Mechanical and Automation Engineering

MAEG3030 Fluid Mechanics

Lecture 10
External Flows Part 1 – Potential Flow Theory

Lecturer -

Dr. Alex Siu Hong Lau

[email protected]
1
MAEG3030 Fluid Mechanics – Lecture Content
➢ Definition External Flows
➢ Potential Flow Theory for Incompressible Inviscid Flows
• Governing Equations
• Flow Building Blocks:
1) Point (or Ideal or Irrotational) Vortex
2) Point Source and Sink
3) Uniform Flow
• Tools for Analysis:
1) Stream Function 𝜓
2) Velocity Potential 𝜙
• 2D Steady Irrotational Flows:
1) Source and Sink Pair
2) Doublet
3) Rankine Oval
4) Circular Cylinder
5) Lifting Circular Cylinder
• Magnus Effect
• Kutta-Joukowski Lift Theorem
• Joukowski Transformation 2
10.1 Introduction
What is an External Flow?
➢ The flow of an unbounded fluid over a surface such as
• a flat plate
• a wing, an aerofoil
• a circular cylinder, a wire, a pipe, a chimney

The wake of a Boeing 767

disrupts the top of a cumulus
cloud and clearly shows the
counter-rotating trailing vortices

3
 10.1 Introduction
Methods to Solve External Flow Problems
➢ Without the constraints such as those provided by internal flows, the non-linear advective
acceleration term of the N-S momentum equations, i.e. (𝛻 ∙ 𝑉)𝑉, cannot be eliminated

➢ Hence, analytical solutions of real (or even for incompressible viscous) external flows do not exist

➢ We rely on physical experiments and computational simulations, from which empirical relations are
derived or the effects of certain variables on some flow phenomena are found

➢ We can also analyse simplified flows, from which the behaviour of the corresponding real flows can
be inferred:

1) First, we will analyse inviscid incompressible flows by using potential flow theory

2) Then, we will learn about boundary layer theory for viscous flows, and find out how some
solutions to the simplified BL equations can be derived

3) After that, Von Karman momentum integral equation, which is valid for laminar boundary
layers (BL) and for time-averaged flow in turbulent BL, will be derived and we will learn how
to find approximate solutions to it

➢ We will close this chapter by learning about drag and lift on an object with relative motion in a fluid
4
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Governing Equations in Potential Flow Theory

➢ In potential flow theory, a steady incompressible inviscid flow is considered, hence,

𝜕𝑉
=0 Steady flow
𝜕𝑡

➢ Also, we have the divergence-free condition for incompressible flow

Continuity equation for
∇⋅𝑉 = 0 incompressible flows

➢ Note that an incompressible inviscid flow is also called an irrotational or a potential flow

➢ For an irrotational flow, vorticity vector 𝜁Ԧ is zero, i.e.

Zero vorticity (fluid rotation) for
𝜁Ԧ = ∇ × 𝑉 = 0 irrotational flows

➢ These three equations are the governing equations in potential flow theory
5
10.2 Potential Flow Theory for Incompressible Inviscid Flows
The Applications of Potential Flow Theory
➢ Potential flow theory is useful for developing a basic understanding of real flows

➢ It is directly applicable to low-Mach-number irrotational flows of homogeneous fluids away from

solid boundaries

𝜌𝑉𝐿
➢ Reynolds number Re = where 𝑉 is a representative velocity magnitude
𝜇 𝐿 is a representative length scale

➢ When Re is large enough (≳ 103 ), the combined influence of fluid viscosity and fluid element
rotation is confined to the thin BL adjacent to solid surfaces

➢ Potential flow theory does not predict viscous effects like skin friction or energy dissipation, so it is
not directly applicable to internal flows in pipes and ducts, to BL flows, or to any rotational flow
region

6
10.2 Potential Flow Theory for Incompressible Inviscid Flows
The Applications of Potential Flow Theory
➢ At a high Re, the BL is thin and provided it stays attached to the whole surface of the body, the
outer flow can be approximately predicted by ignoring the existence of the viscous BL using
potential flow theory

➢ The viscous BL flow is predicted by the viscous flow equations

➢ The two solutions can be adjusted until they match in a suitable region of overlap

➢ Potential flow theory cannot yield satisfactory result when the shape or orientation of the body
is such that BL separates from its surface before reaching the downstream edge, giving rise to
separated flow and a wake, which are rotational

Potential Flow Real Flow

Irrotational
outer region
Viscous
boundary layer

Wake 7
10.2 Potential Flow Theory for Incompressible Inviscid Flows

➢ Examples of flow separation:

Flow over an aerofoil at a high angle of attack

Flow over a circular cylinder

Flow over a step 8

10.2 Potential Flow Theory for Incompressible Inviscid Flows
The Ingredients of Potential Flow Theory
➢ In potential flow theory, the 2D steady irrotational flow desired is constructed from ‘flow
building blocks’ using the idea of superposition

➢ 3 types of flow building block are introduced here:

1) Point vortex

2) Point source and point sink

3) Uniform flow

➢ They are glued together by using either stream function 𝜓 or velocity potential 𝜙

9
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Flow Building Block 1 – a Point Vortex with Circulation 𝚪
➢ In Lecture Notes 4, we have considered point (also called ideal or irrotational) vortex

➢ It can be shown that for a point vortex, all the vorticity is concentrated at the location of the
point (so it is a singularity!)

➢ That is, vorticity is non-zero at the location of the point only, and is zero everywhere else
outside the point

➢ Hence, although the induced flow is moving around the point vortex, it is irrotational

➢ Therefore, any closed curve that includes the point has the same amount of vorticity enclosed,
i.e. the same circulation Γ

➢ So, when we are analysing the flow around a point vortex, we can say the point vortex has
circulation Γ, which is a constant scalar
𝑦
➢ You may think of Γ being the ‘strength’ of a point vortex

𝑢𝜃 𝑟
The flow induced by a point vortex at a 𝜃
radial distance 𝑟 away from the point 𝑥

10
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Flow Building Block 1 – a Point Vortex with Circulation 𝚪
➢ Now, consider the flow induced by a point vortex with constant circulation Γ

➢ 𝑉 ⋅ 𝑇 = 𝑢𝜃 , and 𝑢𝜃 is constant at a fixed radial distance 𝑟 from the point vortex, then

𝑦
Γ = ර 𝑉 ⋅ 𝑇 𝑑𝑠 = ර 𝑢𝜃 𝑑𝑠
𝑐 𝑐

= 𝑢𝜃 ර 𝑑𝑠 𝑢𝜃 𝑟
𝑐 𝜃
𝑥
Γ = 𝑢𝜃 (2𝜋𝑟)

Γ
𝑢𝜃 =
2𝜋𝑟

➢ Circulation Γ of the point vortex in a steady flow is constant, hence 𝑢𝜃 decreases with increasing 𝑟

11
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Flow Building Block 2 – Point Source and Point Sink
➢ A point source (or sink) is like a hole on the 2D plane that emits (or absorbs) incompressible and
inviscid fluid isotropically at a steady mass flux 𝑚ሶ in the radial direction
Source Sink
𝑐 𝑢𝑟 𝑐 𝑢𝑟

𝑟 𝑟
𝜃 𝜃

The steady flow patterns are isotropic (same in all directions), and directed radially
away from the centre

➢ The strength of a point source 𝑚 is defined as

𝑚 = ර 𝑉 ⋅ 𝑛 𝑑𝑠
𝑐
where 𝑛 is the unit vector normal to the closed curve 𝑐 on the same 2D plane 12
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Flow Building Block 2 – Point Source and Point Sink Source
𝑐 𝑢𝑟
➢ Hence, strength 𝑚 is the volume flow rate at a radial
distance 𝑟 from the centre of a point source

➢ The radial velocity 𝑢𝑟 is constant at a radial distance 𝑟

from the point, hence

𝑚 = ර 𝑉 ⋅ 𝑛 𝑑𝑠 = 𝑢𝑟 ර 𝑑𝑠 where 𝑉 ⋅ 𝑛 = 𝑢𝑟
𝑐 𝑐

𝑚 = 𝑢𝑟 (2𝜋𝑟)

𝑚
𝑢𝑟 = for a point source
2𝜋𝑟

Sink
➢ Same analysis for a point sink, but with strength being – 𝑚, then we will get 𝑐
𝑢𝑟

−𝑚
𝑢𝑟 = for a point sink
2𝜋𝑟

➢ Strength 𝑚 of a point source (or −𝑚 for a point sink) is constant, hence

magnitude of 𝑢𝑟 decreases with increasing 𝑟 13
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Flow Building Block 3 – Uniform Flows
➢ Uniform flows on 2D 𝑥𝑦 plane:

𝑢 0 𝑢
𝑉= 𝑉= 𝑉=
0 v v
𝑦

14
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Glue 1 – 2D Stream Function 𝝍
➢ 𝜓 is a scalar function of 𝑥 and 𝑦, i.e. 𝜓(𝑥, 𝑦), such that

𝜕𝜓 𝜕𝜓
≡𝑢 and ≡ −v
𝜕𝑦 𝜕𝑥

➢ 𝜓 satisfies the incompressible (divergence-free) condition automatically:

𝜕𝑢 𝜕v
∇⋅𝑉 = +
𝜕𝑥 𝜕𝑦

𝜕2𝜓 𝜕2𝜓
= −
𝜕𝑥𝜕𝑦 𝜕𝑥𝜕𝑦

∇⋅𝑉 = 0

➢ Hence, 𝜓 is defined for 2D incompressible inviscid (irrotational) flows only

➢ 𝜓 is not defined for a 3D flow field 15

10.2 Potential Flow Theory for Incompressible Inviscid Flows
Glue 1 – 2D Stream Function 𝝍
𝜕v 𝜕𝑢
➢ In an irrotational flow field, 𝜁𝑧 = − =0
𝜕𝑥 𝜕𝑦

𝜕2𝜓 𝜕2𝜓
+ =0
𝜕𝑥 2 𝜕𝑦 2

∇2 𝜓 = 0

➢ ∇2 𝜓 = 0 is a Laplace’s equation

➢ Since Laplace’s equations are linear equations, we can superpose two or more solutions together
and get another solution

16
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Glue 1 – 2D Stream Function 𝝍
➢ Along a curve of 𝜓 = constant, 𝑑𝜓 = 0, hence

𝜕𝜓 𝜕𝜓
𝑑𝜓 = 𝑑𝑥 + 𝑑𝑦
𝜕𝑥 𝜕𝑦

0 = −v𝑑𝑥 + 𝑢𝑑𝑦

𝑢𝑑𝑦 = v𝑑𝑥

𝑑𝑦 v
∴ ቤ =
𝑑𝑥 𝜓=constant 𝑢

➢ Therefore, a curve of constant 𝜓 on the 𝑥𝑦 plane is tangential to the velocity vector of an

infinitesimal fluid parcel moving in the flow field

➢ That is, a curve of constant 𝝍 is a 2D streamline

➢ Hence, 𝜓 is called stream function 𝜓1

𝜓2 𝜓4
𝜓3 17
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Glue 1 – 2D Stream Function 𝝍 and Point Vortex
Γ
➢ For a point vortex, the induced circumferential velocity 𝑢𝜃 =
2𝜋𝑟

𝜕𝜓 No need to memorize expressions

➢ In cylindrical coordinates, 𝑢𝜃 = − 𝜕𝜓
𝜕𝑟 in cylindrical coordinates 𝑢𝜃 = −
𝜕𝑟
𝜕𝜓 Γ 1 𝜕𝜓
➢ Hence, − = 𝑢𝑟 =
𝜕𝑟 2𝜋𝑟 𝑟 𝜕𝜃

Γ
𝜓=− ln(𝑟) Set arbitrary constant to be 0
2𝜋

➢ The flow field induced by a point vortex is irrotational and can be described by streamlines,
each of which associated with a constant stream function 𝜓

𝑦 𝑦
Circulation Γ Circulation −Γ

𝜓3 𝑟3 𝜓3
𝜓2 𝜓1 𝜓2 𝜓1
𝑥 𝑥

18
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Glue 1 – 2D Stream Function 𝝍 and Point Source
𝑚
➢ For a point source, the induced radial velocity 𝑢𝑟 =
2𝜋𝑟

1 𝜕𝜓 No need to memorize expressions

➢ In cylindrical coordinates, 𝑢𝑟 =
𝑟 𝜕𝜃 in cylindrical coordinates

1 𝜕𝜓 𝑚
➢ Hence, =
𝑟 𝜕𝜃 2𝜋𝑟

𝑚
𝜓= 𝜃 Set arbitrary constant to be 0
2𝜋

➢ The flow field induced by a point source is irrotational and can be described by streamlines,
each of which associated with a constant stream function 𝜓
𝜓2 𝜓2
𝜓3 𝜓1 𝜓3 𝜓1
Source of Sink of
strength 𝑚 𝜃1 strength −𝑚
𝜓4 𝜓8 𝜓4 𝜓8

𝜓5 𝜓7 𝜓5 𝜓7
19
𝜓6 𝜓6
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Glue 1 – 2D Stream Function 𝝍 and Uniform Flow
➢ In irrotational flows, we have ∇2 𝜓 = 0

➢ A solution to this Laplace’s equation in 𝜓 is

𝜓 = −v𝑥 + 𝑢𝑦

𝑢
where 𝑉 = and 𝑢 and v are constant, hence this solution is a uniform velocity field
v

➢ A uniform velocity field can be described by streamlines, each of which associated with a
constant stream function 𝜓

𝜓3 𝑢
𝑦 𝑉=
v
𝜓2
𝜓1
𝑥

20
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Glue 2 – Velocity Potential 𝝓 in 2D
➢ 𝜙 is a scalar function of space and time such that 𝜕𝜙
𝜕𝑥
𝜕𝜙 𝜕𝜙 𝑢
𝜕𝜙 𝜕𝜙
≡𝑢 , ≡v and ≡𝑤 , i.e. ∇𝜙 = = v =𝑉
𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜕𝑦 𝑤
𝜕𝜙
Only steady flows are considered, so dependence on 𝑡 is omitted 𝜕𝑧

➢ 𝜙 satisfies the irrotational condition automatically:

𝜕v 𝜕𝑢
𝜁𝑧 = −
𝜕𝑥 𝜕𝑦

𝜕2𝜙 𝜕2𝜙
= −
𝜕𝑥𝜕𝑦 𝜕𝑥𝜕𝑦

𝜁𝑧 = 0

➢ Hence, 𝜙 is defined for incompressible inviscid (irrotational) flows only

𝜕
➢ We only consider 2D irrotational flow here, so 𝑤 = 0 and =0
𝜕𝑧 21
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Glue 2 – Velocity Potential 𝝓 in 2D
𝜕𝑢 𝜕v
➢ In an incompressible flow field, ∇ ⋅ 𝑉 = + =0
𝜕𝑥 𝜕𝑦

𝜕2𝜙 𝜕2𝜙
+ =0
𝜕𝑥 2 𝜕𝑦 2

∇2 𝜙 = 0

➢ ∇2 𝜙 = 0 is a Laplace’s equation

➢ Since Laplace’s equations are linear equations, we can superpose two or more solutions together
and get another solution

22
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Glue 2 – Velocity Potential 𝝓 in 2D
➢ Since the gradients of 𝜙 are the velocity components, 𝜙 is the scalar potential of the velocity
field

➢ Hence, 𝜙 is called velocity potential

𝜕𝜙 𝜕𝜙
➢ Along a curve of 𝜙 = constant, 𝑑𝜙 = 0, hence 𝑑𝜙 = 𝑑𝑥 + 𝑑𝑦
𝜕𝑥 𝜕𝑦

0 = 𝑢𝑑𝑥 + v𝑑𝑦

𝑢𝑑𝑥 = −v𝑑𝑦
𝑑𝑦 𝑢
∴ ቤ =−
𝑑𝑥 𝜙=constant v

𝑑𝑦 v
➢ Recall that a curve of constant 𝜓, i.e. a streamline, has a gradient of ቤ =
𝑑𝑥 𝜓=constant 𝑢

𝑑𝑦 𝑑𝑦
∵ ቤ ቤ = −1
𝑑𝑥 𝜙=constant 𝑑𝑥 𝜓=constant

∴ Curves of constant 𝜙 are perpendicular to streamlines 23

10.2 Potential Flow Theory for Incompressible Inviscid Flows
Glue 2 – Velocity Potential 𝝓 in 2D and Point Vortex
Γ
➢ For a point vortex, the induced circumferential velocity 𝑢𝜃 =
2𝜋𝑟

1 𝜕𝜙 No need to memorize expressions

➢ In cylindrical coordinates, 𝑢𝜃 =
𝑟 𝜕𝜃 in cylindrical coordinates
1 𝜕𝜙
𝑢𝜃 =
1 𝜕𝜙 Γ 𝑟 𝜕𝜃
➢ Hence, =
𝑟 𝜕𝜃 2𝜋𝑟 𝜕𝜙
𝑢𝑟 =
𝜕𝑟
Γ
𝜙= 𝜃 Set arbitrary constant to be 0
2𝜋

𝜙2
𝜙3 𝜙1
➢ The flow field induced by a point vortex is irrotational
and can be described by curves of constant velocity
potential 𝜙
𝜙4 𝜙8

𝜙5 𝜙7
𝜙6
24
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Glue 2 – Velocity Potential 𝝓 in 2D and Point Source
𝑚
➢ For a point source, the induced radial velocity 𝑢𝑟 =
2𝜋𝑟

𝜕𝜙 No need to memorize expressions

➢ In cylindrical coordinates, 𝑢𝑟 =
𝜕𝑟 in cylindrical coordinates

𝜕𝜙 𝑚
➢ Hence, =
𝜕𝑟 2𝜋𝑟

𝑚
𝜙= ln(𝑟) Set arbitrary constant to be 0
2𝜋

𝜙3
➢ The flow field induced by a point source is irrotational
𝜙1 𝜙2
and can be described by curves of constant velocity
potential 𝜙

➢ The analysis for a point sink is identical except for using a

strength of −𝑚

25
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Glue 2 – Velocity Potential 𝝓 in 2D and Uniform Flow
➢ In incompressible flows, we have ∇2 𝜙 = 0

➢ A solution to this Laplace’s equation in 𝜙 is

𝜙 = 𝑢𝑥 + v𝑦

𝑢
where 𝑉 = and 𝑢 and v are constant, hence this solution is a uniform velocity field
v

➢ A uniform velocity field can be described by curves of constant velocity potential 𝜙

26
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Construction of 2D Steady Irrotational Flows by Superposition
➢ Since Laplace’s equations (e.g. ∇2 𝜓 = 0 and ∇2 𝜙 = 0) are linear equations, we can superpose
two or more solutions together and get another solution to describe a more complex potential
flow field

➢ As a summary, the building blocks are:

Point Vortex

𝑦 𝑦
−Γ 𝑢𝜃 Γ
𝑢𝜃 𝜓= ln(𝑟) 𝜓= ln(𝑟)
𝑟 2𝜋 𝑟 2𝜋
𝜃 𝜃
𝑥 𝑥
Γ −Γ
𝜙= 𝜃 𝜙= 𝜃
2𝜋 2𝜋

Circulation Γ Circulation −Γ

Only the streamlines are shown

27
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Construction of 2D Steady Irrotational Flows by Superposition

𝑟 𝑟
Source 𝑢𝑟 𝑚 Sink 𝑢𝑟 −𝑚
𝜓= 𝜃 𝜓= 𝜃
2𝜋 2𝜋
𝜃 𝜃
𝑚 −𝑚
𝜙= ln(𝑟) 𝜙= ln(𝑟)
2𝜋 2𝜋

Strength 𝑚 Strength −𝑚
Only the streamlines are shown

Uniform flow 𝑢 0 𝑢
𝑉= 𝑉= 𝑉=
0 v v
𝑦

𝑥
𝜓 = 𝑢𝑦 𝜓 = −v𝑥 𝜓 = −v𝑥 + 𝑢𝑦

𝜙 = 𝑢𝑥 𝜙 = v𝑦 𝜙 = 𝑢𝑥 + v𝑦
28
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Construction of 2D Steady Irrotational Flows – 1) A Source and Sink Pair
➢ Superpose a point source and a point sink on the 𝑥𝑦 plane
𝑦 𝐴(𝑥, 𝑦)
➢ Consider a general point 𝐴(𝑥, 𝑦) on the 𝑥𝑦 plane:
𝑚
𝜓= 𝜃 − 𝜃2
2𝜋 1 𝑟
2𝜋𝜓 𝜃1 𝜃
= 𝜃1 − 𝜃2 𝜃2
𝑚 𝑥

2𝜋𝜓
tan = tan(𝜃1 − 𝜃2 )
𝑚
𝑎 𝑎
2𝜋𝜓 tan 𝜃1 − tan(𝜃2 )
tan =
𝑚 1 + tan(𝜃1 ) tan(𝜃2 )

𝑦 𝑦
• From the figure, we can see that tan(𝜃1 ) = and tan(𝜃2 ) =
𝑥+𝑎 𝑥−𝑎

𝑟 sin 𝜃 𝑟 sin 𝜃
• Express these in terms of 𝑟 and 𝜃: tan(𝜃1 ) = and tan(𝜃2 ) =
𝑟 cos 𝜃 + 𝑎 𝑟 cos 𝜃 − 𝑎
29
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Construction of 2D Steady Irrotational Flows – 1) A Source and Sink Pair
➢ Then, we have
𝑟 2 sin 𝜃 cos 𝜃 − 𝑎𝑟 sin 𝜃 − 𝑟 2 sin 𝜃 cos 𝜃 − 𝑎𝑟 sin 𝜃
2𝜋𝜓 𝑟 2 cos2 𝜃 − 𝑎2
tan =
𝑚 𝑟 2 sin2 𝜃
1+ 2 𝑦
𝑟 cos2 𝜃 − 𝑎2 𝐴(𝑥, 𝑦)

−2𝑎𝑟 sin 𝜃
2𝜋𝜓 2 2 𝜃 − 𝑎2
tan = 𝑟 cos
𝑚 𝑟 2 − 𝑎2 𝑟
𝑟 2 cos2 𝜃 − 𝑎2
𝜃1 𝜃 𝜃2
𝑥
2𝜋𝜓 −2𝑎𝑟 sin 𝜃
tan = 2
𝑚 𝑟 − 𝑎2

𝑎 𝑎

𝑚 −2𝑎𝑟 sin 𝜃
𝜓= arctan
2𝜋 𝑟 2 − 𝑎2

30
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Construction of 2D Steady Irrotational Flows – 2) A Doublet
➢ When the source and the sink get closer to each other, 𝑎 becomes smaller, hence the angle in
arctan becomes smaller, then by small angle approximation
𝑦
−𝑚𝑎𝑟 sin 𝜃 Streamlines
𝜓= 2 2
𝜋(𝑟 − 𝑎 )
Curves of 𝜙

➢ In the limit of 𝑎 → 0, we get a doublet: 𝑥

𝑟 1
• When 𝑎 → 0, →
𝑟 2 − 𝑎2 𝑟

𝑚𝑎
• When 𝑎 → 0, if 𝑚 is finite, →0
𝜋
𝑚𝑎
• Hence, we set 𝑚 → ∞ as 𝑎 → 0 so that is non-zero and finite
𝜋
𝑚𝑎
• Let 𝐾 = be the strength of the doublet, then
𝜋

−𝐾 sin 𝜃
𝜓=
𝑟
31
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Construction of 2D Steady Irrotational Flows – 2) A Doublet
➢ Differentiate 𝜓 with respect to 𝜃:
𝑦
−𝐾 sin 𝜃 Streamlines
𝜓=
𝑟

𝜕𝜓 −𝐾 cos 𝜃 Curves of 𝜙
=
𝜕𝜃 𝑟
𝑥

➢ The radial velocity 𝑢𝑟 in cylindrical coordinates:

1 𝜕𝜓 −𝐾 cos 𝜃
𝑢𝑟 = =
𝑟 𝜕𝜃 𝑟2

➢ The radial velocity 𝑢𝑟 can also be written in terms of 𝜙, i.e.

𝜕𝜙 −𝐾 cos 𝜃
=
𝜕𝑟 𝑟2

𝐾 cos 𝜃
𝜙=
𝑟
32
 10.2 Potential Flow Theory for Incompressible Inviscid Flows
Construction of 2D Steady Irrotational Flows – 3) Rankine Oval
➢ Superpose a uniform horizontal flow field 𝑉 = (𝑢∞ , 0), a point source and a point sink:

𝑢∞ 𝐴(𝑥, 𝑦)
𝑦
𝑟1 𝑟2
𝑟
𝜃2
𝜃1 𝜃
𝑥

𝑎 𝑎

➢ The stream function for the induced flow field at a general point 𝐴:

𝑚 −2𝑎𝑟 sin 𝜃 𝑚 −2𝑎𝑦

𝜓 = 𝑢∞ 𝑟 sin 𝜃 + arctan 𝜓 = 𝑢∞ 𝑦 + arctan 2 Eq. 1
2𝜋 𝑟 2 − 𝑎2 2𝜋 𝑥 + 𝑦 2 − 𝑎2
Using 𝜓 for a source and sink pair

𝑚 𝑚 𝑚 𝑦 𝑚 𝑦
𝜓 = 𝑢∞ 𝑟 sin 𝜃 + 𝜃1 − 𝜃 𝜓 = 𝑢∞ 𝑦 + arctan − arctan Eq. 2
2𝜋 2𝜋 2 2𝜋 𝑥+𝑎 2𝜋 𝑥−𝑎
33
Consider source and sink separately
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Construction of 2D Steady Irrotational Flows – 3) Rankine Oval
➢ Differentiate Eq. 2, which is a 𝜓 expression in Cartesian coordinates, with respect to 𝑦 to find
the 𝑥-velocity component:

𝜕𝜓 𝑚 1 1 𝑚 1 1 𝑑 1
𝑢= = 𝑢∞ + − arctan 𝑥 =
𝜕𝑦 2𝜋 𝑦2 𝑥+𝑎 2𝜋 𝑦2 𝑥−𝑎 𝑑𝑥 1 + 𝑥2
1+ 2 1+ 2
𝑥+𝑎 𝑥−𝑎

𝑚 𝑥+𝑎 2 1 𝑥−𝑎 2 1
= 𝑢∞ + −
2𝜋 𝑥 + 𝑎 2 + 𝑦2 𝑥+𝑎 𝑥 − 𝑎 2 + 𝑦2 𝑥−𝑎

𝑚 𝑥+𝑎 𝑥−𝑎
𝑢 = 𝑢∞ + −
2𝜋 𝑥 + 𝑎 2 + 𝑦2 𝑥 − 𝑎 2 + 𝑦2

𝑢∞ 𝐴(𝑥, 𝑦)
𝑦

𝑎 𝑎 34
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Construction of 2D Steady Irrotational Flows – 3) Rankine Oval
➢ This potential flow has top and bottom symmetry, as well as left and right symmetry

➢ The axes of symmetry are the 𝑥-axis and the 𝑦-axis

➢ The stagnation points are at 𝑦 = 0 where the 𝑥-velocity component is zero, i.e.

𝑚 𝑥+𝑎 𝑥−𝑎
0 = 𝑢∞ + −
2𝜋 𝑥+𝑎 2 𝑥−𝑎 2
𝑚𝑎 𝑚𝑎
𝑚 1 1 − + 𝑎2 + 𝑎2
0 = 𝑢∞ + − 𝜋𝑢∞ 𝜋𝑢∞
2𝜋 𝑥 + 𝑎 𝑥 − 𝑎

𝑚 −2𝑎
0 = 𝑢∞ +
2𝜋 𝑥 2 − 𝑎2 𝑦
𝑢=0 𝑢=0
𝑥
𝑚𝑎
𝑥=± + 𝑎2
𝜋𝑢∞
𝑎 𝑎
𝑢∞ 𝑢∞

➢ Hence, one stagnation point is to the left of the source, the other is to the right of the sink 35
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Construction of 2D Steady Irrotational Flows – 3) Rankine Oval
➢ The stagnation points are at 𝑦 = 0, put 𝑦 = 0 in either Eq. 1 or Eq. 2, we get

𝜓𝑦=0 = 𝜓stag = 0

➢ We treat the curved streamlines with 𝜓stag = 0 as the surface streamlines , and ignore the
streamlines enclosed by them (shaded in orange below)
• No flow crosses a streamline
➢ The solid body formed is called a Rankine oval • Hence, any streamline of the flow
field can be treated as a solid surface
since there is no flow across it
𝑢∞ 𝑦

𝜓stag = 0 𝜓stag = 0
𝑥

𝑎 𝑎
Note that along the streamlines with 𝜓stag = 0, including the surface streamlines,
velocity is non-zero except at the stagnation points 36
 10.2 Potential Flow Theory for Incompressible Inviscid Flows
Construction of 2D Steady Irrotational Flows – 3) Rankine Oval
➢ The distance between the stagnation point and the origin is known as the half body length 𝒍, i.e.
𝑚𝑎 𝑙 𝑚
𝑙= + 𝑎2 = +1
𝜋𝑢∞ 𝑎 𝜋𝑢∞ 𝑎

➢ The half width 𝒉 can be found by estimating the intersection between the surface streamline
𝜓stag = 0 with the 𝑦-axis (where 𝑥 = 0) at 𝑦 = ℎ:

𝑚 2𝑎𝑦
𝜓 = 𝑢∞ 𝑦 − arctan 2
2𝜋 𝑥 + 𝑦 2 − 𝑎2

𝑚 2𝑎ℎ 𝑦
𝑙 𝑙
0 = 𝑢∞ ℎ − arctan 2
2𝜋 0 + ℎ2 − 𝑎 2 𝑢∞
ℎ
2𝑎ℎ 2𝜋𝑢∞ ℎ
arctan 2 = 𝑥
ℎ − 𝑎2 𝑚
ℎ2 − 𝑎 2 2𝜋𝑢∞ ℎ
ℎ= tan 𝑎 𝑎
2𝑎 𝑚
2
ℎ 1 ℎ 𝜋𝑢∞ 𝑎 ℎ
= − 1 tan 2 which is solved iteratively
𝑎 2 𝑎 𝑚 𝑎 37
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Construction of 2D Steady Irrotational Flows – 3) Rankine Oval

𝑙 𝑚
➢ Half body length to 𝑎 ratio: = +1
𝑎 𝜋𝑢∞ 𝑎

2
ℎ 1 ℎ 𝜋𝑢∞ 𝑎 ℎ
➢ Half width to 𝑎 ratio: = − 1 tan 2
𝑎 2 𝑎 𝑚 𝑎

𝜋𝑢∞ 𝑎
➢ By changing the dimensionless parameter , the length to width ratio of the Rankine oval can
be changed 𝑚

➢ For ovals with small width compared to its 𝑦

𝑙 𝑙
length, the solution obtained could be a
𝑢∞
good approximation of the real flow
ℎ
𝑥

𝑎 𝑎

38
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Construction of 2D Steady Irrotational Flows – 4) Circular Cylinder
➢ Superpose a uniform horizontal flow field 𝑽 = (𝒖∞ , 𝟎) and a doublet:
𝑦
Doublet strength is 𝐾
𝑢∞
𝑟
𝜃
𝑥

➢ The stream function and velocity potential for the induced flow field:

𝐾 sin 𝜃
𝜓 = 𝑢∞ 𝑟 sin 𝜃 −
𝑟

𝐾 cos 𝜃
𝜙 = 𝑢∞ 𝑟 cos 𝜃 +
𝑟
39
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Construction of 2D Steady Irrotational Flows – 4) Circular Cylinder
➢ The circumferential and radial velocity components of the induced velocity field:

1 𝜕𝜙 𝜕𝜓 𝐾 sin 𝜃 𝐾
𝑢𝜃 = =− = −𝑢∞ sin 𝜃 − 𝑢𝜃 = − sin 𝜃 𝑢∞ +
𝑟 𝜕𝜃 𝜕𝑟 𝑟2 𝑟2

𝜕𝜙 1 𝜕𝜓 𝐾 cos 𝜃 𝐾
𝑢𝑟 = = = 𝑢∞ cos 𝜃 − 𝑢𝑟 = cos 𝜃 𝑢∞ −
𝜕𝑟 𝑟 𝜕𝜃 𝑟2 𝑟2
𝑦
𝑏 𝑏
𝑢∞
𝑟
𝜃
𝑥

𝐾
➢ From the expression for 𝑢𝑟 , we can see that at the point where 𝑢𝑟 = 0, 𝑢∞ =
𝑟2
𝐾
𝑟2 =
𝑢∞
𝐾
• Let the radial distance 𝑟 where 𝑢𝑟 = 0 be 𝑏, then 𝑏2 =
𝑢∞ 40
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Construction of 2D Steady Irrotational Flows – 4) Circular Cylinder
➢ Replace the strength of doublet 𝐾 in the expressions for 𝜓, 𝜙, 𝑢𝜃 and 𝑢𝑟 by 𝑏 2 𝑢∞ , we get

𝑏2 𝑏2 where 𝑏 is the radial distance 𝑟

𝜓 = 𝑢∞ sin 𝜃 𝑟 − 𝑢𝜃 = −𝑢∞ sin 𝜃 1 + 2
𝑟 𝑟 of the point at which 𝑢𝑟 = 0
𝑏2 𝑏2 (Sub. 𝑟 = 𝑏 in the expression
𝜙 = 𝑢∞ cos 𝜃 𝑟 + 𝑢𝑟 = 𝑢∞ cos 𝜃 1 − 2 for 𝑢𝑟 and will get 𝑢𝑟 = 0)
𝑟 𝑟

➢ Like that for a Rankine oval, the two stagnation points are at 𝑦 = 0 where 𝜃 = 0 and 𝜃 = 𝜋

➢ Hence, the streamlines connecting the two stagnation points have 𝜓stag = 0
𝑦
➢ At the stagnation point where 𝜃 = 𝜋, 𝑢∞
𝑏2 𝑟
𝑢𝑟 = 0 = −𝑢∞ 1− 2 𝜓stag
𝑟 𝜓stag 𝜃 𝜓stag
𝑥
𝑏2 𝜓stag
0=1− 2
𝑟

𝑟 2 = 𝑏2 (Same can be found by using 𝜃 = 0)

𝑥 2 + 𝑦 2 = 𝑏2 Hence, the curved streamlines with 𝜓stag form a circle 41

10.2 Potential Flow Theory for Incompressible Inviscid Flows
Construction of 2D Steady Irrotational Flows – 4) Circular Cylinder
➢ The curved streamlines with 𝜓stag are treated as the surface streamlines

➢ 𝑢𝑟 = 0 on the surface streamlines

➢ Note that 𝑢𝜃 on the surface streamlines are non-zero except at the stagnation points

➢ To find 𝑢𝜃 on the surface, sub. 𝑟 = 𝑏 into the expression for 𝑢𝜃 and get

𝑢𝜃, 𝑠 = −2𝑢∞ sin 𝜃 𝜃 𝑢𝜃, 𝒔

where subscript 𝑠 denotes surface 0 0
➢ Hence, 𝑢𝜃, 𝑠 = 2𝑢∞ 𝜋 Τ2 −2𝑢∞
max
𝜋 0
𝑦 𝑢𝜃, 𝑠 = −2𝑢∞
3𝜋Τ2 2𝑢∞
𝑢∞ Anticlockwise positive
𝑟
𝜃
𝑥

𝑢𝜃, 𝑠 = 2𝑢∞ 42
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Construction of 2D Steady Irrotational Flows – 4) Circular Cylinder
➢ The surface pressure distribution is calculated by using Bernoulli equation:

1 2 1 2 Steady body forces like gravity in incompressible flow

𝑃∞ + 𝜌𝑢∞ = 𝑃𝑠 + 𝜌𝑢𝜃, 𝑠
2 2 are commonly omitted from the momentum equations

1 2 2
𝑃𝑠 = 𝑃∞ + 𝜌 𝑢∞ − 𝑢𝜃, 𝑠
2
1 2 2
𝑃𝑠 = 𝑃∞ + 𝜌 𝑢∞ − 4𝑢∞ sin2 𝜃 𝑢𝜃, 𝑠 = −2𝑢∞ sin 𝜃
2
1 2
𝑃𝑠 = 𝑃∞ + 𝜌𝑢∞ 1 − 4sin2 𝜃
2

➢ Express the surface pressure in the form of the coefficient of pressure 𝐶𝑝 :

𝑃𝑠 − 𝑃∞
𝐶𝑝 =
1 2
𝜌𝑢
2 ∞

𝐶𝑝 = 1 − 4sin2 𝜃

43
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Construction of 2D Steady Irrotational Flows – 4) Circular Cylinder
➢ The surface 𝐶𝑝 over a circular cylinder predicted by
potential flow theory and that measured in an
experiment are shown on the left

➢ A circular cylinder is a bluff body 𝜃

➢ The flow over a bluff body has a significant portion of
separated flow region
180 150 120 90 60 30 0
➢ The discrepancy between predicted and measured 𝐶𝑝 ,°
𝜃, °
is large where the flow is separated from the surface 𝑦

𝑟
𝑢∞
𝜃
𝑥

The location of the flow separation

points depends on flow and surface
conditions
44
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Construction of 2D Steady Irrotational Flows – 5) A Lifting Circular Cylinder
➢ Superpose a uniform horizontal flow field 𝑽 = (𝒖∞ , 𝟎), a doublet and a point vortex:
𝑦 Doublet strength is 𝑲

𝑢∞ Point vortex has circulation −𝚪

𝑟
𝜃
𝑥

➢ The stream function and velocity potential for the induced flow field:

𝑏2 Γ
𝜓 = 𝑢∞ sin 𝜃 𝑟 − + ln(𝑟)
𝑟 2𝜋

𝑏2 Γ
𝜙 = 𝑢∞ cos 𝜃 𝑟 + − 𝜃
𝑟 2𝜋
45
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Construction of 2D Steady Irrotational Flows – 5) A Lifting Circular Cylinder
➢ The circumferential and radial velocity components of the induced velocity field:

1 𝜕𝜙 𝜕𝜓 𝑏2 Γ
𝑢𝜃 = =− 𝑢𝜃 = −𝑢∞ sin 𝜃 1 + 2 −
𝑟 𝜕𝜃 𝜕𝑟 𝑟 2𝜋𝑟

𝜕𝜙 1 𝜕𝜓 𝑏2
𝑢𝑟 = = 𝑢𝑟 = 𝑢∞ cos 𝜃 1 − 2
𝜕𝑟 𝑟 𝜕𝜃 𝑟

➢ On the surface of the cylinder, 𝑟 = 𝑏, hence

Γ
𝑢𝜃, 𝑠 = −2𝑢∞ sin 𝜃 − and 𝑢𝑟 = 0
2𝜋𝑏

where subscript 𝑠 denotes surface

𝑦

𝑢∞
𝑟
𝜃 𝑥

46
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Construction of 2D Steady Irrotational Flows – 5) A Lifting Circular Cylinder
Γ
➢ At the stagnation points, 𝑢𝜃, 𝑠 = 0, hence 0 = −2𝑢∞ sin 𝜃 −
2𝜋𝑏
−Γ
𝜃 = asin
4𝜋𝑏𝑢∞

−Γ
𝛽 = asin where 𝛽 is 𝜃 of the stagnation points
4𝜋𝑏𝑢∞
−Γ
➢ Γ, 𝑏 and 𝑢∞ are positive, hence is negative and the acute angle 𝛽 is negative
4𝜋𝑏𝑢∞
➢ The rear stagnation point is at the acute angle 𝛽, which is negative and is hence clockwise from the
positive 𝑥-axis as shown below

➢ The front stagnation point is at 𝜋 − 𝛽, which is greater than 𝜋 since 𝛽 is negative

𝑦

𝑢∞
𝜋−𝛽
𝑥
𝛽

47
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Construction of 2D Steady Irrotational Flows – 5) A Lifting Circular Cylinder
−Γ
➢ There are three possibilities for the location of the stagnation points since 𝛽 = asin :
4𝜋𝑏𝑢∞
Stagnation point denoted by 𝑆

Γ Γ Γ
<1 =1 >1
4𝜋𝑏𝑢∞ 4𝜋𝑏𝑢∞ 4𝜋𝑏𝑢∞

• 𝛽 is acute and negative • 𝛽 = 3𝜋Τ2 and −𝜋Τ2 • No stagnation point on the

cylinder surface
• Two stagnation points at • Hence, there is only
𝜃 = 𝛽 and 𝜃 = 𝜋 − 𝛽 one stagnation point • The stagnation point is at a
larger radial distance from
the centre of the cylinder
➢ Hence, the fore and aft stagnation points move downward and is detached from the
along the surface of the cylinder as Γ increases surface
48
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Construction of 2D Steady Irrotational Flows – 5) A Lifting Circular Cylinder
Γ
➢ On the surface of the cylinder 𝑢𝜃, 𝑠 = −2𝑢∞ sin 𝜃 − and 𝑢𝑟 = 0
2𝜋𝑏
Γ
𝑢𝜃, 𝑠 = −2𝑢∞ sin 𝜃 +
4𝜋𝑏𝑢∞

➢ The surface pressure distribution is calculated by using Bernoulli equation:

1 2 1 2
𝑃∞ + 𝜌𝑢∞ = 𝑃𝑠 + 𝜌𝑢𝜃, 𝑠
2 2
2
1 2 1 2 Γ
𝑃∞ + 𝜌𝑢∞ = 𝑃𝑠 + 𝜌𝑢∞ 4 sin 𝜃 +
2 2 4𝜋𝑏𝑢∞

2
1 2 Γ
𝑃𝑠 = 𝑃∞ + 𝜌𝑢∞ 1 − 4 sin 𝜃 +
2 4𝜋𝑏𝑢∞

𝑃𝑠 − 𝑃∞
➢ The surface pressure coefficient is 𝐶𝑝 =
1 2
𝜌𝑢
2 ∞
2
Γ
𝐶𝑝 = 1 − 4 sin 𝜃 +
4𝜋𝑏𝑢∞ 49
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Construction of 2D Steady Irrotational Flows – 5) A Lifting Circular Cylinder
−Γ
𝛽 = asin
4𝜋𝑏𝑢∞

−Γ
sin 𝛽 =
4𝜋𝑏𝑢∞

➢ Hence, the surface pressure distribution and coefficient can also be written as

2 2
1 2 Γ Γ
𝑃𝑠 = 𝑃∞ + 𝜌𝑢∞ 1 − 4 sin 𝜃 + 𝐶𝑝 = 1 − 4 sin 𝜃 +
2 4𝜋𝑏𝑢∞ 4𝜋𝑏𝑢∞

1 2 2 2
𝑃𝑠 = 𝑃∞ + 𝜌𝑢∞ 1 − 4 sin 𝜃 − sin 𝛽 𝐶𝑝 = 1 − 4 sin 𝜃 − sin 𝛽
2

50
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Construction of 2D Steady Irrotational Flows – 5) A Lifting Circular Cylinder
➢ The streamlines are asymmetric about the 𝑥-axis, hence there is a net vertical force acting
on the cylinder
𝑦

𝑢∞
𝑟
𝜃 𝑥

2
Γ Γ
➢ From 𝐶𝑝 = 1 − 4 sin 𝜃 + and 𝑢𝜃, 𝑠 = −2𝑢∞ sin 𝜃 − , we can see that
4𝜋𝑏𝑢∞ 2𝜋𝑏

• At 𝜃 = 𝜋Τ2, 𝑢𝜃, 𝑠 is the highest and 𝐶𝑝 is the lowest

• Hence, the net vertical force is an upward lift force

51
10.2 Potential Flow Theory for Incompressible Inviscid Flows
A Lifting Circular Cylinder – Kutta-Joukowski Lift Theorem 𝑦
𝑑𝐹
➢ Consider a tiny sector with angle 𝑑𝜃 at a positive 𝜃 𝜃
𝑑𝜃
➢ The pressure force acting on the arc of the sector is 𝜃
𝑥
𝑑𝐹 = 𝑃𝑠 (𝑑𝜃)𝑏
𝑏
➢ The upward component of 𝑑𝐹 is
−𝑃𝑠 𝑑𝜃 𝑏 sin 𝜃

➢ Sum the contributions by all these tiny sectors over the whole cylinder:
2𝜋
′
𝐿 = න −𝑃𝑠 𝑏 sin 𝜃 𝑑𝜃 where 𝐿′ is the 2D lift force
0

2𝜋 2
1 2 Γ
= −𝑏 න 𝑃∞ + 𝜌𝑢∞ 1 − 4 sin 𝜃 + sin 𝜃 𝑑𝜃
0 2 4𝜋𝑏𝑢∞
Circulation Γ = 2𝜋𝑏𝑢𝜃
𝐿′ = 𝜌𝑢∞ Γ Kutta-Joukowski lift theorem

named after Kutta and Joukowski (or Joukowsky or Joukovsky or Zhukovsky)

who first developed its key ideas in the early 20th century 52
10.2 Potential Flow Theory for Incompressible Inviscid Flows
A Lifting Circular Cylinder – Kutta-Joukowski Lift Theorem
➢ In the real world, the development of circulation requires viscosity

➢ However, the magnitude of circulation is independent of viscosity but depends on the flow speed,
the shape and orientation of the object

➢ Kutta-Joukowski theorem is very general and is valid for any closed body placed in a uniform
stream

➢ It is useful for calculating lift generated by an aerofoil or a wing

➢ It predicts zero drag and non-zero 𝑢𝜃 on the body surface

𝑢∞
−𝚪
𝑥

53
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Magnus Effect
➢ Magnus effect is the phenomenon of producing lift by the rotation of a solid body

➢ When the body, such as a circular cylinder or a sphere, is not spinning, no lift is produced because
of top-bottom symmetry of the flow field

➢ When the body rotates about an axis, the body drags some fluid around because of the no-slip
condition and the resulting flow has no top-bottom symmetry, hence lift is produced

Real flows Potential flows

𝑦
Drag Lift Drag Lift
𝑟
𝑢∞
𝜃
𝑥

Drag Lift Drag Lift 𝑦

𝑢∞

54
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Magnus Effect

The variation of lift and drag coefficients of a

smooth sphere with the non-dimensional
rate of rotation for Re = 𝑉𝐷 Τ𝜈 = 6 × 104

55
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Example 1
A circular cylinder or radius 𝑏 = 0.5 m and length 𝑙 = 3 m is spinning at a rate of 𝑠 = 5 revolutions
per second in a flow field with freestream velocity 𝑈∞ = 2 m/s. Estimate the life force 𝐿 generated.

Take air density 𝜌 = 1.225 kgΤm3 .

𝑦
𝐿
𝑢∞
𝑠
𝑥
𝑏

56
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Example 1
➢ The circulation Γ = 2𝜋𝑏𝑢𝜃
where 𝑢𝜃 is the circumferential velocity of the fluid adjacent to the surface of
the cylinder and is given by 𝑢𝜃 = 𝜔𝑏, where 𝜔 is the angular velocity of the
rotating cylinder
𝑦
➢ ∵ 𝜔 = 2𝜋𝑠 unit: rad/s 𝐿
𝑢∞
∴ Γ = 2𝜋𝑏(2𝜋𝑠𝑏) 𝑠

Γ = 4𝜋 2 𝑏 2 𝑠 𝑏 𝑥

➢ By Kutta-Joukowski lift theorem, the 2D lift (lift on the 𝑥𝑦 plane) is

𝐿′ = 𝜌𝑈∞ Γ

➢ Hence, the lift force generated is

𝐿 = 𝐿′ 𝑙
= 𝜌𝑈∞ (4𝜋 2 𝑏 2 𝑠)𝑙

𝐿 = 363 N 57
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Conformal Mapping – Joukowski Transformation
➢ Joukowski used a conformal map to transform a circular cylinder to an aerofoil, this is known as
Joukowski transformation

➢ A conformal map preserves angles between the intersecting curves, but length and orientation
of curves might not be preserved

➢ Conformal maps in 2D are locally invertible complex analytic functions

𝑧-plane 𝑤-plane
𝑦 𝑓

𝑥 𝑒
An example of a conformal map

58
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Conformal Mapping – Joukowski Transformation
➢ For simplicity, the cylinder plane is referred to as the 𝑧-plane, while the aerofoil plane is referred
to as the 𝑤-plane

➢ In the ‘original’ 𝑧-plane, let the value of 𝑧 of a point be 𝑧 = 𝑥 + 𝑖𝑦, where 𝑖 = −1

➢ In the ‘transformed’ 𝑤-plane, let the value of 𝑤 of a point be 𝑤 = 𝑒 + 𝑖𝑓

1
➢ The Joukowski transformation is given by 𝑤 =𝑧+
𝑧

➢ By substituting the coordinates of a point 𝑧 = 𝑥 + 𝑖𝑦 into this equation, the transformed

point in the 𝑤-plane is found to be

1 1
𝑤 =𝑥 1+ + 𝑖𝑦 1 −
𝑥2 + 𝑦2 𝑥2 + 𝑦2

1
➢ Hence, the transformed real part 𝑒 = 𝑥 1 + , and the transformed imaginary part
𝑥2 + 𝑦2
1
𝑓 =𝑦 1−
𝑥 2 + 𝑦2 59
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Conformal Mapping – Joukowski Transformation
➢ By plotting the transformed shape, it can be found that

• a circle with a radius 𝑟 = 1 (a unit circle) centred at the origin of the 𝑧-plane is transformed
into a horizontal flat plate between 𝑤 = −2 and 𝑤 = 2

𝑧-plane 𝑤-plane
1
𝑤 =𝑧+
𝑧

• when 𝑟 ≠ 1, the circle centred at the origin in the 𝑧-plane transformed into an ellipse in the 𝑤-
plane
𝑧-plane 𝑤-plane
1
𝑤 =𝑧+
𝑧

60
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Conformal Mapping – Joukowski Transformation
➢ By plotting the transformed shape, it can be found that

• a circle centred at 𝑧 = 1 with its circumference touches the point 𝑧 = −1 (hence 𝑟 = 2 and
the circle is shifted right) is transformed into an aerofoil which has a cusped trailing edge at
𝑤 = −2
𝑧-plane 𝑤-plane
1
𝑤 =𝑧+
𝑧

• a circle centred at 𝑧 = −1 with its circumference touches the point 𝑧 = 1 (hence 𝑟 = 2 and
the circle is shifted left) is transformed into an aerofoil which has a cusped trailing edge at
𝑤=2
𝑧-plane 𝑤-plane
1
𝑤 =𝑧+
𝑧

61
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Conformal Mapping – Joukowski Transformation
➢ By plotting the transformed shape, it can be found that

• a circle centred at 𝑧 = 0.5𝑖 (hence the circle is shifted up) is

transformed into a cambered flat plate with positive camber

𝑧-plane 𝑤-plane
1 Chord
𝑤 =𝑧+ Camber line
𝑧
Max
thickness Max camber

The camber line is the curve

• a circle centred at 𝑧 = −0.5𝑖 (hence the circle is shifted down) is that is halfway between the
transformed into a cambered flat plate with negative camber upper and lower surfaces

𝑧-plane 𝑤-plane
1
𝑤 =𝑧+
𝑧 Here, the circle in the 𝑧-plane
touches 𝑧 = ±1, hence the
transformed shape, which is a
curve, is between 𝑤 = −2 and
𝑤=2
62
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Conformal Mapping – Joukowski Transformation
➢ Therefore, a family of aerofoils called Joukowski aerofoils, all with cusped trailing edge, can be
obtained from Joukowski transformation, e.g.
𝑧-plane 𝑤-plane
1
𝑤 =𝑧+
𝑧

➢ The location of the aft stagnation point on the surface of a lifting cylinder changes with
circulation Γ

➢ Hence, by changing Γ so that the aft stagnation point is at 𝑧 = 1, which transformed to the
cusped trailing edge of the Joukowski aerofoil at 𝑤 = 2, the resulting smooth flow near the
aerofoil trailing edge is more realistic

Finite angle trailing edge Cusped trailing edge

63
Very closed to point 𝑎: 𝑉1 = 𝑉2 ≠ 0
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Conformal Mapping – Joukowski Transformation
➢ Define complex potential in the 𝑧-plane as

𝑇 𝑧 = 𝜙(𝑥, 𝑦) + 𝑖𝜓(𝑥, 𝑦)

which is analytic in the complex 𝑧-plane, and hence by Cauchy-Riemann equations

𝑑𝑇 𝜕𝜙 𝜕𝜓
𝑧 = −𝑖
𝑑𝑧 𝜕𝑥 𝜕𝑦

𝑇′ 𝑧 = 𝑢𝑧 − 𝑖v𝑧 where subscript 𝑧 shows the velocity

components are in the 𝑧-plane

➢ Similarly, in the 𝑤-plane, 𝑇′ 𝑤 = 𝑢𝑤 − 𝑖v𝑤 , where subscript 𝑤 shows the velocity components
are in the 𝑤-plane, and it can be shown that

𝑢𝑧 − 𝑖v𝑧
𝑢𝑤 − 𝑖v𝑤 =
𝑑𝑤
𝑑𝑧
1 𝑑𝑤 1
➢ Since 𝑤 = 𝑧 + , hence =1− 2
𝑧 𝑑𝑧 𝑧
𝑑𝑤
➢ 𝑧 = ±1 are the critical points, at which = 0 , hence it is required that 𝑢𝑧 = v𝑧 = 0 at the
𝑑𝑧
critical points if they are on the aerofoil surface like in a Joukowski aerofoil 64
10.2 Potential Flow Theory for Incompressible Inviscid Flows
Conformal Mapping – Joukowski Transformation

𝑧-plane 𝑤-plane
1
𝑤 =𝑧+
𝑧

No circulation

𝑧-plane 𝑤-plane
1
𝑤=𝑧+
𝑧

Circulation such that

rear stagnation point is
at cusped trailing edge
65