quadrupole sound generated by a system of image quadrupoles in the wall! Adding (2. (2. Problems 2 1. The surface integral term in this formula is equal but opposite in sign to that in the original solution (2.5.Problems 2 39 ¯ (2. and therefore 0≈ ¯ ¯ xi x j ∂2 2 4πc0 x3 ∂t 2 − ∂ x2 4πc0 x2 ∂t Ti j y. (2. we ﬁnd p(x.
Show that Powell’s solution (2. 2.8) determined by the modiﬁed Green’s function G(x. −x2 .3.40 where a 2 Lighthill’s Theory = π ω R3 2 ρ0 U0 6 ωR c0 3 . Explain the signiﬁcance of averaging only over 0 < ωt < π/2. Derive an approximate formula for the farﬁeld acoustic pressure radiated by the body when U contains a small amplitude. 4. timeharmonic component such that U = U0 + u cos(ωt). t − τ ) = 1 x − y δ t −τ − 4π x − y c0 + 1 ¯ − y x . x3 ). and A is the projected crosssectional area of the body in the ﬂow direction. δ t −τ − 4π ¯ − y x c0 ¯ where x = (x1 .3) for the sound generated by turbulence adjacent to a rigid plane wall is identical with the solution of Curle’s differential equation (2. y. y. 0 = 2ω R 3 2 ρ0 U0 3 are respectively the average acoustic and hydrodynamic powers fed into the ﬂuid during the quarter cycle 0 < ωt < π/2. G(x. where U0 and u are constant and u U0 . where CD 2 is the drag coefﬁcient (which may be regarded as constant). t − τ ) is the solution of what problem of linear acoustics? . The wake behind a bluff body ﬁxed in a nominally steady.5. What is the efﬁciency in Problem 1 when the sphere pulsates with small amplitude normal velocity vn = U0 sin(ωt)? 3. low Mach number ﬂow at speed U produces a drag force equal to CD A 1 ρ0 U 2 .
2) where G(x. y. t) 2 c0 ∂t 2 (3. This equation determines ϕ in terms of a speciﬁed source distribution q(x. y.3 The Compact Green’s Function 3. t − τ ) is the outgoing wave solution of 1 ∂2 − ∇ 2 G = δ(x−y)δ(t −τ ).3.9) 1 ∂2 − ∇ 2 ϕ = −q(x. t). t − τ ) is the free space Green’s function G(x.1 The Inﬂuence of Solid Boundaries Let us return to the general problem of linear acoustics.3.1.6 enable us to represent ϕ in the form ϕ(x.8). where G = 0 2 c0 ∂t 2 for t < τ. we shall frame the present discussion in terms of Equation (1. To ﬁx ideas.3. it was found that the presence of a solid boundary S in the vicinity of the turbulence quadrupole sources Ti j resulted in the appearance of additional dipole and monopole sources distributed over S. y.1. Curle’s solution (2. 4π x − y c0 (3. represented by the second and third terms on the righthand side of Equation (2.1.9) of this 41 . (3. (3.1. t − τ ) = x − y 1 δ t −τ − . but our conclusions will be applicable quite generally. t) = ∞ −∞ −q(y.1) for the velocity potential.3) that is. t − τ ) d 3 y dτ. G(x. τ )G(x.4) In our discussion of Curle’s extension of Lighthill’s theory in Chapter 2. y. In the absence of solid boundaries (in free space) the results of Section 1.
6).42 3 The Compact Green’s Function Fig.1.1. y. 3. To do this.1) using Green’s function (3. and halfplanes).1.2) in terms of the modiﬁed Green’s function. t) is large compared to one or more principal dimensions of the solid body S.1. A similar representation involving surface distributions of dipoles and monopoles is obtained for ϕ when we attempt to solve Equation (3.3) in the situation illustrated in Fig. exact analytical representations are known only for solid bodies of very simple shapes (such as spheres. inasmuch that no surface integrals occur in the ﬁnal solution. obtained by use of the free space Green’s function (3.1. The solution ϕ of (3. Although it is always possible in principle. we must ﬁnd a solution of Green’s function equation (3.6. This is called the compact Green’s function. 3. However. equation was derived by using the retarded potential formula (1.3).1. it turns out that a relatively simple and general approximate formula can be found for the modiﬁed Green’s function for those problems where it is known that the typical wavelength of the sound produced by the source distribution q(x.4) that satisﬁes appropriate boundary conditions on S. there being no additional surface integrals to evaluate. t) is adjacent to a rigid boundary S on which the normal derivative ∂ϕ/∂ xn = 0. The main practical difﬁculty is the calculation of the modiﬁed Green’s function.1.1.1) is then once again given by Formula (3. t − τ ) so that it automatically takes account of the contributions from the dipole and monopole sources on S. .1. It would be very convenient if we could modify the functional form of G(x.1. where the source distribution q(x. circular cylinders.
4) ˆ shows that. y.1 The Inﬂuence of Solid Boundaries 43 To simplify the calculation of the compact Green’s function. y. (3. 1958). Sound of frequency ω has wavelength λ= 2π .6) then the substitution of this and (3. for each frequency ω.7) where κ0 = ω/c0 is called the acoustic wave number.1.5) which expresses the δ function as a linear combination of timeharmonic oscillations of frequency ω. κ0 is compact for waves of Thus.5) into the Green’s function equation (3. . (3. ω) is the solution of 2 ˆ ∇ 2 + κ0 G(x. (3. we use the Fourier integral formula δ(t − τ ) = 1 2π ∞ −∞ e−iω(t−τ ) dω. t − τ ) = −1 2π ∞ −∞ ˆ G(x. 1 2π ∞ −∞ e−iω(t−τ ) dω ≡ lim →+0 1 2π π[ 2 ∞ −∞ e−iω(t−τ )− .1.8) This condition will be used below in Section 3.1. G(x. ω dω = lim →+0 + (t − τ )2 ] The ﬁnal term on the right is the usual deﬁnition of δ(t − τ ) as the limit of an ‘ sequence’ (Lighthill. Then.3.1. ω) = δ(x − y).4 to calculate the compact Green’s function. (3.1. y.1. y. ω)e−iω(t−τ ) dω. a solid body of characteristic dimension frequency ω provided that λ = κ0 2π 1. The formula is proved by observing that no real system can oscillate at inﬁnitely large frequencies. and therefore in all practical problems e−iω(t−τ ) can be replaced by e−iω(t−τ )− ω for arbitrarily small > 0. If we now put G(x.
44 3 The Compact Green’s Function 3.5 for the wave equation. ω)e−iωt dω. ω) ˆ (3. t) = ∞ −∞ ˆ q(x.2.2 The Helmholtz Equation The equations 2 ∇ 2 + κ0 ϕ = 0. y.2.5) for r = x > 0. we have ∂2 ˆ 2 ˆ (r G) + κ0 (r G) = 0 ∂r 2 and therefore Aeiκ0 r Be−iκ0 r ˆ G= + .2.2) Therefore.4) In the usual way (see Section 1.1. because (differentiating under the integral sign) 1 ∂2 2 c0 ∂t 2 ∞ −∞ ϕ(x.2. The source term q(x.1. .1) by ϕ(x.1) are known respectively as the Helmholtz equation and the inhomogeneous ˆ Helmholtz equation. ω) for the inhomogeneous Helmholtz equation. t) of Equation (3. ˆ (3.7) determines Green’s function G(x. The free space Green’s function can be found by the method used in Section 1.3) 3. we have to ﬁnd the radially symmetric solution of 2 ˆ ∇ 2 + κ0 G = δ(x).2. B remain to be determined. ω) represents one frequency component of the source q(x. ω) of the inhomogeneous equation is related to the solution of ˆ (3. ω)e−iωt dω. ˆ the solution ϕ(x.1). t) = ∞ −∞ ϕ(x. ω)e−iωt dω. ω)e−iωt dω = ˆ 1 2 c0 ∞ −∞ ∞ −∞ −ω2 ϕ(x. If we temporarily set y = 0. ω)e−iωt dω ˆ ≡− 2 κ0 ϕ(x.1 The Point Source ˆ Equation (3.2. so that q(x. (3. (3. ˆ 2 ˆ ∇ 2 + κ0 ϕ = q(x.1.5). r r where A. (3.
8) we ﬁnd A = −1/4π.1.1) can be expressed as a superposition of point sources by means of ˆ q(x.2 The Helmholtz Equation 45 To do this recall that our solution represents one component of a timedependent acoustic problem of frequency ω. ω) ∂x j ˆ and q(x. the solution of the inhomogeneous Helmholtz equation in an unbounded medium can be written ϕ(x.4. ω) = ∞ −∞ ˆ q(y. The free space Green’s function for the inhomogeneous Helmholtz equation (the solution of (3. ω) ∂ xi ∂ x j (3. By substituting the solution into (3. ω) d 3 y ≡ 4π −∞ ∞ ∞ −∞ ˆ q(y. x − y (3.7 can be used to show that the corresponding solutions for the dipole and quadrupole sources ˆ q(x.2.2.2 Dipole and Quadrupole Sources The method of integration by parts described in Section 1. B = 0.2. Since the time factor is e−iωt . and must therefore be rejected because of the radiation condition. ω) = −1 ˆ ˆ G(x.2. the two terms on the righthand side of (3.7)) is now obtained by replacing r = x by x − y −eiκ0 x−y ˆ G(x.4) and using the Formula (1.2.2.3. ω)eiκ0 x−y 3 d y.7) 3. y.2. ω)q(y. The value of the remaining constant A is found by extending the solution to include the region occupied by the source at r = 0 by writing ˆ G = lim Aeiκ0 r (r 2 + 2) 2 1 →0 .5) correspond to propagating sound waves of the form Ae −iω t− cr 0 r + Be −iω(t+ cr 0 ) . y. ω) = ∂ 2 Ti j (x. ω) = ∂fj (x. ω) in the second of Equations (3. Hence. ω)δ(x − y) d 3 y. r the second of which represents waves converging on the source from inﬁnity.6) ˆ Because the source q(x. ω) = . 4π x − y (3.8) .
ω)eiκ0 x−y 3 d y. x → ∞. y.2. G(x. 3.1. ω)eiκ0 x−y 3 d y. x − y (3.10) 3. using the expression (3. ω). ∂ x1 f 1 δ(y)eiκ0 x−y 3 −iκ0 x1 f 1 eiκ0 x d y≈ . y. ω) = Therefore. We need to consider only a special case of .46 are respectively 3 The Compact Green’s Function ϕ(x. ω) = ˆ and ϕ(x. According to this formula we ﬁnd. y.3). ˆ t−τ ) and G(x. t − τ ) = = = −1 2π ∞ −∞ ˆ G(x. 4π x − y c0 which is precisely Equation (3. y.6) between the Green’s functions G(x. ω) = ˆ −1 ∂ 4π ∂ x j ∞ −∞ f j (y.3 The Reciprocal Theorem The calculation of the compact Green’s function is greatly simpliﬁed by application of the reciprocal theorem.2.6) for the free ˆ space Green’s function G. x − y −1 ∂ 2 4π ∂ xi ∂ x j ∞ −∞ Ti j (y. ω)e−iω(t−τ ) dω ∞ 1 2 x − y 8π e −∞ −iω(t−τ − x−y ) c 0 dω 1 x − y δ t −τ − .9) Example For a point dipole at the origin orientated in the x1 direction ˆ q(x.2.2.3 Green’s Function for the Wave Equation Let us verify the general relation (3. respectively. ω) = ˆ −1 ∂ 4π ∂ x1 ∞ −∞ ∂ ( f 1 δ(x)).1. x − y 4π x2 (3. ϕ(x. for the wave equation and the inhomogeneous Helmholtz equation in the special case in which there are no solid boundaries.
1) (3.3. ω).2) ˆ ˆ In addition G(x. ω) = ∂ xn Z(x.3 The Reciprocal Theorem 47 Fig. ω) = δ(x − x A ). x A .3. ω) ˆ ˆ ∂G G(x.1. ω) = δ(x − x B ). which was ﬁrst used with great effect in acoustics by Lord Rayleigh (1945).3. ω) for x on S. x A . in which sound of frequency ω is generated by two unit point sources at x = x A and x = x B in the presence of a solid body S. (x. (3. ω) = ∂ xn Z(x. x A . ω) . x B .3.1. We denote the functional forms of the respective ˆ ˆ velocity potentials generated by these sources by G(x. x B . (3. x B . ω) and G(x. Consider the two acoustic problems indicated in Fig. this very general theorem of mechanics. where 2 ˆ ∇ 2 + κ0 G(x. 2 ˆ ∇ 2 + κ0 G(x.3. x B . 3. x B . We take these to have the same general linear form ˆ ˆ G(x. ω) ∂G . ω) and G(x. x A . 3.3) . x A . ω) must satisfy appropriate mechanical boundary conditions on S. (x.3.
ω) (x. (3. ω) − G(x. This procedure gives ˆ ˆ ∂G ∂G ˆ ˆ (x. ω)). Z(x. x B . x B . (3. φ deﬁning the orientation of the far ﬁeld point x.3. such that (with implicit time dependence e−iωt ) f A (θ. The angular dependencies of the farﬁeld radiations from the two sources are determined by the factors f A (θ. r r ≡ x → ∞. ω)∇ 2 G(x.3. ω)∇ G(x. x A . ω). ω) − G(x. ω) dS G(x. ω) ˆ ˆ ˆ ˆ = div(G(x.48 3 The Compact Green’s Function where xn is measured in the normal direction from S into the ﬂuid and Z(x. ω) = G(x B . x A . Multiply Equation (3. in the acoustic far ﬁeld. x A . ω) − G(x A . . The reciprocal theorem states that ˆ ˆ G(x A . and generally depend strongly on the details of the interaction of the volume ﬂows from each source with S.3. ω) = ∞. ω)∇ G(x. Green’s identity ˆ ˆ ˆ ˆ G(x. ω) is the surface impedance. x B . x A . both solutions are assumed to exhibit the characteristics of outgoing sound waves. ω).1) by G(x. the potential at x A produced by the point source at x B is equal to the potential at x B produced by an equal point source at x A . φ) and f B (θ. ω) and Equation (3. ω). x A . φ)eiκ0 r ˆ G(x. x B . x B . x A . subtract the resulting equations and integrate over the volume bounded by the surface S and by a large surface in the acoustic far ﬁeld. x B . For a rigid surface. x B . ω) ∼ . At large distances from S.2) by ˆ G(x. x A . x A . ω) ∂ xn ∂ xn S+ ˆ ˆ = G(x B . φ). ω) − G(x. x B . x B . r f B (θ. ω) ∼ . whereas the integrals involving the δ functions can be evaluated explicitly. and the divergence theorem permit the volume integral of the term obtained from the lefthand sides to be expressed as surface integrals over S and .4) where it may be supposed that the coordinate origin is in the neighborhood of S. x B . which are functions of the polar angles θ.5) That is. φ)eiκ0 r ˆ G(x. x A . x A . ω)∇ 2 G(x. ˆ Proof.3.
. as r → ∞. ω) ∼ f A (θ.4 TimeHarmonic Compact Green’s Function 49 The surface integral over S vanishes because of the impedance conditions (3. x. ˆ ∂G =0 ∂ xn on S.1.6) 3. x A .3. x B .3.3.1) where the rigid body S is assumed to be acoustically compact. ∂ xn ∂r . ω) = δ(x − y). and therefore ˆ ∂G iκ0 eiκ0 r (x. (3.4 TimeHarmonic Compact Green’s Function ˆ We are now ready to derive the compact Green’s function G(x.3). y. The result is usually expressed as the simple reciprocal relation ˆ ˆ G(x. y. The inﬂuence of Fig.3. We have to solve 2 ˆ ∇ 2 + κ0 G(x.4. ∂ xn This proves the theorem. 3. ω) = G(y. ω) for the problem depicted in Fig. 3. y.1. φ) ∂ xn r ˆ ∂G iκ0 eiκ0 r (x. ω) ∼ f B (θ. (3.4. φ) ∂ xn r ∂r . ω).4) and because ∂θ/∂ xn and ∂φ/∂ xn are each of order 1/r as r → ∞. The surface integral over vanishes because of conditions (3.4.
x. Let denote the characteristic diameter of the body. ω) (the potential G(x. x. (3.1) is then given by the reciprocal ˆ ˆ ˆ theorem (Section 3. and G(y. obviating the need to evaluate surface integrals. ω) = G(y. y. x. The term G (y. y. ω) produced at the nearﬁeld point y by an equal point source at the farﬁeld point x). ω) = δ(y − x). ω) ≡ −eiκ0 x−y ˆ + G (y.4. 2 ∂ y1 ∂ y2 ∂ y3 ˆ ∂G =0 ∂ yn on S. x. ω) is the spherically spreading wave generated by the point ˆ source at x when the presence of the solid is ignored.2).2) ˆ where the source is at the farﬁeld point x. ˆ In these circumstances the compact approximation for G(x. To solve (3. ω) can be found very easily from the solution of the reciprocal problem: ∂2 ∂2 ∂2 2 ˆ + 2 + 2 + κ0 G(y. x. x. the approximations x − y ≈ x − x j yj x·y ≡ x − x x and 1 1 x·y 1 ≈ ≈ + x − y x x3 x . we put ˆ ˆ ˆ G(y. ω) at the farﬁeld point x produced by the point source at y is exactly equal to the potential ˆ G(y. The solution of (3.50 3 The Compact Green’s Function a solid body on the production of sound by neighboring sources is equivalent to an additional distribution of monopoles and dipoles on S. y. ω) + G (y. x. x. The compact Green’s function includes a ﬁrst approximation for the net effect of these monopole and dipole distributions.8) therefore implies that κ0 1 and κ0 y 1. When x → ∞. In practice. ω) is the velocity potential of the motion produced in the ﬂuid when this wave impinges on S.3) G(x. we are interested primarily in calculating the sound in the far ﬁeld of the body. and take the coordinate origin at O within S. The compactness condition (3. The source point y is assumed to be close to S (so that y ∼ ) and the observer at x is taken to be in the acoustic far ﬁeld. ω) = G 0 (y. ω) 4π x − y ˆ where G 0 (y.4.1.4. ω) is determined as a function of y close to S. x. x.
j . ω) = j 4πx (3. The function ϕ ∗ has the dimensions of length and ∼ in order of magnitude j (Batchelor 1967). G 0 (y. ω) satisﬁes (3. ω) = G 0 (y. G(y. x. and therefore. Then. x. x.2) with the righthand side replaced by zero (because the source is in the far ﬁeld). ω) = 4πx 4π x x where U j = eiκ0 x iκ0 x j . −eiκ0 x ˆ ˆ ˆ + U j (y j − ϕ ∗ (y)) + · · · . Hence.4.3. the distortion of this ﬂow produced by the body must be small. x. ω) = −U j ϕ ∗ (y) + O(κ0 )2 .4.5) where the terms shown explicitly represent a potential ﬂow past the body. Thus. ∇ 2 ϕ ∗ (y) = 0. j But U j (y j − ϕ ∗ (y)) = O(κ0 ). j i.4 TimeHarmonic Compact Green’s Function and the condition κ0 y j ∼ κ0 1 imply that 51 iκ0 x j y j −eiκ0 x−y −eiκ0 x ˆ G 0 (y. At distances y from S.4) can be regarded as the velocity potential of a uniform ﬂow at velocity U j impinging on the solid. ω) + G (y.4. x.3) The linear dependence on y j in the second line of this formula represents the ﬁrst approximation (of order κ0 ) in a power series expansion of rapidly decreasing terms that describes the variation of the incident spherical wave close to the body.e. 4πx x (3. correct to the neglect of small j terms of order O(κ0 )2 . regarded as a function of y. x.4. ˆ Near the body G(y. Let it be represented by the velocity potential ˆ G (y. j where ϕ ∗ (y) → 0 j when y . the terms shown explicitly in eiκ0 x iκ0 x j y j −eiκ0 x ˆ + + · · · ≡ constant + U j y j + · · · . U j ∇ 2 (y j − ϕ ∗ (y)) + O(κ0 )2 = 0. ∇ 2 (y j − ϕ ∗ (y)) = 0. x. ω) ≡ ≈ × e− x 4π x − y 4π x ≈ −eiκ0 x 4π x 1− iκ0 x j y j + O(κ0 )2 x (3.
and can be interpreted as the velocity potential of an incompressible ﬂow past S that has unit speed in the j direction at large distances from S. (3. y. y ∼ O( ). x → ∞. The function ϕ ∗ (y) decays j with distance from S. ω) = 4π x 1− iκ0 x j (y j − ϕ ∗ (y)) + O(κ0 )2 .52 3 The Compact Green’s Function where the rigid surface condition requires ∂ (y j − ϕ ∗ (y)) = 0 on S. j ∂ yn (3.4. and satisﬁes ∂ϕ ∗ j ∂ yn (y) = n j on S.10) is called the compact Green’s function for source points y near the body and observer positions x in the acoustic far ﬁeld. The next term is O(κ0 ) and includes a component −iκ0 x j y j /x from the incident wave plus a correction iκ0 x j ϕ ∗ (y)/x j produced by S.4. .4. Deﬁnition −eiκ0 x ˆ G(x. Hence. The vector ﬁeld Y(y) ≡ y − ϕ∗ (y) is called the Kirchhoff vector for the body.4. ϕ ∗ (y) is just the instantaj neous velocity potential of the motion that would be produced by translational motion of S as a rigid body at unit speed in the j direction.6): When x is in the acoustic far ﬁeld. ω) evaluated at y = 0.4)–(3. j x (3.8) because ∂ y j /∂ yn ≡ n i ∂ y j /∂ yi = n i δi j = n j .4.9) (3. (3.4. j x y ∼ O( ). and y is close to the body −eiκ0 x ˆ G(x. x → ∞.6) Summarizing our conclusions from Equations (3. y.7) The ﬁrst term in the large brackets represents the contribution from the spherical ˆ wave G 0 (x.4. y. the jth component Y j (y) ≡ y j − ϕ ∗ (y) j satisﬁes Laplace’s equation ∇ 2 Y j = 0 with ∂Y j /∂ yn = 0 on S. ω) = 4πx 1− iκ0 x j (y j − ϕ ∗ (y)) .
we shall introduce a very much more elegant representation of the compact Green’s function that greatly expands its utility.5 Compact Green’s Function for a Rigid Sphere 53 In Section 3. Consider the case j = 1 shown in the ﬁgure. φ) with ϑ measured from the positive y1 axis.5. Fig. 3. Take spherical polar coordinates (r. as illustrated in Fig. 3. 3.5 Compact Green’s Function for a Rigid Sphere Let the sphere have radius a and take the coordinate origin O at its center. We have to determine the Kirchhoff vector whose jth component Y j (y) = y j − ϕ ∗ (y) j for j = 1.1.7. 3 is equal to the velocity potential of incompressible ﬂow past the sphere having unit speed in the j direction at large distances from the sphere.1.3. which satisﬁes the axisymmetric form of Laplace equation 1 ∂ r 2 ∂r r2 ∂ ∂r + r2 ∂ 1 sin ϑ ∂ϑ sin ϑ ∂ ∂ϑ (r ) cos ϑ = 0. The ﬂow is evidently symmetric about the y1 axis. (3. ϑ.1) The axisymmetry of the problem suggests that we look for a solution of Laplace’s equation in the form ∗ ϕ1 = (r ) cos ϑ. 2. y1 = r cos ϑ and the condition (3.5.5.4. . Then.9) to be satisﬁed on the sphere is ∗ ∂ϕ1 = cos ϑ ∂r at r = a.
.4.2) This represents the farﬁeld acoustic potential produced by a point source at y close to the sphere. after all.e. second term that determines the leading order approximation for the farﬁeld sound. y ∼ O(a). i. (3. Because of the symmetry of the sphere it is clear that we also have Y2 = y2 1 + a3 2r 3 . This appears to suggest that. Therefore. Thus.1) supplies B = −a /2. 3. ∗ where A and B are constants. x → ∞.5. The condition that ϕ1 → 0 as r → ∞ implies 3 that A = 0. and in these circumstances we shall see that it is the small. y. but most sources of interest in applications are dipoles or quadrupoles. Y3 = y3 1 + a3 2r 3 . r = y. n = −2.5. ω) = 1− 4π x x 1+ a3 2y3 .54 provided that 3 The Compact Green’s Function r2 d2 d −2 + 2r dr 2 dr = 0.2) to determine the farﬁeld sound generated by a dipole source close to a sphere of radius a λ = acoustic . the sphere has a relatively small effect on the production of sound! This is certainly true for monopole sources. Because κ0 y is small the second term in the brace brackets is always small compared to 1.10) for the sphere is iκ0 x j y j −eiκ0 x ˆ G(x.5. 1. Hence. and condition (3.1 Radiation from a Dipole Adjacent to a Compact Sphere Let us apply the compact Green’s function (3. The solutions of this equation are proportional to r n where n is a root of the quadratic equation n 2 + n − 2 = 0. the compact Green’s function (3. ∗ Y1 ≡ y1 − ϕ1 = r cos ϑ − Ar + B r2 cos ϑ.5. Y1 = r cos ϑ + a3 a3 cos ϑ ≡ y1 1 + 3 2r 2 2r .
2. 0).0) (3. 0). y. 3. as in Fig. 3. To determine the solution in the far ﬁeld given that the sphere is acoustically compact we use the compact approximaˆ tion (3. The dipole is orientated in the x1 direction and lies on the x1 axis at (L . y=(L . y. ω) d 3 y.5.5. ω) = ˆ f1 ∂ ˆ {δ(y1 − L)δ(y2 )δ(y3 )} G(x. y. We see immediately that the differentiation with respect to y1 will be applied only to the small second term in the braces . To evaluate the integral we write ϕ(x. ω) d 3 y.5 Compact Green’s Function for a Rigid Sphere 55 Fig. so that ϕ(x.0.3. ω) = f 1 ˆ − f1 ∂ ˆ {G(x. ∂ y1 ˆ where the integration is over the ﬂuid.2.5. ω)δ(y1 − L)δ(y2 )δ(y3 )} d 3 y ∂ y1 ˆ ∂G δ(y1 − L)δ(y2 )δ(y3 ) (x. wavelength. The source term is zero everywhere except at (L .2) for G(x. ω) = −f 1 ˆ ˆ ∂G (x. y. With the origin at the center of the sphere.5. we consider the outgoing wave solution of 2 ∇ 2 + κ0 ϕ = f 1 ˆ ∂ϕ ˆ ∂ {δ(x1 − L)δ(x2 )δ(x3 )}. ∂ y1 The ﬁrst integral is zero because δ(y1 − L) = 0 on the boundaries of the region of integration. y. 0. ω). ω) ∂ y1 . and ∂ G/∂ xn = 0 on the sphere. The solution is given by ϕ(x. 0. where =0 ∂ x1 ∂ xn on x = a.3) Thus far the calculation is exact.
3.0.56 of (3. y. where θ is the angle between the x1 axis and the x direction (so that x1 = x cos θ). The net radiation is therefore equivalent to that produced by a quadrupole source.5. x → ∞. ω) per unit area of S when S is Fig. but it is also true for any compact rigid surface when a dipole orientated in the direction of the local surface normal approaches the surface. we recover the far ﬁeld (3. ω) (Fig. This conclusion applies only to dipoles orientated radially with respect to the sphere (see Problem 1). By setting a = 0 in this formula.3). .2. because in this limit the surface of the sphere is effectively plane in the vicinity of the dipole and an equal and opposite image dipole is formed in the sphere.2 Sound Produced by a Vibrating Sphere Let the surface S of a ﬁxed body execute small amplitude vibrations with normal velocity vn (x. 3.5.5.2).3.10) of a dipole source in the absence of the sphere. 3. ω). and to calculate the sound in this case it would be necessary to use a more accurate approximation to ˆ G(x.5. ω) = ˆ = = 3 The Compact Green’s Function −iκ0 f 1 x j eiκ0 x 4πx2 −iκ0 f 1 x1 eiκ0 x 4π x2 ∂ a3 yj 1 + ∂ y1 2y3 1− a3 L3 a3 L3 y=(L . The presence of the sphere accordingly reduces the amplitude of the sound relative to that produced by a freeﬁeld dipole. The ﬂuid motion is the same as that generated by a distribution of monopoles of strength vn (x. The amplitude is zero when L → a.0) −iκ0 f 1 cos θ eiκ0 x 4π x 1− . κ0 a 1. giving ϕ(x.
1) is ˆ q(x. The second integral is nonzero only for j = 1. ˆ vn (y. ω) ≈ ˆ ˆ ˆ iκ0 U (ω)a 3 x1 eiκ0 x iωU (ω)a 3 cos θeiκ0 x . The ﬁrst integral represents the net volume ﬂux through S and vanishes identically for rigid body translational motion. ω) = vn (x. when y1 = a cos ϑ and y = a on S.7. ω) = U (ω) cos ϑ. This . ω) in the inhomogeneous Helmholtz equation (3.1). ω) is therefore ˆ ϕ(x. and > 0 places the sources just within the ﬂuid adjacent to S.2) for G(x. ω) = ˆ ﬂuid ˆ vn (y. x → ∞. y. Then. ( → +0).4) can ˆ be evaluated using the compact approximation (3. ω)G(x. ω) ˆ ≈ −eiκ0 x 4πx vn (y.1 vibrates with small amplitude about its undisturbed position centred at the origin with velocity ˆ U (ω)e−iωt along the x1 axis. ω): ϕ(x. ∂ xn (3. Hence.2. y. where s⊥ is distance measured in the normal direction from S into the ﬂuid.4) Consider the sound produced when the sphere of Fig. ≡ 2x2 2c0 x where θ is the angle between the x1 axis and the radiation direction x (see Fig. and when the observer at x is in the acoustic far ﬁeld the integral in (3.5. ω)δ(s⊥ − ). The velocity potential ϕ(x.5. ω) d S(y) . ω) d S(y) − S iκ0 x j x yj 1 + S a3 2y3 vn (y. The corresponding source strength q(x.5 Compact Green’s Function for a Rigid Sphere 57 ˆ assumed to be stationary (rigid). ω) = 0 on S. ω)δ(s⊥ − )G(x. The solution for a sphere oscillating at an arbitrary time dependent velocity U (t) can be derived from this result provided the sphere remains compact. 3.5. If the vibrations are at sufﬁciently low frequency. the sphere will be compact.3. 1. and we can take d S = 2πa 2 π ˆ sin ϑ dϑ (so that the surface integral becomes 3πa 3 U (ω) 0 cos2 ϑ sin ϑ dϑ = 3 ˆ 2πa U (ω)). ω) d 3 y ˆ vn (y. ϕ(x. y. ω) d S(y) S ( → +0) = where ˆ ∂G (x. y.5.
and conﬁrms the model used there in which the vibrating sphere was replaced by a point dipole of strength 2πa 3 U (t) at its center. y.7. Hence.2) and (3. 3.1 illustrates the situation for an inﬁnite circular cylinder of radius a whose axis lies along the y3 axis. 3.7) remains valid ∗ with ϕ3 (y) ≡ 0.2) 3.4.6. (3. because the impinging ﬂow described by the velocity potential (3.2. Y3 = y3 . ˆ 1. −a 3 cos θ ∂ = 2c0 x ∂t = −a 3 cos θ ∂U (t − x/c0 ). we can take −eiκ0 x ˆ G(x. t) ≈ a 3 cos θ 2c0 x ∞ −∞ ∞ −∞ ˆ iωU (ω)e−iω(t−x/c0 ) dω ˆ U (ω)e−iω(t−x/c0 ) dω x → ∞.4. Figure 3. ∗ ϕ2 (y) . If this condition is satisﬁed we can use the then U (ω) = 0 only for κ0 a Formulae (3.2. 2c0 x ∂t This agrees with the farﬁeld result obtained in Section 1.1 Circular Cylinder for the circular cylinder of radius a can be found The potentials by the method of Section 3.1) where the Kirchhoff vector Y has the components ∗ Y1 = y1 − ϕ1 (y). ω) = 4π x 1− iκ0 x j Y j x .6. The source point y is adjacent to the cylinder and for the moment (see Section 3.4.4) for j = 3 is unaffected by the cylinder.7) is assumed to be within an axial distance y3  λ from the coordinate origin O.6.6. ∗ ϕ1 (y).4 for the compact body in Fig.6 Compact Green’s Function for Cylindrical Bodies The reciprocal calculation of the Green’s function described in Section 3. and whose diameter 2a ∼ is acoustically compact.1 can be immediately extended to the case of a cylindrical body of compact cross section.58 3 The Compact Green’s Function ∞ −∞ means that if we write U (t) = ˆ U (ω)e−iωt dω.5. y ∼ O( ). x → ∞. ∗ Y2 = y2 − ϕ2 (y).3) to obtain the timedependent velocity potential in the form ϕ(x. (3. In this region the Expansion (3.
3) Fig. Using polar coordinates (y1 .4.9) to be satisﬁed on the cylinder is ∗ ∂ϕ1 = cos ϑ ∂r at r = a. 3.1.6. 3. the condition (3.6 Compact Green’s Function for Cylindrical Bodies 59 Fig. (3.6. sin ϑ).3. y2 ) = r (cos ϑ. For j = 1 the ﬂow is symmetric about the y1 axis and is independent of the spanwise coordinate y3 (Fig.2.6. 3.6. .2).
3) yields B = −a 2 . Y2 = y2 1 + a2 r2 .3 the airfoil occupies −a < y1 < a.2 Rigid Strip The rigid strip of chord 2a and inﬁnite span provides a simple model of a sharpedged airfoil.6. a2 a2 cos ϑ ≡ y1 1 + 2 r r .6. y. which satisﬁes the polar form of Laplace’s equation 1 ∂ r ∂r provided that r2 d2 d − +r dr 2 dr = 0. 2.60 3 The Compact Green’s Function As in the case of the sphere. In Fig. is −eiκ0 x ˆ G(x. Therefore. Y3 = y3 .4) 3. (3. y2 = 0. r and condition (3. y ∼ O( ). so that potential functions ϕ1 (y) ≡ 0 ∗ and ϕ3 (y) ≡ 0. r ∂ ∂r + 1 ∂2 r 2 ∂ϑ 2 (r ) cos ϑ = 0. the compact Green’s function for a circular cylinder. The component Ar must be rejected because it does not decay as r → ∞. with source near the origin.6. Therefore. −∞ < y3 < ∞. The airfoil has no inﬂuence on a uniform mean ﬂow in the y1 ∗ direction. x → ∞. ∗ Y1 ≡ y1 − ϕ1 = r cos ϑ − B cos ϑ.5) 1− iκ0 x j Y j x . (3. Y1 = r cos ϑ + Similarly. The general solution is = Ar + B/r . . 3. we try a solution of the form ∗ ϕ1 = (r ) cos ϑ. j = 1. ω) = 4π x where Yj = yj 1 + a2 2 2 y1 + y2 .6. Hence.6. nor on one in the y3 direction.
This ﬂow can be found by the method discussed above for the circular cylinder (or see Example 3 of Section 4. where w is the complex potential w(z) = − =− i 2 Z− a2 Z z2 − a2 − a2 √ z + z2 − a2 i z+ 2 = −i z 2 − a 2 . y2 ) can be determined by the method of conformal transformation. Thus. ∗ ∗ The potential ϕ2 (y) ≡ ϕ2 (y1 . ω) = 4π x 1− iκ0 x j Y j x . is given by −eiκ0 x ˆ G(x. z = y1 + i y2 .6. (3. Y2 = Re(−i z 2 − a 2 ). (3. and determines ∗ Y2 = y2 − ϕ2 (y1 . (Readers unfamiliar with this procedure should consult Section 4. y ∼ O(a).5). the compact Green’s function for a strip.5. y2 ) = Re[w(z)]. 3.3. with source near the origin. y.7) .3.) If z = y1 + i y2 . Because Z ∼ 2z as z → ∞ a uniform ﬂow at unit speed in the y2 direction in the z plane at large distances from the airfoil corresponds to a uniform ﬂow at speed 1 in the direction of the imaginary Z axis at large distances 2 from the cylinder.6.6 Compact Green’s Function for Cylindrical Bodies 61 Fig.6) where the components of the Kirchhoff vector are Y1 = y1 . Y3 = y3 .6. x → ∞. the cross section of the airfoil in the z plane is mapped onto the circular cylinder Z  = a in the Z plane by the transformation Z =z+ z2 − a2.
3. The streamlines crowd together and change very rapidly near the sharp edges. κ0 L 1.0.6. ω) = − f 2 ˆ ˆ ∂G (x. Figure 3.6. from (3. ∂ x2 on the airfoil L > a. y.6. The solution is given by the following form of Equation (3.0. This is an indication that edges can be important sources of noise when located in the near ﬁeld of a dipole or quadrupole (or any higher order multipole) source. . −∞ < x3 < ∞.4 depicts the streamline pattern of the ﬂow past the strip deﬁned by the velocity potential Y2 (y). x2 = 0.0) i f 2 κ0 x2 eiκ0 x 4πx2 ∂Y2 ∂ y2 y=(L . Example Calculate the farﬁeld velocity potential when 2 ∇ 2 + κ0 ϕ = f 2 ˆ ∂ [δ(x1 − L)δ(x2 )δ(x3 )] .7).0) where. where ∂ϕ ˆ =0 ∂ x2 −a < x1 < a. The dipole source is orientated in the x2 direction and is positioned just to the right of the edge at y1 = a in Fig. ω) ∂ y2 ≈− y=(L .3) ϕ(x.62 3 The Compact Green’s Function Fig. z = y1 + i y2 .5.3.4. 3.6. ∂ ∂Y2 = Re −i ∂ y2 ∂ y2 = Re √ z z2 − a2 z2 − a2 . because ∇Y2 becomes very large there.
9. 3.2) . √ The amplitude of the sound is increased by a factor L/ L 2 − a 2 relative to that produced by the same dipole in free space. i. that κ0 Y j −eiκ0 x ˆ G(x. 63 where = cos−1 (x2 /x) is the angle between the normal to the strip and the radiation direction (x/x) indicated in Fig.7. and is unbounded as L → a. ϕ(x.3.3. 4π x (3.7 Symmetric Compact Green’s Function The deﬁnition (3.6.2) x − Y ≈ x − x·Y . x Now let X(x) denote the Kirchhoff vector for the body expressed in terms of x. x → ∞. 3. e 4π x −eiκ0 x−Y .10) of the compact Green’s function can be recast to exhibit the reciprocal nature of the source and observer positions y and x. To do this we ﬁrst observe that. let X j (x) = x j − ϕ ∗ (x). x → ∞. ω) ≈ 4π x ≈ ≈ 1. when the dipole approaches the edge. √ 4πx L 2 − a 2 x → ∞. Hence. therefore. Y ∼ O( ). j and. y.4.7.e. 1− iκ0 x j Y j x = −eiκ0 x 4π x 1− iκ0 x · Y x −1 iκ0 x− iκ0 x·Y x . j (3. ω) ≈ − ˆ =− L i f 2 κ0 x2 eiκ0 x √ 2 2 − a2 4π x L i f 2 κ0 L cos eiκ0 x . Y j (y) = y j − ϕ ∗ (y) ∼ O( ).7 Symmetric Compact Green’s Function Therefore.1) where on the last line we have used the usual farﬁeld approximation (1.. for a body of characteristic diameter .
3) 1 x·Y 1 1 ≈ + ≈ X − Y x x3 x Thus. 4π X − Y Y ∼ O( ).2. j and.1) can be written −eiκ0 X−Y ˆ G(x. The deﬁnition (3.3) x·Y X − Y ≈ x − x when x → ∞. When x is close to the body the source must be ˆ in the far ﬁeld. Our generalized deﬁnition clearly satisﬁes the reciprocal theorem. (3. y. The components X j and Y j are the velocity potential of incompressible ﬂow past the body having unit speed in the j direction at large distances from the body.64 3 The Compact Green’s Function Then. In applications it is necessary to remember also that it is valid for determining only the leading order approximation to the surface monopole and dipole sources induced on the body . The approximation is valid for arbitrary source and observer locations provided that at least one of them lies in the far ﬁeld of the body. x → ∞. therefore. ω) = . This is because for distant sources the amplitude of the sound scattered by a compact rigid object is O((κ0 )2 ) smaller than the incident sound. y.2) and (1. y.7.9. because of the symmetrical way in which x and y enter this formula we may now remove any restriction on the position of the coordinate origin. is of quadrupole intensity (Lighthill 1978.4) where X = x − ϕ∗ (x). ω) ≈ .7. to the same approximation. ω) then determines the modiﬁcation by the body of low frequency sound received by an observer near the body.9. (3. we also have X ∼ x as x → ∞. because ϕ ∗ (x) → 0 as x → ∞.6). Howe 1998a).4) is easily recalled because it is an obvious generalization of the free space Green’s function (3. 4π X − Y (3. Y = y − ϕ∗ (y) are the Kirchhoff vectors for the body expressed respectively in terms of x and y. Also. This result is the basis of our revised deﬁnition of the Compact Green’s Function for the Inhomogeneous Helmholtz Equation −eiκ0 X−Y ˆ G(x.7. ϕ ∗ is the velocity potential of j the incompressible ﬂow that would be produced by rigid body motion of S at unit speed in the j direction. When both x and y are in the far ﬁeld (so that X ∼ x and Y ∼ y) predictions made with the compact Green’s function will be the same as when the body is absent. from (1.7. that is. G(x.
y.5) for the wave equation will now be used to give a complete theory of the lowfrequency sound produced by a vibrating body.5. Y = y − ϕ∗ (y) are Kirchhoff vectors for the body. y. y. ω)e−iω(t−τ ) dω ∞ 1 2 X − Y 8π e −∞ −iω(t−τ − X−Y ) c 0 dω 1 X − Y δ t −τ − .2) with X. except that we now work directly with time dependent quantities. 4π X − Y c0 (3.6) relating Green’s functions for the wave equation and the Helmholtz equation to derive the compact approximation for Green’s function of the wave equation: G(x.7.4).8 LowFrequency Radiation from a Vibrating Body The compact Green’s function (3. its use is subject to the same restrictions as (3.5) where X = x − ϕ∗ (x).6.8 LowFrequency Radiation from a Vibrating Body 65 by neighboring sources in the ﬂuid. In practice this means that when used in ˆ calculations G(x. Y substituted for x. and it will be valid only when applied to timedependent source terms producing sound whose wavelength is large compared to the characteristic body dimension . ϕ ∗ is the velocity potential of the incompressible ﬂow that would be produced j by rigid body motion of S at unit speed in the j direction. y. With this understanding we can deﬁne the Compact Green’s Function for the Wave Equation G(x. 4πX − Y c0 This remarkable result is formally identical with the classical free space Green’s function (1. The maximum frequency of the vibrations must be small enough to ensure that the body (or its cross section. t − τ ) = X − Y 1 δ t −τ − .1. t − τ ) = ≈ = −1 2π ∞ −∞ ˆ G(x. However. in the case of vibrating a cylinder) is acoustically compact. The argument follows closely the discussion of the vibrating sphere in Section 3. ω) will normally be expanded only to ﬁrst order in the Kirchhoff source vector Y(y).3. The components X j and Y j are the velocity potential of incompressible ﬂow past the body having unit speed in the j direction at large distances from the body. .7. y. We can go further and use the formula (3. 3.7.
where s⊥ is distance measured in the normal direction from S into the ﬂuid.5). t − τ ) in (3. τ )δ t − τ − x x · Y + d S(y) dτ c0 c0 x (X ∼ x as x → ∞) .8.1. y. t − τ ) = 0 ∂ xn on S.8. t) = vn (x. τ )δ(s⊥ − )G(x. The velocity potential ϕ(x. they illustrate the general procedure that should be adopted when using the compact Green’s function (in particular. As before.8.8. t).1) distributed over S regarded as a rigid. 3.2) by its compact approximation (3. t) is given exactly by ϕ(x. t − τ ) d 3 y dτ vn (y. y. t) = − =− ∞ −∞ ∞ −∞ ﬂuid vn (y. t)δ(s⊥ − ) ( → +0) (3. stationary surface.2) At low frequencies the ﬁrst approximation to the farﬁeld sound is obtained by replacing G(x. τ )δ(s⊥ − ) δ t −τ − d 3 y dτ 4π X − Y c0 ( → +0). x → ∞ =− =− X − Y vn (y.66 3 The Compact Green’s Function Fig. y. the technique of expanding to ﬁrst order in Y): ϕ(x.8.1) vibrate with normal velocity vn (x. 4π X − Y c0 ∞ −∞ S 1 4π x vn (y. (3.7. S ( → +0) where ∂G (x. Let the closed surface S (Fig. t) ≈ − ∞ −∞ ∞ −∞ S ﬂuid X − Y vn (y.1)) is the same as that generated by the distribution of volume sources q(x. τ )G(x. τ ) δ t −τ − d S(y) dτ. t − τ ) d S(y) dτ. The details are given below.1. y. 3. the velocity potential in the ﬂuid (governed by Equation (3.
Performing the integration with respect to τ : ϕ(x.8. c0 x x c0 +δ t −τ − where the prime denotes differentiation with respect to time. t) ≈ − − 1 4π x vn y. j (3. x → ∞. where n(y) is the surface normal directed into the ﬂuid.4) Example: The vibrating sphere Consider a rigid sphere of radius a centred at the origin and oscillating in the x1 direction at velocity U (t) (Fig. 3. say. τ ) δ t − τ − x c0 xjYj d S(y) dτ. only for a pulsating body). In terms of spherical polar coordinates (r.3. t) ≈ − =− xj ∂Ui 4πc0 x2 ∂t xj ∂Ui 4πc0 x2 ∂t t− t− x c0 x c0 n i (y)Y j (y) dS(y) S (3. c0 xj ∂ 4π c0 x2 ∂t vn y.8. τ ) = n(y) · U(τ ) = n i (y)Ui (τ ).3) S [n i y j − n i ϕ ∗ ] dS.8. ϑ. and is nonzero only if the volume enclosed by S changes with time (i. and it is only necessary to take i = 1 in (3.3) or (3. 0. sin ϑ cos φ. 0). t − S The ﬁrst integral represents an omnidirectional monopole sound wave. vn (y. The monopole term vanishes for a rigid body executing small amplitude translational oscillations at velocity U(t). Making the substitution Y j = y j − ϕ ∗ (y) in the second integral we obtain an acoustic ﬁeld of dipole j type.8. given by ϕ(x.4). Then U = (U.e. sin ϑ sin φ) . It is then the most important component of the far ﬁeld sound – 1 (because ∂/∂t ∼ ω the second integral is smaller by a factor ∼O(ω /c0 ) and Y j ∼ ).2).8 LowFrequency Radiation from a Vibrating Body =− 1 4πx ∞ −∞ S 67 vn (y. t − S x d S(y) c0 x Y j (y) d S(y)..8. φ) we have y = r (cos ϑ. Then.
(3.3) becomes ϕ(x. j = 1.8. x → ∞. j = 2.2. 0. 2c0 x ∂t and x1 = cos θ.68 3 The Compact Green’s Function Fig. 3.8.5 using the solution derived from the Helmholtz equation. therefore. sin ϑ sin φ) 2 on the sphere and n 1 = cos ϑ Hence. sin ϑ sin φ) cos ϑ sin ϑ dϑ dφ S = 2πa 3 .8. deﬁned by the surface integral Mi j = −ρ0 S n i ϕ ∗ d S. sin ϑ cos φ.5) which is the result already obtained in Section 3. sin ϑ cos φ. n1Y j d S = S 3a 3 2 (cos ϑ. j . 3 and. 3.8. t) ≈ −a 3 cos θ ∂U (t − x/c0 ). Y≡y 1+ a3 2y3 = 3a (cos ϑ. Therefore.1 Far Field Pressure Produced by a Vibrating Body A more general and illuminating discussion of the lowfrequency sound produced by a vibrating rigid body can be given in terms of the added mass tensor Mi j (Batchelor 1967). x (3.
8) . Let us now apply these concepts to determine from (3.4. j (y) ∼ O 1 y2 as y → ∞. (3.3. j x → ∞.8. however. 3.8. must also be overcome by the force F when the body accelerates. t) = xj ∂ 2 Ui 2 ∂t 2 4πc0 x t− x c0 ρ0 S n i y j d S − ρ0 S ni ϕ∗ d S . j j The integration over 1967) because vanishes as the surface recedes to inﬁnity (Batchelor ∗ ϕi. dt The added mass tensor determines the effective mass of ﬂuid dragged along by the body in its accelerated motion.8.4) the sound pressure produced by a rigid compact body executing small amplitude translational oscillations at velocity U(t). and Mi j = −ρ0 S n i ϕ ∗ d S = −ρ0 j S ∂ϕi∗ ∗ ϕ d S ≡ −ρ0 ∂ yn j S ϕi∗ ∂ϕ ∗ j ∂ yn d S = M ji .8.3. incompressible ﬂuid. recalling that ∇ 2 ϕ ∗ = ∇ 2 ϕi∗ = 0. a couple must also be applied to the translating body to counter a rotational torque also exerted on the body by the ﬂuid (see Batchelor (1967) for further discussion). in addition to that of the body.8 LowFrequency Radiation from a Vibrating Body 69 The Condition (3. By evaluating the net force on S produced by the unsteady surface pressure (or by the method described below in Section 4. (3.9) satisﬁed by ϕ ∗ on S implies that Mi j = M ji . this means that when an external force Fi acts through its centre of mass.4) it can be veriﬁed that when the body translates at velocity U(t) without rotation in an ideal. the equation of motion of the body can be written Fi (t) = Mi j (mδi j + Mi j ) dU j = Fi .3). and therefore p(x. The acoustic pressure is given in the far ﬁeld by p = −ρ0 ∂ϕ/∂t (see Section 1.6) The ﬁnal integral is deduced from the second by referring to Fig. and applying the divergence theorem as follows: j ∂ϕ ∗ ∂ϕi∗ ∗ j ϕ j − ϕi∗ ∂ yn ∂ yn dS = ﬂuid S+ (ϕi∗ ∇ 2 ϕ ∗ − ϕ ∗ ∇ 2 ϕi∗ ) d 3 y ≡ 0.1. The inertia of this ﬂuid. it exerts a force on the ﬂuid in the i direction given by dU j . (3. In general. because n i j can be replaced in the integrand by ∂ϕi∗ /∂ yn .7) dt For a body of mass m.
x → ∞. ∂ yi (3.9 Compact Green’s Function Summary and Special Cases 3.5). where Fi is the force exerted by the body on the ﬂuid in the i direction. c0 t− x . t) ≈ = ∂ 2U j xi (m 0 δi j + Mi j ) 2 2 4π c0 x ∂t xi 4π c0 x2 m0 ∂ 2 Ui ∂ Fi + 2 ∂t ∂t t− t− x c0 x . y. For a sphere of radius a oscillating at speed U (t) in the x1 direction m 0 = 4 πa 3 ρ0 3 Therefore. t) ≈ = ∂ 2U j xi 1 m 0 δi j + m 0 δi j 4π c0 x2 2 ∂t 2 ρ0 a 3 cos θ ∂ 2 U 2c0 x ∂t 2 t− x .70 3 The Compact Green’s Function The ﬁrst integral is evaluated by applying the divergence theorem. 3.9.8) is just the added mass tensor Mi j .9. (3. which transforms it into an integral over the volume Vs of the body: ρ0 S n i y j dS = ρ0 Vs ∂yj 3 d y = ρ0 Vs δi j ≡ m 0 δi j .8. t − τ ) = X − Y 1 δ t −τ − 4π X − Y c0 Kirchhoff vectors for the body. The second term in the brace brackets of (3.9) where m 0 is the mass of the ﬂuid displaced by the body. p(x. x → ∞.8.8.1 Compact Bodies and Cylindrical Bodies of Compact Cross Section General Form G(x. c0 and Mi j = 1 m 0 δi j 2 which is equivalent to (3.8.10) c0 where the second line follows from (3. Thus. (3.8.1) X = x − ϕ∗ (x) Y = y − ϕ∗ (y) The vector components X j (x) and Y j (y) are the velocity potentials of . the acoustic pressure can be expressed in either of the forms p(x.7).
3. For a cylindrical body of compact cross section parallel to the x3 direction.6) and (3.1. 3. .3 for the strip airfoil. 3. Fig. ϕ ∗ is the velocj ity potential of the incompressible ﬂow that would be produced by rigid body motion of S at unit speed in the j direction. with centre at origin Circular cylinder of radius a coaxial with the x3 axis Strip airfoil −a < x1 < a.9.9. −∞ < x3 < ∞ X1 x1 1 + x1 1 + a3 2x3 a2 2 2 x1 +x2 71 X2 x2 1 + x2 1 + a3 2x3 a2 2 2 x1 +x2 X3 x3 1 + x3 x3 a3 2x3 x1 √ Re(−i z 2 − a 2 ) z = x1 + i x2 incompressible ﬂow past the body having unit speed in the j direction at large distances from the body (special cases are listed in Table 3. variable chord airfoil illustrated in Fig.1. x2 = 0. Y3 = y3 . with y2 normal to the plane of the airfoil and y3 in the spanwise direction. we take X 3 = x3 .6.1.6.2 Airfoil of Variable Chord The compact Green’s function deﬁned by (3. 3. The airfoil span is assumed to be large.9 Compact Green’s Function Summary and Special Cases Table 3. The coordinate axes are orientated as in Fig.9.9.1). and the chord 2a ≡ 2a(y3 ) is a slowly varying function of y3 .3.6. The potential Y2 of ﬂow past the airfoil in the y2 direction may then be approximated locally by the formula for an airfoil of uniform chord 2a(y3 ).9.7) for a rigid strip can be generalized to include the ﬁnite span. Standard Special Cases Body Sphere of radius a.
y3  < 1 L 2 .2. Y2 = Re − i z 2 − a(y3 )2 .9. (3. y3  > 1 L y2 .2). y.9. This model has been found to give predictions within a few percent of those based on the exact value of Y2 (y) in the case of an airfoil of elliptic planform whose aspect ratio = airfoil span > 5.72 3 The Compact Green’s Function Therefore.9. midchord 3. . 2 z = y1 + i y2 . t − τ ) = x − y 1 δ t −τ − 4π x − y c0 + ¯ − y x 1 .3 Projection or Cavity on a Plane Wall Let the plane wall be rigid and coincide with x2 = 0 (Fig.2) Y3 = y3 . 3. a ﬁrst approximation to the compact Green’s function (3. When the projection or cavity is absent the Green’s function with vanishing normal derivative on the wall is G 0 (x. δ t −τ − 4π¯ − y x c0 Fig.1) for an airfoil of span L occupying the interval − 1 L < y3 < 1 L is obtained by taking 2 2 Y1 = y1 . 3.9.9.
It may now be veriﬁed that (in the usual notation) the required compact Green’s function is G(x.e. . ∗ Y3 = y3 − ϕ3 (y) ∗ X 3 = x3 − ϕ3 (x) . y. but the following discussion applies without change to compact (but nonresonant) wall cavities.3) where ∗ Y1 = y1 − ϕ1 (y). The ﬁgure illustrates the case for a projection.9 Compact Green’s Function Summary and Special Cases 73 ¯ where x = (x1 . y. ¯ − Y c0 4π X (3. t − τ ) ≈ ≈ 1 x x·y + δ t −τ − 4πx c0 c0 x 1 x 2δ t − τ − 4πx c0 + +δ t −τ − ¯ x·y x + c0 c0 x t −τ − x c0 .9.9.. X 3 = x3 . By inspection. Y2 = y2 . X 3 ). ∗ X 1 = x1 − ϕ1 (x).4) ¯ and X = (X 1 . −X 2 . Let x → ∞ (noting that ¯  = x) and x expand G 0 near the projection to ﬁrst order in y (i. 2(x1 y1 + x3 y3 ) δ c0 x We require a corrected expression that has vanishing normal derivative (as a function of y) on the wall and on the projection. x3 ) is the image of the observer position x in the plane wall.3. (3. correct to dipole order) G 0 (x. X 2 = x2 . say. To complete this discussion of compact Green’s function. in the x3 direction simply by setting Y3 = y3 . Assume ﬁrst that the origin is close to the projection. −x2 . we now give without proofs a selection of useful examples. c0 x ∗ ∗ where Y1 = y1 − ϕ1 (y). Y3 = y3 − ϕ3 (y) are the velocity potentials of horizontal ﬂows past the projection that are parallel to the wall and have unit speeds respectively in the y1 and y3 directions as y → ∞. this is obtained simply by replacing the factor 2(x1 y1 + x3 y3 ) c0 x by 2(x1 Y1 + x3 Y3 ) . t − τ ) = X − Y 1 δ t −τ − 4π X − Y c0 ¯ X − Y 1 δ t −τ − + . These formulae can be used also for a twodimensional projection or cavity that is uniform.
4π x − y3 i3 √ κ0 ϕ ∗ (x)ϕ ∗ (y) iκ0 x−y3 i3  −1 ˆ 1 (x. for κ0 (y1 + y2 ) 2 do this. x3 ). ω) are known for a rigid halfplane x1 < 0.6) and ϕ ∗ (x) = ϕ ∗ (y) = (3.9. To small compared to the acoustic wavelength. However.5) 1. 3. 2 2 where. if i3 is a unit vector in the x3 direction (parallel to the edge). Then.3) but are of limited use in applications.9.9. π 2πi x − y3 i3 3/2 √ r sin(θ/2). 1975a) ˆ Analytical representations of the exact Green’s function G(x. √ r0 sin(θ0 /2). y3 ). 3. (3. ˆ G 0 (x. ˆ ˆ ˆ G(x. y. we introduce cylindrical polar coordinates x = (r cos θ.9. 3.4 Green’s Function for a HalfPlane (Howe. r sin θ. y. that is. y. Fig. ω) + G 1 (x. y. ω) + · · · . ω) = G 0 (x.9.7) . ω) = √ G e . y. y.74 3 The Compact Green’s Function Fig. ω) = −1 eiκ0 x−y3 i3  . r0 sin θ0 . for x − y3 i3  → ∞ and κ0 y1 + y2 (3. y = (r0 cos θ0 . x2 = 0 (which is inﬁnite in the x3 direction.3.9. we can deﬁne a compact Green’s function for a source at y whose distance from the edge is 2 2 1 1.
Z = y2 + i y1 . y. y.9. y. t − τ ) ≈ ϕ ∗ (x)ϕ ∗ (y) δ(t − τ − x/c0 ).9. t − τ ) ≈ G 0 (x. t − τ ) + · · ·. y = (y1 . x2 ) is in the acoustic far ﬁeld. and then using the Formula (3.1. satisﬁes 1 ∂2 − ∇ 2 G = δ(x1 − y1 )δ(x2 − y2 )δ(t − τ ). where conditions are uniform in the x3 direction.9. 1975a) The Green’s function for the wave equation in two dimensions. The twodimensional compact Green’s function (the solution of (3. The ydependent part of the compact Green’s function is √ G(x.6) over −∞ < y3 < ∞. 3.7).9 Compact Green’s Function Summary and Special Cases 75 ϕ ∗ (x) is a velocity potential of incompressible ﬂow around the edge of the halfplane expressed in terms of polar coordinates (x1 . 3.4). G 1 is the ﬁrst correction due to presence of the halfplane.9. y.3.3 the corresponding compact Green’s function is obtained by integrating (3.9. y2 ) well within an acoustic wavelength of the aperture on either side of the plane.8)) is applicable for a source at y = (y1 . y.9. (3.5 TwoDimensional Green’s Function for a HalfPlane (Howe. x → ∞. x2 ) = r (cos θ. 3.9. y2 ) and ϕ ∗ is deﬁned as in (3.9. y2 ) is close to the edge of the halfplane in Fig.6 TwoDimensional Green’s Function for a Plane with an Aperture A rigid plane x1 = 0 is pierced by a twodimensional aperture occupying −a < x2 < a (Fig.9) where x = (x1 .9. and is found to be G 1 (x. The observer at x = (x1 .2). (3. x2 ).9. π x (3.10) .8) When a line source at y = (y1 . t − τ ) is the ﬁrst term in the expansion that involves y. t − τ ) + G 1 (x. sin θ) (and ˆ ˆ similarly for ϕ ∗ (y)).6) to calculate G(x. t − τ ) ≈ − c0 sgn(x1 ) χ(t − τ − x/c0 ) Re ln √ √ t − τ − x/c0 π 2π x Z + a Z2 −1 a2 . using the method of stationary phase for 2 2 κ0 x1 + x2 → ∞ (see Example 2 of Section 5. 3. 2 c0 ∂t 2 where G = 0 for t < τ. The component G 0 of G represents the radiation from a ˆ point source at y when scattering is neglected. In particular G 1 (x.
76 3 The Compact Green’s Function Fig. .4. Note the deﬁnition Z = y2 + i y1 .9.7 Green’s Function for Long Waves in a Rigid Walled Duct (Howe. t − τ ) = x1 − y1 c0 H t −τ − .11) where H is the Heaviside step function. 3. x − y 2A c0 √ A. .9. even if the source region is highly three dimensional. 1975b) Only plane waves can propagate in a cylindrical duct of crosssectional area A when the characteristic wavelength of sound is large compared with the √ diameter ∼ A. 3. . provided the crosssectional area is uniform.5a). we have G(x.781072. 3. For a uniform duct with an acoustically . [ln( aξ 2 /4c0 t)]2 + π 2 2 and = 1. (3.9. y.9. When this condition is satisﬁed the corresponding compact Green’s function satisﬁes the onedimensional wave equation. Taking the x1 direction along the axis of the duct (Fig. where χ(t) = H (t) 0 ∞ ln( aξ 2 /4c0 t)e−ξ dξ .
Then. t − τ ) = A. and not necessarily that of the uniform cylinder shown in the ﬁgure.9. when the . Within the duct.9. with the negative x1 axis lying along the axis of the duct.5.6a. several diameters from the entrance. However. 3.9. Take the coordinate origin in the entrance plane of the duct. there are two cases: (i) Propagation within the Duct This is applicable for the case shown in Fig.9 Compact Green’s Function Summary and Special Cases 77 Fig. y.8 Compact Green’s Function for a Duct Entrance (Howe. 3. (3.9. compact section of variable cross section. 3.b) The typical geometry is illustrated in Fig.5b.6a. the geometry of the entrance can be arbitrary.9. involving a source at y near the duct entrance and an observer at x within the duct (or vice versa). x − y H t −τ − 2A c0 √ G(x. the crosssectional area is uniform and equal to A. the compact Green’s function becomes X 1 − Y1  c0 . such as the neck in Fig. 1998a. 3. 3.3.9.12) where X 1 (x) and Y1 (y) are the velocity potential of incompressible ﬂow in the duct having unit speed at large distances from the neck.
3.9. (3.6b) When either the source or observer is located at a large distance from the duct entrance in free space G(x.6. (ii) Propagation in Free Space (Fig.9.14) in which is the end correction (Rayleigh. (3.9.9. G(x. 3.14). ϕ ∗ (x) ≈ x1 − √ ≈ −A/4π x when x A outside the duct. y. t − τ ) ≈ c0 2A H t −τ − ϕ ∗ (x) − ϕ ∗ (y) c0 . 1945) of the duct opening (≈ 0.78 3 The Compact Green’s Function Fig.9. y.15) where ϕ ∗ is deﬁned as in (3.61R √ for an unﬂanged circular cylinder of radius R = A/π ). and satisﬁes √ when x1  A within the duct. t − τ ) = (X(x) − Y(y) − [ϕ ∗ (x) + ϕ ∗ (y)]) 1 . Y(y) denote the Kirchhoff vector whose i component is the velocity potential of ﬂow past the stationary surface formed by the duct entrance having unit speed in the i direction at large distances from the entrance outside the duct (they become exponentially small . δ t −τ − 4π X − Y c0 (3. characteristic acoustic wavelength is large compared to the duct diameter. and X(x).9.13) −H t − τ + ϕ ∗ (x) + ϕ ∗ (y) c0 where the velocity potential ϕ ∗ (y) describes incompressible ﬂow from the duct.
. If the source coordinate y1 → −∞ within the duct. weak monopole sound generated by a source near the duct entrance. X 2 (x). y. The amplitude of this monopole is of the same order as the usual dipole sound determined by the compact Green’s function. ∂ x1 where ∂ϕ =0 ∂ xn on 2 2 x1 + x2 1 2 = a. Use the compact Green’s function to solve 2 ∇ 2 + κ0 ϕ = f 1 ˆ ∂ [δ(x1 − L)δ(x2 )δ(x3 )].Problems 3 79 with distance x1  or y1  into the duct). Y2 (y). rigid sphere. ϕ ∗ (y) account for the additional. The terms ϕ ∗ (x). For a uniform. 2. 4πx (3. X 3 (x)). the source compresses the ﬂuid in the duct mouth producing a sound wave in the duct whose reaction on the mouth causes a volume ﬂux equal to the monopole source strength. for the sound radiated by an azimuthally orientated dipole adjacent to a compact. This limiting form of G can be used to calculate the lowfrequency free space radiation generated by internal sources far from the entrance. for the sound radiated by a radially orientated dipole adjacent to a rigid circular cylinder of compact cross section. x → ∞. G(x.16) This represents a monopole wave centered on the duct entrance. thinwalled cylindrical duct we can take X(x) ≡ (x1 − ϕ ∗ (x). Y3 (y)). Use the compact Green’s function to solve 2 ∇ 2 + κ0 ϕ = f 2 ˆ ∂ [δ(x1 − L)δ(x2 )δ(x3 )] . Problems 3 1. t − τ ) = 1 δ(t − τ − (x − y1 )/c0 ). Y(y) ≡ (y1 − ϕ ∗ (y).9. ∂ x2 where ∂ϕ =0 ∂ xn on x = a.
∂ x3 4. t) = q0 δ(x1 −U t)δ(x2 )δ(x3 ).1) the far ﬁeld acoustic pressure p = −ρ0 ∂ϕ/∂t given that M = U/c0 1.5 at constant.1) is given by ϕ≈ f 2 Ma 2 2πx(h 2 + U 2 [t]2 )3 x1 h x2 U [t] (3U 2 [t]2 − h 2 ) + (3h 2 − U 2 [t]2 ) . S(ξ ) to calculate the acoustic pressure radiated from the neck during the passage of the source. 3.1. where S(x1 ) → A. t) near a rigid body is determined by the solution of ∇ 2 ϕ = q(x. Repeat Question 2 for the dipoles f2 ∂ [δ(x1 − L)δ(x2 )δ(x3 )] . 6. x x where [t] = t − 5. c0 dξ . t) = f 2 ∂ (δ(x1 − U t)δ(x2 − h)δ(x3 )). . If the crosssectional area of the duct is denoted by S(x1 ). S(ξ ) Y1 = A 0 y1 dξ .9. In incompressible ﬂow the velocity potential generated by a distribution of sources q(x. The constant strength dipole q(x. ∂ x2 f 2 = constant translates at constant velocity U past a ﬁxed rigid cylinder of radius a < h whose axis coincides with the x3 axis. The point source q(x.1. x1 → ±∞. The volume source q(x. ∂ x2 f3 ∂ [δ(x1 − L)δ(x2 )δ(x3 )]. Determine from Equation (3. use the approximations X1 = A 0 x1 x . (q0 = constant) convects along the axis of symmetry of the necked duct shown in Fig.80 3 The Compact Green’s Function 3. the farﬁeld acoustic potential determined by Equation (3. 7. t) = q0 δ(x1 − U t)δ(x2 − h)δ(x3 ). low Mach number speed U . t). q0 = constant translates at constant velocity U past a ﬁxed rigid sphere of radius a < h whose center is at the origin. Show that when M = U/c0 1.
7) for the force exerted on the ﬂuid by the body. 3π c0 x ∂t 2 cos θ = x1 . 4π X − Y 8. 9. Show that the velocity potential of the ﬂuid motion is = U j ϕ∗. A rigid body translates without rotation in the j direction at velocity U j (t) in an ideal.8. t) ≈ 2ρ0 a 3 cos θ ∂ 2 U (t − x/c0 ). the pressure can be calculated from Bernoulli’s equation: p ∂ 1 + + (∇ )2 = 0. x → ∞. If ϕ1 (x) = ∓(2/π ) a 2 − x2 − x3 on the 2 2 faces x1 = ±0. j In a ﬁxed reference frame.Problems 3 81 Show that the monopole and dipole components of the solution at large distances from the body (in the hydrodynamic far ﬁeld) can be calculated using the following incompressible limit of the compact Green’s function G(x. t − τ ) = −δ(t − τ ) . x . In the undisturbed state it lies in the plane ∗ 2 2 x1 = 0 with its center at the origin. ∂t ρ0 2 Use these results to prove formula (3. A compact rigid disc of radius a executes small amplitude vibrations at velocity U (t) normal to itself. incompressible ﬂuid at rest at inﬁnity. show that the acoustic pressure generated by the motion is given by p(x. x2 + x3 < a of the disc. y.
inviscid) incompressible ﬂow produced by arbitrary motion of a solid body with surface S (Fig. .1. where ∇ 2 ϕ = 0.1. t) = ∇ϕ. There are no sources within the instantaneous region V occupied by the ﬂuid.1) where the divergence theorem has been used to obtain the second line (there is no contribution from the surface at inﬁnity in Fig.1 Vorticity and the Kinetic Energy of Incompressible Flow 4. To understand this consider the ideal (i. This formula implies that if S is suddenly brought to rest (Un → 0) the motion everywhere in the ﬂuid ceases instantaneously. because V (∇ϕ)2 d 3 x 82 .4 Vorticity 4.e. = − ρ0 ϕ 2 2 S ∂ xn S (4. He gave the following deﬁnition of a vortex in a homogeneous incompressible ﬂuid. ∂ϕ = Un ∂ xn on S. the kinetic energy T0 of the ﬂow is T0 = 1 1 ρ0 (∇ϕ)2 d 3 x = ρ0 (div(ϕ∇ϕ) − ϕ∇ 2 ϕ) d 3 x 2 2 V V ∂ϕ 1 1 d S ≡ − ρ0 ϕUn d S. a portion of ﬂuid having any motion that it could not acquire by ﬂuid pressure transmitted from its boundary.1).. 4.1. The motion generated from rest by ‘ﬂuid pressure transmitted from its boundary’ can be described by a velocity potential ϕ such that v(x.1. .1. 4. Therefore.1 Kelvin’s (1867) Deﬁnition Kelvin was responsible for much of the pioneering work on the mechanics of incompressible ﬂow. where Un is the normal component of velocity on S. . where ϕ ∼ 2 O(1/x )).
1. the kinetic energy becomes T = 1 1 ρ0 (∇ϕ + u)2 d 3 x = ρ0 ((∇ϕ)2 + 2∇ϕ · u + u 2 ) d 3 x 2 2 V V 1 div(ϕu) d 3 x = ρ0 ((∇ϕ)2 + u 2 ) d 3 x + ρ0 2 V V 1 2 2 3 = ρ0 ((∇ϕ) + u ) d x − ρ0 ϕu n d S 2 V S 1 1 1 u 2 d 3 x ≡ T0 + ρ0 u 2 d 3 x. This unphysical behavior is never observed in a real ﬂuid because (i) no ﬂuid is perfectly incompressible. 4. and signals generated by changes in the boundary conditions propagate at the ﬁnite speed of sound.1 Vorticity and the Kinetic Energy of Incompressible Flow 83 Fig.4.2) = − ρ0 ϕUn d S + ρ0 2 2 2 S V V . and (ii) diffusion of vorticity from the boundary supplies irrecoverable kinetic energy to the ﬂuid. (4.1.1. If ϕ is taken to be deﬁned as above (for ideal ﬂow) then. For an incompressible. real ﬂuid we write v = ∇ϕ + u. because div u = 0 and the normal component u n = n · u = 0 on S. can vanish only if ∇ϕ ≡ 0. and deﬁne the vorticity ω by ω = curl u ≡ curl v.
and determines the motion that persists in an incompressible ﬂuid when the boundaries are brought to rest. the distortion corresponds to a deformation into an ellipsoid whose principal axes correspond to the principal axes of ei j . In other words. A vortex line is tangential to the vorticity vector at all points along its length. vortical motions proceed irrespective of whether or not the ﬂuid continues to be driven by moving boundaries or other external agencies. the ﬂow described by the velocity potential ϕ stops instantaneously. For a tube of small crosssectional area δS the product . being a gradient. The vorticity may therefore be regarded as a measure of the angular momentum of a ﬂuid particle. If a spherical ﬂuid particle is suddenly solidiﬁed without change of angular momentum. 2 2 where ei j is the rateofstrain tensor (2. it will rotate at angular velocity ω/2. The crucial difference between rotational and irrotational ﬂows is that. the irrotational motion represents the least possible disturbance that can be produced in the ﬂuid by the moving boundary. inasmuch as the conservation of angular momentum suggests that vorticity is associated the intrinsic kinetic energy of the ﬂow. 2 represents an irrotational distortion of the ﬂuid in the neighborhood of A. If we consider a small spherical ﬂuid particle with center at A. the last term. 4. but that associated with the rotational velocity u persists.1. once established. so that ω may be deﬁned as twice the initial angular velocity of the solid sphere when an inﬁnitesimally small sphere of ﬂuid with center at A is suddenly solidiﬁed without change of angular momentum (but this is not true for arbitrarily shaped volume elements).2) also establishes Kelvin’s theorem that T ≥ T0 : the kinetic energy of the real ﬂow (for which u = 0) always exceeds that of the corresponding ideal.2 The Vorticity Equation Let v A denote the ﬂuid velocity at a point A at x.84 4 Vorticity When the surface motion is arrested. Vortex lines that pass through every point of a simple closed curve deﬁne the boundary of a vortex tube. 1960) v B ≈ v A + (δx · ∇)v = v A + 1 ω ∧ δx + 1 ∇(ei j δxi δx j ). consisting of a translation at velocity v A together with a rotation at angular velocity 1 ω. The ﬁrst two terms on the second line represent motion of A and B as a rigid body. Equation (4.1.3) and the gradient in the second line is taken with respect to δx. The velocity v B at a neighbouring point B at x + δx can then be written (Goldstein. This is consistent with our conclusion above regarding kinetic energy. irrotational ﬂow.
2. The vector ω ∧ v is sometimes called the Lamb vector. The noslip condition (Batchelor 1967) requires the velocity at a boundary to be the same as that of the boundary. and is constant because div ω = div(curl v) ≡ 0. we ﬁrst use the identity curl curl A = grad div A − ∇ 2 A to write the momentum equation (1.2.2 The Vorticity Equation ω δS is called the tube strength. ∂t 3 where B= dp 1 2 + v ρ 2 (4. In the absence of body forces F. 3 . (4.4) (4.3) for homentropic ﬂow in the form ∂v + (v · ∇)v + ∇ ∂t By using the vector identity (v · ∇)v = ω ∧ v + ∇ 1 2 v 2 dp ρ 4 = −ν curl ω − ∇(div v) . A vortex line must therefore form a closed loop.2.1) where Ω is the angular velocity of S.3) is the total enthalpy in homentropic ﬂow.4. or end on a rotating surface S at which n · ω = 2n · Ω. We shall show that vorticity is transported by convection and molecular diffusion. (4. and in particular for the surface formed by two cross sections of the tube and the tube wall separating them. ∂t in which case dissipation occurs only where ω = 0. on the latter of which ω · dS = 0.5) .2.2. Therefore an initially conﬁned region of vortex loops can frequently be assumed to remain within a bounded region. When the ﬂuid is incompressible (or when the term in div v representing the small effect of compressibility on viscous dissipation is neglected) Crocco’s equation reduces to ∂v + ω ∧ v + ∇ B = −ν curl ω.2. (4. 85 and the divergence theorem therefore implies that ω · dS = 0 for any closed surface.2) the momentum equation can be cast into Crocco’s form ∂v 4 + ω ∧ v + ∇ B = −ν curl ω − ∇(div v) . It follows that vortex tubes and lines cannot begin or end within the ﬂuid.
therefore the vorticity equation can also be written Dω = (ω · ∇)v + ν∇ 2 ω. 4. Thus. the ﬂuid particles and the vortex line through A and B have deformed and convected in the ﬂow in the Fig. the term (ω · ∇)v in the vorticity equation causes the vorticity of the ﬂuid particle initially at A to change from ωn at A to ω(n + (n · ∇)v δt) at A .2. During this time. and consider a neighboring particle at B a small distance s from A in the direction n. Let the vorticity at A be ω = ωn. that is. Consider a ﬂuid particle at A in Fig.86 4 Vorticity The curl of (4. Dt (4.1 with velocity v at time t. 4.1.2. A has moved a vector distance v δt to A and B has moved to B whose position relative to A is s(n+(n · ∇)v δt). At time t the points A and B lie on the vortex line through A. the vortex line through A lies along the relative vector s(n + (n · ∇)v δt) from A to B .2.2. where n is a unit vector.7) (4. and the velocity at B is v + s(n · ∇)v. Therefore.3) and the relation curl curl ω ≡ −∇ 2 ω yield the vorticity equation for a Stokesian ﬂuid of constant shear viscosity: ∂ω + curl(ω ∧ v) = ν∇ 2 ω.2. and curl(ω ∧ v) ≡ (v · ∇)ω + ω div v − (ω · ∇)v − v div ω = (v · ∇)ω − (ω · ∇)v. sn is the position of B relative to A. After a short time δt.6) The terms on the right represent the mechanisms that change the vorticity of a moving ﬂuid particle in incompressible ﬂow: (i) (ω · ∇)v. ∂t For an incompressible ﬂuid div ω = div v = 0. .
the crosssectional area δS decreases and therefore ω must increase to preserve the strength of the tube. they are rotated and stretched in a manner determined entirely by the relative motions of A and B. In the absence of viscosity (when ν∇ 2 ω does not appear on the right of (4. where S is any twosided surface bounded by C.2 The Vorticity Equation 87 same way. but the amount of the vorticity available for diffusion from the surface is independent of ν. and viscosity is responsible for its diffusion into the body of the ﬂuid. ρ 2 (4. The rate of diffusion would depend on the value of ν. It follows that vorticity can neither be created nor destroyed in a body of inviscid and homentropic ﬂuid. The equation then reduces to the classical diffusion equation ∂ω = ν∇ 2 ω. in particular near solid boundaries.7) can be neglected. In an ideal ﬂuid the slipping of the ﬂow over the surface creates a singular layer of vorticity at the surface called a vortex sheet whose strength is determined by the tangential velocity difference between the surface and the ideal exterior ﬂow.4. . where it can subsequently be convected by the ﬂow.2. When a vortex tube is stretched. It should be understood that viscosity merely serves to diffuse the vorticity into the ﬂuid from the surface. (ii) ν∇ 2 ω: Molecular Diffusion of Vorticity This term is important only in regions of high shear.2.8) This is Kelvin’s circulation theorem. The circulation with respect to a closed material contour C is deﬁned by = C v · dx = S curl v · dS ≡ S ω · dS. in their new positions A and B continue to lie on the same vortex line. When ν = 0 the motion in homentropic ﬂow evolves in such a way that the circulation around the moving contour remains constant: D D = Dt Dt v · dx = C C ∇ − dp 1 2 + v · dx ≡ 0.7)) vortex lines are therefore said to move with the ﬂuid. Very close to a stationary wall the velocity becomes small and nonlinear terms in the vorticity equation (4. ∂t Vorticity is generated at solid boundaries.2. and does not generate the vorticity. The magnitude of ω increases in direct proportion to the stretching of vortex lines. This vorticity stays on the surface. it would start to diffuse into the ﬂuid if the ﬂuid were suddenly endowed with viscosity.
but the tangential components are discontinuous (n ∧ v− = n ∧ v+ ). ω = ∇ H ∧ (v+ − v− ) = n ∧ (v+ − v− )δ(s⊥ ). it can be assumed that curl v± = 0.2. (4.2) across which the velocity changes rapidly from v− to v+ .9) where ∇ H ≡ ∇ H ( f ) = −∇ H (− f ) = nδ(s⊥ ). 2 where v± are evaluated just above and below the sheet.2. where div A = 0. t) = 0 with unit normal n across which the normal components of velocity are equal (n · v− = n · v+ ). which is v = 1 (v+ + v− ). 4.2.88 4 Vorticity Fig. on either side. Near < the sheet. (4. The equations determining ϕ and A are found by taking in turn the divergence .1 Vortex Sheets A vortex sheet is a useful model of a thin layer of vorticity when viscous diffusion can be neglected. and s⊥ is distance measured in the normal direction from the sheet. In an ideal ﬂuid the sheet is subject only to convection and stretching by the ﬂow at the local mean velocity. Let f > 0 respectively on the ± sides of the surface.2.3 The Biot–Savart Law In an unbounded ﬂuid the velocity v can always be expressed in terms of scalar and vector potentials ϕ and A such that v = ∇ϕ + curl A. and we can set v = H ( f )v+ + H (− f )v− Hence. We approximate the layer by a surface f (x. Imagine a thin shear layer (Fig.2.10) 4. 4. In a real ﬂuid the vorticity would diffuse out from the sheet and it could not therefore persist indeﬁnitely.2. 4.
t)) d 3 y = 0. when κ0 = 0) to obtain A= ω(y. 4π x − y The velocity is then given by the Biot–Savart formula v(x. t) ≈ curl xj 4π x3 y j ω(y. so that it may be assumed that ω → 0 as x → ∞. the divergence theorem also implies that div(yi y j ω(y. t) d 3 y = − yi ω j (y. 4π x − y (4. (4. 89 We can take ϕ = 0 for incompressible.2) Furthermore. By using this result and the expansion (1.6). t) ≈ curl curl I 4π x = grad div where I = I . 4π x 1 2 y ∧ ω(y.3 The Biot–Savart Law and curl (using the formula curl curl A = grad div A − ∇ 2 A): ∇ 2 ϕ = div v. t)n j d S(y) ≡ 0. (4. ∇ 2 A = −curl v ≡ −ω. where the surface (with inward normal n) is large enough to contain all the vorticity.3. t) d 3 y .3.6) (i.9. and therefore that (yi ω j (y.. by the incompressible limit of (3. t)) d 3 y = 0. x → ∞. To ﬁnd A we use the Green’s function for Laplace’s equation determined by (1.3.e. t) d 3 y.1) the following approximation in the hydrodynamic far ﬁeld: v(x.3) for x → ∞ we derive from (4.2) in either of the following equivalent forms v(x. Because vorticity is transported by convection and diffusion.3.4.2. The divergence theorem then shows that ωi (y. This can be used to express (4.3) . t) = curl ω(y. t) d 3 y ∼ O 1 x3 . x → ∞. t) + y j ωi (y.1) This is a purely kinematic relation between a vector v that vanishes at inﬁnity and ω = curl v.4. t) d 3 y . an initially conﬁned region of vorticity will tend to remain within a bounded domain. unbounded ﬂow which is at rest at inﬁnity.3.
6) where U0 = dx0 /dt is the velocity of its center of volume x0 (t). These formulae supply alternative representations of v in the hydrodynamic far ﬁeld (where the motion is entirely irrotational) in terms of either the vector potential A = curl (I/4πx) or the scalar potential ϕ = div (I/4π x). The vector I is called the impulse of the vortex system. But H ( f )v + H (− f )U = ∇ϕ + curl A.2 Incompressible Flow with an Internal Boundary Let the rigid body in Fig.3.90 4 Vorticity (see Question 2 of Problems 4).4) The following representation can also be derived (using (4.3. In the usual way let f (x. (4. . in both the ﬂuid and solid (where it equals U(x. t)). x − y (4.1 Kinetic Energy Using the Biot–Savart formula it can be veriﬁed that the kinetic energy of an unbounded (threedimensional) incompressible ﬂow is given in terms of the vorticity by T = ρ0 8π ω(x.3.3. Now. 1986) uses G. and is an absolute constant in an unbounded ﬂow (see Section 4. and Ω(t) is its angular velocity.) 4. t) · ω(y. t) 3 3 d x d y. Then H ( f )v + H (− f )U is the velocity everywhere. with f > 0 in the ﬂuid.3.1.2. The body has constant volume (div U = 0). but curl U = 2Ω.4). 4. (Batchelor (1967) denotes I by P/ρ0 . t) vanish on S. t) d 3 x.1 have volume ﬂuid at rest at inﬁnity with velocity and move in an incompressible U = U0 + Ω ∧ (x − x0 (t)). Lighthill (1978. (div A = 0). the noslip condition on S implies that div(H ( f )v + H (− f )U) = ∇ H ( f ) · (v − U) ≡ 0 curl(H ( f )v + H (− f )U) = H ( f )ω + H (− f )2Ω.5) 4. (4.2)) T = ρ0 x · (ω ∧ v)(x.
This formula predicts that v = U when x lies in the region occupied by the body. . 4π x − y (4. The formula is also applicable in inviscid ﬂow.3 The Biot–Savart Law 91 Hence. This ﬂow persists only while the piston is moving. In an ideal ﬂuid the motion outside is irrotational and resembles at large distances from the exit a radially symmetric source ﬂow. and that the air is forced out at constant speed V by impulsive movement of the piston over an axial distance L. Vortex lines may be imagined to continue into the solid. Suppose the tube has radius R. where a puff of air is ejected from the tube and directed at the ﬂame of a candle.3. ϕ ≡ 0.1.4.3. Similarly. t) d 3 y + curl 4π x − y 2Ω(t) d 3 y . the asymptotic representations (4. but the contribution from the bound vorticity in the vortex sheet on the surface of the body must be included in the integrals. 4. and the velocity everywhere is given by the following modiﬁcation of the Biot–Savart formula (4.1): v(x.3.3.7) where V is the volume occupied by the ﬂuid. 4.1. As for an unbounded ﬂow. the identity curl (H ( f )v + H (− f )U) d 3 x = 0 implies that v ∼ O(1/x3 ) as x → ∞.3.3.3) remain valid provided the integration includes the region occupied by the body (where ω = 2Ω). t) = curl V ω(y.3 Blowing Out a Candle (Lighthill 1963) An amusing illustration of the signiﬁcance of vorticity is depicted in Fig. during which time the velocity potential at a large distance r from the Fig. 4.
The following numerical estimates conﬁrm this conclusion.2R0 is the radius of its core (assumed to be of circular cross section).12 cm.13 s.5 cm. 4 where R0 ∼ 1.0007 m/s. Take V = 10 m/s.3 m.2 m/s. and t ≈ 0. . In addition.1)) at speed estimated by Kelvin to equal Vt ∼ 4π R0 ln 8R0 σ − 8R0 1 VL ln ≈ 4 8π R0 σ − 1 . Vj ≈ 4. According to this sequence of events. however.6 cm. In its absence the ﬂame would barely ﬂicker under the inﬂuence of blowing by the potential velocity ﬁeld Vϕ . vorticity leaves the tube within a circular cylindrical vortex sheet at the periphery of the slug. and they translate at the local mean air velocity on the sheet equal to 1 V . whose displacement from the tube forces the potential ﬂow ϕ. 2 Shortly after leaving the tube the cylindrical vortex rolls up to form a vortex ring of circulation which translates by selfinduction (as determined by the Biot–Savart law (4. the candle is only blown out because of the presence of the vortex. The circulation of these rings per unit length of the jet is V . 4r so that the air blows against the ﬂame at distance at speed Vϕ ∼ V R2 .3. vorticity is generated at the tube wall. and the exiting ﬂuid is initially contained within a cylindrical slug of air of length L.4 m/s. R = 0.2R is the radius of the vortex ring and σ ∼ 0. Vt ≈ 2. σ ≈ 0. The air on the axis of the vortex ring at its center forms a localized jet with velocity on the centerline equal to Vj ∼ VL . across which the axial velocity changes from V within the jet to 0 outside. Then. The air leaves the tube in the form of a jet. 4R0 If the ﬂame is extinguished it is because the vortex jet blows away the hot combusting gases from newly vaporized wax.92 4 Vorticity exit resembles that produced by a monopole of strength q = π R 2 V : ϕ∼− V R2 . The sheet may be pictured as a succession of vortex rings of radius R and inﬁnitesimal core radii. The total circulation ejected from the tube 2 during the time L/V in which the piston moves is therefore = 1 L V . and = 0. The ring arrives at the ﬂame after a time t ∼ /Vt . Vϕ ≈ 0. R0 ≈ 0. 4 2 In reality. L = 1 cm.
let F = external force applied to the body to maintain its motion.1. (4. t) d 3 x − V+ p(x. t) is deﬁned by the Biot–Savart integral (4.3). 4. t) dS m dU = F −F dt where U is the velocity of the centre of mass of the body. which includes the region occupied by the body. t) = curl A.4 Surface Force in Incompressible Flow Expressed in Terms of Vorticity 93 We now derive the following formula for the force F exerted on an incompressible ﬂuid by a rigid body with surface S whose centre of volume has velocity U0 : F + m0 ρ0 d dU0 dI = ρ0 ≡ dt dt 2 dt x ∧ ω(x. and m 0 is the mass of ﬂuid displaced by the body. The ﬂuid is assumed to be at rest at inﬁnity. t) d 3 x = F + V p(x. t) d 3 x.4. we can also write F + m0 dU0 d = ρ0 dt dt v(x. 4π x . Then.4 Surface Force in Incompressible Flow 4.1. The center of volume of the body is assumed to be in motion at velocity U0 (t). t) d 3 x = V+ V+ curl A d 3 x = − n ∧ A dS → − n ∧ curl I(t) d S. t) dS.3. so that v(x.1) where I is the impulse deﬁned by the integral in (4.3) implies that A = curl(I(t)/4πx) on . m = mass of the body. but in general the body may also be rotating at some time dependent angular velocity Ω. where the vector potential A(x.2) Now let v(x. (4.3. The global equations of motion are m d dU + ρ0 dt dt v(x.4. Subtracting these equations and extending the volume integral to include the volume occupied by the rigid body (where v d 3 x = U0 ).3. including any contributions from bound vorticity within S. With reference to Fig. and let V+ denote the interior of including the volume of the body.4.1) taken over V+ . Let V denote the ﬂuid between a large closed surface containing all of the vorticity and the surface S of the body. (4.
. ideal incompressible ﬂuid in the absence of vorticity (Fig. the integral is therefore just equal to I(t).2) reduces to the desired representation (4. We have seen previously (Section 3. 4. dt Fig. 4π x The righthand side of (4. and the integrand equals I(t)/4π R 2 . On the sphere n = −x/x. − p(x. p → −ρ0 ∂ϕ/∂t on .3)).4.4.4. 4. (4.1. Thus.94 4 Vorticity Similarly. 4. but arbitrary surface can be replaced by an integration over the surface of a large sphere x = R. Hence. where ϕ = div(I(t)/4π x) (see (4.1).4.4. the integral in this expression over the large. t) dS → ρ0 d dt ϕn d S = ρ0 d dt n div I(t) d S. By the divergence theorem.2) can therefore be written ρ0 d dt −n ∧ curl I(t) 4π x + n div I(t) 4π x d S.4.1 Bound Vorticity and the Added Mass Consider the particular case of a rigid body accelerating without rotation at velocity U(t) in an otherwise unbounded. because {curl curl − ∇div}(I(t)/4πx) = −∇ 2 (I(t)/4π x) ≡ 0 for x > 0.8) that force exerted on the ﬂuid can be written Fi = Mi j dU j .3.1).
To calculate the bound vorticity we need an expression for the velocity everywhere in space. with f > 0 in the ﬂuid and f < 0 within S.3) therefore imply that Fi = −m 0 dUi d + ρ0 dt dt dUi d + ρ0 = −m 0 dt dt U j (x j − ϕ ∗ ) j ∂H 3 d x ∂ xi S U j (x j − ϕ ∗ )n i d S. j and the vorticity is ω = curl{U j H ( f )∇ϕ ∗ + U j H (− f )∇x j } j = −U j ∇ H ∧ ∇(x j − ϕ ∗ ) j ≡ curl {U j (x j − ϕ ∗ )∇ H }. The identities x ∧ curl A = 2A + ∇(x · A) − ∂ (x j A).6). where v ≡ U. we can take v = U j ∇ϕ ∗ . ∂x j (·)∇ H d 3 x = S (·) dS (4.4. This result will now be derived from the integral formula (4. the required formula for the velocity is v = U j H ( f )∇ϕ ∗ + U j H (− f )∇x j .4.1). by introducing a control surface f (x. In the ﬂuid. j Now the vector A = U j (x j − ϕ ∗ )∇ H vanishes except on the surface S of the j body.1) accordingly gives the force in the form F = −m 0 dU ρ0 d + dt 2 dt x ∧ curl{U j (x j − ϕ ∗ )∇ H } d 3 x. j Therefore.4 Surface Force in Incompressible Flow 95 where Mi j is the added mass tensor (3. but the slipping of the ideal ﬂow over S generates a singular distribution (a vortex sheet) of bound vorticity that must be used to evaluate the integral. . j But ρ0 S x j n i d S = ρ0 δi j ≡ m 0 δi j . j (where ∇ H = ∇ H ( f ) = −∇ H (− f )) Equation (4. including the region occupied by the body.4.4. t) = 0 that coincides with the surface S of the body.8. There is no vorticity in the ﬂuid.
4. To derive this formula from (4. Using the formula ∂H DH ≡ + v · ∇ H = 0. the ﬁnal term arises from frictional effects on S. Dt ∂t . j where Mi j is the added mass tensor (3.4) where vrel is the ﬂuid velocity relative to the translational velocity of S. X i represents the velocity potential of an ideal ﬂow past S that has unit speed in the i direction at large distances from S (it depends on t because a ﬁxed coordinate system is being used). t) is the Kirchhoff vector already encountered in the deﬁnition of the compact Green’s function (Section 3. any attempt to recast dI/dt must be strongly inﬂuenced by both the shape and motion of S. 4. The contribution from free vorticity is furnished by the volume integral. associated with the added mass. We consider only the important special case of a body in translational motion without rotation at velocity U(t).1) we introduce the usual control surface f (x.4) represents the inviscid component of the force. and Mi j = M ji = −ρ0 S n j ϕi∗ d S is the added mass tensor.8). Therefore. Thus.4.4. and take the curl of the resulting equation. X i = xi − ϕi∗ (x.2 Force Exerted on an Incompressible Fluid by a Moving Body The integral in (4.4. and ρ0 S − ϕ ∗ n i d S = Mi j . force on ﬂuid in irrotadU tional ﬂow ≡ Fi = Mi j dt j . and therefore the contribution to the volume integral from vorticity close to and on S is negligible. which are relatively small at large Reynolds numbers.4). The ﬁrst term on the right of (4. vrel = v − U. because ∇ X i and the relative Lamb vector ω ∧ vrel are orthogonal on S.8. t) = 0 enclosing S.96 where 4 Vorticity = volume contained by S.1) deﬁning the value of dI/dt can be transformed to remove the strong dependence of the integrand on the bound vorticity on S. with f > 0 in the outer ﬂuid region. This vorticity is produced both by motion of S and by relative motion between S and the ﬂuid induced by free vorticity in the ﬂow. indeed. (4.5) by H ≡ H ( f ). and show that the ith component of the force F exerted on the ﬂuid can also be written Fi = Mi j dU j − ρ0 dt ∇ X i · ω ∧ vrel d 3 x − η V S ∇ X i · ω ∧ dS.4. even in the inviscid limit there is no contribution to the integral from the surface vortex sheet forming the bound vorticity. multiply Crocco’s homentropic momentum equation (4.2. Now vrel = 0 on S.
the divergence theorem shows that S Bn i d S = − V div(∇ϕi∗ B) d 3 x ≡ − V ∇ϕi∗ · ∇ B d 3 x. ∇ 2 ϕi∗ = 0.2. d dt x ∧ ω(x. because ω = 0 within S.3). The ﬁrst and last integrals on the right are transformed further by the divergence . Then.4. adopting sufﬁx notation. Hence.4.4. we ﬁnd ∂ ∂ (H ω) = − (∇ H ∧ U) − curl((∇ H · U)U) − ∇ H ∧ ∇ B ∂t ∂t − curl(H ω ∧ v) − ν curl(H curl ω). Then. The surface integral S Bn i d S can be eliminated by recalling that ∂ϕ ∗ ∂ϕi∗ ≡ n j i = ni ∂ xn ∂x j on S. using Crocco’s equation (4.4 Surface Force in Incompressible Flow and the noslip condition on S.5) ∇xi · ω ∧ dS. Thus.5) Bn i d S = S V div ϕi∗ ∂v d 3x + ∂t V ∇ϕi∗ · ω ∧ v d 3 x −ν V div(∇ϕi∗ ∧ ω) d 3 x. where the last line follows by use of the identities (4. t) d 3 x = =− − −ν =2 x∧ d dt x ∧ (H ω) d 3 x = x∧ ∂ (H ω) d 3 x ∂t 97 ∂ (∇ H ∧ U) d 3 x − ∂t x ∧ curl ((∇ H · U)U) d 3 x x ∧ curl (H ω ∧ v) d 3 x x ∧ (∇ H ∧ ∇ B) d 3 x − x ∧ curl (H curl ω) d 3 x dU +0+2 dt B dS − 2 S V ω ∧ v d 3 x − 2ν S ω ∧ dS. d Ii 1 d ≡ dt 2 dt (x ∧ ω)i d 3 x = dUi + dt −ν S Bn i d S − S V ∇xi · (ω ∧ v) d 3 x (4. because ∇ϕi∗ ∼ O(1/x3 ) as x → ∞.
4) is the force necessary to accelerate the added mass of the body.8.4.4. d 3x = ∂t ρ0 dt where Mi j = M ji = −ρ0 For the last −ν V n j ϕi∗ d S is the added mass coefﬁcient of (3. we ﬁnd ∇ X i · ω ∧ v d 3x − ν V S dUi Mi j dU j + − dt ρ0 dt ∇ X i · ω ∧ dS. div(∇ϕi∗ ∧ ω) d 3 x = ν S S Thus.6) can also be written ∇ X i · ω ∧vrel d 3 x−ν V S dUi Mi j dU j + − dt ρ0 dt ∇ X i · ω ∧dS (4. 0).3 Stokes Drag on a Sphere The ﬁrst term on the righthand side of (4. The ith component of the viscous skin friction is −η S (ω ∧ dS)i ≡ −η S ∇xi · ω ∧ dS. so that (4.7) into (4.7) where vrel = v − U.4. Thus. for the ﬁrst div ϕi∗ S 4 Vorticity V Mi j dU j ∂v . U > 0. and is produced by the pressure ﬁeld established by the surface shear stress. Let the sphere have radius a and translate at constant velocity U = (U.4. At very small Reynolds numbers Re = aU/ν 1 .4.4.98 theorem.4.4. The second. The necessity for such a term is vividly illustrated by the Stokes drag on a sphere.6). (4. along the x1 axis.4) is now obtained by substituting from (4. ∇ϕi∗ · ω ∧ dS.4. viscous component is comparable in magnitude to the skin friction. (because X i = xi − ϕi∗ ) the net contribution of the normal pressure forces on S is represented in (4. Equation (4. substituting for d Ii = dt Bn i d S in (4.1) (recalling that m 0 = ρ0 ).6) But the identity ∇ X i · ω ∧ U = div(U(v · ∇ X i ) − v(U · ∇ X i ) − (v · U)∇ X i ) implies that d Ii = dt V ∇ X i · ω ∧ U d 3 x = 0.5).4) by the terms −ρ0 V ∇ X i · ω ∧ vrel d 3 x + η S ∇ϕi∗ · ω ∧ dS.4. 4. 0.
The value of C is most easily found by substituting this expression for ω into the Biot–Savart formula (4.9) The net force F1 on the ﬂuid is in the x1 direction.8) in a reference frame moving with the sphere. By symmetry p must vary linearly with η and U · x.2.4 Surface Force in Incompressible Flow 99 the inertial terms ω ∧ v and ∇( 1 v 2 ) can be discarded from Crocco’s equation 2 (4. Both the pressure and the vorticity therefore satisfy Laplace’s equation ∇ 2 p = 0. Dp = −η S ∗ ∇ϕ1 · ω ∧ dS = −2π ηU a. For the sphere ∗ ϕ1 = −a 3 x1 /2x3 .4. which (for incompressible ﬂow) reduces to the creeping ﬂow equation ∇ p = −η curl ω (4. 2x (4. This yields C = 3a/2. Dp = − S pn 1 d S ≡ − S ∗ p∇ϕ1 · dS = V ∗ ∇ p · ∇ϕ1 d 3 x = −η V ∗ curl ω · ∇ϕ1 d 3 x = −η S ∗ ∇ϕ1 · ω ∧ dS. Because n 1 = n · ∇ϕ1 on S we ﬁnd. and is given by the ﬁnal integral on the right of (4.4. . and therefore ω = curl 3aU . The pressure drag is therefore equal to half the skinfriction drag. where v ≡ U.1) and evaluating the righthand side at the centre x = 0 of the sphere. where Ds and Dp are the respective components of the Stokes drag on the sphere produced by the skin friction and the viscous surface pressure. ∇ 2 ω = 0. This interpretation of Dp as the component of drag attributable to the normal pressure forces on the sphere can be conﬁrmed directly using the creeping ﬂow ∗ approximation (4.5). and the condition that p should vanish at large distances from the sphere supplies the dipole solution p = Cη U U·x ≡ −Cη div .4. and we readily calculate F1 = 6π ηU a.8).4).8) imply that ω = C curl (U/x) ≡ C(U ∧ x)/x3 (x > a).4. x > a. ω must be a linear function of U ∧ x: the identity curl curl(U/ x) = grad div(U/x) and Equation (4.4. using the divergence theorem.4. x3 x where C is a constant. and Ds = η S (ω ∧ dS)1 = −4π ηU a. Similarly. x > a.3. It is equal in magnitude to Ds + Dp .
These results have the following signiﬁcance. Y (x. y) and ψ(x. with derivative w (z) = f (z)W ( f (z)). z = x + iy is regular (analytic) in a region D of the zplane. incompressible ﬂows and its application to determine the equation of motion of a line vortex in such ﬂows. y) = (X (x. y)). The solution of Laplace’s equation within a given twodimensional bounded region D is equivalent to the solution of Laplace’s equation within the transformed region D . If it is possible to solve the latter problem.100 4 Vorticity where (4. the solution to the original problem in D can be found . Y ). the solutions and of Laplace’s equation in D are also solutions of Laplace’s equation in D. ∂ X2 ∂Y 2 The transformation Z = f (z) permits us to deﬁne a corresponding function w(z) ≡ ϕ(x. y) + iψ(x. Y (x. Let W(Z ) be regular in D with real and imaginary parts (X. ∂ X2 ∂Y 2 ∂2 ∂2 + = 0 in D . ∂ 2ϕ ∂ 2ϕ + 2 = 0. y) = W( f (z)). The real and imaginary parts ϕ(x. y). The results will be applied in later chapters to investigate simple models of sound production by vortices interacting with surfaces.5 The Complex Potential The remainder of this chapter is devoted to a brief outline of the complex potential representation of twodimensional. 4. Y ). y)). 2 ∂x ∂ y2 Let f (z) be regular in D.4.8) and the identity div(A ∧ B) = curl A · B − A · curl B have been used on the second line. ψ(x. (X. and deﬁne a (conformal) transformation Z = f (z) of D into a region D in the plane of Z = X + iY .5. ∂2 ∂2 + = 0. 4. y) + iψ(x. y). y) = (X (x. 2 ∂x ∂y ∂ 2ψ ∂ 2ψ + = 0 in D. y) are solutions of Laplace’s equation. which is regular in D. For corresponding points in D and D we have ϕ(x.1 Laplace’s Equation in Two Dimensions Suppose that w(z) = ϕ(x. y). In other words. Then.
Both ϕ(x. y). but these can usually be dealt with by careful examination of the behavior of the transformation near such points. The velocity v = ∇ϕ = (∂ϕ/∂ x. Difﬁculties may arise at isolated points where f (z) = 0 and at points where f (z) ceases to be regular. Method 1 The real and imaginary parts of every regular function w(z) determine the velocity potential and stream function of a possible ﬂow.2 Hydrodynamics in Two Dimensions Irrotational motion of an ideal. For steady motion the velocity at (x. The fact that w(z) is a regular function of z can greatly simplify the solution of many problems. A catalog of ﬂows can therefore be constructed by studying the properties of arbitrarily selected w(z). ∂y ∂x which imply that ∇ϕ · ∇ψ = 0. . i. y) and ψ(x. 0). then the complex velocity w (z) = ∂ϕ ∂ϕ −i ≡ u − iv ∂x ∂y is also regular. ∂x ∂y ∂ϕ ∂ψ =− .4. ∂ϕ/∂ y). If v = (u. v = (U. and the ﬂuid particles travel along a ﬁxed system of streamlines each of which is a member of the family of curves ψ(x. 4. y) + iψ(x. y) does not change with time.. Example 1 w = U z. Thus. v). U = real constant: ϕ = U x.e. ψ = U y. The function ψ is called the stream function.5. y) = constant. This will be illustrated by consideration of two methods based on the theory of complex variables. that the streamlines intersect the equipotentials ϕ = constant at right angles.5 The Complex Potential 101 by transforming back to the z plane. y) are solutions of Laplace’s equation and satisfy the Cauchy–Riemann equations: ∂ψ ∂ϕ = . incompressible ﬂuid in planes parallel to the x y plane can be investigated by introducing the complex potential w(z) = ϕ(x.
ϕ = U sin θ r + a2 r . The radial component of velocity a2 ∂ϕ = U cos θ 1 − 2 ∂r r vanishes at r = a. U = real constant. Example 2 w=U z+ a2 z . z > a > 0..5.5.2) describes potential ﬂow in the y direction past a rigid cylinder of radius a with center at the origin. The motion therefore represents steady ﬂow in the x direction past a rigid cylinder of radius a with centre at the origin (Fig. and the motion becomes uniform at speed U parallel to the x axis.1. (4. z > a.5. U = real constant.102 4 Vorticity The motion is uniform at speed U along streamlines parallel to the x direction.f. Fig. . c. In terms of the polar form z = r eiθ . (4. ϕ = U cos θ r + r r . Thus. a > 0. 4. w = U r eiθ + a 2 −iθ a2 e .6). Section 3. Example 3 w = −iU z − a2 z . 4.5.1.1) At large distances from the origin w → U z.
ψ = . 2π ϕ= 1 1 ln z − z 0  = ln (x − x0 )2 + (y − y0 )2 . . z = r eiθ . The streamlines are circles centered at z = 0.5. in the region y > 0 the potential also describes the ﬂow produced by a point source at z 0 adjacent to a rigid wall at y = 0 (the presence of the wall is said to be accounted for by an image source). 2π 2π is regular except at z = 0. 2π −i ln(z − z 0 ). z = r eiθ . 2π 2π 103 is regular except at z = 0. ∂r The origin is therefore a simple source of unit strength. and ∂ϕ/∂ y = 0 on y = 0. 4. The origin is a singularity of the ﬂow where ﬂuid is created at a rate equal to C ∇ϕ · n ds. and the ﬂow speed is ∂ϕ/r ∂θ = /2πr in the anticlockwise direction (for > 0). and describes the irrotational ﬂow outside a line vortex of strength concentrated at z = 0. In particular. where C is any simple closed curve enclosing the origin with outward normal n.2). where C is any contour encircling the vortex once. The ﬂow is radially outward from the origin along streamlines θ = constant.4. 2π 2π Example 5 The function w= −i ln z. and ds is the element of arc length on C. taking C to be a circle of radius r . 2π ϕ= 1 θ ln r. 2π represents the ﬂow produced by two unit point sources located at z 0 = x0 + i y0 ∗ and z 0 = x0 − i y0 (Fig. The circulation C v · dx = . When the vortex is at z 0 = x0 + i y0 w= Example 6 The function w= 1 ∗ (ln(z − z 0 ) + ln(z − z 0 )). 2π ϕ= 1 (ln r1 + ln r2 ) . at speed ∂ϕ/∂r = 1/2πr . The motion is symmetric with respect to the x axis. 2π ϕ= θ . Therefore. and the contour is traversed in the positive direction (with the interior on the left). When the source is situated at z 0 = x0 + i y0 w= 1 ln(z−z 0 ).5 The Complex Potential Example 4 The function w= 1 ln z.ψ = − ln r. ∇ϕ · n ds = C 0 2π ∂ϕ r dθ = 1.
2π 2π represents the ﬂow produced by two line vortices of circulations ± respectively ∗ at z 0 = x0 + i y0 . The stream function ψ = Im w = 0 on the x axis. The mean value of the local rotational ﬂow produced by the selfpotential of each vortex (Example 5) vanishes on the vortex axis. The transformation is usually chosen to simplify the boundary Fig. on which ∂ϕ/∂ y = 0. 4.3).5. . and cannot therefore cause it to translate.5. Method 2 The ﬂow past a system of rigid boundaries in the z plane is represented by means of a conformal transformation Z = f (z) by an equivalent ﬂow in the Z plane. Example 7 The function w= −i i ∗ ln(z − z 0 ) + ln(z − z 0 ).2. 4. Each vortex translates parallel to the wall at speed u = /4π y0 determined by the velocity potential of its image.3. In the region y > 0 the potential describes the ﬂow produced by a vortex of strength at z 0 adjacent to a rigid wall at y = 0 (which is accounted for by an equal and opposite image vortex). 4.5. z 0 = x0 − i y0 (Fig.104 4 Vorticity Fig. which is therefore a streamline of the ﬂow.
In the z plane this becomes √ √ θ w = iU z ≡ −U r sin 2 √ θ + iU r cos . Point source and vortex singularities of the ﬂow are preserved under the transformation.e. if Z = Z 0 is the image of a vortex of strength at z = z 0 . ∂r r ∂θ −U = √ 2 r θ θ sin . The polar representation of the velocity is therefore v = (vr . y = 0 in terms of polar coordinates (r. 2π The vortex in the z plane therefore maps into an equal vortex at the image point in the Z plane. The complex potential of ﬂow in the positive X direction parallel to the boundary Y = 0 in the Z plane corresponds to ﬂow around the edge of the halfplane in the clockwise sense. The streamlines of the ﬂow are the parabolas √ θ r cos 2 = constant. i. y = ±2β 1 − x . √ The transformation Z = i z maps the z plane cut along the negative real axis (so that −π < arg z < π) onto the upper half of the Z plane. 2 − π < θ < π.5 The Complex Potential 105 conditions. θ): √ θ ϕ = α r sin . where U is real. and has the general representation W = U Z . Indeed. β . 2 α = a real constant. and plot the streamlines. 2 2 This satisﬁes the rigid wall condition on the halfplane because the component of velocity normal to the wall is vθ . the complex potential in the neighborhood of Z 0 (where Z − Z 0 ≈ f (z 0 )(z − z 0 )) is determined by W(Z ) = w(z) = −i ln(z − z 0 ) + terms ﬁnite at z 0 2π Z − Z0 −i ln = + terms ﬁnite at Z 0 2π f (z 0 ) = −i ln(Z − Z 0 ) + terms ﬁnite at Z 0 .. cos . vθ ) = ∂ϕ 1 ∂ϕ . Example 8 Derive the following formula for the velocity potential of irrotational ﬂow around the edge of the rigid halfplane x < 0. which vanishes at θ = ±π.4. thereby permitting the solution in the Z plane to be found in a relatively straightforward manner.
β being a positive constant.6 Motion of a Line Vortex In twodimensional incompressible. as shown in Fig. The streamline for β = 0 corresponds to the upper and lower surfaces of the halfplane.4. When U > 0 ﬂuid particles travel along the parabolic streamlines around the edge in the clockwise direction.106 4 Vorticity Fig. 4.2. The ﬂow velocity becomes inﬁnite like 1/ r as r → 0 at the sharp edge.5. Dt A line vortex is therefore convected without change at the local velocity at its . 4. 4.5. and the vorticity equation (4.4. which maps into the streamline Y = 0 on the √ surface of the wall in the Z plane.7) reduces to Dω3 = 0. where x < β. inviscid ﬂow in planes parallel to x3 = 0 the vortex lines are all parallel to the x3 direction.
4.6. the velocity becomes inﬁnite as the core is approached because of the singular velocity induced by its selfpotential − i ln(z − z 0 ).2) this becomes d x02 i ζ (z 0 ) d x01 −i =− + F (z 0 ). ≈− 2π 2 (4.6. 2π But the rotational ﬂow around the core induced by the vortex cannot induce motion in itself.3) The real and imaginary parts of this equation supply two nonlinear ﬁrstorder ordinary differential equations for the position (x01 (t).6. In applications the complex potential w(z) usually arises in the form w(z) = − i ln(ζ (z) − ζ (z 0 )) + F(z).e. For a vortex of strength at z = z 0 (t) in the plane of z = x1 + i x2 .6 Motion of a Line Vortex 107 core. 2 where the primes denote differentiation with respect to z..2) The complex velocity of the vortex is W (z 0 ) ≡ {W (z)}z=z0 . when z − z 0  is small we have ζ (z) = ζ (z 0 ) + (z − z 0 )ζ (z 0 ) + (z − z 0 )2 ζ (z 0 ) + · · · . dt dt 4π ζ (z 0 ) (4. Thus. subtracting the selfpotential from w(z) we ﬁnd. 2π (4. In particular. i. near the vortex. F(z) are regular functions of z in the neighborhood of the vortex core at z = z 0 .6. x02 (t)) of the vortex at time t. dt dt dt Using (4. ∗ d x01 d x02 dz 0 ≡ −i = W (z 0 ). . and this potential must be removed from the complex potential w(z) before calculating the convection velocity of the vortex.1) where ζ (z). W (z) = w(z) + =− i ln(z − z 0 ) 2π i i ln(ζ (z) − ζ (z 0 )) + ln(z − z 0 ) + F(z) 2π 2π i 1 ln ζ (z 0 ) + ζ (z 0 )(z − z 0 ) + F(z).
Let the vortex at z 0 (t) map into a vortex at ζ = ζ0 (t). 2π 2π In the z plane this becomes w(z) = − i i ln(ζ (z) − ζ (z 0 )) + ln(ζ (z) − ζ ∗ (z 0 )). dt 4π ζ (z 0 ) and let h be a suitably small integration time step (which need not be constant). (4. z = x1 + i x2 . n Assume that at time tn the vortex is at z 0 (tn ) = z 0 . 6 Example 1 Calculate the trajectory of a line vortex of strength > 0 adjacent to a rigid halfplane lying along the negative real axis (x1 < 0.4) maps the ﬂuid region −π < arg z < π into the upper half Im ζ > 0 of the ζ plane (Fig.3) in the form dz 0 i ζ (z 0 ) = f ∗ (z 0 ). Fig. and then ﬁnd n+1 n z 0 = z 0 + 1 (k1 + 2k2 + 2k3 + k4 ). we evaluate n k1 = h f ∗ z 0 . in which case w= −i i ∗ ln(ζ − ζ0 ) + ln(ζ − ζ0 ). Write the equation of motion (4.1b). 1 n k2 = h f ∗ z 0 + k1 . − π < arg z < π. as described in Example 7 of Section 4. x2 = 0.6.5. 4. The velocity potential w(ζ ) of the motion in the ζ plane is found by introducing an image ∗ vortex of strength − at ζ = ζ0 (t). The integration is started from a prescribed point on the trajectory through which the vortex is required to pass. where f (z 0 ) = − + F (z 0 ).6. The time and space variables should ﬁrst be nondimensionalized with respect to convenient time and length scales deﬁned by the problem (several examples are discussed in Chapter 8).6. 2 n k4 = h f ∗ z 0 + k3 . To determine the complex n+1 position z 0 at time tn+1 = tn + h.3) numerically.6.1a).6.6.108 4 Vorticity 4. Let us consider integration by means of a fourthorder Runge–Kutta algorithm. The transformation √ ζ = i z. 2π 2π . 4. 2 1 n k3 = h f ∗ z 0 + k2 .1 Numerical Integration of the Vortex Path Equation In most cases it is necessary to integrate equation (4.
r 1 dθ = 2 cot θ dr 2 dθ =− dt 4πr 2 (4. dt 8πr 2 that is. dt dt dt 8πr Therefore. Let z 0 = r eiθ .5) . the equation of motion (4.6.1. Then the real and imaginary parts of the equation are dr dθ d x01 ≡ cos θ − r sin θ = dt dt dt 8πr sin θ + tan θ . which is of the form (4. + √ √ √ dt dt 8π z 0 4π z 0 [ z 0 + ( z 0 )∗ ] This can be integrated in closed form. Hence.6. 4.3) becomes d x02 i d x01 i −i = .6.6 Motion of a Line Vortex 109 Fig.4. dr θ =− tan .6.1). 2 dr dθ d x02 ≡ sin θ + r cos θ =− (cos θ + 1).
and proceeds along a symmetrical path below the halfplane.6. 2 = constant.6). .6.110 4 Vorticity Fig. Thus. θ = 2 tan−1 − t 8π 2 2 .7) 1 − (U t/ )2 1 + (U t/ )2 −2U t/ 1 + (U t/ )2 where U = 8π . . θ = 2 tan−1 − x02 = Ut . Collecting together these results we have r = x01 = 1+ Ut 2 .2. .2. Thus. 4. 1 r = sec θ . where time is measured from the instant at which θ = 0.6) This is the polar equation of the trajectory plotted in Fig. Substituting for r in the second of equations (4. Thus.6.6. (for > 0) the vortex starts above the halfplane at t = −∞ at x01 = −∞. It crosses the x1 axis at t = 0 at x01 = . (4. 4. The dependence of r on t is now obtained by substituting into (4. initially at speed U parallel to the plane. The constant length is equal to the distance of closest approach of the vortex to the edge of the halfplane. we ﬁnd sec2 1 θ 2 dθ =− dt 4π .6.5). (4. x02 = 2 and translates towards the edge. which occurs at θ = 0.6.
The complex potential is ob∗ tained by placing an image vortex − at the inverse point z = a 2 /z 0 together with a vortex + at the center of the cylinder.8) i a 2 e−iθ dθ d −iθ e = .6 Motion of a Line Vortex 111 Fig. Then w(z) = − a2 i i ln(z − z 0 ) + ln z − ∗ 2π 2π z0 − i ln z. Example 2: Vortex motion outside a cylinder A vortex is located at z 0 = r eiθ outside a rigid cylinder of radius a(< r ) with center at the origin (Fig. ≡ −ir e−iθ dt dt 2πr (r 2 − a 2 ) . and r Therefore. 4. There is no net circulation around the cylinder. − a2 dθ = . The two interior vortices ensure that the total circulation around the cylinder vanishes.6.3).6. 2πr (r 2 − a 2 ) (4. dt 2πr 2 (r 2 − a 2 ) and (for > 0) the vortex trajectory is a circle traversed in the clockwise direction at speed v0 = a2 .3.6. ≡ dt 2π (r 2 − a 2 ) 2π z 0 2π z 0 (r 2 − a 2 ) By multiplying by z 0 and adding the complex conjugate equation we see that r = constant. so that the equation of motion of the vortex is ∗ ∗ i z0 i dz 0 i a2 = − .1) ζ ≡ z). 2π The ﬁrst term on the right is the selfpotential of the vortex (in the notation of (4.6. 4.4.
rigid strip of width 2a. 8. 5. Use the relation (yi ω j (y. U > 0.. where the coordinate origin is taken at the center of the sphere. t)) d 3 y = 0 to show that 1 4πx3 (x · y)ω(y. 0). 7. 0. t) + y j ωi (y.3). 3. ρ 2. 4. 9.112 4 Vorticity Problems 4 1.3. −∞ < x3 < ∞. A gas bubble in water is set into translational motion at velocity U(t) by sound whose wavelength greatly exceeds the bubble radius. Assume the ﬂuid is at rest at inﬁnity and that there is no net circulation around the strip. Calculate the added mass coefﬁcients Mi j for an inﬁnite. 10. t) d 3 y = ∇ 1 4π x ∧ 1 2 y ∧ ω(y. Calculate the path of a line vortex of strength that is parallel to a rigid elliptic cylinder of semimajor and minor axes respectively equal to a and b. along the x1 axis. to deduce the Stokes drag formula D = 6π ηU a. Use the creeping ﬂow approximation ω = curl (3aU/2x). Assume the ﬂuid is at rest at inﬁnity and that there is no net circulation around the cylinder. Assume the motion is ideal and that the net circulation around the cylinder vanishes. If the acoustic particle velocity near the bubble would equal V(t) in the absence of the bubble. Show that in inviscid. when the induced component of the motion of produced by image vortices in the cylinder is neglected). 6. Calculate the added mass coefﬁcients Mi j for an inﬁnite. Repeat Question 5 under the assumption that the vortex is convected solely by the mean ﬂow (i.e.7) takes the form D Dt ω ρ = ω · ∇ v. Calculate the path of a line vortex of strength that is parallel to a rigid strip occupying −a < x1 < a. homentropic ﬂow (where div v = 0) the vorticity equation (4.2. . rigid cylinder of radius a. show that U = 3V when the mass of the air within the bubble is neglected. A rigid sphere of radius a translates at constant velocity U = (U. Deduce the formulae (4. t) d 3 y . Determine the unsteady force on the strip. Calculate the unsteady lift and drag exerted on a rigid circular cylinder of radius a produced by a parallel line vortex of circulation in the presence of a uniform mean ﬂow normal to the cylinder. x2 = 0.
13. Calculate the trajectories of a vortex pair consisting of two parallel line vortices of strengths ± moving under their mutual induction towards a rigid plane parallel to the line of centers of the vortices. Deduce that a line vortex at z = z 0 (t) traverses a path determined by the equation ∗ i dz 0 =− dt 4π 2 z0 z0 2z 0 a2 − 2 2 z0 + a2 − z0 + a2 + a2 . 12. the vortex being parallel to the edge of the corner. Show that the transformation ζ = z 2 /a 2 + 1. Show that the vortex traverses a path with polar representation r sin 2θ = constant. Calculate the trajectory of a line vortex of strength adjacent to the rigid halfplane x1 < 0.Problems 4 113 11. . provided the ﬂuid is at rest at inﬁnity. a > 0 maps the upper z plane cut by a thin rigid barrier along the imaginary axis between z = 0 and z = ia onto the upper ζ plane. x2 = 0 in the presence of a uniform mean ﬂow at speed U in the positive x1 direction. A line vortex of strength is adjacent to a rigid rightangle corner whose sides lie along the positive x1 and x2 axes. 14.
homentropic ﬂow the value of Lighthill’s quadrupole source (2. t) d 3 y . and the continuity equation becomes ∇ 2ϕ + 1 Dρ = 0. Because div u = 0. where the characteristic frequency ∼u/ . Hence. 2.2) M= u .3.1): Ti j ≈ ρ0 u i u j . Dt ρ Dt ρc Dt Hence. 4π x − y (5. and u ∼ O(1/x3 ) as x → ∞ (see Section 4. Thus. Put v = u + ∇ϕ. in order of magnitude ∇ϕ = O(u M 2 ) within the eddy.1) To examine this in more detail. the scalar potential ϕ describes compressible motions.1.3).2.2. and take the coordinate origin within the eddy. = 2 ∼ M 2.2) can be approximated by means of the Biot–Savart induction formula (4. ρ0 u 3 1 Dρ 1 Dp u Dp ∼ . u involves the whole incompressible component of velocity. 114 (5. u(x. where x ∼ .1 The Role of Vorticity in Lighthill’s Theory At low Mach numbers in unbounded. c0 . t) = curl ω(y. ρ Dt But p − p0 ∼ ρ0 u 2 in the eddy.5 Vortex Sound 5.1 consisting of vorticity of characteristic length .1. consider the acoustically compact eddy of Fig.
1) in the form p(x. t − + d 3 y.3) 2 and express the solution p(x.6) x c0 +···.1. x·y ∂ x (ω ∧ u) y.1. because u ∼ O(1/y3 ) as y → ∞. Because div v ∼ O(M 2 ) within the source region. 2 c0 c0 x When retarded time variations x · y/c0 x within the eddy are neglected the identity (5. + c0 c0 x (5. p1 (x. t).9.1. t − x x·y + c0 c0 x = (ω ∧ u) y. t) ≈ −ρ0 xi x j ∂ 2 2 4πc0 x3 ∂t 2 yi (ω ∧ u) j y. (5.1.5. t − c0 x ∂t c0 The order of magnitude of p2 (x.1.1. using (1. where.2. t) = p1 (x. t) = p2 (x.8) as x → ∞.3). t) is estimated by using the momentum equation (4.1 The Role of Vorticity in Lighthill’s Theory Now write ∂ 2 (u i u j ) = div(ω ∧ u) + ∇ 2 ∂ xi ∂ x j 1 2 u 2 115 (5.5) x·y 1 2 x u y. t − + We now ﬁnd p1 (x. To estimate the value of the integral in (5.3) and the divergence theorem imply that ω ∧ u d 3 y ≡ 0. t − x x·y d 3 y.4) it is therefore necessary to expand the integrand to the next higher approximation in the retarded time: (ω ∧ u) y. d 3y ∼ c0 x x → ∞.2. Take the scalar product with u d p 1 2 ∂ϕ + v + ρ 2 ∂t = −u · ω ∧ ∇ϕ − νu · curl ω = −u · ω ∧ ∇ϕ + ν(div (u ∧ ω) − ω2 ) . we can write ∂u +∇ ∂t ∂ ∂t d p 1 2 ∂ϕ + v + ρ 2 ∂t 1 2 u + div u 2 = −ω ∧ u − ω ∧ ∇ϕ − ν curl ω. t) = c0 (ρ − ρ0 ) of Lighthill’s equation given by (2. t) = −ρ0 xi ∂ 4πc0 x2 ∂t ∂2 ρ0 2 4πc0 x ∂t 2 (ω ∧ u)i y.6) and (1.9. t) + p2 (x. t − x ρ0 u 2 M 2 .4) (5.
5) of this equation. In the following we shall actually use the Approximation (4. x Re 1 and Re ∂ 2 (ρ0 vi v j ) ∂ xi ∂ x j 1.2) ∂ ∂t 1 2 u (y. The total enthalpy occurs naturally in Crocco’s form (4. Equation (2. The contributions from the divergence terms vanish because u( d p/ρ + 1 v 2 + ∂ϕ/∂t) tends to zero at least as fast 2 as 1/y3 as y → ∞.1. in Chapter 2.1. All viscous stresses can be ignored in a high Reynolds number source ﬂow except possibly within surface boundary layers on bodies . the principal effect of 3 this term is to attenuate the sound once it has been generated and is propagating to a distant observer in the sourcefree region of the ﬂow. t) d 3 y = − 2 (u · ω ∧ ∇ϕ + νω2 )(y. t) ∼ x ρ0 u 2 M 4 + ρ0 u 2 M 2 . (We have already seen.) Thus. where also ω = 0.7) 2 3 where Re = u /ν typically exceeds 104 in turbulent ﬂow. and therefore that p2 p1 in turbulent ﬂow where M We conclude that the component div(ρ0 ω ∧ v) of the Lighthill quadrupole is principal source of sound at low Mach numbers.1.5) that a more accurate estimate of the radiation damping is 2 u 3 M 5 .5). in place of Lighthill’s c0 (ρ −ρ0 ). when retarded time variations are neglected in (5.1. t) d 3 y ∼ 2 3 u M2 + u .3) of the momentum equation. The two terms on the righthand side nominally represent the dissipation of the turbulent motions respectively by acoustic radiation and by viscous damping.2. but is of no particular interest when studying sound generation mechanisms. Re (5. however. Hence.2 The Equation of Vortex Sound Lighthill’s equation (2. we ﬁnd p2 (x. This attenuation can be signiﬁcant in applications.116 5 Vortex Sound and integrate over the whole of space. Indeed.2. using the estimate (5. 5. in which the viscous term 4 ν∇(div v) is neglected.12) can be recast in a form that emphasizes the prominent rˆ le of vorticity in the production of sound by taking the total enthalpy o B= dp 1 2 + v ρ 2 2 as the independent acoustic variable.2.
5). we use the deﬁnition dp 1 = B − v2. B=− ∂ϕ ∂t in regions where ω = 0. (5.2. t) is the velocity potential that determines the whole motion in the irrotational regions of the ﬂuid.2) To calculate the pressure in terms of B elsewhere in the ﬂow. B is therefore constant in steady irrotational ﬂow.2.2.2.5.2. = Dt The small viscous correction can be ignored in high Reynolds number source ﬂows.1) where ϕ(x.2. the acoustic pressure is given by p = ρ0 B ≡ −ρ0 ∂ϕ .5) reduces to ∂v = −∇ B. But surface friction is dominated by the vorticity term −ν curl ω.2.2 The Equation of Vortex Sound 117 immersed in the ﬂow.2. If the mean ﬂow is at rest in the far ﬁeld. ρ ∂t Dt 5.5) by the density ρ and take the divergence div ρ ∂v ∂t + ∇ · (ρ∇ B) = −div(ρω ∧ v). ∂t (5. and at large distances from the acoustic sources perturbations in B represent acoustic waves. ρ 2 Differentiating with respect to time and using Crocco’s equation (4. In irrotational ﬂow Crocco’s equation (4.5).4) (5.1 Reformulation of Lighthill’s Equation Multiply Crocco’s equation (4.3) . which is retained in (4. where p and B can be taken to be related by 1 ∂p DB = . ∂t In other words. (5.2. we have ∂B ∂v 1 ∂p = − v· ρ ∂t ∂t ∂t ∂B − v · (−∇ B − ω ∧ v − ν curl ω) = ∂t DB + νv · curl ω.
2.2. The differential operator on the left describes propagation of the sound through the nonuniform ﬂow. as in the case of Lighthill’s equation. and there are no sound waves propagating in the ﬂuid. 5.3) has been used on the last line.118 5 Vortex Sound The ﬁrst term on the left is expressed in terms of B by using the continuity equation in the form div v = − and writing div ρ ∂v ∂t = ∇ρ · ∂ ∂v + ρ div v ∂t ∂t ∂ 1 Dρ ∂v −ρ = ∇ρ · ∂t ∂t ρ Dt = ∇ρ · = −ρ = −ρ = −ρ ∂ ∂v −ρ ∂t ∂t 1 ∂ρ ρ ∂t 1 ∂p ρc2 ∂t 1 DB c2 Dt . we obtain the desired vortex sound equation for homentropic ﬂow D Dt 1 D c2 Dt − 1 1 ∇ · (ρ∇) B = div(ρω ∧ v). 1 ∂ρ ρ ∂t − ∂v · ∇ρ − ρv · ∇ ∂t 1 ∂ρ ρ ∂t 1 Dρ . In an unbounded ﬂuid U = constant.5) The vortex source on the righthand side vanishes in irrotational regions.2. when the source region is very extensive it will not normally be permissible to neglect the interaction of the sound with the vorticity through which it propagates. and if there are no moving boundaries. the total enthalpy B is constant.2.4) and dividing by ρ.2 Sound Waves in Irrotational Mean Flow Let an irrotational mean ﬂow be deﬁned by the velocity potential ϕ0 (x). with mean velocity U = ∇ϕ0 . ρ ρ (5. the mean velocity . If acoustic waves cannot enter from inﬁnity. ρ Dt D Dt D Dt D Dt where Equation (5. Substituting into (5.5) may be identiﬁed as the analytical representation of the acoustic sources. and the righthand side of (5. it follows that the (homentropic) ﬂow can generate sound only if moving vorticity is present.2. if ω = 0 everywhere.
Dt ∂t Furthermore. ρ where c ≡ c(x) and ρ ≡ ρ(x) are the local sound speed and density in the steady ﬂow.5. 2 The linearized version of this equation describes the propagation of small amplitude sound waves determined by ϕ (x. Consider an irrotational disturbance ϕ (x. t). t).6) The coefﬁcients of the differential operators in this equation are functions of both mean and perturbation quantities. because the mean ﬂow does not depend on time. and set ϕ(x. However. It can be shown that the general. The linearized ˙ equation then becomes ∂ + U·∇ ∂t 1 c2 ∂ + U·∇ ∂t − 1 ∇ · (ρ∇) ϕ = 0. t) = ϕ0 (x) + ϕ (x.3 Vortex Sound at Low Mach Numbers When the characteristic Mach number M is small the local mean values of the density and sound speed are related to their uniform respective values ρ0 and . we can take the perturbation potential ϕ . which becomes D Dt 1 D c2 Dt − 1 ∇ · (ρ∇) ϕ = 0.2. rather than ϕ. but the linearized equation is obtained merely by replacing these coefﬁcients by their values in the absence of the sound. nonlinear equation satisﬁed by ϕ is 1 ∂ 2ϕ 1 D + 2 2 ∂t 2 c c Dt 1 ∂ 1 (∇ϕ)2 + 2 2 c ∂t 1 (∇ϕ)2 − ∇ 2 ϕ = 0. when ω = 0 the linearized ˙ equation for B = −∂ϕ /∂t ≡ −ϕ is more easily derived from (5. In homentropic ﬂow the mean density and sound speed can be expressed in terms of the variable mean velocity U(x). or externally by the walls of a duct of variable cross section (ϕ0 (x) can be multiplevalued if the boundaries are multiply connected.2 The Equation of Vortex Sound 119 can vary with position only if the ﬂuid is bounded.2. say. ˙ ρ (5. either internally by an airfoil. and ∂ D ≈ + U·∇ . t). as the acoustic variable. 5.2.5). but the mean velocity is always singlevalued).
1) at angular velocity = /4π 2 . 5.2. t).7) and in the far ﬁeld the acoustic pressure is given by the linearized approximation p(x.1. ρ0 The vortex sound equation (5.5) can therefore be simpliﬁed by (a) taking c = c0 . sin t).2. 5. provided the Mach number is small enough for the motion to be regarded as incompressible. Their positions at time t are ¯ x = (x1 . x2 ) = ±s ≡ ±(s1 (t). s2 (t)) = ± (cos t. t) ≈ ρ0 B(x.2. c0 ρ ∼ 1 + O(M 2 ).2. and (b) by neglecting nonlinear effects of propagation and the scattering of sound by the vorticity. The production of sound is then governed by the simpler equation 1 ∂2 − ∇ 2 B = div(ω ∧ v).8) 5. Fig.2. (5. and ρ = ρ0 . .120 5 Vortex Sound c0 at inﬁnity by relations of the form c ∼ 1 + O(M 2 ).4 Example 1 (Powell 1963): Sound Generation by a Spinning Vortex Pair Two parallel vortex ﬁlaments each of circulation and distance 2 apart rotate about the x3 axis midway between them (Fig.2. 2 c0 ∂t 2 (5.
y2 ). (2. by setting ξ = y3 − x3 . ω∧v=− s(t)[δ(¯ − s) − δ(¯ + s)].2 The Equation of Vortex Sound The vorticity distribution is ω = k(δ(¯ − s) + δ(¯ + s)).7): ¯ −x j ∂ ∂ ..5.2. (5. 1 p≈ ρ0 ¯ ¯ xi x j ∂2 2 2πc0 ∂t 2 ∞ −∞ (si s j ) t − (r 2 + ξ 2 ) 2 c0 1 dξ (r 2 + ξ 2) 2 3 . In the acoustic far ﬁeld.7) for this quadrupole source is (c.2.f. where B ≈ p/ρ0 . Hence. we use (1. The vortex convection velocities are v = ± k ∧ s(t) at ¯ x = ±s(t).1)) B= ∂2 1 ¯ ¯ 4π ∂ x i ∂ x j 2 (si s j ) t − x − y c0 δ(¯ ) d 3 y y . where U = c0 . because k ∧ (k ∧ s) = −s.9) .9. we can write in the acoustic far ﬁeld.2. parallel to the vortices. x The solution of the vortex sound equation (5. Thus. ≈ 1 2 + (x − y )2 ) 2 ∂t ¯ ∂x j c0 (r 3 3 2 2 where r = (x1 + x2 ) 2 is the perpendicular distance from the centroid of the vortices (the x3 axis). x − y ¯ y = (y1 . x so that the vortex sound source is equivalent to the quadrupole div(ω ∧ v) ≈ ∂2 (2 ¯ ¯ ∂xi ∂x j si (t)s j (t)δ(¯ )). x x 121 where k is a unit vector in the x3 direction. r → ∞. x x If this is expanded in powers of the radius of the circular orbit (and it can be veriﬁed that this is equivalent to expanding the acoustic pressure in powers of 1) the ith component of the ﬁrst nonzero term is M = U/c0 (ω ∧ v)i ≈ ∂ (2 ¯ ∂x j si (t)s j (t)δ(¯ )).
2. sin θ) cos 2 [t] − sin 2 [t] − π . The sound power must now be calculated by considering the integral p2 dS ρ0 c0 over the surface of a large circular cylinder r = constant. in the limit in which the radial distance r greatly exceeds the acoustic wavelength ∞ −∞ (si s j ) t − 2 (r 2 + ξ 2 ) 2 c0 1 2 1 dξ (r 2 + ξ 2) 2 π 4 π 4 3 ≈ 2r 2 π c0 r cos 2 [t] − sin 2 [t] − sin 2 [t] − −cos 2 [t] − π 4 π 4 . sin θ) we ﬁnd 2 ¯ ¯ xi x j 2r 2 2 π c0 r π c0 r π c0 r 1 2 1 2 cos 2 [t] − sin 2 [t] − π 4 π 4 sin 2 [t] − −cos 2 [t] − π 4 π 4 π 4 π 4 ij = = 2 2 (cos θ. 4 sin 2 [t] − π 4 −cos 2 [t] − π 4 cos θ sin θ 1 2 2 cos 2θ − 2 [t] + and.2. c0 (5. Hence.2. ¯ where [t] = t − r/c0 . 4 √ The amplitude of the sound decreases like 1/ r (instead of 1/r ) because the waves are spreading cylindrically in two dimensions. introducing polar coordinates x = r (cos θ. i.122 In this formula (si s j )(t) = 2 5 Vortex Sound 2 1 + cos 2 t sin 2 t sin 2 t . 1 − cos 2 t but the constant terms in the matrix can be omitted because of the time derivatives in (5. (5. therefore.10) = −4 π ρ0 U 2 M 3/2 cos 2θ − 2 r r c0 π . The integration in (5.2.9).. Taking the time .9) can now be performed by the approximate method described below in Example 2 in the limit that r/c0 → ∞.9) becomes p≈ 3 −ρ0 2 π c0 2 π c0 r 1 2 cos 2θ − 2 t− t− r c0 + + π 4 r → ∞.e.
11) Put ξ = µr .2.11). This differs by a factor of the Mach number M from the power radiated by a compact body of threedimensional turbulence. f (µ) can be replaced by f (0). r c0 → ∞. and is characteristic of the acoustic power produced by twodimensional regions of turbulence in an unbounded ﬂuid. and the main contribution to the integral is from the neighborhood of that value of µ where the oscillations are stationary. As κ0r → ∞ the exponential factor oscillates increasingly rapidly. 1 2 e4. In the ﬁrst approximation. This occurs at µ = 0.5 Example 2 Show that ∞ f −∞ ξ r eiκ0 √ r 2 +ξ 2 dξ ≈ r f (0) 2π κ0 r 1 2 ei (κ0 r + 4 ) . Thus. . (5. 5. In particular.2. I =r ∞ −∞ f (µ)eiκ0 r 2π κ0 r √ 1+µ2 dµ ≈ r f (0)eiκ0 r ∞ −∞ eiκ0 r µ /2 dµ 2 = r f (0)eiκ0 r which yields (5. π κ0r → ∞. iπ (r 2 + ξ 2 ) 2 cos 2 t− − i sin 2 c0 −∞ √2 2 ∞ dξ i2 r +ξ = e−2i t e c0 3 2 + ξ 2) 2 (r −∞ ≈ 1 r2 πc0 r 1 2 ∞ 1 (r 2 + ξ 2 ) 2 t− c0 1 dξ (r 2 + ξ 2) 2 3 e −i [2 (t − r ) c0 − π] 4 . we can show that the acoustic power per unit length of the vortices ∼ ρ0 U 3 M 4 . The integrand is therefore expanded about this point. then I ≡ ∞ f −∞ ξ r eiκ0 √ r 2 +ξ 2 dξ = r ∞ −∞ f (µ)eiκ0 r √ 1+µ2 dµ. and √ 2 2 eiκ0 r 1+µ ≈ eiκ0 r +iκ0 r µ /2 .5.2 The Equation of Vortex Sound 123 average.2.
The body is assumed < to be within S+ .3. 5. and S+ will subsequently be allowed to shrink down to coincide with the body surface S. Introduce a stationary. Multiply equation (5. 4 3 3 ≈ ∞ 1 r2 sin 2 π c0 r cos 2 π 4 −∞ (r 2 + ξ 2 ) 2 t− c0 π c0 r 1 2 ≈ 1 r2 sin 2 t− 5. ∞ 5 Vortex Sound cos 2 −∞ t− 1 2 (r 2 + ξ 2 ) 2 c0 t− 1 1 dξ (r 2 r c0 + ξ 2) 2 − dξ (r 2 r c0 + ξ 2) 2 − π .3 of Curle’s equation.7) by H ≡ H ( f ) and Fig.124 Hence.7) is now applied to determine the sound generated by vorticity in the neighborhood of a ﬁxed body whose surface S may be vibrating at small amplitude (Fig. .2.1. such that f (x) > 0 according as x lies without or within S+ .3 Vortex–Surface Interaction Noise The small Mach number vortex sound equation (5. 5.3. closed control surface S+ on which f (x) = 0.2.1). The development here is analogous to the derivation in Section 2.
2 c0 ∂t 2 This equation is formally valid everywhere. we can make the substitution ∇ H · (∇ B + ω ∧ v) = −∇ H · ≡ −∇ H · Hence the vortex sound equation becomes 1 ∂2 ∂v − ∇ 2 (H B) = −div(B∇ H ) + ∇ H · 2 ∂t 2 ∂t c0 + div(H ω ∧ v) − ν div(∇ H ∧ ω). In Fig.5. Then.5). We use the transformations H ∇ 2 B ≡ H div (∇ B) = div(H ∇ B) − ∇ H · ∇ B = ∇ 2 (H B) − div(B∇ H ) − ∇ H · ∇ B.3. they completely determine B outside this control surface.2) The sources on the right of this equation are either concentrated on the control surface S+ or lie in the ﬂuid outside S+ . (5. .2.3) ∂v + ν curl ω ∂t ∂v + ν div(∇ H ∧ ω). y. The source terms involving ∇ H are concentrated on the control surface. the surface sources constitute a representation ‘to the outside world’ in f > 0 of the various hydrodynamic or acoustic processes that may be occurring within S+ . ∂t (5.3. When x lies in the exterior region these surface terms take account of the presence of the solid body inside S+ .2. 5. (5. t − τ ) that satisﬁes 1 ∂2 − ∇ 2 G = δ(x − y)δ(t − τ ). The solution in this region can therefore be found by using any Green’s function G(x. including the region within S+ where H B ≡ 0.1) (5. and H div(ω ∧ v) = div(H ω ∧ v) − ∇ H · ω ∧ v.3.3 Vortex–Surface Interaction Noise 125 form the inhomogeneous wave equation for the new variable H B.1 the ﬂuid occupies the region V outside the surface S of the solid body.3. the control surface S+ ( f (x) = 0) therefore lies within V .7) becomes 1 ∂2 − ∇ 2 (H B) = −div(B ∇ H ) − ∇ H · (∇ B + ω ∧ v) + div(H ω ∧ v). where G = 0 2 c0 ∂t 2 for t < τ for x and y anywhere within the ﬂuid. Using Crocco’s equation (4. if the body is absent (so that S+ is ﬁlled with ﬂuid).
t) = ∞ −∞ V G(x. τ ) ∂G (x. and is a large. which is understood to vary over the range −∞ < τ < ∞. ∂y where for brevity we have omitted the integration sign for τ . τ ) − V ∂vn ∂G (y. y. where all of the source terms within the brace brackets are functions of y and τ .3) is H B(x. t − τ ) d S(y). ∂ yn ≡ S+ where all vector operators are with respect to the y dependence. t − τ ) + G(x. τ ) ∧ S+ +ν ∂G (x. y. y. τ ) dS(y) dτ (x.126 5 Vortex Sound Thus. . t − τ ) · dS(y) dτ. closed ‘surface at inﬁnity’ where ω = 0. t) = S+ B(y. y. t − τ ) −div(B∇ H ) + ∇ H · ∂v + div(H ω ∧ v) ∂τ − ν div(∇ H ∧ ω) d 3 y dτ. y. The divergence terms are removed by application of the divergence theorem. t − τ ) d 3 y dτ ∂y (5. y. Those involving ∇ H vanish except on the control surface S+ . The general solution in the region f > 0 outside S+ accordingly becomes B(x. Let us illustrate the procedure for the ﬁrst term in the brace brackets of the integrand G{−div(B∇ H )} d 3 y = − V V {div(G B∇ H ) − B∇G · ∇ H } d 3 y G B∇ H · dS + B∇G · ∇ H d 3 y V = S+ = 0+ S+ B∇G · dS B(y. τ ) · ω(y. There are no contributions from S and because ∇ H = 0 everywhere except on S+ .3).3.4) H ( f (y))(ω ∧ v)(y.3. t − τ ) ∂ yn ∂τ ∂G (x. and then using the formula (·)∇ H d 3 y = V S+ (·) dS (see Section 2. for points x within the ﬂuid the solution of (5.
When this is done the control surface S+ is allowed to shrink down onto S (whereupon the ﬁrst term in the ﬁrst integral of (5. and this term in the solution is actually identical with that given previously in (3. y. and in the important case in which the body does not vibrate the acoustic far . recall that the control surface S+ was taken to be ﬁxed in space. t − τ ) = 0 ∂ xn respectively for y. y. When S+ shrinks down to S the implication is that S is also ﬁxed in space. t) = − V (ω ∧ v)(y. t) by p(x. The contribution to the sound from surface friction (the ﬁrst surface integral on the right of (5.8.3. t − τ ) d 3 y dτ + ν ∂y ω(y. y. (5. This means that ﬂuid can ﬂow through the surface. it should also be noted that the reciprocal theorem implies that the normal derivative conditions ∂G (x. t − τ ) (y. ∂y ∂τ S ∂G where (x.3. Green’s function takes full account of the inﬂuence of the body on the efﬁciency with which these sources generate sound. x on S.3. y. At high Reynolds numbers the surface term can therefore be discarded. surface integral involving the surface vorticity is the contribution from frictional forces on S.2) for a vibrating body in the absence of vortex sources. τ ) dS(y) dτ. τ ) · ∂G (x.5)) is nominally of order 1 Re 1. t)/ρ0 . ∂ yn ∂G (x. τ ) S ∧ ∂G ∂vn (x. The second. are always satisﬁed simultaneously.3 Vortex–Surface Interaction Noise 127 We now choose G to have vanishing normal derivative on the surface S of the body.5. However. the normal velocity vn can still be nonzero if the surface of the body is vibrating at small amplitude. The ﬁrst integral represents the production of sound by vortex sources distributed within the ﬂuid. ν relative to the contribution from the volume vorticity (the ﬁrst integral). y. To interpret the ﬁnal term. t − τ ) · dS(y) dτ + G(x.5) ∂ yn In the acoustic far ﬁeld (x → ∞) we can replace B(x. t − τ ) = 0. and the general solution in the ﬂuid becomes B(x. Re = v . t − τ ) = 0 on S. where is the characteristic length scale of the turbulence or body and v is a typical velocity.4) vanishes). In connection with this. y.
t) ≈ −ρ0 x j ∂ 4π c0 x2 ∂t − S (ω ∧ v) · ∇Y j d 3 y − ν S ω ∧ ∇Y j · dS(y) (5. y.8 by expanding Green’s function to ﬁrst order in the retarded time across S: G(x.9. Substituting into (5.4. where the prime denotes differentiation with respect to t.4 Radiation from an Acoustically Compact Body When the surface S is acoustically compact the compact Green’s function (3.128 ﬁeld is then given by p (x. t − τ ) = ≈ X − Y 1 δ t −τ − 4π X − Y c0 x 1 δ t −τ − 4π x c0 + xjYj δ 4πc0 x2 t −τ − x .3. When this happens the resulting monopole radiation from the body is usually large compared to all other sources. τ ) · V ∂G (x. It makes a contribution to the ﬁnal surface integral only if the volume of the body is pulsating. and consider only surface vibrations for which the volume of S is constant. in particular we shall assume that S vibrates as a rigid body.1) ∂Un Y j d S(y) . In this case. and Un is the normal component of velocity of vibration of S. we proceed as already described in Section 3. therefore. When x → ∞ and the origin is within or close to S.1) can be used to evaluate the general solution (5. We shall. y.3. t) = − ρ0 5 Vortex Sound (ω ∧ v)(y. which is independent of y. ∂y (5. we ﬁnd in the acoustic far ﬁeld (where B = p/ρ0 ) p(x. The ﬁrst term in this approximation. t − τ ) d 3 y dτ. ∂t where the large square brackets ([ ]) denote that the enclosed quantity is to be evaluated at the retarded time t − x/c0 .5) in the far ﬁeld.3.5) and performing the integrations with respect to τ . therefore.5). the ﬁrst approximation in the Green’s function expansion can again be discarded.6) 5. . ignore this possibility. c0 x → ∞. clearly makes no contribution to the ﬁrst two integrals in (5.3.
2) − (m 0 δi j + Mi j ) dUi . Re = .1) can therefore be written p(x.4.10) obtained in the absence of vorticity.4. The relative contributions from the volume and surface distributions of vorticity in (5. and Mi j is the added mass tensor of the body (see (3. t) ≈ xj ∂ Fj 4πc0 x2 ∂t t− x c0 + m 0 x j ∂ 2U j 4π c0 x2 ∂t 2 t− x . where n i is the ith component of the surface normal directed into the ﬂuid. and is equivalent to the solution (3. The solution (5. then m 0 = ρ0 is the mass of ﬂuid displaced by the body.4. dt Reference to Equation (4. x Re ν Thus. in high Reynolds number turbulent ﬂows the surface frictional contribution to the dipole force F can usually be neglected. This is just our earlier conclusion (2. Then ρ0 S ∂Un dUi Y j d S = ρ0 ∂t dt S (n i y j − n i ϕ ∗ ) d S j dUi . For a nonvibrating compact body the principal component of the acoustic pressure in the far ﬁeld is therefore .6)). (5.8.5.4) shows that this can also be written p(x.3) where F(t) is the unsteady force exerted on the ﬂuid by the body. with the addition of the ﬂuiddisplacement effect of the vibrating body. c0 x → ∞.4.2) (respectively the ﬁrst and second integrals) for turbulence of length scale and velocity v are estimated respectively by ρ0 v 2 M x and ρ0 v 2 M 1 v .4 Radiation from an Acoustically Compact Body 129 When the body executes translational oscillations at velocity U(t) we have Un = Ui n i .4.2) derived from Curle’s equation.4. t) ≈ −x j ∂ ρ0 4πc0 x2 ∂t (ω ∧ v) · ∇Y j d 3 y − η S ω ∧ ∇Y j · dS(y) (5.8. dt = (m 0 δi j + Mi j ) where if the body has volume .
5. c0 (5. 0).2) For example. t − τ ) = ≈ X − Y 1 δ t −τ − 4π X − Y c0 x − ky3  1 δ t −τ − 4π x − ky3  c0 + ≈ x j Y⊥ j δ 4π c0 x − ky3 2 t −τ − x − ky3  c0 x − ky3  1 δ t −τ − 4π x c0 + x j Y⊥ j δ 4π c0 x2 t −τ − x − ky3  . the monopole term in (5.6. To do this we ﬁrst write the Kirchhoff vector in the form Y = Y⊥ + ky3 .1) where k is a unit vector in the x3 direction. x → ∞.5. and the acoustic pressure given by (5. y. c0 (5. Y2 (y).4) 5. (5. t) ≈ −ρ0 x j ∂ 4πc0 x2 ∂t (ω ∧ v) y.5. Then. t − x · ∇Y j (y) d 3 y.5 Radiation from Cylindrical Bodies of Compact Cross Section An important special case occurs when vorticity interacts with a cylindrical (or approximately cylindrical) surface S of compact cross section (such as the strip airfoil of Fig. x → ∞. t) ≈ −ρ0 x j ∂ 4π c0 x2 ∂t 5 Vortex Sound (ω ∧ v) y.5.3. c0 (5. 3. the compact Green’s function in (5.6) becomes p(x. x → ∞.130 given by p(x. If the vorticity extends over an extensive spanwise section of the body it may be important to account for differences in the retarded times of the sound produced at different spanwise positions. t − x − ky3  · ∇Y⊥ j (y) d 3 y.3.2) can be discarded as before.6) is expanded as follows: G(x.3). when a stationary cylindrical body interacts with high Reynolds number ﬂow at low Mach number.4. because variations in the spanwise source position are not necessarily small compared to the acoustic wavelength. Y⊥ = (Y1 (y).3) The sound is produced by dipole sources orientated in the lift and drag directions .
6.3).7. where is the length scale of the interaction region (the body).4.2) is applicable for a body executing any combination of translations and rotations (Fig.6 Impulse Theory of Vortex Sound 131 only ( j = 1.6. v(x.2) is valid only for a body in translational motion whereas (5.1) shows that this is equivalent to the compact approximation (5. (5. 5. The acoustic dipole is found simply by replacing I(t) in (5.1) This expression for ϕ(x. and this procedure therefore leads to the following formula for the sound in terms of the vorticity: p(x.1.9.1).9)). 3. 2).6. and in cases where the chord of the airfoil is a slowly varying function of y3 (such as the elliptic airfoil of Fig.6.1) by I(t − x/c0 ).5. including the bound vorticity on S. (5.3. t) ≈ −ρ0 ≈ = ∂ ∂x j 1 ∂Ij (t − x/c0 ) 4π x ∂t t− x . where bound vorticity occurs only in the surface integral of the frictional contribution to the sound. . 4.2) involving the Kirchhoff vector. But.4. c0 (5. At large distances from the body (where the undisturbed ﬂuid is stationary) the pressure p(x. x → ∞ c0 x d 3 y. t) = div I(t) 4π x where I(t) = 1 2 y ∧ ω(y. t) in the hydrodynamic far ﬁeld in terms of the impulse I(t).6 Impulse Theory of Vortex Sound An interesting formula for the sound generated by vorticity near a compact body can be derived directly from the representation (4. that ω in (5.7.4) and (1. This should be contrasted with the representation (5. Indeed. t) d 3 y.3) of the velocity v(x. t) ≈ ∇ϕ when x . Note. It is the velocity potential of a hydrodynamic dipole that will be recognized as the acoustic near ﬁeld of an outgoing acoustic dipole representing sound production by the ﬂow (see Equations (1. This approximation is applicable also to a thin airfoil of large but ﬁnite span. t) = −ρ0 ∂ϕ/∂t.2) is the generalized vorticity. however.2) ρ0 x j ∂ 2 I j 4πc0 x2 ∂t 2 ρ0 x j ∂ 2 8πc0 x2 ∂t 2 (y ∧ ω) j y. t) deﬁnes the incompressible motion in the irrotational region far from the body.4.6.4. and ϕ(x.1). t − Equation (4.
M¨ hring (1978) has shown that it is also possible to express the quadrupole o sound as a thirdorder time derivative of a secondorder moment of the vorticity. 2 Show that to ﬁrst order in the vortex is equivalent to the twodimensional quadrupole div(ω ∧ v) ≈ ∂2 (Ti j δ(x1 )δ(x2 )). 2 ∂ xi ∂ x j .6.6. namely p(x. We know that the sound is now generated by quadrupole sources and that it can be represented in terms of the vorticity as in (5. i.132 5 Vortex Sound The impulse I is constant for vorticity in an unbounded ﬂuid (when compressibility is ignored).6). − cos θ + cos(θ − t/2)). x → ∞. t) ≈ ρ0 x i x j ∂ 3 2 12πc0 x3 ∂t 3 yi (y ∧ ω) j (y.2) (see Problems 5).1. t − x/c0 ) d 3 y. where 1 and is the uniform vorticity in the core. v2 ) = − 1 r (sin θ + sin(θ − t/2). Kirchhoff’s spinning vortex: A columnar vortex parallel to the x3 axis has elliptic cross section deﬁned by the polar equation r = a{1 + cos(2θ − t/2)}.3) Problems 5 1. and the velocity distribution within the core 4 is given by v = (v1 . (5. analogous to the ﬁrst order moment in (5. The ellipse rotates at angular velocity 1 . j = 1.
and iθ is a unit vector in the azimuthal direction. x1 ) are cylindrical polar coordinates. where U = 1 a and M = U/c0 . incompressible ﬂuid conservation of energy and momentum implies that n Rn n Rn d Xn d Rn − Xn dt dt ρ0 (3 cos2 2 12c0 x 2 n Rn X n . dt where is the angle between the observer direction and the positive x1 axis and the term in square braces is evaluated at τ = t − x/c0 . = constant. 2 2.7) to calculate the sound produced by the unsteady motions of an acoustically compact system of N vortex rings coaxial with the x1 axis. dt 3 x → ∞. . Rn (t) being the vortex ring radius. θ.Problems 5 where Ti j = π 8 2 4 133 a cos( t/2) sin( t/2) sin( t/2) −cos( t/2) π . Show that p≈ ρ0 (3 cos2 2 8c0 x − 1) ∂2 ∂t 2 n Xn n 2 d Rn . Take the vorticity of the nth vortex to be ω n = n δ(x1 − X n (t))δ(r − Rn (t))iθ . n 2 n Rn = constant. where (r. X n (t) its location in the x1 direction.2. Use these equations to show that p≈ where S(t) = n − 1) d3 S . In an ideal. x → ∞. 4 and that the acoustic pressure is p≈− 8 2πa ρ0 U 2 M 3/2 cos 2θ − r 2 t− r c0 + r c0 → ∞. Coaxial vortex rings: Use Equation (5.
1. c0 = π(a + b)2 . 2 The result now follows by noting that the estimate (5. and use the identity y ∧ curl A = 2A + ∇(y · A) − ∂ (y j A) ∂yj to deduce that the integral in (5. show that p≈ = ρ0 R 2 (3 cos2 − 1) d 3 X 2 dt 3 12c0 x ρ0 U 2 M 2 R(3 cos2 − 1) 8π 3 x . 5. The free space Green’s function in two dimensions – the solution of 1 ∂2 − ∇ 2 G = δ(x1 − y1 )δ(x2 − y2 )δ(t −τ ). – can be derived by integrating the threedimensional Green’s function 1 x − y δ t −τ − 4π x − y c0 over −∞ < y3 < ∞.7) permits the second integral on the right to be discarded.6) can be written yi (ω ∧ v) j d 3 y = − 1 ∂ 3 ∂t 1 yi (y ∧ ω) j d 3 y + δi j 3 1 2 3 v d y. U . . where G = 0 2 c0 ∂t 2 for t < τ. t− x c 0 a−b 2 x t− cos 2 a+b π(a + b) c0 where U = a+b . x → ∞ .134 5 Vortex Sound 3. coaxial with the x1 axis. Calculate the sound produced by a vortex ring of total circulation . ln − + dt 4π R a+b 4 2(a + b) In the notation of Question 2. Derive M¨ hring’s formula (5. multiply by yi . whose core has elliptic cross section of major and minor axes 2a.1. 2b R.3) for the acoustic pressure generated by a o compact region of vorticity in an unbounded ﬂuid.6. Take the cross product of y with the high Reynolds number vorticity equation ∂ω/∂t + curl (ω ∧ v) = 0 (expressed in terms of y and t as independent variables). where R is the mean radius of the ring. Assume that the x1 coordinate X (t) of the vorticity centroid satisﬁes dX 16R 1 3(a − b) = cos 2 t . M= 4.
8. Use the result of Problem 6 to derive the Solution (5. y. x2 ). Use the result of Problem 6 to solve Problem 1. t) ≈ ρ0 xi x j ∂ 3 2 4πc0 x2 ∂t 3 ρ0 x i x j 2 4πc0 x2 t−x/c0 dy1 dy2 yi (y ∧ ω) j (y. Deduce M¨ hring’s (1980) twodimensional representation o p(x. x = (x1 . √ t − τ − x/c0 x → ∞. 2. y = (y1 . τ ) dτ . where x − y ≈ c0 (t − τ ) G(x. y2 ). t − τ ) = H (t − τ − x − y/c0 ) 2 2π (t − τ )2 − (x − y)2 c0 135 . y. Show that yi (ω ∧ v) j dy1 dy2 ≈ − 1 ∂ 2 ∂t yi (y ∧ ω) j dy1 dy2 . . where i. t − τ ) ≈ H (t − τ − x − y/c0 ) . and that near the wavefront.Problems 5 Deduce that G(x.2. j = 1. 7. √ 2π 2x − y/c0 (t − τ ) − x − y/c0 √ 6. where the vorticity ω is parallel to the k direction (the x3 axis). Consider sound production by a compact distribution of vorticity in an unbounded twodimensional ﬂow (independent of x3 ). τ ) dτ 2 (t − τ )2 − x2 c0 t−x/c0 −∞ ≈ c0 ∂ 3 2x ∂t 3 dy1 dy2 −∞ yi (y ∧ ω) j (y.10) for the sound produced by a spinning vortex pair.
2) For twodimensional problems G is a function of y3 − x3 .1 Compact Green’s Function in Two Dimensions In this chapter. y2 and τ . Conditions are assumed to be uniform in the x3 direction.1. so that only the y1 and y2 components of the gradient ∂G/∂y contribute to the integral. x2 = y2 . Let G2 = ∞ −∞ G(x. y. G 2 represents the ﬁeld generated by a uniform line source parallel to the x3 axis extending along the whole of the line x1 = y1 . we apply the general high Reynolds number solution (5. t − τ ) d 3 y dτ ∂y (6. and may be performed prior to any further calculations of the sound. We shall derive the twodimensional analogue of the general solution p (x.3) obtained by integrating the threedimensional Equation (3. (6. parallel to the vorticity. Also.1.3. y. Both the vorticity convection velocity v and the Lamb vector ω ∧ v are parallel to the x1 x2 plane. 2 c0 ∂t 2 where G 2 = 0 for t < τ. τ ) · V ∂G (x. the integration with respect to the spanwise coordinate y3 involves only the Green’s function. and the function G 2 will therefore satisfy the Green’s function equation 1 ∂2 − ∇ 2 G 2 = δ(x1 − y1 )δ(x2 − y2 )δ(t − τ ). τ ) depends only on y1 . because (ω ∧ v)(y. t) = − ρ0 (ω ∧ v)(y.1. In two dimensions.1. (6.6) to determine sound produced by twodimensional interactions of rectilinear vortices with a stationary solid boundary.1) by ﬁrst determining a suitable representation of G in two dimensions. 136 . t − τ ) dy3 .4) over −∞ < y3 < ∞.6 Vortex–Surface Interaction Noise in Two Dimensions 6.
To use this to evaluate (6.6.2 − ϕ1.1) the δ function must be expanded to ﬁrst order in Y G2 ≈ 1 4π + ∞ −∞ δ t −τ − ∞ −∞ ¯ 2 + ξ 2 x c0 dξ ¯ 2 + ξ 2 x (¯ 2 x dξ . 4π X − Y c0 ∗ ∗ X 1. and the corresponding twodimensional compact Green’s function can be found by integrating over −∞ < y3 < ∞.2 (y). X 3 = x3 .2 = x1. t − τ ) = X − Y 1 δ t −τ − . the compact Green’s function for a cylindrical body (with generators parallel to x3 ) G(x. We can therefore take G2 ≈ = ¯ x·Y 4π c0 ∞ −∞ δ ∞ t −τ − δ t −τ − ¯ 2 + ξ 2 x c0 (¯ 2 x dξ + ξ 2) ¯ x·Y ∂ H (t − τ − ¯ /c0 ) x . 2 2 1 ¯ x Set ξ = y3 − x3 . x2 ) and let ¯  = (x1 + x2 ) 2 → ∞. we have G2 ≈ 1 4π ∞ −∞ δ t −τ − ¯ x·Y ¯ 2 + ξ 2 x + c0 c0 ¯ 2 + ξ 2 x dξ ¯ 2 + ξ 2 x as ¯  → ∞. = 2π c0 ∂t (t − τ ) c2 (t − τ )2 − ¯ 2 x 0 ¯ 2 + ξ 2 x dξ 2 + ξ 2) c0 (¯  x 0 ¯ x x · Y ∂ H (t − τ − ¯ /c0 ) = √ x 2π c0 ∂t (¯ 2 + ξ 2 ) ∂ ¯ 2 +ξ 2 √ x 2 ∂ξ c 2 2 0 ¯ x·Y ∂ 2π c0 ∂t ξ= c0 (t−τ ) −¯  x . x = (x1 . Y3 = y3 .1.1.1 Compact Green’s Function in Two Dimensions 137 Similarly. + ξ 2) ¯ x·Y 4πc0 δ t −τ − ¯ 2 + ξ 2 x c0 Only the second term on the right depends on y and therefore contributes to the radiation integral (6.2 − ϕ1.2 (x). y. x ¯ where x · Y = x1 Y1 + x2 Y2 is independent of ξ . is a function of y3 – x3 . Taking the origin of coordinates within the cylindrical body. Y1.2 = y1.1).
after propagating along a ray perpendicular to the source over a distance equal to x. (6. 3.1. the wavefront arrives at time τ + x/c0 . and drop the overbar on x and the subscript 2 on G 2 and soforth. The dipole component of the twodimensional compact Green’s function then becomes x·Y ∂ H (t − τ − x/c0 ) G(x. y2 ). or of a cylindrical wall cavity or projection from a wall (see Fig. √ t − τ − x/c0 (6. The farﬁeld representation (6. x → ∞. and therefore after a time delay x/c0 .1. c0 for t − τ ∼ x . t −τ ) ≈ x 2x c0 (t − τ ) − x. t − τ ) ≈ x1 Y1 ∂ √ 3 2 ∂t π 2c0 x H (t − τ − x/c0 ) . but the observer at x receives sound continuously after the passage of this wavefront.138 6 Vortex–Surface Interaction Noise in Two Dimensions We shall henceforth in this chapter regard all space vectors as twodimen¯ sional. y. y = (y1 .4) can be approximated further by expanding about the wavefront (where t −τ = x/c0 ).9.5) √ t − τ − x/c0 (see Question 1 of Problems 6). x → ∞.3). (6.9.2) is handled by the twodimensional version of the compact Green’s function (3.1. rigid wall at x2 = 0.4) becomes G(x. 2π c0 ∂t (t − τ ) c2 (t − τ )2 − x2 0 (6. At any point x in the far ﬁeld the ﬁrst sound arrives from the nearest point on the line source of (6. x → ∞. such as x = (x1 . c0 ∂ x·Y √ 3 2π 2c0 x 2 ∂t H (t − τ − x/c0 ) . x2 ). t − τ ) ≈ . Just to the rear of the wavefront 2 (t − τ ) c0 (t − τ )2 − x2 ≡ (t − τ ) c0 (t − τ ) + x c0 (t − τ ) − x ≈ Therefore. The procedure described above yields the following expression for the dipole component of the twodimensional compact Green’s function G(x. which contains most of the acoustic energy.1. and at a later time t sound is received from those source points whose distance from x is equal to c0 (t − τ ). generated at more distant sections of the inﬁnitely long line source. G has an inﬁnite peak at the wavefront where t − τ − x/c0 = 0 followed by a slowly decaying tail.4) In contrast to the threedimensional Green’s function.1.6) .3). y. The special case of a cylindrical body adjacent to a plane. y.1. which is nonzero only on a spherically expanding wavefront.
x → ∞.4.1) using the compact Green’s function (6.1.1. 6.5) of the cylinder in the absence of the wall. this can also be written p(x. v= dx0 (t). In these twodimensional problems the vorticity ω is directed along the x3 axis.1 and parallel to the generators of the cylinder. y2 ) is the velocity potential of incompressible ﬂow past the cylinder in a direction parallel to the wall (with unit speed at large distances from the body).1) (ω ∧ v · ∇Y j )(y. p(x. Let k be a unit vector in this direction and consider a line vortex of strength whose position and translational velocity are x = x0 (t). G represents the ﬁeld of a dipole orientated parallel to the wall and perpendicular to the cylinder axis.2. 6.4) applied to a stationary body. t) ≈ xj ∂ √ 3 2π 2c0 x 2 ∂t t−x/c0 −∞ F j (τ ) dτ . and the effect of the wall is to generate an equal image dipole that just doubles the magnitude of the sound relative to the corresponding dipole (6.6. t) ≈ −ρ0 x j ∂ √ 3 2π 2c0 x 2 ∂t × t−x/c0 −∞ √ dτ t − τ − x/c0 (6. rigid cylindrical body of diameter has characteristic wavelength ∼ /M . τ ) dy1 dy2 .1.2 Sound Generated by a Line Vortex Interacting with a Cylindrical Body The sound produced by twodimensional vortex motion at low Mach number M ∼ v/c0 near a stationary. viz. √ t − τ − x/c0 (6.2 Sound Generated by a Line Vortex 139 where Y1 ≡ Y1 (y1 .5). dt If the motion elsewhere is irrotational. x → ∞. out of the plane of the paper in Fig. According to the inviscid form of the Formula (4.2) where F j is the force per unit length of the cylinder exerted on the ﬂuid in the j direction.2.2. we have ω = kδ(x − x0 (t)) . The acoustic pressure is determined by the twodimensional version of (6.
the vortex path have radius r0 .4) 6. the vortex will traverse the circular orbit discussed in Section 4.6 (Example 2). and let the vortex strength be sufﬁciently small for the local motion to be considered incompressible.2. so that ω ∧ v = k ∧ vδ(x − x0 (t)) ≡ k ∧ dx0 (t)δ(x − x0 (t)).1) yields the following general formula for the acoustic pressure as x → ∞: p(x. Then.2.1 Example 1: Sound Produced by Vortex Motion near a Circular Cylinder Assume there is no mean ﬂow. 6. t − τ − x/c0 (6.1.2. If > 0 the vortex moves in the . t) ≈ = −ρ0 x j ∂ √ 3 2π 2c0 x 2 ∂t −ρ0 x j ∂ √ 3 2π 2c0 x 2 ∂t t−x/c0 −∞ t−x/c0 −∞ k∧ dx0 dτ (τ ) · ∇Y j (x0 (τ )) √ dτ t − τ − x/c0 √ x0 (τ ) d x01 ∂Y j d x02 ∂Y j − dτ ∂ y2 dτ ∂ y1 dτ .3) and (6.140 6 Vortex–Surface Interaction Noise in Two Dimensions Fig. and take the coordinate origin at the centre of the cylinder.2. Let the cylinder have radius a.2. dt (6.
4) becomes p(x.6. and k∧ dx0 = dt k ∧ (k ∧ x0 ) = − x0 . t) ≈ x j r0 ∂ ρ0 √ 3 2π 2c0 x 2 ∂t t−x/c0 −∞ ∂Y j dτ (x0 (τ )) √ .2. v= = dx0 (t) = dt k ∧ x0 (t). sin t).5) √ 2 2 where r = y1 + y2 is the radial distance from the cylinder axis. Y2 = sin ϑ r + a2 r . .2. 6.2 at the speed given by (4. clockwise direction in Fig.9. Therefore. 2 2 2πr0 r0 − a 2 x0 = r0 (cos t.6. (6.2 Sound Generated by a Line Vortex 141 Fig.8) when r is replaced by r0 . Then.1 supplies the components of the twodimensional Kirchhoff vector for the cylinder: Y1 = cos ϑ r + a2 r .2.2.2. − a2 . Table 3. 6. ∂r t − τ − x/c0 (6.
t) ≈ ρ0 r0 1 2 2π (2c0 x) 1− a2 2 r0 ∂ ∂t t−x/c0 −∞ cos(θ − τ ) dτ . sin ϑ). sin θ) we ﬁnd so that xj ∂Y j a2 (x0 (τ )) = x 1 − 2 ∂r r0 cos(θ − τ ). (6. At any particular retarded time t −x/c0 . and √ have the characteristic dipole amplitude proportional to ρ0 U 2 M. = ρ0 U √ M πr0 r0 2x a sin θ − x → ∞. but with a phase lag of π/4 radians.2. which is smaller by √ a factor M when M 1. √ The acoustic waves decay like 1/ x with distance. by introducing polar coordinates for x x = x(cos θ. which is appropriate for energy spreading two dimensionally in cylindrically diverging waves. using the formula t−x/c0 −∞ cos(θ − τ ) dτ = √ t − τ − x/c0 π   1 2 cos θ − t− x π − c0 4 the pressure becomes p(x.142 6 Vortex–Surface Interaction Noise in Two Dimensions where (y1 . In three dimensions the pressure would be proportional to ρ0 U 2 M. 1. the peaks of these lobes at a ﬁxed distance x from the cylinder rotate in the clockwise direction at angular velocity   following the orbiting vortex.5) becomes p(x. the acoustic amplitude has the doublelobed directivity pattern illustrated in Fig. y2 ) = r (cos ϑ.6) where U =  r0 is the vortex speed. Thus.1b for a dipole. The increased amplitude in two dimensions is a consequence of the inﬁnite extent of the vortex source parallel to the cylinder.7. The reader can conﬁrm that the instantaneous force exerted on the ﬂuid by the cylinder lies in the direction of the vector x0 (t) . t) ≈ ρ0   2 r0 2(2π c0 x) 2 1 2 3 1− a2 2 r0 2 sin θ − a2 1− 2 r0 2 t− x π − c0 4 t− x π − c0 4 . xj ∂Y j a2 = x 1 − 2 ∂r r cos(θ − ϑ) and (6.2. Because < 0. √ t − τ − x/c0 Hence.
6. It is . Thus. Now ϕ ∗ is the velocity potential of an ideal ﬂow around the edge of the halfplane in the anticlockwise sense (with streamlines as in Fig.2.3) and applying (6. x → ∞. adopting the notation of (6. where v is the vortex translational velocity. (6.2. The wavelength is therefore of order r0 r0 × c0 = v M r0 for M = v c0 1. t) ≈ −ρ0 sin(θ/2) √ π x ×δ t − τ − = k∧ dx0 (τ ) · ∇ϕ ∗ (y)δ(y − x0 (τ )) dτ x dy1 dy2 dτ c0 x → ∞.2.2. 6.2. y2 ) = r0 (cos θ0 . we ﬁnd p(x. the coordinates being deﬁned as in Fig. (6. t − τ ) ≈ sin(θ/2)ϕ ∗ (y) x .7) where x = x(cos θ.2 for ideal motion at low Mach number. so that low Mach number motion is sufﬁcient to ensure that the wavelength of the sound is much larger than the vortex distance from the edge. 4. sin θ0 ) is much smaller than the acoustic wavelength.6. which can be written G 1 (x.9.7)). y. For a line vortex at (r0 .2 Example 2: Sound Produced by Vortex Motion near a HalfPlane (Crighton 1972) The trajectory of a line vortex of strength interacting with a rigid halfplane is shown in Fig. δ t −τ − √ c0 π x √ ϕ ∗ (y) = r0 sin(θ0 /2).2 Sound Generated by a Line Vortex 143 joining the center of the cylinder to the vortex. 4.9). θ0 ) the characteristic frequency ∼v/r0 . This is applicable when the distance r0 from the edge of the source at (y1 . The sound generated by the vortex is calculated using the twodimensional compact Green’s function (3.6. sin θ).8) −ρ0 sin(θ/2) dx0 · ∇ϕ ∗ .1. 6.2. The radiation peak therefore also lags by π/4 the peak in the retarded surface force. k∧ √ dt π x where on the second line the term in the square brackets is evaluated at the retarded position x0 (t − x/c0 ) of the vortex.3).1) in two dimensions (Green’s function being given by (6.2.
dt dt dt dt The acoustic pressure can therefore be put in the form p(x.5) ∂ψ ∗ ∂ϕ ∗ = .3. where ϕ ∗ and the stream function ψ ∗ satisfy the Cauchy–Riemann equations (see Section 4.144 6 Vortex–Surface Interaction Noise in Two Dimensions Fig. x → ∞. ∂ y2 ∂ y1 where [Dψ ∗ /Dt] is evaluated at the retarded position of the vortex. 6. the real part of the complex potential √ w = ϕ ∗ + iψ ∗ = −i z. Dt (6. .2.9) ∂ϕ ∗ ∂ψ ∗ =− . and therefore that dx0 dx0 dx0 dx0 k∧ · ∇ϕ ∗ = − ∧ k · ∇ϕ ∗ = − · k ∧ ∇ϕ ∗ = − · ∇ψ ∗ . z = y1 + i y2 . ∂ y1 ∂ y2 A simple calculation shows that k ∧ ∇ϕ ∗ = ∇ψ ∗ . t) ≈ ρ0 sin(θ/2) dx0 ρ0 sin(θ/2) · ∇ψ ∗ ≡ √ √ dt π x π x Dψ ∗ .2.
4. 6.6.4. t) ≈ (4π )2 1 2 x sin θ 2 t/8π 2 [1 + ( t/8π 2 )2 ]5/4 . Let the circulation of the vortex in . This is found as follows √ θ0 ψ ∗ = − r0 cos 2 Therefore. θ0 are given by (4. Performing the calculations we ﬁnd (see Fig.2).6. dt . t−x/c0 where is the distance of closest approach of the vortex to the edge (where it crosses the x1 axis at time t = 0 in Fig.2. θ0 1 dr0 Dψ ∗ =− √ cos Dt 2 r0 dt 2 √ r0 θ0 sin + 2 2 dθ0 . 6. x → ∞.7) (in which r is replaced by r0 and θ by θ0 ). A vortex that translated along one of these streamlines would be silent.6. The stream function ψ ∗ is constant on each of the parabolic streamlines of the ideal ﬂow around the edge deﬁned by ϕ ∗ (Fig.9) to form a qualitative picture of the inﬂuence of vortex shedding on sound generation.4) ρ0 2 p(x. where r0 .3 Inﬂuence of Vortex Shedding 145 Fig. 6. 6.2.2.3).3 Inﬂuence of Vortex Shedding We can use Equation (6.2. the actual edgegenerated sound depends on the rate at which the trajectory of the vortex cuts across the streamlines of this ideal edge ﬂow.
146
6 Vortex–Surface Interaction Noise in Two Dimensions
Fig. 6.2.3 be in the indicated anticlockwise sense, so that ﬂuid near the edge is induced to ﬂow in a clockwise direction around the edge, as implied by the dashed curve in the ﬁgure. When the Reynolds number is large, inertial forces actually cause the ﬂow to separate from the edge, resulting in the release of vorticity of opposite sign from the edge into the wake. Let us assume for simplicity that this shed vorticity rolls up into a concentrated core of strength s . Equation (6.2.9) then supplies the following prediction for the net acoustic pressure p(x, t) ≈ ρ0 sin(θ/2) √ π x Dψ ∗ Dt + Dψ ∗ Dt , x → ∞,
s
s
where the derivatives are evaluated at the retarded positions of and s respectively. Both vortices translate across the curves ψ ∗ = constant in the direction of decreasing ψ ∗ , and the derivatives therefore have the same sign. Hence, because and s have opposite signs, sound produced by the shed vortex will tend to cancel the edgegenerated sound attributable to the incident vortex alone. This conclusion is applicable to a wide range of ﬂuid–structure interactions. A typical interaction involves a bounded region of vorticity, called a ‘gust,’ swept along in a nominally steady mean ﬂow. The localized velocity ﬁeld of the gust is determined by the Biot–Savart formula (4.3.1). At high mean ﬂow speeds it is sometimes permissible to neglect changes in the relative conﬁguration of the vorticity distribution during its convection past an observation point, the vorticity is then said to be frozen (at least temporarily) and the induced velocity is steady in a frame translating with the gust. If the mean ﬂow carries the gust past the surface S of a stationary body, the free ﬁeld induced velocity determined by the Biot–Savart integral is said to produce an upwash velocity on S; the actual velocity consists of the upwash velocity augmented by the velocity ﬁeld required to satisfy the noslip condition on S. (In an ideal ﬂuid only the normal component of the upwash velocity is cancelled on S.) When a gust convects past a stationary airfoil the high Reynolds number surface force (responsible for the sound) is given by (see (4.4.4)) Fi = −ρ0
V
∇Yi (y) · (ω ∧ v)(y, t) d 3 y,
Yi = yi − ϕi∗ (y).
(6.3.1)
The vector ∇Yi is the velocity of an ideal ﬂow past the airfoil that has unit speed in the i direction at large distances from the airfoil. It is singular (or very large) at the edges of the airfoil. These singularities have the following signiﬁcance, when the vorticity length scale is small compared to the airfoil chord the
6.3 Inﬂuence of Vortex Shedding
147
principal contribution to the integral is from vorticity in the neighborhoods of the singularities. For example, for the strip airfoil of Fig. 3.6.3 Y2 = Re(−i z 2 − a 2 ), z = y1 + i y2 ,
and ∇Y2 becomes inﬁnite at the leading and trailing edges z = ∓a. An incident, smallscale gust convecting in the y1 direction in a mean ﬂow at speed U would in practice induce shedding from the trailing edge at z = a. When this shedding is ignored the force calculated from (6.3.1) has two principal components, respectively from gust elements near the leading and trailing edges. To calculate the overall force, however, it is necessary to include the contribution from the shed vorticity, which affects the motion only near the trailing edge when the length scale of the wake vorticity is small. In the linearized treatment of this case (discussed in more detail below), when both the gust and wake vorticity are taken to convect at the same mean velocity U , it is known from unsteady aerodynamics that the force component produced by the wake is equal and opposite to that generated by the gust at the trailing edge (Sears 1941). The effect of this cancellation can be approximated without calculating any details of the shed vorticity. This is accomplished by formally deleting the trailing edge singularity from ∇Y2 , and then ignoring the contribution to the integral (6.3.1) from the shed vorticity. To understand this observe that, because the value of the integral is dominated by vorticity near the edges, it is only the behaviors of Y2 near these edges that must be retained in the integrand, and Y2 can therefore be replaced by the leading order terms in its expansions about the edges. For the strip airfoil of Fig. 3.6.3, we would write √ √ √ √ √ √ Y2 = Re(−i z − a z + a) ∼ Re( 2a z + a) + Re(−i 2a z − a). (6.3.2) The last term is singular at the trailing edge; it is deleted and the following approximation is used in (6.3.1) with the wake vorticity ignored: √ √ Y2 ∼ Re( 2a z + a), where the branch cut for z = +∞. √ (6.3.3)
z + a is taken along the real axis from z = −a to
6.3.1 Example: Surface Force Produced by a Periodic Gust To illustrate the procedure consider incompressible ﬂow parallel to the airfoil of Fig. 6.3.1 at speed U in the x1 direction, in which a time harmonic vortex
148
6 Vortex–Surface Interaction Noise in Two Dimensions
Fig. 6.3.1.
sheet of vorticity ω I = γ kδ(x2 − h)e−iω(t−x1 /U ) , h > 0, ω > 0,
is convected past the airfoil at perpendicular distance h, where γ is the circulation per unit length of the sheet. The vortex sheet can be regarded as an elementary model of a periodic wake behind a small diameter circular cylinder upstream of the airfoil. The gust upwash velocity induces the shedding of vorticity ω S from the trailing edge of the airfoil. When the reduced frequency ωa/U is large the hydrodynamic wavelength 2πU/ω of the gust and wake is much smaller than the airfoil chord, and the surface force is produced primarily by the gust interaction with the leading edge at x1 = −a. The net force F2 (per unit span) on the ﬂuid in the x2 direction can therefore be calculated from (6.3.1) by setting ω = ω I , where ω I ∧ v = γ U jδ(x2 − h)e−iω(t−x1 /U ) , (j being a unit vector in the x2 direction) and by replacing Y2 by the righthand side of (6.3.3): F2 = −ρ0 γ U
∞ −∞
∂Y2 (y)δ(y2 − h)e−iω(t−y1 /U ) dy1 dy2 ∂ y2
∞
≈ −ρ0 γ U 2a = −ρ0 γ U a
√
Re
−∞
1 2
i 2 y1 + i h + a √
e−iω(t−y1 /U ) dy1 (6.3.4)
πU 2iωa
e−ωh/U −iω(t+a/U ) .
The force can also be calculated exactly from linearized thin airfoil theory with proper account taken of vortex shedding. This is the gust loading problem
6.3 Inﬂuence of Vortex Shedding
149
of classical aerodynamics (Sears 1941). The linear theory wake is treated as a vortex sheet downstream of the edge, whose elements convect at the mean stream velocity U. The strength of the vortex sheet is determined by imposing the Kutta condition that the pressure (and velocity) should be ﬁnite at the edge (Crighton 1985). For arbitrary values of the reduced frequency ωa/U it is found that F2 = πiρ0 γ U aS ωa −ωh/U −iωt e , U (6.3.5)
where S(x) is the Sears function, which is deﬁned in terms of the Hankel (1) (1) functions H0 and H1 by S(x) = 2 πx
(1) H0 (x) (1) + i H1 (x)
.
(6.3.6)
In the limit of high reduced frequency S ωa ∼ U iU 2π ωa
1 2
e−iωa/U ,
and the substitution of this into (6.3.5) yields the prediction (6.3.4) determined by the leading edge singularity of Y2 . The plots in Fig. 6.3.2 of the real and imaginary parts of S(ωa/U ) and its asymptotic limit show that the approximation (6.3.4) and the exact value (6.3.5) of the surface force agree when ωa/U > 1.
Fig. 6.3.2.
1) v = U i.1. 6. twodimensional airfoil of chord 2a occupies −a < x1 < a. Suppose the vortex passes above the midchord of the airfoil at time t = 0. when no account is taken of vortex shedding. A line vortex of strength parallel to the airfoil span convects at the mean ﬂow velocity at a constant distance h above the airfoil (Fig.4. (6. which is applicable when U. Hence.4 Blade–Vortex Interaction Noise in Two Dimensions A rigid. The acoustic pressure Fig. As the vortex passes the airfoil new vorticity is shed from the trailing edge into the airfoil wake. Let us ﬁrst consider the potential ﬂow interaction of the vortex and airfoil. where i and j are unit vectors in the x1 and x2 directions.1). .4. which is assumed to consist of a vortex sheet stretching along the x1 axis from x1 = a to x1 = +∞.4. 6. because it requires vorticity shed there to be swept over the rigid surface of the airfoil on which it cannot inﬂuence the force because ω ∧ v · ∇Y2 ≡ 0. ω ∧ v = U jδ(x1 − U t)δ(x2 − h). then ω = kδ(x1 − U t)δ(x2 − h). This is the approximation of linearized thin airfoil theory. 6. h that is. when the inﬂuence on the vortex trajectory of the induced velocity (∼ / h) of image vortices in the body of the airfoil is negligible. x2 = 0 in the presence of a uniform mean stream at speed U in the positive x1 direction.150 6 Vortex–Surface Interaction Noise in Two Dimensions Linear theory does not permit the corresponding force component at the leading edge to be removed by vorticity production at the edge.
1) with Y1 = y1 . where the time scales of the motions ∼h/U . (6.2.4. ∇Y1 = i and ω ∧ v · ∇Y1 ≡ 0. The radiation produced when a vortex passes very close to the airfoil (so that h a) is likely to be particularly intense.2) t−x/c0 −∞ where = cos−1 (x2 /x) is the angle between the radiation direction x and the normal to the airfoil (the x2 axis). t) ≈ = −ρ0 U x2 ∂ √ 3 2π 2c0 x 2 ∂t −ρ0 U cos ∂ √ 1 2π 2c0 x 2 ∂t t−x/c0 −∞ ∂Y2 dτ (U τ.4 Blade–Vortex Interaction Noise in Two Dimensions generated when the wake is ignored is given by (6. the characteristic frequency ω∼ U h and the reduced frequency ωa a ∼ U h 1. and calculate their individual contributions by using the local approximation (6. We may therefore regard the leading and trailing edges as independent sources of sound.2.6. say.3. x → ∞. For the acoustic pressure pLE . The dominant interactions occur as the vortex passes the edges. make the substitution µ = . so that (6. The sound can be attributed to a dipole source orientated in the x2 direction. ∂ y2 t − τ − x/c0 ∂Y2 dτ (U τ. z = y1 + i y2 . and (6.4. h) √ .2) becomes −ρ0 U cos ∂ ≈ √ 1 2π 2c0 x 2 ∂t ×√ t−x/c0 pLE Re −∞ √ i 2a 2(U τ + i h + a) 2 1 dτ . 151 Thus. Y2 = Re (−i z 2 − a 2 ). produced at the leading edge (x1 = −a) we take √ √ Y2 ∼ Re ( 2a z + a).2). h) √ . Thus. t − τ − x/c0 √ t − τ − x/c0 and To evaluate the integral.1) reduces to p(x. ∂ y2 t − τ − x/c0 x → ∞.
The corresponding nondimensional pressure signature √ 1 a 2 ρ0 U M cos pLE 4πa x is plotted in Fig.3) where M = U/c0 .152 6 Vortex–Surface Interaction Noise in Two Dimensions perform the differentiation with respect to time.2 as the solid curve when h/a = 0. (6.2. x → ∞. c0 [t] = t − The additional substitution µ = 1/ξ transforms the integrand into an exact differential.4.4. pLE −ρ0 U cos ≈ √ 2π c0 ρ0 U 2 cos = √ 4π c0 a x a x 1 2 1 2 ∂ ∂t ∞ ∞ Re 0 i (U [t] + i h + a − U µ2 ) 2 i 3 1 dµ Re 0 (U [t] + i h + a − U µ2 ) 2 where dµ. . x .2. Then. 6. leading ﬁnally to pLE ≈ a 2 ρ0 U 2 cos √ 4π c0 x √ ρ0 U M cos a = 4πa x 1 ∞ Re 0 1 2 iξ [(U [t] + i h + a)ξ 2 − U ] 2 U [t] a 3 dξ +1 2 U [t] a +1 + h 2 a . 6.4. The pressure ﬁeld Fig.
2 is therefore representative of the whole of the radiation produced by the blade–vortex interaction. after the ﬁrst arrival of sound from the nearest point on the source. at high reduced frequencies the sound pressure pTE .6. additional contributions to the sound continue to be received indeﬁnitely in time from more distant parts of the source. x → ∞.4. Thus. To determine pTE the calculation described above for pLE is repeated after setting √ √ Y2 = Re(−i 2a z − a) in (6.4) This is large at the retarded times during which the vortex is close to the trailing edge. it is a further consequence of the twodimensional character of the acoustic sources.3. even though they are produced by the vortex interacting with geometrically identical airfoil edges. (6.4.4. 6.4. according to which. which is plotted as the broken line curve in Fig. The leading and trailing edge generated components pLE and pTE have different waveforms.4. say. This is because the integral in (6.2).2. . with characteristic frequency ω ∼ U/ h. the solid curve in Fig. when the contribution from the wake is ignored (which is equal and opposite to pTE ) the overall acoustic pressure signature is given nondimensionally by ( pLE + pTE ) ρ0 U √ M cos 4πa a x 1 2 . leading to pTE ≈ −ρ0 U cos √ 4π c0 ×√ a x 1 2 ∂ ∂t t−x/c0 Re −∞ 1 (U τ + i h − a) 2 1 dτ t − τ − x/c0 √ −ρ0 U M cos = 4πa a x 1 2 h a U [t] a −1 2 + h 2 a . According to Section 6. generated by the potential ﬂow interaction of the vortex with the trailing edge is cancelled by that produced by the wake vorticity. 6.4 Blade–Vortex Interaction Noise in Two Dimensions 153 is generated predominantly as the vortex passes the leading edge of the airfoil at the retarded time [t] = − a/U .2) determines the acoustic pressure at the retarded time [t] in terms of interactions between the vortex and the airfoil at all earlier retarded times.
examine the production of sound for different values of the nondimensional parameter /U h. and applying Formula (5. If the motion occurs at very small Mach number. (¯ 2 + ξ 2 ) x derive the farﬁeld approximation (6. A line vortex of strength is parallel to a rigid airfoil occupying −a < x1 < a.6. x2 = 0. and show that it can be attributed to two dipole sources orientated in the x1 and x2 directions. If the vortex is initially far upstream of the cylinder at a distance h from the x1 axis. in the presence of a mean ﬂow at speed U in the x1 direction. 2.154 6 Vortex–Surface Interaction Noise in Two Dimensions Problems 6 1..f. estimate the inﬂuence on the sound of vortex shedding from the trailing edge of the airfoil. Assume that there is no net circulation around the cylinder. There is no net circulation around the airfoil. The vortex is in periodic motion around the airfoil under the inﬂuence of image vortices in the absence of a mean circulation around the airfoil. . When U / h. 4.11). Section 4. −∞ < x3 < ∞. Explain the signiﬁcance of these sources in terms of the corresponding components of the unsteady force between the ﬂuid and airfoil.2. and that there is a mean ﬂow past the cylinder which has speed U in the x1 direction when x1  a.1. Example 2). A line vortex of strength is parallel to a rigid airfoil occupying −a < x1 < a. Starting from the formula G≈ ¯ x·Y ∂ 2π c0 ∂t ∞ 0 δ t −τ − ¯ 2 + ξ 2 x c0 dξ .5) for the dipole component of the twodimensional compact Green’s function by writing δ t −τ − ¯ 2 + ξ 2 x c0 1 = 2π ∞ e −∞ −iω(t−τ − √ ¯ 2 +ξ 2 x c0 ) dω. calculate the sound produced when the vortex passes the airfoil for different values of the nondimensional velocity ratio /U h. Investigate the production of sound by the low Mach number motion of a line vortex of strength that is parallel to a rigid circular cylinder of radius a whose axis coincides with the x3 coordinate axis (c. Calculate the sound produced when the motion occurs at a very small Mach number. −∞ < x3 < ∞ in ﬂuid at rest at inﬁnity. The vortex is initially far upstream of the airfoil at a vertical standoff distance h above the plane of the airfoil. x2 = 0. 3.
Problems 6 155 5. If the distance of the vortex from the wall is h when the vortex is far upstream of the barrier. thin rigid barrier of length d at right angles to a plane wall at x2 = 0. and how they contribute to the radiation. There is a low Mach number mean potential ﬂow over the barrier that has speed U parallel to the wall at large distances from the barrier. A line vortex of strength traverses a path of the kind illustrated in the ﬁgure past a twodimensional. . calculate the sound produced as the vortex passes the barrier for different values of /U h. Explain what happens when /h U . Discuss the forces exerted on the barrier by the ﬂow.
7.1) 156 . the wake vorticity is conﬁned to a vortex sheet downstream of the trailing edge. that is. quadrupoles are neglected and vorticity convects as a frozen pattern of vortex ﬁlaments at the undisturbed mean stream velocity U = (U. homentropic ﬂow past a stationary rigid airfoil (Fig.1.1). and the airfoil chord can be assumed to be acoustically compact. The vortex sound source div(ω∧v) includes vorticity in the gust together with any vorticity shed from the airfoil. The undisturbed ﬂow has speed U in the x1 direction. with x3 along the span and x2 vertically upward. the linearized form of the vortex sound equation (5.7 Problems in Three Dimensions 7. and the airfoil regarded as a rigid lamina in the plane x2 = 0. or as tip vortices responsible for the mean lift. to be small. 2 c0 ∂t 2 (7. 0. thickness. In particular. where the origin is at a convenient point within the airfoil. twist. In this approximation.5) becomes 1 ∂2 − ∇ 2 B = div(ω ∧ U). and the perturbation velocities caused by airfoil thickness. 0). twist. convected in high Reynolds number. camber. and the angle of attack. When convection of sound by the ﬂow is neglected. camber.1. When div(ω ∧ v) is expanded about the undisturbed mean ﬂow. In other words. The Mach number M = U/c0 is sufﬁciently small that convection of sound by the ﬂow can be ignored. and angle of attack may ignored.2. where curl u = ω. by requiring the gustinduced velocity. either in response to excitation by the gust. The problem can be linearized by assuming that u U .1 Linear Theory of Vortex–Airfoil Interaction Noise Consider an inhomogeneous ﬁeld of vorticity. only the gust vorticity and additional vorticity shed when the gust encounters the airfoil contribute to the acoustic radiation to ﬁrst order. a gust.
4.7.4) reduces to ∂ ∂t ω3 ∂Y2 ∂Y2 − ω2 ∂ y2 ∂ y3 d 3 y.1. with both the leading and trailing edges at right angles to the mean ﬂow (so that ∂Y2 /∂ y3 ∂Y2 /∂ y2 ). t) = p(x. t − τ ) d 3 y dτ . y. Y3 = y3 .1. Let the interaction occur at an inboard location where the chord may be regarded as constant.1 Linear Theory of Vortex–Airfoil Interaction Noise 157 Fig. t) ≈− ρ0 (ω ∧ U)(y. high Reynolds number frozen gust ω(x−Ut) ≡ ω(x1 −U t. t−x/c0 −ρ0 U cos 4πc0 x (7. Equation (5. τ ) · ∂G (x. x3 ) is swept past the airfoil of Fig.1 is therefore given by (5. At sufﬁciently small Mach numbers G may be approximated by Green’s function for an airfoil of compact chord. Thus. 7. and the origin is taken in the airfoil within the interaction region.4. with solution B(x.1. x → ∞.3) where = cos−1 (x2 /x) is the angle between the radiation direction and the normal to the airfoil. The planform in the interaction .2) where ∂G/∂ y2 = 0 on both sides (y2 = ±0) of the projection of the airfoil planform onto the y1 . and the Kirchhoff vector Y by Y1 = y1 .1. ∂y (7. x2 . because p(x. 7. y3 plane. The sound produced when a localized. x → ∞. ω d 3 y ≡ 0.4) with the convection velocity v replaced by U. t) ≈ ∗ Y2 = y2 − ϕ2 (y).1.
Let the vortex have circulation and be orientated with its axis in the direction of the unit vector n. 7.1.1. Any point x on the vortex can then be represented in the parametric . t) ≈ −ρ0 U cos 4π c0 x ∂ ∂t ω3 ∂Y2 ∂ y2 d 3 y. t−x/c0 (7. x → ∞. ∂ y2 where −F2 is the unsteady airfoil lift during the interaction when the motion is regarded as incompressible. this result can also be expressed in the form p(x. The mean ﬂow speed is sufﬁciently large that the vortex can be assumed to maintain its rectilinear form after being cut by the leading edge of the airfoil. Choose the origin on the airfoil midchord such that the axis of the vortex passes through the origin at time t = 0.3) becomes p(x.4) which reveals that only the spanwise component of vorticity contributes to the production of sound.3) (in which dU j /dt = 0 for a stationary airfoil). such that 2a is equal to the local chord of the airfoil.5) F2 (t) = −ρ0 U ω3 (y. x → ∞. (7. region is then locally the same as that of the twodimensional airfoil of Fig.1.2) for constant a ≡ a(y3 ). 7.2. as indicated in Fig.2 Blade–Vortex Interactions in Three Dimensions The calculations can be performed explicitly for a gust in the form of a rectilinear line vortex. 7. Then. c0 (7.2.158 7 Problems in Three Dimensions Fig.4. and Y2 can be approximated as in (3. t) ∂Y2 (y) d 3 y.2. t) ≈ ∂ F2 cos 4πc0 x ∂t t− x . and the vorticity ω3 includes contributions from the impinging gust together with any shed into the vortex sheet wake.1. 7. According to Equation (5.9.2.1.
0 < φ < 2π. Because of vortex shedding from the trailing edge. Equation (7.2. that is. with the branch cut for the square root taken along the z axis from z = −a to z = +∞.1). where the polar angles θ.2 it will be seen that the last line of (7. sin θ sin φ.2. 0 < θ < π. cos θ). 0.4) becomes √ i 2a −n 3 ρ0 U cos ∂ p(x. The inﬂuence of the shed vorticity can be formally included by the procedure described in Sections 6. −∞ (7. Then. t) ≈ Re δ(s⊥ ) √ d 2 s⊥ ds 4πc0 x ∂t 2 y1 + i y2 + a = = = −n 3 ρ0 U cos √ 4 2πc0 x n 3 ρ0 U 2 cos √ 8 2π c0 x −n 3 ρ0 U 2 cos √ 4 2πc0 x √ i a ∂ Re ds √ ∂t U [t] + s(n 1 + in 2 ) + a −∞ √ ∞ i a Re ds 3 −∞ (U [t] + s(n 1 + in 2 ) + a) 2 √ i a Re 1 (U [t] + s(n 1 + in 2 ) + a) 2 (n 1 + in 2 ) ∞ t−x/c0 ∞ . most of the sound is generated when the vortex is cut by the leading edge. −∞ < s < ∞. By referring to Fig.1) where s is distance measured along the vortex from its point of intersection with the plane of the airfoil. z = y1 + i y2 .4 by expanding Y2 about its singularity at the leading edge.2. φ in Figure 7.2 Blade–Vortex Interactions in Three Dimensions form x = (U t. 7. if s⊥ denotes vector distance measured in the normal direction from the vortex axis.2. by setting √ √ Y2 ∼ Re( 2a z + a).3 and 6.2. . 0) + sn.1 deﬁne the orientation of the unit vector n = (sin θ cos φ.1. ω = nδ(s⊥ ).2. 159 (7.2) is zero when U [t] + a < 0.7. Thus.2) where [t] = t − x/c0 is the retarded time. by recalling the Relation (7.
for example. Hence.2. the acoustic pressure becomes p(x. as indicated in the ﬁgure. t) = 3 U M cos /x) 2 π tan θ 2 U [t] a U [t] a +1 +1 (ρ0 for θ = 85◦ .2. +0 < s < ∞.3. At later times the integration along the vortex axis over the inﬁnite range −∞ < s < ∞ must be split.160 7 Problems in Three Dimensions Fig. which is a plot of sinφ H p(x. The inﬁnite singularity in the pressure is absent for a vortex of nonzero core radius R. π R2 2 as a function of distance s⊥ from the vortex axis. when n 2 > 0 the square root is real and positive on the upper surface (s = +0) and real and negative at s = −0. t) ≈ ρ0 U M cos 2 π x 3 2 sinφ H tan θ U [t] a U [t] a +1 +1 .3) The pressure pulse accordingly begins with a singular peak at the instant at which the vortex is severed by the leading edge of the airfoil at the retarded time U [t]/a = −1. φ = 90◦ . for example. (7. 7.2. If. before the vortex is cut by the leading edge of the airfoil. The waveform is illustrated by the dotted curve in Fig. into the two parts −∞ < s < −0. because the square root in the last line of (7.2. the vorticity is assumed to be distributed according to the Gaussian formula ω(x) = ne−(s⊥ /R) .2. the acoustic pressure is found . say. 7.2) is discontinuous across the airfoil.
θ = 85◦ . when the ﬁnite size of the vortex core is .2. and I± 1 are modiﬁed Bessel functions of the ﬁrst kind. t)  sin φ 2 ≈ U M cos /x) 8 tan θ 3 (ρ0 a πR 1 2 (α). x → ∞. It is an odd function of the retarded time [t]. φ = 90◦ is plotted as the solid curve in Fig. The reader can easily show that.1a. 7.e.. The brokenline curve in Fig. (7.2. The pressure signature 4 predicted by (7.3. α= 2a sin φ R U [t] a +1 .2.4) where (α) = α 2 I− 1 4 1 α2 8 + sgn(α)I 1 4 α2 8 e−α 2 /8 . to be given by (Howe 1998a) p(x. and differs negligibly from the line vortex prediction when U [t]/a > −1. 7. 7.2 Blade–Vortex Interactions in Three Dimensions 161 Fig.3 represents the pressure signature produced by the potential ﬂow interaction between the ﬁnite core vortex and airfoil (i. when vortex shedding is ignored).3. for R a.7. The large negative peak produced as the vortex crosses the trailing edge (at U [t]/a = 1) is cancelled by an equal and opposite contribution generated by the wake.4) for R = 0.2.2.
2x3 (7. the vortex element initially on x3 = 0 remains on this plane of symmetry as it convects past the sphere along a mean streamline. The shape of the distorted vortex will be symmetric with respect to the midplane x3 = 0. more a) are unaffected and remain parallel to the distant parts of the vortex (at x3  x3 direction during the whole of the interaction. It is not generally possible. dt ∂ x3 . trailing edge generated pressure pulse is given by p(x.2) Suppose a line vortex of strength is initially far upstream of the sphere and parallel to the x3 axis at a distance h above the plane x2 = 0. The shape of the vortex at time t is determined by the solution of the equations ∂ X1 d x1 =U (x). as illustrated in Fig.1) X i = xi 1 + a3 . tan θ [t] 1 − Ua 7.3 Sound Produced by Vortex Motion near a Sphere The sound generated when a nominally rectilinear vortex is swept past a compact rigid body can also be treated in a linearized fashion. dt ∂ x2 ∂ X1 d x3 =U (x). 7. by assuming that each element of the vortex core is convected along a streamline of the steady undisturbed mean ﬂow at the local mean velocity. To illustrate the procedure. however.9. except perhaps for streamlined body shapes that are amenable to treatment by the strip theory of unsteady aerodynamics. The part of the vortex that passes close to the sphere must evidently be deformed to pass around the sphere. which has the general form (Table 3. the potential ﬂow. The mean velocity at ﬂow which is in the x1 direction at speed U for x x can therefore be written U = U ∇ X 1 (x).3.162 7 Problems in Three Dimensions ignored. dt ∂ x1 ∂ X1 d x2 =U (x). which shows the motion in the plane x3 = 0.1. The vortex is convected toward the sphere by the mean ﬂow.1) where X 1 (x) is the x1 component of the Kirchhoff vector for the sphere.3. to include the inﬂuence of vortex shedding in a satisfactory manner.3. consider a rigid sphere of radius a with center at the coordinate origin in the presence of a low Mach number irrotational mean a. t) ≈ − ρ0 U M cos 2 2 π x 3 U [t] sinφ H 1 − a . (7.
¯n where x3 = 10n/N . say. these equations are to be integrated subject to the initial conditions x1 = U t. ¯ x2 = h . a ¯ ¯n x3 = x3 at T = −10. 0 x3 = x3 t → −∞.7. and N is a suitably large integer. These are solved (for example. 0 where x3 is the initial spanwise location of the vortex element. by the Runge–Kutta method described in Section 4. In terms of the nondimensional variables T = Ut . 7.1. .3. Because the motion is symmetric about x3 = 0 the solutions are required only for the N + 1 vortex elements with respective the initial positions ¯ x1 = T . It can be safely assumed that the sphere has no perceptible inﬂuence on vortex elements initially located ¯ ¯ at x3  > 10. If the undistorted parts of the vortex (at x3  a) are assumed to convect across the plane x1 = 0 at time t = 0. x2 = h. . a the equations of motion of a point on the vortex are found to be ¯ ¯2 ¯2 ¯2 d x1 x2 + x3 − 2x1 = 1+ 5 . a ¯ x= x . 0 ≤ n ≤ N .3 Sound Produced by Vortex Motion near a Sphere 163 Fig.6) by starting the integration at T = −10. for each element of the vortex. dT ¯2 ¯2 ¯2 2 x1 + x2 + x3 2 ¯ ¯ ¯ −3x1 x3 d x3 = dT 2 ¯ ¯2 ¯2 2 x1 + x2 + x3 5 2 ¯ ¯ ¯ −3x1 x2 d x2 = dT ¯2 ¯2 ¯2 2 x1 + x2 + x3 5 2 .
2. (7.3) where v = U ∇Y1 (y). 7. In reality. none of which is accounted for in the present calculation. .3.164 7 Problems in Three Dimensions Fig. Figure 7.2 illustrates successive calculated positions of the vortex with increasing values of the time T = U t/a for the case h/a = 0. and by viscous diffusion of vorticity from the vortex and from the surface of the sphere. t − x c0 · ∇Y j (y) d 3 y.2. the motion would be strongly inﬂuenced by large selfinduced velocities.3.3. the accelerated motion over the upper surface of the sphere is insufﬁcient to counteract the formation of the loop. t) ≈ −ρ0 x j ∂ 4πc0 x2 ∂t (ω ∧ v) y. The hairpin loop is formed because the translation velocities of vortex elements close to the sphere are small in the neighborhood of the stagnation points just in front and just to the rear of the sphere.4. by image vorticity in the sphere.4) when the sphere is compact: p(x. x → ∞. The distortion of the vortex ﬁrst becomes evident at about T = −2. of course. The sound generated during this potential ﬂow interaction can be calculated using the Formula (5.
The sound is accordingly produced by a dipole source orientated in the x2 direction.3. ∂ p(x. Similarly. because s · ∇Y1 ∧ ∇Y2 → 1 as ¯ s → ± ∞.3. t − x d 3 y. ˆ where s is a unit vector locally parallel to ω. but a consequence of the formal operations used in the application of the compact Green’s function. and give no contribution to the sound when differentiated.4) where = cos−1 (x2 /x) is the angle between the observer direction x and the x2 axis. and disappear on differentiation with respect to T . and s is distance measured along the vortex in the direction of ω. however. Typical plots of the calculated nondimensional pressure (7. The divergence is not real. there can be no net sideforce on the sphere because of the symmetric form of the vortex.3. which has the representation p(x. a M= U . suitable for numerical evaluation.3.3 for two values of h/a.5) is divergent. that is. t) =− U M cos /4π x ∂T ∞ −∞ ρ0 ¯ s [ˆ · ∇Y1 ∧ ∇Y2 ] d s. x → ∞.5) are shown in Fig.3) from j = 1 because ω ∧ ∇Y1 · ∇Y1 ≡ 0. The integral (7. because the contributions ¯ at larger values of s are the same for all retarded times. The inﬁnite ¯ contributions to the integral from large values of s are equal at successive retarded locations of the vortex.5) where the integrand is evaluated at the retarded position of the distorted vortex. .3. c0 (7. c0 (7.3 Sound Produced by Vortex Motion near a Sphere 165 There is no unsteady drag contribution to (7. say. d 3 y = d 2 s⊥ ds. and therefore there will be no contribution from j = 3.3.4) can then be cast in the following nondimensional form. −10 < s < 10.7. To evaluate the integral. write s ω = δ(s⊥ )ˆ. ¯ s= s . s⊥ is the vector distance measured in the normal direction from the local axis of the vortex. 7. they illustrate how the sound level decreases as the initial standoff distance h of the vortex increases relative to the radius a of the sphere. the interaction produces an unsteady lift force in this direction which is responsible for the sound. ˆ Now the integral in (7. t) ≈ −ρ0 U cos 4πc0 x ∂ ∂t (ω · ∇Y1 ∧ ∇Y2 ) y.3. The integral can therefore be evaluated numerically by restricting the range ¯ of integration to a ﬁnite interval.
3.1. .3.4 Compression Wave Generated When a Train Enters a Tunnel A train entering a tunnel pushes aside the stationary air. In a long tunnel the compression wavefront can experience nonlinear steepening that is ultimately manifested as a loud. but the buildup of pressure just ahead of the train propagates into the tunnel as a compression wave at the speed of sound. All of these waves are indicated schematically in Fig. 7.4. 7. In addition inaudible lowfrequency pressure ﬂuctuations called infrasound (at frequencies ∼10–20 Hz) are radiated from the tunnel portal into the open air when the train enters and leaves the tunnel. 7.166 7 Problems in Three Dimensions Fig.4. 7. Their effects become pronounced when the train speed U exceeds Fig.1. most of which ﬂows over the train and out of the tunnel portal. impulsive bang or ‘crack’ (called a micropressure wave) radiating out of the distant tunnel exit.
They vary with position and time within the tunnel. where the origin O is at the center of the tunnel entrance plane. ρ and c. density. and the compression wave can be calculated using the vortex sound equation (5. c0 .2. about 200 km/h (125 mi/h).4.7. and their corresponding undisturbed values are p 0 . and the blockage A0 /A ≤ 0.2. In this . the air ﬂow may be regarded as homentropic.4. circular cylindrical duct of radius R and crosssectional area A = π R 2 (Fig. The cross section of the train is assumed to become uniform with constant area A 0 = π h 2 at a distance L from the nose of the train.5) D Dt 1 D c2 Dt − 1 1 ∇· (ρ∇) B = div(ρω ∧ v). The formation of the compression wave can be studied in terms of the simpler problem involving an axisymmetric train entering axisymmetrically a semiinﬁnite. to ensure that ﬂow separation does not occur.1) The air in the compression wave region ahead of the train may be regarded as ¯ linearly perturbed from its mean state. when in particular the micropressure wave and the infrasound can cause vibrations and annoying structural rattles in neighboring buildings. so that the x1 axis coincides with the axis of the tunnel. 7.4.2. In practice the Mach number M = U/c0 does not exceed 0. The aspect ratio h/L of the nose is taken to be sufﬁciently small.2). 7. Let the train be travelling at constant speed U in the negative x1 direction. and the train proﬁle sufﬁciently streamlined.4 Compression Wave Generated When a Train Enters a Tunnel 167 Fig. Denote the pressure. If heat transfer and frictional losses are neglected during the initial stages of wave formation. ¯ and speed of sound in the air respectively by p .4. where h is the uniform maximum train radius. ρ ρ (7. ρ 0 . with B ≈ p/ρ0 ≡ ( p − p0 )/ρ0 .
2.4. The compressibility of the air adjacent to S and within the very low Mach number exterior ﬂow from the tunnel portal can be neglected when M(A 0 /A)2 1 (Howe et al.4. x3 ) = 0.5) reduces to ∂v/∂t = −∇ B − ω ∧ v. and the inﬂuence of the train on its surroundings can be represented in terms of monopole and dipole sources on S. Let f ≡ f (x1 + U t.3. Thus. In the usual way.1) by H ≡ H ( f ) and rearrange (noting that D H/Dt = 0) to obtain D Dt = 1 D c2 Dt − 1 ∇ · (ρ∇) (H B) ρ (7.2) 1 1 div(Hρω ∧ v) − (∇ B + ω ∧ v) · ∇ H − div(ρ B∇ H ). the equation ﬁnally reduces to ∂ 1 ∂2 (U · ∇ H ) + div (v U · ∇ H ) − ∇ 2 (H B) = 2 ∂t 2 ∂t c0 − div p 1 + v 2 ∇ H + div(H ω ∧ v).4. The surface is ﬁxed relative to the train.2). if the nonlinear terms on the left of (7. multiply (7. with f < 0 inside S (in the region occupied by the train) and f > 0 outside. ∂t ∂t ρ ρ where U = (−U.168 7 Problems in Three Dimensions simple model the vorticity ω vanishes everywhere except within the outer shear layer of the exit ﬂow of the air displaced when the train enters the tunnel (see Fig. and the source approximated further by ∂ (U · ∇ H ) + div (v U · ∇ H ) − div ∂t p 1 + v 2 ∇ H + div(H ω ∧ v). x3 ) = 0 be a control surface S contained within the ﬂuid that just encloses the moving train. 0.4. 2000). x2 .3).4. 7. x2 . The two terms on the righthand side involving ∇ H respectively represent monopole and dipole sources distributed over the moving surface f (x1 + U t. ρ0 2 where the relation ∂ H ( f )/∂t = −U · ∇ H ( f ) has been used. (7. so that the source terms can be written 1 1 ∂ ∂H (U · ∇ H ) − v · ∇ − div (ρ B∇ H ) + div(Hρω ∧ v). ρ ρ This is a generalization of Equation (5. 0).2) (which affect the propagation of the compression wave) are also ignored. When frictional losses are neglected Crocco’s equation (4.3) ρ0 2 .
3).4. we can collapse the monopole source distribution over the surface of the train into a line source concentrated on its axis. This source can be simpliﬁed when the aspect ratio h/L 1 by introducing a slender body approximation.4. consider a system of cylindrical coordinates in which r = x2 + x3 is the perpendicular distance from the axis of the train. ∂ x1 Thus. U ∂AT (x1 + U t) = monopole source strength per unit length of the train.4 Compression Wave Generated When a Train Enters a Tunnel 169 Fig.3) by ∂ ∂ (U · ∇ H ) (x. t) ≈ ∂t ∂t U ∂AT (x1 + U t)δ(x2 )δ(x3 ) .4.4.3) the rate at which air is displaced by a section of the train of length d x1 is ∂ f¯ (x1 + U t) d x1 U 2π f¯(x1 + U t) d f¯(x1 + U t) = U 2π f¯(x1 + U t) ∂ x1 ∂AT ≡U (x1 + U t) d x1 . ∂ x1 When h/L 1. 7.4. and approximate the monopole on the righthand side of Equation (7.3. √ 2 2 To do this. 7. x3 ) = 0 can be written r = f¯(x1 + U t). ∂ x1 .1 Linear Theory When the blockage A 0 /A is small it is sufﬁcient to retain only the ﬁrst monopole source on the righthand side of Equation (7. The control surface equation f (x1 + U t. x2 .7. where AT (s) = π f¯2 (s) is the crosssectional area of the train at distance s from the nose. and the nose is assumed to cross the tunnel entrance plane (x1 = 0) at t = 0. As the train moves (to the left in Fig. 7.
y. y1 .f.9.4.170 7 Problems in Three Dimensions The source strength is proportional to the rate at which the train cross section changes with distance along the train.9. 0) + U [t]) − AT (y1 + Mϕ ∗ (y1 .4) where B → p/ρ0 in the linear acoustic region ahead of the train. and [t] = t + (x1 − )/c0 is the effective retarded time.4. as the train nose enters the tunnel. ahead of the train where B = p/ρ0 . This occurs over a time ∼R/U . before the onset of nonlinear steepening. Because nonlinear propagation terms have been ignored. Equation (7. It is therefore applicable within the region several tunnel diameters ahead of the train. and is nonzero only in the vicinity of the train nose (and also the tail).4. t − τ ) dy1 dτ ∂ y1 {AT (y1 − Mϕ ∗ (y1 .4. 0. The compression wave is produced when the near ﬁeld of the source interacts with the tunnel portal. Question 4 of Problems 1 when q0 (t) = constant).13) for a duct entrance G(x. 0. during and just after tunnel entry. so that the characteristic thickness of the wavefront ∼R/M R. t − τ ) ≈ c0 ϕ ∗ (x) − ϕ ∗ (y) H t −τ − 2A c0 −H t −τ + ϕ ∗ (x) + ϕ ∗ (y) c0 .14)). we have p≡ p t+ = ρ0 U c0 2A x1 c0 ∞ −∞ ≈ ρ0 ∂ ∂t ∞ U −∞ ∂AT (y1 + U τ )G(x. 0) + U [t])} dy1 . Thus. at x within the tunnel. The source therefore creates only a nonacoustic near ﬁeld when travelling within the tunnel or in free space far from the tunnel entrance (c. . The corresponding approximation of Equation (7. 0. The monopole in this equation does not depend on time when viewed in a reference frame moving at the uniform subsonic speed U of the train. where ϕ ∗ (x) is the velocity potential of a uniform incompressible ﬂow out of the tunnel portal that has unit speed far inside the tunnel (see (3. ∂ x1 (7.4) can therefore be solved by using the compact Green’s function (3.. this approximation determines the initial form of the compression wave proﬁle.3) is therefore ∂ 1 ∂2 − ∇ 2 (H B) = 2 ∂t 2 ∂t c0 U ∂AT (x1 + U t)δ(x2 )δ(x3 ) . 0. (7.5) where the prime on AT denotes differentiation with respect to the argument.
and may be calculated by temporarily considering a train of semiinﬁnite length. 0) dy1 . 0.4. The data points in the ﬁgure are measurements (made at the point labelled T) of the pressure gradient d p/dt 1 m from the entrance for these three different . 1. This implies that the Approximation (7. 0 < s < L. During the formation of the wave.9. (7. asymptotic pressure rise can also be calculated exactly.4. Wireguided.6) predicts the overall (linear theory) pressure rise to be p ≈ ρ0 U 2 A 0 /A.7.4. 0.4. ∂ y1 ∂ y1 M2 1. and in the particular case in which the Mach number is small enough that terms ∼O(M 2 ) are negligible.2. But the linear theory. The nose aspect ratio h/L = 0. 2000).147 m. the blockage A 0 /A = 0.4.4. s.4 Compression Wave Generated When a Train Enters a Tunnel 171 The main contributions to the Integral (7. with respective crosssectional areas given by AT (s) = A0 s2 .7) This extrapolation of the linear theory to ﬁnite values of M turns out to be applicable for M < 0.116. by expanding to ﬁrst order in Mϕ ∗ and integrating by parts. with no restriction on Mach number. and we then ﬁnd.5) (see Equation (3. the term Mϕ ∗ in the arguments of AT in (7. 0. ∂ y1 ∂ y1 (7.4 illustrates schematically an experimental arrangement used by Maeda et al. 0) dy1 . ∂ϕ ∗ /∂ y1 = 1 in the region occupied by the nose.5) is small.6) After the nose has passed into the tunnel. Figure 7. where the crosssectional area AT is changing. to be p= ρ0 U 2 A0 .5) are from the vicinities of the nose and tail of the train. and (7. and the projection speed U ≈ 230 km/h (M ≈ 0. (1993) to investigate the compression wave. 0) ≈ y1 − in (7. s L2 L L 2− s L .14)).188). s ≥ L.6) can be extrapolated to ﬁnite Mach numbers by writing p≈ ρ0 U 2 A(1 − M 2 ) ∞ −∞ ∂AT ∂ϕ ∗ (y1 + U [t]) (y1 . and the paraboloid and ellipsoid of revolution.4 (Howe et al.4. axisymmetric model trains are projected into and along the axis of a tunnel consisting of a 7m long circular cylinder of internal diameter 0. The train nose proﬁles include the cone. The compression wave is generated as the nose enters the tunnel.4. that p≈ ρ0 U 2 A ∞ −∞ ∂AT ∂ϕ ∗ (y1 + U [t]) (y1 . A(1 − M 2 ) because this is attained when ϕ ∗ (y1 .
172
7 Problems in Three Dimensions
Fig. 7.4.4.
nose proﬁles. The solid curves are predictions of Equation (7.4.7), evaluated using the following formulae for ∂ϕ ∗ /∂ y1 (Howe 1998b): 1 1 ∂ϕ ∗ (y) = − ∂ y1 2 2π Z(ξ ) = 1 π
∞ ∞
I0
0
ξr R
2K 1 (ξ ) I1 (ξ )
1 2
sin ξ
y1 + Z(ξ ) R
dξ,
ln
0
K 1 (µ)I1 (µ) K 1 (ξ )I1 (ξ )
dµ , µ2 − ξ 2
2 2 where r = y2 + y3 < R and I0 , I1 , and K 1 are modiﬁed Bessel functions. The linear theory underpredicts the maximum observed pressure gradients by about 8%. The agreement with experiment can be greatly improved by including contributions from the surface dipoles in Equation (7.4.3) (which in a ﬁrst approximation are determined by the drag force exerted on the nose of the train by the linear theory pressure rise) and, to a lesser extent, by including the vortex sound generated by the tunnel exitﬂow vorticity (the ﬁnal source term on the right of (7.4.3)).
Problems 7 1. The term ω2 ∂Y2 /∂ y3 in the Representation (7.1.3) of the sound produced by a gust interacting with a thin airfoil accounts for the inﬂuence of changes in
Problems 7 the airfoil chord 2a(y3 ) over the interaction region. Show that when da (y3 ) dy3 1
173
the Formula (7.2.3) for the sound produced by a line vortex is given in a ﬁrst approximation by p(x, t) ≈ ρ0 U M cos 2 π x
3 2
sinφ sin θ da dy3 H
0 U [t] a U [t] a
× cos θ − sin θ cos φ
+1 +1
,
where da/dy3 is evaluated at y3 = 0, where the vortex is cut by the airfoil. Show that this result is identical with that given by (7.1.4) provided that in (7.1.4) ω3 is interpreted as the component of the vorticity parallel to the local leading edge of the airfoil and the convection velocity U is replaced by its component normal to the local leading edge. 2. A vortex ring orientated with its axis parallel to the +x1 axis is convected in a low Mach number mean ﬂow at speed U in the x1 direction past the edge of the rigid halfplane x1 < 0, x2 = 0, −∞ < x3 < ∞. Use the compact Green’s function G(x, y, t − τ ) =
∗
ϕ ∗ (x)ϕ ∗ (y) 1 √ 2π 2 2πic0 x − y3 i3 3/2
∞ −∞
√
ωe−iω(t−x−y3 i3 /c0 ) dω,
where ϕ is deﬁned as in (3.9.6), to calculate the sound produced as the vortex passes the edge when the inﬂuence of the halfplane on the motion of the ring is ignored.
3. Calculate the sound produced within and outside a semiﬁnite circular cylindrical rigid pipe when a vortex ring exhausts axisymmetrically from the open end. Neglect the inﬂuence of the pipe walls on the motion of the vortex and ignore any change in the vortex radius at the exit. 4. Determine the (quadrupole) sound produced by the headon collision of two equal ring vortices. Estimate the sound generated when a ring vortex is incident normally on a plane wall.
174
7 Problems in Three Dimensions
5. Use the Green’s function (3.9.15) and Equation (7.4.4) to determine the infrasound generated by a train entering a tunnel modeled by the unﬂanged, circular cylindrical duct in Fig. 3.9.6b. Assume that the train travels along the axis of the duct and show that the acoustic pressure at the far ﬁeld point x outside the tunnel is given approximately by p(x, t) ≈ ρ0 U 2 M 4π x ×
∞ −∞
1−
x1 x
∂AT ∂ 2ϕ∗ (y1 + U [t]) 2 (y1 , 0, 0) dy1 , x → ∞, ∂ y1 ∂ y1
where [t] = t − x/c0 .
8
Further Worked Examples
8.1 Blade–Vortex Interactions in Two Dimensions The linear theory of the low Mach number, twodimensional interaction of a line vortex with an airfoil was discussed in Section 6.4. The interaction will now be examined in more detail, including also the inﬂuence of image vortices on the motion. The general problem to be considered is depicted in Fig. 8.1.1, which shows a vortex of strength moving in the neighbourhood of a rigid airfoil of chord 2a occupying −a < x1 < a, x2 = 0. There is no mean circulation about the airfoil. We shall consider cases with and without a mean ﬂow in the x1 direction and examine the inﬂuence of vortex shedding from the trailing edge.
8.1.1 Equation of Motion of the Vortex At time t let the vortex be at x ≡ (x1 , x2 ) = x0 (t), and translate at velocity v0 = If we set z = x1 + i x2 , z 0 = x01 + i x02 , the transformation ζ = z + a z2 −1 a2 (8.1.1) dx0 (t). dt
maps the ﬂuid region in the z plane of the airfoil into the region ζ  > 1 in the ζ plane. The upper and lower faces of the airfoil (x2 = ± 0) respectively transform into the upper and lower halves of the unit circular cylinder ζ  = 1, and the vortex maps into an equal vortex at ζ = ζ0 (Fig. 8.1.2). In the absence of mean ﬂow (U = 0), and when there is no mean circulation about the cylinder (and therefore about the airfoil), the complex potential of the motion is ob∗ tained by placing an image vortex − at the inverse point ζ = 1/ζ0 together with 175
w(z) = − where i ln(ζ (z) − ζ (z 0 )) + F(z). Because a mean ﬂow in the x1 direction is unaffected by the airfoil.6. 8. 8.1).1. F(z) = ∗ 2π ζ (z 0 ) 2π This is of the form given in (4. we can include its contribution by adding the complex potential U z. Then. 2π i 1 i − ln ζ (z) − ln ζ (z) + U z. 2π The velocity potential of the motion in the z plane is given by setting ζ = ζ (z).176 8 Further Worked Examples Fig.1. Then. a vortex + at the center of the cylinder. w(ζ ) = − 1 i i ln(ζ − ζ0 ) + ln ζ − ∗ 2π 2π ζ0 − i ln ζ. so that the corresponding equation of motion Fig. .1. The two interior vortices ensure that the total circulation around the cylinder vanishes.2.
1. dt dt dt 4π ζ (z 0 ) 177 that is ∗ i dz 0 Z 2 = −1+ + U. In a linearized calculation (in which image effects in the airfoil are neglected) this could be done by assuming shed vorticity to lie in a thin vortex sheet downstream of the trailing edge at x1 = a. we can write p(x.1. t) ≈ ρ0 sin ∂ √ 2π 2c0 x ∂t × ∂ ∂t t−x/c0 −∞ t−x/c0 −∞ for an observer at x in the far ﬁeld as in dτ d x02 ρ0 cos (τ ) √ − √ dτ t − τ − x/c0 2π 2c0 x dτ . Instead. where the effects of vortex shedding are modelled by deleting singularities from the compact Green’s function.2) takes no account of the inﬂuence of vortex shedding.9. 8.2). but we shall not do this.1) Y1 = y1 .2. where Z = .1. t) ≈ t−x/c0 −ρ0 x j ∂ √ 3 2π 2c0 x 2 ∂t −∞ dτ ×√ .2 Formula for the Acoustic Pressure The sound produced by the low Mach number motion of the vortex is calculated from (6.3) and the Formula (6.1. Y2 = Re(−i z 2 − a 2 ).6).4) By deﬁning the radiation angle Fig.3.4): p(x. t − τ − x/c0 d x01 ∂Y j d x02 ∂Y j − dτ ∂ y2 dτ ∂ y1 x0 (τ ) (8.8.3) to be ∗ dz 0 d x01 d x02 i ζ (z 0 ) ≡ −i =− + F (z 0 ). (8.6. √ t − τ − x/c0 d x01 ∂Y2 d x02 ∂Y2 − dτ ∂ y2 dτ ∂ y1 x0 (τ ) The two integrals in this formula represent the acoustic ﬁelds of dipole sources.1. (8. 8. the strengths .3) where the Kirchhoff vector for the strip airfoil has the components (Table 3.4.1.2.3. a Equation (8.1. According to Curle’s theory (Section 2. we shall apply the method discussed in Section 6.1 Blade–Vortex Interactions in Two Dimensions of the vortex at z 0 is found from (4.2) √ √ dt ζ0 2 − 1 4πa Z 2 − 1 Z2 − 1 z0 ζ0 = Z + Z 2 − 1. as in Section 6. This would lead to a solution in terms of the Sears function (6. because it limits the discussion to linearized motions. z = y1 + i y2 .
1. other things being equal. Only the second integral in (8.1. F2 ) exerted on the ﬂuid (per unit span) by the airfoil. Introduce the shorthand notation W= −i Z d (−i z 2 − a 2 ) = √ dz Z2 − 1 (8. because the second integral in (8.6) is dominated by contributions from the neighbourhoods of the singularities.1. and it is understood that Z = Z (τ ). 8.4.1. By deleting the .3 Linear Theory In the linearized approximation the vortex is swept past the airfoil along a trajectory parallel to the x1 direction at precisely the uniform mean ﬂow speed U .2) and (8.6) contributes to the sound (because d Z /dτ = U/a is real and F1 ≡ 0). When the standoff distance h a it was argued in Section 6. (8.1.1.6) for the vortex motion and the acoustic pressure will now be applied to several different special cases.4 that the inﬂuence of vortex shedding from the trailing edge could be estimated by deleting the singularity that occurs at the edge from the Green’s function and ignoring the shed vorticity. and the trailing edge singularity corresponds to the singularity of W(Z ) at Z = 1.5) evaluated at the vortex. Then.1.178 8 Further Worked Examples of these dipoles are determined by the unsteady force (F1 . This is the case illustrated in Fig.2). Equations (8. Im −∞ [t] −∞ dZ dτ √ dτ [t] − τ dZ dτ √ dτ [t] − τ .6) Im W(Z ) where [t] = t − x/c0 is the retarded time. t) ≈ ∂ ρ0 a sin √ 2π 2c0 x ∂t + cos [t] ≡ −a Im W(Z ) x0 (τ ) dZ dτ . In general the integrals must be evaluated numerically using the solution of the equation of motion (8. are therefore the most signiﬁcant sources of sound at high frequencies. the second component F2 is equal and opposite to the unsteady lift experienced by the airfoil during the interaction.1. 6. d x01 ∂Y2 d x02 ∂Y2 − dτ ∂ y2 dτ ∂ y1 and the acoustic pressure becomes p(x. The ﬁrst is aligned with the airfoil chord (the mean ﬂow direction) and represents the inﬂuence of suction forces at the leading and trailing edges (Batchelor 1967). W(Z ) is singular at both the leading and trailing edges (Z = ±1).1. which.
1.1.7) √ where the branch cut for Z + 1 runs along the real axis from Z = −1 to Z = +∞. and √ changing the integration variable to ξ = 1/ t − τ − x/c0 .8) The nondimensional acoustic pressure signature (the righthand side of (8. . which can be written. (8. Making this substitution in (8.6). 8.8)) is plotted as the solid curves in Fig.3).3. ρ0 U √ p(x.0.1 Blade–Vortex Interactions in Two Dimensions 179 contribution from the trailing edge we are asserting that all of the sound is produced by the interaction of the vortex with the leading edge. so that Z = U τ/a + i h/a. Fig. x → ∞.1.1.4. we then recover the result (6. 8. measuring time from the instant that the vortex crosses the midchord x1 = 0 of the airfoil. 1. t) M cos (a/x) 1 2 4πa ≈ U [t] a U [t] a +1 2 +1 + h 2 a .2.3 for h/a = 0.5.1. Near this edge 1 W(Z ) ≈ √ √ .1.8. 0. 2 Z +1 (8.
when the characteristic reduced frequency λ = ωa/U of the motion is relatively small. 8.10) which can be solved for Z by Runge–Kutta integration (Section 4.2.6).1.3 that the linearized problem of determining the unsteady force F2 exerted on the ﬂuid when an incompressible.2. and might be expected to lie outside the range for which (8. and must be determined by numerical integration of Equation (8.3.180 8 Further Worked Examples It was pointed out in Section 6.1.1.8) is valid. say L = 10a.1.3. (8. The agreement with the approximate theory of Equation (8.1.8) is remarkably good even when h/a is as large as unity.2. t) M cos (a/x) 1 2 ρ0 U 4πa ≈ √ ∞ 2π Re 0 (iλ) 2 S(λ)e−λ{h/a+iU [t]/a} dλ. If the initial standoff distance is h at x1 = −∞. h = 0. To do this it is convenient to introduce a dimensionless velocity ratio and time T deﬁned by = . a . (8. 1 x → ∞. 8. 8.2). which in the present case can be shown to predict that √ p(x.2) becomes d Z∗ i =√ dT Z2 − 1 √ Z Z2 −1 −1+ 2 ζ0 − 1 2 + 1.9) The corresponding pressure signatures are plotted as the dotted curves in Fig.1.4 shows a calculated trajectory for = 0.2) (because F1 ≡ 0). by prescribing the initial position of the vortex to be Z = −L/a + i h/a.1. This force also determines the low Mach number acoustic radiation by Equation (6.6). a 4πaU in terms of which (8. sinusoidal gust convects past the airfoil can be solved exactly with full account taken of vortex shedding.4 Nonlinear Theory When account is taken of image vortices in the airfoil the trajectory of the vortex in the neighbourhood of the airfoil is no longer parallel to the mean ﬂow direction. in terms of the Sears function (6. the integration is started at a large distance L upstream of the airfoil midchord. The upper part of Fig.1. T = Ut .1.
The nonlinear inﬂuence of the image vorticity is to shift the initially rectilinear trajectory of the vortex away from the airfoil in the direction of the vortex force ω ∧ U (U = U i). ˆ dT Uτ ˆ . The vortex is closest to the airfoil at U t/a = 0.4.2 U ∼U 1+ x02 /a 0.28a.71U.28 = 1. where T = a . v(T )) by dZ ˆ ˆ = u(T ) + iv(T ).1 Blade–Vortex Interactions in Two Dimensions 181 Fig. 8.1.7) for W(Z ). where time is measured from the instant that the vortex passes the midchord of the airfoil. The integrals must be evaluated numerically. where x02 ∼ 0.8. 4π x02 The sound generated as the vortex passes the airfoil is given by (8. and this ˆ ˆ is done by deﬁning a dimensionless vortex convection velocity (u(T ). and where convection by the images increases the translation speed of the vortex from U to approximately U+ =U+ 0.6).1.1. The inﬂuence of vortex shedding into the wake is included by using the approximation (8.
and the pressure signatures have proﬁles similar to those depicted by the solid curves in the ﬁgure. ρ0 U √ 1 8 Further Worked Examples p(x. obtained by interpolating smoothly between the calculated pressures on either side of the blips. An interesting nonlinear interaction occurs when the initial standoff distance of the vortex h = 0 (Fig. However.182 Then. t).1. the effect of shedding was not included in the calculation of the vortex trajectory. their nondimensional forms ρ0 U √ p1 (x. But the calculated pressures exhibit blips shown as dotted curves in the ﬁgure. The components p1 (x. √ √ 1 1 U M(a/x) 2 4πa ρ0 U M(a/x) 2 4πa (8. at which time the . the smoothing inﬂuence of shedding at a sharp edge acts to remove the blips. Vortex shedding should smooth out the pressure signatures at the retarded times when the vortex is close to the trailing edge.7)).11) correspond respectively to the dipole sound produced by the unsteady suction and lift forces. and the integration variable T has been reˆ placed by λ = [T ] − T . [T ] = U [t]/a. In the linearized approximation. although our calculation has accounted for vortex shedding in evaluating the dipole source strengths (by means of the approximation (8. t) of (8. These arise because. the vortex would strike the leading edge of the airfoil at U [t]/a = −1. The ﬁnal integrals are easily evaluated numerically when the path of the vortex has been determined. t) M(a/x) 2 4πa [T ] −∞ ∞ 0 1 ≈ 22 3 d sin dT d dT ˆ ˆ v(T ) d T ˆ [T ] − T + cos [T ] −∞ ˆ ˆ Im(W(Z )(u + iv))(T ) d T ˆ [T ] − T = 2 2 sin + 2 2 cos ≈ ρ0 3 v([T ] − λ2 ) dλ ∞ d Im(W(Z )(u + iv))([T ] − λ2 ) dλ dT 0 p1 (x.1.5). p2 (x. 8.11) ˆ where T = U t/a.4. 8. t) p2 (x. because the source terms must be set to equal zero as soon as [T ] − λ2 reduces to the nondimensional time at which the computation of the vortex path begins (where the vortex is sufﬁciently far upstream that it effectively produces no sound by interaction with the airfoil). t) M cos (a/x) 2 4πa 1 are plotted in Fig.1. t) M sin (a/x) 1 2 4πa and ρ0 U √ p2 (x. t) + .1. The upper limit of integration is actually ﬁnite.1.
3U. The maximum convection velocity of the vortex (at U t/a = 0) is now more than twice the mean stream velocity: U+ 4π(0.5 illustrates this for the same value of the velocity ratio = /4πaU = 0.5.1.15 = 2.1. a . linear theory acoustic pressure (8.8.5 Periodic Vortex Motion When there is no mean ﬂow (U = 0) the characteristic velocity and dimensionless time become V = 4πa . 8.2 considered above.and liftdipole acoustic pressures p1 and p2 shown in the ﬁgure are also greatly increased. This singular event does not occur because the vortex trajectory is deﬂected around the airfoil by the image vorticity (for a rounded nose the possibility of additional vortex shedding from the leading edge may be ignored). The upper part of Fig.1 Blade–Vortex Interactions in Two Dimensions 183 Fig.1. 8.1.15a) =U 1+ 0.8) is inﬁnite.2 0. T = Vt . The corresponding suction. 8.
and the vortex equation of motion (8. the nondimensional .1.6. where x01 = −2a.184 8 Further Worked Examples Fig. for the case where the trajectory passes through the point labelled 0. The solutions are closed trajectories orbiting the airfoil periodically.6 for > 0.1.1.84.6) for the acoustic pressure. The calculated period is T0 ≡ V t0 /a ≈ 35. 8. x02 = 0.2) reduces to i d Z∗ =√ dT Z2 − 1 √ Z Z2 − 1 −1+ 2 ζ0 − 1 2 . An orbiting vortex motion of this kind cannot be realized in practice (because of diffusion from the vortex core and the continual shedding of additional vorticity from the airfoil). By writing Vτ ˆ T = a and dZ ˆ ˆ = u(T ) + iv(T ) ˆ dT in the general formula (8. 8.1. A typical orbit is plotted in the upper half of Fig. but it is still instructive to calculate the sound produced by the motion.
12) where M = V /c0 .6.1.11) p2 (x. 8. 2πnT T0 dT. the suction and dipole source strengths have period T0 . and possess Fourier series expansions of the form v(T ) = ∞ an cos n=1 2π nT T0 .5) in the absence of vortex shedding. bn can be calculated by using the numerical solution for the orbit to evaluate an = bn = 2 T0 2 T0 T0 0 T0 0 v(T ) cos 2π nT T0 dT. the right −an n sin 0 ∞ sin 2πn ∞ [T ] − λ2 T0 [T ] − λ2 T0 dλ dλ .8.1. The function W(Z ) in the integrand is given by (8. Im(W(Z )(u+iv))(T ) = ∞ bn sin n=1 2πnT T0 . ˆ [T ] − T (8. where the coefﬁcients an .1. + bn n cos 0 cos 2πn The integrals are evaluated from the real and imaginary parts of ∞ 0 1 1 2 e2πin{[T ]/T0 −λ } dλ = √ e{2n[T ]/T0 − 4 }πi . that when T is measured from Z = −2. 2 2n .1.1 Blade–Vortex Interactions in Two Dimensions 185 suction and lift acoustic pressures are found to be given by the following modiﬁed form of (8. t) p1 (x. Im(W(Z )(u + iv))(T ) sin By making the change of integration variable λ = hand side of (8. It follows by inspection and from the numerical solution.1. as indicated in Fig. t) √ √ 1 + 1 2 M(a/x) 2 M(a/x) 2 ρ0 V 1 ρ0 V2 ≈ 22 d sin dT [T ] −∞ ˆ ˆ v(T ) d T ˆ [T ] − T + cos [T ] −∞ ˆ ˆ Im(W(Z )(u + iv))(T ) d T .12) now becomes √ 4 2π √ T0 ∞ n=1 √ ˆ [T ] − T / T0 .
t) ρ0 V 2 ρ0 V2 √ 2nπt π an n sin − t0 4 n=1 ∞ √ 2π ≈ √ T0 M cos (a/x) √ 2nπt π . and exhibit rapid variations at the retarded times at which the vortex is directly above and below the airfoil. .2 Parallel Blade–Vortex Interactions in Three Dimensions We now examine to what extent the simple twodimensional methods of the previous section can be adapted to wings of ﬁnite span and variable chord for problems of the kind shown in Fig. when Fig.1. both have similar orders of magnitude. 8. 8. t) 1 2 p1 (x.2. 8.2.1.186 8 Further Worked Examples Hence.6 (taking the ﬁrst 26 terms in the series). The corresponding nondimensional pressures are plotted in Fig. bn n cos − t0 4 n=1 ∞ where [ ] denotes evaluation at the retarded time t − x/c0 . x → ∞.1.1. the suction and lift force dipole ﬁelds are given respectively by √ 2π 1 ≈ −√ T0 M sin (a/x) 2 p2 (x. 8. The general representation of the sound produced by vortex–airfoil interactions is discussed in Section 7.
2. Y3 = y3 . For the rectangular ˆ ˆ airfoil a(y3 ) ≡ a = constant. 8. This smooths out conditions at the trailing edge.4): p(x. A spanwise line vortex of strength is swept past the airfoil at an initial standoff distance h above the airfoil.1. (8. so that sound is generated primarily as the vortex passes over the leading edge of the airfoil.4.2 Parallel Blade–Vortex Interactions in Three Dimensions 187 the airfoil chord can be regarded as compact.1) It is assumed that the section of the line vortex that interacts with the airfoil remains rectilinear. t − x c0 · ∇Y j (y) d 3 y. 0). ˆ where 2a(y3 ) is the airfoil chord at the spanwise location y3 . x02 .1 at zero angle of attack to a mean ﬂow at speed U in the x1 direction. t) ≈ −ρ0 x j ∂ 4πc0 x2 ∂t (ω ∧ v) y. the Kirchhoff vector 2 2 Y has the components Y1 = y1 . Y2 = ˆ Re(−i z 2 − a(y3 )2 ). but we shall consider only the case where the span is compact. y3  < 1 L 2 y2 . z = y1 + i y2 . where x0 = (x01 . When h = 0 it will be necessary to take account of nonlinear interactions with the airfoil. y3  > 1 L 2 . 8. as indicated in the side view of Fig. with the representation ω = kδ(x1 − x01 (t))δ(x2 − x02 (t)). this can be dealt with in a ﬁrst approximation by ignoring the shed vorticity and deleting the trailing edge singularity of .1b. for the elliptic airfoil a(y3 ) assumes a maximum value of a at y3 = 0. predictions for a noncompact span will be intermediate between those discussed here and those in Section 8. 2 L 2 (8.2) Vorticity is shed into the wake of the airfoil in accordance with the Kutta condition of unsteady aerodynamics.2. For an airfoil of compact chord and span the acoustic pressure produced by the interaction is given by Equation (5.8. x → ∞.2. and we shall write ˆ a(y3 ) = a 1− 2 4y3 1 . y3  < L . The general solution is applicable to airfoils of arbitrary span. orientated as illustrated in Fig. Consider a planar airfoil of either rectangular or elliptic planform.2. As before. For an elliptic airfoil of span L (between − 1 L < x3 < 1 L).
It is assumed to be nonzero only over the section − 1 L < y3 < 1 L 2 2 of the vortex. (8.1 Linear Theory When there is no backreaction of the airfoil on the vortex the convection velocity of the vortex is equal to the mean stream velocity d x01 = U.3). ˆ (8. t) 1 ≈− 3 ρ0 U M cos (L/x)/4πa 22 1 2 Im U [t] a ˆ a a 3 2 −1 2 + ˆ a a h +ia d y3 . The second term is the conventional lift dipole radiation. z 0 (t) = x01 (t) + i x02 (t). by using the following modiﬁcation of the x2 component of Y: ˆ ˆ Y2 = Re( 2a(y3 ) z + a(y3 )).3) √ dt z 0 (t) + a(y3 ) ˆ −L 2 L 2 where . t) + p2 (x.f.1a between the x1 and x2 directions and the radiation direction.2. dt The radiation is produced entirely by the lift dipole. aligned with the airfoil chord. and have been discarded (c.2. Section 7.2. (8. are respectively the angles shown in Fig. and quantities in square braces are evaluated at the retarded time [t] = t − x/c0 . 8. where d x02 /dt = 0 because of nonlinear interactions with the airfoil. Then. Note that ‘inﬁnite’ contributions to the integrals from y3  > L/2 are constant because ω ∧ v is constant for y3  > L/2.2. 8. t) ≈ ρ0 cos ∂ 4πc0 x ∂t × ∂ ∂t L 2 y3  < 1 L. and if the vortex crosses the midchord of the airfoil at time t = 0 (8..2.4) . dt d x02 = 0.3) reduces to p2 (x. The ﬁrst term on the right is the suction force dipole.188 8 Further Worked Examples Green’s function. t) ≡ p1 (x. whose strength is determined by the x2 component of the vortex convection velocity.2. 2 −L 2 d x02 ρ0 cos dy3 + √ dt 4 2π c0 x √ ˆ dz 0 a(y3 ) Im dy3 x → ∞.1) becomes p(x.
2a.2. and is given by (8.4)) for the rectangular and elliptic airfoils are plotted in Fig.3 for an airfoil of inﬁnite span.1.2. 8. The proﬁles are qualitatively similar to the corresponding plot in Fig.2. 8.2) for the elliptic airfoil.8. ˆ ˆ where y 3 = y3 /L. image vorticity in the airfoil prevents the vortex from impinging on the leading edge.2. although in three dimensions the amplitude decreases much more rapidly with increasing retarded distance of the vortex from the leading edge. and 2 2 that the distorted path can be approximated by that for locally twodimensional ﬂow. a/a = 1 for the rectangular airfoil.2.2 Nonlinear Theory When the standoff distance h = 0.2.2 for a vortex standoff distance h = 0. . For a rectangular airfoil the integral evaluates to Im U [t] h +1+i a a −3 2 . If the leading edge of the airfoil is suitably rounded (so that no additional vortex shedding occurs) this case can be treated for a rectangular airfoil by assuming that only the section of the vortex within the span − 1 L < x3 < 1 L of the airfoil is affected in this way. The peak amplitude is larger and the acoustic pulse is of smaller duration ∼h/U for the rectangular airfoil because different sections of the vortex interact with the leading edge of the elliptic airfoil at different times during a total interaction time ∼ a/U > h/U . 8. and causes the trajectory to be locally deﬂected above the airfoil (for > 0).2 Parallel Blade–Vortex Interactions in Three Dimensions 189 Fig. The acoustic pressure signatures (the lefthand side of (8. 8.
1.2 when the vortex is released upstream with h = 0. The upper part of the Fig. dT 1 W(Z ) = √ √ .190 Let = .2. a Z= z0 . t) ≈ U M cos (L/x)/4πa dT ρ0 ρ0 ∂ p2 (x. 8 Further Worked Examples 4πaU T = Ut . 8. where [ ] denotes evaluation at the retarded time t − x/c0 .3 for a velocity ratio = 0. a dZ = u(T ) + iv(T ). . 2 Z +1 Then. These nondimensional pressures are plotted in Fig.2.3.1. and the suction and lift dipole radiation pressures are given by dv p1 (x. t) ≈ [Im (W(Z )(u + iv))] U M cos (L/x)/4πa ∂T x → ∞.10) (where ζ0 is deﬁned in terms of z 0 2 as in (8. 8.1)). the motion of the section of the vortex within the airfoil span (− 1 L < 2 x3 < 1 L) is governed by Equation (8.
0). 8.3.3 Vortex Passing over a Spoiler 191 ﬁgure shows the path followed by those sections of the vortex inboard of the airfoil tips.1. The wall coincides with the plane x2 = 0.3 Vortex Passing over a Spoiler The sound produced when vorticity interacts at low Mach number with surface irregularities on a nominally plane.3. . it is the same as that depicted in Fig.f. 8. is parallel to the spoiler. 8. Fig. z 0 = x01 + i x02 . and is assumed to convect over it in a low Mach number. The mean ﬂow is parallel to the real axis in the ζ plane. The transformation ζ = z2 +1 a2 (8.1. Fig. 8. where x0 = (x01 . A simple canonical interaction (Kasoev.4).1.8. rigid wall is produced by dipoles orientated in the plane of the wall. As in that case. Deﬁne z = x1 + i x2 . 1976) involving a line vortex near a thin vertical spoiler is illustrated in Fig. and the spoiler extends along the x2 axis from x2 = 0 to x2 = a > 0 for −∞ < x3 < ∞.3. 8. small acoustic blips spuriously predicted during the passage of the vortex past the trailing edge have been removed (c.. and the vortex maps into an equal vortex at ζ = ζ0 . by the unsteady wall drag.5 for the inﬁnite span airfoil. irrotational mean stream having uniform speed U in the x1 direction.1. The threedimensional acoustic pulses are narrower than those predicted in two dimensions. x02 . The vortex ω = kδ(x − x0 (t)). that is.1) maps the ﬂuid region onto the upper half Im ζ > 0 of the ζ plane. The left and right faces of the spoiler (x1 = ∓ 0) transform respectively into the intervals −1 < ζ < 0 and 0 < ζ < 1 of the real ζ axis.
V The compact Green’s function for this problem (applicable when the acoustic wavelength ∂) is given by (6.3)) ∗ d x01 d x02 i ζ (z 0 ) dz 0 ≡ −i = − + F (z 0 ). = U . a V = 4πa . (8. d Z∗ = −i dT where Z= z0 . (8.3.3.1). 2π i ln(ζ (z) − ζ (z 0 )∗ ) + U aζ (z).192 8 Further Worked Examples with complex potential U aζ .1. √ t − τ − x/c0 (8. dt dt dt 4π ζ (z 0 ) that is. and is given by w(ζ ) = − i i ∗ ln(ζ − ζ0 ) + ln(ζ − ζ0 ) + U aζ.6).4) .4) for the farﬁeld acoustic pressure becomes p(x.6. setting ζ = ζ (z) the motion in the z plane is deﬁned by w(z) = − where F(z) = i ln(ζ (z) − ζ (z 0 )) + F(z).3. The complex potential of the whole ﬂow in the ζ plane is obtained by introducing an image vortex of strength − at the ∗ complex conjugate point ζ = ζ0 .2. t) ≈ = −ρ0 x1 ∂ √ 3 π 2c0 x 2 ∂t −ρ0 x1 ∂ √ 3 π 2c0 x 2 ∂t t−x/c0 −∞ t−x/c0 −∞ k∧ dτ dx0 (τ ) · ∇Y1 (x0 (τ )) √ dτ t − τ − x/c0 dτ . a 1 2Z Z − 2 .2) +√ Z (Z 2 + 1) Z + 1 − Z 2 + 1 Z2 + 1 T = Vt . so that the analogue of Equation (6. 2π which is in the form (4.6. The equation of motion of the vortex at z 0 is therefore (see (4.3) d x01 ∂Y1 d x02 ∂Y1 − dτ ∂ y2 dτ ∂ y1 x0 (τ ) where the Kirchhoff vector Y1 = Re(aζ ) = aRe( Z 2 + 1 ) at z = z0. 2π 2π Hence.
the corresponding nondimensional acoustic pressures (8. and this is responsible for more than doubling the amplitude and the effective frequency of the sound. in the linearized approximation in which the vortex is assumed to translate at the local mean stream velocity. ρ0 V 2 M sin (a/x) √ p(x.6) where [T ] = V [t]/a is the nondimensional retarded time and M = V /c0 .2 shows the vortex trajectories when the initial distance of the vortex from the wall is h = 0. introduce the notations dZ = u(T ) + iv(T ).95V when U = V . because in that case dx0 = U ∇Y1 (x0 ) dt and k ∧ ∇Y1 · ∇Y1 ≡ 0. taking the initial position of the vortex to be several spoiler heights a upstream. dt This force vanishes. dT W= d Z ( z2 + a2 ) = √ dz Z2 + 1 (8.75a for the two cases (i) of no mean ﬂow. U = 0. and therefore no sound is generated.6) are evaluated numerically.8.98V when U = 0 to 3. t) 1 2 ρ0 V M sin √ aπ 2 √ a x 1 2 ∂ ∂T [T ] −∞ Im W(Z ) dZ ˆ dT ˆ dT ˆ [T ] − T .5) ˆ evaluated at the vortex.3. 8.6) are plotted in the lower part of the ﬁgure.3 Vortex Passing over a Spoiler 193 The radiation is produced by the unsteady drag force F1 exerted on the ﬂuid by the spoiler. . and (ii) U = V ≡ /4πa.3. where its motion is unaffected by the spoiler. The effect of mean ﬂow is to draw the trajectory marginally closer to the spoiler as it passes the tip of the spoiler where the interaction is strongest. ≈ 22 5 ∂ ∂T ∞ 0 Im(W(Z )(u + iv))([T ] − λ2 ) dλ.3.3) to obtain the acoustic pressure in the form p(x.3. Following the procedure of Section 8. The upper part of Fig. and make the substitution T = V τ/a in (8.3. The convection velocity at this point is also increased from about 1.2) and the acoustic pressure integral (8. where time is measured from the instant that the vortex passes the spoiler. (8. The vortex path equation (8.3.3. given (per unit span) by F1 = −ρ0 ω ∧ v · ∇Y1 dy1 dy2 = −ρ0 k ∧ dx0 · ∇Y1 (x0 ). t) ≈ that is.1.
8.3..1): w(z) = − a2 a2 i i i ln(z −z 0 )+ ln z − ∗ − ln z + U z + 2π 2π z0 2π z .f. twodimensional interactions of a line vortex with a circular cylinder provide an interesting contrast to the sharpedge problems discussed above. Let the cylinder have radius a and be coaxial with the x3 axis. and let there be an irrotational mean ﬂow at speed U past the cylinder in the x1 direction.1) The velocity potential governing the motion of the vortex at z 0 is obtained by deleting the selfpotential − i ln(z − z 0 ). 2π . and by adding the potential for the uniform mean ﬂow past the cylinder (c. Section 8.194 8 Further Worked Examples Fig. Set z = x1 + i x2 and let the vortex of strength have the complex position z 0 = x01 + i x02 at time t.2. a vortex + at the centre to make the circulation vanish. with no mean circulation about the cylinder. 8. (8.4.4 Bluff Body Interactions: The Circular Cylinder Low Mach number. The complex potential w(z) is found by placing an ∗ image vortex − at the inverse point z = a/z 0 within the cylinder.
dt dT 8. t) ≈ t−x/c0 −ρ0 x j ∂ √ 3 2π 2c0 x 2 ∂t −∞ dτ ×√ . c0 [T ] = [T ] −∞ ˆ Im(W j (u + iv))(T ) x . Equation (8.4) ˆ evaluated at z 0 . say. c0 ˆ dT ˆ [T ] − T .5) Y1 = Re z + By deﬁning W1 = d dz z+ a2 z ≡1− d 1 a2 . (8. F2 ) exerted on the ﬂuid (per unit span) by the cylinder. V = . Z2 = U .4. V 195 (8. Y2 = Re −i z − a2 z .4. The integrals must be evaluated numerically using the numerical . T = . and making the change of integration variable T = V τ/a.3) can be written √ ρ0 V M x j ∂ p(x.2) z0 Vt . V a t− The subscripts j = 1.4 Bluff Body Interactions: The Circular Cylinder Hence we arrive at the equation of motion i d Z∗ = + dT Z (Z 2 − 1) where and Z = 1− 1 .4. 2 in this formula respectively correspond to the acoustic pressures p1 .3) where the components of the Kirchhoff vector can be written (see Section 4. z = y1 + i y2 .4.1 The Acoustic Pressure The farﬁeld sound produced by the vortex is calculated from (6. t) ≈ √ 3 2π 2ax 2 ∂ T where M= V . t − τ − x/c0 d x01 ∂Y j d x02 ∂Y j − dτ ∂ y2 dτ ∂ y1 x0 (τ ) (8.4.4): p(x. W2 = −i z − Z2 dz z ≡ −i 1 + 1 Z2 a2 z . p2 . a 2πa a dZ dz 0 =V ≡ V (u + iv).8. produced by drag and lift dipoles.2. whose strengths are determined by the force (F1 .
The upper limits of integration are then ﬁnite because the source terms vanish as soon as [T ] − λ2 < T .4. The amplitude of the sound decreases rapidly with increasing distance of closest approach of the vortex to the cylinder.1 illustrates the typical nondimensional waveforms produced when V ≡ /2πa = 2U and for h/a = ±0.4. Im(W2 (u + iv))([T ] − λ2 ) dλ. say.2) for the vortex path. 8.1.196 8 Further Worked Examples Fig.4. t) 1 ρ0 V 2 M sin (a/x) 2 ρ0 V 2 M cos (a/x) √ p2 (x. This is done by making the ˆ further change of integration variable λ = [T ] − T . by taking the initial position of the vortex a is sufﬁciently to be far upstream of the cylinder at z 0 = −L + i h. 8.4. The calculation begins at time T . t) 1 2 ≈ 22 ≈ 22 1 1 ∂ ∂T ∂ ∂T ∞ 0 ∞ 0 Im(W1 (u + iv))([T ] − λ2 ) dλ.1). Figure 8. solution of Equation (8. time being measured from the instant that the vortex crosses x1 = 0.7. in which case √ p1 (x. where L large that the source strengths are negligible for T < T (Fig. near the cylinder the translational velocity of the vortex is increased because the .
the lift dipole vanishes identically.4.3.2 WallMounted Cylinder The case of ideal motion of a vortex translating past a cylindrical.8. 8. The velocity potential of the unsteady motion is given by augmenting the complex potential (8.4.4 Bluff Body Interactions: The Circular Cylinder 197 mean ﬂow velocity is larger. the lift dipole will tend to predominate because convection by the image vortices can then be neglected in a ﬁrst approximation.4. In cases where U V . semicircular projection on a rigid wall (Fig. 8. The problem is equivalent to that in which a vortex pair. is incident symmetrically on a circular cylinder. 2π Fig. consisting of a vortex at z 0 accompanied by an image of strength − at ∗ z 0 . and the drag ∼ (ω ∧ U ∇Y1 ) · ∇Y1 ≡ 0.4. . 8.2.1) by the terms a2 i i ∗ ln(z − z 0 ) − ln z − 2π 2π z0 + i ln z. In this case.2) can be treated by the method used for the spoiler in Section 8. and also because of the increased inﬂuence of the image vorticity.
4. when the vortex is assumed to convect passively at the local velocity of the undisturbed mean stream. 8. t) ≈ = −ρ0 x1 ∂ √ 3 π 2c0 x 2 ∂t −ρ0 x1 ∂ √ 3 π 2c0 x 2 ∂t t−x/c0 −∞ t−x/c0 −∞ k∧ dx0 dτ (τ ) · ∇Y1 (x0 (τ )) √ dτ t − τ − x/c0 √ x0 (τ ) d x01 ∂Y1 d x02 ∂Y1 − dτ ∂ y2 dτ ∂ y1 dτ . and M = V /c0 . [T ] = V [t]/a is the nondimensional retarded time (T = 0 when the vortex is at x1 = 0). 8.4. (8. Z2 (8. set dZ = u(T ) + iv(T ). Then.9) where the angle is deﬁned as in Fig.5) and the acoustic pressure integral (8.6).4.4.4. t) 1 2 W1 = d dz z+ a2 z =1− 1 Z2 (8. ρ0 V 2 M sin (a/x) √ p(x. Then.4.2 shows the vortex trajectories when the .9) must be evaluated numerically. As before.7) The radiation is produced by the unsteady drag force exerted on the ﬂuid by the cylinder.5) z0 .198 8 Further Worked Examples which correspond to the net potential produced by the image. a The compact Green’s function is given by (6. which vanishes in the linearized approximation. dT evaluated at the vortex. a = U .8) ≈ 22 3 ∂ ∂T ∞ 0 Im(W1 (u +iv))([T ]−λ2 ) dλ.1.1.4. V 1− 1 . The vortex path equation (8. and the farﬁeld acoustic pressure by p(x. taking the initial position of the vortex to be several cylinder radii a upstream where its motion is unaffected by the cylinder.6) where the Kirchhoff vector Y1 = Re z + a2 z . (8.4. t − τ − x/c0 (8.4. the equation of motion becomes d Z∗ =i dT where Z = 1 Z − Z∗ + + Z − Z∗ (Z 2 − 1)(Z 2 − 1) V = 2πa . T = Vt . The upper part of Fig.
The waveforms and these general conclusions are qualitatively similar to those discussed in Section 8. σ ) = ln 8r0 σ − 1 . selfinduction and image vorticity in the sphere (Fig.3 for the sharpedged spoiler. 1967) u (r0 . 8.1).23V when U = 0 to 3. 8. The selfinduced velocity of the ring (in inviscid ﬂow) is parallel to the x1 axis at speed u given approximately by Kelvin’s formula (Batchelor.1. .8. threedimensional vortex– surface interaction amenable to analysis is the axisymmetric motion of a ring vortex over a sphere. inviscid. The vortex convection velocity at y1 = 0 is increased from about 1. A vortex ring of radius r0 (t) and circulation is coaxial with the x1 axis and translates in the positive x1 direction under the inﬂuence of the mean ﬂow.5 Vortex Ring and Sphere 199 initial standoff distance of the vortex from the wall h = 0.5. and (ii) U = V . Let a sphere of radius a be placed with its centre at the origin in the presence of a uniform mean ﬂow at speed U in the x1 direction.5. 4 (8.07V when U = V .5 Vortex Ring and Sphere Perhaps the simplest low Mach number. 8.1) 4πr0 Fig. U = 0.5. this is responsible for the increased acoustic amplitude and for more than doubling the effective frequency.5a for the two cases (i) of no mean ﬂow. We shall assume the vortex core is circular (and remains circular throughout the interaction) with radius σ (t) r0 .
5. and axial location x01 are given by =− 2 2 r0 + x01 a 1 2 . x01 ). x01 ).200 8 Further Worked Examples . 2 2 r0 + x01 (8.5. E( ) are respectively complete elliptic integrals of the ﬁrst and second kinds. x1 ) (Batchelor.5. The image vorticity consists of a coaxial ring vortex whose circulation radius r0 . where r denotes perpendicular distance from the x1 axis. r0 = a 2 r0 . Ting and Klein. 1967. because the vortex lines move with the ﬂuid particles. i. 1991) ψ(r. x01 (t).5) . then at any time t 2 (2πr0 )π σ 2 = (2π h)π σ0 .5. and K ( ). (8.e. If r0 = h and σ = σ0 are the initial values when the vortex ring is far from the sphere.5. x1 ) = a Ur 1− 2 2 r 2 + x1 2 3 3 2 + 2π ( + + − ){K ( ) − E( )}.1) becomes u = 8h r0 (t) ln σ0 h 3 2 4πr0 − 1 . σ (t) = σ0 h r0 (t) so that the selfinduced velocity (8. 4 (8. σ ) + (r0 . The radius r0 (t) and axial position x01 (t) of the ring vortex are then determined by the equations of motion 1 ∂ψ dr0 =− (r0 . 2 K( ) = 0 dµ 1− 2 sin2 µ .2) where the planes of symmetry of the ring vortex and its image cut the x1 axis respectively at x01 (t).4) The core radius σ decreases when r0 increases.. 2 2 r0 + x01 x01 = a 2 x01 .3) where ± = (r ∓ r0 )2 + (x1 − x01 )2 . dt r0 ∂ x 1 d x01 1 ∂ψ = u (r0 . The motion of the vortex produced by the combined induction by the image and the mean ﬂow can be expressed in terms · of the Stokes stream function ψ( r . dt r0 ∂r (8. π 2 = E( ) = 0 − −+ − π 2 + + . 1− sin2 µ dµ.
5.05h.2. h = 0.8a. a R= r0 . 4 (8. . X ) + ln dT R ∂R 2R σ0 1− 1 (R 2 + X 2 ) 2 3 x01 . V aR h 1 2 3 2 − 1 . R2 + X 2 = ˆ− − ˆ+ ˆ− + ˆ+ . In both cases the integration is started ﬁve sphere diameters upstream with the following initial values for the vortex ring radius and core radius.7) The solid and brokenline curves in the ﬁgure correspond respectively to = 0 Fig. dR 1∂ =− (R. 8.6) R2 = 2 − (R 2 + X 2 ) × ( ˆ + + ˆ − ){K ( ) − E( )}. dT R ∂X where dX 1∂ 1 8h = (R. Figure 8. σ0 = 0.5.5. R2 + X 2 X = X .5.2 illustrates the sections in the vertical plane of symmetry of the sphere of two typical vortex trajectories predicted by Equations (8.5 Vortex Ring and Sphere 201 The equations of motion of the vortex are cast in nondimensional terms by deﬁning X= Then. T = Vt . X ). a V = 2πa .8.6). (8. R . ˆ± = R = (R ∓ R )2 + (X − X )2 . a = U .5.
The force on the sphere is in the mean ﬂow direction – the effective acoustic source is the unsteady drag – and only the component Y1 = y1 a3 1+ 2y3 ≡ y1 1 + a3 2 2 r 2 + y1 3 2 (from Table 3. x → ∞. 8.5. we ﬁnd p(x.5.x01 ) . the amplitude and characteristic frequency of the sound both increase with U . (8.1) of the Kirchhoff vector makes a nontrivial contribution to (8. (8.5. equation (5.5.4. t − x c0 · ∇Y1 (y) ≡ 0.8) and evaluating the integral.5. t − x c0 · ∇Y j (y) d 3 y. However.1. although the convection speed of the ring past the sphere is greatly increased. Substituting (8.8).1 Acoustic Pressure When the sphere is acoustically compact. the mean ﬂow velocity U = U ∇Y1 is absent because (ω ∧ U ∇Y1 ) y. t) ≈ −ρ0 x j ∂ 4π c0 x2 ∂t (ω ∧ v) y.4) gives p(x. t) ≈ −ρ0 cos 2c0 x ∂ r0 ∂t d x01 ∂Y1 dr0 ∂Y1 − dt ∂r dt ∂ y1 . locally tangential to the vorticity ω and orientated in the clockwise direction when the vortex ring is viewed from upstream.9) into (8. The latter value is chosen to make the mean stream velocity U approximately the same as the selfinduced velocity u at large distances from the sphere. This has a relatively small effect on the trajectory. (r0 .8) For the purpose of evaluating the integral we may neglect the ﬁnite core size of the vortex.9.5. and set ˆ ω = θδ(r − r0 (t))δ(x1 − x01 (t)).9) ˆ where θ is a unit azimuthal vector. 8. The production of sound is therefore a nonlinear event – the source explicitly involves only the selfinduced velocity and the velocity induced by the image vortex.202 8 Further Worked Examples (no mean ﬂow) and = 3. x → ∞. as indicated in Fig.5.
The nondimensional acoustic pressure signatures plotted in Fig. Expressing this result in nondimensional form.5.5 Vortex Ring and Sphere 203 Fig.5.5. The thick solid curve is the pressure proﬁle in the .2.5. 8. dT where M = V /c0 . and is the angle between the radiation direction and the x1 axis illustrated in the upper part of Fig. 8.3. t) ≈π ρ0 V 2 M cos (a/x) ∂T 3R 2 X 2(R 2 + X 2) 2 2(R 2 5 dX dT + X 2) 2 5 + 1+ R 2 − 2X 2 R dR . 8.3.8. where the quantity in the square braces is evaluated at the retarded position of the vortex ring.7) considered above for the vortex ring trajectories in Fig. and R and X are the solutions of the vortex equations of motion (8. we have ∂ p(x. 8.5.5.6).3 are for the same the initial conditions (8.
5. To integrate the equation. The positions of the vortex ring in this case at several different retarded times V [t]/a are marked on the thick curve in the upper part of Fig. Both the amplitude and frequency of the sound are increased because of the increased convection velocity of the vortex past the sphere. 1975) 3i Z 2i d Z∗ = 2 + . when the selfinduction velocity u ≈ U at large distances from sphere. 8. The motion is evidently symmetric with respect to the x1 axis.6. A vortex pair aligned with the x3 axis.204 8 Further Worked Examples absence of mean ﬂow (U = 0). The rigid wall coincides with the plane x1 = 0. 8. T = a a 4πa dZ dz 0 =V ≡ V (u + iv). The upper part of Fig.1 depicts a simple model of such a source. and the transformation ζ =√ z z2 + a2 (z = x1 + i x2 ) maps the region Im z > 0 cut along the upper section x2 > a of the wall onto the upper half of the ζ plane. we accordingly obtain the equation of motion of the vortex pair in the form (Karweit. and is pierced by a twodimensional slit aperture of width 2a whose centerline extends along the x3 axis. V = . By the usual method. and dt dT (8. consisting of vortices of strengths ± at the respective complex positions z 0 = x01 + i x02 and ∗ z 0 = x01 − i x02 is incident on the aperture from the left (x1 = −∞). where Z = . Similarly. 8.6. we can set z 0 = − L + i h at a convenient initial (but arbitrary) time .3 (time being measured from the instant that the ring crosses the centre of the sphere). which is acoustically equivalent to a monopole source when the aperture is compact. the thinline curves in the ﬁgure give the pressure signature and retarded positions for U = 3V . √ √ 3 dT Z + 1 (Z 2 + 1) 2 {Z / Z 2 + 1 − (Z / Z 2 + 1)∗ } Vt z0 .6 Vortex Pair Incident on a Wall Aperture Hydrodynamic motion in the vicinity of an aperture in a large thin wall generally produces an unsteady volume ﬂux through the aperture.1) Let the initial separation of the vortices at x1 = − ∞ be 2h.
When h/a is smaller than 2/3 2 ≈ 0.385 the vortex pair passes through the aperture in the manner indicated in Fig. t) ≈ −ρ0 ω∧v· ∂G 2 d y dτ.8.3) . The acoustic pressure in the far ﬁeld is given by p(x. ˜ z = y2 + i y1 . t − τ )≈ − √ √ a t − τ − x/c0 π 2πx ˜ z2 −1 a2 .6. ∂y (8. 8. y.6. T = T .9.6. For larger values of h/a the trajectories of the two vortices separate. where L a. as illustrated for h/a = 0.35.6.10) for the wall aperture √ ˜ c0 sgn(x1 ) χ (t − τ − x/c0 ) z Re ln + G(x.1 for h/a = 0. the vortices travel along symmetric paths parallel to the wall on either side of the aperture. 8. (8.1.6 Vortex Pair Incident on a Wall Aperture 205 Fig.2) 3 where G is the compact Green’s function (3.6.
206 and χ (t) = H (t) 0 ∞ 8 Further Worked Examples ln( aξ 2 /4c0 t)e−ξ dξ . T = a we ﬁnd 2 2 ρ0 V 2 sgn(x1 ) p(x. by setting λ = p(x.3) is contained entirely in the logarithmic term.4) where χ(λ ) = ˆ 2 0 ∞ ln( Mξ 2 /4λ2 )e−ξ dξ . [ln( Mξ 2 /4λ2 )]2 + π 2 2 . If we deﬁne ˜ W(Z ) = and put Vτ ˆ in the integral (8. The dependence on source position y in (8. t) 7 ˆ [T ] − T we can write ∞ 0 22 1 ≈ √ πM ρ0 V 2 sgn(x1 )(a/x) 2 ˜ Re(W ∗ (Z )(u + iv))([T ] − λ2 )χ (λ2 ) dλ. ˆ x → ∞.2). ˆ ˜ Z =i Z ∗ M= V . When vortex shedding from the aperture edges is ignored. ω = kδ(y1 − x01 )δ(y2 − x02 ) − kδ(y1 − x01 )δ(y2 + x02 ). a Therefore.6.6. t) ≈ √ πM × 5 1 ˜2 − 1 Z . 8. (8.6.1). [ln( aξ 2 /4c0 t)]2 + π 2 2 = 1. c0 a x 1 2 [T ] −∞ ˜ ˆ Re(W ∗ (Z )(u + iv))(T ) ˆ ˆ ˆ χ([T ] − T ) d T . χ (T ) = χ (t). which represents the velocity potential of the ideal ﬂow that would be produced through the aperture (from left to right) by a uniform pressure drop across the wall. where k is a unit vector in the x3 direction (out of the plane of the paper in Fig.781072.6. ˆ [T ] − T [T ] = V [t] .
8.6.6 Vortex Pair Incident on a Wall Aperture 207 √ As before. Further increases of h/a beyond 0.03 for the farﬁeld acoustic pressure signatures plotted in Fig.35. when the vortex trajectories pass very close to the aperture edges. The radiation therefore has the characteristics of an acoustic monopole source for x1 > 0 and a sink for x1 < 0. Numerical results are illustrated in the ﬁgure for h/a = 0. 0. which occurs when the vortices pass close to the edges of the aperture. The maximum acoustic pressure amplitude is found to occur when h/a just exceeds the critical value (∼0. .e.1. We have taken M = 0.35. forcing ﬂuid through the aperture into the region x1 > 0. c0 4πac0 This is just the selfinduced convection Mach number of the vortex pair when separated by a distance 2a. When h/a = 0.4) is actually λ = [T ] − T . directed toward the wall.8.. the time origin has been adjusted to correspond approximately with the peak in the radiated acoustic pressure. in air this would imply that V ∼10 m/sec. The ﬂow induced by the vortex pair approaching the wall forms a localized twodimensional jet between the vortices. The resistance of the wall to this ﬂow causes the pressure just to the left of the wall aperture to rise. a decrease in the characteristic frequency of the sound). and a corresponding increase in the width of the acoustic pulse (i.6. The value of the integral depends weakly on the characteristic Mach number M= V ≡ . this produces a relatively larger pressure rise than for h/a = 0.6. In each case.6 result in a gradual reduction in the amplitude of the sound.6 the vortices do not penetrate the aperture but are deﬂected by the wall. where T is the nondimensional initial time from which the motion of the vortex pair is calculated.385). where the vortices pass through the aperture. the upper limit of integration in (8.
.
E... Crighton.. 1981. J. Philosophical Transactions of the Royal Society of London A255: 469–503.S. S. Lectures on Fluid Mechanics. New York: Interscience.P. Ffowcs Williams. Journal of Fluid Mechanics 66: 791–816. Journal of Fluid Mechanics 51: 357–362. Ffowcs Williams.E. A. Interaction between a vortex ﬁlament and an approaching rigid sphere. Goldstein. Radiation from line vortex ﬁlaments exhausting from a twodimensional semiinﬁnite duct. Sound and Sources of Sound. Progress in Aerospace Sciences 16: 31–96.G. F. Journal of Fluid Mechanics 58: 65–80. J. Journal of Fluid Mechanics 110: 129–147.Bibliography Batchelor. N. 1960. J.H. 1975a. Ffowcs Williams. M. Curle. and Ffowcs Williams. Crighton.P. Radiation from vortex ﬁlament motion near a half plane. L.. and Hawkings. D. 1973. J.G. D. Annual Reviews of Fluid Mechanics 17: 411–445. and Ffowcs Williams. 1972.G. London: SpringerVerlag.E. G. Dowling. Howe. Sound generation by turbulence and surfaces in arbitrary motion. Crighton.E. Chichester: Ellis Horwood.E. An Introduction to Fluid Dynamics. Dhanak. Ffowcs Williams. 1992. Ffowcs Williams. Journal of Fluid Mechanics 40: 657–670. M. 1974. A.G. New York: McGrawHill. J. Modern Methods in Analytical Acoustics (Lecture Notes). 1985. Heckl. Cannell. M. P.R. Philosophical Transactions of the Royal Society of London A264: 321–342. Cambridge: University Press. Goldstein. 1963. Contributions to the theory of aerodynamic sound with 209 . The Kutta condition in unsteady ﬂow. D. J. 1970. Basic principles of aerodynamic noise generation. and Leppington. The noise from turbulence convected at high speed. M. 1976. and Hall. J.G. 1975.E. 1955. 1967. D.E. The inﬂuence of solid boundaries upon aerodynamic sound.L. 1983. Sound production at the edge of a steady ﬂow. Aerodynamic sound generation by turbulent ﬂow in the vicinity of a scattering halfplane. Aeroacoustics. D. Dowling. Crighton.E. Proceedings of the Royal Society of London A231: 505–514. 1969.K.
210 Bibliography application to excess jet noise and the theory of the ﬂute. Modelling low Mach number noise. Rosenhead. 1998a. J.. Ting. Aerodynamic noise and the plane boundary. O. Journal of the Acoustical Society of America 103: 83–98. Journal of Fluid Mechanics 425: 111–132.J. Landau. T. 1976. On sound generated aerodynamically. Mechanics of Sound o Generation in Flows.S. Journal of Engineering Mathematics 41: 367–400. Howe. Japan 22–26 November) pp. 1993. Nakatani. M. M. MilneThomson. Proceedings of the International Conference on Speedup Technology for Railway and Maglev Vehicles. 1932. 1958. Iida. An Informal Introduction to Theoretical Fluid Mechanics. 1952. 1975b. The velocity of a vortex ring with a thin core of elliptical crosssection. Journal of Fluid Mechanics 67: 579–610. M. 5th ed.G. 85–96. Laminar Boundary Layers. 2. 1987. pp. On vortex motion. Japan Society of Mechanical Engineers (Yokohama. 1867.. W. Horace. Part I: General theory. M¨ hring.S. Oxford: University Press.J.M. 1989. An Introduction to Fourier Analysis and Generalised Functions. 315–319. .M. Proceedings of the Royal Society of London A370: 407–415. Howe. R..W. Proceedings of the Royal Society of London A211: 564–587. 1975. and Maeda. Lighthill. Lamb. James. 2001. Proceedings of the Cambridge Philosophical Society 52: 317–321. Cambridge: Cambridge University Press. 1978. Journal of Fluid Mechanics o 85: 685–691. M¨ ller. Lord. On vortex sound at low Mach number. M. Cambridge: Cambridge University Press. Physics of Fluids 18: 1604–1606. 1980. u Powell. T. Cambridge: Cambridge University Press. Chs. Kasoev. K. Berlin: SpringerVerlag. S. 1998. The image system of a vortex element in a rigid sphere.D. M. M.J. M. Proceedings of the Royal Society A454: 1523–1534. London: Macmillan. Howe.S. Journal of the Acoustical Society of America 32: 962–990. Journal of Fluid Mechanics 206: 131–153. Lighthill. Howe. Acoustics of Fluid–Structure Interactions. 1956. Matsumura. Waves in Fluids. Hydrodynamics. Kelvin.M. On unsteady surface forces. Theoretical and experimental investigation of the compression wave generated by a train entering a tunnel with a ﬂared portal.S. L. Effect of shape of train nose on compression wave generated by train entering tunnel. Journal of Fluid Mechanics 71: 625–673. 1998b. Edited by E.. D. L. Soviet Physics Acoustics 22: 71–72. Edited by L. Moore. 6th ed. K. Oxford: Clarendon Press. L.A. Iida.. 1980.. Cambridge: Cambridge University Press. 2nd ed. Theoretical Hydrodynamics. and Uchida. Interaction of a slender vortex with a rigid sphere: Dynamics and far ﬁeld sound. T. Lighthill. Lighthill. Howe. M. Transactions of the Royal Society of Edinburgh 25: 217–260.J.S. 1. M. and Klein. and sound produced by the normal chopping of a rectilinear vortex. 1968. and Lifshitz. Fukuda. Lighthill. M.S. Knio. Sound radiation from a linear vortex over a plane with a projecting edge. 1960. Vorticity and the theory of aerodynamic sound. 2000. W. E. 1978. The generation of sound by aerodynamic sources in an inhomogeneous steady ﬂow. 1986. Karweit... A. M. 1963. M. T. Maeda.. Howe. Motion of a vortex pair approaching an opening in a boundary. Fluid Mechanics. The compression wave produced by a highspeed train entering a tunnel. M¨ hring. M. Oxford: Pergamon. Lighthill.
Lord. L. Rayleigh. Ting. A. R. New York: Dover. 2 vols. 374. P.Bibliography 211 Powell. AGARD Report No. Powell. . 1945. 1993. W. Theory of vortex sound. Saffman. 466. 1963. New York: SpringerVerlag.G.R. Cambridge: University Press. Mechanisms of Aerodynamic Sound Production. 1991. Journal of the Aeronautical Sciences 8: 104–108. Some aspects of nonstationary airfoil theory and its practical applications. Journal of the Acoustical Society of America 36: 177–195. Viscous vortical ﬂows. 1941. Lecture Notes in Physics. and Klein. Theory of Sound. Vortex Dynamics. Vol. A. Sears. 1964.
.
generated by train. elliptic. 41 bubble. 103 complex velocity. preliminary deﬁnition. 65 duct (tunnel) entrance. 103 vortex. 38 compact Green’s function. theory. 154 uniform duct. 186 blockage. 149 Crocco’s equation. 77. 58 cylindrical bodies. 170 duct with neck. 99 formula. 18 body. 158. 32 differential equation. blowing out. 52 airfoil of variable chord. 105 source. 74 rigid strip. 189 threedimensions. 102 halfplane. 32 creeping ﬂow. 36. 43 added mass tensor. 91 causality. 25 airfoil theory. 85 Curle. 136. 91 boundary. 77. 194 bound vorticity. 175 linear theory. 188 nonlinear. 63 twodimensional. 60 sphere. 114 with bound vorticity. 35 integral equation. 64 deﬁnition for wave equation. 36 213 . 100 circular cylinder. 87 per unit length of wake. deﬁnition. 58 deﬁnition for Helmholtz equation. 167 bluff body. 2 linearized.Index acoustic wave number. 187. linearized. 148 compact. acoustically. deﬁnition. application to creeping ﬂow. 105 strip 61. 128 turbulence eddies. 187 rectangular. 166 compressive stress. solid. 30. 99 Crighton. 178. 80 for incompressible ﬂow. 5 with source. 94 relation to bound vorticity. in sound ﬁeld. 148 airfoil. 87 Kelvin’s theorem. 70 halfplane. 69. 5 control surface. 75 wall cavity. 94 sphere. 76 wall aperture. 72 complex potential. 190 analogy. 11 circulation. 49. 88 used to calculate sound source. 68 contribution to surface force. 143. twodimensions 150. deﬁnition. 70 aerodynamic sound. 53 symmetric. 91 bladevortex interactions. 25 BiotSavart. 1 acoustic analogy. 71 circular cylinder. continuity equation. 112 candle. 72 wall projection. 100 halfplane. acoustic. 35 conformal transformation. 81 general form. 180. 101 compression wave.
94 Sears’ formula. acoustic. 185 lift. 89 Ffowcs Williams. 78 energy. circulation theorem. 185 produced by gust. 4 delta function Fourier integral for. 57 homentropic. 44. 4 Index hydrodynamic. 96 infrasound. 191 Kelvin. 2 values for air and water. 156 Heaviside unit function. 4 Lighthill’s equation. 87 deﬁnition of vorticity. acoustics. 89 wavelength. 179 special cases. 84 kinetic energy. 45 freespace. 85. impulse formula. 21 Green’s function. 166. time harmonic. 93 added mass contribution. 93. 82 in terms of vorticity. 148 ideal acoustic medium. 32 intensity. 131 incompressible ﬂuid. deﬁnition. 12 for Helmholtz equation. 7. equation. 44 Green’s function. 54 KraichnanPhillips theorem. 17 acoustic. 37 end correction. 146. 32 of quadrupole radiation. transformation using Heaviside function. 182. 60 rigid strip. 16 drag. 28 equation reformulated in terms of vorticity. 9 density. 82. 45 lift. 4 far ﬁeld. 35 in Helmholtz equation. 92. 60 two dimensions. 90 Kelvin’s theorem on. 182. deﬁnition. 94 formula for translating body. 147 role of bound vorticity. equation. 84 Kirchhoff spinning vortex. 195 Lighthill. 195 efﬁciency. 96 lift. 1 acoustic analogy. 188. 27 momentum equation. 19 irrotational ﬂow. 195 far ﬁeld form. 117 stress tensor. 188. 4 ﬂux. 46 gust. 62 drag. see total enthalpy entropy. 3 diffusion. deﬁnition. 29 Maeda. 149 Fraunhofer approximation. 82 formula for vortex ring speed. 39 Kutta condition. 171 . 53 polar from. 204 Kasoev. 19 calculation of. 191. 54 near edge. 27 impedance. 90 enthalpy. 25 eighth power law. 48 impulse. 132 Kirchhoff vector. 52 circular cylinder. 61 singularities of. 136 Laplace equation axisymmetric. 15 in Curle’s equation. 159. 20 hydrodynamic. 118 Karweit. 195 near compact sphere. 71 sphere. see also compact Green’s function deﬁnition. 28 linear. 32 of surface dipoles. 190. 29 Force. 45 inhomogeneous equation. acoustic. 21 ﬂuid volume. 90 theory of vortex sound. 43 in three dimensions. 38 far ﬁeld. 87 dipole. 190. 29 equation. deﬁnition. deﬁnition. 100 leading edge. 12. 146. 96. of vorticity. 148 vortex. 158 lift dipole. pressure. 199 theorem on kinetic energy. 5 Mach number.214 decibel. 25. sound waves in. 14 directivity. 149 see also singularities Lamb vector. 89 in terms of impulse. 174 integrals. 32 Helmholtz. 150 suction. Kutta condition. 18 kinetic. 178. 90.
98 stream function. 47 end correction. deﬁnition. 103 equation of motion.Index micropressure wave. 147 Sears function. 78 reciprocal. 13 retarded time. 3 volume source. deﬁnition. 38 power. 161. 133 complex potential. 50 reduced frequency. 135 o momentum equation. 117 relation to pressure. 28 radiation condition. 162. 4 speed of sound. 16. 149 self potential. 3 Crocco’s form. coaxial rings. deleting from Green’s function. 3 near ﬁeld. 160 line. deﬁnition for homentropic ﬂow. 84. 180. compact. 57. 5 Reynolds’ form. 45 Lighthill’s. 67 vibrating. 75 Stokes. 166 M¨ hring. 173. 67 slender body approximation. 88 potential ﬂow interaction. 85 linearized. 200 Stokesian ﬂuid. 190 surface. 6 values for air and water. 85 acoustic variable. 22 image in plane wall. 28 viscous. 204 ﬁnite core. 156. 158 motion near halfplane. 117 tunnel. 84 . 189 periodic. 29 RungeKutta integration. 85 nonlinear steepening in tunnel. 107 singularities. 5 dipole. 150. 129 vibrating sphere. acoustic. 198. 56 low frequency radiation from. 65. acoustic particle. 175. 7. 7 sphere. 18 in terms of impulse. Second Law. 199 modelled by point source. 36 Reynolds stress. 183 threedimensional. 108 outside cylinder. 46. 14 Reynolds number. 3 kinematic. drag. added mass. 150 vortex. 68 interacting with a vortex. 169 spinning vortex pair. hydrodynamic. 11. 179. 187 speciﬁc heats. 166. 9 thermodynamics. 19. 13 in Curle’s equation. linear theory. 106. 16. 26 suction force. 127 problem for compact Green’s function. vortex. 9 potential. 27 monopole. 107 shedding. 19 quadrupole. 29 linear. vortex sound formula. 10. 3 values for air and water. 13 rate of strain tensor. 15 far ﬁeld form. 44 monopole. 25 velocity. 148 retarded potential. 146. 9. 39 in Helmholtz equation. 92. 9 pulsating. 15 vortexairfoil interaction. 26. 35 NavierStokes equation. 9 impulsive. 132. 166 plane wave. 84 Rayleigh. 158. 195. 180. 57. 182. 192. 6 vibrating body. 12 in Helmholtz equation. 28 in Lighthill tensor. 17 no slip condition. 37 215 test function. 188. deﬁnition. 11 total enthalpy. 111 ring. 108. deﬁnition. 174 turbulent nozzle ﬂow. 178. theorem. 159 tube. 145. 159. 6 vector. 185 dipole. 5. deﬁnition. deﬁnition. deﬁnition. 120 stationary phase. 27 momentum ﬂux tensor. 199 self potential. 188 nonlinear. 164 Powell. 116 in terms of velocity potential. deﬁnition. 163 Sears. 48 noncompact. velocity. 23 point source. 186 twodimensional. ratio. 67 viscosity. 178. 36 impedance. 89 potential. 3 stress tensor Lighthill’s.
10 wavelength acoustic. 82 molecular diffusion. cylinder. 139 from vortex near halfplane. 194 for nonvibrating body. 148 wave number. 13 radially symmetric. 150 wall drag. 96 Kelvin’s deﬁnition. 162 Index vortex near spoiler. 130. 116. 7 hydrodynamic. 84 equation for Stokesian ﬂuid. 197 vorticity. 91 wake. 191 wall mounted cylinder. 17. 204 spinning. 191 from vortex interacting with cylinder. 114 wake. 128 inﬂuence of vortex shedding. at low Mach numbers. bound. 119 for cylindrical bodies. 162.216 vortex pair. 120 vortexsurface interaction noise. 199 from wall aperture. 87 source of sound. 120 from spoiler. 88. 129 vortex sound equation. 118 linearized. 150 vortex sound. 148 vortex sheet. 204 in terms of surface force. 127 high Reynolds number form. 156 low Mach number approximation. 6 in irrotational mean ﬂow. 124 general formula. 93. 118 inhomogeneous. 43 . 143. 191 wave equation classical acoustics. 143 vortex near sphere. 87. 120 vortex sheet. 7. 86 in force formulae. 130 from spinning vortices. 43 at 1 kHz. 91 equation. 173 from vortex near sphere. 6 for pressure.
This action might not be possible to undo. Are you sure you want to continue?