Professional Documents
Culture Documents
Ebook Introduction To Simple Shock Waves in Air With Numerical Solutions Using Artificial Viscosity 2Nd Edition Sean Prunty Online PDF All Chapter
Ebook Introduction To Simple Shock Waves in Air With Numerical Solutions Using Artificial Viscosity 2Nd Edition Sean Prunty Online PDF All Chapter
https://ebookmeta.com/product/introduction-to-graph-theory-with-
solutions-to-selected-problems-2nd-edition-khee-meng-koh/
https://ebookmeta.com/product/solutions-manual-to-accompany-an-
introduction-to-numerical-methods-and-analysis-3rd-ed-3rd-
edition-epperson/
https://ebookmeta.com/product/introduction-to-the-mathematical-
physics-of-nonlinear-waves-2nd-edition-minoru-fujimoto/
https://ebookmeta.com/product/introduction-to-statistics-and-
data-analysis-with-exercises-solutions-and-applications-in-r-2nd-
edition-christian-heumann/
Introduction to AWS IaaS Solutions 1st Edition
Introduction To Aws Iaas Solutions
https://ebookmeta.com/product/introduction-to-aws-iaas-
solutions-1st-edition-introduction-to-aws-iaas-solutions/
https://ebookmeta.com/product/introduction-to-python-for-
engineers-and-scientists-open-source-solutions-for-numerical-
computation-1st-edition-sandeep-nagar/
https://ebookmeta.com/product/simple-solutions-to-energy-
calculations-6th-edition-vaillencourt/
https://ebookmeta.com/product/mediterranean-cooking-101-a-simple-
introduction-to-mediterranean-cooking-with-classical-healthy-
recipes-2nd-edition-booksumo-press/
https://ebookmeta.com/product/an-introduction-to-artificial-
intelligence-in-education-yu/
Shock Wave and High Pressure Phenomena
Seán Prunty
Introduction
to Simple Shock
Waves in Air
With Numerical Solutions Using
Artificial Viscosity
Second Edition
Shock Wave and High Pressure Phenomena
Founding Editor
Robert A. Graham, USA
Honorary Editors
Lee Davison, USA
Yasuyuki Horie, USA
Editorial Board
Gabi Ben-Dor, Israel
Frank K. Lu, USA
Naresh Thadhani, USA
L.L. Altgilbers, M.D.J. Brown, I. Grishnaev, B.M. Novac, I.R. Smith, I. Tkach, and
Y. Tkach: Magnetocumulative Generators
T. Antoun, D.R. Curran, G.I. Kanel, S.V. Razorenov, and A.V. Utkin: Spall Fracture
J. Asay and M. Shahinpoor (Eds.): High-Pressure Shock Compression of Solids
S.S. Batsanov: Effects of Explosion on Materials: Modification and Synthesis Under
High-Pressure Shock Compression
G. Ben-Dor: Shock Wave Reflection Phenomena
L.C. Chhabildas, L. Davison, and Y. Horie (Eds.): High-Pressure Shock Compres-
sion of Solids VIII
L. Davison: Fundamentals of Shock Wave Propagation in Solids
L. Davison, Y. Horie, and T. Sekine (Eds.): High-Pressure Shock Compression of
Solids
V.L. Davison and M. Shahinpoor (Eds.): High-Pressure Shock Compression of
Solids III
R.P. Drake: High-Energy-Density Physics
A.N. Dremin: Toward Detonation Theory
J.W. Forbes: Shock Wave Compression of Condensed Matter
V.E. Fortov, L.V. Altshuler, R.F. Trunin, and A.I. Funtikov: High-Pressure Shock
Compression of Solids VII
B.E. Gelfand, M.V. Silnikov, S.P. Medvedev, and S.V. Khomik: Thermo-Gas Dynam-
ics of Hydrogen Combustion and Explosion
D. Grady: Fragmentation of Rings and Shells
Y. Horie, L. Davison, and N.N. Thadhani (Eds.): High-Pressure Shock Compression
of Solids VI
J. N. Johnson and R. Cherét (Eds.): Classic Papers in Shock Compression Science
V.K. Kedrinskii: Hydrodynamics of Explosion
C.E. Needham: Blast Waves
V.F. Nesterenko: Dynamics of Heterogeneous Materials
S.M. Peiris and G.J. Piermarini (Eds.): Static Compression of Energetic Materials
M. Súceska: Test Methods of Explosives
M.V. Zhernokletov and B.L. Glushak (Eds.): Material Properties under Intensive
Dynamic Loading
J.A. Zukas and W.P. Walters (Eds.): Explosive Effects and Applications
Seán Prunty
Second Edition
Seán Prunty
Ballincollig, Cork, Ireland
© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland
AG 2019, 2021
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the
material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.
This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface to the Second Edition
v
Preface to the First Edition
involving shock waves that were of interest to the author. Accordingly, the reader is
strongly advised to consult the standard texts on fluid or gas dynamics for a
comprehensive account of fluid motion as well the rapidly growing area of CFD
by the experts in the area. Several textbooks and articles that I found useful can be
found in the references at the end of each chapter.
The book is suitable for both graduate and advanced undergraduate students in
applied mathematics, engineering and physics who are taking courses in fluid
dynamics and who require an introduction to a specific numerical technique for
dealing with shock waves. I decided very early on that the program listings would
not be included for the very good reason that there are good and bad programming
techniques and I would probably fall into the latter category. Nonetheless, the finite
difference representation of the differential equations is presented, and it is hoped
that this will be a starting point to encourage interested students to obtain solutions to
similar problems in compressible flow using their preferred programming language.
It is hoped that the material presented here will renew interest in gas dynamics and
shock waves in the undergraduate physics curriculum.
The book is structured in such a manner as to allow the reader to review some
basic material in relation to the equations of fluid flow. In this respect, a brief review
of the one-dimensional form of the equations is presented in Chap. 1 together with
the propagation of small amplitude disturbances. Chapter 1 also includes some basic
thermodynamic relationships that are relevant to gas dynamics. Waves of finite
amplitude and the formation of shock waves are discussed in Chap. 2. This chapter
also includes a basic introduction to the method of characteristics and to Riemann
invariants. Some important relationships arising from the conservation of mass,
momentum and energy across the shock front are presented in Chap. 3, and these
relationships are used in the subsequent chapters to ascertain the accuracy of the
numerical results obtained. Chapter 4 outlines the numerical procedure employed to
solve some examples of plane shock waves using artificial viscosity: The differential
equations in Lagrangian form are derived, and the corresponding difference equa-
tions are presented. Stability issues in relation to these equations are briefly
discussed as well as the choice of grid interval to be used in the numerical procedure.
Several simple examples of plane shocks arising from piston motion are also
presented and discussed. The remaining chapters deal with spherical shock waves:
Chap. 5 has an almost independent character and deals exclusively with the strong-
shock, point-source solution and its ramifications. Chapter 6 describes the numerical
procedure used when dealing with spherical shock waves, and the appropriate
differential equations in Lagrangian form are derived, and the difference form of
these equations incorporating artificial viscosity is presented. The equations are
numerically integrated to predict the pressure, density and particle velocity as
functions of position outside the strong-shock regime by using the strong-shock,
point-source solution as initial conditions. Finally, the shock waves generated
following the sudden expansion of a high-pressure, high-temperature sphere of air
into the surrounding atmosphere are presented and discussed.
Preface to the First Edition ix
It is assumed that the reader is familiar with differential and integral calculus and
has a solid understanding of the physical principles underpinning the basic equations
of fluid dynamics. Some knowledge of numerical methods would be an advantage,
particularly, to do with the stability issues in relation to the difference equations
involved.
I wish to acknowledge the help received from Diarmuid O’Riordáin in relation to
a computer file storage issue with embedded graphics.
xi
xii Contents
Appendix A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
Appendix B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339
Chapter 1
Brief Outline of the Equations of Fluid Flow
1.1 Introduction
Those who are familiar with compressible fluid flow are aware that the equations of
motion are nonlinear and, as such, it is very difficult to obtain analytical solutions.
As a consequence, numerical methods are generally employed and the differential
equations are approximated by finite difference equations and these in turn are
solved in a stepwise manner. In many examples involving compressible fluid flow
shocks appear and their presence is a complicating factor since they are characterized
by very steep gradients in the variables describing the flow, such as, in the velocity,
density, pressure and temperature. In fact, the gradients become infinitely steep when
the effects of viscosity and thermal conduction are neglected: this introduces dis-
continuities in the solutions and, as a result, it requires the application of boundary
conditions connecting the values across the shock front but the implementation of
this technique can be quite complex. However, the need for any boundary conditions
can be avoided by using a method proposed in 1950 by Von Neumann and
Richtmyer [1] where an artificially large viscosity is introduced into the numerical
calculations: the present text utilizes this technique. Instead of obtaining a discon-
tinuous solution at the shock front, the shock acquires a thickness comparable to the
spacing of the grid points used in the numerical procedure so that the shock appears
as a near-discontinuity and across which velocity, pressure etc. vary rapidly but
continuously.
A brief review of the fundamental equations of fluid dynamics [2–6] is provided
in this chapter so that the reader can have to-hand the appropriate governing
equations. The one-dimensional form of the equations is presented as they apply
to non-viscous1 flow, so that any physical effects involving friction and thermal
conduction are neglected. A treatment involving the derivation of the full
1
Although the term “non-viscous” implies the absence of viscosity or friction, it also implies “non-
conducting” as well.
3-dimensional fluid dynamic equations can be found in the excellent text by Ander-
son [7] where viscous forces and thermal conduction are included. However, as this
present text deals with fluid motion where shocks occur, the neglect of friction and
thermal conduction in the neighbourhood of the shock must be included and,
accordingly, an artificial viscosity is introduced, as we shall see in due course, into
the set of difference equations that approximate the differential equations for
the flow.
The equations governing a fluid or gas in motion are mathematical expressions of the
laws of conservation of mass, momentum and energy. In the case of an ideal gas, for
example, these equations are supplemented by thermodynamic equations; one called
the equation of state and the other called the caloric equation of state. The equations
describing the motion of the fluid can be written in one of two coordinate systems
[6]; one called the Eulerian system and the other called the Lagrangian system (each
named after the mathematicians, Euler and Lagrange). In the Lagrangian system we
follow the path taken by individual particles of fluid and determine the velocity,
pressure, density etc. as a function of the path taken. Accordingly, the Lagrangian
description of motion is connected with a definite particle or mass element of the
fluid; each particle is assigned a symbol x0, for example, indicating its position at
some initial time which is usually taken at time t ¼ 0, while at a later time t the
position of the particle is x(x0, t) and, clearly, x(x0, 0) ¼ x0. The motion of the particle
satisfies Newton’s second law of motion, namely; m(d2x/dt2) ¼ F where m is the
mass of this particle of fluid lying between x0 and x0 + dx0 and F is the force acting
on the particle. Suppose, for example, we wish to determine the flow of air over a
fixed surface, we could imagine a weightless soap bubble released into the air. The
path taken by this bubble as it moves over the surface provides a Lagrangian
description of the flow. To indicate the time-rate of change following the fluid
particle the material derivative, D/Dt rather than the usual d/dt has become standard
notation in fluid dynamics. However, in the present text we will adopt a slightly
different notation as we follow specific particles of fluid and we will be returning to
this aspect in Chaps. 4 and 6.
In the Eulerian description, on the other hand, one is not interested in the motion
of individual particles; instead, one is interested in, say, the velocity at points in
space. This is analogous to setting up a fixed and very fine grid throughout space and
noting, for example, the velocity, pressure or density etc. at each grid point, so in the
case of the Eulerian system we are interested in the properties of the fluid, such as,
velocity, pressure, density etc. as they pass fixed points in space.
Another good example, the source of which cannot be recalled, that distinguishes
the Lagrangian and Eulerian systems involves the acceleration of a log on a steadily
flowing river which has a section of rapids on it. By concentrating on the log one
observes that it accelerates as it enters the rapids which is the Lagrangian
1.3 Some Elements of Thermodynamics 3
acceleration (following the particle). However, an observer on the bank of the river
who concentrates on the velocity at fixed points in the flow will not see any
acceleration as a succession of logs that pass the same point will do so at the same
velocity since the river as a whole is not accelerating.
In mathematical terms, Du/Dt is the Lagrangian acceleration, while ∂u/∂t or more
specifically, (∂u/∂t)x is the Eulerian acceleration, where the subscript indicates that
the acceleration is measured at a particular point x. However, u u(x, t), so that
du Du ∂u ∂u
¼ þu :
dt Dt ∂t ∂x
In relation to the log entering the rapids we have, ∂u/∂t ¼ 0, so that Du/Dt ¼ u
(∂u/∂x); implying that the river exhibits a spatial variation in velocity, thereby
accounting for the Lagrangian acceleration.
m
pV ¼ ℜT, ð1:1Þ
M
where p is the pressure, V is the volume of gas of mass m, M is the molecular weight,
ℜ is the universal gas constant (ℜ ¼ 8.31JK1mole1) and T is the temperature. In
this text we will assume that air behaves as an ideal gas. One mole of air has an
approximate molecular weight (based on its composition which contains largely
nitrogen and oxygen) of 28.97 103kg, so that ℜ/M ¼ 287Jkg1K1. Accord-
ingly, Eq. (1.1) for air can be written as
pV ¼ mRT, ð1:2Þ
p ¼ ρRT or pυ ¼ RT ð1:3Þ
4 1 Brief Outline of the Equations of Fluid Flow
where ρ is the density and υ is the specific volume (that is, the inverse of the density).
For example, the density of air is ρ ¼ 1.29kgm3 at a pressure of one atmosphere
(equal to 1.013 105Nm2) and at a temperature of 273 K according to Eq. (1.3).
The first law of thermodynamics states that the energy of an isolated system is
conserved; however, the energy may be transformed from one form to another but
the energy cannot be created nor destroyed. The mathematical form of the first law of
thermodynamics for an infinitesimal change is given by the equation,
dQ ¼ dE þ dW, ð1:4Þ
where dQ is the quantity of heat exchanged between the system and its surroundings,
dE2 is the change in the internal energy of the system and dW is the work done. For a
system of constant mass m that exerts a uniform pressure p on its surroundings, the
first law can be written as
dQ ¼ mde þ pdV,
where de is the change in the specific internal energy, that is, the change in the
internal energy per unit mass and dV is the change in volume. Dividing across by m,
gives,
dq ¼ de þ pdυ, ð1:5Þ
where dq is the quantity of heat transferred per unit mass and dυ is the change in the
specific volume.
2
The symbol E is used to represent the internal energy of a system rather than the more commonly
used symbol U, as this symbol is used in this text to represent the velocity of air motion and the
velocity of shock waves.
1.3 Some Elements of Thermodynamics 5
dq
c¼ :
dT
If the process takes place at constant volume, then the heat capacity at constant
volume is
dq
cV ¼ ð1:6Þ
dT V
and if the process takes place at constant pressure, the heat capacity is
dq
cP ¼ : ð1:7Þ
dT P
hence,
∂e
cV ¼ ð1:9Þ
∂T υ
h ¼ e þ pυ, ð1:10Þ
then
dh ¼ de þ pdυ þ υdp
¼ dq þ υdp,
hence,
6 1 Brief Outline of the Equations of Fluid Flow
dq ¼ dh υdp: ð1:11Þ
hence,
∂h ∂h
dq ¼ dT þ υ dp, ð1:13Þ
∂T p ∂p T
Also, for an ideal gas the specific internal energy is a function of the temperature
only, hence, E ¼ f(T ) and, therefore,
∂e de
cV ¼ ¼ ,
∂T υ
dT
hence,
e ¼ cV T ð1:15Þ
and similarly,
h ¼ cP T: ð1:16Þ
cP T ¼ cV T þ RT,
so that
cP ¼ cV þ R: ð1:17Þ
1.3 Some Elements of Thermodynamics 7
Work in thermodynamics is concerned only with the changes that take place between
a system and its surroundings. An infinitesimal amount of work dW is said to be done
by a system when the system undergoes a change in volume dV under the action of
the pressure p that the system exerts on its surroundings, hence,
dW ¼ pdV ð1:18Þ
and if the volume changes from an initial value Vi to a final value Vf the work done by
the system is given by
ZV f
W¼ pdV: ð1:19Þ
Vi
This latter equation cannot be integrated until the pressure is known as a function
of V. This means that dW is not an exact differential; it depends on the path, unlike,
for example, the internal energy function, E, which is an exact differential and only
depends on the initial and final states and therefore independent of the path taken in
going from, say, Ei to Ef.
In the case of an isothermal expansion or compression of an ideal gas whose
equation of state is given by
pV ¼ mRT:
The work done can be calculated by substituting for the pressure, p, hence,
ZV f
mRT
W¼ dV: ð1:20Þ
V
Vi
where ln in the latter equation denotes the natural logarithm. If Vf>Vi the system does
work on its surroundings and, on the other hand, if Vf<Vi, the surroundings does
work on the system, that is, the system is compressed.
8 1 Brief Outline of the Equations of Fluid Flow
Let us now consider a change in a system that undergoes an adiabatic process, that
is, changes take place to the system without the addition or removal of heat, hence,
dq ¼ 0. During the process the system departs very little from thermodynamic
equilibrium, so that the change taking place in going from some initial state to
some final state goes through a sequence of equilibrium states. The mathematical
statement of the first law of thermodynamics in infinitesimal form is
dq ¼ de þ pdυ
¼ cV dT þ pdυ: ð1:22Þ
1
dT ¼ ðpdυ þ υdpÞ
R
cV
dq ¼ ðpdυ þ υdpÞ þ pdυ
R
cV c
¼ υdp þ P pdυ:
R R
dp dυ
þ γ ¼ 0, ð1:23Þ
p υ
where γ ¼ cP/cV is the ratio of the specific heats. Integrating Eq. (1.23), gives
p
¼ constant: ð1:25Þ
ργ
By using the equation for an ideal gas, one can show that the following relation-
ships apply for an adiabatic process;
T
γ1 ¼ constant and Tυγ1 ¼ constant: ð1:26Þ
p γ
1.3 Some Elements of Thermodynamics 9
dq ¼ cV dT þ dW: ð1:27Þ
dW ¼ cV dT ð1:28Þ
where Ti and Tf are the initial and final temperatures, respectively. Since the
process in going from Ti to Tf passes through equilibrium states, we can use the
ideal gas equation and write the latter equation as
cV
W¼ pi υ i p f υ f ð1:30Þ
R
and using the fact that R ¼ cP cV in conjunction with γ ¼ cP/cV Eq. (1.30) becomes
pi υi p f υ f
W¼ : ð1:31Þ
γ1
cV pυ
e ¼ cV T ¼ pυ ¼ , ð1:33Þ
R γ1
p
e¼ : ð1:34Þ
ðγ 1Þρ
Similarly,
γp
h¼ : ð1:35Þ
ðγ 1Þρ
Let us now consider the values for cV and cP or, more specifically, the value of γ for
air. In relation to the Kinetic Theory of Gases, the principle of equipartition of
energy states that the energy associated with each degree of freedom of an atom or
molecule is (1/2)kBT, where kB is Boltzmann’s constant (kB ¼ 1.38 1023JK1).
Each atom or molecule has three translational degrees of freedom; namely, in the
x, y and z-directions, giving (3/2)kBT for its internal energy.
One mole, corresponding to the molecular weight M, contains NA atoms or
molecules (NA ¼ 6.02 1023), so that the specific internal energy of one mole is
(3/2)NAkBT ¼ (3/2)ℜT, where ℜ ¼ NAkB is the universal gas constant.
Air comprises largely N2 and O2 molecules and each molecule contributes two
rotational degrees of freedom in addition to the translational degrees, hence, the
specific internal energy for air amounts to (5/2)RT and the specific enthalpy amounts
to (7/2)RT, hence,
5 7
e ¼ RT and h ¼ RT ð1:36Þ
2 2
dQrev
dS ¼ , ð1:37Þ
T
reservoir T
12 1 Brief Outline of the Equations of Fluid Flow
where the subscript rev on dQ implies that the heat transfer must occur reversibly.3
In the case of a finite change from some initial (i) state to a final ( f ) state, the entropy
change of the system is given by
Zf
dQrev
S f Si ¼ ð1:38Þ
T
i
and the integral must be evaluated along any reversible path connecting the initial
and final states. Entropy, like internal energy, is a property of the system; hence, the
change Sf Si is independent of the path connecting the two states. For a reversible
adiabatic process, dQrev ¼ 0, so that dS ¼ 0, hence, the entropy remains constant and
the process is called isentropic. An adiabatic process is one in which no heat
exchange occurs between the system and its surroundings, so, in general, an adia-
batic process is not necessarily isentropic; the process must be reversible for it to be
isentropic.
If a system undergoes an irreversible process between an initial and a final state as
illustrated in Fig. 1.2, then the change in entropy ΔS is given by
ΔS ¼ S f Si ,
where it is assumed that these initial and final states are equilibrium states. One can
calculate this change in entropy by replacing the irreversible path (broken line in
Fig. 1.2) by any reversible path connecting these initial and final states. The change
in entropy can be obtained by evaluating the integral,
irreversible
final
process state
f
V
3
A reversible process is one in which the system and its surroundings can be restored to their initial
states following the conclusion of the process without causing any changes elsewhere.
1.4 Conservation Equations in Plane Geometry 13
Z
dQ
T
iAf
where, for example, the integration is carried out for the reversible path iAf that is
shown in Fig. 1.2: the path, i ! A is taken as a constant pressure process and the path
A ! f is taken as a constant volume process.
In calculating any entropy changes one must also include the entropy change of
the surroundings that interact with the system. Hence, the total entropy change is
specified as the entropy change of the universe which is equal to the entropy change
of the system plus the entropy change of the surrounding environment and is given
by
When the second law for a closed system is expressed in terms of entropy change,
we have
ΔSTotal 0, ð1:40Þ
where the equality sign applies to a reversible process and the inequality sign applies
to an irreversible process. Therefore, reversible processes produce no change in the
entropy of the universe while all irreversible processes result in an increase in the
entropy of the universe.
Let us now provide a brief review of the one-dimensional equations of fluid flow for
plane geometry. These are the conservation equations for mass, momentum and
energy.
The continuity equation can be obtained by considering the mass of fluid entering
and leaving a small element of volume Adx lying between x and x + dx as shown in
Fig. 1.3. It is assumed that the motion is one-dimensional so that velocity, density
and pressure are constant over the cross-sectional area A. Mass conservation implies
that the difference between the mass flow rate into the volume element and the mass
flow rate out of the volume element is equal to the rate of accumulation of mass
within the volume element.
14 1 Brief Outline of the Equations of Fluid Flow
and this must be equal to the rate of increase of mass within this element, that is, Adx
(∂ρ/∂t), hence the one-dimensional continuity equation in Eulerian form is
∂ρ ∂
þ ðρuÞ ¼ 0: ð1:41Þ
∂t ∂x
∂ρ ∂ρ ∂u
þu þρ ¼ 0,
∂t ∂x ∂x
Dρ ∂u
¼ ρ , ð1:42Þ
Dt ∂x
Dρ ! !
¼ ρ∇ V : ð1:43Þ
Dt
with
!
V ¼ ðu, v, wÞ
where u, v and w are the velocity components in the x, y and z-directions, respec-
tively, and the gradient operator is
! ∂ ∂ ∂
∇¼ , , :
∂x ∂y ∂z
1.4 Conservation Equations in Plane Geometry 15
or
By writing the derivative in Eq. (1.42) in terms of the specific volume υ ¼ 1/ρ
rather than the density ρ we have,
Dυ ∂u
ρ ¼ : ð1:44Þ
Dt ∂x
The law of the conservation of momentum implies that the rate of change in
momentum of the fluid in a volume element between x and x + dx is equal to the
rate at which momentum flows into the volume element at x minus the rate at which
momentum flows out at x + dx plus the net force acting on the volume element.
Expressing this mathematically, we have [8],
∂
ðρuÞAdx ¼ ρðx, t Þu2 ðx, t ÞA ρðx þ dx, t Þu2 ðx þ dx, t ÞA þ pðx, t ÞA
∂t
pðx þ dx, t ÞA
hence,
∂ ∂ ∂pðx, t Þ
ðρuÞAdx ¼ ρðx, t Þu2 ðx, t Þ Adx Adx,
∂t ∂x ∂x
so that
∂ ∂ 2 ∂p
ðρuÞ þ ρu ¼ , ð1:45Þ
∂t ∂x ∂x
which is the momentum equation. By using the continuity equation it is easy to show
that the latter equation can be written in final form as
16 1 Brief Outline of the Equations of Fluid Flow
∂u ∂u 1 ∂p
þu ¼ , ð1:46Þ
∂t ∂x ρ ∂x
Du ∂p
ρ ¼ , ð1:47Þ
Dt ∂x
and for the sake of completeness, the 3-dimensional form of Eq. (1.47) can be
obtained by noting that u, in general, is a function of x, y and z, hence,
u ¼ uðx, y, z, t Þ
and, therefore,
∂u ∂u ∂u ∂u
du ¼ dt þ dx þ dy þ dz,
∂t ∂x ∂y ∂z
accordingly,
Du du ∂u ∂u ∂u ∂u
¼ þu þv þw ,
Dt dt ∂t ∂x ∂y ∂z
Du ∂u ! !
¼ þ V ∇u,
Dt ∂t
!
where V is given by
!
V ¼ ubx þ vby þ wbz
and (u, v, w) are the velocity components in the x, y and z-directions, respectively,
and ðbx, by,bzÞ are unit vectors in these directions, hence, Eq. (1.47) becomes,
Du ∂u ! ! ∂p
ρ ¼ρ þ ρV ∇u ¼ ð1:48aÞ
Dt ∂t ∂x
Dv ∂v ! ! ∂p
ρ ¼ ρ þ ρV ∇v ¼ ð1:48bÞ
Dt ∂t ∂y
Dw ∂w ! ! ∂p
ρ ¼ρ þ ρV ∇w ¼ ; ð1:48cÞ
Dt ∂t ∂z
∂v ∂v ∂v ∂v ∂p
ρ þu þv þw ¼ ð1:49bÞ
∂t ∂x ∂y ∂z ∂y
∂w ∂w ∂w ∂w ∂p
ρ þu þv þw ¼ ð1:49cÞ
∂t ∂x ∂y ∂z ∂z
Let us now consider how the energy within the small volume element Adx changes
with time, this energy is comprised of two parts; kinetic energy due to gas motion
and internal energy due to molecular motion. The rate of change of this energy with
time is
2
∂ u
ρ þ e Adx
∂t 2
where e is the internal energy per unit mass (for an ideal gas e ¼ p/(γ 1)ρ). This
rate is equal to the rate of flow of energy into this volume element at x minus the rate
of flow of energy out of the volume element at x + dx, plus the rate at which pressure
forces does work at x minus the rate at which pressure forces does work at x + dx.
Expressing this mathematically, we have [8],
2 h i
∂ u 1
ρ þ e Adx ¼ ρðx, t Þu2 ðx, t Þ þ ρðx, t Þeðx, t Þ uðx, t ÞA
∂t 2 2
18 1 Brief Outline of the Equations of Fluid Flow
h i
1
ρðx þ dx, t Þu2 ðx þ dx, t Þ þ ρðx þ dx, t Þeðx þ dx, t Þ uðx þ dx, t ÞAþ
2
pðx, t Þuðx, t ÞA pðx þ dx, t Þuðx þ dx, t ÞA:
By carrying out a Taylor expansion to first order in dx, namely, for example,
ρðx þ dx, t Þ ¼ ρðx, t Þ þ ∂ρ
∂x
dx with similar expansions for u(x + dx, t) and e(x + dx, t),
it is straightforward to show (after neglecting terms of the order of dx2) that
2
∂ u 3 2 ∂u u3 ∂ρ ∂e ∂u ∂ρ
ρ þ e Adx ¼ Adx ρu þ þ ρu þ ρe þ ue
∂t 2 2 ∂x 2 ∂x ∂x ∂x ∂x
∂ðpuÞ
Adx
∂x
∂ ρu3 ∂ ∂ðpuÞ
¼ Adx þ ðuρeÞ Adx
∂x 2 ∂x ∂x
2
∂ u ∂ðpuÞ
¼ Adx ρu þ e Adx :
∂x 2 ∂x
Hence,
2 2
∂ u ∂ u
ρ þe þ ρu þ e þ pu ¼ 0, ð1:50Þ
∂t 2 ∂x 2
which is the one-dimensional energy balance equation in Eulerian form. The three
equations; continuity, motion and energy are supplemented by two other equations;
an equation of state of the form, p ¼ p(ρ, T) and a caloric equation of the form, e ¼ e
(ρ, T ). For an ideal gas we have already seen that these equations are p ¼ ρRT and
e ¼ cVT ¼ p/ρ(γ 1), respectively. These five equations are sufficient to determine
all five quantities, u, p, ρ, Tand e. It is important to note that external forces such as
gravity as well viscous forces have been neglected in the momentum equation;
similarly, heat transport arising from temperature gradients as well as viscous forces
have also been neglected in the energy balance equation. In real fluids, however,
these quantities are never quite zero but in the case of idealized flow the neglect of
these quantities forms a substantial and useful part of fluid dynamics.
Writing Eq. (1.50) as
∂ ∂
½ρϖ þ ½ρuϖ þ pu ¼ 0
∂t ∂x
2
where ϖ ¼ u2 þ e. Hence,
1.4 Conservation Equations in Plane Geometry 19
∂ ∂
½ρϖ þ ½uðρϖ þ pÞ ¼ 0
∂t ∂x
∂ ∂ ∂u
ðρϖ Þ þ u ðρϖ þ pÞ þ ðρϖ þ pÞ ¼0
∂t ∂x ∂x
∂ ∂ ∂p ∂u ∂u
ðρϖ Þ þ u ðρϖ Þ þ u þ ρϖ þp ¼ 0:
∂t ∂x ∂x ∂x ∂x
Hence,
∂ϖ ∂ρ ∂ϖ ∂ρ ∂p ∂u ∂u
ρ þϖ þ uρ þ uϖ þu þ ρϖ þp ¼0
∂t ∂t ∂x ∂x ∂x ∂x ∂x
The quantity in square brackets in the latter equation goes to zero as we recognise
it as the terms appearing in the continuity equation, hence,
∂ϖ ∂ϖ ∂p ∂u
ρ þu þu þp ¼0 ð1:51Þ
∂t ∂x ∂x ∂x
2
Now using; ϖ ¼ u2 þ e, we have
∂ϖ ∂u ∂e ∂ϖ ∂u ∂e
¼u þ and ¼u þ
∂t ∂t ∂t ∂x ∂x ∂x
Similarly, the quantity in square brackets in this latter equation goes to zero
according to the momentum equation; hence,
∂e ∂e ∂u
ρ þu þp ¼ 0:
∂t ∂x ∂x
By substituting for ∂u/∂x using the continuity equation this latter equation can be
written in the following form;
∂e ∂e p ∂ρ ∂ρ
ρ þu þu ¼0 ð1:52Þ
∂t ∂x ρ ∂t ∂x
De p Dρ
ρ ¼ 0: ð1:53Þ
Dt ρ Dt
As ρ ¼ 1/υ, where υ is the specific volume (that is, the volume per unit mass of
material), then
Dρ Dυ
¼ ρ2 ,
Dt Dt
De Dυ
¼ p : ð1:54Þ
Dt Dt
pυ
e¼
γ1
Dðpυγ Þ
¼ 0,
Dt
so that
∂ ∂
þu ðpυγ Þ ¼ 0:
∂t ∂x
It is important to appreciate that this result has been obtained by assuming that the
fluid element is non-conducting, devoid of viscosity and obeys the ideal gas equation
1.5 Constancy of the Entropy with Time for a Fluid Element 21
with constant heat capacities. The latter equation implies that the product pυγ
remains constant as we follow the fluid element in its motion. If, for example, the
fluid element has pressure p0 and specific volume υ0 at some instant in time, then its
pressure p and specific volume υ at later times are related according to the equation,
pυγ ¼ p0 υγ0 :
Consequently, the quantity pυγ remains constant in the flow and, as such, it
assumes the status of a state variable which, we will see in the subsequent discussion,
is related to another state variable called the specific entropy s, where we show in
Sect. 1.6 that
s ¼ cV ln ðpυγ Þ þ constant:
Provided the fluid motion experiences no abrupt changes in any of the quantities, u,p
or ρ etc., the conservation of energy implies the constancy of entropy [9, 10]. Accord-
ingly, if the fluid element is in thermodynamic equilibrium during the motion the
entropy of a fluid element will remain constant with time. In order to see this we need
to consider reversible changes taking place in thermodynamic systems. For a
reversible thermodynamic change we have [11],
TdS ¼ dE þ pdV
where dS is the change in entropy, dE is the change in internal energy and pdV
represents the work done in the process. Writing the latter equation as
dE ¼ TdS pdV
and dividing both sides of this latter equation by the mass m of a fluid element we
have
de ¼ Tds pdυ
where e is the internal energy per unit mass, ds is the entropy change per unit mass
and υ is the specific volume. As υ ¼ 1/ρ, we have dυ ¼ (1/ρ2)dρ, so that
p
de ¼ Tds þ dρ: ð1:55Þ
ρ2
22 1 Brief Outline of the Equations of Fluid Flow
e ¼ eðs, ρÞ,
so that
∂e ∂e
de ¼ ds þ dρ: ð1:56Þ
∂s ρ ∂ρ s
and
∂e ∂e ∂s ∂e ∂ρ
¼ þ ,
∂x ∂s ρ ∂x ∂ρ s ∂x
hence
∂e ∂s p ∂ρ
¼T þ
∂t ∂t ρ2 ∂t
and
∂e ∂s p ∂ρ
¼T þ ,
∂x ∂x ρ2 ∂x
after using Eq. (1.57). Substituting these latter two equations in Eq. (1.52) we obtain,
∂s p ∂ρ ∂s up ∂ρ p ∂ρ ∂ρ
ρ T þ þ uT þ þu ¼ 0,
∂t ρ2 ∂t ∂x ρ2 ∂x ρ ∂t ∂x
so that
1.6 Entropy Change for an Ideal Gas 23
∂s ∂s
ρT þu ¼0
∂t ∂x
or
Ds
¼ 0: ð1:58Þ
Dt
This equation implies that the entropy of the fluid element or particle does not
change with time along the element’s path of motion and, as a result, the flow is
called isentropic. However, if discontinuities, such as, shock waves occur in the
flow, even in the case of an ideal fluid, there will be an increase in entropy across the
discontinuity and the equation Ds/Dt ¼ 0 no longer applies.
If the fluid is an ideal gas, we can calculate the entropy change (per unit mass) as
follows;
Hence,
cV dT pdυ
ds ¼ þ ,
T T
where cP ¼ cV + R, hence,
dυ dp
ds ¼ cP þ cV ,
υ p
therefore,
24 1 Brief Outline of the Equations of Fluid Flow
ds dυ dp
¼γ þ ,
cV υ p
Zs Zυ Zp
ds dυ dp
¼γ þ ,
cV υ p
s0 υ0 p0
hence,
s s0 υ p pυγ
¼ γ ln þ ln ¼ ln ,
cV υ0 p0 p0 υ 0 γ
so that
pυγ ss0
γ ¼ e cV :
p0 υ 0
Accordingly, constancy of the entropy as expressed by Eq. (1.58) implies that the
fluid element lies on the same adiabatic curve when it commenced its motion, hence,
∂ ∂
þu ðpυγ Þ ¼ 0, ð1:59Þ
∂t ∂x
Let us now turn our attention to the conservation equations in spherical geometry as
we will be requiring these equations later on when spherical motion is discussed.
Initially, we will present the general form of these equations in spherical coordinates
before considering the case where the motion is confined to take place in the radial
direction and the equations reduce to their one-dimensional form.
1.7 Spherical Geometry 25
Initially, we will spend some time developing various vector relationships that
will be required for writing the conservation equations in spherical geometry.
Relationships Between the Unit Vectors
!
In a Cartesian coordinate system the position vector r is given by the equation,
!
r ¼ xbx þ yby þ zbz,
where bx, by and bz are unit vectors in the x, y and z directions, respectively. In terms of
the spherical coordinate system as shown in Fig. 1.4, the latter equation becomes,
!
r ¼ ðrSinθCosϕÞbx þ ðrSinθSinϕÞby þ ðrCosθÞbz,
since x ¼ rSinθCosϕ, y ¼ rSinθSinϕ and z ¼ rCosθ from Fig. 1.4. Tangent vectors in
! ! !
the r, θ, ϕ directions are given by ∂ r =∂r, ∂ r =∂θ and ∂ r =∂ϕ, respectively, and
unit vectors (br , b b in these directions as shown in Fig. 1.4 are given by
θ, ϕ)
! ! !
∂ r =∂r b ∂ r =∂θ b ¼ ∂ r =∂ϕ :
br ¼ ! , θ¼ ! and ϕ !
∂ r =∂r ∂ r =∂θ ∂ r =∂ϕ
!
By carrying out the differentiation of the vector r as prescribed by these latter
equations, we obtain the following set of relationships for the unit vectors in the
spherical coordinate system in terms of the unit vectors in the Cartesian coordinate
system,
b
θ ¼ ðCosθCosϕÞbx þ ðCosθSinϕÞby ðSinθÞbz
b ¼ ðSinϕÞbx þ ðCosϕÞby
ϕ
∂br ∂b
θ b
∂ϕ
¼ 0, ¼ 0 and ¼0
∂r ∂r ∂r
∂br
¼ ðCosθCosϕÞbx þ ðCosθSinϕÞby ðSinθÞbz ¼ b
θ
∂θ
∂b
θ
¼ ðSinθCosϕÞbx ðSinθSinϕÞby ðCosθÞbz ¼ br
∂θ
b
∂ϕ
¼0
∂θ
1.7 Spherical Geometry 27
∂br b
¼ ðSinθSinϕÞbx þ ðSinθCosϕÞby ¼ ϕSinθ
∂ϕ
∂b
θ b
¼ ðCosθSinϕÞbx þ ðCosθCosϕÞby ¼ ϕCosθ
∂ϕ
b
∂ϕ
¼ ðCosϕÞbx ðSinϕÞby ¼ ðSinθÞbr ðCosθÞb
θ
∂ϕ
! ∂br ∂br ∂br
d r ¼ dðrbr Þ ¼ br dr þ rdbr ¼ br dr þ r dr þ dθ þ dϕ
∂r ∂θ ∂ϕ
and by using the previous relationships for the derivatives of these unit vectors we
can write the previous equation in the form,
!
d r ¼ br dr þ b b
θrdθ þ ϕrSinθdϕ:
→
The Vector Differential Operator del, Written as — , in Spherical Coordinates
!
The differential operator ∇ in a Cartesian coordinate system is
! ∂ ∂ ∂
∇ ¼ bx þ by þ bz ,
∂x ∂y ∂z
and we need to express this operator in terms of the unit vectors br , b b in the
θ, ϕ
spherical coordinate system. In order to so this, let us consider a function Ψ that
depends on r, θ, ϕ, that is, Ψ ¼ Ψ(r, θ, ϕ), then the derivative dΨ is
∂Ψ ∂Ψ ∂Ψ
dΨ ¼ dr þ dθ þ dϕ
∂r ∂θ ∂ϕ
! !
dΨ ¼ ∇Ψ d r
28 1 Brief Outline of the Equations of Fluid Flow
! ! !
¼ ∇Ψ dr þ ∇Ψ rdθ þ ∇Ψ rSinθdϕ,
r θ ϕ
!
after substituting for d r , hence,
! ∂Ψ ! 1 ∂Ψ ! 1 ∂Ψ
∇Ψ ¼ , ∇Ψ ¼ and ∇Ψ ¼
r ∂r θ r ∂θ ϕ rSinθ ∂ϕ
! ∂ b1 ∂ b 1 ∂ :
∇ ¼ br þθ þϕ ð1:61Þ
∂r r ∂θ rSinθ ∂ϕ
→
Divergence of a Vector A in Spherical Coordinates
! !
Performing the following dot product of ∇ with the vector A we have
! ! ∂ 1 ∂ b 1 ∂ Arbr þ Aθ b
∇ A ¼ br þ b
θ þϕ b
θ þ Aϕ ϕ
∂r r ∂θ rSinθ ∂ϕ
∂ ∂ ∂
¼ br ðArbr Þ þ br Aθ b
θ þ br b
Aϕ ϕ
∂r ∂r ∂r
1 ∂ 1 ∂ 1 ∂
þb
θ ðA br Þ þ b θ Abθ þb θ A ϕb
r ∂θ r r ∂θ θ r ∂θ ϕ
b 1 ∂ ðArbr Þ þ ϕ
þϕ b 1 ∂ Aθ b θ þϕ b 1 ∂ Aϕ ϕb
rSinθ ∂ϕ rSinθ ∂ϕ rSinθ ∂ϕ
Carrying out the differentiation of the various terms and by using the previously
developed derivatives of the unit vectors, we find that the latter equation reduces to,
! ! 1 ∂ 2 1 ∂ 1 ∂Aϕ
∇A ¼ r Ar þ ðAθ SinθÞ þ : ð1:62Þ
r ∂r
2 rSinθ ∂θ rSinθ ∂ϕ
This completes our brief look at the various relationships that are required for the
subsequent discussion of the conservation equations in spherical geometry.
1.7 Spherical Geometry 29
∂ρ ∂ρ ∂ρ ∂ρ ! !
þu þv þw þ ρ∇ V ¼ 0
∂t ∂x ∂y ∂z
∂ρ ! ! ! !
þ V ∇ρ þ ρ∇ V ¼ 0
∂t
∂ρ ! !
þ ∇ ρV ¼ 0:
∂t
This is the vector form of the continuity equation. Using Eq. (1.62) in this latter
equation we obtain the following continuity equation in spherical geometry;
∂ρ 1 ∂ðρr 2 ur Þ 1 ∂ðρuθ SinθÞ 1 ∂ ρuϕ
þ þ þ ¼ 0,
∂t r 2 ∂r rSinθ ∂θ rSinθ ∂ϕ
!
with V ¼ urbr þ uθ b b where ur, uθ and uϕ are the components of the velocity
θ þ uϕ ϕ,
b b
vector in the br , θ and ϕ directions, respectively. If the motion takes place purely in
the radial direction (uθ ¼ 0 and uϕ ¼ 0), then the latter equation becomes
∂ρ 1 ∂ðρr 2 ur Þ
þ ¼ 0,
∂t r 2 ∂r
and by carrying out the differentiation we obtain the following continuity equation,
∂ρ ∂ρ ∂u 2u
þu þρ þ ¼ 0, ð1:63Þ
∂t ∂r ∂r r
Recalling Eq. (1.48) and writing the components of the equation of motion in the
form,
∂u ! ! 1 ∂p
þ V ∇ u¼
∂t ρ ∂x
∂v ! ! 1 ∂p
þ V ∇ v¼
∂t ρ ∂y
∂w ! ! 1 ∂p
þ V ∇ w¼ :
∂t ρ ∂z
Multiplying the first of these equations above by bx, the second equation by by and
the third equation by bz, yielding,
∂ubx ! ! 1 ∂p
þ V ∇ ubx ¼ bx ,
∂t ρ ∂x
∂vby ! ! 1 ∂p
þ V ∇ vby ¼ by
∂t ρ ∂y
∂wbz ! ! 1 ∂p
þ V ∇ wbz ¼ bz ,
∂t ρ ∂z
by adding all three components of the above equation we can write it as a single
equation in the form,
∂ ! ! 1 ∂ ∂ ∂
ðubx þ vby þ wbzÞ þ V ∇ ðubx þ vby þ wbzÞ ¼ bx þ by þ bz p,
∂t ρ ∂x ∂y ∂z
which gives the following momentum equation for inviscid flow in vector notation,
!
∂V ! ! ! 1!
þ V ∇ V ¼ ∇p, ð1:64Þ
∂t ρ
! !
where V ¼ ubx þ vby þ wbz and ∇ ¼ bxð∂=∂xÞ þ byð∂=∂yÞ þ bzð∂=∂zÞ.
1.7 Spherical Geometry 31
!
We have already noted that the velocity vector V in terms of the orthogonal
spherical velocity components, ur, uθ, uϕ, is given by
!
V ¼ urbr þ uθ b b
θ þ uϕ ϕ
!
and we have shown that the vector differential operator ∇ for this spherical
coordinate system is
! ∂ b1 ∂ b 1 ∂ ,
∇ ¼ br þθ þϕ
∂r r ∂θ rSinθ ∂ϕ
! !
hence V ∇ in Eq. (1.64) becomes,
! ! ∂ uθ ∂ uϕ ∂
V ∇ ¼ ur þ þ
∂r r ∂θ rSinθ ∂ϕ
! ! !
and forming the product V ∇ V we have
! ! ! ∂ uθ ∂ uϕ ∂
V ∇ V ¼ ur þ þ urbr þ uθ b b
θ þ uϕ ϕ
∂r r ∂θ rSinθ ∂ϕ
∂ ∂ ∂ b þ uθ ∂ ðurbr Þ þ uθ ∂ uθ b
¼ ur ðurbr Þ þ ur uθ b
θ þ uϕ ϕ θ
∂r ∂r ∂r r ∂θ r ∂θ
u ∂ b þ uϕ ∂ ðurbr Þ þ uϕ ∂ uθ b uϕ ∂ b :
þ θ uϕ ϕ θ þ u ϕ
r ∂θ rSinθ ∂ϕ rSinθ ∂ϕ rSinθ ∂ϕ ϕ
Carrying out the differentiation in this latter equation and using the previously
established results for the derivatives of the unit vectors, we can write the r-compo-
nent, the θ-component and the ϕ-component of Eq.(1.64), in the following form,
FAR back in the long ago time New Zealand was a crowded happy
land. Big Maori fortress villages crowned the hilltops, broad farms
covered the hillsides; the chiefs kept a good table, cooking was
excellent, and especially when prisoners were in season, the people
feasted between sleeps, or, should provisions fail, sacked the next
parish for a supply of meat. So many parishes were sacked and
eaten, that in the course of time the chiefs led their tribes to quite a
distance before they could find a nice fat edible village, but still the
individual citizen felt crowded after meals, and all was well.
Then came the Pakehas, the white men, trading, with muskets
for sale, and the tribe that failed to get a trader to deal with was very
soon wiped out. A musket cost a ton of flax, and to pile up enough to
buy one a whole tribe must leave its hill fortress to camp in
unwholesome flax swamps. The people worked themselves thin to
buy guns, powder and iron tools for farming, but they cherished their
Pakeha as a priceless treasure in special charge of the chief, and if a
white man was eaten, it was clear proof that he was entirely useless
alive, or a quite detestable character. The good Pakehas became
Maori warriors, a little particular as to their meat being really pig, but
otherwise well mannered and popular.
Now of these Pakeha Maoris, one has left a book. He omitted his
name from the book of Old New Zealand, and never mentioned
dates, but tradition says he was Mr. F. C. Maning, and that he lived
as a Maori and trader for forty years, from 1823 to 1863 when the
work was published.
In the days when Mr. Maning reached the North Island a trader
was valued at twenty times his weight in muskets, equivalent say, to
the sum total of the British National Debt. Runaway sailors however,
were quite cheap. “Two men of this description were hospitably
entertained one night by a chief, a very particular friend of mine,
who, to pay himself for his trouble and outlay, ate one of them next
morning.”
Maning came ashore on the back of a warrior by the name of
Melons, who capsized in an ebb tide running like a sluice, at which
the white man, displeased, held the native’s head under water by
way of punishment. When they got ashore Melons wanted to get
even, so challenged the Pakeha to a wrestling match. Both were in
the pink of condition, the Maori, twenty-five years of age, and a
heavy-weight, the other a boy full of animal spirits and tough as
leather. After the battle Melons sat up rather dazed, offered his hand,
and venting his entire stock of English, said “How do you do?”
But then came a powerful chief, by name Relation-eater. “Pretty
work this,” he began, “good work. I won’t stand this not at all! not at
all! not at all!” (The last sentence took three jumps, a step and a turn
round, to keep correct time.) “Who killed the Pakeha? It was Melons.
You are a nice man, killing my Pakeha ... we shall be called the
‘Pakeha killkillers’; I shall be sick with shame; the Pakeha will run
away; what if you had killed him dead, or broken his bones”.... (Here
poor Melones burst out crying like an infant). “Where is the hat?
Where the shoes? The Pakeha is robbed! he is murdered!” Here a
wild howl from Melons.
The local trader took Mr. Maning to live with him, but it was
known to the tribes that the newcomer really and truly belonged to
Relation-eater. Not long had he been settled when there occurred a
meeting between his tribe and another, a game of bluff, when the
warriors of both sides danced the splendid Haka, most blood-
curdling, hair-lifting of all ceremonials. Afterward old Relation-eater
singled out the horrible savage who had begun the war-dance, and
these two tender-hearted individuals for a full half-hour, seated on
the ground hanging on each other’s necks, gave vent to a chorus of
skilfully modulated howling. “So there was peace,” and during the
ceremonies Maning came upon a circle of what seemed to be Maori
chiefs, until drawing near he found that their nodding heads had
nobody underneath. Raw heads had been stuck on slender rods,
with cross sticks to carry the robes, “Looking at the ’eds, sir?” asked
an English sailor. “’Eds was werry scarce—they had to tattoo a slave
a bit ago, and the villain ran away, tattooin’ and all!”
“What!”
“Bolted before he was fit to kill,” said the sailor, mournful to think
how dishonest people could be.
Once the head chief, having need to punish a rebellious vassal,
sent Relation-eater, who plundered and burned the offending village.
The vassal decamped with his tribe.
“Well, about three months after this, about daylight I was
aroused by a great uproar.... Out I ran at once and perceived that M
—’s premises were being sacked by the rebellious vassal who ...
was taking this means of revenging himself for the rough handling he
had received from our chief. Men were rushing in mad haste through
the smashed windows and doors, loaded with everything they could
lay hands upon.... A large canoe was floating near to the house, and
was being rapidly filled with plunder. I saw a fat old Maori woman
who was washerwoman, being dragged along the ground by a huge
fellow who was trying to tear from her grasp one of my shirts, to
which she clung with perfect desperation. I perceived at a glance
that the faithful old creature would probably save a sleeve.
“An old man-of-war’s man defending his washing, called out, ‘Hit
out, sir! ... our mob will be here in five minutes!’
“The odds were terrible, but ... I at once floored a native who was
rushing by me.... I then perceived that he was one of our own people
... so to balance things I knocked down another! and then felt myself
seized round the waist from behind.
“The old sailor was down now but fighting three men at once,
while his striped shirt and canvas trousers still hung proudly on the
fence.
“Then came our mob to the rescue and the assailants fled.
“Some time after this a little incident worth noting happened at
my friend M—’s place. Our chief had for some time back a sort of
dispute with another magnate.... The question was at last brought to
a fair hearing at my friend’s house. The arguments on both sides
were very forcible; so much so that in the course of the arbitration
our chief and thirty of his principal witnesses were shot dead in a
heap before my friend’s door, and sixty others badly wounded, and
my friend’s house and store blown up and burnt to ashes.
“My friend was, however, consoled by hundreds of friends who
came in large parties to condole with him, and who, as was quite
correct in such cases, shot and ate all his stock, sheep, pigs, ducks,
geese, fowls, etc., all in high compliment to himself; he felt proud....
He did not, however, survive these honors long.”
Mr. Maning took this poor gentleman’s place as trader, and
earnestly studied native etiquette, on which his comments are
always deliciously funny. Two young Australians were his guests
when there arrived one day a Maori desperado who wanted
blankets; and “to explain his views more clearly knocked both my
friends down, threatened to kill them both with his tomahawk, then
rushed into the bedroom, dragged out all the bedclothes, and burnt
them on the kitchen fire.”
A few weeks later, Mr. Maning being alone, and reading a year-
old Sydney paper, the desperado called. “‘Friend,’ said I; ‘my advice
to you is to be off.’
“He made no answer but a scowl of defiance. ‘I am thinking,
friend, that this is my house,’ said I, and springing upon him I placed
my foot to his shoulder, and gave him a shove which would have
sent most people heels over head.... But quick as lightning ... he
bounded from the ground, flung his mat away over his head, and
struck a furious blow at my head with his tomahawk. I caught the
tomahawk in full descent; the edge grazed my hand; but my arm,
stiffened like a bar of iron, arrested the blow. He made one furious,
but ineffectual attempt to wrest the tomahawk from my grasp; and
then we seized one another round the middle, and struggled like
maniacs in the endeavor to dash each other against the boarded