Flight dynamics-I Prof. E.G.

Tulapurkara
Chapter-3
Dept. of Aerospace Engg., Indian Institute of Technology, Madras
1
Chapter 3
Lecture 7

Drag polar 2

Topics

3.2.3 Summary of lift coefficient, drag coefficient, pitching moment
coefficient, centre of pressure and aerodynamic centre of an airfoil
3.2.4 Examples of pressure coefficient distributions
3.2.5 Introduction to boundary layer theory
3.2.6 Boundary layer over a flat plate height of boundary layer,
displacement thickness and skin friction drag

3.2.3 Summary of the lift coefficient, drag coefficient, pressure coefficient,
pitching moment coefficient, centre of pressure and aerodynamic centre of
an airfoil
In order to understand the dependence of pressure drag and skin friction
drag on various factors, it is appropriate, at this stage, to present brief
discussions on (I) generation of lift, drag and pitching moment from the
distributions of pressure (p) and shear stress (t ) and (II) outline of boundary
layer theory. These and the related topics are covered in this subsection and in
the subsections 3.2.4 to 3.2.10. In subsections 3.2.11 to 3.2.13 the airfoil
characteristics and their nomenclature are dealt with. Subsequently, the
estimation of the drags of wing, fuselage and the entire airplane at subsonic
speeds are discussed(sections 3.2.14 to 3.2.21).
Figure 3.7 shows an airfoil at an angle of attack ( )kept in a stream of
velocity V

. The resultant aerodynamic force (R) is produced due to the

distributions of the shear stress(t ) and the pressure (p). The distributions also
produce a pitching moment (M). By definition, the component of R perpendicular
to the free stream direction is called lift (L) and the component along the free
stream direction is called drag (D). The resulant aerodynamic force (R) can also
Flight dynamics-I Prof. E.G. Tulapurkara
Chapter-3
Dept. of Aerospace Engg., Indian Institute of Technology, Madras
2
be resolved along and perpendicular to the chord of the airfoil. These
components can be denoted by C and N respectively(Fig.3.7). From the
subsequent discussion in this section, it will be evident that it is more convenient
to evaluate N and C from the distributions of shear stress (t ) and pressure (p)
and then evaluate L and D.

Fig.3.7 Aerodynamic forces and moment on an airfoil
From Fig.3.7 it can be deduced that :
L = Ncos - Csin (3.7)
D = Nsin+Ccos (3.8)
Figure 3.8 shows elements of length and ds
u
and ds
l
at points P
u
and P
l

on the upper and lower surfaces of the airfoil respectively. The cartesian
coordinates of points P
u
and P
l

are (x
u
,y
u
) and ( x
l
, y
l
) respectively. Whereas s
u

and s
l
are respectively the distances along the airfoil surface, of the points P
u

and P
l
measured from the stagnation point (Fig.3.8).

Flight dynamics-I Prof. E.G. Tulapurkara
Chapter-3
Dept. of Aerospace Engg., Indian Institute of Technology, Madras
3

Fig.3.8 Pressure and shear stress at typical points on upper
and lower surfaces of an airfoil

To obtain the forces at points P
u
and P
l
,
the local values of p and t are
multiplied by the local area. Since the flow past an airfoil is treated as two-
dimensional, the span of the airfoil can be taken as unity without loss of
generality. Hence, the local area is (ds x 1) and the quantities, L, D, N and C, on
the airfoil, are the forces per unit span. Keeping these in mind, the local
contributions, dN
u
and dC
u
, to N and C respectively, from the element at point P
u

are obtained as:
t
u u u u u u u
dN = -p ds cos - ds sin (3.9)
u u u u u u u
dC = -p ds sin + ds cos t (3.10)
Note that the suffix u denotes quantities at point P
u
and the positive direction of
the angle
u
is as shown in Fig.3.8 .
Expressions similar to Eqs.(3.9) and (3.10) can be written down for the
contributions to N and C from element at point P
l
.
Integrating over the entire airfoil yields :
( ) ( )
u u u u u
upper surface lower surface
N = - p cos + sin ds + p cos - sin ds
l l l l l
t t
} }
(3.11)
( ) ( )
u u u u u
upper surface lower surface
C = -p sin + cos ds + p sin + cos ds
l l l l l
t t
} }
(3.12)
Flight dynamics-I Prof. E.G. Tulapurkara
Chapter-3
Dept. of Aerospace Engg., Indian Institute of Technology, Madras
4

Proceeding in a similar manner, it can be shown that M
le
, the pitching
moment about leading edge of the airfoil, per unit span, is :
( ) ( )
le u u u u u u u u u u u
upper surface
M = p cos + sin x - p sin - cos y ds t t (
}

( ) ( )
lower surface
+ -p cos + sin x + p sin + cos y ds
l l l l l l l l l l l
t t (
}
(3.13)

Note: Once N and C are known, the lift per unit span (L) and drag per unit
span (D) of the airfoil can be obtained using Eqs.(3.7) and (3.8).
It is convenient to work in terms of lift coefficient ( C
l
) and drag coefficient
(C
d
). The definitions of these may be recalled as :

l

2
L
C =
1
V c
2
(3.14)
and

d
2
D
C =
1
V c
2
(3.15)
It may be pointed out, that integration of a constant pressure, say

p ,
around the body would not give any resultant force i.e.
p ds = 0

}
(3.16)
Hence, instead of p the quantity ( )

p-p can be used in Eqs.(3.11), (3.12)

and (3.13). At this stage the following quantities are also defined.
pressure coefficient :
p
2
p-p
C =
1
V
2

(3.17)
skin friction drag coefficient :
t

f
2
c =
1
V
2
(3.18)

Flight dynamics-I Prof. E.G. Tulapurkara
Chapter-3
Dept. of Aerospace Engg., Indian Institute of Technology, Madras
5
:

)
n
2
c
2
le
mle
2 2
N
Normal forcecoefficient C =
1
V c
2
C
Chordwiseor axial forcecoefficient: C = (3.19)
1
V c
2
M
Pitchingmoment coefficient: C =
1
V c
2

It may be noted that dx = ds cos and dy = -ds sin , where ds is an
elemental length around a point P on the surface and is the angle between the
normal to the element and the vertical (Fig.3.8). Note that is measured positive
in the clockwise sense. It can be shown that :

( )
( )
(
(
(

`
(

(

(
)
} } }
} } }


n p pu fu f
0 upper surface lower surface
c
c fu f pu p
0 upper surface lower surface
c
1
C = C - C dx + c dy + c dy
c
(3.20)
1
C = c +c dx + C dy - C dy
c
l l
l l

Following section 10.2 of Ref.1.4, the expressions for C
n
, C
c
and C
mle
can be
rewritten as:

( )
( )
( )
2
1
1
1
(
| |
( |
\ .

(
| |
( |
\ .

(
| |
( |
\ .

| |
+
|
\ .
} }
} }
} }
} }
u
n p pu fu f
0 0
u
c pu p fu f
0 0
u
mle pu p fu f
0 0
u
pu fu u p 2
0 0
c c
c c
c c
c c
dy dy
C = C - C dx+ c +c dx
c dx dx
dy dy 1
C = C - C dx + c - c dx
c dx dx
dy dy
C = C - C xdx- C +c xdx
c dx dx
dy d
+ C +c y dx+ - C
c dx
l
l l
l
l l
l
l l
l

(
| |

( |
\ .
)
f
(3.21)
y
+c y dx
dx
l
l l



Flight dynamics-I Prof. E.G. Tulapurkara
Chapter-3
Dept. of Aerospace Engg., Indian Institute of Technology, Madras
6
Remarks:
(i) From C
n
and C
c
the lift coefficient ( C
l
) and drag coefficient (C
d
) are obtained
as :

n c
C = C cos - C sin
l
(3.22 )

d n c
C = C sin + C cos (3.23)
(ii) Centre of pressure : The point on the airfoil chord through which the
resultant aerodynamic force passes is the centre of pressure. The aerodynamic
moment about this point is zero. It may be noted that the location of centre of
pressure depends on the angle of attack or the lift coefficient.
(iii) Aerodynamic centre: As the location of the centre of pressure depends on
lift coefficient ( C
l
) the pitching moment coefficient about leading edge (C
mle
) also
changes with C
l
. However, it is found that there is a point on the airfoil chord
about which the pitching moment coefficient is independent of the lift coefficient.
This point is called Aerodynamic centre. For incompressible flow this point is
close to the quarter chord point of the airfoil.
(iv) If the distributions of C
p
and c
f
are obtained by analytical or computational
methods, then the pressure drag coefficient (C
dp
) and the skin friction drag
coefficient(C
df
) can be evaluated.
In experimental work the pressure distribution on an airfoil at different angles of
attack can be easily measured. However, measurement of shear stress ( ) t on
an airfoil surface is difficult.The profile drag coefficient (C
d
) of airfoil, which is the
sum of pressure drag coefficient and skin friction drag coefficient, is measured in
experiments by Wake survey technique which is described in Chapter 9, section
f of Ref.3.10. In this technique, the momentum loss due to the presence of the
airfoil is calculated and equated to the drag (refer section 7.5.1 of Ref.3.11 for
derivation).
3.2.4 Examples of pressure coefficient distributions
Though the expression for lift coefficient ( C
l
) involves both the pressure
coefficient (C
p
) and the skin friction drag coefficient (c
f
), the contribution of the
Flight dynamics-I Prof. E.G. Tulapurkara
Chapter-3
Dept. of Aerospace Engg., Indian Institute of Technology, Madras
7
former i.e. C
p
is predominant to decide C
l
. On the other hand, the pressure drag
coefficient (C
dp
) is determined by the distribution of C
p
and the skin friction drag
coefficient (C
df
) is decided by the distribution of shear stress( ) t .
In this subsection the distributions of C
P
in typical cases and their implications for
C
l
and C
dp
are discussed.
The distribution of the pressure coefficient is generally plotted on the outer side
of the surface of the body (Fig.3.9a). The length of the arrow indicates the
magnitude of C
p
. As regards the sign convention, an arrow pointing towards the
surface indicates that C
p
is positive or local pressure is more than the free stream
pressure ( )

p . An arrow pointing away from the surface indicates that C

p
is
negative i.e. the local pressure is lower than ( )

p .

(a) Ideal fluid flow (b) Real fluid flow
Fig.3.9 Distribution of C
p
around a circular cylinder
Figure 3.9 shows distributions of C
p
in ideal fluid flow and real fluid flow past a
circular cylinder. It may be recalled that an ideal fluid is inviscid and
incompressible whereas a real fluid is viscous and compressible. From the
distribution of C
p
in ideal fluid flow (Fig.3.9a) it is seen that the distribution is
symmetric about X-axis and Y-axis. It is evident that in this case, the net forces in
vertical and horizontal directions are zero. This results in C
l

= 0, C
dp
= 0. These
results are available in books on fluid mechanics and aerodynamics. In the real
Flight dynamics-I Prof. E.G. Tulapurkara
Chapter-3
Dept. of Aerospace Engg., Indian Institute of Technology, Madras
8
fluid flow case, shown in Fig.3.9b, it is seen that the flow separates from the body
(see description on boundary layer separation in section 3.2.7) and the pressure
coefficient behind the cylinder is negative and nearly constant. However, the
distribution is still symmetric about horizontal axis. Thus in this case C
l
= 0 but
dp
C > 0.
The distributions of C
p
over symmetrical and unsymmetrical foils at C
l

= 0
and C
l
> 0 are shown in Figs.3.10 a to d. Note also the locations of centre
pressure and the production of pitching moment for the unsymmetrical airfoil.
Flow visualization pictures at three angles of attack( ) are shown in Figs.3.36 a,
b and c. An attached flow is seen at low angle of attack. Some separated flow is
seen at moderate angle of attack and large separated flow region is seen near
close to the stalling angle (
stall
). It may be pointed out that theoretical calculation
of skin friction drag using boundary layer theory can be done, when flow is
attached. This topic is discussed in the next subsection.




(a)Distribution of pressure coefficient on symmetrical airfoil at C
l
= 0 and = 0
Note : L
u
= L
l



Flight dynamics-I Prof. E.G. Tulapurkara
Chapter-3
Dept. of Aerospace Engg., Indian Institute of Technology, Madras
9



(b) Distribution of pressure coefficient on symmetrical airfoil at C
l
> 0 and > 0



(c) Distribution of pressure coefficient on cambered airfoil at C
l
= 0, < 0 ;
Note: L
u
and L
l
form a couple; centre of pressure is at infinity, C
mac
< 0,
Flight dynamics-I Prof. E.G. Tulapurkara
Chapter-3
Dept. of Aerospace Engg., Indian Institute of Technology, Madras
10

(d) Distribution of pressure coefficient on cambered airfoil at C
l
> 0, > 0
Note : C
mac
same as in Fig.(c)

Fig.3.10 Distributions of pressure coefficient on symmetrical and unsymmetric
airfoils at C
l
= 0 and C
l
> 0
3.2.5 Introduction to boundary layer theory
Under conditions of normal temperature and pressure a fluid satisfies the
No slip condition i.e. on the surface of a solid body the relative velocity between
the fluid and the solid wall is zero. Thus, when the body is at rest the velocity of
the fluid layer on the body is zero. In this and the subsequent subsections, the
body is considered to be at rest and the fluid moving past it. Though the velocity
is zero at the surface, a velocity of the order of free stream velocity is reached in
a very thin layer called Boundary layer. The velocity gradient normal to the
surface
| | c
|
c
\ .
U
y
is very high in the boundary layer. Hence even if the coefficient of
viscosity( ) is small, the shear stress,
| | c
|
c
\ .
U

y
,in the boundary layer may be
large or comparable to other stresses like pressure. Outside the boundary layer
Flight dynamics-I Prof. E.G. Tulapurkara
Chapter-3
Dept. of Aerospace Engg., Indian Institute of Technology, Madras
11
the gradient ( ) / c c U y is very small and viscous stress can be ignored and flow
treated as inviscid. It may be recalled from text books on fluid mechanics, that in
an inviscid flow the Bernoullis equation is valid.
Features of the boundary layer over the surface of a streamlined body are shown
in Fig.3.11a. On the surface of a bluff body the boundary layer develops upto a
certain extent and then separates (Fig.3.11b). The definitions of the streamlined
body and bluff body are presented at the end of this subsection.







(a) boundary layer over a streamlined body



(b) Boundary layer over a bluff body
Fig.3.11 Boundary layer over different shapes (not to scale)

Flight dynamics-I Prof. E.G. Tulapurkara
Chapter-3
Dept. of Aerospace Engg., Indian Institute of Technology, Madras
12
The features of the flow are as follows.
1.Near the leading edge (or the nose) of the body the flow is brought to rest.This
point is called the Stagnation point. A laminar boundary layer develops on the
surface starting from that point. It may be recalled, from topics on fluid
mechanics, that in a steady laminar flow the fluid particles move downstream in
smooth and regular trajectories; the streamlines are invariant and the fluid
properties like velocity, pressure and temperature at a point remain the same
with time. In an unsteady laminar flow the fluid properties at a point may vary but
are known functions of time. In a turbulent flow, on the other hand, the fluid
properties at a point are random functions of time. However, the motion is
organized in such a way that statistical averages can be taken. In a laminar
boundary layer the parameter which mainly influence its development is the
Reynolds number ( )
x e
R = U x/ ; x being distance along the surface, from the
stagnation point.
2.Depending on the Reynolds number (R
X
), the pressure gradient and other
parameters, the boundary layer may separate or become turbulent after
undergoing transition. The turbulent boundary layer may continue till the trailing
edge of the body (Fig.3.11a) or may separate from the surface of the body (point
S in Fig 3.11b). It may be added that the static pressure across the boundary
layer at a station x, is nearly constant with y. Hence the pressure gradient
referred here is the gradient (dp/dx) in the flow outside the boundary layer.
3.Nature of boundary layer decides the drag and the heat transfer from the body.
If the boundary layer is separated, the pressure in the rear portion of the body
does not reach the freestream value resulting in a large pressure drag (Fig.3.9b).
Incidently a streamlined body is one in which the major portion of drag is skin
friction drag. For a bluff body the major portion of drag is pressure drag. A
circular cylinder is a bluff body. An airfoil at low angle of attack is a streamlined
shape. But, an airfoil at high angle of attack like
stall
is a bluff body.
Remark:
General discussion on boundary layer is a specialised topic and the
interested reader may consult Ref.3.11 for more information. Here, the features
Flight dynamics-I Prof. E.G. Tulapurkara
Chapter-3
Dept. of Aerospace Engg., Indian Institute of Technology, Madras
13
of the laminar and turbulent boundary layers on a flat plate are briefly described.
While discussing separation, the boundary layer over a curved surface is
considered.

3.2.6 Laminar boundary layer over flat plate height of boundary layer,
displacement thickness and skin friction drag
The equations of motion governing the flow of a viscous fluid are called
Navier-Stokes (N-S) equations. For derivation of these equations refer to
chapter 15 of Ref.3.12. Taking into account the thinness of the boundary layer,
Prandtl simplified the N-S equations in 1904. These equations are called
Boundary layer equations (Chapter 16 of Ref.3.12). Solution of these equations,
for laminar boundary layer over a flat plate with uniform external stream, was
obtained by Blasius in 1908. Subsequently many others obtained the solution.
The numerical solution by Howarth, presented in Ref.3.10, chapter 7, is given in
Table 3.2. In this table U is the local velocity, U
e
is the external velocity (which in
this particular case is V

), and is the non-dimensional distance from the wall

defined as :

v
e
U
= y
x
(3.24)
Flight dynamics-I Prof. E.G. Tulapurkara
Chapter-3
Dept. of Aerospace Engg., Indian Institute of Technology, Madras
14
Table 3.2 Non-dimensional velocity profile in a laminar boundary layer over a
flat plate
Height of boundary layer
It is seen from table 3.2 that the external velocity (U
e
) is attained very
gradually. Hence the height at which U/U
e
equals 0.99 is taken as the height of
the boundary layer and denoted by
0.99
. From table 3.2, ~
e
U/U 0.99 is attained
at = 5 . Noting the definition of in Eq.(3.24) gives :

e
v
0.99
U
5 =
x

Or
0.99 e
x
e x
U x 5 5
= = ; R =
x U x R v
v
(3.25)
Flight dynamics-I Prof. E.G. Tulapurkara
Chapter-3
Dept. of Aerospace Engg., Indian Institute of Technology, Madras
15
Figure 3.12 shows a typical non-dimensional velocity profile in a laminar
boundary layer. While presenting such a profile, it is a common practice to plot
U/U
e
on the abscissa and (
0.99
y/ ) on the ordinate.

Fig.3.12 Non-dimensional velocity profile in laminar and turbulent boundary
layers on a flat plate

It is seen from Eq.(3.25) that
0.99
grows in proportion to
1
2
x (see Fig.3.13). It
may be added that in this special case of laminar boundary layer on flat plate, the
velocity profiles are similar at various stations i.e. the non-dimensional profiles of
U/U
e
vs (y/
0.99
) are same at all stations.

Flight dynamics-I Prof. E.G. Tulapurkara
Chapter-3
Dept. of Aerospace Engg., Indian Institute of Technology, Madras
16

Fig.3.13 Schematic growth of boundary layer
Displacement thickness and skin friction drag coefficient
The presence of boundary layer causes displacement of fluid and skin
friction drag. The displacement thickness ( )
1
is defined as :
1
e 0
U
= 1- dy
U

| |
|
\ .
}
(3.26)
The local skin friction coefficient (
f
C' or c
f
)is defined as :
wall
f f wall wall
2
y=0
e
u
C = c = ; = ;Note: is a function of 'x'.
1
y
U
2
t
t t
| | c
'
|
c
\ .
(3.27)
If the length of the plate is L, then the skin friction drag per unit span of the
plate (D
f
) is :
f wall
0
L
D = dx t
}

Hence, skin friction drag coefficient C
df
is given by:

f
df
2
D
C =
1
V L
2
(3.28)
From the boundary layer profile (table 3.2) it can be shown that for a flat
plate of length, L, the expressions for
1
and C
df
are:
1
L
1.721
=
L R
(3.29)
Flight dynamics-I Prof. E.G. Tulapurkara
Chapter-3
Dept. of Aerospace Engg., Indian Institute of Technology, Madras
17
df L
L
V L 1.328
C = ; R =
R v

(3.30)
Remark :
Reference 1.11, chapter 6 may be consulted for additional boundary
layer parameters like momentum thickness (
2
), shape parameter (H =
1
/
2
)
and energy thickness (
3
) of a boundary layer.
Example 3.1
Consider a flat plate of length 500 mm kept in an air stream of velocity 15 m/s.
Obtain (a) the boundary layer thickness ( )
0.99
and the displacement thickness
( )
1
at the end of the plate (b) the skin drag coefficient. Assume
-6 2
= 1510 m /s v
and the boundary layer to be laminar.
Solution:
L = 0.5 m, V = 15 m/s

,
-6 2
= 1510 m /s v
Hence,
5
L -6
0.515
R = = 510
1510

Consequently, from Eq.(3.25):

-3 0.99
5
L
5 5
= = = 7.0710
L R
510

Or
-3 -3
0.99
= 7.0710 0.5 = 3.5410 m = 3.54mm
From Eq.(3.29):

-3 1
5
L
1.721 1.721
= = = 2.43410
L R
510

Or
-5 -3
1
= 2.43410 0.5 = 1.21710 m = 1.217 mm
From Eq.(3.30):
df
5
L
1.328 1.328
C = = = 0.00188
R
510

Remark:
( )
0.99
/ L is found to be 7.07 x 10
-3
. Hence the assumption of the thinness of
boundary layer is confirmed by the results.