Escolar Documentos
Profissional Documentos
Cultura Documentos
N. F. Krasnov 1
Fundamentals of Theory.
Aerodynamics of an
Airfoil and a Wing
Ha anMUUCKOM Rablne
AapOAVlHOMVlKO
YaCTb I
OCHOBbl TeOpHH
AspOAHHaMHKa npO~HnH H Kp~na
j.13AaTenbCTBO
«Bblcwafl WKOna»
MOCKBa
Preface
Preface 5
Introduction 13
Chapter 1 25
Basic Information 1.1. Forces Acting on a Moving Body 25
from Aerodynamics Surface Force 25
Property of Pressures in an Ideal Fluid 26
Influence of Viscosity on the Flow
of a Fluid 28
1.2. Resultant Force Action 36
Components of Aerodynamic Forces
and Moments 36
Conversion of Aerodynamic Forces
and '\[oments from One Coordinate
System to Another 40
1.3. Determination of Aerodynamic For-
ces and Moments According to the
Known Distribution of the Pressure
and Shear Stress. Aerodynamic Coef-
ficients 41
Aerodynamic Forces and Moments and
Their Coefficients 41
1.4. Static Equilibrium and Static Stab-
ility 52
Concept of Equilibrium and Stability 52
Static Longitudinal Stability 53
Static Lateral Stability 57
1.5. Features of Gas Flow at High Speeds 58
Compressibility of a Gas 58
Heating of a Gas 59
State of Air at High Temperatures 65
References 493
Supplementary Reading 494
Name Index 495
Subject Index 497
Introduction
modern aerodynamics for two basic kinds of motion: free (in viscid)
flow and flow in a thin layer of the gas adjacent to a wall or bound-
ary-in the boundary layer, where motion is considered with account
taken of viscosity. This division of a flow is based on the hypothesis
of the absence of the reVf'rse influence of the boundary layer on the
free flow. According to this hypothesis, the parameters of inviscid
flow, i.e. on the outer surface of the boundary layer, are the same as
on a wall in the absence of this layer.
The linding of the aerodynamic parameters of craft in unsteady
motion characterized by a change in the kinematic parameters with
time is llsually a very intricate task. Simplified ways of solving this
problem are used for practical purposes. Such simplification is pos-
sible when the change occurs sufficiently slowly. This is characteristic
of many craft. ·When determining their aerodynamic characteristics,
we can proceed from the hypothesis of steadiness in accordance
with which these characteristics in unsteady motion aro assumed to
be the same as ill steady motion, and are determined by the kinematic
parameters of this motion at a given instant.
\Yhen performing aerodynamic experiments and calcul ations,
account must be taken of variOllS circumstances associated with the
physical similitude of the flow phenomena being studied. Aerodyna-
mic calculations of full-scale craft (rockets, airplanes) are based on
preliminary widespread investigations (theoretical and experimental)
of now over models. The conditions that must be obserYed in such
investigations on models are foulld in the theor~' of dynamic simili-
tude, and typical and convenient parameters determining the basic
conditions of the processes being studied are established. They are
called dimensionless numbers or similarity critNia. The modern
problems of similarity and also the theor;· of dimellsions widely
used in aerodynamics are set out in a fundamental work of academi-
cian L. Sedov titled "Similarity and Dimensional Methods in Mech-
anics" [1 J.
Aerodynamics, figuratively speaking, is a multi branch science. In
accordance with the needs of the rapidly developing aviation, rocket,
and cosmic engineering, more or less clearly expressed basic scientific
trends have taken shape in aerodynamics. They are associated with
the aerodynamic investigations of craft as a whole and their indivi-
dual structural elements, and also of the most characteristic kinds
of gas flows and processes attending the flow over a body. It is quite
natural that any classification of aerodynamics is conditional to
a certain extent because all these trends or part of them are inter-
related. Nevertheless, such a "branch" specialization of the aero-
dynamic science is of a practical interest.
The two main paths along which modern aerodynamics is deve-
loping can be determined. The first of them is what is called force
aerodynamics occupied in solving problems connected with the
2-01715
18 Introduction
------------- ---,
24 Introduction
faces can also be chosen for special purposes ensuring a definite aero-
dynamic effect. The shape of blunt bodies ensures a minimum trans-
fer of heat to the entire body. Consequently, a blunt surface can be
considered optimal from the viewpoint of heat transfer. In designing
craft, the problem appears of choosing a shape with the minimum
force action. One of these problems is associated, particularly, with
determination of the shape of a craft head ensuring the smallest drag
at a given airspeed. Problems of this kind are treated in a branch of
aerodynamics called aerodynamics of optimal shapes.
Basic Information
I
from Aerodynamics
Fig. 1.1.1
Forces acting on a surface ele-
ment of a moving body
--------- -------------
--------------------
:Fig. 1.1.2
Designa tion of parameters of ~------------------
,disturbed and undisturbed flows
fig. 1.1.3
Normal stresses acting on a face
of an elementary fluid particle
having the shape of a tetrahed- z
ron
Another force acting on the isolated fluid volume is the mass (body)
force proportional to the mass of the particle in this volume. Mass
forces include gravitational ones, and in particular the force of
gravity. Another example of these forces is the mass force of an
electromagnetic origin, known as a ponderolllotive force, that
appears in a gas if it is an electric conductor (ionized) and is in an
electromagnetic field. Here we shall not consider the motion of a
gas under the action of SHC h forces (see a special cOllrse in magneto-
gasdynamics).
In the case being considered, we shall write the projection of the
mass force onto the x-axis in the form of XPay~W. denoting by X
the projection of the mass force related to a unit of mass. \Vith ac-
count taken of these val lies for the projections of the surface and
mass forces, we obtain an equation of motion
Pay
AWdFx_X
U dt - AW'
(la" u
AS
,- Px u x - p" ~ S11 COS
(/'-
n,x)
where ~Sx and ~S1I are the areas of faces ;11 0 .1[2;1[3 and JJ 1 J1[2 J1 3,
, '-
respectively, cos (n,x), is the cosine of the angle between a normal n
to face J1f 1 Jf 2 Jl:J and the x-axis, and Px' and P1I are the pressures
acting on faces ;1r oJ1 2 JI 3 allfi JlljI~JI3' respectively.
Dividing the equation obtaillerl hy ~S~ and haYing ill view that
~SX ~= ~S" cos (,;::r). let us pass over to the limit \yiLlt ~x, ~y. and
~z tending to zero. Consequently. the terms con taining ~ W; ~Sx
will also telld to zero because L1 W is a small quantity of the third
order, while tiS x is a small qllantity of the second order in com-
parison with the linear dimensions of the surface dement. As a result,
we have p" - p" = 0, and, therefore, p" ~= Pll'
\Vhen considering the equations of motion in projections onto the
y- anrl z-axes. ~\\'e find that Py = p" anri pz = fi,,·
Since 0111' sllrface elemen t wi t h tIll' II ormal 11 is oriented arbi trarily,
we can arrive at the following condllsion from the results obtained.
The pressure at any point of a flow of an ideal t1uid is identical on
28 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
all surface elements passing through this point, i.e. it does not
depend on the orientation of these elements. Consequently, the
pressure can be treated as a scalar quantity depending only on the
coordinates of a point and the time.
Influence of Viscosity
on the Flow of a Fluid
---------------------.
30 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
(IL)
( b)
~i~
:t!Lx '('J 1~"-=': : =-/r-+v~;,
Y~tf Vx
D
/X
laminar Tvrbulent y2
l.a!Jer tayer 2
Fig. 1.1.4
Flow of a viscous fluid over a body:
a - schrmatic view of flow; I-laminar boundary layer;
2-viscous sublayrr; 3-turbulent boundary layer; 4-surfacp in the flow; 5-wake; 6-free
flow; 7-wake vortex; b--velocity profile in the boundary layrr; c-diagram defining the
concept of a two-point correlation coefficirnt; Vb is the velocity component at the outer edge
of the boundary layer
i
00
R dr L= (1.1.4)
o
The turbulence scale is a linear dimension characterizing the
length of the section of a flow on which fluid particles move "in
Ch. 1. Basic Information from Aerodynamics 311
(1.1.5)
This coefficient characterizes the statistical association between
fluctuations at a point and, as will be shown below, directly deter-
mines the shear stress in a turhllien t flow.
Turbulence will be homogeneous if its averaged characteristics
found for a point (the level and ill tensity of turbulence, the one-
point correlation coefficient) are the same for the entire flow (illvari-
ance of the characteristics of tnrblilence in transfers). Homogeneolls
t1ll'bulence is isotropic if its characteristics do not depend 011 the
direction for which they are caiclilated (inYariance of the characteris-
tics of turbulence in rotation and reflection). Particularly. the
following condition is satisl'led for an isotropic flow:
F'~
x = T7'2 =~=
Jt Y - V'2
z
(1.1. 7)
I 'tt I = p V~ (V x + V~)
Ch. 1. Basic I nformation from Aerodynamics 33
ITtl=~
t 2 -t1 J
I. V;dt+-P- \'
t - t .l 2 1
V;V~dt=pVxV~--l-pV~V~
11 11
12
------_._----------_._-_.--
34 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
l:ttl=pV~V~=-P- (dVx)2
t2 - t1 dy JI' l'2adt=Pl2( dVX)2
dy (1.1.10>
11
same order as all the other forces (for example, the forces of inertia
an d pressure) governing motion and, therefore, taken in to account
in the equations of motion.
A physical notion of the boundary layer can be obtained if we-
imagine the surface in the flow to be coated with a pigment soluble-
in the fluid. It is obvious that the pigment diffuses into the i1uid
and is simultaneously carried downstream. Consequently, the colo-
ured zone is a layer gradually thickening downstream. The coloured
region of the i1uid approximately coincides with the boundary layer.
This region leaves the surface in the form of a coloured wake (see-
Fig. 1.1.4a).
As shown by observations, for a turbulent flow, the difference of
the coloured region from the boundary layer is comparatively small,
whereas in a laminar flow this difference may be very significant.
According to theoretical and experimental investigations, with an
increase, in the velocity, the thickness of the layer diminishes, and
the wake becomes narrower.
The nature of the velocity distribution over the cross section of
a boundary layer depends on whether it is laminar or turbulent.
Owing to lateral mixing of the particles and also to their collisions,
this distribution of the velocity, more exactly of its time-averaged
value, will be appreciably more uniform in a turbulent flow than in
a laminar one (see Fig. 1.1.4). The distribution of the velocities
near the surface of a body in a flow also allows us to make the con-
clusion on the higher shear stress in a turbulent boundary layer
determined by the increased value of the velocity gradient.
Beyond the limits of the boundary layer, there is a part of the
flow where the velocity gradients and, consequently, the forces of
friction are small. This part of the flow is known as the external
free flow. In in vestigation of an external flow, the influence of the
viscous forces is disregarded. Therefore, such a flow is also considered
to be inviscid. The velocity in the boundary layer grows with an
increasing distance from the wall and asymptotically approaches
a theoretical value corresponding to the flow over the body of an
inviscid fluid, i.e. to the value of the velocity in the external flow
at the boundary of the layer.
We have already noted that in direct proximity to it a wall hinders
mixing, and, consequently, we may assume that the part of the
boundary layer adjacent to the wall is in conditions close to laminar
ones. This thin section of a quasilaminar boundary layer is called 3.
viscous suhlayer (it is also sometimes called a laminar sublayer).
Later investigations show that fluctuations are observed in the
viscous sublayer that penetrate into it from a turbulent core, but
there is no correlation between them (the correlation coefficient
R = 0). Therefore, according to formula (1.1.9), no additional shear-
stresses appear.
a·
36 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
2
.1._
Fig. U.5
Boundary layer:
J-wall of a body in the flowl x
2-outer edge of the layer..;
The main part of the boundary layer outside of the viscolls sub-
layer is called the turbulent core. The studying of the motion in
a boundary layer is associated with the simultaneous investigation
of the flow of a fluid in a turbulent core and a viscous sublayer.
The change in the velocity over the cross section of the boundary
layer is characterized by its gradually growing with the distance
from the wall and asymptotically approaching the value of the
velocity in the external flow. For practical purposes, however, it is
convenient to take the part of the boundary layer in which this
change occurs sufficiently rapidly, and the velocity at the boundary
of this layer differs only slightly from its value in the external flow.
The distance from the wall to this boundary is what is conventionally
called the thickness of the boundary layer 6 (Fig. 1.1.5). This thick-
ness is usually defined as the distance from the contour of a body to
a point in the boundary layer at which the velocity differs from its
value in the external layer by not over one per cent.
The introduction of the concept of a boundary layer made possible
effective research of the friction and heat transfer processes because
owing to the smallness of its thickness in comparison with the dimen-
sions of a body in a flow, it became possible to simplify the differen-
tial equations describing the motion of a gas in this region of a flow,
which makes their integration easier.
Fig. 1.2.t
Aerodynamic forces and moments acting on a craft in the flight path (xa, Ya,
and za) and body axis (x, Y, and z) coordinate systems
a+ a+ + +
M = Mx My M z = Mx My Mz (1.2.2)
a
We shall consider a moment about an axis to be positive if it
tends to turn the craft counterclockwise (when watching the motion
from the tip of the moment vector). In accordance with the adopted
arrangement of the coordinate axes, a positive moment in Fig. 1.2.1
increases the angle of attack, and a negative moment reduces it.
The magnitude and direction of the forces and moments at a
given airspeed and altitude depend on the orientation of the body
relative to the velocity vector V (or if inverted flow is being con-
sidered, relative to the direction of the free-stream velocity V(0).
This orientation, in turn, underlies the relevant mutual arrangement
()f the coordinate systems associated with the flow and the body.
This arrangement is determined by the angle of attack ex and the
sideslip angle P (Fig. 1.2.1). The first of them is the angle between
Ch. 1. Basic Information from Aerodynamics 39
y*
g
~~~~~:=TIC~~
x*g
Fig. t.2.2
Determining the position of a craft in space
the axis Ox and the projection of the vector V onto the plane xOy,
and the second is the angle between the vector V and the plane xOy.
The angle of attack is considered to be positive if the projection
of the air velocity onto the normal axis is negative. The sideslip
angle is positive if this projection onto the lateral axis is positive.
When studying a flight, a normal earth-fixed coordinate system is
used relative to which the position of a body moving in space is
determined. The origin of coordinates of this system (Fig. 1.2.2)
coincides with a point on the Earth's surface, for example with the
launching point. The axis Ooyg is directed upward along a local
vertical, while the axes OOXg and OOZg coincide with a horizontal
plane. The axis OOXg is usually oriented in the direction of flight,
while the direction of the axis OOZg corresponds to a right-handed
coordinate system.
If the origin of an earth-fixed system of coordinates is made to
coincide with the centre of mass of a craft, we obtain a normal earth-
fixed coordinate system also known as a local geographical coordi-
nate system Ox~y~z~ (Fig. 1.2.2). The position of a craft relative to
this coordinate system is determined by three angles: the yawing
(course) angle ~J, the pitching angle &, and the rolling (banking)
angle "(.
The angle ~ is formed by the projection of the longitudinal body
axis Ox onto the horizontal plane x~Oy~ (Ox *) and the axis Ox~;
this angle is positive if the axis Ox~ coincides with the projection
of Ox* by clockwise rotation about the axis Oy~.
The angle & is that between the axis Ox and the horizontal plane
x~Oz; and will be positive if this plane is below the longitudinal
where cos (;;a), cos (;;;:a), cos (;;a) are the cosines of the angles
between the axis OXa and the axes Ox, Oy, and Oz, respectively.
The expressions for the other components of the force vector, and
also for the components of the moment vector, are written in a
similar way. The values of the direction cosines used for converting
forces and moments from one coordinate system to another are
given in Table 1.2.1.
Table 1.2.1
Flight path system
Body axis
sy,tem QXa oVa oZa
I I
Ox cos a cos B sin a -cos a sin B
Oy -sin a cos B cos a sin a sin B
Oz sin B 0 cos B
Ch. 1. Basic Information from Aerodynamics 41
1.3. Determination
of Aerodynamic Forces
and Moments According to
the Known Distribution
of the Pressure and Shear Stress.
Aerodynamic Coefficients
Aerodynamic Forces
and Moments and Their Coefficients
Assume that for a certain angle of attack and sideslip angle, and
also for given parameters of the free stream (the speed V 00, static
pressure p 00, density p 00, and temperature Too), we know the distrib-
ution of the pressure p and shear stress 'L over the surface of the body
in the flow. We want to determine the resultant values of the aero-
dynamic forces and moments.
The isolated surface element dS of the body experiences a normal
force produced by the excess pressure (p - p 00) dS and the tangential
* We shall omit the word "components" below for brevity, but shall mean
it and use formulas for scalar quantities.
42 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
Fig. t.l.t
Action of pressure and friction
'(shear) forces on an elementary
. area SIJ
force 1: dS. The sum of the projeetions of these forces onto the x-axis
,of a wind (flight path) coordinate system is (Fig. 1.3.1)
/\ /\
[(p - poo) cos (n,x a) + 1: cos (t,xa)] dS (1.3.1)
where nand t are a normal and a tangent to the element of area,
respectively.
The other two projections onto the axes Ya and Za are obtained by
a similar formula with the corresponding cosines. To find the resul-
tant forces, we have to integrate expression (1.3.1) over the entire
surface S. Introducing into these equations the pressure copfficient
p = (p - p oo)/q and the local friction factor Cj,x = 1:/q
00 where 00,
Xa = qooSr
rI [p- cos (n,x
/\ /\ (1.3.2)
a) +Ct, x cos (t,Xa)] dS/Sr
(8)
r _ /\ /\
Ya=qooSr J [-pcos(n,Ya)+Cr,xCOS(t,Ya)]dS/Sr (1.3.3)
(8)
, - /\ /\
Za= -qooSr) [pcos(n,Za)+Ct,xCOS(t,za)]dS/Sr (1.3.4)
(8)
In the other two formulas, the corresponding symbols cYa and cZa
are introduced. The relevant quantities are known as the aerodyna-
mic lift coefficient and the aerodynamic side-force coefficient. With
,it view to the above, we have
- 1"- /"'- dS
- [p cos (n,x a) + Ct,x cos (t,xa)] Ya} Sr L ( 1.3.6)
in which L is a characteristic geometric length, we obtain a formula
for the pitching moment:
M z = m z qooS rL (1.3.7)
a a
The parameter mZa is called the aerodynamic pitching-moment
coefficient. The formulas for the other components of the moment
are written similarly:
JVf x = tn x q ooS rL and 111 y = tn y q ooS rL (U3.8)
a a a a
The dimensionless parameters tn x a and tn y a are called the aerody-
Ilamic rolling-moment and yawing-moment coefficients, respectively.
The relevant Goefficients of the aerodynamic forces and moments
can also be introduced in a body axis coordinate system. The use of
these coefficients allows the forces and moments to be written as
follows:
x = cxq ooS r, M x = mxq ooS rL
Y=cyqooSr, My=7nyqooSrL (1.3.9)
Z = czqooSr, M z = 7nzqooSrL
44 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
Fig. 1.3.2
Schematic view of a wing:
bo-centre chord. bt-tip chord.
and b-local chord x
CYan
ton-' ffmax:
/---
Fig. t.3.3
Constructing a polar of the first kind of a craft:
a-c vs. a; b-C VS. a; c-polar of first kind
xa Ya
of the forces and moments acting on the same craft. These relations
are used to find the coefficients C 2 and m 2 , respectively, converted to
the new characteristic dimensions 8 2 and L 2 :
c2 = c1 (8 1 /8 2 ), m2 = m 1 (8 1 L 1 /8 2L 2 )
where the previous dimensions 8 1 , L1 and aerodynamic coefficients
c1 ' m 1 , as well as the new dimensions 8 2 , L2 are known.
Polar of a Craft. A very important aerodynamic characteristic
that has found widespread use is what is known as the polar of
a craft. It establishes a relation between the lift and drag forces or,
which is the same, between the lift and drag coefficients in a flight
path coordinate system. This curve, called a polar of the first kind
(Fig. 1.3.3c) is the locus of the tips of the resultant aerodynamic force
vectors Ra acting on a craft at various angles of attack [or of the
vectors of the coefficient CR a of this force determined in accordance
with the relation CR a = Ra/(8 rq 00)].
A polar of the first kind is constructed with the aid of graphs of
Cx
a
versus ex and c ya versus ex so that the values of C x a and c ya are
laid off along the axes of abscissas and ordinates, respectively. The
relevant angle of attack ex, which is a parameter of the polar in the
given case, is written at each point of the curve.
A polar of the first kind is convenient for practical use because it
allows one to readily find for any angle of attack such a very import-
ant aerodynamic characteristic of a craft as its lift-to-drag ratio
(1.3.10)
If the scales of cYa (or Y a ) and cXa (or X a) are the same, the quan-
tity K equals the slope of a vector drawn from the origin of coordi-
Ch. 1. Basic Information from Aerodynamics 47·
Fig. f.l.4
Drag polar of the second kind
nates (the pole) to the point of the polar diagram corresponding to the
chosen angle of attack.
We can use a polar to determine the optimal angle of attack aopt
corresponding to the maximum lift-to-drag ratio:
K max = tan aopt (CLiO'}>
if we draw a tangent to the polar from the origin of coordinates.
The characteristic points of a polar include the point cYamax
corresponding to the maximum lift force that is achieved at the
critical angle of attack aero We can mark a point on the curve deter-
mining the minimum drag coefficient Cx a min and the corresponding
values of the angle of attack and the lift coefficient.
A polar is symmetric about the axis of abscissas if a craft has,
horizontal symmetry. For such a craft, the value of ex a min corre-
sponds to a zero lift force, cYa = O.
In addition to a polar of the first kind. a polar of the second kind;
is sometimes used. It differs in that it is plotted in a body axis
coordinate system along whose axis of abscissas the values of the
longitudinal-force coefficient C x are laid off, and along the axis of
ordinates-the normal-force coefficients cY (Fig. 1.3.4). This curve
is used, particularly, in the strength analysis of craft.
Theoretical and experimental investigations show that in the
most general case, the aerodynamic coefficients depend for a given
body confIguration and angle of attack on dimensionless variables
such as the Mach number Moo = V oofa oo and the Reynolds number
Re = V ooLp oo/ftoo. In these expressions, a is the speed of sound
00 00
in the oncoming flow, p and ft are the density and dynamic vis-
00 00
._--------_.._---_.
48 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
(a) !:I
(b)
Y(e",) !:i
Mz(mz} Y
CP eM
0'-
/' --F---- x
X(C x ) I
Xp
xc..
b
--------
xn (XFo<)
b
1-
Fig. 1.3.5
Determination of the centre of pressure (a) and aerodynamic centre (b)
velocity Moo. The curves in Figs. 1.3.3 and 1.3.4 correspond to a fixed
value of Re and determine the relation between c yB and C x a for
00
low-speed flights (of the order of 100 m/s) when the aerodynamic
coefficients do not depend on Moo.
Centre of Pressure and Aerodynamic Centre. The centre of pressure
(CP) of a craft is the point through which the resultant of the
aerodynamic forces passes. The centre of pressure is a conditional
point because actually the action of fluid results not in a concen-
trated force, but in forces distributed over the surface of the moving
body. It is customarily assumed that for symmetric bodies or ones
close to them this conditional point is on one of the following axes-
the longitudinal axis of the craft passing through the centre of mass,
the axis of symmetry of a body of revolution, or on the chord of an
airfoil.
Accordingly, the longitudinal force X is arranged along this axis,
while the centre of pressure when motion occurs in the pitching
plane is considered as the point of application of the normal force Y.
The position of this centre of pressure is usually determined by the
coordinate xp measured from the front point on the contour of the
body in a flow. If the pitching moment M z about this point and the
normal force Yare known (Fig. 1.3.5a), the coordinate of the centre
of pressure
(1.3.11)
A moment M z tending to reduce the angle of attack is considered
to be negative (Fig. 1.3.5a); hence the coordinate xp is positive.
Taking into account that
Mz = mzq ooS rb and Y = cyq ooS r
we obtain
Ch. 1. Basic Information from Aerodynamics 49
whence
xplb = cp = -mzlc y (1.3.11')
The dimensionless quantity c p defined as the ratio between the
distance to the centre of pressure and the characteristic length of a
body (in the given case the wing chord b) is called the centre-of-
pressure coefficient. \Vith small angles of attack, when the lift
and normal-force coefficients are approximately equal (c Ya ~ cy )"
we have
(1.3.12)
In the case being considered of a hvo-dimensional flow past a
body, the pitching-moment coefficients in wind (flight path) and
body axis coordinate systems are the same, i.e. mZa = m z •
For a symmetric airfoil when at a-+-O the quantities c y and m z
simultaneously take on zero values in accordance with the expres-
sions
Cy = ({Jcyloa) a, m z = (omz/oa) a
holding at small angles of attack (llOre the derivatives iJcy/oa and
omz/oa are constant quantities that can be determined for an angle
of attack of a ~ 0), the coefficient c p equals a constant value not
depending on the angle of attack:
c p = -omzloc y (1.3.13)
The value of the coefficient c p and of tIte dimensionless coordinate
of the centre of mass XC;\I -= :rC:II 1) can be used to determine the
pitching-moment coeflicienls abollt this centre:
(1.3.14)
Investigations show that in real conditions of flow, all appreciable
displacement of the centre of pressure can be observed in craft even
upon a slight change in the angles of attack. This is especially notice-
able in craft with an asymmetric confIguration or upon deflection of
an elevator, which disturbs the existing symmetry. In these con-
ditions, the centre of pressure is not convenient for use as a charac-
teristic point in estimating the position of the resultant of the aero-
dynamic forces and the appearing pi tching moment ahou t the COll tre
of mass. In these conditions, it is more cOIlvenient to assess the
fligh t properties of a craft according to the aerodynamic centre
location. To reveal the meaning of this concept, let us consider all
asymmetrie airfoil and evaluate the moment J1zn about all arbiLrary
po in t F ex ,vitll th e coordinate Xn OIl the ehord of the airfoil. I t follows
from Fig. 1.3.5b that
50 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
r-
1.2
1.0
I\. 0.8
\ 0.6
0.4
Fig. 1.3.6
\ 0.2
Relation between the moment \~
coefficient m z and normal-force -0.2 \0 0.2 mz
coefficient cy for an asymmet-
ric aircraft ·O}
or
(1.3.17)
where n1Z0 is the coeffieient of the moment about a point on the
leading edge at c y = 0 (Fig. 1.3.6).
The second term in (1.3.17) determines the increment of the
moment associated with a change in the normal-force coefficient.
If we choose the point Fa on a chord whose coordinate Xn = XF tl
is determined by the condition (see Fig. 1.3.5b)
(1.3.18)
the coefficient of the moment about this point will not depend on
cy ,and at all (small) angles of attack it will be a constant quantity.
This pcint is Cil1led the aerodynamic centre (AC) of the given body.
The aerodynamic centre is evidently the point of applieation of the
additional normal foree pl'oduced by the angle of attack [the eoef-
ficient of this force is (ocy/oa) a = c~al. The pitching moment
about an axis passing through this point does not depend on the
Ch. 1. Basic Information from Aerodynamics 51
(1.3.19)
(1.3.22)
('1 .o.~
.. ?3)
The point of application of the normal force dlle to the elevator
deflection angle alld proportional to this angle is known as the
eleyator-del'lection aerodynamic centre. The moment of the forces
about a lateral axis pas~ing through this centre is evidently inde-
pendent of the angle 0e' In the general case for an asymmetric con-
figmation, its centre of pressure coincides with nOlle of the aero-
dynamic centres (based on a or (je). In a particular case, in a sym-
metrie craft at a = 0, the centre of pressure coincides with the aero-
dynamic centre based on oe.
Using' the defmition of the aerodynamic centres based Oil the
angle of attack and the elevator deflection angle and introducing the
corresponding coordinates XF a and Xp.,,)
\ve find the coefficient of the
moment about the centre of mass. This coeffIcient is evaluaterl by
formula (1.3.22), in which
(1.3.24)
Fig. 1.4.1
Dependence of the aerodynamic
moment 111 z on the angle of
attack Ci. of a craft and the de-
flection of the elevators 6e:
1, 2, 3, 4-points of intersection of
the moment curve with the (X-axis
that determine the balance angles of
attack; 5-region of a linear change
in tile pitching moment with the
angle of atlack and the elevator
angle
In the Jirst case, free rotation of the craft continues unLil it occupies
the equilibrium position at point 1. and in the second case-at
po in t 3.
At point 2 (u 2bal ) the equilihrium is ullstable. lmleed, examin-
atioll of Fig. 1.1l.1 reyeals that at ntllles of the angle u larger or
smaller than Ci. 2bal , moments are induced, positiye or negative. re-
spectiYely, that tend to increase (or redllCe) u. lIence, these moments
are destabilizing. and the craft will be statically llnstable.
Static stability is schematically divided illto longitudinal and
lateral stability. For static longi t udinal stability, it is assumed
that all the distlll'hing forces anel moments act in the 100Igitudillal
plane of the body axes :l'Oy. Ilence, only snell moyements of a craft
are investigated that occur in its plane of symmetry in the absence of
roll and slip. When analysing static lateral stability, the distmbed
movements of a craft are considered that are associated with a change
in the rolling and sideslip angles at a constant angle of attack. Such
moyements are ah\-ays mlltually related. Deflectioll of the ailerons
causes not only roll. 1111t alfo slip. At the same time, t11rnillg of the
rlldders alED leads to rolling. Therefore in"e~tigation of latE'ral
stability is associated with an analysis of both rolling ami yawing
moments.
Static Longitudinal Stability
-,>ey
(cj /y\
(c)
Fig. 1.4.2
Action of a force and moment in analysing the static stabili ty of a craft:
II-static stability; b-static instability; c-neutrality relative to static stability
The values of c yo are usually very small even for asymmetric con-
figurations (at low 6e and a) and equal zero exactly for craft with
axial symmetry. Hence, with a sufficient degree of accuracy, we can
write
(1.4.6 »
Static Lateral Stability
direction, but like a weathercock, turns with its nose part in the
·direction of the new vleocity vector V.
Similar to the aerodynamic centre based on the angle of attack,
we can introduce the concept of the aerodynamic centre based on
the sideslip angle whose coordinate we shall designate by XFj3'
Using this concept, we can represent the degree of static directional
stability in the form
m~z = -(XFf; - XCM) (1.4.7)
where XFfj = xF(/Z and ~M = xcMIl (l is a characteristic geometric
dimension that can determine the wing span, fuselage length, etc.).
Hence, the static directional stability or instability depends on
the mutual position of the centre of mass and the aerodynamic centre.
A rear arrangement of the aerodynamic centre (XFj3 > ~M) deter-
mines static directional stability (m~l < 0), while its front arrange-
ment (XF 11 < :.reM) determines static instability (myC z > 0). When
the two centres coincide (xF = XCM), the craft is neutral as regards
tl
static directional stability (m~z = 0).
A particular case of motion of a craft in the plane of the angle of
attack can be characterized by a constant angular velocity (Q y =
= const) and stabilization with respect to rolling with the aid of
an automatic pilot (Q x ;:::::; 0). The condition for such steady motion
is lateral trim of the craft when the yavdng moment vanishes, i.e.
my = myo +mg~ +m~rl') r +
m;y Q y = 0
All craft customarily have longitudinal symmetry, therefore
myo = O. When this condition holds, the equation obtained allows
us to determine the balance rudder angle I') r = I') r.bal corresponding
to the given values of the balance sideslip angle Bba) and the angular
velocity Qy. Most craft have a sufficiently high degree of static
<lirectional stability at which the term m~yQy is negligible. Therefore
I')r.bal = (-m~!mBr ) Bbal (1.4.8)
This relation, like (1.4.4), will be accurate for conditions of rec-
tilinear motion.
Heating of a Gas
The appreciable illcrease in the i'peeds of craft ll1f\{!e it necessary
to take into account the featurps of gas :"treams rille to the change
in the ph.\'sicoc hemic al properties of t he air in aeror! .\'!lam ic in \"ef'tiga-
tions. In "conventional" sllper:..;onir aerodynamics. the compressibili-
ty was taken into cOllsideratioll as thp mo:-t important manifestation
of a featllre of flow at high f'peed:-, while the infillpllce of the tem-
perature Oil the thermodYllRmic paralllpters and kinetic coefficients
of air. and also 011 tile physicochemical procel"ses that Illay proceed
in it was disregarded. For \'Cry high (hypprsollic) speeds. howeypI',
the featmes associated with thp influencp of high temperatures come
to the forefront.
High temperatures appear owing to clpceleration of thp gas stream
whell the kinetic energ',\' of ordered motion of the particles transforms
into the internal energy of the gas.
A tat em peral me of Ihe order of 1500 K, excitation of the vibra-
tional l!'vels of till' intprnal ellNgy of the oxygPI1 and nitrogen
molecllles in the air become:- noticeable. At a tpmperatnre of about
3000 K and a pressurp of t0 5 Pa. the yibrational clegrpes of freedom
·of the oxygPll molecules are complpte!y excited. and further eleva-
tion of the temperature allows the atoms to f'llrmonnt the action
60 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
N2 + O 2 +± 2NO, N2 + 0 +± NO + N, or N + O 2 +± NO + 0
xln,'~~__r - _ _~_ __ _
1.0
~
\' - - 1 ' I !J
i
0.9 ---~~~==r~-:- -I~!-
O.B n"",- "i7
rn
I I I II
Fig. 1.5.2
Change in the specifIc heat cp of air at high temperatures
t.2
t.1 t:::::::±=::±:±.::f±:::±:J±::::::t±==iti±±J:±t.::li::=±:==:;t::=::±j
tOeD 2000 3000 4000 5000 6000 7000 8000 goOO ICC~J .<:OCO T, K
Fig. 1.5.3
Cliange in the ratio of the specifIc heats k = cplc v for air at high temperatures
WiD; PO's
28
20 -
15 1--+---+__---+---1
12
----------. __._----_._.__._---
64 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
'AxZ.5 W;'lm·l(j
-'---~~--~------~I ---:
7 r--+--_-...J-_-+_---'_~i'---
6r----+--+-~8_--~--~--~
5r--+---+-+-*-~~--~--~
4r---+---~--4L--~~~--~
Jr---+-~~~~--~~~~~
2 f----l---I--~
Fig. 1.5.5
,Change in the thermal conduc-
tivity of air at high tempera-
tures D~~~--~--~ __~__-L__~
4000 5000 80DO 10000 T,II
P'1O-~ PrL
T=10000 K
Z5"I------I----+---+~..q.~ 9000
Fig. 1.5.6
An i-S diagram for dissociating air
J-lm,kf!lmOl'iimlmlmi~i
JO t
Z5~ii
ffi • •~'~··~m -.---1-
....... p=!J7pa.
"'.~
.+
!5
'0
2000 JOOO 4000 5000 5000 7000 1, K
Fig. 1.5.7
Change in the mean molar mass of air at high temperatures
A+A (1.5.9)
5*
(a) 15 15.5
a,m/s C!)
,Ii!'"': ;.'":;,
fOOO 2000 JOOO '1000 T, K r-r-~'\,~"'" /. ~~~ % /~V.V::~ /~~~~::.::::~ ID
0
(b) ~/~~~~~%/:::~~~%~V~~~ a..
~'/ ~ ~ ~v. ~:..-: %:~~t:/~~.
'l.~ '2,35 r---
"<
a,m/s f-+-+-+-I-+-+-+-1I-+-l-+-+-+--+-+-+-I '\, OJ
'3"
2500
t--t--t--t-f-+-+--t-1f-+-I--I--I-+-+-v"'7!"~'T.A"~7f~~V..-.::
'<''iJ' " ,
;;'~V
;-:;f ~p 't\ -l-+'-+-+-+-+-+-+-+--+-+-+-i n
III
, :%:I?':;,.j . . .: '1,\).15 -, g"
",f- --~~% ~v.: ~ \~~ -'/) +"-1-+--+-+-+-+---+-+-+--+-+-+-+-+-+-+
t--t--t--t-f-+-+--t-1f--:.'\.~ I I/:~I": ~~ '"
OJ
2000
0~~~~~V~~~ I ~
f-+-+-+-I-+-+-S' ~~~ '/t%:!% ~~~ §:
~~~~I/:~r'l~~
~~~~~~0r/' 'a.."
OJ
~~~~~~~V
'"
tM
f500 ~f0~~ ~
~4-~ ~y~~V--t--t-~--t--t-f-+-f-+-f-+-I--t-1-+-+-~--t-+-+-+-f-~I-++-+-+--+-~ OJ
10
~--+-j-,fj
,qJ'B,
, .,~ I
fOOD
H
n /0 20 JO 40 ,50 50 70 Z-10- 5m2/(S2/()
Fig. 1.5.8
Speed of sound in air at high temperatures:
a-against the temperature and pressure; b-against the entha py and entropy
Ch. 1. Basic Information from Aerodynamics 69
the rates of the direct reaction rD and of the reverse one rR (the
rates of dissociation and recombination respectively) are identical.
Investigations of a dissociating flow are connected with the deter-
mination of the degree of equilibrium dissociation a. Its value for
the diatomic model of the air being considered is given in chemical
thermodynamics by the expression
(1.5.10)
where
(1.5.11)
Pd and T d are the characteristic density and temperature for dissocia-
tion, respectively, nA is the number of atoms of the element A in
a certain volume, and nA, is the number of molecules of the gas A z
in the same volume.
The characteristic temperature Td = D/!c, where D is the dissocia-
tion energy of one molecule of A2 and k is a gas constant related
to one molecule (the Boltzmann constant). Investigations show that
for the temperature interval from 1000 to 7000 K, the values of Td
and Pd can be assumed to be approximately constant and eqnal to
Td = 50000 K, Pd = 150 g/cm 3 for oxygen, and to Tr] = 113000 K,
Pd = 130 g/cm 3 for nitrogen.
To obtain an equation of state for the gas mixture appearing as
a result of dissociation of diatomic molecules, we must use the expres-
sions for determining the pressure P and molar gas constant R for
the mixture of gases and the partial pressure Pi of a component:
Lagrangian Approach
Eulerian Approach
dx ) dy dz
Tt=11(x, y, z, t; at = 12 (x, y, z, t); Tt= la (x, y, z, t)
(2.1.3)
(a) y
o,~-------------------vx
z
fig. 2.1.1
Construction of a streamline (a) and a stream tube (b):
1-streamlines; 2-contour
1/
Vyc Vye
Vxc
0
Vxc G
~A
dy
VxA
,> VxB
~
X
A dx B
H
fig. 2.2.1
Motion of a fluid particle o x
Fig. 2.2.2
Angular strain of a fluid .
particle o~------------------------~x
velocity (8V x/8x) dx. The result is a linear strain of segment AB.
The rate of this strain is ex = 8V x/fJx.
Point B moves in the direction of the y-axis at the velocity V 1r
together with point A and simultaneously moves relative to this
point at the linear velocity (fJVy/fJx) dx determined by the angular-
velocity of segment AB.
Considering point D, we can by analogy with point B determine
that the relative linear velocity of this point in the direction of the
y-axis is (8Vy/fJy) dy and, consequently, the rate of linear strain of
segment AD will be e y = fJV /fJy. The angular velocity of the seg-
ment about point A is -fJv.}fJy (the minus sign takes into account
that point D rotates about point A in a direction opposite to that of
rotation of point B). Rotation of segments AD and AB causes distor-
tion of the angle DAB (Fig. 2.2.2), i.e. angular strain of the particle
is produced. Simultaneously, bisector AM of angle DAB may turn,
the result being the appearance of an angle d~ between it and bisec-
tor AN of the distorted angle D' AB'. Hence, the particle rotates
additionally.
The angle of rotation of the bisector (Fig. 2.2.2) is
d~ = Y- 't
The quantity
(2.2.4')
is called the semi-rate of downwash of the right angle DAB.
Let us apply this reasoning to three-dimensional flow and consider
point C belonging to a particle in the form of an elementary parallel-
€piped with edge lengths of dx, dy, and dz. The velocity at this
point at the instant t = to is a function of the coordinates x dx, +
y + dy, and z + dz. Writing the velocity components in the form of
.a Taylor series in which only small terms of the first order are re-
tained, we have
V,c = Vx + (oV)ox) dx + (oVxloy) dy + (oVxloz) dZ}
VyC -;-, Vy + (oVylox) dx + (8Vyloy) dy + (oVyI8z) dz (2.2.5)
V zc = V z (oVzlox) dx + (oV zI8y) dy + (oV)oz) dz
Let us introduce a notation similar to that adopted when analysing
the motion of a two-dimensional particle. We shall assume that
{lz = oVzloz. This quantity determines the rate of linear strain of
a three-dimensional particle in the direction of the z-axis. Let us
also introduce the notation
Ul x = 0.5 (8V z/8y - 8V y/8z), Ul y = 0.5 (oV"loz - 8V)ox) (2.2.6)
The quantities Ul x and Ul y are the components of the angular veloc-
ity of a particle along the x- and y-axes, respectively. The components
of the angular velocity of a particle Ul x , Ul y , and Ul z are considered
to be positive upon rotation from the x-axis to the y-axis, from the
y-axis to the z-axis, and from the z-axis to the x-axis, respectively.
Accordingly, tho signs of the derivatives 8Vy/Dx, DVz/Dy, and DV)Dz
coincide with those of the angular velocity, while the signs of the
derivatives 8V)8y, 8Vy/8z, and 8V z/8x are opposite to those of the
angular velocity.
By analogy with (2.2.4'), we have the values
ex = 0.5 (oVz/oy + oVyfoz) , e y = 0.5 (oV)oz + 8Vzlox) (2.2.7)
that equal the semi-rate of downwash of the two right angles of
the parallelepiped in planes yOz and xOz, respectively.
By performing simple transformations, we can see that
8T'z/oy = ex + Ul x ; oVx/oz = lOy + Ul y ; oVylox = ez + Ul z
ol'yloz = ex - Ul x ; oVzlox = lOy - Ul y ; oV)oy = ez - Ul z
VyC = +++
V xc = Vx
Vy
V zc - V z
8 x dx
8y dy
8 z dz
+ lOy dz + ez dy
+ e z dx + ex dz + Ul z dx -
+ ex dy + lOy dx + Ul x dy -
+ Ul y dz - Ul z
Ul x
Ul y
dy }
dz
dx
(2.2.8)
78 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
2.3. Vortex-Free
Motion of a Fluid
---------~------.-- ..------
'80 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
°
relation characterizing the motion of a fluid at each point, let us go
over to the limit with • -+ in the last equation, which signifies
contraction of this volume to an internal point. If the condition is
observed that the moving fluid completely fills the space being stud-
Ch. 2. Kinematics of a Fluid 81
---T--------------~~~,
I
1
I
\ I
Vx ________\ __ . !
dx rd r~' '. '- \ I
I
\ \ \\ !I
Fig.2.4.f
An elementary fluid particle in \~\!\',1
a cylindrical coordinate system OL-----~------------~~__.
Fig. 2.4.2 x
An elementary fluid particle in
a spherical coordinate system
where iI' i 2, and ia are unit vectors along the relevant coordinat&
lines.
S4 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
._-------------_.-...._--------_.
86 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
We shall write the total derivative for the density dp/dt in accord-
ance with (2.4.27) in the following form:
~=~+.!!.:..~+~.~+ d1jl .~
dt at dt ar dt a8 dt a1jl
_~ V~ ~.~_ vljJ ap (2.4.35)
- at + r ar + r a8 -t r sin 8 . a1jl
IIntroducing the values of div V from (2.4.34) and the derivative
dp/dt from (2.4.35) into (2.4.1) and grouping terms, we have
~+_1_.a(pVrr2)+ 1 .a(pV e sin8)+ 1 .a(pV~,)=O
at r2 ar rsin8 a8 rsin8 aljJ
(2.4.36)
.For steady motion, the partial derivative op/ot = 0, hence
~. a(pVrr2) , _1_ . a (pVesin8) +_1_. a(pVIjJ) =0 (2.437)
r ar T sin 8 a8 sin S a1jl .
In the particular case of an incompressible fluid (p = const)
~.
r
a (Vrr2) +_1_. a (Ve sin 8)
ar sin 8 a{:J
+~S .
sin
av tI, =
aljJ
° (2.4.38)
!:I
Vy
fig. 2.4.3
To the derivation of the con
tinuity equation in curvilinea r
coord ina tes
dx dy dz
+ k (w x dy - Wy dx) = °
Ch. 2. Kinematics of a Fluid 91
r= j V·ds (2.7.1)
(K)
--_._--_..
92 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
where V· ds is the dot product of the vectors V and ds. The quantity r
determined in this way is known as the circulation of the velocity
around a closed contour. Since
V = V xi + V yj + V zk and ds = dx i + dy j + dz k
we have
r =-- ~ V x dx + V y dy + V z dz (2.7.2)
(K)
f= ) dqJ=qJA-CVB (2.7.5)
(L)
Stokes Theorem
or
dfz = (oVy/OX - oV x/oy) dx dy
According to (2.2.3), the quantity in parentheses equals the double
value of the angular velocity component 2w z • Consequently,
dfz = 2w z dx dy (2.7.6)
We can prove similarly that
df x = 2w x dy dz and df y = 2w y dx dz
In these expressions, the products of the differentials are the
areas confined by the relevant elementary contours. With a view
to the results obtained, for the area element d(J oriented in spflce
arbitrarily and confined within an elementary contour, the circula-
tion is
df n = 2w n d(J (2.7.7)
where Wn is the component of the angular velocity along the direction
of the normal n to the surface element d(J. In accordance with (2.7.7),
the velocity circulation around an elementary closed contour equals
the double strength of a vortex inside the contour.
Relation (2.7.7) can be used for a contour L of flllite dimensions
confllling a surface S at each point of which the value of Wn is known.
Here the velocity circulation around the conionr is
x = 2 \ \ «) Ii da (:2 .7 .g' )
«8)
f( (r;,)
Fig. 2.7.1
Simply and triply connected regions on a plane:
a-simply connected region (r X = circulation over the contour X); b-triply connected re-
gion (r[( = circulation of the external contour, rX , and rx.= circulations over the in-
ternal contours X, and X .. respectively)
whence the circulation over the contour K of the region being consid-
ered is
1'K = 1'Kl + 1'X2 + 2 ~ ~ (Un do
(8)
Using this formula when we have n internal contours, we find
i=n
l'K = ~ l'K i + 2 ~ i
(8)
(Un do
Vortex-Induced Velocities
Appearing vortices produce additional velocities in the fluid-
filled space surrounding them. This effect is similar to the electro-
magnetic influence of a conductor carrying an electric current. In
Ch. 2. Kinematics of a Fluid 95
Fig. 2.7.2
Influence of a vortex
i
0:2
._-_._--_._--
96 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
Fig. 2.7.3
Influence of a line vortex on
a fluid particle at point A
Fig. 2.7.4
Interaction of vortices: h
-a-vortices with an identical direc-
tion of rotation; b-vortices with
opposite directions of motion
For a vortex, one end of which extends to infinity, and the other
has its origin at point A (a semi-infinite vortex), we have a l = 0
and a 2 = n/2. Consequently
w = r/(4nh) (2.7.13)
If a fluid accommodates two or more vortices, they interact with
one another, and as a result the vortex system is in motion. The veloc-
ity of this motion is determined with the aid of the Biot-Savart
relation. Let us take as an example two infmite vortices with the
same strength and direction of rotation (Fig. 2.7 .4a). These vortices
impart to each other the velocities V 2 = -r/(2nh) and VI =
= r/(2nh) that are equal in magnitude and opposite in direction.
As a result, both vortices rotate about an axis passing through the
middle of the distance between them. If of two vortices one has
a strength of the opposite sign (Fig. 2.7 .4b), the induced velocities
are of the same direction and, consequently, the system of vortices
moves translation ally at the velocity V = r/(2nh) in a direction
perpendicular to the straight line connecting the vortices.
we obtain
dWlda = Ve- i6 (2.8.4)
7-01715
98 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
Parallel Flow
x
Fig. 2.9.1
General view of a forward
plane-parallel flow
y
Fig. 2.9.2
A two-dimensional point source
Doublet
~8
ret = r (cos tl~··
I SIn
6) =.!-.(cosO-isinO)
r
Ch. 2. Kinematics of a Fluid 101
(a)
y (6)
y M(X,y)
x
Fig. 2.9.3
To the definition of a doublet:
a-doublet streamlines; b-formation of a doublet
we obtain
cP + itjJ = (M 12nr) (cos a- i sin a)
Hence
cP = (MI2n) cos aIr (2.9.16)
tjJ = - (M 12n) sin SIr (2.9.17)
Assuming that tjJ = const and having in view that r = x2 yZ V +
and sin a = ylr = ylV x + y2, we obtain an equation for a family
2
then (J2ffJ/f)x2 +
f)2ffJ/f)y~ identically equals ·zero. Consequently, the
resultant function ffJ satisfies the contitIUity equation. The resultant
potential from the source and the. sink is
. ffJ = (q/2n) In (rIlr?)
Assume that the source and sink approach each other (e --+ 0)
and simultaneously their strengths increase so that the product
q2e at the limit upon coincidence of the source and sink tends to
a finite value M. The complicated flow formed is called a doublet,
the quantity M characterizing this flow is called the moment of the
doublet and the x-axis-the axis of the doublet. Going over to the
limit in (2.9.19) for ffJ at e --+ 0 and 2ge --+ M, we obtain the follow-
ing expression for a doublet
M x M cose
ffJ = 2n . x 2+y2 = 2n . - r -
ffJ = 4n
q(1R2 -
1) = 4nq[ -l~/;-::(x=-=e:::::;)2:;=+=r:;:-2
Rl
1
where r2 = y2 + Z2.
Ch. 2. Kinematics of a Fluid 103
!J
Fig. 2.9.4
A circulation flow (point vor-
tex)
3. t. Equations of Motion
of a Viscous Fluid
whence
(3.1.1)
where dV x1dt is the total acceleration in the direction of the x-axis.
We obtain similar equations in projections onto the y- and z-axes:
dVyldt = Y + (1/p) P y (3.1.2)
dVzldt = Z + (lIp) P z (3.1.3)
Ch. 3. Fundamentals of Fluid Dynamics 107
0r---~ ____+-__~______________~
X x
z z
Fig. 3.1.1
Surface forces acting on a fluid particle
Account must be taken, ho.weverj ·of the shear stresses .zx and .yx.
The rear face experiences the force: -.zx dx dy, and the front one-
the force [.zx + (o.zx/{)z) dzl dx dy. The resultant of these forces is
(o.zx/oz) dx dy dz. In a similar way, we find the resultant of the
forces acting on the bottom and top faces. It is (o.yx/oy) dx dy dz.
Hence, the projection onto the x-axis of the surface force related
to unit volume is
(3.1.4)
Similarly, the projections onto the other coordinate axes of the
surface force related to. unit volume are
P y = O.XII/OX + OPyy/oy + o.Zy/oz }
(3.1.4')
+
P z = o.xz/ ox + o.yz/oy oPzzloz
According to the property of reciprocity of shear stresses, the
values of these stresses acting along orthogonal faces equal each
other, i. e . • zx = .xz, .zy = .yz, and .yx = .xy·
Hence, of six shear stresses, three are independent.
To determine the values of the shear stresses, we shall use the
hypothesis that stresses are proportional to the strains they produce.
The application of this hypothesis is illustrated by Newton's formula
for the shear stress appearing in the motion of a viscous fluid relative
to a solid wall. By this formula, .yx = j.t (oVy/ox) , i.e. the stress
is proportional to the half-speed 8 z = 0.5 (oVylox) of angle distor-
tion in the direction of the z-axis, whence .yx = 2j.t8 z (here j.t is
the dynamic viscosi ty). This relation covers the general case of three-
dimensional motion when the angular deformation in the direction
of the z-axis is determined by the half-speed of distortion 8 z =
= 0.5 (oVy/ox + oVx/oy). We shall write the other two values of
the shear stress in the form .yz = 2j.t8 x and .zx = 2j.t8 y.
Consequently,
.yx = 2j.t8z = j.t (oVy/ox -I- oV x/Oy)}
.YZ = 2~L8x -j.t (oVy/oz+ oVz/oy) (3.1.5)
.zx = 2j.t8y =.= j.t (oV)ox + oV x/ oz )
We shall use the above-mentioned proportionality hypothesis to
establish relations for the normal stresses Pxx, pYY' and Pzz. Under the
action of the stress Pxx, a fluid particle experiences linear strain or
deformation in the direction of the x-axis. If the relative linear de-
formation is e~, then Pxx = EfY", where E is a proportionality factor
or the modulus of longitudinal elasticity of the fluid. The normal
stresses Pyy and pzz cause the particle to deform also in the directions
of the y- and z-axes, which diminishes the deformation in the direc-
tion of the x-axis.
Ch. 3. Fundamentals of Fluid Dynamics 109
+~[ z
az !-l (av az )J=-~+
ax + avx ax!-l ~V x +J::.~divV
3 ax
1
y
all (2 OV x -.2.divV)+~( aV x + ov )-L~( avx +avz)
-r ax ax 3 ay ay ax az az ax 1
+ ~ ( av x + of y
oyay ax
) + ~ ( oV x + oY z )
az az ox
J
avy 1 oVy avy aFlJ 1 op
---rV
at - + V y -oy+ Vz -az= y - _
x ox p .Oy
-
+v~V y +2...~
3 ay div V +-.!...
p
rL~
ay ,(2 ov
y
ay _2
3 div v) (3.1.17)
+~
az
( avazy + av Z
oy
) -L ~
I dx
( avaxy + avoyx ) J
112 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
iW z + V iW z + V oV z + V iW , =Z--.!.. Op
ot x iJx Y oy z OZ p oz
+vAVZ +~.~divV+~[~(2
3 iJz P . OZ
o~z
dz
-';'divV)
::I
+~ ( oV z + oV x ) + oil ( iW z + oV
y
)]
ox ox OZ iJy dy iJz J
where v = /1lp is the kinematic viscosity.
The differential equations (3.1.17) form the theoretical foundation
{)f the gas dynamics of a viscous compressible fluid and are known
as the Navier-Stokes p-quations. It is assumed in the equations that
the dynamic viscosity /1 is a function of the coordinates x, y, and
z, i.e. /1 = f (x, y, z). Presuming that /1 = const, the Navier-
'stokes equations acquire the following form:
dV x =X-!...!.E..+vAV +~o~divV )
dt p iJx x 3 ax
dV!I
at = Y 1 0p
-p'ay va. V
+ vAVy +-g-.a;dlV JI
(3.1.18)
dV z 1 op
Z --'-+VLl AV' v a- d'IV V
--=
dt P iJz
-r13
- 'iJz Z
When studying gas flows, the mass forces may be left out of ac-
-count, and, therefore, we assume that X = Y = Z = O. In this
·case, we have
dV x = _!...!L+ vAV +~. ~div V)
dt P iJx x 3 ax
- = - - · p1- +op
-dVy
dt
v 0
v A VY +-.-divV
iJy 3 iJy
t (3.1.19)
iW x
iJt
=-~.~+1-'~[/1(2
p iJx p
2
ax
a
iJV x
iJx
-2..diVV)]
3 I
I
+p0ay- (fJ£z)
av y
--=--._+_.-
iJt
1 ap
p iJy
1 iJ
P iJy _
/1 r( av 2
2----divV
iJy
y
)]
3
I~
(3.1.20)
2 a
+ p. ax (/1ez) J
Ch. ~. Fundamentals of Fluid Dynamics 113
Vector Form
of the Equations of Motion
Curvilinear Coordinates
The first vector on the right-hand side of (3.1.33) has been deter-
mined in the form of (3.1.25), while for calculating the second one
Eq. (3.1.28) should be used. Taking the curl of both sides of this
equation, we obtain
3
curl curl V = 2J (curl curl V)nin
n=l
3
= h _1_ { 0 [hJ (curl V)jJ o [h m (~url V)mJ} in (3.1.34)
hmhj oqm qJ
n=l
Cylindrical Coordinates
In accordance with (2.4.12), (2.4.16), and (2.4.25), we have
ql = X, q2 = r, qa = y, hl = 1, h2 = 1, ha = r,
VI = V x , V 2 = Vr. Va = V,\,
Consequently,
W1= OV x ' +V
at . x
oV x
,ox + V r 'OVarx +..!2..
r
av x
ay'• {grad P)l = op/OX
Next we find
(grad div Vh = 0 div V/ox
Since div V is determined by formula (2.4.26), we have
odivV =~
ox ox
(OV x
ox
+ aV0 rr +..!...
r
OV,\,
ay
+ Vr)
r
(cur1V)3= oV
axr oV x
--or (cur1 V) 2 =1r- (av
--
oy
x - roV,\,
- )
ax
Introducing these expressions into (3.1.34), we obtain
ax 2 ar 2 r2 ay2 r or
vVe thus compile an equation of motion in a projection onto the
x-axis of a cylindrical coordinate system:
iJV x +V aV x LV aV x + .£2... iJV x = __1_ • ~
at x ax ' 'a r r ay fl ox
v a div V (3.1.35)
3' ax
We obtain the other two equations in projections onto the coor-
dinate lines r and 'V in a similar way:
1
iJV r+
at v x iWr+v
ax r
avr+v,\,. aV,_ V~
or r a,' r I
= _..!.. ~+v (~V _2-.iJV'\'_.!::2:.)+~. adivV
p or r2 ay r2 3 or l
I
T
r (3.1.35')
ax + V aV,\,
at + V x ov"
av"
or
+.£2...
r T
OF,\, + VrV'\'
ay r
1 ap ( 2 aV r v'\') vadivV
= -pr'ay-+v ~V'\'+~'ay--;:2 + Sr' ay J
In these equations, we have introduced a symbol for the Laplacian
operator in cylindrical coordinates:
02 02 1 02 1 a
~ = ax2+ ar2 + ~. ay2 + -r-' 7ii
We determine the divergence of the velocity by formula (2.4.26),
For an axisymmetric flow, the equations of motion are simplified:
iJV x
ot
+V
x
iJV x
ax
+V
T
aV x = _ ~ . 2
+~. adivV
or p ax
+ v ~V x 1 I
ax
I
3 l
r (3.1.36)
iJV = -~.~+v~V
ot +V x r'lFr+v
r
aV T
ax T or p or T
+~. adivV
3 or
J
where div V = aVxlox + aVrlor + VTlr; ~ = a2 1ax2 + a 1ar +
2 2
+ (1/r) alar.
For a steady flow, one must assume in the equations that
aVxlat = aVrlat = aV,\,lat = 0
118 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
Spherical Coordinates
(curl
+ 1
. a- Vr
sin 2 (} 01jJ2
r2
2
(3.1.42)
Ch. 3. Fundamentals of Fluid Dynamics 119
i1 = ~
r" . -:-
(lr
(r2 -:-) + r-sln.1 U . :8
ur
.>
u
(sin 8 :8 )+
2 .1
urSI n 28' "~1:2 (3.1.44)
u'!'
+ r r2 a8 r2 r2 3 ar l
1_.!!.L I
( (3.1.45)
x aVe+V DVs+Ve.oVs+VrVe= __
at' r Dr r aB r rp 08
2 D1' r Ve) v DdivV
+\1 ( i1Ve +72'ae-r2sin28 +y,:-' 2B J
120 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
where
A 10(oa)
ar r~ ar + r2 sin1 8 . aBa('8a)
u = "';2 • sm as , (3.1.46)
divV=....!.-. a (Vrr2) + 1 a(Ve sin8 ) (3.1.47)
r2 iJ r r sin tl a8
+ 3
~. adivV
ax I
~ (3.1.59}
iJVy --LV av +V iJV
y II = __l_.~+v(~V),
at ox y ay oy'
I x
+
~.advV
3 oy
p 2
I
J
where (~V)l and (~V)z are determined by formulas (3.1.54) and
(3.1.58), respectively.
We have thus obtained various forms of the equations of motion
for a viscous liquid. Experience shows that this is needed because
in some cases, when studying the laws of interaction of gas flows
with bodies in them, it is convenient to use one form of the equa-
tion, and in others, a different form.
Diffusion Equation
r
v,.
Fig. 3.2.1
An elementary gas particle in
an axisymmetric three-dimen-
sional flow
Energy Equation
(3.2.6")
Hence,
_ __' _x + --.!...ll.
iJqd' iJqd iJ q
+~ ) 't dt = _ ( " i oQi, d. x
( iJx iJy iJz L.J i ox
i
+ L.J
'V i.I iJQi. d.1I
iJy
+ "L.J i.I iJQi. d.
oz
~ i.I divQ.I, d ) 'tdt
z ) 'tdt= (_ L.J
i i
+ ~ iidiv(pDgradci)+e (3.2.10)
where
V = Vxi + Vyj, div V = iJVx/iJx + iJVyliJy,
grad T = (iJT/iJx) i + (iJT/iJy)j, grad C; = (iJcu'iJx)i + (iJc)iJy)j
T
I a [ I-"
V xa; (2 ~-3
av." 2 d' V) ] 'V yIV T
a [ /1
ay (2 ---ay--:;;
avl,l 2 diVV)),
ax I fl
vx ~ L
(2 aVx_.2.div
ax 3
V)J +V y~
ay [fl (2 av,/_2.
ay 3
dlv V)J
=~(
ax ,fl av~)+~(
ax ay fl aV~)_2
ay fl
[(aVx)2+(~)2J
ax ay
- --} div (flV div V) +-} fl (div V)2
p~=
dt
dp
dt
+2fl{[(aV x
ox )2 + (!8V
1I
ay )2J_2.
3
(divV)2+4e~}
(3.2.14)
p ~=~+2 av
ox )2+( oV"
x
dt dt fl {[( oy )2J_2.(d·
3 IV
V)2+42}
8,
p = RpT, A pk- 1 = RT J
dV z = - _1
dt
Ro
P
.!..E...
oz' ie..
dt
di
+ p di v V = 0
dp
I (3.3.4)
2 i)
+ P dY (!tf:z)
= - _1_.~+_1_
p ay p
. ..!. [ft
oy
(2 aV'1 -- ~ div
Dy:3
V)]
2 a
+p- . ax (ftE z ) (3.3.10)
~~ + p div V = 0; -pT
p =Ro
Itm
+2!t {[(avx)? -L(aV )2J_2.(d· V)Ll_4 z
p .!!i=!:J!.
y 2
}
dt dt ax 'Dy :3 IV ,E
The last two quantities are not contained explicitly in Eqs. (3.3.10),
but they are nevertheless used in solving them because when stu-
dying the flow of a gas its thermodynamic characteristics are deter-
mined. Since the energy equation also takes into account heat
transfer by diffusion, equation (3.2.5) has to be included addition-
ally. It must be taken into account simultaneously that the con-
centration Ci in the energy and diffusion equations is a function
of the pressure and temperature, and it can be written in the form
of the general relation
I t is general knowledge that the vector grad F [here F (ql' q2, q3) =
= 0 is the equation of the sllfface in the flow, and qI' Q2, q3 are
the generalizerl curvilinear coordinates] coincides in direction with
a normal to the surface. Hence, for conditions of flow without sep·
aration, the dot product of this yector and the velocity vector V
is zero.
Consequently, the condition of Dow without separation can be
written in a mathematical form as follows:
V grad F = 0
Taking into account that
1 of. . 1 of. 1 of.
gra d F = - ' - - 11 + - , - 1 , + - . - - 1 (3.3.15)
hI oqI 'h2 oq2 ~ h3 oq2 3
the condition of flow without separation can be written as
_1_. aF V
hI aqi 1 +-
1 . of V2+~.~V3=O
h2 Oq2 h3 oQ3
(3.3.16)
For Cartesian coordinates, we have
grad F = (oF/ox) i 1 + (rJF/oy) i2 + (oF/oz) i3
Consequently, f6lr (3.3.16)
Vx oF/ox + Vy oF/oy + V z of/oz = 0
For two-dimensional plane flow
VII of/ax
V; - aF/ay (3.3.17')
grac1 F o f . , of.
= - 1 1 - - 12
+ 1
-'-,-13
aF.
ax 'Dr r (ii
therefore, the condi tion of flow wi th011 t separation has the form
~
D.<
V x +~
ar
V r ,+-_1
r
.~
oy
v-/=O (3.3.18)
a viscous gas differing from the conditions for an ideal fluid. Particu-
larly, when studying the flow of a viscous gas in a boundary layer,
the solutions of the pertinent equations must satisfy the conditions
on the surface of the body and at the edge of the boundary layer.
According to experimental data, the gas particles adhere, as it
were, to the surface, and therefore the velocity on it is zero. At
the boundary layer edge, the velocity becomes the same as in free
(inviscid) flow, and the shear stress equals zero.
Going over from the relation for the gradients to one between
the corresponding scalar functions. we find
{}rpiOt + V /2 + P + u =
2
C (t) (3.4.4)
where
(3.4.5)
Expression (3.4.4) is known as the Lagrange equation or integral.
The right-hand side of (3.4.4) is a function that depends on the time,
but does not depend on the coordinates, i.e. is identical for any point
of a potential flow. The terms on the left-hand side of (3.4.4) have
a simple physical meaning: V 2 /2 is the kinetic energy, P = dplp f
is the potential energy due to the pressure for a unit mass, and U
is the potential energy due to the position of the fluid particles and
related to their mass.
To reveal the physical meaning of the first term, let us use the
expression for the potential function {}rpl{}s = Vs, where Vs is the
projection of the velocity vector onto a certain direction s. We can
s
determine the function rp from the condition (p = f Vs ds (where
So
So and s are the coordinates of a fIxed and an arbitrary point, respect-
s
of the body in the flow and known values of the angle of attack and
the sideslip angle, as well as of the rudder and elevator angles, will
be functions of the free stream velocity V 00, the pressure P 00, the
density p 00, the dynamic viscosity Il 00, the specific heats Cp 00 and
Cv 00 of the gas, as well as of a certain characteristic (reference) linear
dimension of the body L. Consequently, the drag coefficient will
also depend on these parameters, and we can compile a functional
relation for it in the form Cx = f (V 00, P 00, P 00, Il 00, c]l 00, C" 00, L).
Since this coefficient is a dimensionless quantity, it must also be
a function of dimensionless parameters. From the general consider-
ations of the dimensional method, it follows that the seven different
arguments of the function C x can be reduced to three. The latter are
dimensionless combinations compiled from l' 00, P 00, P 00, f! 00, c]l 00,
C,' 00, and L because there are four independent units of measurement,
namely, mass, length, time, and temperature. These dimensionless
combinations have the following form: V ooiVkoopoo/poo = V oo/a oo =
= Moo-the Mach number for an undisturbed flow; V oop ooL/1l 00 =
= Re 0 0 - the Reynolds number based on the parameters of an
undisturbed flow and the characteristic linear dimension L; cp no/cvo<>
= k oo-the adiabatic exponent.
In the expression for Moo, it is assumed that Vkoopoo/poo = a oo
is the speed of sound in the undisturbed flow. Indeed, in accordance
with the general expression for the speed of sound a 2 = dPidp, and
also with a view to the adiabatic nature of propagation of sonic-
disturbances in a gas, according to which p = Apk, we have a 2 =
= d (Apk)/dp = kp/p. Accordingly, the square of the speed of
sound in an undisturbed flow is a;' = k oop oo/p 00' Hence, the ratio
V oo/Vk oop oo/p 00 = V ooia 00' All other dimensionless combinations
except for M "'" Re 00, and k 00 formed from the seven parameters.
indicated above or in general from any quantities that can be deter-
mined by them are functions of the combinations Moo, Re oo , and
k 00' Consequently, the drag coefficient is
Cx = f (M 00, Re oo , k oo ) (3.5.3)
Similar expressions can be obtained for the other aerodynamic
coefficients. It follows from these expressions that when the num-
bers Moo, Re 00 and the parameter k 00 for a model and a full-scale
flows are equal, the aerodynamic coefficients for geometrically
similar bodies are the same. Hence, an important conclusion can
be made in the dimensional analysis and similarity method in
accordance with which the necessary and sufficient condition for
aerodynamic similarity is the constancy of the numerical values
of the dimensionless combinations forming what is called a base,
i.e. a system of dimensionless quantities determining all the other
parameters of a flow. These dimensionless combinations are called
similarity criteria.
Ch. 3. Fundamentals of Fluid Dynamics 141
Voo.aV x + v!(v.a~x+v
too at L x ax Y
av!
ay
+v Z
a~x)== X
az g
flow (k and Too are the adiabatic exponent and the temperature of
00
+ D~
Dz
( DV!
Dz
+ D~z
Dx
)] } (3.5.7)
- - ( DT - DT - DT - DT)
pcp Sh-i-_-+l'x-:::--+Vy-_ +l'z--
Dt Dx oy Dz
10-01715
146 Pt. I. Theory. Aerodynamics 01 an Airfoil and a Wing
and the ratio of the specific heats are due to definite requirements
to the physical properties of the gases in the full-scale and model
flows. The gases may differ, but their physical properties must
observe the equalities Pr 1 = Pr 2 and k 001 = k 002' The Prandtl
number depends on the dynamic viscosity and the heat conductiv-
ity. The dynamic viscosity reflects the properties of a gas which
the molecular transfer of the momentum depends on, while the
heaL conductivity characterizes the intensity of the molecular
transfer of heat. Consequently, the Prandtl number Pr = ~lCpoo.Aoo
is the measure of the transformation of the energy of molecular
transfer into heat. For a gas, Pr < 1.
A dimensionless variable of the aerodynamic force or heat trans-
fer is a composite function of a number of similarity criteria, each
of which reflects the influence of a definite physical process. Complete
similarity of a fltll-scale and a model flows can be ensured only
when equality of all tlte similarity criteria is observed. In practice,
this cannot be done because some of these criteria are contradictorv.
Let us consider, for example, the Reynolds, Froude, and Ma~h
numbers. For the observance of similarity with respect to the skin
friction forces, it is essential that l\Ll/V 1 = 1/2L2'\'2' If ,ye assume
that for a full-scale and model flows the coefficients VI = "2' then
the speed of the model flow 1/ 2 == l\ (L 1/L2) , i.e. it is greater than
the speed of the full-scale flow the same number of times that the
model of the body in the flow is less than the full-scale one.
To ensure similarity with respect to the forces of gravity, it is
necessary to ensure equality of the Froude numbers, i.e. V~/(Llgl) =
= V; (L 2 g 2 ), whence it follows that if experiments were 1'11l1 at
identical values of g, then the speed of the model flow is 1/ 2 =
= 1/1 V L 2 /Ll' \Ve can see that in the given case the speed V 2 for
the small-size model mnst be smaller than VI instead of greater as
in the previous example.
Upon equality of the Mach numbers, we have V 1 1a 1 = V 2 /a 2 •
Assuming for simplification that a 2 = aI' we obtain the condition
for eqnality of the speeds of the model and full-scale flows.
I t is natural that all these conditions for the speed cannot be
observed simultaneously, therefore we can consider only incomplete
similarity. vVe must note, howe\'er, that in practice there is no
need to satisfy all the similarity criteria because their influence in
a specific case of motion is not the same. For example, the forces
of skin friction and pressure have a more signiflcant influence on
the flow of a gas over a body than the forces of gravity, and accord-
ingly the numbers Re and M are more signifIcant than Fr. In this
connection, the Froude number is not taken into consideration as
a similarity criterion in such cases.
10*
148 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
If at the same time the speeds are not great, then the influence
of the pressure forces due to compressibility of the gas is negligibly
small, and, consequently, no account may be taken of the similarity
criterion with respect to the Mach number, assuming that the aero-
dynamic coefficient depends on the Reynolds number.
The aerodynamic force, moment, or heat flux from a gas to a sur-
face is the result of the action of a moving gas on a body. Various
processes occur simultaneously in the gas: skin friction, compression
(or expansion), heating, a change in the physical properties, etc.
Therefore, one must try to satisfy the maximum number of simi-
larity criteria. For example, it is expedient that the equality of the
Reynolds and Mach numbers for a full-scale and model flows be
retained simultaneously, i.e. ReI = Re 2 and MI = M 2' This is
especially important when studying aerodynamic forces, which for
bodies with a large surface may consist of equivalent components
depending on the friction and pressure due to compressibility.
This condition can be ensured when running experiments in variable
density wind tunnels.
If tests are being performed in a gas flow in which the speed of
sound is the same as in the full-scale flow (a 2 = al ), it follows from
the equality of the Mach numbers that V 2 = VI' Having this in
view and using the equality ReI = Re 2 or, which is the same,
V2P2L2/!t2 = VIPILI/!tl> we obtain L 2P2/!t2 = LIPI/!tl' Assuming
that !t2 = ~l' we find that the density of the gas in the wind tunnel
flow is P2 = PI(LIIL 2). Assuming that the temperature of the full-
scale and model flows is the same (T 2 = T 1 ) and using an equation
of state, we obtain the condition P2 = PI (L11L2)' Hence, to simul-
taneously ensure similarity with respect to the forces of skin fric-
tion and of pressure with account taken of compressibility, i.e. to
observe the equalities ReI = Re 2 and M1 = M 2' it is essential
that the static pressure in the flow of a gas produced by a wind
tunnel be greater than the pressure in the full-scale flow by the
same number of times by which the model is smaller than the fu11-
scale body. The design of a wind tunnel makes it possible within
known limits to control the static pressure in the model flow of a gas
depending on the size of the model.
With a known approximation, the influence of heat transfer
may not be taken into account when determining the force interac-
tion. Here the aerodynamic coefficients will depend on the numbers
Re, 1Jl, and Sit. If in addition the tests are conducted in a gas for
which k 002 = k 001' we have
Cx = f (Re, M, Sit) (3.5.23)
For a steady flow
Cx = f (Re, M) (3.5.24)
Ch. 3. Fundamentals of Fluid Dynamics 149
----------------------
160 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
V,p,i
v~ p, T, a
P~--+----------I-'---~
p~
T_
Fig. 3.6.1
Parameters of a gas flowing over a body
flow with a diminishing velocity (M < 1, dV < 0), the cross section
increases, which is indicated by the inequality dS >
from (3.6.4).
following °
Let us consider a supersonic flow (M> 1). If the velocity decreases,
then, as can be seen from (3.6.4), the differential dS < 0, and,
consequently, the jet converges. Conversely, when the velocity
grows, the value of dS > 0, i.e. the jet diverges.
Let us take a nozzle that first has the shape of a converging, and
then of a diverging channel. In definite conditions in the converging
part of the nozzle, a subsonic flow is accelerated, reaching the speed
of sound in the narrowest cross section [here dS = and, as follows
from (3.6.4), M = 1], and then becomes supersonic. This is how
°
nozzles are designed in rocket engines, gas turbines, and wind
tunnels intended for obtaining supersonic flows.
Flow Velocity
Let us consider a gas jet flowing over a surface (Fig. 3.6.1). We
shall denote the free-stream parameters by V 00, p 00, p 00, Too, i 00,
and a 00, and the parameters for the part of the jet in the disturbed
region by the same symbols without subscripts. To find the velocity
in an arbitrary cross section of the jet, we shall use Eq. (3.4.14) in
which we shall determine the constant C according to the preset
parameters of the free stream:
C = V;"/2 ioo + (3.6.5)
With this in view, we have
V 2/2 +i= V;,/2 + icc
whence
V = v V;" + 2 (ioo - i) (3.6.6)
At the stagnation point, V = 0, consequently the enthalpy is
i=i o =V;"/2+i oo (3.6.7)
Ch. 3. Fundamentals of Fluid Dynamics 151
k-1 L Po
r
V = ao l/-2- 1- (L)(k-1)/k]
_
(3.6.13)
consequently,
P = Po ( 1 +
k
-;- 1 ilF
)-k/(k-l)
= Port (M) (3.6.28)
Consequently,
_ {1,[(k-l)/2 j Jl;'\h/(k-l)_ J1(.l1)
p-poo 1,[(k--l):2jJI2 J -poo n(JIoo) (3.6.30),
and also Eqs. (3.6.26) and (3.6.28) for the pressure and (3.6.31).
for the density, we have
Po
=
poo
(1+ k - 1 M2)1/(k-l)=~
2 00 e (JIoo)
(':><::31.)
.J.C/. 'f.
T 0= T 00 (1 . ~
k- 1
' lu
,-
U2
oo
)
=
Too
T (JIx) (3.6.35)
154 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
(3.6.37)
(3.6.38)
where rc (1), e (1), and 't (1) are the values of the gas-dynamic func-
tions at M = 1.
The above formulas, suitable for any speeds, can give us approxi-
mate relations for cases when the numbers M are very large.
A glance at (3.6.30) reveals that when M ~ 1 and M ~ 1, we 00
have
(3.6.39)
Similar relations for the density and temperature have the form
pip 00 = (M 00IM)2/(k-l) (3.6.40)
TIT 00 = (M 001M)2 (3.6.41)
Using relation (3.6.27) for the velocity of the free-stream flow,
Y;" k-1 -1
1-~= (1+-2-Moo)
2
(3.6.27')
max
S Se
p ~ =p~ Hllid
~P_--,!i =p_
Fig. 3.6.2
Parameters of a gas flowing from a reservoir
to, the Mach number, as can be seen from (3.G.4), equals unity,
i.e. the velocity in this section equals the local speed of sound:
P
=
Po
(1 _ ~)l/(k-l)=
V2 Po
(1- k-1 .~
k ''1a*2
)l/(k-l)
max' T
___- Po (1 - k-1 ~2)1/(k-l)
k+1 '"
Ch. 3. Fundamentals of Fluid Dynamics 167
Fig. 3.6.3
Change in the mass velocity
ratio when a gas flows out from
a reservoir
(e)
( a.)
AIC".:JO:>i
Normal
sllOek
Fig. 4.1.1
Shocks:
a-attached curved shock; b-detached curved shock: c-attached straight shock
-
V=D
-V",a
'fig. 4.t.l
Propagation of disturbances in a gas:
a-gas at rest; b-subsonlc floWI c-supersonic flow
V(M)
Fig. 4.1.3 • A
Origination of a shock
We shall consider the more general case when the gas behind
a shock, owing to substantial heating, experiences physicochemical
transformations and changes its specific heat. Of major significance
when studying shocks behind which oscillations are generated and
dissociation, ionization, and chemical reactions occur are the rates
of the physicochemical transformations.
Processes behind shock waves are characterized by a fraction of
the kinetic energy of the moving gas virtually instantaneously
transforming into the internal energy of the gas. In these conditions,
we cannot ignore the fact that thermodynamic equilibrium is ach-
ieved after a certain time elapses only in conditions of such equi-
librium do all the parameters experiencing discontinuities (the preSSlll'e,
density, temperature) become time-independent. The analysis of
these phenomena is a more involved problem and is assoeiated
primarily with studying of the mechanism of non-equilibrium pro-
cesses, and with a knowledge, particularly, of the rates of chemical
reactions in the air.
The simplest case is characterized by an infinitely high rate of
the physicochemical transformations and, consequently, by the
instantaneous setting in of thermodynamic equilibrium. Such pro-
cesses behind shock waves are possible physically, which is confirmed
by experimental studies.
Let us consider the basic theoretical relations allowing one to
evaluate the equilibrium parameters behind a shock wave.
Ch. 4. Shock Wave Theory 163
Oblique Shock
Shock line
P, Pz
fJ,~~~2
v~ '.
Vrz e
T, Tz
Bs .-[3.
"'.!:.L _ _ _____ -=__ _
{3s
Vi S( S2
a, az
Fig. ".2.1
Oblique shock
The continuity equation (or the mass flow equation) is the second
one. It determines the amount of fluid passing through unit surface
of a shock in unit time:
PIVIO = pzVoz (4.2.3)
here Vnl = VI sin as is the normal component of the velocity ahead
of the shock (Fig. 4.2.1).
Let us use the equation of motion reduced to the form of an equa-
tion of momentum for the conditions of the passage through a shock.
This is the third equation of the system. We shall obtain it by assum-
ing that the change in the momentum of the fluid passing in unit
time through unit surface area of the shock in the direction of a
normal to this surface equals the impulse of the pressure forces:
PIV~l - PzV~ = pz - PI (4.2.4)
With a view to Eq. (4.2.3), we can write this equation in the form
PIVOI (VBI - Vn2) = pz - PI (4.2.4')
Equation (4.2.4) can also be written as
(4.2.4")
In this form, the equation expresses the law of momentum conser-
vation when passing through a shock. If we consider the change in
the momentum in a direction tangent to the shock surface, then,
taking into account that the pressure gradient in this direction is
zero, we obtain the following relation:
pIVnlVn - pzVozV'fZ = 0
whence, when pIVnt = PzVn2 , we have
Vn = V'f2 (4.2.5)
Equation (4.2.5) is the fourth one of the system. I t indicates that
the tangential components of the velocity when passing through
Ch. 4. Shock Wave Theory 165
then
(4.2.18')
Let us find a relation for the flow deviation angle behind a shock .
.By (4.2.2) and (4.2.5'), we have
L\ Vn = 1 - tan (Os - ~s}/tan Os (4.2.19)
Ch. 4. Shock Wave Theory 167
B =
kl
Mi 2
n1
(Tr21 . fAml
fAmz
- 1) (4.2.24)
Normal Shock
We have given the general relations for shocks. Now let us use
them to analyse the nature of flow and the ways of calculating the
parameters of a gas behind shocks for constant specific heats and then
consider in greater detail practical ways of calculating similar
parameters for a dissociating fluid, i.e. for the more general case of
varying specific heats.
System of Equations
= V n1 - V n2
Introducing the values of V; = Vn~ +
Vi and Vi ,~ Vnf Vi +
.and multiplying both sides of the equation by V lIl V n2' we obtain
after simple transformations
k-l-1 *2(T1
---v;-a V' /;-1 V2(V
~nl- n2)-~ '{
V)
n l - n2
+ k;1 VnIVnz(Vni-Vn2)=VniVn2(Vnl-Vn2)
Excluding the trivial solution V nl = V n2 = 0 corresponding
to the absence of a shock, we obtain
V ni V nz -- a *2 - k-1
k +1 lT2
'{ (4.3.10)
where Al = Vl/a*.
Substituting Eq. (3.6.23) for Al and performing transformations,
we find
- 2 (1 -
Ll V n = It + 1
1)
My sin 2 8s (4.3.11)
(a) (c)
P2/P, TzlT,
ITT--'-~'F' r--T-'--'-'---'-~
7
, ,
I -
I
1- - i '---, - --~,~i
---1 -+-1""-
60 H---!
r--- r--- --- r---
6 j 50
40
tT, 6 f---+---+--+---f---L--c
5 f---i----t--t-J'--t--
lL
I
I '+ I--+--I~P--
30 I----
20
/ J ~-+--~--t--.:-----"
2 fO V
V
1 ! J '+ 5 6 Mnr
Fig. 4.3.1
Dependences of the ratios of the gas parameters behind a shock and ahead of it
on the value of Mlll = Moo sin as for k = 1.4 (IS = 1/6):
a-density ratio p,/p,; b-pressure ratio p,/p,; c-temperature ratio T,IT,
At M nf - - -
I
!
,
(Pz/P1 - - )
V
I
40
30 ~
r"f
20
/
10
// j, /
~~
Fig. 4.3.2 ~
Shock adiabat (1) and iseIl-
trope (2) o r--
(k = 1.4, 1\ = 1/6) 2 3 '+
Po
=
PI
(1 + _0_M
1-0
)i 1 +bli 2t
2
1 (4.3.21)
Consequently,
.J?l [(1 0)ltJ~sin28s-0](II-1)/26(Jllsin 8 s )(l+l1n
PI
o )(1+11)/211
X (1_0)-(1+11)/211
( 1-t - -1-6
-}J2
. 1
('4 ..'.).~
,
')2)
0 )(1+11)/26
( 1+ t1 - 0 Mrsin28s,
Po = P~-Pl
ql
= 2(1-0)
(1+0) il1i
{[(1'LO)
1·
7J1~sin28
1 S
-0](6-1)/26
(4.3.23)
'176 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
Fig. 4.3.3
Change in the flow deflection angle ~s depending on the shock angle 8 s for
various numbers Ml
Upon a further growth of the nose angle, the angle ~s may become
critical. According to Fig. 4.3.3, its value grows with an increase
in the number Mi' From the physical viewpoint, this is explained
by an increase in the strength of a shock, a greater density behind
it, and, as a result, by the shock coming close to the surface of the
body, which leads to denection of the now through a larger angle.
At a still larger nose angle, the flow behind an attached shock
becomes unstable, as a result of which the shock moves away from
the nose. Behind such a detached shock, a new stable flow region
appears. It is characterized by deflection through an angle also less
than the critical one. But unlike a snbcritical flow, this one is called
supercritical. This definition corresponds to the fact that the nose
angle of the body in the now exceeds the value at. which a shock is
still attached.
A detached shock changes its shape absolutely, 'which can be seen
especially clearly in the example of a now over a sharp-nosed cone or
wedge (Fig. 4.3.4). As long as the now is subcritical, the shock is
attached to the nose and the generatrix of its surface is straight.
The flow around thick wedges or cones may become supercritical,
upon which the shock detaches and acquires a curved shape. At the
point of intersection of the shock surface with the flow axis, the shock
angle 8 s = n!2 and, consequently, the parameters change according
to the law of a normal shock. In practice, there is a section of sllch
a normal shock near the axis.
'Vith an increase in the distance from the axis, the sllOck angle 8,
in accordance with Fig. 4.3.4 diminishes, remaining on a certain
section larger than the value that a subcritical flow corresponds to.
The change in the flow deflection angle is of the opposite nature.
12-01715
178 Pt. I. Theory. Aerodynamics of an Airfoil and' a Wing
V1 (M,)
Fto~s
Subsonic
F·19· 4.•
34 reqton
Detached shock ahead of a SupersoniC region
sharp-nosed body
tan Ss = cot ~s [{ ( ~: - 1)
+ l/ ~4 (..£!..
PI
- 1 ) 2 - tan 2 ~s ..£!..]
PI
(4.3.27)
two flow conditions behind snell a shock: to the left of the curve the
flow will be supersonic (M 2 > 1), and to the right of it-subsonic
(M~ < 1).
4.4. Hodograph
In addition to an analytical solut.ion of the problem of determin-
ing the flow parameters behind an oblique shock, there is a graphical
method based on the concept of a hodograph .
. A hodograph is a curve forming the locus of the tips of the velocity
vectors in the plane behind a shock. Let us consider the equation
of a hodograph. Let poillt A (Fig. 4.4.1) be the tip of the velocity
veet or V 2 and be loca ted, conseqllcn tly, on a hodogra ph constructed
in a coordinate system whose horizontal axis coincides with the
direction of the velocity VI ahead of a shock. Hence, the inclination
of the velocity vector V 2 is determined by the angle ~s. Let us
designa:te the vertical and horizontal components of this velocity
by 11' and 11, respectively. A glance at Fig. 4.4.1 reveals that u and lJj
can be expresserl in terms of the normal V n2 and tangential V,;
components of the velocity V 2 to the plane of the shock as follows:
u = V, cos 8 s V n2 sin 8 s , w = F,; sin 8 b - V n2 cos 8 8 (4.4.1)
We determine the component Vn2 from formula (4.3.10) in "which
we assume that VI11 = VI sin 8 s and ll,; = VI cos 8 s . Accordingly,
(4.4.2)
Let us eliminate the angle 8 s from this equation. For this purpose f
we shall use Eqs. (4.4.1). Multiplying the first of them by cos 8.,
the second by sin 8 s and summating them, we obtain
u cos 8 s + w sin 8 8 = V,;
Having in view that V" = VI cos 8 8 • we find
tan 8 s = (VI - u)/w (4.4.3)
Substituting for tan 8 s its value from (4.4.3) into the trigonometric
relation cos 2 8 s = (1 -+- tan 2 8 s )-1, we obtain
cos 2 8 s = [1 + ("VI - U)2/W 2]-1
Introduction of this value into (4.4.2) yields
T'I
12*
180 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
Fig.4.4.t
To the derivation of the hodo~
graph equation
It
", ~I
f---~~--- ~, (f ~ j'.~)_+~fI_?~I_~ ___
Fig. 4.4.2
Strophoid (shock polar)
52 - 51 = R In (Po/P~) (4.4.5)
When there is a compression shock, Po > p~ and, (;onsequently,
8 2 - 8 1 > O. This conclusion corresponds to the second law of
thermodynamics, according to which the entropy of an isolated
system with compression shocks increases.
Let us now consider the reverse situation when a gas passes from
a state characterized by the stagnation pressure p~ (the parameters
with the subscript 2) into a state with the stagnation pressure Po
(the parameters with the subscript 1) through an expansion shock.
In this case, by analogy vdth (4.4.5), the change in the entropy is
81 - 52=Rln(p~/po)
flo\v [M1 > 1 (a normal shock), Ml sin 8 s > 1 (an oblique shock)],
while the physically impossible phenomenon of a decrease in the
entropy (S2 - Sl < 0), to a subsonic flow (Ml < 1 and Ml sin 8 s <
< 1). Hence, shocks can appear only in a supersonic flow. We must
note that the relations obtained for the change in the entropy are
valid when an irreversible process of transition through a shock is
attended by an isentropic flow of the gas both ahead of a shock and
behind it.
I t follows from the above that the branches of a strophoid extend-
ing to infinity have no physical meaning. The part of a strophoid
(to the left of point A, Fig. 4.4.2) having a physical meaning is called
a shock polar. Such a curve is constructed for a given number Ai
(or M 1 ). Several curves constructed for various values of Ai form
a family of shock polars allowing one to calculate graphically the
velocity of a flow behind a shock and the flow deflection angle.
Let us consider an attached shock ahead of a wedge-shaped surface
with the half-angle ~s (see Fig. 4.1.1c). To determine the velocity
behind such a shock, we construct a shock polar corresponding to
the given I number Ai (or M 1 ) and draw a straight line from point 0
-
(Fig. 4.4.2) at the angle ~. Point of intersection N with the shock
polar determines the vector ON whose magnitude shows the value of
the velocity ratio A2 behind the shock. By formula (4.4.3), which we
shall write in the form
tan 8 s = (Ai - Au)/Aw (4.4.3')
the angle ANG on the shock polar equals the shock angle 8 sN ' We
must note that this angle can also be determined as the angle be-
tween the horizontal axis and a normal to the straight line connect-
ing the tips of the velocity vectors ahead of a shock and behind it
(points A and N, respectively, in Fig. 4.4.2).
When considering the shock polar, we can arrive at the conclusion
that a decrease in the angle ~s (point N moves along the curve toward
point A) is attended by a decrease in the shock angle 8 sN ' At the
limit, when ~s - 0, point N merges with point A, which physically
corresponds to the transformation of the shock wave into an infini-
tesimal shock wave, i.e. into a line of weak disturbances. The shock
angle for such a shock 8 s = /!1 is determined as the angle between
the horizontal axis and the straight line perpendicular to a tangent
to the shock polar at point A (Fig. 4.4.2).
A growth in the flow deflection angle (in Fig. 4.4.2 this corresponds
to point N moving away from A) leads to an increase in the shock
angle and to a higher strength of the shock. A glance at the shock
polar shows that at a certain angle ~s a straight line drawn from
point 0 will be tangent to the curve at point C. The angle of inclina-
tion of this tangent determines the maximum flow deflection angle,
called the critical one above (~s = ~cr). Assume that the wedge
Ch. 4. Shock Wave Theory 183
angle Bs> Bcr. In the graph, solid line OH drawn from point 0
and not intersecting the shock polar corresponds to this angle.
Therefore, when Bs> Ben we cannot find a graphical solution for
a shock with the aid of a shock polar. This is due to the fact that
the inequality Bs > Be r does not correspond to the assumptions
(on the basis of which we obtained equations for a shock) consisting
in that a shock is straight and should be attached to a nose. Physi-
cally-in the given case of the wedge angle Bs exceeding the critical
deflection angle Ber-the compression shock detaches and becomes
curved.
The determination of the shape of such a curved shock and of its
distance to the body is the task of a special problem of aerodynam-
ics associated, particularly, with the conditions of supercritical
flow past a wedge. If such a problem is not solved, then with the aid
of a polar in the field of definition from point D to A "we can give
only a qualitative appraisal of the change in the parameters in a re-
gion ahead of the surface in the flow. If, on the other hand, the shape
of the shock is determined for preset flow conditions (in addition
to calculations, this can also be done with the aid of blowing in
a wind tunnel), it is possible to establish quantitative correspondence
between the points of a shock polar and the shock surface.
Assume, for example, that we have set the angle Bs and points E
and N on a shock polar (Fig. 4.4.2). The shock angle SsN = LANG
corresponds to point lV, and the angle SsE = LAEK (EK -L OB)
to point E. If the confIguration of the shock wave front is known,
then by direct measurement we can find on it a point N' with the
wave angle Ssw and a point E' with the angle SsE' (see Fig. 4.3.4).
In the :same way, we can find a point C' on the shock that corresponds
to the critical (maximum) deflection angle Be r.
On a preset surface of a detached shock, point D on the shock polar
corresponds to the shock apex (a normal shock), and terminal point A
of the polar corresponds to the remotest part of the shock that has
transformed into a line of weak disturbances.
For an attached shock (Bs < Ber), we can indicate two solutions.
as can be seen on the shock polar. One of them (point E) corresponds
to a lower velocity behind the shock, and the other (point N), to
a higher one. Observations show that attached shocks with a higher
velocity behind them, i.e. shocks with a lower strength are possible
physically.
If we draw on the graph the arc of a circle whose radius is unity
(in the dimensional axes wand u this corresponds to a radius equal
to the critical speed of sound a*), we can determine the regions of
the flow-subsonic and supersonic-which points on the shock polar
to the left and right of the arc correspond to. In Fig. 4.3.4, the
section of the flow corresponding to a subsonic velocity is hatched.
A close look at the shock polar reveals that the velocity is always
184 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
Fig. 4.5.t
Parameters of a gas at the
stagnation point behind a nor-
mal shock
(4.3.20') in which the ratios P2' PI and pz PI are found from (4.5.3}-
and (4.5.4), respectiyely. With this in vie\Y, we have
p~/Po = [(1 +
6) Mi - 0]<0- 1 )/21\ [Mi1(1 - 0 -:- oM;)]<l+6)/2 6 (4.5.7)
Determining Po by (4.3.21), we find
P~/PI = [(1 + 6) M; - 0]<0- 1)/26 M<~+6)/6 (1 - 0)-<1+ 6 )/2 6 (4.5.8)
Knowing the absolute pressure p~, \ve can determine the dimension-
less quantity Po
= (p~ - Pl)/QI-the pressure coefficient at the
stagnation point. Taking into account that the velocity head is
kpIMr 1-+-8 U2
Ql= 2 =2(1_8)Pl 1r.l 1
we obtain
2(1-8) {[(1 ' 6) M2_0](6-1)/~6lI (1+6)/6
Po= (1 + 8) M~ :- 1 ~ 1
(4.6.1)
Consequently, the limiting ratio of the densities by (4.2.13) is
Pl/P2' = 1- ~ Vn = 6 (4.6.2)
Introducing this value into (4.3.27), we obtain the following ex-
iPression at, the limit when Ml sin as -+ 00:
tan as = (cot ~s/26) [1-6 + V (1- 6)2- 46 tan 2
Ps ] (4.6.3)
Let us find the limiting value of the pressure coefficient. For the
-conditions directly behind a shock, as follows from (4.3.15"), when
Ml sin as -+ 00 and Ml -+ 00, we have
(4.6.4)
We obtain the corresponding quantity for the point of stagnation
'from (4.3.23):
Po = 2 (1 - 6)(0-1 )/2 6 (1 + 6)-(1+ 6)/2 0 sin 2 as (4.6.5)
The ratio of the pressure coefficients is
PO/P2 = (1 - ( 2)-(1+ 6)/20 (4.6.6)
In the particular case when 6 = 1/6 (k = 1.4), the ratio = i/P2
= 1.09. The limiting value of the number M 2 can be found from
(4.3.19), using relation (4.3.16) for T 2 /T 1 • A passage to the limit
when Ml sin as -+ 00 and Ml -+ 00 yields
(4.6.7)
To find the limiting parameters behind a normal shock, we must
assume that as = n/2 in the above relations. As a result, from (4.6.4)
and (4.6.5) we have:
P2 = 2 (1 - 6) (4.6.4')
Po = 2 (1 - 6)(0- 1 )/20 (1 + 6)-(1+6)/2 6 (4.6.3')
Ch. 4. Sr.ock Wave Theory 187
We can see that the ratio Po/Ill is the same as for an oblique shock.
The limiting Mach number behind a normal shock is
(4.6.7')
For 6 = 1/6 (k = 1.4), the Humber M 2 = V 1/7 :::::::: 0.38. The
actual "alues of the dimensionless parameLers behind a shock at
finite, although "ery large, Mach numbers depend on MI'
Let us consider the corresponding working relations for the case
when attached shocks originate ahead of slencter ,,,edges, and the
shock angles are therefore low. Assillning in (4.3.2;-)) that tan 8 s ::::::::
.~ 8 s and tan (8 s - ~,) :::::::: 8 s - ~s, ,ye obtaill
(8 s - ~J8s
= (1 - 0 , oM~8~)(Mi8~)
Introducing the symbol K = M 1 f1 s , after trallsformations, ,ye flUd
f)~ 1 8s 1 O' 4. 6 8
~~ - 1- 6 . ~ - [(2 = ( . . )
Solving this equation for 8s/~s and taking into account that the
condition 8s/~s > 1 is physically possible, we find
8s --L ~.r 1 --L _ (4 6 9)
~= 2(1-6) I V 4(1-6)2 , K2 ..
Substituting for the angle fls the value determined from (4.6.9),
we find
(4.6.12)
We can see from (4.3.18) that at low values of 8 8 the second term
on the right-hand side may be ignored, and we can thus consider
that V 2 ~ VI' With this in view, the ratio of the sqllares of the
Mach numbers in accordance with (4.3.19) is M;IM~ = T 1 IT 2 •
Substituting for the ratio TIIT2 here its value from formula (4.3.Hi)
in which we assume that sin fls ~ fls, we ohtain
fl~M; = K~I {[(1 + 8) K~- 8] (1-8 + 8K~I} (4.6.16)
When Ks -+ 00, we have
(4.6.16')
By next using an i-S diagram (see l6, 81), we determine the tempera-
ture T 2' and then from Fig. 1.;).7. the mean molar mass !-lm2' Instead
of the diagram, one may use suitahle tahles of the thermodynamic
fUIlctions of air (see l71), which will increase the accuracy of the
cakulations.
By inserting the found values of PZ' T 2' and flm2 into the equa-
tion of state (1.5.8). we can determine the density pz and defme
the value of tl Vn more precisely by (4.2.21). We next use this value
in a second approximation to fmd the pressure and enthalpy, respect-
ively, by formulas (4.2.15) and (4.2.16). According to these values
and with the aid of tables and graphs, we define the temperature
and mean molar mass more exactly. "~e use the refined values of
P2' T 2, and !-lm2 to [md the density in the second approximation
by the equation of stale. The approximations are terminated when
the preset accuracy is achieved.
\Ve can evaluate the shock angle corresponding to a given velocity
VI by the formula sin 8 s = Vm VI' By introducing the value of 8 s
and also the number MI =-col\ia l into (4.::\.2fi), we determine the
flow deviation angle Bs behind the shock.
An oblique shock can also be calculate([ when the values of the
free-stream parameters (including the number M I ) and the angle Bs
are known. In a fust approximation. we determine the slwck angle
Os for an undissociating gas [see (11.3.27)], and 'hnn find the cor-
responding values of tl iTn , P2' and 12 by formulas (4.2.19), (4.2.15),
and (4.2.16). Using these values. we determine tlte temperature T,..
and the mean molar mass ~tm2 from tables [71 or graphs [6, 81. ~ext
by formulas (4.2.23) and (4.2.24). we define ~ en more precisely,
and by expression (4.2.25), tan 8 s and the angle 8 s . We refine the
other parameters according to the relevant formulas.
We calculate the parameters of a gas behind a normal shock in
a similar way with the use of tables or graphs of the thermodynamic
functions for high temperatures. 'Ve assume that 8 s = :11/2 and
Bs = 0 and, therefore, use (4.2.27)-(4.2.35).
When dissociation and ionization occur, the relative values of the
parameters of a gas behind a shock waye depend not only on the
temperature, which is characteristic of varying specific heats, but
also on the pressure. These relations are shown graphically in
Figs. 4.7.1-4.7.3. The ratios of the temperatures and densities are
calculated for averaged values of the temperature TI of 220 and
350 K. These values equal, respectively, the probable minimum
and maximum that are chosen depending on the change in the air
temperature with altitude for decreased and increased annual average
values. Available data show that dissociation and ionization give
rise to a substantial change in the equilibrium temperature and
density in comparison with constant specifIC heats (k = 1.4 =
= const). The pressure depends to a considerably smaller extent
190 Pt. I. Theory. Aerodynamics 01 an Airfoil and a Wing
TZ/T1
60
50~----~------r-----~----~
40~-----L------+-~
Fig. 4.7.1
Ratio of the air temperatures 20 If--------r-~~A~::::-::::=t==;
behind and ahead of a shock
with account taken of dissocia- 10
tion and ionization:
solid lincs- T t = 220 K, dashed
linrs- T t = 350 K
~/rp.L-__-,____-,____-,_____
20r------,-------T--
Fig. 4.7.2
Ratio of the air densities behind
and ahead of a shock with ac-
count taken of dissociation and
ionization:
solid lines-T, = 220 K, dashed
lines-Ttl = 350 K o5 10 15 20 Mnl
~/Pf
800
H=BO/fm
0=220 X ,//
600
/
40D d
Fig. 4.7.3 200 /~ ~
Ratio of the air pressures
behind and ahead of a shock V
with account taken of dissocia- o /
tion and ionization 5 fO 15
H,lrm
J..- f-f" !
fOO ,// I / 10 2
80
50
/
~
/'
k
""
t--~ I-~
I
-
f---
....,""
I
F
/
2'!J
I-""
-::t-
f--r
/"
Vi'l IL--- ~~ V-
1-$
10 5
r---- 1--." 10'+ /
40
,-" V 1
V
V ,.. L.--J 7I / fOS-
20
...... v
./
---: 7 /
'-
Fig. 4.7.4
Pressure and temperature at the aV .....
V
>,
/ If /
,Point of stagnation 2000 'fOOD 5000 Vr,lnjS
Fig. 4.7.5
Density at the point of stagna-
tion on the surface of ,a sphere
for nitrogen, oxygen, and air
(Poo = 10 3 Pa, Too = 290 K) 17 2f 25' Moo
in Moo, the gas completely dissociates, and the density drops. Next
a growth in Moo is attended by primary ionization of the oxygen.
This leads to an increase in the specific heat and, consequently, to
a certain growth in the density.
The influence of the varying nature of the specific heat on the
change in the density of nitrogen is observed only at very large
numbers Moo when dissociation and ionization take place. These
processes run not consecutively, as in oxygen, but virtually simulta-
neously, which is due to the smaller difference between the energies
of dissociation and ionization of nitrogen. This is why the curve
p~/p 00for nitrogen is more monotonic than for oxygen.
Ch. 4. Shock Wave Theory 193
one of the similar relation for gaseous nitrogen because the latter
is the predominating component in the air.
The calculation of the parameters of a gas behind a shock with
a view to the varying nature of the specific heats is described in
greater detail in [14].
-----------------
194 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
Fig. 4.8.t
Experimental (1, 2) curves of
the vibrational relaxation time
and calculated (3, 4) curves of
the relaxation time for the dis-
socia tion of oxygen and ni tro-
gen, respectively, at atmosp-
heric pressure
EquilibrIum Processes
Equilibrium flows have been studied better than non-equilibrium
ones both from the qualitative and the quantitative standpoints.
The regions of a flow in which equilibrium sets in are different because
of the different relaxation times for the excitation levels. The
duration of the establishment of equilibrium relative to the vibra-
tional degrees of freedom is longer by several orders than relative
to the translational and rotational ones. Equilibrium sets in even
more slowly relative to the composition of a gas mixture when dis-
sociation and ionization occur. Accordingly, the scheme of a non-
equilibrium process is such that the attainment of equilibrium of
one degree of freedom may be attended by the beginning of a relax-
13*
196 Pt. I. Theory. Aerodynamics of an Airfoil and· Bi Wing
T p
t,x
Fig. 4.8.2
Relaxation process in a shock wave:
(the hatched region determines the thickness of the shock wave):
1 and 2-front and rear surfaces of the shock wave, respEctively
,J2/1P'
fO
--
~--
f-.-i -'::'j-"':r-
\ )--./ Siwek rave P2' T2
8 +------1
'"r-:::.::i-- ~- V(J~fl f------ f----
Fig. 4.8.3
I
I---
:
L
Influence of non-equilibrium I
dissociation on the density and I
temperature behind a shock '+
wave
("ll. = 14.2, T. = 300 K, P. = I
120 Pal o 2 4 6 8 s,mm
follows:
(5.1.3)-
When e = 0, this equation coincides with the continuity equation
for a two-dimensional plane flow in the Cartesian coordinates x and y.
If e = 1, we have a continuity equation for a two-dimensional axi-
symmetric flow in the cylindrical coordinates y (r), x. Accordingly,.
for both kinds of flow, we may consider that the equations of motion
(5.1.1) are written in a generalized form.
Having determined the partial derivatives in continuity equation
(5.1. 3), we obtain
(Vx ~~ + Vy ~~ ) y8+ pye ( a~x + a~y ) +pV yeyE-l =ceO (5.1.4}
.!!L=
dx
_L(V
a2 x
iJV x
ax
+VY aFx)
(iy
(5.1.5)
where
2=~ k-l [V2 _(~)2_(~)2J
a 2 max ax ay
(5.1. 9)
where V = V/VJllax '
The calculation of a vortex gas flow consists in solving a differen-
tial equation for the stream function 'IjJ. To obtain this equation, let
us differentiate (5.1.9) with respect to y and to x:
av
_x_
ay
= -
_By-e-l (1- V2t1/(k-l) a·"
_'"
ay
+ y-e _k-1
1_
2
(1_V2 t 1/(k-1)-1 aV2. 81jl +y-e(1_V2)-I/(k-1) 8 1jJ
ay 8y 8y2
z
av" = _ y-e _1_ (1- V2)-I/(k-1)-1 aV2 . aljl _ y-e (1- V2)-I(k-1) 8 Ijl
ax k-l 8x ax 8x 2
Taking into account expression (3.6.22) for the square of the speed
of sound, and also relations (5.1. 9), we can write the expressions
obtained for the derivatives f)V)f)y and f)Vy/f)x as follows:
aV x
ay
= _~ V.+
y x
vx
2a 2
• 8V
8y
2
+ y -e(1_V2)-I/(k-1) a2 1jl2
8x
(5.1.10)
8V rl = ~.
2
8V2 _ -e (1- V2)-I/(k- 1)
2
8 1jl (5.1.11)
8x 2a 8x y 8x 2
- V
y
iJ21jl
ox:!
-+. V x ~
ax ay
= L.
2
aV2
iJx
(1- V2)1/(Ii-l) (1 -.-:.:)
a2
(5 . 1 . 13)
x ( 1 - --;;s- ) +
l'2
J-e yz (1 - V2)1/(k-1) (5.1.14)
Determining the derivatives aV2 /ay and aV2/c]x from these expres-
sions and inserting them into (5.1.10) and (5.1.11), respectively, we
obtain
JL
2
(1 - V2)1/(k-1l ( 1-'-:':)
2 a
iJV x
ay
= _2_l-e
t _ V (1- 0)1/(1<-1)
x y
(1 -1"2) a 1\' 2
X a 1jJ
V [ -V --+V.--
- +_x
2
__t_
yl-e
V2 (1- V2)1/(Ii-l)] +~
2
(1 _ F,2)
a2
a2 1f
ay:!
+Vx ~)-i.(1-~)
ax ay 2 a2
iJ 1p2
rJx
+ Vx ~)
ax rJy
(1- YZt 1
1/(k- 1 ) _ _
l
(1- V 2 t l /(h- 1 )
204 Pt. I. Theory. Aerodynamics of an Airfoil and ~ Wing
_ Vx
a 2 - V2
. _1 (_ V ~
ye Y ax ay
+ V x aay2'l') (1- V2)-1/(k-1)
2
Calculations yield
, 'XV=-V (ilV"y _ i:lVx). +V (i:l~'y _ i:lVx).
cur 1 Y i:lx i:ly 11 x ax i:ly 12
(5.1.21)
Accordingly, for the projection of the vector curl V X V onto
a normal n to a streamline, we obtain the relation
(curl V X V); = (curl V X V)~ + (curl V X V)~
= V 2 (aVy/ax - avx /ay)2 (5.1.22)
Examination of (5.1.20) reveals that this projection can also be
\\Titten in the form
(curl V X V)n = T dS/dn
or, with a view to (5.1.22)
V (aVy/ax - (}Vx/{}y) = T as/dn (5.1.23)
Z
Since a = kRT, then, taking (3.6.22) into account, we find an
expression for the temperature:
T=~= k-1 ._1_(V2ax_V2)= k-1. v;hax (1_V2)
kR 2k R m 2k R
+_8_
yl-e
a2V x (1_V2)1/(k-1)
= -=-.
k 1 V2
2k RV
~
-
(1- V2)k/(k-1l ye (a 2 _ V2) __
dS
dn
(5.1.25)
Equations (5.1.8) for the velocity potential and (5.1.25) for the
stream function are inhomogeneous non-linear second-order partial
differential equations. The solutions of these equations cp = cp (x, y)
and tP = tP (x, y) are depicted geometrically by integral surfaces in
a space determined by the coordinate systems x, y, cp or x, y, tP.
In these systems, the plane x, y is considered as the basic one and
is called the physical plane or the plane of independent variables.
206 Pt. I. Theory. Aerodynamics of an Airfoil and a, Wing
y
B
Fig. 5.2.1
Initial curve AB; the required
function and its first deriva- A
tives with respect to x and yare
known on it o~------------------~x
where u = CPxx (or 'Ijlxx), s = CPxy (or 'Ijlxy), and t = cPyy (or 'IjlyyJ are
second partial derivatives, the values of A, B, and C equal the
coefficients of the corresponding second partial derivatives, and
the quantity H is determined by the free terms in Eqs. (5.1.8) and
(5.1.25).
We shall find a solution of Eq. (5.2.1) in the vicinity of the initial
curve AB (Fig. 5.2.1) in the form of a series. At a point M (xo, Yo).
the required function is
+r [(~x)n ~:~
00
where cp (x, y) is the value of this function at the given point N (x, y)
on the initial curve, /l..x = Xo - x, and l1y = Yo - y. The fUllction
'IjJ may be used in (5.2.2) instead of cp.
Series (5.2.2) yields the required solution if values of the functions
cp (or 'IjJ) and also their derivatives of any order exist on a given curve
and are known. Since the first derivatives on this curve [we shall
denote them by p = CPx (or 'ljJx) and q = CPy (or 'ljJy)l are given. our
task consists in fmding the second derivatives on it, and also deriva-
tives of a higher order. Hence, the solution of the Cauchy problem
is associateu with fmding of the conditions in which the higher deriva-
tives on the given curve call be determined. We shall limit our~elves
to determination of the second derivatives. Since these cieriYatives
are three in number (u, s, and t), we have to compi]p thp same
number of independent equations to flIld them. Equation (5.:2.:2),
which is satisfied on the initial cune AB, is the first of them. The
other two are obtained from the following known relations for the
total differentials of the hmctions of two independent variahlrs on
this curve:
dp == (fJp/8x) dx + (8p/rJy) dy = u dx ~. s dy
dq = (iJq/8x) dx + (8q/8y) dy = s dx + t dy
Hence, we can \vrite the system of equations for determining the
second derivatives in the form
Au+2Bs-i Ct+H =0 }
dxu·+ dys+O.t-dp=O (5.2.3)
O.u+dxs+dyt-dq=O
This system of equations is solved for the unknowns u, s, and t
with the aid of determinants. If we introduce the symbols /I.. and
/1..,,, 11 8 , /l.. t for the principal and partial determinants, respectively,
we have
u = l1u//I.., s= /1.. 8 111, t = /l..tll1 (5.2.4)
where
A 2B C -H 2B C )
/1..= dx dy 0 /I.. u =". dp dy 0
0
A
dx
-H C
dy dq
A
dx
2B
dy
-H
I
)- (5.2.4')
I
/l.. s = 0 , /1.. 1 = dx
dx dp dy dp I
0 dq dy 0 dx dq J
It follows from these relations that if the principal determinant &
does not equal zero on the initial curve AB, the second derivatives u.
s, and t are calculated unambiguously.
~08 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
Let us assume that the curve has been chosen so that the principal
determinant on it is zero, i.e. ~ = O. Hence
A (dy/dx)2 - 2B (dy/dx) C = 0 + (5.2.5)
I t is known from mathematics that when the principal determin-
ant ~ of the system of equations (5.2.3) is zero on the curve given
by Eq. (5.2.5), the second derivatives u, s, and t (5.2.4) are either
determined ambiguously, or in general cannot be determined in terms
()f cr, p, and q.
Let us consider quadratic equation (5.2.5). Solving it for the
derivative dy/dx, we obtain ~
(dy/dx)1.2 = y~ .• = (VA) (B + V B2 - AC) (5.2.6)
This equation determines the slope of a tangent at each point of
the initial curve on which the principal determinant ~ = O. It is
not difficult to see that (5.2.6) is a differential equation of two fam-
ilies of real curves if B2 - AC > O. Such curves, at each point of
which the principal determinant of system (5.2.3) is zero, are called
characteristics, and Eq. (5.2.5) is called a characteristic one.
From the above, there follows a condition in which the unambigu-
ous determination of the second derivatives on the initial curve
is possible: no arc element of this curve should coincide with the
characteristics. The same condition ~ =1= 0 holds for the unam-
biguous determination of the higher derivatives in series (5.2.2).
Consequently, if ~ =1= 0, all the coefficients of series (5.2.2) are
determined unambiguously according to the data on the initial
curve.
Consequently, the condition ~ =1= 0 is necessary and sufficient to
solve the Cauchy problem. This problem has a fundamental signi-
ficance in the theory of partial differential equations, and formula
(5.2.2) can be used to calculate the flow of a gas. But from the view-
point of the physical applications, particularly of the calculation
of supersonic gas flows, of greater interest is the problem of determin-
ing the solution according to the characteristics, i.e. the method of
characteristics. This method can be obtained from an analysis of the
Cauchy problem and consists in the following. Let us assume that
the initial curve AB coincides with one of the characteristics, and
not only the principal determinant of the system (5.2.3) equals zero
along it, but also the partial determinants ~u = ~s = ~t = O. It
can be proved here that if, for example, the determinants ~ and ~t
equal zero, i.e.
Ay'2 - 2By' +
C = 0 (5.2.5')
A (y'q' -p') -2Bq' - H = 0 (5.2.7)
where p' = dp/dx, q' = dq/dx, then the equality to zero of the other
determinants is satisfied automatically.
Ch. 5. Method of Characteristics 209
5.3. Characteristics
Compatibility Conditions
Determination of Characteristics
y
2
Fig. 5.3.1
To the explanation of the
physical meaning of a character-
istic:
1-direction of a characteristic of
the first family at point A; 2-cha-
rac1eris1ic of 1he first family with
the angular coefficient ).,; a-direc-
tion of a characteristic of the second
family at pOint A; 4-characteristic
of the second family with the angu-
lar coefficient"2 o x
It thus follows that ~t is the Mach angle. vVe have therefore estab-
lished an important property of characteristics consisting in that at
every point belonging to a characteristic, the angle between a tan-
gent to it and the velocity vector at this point equals the iVIach angle.
Consequently, a characteristic is a line of weak diosturhances (or a
Mach line) having the shape of a clll've in the general case.
The definition of a characteristic as a Mach line has a direct
application to a two-dimensional plane supersonic flow. If we have
to do with a two-dimensional spatial (axisymmetric) supersonic
flow, the Mach lines (characteristics) should be considered as the
generatrices of a surface of revolution enveloping the Mach cones
issuing from vertices at the points of disturbance (on the character-
istics). A surface confining a certain region of disturbance is called
a wave surface or three-dimensional Mach wave.
We already know that pressure waves appear in a gas whose super-
sonic flow is characterized by a growth in the pressure. But such a
flow may be attended by lowering oj the pressure, i.e. there will be
a supersonic expanding flow, and the Mach lines will characterize
expansion waves. The relevant characteristics, which in the general
case are curved lines (for a plane flow) or surfaces formed by the ro-
tation of these lines (for a spatial axisymmetric flow) coincide ,,,,-itlt
these Mach lines. If a flow contains Mach lines (characteristics) ill
the form of straight lines, then simple expansion waves whose vel-
ocity of propagation has one direction corresponrl to them. 'Vhen
Mach lines correspond to expansion waves, we call them lines of weak
disturbances, using the terminology adopted for weak pressure wayes.
I t must be remembered here that no other expansion ,vaves except
weak ones appear in an expanding supersonic flow, becanse other-
wise we would haye to assume the possibility of the formation of
"strong" expansion waves (expansion shocks) which in real flow con-
ditions cannot exist.
If at a point of a physical plane the flow velocity and speed of sound
are known, the above property of characteristics makes it possible
14*
212 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
Orthogonality of Characteristics
fig. 5.3.2
Property of orthogonali ty of
charac teristics
X (1 - V-2)1/(k-l) - ~
k-l
V (1_112 )1/(h- 1)-1 V- dV
x
(5 3 12)
dx· .
Ch. 5. Method of Characteristics 215
Y' ( dV x
" dx
A _ dry)
2.1 dx T
I ~
k-l
(1- V2)1/(I<-I)-1 V dV (V -l' A, )
dx Y x 2.1
+. 1
V2 _ a2
[e-l
lOy
2V (1_V2)1/(k-l)_
a x
k-l
2k
x
><
V2
max
RV
(1- yz.)H/(I<-1) yE (a 2 - 1'2) .-!:.£]
dn
=0
After cancelling quantities where possible and introducing the di-
mensionless variables
Vx = T'x/Vrnax, fy = l'y/Vmax , iT = V/V max , Via = M
we can 'write this equation in the form
~ (r x"2.1
y 1\,1.2
E ~ - 1.' I')
Y
dV x ~
+ ( dx 1\,2.1 -
dV
d;u )
-l- 21' .Vy-VXf.2.1.dV_!:... Vx
'k-l l-P dx y l-V~/a2
(5.3. '14)
d~
-
- V dx (/"'2,1 sin ~ + cos~) + Iie -11 [
/"'1.2 (/"'2.1 cos ~ - sin~)
_ cos ~ ] _ k -1 . _1_ (1 _ l1Z) cot21-t .~=O
1- cos 2 ~/sin21-t 2k RV l-cos2B/sin2/-t dn
Having in view that [(k - 1)/21 (1 - V 2lVinax) = a2/V~ax =
= V2 sin 2 1-1, we find
~ --+- d~ (1.,2.1 sin ~ + cos ~) tan 2 I-t _ e..!!.!....
V ' A2.1cos~-sin~ y
X [
/..,
1.2-
sin2l-tcos~
(A2'1cos~-sin~)(sin21-t-cos2~)
Jt an "'"
2
+ ~.
kR
sin41-t
(sin2 I-t-COS 2 B) (A2.l cos B-sintl)
~--O
dn - (5.3.17)
Performing the substitutions /"'2,1 = tan (~ + 1-1) and /"'1.2 =
= tan (~ ± 1-1), we transform the individual expressions contained
in (5.3.17) to the following form:
A2.1sin~+cos~ 1
A2,1CusJ)-sin~ =J=tanl-t
(5.3.18)
/.., _ sin 2 I-t cos ~ sin ~ cos I-t
(1.,2,1 cos ~-sin~) (sin21-t--cos2~)
(5.3.19)
1,2 cos (~ ± ~t)
4
sin I-t = + sin 3 I-t (5.3.20)
(sin21-t-cos2~) (Au cos B-sin B) cos (B ± I-t)
With a view to (5.3.18)-(5.3.20), Eq. (5.3.17) for the character-
istics of the first and second families, respectively, acquire the form
dV _ tan dR __ e.!!::.. . si n I-t sin ~ tan I-t
V 1-1 P Y cos (~+ I-t)
that is an angle. We shall express the ratio dVIV in the form d'A/'A.
('A =~ Vla*), and cot fl with the aid of (3.6.23) in the form
(5.3.26),
Consequently,
A
ffi= \' Vr(A2-1)/'(1_ k--l 'A'l.) d,")'" (5.3.27).
• k+2 A
1
(t)
~ /
= V
k+l
k _ '1 ta n -I
v r
k-l ").,,2-'1
k -1- '1 • -1-_---:-k---l-;--").,,-2-
k -1- 1
(5.3.28),
(5.3.29)
Table 5.3.1
Introducing the angle ffi into (5.3.21) and (5.3.22), we obtain the
following equation for the characteristics:
-A)
d (ffi+ dx
p -e-'
sin Bsin f.t + - dx . sin 2 f.t cos f.t . _ _ 0
dS _
y C:JS (~ ± f.t) - kR cos (B ± ft) dn
(5.3.31)
Equation (5.3.31) corresponds to the most general case of super-
sonic two-dimensional (plane or spatial) vortex (non-isentropic)
flow of a gas.
Ch. S. Method of Characteristics 219
(5.3.36)
Fig. 5.3.3
E picycloids-characteristics of
a plane supersonic flow:
1-characteristic of the first family;
2-characteristic of the second fa-
mily
.2_ (_1 _ _
1 ) (5.3.37')
k-1 M2 Ml
222 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
All the above relations have been found for a perfect gas. At very
low pressures, however, a gas is no longer perfect. This is why the
calculated ultimate flow angles are not realized and have only a theo-
retical significance.
(a)
A'
c'
':!A
a xA X3 Xc X
\/~
I~
'j k-I _ _-----"~
Fig. 5.4.1
Calculation of the velocity at the intersection point of two characteri~tics of
different families:
a-physical plane; b-plane of hodograph
(5.4.17)
(a) (0)
Y Ay
Vn
I
I
I
I
I
!/n:
I
I
I
I
I
D~-~-~-----~~----- D'~---1----'------i--
Xn XB X 7.x
Fig. 5.4.2
Calculation of the velocity at the intersection point of a characteristic with
a surface in a flow:
a-physical plane; b-plane of hodograph
or
M R R (PO,E-Pll, D) COS (~D-f.lD)
--,-
PO, D (xD-xB) sin f.lD
(5.4.29)
The coordinates XB and YB of point B are determined by the si-
multaneous solution of the equation for a characteristic of the second
family and the equation of the wall contour:
YB - YD = (XB - xo) tan (~D - ~lD); YB = f (5.4.30)
(xD)
A graphical solution of these eqllations is shown in Fig. 5.4.2a.
We use the found coordinates XB and YB to evaluate the angle
~B from the equation
tan ~B = (dy/dx)B (5.4.31)
The entropy SB (or the stagnation pressure Po, B) at point B is
considered to be a known quantity and equal to its value on the
streamline coinciding with a tangent to the surface. The entropy is
considered to be constant along an element of the tangent.
Upon calculating the increment LlffiD by (5.4.27), we can find
the angle ffiB = LlffiD +ffiD, and then determine the number MB
from Table 5.3.1. How the velocity at point B is determined is shown
graphically in Fig. 5.4.2b, where element D' B' of a characteristic
of the second family in a hodograph corresponds to an element of a
characteristic of the same family in a physical plane.
The third problem consists in calculating the velocity at the
intersection of the characteristics with a shock and in determining
the change in the inclination of the shock at this point. Since a char-
acteristic in its natnre is a line of weak disturbances, this intersec-
tion physically corresponds to the interaction of a weak wave with
a compression shock. Assume that closely arranged expansion waves,
which characteristics of the first family correspond to, fall at
points J and H on compression shock N of a preset shape Y = f (x)
(Fig. 5.4.3a). As a result, the strength diminishes, as does the incli-
nation angle of the shock. Since points J and Hare SOllfces of distllf-
bances, expansion waves appear, and characteristics of the second
family can be drawn through these points. One of such character-
istics, passing throllgh point J, intersects the neighbouring conju-
gate characteristic at point F called the nodal point of the charac-
teristics.
To determine the change in tIle inclination of the shock and in
the velocity behind it, one must use the properties of the character-
istics .J F and FH passing through nodal point F, and also the rela-
tions for calculating a compression shock. Since length JH of the
shock is small, this section may be assumed to be linear. The angle
of inclination of the shock on this section and the corresponding pa-
15·
228 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
(lJ)
y :>'y
Shock
potar
o x
Fig. 5.4.3
Calculation of the velocity at the intersection of characteristics with a shock:
a-physical plane; b-plane of hodograph; I-characteristics of the first family; 2-
characteristics of the second family
d~ (J2.)
I"S PI
= 2 cot S~ (1-8~)
CPI
~.~
PI dfls
(5.4.34)
Ch. 5. Method of Characteristics 229
+ sin (8 s - 1
~s)cos (8~ - ~s)
}
Inserting the value of Ll~F into (5.4.38), we can find A(OF, calcu-
late the angle (OH = A(OF + (OF, and determine more precisely the
number M H . By calculating the angle ~H = A~F +
~F' we use
the values of this angle, and also of the preset number Moo to find
the shock angle 9s , H at point H and, consequently, to determine
the shape of the shock more accurately on section JH. If necessary,
the calculations can be performed in a second approximation, adopt-
ing instead of the parameters at point J their average values between
points J and H. Particularly, instead of the angles (OJ and ~J' we
take the relevant average values of 0.5 «(OJ +
(OH) and 0.5 (~J ~H)' +
Figure 5.4.3b shows how the problem is solved graphically. Point
H' on a hodograph, corresponding to point H on a physical plane, is
determined as a result of the intersection of element F' H' of a first
family characteristic with a shock polar constructed for the given
free-stream number Moo. The vector 0' H' determines the velocity
AH at point H.
Fig. 5.5.1
Nozzle of supersonic tunnel:
I-top wall; 2 and 5-side walls; 3-exit section; 4-bottom wall; 6-critical section; 7-
receiver
Fig. 5.S.l
Unshaped two-dimensional
supersonic nozzle with a radial
flow
The nozzle ensuring such a flow is a mouthpiece whose side walls are
flat, while its top and bottom walls have a specially shaped contour
(Fig. 5.5.1).
In addition to determining the shape of its curved contour, the
design of a nozzle includes calculation of the parameters of the gas
in the receiver (the parameters of stagnation) and in the critical sec-
tion, and also its area S*. The parameters of the gas at the nozzle
exit are usually preset, namely, the number Moo, the pressure po,,,
the area of the exit section S = lb, and the temperature of the gas
in the receiver To. The area of the critical nozz Ie section is found
from flow rate equation (3.6.44) which we shall write in the form
pooV ooS = p*a*S*. Hence S* c.= (pooV oo/p*a*) S = qS.
I t follows from (3.6.46') that the parameter q
is determined hy
the preset number M at the nozzle exit. According to this value of
0()
Moo and the pressure p at the exit, and by using formula (3.6.29),
00
we can find the pressure Po in the receiver needed to ensure the preset
number M"" at the exit.
Next the angle 2y of an unshaped nozzle is set (Fig. 5.5.2). Expe-
rimental investigations show that this angle is generally chosen
equal to 30-35°.
If the section of the inlet part of a nozzle changes sufficiently
gradually, the flow downstream of the critical section can be consid-
flred as an expanding radial flow from a source at point O. Such a
now has the property that its direction coincides with that of radial
lines emerging from point O. The change in the parameters of the
gas in magnitude along each of these lines is of the same nature.
The length of the subsonic portion of the nozzle of uni t width is de-
232 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
Fig. 5.5.3
Construction of a:shaped supersonic two-dimensional nozzle
fig.6.t.t
Thin airfoil in an incompressible flow
where CPb and CPu are the velocity potentials on the bottom and upper
surfaces, respectively. Consequently, the lift force produced by the
pressure acting on an element of area is
dYa,!c = (Pb - Pu) dx = -V oopoo (oCPb/ox - ocpu/ox) dx
while the lift force for the entire airfoil with the chord b is
b'
} T
a.ic = - V
oop"" J r (~
oCfJb oCfJu) d
- ----ax x
o
Assuming the upper limit b' of the integral to be approximately
equal to the chord b (because of the smallness of the angle of attack),
we find the corresponding lift coefficient:
b
C . - Ya,iC 2_
____ r ( oCfJb _ oCfJu ) dx
YR,IC - qoob - V rob J ox ox (6.1.6)
o
Let us consider the circulation of the velocity over a contour that
is a rectangle with the dimensions dx and dy and encloses an element
of the airfoil. In accordance with Fig. 6.1.1, the circulation is
dr =~ (V +00 OfPu ) dx _
ox
(V + ofPb
00
ox
) dx ,_ ( oCfJ]
-, oy
_ OCfJr ) dy
oy
where cp] and CPr are the velocity potentials on the left and right
surfaces, respectively.
Let us introduce the concept of the intensity of circulation (of a vor-
tex) determined by the derivative dr/dx = V (x). The magnitude of
this intensity is
(x) = oCfJu _ oCfJb , !:JL ( oCfJ) _ oCfJr )
V ox ax -r dx ay ay
Since for a thin airfoil, the angular coefficient dyldx is small, the
product of this coefficient and the difference of the vertical component
of the velocities is a second-order infinitesimal and, consequently,
V (x) = ocpu/ox - OCPb /ox (6.1. 7)
Hence, (6.1.6) can be written in the form
b
2 '
Cy".ic= Voob) V(x)dx (6.1.8)
o
We calculate the moment coefficient in a similar way:
b
Substitutions yield
n o o n
~o
"J
r
cot (8/2) sin 8 d8
cos 8 0 - cos 8
A
,,~n
+ : ""
sin (n8) sin 8 d8 = A
cos 80 - cos 8 t' J
r (e)
0
(6 1 17)
••
o n=l 0
Ao = - + n
~ ~ (8) d8
o
(6.1.19)
Integration yields
(6.1.22)
Hence, the lift lOoefficient depends on the first two coefficients of
the series. Introducing into (6.1.22) expression (6.1.19) and for-
mula (6.1.20) in which n = 1, we have
Jt
!I,
n~ ________________L -__~~~~__~X
Fig. 6.1.2
:vIean camber line of an airfoil
Integration yields
mza.lc = - (:11/2) (Ao + Al - A 2/2) (6.1.25}
where
n n
1
eo = --;t
(.
J Bd1-cos e) de; ~o=
1 r
-'4 J ~~ (1-cos 2e) de (6.1.30)
o 0
(6.1.31)
t Fig. 6.2.t
'" Conformal transformation of the flow past a round cylinder (a) into a flow past
a flat plate arranged (b) at right angles to the direction of the free-stream velo-
city, or (c) along the flow
the flow on the plane (J, then the function W = t [F «(J)l is the com-
plex potential for the flow over the plate.
Let us see how we can determine the complex potential for the
flow over a circular cylinder. For this purpose, we shall again revert
to the method of conformal transformation, using the known func-
tion of the complex potential for the flow over a plate arranged
along it. This function has the form
W = qJ + itIJ = -iV (x + iy) = iY(J (6.2.1)
We shall show that the conformal function
(J = ~ - R2 n (6.2.2)
):} transforms a segment of a straight line arranged along the- flow on
the plane (J into a circle on the plane ~ (Fig. 6.2.1c, a). Indeed, exam-
ination of (6.2.2) reveals that since for points satisfying the condi-
tions -a ~ y ~ a, x = 0, and a = 2R, we have (J = iy, then from
the solution of the quadratic equation ~2 - (J ~ - R2 = 0, we obtain
~= iy/2 -+ V R2_y2/4 = ~+ i1] (6.2.3)
Separating the real and imaginary parts, we obtain
11 =y/2, s= -+ V R2- y2/4
whence ~2 + 1]2 = R2.
Consequently, points on the circle in the plane ~ correspond to
points in the plane (J on the vertical segment. Substituting for (J in
(6.2.1) its value from (6.2.2), we obtain the complex potential of
the flow over a circular cylinder of radius R in a plane parallel flow
16-01715
242 Pt. I. Theory. Aerodynamics of an Airfoil end eWing
at the velocity V:
(6.2.4)
To obtain the complex potential for the flow over a plate arranged
across the flow (Fig. 6.2.1b), let us substitute for ~ in (6.2.4) the
following value obtained from the conformal formula a = ~ +
+ R2/ ~ transforming a circle of radius R (plane ~) into a segment of
a straight line across the flow (plane a):
- -fiI n ( -
211:
0 ±
2
V __ R2
0
4
2
) (6.3.6)
16*
244 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
-
y
v~
6=x+iy
a 0
x x
a a
Fig.6.l.t
Flow over a flat plate at an angle of attack
Fig. 6.3.2
Contour in the flow of an in- x
compressible fluid
To determine the force acting on the plate, we shall use the general
expression for the principal vector of the hydrodynamic pressure
forces applied to a stationary cylindrical body of an arbitrary shape
in the steady flow of an incompressible fluid. By analogy with the
complex velocity, let us introduce the concept of the complex force
Ra = X - iY, determining this force as the mirror reflection of
the principal vector Ra of the pressure forces with respect to the real
axis.
The vector Ra
being considered is determined by formulas (1.3.2)
and (1.3.3) in which the friction coefficient Cf.x is taken equal to zero:
- ,h'/"-.... /"-....
Ra=X-iY= -'YP[cos(n, x)-icos(n, Y)1ds
c
+
= - ~ p (sin e i cos e) ds = - i ~ pe- i8 ds (6.3.13)
C C
where n is the direction of an outward normal to contour C of the
body in the flow, and e is the angle between the element ds of the
contour and the x-axis (Fig. 6.3.2).
Since
da = dx + i dy = ds (cos e + i sin e) = e iB ds }
da = dx - i dy = ds (cos e- i sin e) = riB ds (6.3.14)
and the pressure is determined by the Bernoulli equation
P = poo +
pPoo/2 - pV2/2 = C 2 - pV2 /2 (6.3.15)
then
Ra = -iC2 ~ da+ i~ ~ V2 dcr= i~ ~ V2 da
c c c
Taking into account that by (6.3.14) we have da= e- 2iB da, and
also that in accordance with the condition of flow without separa-
Ch. 6. Airfoil and Finite-Span Wing in Incompressible Flow 247
tion the complex velocity at the point of the contour being consid-
ered is
v= Vx - iVy = V cos 8 - iV sin 8 = Ve- i8 (6.3.16)
we find
Ra=X-iY= ~ ~V2da (6.3.17)
c
Since for a potential flow, the complex velocity is V = dW/da, we
have
Ra = X - iY = i ~ ( ~~ ) da 2 (6.3.18)
c
Expression (6.3.18) is called the Zhukovsky-Chaplygin formula.
Integration in (6.3.18) can be performed over another contour en-
veloping the given contour C of the body in the now, for example,
over a circle K whose complex velocity in the plane a is written as
follows:
dW/da = V 00 - ir/(211a) A/a 2 (6.3.19) +
where Voo = V xoo - iVy 00= V oo e- i8 00, and A is a coefficient deter-
mined with the aid of an equation similar to (6.3.8).
The square of the complex velocity is
dW
( dcr
)2 = 1'2 _ ifVoo
00 ncr
+ 2AVoo-f2j(4n2)
cr 2
+ •••
Here the fIrst and third integrals equal zero. The integral ~da!(J
K
is evaluated with a view to the formula (J = x + iy = rei(jl (Fig.
6.3.2) and equals
~
G
dcr = In (J IK = In (rei(/» IK = 211i
K
Hence,
Ra=X-iY=ipvoor or X-iY=ipVoore- i800 (6.3.20)
where V 00 is the magnitude of the free-stream velocity.
248 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
YI
I
y
~
\I'
~ I
where X].e is the abscissa of the leading edge (Xl.e = -a), x is the
running abscissa of points of the plate, and V~ is the longitudinal
component of the disturbed velocity on the upper surface of a wing.
Formula (6.3.28') can be shown to be correct. For this purpose,.
we shall insert exprtlssion (6.3.26) into (6.3.28'):
c2 = lim [V;;"a 2
x~-a
(.!!:-=!..)
a+x
(x+a)]=2aa 2 V;;"
By introducing this value into (6.3.27), we obtain formula (6.3.25).
The conclusions on an incompressible circulatory-forward flow
over an airfoil in the form of a thin plate are used in studying the
aerodynamic characteristics of airfoils encountered in practice, and
also of finite-span wings in flows of an incompressible fluid or of a
compressible gas.
'fig.6.4.t
:System of equivalent vortices \ y
80und Free
for a rectangular wing lFortex vortices
flow. Hence, the theory of this flow is also a foundation of the aero-
dynamics of an infinite-span wing.
By formula (6.3.22), the lift force of a unit-span wing part is
y(l) = pex,voor (Fig. 6.4.1a). Consequently, there is a circulation
flow around the airfoil with the velocity circulation r. If the circu-
lation is clockwise, the velocities on the upper surface of the airfoil
are higher (a circulation flow having the same direction as the on-
-coming one is superposed on it), while on the bottom surface they
are lower (the circulation flow does not coincide with the direction
-of the oncoming flow). Therefore, in accordance with the Bernoulli
.equation, the pressure from above is lower than from below, and the
lift force is directed upward as shown in Fig. 6.4.1b.
Since by (2.7.8) and (2.7.8'), the circulation equals the vorticity
(vortex strength) x, the part of the wing can be replaced with an
·equivalent vortex of the indicated strength passing along its span.
N. Zhukovsky used the term bound to designate this vortex. Hence,
in the hydrodynamic sense, an infinite-span wing is equivalent to
a bound vortex.
Let us now consider an approximate scheme of the flow past a finite-
.span wing with a rectangular planform. As established by S. Chap-
lygin, a bound vortex near the side edges turns and is cast off the
wing in the form of a pair of vortex cores approximately coinciding
with the direction of the free-stream velocity. The distance e
{Fig. 6.4.1c) from a vortex core to the relevant side edge depends on
the geometry of the wing. Consequently, the hydrodynamic effect
-of a finite-span wing can be obtained by replacing it with a bound
vortex and a pair of free horseshoe vortices. This wing pattern is
,called Chaplygin's horseshoe one.
A vortex system equivalent to a finite-span wing induces addition-
.al velocities in the flow and thus causes downwash, which is a fea-
Ch. 6. Airfoil and Finite-Span Wing in Incompressible Flow 251
Rolled-up
[' l/Ortlces
z
Fig. 6.4.1
Vortex sheet and rolled-up vortices behind a wing
Yo F~
Fig. 6.4.3
Flow downwash at a wing and
the appearance of induced drag
- i/2
8 dz.
efficient
(6.4.3)
Ch. 6. Airfoil and Finite-Span Wing in Incompressible Flow 253
-L/2
.3
fig. 6.4.4
Replacement of a finite-span wing with a loaded line:
I-distribution of circulation; 2-Joaded line; 3-distribution of induced velocities or down-
wash angles
(6.4.5)
(i.e. the same for all sections) or may vary along the span if geo-
metric warp of the wing is present.
One of the most favoured ways of solving Eq. (6.4.10) is based on
expanding the required function r (z) into a trigonometric series
(the Glauert-Trefftz method)
00
ex. j =
-1
2rrl';,Sw
\'
J
r (z) [ \'
J
df (z') . _,dz'
dz' z -z
Jdz (6.4.15)
-1/2 -1/2
Introducing the value of f(z) and substituting "Aw for 12/ Sw, we
have
(6.4.16)
where the coefficient 6 taking into account the influence of the aspect
ratio on the drag depending on the lift is
m
6= ~ nA;JA; (6.4.17)
n=2
The coefficients 't and 6 for wings of various planforms can be de-
termined according to the data given in [13, 161.
The results of the general theory of a loaded line obtained can be
seen to be characterized by a comparative simplicity of the aerody-
namic relations, provide a clear notion of the physical phenomena
attending flow past finite-span wings, and allow one to reveal the
mechanism of formation of the lift force and induced drag. The ap-
plication of this theory, however, is limited to wings with a suffi-
ciently small sweep and a relatively large aspect ratio. In modern
aerodynamics, more accurate and more general solutions are worked
out. They are described in special literature.
At the same time, the development of ways of evaluating the aero-
'dynamic properties of wings by constructing approximate models
,of the flow over finite-span wings is of practical significance. Let us
consider one of them based on the representation of the aerodynamic
scheme of a wing in the form of a bound and a pair of free vortex
-cores. This representation is based on experimental data according
to which a vortex sheet is not stable and at a comparatively short
distance from the wing rolls up into two parallel vortex cores (see
Fig. 6.4.2).
The basic element of this problem is the finding of the distance 10
'between the free (rolled up) vortices. We proceed here from the fact
that for a wing with a span of l, the vortex pattern of the wing may
be replaced with a single horseshoe vortex with the constant circu-
lation r 0 corresponding to the root section. We also assume that the
hound vortex (the loaded line) passes through the aerodynamic cen-
ell. 6. Airfoil and Finite-Span Wing in Incompressible Flow 257
tre of the wing with the coordinate XFa' The magnitude of this circu-
lation can be determined by a coupling equation according to ,vhich
r0 = O.5c yao boV 00 (6.4.18)
where cllaO and bo are the lift coefficient and the section chord, re-
spectively.
A similar expression can be compiled for the mean circulation
over the span, the same as in the section with the chord bm:
rm = 0.5e", a bmV 00 (6.4.19)
em = - Cy~ bo In In +1 = _ ~ . _1_ In 10 + l 4 2
4nl..w bm 10 -1 nl..w 1 -1
41 0 0
(6 .. 3)
We can use this downwash angle to estimate the value of the in-
duced drag coefficient with the aid of (6.4.3). In a particular case
for an infinite-span wing ("Aw---+ 00), the downwash angle is absent
and, consequently, the induced drag vanishes.
17-0171:;
258 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
Conversion of Coefficients
Cy and ex, 1 from
One Wing Aspect Ratio
to Another
have r 0 = 0.5c yao b oV00'Now we can use Eq. (6.5.2) to find the
following expression for the chord in a wing section with the optimal
planform:
(6.5.3)
Let us consider a wing at the angle of attack a having section
profiles that are identical along the span. We shall assume that the
sections of the wing are arranged at the same local geometric angle
of attack determined by the given value of a. Since the downwash
angle does not change from one section to another for a wing with an
elliptical distribution of the circulation, the true angles of attack
Ch. 6. Airfoil and Finite-Span Wing in Incompressible Flow 2~9
Fig.6.5.t
Elliptical distribution of the
circulation and a geometric z
interpretation of the constants lj2
z and e L -1- ----,
law (6.5.2).
In the simplest case of a wing with a rectangular planform, b =
= const, this law of circulation distribution is ensured by only yary-
ing the values of c~a and at (by selecting the profiles and warping,
respectively). It must be borne in mind that for wings of a nOll-ellip-
tical planform, including rectangular ones, an elliptical distribution
of the circulation can be achieved only at a definite angle of attack,
and the distribution will be different if the value of the angle is
changed.
It is interesting to appraise by how much wings with a different
planform differ from their elliptical counterparts in their aerodynam-
ic properties. Table 6.5.1 gives the results of evaluating the coeffi-
260 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
Table 6.5.1
Piantorm ot wing
Elliptical o o
Trapezoidal (1]w = 2-3) o o
Rectangular (A.w = 5-8) 0.053 0.179
Rectangular with rounded tips o 0.147
Fig. 6.5.1
Conversion of aerodynamic coef-
ficients of a wing from one
aspect ratio to another
wash angle (e w 2 > eWI ) must have an increased setting angle (a2 >
> al) to obtain the same coefficient cYa '
The induced drag coefficients for two wings with the aspect ratios
AWl and AW2 and with identical coefficients cYa are determined by
the following formulas, respectively:
cx .11 = (cy/nA W1 ) (1 + <')1), C x .12 = (c ya /nAw 2) (1 + <')2)
Accordingly, for a wing with the aspect ratio AW2' the induced drag
coefficient is
(6.5.6)
This formula is used to convert the value of C x .l for a wing with
the aspect ratio AWl to its value for the aspect ratio Aw 2' If this new
ratio is smaller than the given one ("'w2 < AWl)' a large dOWllwash
appears (10 m 2> eml), and, consequently, the induced drag coefficient
grows (C X .12 > cx.H)'
Figure 6.5.2 shows graphically how the coefficients cYa and Cx.l
are converted from the aspect ratio AWl to "'w2 in accordance with
formulas (6.5.5) and (6.5.6). First the curves cYa = 11 (al) and cYa =
= CP1 (ex.u) are plotted for a wing with the given aspect ratio Awl'
Next, setting a number of values of cY ,we use the plot to determine
. a
the correspondlllg values of al and Cx .ll, and then calculate the an-
gles of attack a2 and the coefficients Cx .12 for the second wing by
formulas (6.5.5) and (6.5.6). The corresponding points are laid off,
and the curves cYa = h (a2) and cYa = CP2 (cx.d are plotted.
For wings with smaller aspect ratios (Aw2 < Awl)' these curves
will be to the right of those of the ratio "'wI because with the same
true angle of attack, the increase in the downwash angle ew 2 for the
second wing with a smaller Aw2 is compensated by the growth in the
setting angle of attack (a2 > a l ); in turn, an increased coefficient
cx .12 corresponds to the greater downwash angle ew 2, and this is ex-
actly what is shown in Fig. 6.5.2.
The plots in Fig. 6.5.2 can be used to convert cYa and cx.1 to a
wing with an infinitely large aspect ratio (airfoil). For this purpose,
262 Pt. I. Theory.• Aerodynamics of an Airfoil and a Wing
the second terms in the brackets in formulas (6.5.5) and (6.5.6) must
be taken equal to zero. The converted curves c ya = 12 (a2) and c.ya =
= fP2 (c x • 12) will occupy a position at the extreme left of the corre-
sponding curves CYa = 11 (al ) and c Ya = fP1 (cx.u).
Mean Aerodynamic Chord. When performing aerodynamic calcu-
lations of finite-span wings, the mean aerodynamic chord is selected
as the characteristic geometric dimension of the span. Such a chord
belongs to a conditional wing having a rectangular planform for
which the planform area, the aerodynamic force, and the pitching
moment are the same as for a wing of the given planform. The mean
aerodynamic chord allows us to compare the moment characteris-
tics of various wings with a varying chord along their span. One
of such characteristics is the pitching moment coefficient determined
as mzA = Mz/(bAqSw},' where b A is the mean aerodynamic chord.
The quantity mzA is sufficiently stable when the planform of a wing
and its dimensions change.
The value of the chord b A, and those of the coordinates of its lead-
ing point XA, YA (relative to coordinate axes passing through the
apex of the given wing), are usually determined approximately
assuming equality of the aerodynamic coefficients (of the moment,
drag, and lift) of a wing as a whole and its individual section (pro-
l
file). Accordingly, we have
1/2
2
I xb(,)d,
(6.5.7)
YA = s~ ~ yb (z) dz
o J
where x, yare the longitudinal and vertical coordinates of the leading
point of a section with the varying chord b (z). Evidently, knowing
XA and the equation of the leading edge of a given wing, we can de-
termine the lateral coordinate ZA of the mean aerodynamic chord.
Formulas (6.5.7) make it possible to calculate the magnitude of
the chord bAby numerical or graphical integration and determine
its position. The middle of bA coincides with the centre of gravity
of the wing area.
For a broad range of wings having a trapezoidal planform, we
can find analytical relations for bA, XA, YA, and ZA' For such wings
where rtw = bolb A is the taper ratio of the wing, Xo is the sweep
angle of the leading edge, ~J is the dihedral angle, and
Introduction of (6.5.8) into (6.5.7) yields
= 2zll.z
bA =4Sw [1-'Y)w/(ljw+ 1)2]!(31) }
XA = I (rtw + 2) tan xo/[6 (llw + 1)) (G.5.9)
YA = l ('Y)w + 2) tan '\/'1[6 (llw + 1»)
The lateral coordinate of the chord b A is found from the formula
ZA = XA /tan Xo- If a wing has no dihedral, the coordinate YA is
evidently zero.
---_._----_.
An Airfoil
7
in a Compressible Flow
+[v2-a~+(k-1)Voou] ay~ =0
a2 '
(7.1.3)
The second partial derivatives of the potential cp' with respect to
the coordinates x and yare first-order infini tesimals:
a2cp' au a2cp' au av a2cp' av
ox2 =Tx' ox ay = ay-= B;:' ay2 =Ty
With this in view, in (7.1.3), we can determine the group of terms
that are second- and third-order infinitesimals; disregarding them,
we obtain a linearized equation in the following form:
(7.1.4)
or
(1- M2)
00
aax2cp'
2
+ 02cp'
ay2 = 0 (7 .1 .4')
where Moo = V oo/a oo •
Let us consider the expression for the pressure in a linearized flow.
To do this, we shall use formula (3.6.26), which we shall write in
the form
pip = (1 - V2IV~aJJ k/(k-l) (1 - VSa,,1V~ax)-k/(k-l)
00
Relation Between
the Parameters of a Compressible
and Incompressible Fluid Flow
over a Thin Airfoil
mzaIe = Y ~J
t, px ix == ym,a (7.1.9)
where x= x/b.
Let us establish the relation between the configurations of airfoils
and the angles of attack. For this purpose, we shall first determine
Ch. 7. An Airfoil in a Compressible Flow 267
(a) ( b)
fig. 7.1.1
Airfoils in a nearly uniform incompressible (a) and compressible (b) flows:
A and Ao-stagnation points
where b' is the distance to the trailing edge of the airfoil, and x is
the horizontal coordinate (Fig. 7.1.1).
Hence, by (7.1.11), we have
al c = a')'/V 1 - M~ (7.1.12)
Let us assume that the arbitrary parameter ')' = 1. Therefore,
P=Plc' Y=YoV1-M~, a=alcV1-M~}
(7.1.13)
cUa = Cllalc, m.a = mZaic
268 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
(a) (b)
!d f5
x X'
FIg. 7.1.1
parameters from one number M 001 > 0 to another one M 002 > o.
The Khristianovich method is suitable provided that the velocity
is subsonic over the entire airfoil. This condition is satisfied if the
Mach number of the oncoming flow is less than the critical value
Moo, cr' Consequently, before performing calculations, one mllst find
this critical value and determine the number Moo < Moo, cr for
which calculations are possible. The critical number M "",cr can also
be found by the Khristianovich method.
compressible flow corresponding to the value A00, and for the chosen
value of Pic
from the Bernoulli equation
(7.2.2)
Knowing A, we use Table 7.2.1 to find the local true speed ratio A
of the compressible flow, and we calculate the pressure fJ by for-
mula (3.6.26) in which it is necessary to assume- tha V2/V~ax =
= [(k - 1)/(k+ 1)1 A2. We determine the pressure coefficient by the-
"--_. __._-------
':.272 Pt. I. Theory. Aerodynamics of an Airfoil and fI Wing
x
'fig. 7.2.2
Nature of the pressure distri-
bution on one side of an airfoil
.at different values of Moo
while llsing Table 7.2.1 and the value of "\00' we can determine the
critical speed ratio A cr of the compressible f1ow. The corresponding
00,
k+l k-l 2
M ex;. cr = Aoo , cr ( -2- - -2- Aoo , cr
)-1/2 (7.2.6)
I.Z
.1.1
t.O
0.9 ~
0.8 \
0.7 .~
0.5 r'\.
Fig. 7.2.3 0.5 "t-... ...........
A Khristianovich curve for
determining ,the cri tical Mach D.4
~0.5 -1.0 -1.5
r--
Pic,min
number
Aerodynamic Coefficients
B
(aJ
. I
P,-
P,-
v!co<af_
I>Mc£M... cr
o~------------~
l'
Fig. 7.3.1
Subsonic flow over an airfoil at a supercritical velocity:
a-diagram of the flow used for calculations when a local normal sllOck is present; b-
distribution of th(' pressure over the airfoil when a local A-shaped shock forms
V 2 , P2, P2, and M 2 • Let us introduce two control surfaces I-I and
II-II to the left and right of the airfoil at a sufficiently large distance
from it; let the parameters of the gas along the left plane be VI 'X"
PI"'" and PIOO, and along the right plane be V 2oo , P2oo, and P2oo'
Using the theorem of the change in the momentum of the mass of
a gas when flowing through the control surfaces, we obtain
00 00 00
where PI and P2 oc are the forces acting on the left and "right sur-
00
faces, Xw is the force of the wave drag with which the airfoil acts on
the flow, and dm ioo = PI VI 00 dYl <X>l dm200 = P2 V
00 dY200 are 00 200
the rates of flow of the gas along the stream filaments that according
to the condition of the constancy of the flow rate are equal, i.e.
(7.3.2)
Let us assume (see [161) that levelling out of the velocities occurs
behind the wing at a large distance from it, i.e. V 200 -+ V p ,' Accord-
ingly, and by (7.3.2), we have
r
-00
V 200dm 2 oo- r
-00
V 100dm 1° o = r Pi""Vioo(V2oo-Vioo)dYioo=O
Consequently,
00
We can assume that for stream filaments not intersecting the shock,
P2 oc = PI 00 i.e. the pressure at a large distance from the airfoil be-
hind it is restored to the value of the free-stream pressure. For the
filaments that pass through the shock, P200 < PIoo, Indeed, since
PiCO = Po (1- Vioo!V~ax)k/(k-1) }
(7.3.4)
P2 00 = p~ (1- V~oo!V~ax)k/(k-t) -= p~ (1- Vico!V~ax)k/(k-1)
where p~ < Po, we have
P200/Ploo = p~/Po = Vo < 1 (7.3.5)
According to (7.3.5), formula (7.3.3) can be transformed by taking
into account the flow rate equation PI 00 VI 00 dYIoo = PI VI ds, in
accordance with which
dYIoo = dy = (PI VI/PI oc VI 00) ds
where s is the length of the shock.
Ch. 7. An Airfoil in a Compressible Flow 277
cx•w =
2X w
kPooM~b =
2
Hl~b
r PIV'1
J PIOOV
(l- v o)ds (7.3.6')
IOO
o
Let us expand the function Vo in (4.3.20) into a series in powers of
n = MI - 1:
Vo (n) = \'0 (0) + ( ddvon ')
n=O
n+ 21 ( dd
2V20
n
)
n=O
n2
+ 6"1 ( (h;3
d3vo)
n=O n
3 '
--t- ••• (7.3.7)
In expression (7.3.7), the quantity Vo (0) = 1 because at illi = 1
(n = 0), the shock transforms into a wave of an infmitesimal strength,
and the pressure P~ = Po.
Using (4.3.13), (4.3.15), and (4.3.20), we can also show that
( ~)
dn n=O
= (~)
DM 1 .11 ,=1
=0 and ( d2~,o)
dn" n=O
= (d2V~)
dM 1 .1I,=1
=0
CXa
0.6 V_(Moo)
)IE .~
0.5
0.4-
0.3 V
0.2 J
V
0.1
..-'
Fig. 7.3.2
Drag of an airfoil in a nearly o
sonic flow 0.2 0.4- 0.5 O.BM_
~
y1 y
" \
/
V~(M~)
• 0
x
I
-i
fig. 7.4.t
Supersonic flow over a thin plate:
I-expansion fan; 2-shock
('J
~I
: /
I
I /
/ / /
~/;i.~.
I
fig. 7.4.2
Prandtl-l\leyer flow:
a-physical plane; b-hodograph plane; c-diagram of a nearly uniform flow; l-exJlan-
sion fan: 2-epicycloid
of attack a of the plate (see Fig. 7.4.1). The disturbed region of the
expanding now is limited at the left-hand side by Mach line OD
inclined to the free-stream yelocity vector V 00 at the angle f.t DC =
= sin _1 (11M 00)' \>;here Moo is the Mach number of the undisturbed
now. The expansion process terminates on Mach line OE inclined to
280 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
the disturbed velocity vector Voe at the angle ~Loe = sin _1 (1IMod
determined from the Mach number 0 the disturbed flow along plane
OG. The change in the direction of the flow between Mach lines OD
and OE can be represented as a consecutive set of deflections of the
streamlines through the small angles tlp. A straight Mach line issu-
ing from point 0 corresponds to each of these deflections indicating
the formation of an additional disturbance.
Hence, the turning flow is ftIled with an infinite multitllde of
Mach lines forming a "fan" of disturbance lines that characterizes a
centered expansion wave. This centered wave, sometimes called a
Prandtl-Meyer fan, is defined by straight Mach lines along each of
which the flow parameters are constant, and this is why it belongs to
the class of simple expansion waves.
The problem on the disturbed motion of a gas near an obtuse
angle, which is associated with the formation of a centered expansion
wave, can be solved according to the method of characteristics.
Point F' on an epicycloid -a characteristic in a hodograph of the
same family-corresponds to point F of intersection of a streamline
belonging to the plape parallel oncoming flow (the inclination of a
streamline at this pomt is P = 0) with characteristic OD in a rhys""\
ical plane. To be specific, we can relate each of these characteristics
to those of the fIrst family. The equation P = ffi + ~1 is used for a
characteristic of this family in the hodograph. Since we have assumed
that P = 0, the constant ~1 = -ffi 00 (Moo), where the angle ffi 00 is
found from (5.3.30) according to the known number Moo. Consequent-
ly, the equation for the characteristic has the form ~ = ffi - ffi 00,
whence
ffi=P+ffioo (7.4.1)
By setting the inclination of a streamline on the small angle
P = tlP, we can calculate the corresponding angle ffi = tlp + ffi 00
and find the number M on the neighbouring Mach line inclined to
the new direction of a streamline at the angle ~ = sin _1 (11M).
The Mach number on plane OG with an inclination of P = Poe = a,
i.e. on the upper side of the plate, is determined according to the
angle
ffioe = Poe + ffi 00 (7.4.2)
The found value of the local number Moe makes it possible to de-
termine the Mach angle ~oe = sin _1 (1IM oc). A graphical solution
of the problem on the Prandtl-Meyer flow is shown in Fig. 7.4.2b.
The coordinate of point G' of intersection of the epicycloid with
straight line O'G' parallel to plane OG determines the speed ratio
Aoe of the disturbed flow near plane OG. The point G of intersection
of a streamline with characteristic OE corresponds to point G' in a
physical plane.
Ch. 7. An Airfoil in a Compressible Flow '28t
and introduce the value of a 2 from (7.1.2') into this formula, we obtain
~
a2
= M2 = M2 (1
00
-I-
I VOC"
r
~) 1 - (k -1) u V
.
00 ]
a~
- 1 (7.4.7)
Consequently,
u/V 00 = +~/VM;" - 1 (7.4.8')
Inserting the value of u from (7.4.8') into (6.1.5), we obtain the
pressure coefficient
(7.4.9)
Since we are considering an expanding flow for which p < 0 and
are having in view that the magnitude of the angle ~ is being found,
we must take the minus sign in formula (7.4.9). Accordingly, on the
upper side of the plate inclined at a small angle of attack ~ = a,
the pressure coefficient is
Pu ,,= Poe.u = -2a/V JJl;" - 1 (7.4.10)
Let us consider the bottom side of the plate. The flow over this
side (see Fig. 7.4.1) is attended by the formation of shock OE issuing
from a point on the leading edge and, consequently, by compression
of the flow. To determine the angle 8s ,OE of inclination of the shock,
we should lise formula (4.3.25) in which we must assume that Ml =
= Moo and ~s = a. According to the found value of 8 s 'OE, \\'e find
the Mach number M2 = Moe,b on the bottom side by (4.3.19) or
(4.3.19').
vVhen determining the nature of the flow in the region behind
point C on the trailing edge, we can proceed from the following
-considerations. On the upper side of the plate, the number Moe, u
ahead of shock CD is larger than the number Moo ahead of shock OE
Ch. 7. An Airfoil in a Compressible Flow 283
C ya la=4/VM~-1 (7.4.20)
cx /a 2 = 4/V M~ - 1 (7.4.21)
m za /a= -2/VM~-1 (7.4.22)
In the given case. the number Moo is the similarity criterion.
'Vhen its value is retained, and regardless of the value of the angle
of attack, the corresponding values of -pia, c ya la, C.y ala 2 , and m za la
o
Fig. 7.5.1
Supersonic flow over a sharp-nosed airfoil:
I-expansion fan; 2-Mach Jines
Fig. 7.5.2
Supersonic flow over the bottom side of an airfoil with the formation of a shock:
I-straight part of the contour of the airfoil in the flow; II-curved part of the airfOil con-
tour; III-curved part of the shock; IV-straight part of the shock
Calculation of the flow over the bottom side of the airfoil (Fig. 7.5.2)
begins with determination of the gas parameters at point O-directly
behind the shock. For this purpose, nsing formllla (4.3.25) and the
values Ml = Moo and ~c = ex + ~O,l" we calCldate the shock angle
8 s .0 . We fmd the number Mo,b = M z at point 0 from (4.3. Hl) or
(4.3.19'). We may assllme that this nnmber remains constant within
a very small neighbourhood of point 0 on straight line element OD
of the contour. Straight element OJ of an oblique shock corresponds
to OD. Its length is determined as the distance from point 0 to point J
that is on the intersection of the shock with a fIrst family characteris-
tic issuing from point D.
The flow behind a straight shock is vortex-free, con~equently there
is an isentropic flow over part of the contour behin([ point D. To deter-
mine the velocity of such a flow at point F, we ~hallllse Eq. (7.4.1),
from which we find ffiF = ffiD - (~D - ~F), where ffiD i~ the value
of the angle ffi calculated by formula (5.3.30) for the number M on
part OD of the con tour. The values of the angles ~D and ~F are deter-
mined with a view to the sign (in the given case the angles ~D and ~F
are negative on the leading part of the contour).
The flow near part DF of the contour can be considered as a Prandtl-
Meyer flow, therefore disturbance line F-J-3 iS~lles from point F as
from a perturbation source. It intersects the conlil111atioll of the
normal shock at point 3 and curves it, as a result of which the actual
direction of the shock is determined by point 3 of intersection of the
shock and the characteristic.
Downstream, curving of the shock is due to its interaction with
the characteristics issuing from points G, H, K, etc. Curving of the
shock causes a vortex flow to form for whose calcnlation we must use
relations on the characteristics for a non-isentropic plane flow. The
..... _---._------
288 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
Fig. 7.5.3
Supersonic isentropic' flow over a curved sharp-nosed airfoil:
I-contour of the body in the flow; 2-second family characteristic
iN = 0 (7.5.18)
p~ = 2 (a - ~d2/(1 - 8) (7.5.19)
With a zero angle of atta(;k, formulas (7.5.16), (7.5.17), and
(7.5.19) acquire the following form:
PN/P& = 1/(1-8) - V 1/(1- 8)2 + 4/ Kk (7.5.16')
pclPL = 1/(1- 8) + V 1/(1-8)2+ 4KL (7.5.17')
PL = 2~[I(1 - 8) (7.5.19')
In formulas (7.5.16') and (7.5.17'), KN = M ooPN and KL =
= M OO~L
Ch. 7. An Airfoil in a Compressible Flow 293
X
r-
= goeSr J P cos (n,
/"x) S;'
dS r-
Y = - goeSr .1 p cos (n, y) s;
/" dS
(s) (s)
294 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
Fig. 7.5.4
Aerodynamic forces for an airfoil in a body axis and flight path coordinate
systems
Fig. 7.5.5 Xp
Determination of the moment
of the forces for an airfoil b
With flight angles of attack not exceeding the values of a :::::; 10-12,
we have
(7.5.25')
For the coefficient of the moment about the leading edge of the
airfoil due to the pressure force (Fig. 7.5.5), by analogy with (1.3.6),
we derive the formula
m z = q:!szr b = qOO~ rb (~ x dY - ~ Y dX )
1 { (' -
=b
as - Jr -PY cos (n,/"x) s;
J px cos (n,/"y) s; as}
(8) (8)
or
1 1
m Z = qoo
-b
Mz
2 ~ (Pb - Pu) Xdx - ~ [PbYb ( ~~ )b - PuYu ( :; )
o 0
J dx
(7.5.26)
where Yb = Yb 1b and Yu = y,Jb.
'Ve determine the coefficient of the centre of pressure for the
condition that the point of application of the resultant of the aero-
dynamic forces is on the chord of the airfoil. If its coordinate is x p ,
we have
1 1
Xp m
~ (Pb-Pu)X ix+ ~ [PbYb (~; )b -PuYu (~~
o
)J a;
cp = - = - -z 0
b Cy
(7.5.27)
For thin airfoils, we may disregard the second integral on the
right-hand side of (7.5.26) and in the numerator of (7.5.27). Accord-
296 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
ingly,
1
Cx =
a
2c~ + 2C1 Jr ~2 dx
Cl
(7.5.32)
o
Equation (7.5.32) determines the relation between the drag and
lift coefficients-what is called the polar of an airfoil.
Ch. 7. An Airfoil in a Compressible Flow '297
!J
Fig. 7.5.6
Wedge-shaped airfoil b
z
m = -0.5 (Pb - Pu) - 0.5 (P~ tan 2 ~b - Pu tan 2 ~u) (7.5.41)
while the coefficient of the centre of pressure by (7.5.27) is
(7.5.42)
For a symmetric airfoil, we have
where the angle ~ is chosen with a minus sign (for the bottom side).
The use of formulas (7.5.37), (7.5.38) allows us to determine the
aerodynamic coefficients (7.5.39)-(7 ..5.43) corresponding to hyper-
sonic velocities of a thin airfoil at a small angle of attack.
is parallel to the initial velocity -VI (see Fig. 7.6.1) and the rectangu-
lar wing is turned through the angle x = ~ (Fig. 7.6.2). Such a
wing, which is also a sideslipping one, is usually called an infinite-
span swept wing. The angle x is determined as the sweep angle rela-
tive to the leading edge.
Let us consider some features of the flow over sideslipping wings.
The flow about such wings can be divided into two flows: a longi-
tudinal one (along the span of the wing) characterized by the velocity
V,; = V sin x parallel to the leading edge, and a lateral one
00 00
(b)
-Vz
Fig.7.6.t
Motion of a wing with sideslipping:
a-longitudinal motion of an infmite-span rectangular wing; b-lateral motion; c-resul-
tant motion at the sideslip angle,
Fig. 7.6.2
Sideslipping wing:
I-wing surface; 2-nirfoil in a section along a normal; a-airfoil in a section along the now
(a)
v~
Fig. 7.6.3
Finite-span swept wing:
a-designation of the sweep angles: b-subsonle flow over a swept wing
Aerodynamic Characteristics
01 • SideSlipping Wing Airfoil
:x>;;r/2-;u~
Fig. 7.6.4
Sideslipping wing with subsonic (a) and supersonic (b) leading edges
V cos x (or the Mach number from Moo to Moo cos x). Here, natural-
00
inequality x > :rt/2 - !loo, according to which cos x < sin !l and, 00
to the leading edge is subsonic. Hence, the flow over the sections of
a sideslipping wing is subsonic in its nature. I n the case being con-
sidered, the swept edge is called sullsonic (Fig. 7.6.1a).
At increased flow velocities, the normal velocity component may
become higher than the speed of sound (Vn > a = V sin !l
00 00 00 00)'
(7.6.9)
mZax = -2a n cos2 xiV M2"., cos2 x-1
Upon analysing these relations, we can establish a feature of
swept wings consisting in that in comparison with straight ones
(x = 0), the lift force and drag coefficients, and also the coefficient
of the longitudinal moment of airfoils (in their magnitude) are
smaller at identical angles of attack an along a normal to the leading
edge. The physical explanation is that in flow over a sideslipping
wing not the total velocity head q 00 = O.Sp 00 V?x, is realized 1 but only
a part of it, qn = q 00 cos 2 x, and flow in the direction of the oncoming
stream occurs at a smaller angle of attack than in t.he absence of
sideslip (a < an, an = a/cos x).
Subsonic Leading Edge. The flow over sections ('.orresponding to
the motion of a straight wing with the number Mn 00 < 1 is invest i-
Chi 7. An Airfoil in a Compressible Flow 305
If Mn < Moo c r, then shocks and, therefore, wave drag are absent.
00
20 -01715
_.,---------
306 Pt. I. Theory. Aerodynamics of an Airfoil end a Wing
Fig. 7.6.5
Suction force of a sideslipping
wing
i.e.
(7.6.14)
Therefore, the pressure coefficient for a compressible flow is deter-
mined by the Prandtl-Glauert formula:
(7.6.15)
The suction force T in its physical nature is a force produced by the
action of the normal stress (pressure) and at low angles of attack is
determined from the condition T ~ aY a • The corresponding suction
Ch. 7. An Airfoil in a Compressible Flow 307
or
dT = dT lc
But since dz = dz lc , then
dTldz = dT ic1dzic (7.6.17)
The right-hand side of Eq. (7.6.17) corresponding to an incompres-
sible flow is determined by formulas (7.6.11) and (7.6.12):
dT Ic1dzlc = np ooCI~ V 1 + tan 2
"ic
where
Crc= lim (ule (Xtc-Xs.e.lc)]
Xlc~Xs.e. Ic
In accordance with (7.6.17), (7.6.13), and (7.6.14), we obtain
cre= X-Xs.e
lim [u 2
(x-x s .e )]V1-M;'; Vl+tan 2 "ic
tan 2 x
I-J1~
Consequen tly,
dTldz = np ooc 2 V 1 + tan 2
" - M~ (7.6.18)
where
C2 = lim (u 2 (x - x s.e)] (7.6.19)
X-+Xs.e
20*
--_._-----
8
A Wing in a Supersonic Flow
V
x
oV x
ox
+V Y
8V x . V 8V x -
P iJx
iJy -t- Z iJz -
_..!...!.E.. 1~
8V 8V
V - - Y +V _ _ oV
11 +V __ 1 iJ
11 _ _ _ .--"!!....
x iJx Y iJy . Z iJz - P iJy (8.1.3)
V 8V z +V iJV z +V iJV z = -~.~
I
x iJx Y iJy Z iJz P iJz J
Ch. 8. A Wing in a Supersonic Flow 309
P ( aVx
ax
+ aV"
ay
+ aV8zz )+Vx ~+V~.
()x 8y
_LV y 1 Z
~=O
8z (8.1.4)
(V2.x _ a2 ) ~
ox
~T"
+ (V2Y _ a2)
8V
_ ! I _.I-
ay'
(V:; _ a2) _V_z
-
~lT
Dz
+V :x;
V (8V x
Y ay
_.I-
1
av II
ax
) _I
I
F.V (av x
x Z az
..Lav z
' ax
)
(8.1.5)
vx uy + 2V .xV z --;;-;;--
+ 2V xV"" -:;--::;- vx vz + 2VyV z --;;-;;--
vy vz = 0 (8.1.6)
(a) (6)
y
!I
fig. 8.1.1
Thin wing in a linearized flow
Boundary Conditions
Here
8cp
8;""=
V 1
oo+u=loc+~,
ocr;' oCf.' = V= a<p' oq acp'
jjy oy , Tz=- W= - 0 -
the lift coefficient for an elementary surface equals the value c~a =
= (2/V""b) r (z), whence the circulation in the given section is
r (z) = 0.5c~ V""b (8.1.13)
a
According to the coupling equation (8.1.13) [see (6.4.8)1, upon
moving to a neighbouring section with a different lift coefficient,
the velocity circulation also changes. This change is
dr (z) = (dr /dz) dz = 0.5 V "" [d (c~a b)/dz] dz (8.1.14)
In accordance with the vortex model of a wing treated in Sec. 6.4,
an elementary bound vortex belonging only to the section being con-
----------.
312 Pt. I. Theory. Aerodynamics of an Airfoil and II Wing
components V n are the same, i.e. o<p' (x, -y, z) ()y = o<p' (x, +y, z)/
lay. Consequently, the additional potential fp' is an odd function
relative to the coordinate y, i.e.
<p' (x, -y, z) = -cp' (x, +y, z) (S.1.1S}
Accordingly, the partial derivative a<p'lox on the bottom side of
the vortex sheet equals the value -fJcp'lux on the upper side. But
the equality of the derivatives ocp'lax was established from the con-
dition of pressure continuity. This equality can he observed only
when
ocp'ox = 0 (S.1.19)
on the vortex sheet.
4. To estahlish the last boundary condition, let us consider the-
disturbed regions 8 rand 8 1 (Fig. 8.1.1) that are cut off from.the plane
y = 0 by the Mach wave surface and located outside the wing and
the vortex sheet. Over these regions of the plane y = 0 and within
the limits of the wave zone, the flow is continuolls, therefore the-
potential cp' here is also a continuous function. At the same time,
taking into account that by (S.1.18) the function cp' is odd, we should
adopt the following equation for the plane y = 0:
<p' (x, 0, z) = 0 (S.1.20).
The solution of Eq. (S.1.7) for the additional potential <p' must
correspond to the considered boundary conditions. This solution
can be obtained for a wing with a given planform by summation of
two potentials: the first, <p~, is found for an idealized flat wing 1
(Fig. 8.1.2) of the same planform as the given one, but with a sym-
metric airfoil, and at a zero angle of attack (a = 0). The second
potential, cp~, is evaluated for a different idealized cambered wing 2
of zero thickness, but for the given angle of attack a.
The surface of an idealized symmetric airfoil wing can be given by
the equation
y = +0.5 Uu - fh) (S.1.21)
and of a zero-thickness airfoil wing, by the equation of the surface
of the mean camber lines of the airfoil
y = 0.5 Uu + fh) (S.1.22}
Hence, the total potential of the given wing is
q!' = (P~ + fP;
314 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
Mean c{)'moe'"
[;/li81l
CfI rlol! (2) line
~~
~1 (l'v1J' )1 b ,
~1£:£
.,~ 'I( ... ,
(1) y=D.5(fu-f,;)
-----~~
Fig. 8.1.1
Linearized supersonic flow over a fmite thickness airfoil at an angle of attack ex:
I - a = 0, symmetric airfoil, and giwn thickness distribution; 2-a =1= 0, zero thickness
(the airfoil coincides with the mean camber line); 3-a = 0, zero thickness (the airfoil coin-
cides with the mean camber line); 4-a
the chord)
'*
0, zero thickness (the plate airfoil coincides with
(8.1.27)
Ch. 8. A Wing in a Supersonic Flow 315
Features of Supersonic
Flow over Wings
Fig. 8.1.3
Supersonic flow over a wing:
a-wing with subsonic rdgrs; b-wing with supersonic edges
(Fig. 8.1.3a) and behind them (Fig. 8.1.3b). The arrangement of the
Mach lines at a given wing planform depends on the number Moo.
In the first case, the number Moo is smaller than in the second one t
and the angle of inclination of the Mach line ft > nl2 - x (x is the
00
.n < 1. For subsonic and sonic leading edges, we have n > 1 and
.n = 1. respectively, because in the [1rst case cot [100 < tan x and
in the second one cot [100 = tan x.
By analogy with the leading edges, we can introduce the concept
of subsonic, sonie, and sllpersonic tips (side edges) and trailing edges
of a wing. Tip CD with an angle of inclinaion Yt to the direction of
the free-stream velocity smaller than the Mach angle (Fig. 8.1.3a) is
called subsonic. The velocity component normal to a tip and equal to
Vn 00 = V 00 sin Yt is lower than the speed of sOllnd in the given case.
Indeed, since a oo = V sin ~too and floo > Yb we have V noo < a oo •
00
where qie is the flow rate of the source, i.e. the volume of fluid flowing
out of the source in unit time.
The method of sources deals not with individual point sources,
but with sources continuously distributed over a part of a plane.
usually the coordinate plane xOz.
Let dqlc be the elementary volume flow rate of the fluid produced
by the sources on the small area da Ie = d~ Ie d~ Ic in the plane xOz.
Hence the derivative dqleldalc = Qle' known as the density (or
intensity) of source distribution, determines the strength of the
sources per unit area.
If + v is the vertical com ponen t of the velocity on the area da Ie
(the plus sign signifies that the fluid is discharged upward from the
sources, and the minus sign-downward), the elementary volume
flow rate is evidently dqlc = 2v dale and, therefore,
QiC = 2v (8.2.2)
The following potential corresponds to an elementary source:
dcpic = - QIC da lc /(4n V xfc + Yfc + zrc) (8.2.3)
Using this expression, we can obtain a relation for the elementary
potential of a source in a subsonic compressible flow. To do this, let
us consider Eq. (8.1.7) and introduce the new variables
x!c=x/V1-M;" YIC=Y' Zic=Z (8.2.4)
With the aid of these variables, Eq. (8.1.7) is transformed as follows
o2cp'Ioxf c+ o2cp'IoYrc +o2cp'Iozrc = 0 (8.2.5)
This expression coincides with the continuity equation (2.4.8') for
an incompressible flow. Consequently, the problem on the compres-
sible disturbed flow in the coordinates x, y, and z can be reduced to
the problem on an incompressible disturbed flow in the coordinates
Xl c ' Ylc, and Zlc, both systems of coordinates being related by con-
ditions (8.2.4).
Accordingly, we can go over from potential (8.2.3) for an elemen-
tary source of an incompressible fluid to the relevant potential for a
subsonic source of a compressible fluid. To do this, we shall find
the relation between the small area da Ic on the plane x IcOZ Ic in the
incompressible flow and the area da on a corresponding plane in the
compressible flow. Using (8.2.4) (with the substitution of ~ for Xi c
and ~ for ZIC) and the expression dalc = d~lc d~ic' we find that
daic = d~ d~ (1/V 1 - M;,), whence, taking into account that
d~ d~ = da, we find
(8.2.6)
Let us transform the expression for Qlc in (8.2.3). Thecomponent
Vic = ocp'loy Ic' or with a view to (8.2.4), Vic = ocp'loy. It thus follows
Ch. 8. A Wing in a Supersonic Flow 319>
Fig.8.1.t
Disturbed flow due to a supersonic source:
at the right-a Mach (disturbance) cone in a real supersonic flow, at the left-an "inverted'
Mach cone"
1D_=-LLLl.l..U.J.il.LLLLl.l..U.J.JId="""oA (X, z, Yf 0)
'fig. 8.2.2
Region of influence of supersonic sources
and not half of it, is realized in the flow conftned within the Mach
cone. Accordingly,
dcp' = - Q da/[2n V x 2_ a'2 (y'!. + Z2)] (8.2.10)
In the considered case, the elementary area da = d~ d~ with the
sources is at the origin of coordinates. If it is displaced with respect
to the origin of coordinates to a point with the coordinates x = G
and z = ~, Eq. (8.2.10) has the form
dcp' = - Q da/{2n V(X-S)2_ a '2l y 2 + (Z_~)2J) (8.2.11)
When studying the flow over a wing, its surface is replaced with
a system of disturbed sources. To obtain the potential due to these
sources at arbitrary point A (x, y, z) (Fig. 8.2.2), we must integrate
(8.2.11) over the region a in which only part of the sources are located.
Each of these sources influences point A (x, y, z) if it is inside a Mach
cone with its vertex in the source. Hence, the region affected by the
sources (the region of integration) is in the zone of intersection with
the wing surface of an "inverted Mach cone" with its vertex at point
A (x, y, z) being considered.
In a simpler case, point A and the source are located, as can be
seen from Fig. 8.2.2, in the same plane y = O. In this case, the
affected zone coincides with the region of intersection of the wing
and the Mach lines issuing from point M (x, z), while the region of
integration a is on the wing and is the intersection of the wing sur-
face with the inverted plane Mach wave having its vertex at point
A (x, z).
Having determined the integration region a, we c,an evaluate the
total potential at point A (x, y, z):
(8.2.12)
Ch. 8. A Wing ,in a Supersonic Flow 321
(8.2.13)
!:J
~ /O~~~~~__~__~~~_
Fig. 8.3.1
Triangular wing with a subsonio leading edge
where
(8.3.3)
s
On section OAPH, integration with respect to for each value of
~ = ~1 must be performed from S = sc
= ~ tan x to S = SD =
= Xp - a' (zp - ~), and integration with respect to ~ from 0 to Zp.
On section HPB, integration with respect to S, for which the value
of \; = ~2' must be performed from S = SF = ~ tan x to S = SE =
= Xp - a' (~ - zp), and integration with respect to ~ from Zp to
Ch. 8. A Wing in a Supersonic Flow 323
(P' = - 7 [ ) d~AV' \
J
o 6 tan x
ZE xp-cx.'(\;-zp)
+ i d~ ~ f (~, ~) d~ J (8.3.4)
Zp \; tan x
Taking into account that tan x > a' (cot !l 00), we find as a result
of integration that
AVoo h- xptanx-Cl.'2Zp
u= - Jt ytan 2 X-Cl.'2
c o 1sCI., (xp-Zp tan x ) (8.3.8)
Fig. 8.3.1
Angular parameters for a trian-
gular wing x
Now let us consider point N outside of the wing between the Mach
line OK' and the x-axis (see Fig. 8.3.1) and calculate for it the
velocity that is induced by the sources distributed over the wing
surface. To do this, we shall use formula (8.3.1) determining the
velocity potential. Taking into account that the action of the sources
on point N is confined by the region 0' = OLJ, we obtain the expres-
sion
cp' = - A: 00 .\ ~ f (,£, s) d,£ ds (8.3.11)
OLJ
cp' = - A: Jds
00 ~ f (,£, s) d~ (8.3.12)
o b tan x
where ZJ = (XN + a'zN)/(a' +
tan x).
By integrating and using the main value of the integral, we obtain
,_ AV00
cp - n
1cos
ZJ
I -1
1
XN - ~ tan x
a , IzN-_t I
\,
dr
o
This expression is similar to (8.3.6) with the difference that the
coordinate ZJ is taken as the upper limit of the integral. By cal-
culating the derivative ocp'lox and integrating, we obtain relation
(8.3.9) for the component of the additional velocity. We must assume
in it that 0' < 0 because the coordinate zN is negative. In calculations,
the coordinate ZN may be assumed to be positive, and, consequently,
0' > O. If we take the magnitudes of tan x, then to determine the
Ch. 8. A Wing in a Supersonic Flow 325
y
K
z
K'
Fig. S.3.3
Influence of sources on the velocity outside a wing
(8.3.13)
cr ' = - A~ 00 I d~ I
o 6 tan x
/it ~) ds (8.3.15)
l/X
\
\
\
\
.-A
Fig. 8.3.4 Mach line \
\
Pressure field for a triangular \
where n>u.
We use the value of this velocity to find the pressure coefficient:
2u 2A. n 2 -a
p= - roo = net'
1
y'n2-i cosh- n(o--i) (8.3.17)
Figure 8.3.4 shows the fIeld of pressures for a wing with a triangular
planform having a subsonic leading edge. The pressure coefficient
along the Mach lines is zero. On the leading edge, the theoretical
pressure coefficient equals infinity. The physically possible pressure
can be considered to be sufficiently high, corresponding to the stag-
nation pressure at a subsonic velocity whose direction coincides with
a normal to the leading edge.
U = - net' A.Voo [
y n 2-1,cosh -1 n 2 -0-
n (i -0- ) + cos h-1 n 2 +a
n (i + 0- ) J
Ch. 8. A Wing in a Supersonic Flow 327
y z
Fig. 8.].$
Triangular wing symmetric about the x-axis with subsonic leading edges
(8.3.19)
Point L located between the leading edge and the Mach wave
(Fig. 8.3.5) is influenced by the sources distributed on section OUG'
of the wing. The velocity induced by these sources can be calculated
as the sum of the velocities induced by the sources in region OUG
{formula (8.3.16)1 and by the sources distributed in triangle OGG'
[formula (8.3.13)].
Consequently,
u= _ "Yoo [cosh- 1 n2 - a cosh- 1 n·+ a ] +
na' Vn 2 -1 II (a-I) II (1+a)
or after transformations
21..Voo / n"-l
II = - Vn cosh- 1 1 -.- (8.3.20)
na' 2 -1 a"-1
The corresponding pressure coefficient is
41.. }-1 ,I 112_1
P = :la.' .1
r n 2 -1
cos 1 -2-1
a -'
V (8.3.21)
Semi-Infinite Wing
with a S~personic Edge
. For such a wing (Fig. 8.3.6), Mach line OK issuing from a vertex
is on its surface. Consequently,
'It/2-% > /-toot tan %< a', n = tan x/a' < 1 (a' = cot /-too = VM~-1)
Let us consider the velocity at point L on the wing between the
leading edge and Mach line OK. Since the side edge (tip) of the
wing coinciding wi th the x-axisis ou tside the Mach line drawn through
point L, no influence of the side edge is observed on the flow at this
point. This flow is the same as over a· flat plate in a flow directed
along a normaL to its leading edge at the snpersonic velocity V noo =
V cos % > a
00 By expression (7.6.8) and the formula p =
00'
we have
U = -A V oolJl"a'2 - tan 2 %
whence
U = -AV oo/(a' V1 - n 2 ) (8.3.22)
The relevant value of the pressure coefficient is
p= -2u/VGO = 2A1(a' V1 - n2) (8.3.23)
Formulas (8.3.22) and (8.3.23) may be applied for n < 1 and
1> > (J n.
Let us calculate the velocity and pressure at point P between
Mach line OK and the side edge. If we assume that this point belongs
to a wing whose vertex is at point C (Fig. 8.3.6), the velocity would
be calculated with account taken of the influence of only the leading
edge and, consequently, of the sources distributed in region PCH.
By (8.3.22), this velocity is
(8.3.24)
To find the actual velocity at point P belonging to the wing with
its vertex at point 0, we must subtract from (8.3.24) the velocity
induced by sources distributed in triangle A CO and having a strength
of the opposite sign. The magnitude of this velocity is determined
by means of formula (8.3.7). Substituting Zc = (xp - a'zp)/(tan x-a')
Ch. 8. A Wing in a Supersonic Flow 329'
::'/
.':11
~~I
!
Fig. 8.3.6
Triangular wing with a supersonic edge
(8.3.25),
-
2
p - - -u- -
~'oo - '7..'
~o
~"
-V 1 - n2
[ 1 - -.
1
;-t
0,
CO."-1 - - -n-
It
----
(1-0)
0
J (8.3.28)-
Pig. 8.3.7
Pressure field for a semi-infinite
triangular wing with a super--
sonic leading edge
Fig. S.3.S
Triangular wing symmetric about the x-axis with supersonic leading edges:
I-Mach line; 2-maximum thickness line
The velocity at point P that is within the Mach angle is affected not
only by the leading edge, bllt also by the trailing and side edges of
the wing. The velocity due to the influence of the leading edge and
part OA of the maximum thickness line (Fig. 8.3.8) is determined
by formula (8.3.27), while the velocity induced hy the sources dis-
tributed over region OA' A is evaluated by expression (8.3 31). Adding
(8.3.27) and (8.3.:31), we obtain the total velocity at point P of a sym-
metric wing:
u= - Ali 00 [ 1- -cos-1
1 o-n 2 1
+-CO,,-1
n 2 -La
I
]
a' 1(1-n2 _ :t 11 (1-~0).Tl n (1+a)
or after transformations,
,,1-
u= - 00 (1 L
- -;;- Si1C 1 1/
' /1 2 - 0 2
(8.3.33)
a' 1/1-/12 ., JI 1-a 2
P --- -
2[1
1- 00 =
211,
a' 1(1-1/2
(. 2.
1 - -.TlS.l n -
1
V
.- IZ 2 -
1--0 2
0 2
(8.3.34)
Fig.8.4.f
Tetragonal wing with a symmetric airfoil and subsonic edges
trary planform. Let us consider the tetragonal wing shown in Fig. 8.4.1.
A left-hand system of coordinates has been adopted here and in some
other figures to facilitate the spatial depicting of the wing and the
arrangement of both the sections being considered and the required
notation. We shall assume that for such a wing the leading and trailing
edges, and also maximum thickness line CBC' are subsonic. According-
ly, the sweep angles Xl' X3 of the leading and trailing edges and the
angle X2 of the maximum thickness line are larger than n/2 - fL "".
The distribution of the velocity and pressure over an airfoil depends
on where the latter is along the wing span, i.e. on the lateral coor-
dinate z of the section.
Airfoil F L (z = z,). Four flow regions, namely, FG, GH, HJ,
and J L, should be considered on the airfoil. Region FG is confined
by point F on the leading edge and point G at the intersection of a
Mach line with the coordinate plane z = Zl' Point G is considered
to be on plane zOx and is determined, consequently, as the point of
intersection of the Mach line issuing from the projection B' of point
B onto plane zOx and of the straight line z = Zl (Fig. 8.4.1). The
velocity and pressure coefficient in region FG behind Mach line OKo
on the surface of the wing are determined with the aid of the distri-
bution of the sources in triangleOCC' by the relevant formulas (8.3.18)
and (8.3.19).
Ch. 8. A Wing in a Supersonic Flow 333
PFG=
4i..l
~COS1,'
1- 1 ~/ nt-Or
1 ., (8.4.1)
na'l'ni-l - 01
where 111 = tan x/a 1 • (Jl = Zl tan x l /.1\, and Xl is the running coor-
dinate of the point.
The drag coefficient of the airfoil corresponding to region FG is
XG
ex, PG = 1~1 ~ (Pu + Pb) dX 1 (8.4.2)
xp
we have
(8.4.3)
Using formulas (8.4.2) and (8.4.4), let us determine the drag coeffi-
cient corresponding to region GH:
cos h-1 V _ __
n~-1 d0 2
-2-1 ' - 2
O2 - O2
(
845)
..
O"zG
we obtain
ex, FH -= ex, FG + ex, GH
Zl tan Xl
=-
8Aiz] tan Xl
B
X +l
1 an
)t X2
COS
h- 1 .V- ni-oi
1 - 02
bna.' V ni-1 1
1
1 _ __
X dOl _ 8 (1.2 - A1 ) AjZ j tan x2 r cosh-1 1 jr n~ -1 • d0 2 (8.4.6)
01 bM' V/1~-1
/12
J II o~-l O~
Let us assume that part of chord HL (we presume that point H
r
is on line BC) equals ;:-b, where is a dimensionless coefficient of
proportionality determined from the condition;:- = B'Dlb o (here bo
is the central chord). Therefore, for surface OBC, the part of chord
FH will be (1 - ;') b.
Since a part of central chord B'D eqnals rb o , the remaining part
OB' equals (1 - r) boo The angles are
Al = A/2 (1 - r), A2 = -~/(2;=) (8.4.7)
~;-~ = !J./b o is the relative thickness of the airfoil.
Taking into account the values of Al and A2 we can write formula
(8.4.6) as:
(l-T)bo+Zl tan X2
ex, F'H=-
2X2 Z 1
[tan Xl
b (1-r) na.' - ni-1 V )
1
cosh-1
V n 21 - 012
1-021
V
/12 _ __
+
~
X
do]
-0 _ ,/
ta x 2
COS h-l n~-1
---.--
d0 2 ]
(8.4.8)
2
1 r V n~-l o~ -1 o~
1
Ch. 8. A Wing in a Supersonic Flow 336
+ 4(1.2-1.1)
--~~====~COSl
1-1 V r
Il~-O~
2 (S.4.9)·
:rtcx.' Vn~-1 1-0 2
Using formula (8.4.5) in which we must replace the value of At
with A2 and taking into account that
Al = ~/[2 (1 -;=)1, A2 - Al = - X~/[2r (1 - r)l, A2 = - ~(2r)
n l = tan 'Xl/a', n2 = tan xz/a', a l = Zl tan Xl/Xl
a z = Z2 tan 'X2Ix~, XH:::;; Xl -:( XJ, x; = Xl - .r"ll
we obtain
(J1J
ex, HJ =
-2"X 2Z 1
---==---=-''---
[tan Xl
'//,2 -1
)'
cos
h- 1 Vi'" ni-or
1- 0 12 •
dOl
oc1'
br (l-r) ~a' r '1
(ill!
- tan X 2
Vn~-1
G:>J
)
cos } 1-1 V n~- o~
1-0 2
2
• d0 2
o~
]
(S.4.1O)
(i2H
(SA 11)
336 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
where na = tan xa/a'; the values of 0'1' 0'2, and 0'3 are calculated
:relative to points 0, Band D.
Introducing the value of P-;L into formula (8.4.2) and substituting
.Az for Al in it, we obtain
2"K2Zl [ tan XI
ex JL= - - - - -
• bma' . (1-r) Yn~-1
+ tan X3 (8.4.12)
~ yn~-1
•
_
ex HL- - - - - -
bma'
_
2LlA2 Zt [ tanx
(l-r) Vn~-1
l
GIL
J. co~
'h-1 V n2t - a t2
1-a l
2
d al
.--
ai
GlH
tan X2
r (1-r) yn~-1
. cosh-l V -n~-1
(J3L _ __
tan
r ynl-1 j
X3 da-a]
+-=----........:!.=- --.-
ai a~-1
(8.4.13)
(J3J
In this expression
tan Xl
Zl
O'iH = ---:~---,:,,---
·(1- r) bO+ZI tan X2
0'2H = 1, (8.4.13 1)
Zt tan Xs Zl tan X3
0'3J = x' = n3' 0'3L= X' =1
J L
in triangle OCC'.
Accordingly, we evaluate the pressure coefficient by formula
(8.3.19). Assuming that 0' = Zl tan X/Xl = 0, we obtain
- 4Al h- 1 (S.4.17)
POE = ,~ cos nj
:ta'y ni-- 1
Airfoil section DB with a length of tb o experiences the action of
sources with a strength of Q = 2"1 Voo distributed over region OCC'
and with a strength of Q = 2 ("-2 - AI) V 00 distributed in triangle
BCC'. In accordance with this, we determine the pressure coefficient:;:.
Using formula (8.3.19) at 0' = 0, \ve find
- 4AI h-1 4 (1.2 - AI) I
PBD =
na'
V ni-1
cos nj + :rta' -vnf=1
n~-1
cos 1
-1
no
"
(8.4.18)
The drag coefficient of the airfoil with account taken of the llpper
and bottom surfaces related to the central chord lin is
2
cx = - - (S.4.1H)
bo
z
-------Ka
c' c
Fig. 8.4.2
Arrangement of ~[ach lines
(8.4.21)
(8.4.22)
where 0'1' 0'2. and 0'3 are determined relative to points 0, E, and
D. respectively.
By using the formula
(8.4.23)
we can calculate the drag coeffIcient for the section of the airfoil
being considered related to the length bo of the centre chord.
The flow over the last airfoil section J 2L2 is the result of induction
by the sources distributed in the same regions of the wing as for
airfoil section H 2J 2' Here account must be taken of the feature
that the velocity induced by the sources in region BCC' is determined
by formula (8.3.18), where A is replaced by the angular coefficient
A2 - AI' Therefore, the pressure coefficient should be calculated by
formula (8.4.11), and the drag coeffIcient ex, J,L, from expression
(8.4.12) in which we take the integrals between the limits O'nJ.
and O'nL, (n = 1, 2, 3).
We obtain the total drag coefficient by summation of the coeffi-
cien ts for all four sections of the airfoil:
ex, F,L, = ex, FIG, + ex, G,H, + ex, H,J, + ex, J,L I
provided that the value of Zl satisfies the inequality
rb o < (1-;:)b o
tan Xl-a' Z1 < tan Xl-a'
22*
840 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
Airfoil FaL3' Let us consider section F3L3 (Fig. 8.4.2) with the
coordinate Z1 satisfying the inequality
(8.4.24)
where
(8.4.25)
The pressure on section F aJ a of the airfoil is due to the action
of sources distributed on the triangular surfaces aGG' (Q = 2A.1 V 00)
and BGG' [Q = 2 (1.,2 - AI) V001. We shall therefore calculate the
pressure coefficient by expression (8.4.4), and the drag coefficient
by formula (8.4.5) in which we replace the limits an, G and an, R with
the values an, F. and an, J. (n = 1, 2).
Airfoil section J aH 3, in addition to the indicated source distribu-
tions in regions aGG' and BGG', also experiences the action of sources
having a strength of Q = -21.,2 VL distributed in triangle DGG'.
Consequently, the pressure coefficient equals the value calculated
by formula (8.4.4) plus the additional value calculated by formula
(8.3.21) in which we assume that A = -1.,2'
Flow over section H aL3 is characterized by the induction of sources
distributed in three regions of the wing, namely, aGG', BGG', and
DGC'. Accordingly, the pressure coefficient on this section should be
calculated with the aid of formula (8.4.11).
By calculating the relevant components of the drag coefficient
for all three sections and summing them, we obtain the total drag
coefficient:
ex, F.L. = ex, F,J, +ex, J,R, +
ex, R,L I
The relevant expression is suitable for calculating the drag coeffi-
cient of the airfoil between points Dl and D2 (Fig. 8.4.2) with the
coordinate Z1 that satisfies the inequality
Fig. 8.4.3
Distribution of the drag coeffI-
cient over the span of a swept
constant chord wing with sub-
sonic edges
(the dashed line shows an unswept
wing)
JV~(M..,>1)
-------------=~O~~~,_r_._-----Z
Fig. 8.4.4
Wing with side edges
where 0'1,0'2 and 0'3 are calculated relative to points 0', B', and D'.
If Mach line O~K; intersects chord F 2 L 2 , then simultaneously
with the action of sources 0' CC" account must also be taken of the
influence of the sources distributed in triangle O~ C' C~ at the opposite
side of the wing. The same formula (8.3.13) is used to calculate the
induced velocity.
(b) (",
~l
Fig. 8.5.1
Tetragonal wing in a supersonic flow:
a-the leading and middle edges are subsonic, the trailing edge is supersonic; b-the lead-
ing edge is subsonic, the middle and trailing edges are supersonic; c-all the edges are super-
sonic
are replaced with the values on, H, and 0 11 , L.(n = 1,2). respectively.
The total drag coefficient for airfoil FILl is
(8.5.2)
A feature of the flow over the wing (Fig. 8.5.1b) consists in that
the sources distributed in region BDDI do not affect the distribution
of the velocities and pressures on the remaining part of the wing
above Mach line BK B •
Let us consider airfoil F L with the coordinate Zl < ZD •• Section
FH of this airfoil is acted upon by the sources distributed in triangle
OCC' (Fig. 8.5.1a); consequently, the pressure distribution can be
found from (8.4.1), and the corresponding drag coefficient Cx. FH,
from (8.4.3) in which aIR is substituted for the upper limit 0IG'
The second section HG is influenced by subsonic leading edge OC
(and, therefore, the distribution of the sources in OCC') and by
supersonic middle edge BC. The corresponding pressure coefficient
is determined as the sum of two coefficients, the first of which is
evaluated by expression (8.4.1), and the second by (8.3.23) wherl~
we assume 'A = 'A 2 - 'AI and n = n 2 • Consequently,
(8.5.3)
(8.5.4)
For such a wing (Fig. 8.5.1c), Mach lines OKo, BKB • and DKo
drawn from points 0, B, and D are below the corresponding edges
OC, BC and DC, therefore formulas (8.3.23) and (8.3.34) are used to
calculate the pressure coefficient.
Let us consider airfoil F L with the coordinate 0 < ZI < ZD ••
Section FH is between leading edge OC and Mach line OK o. There-
fore, the other edge OC' (Fig. 8.5.1a) does not affect the flow on this
section, which is considered as plane supersonic. Taking into account
that section F H is influenced by the sources in triangle OCC' having
a strength of Q = 2A,1 V the pressure coefficient PFH can be deter-
00,
-
PGJ=
21.,1
ex' Y1-ni
(1 - 2.
-SIn
It
-I V n~-ai )
l-a t
2
+ 2(1.,2-1.1)
ex' yl-n~
(8.5.7)
-
PJL= ,,/1
2At
2
(1 -n2Slll
. -1" /
V
n~-ai )
l-a 2
ex r -~ 1
(8.5.8)
Ch. 8. A Wing 'in a Supersonic Flow 347
xJ XL
ex. FjL I = ! ()
ICPI
PF]R] AI dx + J PRIGIAI dx + J PG
xH l xGl
IL1 A2 dx) (8.5.10)
The flow near airfoil F 2L2 in the section ZD, < Zl is plane super-
sonic, therefore the pressure coefficient for it is determined by for-
mula (8.3.23). The sources in triangle aGC' having a strength of
Q = 211,1 V oc act on section F 2G2' therefore the pressure coefficient
PP,G, is found by formula (8.3.23) in which A = Al and n = 121,
The second section G2 L 2 is additionally influenced by the sources
having a strength of Q = 2 (11,2 - AI) V 00 distributed in triangle
BGG'. Therefore, the pressure coefficient on this section is
2Al , 2 (A 2 -A 1 )
(8.5.1'1)
PG2L2= 0.' l/l-n§ -t- 0.' Vl-n~
The drag coefficient for airfoil F 2L2 is
XG2 XL2
Cx .F 2 L2 = f ( ) PF G)'1 dx + ~
2 PG2L2A2 dx ) (8.5.12)
ICP2 XGt
ex
7l w J2
2. I; f\--+---<;~---r
2. D ~-+----h+----f
Fig. 8.5.2
Drag coefficient of wings with I. 2 1---+------1~~~~--+----j----;
this section are calculated by the method set out in Sec. 8.4 with
account taken of the kind of the leading and middle edges (i.e.
depending on whether they are subsonic or supersonic) by means of
the relevant relations similar to (8.4.31).
By integrating over the span, we can determine the wave-drag
coefficients in each of the cases of flow over a hexagonal or tetra-
gonal rhombiform airfoil wing considered in Secs. 8.4 and 8.5.
According to the relations obtained, these coefficients depend on the
number Moo and on the configuration and relative dimensions of the
wing:
(8.5.13)
Here in addition to the known notation, we have introduced the
quantity Xc = xc1b-the dimensionless distance to the spot with
the maximum airfoil thickness.
The number of independent variables in (8.5.13) can be reduced
by using the relation
(8.5.14)
where tan 1(2 is the tangent of the sweep angle along the maximum
thickness line (i.e. of the middle edge).
Figure 8.5.2 shows a family of curves constructed in accordance
with formula (8.5.14) for the following conditions: 1]w = bolb t = 5
and Xc = 0.5. The salient points of the curves correspond to sonic
edges. Particularly, for the curve corresponding to the value of
'Aw tan 1(2 = 3, the salient point for the smallest value of
'AwV M~ - 1 corresponds to a sonic trailing edge, and the second
and third points to a sonic middle (the maximum thickness line)
and trailing edges.
Ch. 8. A Wing in a Supersonic Flow 349
ex
2.0.--.~~~,-----·-----,----,----,----'
1.5
1.0
Fig. 8.5.3
Drag of triangular wings with a symmetric rhombiform airfoil in a supersonic
flow
(n, - tan x,/a')
General Relation
for Calculating the Drag
(8.5.15}
(8.5.16)
I
Maximum
0.8 tlllC If Tless
II Tle
0.4- r---!---t
Fig. 8.5.4
Change in drag function f~ by formula (8.5.16)
x x
fig. 8.6.1
Wing with a finite section of a supersonic leading edge
Fig. 8.6.1
Wing with subsonic leading
edges x
(8.6.3)
in the form
L1 QM [ -1
CP=2i1l V(X-S)2_ a '2 [(y-e)2+(z-~Yl
+ 1 }
y(x- £)2 _a'2 [(y+e)2+ (Z_~)2J
by this surface.
fig. 8.8.1
Plane triangular wing with subsonic edges
x
fig. 8.8.2
Region of doublet influence on
the flow over a wing with sub-
sonic leading edges:
I-region of doublet influence; 1 -
triangular wing
Cjldoub = r
J
m (h) dh day r
J0
~2d~
V(x-~)2-a/2 [y2+(z_h~)2]
(8.8.4)
-cot x
Ch. 8. A Wing in a Supersonic Flow 357
C=X2 , 2 -'-_"_
[ 1-a'2 (-~- .2) ]
2 (8.8.7)
x x ' 2
(8.8.8)
~.Taking into account that a = 1 - a'2 and h 2 > 0, \ve find the
integral on the right-hand side of (8.8.8):
;,1
V
d£
=
1
,r In[2Va(as2+b~ c)+2a~+b]o.-1
I" (8.8.9)
• a£2+b£+c va
o
By (8.8.5), the consequen tly,
o
2
+ 3b -4ac [In (2aS1+b)-ln(2Vac+b)] (8.8.10)
Sa 2 ya
Sol ving Eq. (8.8.5) in the form
a~~ + b~l + c = 0, (8.8.11)
we obtain
S1 = (-b +V b2 - 4ac)/2a (8.8.12)
Accordingly,
In (2a~1 + b)= In V b2 - 4ac (8.8.13)
------------------------------------------
358 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
= 1 .~ ( b . 8ac + th- 1
2a Va oy 4 vac 4ac-b 2
Introducing the symbol
" = bl(2 Vac) (8.8.16)
and having in view that fJc/oy = _2ya'2 [see formula (8.8.7)],
we obtain
N= -ya'2 (_v_+th-1 v) (8.8.17)
a Va I--v 2
We introduce this expression into (8.8.4):
cot ><
Cj)dnub ,= ) Nm (h) dh (8.8.18)
-cot ><
To determine the form of the doublet distribution function, we
shall use the condition of flow without separation, in accordance
with which
(8.8.19)
The partial derivative (OCj)doub/oy)y=o is determined in accordance
with (8.8.18) in the form
cot ><
( OfPg~ub ) y=o = J
-cot x
(~~) y=o m (h) dh (8.8.20)
Fig. 8.8.3
Incompressible two-dimensional cross flow at the velocity V ooa over flat plate
AB belonging to a triangular wing
We shall show that this equation has the same form as the one
obtained when using the doublet distribution method to solve the
problem on the flow of an incompressible fluid over a flat plate of
infinite length installed at right angles to the direction of the free-
stream velocity. We can determine the velocity potential for a plate
that is a part of a wing in section AB with the coordinates ZA = -c,
ZB = C, and over which an incompressible fluid flows in a lateral
direction at the velocity v = V ooa (Fig. 8.8.3) with the aid of Eq.
(2.9.16). We shall write the latter in accordance with the coordinate
system chosen in Fig. 8.8.3 in the form
(jldoUb = y/[(z - "Y})2 + y2] (8.8.25)
This expression determines the potential at point P (y, z) pro-
duced by a two-dimensional point doublet with a unit moment
360 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
(M = 1). If the moment of the doublet differs from unity and the
doublet distribution over the span of the plate is set by the function
M(lJ) , the velocity potential induced by the doublets on the wing
spction dlJ is ~
(8.8.26)
The velocity potential at point P due to the influence of the doub-
lets located along the span of the plate on the section from ZA = -c
to ZB = c is
(8.8.27)
-c -c
Since the component vy=o does not change over the span of the
plate, we have
(8.8.29)
-c
Differentiation yields
(8.8.30)
-c
and on the section between the leading edge and a Mach cone [see
(8.3.23)] it is
P2
= +2a/(a' V 1 - n 2 ) (S.S.50}
where n = tan x/a' , and cr = z tan x/x = h tan x.
364 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
(c) (d)
a z
x x
x
Fig. 8.8.4
Wings with supersonic trailing and side edges (xo = bo):
a-tetragonal wing with a dovetail; b-tetragonal wing with a vee-shaped appendage (rhom-
biforrn platp); c-pentagonal wing; d-hexagonal wing
__ -4acotxl [2+e 2
-I e(4-E)2]
(8.8.52)
mZa - J(1-e) E (k) (1+e)3/2 cos e- 1-e2
Ch. 8. A Wing in a Supersonic Flow 365
Fig. 8.8.5
Curves characterizing the change in the derivative of the lift coefficient (a) and
centre-of-pressure coefficient (b) for a tetragonal wing
Cy
_
- (1
8a
) ,.!
(1 cos -1 _n1,/
8 cos -1)
n (8.8.53)
" n a " 1-nl
-E " 1-n2
_ - 8a [ 2 - nf (1 + 8 2) -1
mZa - 3 (1-E) na' (1 +8)2 (1- nn3/" cos n 1
_ _ _8--'..,(3_--"7:8=2)==- -:-n....;l,-;;-l
cos -1 n--c (8.8.54)
(1+E)2 V1-n2 1-ni J
wings in Fig. 8.8.5) alters the lift force of the wing almost in pro-
portion to the change in the area of the plate.
It is not difficult to see that for e = 0 formulas (8.8.51)-(8.8.54)
yield the values of the relevant coefficients for a triangular wing:
c ya = cya ~ and m za = mz~. a
The calculations by these formulas
can be simplified if the tetragonal wings differ only slightly from
triangular ones. In this case, provided that I e I ~ 1, we have
(8.8.55)
a
~ -:-yr:;(=x=_::;::S::=;;)2=~=S=~:;:;;'2:=;:(Z=_==;::::~)~2 r 1 (8.9.3)
Fig.8.9.t
Hexagonal wing with subsonic
leading and supersonic trailing
edges
directions of the Mach lines issuing from the vertex of the wing
(Fig. 8.9.1):
r = (M 00/2a') (x - a'z), S = (M 00/2a') (x + a'z) (8.9.4)
The characteristic coordinates of point A(XA' ZA) are as follows:
rA = (Moo/2a') (XA - a'ZA), SA = (M oo /2a') (XA + a'zA) (8.9.4')
Let us convert Eq. (8.9.2) to the characteristic coordinates rand s.
From (8.9.4), we fmd:
r +S = (M oo/a')x, r - s = -M ooZ (8.9.4")
Consequently,
x - S = XA - X = (a' 1M 00) [(rA + SA) - (r + s)] }
a ' (z - ~) = a ' (ZA - z) = -(a'IM 00) [(rA - SA) - (r _ s)] (8.9.5)
(8.9.6)
,
IPn =
aVoo (n+l)
4nMooE (k)
TA
j
Y
dr
rA-r
I
SA
ds
(8.9.15)
TK SB'
Points K' and G' are on the tips whose equations are Z = ±lI2
(the plus sign is for the starboard and the minus sign is for the port
tip). In the coordinates rand s, the equations of these tips in accord-
ance with (8.9.4) have the form:
r- S = -M 00l/2 for the starboard tip }
(8.9.22)
r - S = M ool/2 for the port tip
}- ConsequentlyI for points K' and G', we have, respectively:
rK' = SK' - M ool/2 = SA - M ool/2 }
(8.9.23)
SG' = rG' - M ool/2 = rA - M ool/2
After the integration of (8.9.21), with a view to the values of the
limits (8.9.23), we obtain
, aV oo (n+1) , , / ( Mool)( SA-rA-2-
Mool) (8924)
<PIlI= nMooE(k) V rA-sA 2-
-1-- .,
(a) (b)
Fig. 8.9.2
Wings with subsonic leading
edges and tips and with super-
sonic trailing edges:
a-hexagonal wing with a dovetail;
b-plate with a straight trailing
edge (a pentagonal wing)
x
4DO
81----++_
6 1---1H---+
41--~~--~---+--~--~
Fig. 8.9.3
Distribution of the quantity 21--~~--~---+--~--~
IIp/rx = (Pb - Pu)/rx near a tip
of a pentagonal wing with oL~::±=::i=±==1
subsonic leading edges (sec- 20 40 60 80 100"/..
tion AA) x=(xjb)fOO
-
PII=-
Voo
- - 'i}xA
2
--=+
oXA - E (k)
oCPII I
)- (8.9.26)
VOO
-
PIlI = ---.
2 oCPJII
=0
I
)
Voo oXA
Here the plus sign on the right-hand side of the equations corre-
sponds to the bottom side, and the minus sign, to the upper one.
The above relations for calculating the pressure coefficients can
be used not only for triangular wings with a V-shaped appendage
(Fig. 8.9.1), but also for similar wings with a dovetail and wings
with a straight trailing edge (pentagonal plates shown in Fig. 8.9.2).
The change in the quantity /).p/a = (Pb - pu)/a near a tip as
a function of the distance from the leading edge (in per cent of the
Ch. 8. A Wing in a Supersonic Flow 371
~LL
I I I
O.751---j---AF-----1
0.7
H-- 7. w tan'J4=J
'JJ-
0.5 ~
I I
D.5 I 1_
"I ill~I"·
I
- 1
8.4
I I ~.
I
TiT I TT1"iTT1*
0.3 D 5 fO
Fig. 8.9.4
Curves characterizing the change in the lift (a) and centre-oI-pressure (b)
coefficients for a pentagonal wing ~
(8.9.28)
where Sw and b o are the area and the root chord of the wing, re-
spectively, andpb and Pu are the coefficients of the pressure on the
bottom and upper sides determined by formulas (8.9.26).
According to the values of m z a and c y a , we can determine the
centre-of-pressure coefficient c p = xp/b o = -mz,,'c y,,' Figure 8.9.4
shows theoretical curves characterizing the change in c Ya and c p
for a wing in the form of a pentagonal plate. In this figure, the parts
of the curves for 'which AwV' M;, - 1 ~ Aw tanx correspond to
subsonic leading edges. A feature of the graphs characterizing the
change in cYa is the absence of noticeable salient points on the cor-
responding curves in the transition to supersonic leading edg'es (i.e.
when "'wV'M;" - 1 = "'w
tan x).
24*
372 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
(8.10.3)
(d)
(e)
x EnLarged x
view 01' 8
Fig. 8.tO.t
Hexagonal wing with supersonic leading and trailing edges
374 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
, -aV 00 I I d~ d~
<PIlI = n ~IJ y(x _ ~)2_a'2 (Z_\;)2
aVec
- -n- tI Jr y(x -
d~ d~
~)2_a'2 (Z_~)2
1 r\ Q (£. ~) d~ d~
(8.10.4')
- 2" J8 J l/(x - ~)2_a'2 (X __ ;;)2
3
~
-aV",
~
dr ds
<PIII = (8.10.5)
nMoo -YrA-r YSA-S
r A' sD3
Ch. 8. A Wing in a Supersonic Flow 371)
(8.10.10)
We determine the integration limits ST' and SJ' with the aid of
(8.9.8). For point T' on the starboard leading edge and point J'
on the port one, we have
ST' = rT'/m = rim, SJ' = rJ,m = rm (8.10.11)
With a view to the:se values of the limits, we calculate the inte-
grals:
SA
2
J\ ds
YSA-S
= V m- -Vr- sAm
'T
·Jr YSA-S
ds = 2V m .. /
V
r _ SmA
sJ,
Accordingly,
o
I ___ 2a_V-;-:oo=- r .. /r-SAm dr
IJlIV=
nMoo -V m J V rA-r
rR
_2aVoo V~ rt 1/ r-sAlm dr
nMoo J Y r A-r
o
where rR is determined by formula (8.9.23):
rR = SR - M ool/2 = SA - Mool/2
Integration and substitution of the values for SA and rA from
(8.9.4'), and also m = -iii = (n - 1)/(n + 1) yield
, _
IJlIV-,/
a'n y 1--n 2
-2aVoo {(
XA-zA tanx 1) [ -tan-1 V (n+1) (xA-a'zA\:
(1)( +
-n XA a I ZA )
+t -1"/ a' (n+1)(l+2z A) ]
an V 2 (xA+a'zA)-a'l(n+1)
+ {- V(l- 2z A) [2 (XA +a'zA) -a'l (n+ 1)] (n + 1) a'
(8.10 12)
Ch. 8. A Wing in a Supersonic Flow 377
tis
(8.10.14)
Consequently,
rA
~
-aVoo dr
-
nMoo VrA - r
rBf
SA rB{ sA
ds
X \
J
ds
II SAc - s -
av",,)
nM 00
dr \
VrA-r ~ II SA -s
(8.10.17)
sEl rBl S "
El
rBl = SA - M ool/2
-
PYI =
±40; [t V an
-1 a'(n-t)(ZA-I/2)
na' V1-n 2
_ xA+a'zA+ (+)(n-1)0;'
(8.10.20)
(8.10.22)
Let us now consider region VII I (Fig. 8.10. th) for whose points
the region of integration simultaneously intersects regions Q 1 and
Q2; disturbances manifest themselY(~s in these regions that are pro-
duced by the flow over both tips DE and GH. To evaluate the velocity
potential for point A that belongs to region VI I I, it is sufficient
to extend integration in formnla (8.1O.3) to the region S = Sl S2 +
hatched in Fig. 8.10.1h. Here the integral calculated over region S 2
in formula (8.10.3) mnst be taken with the opposite sign, i.e. with
the plus sign. Accordingly, we haye
-'Xl'""
JliU x
(8. '10.23)
Here
ds
/ (8.10.24)
l sA-5
380 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
\ r dr ds _ \ (' dr ds
. J V(rA-r) (SA-S)
82
-- • J
01
V(rA-r) (SA-S)
-L (' \ . dr ds + r (' dr ds
, J02J V(rA-r) (SA-S) JJ
03
V(rA-r) (SA-S)
rF , SF{ 0 s F~
o
= f vr:r_r
sL 1 rFI
I Vs~-s + .~, vr:~r
SKI
) vsd~_s (8.10.25)
provided that
SFI = rA- M ooll2, rHI = SA - M 00l/2 (8.10.26')
With a view to these values of the limits, we calculate integrals
(8.10.24) and (8.10.25), and then find the velocity potential (8.10.23).
The following pressme coefficient corresponds to this potential:
- ±4a [t -1 11+n .. /(xA-a'zA)+(l/2)a'(n-l)
PVlII = a'n Vi-n2 an V~ V a' (1/2+z A ) (n+1)
- t an -1 vi 1+n
- - V'XA-a'ZA
1-n XA+a'ZA
+ t an -1 vi - - vi 1-n
1+n
XA-a'ZA
XA+a'ZA
t -1"I 1-. n -. I a' (1/2-z A ) (n+l) ]
- an V 1-rn V xA+a'ZA+(1/2) a' (n-l) (8.10.27)
The above method of calculating the pressure distribution can
be related not only to hexagonal wings with a V-shaped appendage
(see Fig. 8.9.1), but also to other planforms, namely, to a wing with
a dovetail and a pentagonal plate (see Fig. 8.9.2) provided that the
tips are subsonic, while the leading and trailing edges are supersonic.
The corresponding change in the quantity !1pla = (p h - Pu)/rx
near a tip with the distance from the leading edge (in per cent of the
chord) evaluated for a wing with a pentagonal planform at Moo =
= 1.61 is shown in Fig. 8.10.2. The coefficient !1pla is constant
in the region between the leading edge and the Mach cones issuing
from the vertex and tips of the wing, on the section confined by the
tips and trailing edges, and also by the Mach wave reflected from a
tip. A break in the curve characterizing a change in the pressure
drop coefficient occurs on the Mach lines and is a result of the break
in the configuration of the \ving tip.
The found pressure distribution and formulas (8.9.27) and (8.9.28)
are used to determine the lift, moment and centre-of-pressure coeffi-
cien ts. Figure 8.9.4 shows curves characterizing the change in the
values of cYa and c p for a pentagonal wing. These values should be
Ch. 8. A Wing in a Supersonic Flow 381
fig. 8.10.2
Distribution of the quantity
IIp;a=(Pb-Pu)/a near a tip
of a pentagonal "ing \\ith aL--L--L--L--~~--~
supersonic leading edges (sec- 20 '+0 50 80 100. %
tion AA) X=(X/b)100
determined provided that t,wV lJ1;, - 1 > )~w tan Xl. i.e. if the
leading edge is supersonic. If a wing has subsonic leading edges (see
Fig. 8.9.1), the flovv over the section of the surface between the edges
and the Mach lines issuing from points E and H is affected by a vortex
sheet. The calculation of this flo,,' is associated with solution of
integral equation (8.2.16) anci the use of boundary conditions (8.1.15)
and (8.1.16). Such a solution is treated in detail in [171.
Fig. S.t1.1
To the calculation of the suc-
tion force for a triangular wing
with subsonic leading edges
x
T = 2JIpoo V-:-1-+-t-an-~:-x--:M-::-';,:- i
1/2
c'2 dz (8.11.3)
o
where c is a coefficient determined from (7.6.19).
Let us determine the drag for a triangular wing. For this purpose,
we shall use expression (8.8.39) for the disturbed velocity. Assuming
that cotx = z/XI. e , we find
a V 00 (z/Xl.e)2
U = -;"[E-::-:-(k=-)-y""7=:=1-=(x=]=.e=/x=)~2]-
where X].e is the distance to the leading edge, Z and x are the coor-
dinates of a point on the wing (Fig. 8.11.1).
Introducing the value of u 2 into (7.6.19) and taking into account
that lim (X/XI. e ) = 1, we have
X-+X].e
2
C =
[a:Voo
E (k)
J2 I'
lID
Z2 (xj X l. e )2 (x-xJ.e)
(X-XI. e ) (X+XI. e ) =:2"
1 [ a:Voo
E (k)
J2 Z cot X
X-X].e
.il r
1.0 r---.---,-,--,---r-----=r--,----,
0.4
0.2
Fig. 8.11.2
Cl:iange in the factor c\T in
formula (8.11.5) for calculating
the suction force
leading edge becomes sonic (tan x = cot !loo = V M'Zx, - 1), the
suction force coefficient equals zero. \Vith a subsonic leading edge
(M cos x < 1), the quantity CX,T = 0, btlt experimental investiga-
00
tions show that its actual value is lower than the calculated one.
This is especially noticeable at large angles of attack or sweep when
local separation of the flow (burble) occurs ill the vicinity of the
leading edge and no further growth of the suction is obsefYed. TJle
decrease in the suction force can be taken into account bv the cor-
rection factor,1T in accordance with which .
(8.11.5)
Experimental data on the change in the quantity ~T are contained
in Fig. 8.11.2. According to these data, the suction force at subsonic
velocities lowers less than in supersonic flow upon an increase in the
384 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
(8.11.8)
<piI =
-cxFoo rr dr ds
,ll~l 00 JJ
.\HLJE
11 (1'.\-1')
.
(-,\-s)
.
SA l'A
-
-cxFoo
nil/ oo l
sB
ds
liSA -s i
rF
dr
r'r.\ -- r
(8.12.2)
-------_._--_ ------ ..
386 Pi. I. Theory. Aerodynamics of an Airfoil and a Wing
Fig. 8.12.1
Rectangular wing in a linearized supersonic flow
(8.12.7)
Let us calculate the corresponding pressure coefficient:
PIl = - (2/V 00) dCPlliOXA = ± ~C( tan- 1 1/ C('z~ (8.12.8)
"ex. , x . \ --ex. ZA
On the boundary of the two regions I and I I separated by a Mach
line, the pressnre coefficient determined by (8.12.8) eqnals the
value (8.12.1). Indeed, since the equation of this boundary is ZA =
= XA tan !L = xA/a', from (8.12.8) we [md
00
(8.12.10)
·v
we have
4'1.b 2 floc Fz., '1.' tan" dy
ell • ~-~. 2
tan- 1. '
J-'1.'tan"I
.--
C()S2"
I
(8.12.13)
= 4'1. (
Cy
a Cyll Cy I ~~ '/:::::M::::2=:::J=- 1- / (8.12.111)
l . 00 _ . 2/." 1 .lIz., --1
The wave drag coefficient i;;::
2
C II' V = acya = 4ex
--;::;:;:;==;=- ( 1- ) (8 ....
P , 15)
Y. a e VrM~-l 2;"\\ {JJ;,--l
23*
388 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
Fig. 8.12.2
To the determination of the
aerodynamic moment for a rec-
tangular wing
where Yq), Y<p, and Y II are the values of the lift force calculated
for parts CDB (HEF), EFCB, and ADB (GHE), respectively. By
(8.12.11) and (8.12.12), we have
"(,-Ill ,pooV~
1 I = Cy 2
(8.12.18)
poo y~ 2a poo V~
YII=C~ 2 SII=-2-' 2 Sn
Introducing (8.12.7) and (8.12.18) into (8.12.16) and taking into
account the value S II = Sw/(2Aw a'), we have
-2a
.. rM" -1
(1 - 2)
3'A. w 'IM2_1
(8.12.19)
V 00 r 00
Ch. 8. A Wing in a Supersonic Flow 389
If the number 1J1 "" for the flow over the wing shown in Fig. 8.12.1a
is reduced, then at a certain value of this number the tip Mach cones
intersect inside the wing. Here region II I appears (see Fig. 8.1~.1b)
in which the disturbed flo\\' is affected by the leading supersonic edge
and both subsonic tips. The natme of the flow ill regions I and II
(II') is the same as in the corresponding zones I and II (J 1') of the
wing whose diagram is shown in Fig. 8.12.1a.
Let us consider tl18 flow at point A of region III (see Fig. 8.12.1b).
The region of influence of the SOllrces on this flow coincides with
wing section ABDD' 15'. The latter can be represen Led as the sum of
areas HBDD' and AHD'B'. 'VitI! this in view and according to
(8.10.3), the velocity potential at point A is
(PiII~ -aV""
:dfoo
[rJI1((rA-r)(s_\-s)
r drds + (' ('
JJ
drds
l((rA-r) (s.\-s)
J
REDD' ~ . ARD'B'· .•
+ (' . ds
J l/sA-s
.\ dr
VrA-r
J
-'H rB'
(8.12.25)
The first and second terms in this expression are the pressure
coefficients Pu and PII' for the tip parts of the wing, respectively,
while the third term is the pressure coefficient for region I where Pr
there is no influence of the tips.
The flow over a rectangular wing is of a still more intricate nature
if it has a low aspect ratio and the quantity A,wVM~ - 1 < 1.
In this case (see Fig. 8.12.1c), new wave regions appear that are
formed as a result of intersection of the disturbance waves incident
on and reflected from the tips. For example, the disturbances travel-
ling from section OH of the port tip reach the starboard one on section
O'H' and over it propagate along Mach linesO'O"-H'H" in the opposite
direction. This gives rise to new wave zones IV and V in which the
nature of the flow changes substantially. This flow can be calculated
by using the method of sources set out above and taking into con-
sideration the indicated intricate nature of formation of the dis-
turbance zones.
Ch. 8. A Wing in a Supersonic Flow 391
(a) IX JM2.,.,-1
Cya- --
4
.--;- l
0.8
~
r-~i- -"" I I
0.5
t--
/ v lV_(M~>2-
_. . -- V~ +--
/
~t
0.4 ! f----
/ f----
V I
0.2 f----
J i t---
aV
2 J
Cp=Xp/b
0.5 I
! j..-t-
-"" I
1/ I
4=D
I
Fig. 8.12.3
Curves characterizing the change 7 L
~ ~1: TH
0.4
in the derivative of the lift ]1 '
force coefficient (a) and the
I
cen tre-of-pressure coefficient (b)
for a rectangular wing (the I i
I
aspect ratio 'AI\' = lib) 0.3
D 5
---.--.--------.-----.~
892 Pt. !. Theory. Aerodynamics of an Airfoil and a Wing
21: 1 2
21 z
3
~~ ~V~
J~X~a-
p"IX, ?",lXz
Fig. s.n.t
To the explanation of the reverse-flow method:
1-forward flow; 2-reverse flow; 3-combined flow
while the pressure increment between the bottom and upper sides is
i1p = Pb - Pu = poo"v (u u - Ub) = poo"V i1u
00 (8.13.1) 00
s
X a2 = ~
i1p za 2 dS I (8.13.3)
Hence the pressure increment between the bottom and upper sides-
of this wing is f'l.P3 = P l' oof'l.u 3 = [J ",Yox; (,:lU l - LlUz), or by
00
(8.1:3.1):
f'l.P3 = f'l.Pl - f'l.pz (K.13.5)
The drag of wing 3 is
Xa3= ~
s
i (f'l.Pl-f'l.PZ) (cx,t cx,J clS (S.LUi)
s
~~
f'l.Plcx,Z dS = ~ f'l.PZcx,l dS
s
i (8.13.8)
-------------
9
Aerodynamic
Characteristics of Craft
in Unsteady Motion
QyllV,xl)
.
(/.,Z/V
(()z
..
OCl)
=
7:i = BliV
QzllV 00, 00
00
V ~= Vocz.V 00
I (9.1.2)
(9.1.3)
The variables Moo, (/", ~, {)r' {)e, {)a, (()x, (()y, and (()z determine
the coefficients of the aerodynamic forces and moments due to the
instantaneous distribution of the local angles of attack. Here the
variables Moo, (/", ~, ()" {)e' and {)a characterize the forces and
moments depending on the velocity, the angle of attack, the sideslip
angle, and the control surface angles, while the variables (()x, (()y,
and (()z determine the forces and moments due to the change in the
local angles of attack caused by rotation.
. .. . .
The time derivatives V Ct, ~, 6" 8e , 8a , ~x, (()y and ~z deter-
00,
(9.1.4)
9.2. Analysis
of Stability Derivatives
and Aerodynamic Coefficients
The first seyell terms ill Eq. W.1.4) dptermine the static, anci the
remaining ones, the dynamiC components of the aerodynamic coeffi-
cients. The static compollents correspond to ste1dy conditions of flow
over a craft in which its velocity is constant, while the angle of
attack, the sidpslip angle, and also the control surface angles are
jlxed. The dynamic components appear in non-stationary motion
attended by acceleratioIl or deceleration of the flow over a craft, its
rota1 ion, and a change in time of the control surface angles.
In accordance with the above , the dprivatives Moo ca cl:l and CO,.
' . cU U' Y ~.I/'
a.re ca:led. static stability derivatives, and the remaining ones, (cj1 x
c~oo, c~, CjJr, etc.), the dynamic stability derivatives.
A:- can be seen from (9.1.4), the aerodynamic coefficients are deter-
mined by the corresponding first and second order stability derivatives.
. .
The first order ones include the values of cel "" , c~, c~, c~, and
t
cf,
th e secon(I or d er ones, ca(:l _:::>2 /:::> :::>R aw . - :::>2 I:::>:::>
y - U cylUa uIJ, c y x - U cy,ua uffi x , e c.
An analysis of (9.1,4) reveals that the aerodynamic coefficients
are determined by two groups of derivatives, one of which depends
on the control surface vari~bles (ce,., cZ r , • • • ), while the other
is not related to them (c~, c~, ... ). The magnitude of a coefficient
depends on the action of these controls. This is due to the change
Ch. 9. Aerodynamic Characteristics in Unsteady Motion 399'
in the angle of attack and sideslip angle upon deflection of the control
surfaces. The values of the angles ex and Bcorrespond to the position
of static equilibrium of a craft.
To retain the given flight conditions, the control sllrfaces must be
fixerl in place. Such a flight with "fixed" control surfaces becomes un-
controllable. Its conditions are determined completely by the sta-
bility derivatives that dE'pE'nd on the intrinsic aerodynamic
properties of a craft if controls are absent or are fIxed in place.
Derivatives Stich as c3, ce, cel', ... ,
rielE'rmined as a result of
differentiation of the aerodynamic coefficiellLs with respect to
ex, B, 8,. (static stability derivatives) relate to the first grollp of
deri vaU yes, The partial derivat i ves of the coelfici8nts with respect
to one of the variables w x , w iI ' W z form the second group of what
we call rotary dE'riYatiyE's (for example, c!~'(, C,;'I, m"z':, ... ). This
, . .
group also inclndes derinltiYes similar to c~, C;]~ c~~, and also acce-
leration deriuatiues determined as the partial derivatives of tile aero-
dynamic coefficients \\"itll respect to one of the parameters V
.. ,
00,
Wx, w y , or W z •
If the non-stationary disturbances are Ilnite, nOll-linearity becomes
a noticeable feature. It is reflected in expansion m.1.4) where there
are several second-order terms. The latter are determined by stability
derivatives of the second order (for pxarnple, c~i3 and c~(')\) that are
in the fourth group. ."
InYestigations show that the influence of the derivatives on the
aerodynamic coefficients is not the same and that a practical sig-
nificance is a featllre of only a part of such derivatives, among which
second order derivatives form a quite small fraction. This influence
is analysed ill each specific case dependillg" Oil lhe aerodynamic COll-
f"lgnration of a craft and the conditions of its motion. A.s a result, we
lind the derivatives which the aerodynamic coefficients (or the cor-
responding forces and moments) de'pend Oil and the influence of
which of them may be ignorer!. For each coefficient, we can reveal
the characteristic trend associated with such a relation. Let ItS con-
sider, for example, the longitudinal (axial) force coeHicient. the
expression for which will be written in the form
(\1.2.1)
where C xo is the valne of C x at ex .~ ~ = 8" = 8 r = 0, C~2, • . , c~(j"
are the seconrl-order partial derivatives of tlw iongitll<iinal foro>
coefficient with respect to ex 2 , . . . . a8 e (for example, ('~" =
= fi 2 cx /Da 2 ).
In accordance with expression (D.2.1), the longitudinal force coeffi-
cient is a quantity not depending on the rotary derivatives or the
acceleration derivatives. This quantity is determine(l in the form
400 Pt. I. TI->eory. Aerodynamics of an Airfoil and a Wing
(9.2.4)
------------------------_.--_.
404 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
Fig. 9.3.1
To the conversion of the aerody-
namic coefficients and their
stability derivatives from one x
reduction centre to another
9.3. Conversion
of Stability Derivatives
upon a Change in the Position
of the Force Reduction Centre
(9.3.2)
For this purpose, let us first consider the relations between a, a. (oz.
~z, and aI' aI' (OZl' (OZl' A glance at Fig. 9.3.1 reveals that
a1 = a + ~a = a - (Oz';
;. .
where x= xll, (oz = QzllV 00, and, consequently, a 1 = a - (OzI.
The angular velocities and their derivatives do not change upon
a change in the reduction centres, i.e. (On = (oz, and ~Zl = ~z'
Using these relations, let us reduce the formula for the moment
coefficient:
(9.3.3)
expressed in terms of the new kinematic parameters (for point OJ)
to the form
. . . .
mz = m~la + m~la + (m~"Zl - m~lx) (Oz + (m~)Zl- m~x) ~z (9.3.4)
By comparing (9.3.1) and (9.3.4) we find
0:;1 0:;
.
a1 = m ex '
. (U Ul a- (v 1 til
mz = m z , m,z z m z zl = m,_ Z --r m z x,
I
n~ z z = 1n z Z ,1nz·x
I :;.-
(9.3.5)
The obtained moment coefficient can be expressed in terms of
its value found for the new centre 0 1 by means of the formula In z =
= m Zl + cyx (Fig. 9.3.1). Here formula (9.3.2) can be substituted
for mw and c y written in the form of a relation similar to (9.3.5):
(9.3.6)
By comparing the expression obtained for nL z = mZl -1- cyx with
the corresponding relation (9.3.6), we find formulas for the stability
derivatives relative to point 01:
mal
..
-
zl
= ma -
-
z
--
cax mUlZl
y' zl
.
= m z
(i)
z
+ (maz -
...
- W
cy z)
-
X-
--
CCl.X2
y
} (9.3.7)
(9.3.8)
I
Cy=C y a, Cz=C z
The division of the total motion of a craft into these two modes
is possible if we assume that the control system functions perfectly
and during the entire t1ight ensures the equality of the moments
M", My, and M z to zero. Such a craft and its control system are
considered as inertiafree (fast-response). The assumption on the absence
Qf inertia signifies that when control surfaces are det1ected, the angle
Qf attack and the sideslip angle instantaneously (or sufficiently
rapidly) take on values corresponding to a statically stable position
Qf the craft. In these conditions, the motion of its centre of mass
in the plane of t1ight is inyestigated independently. When perform-
ing such investigation, we shall write the aerodynamic coefficients
in the form
cx=cx(MQG, a, B, 6e, 6 r), cy=c{/(Moo, e.G, 6e)
Cz = Cz (Moo, B, 6 r ) (9.4.3) 1
lny=rtly(M oe , B, 6 T ) , mz=mz(Moc, a, 6e) f
Assuming that my = m, = 0, we can fmd the elevator and rudder
angles corresponding to the required angle of attack and sideslip
angle, and to the given trajectory.
The possibility of using such a method of investigation of the
trajectory is based on the low sensitivity of displacement of the centre
of mass to the rotation of the craft about this centre up to the instant
Table 9.4.1
Variables
Motion
CG
I B
I
Q
x
I
Qy
I Qz
I <'Ie
I <'I r
I <'Ia
pitching
Motion without
- - + - - - - +
pitching and rolling - - - - - - - -
I
408 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
Pitching Motion
x~()
~~~~~~~-~O::--
Fig. 9.4.1
Particular cases of motion of a craft:
a-at a zero angle of attack; b-with constant orientation of the craft (~ = const); c-ale,ng
a straight trajectory
(a) J V .
II~? (6)
, w"
-\~"J
Fig. 9.4.2
Modes of motion of a craft:
a-perfect loop; b-free fall
are simpler, but are encountered less frequently. Each of them allows
one to understand the distinction between the modes of motion
. .
°
characterized by two conditions: a = 0, Qz =1= aud a =1= 0, Qz = 0,
and accordingly to gain an idea of the derivatives m~ and m~l.
The three modes of sinusoidal motions indicated above can be-
treated as combinations of these two modes.
The first of.them (Fig. 9.4.2a) is characterized by the fact that the
angle of attack between the instantaneous direction 3 of the vector
410 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
Fig. 9.4.3
.- v
I X= a
Banking of a craft
- V 00 and body axis 1 does not change and, consequently, the deri-
vative ~ = O. In this motion, however, a varying angle {t forms
between fixed direction 2 and axis 1. Therefore Qz = dfr/dt =1= O.
In a particular case when Qz = const, the trajectory in Fig. 9.4.2a
<;haracterizes the motion of a craft performing a perfect loop.
Figure 9.4.2b shows a second mode of motion corresponding to
the conditions ~ =1= 0 and Qz = O. This motion may occur if free
fall with a velocity of w = gt (where g is the acceleration of free
fall) is superposed on the longitudinal motion. Here the angle {t
between the fixed direction 2 and body axis 1 remains constant,
while the angle ex changes (between the instantaneous flight direc-
tion 4 and axis 1). Consequently, the derivatives -it = Qz = 0 and
. .
a =1= 0; in the given case the derivative ex = const.
.
In a more general case (for example, in sinusoidal motions), it is
necessary to use both derivatives m~ and m~z. If one of the consid-
ered modes of motion dominates, only one derivative is signifi-
<;ant, and it is exactly what should be used. Figure 9.4.3 shows
a wing for which translational motion along its longitudinal axis,
at a zero angle of attack. is supplemented with rotation about this
axis at the angular velocity Q x = dr/dt. The non-zero variables
<Ox and Wx correspond to this mode of motion. If the rotation about
Definition
nents of the angular velocity [Qx (t), Q y (t), Qz (t»), and the pitch-
ing, yawing (course angle) and rolling angles [it (t), 1p (t), and
l' (t)1.
412 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
y y
x o~---------------------x~
Fig. 9.5.1
Modes of motion of a craft:
1-stable; 2-unstable; A-amplitude of oscillations
Let us assume that the values of the same variables V?x, (t), aO (t)
etc. correspond to a given undisturbed motion. If at sufficiently
small initial deviations (at the initial disturbances) ~ V?x" ~ao, ...
(here ~ V~ = V~ - V~, ~ao = a'o - aO, ... ) the subsequent
deviations ~ V = V
00 00 V~, ~a = a - aO do not exceed certain
-
Stability Charaderistics
Let us consider the characteristics of dynamic stability, and also
the role and place of the aerodynamic coefficients (stability deriva-
tives) in investigating the dynamics of a flight using the example of
translational motion of a craft also performing rotation (oscilla-
tions) about the axis Oz (Fig. 9.4.1c).
We shall write the equation of such disturbed motion in the form
(9.5.1)
where
(9.5.2)
aa = -maqooSl/J
z z
Here J z is the principal central moment of inertia of the craft
about the axis Oz that is one of the principal central axes.
Taking into account that m~z = m~, we obtain for the sum of the
coeffic i en ts
(9.5.3)
With a view to the value of the coefficient a 3 (9.5.1) can be
written as
.. . .
aJz/(qooSl) - 2amCl.z - am az = 0 (9.5.4)
Assuming that the coefficients ai are constant, we find a solution
of Eq. (9.5.4) in the following form:
a=ClePlt+C2eP2t (9.5.5)
Here PI and P2 are the roots of the characteristic equation p2 +
+ (al + a2) P + a 3 = 0 determined by the formula
(9.5.6)
We find the constants CI and C 2 from the conditions' that can be
determined by the derivative a = ao and the zero disturbance
a = 0 at the instant t = O. For these conditions, the constants are
Cl=C2=0.5~o[0.25 (a j +a2)2-aarl/2
and the solution becomes
a = 0.5~ob-je-"'lt (e bt _ e- bt ) (9.5.7)
where b = VA,~ - a 3 , A, = 0.5 (a l + a2)'
-------------------
414 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
in which b = VI aa 1 - Ai·
Examination of (9.5.8) reveals that the change in the angle of
attack has the nature of periodic oscillations. Since the quantity Ai
is always positive, these oscillations are damped ones; consequent-
ly, we have to do with oscillatory stability.
Let us consider the characteristics of this stability. The period
of the oscillations is
T = 2n (Im~lqooSlJ;1- Ai) -1/2 (9.5.9)
and their frequency is
ffi = "2n/T = (I m~ 1qooSlJ; 1 - AI) 1/2 (9.5.10)
The oscillation frequency is influenced mainly by the degree of
static stability, whereas the influence of damping is not great.
The nature of action of these factors is different. An increa~e in the
degree of static stability leads to a higher.frequency, while an increase
in damping, conversely, leads to its diminishing somewhat.
The logarithmic decrement is
(9.5.11)
where
(9.5.12)
The larger the coefficient me; (or m~z), the greater is the logarith-
mic decrement and the more rapidly do the oscillations damp in
time.
One of the important characteristics of oscillatory motion is the
time to damp to half amplitude:
(9.5.13)
\Vhen investigating oscillatory stability, the quantity AI can be
considered as an independent characteristic of this stability called
the damping coefficient. A glance at (9.5.12) and (9.5.13) reveals that
its magnitude does not depend on the degree of static stability m~.
Let us consider the wavelength of oscillations
Aerodynamic Coefficients
M x = - .\ ) 1'1 pz dx dz (9.6.1)
(8)
(9.6.3)
Ch. 9. Aerodynamic Characteristics in Unsteady Motion 417
\~
\ \
Fig.9.6.t
To the determination of the
aerodynamic coefficients for a
lifting surface z
• 0
(9.6.5)
(9.6.7)
. .
where pO: = I1pO:, prX = I1pa, etc. are the derivatives of the differ-
ence of the pressure coefficients.
By (9.6.7), the aerodynamic coefficients can be written as
cy =
3
2J (C~iqi + C'I i qi )
1=1
3
• •
••
l
")
1nUlX m Wz
z z, z
.
mqi = ma mWx mWz
x x' x' x
Assuming that the craft length Xk has been adopted as the charac-
teristic geometric dimension, we obtain from (9.6.3), (9.6.7) and
(9.6.8) expressions for the derivatives of the aerodynamic coefficients
in terms of the derivatives of the difference of the pressure coeffi-
Ch. 9. Aerodynamic Characteristics in Unsteady Motion 419
dents:
60
~ pqi~d'Sd~,
e
"1
(9.6.9)
1/(2x,) ~O
1/(2x k ) 60 1/(2x k ) £0
mqi = 2xt
z s Jr ~\. p flit.- dt.- dt-' m';iz = 2x'f,
s J
\ r
JP
~it. dt. dt
- - -
o £1 0 ~l
27*
420 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
Cauchy-Lagrange Integral
Accordingly,:
(1!p) grad p = grad P (9.6.12)
We can therefore write (9.6.10) as
grad (8c[J/8t V2/2 + + P) = 0
Hence we obtain an equation suitable for the entire space of
a disturbed flow near a lifting surface including the vortex wake:
8<D18t V2/2 +
P = F (t) + (9.6.13)
where F (t) is an arbitrary function of time determined with the
aid of the undisturbed flow variables for which 8<D18t = O.
AssumiNg that V2 = V~ and P = P 00, we find
F (t) = V~/2 +p 00
Consequently,
f)rJ)/8t + V2/2 + P = V!,/2 P 00 (9.6.14) +
This equation is known as the Cauchy-Lagrange integral. For an
incompressible medium, we have p = p = const. Therefore, P =
00
point in space is also small. With this in view, we find the linearized
value of the difference of the functions P - P 00. For an isentropic
flow, this difference is
p - P 00 = [kl(k - 1)] (pip - p oo/p 00)
We determine the ratio pip from the adiabat equation pip" =
= p oo/p~, in accordance with which
pip = (poo/poo) (plpoo)(I<-l)/k
Introducing p = p 00 +
p' into this expression, we obtain
pip = (p oo/p 00) (1 p'lp 00)(1<-1)/1< +
Expanding this expression into a binomial series, we find with
an accuracy within linear terms that
..J!... =..E::::..
P poo
(1 + k-l
k
L)
poo
Accordingly,
p - Poo = p'/pOO = (p - poo)/poo (9.6.17)
Hence, a linearized expression of the Cauchy-Lagrange integral
suitable for investigating nearly uniform unsteady flows has the
form
p - p 00 = -p 00 (V 00 a~/ax a~/at) (9.6.18) +
where ~ = <D - <D is the additional disturbance potential (<D is
00 00
i=1 i=1
3 , •
+~
i
((jilllqi+q;qi ~~) J (9.6.23)
With a view to the obtained value for qf, we find the following
expressions for the derivatives of the difference of the pressure
coefficients from (9.6.23):
q. -
p , = - 4 [o<pQi/8"£ + (pf)2<pQI1;
-"
pql = - 4 (o<pqi/o"£ + <pqi) (9.6.24)
-'-
Ch. 9. Aerodynamic Characteristics in Unsteady Motion 423
Wave Equation
(9.6.26)
Let 11S exclude the dynamic variables from this equation, retain-
ing only the kinematic ones. To do this, we shall use the Cauchy-
Lagrange integral (9.G.18') from which we shall nnd expressions for
the pressure derivatives with respect to the corresponding coor-
dinates x, y, z, and also the time t related to the density:
(1/p) fJp/fJx, (1/p) fJp/fJy, (1/p) fJp/fJz
For this purpose, we shall differentiate (9.6.18) consecutively
with respect to x, y, Z, and t:
_1_.
f'
ap _
ax -
p""[V
p . 00
02(p _ ~~)
ax 2 ax at 1
_1_.~
p ay
=poop (Voo~-~)
ax ay oyat I
_1_. Dp _poo (V ~-~)
p Dz - P ax ot OZ at
00
I
( (9.6.27)
af{l/ay +
Q x (t) z Qz (t) x +
ex. (t) V = 0 (9.6.32) + 00
Method of Sources
Fig. 9.7.1
'To the determination of the velocity potential of non-stationary sources
(9.7.2)
-where
r 1 =V(x-Moor2 )2- a '2(y2+ Z 2) }
(9.7.4)
Moo = Voo/a oo , a'=VM~-1
The upper limit of the integrals r; = a 00 (t - r' is deter- tn -
mined by the instant t;
at which point 0 moves to point E on nor-
mal P E to the Mach line. The first term in (9.7.3) takes into account
the influence on point P of the sources on section AE (the leading
front of the sound wave), and the second, on section BE the
trailing front of the wave).
Let us consider the case when a source is located only at moving
point 0, and at the instant t point P is influenced by the disturban-
Ch. 9. Aerodynamic Characteristics in Unsteady Motion 427
~es from the spherical waves occurring- at the instants t~ and t~.
We shall find the potential at point P by formula (9.7.3) as a result
of a limiting process, assuming- r 2 -+ 0. We shall presume that
r~ r(,
q(tt)=_1-
aoo Jr qt(..!.L)dr
aoo 2, q(t2) =_1_ . q2 (..!2.)dr z
aoo , aoo
(9.7.5)
o 0
whflre tl = t - iltl ; t2 = t - ilt 2; iltl , ilt 2 are intervals during
whieh the signal is transmitted from the origin of coordinates to a
point (x, y, z) with the wave front and its rear surface.
With a view to (9.7.5), we have the following expression for the
velocity potential:
q(x, y, z, t)=q(tt)!r q(t2)lr; r=Vx2-a'2(y2+z:!) (9.7.6)
Here q (tl)lr determines the potential due to the source at point A
and q (t 2 )lr, at point B (Fig. 9.7.1).
We consider that at the instant t = °
sources with a strength of
q Cr, z, t) appear simultaneously on the entire area occupied by
the lifting surface. The potential produced by such sources at a point
with the coordinates Xl' YI' Zl can be written as
°
When solving problems on the unsteady flow over flat lifting
surfaces, we may assume that Yl = because to determine the aero-
dynamic characteristics, we must find the values of the potential
function on this surface.
The velocity potential qJ [see (9.7.7)], in addition to the conditions
on the surface of bodies, must satisfy the conditions on the vortex
sheet and at infinity, and also the Chaplygin-Zhukovsky condition
in accordance with which a flow does not bend around a trailing sub-
sonic edge, but is cast off it. In accordance with this condition, dis-
turbed velocities near such edges change continuously, therefore the
derivatives oqJ/ox, oqJ/oy, and oqJ/oz are continuous. The Chaplygin-
Zhukovsky condition for the supersonic flow over a lifting surface
with subsonic trailing edges is used in the form
Vortex Theory
fig. 9.7.2
Vortex system:
1-lifting surface; 2-bound vortex
with varying circulation; .3-sheet
of non-stationary vortices
!:I
r(x,t)
x
fig. 9.7.3
Vortex pattern of rectangular
lifting surface
passing through the same particles does not depend on the time.
Consequently, when the circulation r (t) changes, a system of free
vortices with the circulation -L1r appears behind the lifting surface
that compensates the change in the circulation up to its initial
value. Accordingly, a vortex sheet forms behind the lifting surface
that consists both of longitudinal vortices parallel to the vector V 00
and moving along with the flow, and of lateral (bound) vortices that
are stationary relative to the lifting surface.
The intensity of vortex filament distribution in the sheet along
the longitudinal axis Ox is y = -dr/dx = - (1!V (0) dr/dt. It fol-
lows from coupling equation (6.4.8) that the circulation is propor-
tional to the lift force coefficient Cy' With a linear dependence of c y
on the angle of attack (c y = cija) , the circulation is also a linear
function of a, i.e. r = raa (where r a = dr/da = const). Therefore,
the strength of the vortex sheet y = - (ra/V (0) da/dt depends
on the rate of change in the angle of attack.
Figure 9.7.3 shows a vortex pattern modelling a rectangular
lifting surface. It consists of straight bound vortices with a vortex
sheet cast off them.
For a lifting surface of an intricate planform, the vortex pattern
consists of a number of bound vortex filaments each of which is
replaced with several discrete bound vortices from which a pair
430 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
V_ J
o .Ir-_ _ _ _ _ _z_
Fig. 9.7.4 z
Vortex model of intricate lifting
surface:
l-discrete oblique horseshoe 'vor- J
tex; 2-vortex filaments consisting
of discrete oblique vortices; 3-vortex
sheet x
of free vortex filaments is shed. Such vortex patterns are called ob-
lique horseshoe vortices (Fig. 9.7.4)
Let us consider the Kutta-Zhukovsky theorem allowing us to
determine the aerodynamic loads acting on a surface element in
unsteady flow. This theorem relates to circulation flow that is at-
tended by the appearance of a trailing vortex and circulation over the
contour enveloping the lifting surface.
According to the Kutta-Zhukovsky theorem, the pressure differ-
ence on the bottom and upper sides of a surface element in a linear-
ized flow is
(9.7.11)
where I' = (1/V 00) iJflat is the linear strength of the bound vortices,
and V 00 is the velocity of translational motion.
Formula (9.7.11) is an application of the Kutta-Zhukovsky theo-
rem to an arbitrary unsteady flow over a thin lifting surface and
indicates the absence of the influence of free vortices on the aero-
dynamic loads. From this theorem, particularly, there follows the
absence of a pressure difference on a vortex sheet consisting of free
vortices.
Let us write the strength of a vortex layer in the form of a series:
3 •
"\:1
I' = V 00 L: (I' iqi
i=1
qq'
+y iqi) (9.7.12)
Fig. 9.7.5
Vortex model for non-circula-
tory flow:
l-lifting surface; 2-discrcte closed x 2
a ttacbed vortices
(9.7.15)
The above relations are exact if the kinematic variables change·
harmonically.
For circulation problems, the Chaplygin-Zhukovsky condition on
the passing off of the flow from the trailing edge of a surface and
on the finiteness of the velocity at this edge is satisfied. According
to this condition, the strength of the bound vortices on the trailing-
edges is zero, i.e.
y (x, z, t) = 0 (9.7.16)
\Vhen bodies move at very low speeds, or when oscillations occur in
the absence of translational motion, a scheme of non-circulatory flow
is realized. Here no wake forms behind a body, and the circulation
over an arbitrary contour enveloping the lifting surface is zero.
Accordingly, a vortex layer equivalent to the lifting surface is
represented as a system of closed vortex filaments of constant strength
(Fig. 9.7.5). Some results of studying non-circulatory flow, in partic-
ular the derivation of Zhukovsky's theorem for such flow, are-
given in [4, 19].
8 f (La/2)tanx, D,-(LaM] , v=
~2 0 A [(L a/2jto1l')(:,P, {La/2}]
x zU")
,/ r'"~")J~
L 2
x(~) I
Lo/2 I
'Fig. 9.8.t
Oblique horseshoe vortex in an unsteady flow:
1, 2-free vortices
where 70 = lo/b.
The total velocity induced by the bound and free vortices is
W = W' +
WIt. It will be more convenient in the following to
Ch. 9. Aerodynamic Characteristics in Unsteady Motion 433
Fig.9.S.2
To the determination of the
velocities induced by a vortex
sheet and free vortex filaments
of varying strength:
1. 2-free vortices
~ x ) = T·V"
w "(!:so' ~o, v ('"In IV 00 r) = - (i+COS
_ ~ Bl
i~cos ~2) (9.8.4)
lo-("o lo+ ~o
r1( x, ° - 0 0 - x-o.5l
t)-r(t Voo
o tanx)}
(9.8.7)
r 2 (x, t )- r °(t _ x + 0.V51 tan x )
0 - 0
0
00
r
00
I
} (9.8.9)
V"- __1_
00
J
l'2(X, to)(()·510+~o)d~ I
2- 4:rtb
- o.5lotan x
[(~o - ~)2 +
(0.5/d- ~O)2J3/2
J
The total velocity induced by the bound vortex, the vortex sheet,
and the free vortices is
V = W' + V' + + V; V;
(9.8.10,
Let us consider a harmonic change in the circulation
f 0 (t) = V oobf sin pt (9.8.11)
where p is the angular frequency and f is a dimensionless constan t.
In accordance with this law, we shall write the circulation (9.8Ji}
in the form
y (x, to) = -bfp cos ptl = - V oop*f cos (pt o - p*~) (9.8.12)
Ch. 9. Aerodynamic Characteristics in Unsteady Motion 435
I~ IX,
( t 0 )="
t.x
·bl' SIll (t -
. [ Po X-I,.SlntanX)]
1'00
~v.P = ~0.p+O.5bp
- ( 1-cos7
vn ) )
v = 0, 1, 2, ... , n
The cells arranged along the line v separating them along the span
of the wing form panels whose number is the same as that of the
cells in a strip (n). We shall designate the boundaries of the panels
Ch. 9. Aerodynamic Characteristics in Unsteady Motion 437
Fig.9.S.3
Vortex model of a wing in an unsteady circulatory now:
I-wing; 2-obliquc horseshoe vorl<'x; 3-cells; 4-strip; 5-pands; 6-vortrx \\'ake
Points wilh identical u's coincide with the ends of discrete vor-
tices. vVe have the follo\~ingrelations for the dimensionless coordi-
nates of the middles of the oblique bound vortices, their span lk,k -1'
and sweep ana-Ies
,.,
tan x[J., h
f,l, k-l
:
tan xt::
::-1 = (£[1,1;-1- Sfl, h)/(SI1, /;-1 - S[1, It)
I
fl=1,2, ... , n; k=1, 2, ... , N I
where the variables for the port side of the wing are designated by
the factor a.
For a wing with rectangular edges, the sweep angles of the bound
vortices are constant (x" = const), and each of them transforms
into an ordinary straight horseshoe vortex. If a wing plane has
breaks in its contour, then the corresponding breaks are present in
the vortex filaments too.
The circulation of oblique and conventional (straight) horseshoe
vortices along the span is constant, while free filaments parallel
to the axis Ox are shed off the ends of the bound vortices. The free
filaments propagate downstream, and when the flow is circulatory
they pass away to inflllity. In addition, when the circulation of
the bound vortices changes with time, free vortices of the relevant
strength are shed off them.
The vortex model being considered is very convenient for comput-
erized calculations of a flow. This is due, first, to the sufficiently
simple relations describing the disturbed flow near a wing, and,
second. to a number of important properties of the system of alge-
braic equations which the solution of the problem is reduced to.
One of these properties is that the diagonal terms in the matrix of
the equation coefficients play a dominating role; the solutions'
themselves have a great stability relative to the initial conditions.
A significant feature of computerized calculations is also the fact
that the use of oblique horseshoe vortices instead of the convention-
al ones leads to suhstantial simplifIcation of the calculations and
to more accurate results.
Ch. 9. Aerodynamic Characteristics in Unsteady Motion 439
I
I
Pi*r fl,
qi
Ii, k - 1 cos [ Pi* (T - tfl, k
"'fl, h - 1
) l} (9.8.29)
where ~~:"k-1 is the longitudinal coordinate of the middle of an
oblique bound vortex.
The expressions for the circulation of the free vortices (9.8.7)
have the form
3
r 1 (2) (~r.: L I,T) = V ",b oi=1
~ q"( [r~~ 1<, k-1 cos {pi [ T - (~r.: Ll
-or:-z
+ t .fl,k
.J k,lI_1auX;l, ;1-1
)l} - f qi
fl, k, k-1Pi* sln
. {*[
Pi T - (1Ofl,1i
Sfl,1I-1
-----------------_._------_.----------
440 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
less velocities V(l) and V(2) that are evaluated in accordance with
general relations similar to (9.8.14) and (9.8.15).
Let us designate by (JU(l) and (JV(2) the additional velocities at
the control point being considered produced by an oblique vortex
located on the port half of the wing symmetrically about the plane
xOy. The coordinates of the middle of such a vortex are S~:~ -1 and
-~~:~ -1. 1 The values of (JV(l) and (JV(2) are determined with the aid
of (9.8.1), (9.8.8) and (9.8.9), account being taken of the ehange in
the direction of the coordinate axes and the transfer of their origin.
We shall add these values of the corresponding dimensionless
q.
.
velocities V(I) and V(2) for symmetric motion (for qi = ex, WZ, and
q.
even functions r ~~k.k -1' r ~~k.k -1) and subtract them for asymmetric
motion (for qi = Wx and odd derivatives of the dimensionless circu-
lation). We shall determine the total velocity at a control point with
account taken of the influence of all the other vortices, i.e. calculate
it by double summation of the induced velocities over the number
of strips N and panels n. This velocity is determined by the values
of the derivatives of the dimensionless circulation r~~k.k -1 and
q.
r ~:k.k-l'
To find them, we must use the condition of flow without separa-
tion in accordance with which the total dimensionless velocity at
a control point induced by the entire vortex pattern of the wing
equals the undisturbed component determined by Eq. (9.6.32'):
_ VV,P,P-l _ (W'+v'+vi'+v;)V,P,P-l
Vv. p. P-l- Voo - Foe
1'v,P tV,P
= - Wx\,v, p-1 - wzsv • p-1 - ex (9.8.32)
Ch. 9. Aerodynamic Characteristics in Unsteady Motion 441
(9.8.34).
"" )1,
L.J.......
(V(2»).1, h, h-l +av(2»).1, h. h-I) r qi
v, P. p-l - v, 1', 1'-1 ).1, h. h-I
h=1 ).1=1
N n
+p ~ t
"" ""
L.J L.J (V(I)Il,h,h-l+ av (I),I,h,h-l)r Qi
v. p, p - 1 - v, 1', l' - I ,I, h, h - I -
-0
h=I).1=1
p=O, 1,2, ... ,N; v=O, 1,2, ... , n-l;
k = 1, 2, 3, ... , "v J
The quantity v~~P'P-l is determined in accordance with Eq. (9.8.32)
as follows:
a
V v, p, p - l -
__ '1', VV
<Ox
•
_ rV, p • <00
p, p - l - - " v , 1'-1, V v , p, )1-1--Sv, ]1-1
- f" (9.. 0.
Q 3-)
:J
(9.8.36)1
rV , p +rfL, k
~= 'ov, 1'-1 'o}.t, k-l I
N "
_L~ * ~ (V(2)Il,k,h-l+ av (2)fL,h,k-l)r a - =1
4[[ L...J L.J v, p, p - 1 v, p, p - 1 Il, 1<, il - 1
il=1 1l=1
N n
~ )' (v(2)/J, h, k-l -l-av(2)fl, h, k-l) r afL, (9.8.39)
L.J ~ v, p, 1'-1 I v, 1', )J-l k, k-l
I
h=lfL=1
J\T n .:...
+ p* :6 ~ h 1
(vV,)/:,'ph_'l - +av~I,)bt::.:}-I) r~, h, h-l =0
1<=1/1=1
r~. h. h-l!/1=Il¥ =
p==O, 1, 2, ... , N; v==O, 1,2, ... , n-l;
0, r~. h. h-l !1l=fL¥ = °
k = 1, 2, 3, ... , N
Ch, 9, Aerodynamic Characteristics in Unsteady Motion 443
lY
+
II
!Y n •
I 1<, I l - l + Uv (,1)J.1, 11, "-I)f U
--1--,
P *)'
_ V \, l',l,
_ (v(,1) [J - 1 \ , )', J> - 1
)', ,
Il, I, , I, - 1
=0
/(=11l =1
_I:....
4.'1
~ ~
L.i L.i
(v(2)J.1,h,I1-t_
\', [', J> - 1
(2JJ.1,h,h-l)r;')(
uv \', )1, I' - 1 Il, Ii, I, - 1
1<=1 w~1
(9.8.41)
rU)e
Il, 1<,1<-1
1 °
,ll=J.1¥-~'-' r(U,-
p, 11, h-l I"=",,, =(J
p=O, 1, 2, ... ,1\'; \'=0, 1,2, 11-1;
k c-= 1, 2, 3, ... , iV
444 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
k=II-'=1 q. (9.8.43)
rl-", k, P= 1, 2, ••• , N;
k-l 11-'=1-1,,=0,
q,
r /.t', k, k-l = 0;
P = 1, 2, ... , N;
II-I=/.t"
v=1,2, ... , n-1 k=1, 2, ... , N
Consequently, the problem on the oscillations of a wing with
low Strouhal numbers reduces to solving six (instead of three)
systems of equations but simpler ones than for arbitrary values
of pi. The first system of equations (9.8.43) contains no derivatives
ra:k,k-1 and is solved independently of system (9.8.44) for the
derivative r~:k,k -1' Its value is used to calculate the right-h~nd
side of the equations of the second system (9.8.44), the solutIOn
yielding the derivative r~:k,k-l'
The dimensionless coefficients of equation (9.8.43) and (9.8.44)
are determined from the relations
VI-l,k,k-l=V(1:I-I,Il,k-l .;..J,l, l~, k-l l
v, p, p-l _v, p, 1'-1, sv, p, l' - I, k, Ii - 1 •
UVI-l,k,k-l_ (tl-l,h,h-I rl-l,k,k-I l /.t,k)
v, p, p_!-V Sv, p, p-I, U sv , p, fl-l, k, k-l, -XI-I,I<-1
rlv(2)I-I, k, k-l ~,,(2)
v, /J. p-l Uv
(9.8.45)
Ch. 9. Aerodynamic Characteristics in Unsteady Motion 445
Let us write Eqs. (9.8.43) and (9.8.44) ~with account taken of the
specific values of qi and qil and also of the nature of motion, Le.
symmetric or asymmetric.
The system of equations for determining the derivatives r~,k'''-l
has the following form:
N n
1
4n:
)'
.;;...J
h=1 f.L=1
"" ( ,,",
.L.J vv,
h, h - 1.
p, p-l ,avv,,",, p,
k, Ii --1)
1'-1
ref-,,", k, k-1 =-
1
II.
I (9.8.46)
r~, h, h-llf.L=fl .. =O; p= 1,2, ... , II{; V= 1,2 .... , n-1; I
k= 1, 2, .. "' JV J
-
We find the derivative r~.R.k-l by solving a second system of
equations:
lV 11.
v .=
rt. k
1, 2, ... , n - 1;
,~h-1 1f.L=r t .. =
k
0; p = 1,
= 1, 2, ... ,
2, ... , N;
j\j
N n
1 "" "" ( f.L, k, k-1, f.L, 1<, 1<-1) rWz _ 1:;v, p
4n: .L.J .L.J Vv , p, p-l--r(i Vv, p, p-1 f.L, k, k-1--"'V, p-1
k=1 f.L=1
(9.8.48)
r~~ k, k-11f.L=I1 .. ~-" 0;
p=1, 2, ... , IV; v=1, 2, ... , n-1; k=1, 2, .. -,
446 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
I
N n
4n
1 "
.LJ
'Y (vv,
1.1,
..:...J
k, 11-1 1.1,11, Ii-I)
p, p_I-(JV v , p, p-J
w,
rl.1, 11, k-l=-Sv,
v, p
p-l
R=II.1=1 } (9.8.50)
r~~ k, k-III1=I.1¥=O; p=1, 2, ... , N; v=1, 2, ... , n-1; I
k= 1, 2, ... , N )
N n
"
.LJ " ( 11,11, k-l k, k-I)
.LJ v v , p, p_l-(Jv v1.1,, p, p-l
rJJX
1.1, k, k-l
k=1 11=1
Aerodynamic Characteristics
t'I1, k
'\' ( £11,k-l,
til, k
_11,k-l
)
=
r (t'fl, It
':;11,k-J,
~~t, II ) . A"
\'11,k-I/Ll~~t,k,k-l
(9 "8 54)
Let us write the strength of the vortex layer and the circulation
of a discrete vortex in the form of the series (9.7.12) and (9.8.25):
(9.8.55)
(9.8.56)
With account taken of these derivatives, the expressions for the
Strength of the vortices are determined by the quantities pJ: k.k-l
and P~~k,k-l (9.8.53). Let us introduce them into formulas (9.6.5)
in which ~p is replaced 'with series (9.8.52), the value of Xk is taken
equal to b o and x' to the local chord bk ,k-l:
(9.8 ..57}
448 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
(9.8.58)
(9.8.60)
,qi 2bg
iJ,=~l ..
q. iJ, k 1
m z , k, k-1 = 2
bk , k-1
~ riJ,', k, k-1SiJ,: k-1
I
iJ,=iJ,o
iJ,=iJ,,, . } (9.8.61)
,qi
m z , k, k-1 = 2
2bg
~
q. iJ, k
riJ,', k, k-1SiJ,: k-1 I
bk ,k-1 iJ,=iJ,o J
(9.8.62)
tics of a wing:
lo/(2b ol
Cy = ~~ .1
o
10/(2 bol
nl z = 2
Sw .1' (9.8.63)
o
lo/(2b ol
mx = o
_-_2b_ \' C~ b (~) d~
Su,
o ,
""7e can express the coefficients cY ' m z , and mx for a wing in the
form of series (9.6.8), and their corresponding values c~, m~, and
m~ for the sections in the form of (9.8.58). After using the deriva-
tives (9.8.60)-(9.8.62), we obtain the stability derivatives for a wing
. .
qi cqi m qi m qi m qi and m qi )
as a whole (c Y' Y' z' z' x' x·
Going over from integrals to sums, we obtain for symmetric
motions (qi = ex, (j}z):
(9.8.64)
(9.8.65)
(9.8.66)
(9.8.67)
q. q.
mz'=m·z'=O (9.8.68)
(9.8.69)
29-01715
450 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
cy == Sw
4b~
N
L; -
Zh,h-1
n.
L; f a
ll ,k,h-1
I
h=1 1l=1 -'
cy
Wz
=
4b~ ~
Sw LJ
-Z
h, h-1.LJ
~ fWZ
1-', h, h-1
II
h=l 1-'=1 l
N n ( (9.8.72)
~z
cy =
4b8 ~ -
S w .LJ h, h - 1
~
.LJ f
~z
Il, h, h -1
I
J
h=1 1l-1
N n )
4b~ ~ - ~ fet I
mz
ex
= Sw .LJ Zh, h-1.LJ
til, h
Il, h, h-1'<>Il, h, h-1 I
h=1 1l=1 }
az
•
4b5 ~
N
I h, h-1.t::J
~
n.
fa tl-', h
I (9.8.73)
m = Sw .LJ 1-', h, h-l"ll, h, h-l J
h=1 1l=1
mz
Wz
=
N -
4bij ~ ~n Wz til, h
Sw .LJ Zh,h-1.LJ f ll ,h,h-l"l-',h-1
1
h;1 ~t:1 I (9.8.74)
~z4b~ ~ - ~ ~z
m z = Sw .LJ lh, h-1 LJ fll'
Il. h
h, h-1£I-', h, h-1 I
h=l 1l=1 ,
~ = -a(,;)-Ulx(';)S-- Ulz(,;}G
+~ off,
{j (,;) h (~,
j
I ..:.
~) (j (,;) (9.8.7U)
----,,-----------
452 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
Influence of Compressibility
(the Number Moo)
on Non-Stationary Flow
I
Let us introduce the functions 'IjJ, 'ljJqi, and 'ljJqi related to the
velocity potential and its derivatives by the following expressions:
qJqi ='ljJqicosw~+ptk-2'IjJqisinw~
(9.8.81)
<Pq i = k-z'ljJq i cos w~ - p!-l'IjJq i sin w~
<Po = 'ljJo, w= M~k-2pi
where k = V1 - M~.
After the corresponding substitution for the derivatives in (9.8.80)
of their values from (9.8.81) and then introducing cP into (9.6.31),
we obtain the following three equations:
+ +
I
2 2 2
k2 iJ 1jJo iJ 1jJo iJ 1jJo = 0 (9.8.82)
iJ~2 iJ1']2 iJ~2
2
q 2 q 2 q
k 2 iJ 1jJ i
iJ~2 + iJ 1jJ i
iJ1']2 + iJ 1jJ i
iJ~2 + (p!)2 M~k-2'IjJq~ = 0
(9.8.83)
iJ2,I,qi iJ2,,,qi iJ 2 ,,,Qi
k2 iJ'Y~2 + iJ~2 + iJ~2 + (p!)Z M~k-2'IjJqi = 0
Ch. 9. Aerodvnamic Characteristics in Unsteady Motion 453
.
iJ21j;(!;
-'- iJ21jJ'1;
0112
.
oZ1jJ'1 i
+ 021j.<1;
0~2-
.
iJ2\jJ'! i
- U I (9.8.84)
k
2
iJ~2 + iJ'l2 + iJ~2 = 0
-Let .us consider a rigid lifting surface. Instead of the flow of a com-
pressi.ble fluid over it, we shall study the flow of an incompressible
fluid over the transformed surface. The coordinates of points in the
incompressible flow are related to the coordinates of the space of the
compressible gas by the expressions
XiC = x/k, YiC = Y, ZIC = Z (9.8.85)
or in the dimensionless form
~IC = xlc/bo,lc =~, Tjlc = Ylc!b o,IC = 11k,
~IC = zic/bo. Ic = ~k (9.8.86)
Let us change the planform of the given lifting surface, using the
relations between the coordinates in the form of (9.8.86). In accord-
ance with these relations, the dimensionless geometric variables
of the transformed surface can be written as follows:
~O,IC = ~ok, ~O,IC, = ~o blc (~Ie) = b(~) (9.8.87)
where ~o and ~o are the dimensionless coordinates of the leading
edge;
blc = ble/bo,IC, b = blbo, b le = b/k
For the centre and running chords in (9.8.87), we have introduced
the symbols bo and b, respectively.
For wings with straight edges, the planform is usually set by
dimensionless parameters, namely, the aspect ratio Aw = lolb m•
the taper ratio llw = bo/bt, and the sweep angle of the leading edge
Xo. The corresponding dimensionless geometric variables of the
transformed wing in an incompressible flow are:
Atc,w = kAw. lllc,w = llw, Alc,w tan Xo,lc = Aw tan XO (9.8.88)
where Alc,w = lle/bm,le, Aw = lIb m, 111c,w = bo,le/buc" and Tjw =
= bo/b t ·
Figure 9.8.4 shows the given and transformed wing planes. The
transformed plane retains its lateral dimensions and is extended in
a longitudinal direction in accordance with the relation Xic =
= x/Vi - M!,.
454 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
(aJ
lle
Fig. 9.8.4
Geometric dimensions of a wing:
a-given wing In a compressible flow; b-transformed wing in an incompressible flow
(9.8.90)
Let us now consider the motion of the transformed lifting surface
in an incompressible fluid. Let us write the disturbing potential
of the flow over the surface as follows:
3 •
~ (rn qi Ie
rn
'-Vic = k[rn
'-Vo +.l.J '-Vic' qi.
rnqi ic·
ie + \.j,'!c· qi. ic
)] (9.8.91)
i=1
flow over the surface, the potential must satisfy the continuity
equation
(9.8.92)
(9.8.93), we find that the functions '¢o, "ljJqi, and "ljJqi determining
the velocity potential for a wing in a compressible flow, and the
functions <Do, <D?J,ic, and <D7J,iC determining the potential for
a transformed wing in an incompressible flow satisfy the same
differential equations. It is shown in [19J that the boundary condi-
tions are identical. Accordingly, the values of these functions at
the points related by conditions (9.8.86) are also identical, i.e.
m~i=(mqi,.iC_lJ12
z
J )lk
Z, Ie 00 z
3 ,I
where
I _ 2b~,
X- Sw. ic
ic
.\.\ p~~' iCSic~ic dsic d~lc 1
Sw,k }
(9.8.98)
2b~ Ic
I z = --,;--'----
"
J'.\ p.IC
q i, icr:~
'::llC
dE·_lC d"t.
'::lIC
I
C W, ic S ,I
w, ic
(9.8.100)
J\ J\ r1_ iuy
:,CP
1 iPt
(P (XI. YI' ZI' t) = --2
IT
Re e (x, y, z)] y=o
a(x,. z,)
(x, - x).ll oc T r
- ip a 0:' (iP~ -iP~) dx dz
xe 00 e "" +e 00 -r-
x (9.9.5}
where
q, = qje iPt , qj = dqj/d. = (dqj/dt) bolV = ipjqje iPt 00 (9.9.7)
In accordance with (9.9.6), the partial derivative of the potential
function with respect to y is
q.
+ . * ay
flfP
lp J (
X, y, Z )] (9.9.8)
(9.9.9)
X cos W(X-X)
I
bo
+ p* [flCJj
_fP_ (x,
fly
y, Z) ]
y=o
(O(X-X)}
sin---,-....!.l--,-
bo
(Or ) dx dy
X cos ( boMoo r l
( (9.9.10)
•
<Pqj (x, y, z) =
1
-n- J J {[
\ r .
flCPqj (x, y, z) ]y=o
a(X1, Z1) fly
xcosW(XI-X) __l_[~(x, y,
~ ~ ~
q.
z)J
_0
sinW(XI-X)}
~
j
(Or dx dz
Xco s---
boMoo r
-.
Ch. 9. Aerodynamic Characteristics in Unsteady Motion 459
(8.9.12)
0"
x [ a~fJj ] } d~_d~
til'] 1) -0 r )
(b)
0'
fig. 9.9.1
Adopted designations of the coordinate systems:
a-conventional (~, 0, \;); b-transformed (§t, 0, St); s, r-characteristic ~oordinates
cp
qj (I:
.,1, t,
0 r
,.,1, t
)= _.2....
n 1•
r J [ on
oq>q j
"t
]
-0
11t-
d~: d~t
rt
\
at
~ (9.9.24)
462 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
After inserting the value of qJqj (9.9.25) into the first of Eqs.(9.9.24),
we obtain an integral equation for determining the derivative Fqj:
Fqj (s 1, t, °~
, 1, t
)= -~
n
rr[
j j
f)Fqj]
f)Tjt 1'1t=O
dStd~t
Tt
(9.9.26)
crt
(9.9.27)
Fig. 9.9.2
To the numerical calculation
of a wing with combined leading
and supersonic trailing edges in
an unsteady flow:
I, II-regions for finding the down-
wash
o 0 )
(9.9.31}
where
{)Fqj ]
D·=
]
[ --
ol]t 11t=O '
(9.9.32)
Let h stand for the width of a cell in characteristic axes. This quan-
tity is the dimensionless distance along the r ~or s) axis between the
-edges of a cell for points n, n 1 (or i, i +
1), respectively. It is +
equal to
hi = l/(2Nb o) (9.9.33)
where N is the number of parts into which the wing half-span l/2
is divided.
Let rand sbe the coordinates of a fixed point (point PI in Fig. 9.9.2)
at which we are finding the velocity potential (or its derivatives),
and m and i be the coordinates of the running points used to perform
numerical integration. Accordingly,
r1 = rh, 81 = sh, r = mh, 8 = ih (9.9.34)
Fq'(r p 0, 8 1)= - ~ ~ ~
m=O 2=0 V
L L mh Vill
DjdVdul
where
Dj,m,i=Dj[(m-1I2)h,
Ej,m,i=Ej[(m-1I2)h,
B-r,
(i-1/2)h]
(i-1/2)h]
- 1..=(Vr-m+1-Vr-m)(Vs-i+1-Vs-1)
I (!)'\:).37)
m, s,
I
D)I'~i = -1 (j = 1); D~I~i = h (m- i)/2 (j = 2) }
(I) ) (9.9.38)
D 3mi = -a'h (m+i-1)/2-a'£o, 0, t (j=3
ElI'~i=M;,[h(m+i-1)/2+£o,o,tl/a' (j=1)
E~%i = {M;, [h (m + i - 1)/2 + £0,0, tl/(2a')}
X h (m -1) (j = 2) (9.9.39)
E~I~i = - M;, [h (m+ i -1)/2 + £0,0, t]2 (j = 3)
Section II (Fig. 9.9.2) between the characteristic axes, the leading
edge and the Mach lines issuing from the vertex of the wing is occu-
pied by an undisturbed flow region; hence, all the downwashes equal
zero and, consequently,
DJ~~, • i = EJI~ " = 0
I ,
(9.9.40)
Let us determine the flow downwashes in disturbed region III
between the leading edge, the tip, and the Mach cone issuing from
the points of a break in the wing contour (Fig. 9.9.3). The velocity
potential in this region (outside of the wing and vortex wake behind
30-01715
466 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
Fig. 9.9.3
To the numerical calculation of
a wing with subsonic edges in
an unsteady flow:
III, IV-regions for finding the
downwash
~t
J JD
81
j
dr ds
V(r1- r) (S1-S)
=01
,
~ (9.9.41)
0 0
rl SI
q.
2~ J~ F dr ds
F (ri' 0,
J 8 1) = - j = 0J
V(rl-r) (S1- S)
0 0
J
0
dr
V r 1- r
iD
0
j
ds
V S1- S
=0; .\'
0
dr
V r 1- r .l E
0
j
ds
V S1- S
==0 (9.9.42')
l
Sl-
~ E)III) ds
- ds
s,-h
YSl-S JE
0
j
YSI-S
T,
~ E)III)
T,-h
~ dr I (9.9.44)
T,-h
dr
Yrl-r
= _
0
Ej
Yrl-r ,I
Assuming the values of the downwashes in the cells to be constant
and going over from integrals to sums, we find in accordance with
the variables (9.9.30) the following relations for determining the
downwashes in a cell (point) with the coordinates r 1 = and $1 = rh
= sh:
B~.
7-1
D}IlI) (rh, sh)= - ~ Dj[(m-0.5)Jz, (i-O.5) h] m 1
m=1
~ (9.9.45)
7-1
(III) -
Ej (rh,
-
sh)= -
~
L..J
m=1
Eif(m-O.5)h, (i-O.5)h]B-
r, m
I
J
(for points to the left of the axis ~d;
s-1 \
(III) - - ~
I
Dj (rh, sh) = - L..J D j [(m - 0.5) h,
;-1
i=1
(i - 0.5) h] B- .
s.' I (9.9.46)
,I
(III) - - ~
Ej (rh, sh)=-L..JE j [(m-O.5)h, (i-O.5)h]B-s ••.
i=1
(for points to the right of the axis ~t).
In Eqs. (9.9.45) and (9.9.46), we have introduced the notation
B-r m =Vr-m+1-V;:-m , B-.
s 1 =Vs--m+1-Vs-i
. . (9.9.47)
According to the above relations, the flow downwash at a point
is determined by summation of the downwash values in all the
cells in the corresponding strip m = const (or i = const). The
calculations are performed consecutively beginning from the cell
at the vertex of the wing where the downwash is determined by
formulas (9.9.38) and (9.9.39). Let us now calculate the downwashes
on the vortex sheet behind the wing (region IV in Fig. 9.9.3). We
shall find them from condition (9.9.20), \vhich we shall write in
the transformed coordinates
CPqj(Sb 0, ~t}=(Pqj(st. 0, ~n }
(9.9.48)
CPqj (st. 0, ~t)=cpqj (sf, 0, ~n-a'(st-'St)cpqj
30*
._---_._._-_._----_._----------
468 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
1 (9.9.49)
__
T1 S1
1 )
211:
D i j
dr ds qj ( r*+s*
2'
s* -; r* )
)
I}
Y(r1-r) (s1- s) F
0 0
rl Sl
-7J1
1 ~ ia E j
dr ds
- F q.J ( r*+s* s*-;r* ) (9.9.52)
l/(r1-r) (S1- S) -
I
2'
0
+ a' [(r1+s1)-(r*+s*1l
2
Fqj ( r*t s*
, r*-s* )
2 J
Ch. 9. Aerodynamic Characteristics in Unsteady Motion 469
+
Sl-
.1
h
ds
V S1- S I
Tl-h
Dj
dr
1/ r1- r
(9.9.53)
(9.9.55)
i=1 m=1
where B-;',m,-;,i is an influence function determined by the third
formula (9.9.37).
For the second and third integrals on the right-hand side of (9.9.53)"
we obtain the relations:
T1-h r-1
Jr =4h ~
ds
D.] -.1
dr
LJ D]·(mh, sh)B-T, m--
,s, s
(9.9.56)
o II r l - r m=1
~1 D
1
ds 'r. D. dr = 4h (rh, ih) B- __ . (9.9.57)
o
VSl-S J
r1-h
J Vrl-r
;=1
j
T, T, S, l
470 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
where
- -= Vr-m+ 1-Vr-m',
B-r,m,s,s B-r,r,s,'l.
- - .
=Vs-1+1-Vs-i (9.9.58)
For the fourth integral (9.9.53), we have the formula
rl d~ Trl D j dr = 4hD (rh, Sk) (9.9.59)
J
sl-h
VSl-S J
rt-h
rl-r
j
P
qj (1: r) = _ ~
:" '" a'
aF ] (st. ~t)
a~t
1
1
p;j (~, s) = - a\ { -a'Fqj(~t, St) ~ (9.9.63)
+M 2
00
a' [~
M;'"
aF'Ij(st,
aSt
stl - t
_t
aFQj(St,
a~t
~tl]} JI
4 {
pqj (~, S) = (?3 q. -
- a' F ] (r, s)
. } (9.9.65)
+ a~2 [Fqj (r+ 1, s+ 1) -Fqj (;, s)l
472 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
(9.9.68)
" ,
M2
a m· x = m x, t, a mx = m x 1,t +
'2 OOx OOx '4 OOx OOx OOx (9.9.69)
""mx2, t
lj( 2b o) st
~ ~ Fa (St, ~t) dS t d~t
-lj( 2b o) 60 , t
~ (9.9.70)
lj(2bo) st ~a
[~ \' r aF (~t. ~t) dS t d~t
M~ J J a~t
-1/(2bo) 6o. t
Ch. 9. Aerodynamic Characteristics in Unsteady Motion 473;
1/(2bo)
.*
·t
Ill z 4b5a'Aw (' \ aFlllz (~t, ~t)
Cy,t= 12 J J a~t
-1/( 2b ol 60 • t
1/(2bol st
J ~ F"lz (St. ~t) dS t d~t
-1/( 2b ol 60 , t ~ (9.9.71)
• Ij(2bol sf .
III z 4b5a' 2A,W [ a' r r aF'''z (£t,
ii~t
~t)
~d£t d~t
C y 2, t = /2 M!,.1 J
-1/(2b o) ~o, t
1/( 2b ol ~t III
~ ) aF za~~' ~) Sds d~ ]
-lj( 2b ol;;o,t
J
1/(2bo)
.*
6t
.\
- Ij( 2b o) ~o. t
)
1/(2bo) ~t
i J Fa (£t,
-1/( 2b ll) So. t
} (9.9.72)
1/(2bo) si
~"
2
a 200'
m~2,
- t=- _4_b,:",.oa"",'3_A,-"w_
12
[ M )
-1/( 2b o) ~o, t
J
Ill
mz. Zt =
(9.9.73,
1/(2bu) £t
~
- 1/( 2b o) ~o, t
J
474 Pf. I. Theory. Aerodynamics of an Airfoil and a Wing
1/(2bo) ~t .
4blla'3A. w
[2
[~
M2
r Jr
.1
aFwx (st. ~tl ~t dS t d~t
aSt
00 -1/(2bo) SO. t
1/(2b o) st w
J J aF Xaitt' ~t) ~tSt ds t d~t ] (9.9.75)
-1/( 2b o) So. t
(9.9.76)
Accordingly,
st
JF q
j (St. ~t) dS t
So. t
} (9.9.77)
st
JF q
j (St, ~tHt ds t
So. t
Ch. 9. Aerodynamic Characteristics in Unsteady Motion 475
St
- ~
So. t
~
Cy l,t= -
4bija' 2A w
12
h2 ~ ~ Fa(-r,
LJ LJ s)
s
~ (9.9.78)
ex 4bija'2t..w
Cy , 2t = 12
~ F z (- * -s*)
<U 4b2N'~ Ul
C y, t
Z
= o~ --,,'
J2 h LJ r ,
;:.
Ul Z 4blla'2t..w h 2 ~ ~ FUlz (- s)
Cyl, t = - l2 L.J L.J r,
s
Ul z 4b 20a'2t..... {a' "" Ul - -
(9.9.79)
cy2, t = 12 " h M!, ..:....i
y F z (r*, s*)
r*
476 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
s
.
m~2. t = 4b67~3AW h {+ :~ ~ 00 _
r(r* + s*) h/2 (9.9.80)
r*
r*
(9.9.83)
+ ~o. o. t[ FO (r*,
8*) + h ~ ~ [(8- i) h/2] F"'x (i, s)}
lX
s
In the above relations, the symbol 2J signifies summation over
r"
the cells along the trailing edge, and the symbol ~ double sum- 2J,
s
mation over all the cells on the lifting surface.
When performing calculations of the total aerodynamic character-
istics, one must have in view that these calculations are carried
out in the system of coordinate axes depicted in Fig. 9.6.1 (with
the rule of signs indicated in this figure). All the geometric quan-
tities are measured and the pressure coefficients are found in the
system of coordinates whose axis Oz passes through the vertex of
the wing (see Fig. 9.11.1). When the mean aerodynamic chord b o
is taken as the characteristic dimension, the aerodynamic coefficients
and derivatives are found by converting the values of cY ' m x , m z ,
and their derivatives by the following formulas:
. .
Cy , A a, A = a' a
Cy. A = ab0/b A }
.ceo cY' Cy
.
CY
. cy
(9.9.85)
"'z. A "'Zb 0 /b A,
Cy , A = Cy
"'z. A
CY. A =
"'z
Cy
(b 01 bA )2
478 Pf. I. Theory. Aerodynamics of an Airfoil and a Wing
- .
mz,A = mzbolbA, m~ A ~ mzbolbA.' m~~ = m~ (b olbA)2 }
(9.9.86)
m~ZAA = m~Z (b olb A)2, m~ZAA = m:Z (bolb A)
. .
mx, A = mx b0/b A, mx, A = mx (b 0/b)2
Ulx, A Ul x , A
A , mx, Ul X (b /b )3
A = mx GA (9·9' 87)
In the above expressions, it is assumed that
9.10. Properties
of Aerodynamic Derivatives
f----+-----+-------t------j
Fig. 9.tO.2
Change in the quantity kc~,z'AA
determining the derivative
cWz.Afor
y.A
a wing at subsonic
velocities
Fig. 9.tO.3
- 0,2 ~+~~~-+-+--+-+---+-~-l
'Fig. 9.10.5
"Change in the variable kmx"ix1
determining the derivative of
the rolling moment coefficient
with respect to W X1 for a wing
at Moo < 1
-0.8 J----t-~~---+--_+--_1
Fig. 9.10.7
Function km~ A detennining
the derivative m~ A for a wing
in a subsonic flow
Fig. 9.tO.8
-
Valuesofk3m:'~-\1 and k3m~AA2
determining the derivative
- -0.5 f---"'...;.:,""!c-::::==----'---'-----;----j
m 0;z,AA for a wing at subsonic
velocities -1. (l '--_ _-'--_ _ _ _ _-'-_ _.....L_ _---.J
Fig. 9.10.9
Ul
Values of km"zAA determining
the derivative m~zAA for a wing
atM oo <1
31-01715
482 Pt. I. Theory. Aerodynamics of an Airfoil and a Wing
Fig. 9.tO.tO
Variables and
(9.10.3)
-
. -. -.
m~ A = (km~, A)/k, m:AA = [(k 3m:"lt) + ill~ (k3m~,AA2)]/k3 (9.10.4)
Ul UlA Ul Ul
m Z,z,AA = (km z, z.A )/k , m z,z.A A = [(k3m z,z,AiA) + iJ'J2
00
UlA
(k 3m Z,z.A2 )1/k3 (9.10.5)
pressible fluid (ill 00 < 1). Here the stability derivatives at ill 00 < 1
can be calculated sufficiently accurately according to the Prandtl-
Glauert compressibility rule c = CiC (1 - ill~)-1/2, where c and Clc
are aerodynamic variables in a compressible and incompressible
fluid, respectively. The lower the aspect ratio, the smaller is the
action of the compressibility, i.e. the smaller is the influence of
the number ill < 1. This is due to the fact that the disturbances
00
a 'cry
4-
5
/ :--
J
;!-/ ?tU/t{!llXo~rf ~ £
Fig. '.to.tt
z /1/ IJru:Z
Values of a' c~ determining the ,fJL
stability derivative c~ for a ~
wing in ta supersonic flow o 5 10 15
1.OH+----v=-~.q-_+-+---_I
1.0 f-H'--JI.----f----+----f----I
Fig. '.to.tl
Change in the quantity
05\-fhf------f-----f-----I----I
a' eOOzAA
y,
determining the deriv-
ative C:,ZA A for a wing at su-
personic velocities o 5 10 15
1.0
0.5
,-., ",.-
--
-1.0
an d ~\ \ /" y,A2
He Iations f or a
'3 eGA
Cy , At
-1.5
;'
1.0
/
,- ,,
O.B arJcWZA
I ,,;' y,AZ
0.6
0.4-
/1 '..,. .... 1w~2
5 10 ,5 a'ilur
-1).3 f-----\l~H-----+------1-~~___j
I
-0. If ----":-1''<------+----+-----1
Fig. 9.tO.15 I
f-:
ltU!~llXo =0 4- 8
.
F unctIOn , (')'1 I -0.5 f
a I/[x 1\C 1aracteriz-
I
ing the change in the deriva-
tive of the rolling moment -O.~ f------+-~-!---=
coefficient with respect to (J) 1
for a wing at Moo> 1 x a'mflX1
Fig. 9.10.16
Values of a'2(nt").Y1) ,,-- --r---- ____ _
.d 1 and J~+--_+7~==~~~_r--~~~9
1,,-:;::; 10 / '15 a'l1zv
a'2 (m;j'l:1)2 characterizing
'.----r
the -0.0'i f-----+iltw tall Xa=O B- - - 1 - - - - - 1
I, /
10
-1~~~~~----~------~-------~
Fig. 9.10.17
Variable a'm~A determining
the deri vative m~A for a wing
at Moo> 1
3.0
a'~ci"f'......
I, /\ 'lw: 2
2.5
2.0
I \
Fig. 9.10.18
1.5 II- ...... /- ... ~ ?twtanxo:O t /8
supersonic flow
o 10
~ 'h/2
1~
-2 /
-3 '\ Ilw taT/ X o:b I
-4 f\.. 4
-
"
8~
-5
Fig. 9.10.19
w -6 ~ ............ r-
Function a'mz,zAA determining I-
the derivative m~zAA for- a -7
wing at ;}loo > 1 '
Ch. 9. Aerodynamic Characteristics in Unsteady Motion 487
2.5
..--1 a'Jm~H j
Z.O
1.5 ./ ", '7UJ" =Z
1
fig. 9.tO.20 1.0
/ ~w tan Xo=O 4 8
0.5 I?
Variables a
•
3 m WzA,A
Z, 1
and
o 7T i I
a,JmWIA
I,M.
'3 (I)
a m z ,zA2 A
c h aracterizlllg
.
-1.5
~/~
'+
1----""
~-~-~=/7r
8
Au,tWI Xu 0
./'
0 --- i--.,-
iJll !! _-J.-a'!lzv
Hypotheses of Harmonicity
and Stationarity
Tangent-Wedge Method
(9.11.8)
Knowing the quantity (9.11.7), we can find the stability deriva-
Itives:
ixt iXI
c~ = Awlb o ~ j pa dx iz; c~z = A~bo J ) lUz dx d~
o Xo Xo
iii
m:t 1
= - A; ) i P OOX1
Zdx dz (9.11.9)
o Xo
iii
a Abr[ ---
mz = ~ J J pa.X dx dz;
o XO
We locate the origin of the coordinates at the nose of the centre
.chord of a wing (Fig. 9.11.1). We adopt the centre chord b o as the
characteristic dimension for m z and WZ, and the span land half-
span 112, respectively, for m X1 and W X1 • For such a wing, the equa-
tions of the leading and trailing edges in the dimensionless form are
as! follows:
(9.11.10)
Aw tan Xo
4
Ch. 9. Aerodynamic Cf.aracteristics in Unsteady Motion 491
Fig. 9.H.t
Diagram of a wing in the calcu-
lation of stability derivatives
according to the method of
local wedges
Let us write the ratio of the centre chord to the span and the
aspect ratio of a wing:
holl= 2Ih./[Aw (f]w + i)], Aw = 12/8 = 211wl/[(11w+ 1) b ol (9.11.11)
The tangent of the sweep angle along the trailing edge is
tan Xl = tan XO - (2b oll) (1 - t/1h·) (9.11.12)
(9.11.14)
(9.11.15)
cay -= 41'
a, cyo)z = 2/'
a (9.11.16)
14. Krasnov, N.F., Koshevoi, V.N., Danilov, A.N., and Zakharchenko, V.F.
A erodinamika raket (Rocket Aerodynamics), Vysshaya shkola, Moscow
(1968).
15. Rakhmatulin, Kh.A., Sagomonyan, A.Ya., Bunimovich, A.I., and Zve-
rev, LN. Gazovaya dinamika (Gas Dynamics), Vysshaya shkola, Moscow
(1965).
16. Arzhanikov, N.S. and Mal'tsev, V.N. Aerodinamika (Aerodynamics).
Oborongiz, Moscow (1956).
17. Krasilshchikova, E.A. Krylo konechnogo razmakha v szhimaemom potoke
(A Finite-Span Wing in a Compressible Flow), Gostekhizdat, Moscow
(1962).
18. Lebedev, A.A. and Chernobrovkin, L.S. Dinamika poleta (Flight Dynamics).
Mashinostroenie, Moscow (1973).
19. Belotserkovsky, S.,\1., Skripach, B.K., and Tabachnikov, V.G., Krylo
v nestatsionarnom potoke gaza (A Wing in an Unsteady Gas Flow), Nauka,
~Ioscow (1971).
Supplementary Reading
Courant, R. and Friedrichs, K.O. Supersonic Flow and Shock Waves. New York,
Interscience (1948).
Ginzburg, I.P. Aerodinamika (Aerodynamics). Moscow, Vysshaya shkola (1966).
Hayes, W.D. and Probstein, R.F. Hypersonic Flow Theory. New York, Academic
Press (1959).
Krasnov, N.F. (Ed.). Prikladnaya aerodinamika (Applied Aerodynamics).
Moscow, Vysshaya shkola (1974).
Kuethe, S. Foundations of Aerodynamics, 3rd ed. New York, Wiley (1976).
Landau, L. D. and Lifshits, E .}J. Jfekhanika sploshnykh sred (Continuum Me-
chanics). Moscow, Gostekhizdat (1968).
Liepmann, H.W. and Puckett, A.E. Introduction to Aerodynamics of a Com-
pressible Fluid. New York, Wiley; London, Chapman & Hall (1948).
Liepmann, H.W. and Roshko, A. Elements of Gasdynamics. New York, Wiley
(1957).
:\Iilne-Thomson, L.l\I. Theoretical A erodynamics. London, MacMillan (1958).
:\Iironer, A. Engineering Fluid Mechanics. New York, McGraw-Hill (1979).
Oswatitsch, K. Gas Dynamics (English version by G. Kuerti). New York,
Academic Press (1956).
Schlichting, H. Boundary Layer Theory, 6th ed. New York, McGraw-Hill (1968).
Sedov, L. Tu:o-Dimensional Problems in Hydrodynamics and Aerodynamics.
New York, Wiley (1965).
Name Index
Mal' tsev, V. N., 238, 255, 256, 275, Rozhdestvensky, LB., 61, 62, 63,
276, 494 67, 165, 188, 493
Melnikov, A, p" 19
Meyer, T" 279 Sadekova, G. S., 111, 493
Milne-Thomson, L. M" 494 Sagomonyan, A. Ya, 209, 494
Mironer, A., 494
Mkhitaryan, A, M., 165, 255, 256, Samuilov, E. V., 61, 62, 66, 67, 165,
493 188, 493
Schlichting, H., 494
Sedov, L. L, 17, 20, 21, 22, 493, 494
Navier, A., 19 Shumyatsky, B. Ya., 62, 63, 67, 188,
Nekrasov, A. L, 22 493
Newton, L, 32 Skripach, B. K., 23, 398, 419, 427,
431, 433, 436, 451, 452, 455, 460,
488, 493, 494
Oswatitsch, K., 494 Stakhanov, 1. P., 61, 62, 66, 67, 165,
188, 493
Stepanov, E. A., 153, 157,493
Pavlukhin, B. N., 153, 157, 493 Stokes, G. G., 19
iPetrov, G, L, 19 Struminsky, V. V., 19
Pleshanov, A. S., 61, 62, 66, 67, 165, Stupochenko, E. V., 61, 62, 66, 165,
188, 493 188, 493
Porodenko, V. V., 153, 157, 493
Prandtl, L" 19, 21, 32, 279
Predvoditelev, A, S., 61, 62, 66, Tabachnikov, V. G., 398, 419, 427,
67, 165, 188, 493 431, 451, 452, 455, 460, 488, 494
Probstein, R. F" 494 Taylor, G. L, 32
Puckett, A, E., 494
Voskresensky, G. P., 22
Rakhmatulin, Kh. A., 209, 494
Reynolds, 0., 19 Zakharchenko, V. F., 193, 494
Roshko, A., 494 Zhukovsky, N. E., 15, 21, 250
Roze, N. V., 82, 111, 493 Zverev, I. N., 209, 494
Subject Index
Ablation, 15 Aerodynamics,
Acceleration. unsteady. 20, 436
resultant, '72f wing. 21£
yector, 72, 113 Aerohydrodynamics, 14
Adiabat, shock, 172, 184 Aerothermodynamics, 18
Aerodynamic coefficients, Air, see also Flo\\'(s), Gas
airfoil, 294 diatomic model, 67, 69
conversion from one aspect ratio dissociating, equation of state, 65
to another, 258ff dissociation and ionization, 192f
dynamic components, 398 motion at high speeds, 14
and similarity method, 139ff structure in dissociation, 67
static components, 398 Aircraft, 20, see also Craft
unsteady flow, 416[£ Airfoil,
Aerodynamics, aerodynamic forces, 293£
ablating surfaces, 23 arbitrary configuration, 285fi
blunt-nosed bodies, 23 in compressible flow, 26!df
bodies of revolution, 22 drag, 278
boundary layer, 19 fictitious, 270£
classification, 17 in hypersoniC flow, 29if, 297
continuum, 1!lf longitudinal force, 293
controls, 22 in mid-span section, 337
definition, 13 polar, 296
development, 13 sharp-nosed, 285f
experimental, 15 curved, 290f
force, 17f sideslipping wing, 299ff
high-speed, 1M, 18, 59 in supersonic flow, 285f£
hulls, 22 symmetric. 297ff, 3t7ff, 321ft
hypersonic, 18 thin,
ideal fluid, 18 in incompressible flow, 234f£
incompressible fluid, 18 in nearly uniform flow, 293
interference, 23 in subsonic flow, 264ff,
low-speed, 14 total drag, 275
optimal shapes, 24 wedge-shaped, 297£
radiating gas, 131 Angle(s),
rarefied gases, 20 attack, 38f
steady-state, 20 balance, 52
subsonic, 18 in normal section, 301£
supersonic, 18, 59 optimal, 47
transonic. 18 true. 252
32-01715
498 Subject Index
Angle(s), Centering,
balance rudder, 58 neutral, 56
banking, 39 Centre,
control surface, 407f aerodynamic, 50f, 55
course, 39 angle-of-attack, 51
downwash, 251£ coordinate, 240
span-averaged, 252f, 255, 257 elevator deflection, 51
total, 253 sideslip angle, 58
elevator, 407 force reduction, 404
flow deflection, 220f moments, 36f
behind shock, 166, 176f, 189 pressure, 48
critical, 176f, 182 and aerodynamic centre, 51
hypersonic velocities, 221 conical flow, 363
and shock angle, 182f Centre-of-pressure coefficient, 49,
flow deviation, see Angle, flow de- 295, 297, 389, 483
flection compressible flow, 274
Mach, 161, 212 nearly uniform flow, 240
pitching, 39f pentagonal \ving, 371
rolling, 39f sideslipping wing, 302
rudder, 407 symmetric airfoil, 297ff
setting, 252, 260f tetragonal wing, 365
shock, 167f wedge-shaped airfoil, 298
oblique, 176ff Characteristics, 208ff
and velocity, 189 aerodynamic,
sideslip, 38f, 52 rectangular wing, 385ff
sweep, 299f, 363 triangular wing, 349
ultimate flow, 221£ unsteady motion, 394ff
yawing, 39f conjugate, 210, 212
zero lift, 240 determination, 209f
Approach, in hodograph, 214, 282
Eulerian, 72 first family, 210ff, 21G, 219, 222f
Lagrangian, 71£ kind, 209f
Axis, nodal point, 227
doublet, 102 orthogonality, 213f
flight path, 37 in physical plane, 210f
lateral, 37 second family, 210ff, 216, 219, 222f
lateral body, 38 stability, 413ff
lift, 37 Chord,
longi tudinal body, 38 local, wing, 259
normal, 38 mean aerodynamic, 45, 262
mean geometric, 45
wing, 45
in wing section with optimal plan-
Backpressure, 155f form, 258
Balancing, longitudinal, 56f Circulation,
Boundary conditions, 132ff, 424 flow, 103f, 243, 430
dimensionless equations, 145 intensity, of vortex, 236f
linearized flow, 310ff velocity, 91ff, 236, 311, 428f, 434
unsteady flow, 451 in vortex-free flow 92
on vortex sheet, 312," 424 vortex, 439, 447
and wall temperature, 145 Coefficient(s) ,
Boundary layer, 34ff aerodynamic, see Aerodynamic co-
thickness, 36 efficients
turbulent core, 36 centre-of-pressure, see Centre-of-
pressure coefficient
correlation, 30
centerinf' 55 one-point, 31
centra, 55 two-point, 30
Subject Index 499
Coefficient(s) , Cone,
damping, 4141" l\lach, 160, 315, 31\)
longitudinal. 4G:l inverted, 31\)
rolling, 403 tip, 371
yawing, 403 Contour,
drag, see Drag coefficient multiply connected, 94
Lame's, 82f, 85, 87 simply connected, \)4
lateral-force, 44, 400 Controllability, craft, 395
lateral linear deformation, 109 Controls,
lateral motion, 406 fast-response, 407
lift, see Lift coefficient inertiafree, 407
longitudinal-force, 44, 296, 399 Coordina te systems,
longitudinal moment, 297 body axis, 37
compressible flow, 303 curvilinear, 82ff
longitudinal motion, 406 cylindrical, 82f
moment, see Moment coefficient fixed, 37
normal-force, 4/1, 51, 56f, 296, 397, flight path, 37
448f, 456 local geographic, 39
compressible flow, 303 normal earth-fixed, 39
and stability derivatives, 400, spherical, 82£
405f wind, 38
pitching-moment, 43, 49, 51, 363, Core(s),
401, 404f, 416ff, 448f, 456 turbulent, 36
body axis, 44 vortex, 256
pressure, see Pressure coefficient semi-infinite, 253
pressure-drop, 284 Correction, Prandtl-Glauert, 268
rolling-moment. 43. 401, 416ff, Cosines, direction, 40
448f Craft,
body axis, 44 banking, 410
spiral, 403 controllability, 395
side-force, 43 fast-response, 407
stagnation pressure, behind shock, inertiafree, 407
175 manoeuvrability, 395
static lateral stability, 57 thermal protection, 23
static longi tudinal stability, 54£ Criterion(a), similarity, 17, HOff, 187
suction force, 306f, 383£ hypersonic, 284f
vortex drag, 315 Curl, velocity, 78
wave drag, 277f, 283, 350, 387 Curves,
induced, 315 characteristic, see Characteristics
wing drag, 315 equipotential, 90
yawing-moment, 43, 401
body axis, 44
spiral, 403 Decrement, logarithmic, 414
Compressibility, gas, 14, 16, 58f Deformation, see also Strain
influence on non-stationary flow, relative linear, 108f
452ff relative volume, 109
Conditions, wing surface, 451
boundary, see Boundary conditions Degree(s),
Chaplygin-Zhukovsky, 428, 431 dissociation, 60
compatability, 209 equilibrium, 69
at infinity, 425 and pressure, 60
initial, 132 and temperature, 60
Cauchy's, 206 freedom, inert, 194
Conductivity, thermal, 63f ionization, 60f
and pressure, 63 static lateral stability, 57
and temperature, 63f static longitudinal stability, 54£
Cone, Oensity(ies),
disturbance, 160, 315, 319 characteristic, for dissociation, 69
32*
500 Subject Index
Eq uation(s), Equation(s),
continuity, steady now, 13G
steady flow, 81, 85, 88f stream function, differential, 205
two-dimensional now, 81, 88 two-dimensional now, near curved
unsteady now, 81 surface, 120f
coupling, 254, 257 velocity potential, 20lf, 264, 309f,
diffusion, 121 372ff
in Cartesian coordinate system, linearization, 264f
123 vortex, 203f
in cylindrical coordinate system, vortex lines, ~11
122f wave, 423ff
dimensionless, 1421' Equilibrium,
and boundary conditions, 145 stable, 52
energy, 124ff, 175 static, 52£
conserva tion, 165 trim, of craft, 52
in dimensionless variables, 143 unstable, 52
t\yo-dimensional plane motion, Exponent, adiabatic, 26, 6ti
127
Euler, 113, 136
now rate, 89 Factor, local friction, 42
gas dynamics, Fall, free, 409f
fundamental, 129, 201, 309 Fan, Prandtl-l\leyer (expansion), 280,
system, 129fl, 144 285f
Gromeka's, 134, 36 Field,
hodograph, 180 pressure, 326, 330
Hug-oniot, 172f veloci ty, 71
in tegro-d i fleren tial, 254 Filament,
kinpmatics, fundamrntal, 201 stream, 74
Lagrange, '135 vortex, 429
Laplace, 8ir, 266 strrngth, 433f
mass now rate, 156ff Fineness, 284, see also Ratio, lift-
momentum conservation, '164 to-dra(:
motion, Flow(s), see also l\Iotion(s)
axisymmetric form, 1'17 at angle of attack, 243ff, 351ff
curvilinear coordinates, '114f axisym metric. 117, 200\'. 222
cylindrical coordinates, 1'16f basic kinds, 1,
in dimensionless variables, 142 boundary coneli tion, 451
disturbed, 413 in boundary layer, 17, 19
i deal gas, 134 circulation (circulatory), 103[, 243,
inviscid fluid, 113, 200f 430
linl'arized flow, ;'308 calculations, 4391'1
particle, lOG compressibk 2G4H, 274, ;)O:2f, 306
potential, 81, 85f over circular cylinder, 241
spherical coordinates, 118ff conical, 355fl, 380
;;teady now, 117 cross, 359
t\vo-dimensional, 119, 200f along curved surface, 88
vector form, 113[ disturbed, 420
vi ,",cous nu i d, 11:l from supersonic source, 319[
;'\avier-Stokes, 112f downwash, 457, see also Downwash
oblique shock, 1G3ff, 169f equilibrium, 194f[
basic, 171 expanding radial, 231
pathline, 71 expansion, 220
potential function, 367ft fictitiolls incompressible, 270f
speed of sound, 264 forward, 301f
state, plane-paralleL 98f
diatomic gas mixture, 69f free, 17
diSSOCiating gas, 65 externaL 35
perfect gas, fi5 streamline, 13:2
502 Subject Index
Flow(s), Flow(s),
over hexagonal wing, 366ff stability derivatives, 483ff
hypersonic, over thin airfoil, 291£ over tetragonal wing, 344
incompressible, 90, 240ff, 249ff, over thin plate, 269ff
360f over thin wing, 315f, 385
cross, 359 unsteady, 425, 4.56ff
over flat plate, 243ff over wings, 315f
velocity potential, 105 over symmetric airfoil wing, 312
inviscid, 17, 35, 200f tetragonal, 331ff, 343ff
isentropic, 138, 149ff, 222, 290f, over tetragonal wing, 331ff, 343ff
420£ over thin plate, 243
isotropic, 31 three-dimensional,
laminar, 28 disturbed, 308
lateral, 299 steady, 309
linearized, 264£, 297, 308ff, 360 transverse, 359
longitudinal, 299 over thin plate, 240ff
nearly uniform, 235, 282, 293 over triangular wing, 355ff
at angle of attack, 351 ff two-dimensional, 89f
linearized, 264 axisymmetric, 200f
pressure, 235 near curved surface, 120f
velocity, 235 isentropic, 222
non-circulatory, 242f, 431 non-isentropic, 222
non-equilibrium, 193ff plane, 200f, 211, 264
behind shock, 197f spatial, 119, 211
non-isentropic, 219, 222 supersonic, 211
one-dimensional, 158 vortex, 202
parallel, 98, 249f vortex-free, 219
stream function, 98 turbulent, 28ff
streamlines, 98 quasi-steady, 31
velocity potential, 98 unsteady, 73, 146, 394, 416ff, 425,
plane, 123, 222 456ff
over plate, 241ff deformable wing, 451
potential, 79, 104, 201£, 356 nearly uniform, 421
Prandtl-ilfeyer, 279, 285 velocity, in jet, 150f
at hypersonic velocities, 281£ viscous; 17, 19
limiting case, 286 in b6undary layer, 134
purely subsonic, 275 pressure, 110
from reservoir, 154ff vortex, 78, 202, 309
reverse, 391£ steady, 89
without separation, 132f, 145 vortex-free, 79ff, 92, 219, 309
two-dimensional, plane, 133 Fluid,
velocity ratio, 133 barotropic, 420, 423
behind shock, 174f ideal,
steady, 73, 117, 122£, :WOf integrals of motion, 134ff
non-potential, 136f pressure, 26ff
potential, 136£ incompressible, 158
subcritical, 176 inviscid, see Fluid, ideal
subsonic, 302f viscous, 127
stability derivatives, 478ff Fluid mechanics, 14, 18
over thin airfoil, 264f Force(s),
supercritical, 177, 274ff aerodynamic, airfoil, 293f
supersonic, 211, 282, 308ff, 353ff body, 27
over airfoil, 285ff complex, 246f
disturbed, 222ff conversion to another coordinate
over finite-span wing, 308ff, 385 system, 40f
Over rectangular wing, 385 dissipative, 127
over sharp-nosed airfoil, 285f, drag, 38, 42, see also Drag
290f gyroscopic, 403£
Subject Index 503
--------------------------------------~
Force(s), Gas,
lateral, 38 equations, 12\)0, 1!14
lift, 38, 42, 250f, 254 flow from reservoir, 154fl
fiat plate, 248, 393 heating, 59ff
maximum, 47 ideal, 134, 149, see also Gas, per-
nearly uniform flow, 235 fect
rectangular wing, 385 interaction with body, 15f, 18
total, 315 ionization, 189f
triangular wing, 362 jet, configuration, 149f
longitudinal, 38, 44, 293 kinetic coefficients, 63f
mass, 27 mixture, 57
external, work, 124 mean molar mass, 70
on moviug body, 25fi parameters, at stagnation point, 191
normal, 38, 416£ perfect, see also Gas, ideal
produced by pressure, 293 calorically, 55
ponderomotive, 27 equation of state, 65
side, 38, 42 thermally, 65
suction, 249, 382ff recombination, 60
correction factor, 383£ stream, configuration, 149£
sideslipping wing, 305ff viscous, flow in boundary layer,
triangular wing, 362 134
surface, 25f, 107f Gradient,
work, 124£ entropy, 216
viscous, 146 normal velocity, 32
Formula, see also Equation(s) potential function, 80
Biot-Savart, 95 pressure, 107
conversion, coordinate systems, velocity, 34
82fi
Euler, 97
Karman-Tsien, 268
Prandtl-Glauert, 258, 302, 306 Half-wing, inuni te triangular, 321
Reynolds generalized, 33f Heat, specifIc, 61£
Sutherland's, 63 Heating, aerodynamic, 15, 18, 23
Zhukovsky, 248 Hodograph, 179ff
Zhukovsky-Chaplygin, 247 Hypothesis,
Frequency, oscillations, 414 absence of reverse influence, 17
Friction, see also Viscosity continuum, 16, 20
in turbulent flow, 32ff harmonicity, 488£
Function, Khristianovich's, 273
conformal, 240f Newton's, 32
dissipative, 127 plane sections, 253f
doublet distribution, 354, 358, 360f Prandtl's, 32£
potential, 79, 243, 354£, 357ff,457, steadiness, 17, 489
see also Velocity potential expanded, 489
derivatives, 458 stress-strain proportionality, 108
doublet, 354£ Zhukovsky-Chaplygin, 244
gradient, 80
stream, 89£, 98, 202
Instability,
dynamic, 394
Gas, see) also Air, Flow(s), Fluid static, 53, 415
compressibility, 14, 15, 58f, 452ff directional, 57f
conduction of electricity, 19 lateral, 57
diatomic, mean molar mass, 70 longitudinal, 54£
diffusion, 121 Integral(s) ,
dissociation, 60, t89ff Bernoulli, 138
dynamics, 14f, see also Aerodynam- Couchy-Lagrange, 420ff
ics, high-speed linearized expression, 421
504 Subject Index
Integral(s) Line(s),
Lagrange, 135 disturbance, 160
motion, 134ff Mach, 160, 211
Intensity, waximum thickness, 348f
source distribution, 318 vortex, 90f
turbulence, 29 weak disturbances, 161, 182, 211
vortex circulation, 236f Loop, perfect, 409f
Interaction,
body-plasma, 19
chemical, 15, 18 Magnetogasdynamics, 19
force, 15 Manoeuvrability, craft, 395
mechanical, 15 Margin, static stability, 55
terms, 402 Method,
thermal, 15, 18 Burago's, 275ff
Interference, aerodynamic, 20f characteristics, 22, 200ff
Ionization, 60, 189f calculation of supersonic flow
thermal, 60 parameters, 285ff
Isobars, 66f and wind tunnel nozzle shaping,
Isochors, 66f 230ff
Isotherms, 66f conformal transformations, 240ff
doublet distribution, 353ff
Glauert-Trefftz, 255
Khristianovich, 269ff
Law, mapping, 240
Dalton's, 69 reverse-flow, 391ff
elliptic, circulation distribution, similari ty, 139
258f small perturbations, 412
energy conservation, 124, 165 sources, 317ff, 425ft
Fourier, 125 field of application, 351ff
mass conservation, 80 tangent wedges, 292, 489ff
momentum conservation, 164 stability derivatives, 490f
Newton's friction, 32 Model, vortex,
thermodynamics, second, 181 intricate lifting surface, 430
Layer, boundary, see Boundary layer non-circulatory flow, 431
Leading edge(s), 367 Modulus,
sonic, 316, 363, 365, 384 longitudinal elasticity, 108
subsonic, 303ff, 316, 326f, 331ff, shear, 109
351ff, 355ff, 366ff, 381ff Moment(s),
supersonic, 303f, 316, 330f, 35if, con version to another coordinate
372ff, 385 system, 40f
sweep parameter, 316f destabilizing, 53f
Length, doublet, 102
mixing, 34 gyroscopic, 403f
relaxation, 198 pitching, 38, 400f, 404f, 416f
Level, turbulence, 29 positive, 38
Lift, 38, 42, see also Force, lift rolling, 38, 401£, 416£
Lift coefficient, 43, 253ff, 296, 315 stabilizing, 52
compressible flow, 274 tilting, 53
and drag coefficient, 296 yawing, 38, 401
nearly uniform flow, 236, 238 Moment coefiicient, 284, 295, 388
pentagonal wing, 371 see also Coefficient(s): longitudinal
sideslipping wing, 302, 304 moment, pitching-moment, rolling-
symmetric airfoil, 298f moment, yawing-moment
tetragonal wing, 364£ com pressible flow, 274
total, 387 nearly uniform flow, 236, 239
triangular wing, 362ff pentagonal wing, 371
wedge-shaped airfoil, 298f sideslipping wing, 302, 304
wing tips, 387 symmetric airfoil, 298£
Subject Index 50S
Shock(s) Source,
adiabat, 172, 184 two-dimensional, 99
angle, 159 region of influellce, 319f
attached, 183f, 187 strength, 100. 427
curved, 159f varying, 425f
straight, 159f vortex, t \\'o-u i mensional, 103
curved. 159f, 178£ Span, characteristic dimension, 262
density ratio, 165, 172£ Specific heats,
detached, and pressure. 6lE
curved, 159£ and temperature, 61f
now over sharp-nosed cone, 177 Speed, see also Velocity
expansion, 181, 211, L81 fluctuation, 29
at hypersonic velocities, 18611 sound, see Sound, speed
lambda -slla ped, 275 Stability,
local, 275 dynamic. 3\)4f, 410ff
normal. 159, 168f, 184 ff free motion, !t12
in dissociated and ionized gas, motion, 411
189ff neutral. !112
oblique. 159, 163fl, 171 longitudinal, oscillatory, 415
in dissociated and ionized gas, oscilla tory. 111 Ii
189 static, 52ff, li15
formation, 16if axisymmetric craft, 54
polar, 180, 18:! lateral. 5H. 57f
possibility, 182 directional, 57f
pressure ratio, 166, 172f rolling. 57
strength. 172 longi tudinal, 53[1
temperature behind, 170, 172 and elevator deflection, 56
margin, 53
temperature ratio, 166, 172f Stability derivativps, 397f, 418£
thickness, 163 accelpra tion. 399
and ~lach number, 163 and aerodynamic coefficients, 399
velocity behind, 166, 171 and control surfaces, 402f
Similari ty, conversion, 404£f
aerodynamic, 138ff dynamic, 3\18
dynamic, 139, 144f first order, 398
full-scale and model flows, 146£ groups, 399
complete, 147 gyroscopic, 1104
partial. 146 Magnus, 403
geometric, 139 potential fllnction, 458
Sink, rotary, 399
point, second ordl'r, 3G8
three-dimensional, 100 static. 398, 402f
two-dimensional, 100 in subsonic flow, 478ff
strength. 100 in supersonic flow, 483ff
Sound, speed, 67f, 142f, 151f, 155, wing, 449
26* Strain. see also Deformation
critical. 15H, 155 angular 76, 78
local, 152 linear, 76, 78
and pressure, 67£ specifIc volume, rate, 78
stahnation conditions, 151 Stream.
and temperature, 67£ filament, 74
in undisturbed flow, 142f free, 25
Source, function, 89[, 98, 202
distribution density, 318£ Streamlines, 73f, 98, lOU
disturbances, 159 family, 74, 10'1
elementary, potential, 318£1 zero, 10M
point, Strength,
three-dimensional, 100 bound vortex, 435
508 Subject Index
Strength, Theory,
frpe vortex filaments, 433f shock wave, 159f£
shock, 172 vortex, 428ft
sink, 100 Time,
source, 100, 425ff relaxation, 194, 199
vortex, 430, 433, 435, 447 dissociative, 194f
vortex layer, 430, 433, 447 vibrational, 194f
vortex sheet, 429f Tip(s),
vortex tube, 91 influence on flow over wing, 342f
Stress, rounding, 260
friction, see Stress, shear sonic, 317
normal, 107f£ subsonic, 317, 385
shear, 31ft, 107f supersonic, 317, 364
in laminar flow, 31£ Trailing edge(s),
in turbulent flow, 32f£ sonic, 317, 365
Strophoid, 180f subsonic, 317. 428
Sublayer, supersonic, 317. 364, 366ff, 372ff
laminal', 35 Trajectory, particle, see Pathline
viscous, 35 Triangle, definiteness, 232
Surface(s), Trim, lateral, 58
con trol, 402f Trimming, 55
angles, 407f Tube,
discontinuity, 156, see also Shock(s) stream, 74
lifting, vortex, 91
flat, 428 strength, 91
intricate planform, 429f Tunnels, wind, 15
rectangular, 428f aerodynamic, 29
wave, 211 Turbulence,
wing, deformation, 451 homogeneous, 31
System(s), initial, 29f
coordinate, see Coordinate systems intensity, 29
disturbed sources, 320 isotropic, 31
vortex, 429 scale, 30f
Temperature, Variables,
characteristic, dissociation, 69 dimensionless, 141f, see also Crite-
cri tical, 154 ria, similarity
ratio, 166, 172f, 184, 189£[ Eulerian, 72
behind shock, 170, 172 kinematic, 395f, ~ 441
stagnation, 153, 191£ Vector,
Terms, interaction, 402 acceleration, 72
Theorem, total, 113
equivalence, 209 aerodynamic forces, resultant, 36f,
Helmholtz, 78, 91 252
Kutta-Zhukovsky, 430 moment of aerodynamic forces, re-
Stokes, 93f suI tan t, 36f
Theory, principal, hydrodynamic pressure
aerodynamic, of wing, 21 forces, 246
boundary layer, 19 velocity, 72
finite-span wing, 21 divergence, 78, 86
heat transfer, gas-dynamic, 19 Velocity(ies),
infinite-span wing, 21 actual, 28
I inearized, supersonic flow over additional, 323, 325f, 328ff, 355
finite-span wing, 308ft average, 28
"Joadpd lint''', 253 calculation, 226ff
s('cond-order aerodynamic, 293,296, at characteristic-shock intersec-
:.!\J8 tion, 227ff
Subject Index 509
Velocity, Vortex(ices),
at characteristic-surface intersec- circulation, 43\). 447
tion, 227ff intensity, 236£
circulation, 91£f components, 78
complex. 97, 245, 247 core(s), 256
curl, 78 semi-infinite. 253
dimensionless, 433, 435 strength, 253
divergence, 78, 86 curvilinear, 95
ficti tious, 266 equi valent. 250
fluctuation, 28 filament, distribution, 429
component, 28 free, 250, 429
free-stream, 25 strength. 43;~f
in gas jet, 150f velocity induced by, 432, 434
induced, 251, 440 horseshoe, 256. 438
by bound vortex, 432, 434 oblique, 430[[' 437
compound due to doublets, 362 infinite, 95[
by free vortices, 432, 434 interaction. 96
by vortex sheet, 434 layer, strength. 430, 433, 447
local, 154 line, 95£ -
nearly uniform flow, 235 model (pattern), 428H
potential. see Velocity potential point, 104
ratio, 152, 156, 271 semi-intini te. D6
fictitious, 266 sheet, see Vortex sheet
flow without separation, 133 source, two-dimensional, 103
and Mach number, 152 strength, 430. 433, 435, 447
relative, 152, see also Velocity, ratio system, 429
behind shock, 166, 171 theory, 428ff
supersonic, 275, 303f tube, 91
total. 2!t5 strength. 91
vector, 72 Vortex sheet. 251, 256, 312, 429
vortex now, 89 boundary conrli tions, 424
vortex-induced, ()4ff strength. 429f
Velocity potential. 79, 98, 103, 105, velocity inrluced by, 312, 434
20if, 26M, 372ff, 457 velocity potential, 460
dimensionless, 452 Vorticity. 91. 93
doublet. 103
elementary source, 318f£
hexagonal wing, 366, 368f Warp, geometric, 259
incompressible flow, 266 Wavers),
induced by doublets, 360 Mach, 161
linearized now, 264£ three-dimensional, 211
non-sta tionary sources, 426f shock, 159, see also Shock(s)
on platp, 243 formation, 159ff
rectangular wing, 385f, 389 infmitesimal. 182
supersonic unsteady flow, 425 weak, 161
two-dimensional flow, 264 simple pressure. 161
on vortex sheet, 460 spherical, 427
Viscosity. stationary, 15D
dynaI~ic. 32, 63 surface, 211
and pressure, 63 Wavelet. see Line(s), weak disturb-
and temperature, 63f ancf'S
and nuid flow, 28££ Wing-Is).
Volume. relative rate of change, 78 conditional. with toothed edges,
Vortex(ices), 103 464
bound, 250, 311, 429 downwashes. 465
strength, 435 finite-span, 249ff, 254, 258££, 300f,
linpar, 430 308H
velocity induced by, 432, 434 in incompressible flow, 249ff
510 Subject Index
Wing(s), Wing(s),
finito-fpan 301 taper ratio, 263
in supf'rsonic flow, 308ff tetragonal, 364£
hexagonal, 364, 366ff, 372ff with dovetail, 364f
with dovetail, 370 in supersonic flow, 344
infinite-span, symmetric airfoil, 331ff
in plane parallel flow, 249f with vee-shaped appendage, 364f
swept, 299 thin, 315f
lift force, 21, 362, 386 at angle of attack, 351ff
lifting capacity, in nearly uniform flow, 351ff
and aspect ratio, 482f, 487 symmetric airfoil, 317ff
and taper ratio, 483 trapezoidal, 260, 262
in linearized flow, 310ff aerodynamic characteristics, 349
non-elli ptical planform, 260 lift force, 362f
optimal planform, 258ff pressure field, 326, 330
pentagonal, 364, 370 semi-infinite, 328ff
rectangular, 260, 385f[ suction force, 362
aerodynamic characteristics, symmetric about x-axis, 326f,
385ff 330f
in linearized flow, 386 vortex model, 436ff
in supersonic flow, 386 zero-thickness, 312f
semi-infinite, 328ff Work,
sideslipping, 299f, 302, 304ff external mass forces, 124
in subsonic compressible 110w, 302f surface forces, 124f
in supersonic flow, 308ff, 315, 344
surface deformation, 451
swept, 300, 341
symmetric airfoil, 312, 317ff
irlealized, 313, 317 Zone, non-equilibrium, 197
TO THE READER
Mir Publishers would be grateful for your
comments on the content, translation and
design of this book.
We would also be pleased to receive any
other suggestions you may wish to make.
Our address is:
Mir Publishers
2 Pervy Rizhsky Pereulok
1-110, GSP, Moscow, 129820, USSR