DERIVATION OF THE

SMALL DISTURBANCE POTENTIAL FLOW EQUATION

2.1 Preliminary Remarks
In this chapter, and the next chapter, we develop approximate forms of the
governing equations for subsonic and supersonic flow over thin airfoils, wings
and bodies of revolution. Our ultimate interest is in techniques for predicting the
pressure distribution Cp and the loads (lift Cl, drag Cd and the pitching moment
about the quarter chord cm) over these geometries. For simplicity, we extensively
discuss only 2-D planar flows. We recall that the exact governing equations are:

x x y y 0
(2.1)
and
1

1 2
u2 v 2 1
1
M 1

2

V2
(2.2)
It is possible to eliminate the explicit appearance of density from
equations (3.1) and (3.2), and arrive at the following quasi-linear form of the full
potential equation:

u2 xx 2uv xy a2 v2 yy 0

(2.3)

Equation 2.3 is called quasi-linear because it is linear in its highest

derivatives xx, yy and xy. It is, of course, nonlinear in. It is this nonlinearity
that allows us to model shock waves, a very non-linear phenomenon. From a
mathematical theory called the method of characteristics, one can show that
equation (3.3) is elliptic if
(u2+v2)/a2 < 1
and hyperbolic if

(u2+v2)/a2 > 1
and parabolic if
(u2+v2)/a2 = 1

2.2 Loads over the Airfoil

We define the surface pressure coefficient Cp as,

p
1
1

p p
p
Cp 1


V 2
M2
M2
2
2
2

(2.4)

The airloads over the airfoil may be found once Cp is known as follows:
CY

x c

p, lower

x0

CX

x
C p, upper d
c

dY
Cp,upper

dx
x 0
xc

dY
Cp,lower
d
dx lower
upper

x
c

Cl CY cos CX sin
Cd CX cos CY sin
Cm,c / 4

x c

x0

p, upper

x 1
x
Cp,lower d
c 4 c
(2.5)

Normal
Force, N
Lift

Drag

Chordwise Force, X

where Y(x) is the airfoil shape and is the angle of attack.

We are interested in the solution of the governing equations, in the entire
region between the airfoil and infinity. Because of the nonlinearity of the exact
form of the governing equations, we can not analytically solve the governing
equations except in supersonic flows, using the method of characteristics. Thus,
further approximations to the governing equations are necessary.
2.3 Small Disturbance Assumptions
The assumptions we make are:
(i)
The body is thin, has a small angle of attack, and has only a mild camber.
As a result, the body slope dY/dx, in a coordinate system attached to the
freestream (known as the wind tunnel coordinate system) is small.
(ii)
As a consequence, we assume that the local flow velocity components u
and v are not significantly different from their freestream values.
Disturbance Potential :
We introduce a disturbance potential, , related to the full potential as
follows:

V x +
(2.6)
It must be remembered that in this chapter we are using the wind tunnel
coordinate system, and the freestream velocity is parallel to the x- axis. From
equation (3.6), we obtain the flow velocity components terms of as follows:
u x V x
v y y
(2.7)
From the small disturbance assumptions, it then follows that

x
1
V
y
1
V
(2.8)
3.4 Derivation of the Small Disturbance Equation
Our starting point is the quasi-linear form of the full potential equation 2.3.
From equation 2.6 we note that

xx xx ; yy yy ; xy xy
(2.9)
Next, consider the second term in equation (2.3), -2uvxy . This term may
be viewed as
2uv xy 2V x y xy
2

x y
V 1
0

V V 2 x
2

small
(2.10)
2 2
Next, consider the coefficient a -u , in front of the first term in (2.3).
Noting the fact that the speed of sound is related to the flow speed by the energy
equation:

a2
u2 v 2
a2
V2

1
2
1 2
(2.11)
We can approximate this coefficient as follows:
a 2 u2 a2

1 2
V u 2 v 2 u2

2
2
2

1 2

2
x
x

a
V 1 1
V 1 V
2

a2 V 2 1V 2 x
V

a2 1 M2 1M2 x

V
(2.12)
Examine the above approximation carefully. Note that we have neglected
second powers of the "disturbance velocities" as small. We have, however, kept
2
the first power of the term x/V . This is because the term (1-M ) itself may

be small in transonic flows. Thus, in transonic flows only the last term in the
2
above approximation for (a2-u2) may be comparable in magnitude to (1-M ) ,

and can not be neglected.

In a very similar manner, we can show that
a 2 v 2 a2

(2.13)
With these approximations, the quasi-linear form of the full potential
equation takes on the following simpler form for transonic flows:

1 M

1M2 x xx yy 0
(2.14)

In subsonic and supersonic flows, the small disturbance equation

becomes:

1 M
2

xx

yy 0
(2.15)

Exercise 2.1
Show that the axi-symmetric form of the transonic small disturbance
equation is given by

1 M

1M2 x xx

1
r r r 0
r

2.5 Mathematical Characteristics of the Small Disturbance Equation:

(a)
The transonic small disturbance equation is still nonlinear, in spite of the
approximations that were made to arrive at this equation. In subsonic and
supersonic flows, for thin airfoils, wings and bodies, the governing equation
(2.15) is linear.
(b)
Equation (2.15) may be elliptic, hyperbolic or parabolic. This was given as
a homework problem earlier.
2.6 Small Disturbance Approximation for Surface pressure Coefficient., C p:
We next develop a small disturbance approximation for the surface
pressure. Our starting point is equations (2.2) and (2.4), which may be combined
to yield:

1 2
u2 v 2 1
1
1 2 M 1 V2
Cp
2
M
2

V x y2 1
1 2

1
2
1 2 M 1
V

2
M
2

1 2
1
1
M 2 x
1

2
M
2
In arriving at the above form, we have neglected second powers of terms
such as x/V and y/V .
Next, we use the binomial expansion

1 n 1 n
(2.16)
where is any small quantity << 1 and n is any real number. Then,
1 M 2 1

2 x
Cp
2
M
2
Simplifying, the following small disturbance approximation to surface pressure
coefficient results:
2
Cp x
V
(2.17)
2.7 Boundary Conditions
Before the transonic small disturbance equation (2.15) may be solved, we
need to specify the boundary conditions. Of course, the boundary conditions

must take into account the physics of the problem, and the mathematical
characteristics of the equation.
Boundary Conditions at the Solid Boundary:
y = Y(x)

At nay point on the body surface, the flow

r must be tangential to the body.
In other words, the slope of the velocity vector V must equal the body slope.

y
v
dY

u V x dx
Or,

y V x

dY
dY
V 1 x
dx

V dx

Neglecting the disturbance velocity contribution x/V in comparison to

unity, we get the following boundary condition at the airfoil surface.

y V

dY
dx

(2.18)

Where should this boundary condition be applied? We have two choices.

This boundary condition may be applied either at the actual airfoil surface, or on
a slit along the chord line, located on the x- axis. The latter choice makes the
solution procedure simpler because we can use a Cartesian coordinate system,
rather than a curvilinear coordinate system that is wrapped around the body.

Within the assumptions built into the small disturbance theory, these two
approaches may be shown to be equivalent. Note that this slit is a discontinuity,
across which both the disturbance velocity potential and its y- derivative are
discontinuous.

Boundary conditions
may be applied at
y=Y(x)

Or on a slit
y=0

x
x