You are on page 1of 253

FDeg Year 2

Aerodynamics
2009/2010
Prof Andrew Rae
Learning Outcomes and Assessment Methods
Learning Outcomes - On successful completion of thisAssessment Methods
module, the student will be able to:

1. Identify and analyse the aerodynamic forces on anExam


aircraft. Explain the effects of airflow at subsonic,
transonic and supersonic speeds.

2. Discuss the different types of aerodynamic experimentalAssignment


methods and the advantages and disadvantages of each
method.

3. Describe the factors leading to flow separation and solveExam


simple boundary layer and skin friction problems, using
standard basic results.

4. Discuss the relative merits of standard wing planformsExam/Assignment


and explain the use and benefits of lift augmentation
devices.

Aerodynamics 2009/2010 Page 2


Content (1)

Fundamentals:
• Static dynamic and total pressure; Bernoulli’s principle; Speed of
sound and Mach number; ISA tables.

Lift generation:
• Circulation, lift and downwash; Kutta condition; Kutta-Joukowski
theorem; spanwise lift distribution, loads and bending moment.

Subsonic Flows:
• Contributions to subsonic drag; zero-lift drag, skin-friction;
‘Horseshoe’ vortex system; wing planforms in subsonic flow;
induced drag; span efficiency; tip devices; wing design through twist
and camber including “wash-out” and “wash-in”.

Aerodynamics 2009/2010 Page 3


Content (2)
Viscosity:
• Definition; boundary layers; Reynolds number; velocity profiles; no-slip
condition; effect of surface roughness on skin friction; laminar and
turbulent flows; local and global skin friction calculations; boundary layer
thickness definition; momentum and displacement thickness; equivalent
body in inviscid flow; transition and flow separation.

Aerodynamic methods;
• History of aerodynamic testing; Wind tunnel types; low speed and high
speed testing; Open and closed circuit (Eiffel/Goettingen) type tunnels;
Open, closed and slotted/porous working section type tunnels; Flight
testing; Model mounting systems; upwash, buoyancy and blockage
correction methods; Mach similarity; Methods of increasing Reynolds
number; Powered wind tunnel models; Pressurised, cryogenic, heavy gas
and water tunnels; Introduction to CFD; Description of CFD; Advantages
and disadvantages of CFD; Examples/demonstration of CFD usage.

Aerodynamics 2009/2010 Page 4


Content (3)
Lift augmentation and flow control devices;
• The need for high lift; history of high lift; slats, flaps and other high lift devices;
the effects of slots; Coanda effects and blown devices; powered high lift
devices; vortex generators.

Supersonic Flows:
• Critical Mach Number; formation of shockwaves; Normal and oblique
shockwaves; Effect of wing thickness and camber; Wave drag and methods
of reducing wave drag (Wing Sweep, Transonic Area Ruling, Supercritical
Aerofoil design, Wing design); Shockwave control and the Shock-induced
separation.

Swept wings:
• Swept wing flows; Effect of spanwise and normal velocity components;
qualitative description of 3D boundary layers on swept wings; Forward,
rearward and variable sweep wings; control surface effects; delta wings and
vortical flows; vortex flap; aerodynamics of aircraft at high incidences.

Aerodynamics 2009/2010 Page 5


Fundamentals
• Static dynamic and total pressure;
Bernoulli’s principle; Speed of sound and
mach number; ISA tables.

Aerodynamics 2009/2010 Page 6


Lift Generation
• Circulation, lift and downwash; Kutta condition;
Kutta-Joukowski theorem; spanwise lift distribution,
loads and bending moment.

Aerodynamics 2009/2010 Page 7


Lift generation (1)
Circulation
– A term meaning rotation, which in aerodynamics is usually associated with
vorticity.

– An commonly seen example is a type of forced circulation called the Magnus


effect

• If a cylinder or sphere is made to


rotate as it travels through air,
friction causes:
– the air on the forward moving side
to slow down
– the air on the rearward moving
side to speed up
– a differential pressure (Bernoulli)
and a lift force

 The object moves sideways

Aerodynamics 2009/2010 Page 8


Lift generation (2)
Spin on a tennis ball
Some you might have seen…..

A David Beckham free


kick

A ‘curve’ ball
Slice on a golf ball
Purposely ignoring cricket - polishing, seam, boot studs, etc. are all separation control (dimples on
a golf ball)….see later….

Aerodynamics 2009/2010 Page 9


Lift generation (3)
• An aerofoil is a body that induces the same effect through shape only.

Note: Pictures from ‘An Album of Fluid Motion’ (Parabolic


• Consider first inviscid flow (no friction) Press)

Dye injection shows the streamlines in water flowing at 1mm/s


between glass plates spaced 1mm apart. It is interesting that
the best way of showing the unseparated patterns of inviscid
flow (which would be spoiled by separation in a real fluid of even
the slightest viscosity) is to go to the extreme of creeping flow in
a narrow gap, which is dominated by viscous forces, i.e.
boundary layers.

• The flow around a body produces changes in velocity and thus changes in pressure
– but the pressure variations are symmetrical, i.e. no lift

Aerodynamics 2009/2010 Page 10


Lift generation (4)
• So, in inviscid flow, the pressures on the upper and lower surfaces of
an aerofoil are equal and thus so are the velocities:

• In real life (air, water) the flow is viscous:


– The flow on the lower surface will not traverse the sharp trailing edge –
there is a limiting curvature (pressure gradient) round which a viscous
fluid will flow (spoon under a tap)
– By not doing so it creates a creates a partial vacuum (low pressure) on
the trailing-edge upper surface
– This draws the upper-surface flow down to the trailing edge too
– Both upper and lower-surface flows leave smoothly at the aerofoil at the
trailing-edge
– the Kutta Condition (M.Wilhelm Kutta, Germany, 1902)
Aerodynamics 2009/2010 Page 11
Lift generation (5)
• So, the combination of pressure gradients and the Kutta condition
results in higher velocity air over the upper surface and lower
velocity air over the bottom surface and hence lift

• A mathematical way of representing this is to take the inviscid


flow…

Aerodynamics 2009/2010 Page 12


Lift generation (6)
• So, the combination of pressure gradients and the Kutta condition
results in higher velocity air over the upper surface and lower
velocity air over the bottom surface and hence lift

• …and add circulation (rotation, a vortex)

Aerodynamics 2009/2010 Page 13


Lift generation (7)

Aerodynamics 2009/2010 Page 14


Lift generation (8)
• The vorticity is not just a mathematical device, it is a real effect and
can be seen most obviously when it is shed from a wing tip.
• In addition, the velocity gradients in the boundary layer produce
vorticity that is thus distributed along the aerofoil surface.

Aerodynamics 2009/2010 Page 15


Lift generation (9)

• The circulation (vortex) has a strength (Γ ) and the lift generated


by a 2-D aerofoil (or per unit span for a 3-D wing) is given by:

L = ρU Γ (the Kutta-Joukowski Theorum)

where ρ is the density of the air and U is the velocity of the


aerofoil.
– Thus for a given speed and altitude, higher lift means stronger
vorticity.

• So, could there be lift without friction?


– Lift is a result of the surface pressure distribution
• an inviscid phenomenon
– But the differential pressure between upper and lower surfaces is a
result of the Kutta condition
• a viscous phenomenon

Aerodynamics 2009/2010 Page 16


Lift generation (10)
• For a finite wing, the combination of
bound and shed (free) vorticity is called
the horseshoe vortex system
• A simple, finite, rectangular wing can be
represented as a single bound vortex of
constant strength, and a pair of semi-
infinite trailing vortices
• The bound vortex is located at the
centre of pressure (~c/4)

Anderson ‘Fundamentals of Aerodynamics’

Aerodynamics 2009/2010 Page 17


Lift generation (11)
• The starting vortex
– When an aerofoil starts
moving the flow tried to curl
around the trailing edge
– In so doing the flow velocities
there become very large
– Consequently a thin region of
very large velocity gradient
(and thus high vorticity) is
formed at the trailing edge
– Once the flow is established,
the flow leaves the trailing
edge smoothly (Kutta
condition) and the velocity
gradients disappear
– The shed vorticity rolls up into
a starting vortex

Aerodynamics 2009/2010 Page 18


Lift generation (12)
• Trailing vorticity – an aside (#1)
– Crow instability
– Condensation trails from a B-
47 taken at 15s intervals after
its passage

Aerodynamics 2009/2010 Page 19


Lift generation (13)
• Trailing vorticity – an aside (#2)
– Wake vorticity and wake
encounters
– Especially on descent (ρ U
Γ)

Aerodynamics 2009/2010 Page 20


Lift generation (14)
• The bound and trailing vortices are examples of vortex
filaments:
– Lines of constant strength (point) vorticity
– They can be curved but for current purposes we will consider only
straight filaments
• The vortex filament will induce flow around it, depending on its
strength and direction of circulation
– E.g. the bound vortex….

– ….and the trailing vortices too


Aerodynamics 2009/2010 Page 21
Lift generation (15)

• Based on the Biot-Savart Law (see Anderson, ‘Fundamentals of


Aerodynamics’ Section 5.2), the magnitude of the velocity (V) at a point (P)
that is at a perpendicular distance (h) from a vortex filament of strength Γ is
given by the equation:

Γ Γ ∞
V=
4πh

─∞ V

• The magnitude of the induced velocity decreases with increasing distance


from the vortex
Aerodynamics 2009/2010 Page 22
Lift generation (16)

• So from before, with the horseshoe vortex system, the downwash


(w) induced along the span of the wing by the trailing vortices can
be shown as:

Anderson ‘Fundamentals of Aerodynamics’

Aerodynamics 2009/2010 Page 23


Lift generation (17)
• The angle of attack
(α) between the
chord line and the
free stream (V∞) is
the geometric angle
of attack

• The downward component of velocity generated by downwash at the wing is w, producing a


local relative wind inclined from the below V∞ the induced angle of attack αi. This has two
effects:
– The angle of attack seen by the aerofoil is less than the geometric angle of attack and is known as
the effective angle of attack
αeff = α – αi
– The local lift vector is perpendicular to the local relative wind and thus is now inclined behind the
vertical by the angle αi and thus has a longitudinal component which contributes to drag, Di

Aerodynamics 2009/2010 Page 24


Lift generation (18)
• But what is wrong with this picture?
– Only applicable to rectangular wings
• No taper, no twist, no sweep
– Not even brilliant for that:
• Downwash of infinite value at the wing tip?

• Consider the concept of lift distribution


– The variation of lift along the span

• For the horseshoe vortex system, the bound vortex is of constant


strength and the lift is thus constant across the span
BUT
• The equalisation of pressure at the wing tip (bleed between upp an
lower surfaces) means that in reality there is zero lift at the wing tip

Aerodynamics 2009/2010 Page 25


Lift generation (19)

• The lift distribution of a simple horseshoe vortex system would thus


be draw as:

y
• Whereas it should be:

(The local lift is the local height of the lift distribution and the total lift is the area under the lift distribution curve)

Aerodynamics 2009/2010 Page 26


Lift generation (20)
• To model reality better we introduce a concept called the lifting line
– Ludwig Prandtl (Göttingen, 1911-1918)
• What happens if we take a single horseshoe vortex and add
another of smaller span on same chordline?

Γ2 b
Γ 1+
Γ1 2 Γ1
Γ2 Γ2 ∞

Γ2
Γ1
Γ1
Γ2 ∞
b
− Γ1 ∞
2

Aerodynamics 2009/2010 Page 27


Lift generation (21)

• By playing games with the number, span and strength of the


horseshoe vortices we can achieve any lift distribution to define any
planform
– Still used today for preliminary calculations, bearing in mind its
limitations (inviscid, incompressible)

Aerodynamics 2009/2010 Page 28


CL = lift coefficient
ρ = pressure

Lift generation (22) V


S
= velocity
= wing area
M = aircraft mass
g = acceleration due to gravity
WS = wing loading
• Wing loading
– The mass of an aircraft divided by the area of its wing planform
• A useful indicator of an aircraft’s handling and performance

& L
CL =
L 1 Mg
1 = ρ V 2C L = = WS g
ρ V 2
Take-off and landing S performance S 2 S

2
Rearranging the equation for lift coefficient gives

and therefore

So for aircraft2with
g Wthe
S same lift coefficient at take-off and landing (and in the same atmospheric

V =
conditions) the aircraft with the bigger wing will need lessV α at
speed Wtake-off and landing, or less lift
ρ C L
coefficient at the same speed
S

Aerodynamics 2009/2010 Page 29


CL = lift coefficient
ρ = pressure

Lift generation (23) V


S
= velocity
= wing area
M = aircraft mass
g = acceleration due to gravity
• Initial climb rate WS = wing loading
av = vertical acceleration
– Newton’s Second Law; F = Ma
– At rotation, the lift generated is greater than that needed to balance the aircraft
weight otherwise the aircraft would not get airborne.
– The vertical force due to the difference between the lift generated and the
aircraft weight is thus
Lift – Weight = L – Mg

=
1
ρ V 2 S C L − Mg
2
– And the vertical acceleration can then be found from

1 1
M av = ρ V 2 S CL − M g av = V 2 ρ CL − g
– So for the2 same lift coefficient the aircraft with the
2WSlower wing loading will have
the greater initial climb

Aerodynamics 2009/2010 Page 30


CL = lift coefficient
ρ = pressure

Lift generation (24) V


S
= velocity
= wing area
M = aircraft mass
= acceleration due to gravity
• Turning performance g
WS = wing loading
– An aircraft performing a constant-speed, constant av = vertical acceleration
radius turn obeys the mathematical rules of circular ac = centripetal acceleration
motion
• The velocity (rate of change of position) is the distance
travelled around the circle (the circumference) divided by
the time taken to complete a rotation. i.e.
V
ac

• The acceleration is rate of 2 π R of velocity divided by


change
V = a rotation. The velocity rotates
the time taken to complete R
by 2π in time T. T

• Rearranging and substituting gives


2π V
ac =
T

V2
ac =
R

Aerodynamics 2009/2010 Page 31


CL = lift coefficient
ρ = pressure

Lift generation (25) V


S
= velocity
= wing area
M = aircraft mass
g = acceleration due to gravity
• Turning performance (continued) WS = wing loading
av = vertical acceleration
– The centripetal force is given by
ac = centripetal acceleration
Newton’s 2nd Law and is equal to the θ = bank angle
horizontal component of lift, so

L
MV2 1
= ρ V 2 C L S sin θ
R 2
θ
M 2 2 WS
R = =
S ρ C L sin θ ρ C L sin θ

– And thus turn radius (R) is proportional


to wing loading

Aerodynamics 2009/2010 Page 32


CL = lift coefficient
ρ = pressure

Lift generation (26) V


S
= velocity
= wing area
M = aircraft mass
g = acceleration due to gravity
WS = wing loading
av = vertical acceleration
ac = centripetal acceleration
θ = bank angle

∆L S ∆L
av = =
M WS

N.B. – These analyses are good indicators of aircraft performance


and handling but should be treated with care as they are
essentially static assessments of what are, in reality, dynamic
manoeuvres
– Mean wind direction, turbulence, aircraft pitch, roll, yaw, etc.

Aerodynamics 2009/2010 Page 33


Lift generation (27)
• A lot of mathematical analysis, but what does it mean in practice?

• Landing and take-off performance


– Airbus aircraft have lower wing loading than Boeing aircraft and thus can
trade this for lower CL at take-off and landing
• Simpler and lighter high-lift devices (see later)

Aerodynamics 2009/2010 Page 34


Lift generation (28)
• Initial climb rate
– The original A340 is underpowered but gets way with it because of low
wing loading
– It can achieve a certifiable climb rate even with relatively small engines
• CFM56 on A340-200 and -300 (34,000 lbf each)
• RR Trent 500 on A340-500 and -600 (60,000 lbf each)

Aerodynamics 2009/2010 Page 35


Lift generation (29)
• Turning performance
– Fighter aircraft
• The difference between fighter and interceptor

Aerodynamics 2009/2010 Page 36


Lift generation (30)
• Turning performance
– Fighter aircraft
• The Spitfire had much lower wing
loading
BUT
• The Bf109 had automatic
leading-edge slats
 Higher CL

Aerodynamics 2009/2010 Page 37


Lift generation (31)
• Gust response
– Airbus (low wing loading)
• Lower landing and take-off speeds, better initial climb
vs
• Boeing (high wing loading)
• Smoother cruise

Aerodynamics 2009/2010 Page 38


Lift generation (32)
• Gust response
Low-level bomber vs high-level bomber

Aerodynamics 2009/2010 Page 39


Lift generation (33)

Moment arm

Aerodynamics 2009/2010 Page 40


Lift generation (34)
• Wing root bending moment
– 0g to 1g flight
– 2.5g gust and manoeuvre load

Aerodynamics 2009/2010 Page 41


Lift generation (35)
• Ways to alleviate wing root bending moment
– Engines
– Winglets

Aerodynamics 2009/2010 Page 42


Subsonic Flows
• Contributions to subsonic drag; zero-lift drag, skin-friction; ‘Horseshoe’
vortex system; wing planforms in subsonic flow; induced drag; span
efficiency; tip devices; wing design through twist and camber including
“wash-out” and “wash-in”.

Aerodynamics 2009/2010 Page 43


Viscosity
Definition; boundary layers; Reynolds number; velocity profiles; no-slip condition; effect
• of surface roughness on skin friction; laminar and turbulent flows; local and global skin
friction calculations; boundary layer thickness definition; momentum and displacement
thickness; equivalent body in inviscid flow; transition and flow separation.

Aerodynamics 2009/2010 Page 44


Viscosity (1)
• What is viscosity?
– The material property that measures a fluid's resistance to flowing
• It concerns the transport of mass, momentum and energy when the molecules move.
– It results in friction between air and any surface over which it flows

Aerodynamics 2009/2010 Page 45


Viscosity (2)
• Boundary Layers
– A thin region of the flow adjacent to a surface, where the flow is retarded
by the influence of friction between a solid surface and the fluid
(Anderson, ‘Fundamentals of Aerodynamics’)
– Its characteristics (size, composition, etc.) are determined by a variety of
things
• viscosity (friction)
• Shape (pressure gradient)
• Reynolds number (density, velocity)
– Extremely difficult to measure u=U∞
U∞
– Still not fully understood y (v)
• Transition mechanisms
• The limit on many numerical x (u) y=0, u=0, v=0
methods

Aerodynamics 2009/2010 Page 46


Viscosity (3)
• Boundary Layers
– Inviscid vs Viscous

Aerodynamics 2009/2010 Page 47


Viscosity (4)
• Reynolds number
– A non-dimensional measure of the ratio of inertia forces (ρ U2) to viscous forces
(μU/d)
Stokes
• Where ρ = density
U = velocity
d = reference length
μ = viscosity

– Which gives ρU d
Re =
µ
– One of the most powerful parameters in fluid
dynamics
• Helps assess the similarity or
equivalence of differing flow conditions
• Scale effect
– High Re flows approach inviscid conditions
(thin boundary layers)

Reynolds
Aerodynamics 2009/2010 Page 48
Viscosity (5)
• Reynolds number

Aerodynamics 2009/2010 Page 49


Gliders
Viscosity (6) Re <20000

• Reynolds number
– Some examples

X15
Re = 6 million

A380
Re = 80 million

Aerodynamics 2009/2010 Page 50


European
EWA Wind Tunnel Performance Envelopes
25

Wind Tunnels
QinetiQ 5m

Viscosity (7)
ONERA S1MA
20

Reynolds number (millions per metre)


CIRA IWT
VZLU 3 m
ONERA F1
FOI LT1
15
DNW-LLF

Low-speed w/t 10

• Reynolds number 5

– Wind tunnel coverage 0


0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50
Mach Number

• (see ‘Aerodynamics Methods’)


EWA Wind Tunnel Performance Envelopes

ONERA S2MA
100
DNW-HST
ONERA S1MA
90 ARA-TWT

Reynolds number (millions per metre)


CIRA PT-1
80 VZLU A1
CIRA IWT
70 FOI T1500
FOI S4
FOI S5
60 FOI TVM500
50

Transonic w/t 40

30

20

10

0
0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40
Mach Number

EWA Wind T unnel Performance Envelopes


ONERA S2MA
140
DNW-SST
FOI T1500
120 FOI S4

Reynolds number (millions per metre)


FOI S5
100 FOI TVM500

80

Supersonic w/t 60

40

20

ETW (Cryogenic)
1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00
Mach Number

Aerodynamics 2009/2010 Page 51


Viscosity (8)

• Velocity profiles
– The rate at which the velocity increases from zero at the wall to freestream
velocity at the edge of the boundary layer

δ u=0.99U∞
Boundary-layer
U∞
y (v) velocity profile

x (u) y=0, u=0, v=0

– The edge of the boundary layer can be difficult to define so a (sometimes


artificial) definition is imposed:
u = 0.99 U∞

– The boundary-layer depth (the distance from the surface where u = 0.99 U∞)
is defined as δ

Aerodynamics 2009/2010 Page 52


Viscosity (9)

• Shear Stress
– The assertion that the velocity at the surface is zero
– The action of viscosity tugs at the surface (rubbing hands together)
• Generates shear stress (τxy )

– You can imagine this as two adjacent layers of


fluid, each at different velocities rubbing
against each other
– The shear stress is thus related to the
difference in velocity between the two layers
and that is defined by the velocity gradient
(du/dy), so
du du
τ xy ∝ =µ
dy dy
• In reality there is a semi-infinite number of
adjacent layers (solid boundary)
Aerodynamics 2009/2010 Page 53
Viscosity (10)

• No-slip condition
– The no-slip condition maintains
that the flow at the surface is
stationary
• i.e. that u = 0 at y = 0
– This is difficult to justify
theoretically and is
demonstrably not true in many
cases
– But it is close enough to the
truth (and the convenience of it
as a boundary condition so
large) that its consequences
are accepted

Aerodynamics 2009/2010 Page 54


Viscosity (11)
• Laminar and Turbulent Boundary Layers
– When a boundary layer starts on a surface it is laminar, i.e. smooth with the stream lines roughly parallel to the surface
– At some point a transition occurs (due to roughness, contamination, pressure gradients, etc.) to a turbulent boundary layer

– There is a general mean motion roughly parallel to the surface, but in addition there are local rapid, random fluctuations in velocity
direction and magnitude

– These fluctuations provide a powerful mechanism for mixing within the layer
– Just as viscosity give rise to shear stress, the turbulent fluctuations give rise to eddy shear stresses
– Consequently there a important differences between the characteristics of laminar and turbulent boundary layers

Aerodynamics 2009/2010 Page 55


Viscosity (12)

• Laminar and Turbulent Boundary Layers


– Boundary-layer profile

Steeper profile

Aerodynamics 2009/2010 Page 56


Viscosity (13)
• Laminar and Turbulent Boundary Layers
– Boundary layer on a flat plate

– Boundary layers on a wing combine to form the wake (profile drag, CDo)

Aerodynamics 2009/2010 Page 57


Viscosity (14)

• Laminar and Turbulent Boundary Layers (see also later)


– Transition

Aerodynamics 2009/2010 Page 58


Viscosity (15)
• Laminar and Turbulent Boundary Layers
– Characteristics which encourage transition
• Increased surface roughness
– Boundary-layer tripping on wind tunnel models
– Dimples on a golf ball

• Increased freestream turbulence


– Wind tunnel comparisons

• Adverse pressure gradients


– Amplification of instabilities
– Sailplane wing profiles

• Heating of the fluid by the surface


– Amplification of instabilities

– The inverse of all these encourage laminar boundary layers

Aerodynamics 2009/2010 Page 59


Viscosity (16)
• Reynolds Experiment
– Classic experiment examining laminar
and turbulent flow in pipes
• Flow through a pipe metered by a
stopcock
• Dye injection at the centreline of the tube
mouth
• Reynolds noted that low speed the dye
filament remained smooth and narrow
• At higher speed the filament broke up
and diffused throughout the cross section
• The speed at which it occurred was
different for pipes of different diameter

– Relationship to U∞ and d, i.e Re

Aerodynamics 2009/2010 Page 60


Viscosity (17)
• Reynolds Experiment

a & b – laminar
c - turbulent

Aerodynamics 2009/2010 Page 61


Viscosity (18)
• Reynolds Experiment
Results
Laminar
– Recreation at the
University of
Manchester using
Reynolds original
apparatus a century
Transitional
later

Turbulent

Turbulent

Aerodynamics 2009/2010 Page 62


Viscosity (19)
• Blasius’ Equation
– Uses the boundary layer equations ∂u ∂v
• Reduction of the Navier-Stokes equations to + =0
simpler forms which apply to boundary layers ∂x ∂y
– Continuity
– x momentum ∂u ∂v ∂ 2u
– y momentum u + v =υ 2
∂x ∂y ∂y
– Uses a function to turn a set of partial differential
equations into a single ordinary differential ∂p
equation =0
∂y

Where υ is the kinematic


viscosity , defined as
2 f '' ' + f f '' = 0 υ≡µ ρ
– See Anderson ‘Fundamentals of Aerodynamics’, Chapter 18.2

V∞ u
η=y f ' (η ) =
υx V∞

Aerodynamics 2009/2010 Page 63


Viscosity (20)

• Blasius’ Equation
– The important result is that the solution of the
equation is a velocity profile and that it is a function
of η only
– This form of the velocity profile is independent of the
distance along a surface (x)
– Self-similar solutions
– If f’=u/Uo, the b.l. edge is at f’=0.99 and η =5.0
V∞ V∞
η=y =δ = 5.0
υx υx
5.0 x
δ=
Re x
– The reduction of the boundary layer equations to an
ODE is only valid for certain conditions
• E.g. flow on a flat plate

Aerodynamics 2009/2010 Page 64


Viscosity (21)
• Local Skin Friction
– Remember that shear stress within the boundary
layer was defined as

du
τ xy = µ
– dyat the wall (skin friction) is given
So shear stress
by

 du 
τ = µ
w local 
coefficient
– And the
 dy  y =0 of skin friction (cf), the
skin friction at a point x along a surface, is given
by

– And is the local coefficient of drag due to viscosity


τw
cf = 1
ρ
2 ∞ ∞V 2

Aerodynamics 2009/2010 Page 65


Viscosity (22)

• Local Skin Friction


– From the boundary layer and Blasius equations1 it can be shown that

 du  V∞
  = V∞ f '' ( 0 )
2υ x
 dy  y =0local shear
– And so the stress is given by

τw 2 V∞
cf = = µ V∞ f ''
( 0)
1
ρ
2 ∞ ∞V 2
ρ V
∞ ∞
2
2 υ x

2 µ 2 f '' ( 0)
– Reference
''
( 0) = solution of the Blasius equation and tabulates the
c f = [2] describes af numerical
2 ρ V
results, giving f’’(∞0) = x
∞ 0.4696, so 2 Re x

– Where Re0x.664
is the local Reynolds
c f
number
=
Re x
1 Anderson ‘Fundamentals of Aerodynamics’, 3rd Edition,
Chapter 18.2
2 Schlichting, ‘Boundary Layer Theory’, 8th Revised and
Enlarged Edition, Page 158 and Table 6.1

Aerodynamics 2009/2010 Page 66


Viscosity (23)

• Global Skin Friction


– If the local skin friction drag coefficient is cf, the global or total drag due to friction on
that surface (say the chord of a wing, c) is found by integrating the local skin friction
over the length of that surface. i.e.

1 c
C f = ∫ c f dx
c 0
– Substituting the previous expression for local skin friction coefficient gives

1 µ c 1.328 µ c

−1
C f = (0.664) x dx =2

– And
c ρ ∞ V∞ 0 c ρ ∞ V∞

1.328
Cf =
– Where ReRe
c is the Reynolds number based on the wing chord
c

Aerodynamics 2009/2010 Page 67


Viscosity (24)

• Boundary layer thickness and displacement thickness


– Remember we described the boundary layer thickness, δ, as being the
distance from the surface where u = 0.99U∞
δ u=0.99U∞
U∞
y (v)

x (u) y=0, u=0, v=0

δ u=0.99U∞ u=0.99U∞

y (v)
δ*
– We now introduce the concept of displacement thickness, δ*
x (u)

Aerodynamics 2009/2010 Shaded regions have equal area Page 68


Viscosity (25)

• Boundary layer displacement thickness


– The displacement thickness can be thought of in two ways:
– (a) The thickness representing the missing mass flow if it were crammed into a
flow with the free stream characteristics (cf. inviscid)

δ u=0.99U∞ u=0.99U∞

y (v)
δ*
x (u)

Shaded regions have equal area

– (b) The boundary layer displaces the flow around an object by acting as
additional volume

Aerodynamics 2009/2010 Page 69


Viscosity (26)

y1
∫ ρ u dy
The actual mass flow between y=0 and y=y1 is

The hypothetical mass flow between y=0 and y=y1 if the boundary layer were not present is
– 0
The difference between the two is the missing mass flow

y1
∫ ρ e ue dy
And this can be expressed in terms of δ*

∫ (ρ ue − ρ u ) dy
y1
e
0

ρ e ue δ *
Aerodynamics 2009/2010 Page 70
Viscosity (27)
• Boundary layer displacement thickness derivations
– (a) Missing mass flow (continued)

– So

∫ (ρ ue − ρ u ) dy
y1
ρ e ue δ = *
e
0
– Or

y1  ρu 
δ = ∫ 1 −  dy
*
0
 ρ e ue 

Aerodynamics 2009/2010 Page 71


Viscosity (28)
Boundary layer displacement thickness derivations

(b) Displacement of external streamlines

. y1
– The mass flow at Station 1 m= ∫0
ρ e ue dy
– At Station 2, the mass flow between the surface and the . y1
m= ∫ ρ u dy + ρ e ue δ *
same streamline is

– Since the surface and the streamline form the boundaries 0


of a stream tube, the mass flow must be constant, i.e.
y1 y1

Or ∫
0
ρ e ue dy = ∫0
ρ u dy + ρ e ue δ *

y1  ρu 
δ = ∫ 1 −  dy as before
*
0
 ρ e ue 

Aerodynamics 2009/2010 Page 72


Viscosity (29)

Aerodynamics 2009/2010 Page 73


Viscosity (30)
• Boundary layer momentum thickness
– While the displacement thickness accounted for the ‘missing mass flow’ another important
boundary-layer characteristic accounts for loss of momentum within the boundary layer

– The mass flow across dy (dm) = ρ u dy


– Now, momentum flow across dy in the b.l. = dm u = ρ u2 dy
– Momentum across dy if it were in the freestream = dm ue = (ρ u dy)ue

– Therefore, the loss in momentum associated with dm = ρ u(ue –u)dy

– So the total momentum deficit from y=0 to y=y1 =

ρ u ( ue − u ) dy
y1

Aerodynamics 2009/2010
∫0
Page 74
Viscosity (31)
Boundary layer momentum thickness
We can now introduce a thickness (θ) representing the missing momentum if it were crammed into a flow with the free stream characteristics, i.e.

– The missing momentum flow = ρ eue θ =
2

And so

ρ u ( ue − u ) dy
y1



This momentum thickness (θ) is the height of a hypothetical streamtube carrying the missing momentum flow at freestream conditions
The momentum thickness can be used to generate a similar effective or equivalent body, this time representing a body exhibiting an equivalent momentum loss

0

y1 ρu  u
θ =∫ 1 −  dy
0 ρ e ue  ue 

Aerodynamics 2009/2010 Page 75


Viscosity (32)
• Methods of measuring boundary layer state
– Hot wires
– Hot films
• Measurement element is one arm of a Wheatstone
bridge
• Measure the voltage changes required to keep a
constant current (CCA), or the current changes
require to keep a constant temperature (CTA)
– Sublimation/Evapouration
– Heat transfer (IR)
– All have difficulties
• Intrusion, chemicals, temperature gradients, viewing
angle, calibration, interpretation

Aerodynamics 2009/2010 Page 76


Viscosity (33)
• Boundary layer transition

– From Reynolds’ pipe flow experiment a critical Reynolds number was observed, below which the flow was laminar and above which it
was turbulent.

– The discovery that transition occurred on surfaces did not come until much later
ud 
– Transition on a flat plate
Re crit =   = 2300
 υ  crit

U∞ x 
Re x crit =  = 3.5 ×10 → 10
5 6

Aerodynamics 2009/2010
 υ  crit Page 77
Flow direction
Viscosity (34)

Factors affecting
transition
• Pressure gradient
– Positive pressure
gradients suppress Disturbances suppressed by positive pressure gradient

turbulence
– Adverse pressure
gradients amplify
them

Disturbances amplified by negative pressure gradient

Shape of the profile


Aerodynamics 2009/2010 Page 78
Viscosity (35)
Factors affecting transition
• Pressure Gradient
– On a wing transition generally occurs at or just after CPmin where the pressure
gradient changes from +ve to –ve (e.g. laminar separation bubble): 5% is a rule
of thumb

Aerodynamics 2009/2010 Page 79


Viscosity (36)
Factors affecting transition
• Pressure Gradient
– Turbulent vs Laminar

Aerodynamics 2009/2010 Page 80


Viscosity (37)
• Factors affecting transition
– Surface roughness
• Imperfections or roughness elements act like little bluff bodies, shedding eddies
which disturb a laminar boundary layer and can induce transition

• It is possible to reduce drag by inducing transition through roughness


– Only if separation is normally of a laminar boundary layer
– A turbulent boundary layer is more resistant to separation than a laminar one
– Triggering transition before separation means the boundary layer will separate later
 Golf Ball

Aerodynamics 2009/2010 Page 81


Viscosity (38)
• Factors affecting transition
– Surface roughness

Laminar
separation

Transition
– Addition of a trip wire
at the wire
Turbulent
separation

Aerodynamics 2009/2010 Page 82


Viscosity (39)
• Factors affecting transition
– Surface roughness
– Ice
– Liquids (rain & de-icing fluid)

Aerodynamics 2009/2010 Page 83


Viscosity (40)
Factors affecting transition

Freestream turbulence (and noise)

E.g. on a flat plate, transition starts with the formation of Tollmien-Schlichting (T-S) waves

External excitation (especially of matching frequencies) can amplify the waves and hasten

transition

3-D geometries (e.g. swept wings)



In addition to chordwise disturbances we now have spanwise flow and spanwise disturbance.

Traditional civil aircraft-type wings are fully turbulent

Attachment line transition

Aerodynamics 2009/2010 Page 84


Viscosity (41)
Boundary layer separation
Once the pressure gradient on a surface becomes positive the pressure rises with distance

The effect of which is shown [top right]

Loss of kinetic energy which is only partially compensated for by mixing within the boundary layer

The velocity profile becomes less full with the inner part of the layer slowing down w.r.t. the outer

The shear stress at the wall reduces

With a sufficiently large pressure gradient a point where the shear stress becomes zero and

the flow on the surface is on the point of reversing

Aerodynamics 2009/2010 Page 85


Viscosity (42)
Boundary layer separation
• The reversed flow forms a large eddy under the outer part of the boundary layer (wakes)

• Open separations are generally unstable and highly dynamic

• Closed separations exist (laminar bubbles) but even these are unsteady

• Classic wind tunnel surface flow visualisation can indicate these regions
– Oil flow (time averaged)
– Tufts (point data)

Aerodynamics 2009/2010 Page 86


Viscosity (43)
Boundary layer separation
• Bluff body separations
– E.g. delta vortices

Aerodynamics 2009/2010 Page 87


Viscosity (44)

• Boundary layer modelling


– It is important to capture the boundary layer with sufficient fidelity in
viscous CFD methods
• Concentration of mesh points close to the surface

Aerodynamics 2009/2010 Page 88


Aerodynamic Methods
• History of aerodynamic testing; Wind tunnel types; low speed and high speed testing; Open and
closed circuit (Eiffel/Goettingen) type tunnels; Open, closed and slotted/porous working section
type tunnels; Flight testing; Model mounting systems; upwash, buoyancy and blockage
correction methods; Mach similarity; Methods of increasing Reynolds number; Powered wind
tunnel models; Pressurised, cryogenic, heavy gas and water tunnels; Introduction to CFD;
Description of CFD; Advantages and disadvantages of CFD; Examples/demonstration of CFD
usage.

Aerodynamics 2009/2010 Page 89


Lift Augmentation and Flow
Control Devices
• The need for high lift; history of high lift; slats, flaps and other
high lift devices; the effects of slots; Coanda effects and
blown devices; powered high lift devices; vortex generators.

Aerodynamics 2009/2010 Page 90


Lift augmentation and flow control devices (1)
• Early (propellor-driven) aircraft had low wing loadings and thus
low take-off and landing speeds
– High-lift systems were either not needed or were relatively simple
• Leading-edge devices, at least on civil aircraft, were rare
• Trailing-edge devices were primarily plain or split flaps

Aerodynamics 2009/2010 Page 91


Lift augmentation and flow control devices (2)

Aerodynamics 2009/2010 Page 92


Lift augmentation and flow control devices (3)
• Handley-Page
– In 1911 Sir Frederick Handley-Page noted that square wings (AR=1)
maintained lift to a much higher incidence than more conventional
rectangular wings (AR≈6)
– In 1917 he and his aerodynamicist (R.O.Boswell) tried to combine the low
drag characteristics of high aspect ratio with the delayed stall of low aspect
ratio by incorporating chordwise slots in a conventional wing

– Wind tunnel test results


were disappointing
– Despite many variations
in shape, gap and
proportion the idea could
not be made to work

Aerodynamics 2009/2010 Page 93


Lift augmentation and flow control devices (4)
• Handley-Page
– At some point someone (whether Handley-Page, Boswell or one of the
carpenters, it is not clear who) had the idea of cutting spanwise slots
• Parallel to the leading edge, at about c/4 and sloping upwards and rearwards

– The initial tests on a RAE15 aerofoil


gave a spectacular 25% increase in
maximum lift
– An improved slot shape in a RAE6
aerofoil gave a 50% increase, with
only a slight increase in drag
– Various test throughout 1918 and
1920 showed that chordwise location
was crucial

Aerodynamics 2009/2010 Page 94


Lift augmentation and flow control devices (4)
• Lachmann
– Independently, a parallel investigation was being conducted by a German engineer-pilot,
Gus Lachmann.
• He transferred to the flying corps from the cavalry in 1917 but stalled and spun-in during an early
training flight, breaking his jaw
– In hospital he pondered the cause of his accident and how stall could be prevented,
concluding that a cascade of small aerofoils within a normal wing profile might be better

– He took his idea to the


German patent office in
February 1918 but this was
rejected unless he could
prove experimentally that the
idea would work

Aerodynamics 2009/2010 Page 95


Lift augmentation and flow control devices (5)
• Lachmann
– He approached Prof Ludwig Prandtl at Gottingen who agreed to do the
tests for £50
– Lachmann had no money and so borrowed it from his mother
– The results convinced the patent office to grant his application
– Lachmann ended up working for
Handley-Page after the end of WW2
-
– The consensus was that the slot
behaves as a boundary-layer control
device
• The jet through the slot
– It was not until 1972 (A.M.O. Smith)
that the correct physical principles
underlying its operation were finally
understood

Aerodynamics 2009/2010 Page 96


Lift augmentation and flow control devices (6)

Aerodynamics 2009/2010 Page 97


Lift augmentation and flow control devices (7)
• Types of high-lift device - leading edge

Aerodynamics 2009/2010 Page 98


Lift augmentation and flow control devices (8)

Aerodynamics 2009/2010 Page 99


Lift augmentation and flow control devices (9)

Aerodynamics 2009/2010 Page 100


Lift augmentation and flow control devices (10)

HIGH-LIFT SYSTEM TERMINOLOGY

Fixed leading
edge (D-nose)
Shroud
Vane
Main element

Cove Main flap


Slat

Aerodynamics 2009/2010 Page 101


Lift augmentation and flow control devices (11)
• Take-off
– Slat deployment and take-off flap setting (large Fowler motion and a little
deflection)

Aerodynamics 2009/2010 Page 102


Lift augmentation and flow control devices (12)
• Landing
– Slat deployment, landing flap setting (large Fowler motion and large
deflection) and spoilers (shroud) after weight on wheels

Aerodynamics 2009/2010 Page 103


Lift augmentation and flow control devices (13)
BOEING TRAILING-EDGE SYSTEM TERMINOLOGY

Aerodynamics 2009/2010 Page 104


Lift augmentation and flow control devices (14)

Aerodynamics 2009/2010 Page 105


Lift augmentation and flow control devices (15)
So what effects do high-lift systems have?
• Flaps shift the CL-α curve
– Greater lift at a given incidence and greater CLmax , but
– Reduction in maximum α
• Slats extend the CL-α curve, increased αmax

Aerodynamics 2009/2010 Page 106


Lift augmentation and flow control devices (16)

Aerodynamics 2009/2010 Page 107


Lift augmentation and flow control devices (17)
• An example of how certification requirements lead to design choices
– e.g. Airbus A340

Aerodynamics 2009/2010 Page 108


Lift augmentation and flow control devices (18)
• An example of how each design choice affects others
– e.g. fuselage length, fuselage shape and undercarriage height

(γ = climb angle)

Aerodynamics 2009/2010 Page 109


Lift augmentation and flow control devices (19)
Boeing 737
• The National High-Lift
Programme (NHLP)
– Instigated by the RAE in the late
1960s in response to a perceived
American lead in high-lift system
design
• Current UK designs (BAC1-11,
Trident, VC-10) had much
simpler leading and trailing-edge
devices than their US
equivalents (707, 727, 737, 747, HS Trident
DC-8, DC-9, DC10)
– It was considered that more
complex meant more powerful
– UK industry had been combined
into BAC and HAS which was
contemplating the next
generation civil transport

Aerodynamics 2009/2010 Page 110


Lift augmentation and flow control devices (20)
• The National High-Lift
Programme (NHLP)
– Lasted through to the late 1970s
and its legacy continues today
– Combinations and permutations of 8
different leading-edge devices and
11 different trailing-edge devices
– Defined the design philosophy for
the Airbus A320 and all subsequent
Airbus aircraft
• Flap chord, slat chord, shroud
length, deflections, etc.
– Identified Reynolds number (scale
effect) and testing fidelity as
important design parameters
• RAE (QinetiQ) 5m Pressurised Wind
Tunnel

Aerodynamics 2009/2010 Page 111


Lift augmentation and flow control devices (21)
• So how do high-lift systems really work?
– A.M.O. Smith (Douglas) illuminated the aerodynamics community in 1972
• The low-speed aerodynamicists equivalent of Darwin’s ‘Origin of the Species’ but nowhere near as easy to read!

– Prior to 1972:
• Fresh momentum through the
slot
• High energy air from lower
surface to upper surface
– A.M.O. Smith
• Leading-Edge Slat Effect
• Circulation Effect
• Dumping Velocity
• Off-Surface Recovery
• Fresh Start for the boundary
layer on each element

Aerodynamics 2009/2010 Page 112


Lift augmentation and flow control devices (22)
• Leading-Edge Slat Effect

– The leading-edge of a downstream


element benefits from the
circulation of an upstream element
– The velocity induced by the
upstream element runs counter to
that of the downstream element
 Reduction of the pressure peak on
the downstream element and a
resilience to high angle of attack
– Works mainly at high angle of
attack when the slat is generating a
lot of lift and where the main
element is highly loaded
– The main element has a similar
effect on the flap Load on the main element is reduced, but
the combination of slat and main element
is positive

Aerodynamics 2009/2010 Page 113


Lift augmentation and flow control devices (23)
• Circulation (Flap) Effect
– The trailing-edge of an upstream element benefits from the circulation of
a downstream element
– The velocity induced
by the downstream
element reinforces
that of the
downstream element

– Increased resistance
to main element
trailing-edge flow
separation and a
resilience to high
angle of attack

Aerodynamics 2009/2010 Page 114


Lift augmentation and flow control devices (24)
• Dumping Effect
– The circulation effect not only improves the performance of the upstream element
– The increased trailing-edge velocity due to the circulation of the downstream
element means that the boundary layer (wake) from the upstream element is
accelerated
 Relieves the pressure rise on the trailing-edge of the downstream element and
improves its ability to resist flow separation

• Off Surface Pressure Recovery


• The boundary layer of the upstream element is
dumped at higher velocity and impinges upon
the boundary layer of the downstream element
• The deceleration of the upstream wake is thus
more efficient than if it were in contact with a
solid boundary

Aerodynamics 2009/2010 Page 115


Lift augmentation and flow control devices (25)
• Fresh Boundary Layer Effect
– Each element starts with a fresh boundary layer at the leading edge
– Thin boundary layer can withstand stronger adverse pressure gradients than thick
ones

• Viscous Effects and Separation


– Confluent boundary layers modify pressure gradients and boundary layer velocity
gradients

Aerodynamics 2009/2010 Page 116


Lift augmentation and flow control devices (26)
• High-Lift Optimisation
• All of the above means that each element has an optimum position relative to its neighbour
• That position will change with the position of further upstream or downstream elements
• In practice, for a 3-element system (slat, main element, flap) the flap position is optimised
using a (e.g. 9 point) optimisation matrix, then the slat is optimised in a similar way. The flap
is then re-optimised with the slat in its new position and ditto for the slat
• The optimisation of a system with a triple-slotted flap is not trivial!

Aerodynamics 2009/2010 Page 117


Lift augmentation and flow control devices (27)

Aerodynamics 2009/2010 Page 118


Lift augmentation and flow control devices (28)
Blown devices (boundary layer control)
• Reduces engine performance
– Requires internal ducting
– Weight
• Volume
• Maintenance

Aerodynamics 2009/2010 Page 119


Lift augmentation and flow control devices (29)
• Flow control
– Which for high lift usually means separation control
• Entraining or redistributing higher momentum air close to the flap surface and delay flow separation

– Vortex generators
• Small blades (rectangular or triangular) that create vortices close to the surface

Aerodynamics 2009/2010 Page 120


Lift augmentation and flow control devices (30)
Flow control – An example of Aerodynamics Research
– Conventional vortex generators can cause a significant drag penalty
– Hence the concept of sub-boundary-layer vortex generators (SBVGs)

2-D separation control using SBVGs

m
m

m
m
a)
NoVBasic
G'
s flow (no 1
0
m
m
w
e
dg
e
s
b) Forwards wedges
a) Basic flow (no b)
x=5Forwards
2hf
ro
msep
ar
ati
on
3
0
0
control) spac
in
g=12h

z
control) wedges

)
)
U
e
a
n
:
-
5
.
m
05
.
0 1
5
.
0 2
5
.
0 3
5
.
0 3
0
0

(
(
2
5
0

2
0
0
2
0
0

1
5
0

1
0
0
1
0
0

5
0

-
3
0
0 -
2
0
0 -
1
0
0 01
0
0 -
4
0
0 -
3
0
0 -
2
0
0 -
1
0
0 01
0
0
x
(
m
m
) x
(
m
m
)

Regions of constant streamwise velocity ‘above’ the bump

m
m

m
m

in vertical plane of symmetry.


1
0
m
m
W
h
ee
l
e
r
We
d
ge d)
d)
10m
Counter-rotating
Counter-rotating
mWhee
ler
W ed
gewi
th1hga
p
c) Joined Counter-
4
0
0 4
0
0
x=52h
fr
omsep
ar
ati
on
c)x
sJoined
=
p
5
a
2
c
i
h
n
g
f
rom
Counter-
s
e
vanes
s
=p
1a
2c
i
hn
g
p
a
=
r
at
i
vanes
o
n
spaced
spaced
12
h apart
apart
rotating vanes
z

z
)

)
3
0
0
rotating vanes 3
0
0by by
1h1h
(

(
2
0
0 2
0
0

1
0
0 1
0
0

-
4
0
0 -
3
0
0 -
2
0
0 -
1
0
0 01
0
0 -
4
0
0 -
3
0
0 -
2
0
0 -
1
0
0 01
0
0
x
(
m
m
) x
(
m
m
)

Aerodynamics 2009/2010 Page 121


Lift augmentation and flow control devices (31)
Extension to 2.75D (sweep and taper)

Aerodynamics 2009/2010 Page 122


Lift augmentation and flow control devices (32)
Extension to fully 3D and to high
Reynolds number

340 flap flow separation (AWIATOR)

No VGs

A380 in QinetiQ 5m Wind Tunnel

VGs on

Aerodynamics 2009/2010 Page 123


Lift augmentation and flow control devices (33)
Flow control
• Alternative flow control devices
– Air jet, Synthetic jets
• Thousands of devices
– Networked, Sequenced
– Rapid response
• Manufacturing
– Embedding in composite/metal structures
– Power supply
– Calibration
• Maintenance
– Robustness
• Certification
– Consequences of failure
• Flow control
• Loads control
– Performance degradation

Aerodynamics 2009/2010 Page 124


Lift augmentation and flow control devices (34)
Flow control
• Also used for as separation
control on:
– Pylon/slat junction
• Douglas invention
• Big effect on CLmax

Aerodynamics 2009/2010 Page 125


Lift augmentation and flow control devices (35)
Flow control
• Pylon/slat junction

Aerodynamics 2009/2010 Page 126


Lift augmentation and flow control devices (36)
Flow control
• Also used for as separation control on:
– Flight control devices

Boeing 727

Gloster Javelin

Aerodynamics 2009/2010 Page 127


Lift augmentation and flow control devices (37)
Flow control
• Also used for as separation control on:
– Afterbody

Rockwell B-1 Lancer

Aerodynamics 2009/2010 Page 128


Lift augmentation and flow control devices (38)
Flow control
• Also used for as separation control on:
– Shock/boundary layer interaction

Boeing 737

Aerodynamics 2009/2010 Page 129


Lift augmentation and flow control devices (39)
• Factors affecting high-lift
performance
– Surface roughness
– Ice
– Liquids (rain & de-icing fluid)

Aerodynamics 2009/2010 Page 130


Lift augmentation and flow control devices (40)
• Abnormal use of high-lift devices
– British Airways Boeing 777 crash at Heathrow
(January 2008)

– Glide approach due to loss of power


– Correct selection of take-off setting
– Maximum L/D

Aerodynamics 2009/2010 Page 131


Lift augmentation and flow control devices (41)

Aerodynamics 2009/2010 Page 132


Supersonic Flows
• Critical Mach Number; formation of shockwaves; Normal and oblique
shockwaves; Effect of wing thickness and camber; Wave drag and methods of
reducing wave drag (Wing Sweep, Transonic Area Ruling, Supercritical Aerofoil
design, Wing design); Shockwave control and the Shock-induced separation.

Aerodynamics 2009/2010 Page 133


Supersonic flows (1)
• Incompressible flow
– Most of our analysis so far has assumed that the flow is incompressible

– We do this because it allows us to simplify things in two important ways


• The density is known and can be treated as a constant in, for example, the
continuity equation
ρ 1u1A1 = ρ 2u2A2

• The interaction between mechanical and thermal energy is weak which permits
use of a simplified version of the energy equation
– This assumption does not, however, match reality
• All fluids are compressible, even liquids
– If the pressure changes in a flow are sufficient to cause significant density
changes we have to abandon the incompressible flow assumption
– It is more likely to be of concern in a gas than in a liquid
• A pressure change of 500kPa (~72psi) causes a density change of 0.024% in
water but 250% in air

Aerodynamics 2009/2010 Page 134


Supersonic flows (2)
• Forward influence
– In subsonic flow pressure changes are
propagated through the fluid at the speed
of sound through pressure waves
– The pressure changes caused by a body
moving through a fluid are thus
transmitted through the fluid
– The air ahead of a subsonic aircraft, for
example, therefore ‘knows’ that it is
coming because the pressure changes
are transmitted forward
– The degree to which the flow ahead of
body is altered depends greatly on the
pressure changes on the body itself and
these are dependent on its shape and
speed
– This effect is known as forward
influence and can be a very important
consideration

Aerodynamics 2009/2010 Page 135


Supersonic flows (3)
• Compressible flow
– As an aircraft approaches the
speed of sound, the difference
between the speed at which the
pressure waves are transmitted
ahead of it and that of the aircraft
reduces
– So the time between the
pressure wave and aircraft
passing through the same point
reduces
– When the aircraft reaches the
speed of sound (Mach 1) the air
receives no ‘warning’ and so has
to react instantly to the presence
of the aircraft
– This instantaneous change can
lead to shock waves

Aerodynamics 2009/2010 Page 136


Supersonic flows (4)
• Compressible flow
– An aircraft does not have to be travelling at the speed of sound to
generate shock waves
– The wings are designed to accelerate the air to produce lift so shock
waves will form at relatively low Mach number
– Shock waves can be a
problem even at landing
speed
• the acceleration generated
by a slat can be so severe
as to cause near sonic flow
on its upper surface as is a
design limit
• CL for Cp-10

Aerodynamics 2009/2010 Page 137


Supersonic flows (5)
• Critical Mach number
– The speed at which sonic flow occurs on a body is called the critical
Mach number
– Shock waves cause drag (wave drag) (see later) and so the speed at
which they start to form is important
– For example, the economy of a civil aircraft reduces dramatically once
strong shock waves form its wings and the point at which this occurs is
the drag rise Mach number
• the wing section must be designed so that the design cruise speed is below
the drag rise Mach number

Aerodynamics 2009/2010 Page 138


Supersonic flows (6)
• Mach wave
– If we imagine a stationary disturbance emitting pressure waves (e.g.
sound) the waves will propagate uniformly from the source
• Ripples in the surface of water
– The waves will travel at the speed of sound (c) so that after a time interval
(Δt) the waves will have travelled a distance of cΔt

Aerodynamics 2009/2010 Page 139


Supersonic flows (7)
• Mach wave
– Now suppose that there is a flow over the disturbance travelling at Mach 0.5, i.e. V = ½c
– In addition to spreading into the fluid, the waves will also be swept downstream by the flow
– As a result the waves bunch up on the upstream side and spread out on the downstream side and the rate at
which the fluid experiences the disturbances is greater on the upstream side than the downstream
• If the disturbance was sound an upstream listener would here a different frequency to one downstream
 Doppler effect

Aerodynamics 2009/2010 Page 140


Supersonic flows (8)
• Mach wave
– Now suppose that the flow past the disturbance is sonic, i.e. at a speed exactly
equal to the speed of sound (M =1, V = c)
– In this case the waves are swept downstream at exactly the same speed at which
they spread
– The waves cannot propagate upstream and the fluid is not affected by the
disturbance until it arrives at it

– The upstream waves will sit on


top of each other forming an
envelope
– The fluid experiences the total
effect of all of the waves at once
upon crossing the envelope

Aerodynamics 2009/2010 Page 141


Supersonic flows (9)
• Mach wave
– Finally suppose that the fluid flows past the disturbance at supersonic speed, i.e. a speed greater than the
speed of sound (M > 1, V > c)
– At supersonic speed the disturbances are swept downstream faster than they can spread
– All the waves are confined to a triangular (in 2-D) or conical (in 3-D) region extending downstream from
the disturbance
– The waves form an envelope of half angle μ

– Only the fluid inside the


envelope is affected by the
disturbance
• If it were sound it would be
silent outside the envelope
– The envelope is called the Mach
wave (in 2-D) or the Mach cone
(in 3-D)

Aerodynamics 2009/2010 Page 142


Supersonic flows (10)
• Mach angle
– The half angle of the envelope can be calculated
c ∆t c 1
sin µ = = =
V ∆t V M

– So the Mach angle (μ) is given by:

1
µ = sin-1
M

– Because the fluid experiences


the combined effects of a
disturbance almost
instantaneously, the fluid
property and velocity variations
may be discontinuous
– E.g. density changes
• Schlieren technique

Aerodynamics 2009/2010 Page 143


Supersonic flows (11)
• Shock waves
– A finite strength disturbance (e.g. a sharp wedge or cone) in supersonic
flow creates a finite strength wave which is stationary with respect to the
disturbance
– These finite strength disturbances are known as shock waves
– They are extremely thin and fluid properties change dramatically across
them

Aerodynamics 2009/2010 Page 144


Supersonic flows (12)

Shock wave on an A320 in cruise


Super-critical wing design (see later)
Aerodynamics 2009/2010 Page 145
Supersonic flows (13)
• Normal and oblique shock waves
– A shock wave that is perpendicular to the upstream flow is a normal
shock
– One that is inclined at a constant angle to the upstream flow is an oblique
shock
– A curved shock has a varying angle between it and the upstream flow

Aerodynamics 2009/2010 Page 146


Supersonic flows (14)
Normal shock waves
• Consider a normal shock wave of zero thickness as shown in the picture
– The continuity equation along a streamline is ρ 1u1A1 = ρ u A2
2 2

– For a section of the shock wave, the area before and after the shock will be equal, i.e. A = A = A, so
1 2

– ρ 1u1 = ρ 2u2
The ideal gas law is

And we can write the velocity as

So ρ 1u1 = ρ 2u2 becomes


p
ρ=
→ or RT

u = Mc = M γ R T

p1 p2
M 1 γ R T1 = M 2 γ R T2
R T1 R T2

p1M 1 p2 M 2 M 2 p1 T2
= =
T1 T2 M 1 p2 T1
Aerodynamics 2009/2010 Page 147
Supersonic flows (15)
Normal shock waves
• Now consider the forces acting across the shock wave
– The momentum equation gives

The force is equal to the pressure difference across the shock wave
– = pressure x area = p A = p A
1 2

ΣFx = ρ 2u22 A − ρ1u12 A


Therefore

or

and

p1 A − p2 A = ρ 2u22 A − ρ1u12 A

p1 − p2 = ρ 2u22 − ρ1u12

p1 (1 + γ M 12 ) = p2 (1 + γ M 22 )
p2 (1 + γ M 12 )
= u = Mc = M γ R T
p1 (1 + γ M 22 ) ρ=
p
RT

Aerodynamics 2009/2010 Page 148


Supersonic flows (16)
• Normal shock waves
– We also need to make use of the energy equation
– The state of a gas is defined by several properties including the temperature, pressure,
and the volume which the gas occupies.
– From the first law of thermodynamics (conservation of energy) we find that the internal
energy of a gas is also a state variable
• That is, a variable which depends only on the state of the gas and not on any process that
produced that state
– We are free to define additional state variables which are combinations of existing
state variables

Aerodynamics 2009/2010 Page 149


Supersonic flows (17)
• Normal shock waves
– The new variables often make the analysis of a system much simpler
– For a gas, a useful additional state variable is the enthalpy (H) which is defined to be the
sum of the internal energy E plus the product of the pressure p and volume V, i.e. H = E
+ pV
– The enthalpy can be made into a specific variable ( ) by dividing
~ by the mass
– Propulsion engineers use the specific enthalpy (or more oftenhthe change in specific
enthalpy) in engine analysis more than the enthalpy itself

Aerodynamics 2009/2010 Page 150


Supersonic flows (18)
• Normal shock waves
– How do we use this new variable called enthalpy?
– Let's consider the first law of thermodynamics for a gas
– For a system with heat transfer Q and work W, the change in internal energy E from State 1 to State 2 is equal to the difference in the
heat transfer into the system and the work done by the system:

E2 - E1 = Q - W

– The work and heat transfer depend on the process used to change the state.
– For the special case of a constant pressure process, the work done by the gas is given as the constant pressure p times the change in
volume V. i.e.

W = p (V2 - V1 )

– Substituting into the first equation, we have:


E2 - E1 = Q - p (V2 - V1 )

Aerodynamics 2009/2010 Page 151


Supersonic flows (19)
• Normal shock waves
– Let's group the conditions at State 2 and the conditions at State 1 together:

E2 - E1 = Q – p (V2 - V1 )
becomes

(E2 + p V2) - (E1 + p V1) = Q

– The (E + pV) can be replaced by the enthalpy H

H2 - H1 = Q
– From the definition of the heat transfer, we can represent Q by some heat
capacity coefficient Cp times the temperature T

H2 - H1 = Cp (T2 - T1)

Aerodynamics 2009/2010 Page 152


Supersonic flows (19)
• Normal shock waves
– We have previously divided by the mass of gas to produce the specific
enthalpy equation version

~ ~
h2 − h1 = c p ( T2 − T1 ) ( ~
h2 = c p T2 )
– The specific heat capacity (cp) is called the specific heat at constant
pressure
– This final equation is used to determine values of specific enthalpy for a
given temperature
– Across shock waves, the total enthalpy of the gas remains a constant

Aerodynamics 2009/2010 Page 153


Supersonic flows (20)
• Normal shock waves
– The energy equation for this situation is:

   ~ ~ u 2
u 2

Q − Ws = m  h2 − h1 + − 
2 1

 2 2
where Q is the heat transfer rate,

W s is power (rate of work),


m is mass flow rate,
~
and h is specific enthalpy.

– Work is the energy transfer by the action of a force through a distance


– Because we have defined the thickness of the shock wave to be zero, there
can be neither heat transfer or work as they require finite volumes
– So the equation reduces to:

~ u12 ~ u22
h1 + = h2 +
2 2
Aerodynamics 2009/2010 Page 154
Supersonic flows (21)
• Normal shock waves
– So we now have five equations:
Continuity ρ 1u 1 = ρ 2u 2 (1)

Momentum
p1 − p2 = ρ 2u 22 − ρ1u12 (2)

~ u12 ~ u22
Energy h1 + = h2 + (3)
2 2
~
Enthalpy h2 = c pT2 (4)

Equation of state p2 = ρ 2 R T (5)

– And five unknowns, the flow conditions after the shock wave:
~
ρ 2 , u2 , p2 , h,2and T2

– These equations then are sufficient to calculate these unknown conditions


in an ideal gas
Aerodynamics 2009/2010 Page 155
Supersonic flows (22)
• Normal shock waves
– Using the continuity and momentum equations
p1 p
+ u1 = 2 + u2
ρ1u1 ρ 2u 2
p1 p
− 2 = u 2 − u1
ρ1u1 ρ 2u2
a12 a22
− = u 2 − u1 a= γ p ρ
γ u1 γ u2
u
– We introduce a characteristic Mach number M∗ =
a∗
where a* is the value of the speed of sound at sonic conditions, not the

actual local value, and a ∗ = γ RT ∗

Aerodynamics 2009/2010 Page 156


Supersonic flows (23)
• Normal shock waves
– The energy equation we had as
~ u12 ~ u22
h1 + = h2 +
2 2
where u1 and u2 are velocities at any two points along a 3-D streamline
~
– We had that for a perfect gas, h = c pT , so
u12 u 22 cp - specific heat at constant pressure
c pT1 + = c pT2 +
2 2 cv - specific heat at constant volume

– Also for a perfect gas, cp – cv = R


– Which we can modify by dividing through by cp to give
Rcp cp 1 R
1− = and γ = so 1 − =
cp cp cv γ cp
γR
cp =
or γ −1

Aerodynamics 2009/2010 Page 157


Supersonic flows (24)
• Normal shock waves

u12 u22 γ R T1 u12 γ R T2 u22


– So c pT1 + = c pT2 + becomes + = +
2 2 γ −1 2 γ −1 2

remember a = γ RT

so a12 u12 a22 u 22


+ = +
γ −1 2 γ −1 2
– If we make the Point 2 on the streamline
represent sonic flow, then u = a* so that
2 2
a2 u2 a* a*
+ = +
γ −1 2 γ −1 2
or
a2 u2 γ + 1 *2
+ = a
γ −1 2 2(γ − 1)
Aerodynamics 2009/2010 Page 158
Supersonic flows (25)
• Normal shock waves
a2 u2 γ + 1 *2
– Rearranging the equation + = a and applying it
γ −1 2 2(γ − 1)
first ahead of the shock wave and then behind it, we get
γ + 1 *2 γ − 1 2 γ + 1 *2 γ − 1 2
a12 = a − u1 and a22 = a − u2
2 2 2 2
a* is the same constant value because the flow is adiabatic, i.e. one in
which no heat is added or removed from the system

a12 a22
– Substituting this pair into − = u2 − u1
γ u1 γ u2

2 2
– Gives γ + 1 a* γ − 1 γ + 1 a* γ − 1
− u1 − − u2 = u2 − u1
2 γ u1 2γ 2 γ u2 2γ

Aerodynamics 2009/2010 Page 159


Supersonic flows (26)
• Normal shock waves
*2 *2
– Rearranging γ + 1 a −
γ −1
u1 −
γ +1 a

γ −1
u2 = u2 − u1
2 γ u1 2γ 2 γ u2 2γ
γ +1 2 γ −1
– Gives (u2 −u1 ) a * + (u2 −u1 ) = u2 − u1
2γ u1 u 2 2γ

– Dividing by u2 - u1 gives

γ + 1 *2 γ − 1
a + = u2 − u1
2γ u1 u 2 2γ

– Which can be rearranged and solved for a* to give

a* = u1u2

– This is called the Prandtl relation and is a useful intermediate relation for
shock waves
Aerodynamics 2009/2010 Page 160
Supersonic flows (27)
• Normal shock waves
– The usefulness of the Prandtl relation is shown if we recall the equation
a2 u2 γ + 1 *2
+ = a
γ −1 2 2(γ − 1)
– Dividing through by u2 gives
2
(a / u ) 12
γ + 1  a* 
+ =  
γ −1 2 2(γ − 1)  u 
– And converting to Mach number
2
(1 / M ) 2 γ +1  1  1
=  * −
γ −1 2(γ − 1)  M  2
– And rearranging gives

2 *2 (γ + 1) M 2
M 2
= or M =
*2
((γ + 1) / M ) − (γ − 1) 2 + (γ − 1)M 2
Aerodynamics 2009/2010 Page 161
Supersonic flows (28)
• Normal shock waves
– We now take the Prandtl relation and incorporate the characteristic Mach
number (M* = u/a*)
u1 u2 1
1= ∗ ∗ 1= M M *
1
*
2 M = *
*
2
a* = u1u2 becomes →
a a or M1
– On the previous page we derived the equation
2 (γ + 1) M
2
M* =
2 + (γ − 1)M 2
1
M = *
*
2
– Substituting this into M1
−1
(γ + 1) M 22  (γ + 1) M 12 
= 
– Gives 2 + (γ − 1) M 2  2 + (γ − 1) M 12 
2

1 + [ (γ − 1) / 2] M 12
– And solving for M22 M =
2
2
γ M 12 − (γ − 1) / 2
Aerodynamics 2009/2010 Page 162
Supersonic flows (29)
• Normal shock waves

1 + [ (γ − 1) / 2 ] M 2

– The equation M 2is=an important


2 1
result
γ M 1 − (γ − 1) / 2
2

– It shows that the Mach number after a normal shock wave is dependent only
upon the Mach number before it
– If M1=1, then M2=1 and this is an infinitely weak shock wave, or Mach wave
– If M1>1, then M2<1, i.e. the flow after the shock wave will be subsonic
– As M1 increases above 1 the shock wave becomes progressively stronger
and M2 becomes progressively less than 1
– As M1 → ∞ , M2 approaches a finite minimum value
M 2 → (γ − 1) / 2γ

which for air (γ = 1.4) is 0.378

Aerodynamics 2009/2010 Page 163


Supersonic flows (30)
• Normal shock waves
– Now we have a way of determining the relationship between the Mach
numbers before and after a normal shock wave
– We also need to determine the relationships between the other flow
parameters ρ 2 /ρ 1, p2 /p1, and T2 /T1
– Using the continuity equation ( ρ 1 u1 = ρ 2 u2) and the Prandtl relation
(a* = u1u2) we get
ρ 2 u1 u12 u 2
= = = 22 = M 1*
ρ1 u2 u1u2 a *

*2 (γ + 1) M 2
– Substituting M = into the above equation
2 + (γ − 1)M 2

– Gives ρ 2 u1 (γ + 1) M 12
= =
ρ1 u2 2 + (γ − 1)M 12

Aerodynamics 2009/2010 Page 164


Supersonic flows (31)
• Normal shock waves
– To obtain the pressure ratio we combine the continuity equation with the
momentum equation
ρ 1u1 = ρ 2u2 and p1 − p2 = ρ 2u2 − ρ1u1
2 2

– To give
 u 
p2 − p1 = ρ1u12 − ρ 2u22 = ρ1u1 (u1 − u2 ) = ρ1u12 1 − 2 
 u1 
– Dividing by p1 and recalling that a1 = γ p1 ρ1 or a12 = γ p1 ρ1

p2 − p1 γ ρ1u12  u2  γ u12  u2   u 
= 1 −  = 2 1 −  = γ M 12 1 − 2 
p1 γ p1  u1  a1  u1   u1 

– For u2/u1 in this equation we can substitute


ρ 2 u1 (γ + 1) M 12
= =
ρ1 u2 2 + (γ − 1)M 12

Aerodynamics 2009/2010 Page 165


Supersonic flows (32)
• Normal shock waves
p2 − p1  2 + (γ − 1)M 12 
– The substitution gives = γ M 12 1 − 
p1  (γ + 1) M 1
2

– Which simplifies to
p2 2γ
= 1+ ( M 12 − 1)
p1 γ +1

– To get the temperature ratio we use the gas equation p = ρ R T, i.e.

T2  p2  ρ1 
=   
T1  p1  ρ 2 
– Which gives

T2 h2  2γ  2 + (γ − 1)M 12
= = 1 + ( M 1 − 1)
2

T1 h1  γ + 1  (γ + 1) M 1
2

Aerodynamics 2009/2010 Page 166


Supersonic flows (33)
• Normal shock waves
– So now we have

1 + [ (γ − 1) / 2] M 12
M =
2
2
γ M 12 − (γ − 1) / 2

ρ 2 u1 (γ + 1) M 12
= =
ρ1 u2 2 + (γ − 1)M 12

p2 2γ
= 1+ ( M 12 − 1)
p1 γ +1

T2 h2  2γ  2 + (γ − 1)M 12
= = 1 + ( M 1 − 1)
2

T1 h1  γ + 1  (γ + 1) M 1
2

Aerodynamics 2009/2010 Page 167


Supersonic flows (34)
• Normal shock waves
– Note that all these relationships are in terms of upstream Mach number
(M1) only
– M1 is the determining parameter for changes across a normal shock wave
in a perfect gas
– This is a good example of the power of the Mach number as a governing
factor in compressible flow
– As before, at M=1 p1=p2, ρ 1=ρ 2, and T1=T2 and we have a normal shock
wave of vanishing strength; a Mach wave
– As M1 increases above 1, p2, ρ 2 and T2 all progressively increase

Aerodynamics 2009/2010 Page 168


Supersonic flows (35)
• Normal shock waves
– In the limiting case, where M1 → ∞

γ −1 p2
lim M 2 = = 0.378 lim =∞
M 1 →∞ 2γ M 1 →∞ p1
ρ2 γ + 1 p2
lim = =6 lim =∞
M 1 →∞ ρ1 γ − 1 M 1 →∞ p1

– So as the upstream Mach


number increases towards
infinity
• the downstream Mach number
decreases to a finite value
• density increases to a finite
number
• but temperature and pressure
can increase without bound

Aerodynamics 2009/2010 Page 169


Supersonic flows (36)
• Measurement of velocity in compressible flow
– In low-speed, incompressible flow, the velocity can be measured using a
Pitot-static tube
• The total pressure is measured by the Pitot tube and the static pressure from a static
pressure orifice
• Bernoulli gives the dynamic pressure as the difference between the two

Aerodynamics 2009/2010 Page 170


Supersonic flows (37)
• Measurement of velocity in compressible flow
– The same is true in high-speed, compressible flow if we use Mach number
instead of velocity, although the formulae are different for each Mach-
number regime
• Subsonic compressible
• Supersonic compressible

– For Region 1 the isentropic flow


relationships hold
γ (γ −1)
p0,1 γ −1 2 
= 1 + M1 
p1  2 
– (see slides ‘A’ at the end of the section
for derivation if desired)
– Solving for M12 gives

2   p 
(γ −1) γ

M1 =
2
 0 ,1
 − 1
γ − 1  p1  

Aerodynamics 2009/2010 Page 171
Supersonic flows (38)
• Measurement of velocity in compressible flow
– Using the relationship for Mach number


2  0,1 
p
( γ −1) γ

M1 =
2
  − 1
γ − 1  p1  

– becomes

2  
( γ −1) γ
2 a  p 
u1 =
2 1
 0 ,1
 − 1
γ − 1  p1  

– So, unlike incompressible flow, a


knowledge of the total and static
pressures is not enough
– We also need to know the freestream
speed of sound, a1

Aerodynamics 2009/2010 Page 172


Supersonic flows (39)
• Measurement of velocity in compressible flow
– Now in supersonic flow the Pitot tube creates a stagnation region and the
flow is brought to rest at the throat
– However, because the upstream flow is supersonic and the Pitot tube is an
obstruction, there will be a bow wave ahead of it
– The centreline streamline crosses the
normal portion of the bow shock
– So the flow is decelerated to subsonic
speed (non-isentropically) through the
shock wave
– And then (isentropically) to zero velocity
at the throat
– The total pressure measured by the
Pitot tube is (p0,1) but of the flow behind
a normal shock wave (p0,2)

Aerodynamics 2009/2010 Page 173


Supersonic flows (40)
• Measurement of velocity in compressible flow
– However, knowing the freestream static pressure (p1) and the throat total
pressure (p0,2) is still enough to calculate the freestream Mach number (M)
p0 , 2 p0 , 2 p 2
=
p1 p2 p1
– Where p0,2/p2 is the ratio of total and
static pressure after the shock and p2/p1
is the static pressure ratio across the
shock
γ ( γ −1)
p0 , 2  γ − 1 2 
= 1 + M2 
p2  2 
1 + [(γ − 1) / 2]M 12
M2 =
2

γ M 12 − [(γ − 1) / 2]
p2 2
= 1+ ( M 12 − 1)
p1 γ +1

Aerodynamics 2009/2010 Page 174


Supersonic flows (41)
• Measurement of velocity in compressible flow
– Substituting rearranging and simplifying gives
γ (γ −1)
p0 , 2 (γ + 1) M2 2
 1 − γ + 2γ M 12
=  1

p1  4γ M 1 − 2(γ − 1) 
2
γ +1

– This is called the Rayleigh Pitot tube formula


– It relates the Pitot pressure measured by the
tube (p0,2) and the freestream static pressure
(p1) to the freestream Mach number (M1)

Aerodynamics 2009/2010 Page 175


Supersonic flows (42)
• Oblique shock waves
– Most shock waves form an oblique angle with the upstream flow
– Normal shock waves are just a special case of oblique shock wave where
the angle is 90°
– In addition to oblique compression waves where the pressure increases
discontinuously across the shock wave, there are expansion waves
where the pressure decreases continuously across the shock wave

Aerodynamics 2009/2010 Page 176


Supersonic flows (43)
• Oblique shock waves
– Consider supersonic flow encountering a concave corner
– The wall is turned upwards at an angle θ
– An oblique shock wave will form at the corner where the streamlines before
and after the shock are all parallel, deflected through the angle θ
– The Mach number suddenly
(discontinuously) decreases while
the pressure, density and
temperature all suddenly increase

Aerodynamics 2009/2010 Page 177


Supersonic flows (44)
• Oblique shock waves
– Now consider the same supersonic flow encountering a convex corner
– The wall is turned downwards at an angle θ
– A series oblique shock waves form an expansion fan will form at the
corner
– The fan opens continuously away from the corner
– Again the streamlines before and after are all parallel, deflected
continuously and smoothly through the expansion angle θ
– The Mach number smoothly
(continuously) increases while the
pressure, density and temperature
all smoothly decrease
– In contrast to essentially 1-D normal
shock waves, oblique shock and
expansion waves are inherently 2-D
• i.e. the flow field properties are a
function of x and y

Aerodynamics 2009/2010 Page 178


Supersonic flows (45)
• Oblique shock waves
– If we return to the oblique compression
wave, this is representative of supersonic
flow past a wedge
– The wedge semi-angle is now θ and the
angle of the oblique shock wave is β
– The relationship between the shock
angle (β), the wedge angle (θ) and the
upstream Mach number (M1) is given by
 1  γ −1 2 2 1 
θ = β − tan  −1
 sin β + 
2 
 sinβ cosβ  γ +1 γ + 1 M 1 
– (See slides ‘B’ at the end of the section for derivation if desired)
– This is the θ-β-M relation and it specifies θ as a unique function of M1
and β
– The results from the equation are plotted graphically on the next slide
and the graph is used to solve oblique shock problems
Aerodynamics 2009/2010 Page 179
Supersonic flows (46)
Oblique shock
waves
The wedge angle
(θ) plotted against
shock angle ( β)
for varying
upstream Mach
number (M1)

Aerodynamics 2009/2010 Page 180


Supersonic flows (47)
• Oblique shock waves
– The graph illustrates a lot of physical
phenomena associated with oblique
shock waves:
– For any given upstream Mach number
M1, there is a maximum deflection angle,
θmax
– If the physical geometry is such that
θ > θmax then no solution exists for a
straight oblique shock wave
– Nature establishes a curved shock wave
detached from the corner or the nose of
a body
– Note that as the freestream Mach
number increases, θmax also increases
• straight oblique shock waves can exist at
higher deflection angles at higher speed
– But there is a limit (for gamma = 1.4)
• θmax → 45.5° as M1 → ∞
Aerodynamics 2009/2010 Page 181
Supersonic flows (48)
• Oblique shock waves
– For any given θ less than θmax there are two
straight oblique shock solutions for a given
upstream Mach number
– E.g. if M1=2.0 and θ =15° then β can equal
either 45.3 or 79.8°
– The smaller value is called the weak shock
solution and the larger value is the strong
shock solution
– The terms “weak” and “strong” derive from
the fact that the for a given upstream Mach
number (M1), the larger the wave angle the
larger the normal component of upstream
Mach number (Mn,1) and thus the larger the
pressure ratio p2/p1
– That is, the higher-angle shock wave will
compress the air more than the lower-angle
shock wave
– In nature the weak shock solution usually
prevails
Aerodynamics 2009/2010 Page 182
Supersonic flows (49)
• Oblique shock waves
– Whenever you see straight, attached oblique
shock waves (as shown in the bottom picture)
they are almost always the weak shock
solution
– It is safe to make this assumption unless you
have information to the contrary
– Note that the locus of points connecting all the
values of θmax divides the weak and strong
shock solutions
• Above the curve the strong shock prevails
• Below the curve the weak shock prevails
– There is another curve just below this one
– This is the dividing line above which the
downstream Mach number is subsonic (M2 <
1) and below which it is supersonic (M2 > 1)
– For the strong solution the downstream Mach
number is always subsonic
– For the majority of weak shock solutions the
downstream Mach number is supersonic
Aerodynamics 2009/2010 Page 183
Supersonic flows (50)
• Use of oblique shock waves
– Engine inlets

Aerodynamics 2009/2010 Page 184


Supersonic flows (51)
• Critical Mach number
– We saw before that the
freestream Mach number at
which the flow on an aerofoil
first becomes sonic is the
critical Mach number, Mcr
– If we define the static pressure
in the freestream as p∞ and
that at a point A on an aerofoil
as pA, we can use the
isentropic pressure ratio (slide
169) to give us

γ ( γ −1)
p A p A p0  1 + [(γ − 1) / 2]M ∞2 
= =  
2 
p∞ p∞ p0  1 + [(γ − 1) / 2]M A 

Aerodynamics 2009/2010 Page 185


Supersonic flows (52)
• Critical Mach number
– Recall that pressure coefficient is given by

p − p∞ where q∞ = 12 ρ ∞V∞2
Cp =
q∞

1 1 γ p∞
q∞ = ρ ∞V∞2 = ρ ∞V∞2
2 2 γ p∞
γ  ρ∞  2
= p∞  V∞
2  γ p∞ 
γ
= p∞ M ∞2 (a∞2 = γ p∞ ρ ∞ )
2
– So
2  p 
Cp =  − 1
γ M ∞2  p∞ 

Aerodynamics 2009/2010 Page 186


Supersonic flows (53)
• Critical pressure coefficient
– So the pressure coefficient at point A is
given by
2  1 + [(γ − 1) / 2]M 2 
γ (γ −1)

C p, A =  ∞

2 
− 1
γ M ∞2  1 + [(γ − 1) / 2]M A  
– This is the compressible equivalent of
the Bernoulli equation, relating local
pressure to the local Mach number
– The critical pressure coefficient (Cp,cr)
is the pressure coefficient at the point
where the flow on the aerofoil first
becomes sonic, i.e. MA=1

2  2 γ
1 + [(γ − 1) / 2]M ∞ 
( γ −1)

C p ,cr = 
2 
 − 1
γ M ∞  1 + [(γ − 1) / 2]  

– This equation allows us to calculate the
pressure coefficient at any point where
the local Mach number is 1 (i.e. along
the sonic line)

Aerodynamics 2009/2010 Page 187


Supersonic flows (54)
• Critical pressure coefficient
– When the freestream Mach number is
precisely equal to the critical Mach number,
there is only one point on the aerofoil where
M=1, namely point A
– In this case M∞ = Mcr and

2  1 + [(γ − 1) / 2]M cr2 


γ (γ −1)

C p ,cr = 2 
  − 1
γ M cr  1 + [(γ − 1) / 2]  

– This equation has no connection with aerofoil
shape and is thus a universal relationship
which can be used for all aerofoils
– The Prandtl-Glauert rule relates the
incompressible pressure coefficient (Cp,0) to a
compressible one:
C p,0
Cp =
1 − M ∞2
– (Other approximations exist, but this is the
simplest and most common)
Aerodynamics 2009/2010 Page 188
Supersonic flows (55)
• Critical pressure coefficient
– To estimate the critical Mach number we need
to:
• By some means (either experimental or
theoretical) obtain the low-speed,
incompressible value of Cp,0 at the minimum
pressure point on the aerofoil
• Using a compressibility correction (e.g. Prandtl-
Glauert) plot the variation of Cp with M∞ (curve
B)
• The point where curve B crosses the line
representing

2  1 + [(γ − 1) / 2]M cr2 


γ (γ −1)

C p ,cr = 2 
  − 1
γ M cr  1 + [(γ − 1) / 2]  

• Is the point where sonic flow occurs at the
minimum pressure location on the aerofoil
• The value of M∞ at this intersection is thus the
critical Mach number Mcr

Aerodynamics 2009/2010 Page 189


Supersonic flows (56)
• Critical pressure coefficient
– The graph is not an exact determination of Mcr
• The curve for Cp.cr is exact, but curve B is only an approximation
– Hence the value of Mcr obtained is only approximate
– However, such an estimate is useful for preliminary design and the results
are accurate enough for most applications

Aerodynamics 2009/2010 Page 190


Supersonic flows (57)
• Effect of thickness
– Thicker aerofoils perturb the airflow more and create greater suction on
the top surface than thinner aerofoils
– That is, on a thick aerofoil the value of the pressure coefficient at the
minimum pressure location will be a larger negative number than the
equivalent value on a thin aerofoil

– Plotting this results


shows immediately that
a thick aerofoil has a
lower critical Mach
number than a thin one
– For high-speed aircraft it
is desirable to have a
high value of Mcr and this
drives the designer
towards a thinner wing

Aerodynamics 2009/2010 Page 191


Supersonic flows (58)
• Effect of thickness
– For example a Lear Jet has a 9% thick aerofoil, while the Piper Aztec has
one that is 14% thick

Aerodynamics 2009/2010 Page 192


Supersonic flows (59)
• Drag-divergence Mach number
– As we increase the freestream Mach number (M∞), from a to b, for a
given aerofoil the drag remains virtually constant
– We then encounter the critical Mach number where the flow on the
aerofoil first becomes sonic, point c
– As we increase M∞ to slightly above Mcr (to point d) a finite region of
supersonic flow appears on the aerofoil
– As we nudge M∞ still
higher we encounter
point e where the drag
suddenly starts to
increase.
– The value of M∞ where
this sudden increase in
drag starts is called the
drag-divergence Mach
number

Douglas dCd/dM > 0.1


Boeing ΔCd = 0.002
Aerodynamics 2009/2010 Page 193
Supersonic flows (60)
• Drag-divergence Mach number
– Beyond the drag-divergence Mach number the drag coefficient can be
very large, typically increasing by a factor of 10 or more
– This drag increase is associated with an extensive region of supersonic
flow over the aerofoil terminating in a shock wave
– For an aerofoil design for low-speed application, the local Mach number
can reach 1.2 or higher and the terminating shock can be very strong
– These shocks generally cause severe flow separation, with an attendant
increase in drag (wave drag)

Aerodynamics 2009/2010 Page 194


Supersonic flows (61)
• Drag rise reduction
– Research since 1945 has focused on reducing the large drag rise
– Instead of a factor of 10 increase in drag at Mach 1, can reduce it to 2
or 3?
– Several design ploys have been utilised to achieve this
– The first was the use of thin aerofoils
– We have already seen that thinner aerofoils have higher critical Mach
numbers than thicker ones

Aerodynamics 2009/2010 Page 195


Supersonic flows (62)
• Drag rise reduction

Variation of thickness-to-chord ratio for a representative selection of different aircraft


Aerodynamics 2009/2010 Page 196
Supersonic flows (63)
• Drag rise reduction
– The second design ploy was to use
swept wings
– Imagine a straight wing with a
thickness-to-chord ratio (t/c) of 0.15
– If we sweep the same wing at 45°
the flow sees the same physical
thickness but the chord has
extended
c
c2 = = 1.41c
cosΩ

and the thickness-to-chord ratio has


reduced
– Thus by sweeping the wing the flow
behaves as if the aerofoil is thinner
and it has a higher critical Mach
number

North American F-86 Sabre

Aerodynamics 2009/2010 Page 197


Supersonic flows (64)
• Area rule
– Beside using thin, swept
aerofoils to reduce the drag
rise at Mach 1, two other
concepts have been
developed
– The first of these is the area
rule
– The early jets did not have
enough thrust to overcome
the massive drag rise near
Mach 1
– Even the early “century”
series aircraft designed to
provide the UASF with
supersonic fighters in the
early 1950s (e.g. the
Convair F-102 Delta
Dagger) could not at first
penetrate the sound barrier
in level flight
Aerodynamics 2009/2010 Page 198
Supersonic flows (65)
• Area rule
– The picture shows the area distribution (the variation of cross-sectional
area with distance along the aircraft axis) of a typical US aircraft of that
period
• Note the discontinuities in the distribution
– Ballisticians had known for almost a century that bullets and shells with
smoothly varying cross-sections were faster than those with those with
abrupt or discontinuous shape changes

Aerodynamics 2009/2010 Page 199


Supersonic flows (66)
• Area rule
– The NACA engineer Richard Whitcomb applied this knowledge to the
problem of transonic aircraft
– He reasoned that the area distribution should be as smooth as possible
– This meant that, in the region of the wing and tail, the fuselage cross-
sectional area had to reduce to compensate for the additional area of
these structures
– This led to the “coke bottle” shape and the design philosophy is called the
area rule

Aerodynamics 2009/2010 Page 200


Supersonic flows (67)
• Area rule
– The F102 was redesigned and rebuilt in 118 days and achieved M1.22

F-102 F-102A

Straight-sided fuselage Coke bottle fuselage

Aerodynamics 2009/2010 Page 201


Supersonic flows (68)
• Area rule
– Other ways of area ruling
Boeing 747-100

Boeing 747-400
Aerodynamics 2009/2010 Page 202
Supersonic flows (69)
• Area rule

Aerodynamics 2009/2010 Page 203


Supersonic flows (70)
• Area rule
– Some aircraft cannot change the fuselage shape so extra volume is
added to smooth the area distribution
• Kuchemann “carrots”

Convair CV-990

Aerodynamics 2009/2010 Page 204


Supersonic flows (71)
• The supercritical aerofoil
– While thinner aerofoils help reduce the drag rise near Mach 1, there is a
practical limit on how thin an aerofoil can be
• Spar depths and fuel volume
– So is there a way we can delay the drag rise to higher Mach numbers for
an aerofoil of given thickness?
– Increasing Mcr is one
way, but another is to
increase the increment
between the critical
Mach number and the
drag-divergence Mach
number
• i.e. increase the gap
between point c and
point e
– An aerofoil which does
this is known as a
supercritical aerofoil

Aerodynamics 2009/2010 Page 205


Supersonic flows (72)
• The supercritical aerofoil
– The shape of a supercritical aerofoil is compared to a common NACA 64-
series aerofoil in the picture
– The supercritical aerofoil has a relatively flat top leading to a region of
supersonic flow with lower Mach number than the NACA 64-series
– In turn, the terminating
shock is much weaker
and thus creates less
drag
– The picture shows the
NACA section at a
lower Mach number
but the supersonic
region is taller, the
local Mach numbers
higher and the
terminating shock
stronger than the
supercritical aerofoil at
a higher speed

Aerodynamics 2009/2010 Page 206


Supersonic flows (73)
• The supercritical aerofoil
– The picture shows experimental data from the two aerofoils
– The drag divergence Mach number for the NACA 64-series aaerofoil is
0.67 and for the supercritical aerofoil is 0.79

– The relatively flat upper


surface is achieved
through negative camber
for the forward 60% of
the aerofoil
– This lowers the lift which
is compensated by
extreme positive camber
on the rearward 30%
– This produces the cusp-
like shape of the lower
surface near the trailing
edge

Aerodynamics 2009/2010 Page 207


Supersonic flows (74)
• The supercritical aerofoil

Airbus A300

Handley Page Victor

Aerodynamics 2009/2010 Page 208


Supersonic flows (75)
• Shock-induced separation
– The onset of transonic shock-induced
flow separation is not confined to
large increases in drag
– It can also trigger a variety of
aeroelastic instability and response
phenomena including flutter,
oscillations and control surface buzz,
shock-induced oscillations
– That is, the shock is not always static

Aerodynamics 2009/2010 Page 209


Supersonic flows (76)
• Shock-induced separation
– We might wish, therefore, to consider some form of shock wave
control
– This will consist of either
• Modification of the geometry at the foot of the shock wave to either
smear a single shock into a series of weaker ones or fix its location
(shock bump)
• Modification of the boundary layer to withstand the pressure gradient
across the shock (passive or active vortex generators, suction or
blowing)

Aerodynamics 2009/2010 Page 210


Supersonic flows (77)
• Shock-induced separation
– Shock bodies

Aerodynamics 2009/2010 Page 211


Supersonic flows (77)
• Shock-induced separation
– To maintain their operational
ceiling of 70,000 feet
(21,000 m), the U-2A and U-2C
flew very near their maximum
speed
– However, the aircraft's stall
speed at that altitude is only
10 knots less than its maximum
speed
– There was a danger when
turning that the inner wing
stalled because it was going too
slow and the outer wing stalled
because of shock formation and
shock-induced separation
– This point of the flight enveloped
was referred to by the pilots as
"coffin corner"
Aerodynamics 2009/2010 Page 212
Supersonic flows (78)
• Total aircraft drag

Trim Drag
Wave (Wing) Drag
Parasitic Drag
Nacelle Interaction Drag
Unaccounted Drag
Vortex (Induced) Drag
Profile drag

Drag Breakdown of a Representative 4 Engine Civil


Transport Aircraft
At Design Cruise Mach Number & Lift Coefficient

Aerodynamics 2009/2010 Page 213


Supersonic flows (79)

Aerodynamics 2009/2010 Page 214


Supersonic flows (A1)
• Isentropic flow (derivations)
– The basic equations:
Continuity ρ 1u1A1 = ρ 2u2A2

Momentum ρ1u12 A1 − ρ 2u22 A2 + p1 A1 − p2 A2 + 12 ( p1 + p2 )( A2 − A1 ) = 0

~ u12 ~ u22
Energy h1 + = h2 +
2 2
~ ~
Enthalpy
h1 = c pT1 , h2 = c pT2

c pT1 + 12 u12 = c pT2 + 12 u22


i.e.
p1 p
= 2
Equation of state ρ1T1 ρ 2T2
γp u
Speed of sound a= = γ RT = (γ − 1)c pT =
ρ M
Aerodynamics 2009/2010 Page 215
Supersonic flows (A2)
• Isentropic flow (derivations)
– The energy equation c pT1 + 12 u12 = c pT2 + 12 u22

– Thus becomes u12 γ p1 u22 γ p2


+ = +
2 (γ − 1) ρ1 2 (γ − 1) ρ 2

– Or u12 a12 u22 a22


+ = +
2 γ −1 2 γ −1

– If we take p1 as p0 (stagnation conditions) and p2 as p (local conditions) and


rearrange the equations, we get
γ (γ −1)
p0  γ − 1 2 
= 1 + M 
p  2 

Aerodynamics 2009/2010 Page 216


Supersonic flows (B1)
• Oblique shock angle (derivation)
– The basic equations (note the resolution of the velocity vector into components normal to and
tangential to the shock wave):
Continuity ρ 1V1n = ρ 2V2n

Momentum (normal to the shock wave)

Momentum (parralel to the shock wave)


p1 − p2 = ρ 2V22n − ρ1V12n
• i.e. no change in pressure)

0 = ρ 2V2 nV2t − ρ1V1nV1t


Energy (from slide ‘Supersonic flows (25)’)

(a* is the value of the speed of sound at sonic conditions and is constant)
a2 u2 γ + 1 *2
+ = a
γ −1 2 2(γ − 1)

Aerodynamics 2009/2010 Page 217


Supersonic flows (B2)
• Oblique shock angle (derivation)
– Applying the energy equation before and after the shock gives

V12n + V1t2 γ p1 γ + 1 *2
+ = a
2 γ - 1 ρ1 2(γ − 1)
V22n + V22t γ p2 γ + 1 *2
+ = a
– From ρ 1V1n 2= ρ 2V2nγ and
- 1 ρ2 γ − V1)1t =V2t
2(that
it follows
– i.e. the tangential velocity is the same on both sides of the shock
0= ρ V V −ρV V
– Since the tangential velocity doesn’t change we2 just
2 n need
2 t to determine
1 1n 1t the normal velocity after the shock
– Again using continuity we get

– Where p2/ρ 2 and p1/ρ 1 can be eliminated using the equations at the top of the slide
p2 p1
− = V1n − V2 n
ρ 2V2 n ρ1V1n

Aerodynamics 2009/2010 Page 218


Supersonic flows (B3)
• Oblique shock angle (derivation)
– Dropping the station subscript (1 or 2) for Vt (because V1t = V2t ) gives

γ + 1 *2  1 1  γ − 1  Vt 2   Vt 2 
a  −  + V1n +  − V2 n +  = V1n − V2 n
2γ  V2 n V1n  2γ  V1n   V2 n 
– Which we arrange to the form

 γ + 1 a*2 γ − 1 Vt 2 γ + 1 
 − (V − V ) = 0
 2γ V1nV2 n 2γ V1nV2 n 2γ  1n 2 n
 
– This equation is satisfied when either factor is zero
– The solution that the second factor is zero (i.e. V1n – V2n = 0, or V1n = V2n ) corresponds to a
shock wave of zero intensity, or a Mach wave
– Setting the first factor to zero gives a non-trivial solution:

∗2 γ −1 2
V1nV2 n = a − Vt
γ +1

Aerodynamics 2009/2010 Page 219


Supersonic flows (B4)
• Oblique shock angle (derivation)
– From the picture we get V1n = V1 sinβ and Vt = V1 cosβ and we substitute
these into the previous equation to get
2
a∗ γ − 1 cos 2 β
V2 n = − V1
V1sinβ γ + 1 sinβ

– Which can be written as


 2
V1  a a0  γ − 1
∗ 
V2 n =   − cos β 
2

sinβ  a0 V1  γ + 1 

– And we can replace a*/a0 and a0 /V1 for terms involving γ and M to give

V1 γ −1 2 2 1 
V2 n =  sin β +
sinβ γ + 1 γ + 1 M 12 

Aerodynamics 2009/2010 Page 220


Supersonic flows (B4)
• Oblique shock angle (derivation)
– From the picture we can see that V2n is related to the wave angle( β)and
deflection angle (θ) by

V2n = Vt tan(β - θ) = V1 cosβ tan(β - θ)

– Which may be equated to the equation


at the bottom of the previous slide to give
1  γ −1 2 2 1 
tan( β − θ ) =  sin β + 
2 
sinβ cosβ  γ +1 γ + 1 M1 

– Or

 1  γ −1 2 2 1 
θ = β − tan −1
 sin β + 
2 
 sinβ cosβ  γ +1 γ + 1 M 1 

Aerodynamics 2009/2010 Page 221


Swept Wings
• Swept wing flows; Effect of spanwise and normal velocity
components; qualitative description of 3D boundary layers on
swept wings; Forward, rearward and variable sweep wings;
control surface effects; delta wings and vortical flows; vortex
flap; aerodynamics of aircraft at high incidences.

Aerodynamics 2009/2010 Page 222


Swept wings (1)
• Why sweep the wings
– We’ve already seen that wing sweep increases the effective t/c
– But it also moves the wings behind the bow shock

Aerodynamics 2009/2010 Page 223


Swept wings (2)
• Incidence of a shock on a
wing
– Sweeping the wing back alters
the onset Mach number, i.e.
that normal to the leading
edge, M∞ n
– If the wing is swept at the
angle of the Mach line
M∞ n = M∞ cos (90°-μ)
= M∞ sin μ
=1
– That is, the normal Mach
number is unity
– Sweeping it back further
reduces the normal Mach
number to subsonic

Aerodynamics 2009/2010 Page 224


Swept wings (3)
• Bell X-1 and X-2

Aerodynamics 2009/2010 Page 225


Swept wings (4)
• Effect of sweep on pressure
– Recall that the pressure
coefficient is dependent on
onset Mach number only

2  p 
Cp =  − 1
γ M ∞2  p∞ 

– So the normal pressure


distribution is thus related to
the normal Mach number

2  p 
C pn =  − 1
γ M ∞2 n  p∞ 

– And thus varies with sweep

Aerodynamics 2009/2010 Page 226


Swept wings (5)
Velocity components

Sweep introduces a spanwise component of the freestream velocity on
– the wing

Transition of the boundary layer can now occur in each direction



The surface flow is different to the freestream

This creates shear stresses and an additional transition mechanism

Aerodynamics 2009/2010 Page 227


Swept wings (6)
• Cross-flow instabilities
– The shear causes waves and vorticity in a spanwise direction
– In addition to 2-D transition (Tollmien-Schlichting waves) there is additional 3-D transition
mechanism due to these spanwise disturbances or cross-flow instabilities

Aerodynamics 2009/2010 Page 228


Swept wings (7)
• Stall characteristics
– Rectangular wings have larger downwash
angles at the tip than at the root
– The effective angle of attack at the tip is
therefore lower at the tip and it will stall last
– However, rectangular wings are not very
efficient
• They have more induced drag than the ideal
elliptical planform

– A compromise is to taper the wing


– But with a small tip chord come reduced local
Reynolds number, increased effective angle
of attack and thicker boundary layers (due to
spanwise flow)
– This generally leads to swept, tapered wings
being prone to tip stall

Aerodynamics 2009/2010 Page 229


Swept wings (8)
• Tip stall
– Tip stall is not good!
– Ailerons become ineffective
– Loss of lift means the aerodynamic centre moves forward
– Pitch up

Aerodynamics 2009/2010 Page 230


Swept wings (9)
• Methods to
prevent pitch-up
– Wing twist
– By twisting the
wing tip nose
downwards
(wash-out), the
local angle of
attack is reduced

Aerodynamics 2009/2010 Page 231


Swept wings (10)
• Methods to
prevent pitch-up
– Wing fences
– Reduces or stops
the spanwise flow

Aerodynamics 2009/2010 Page 232


Swept wings (11)
• Methods to prevent pitch-
up
– Wing snags, saw teeth,
dog teeth
– Generate discrete, strong
vorticity that helps the flow
remain attached

Aerodynamics 2009/2010 Page 233


Swept wings (12)
• Methods to prevent pitch-
up
– Forward sweep
– The spanwise velocity is in
the other direction – tip to Grumman X-29
root
– Other advantages include:
• Better pilot vision as the
wing root is relatively far
aft
• Wing spars can be placed
behind a weapons bay
rather than through it
• Controllability to much
Schleicher ASK 13
higher angle of attack (67°
for the X-29)

Aerodynamics 2009/2010 Page 234


Junkers Ju 287
Swept wings (13)
• Methods to prevent pitch-
up
– Forward sweep
– The main disadvantage is
structural
Sukhoi Su-47
– The wing tip tends to twist
up, increasing the local load
and thus increasing the twist
even more
– An unfortunate tendency that
can be countered by
strengthening the structure
Hansa HFB320
of a metal wing or using
cunning layups of carbon
fibre
– Aeroelastic tailoring

Aerodynamics 2009/2010 Page 235


Swept wings (14)
• Variable sweep wings
– We have seen that a swept wing is more suitable for high speeds
– An unswept wing is suitable for lower speeds
– A variable-sweep wing allows a pilot (or flight control system) to select the correct wing configuration for the plane's intended
speed

– The variable-sweep wing is most useful for those aircraft that are expected to function at both low and high speed, and for this
reason it has been used primarily in military aircraft

Messerschmitt Me P.1101
Grumman XF10F Jaguar

Bell X-5
Aerodynamics 2009/2010 Page 236
Swept wings (15)
• Variable sweep wings
– But the extra mechanisms are heavy

Tornado

Aerodynamics 2009/2010 Page 237


Sukhoi Su-17

Swept wings (16)


• Variable sweep wings

Tupolev Tu-160
Rockwell B-1

Grumman F-14

General Dynamics F-111

MiG-23
Aerodynamics 2009/2010 Page 238
Swept wings (17)
• Delta wings
– Delta wings are a special form of swept wing pioneered by Lippisch
– The wing leading edge remains behind the shock wave generated by the nose of the aircraft when flying at supersonic speeds
– While this is also true of ordinary swept wings, the delta's planform carries across the entire aircraft which has structural advantages
– Another advantage is vortex lift
– Beyond a certain angle of attack, the wing leading edge generates a stable vortex which remains attached to the upper surface of the wing
– This gives delta wings a relatively high stall angle

Convair XF-92
Lippisch P.13

Aerodynamics 2009/2010 Page 239


Swept wings (18)
• Delta wings
– Types of delta wing

Aerodynamics 2009/2010 Page 240


Swept wings (19)
• Vortex lift
– Leading edge boundary layer rolls up into a vortex

Aerodynamics 2009/2010 Page 241


Swept wings (20)
• Vortex lift
– The wing generates lift like a conventional aerofoil at low angles of attack
– The leading edge vortices form at increasing angle of attack and contribute significant lift
and enable stability and control at relatively high angles of attack

Aerodynamics 2009/2010 Page 242


Swept wings (21)
• Vortex lift
– The solution for low-speed, high-lift performance of Concorde (Kuchemann
at RAE)

Aerodynamics 2009/2010 Page 243


Swept wings (22)
• Vortex lift
– Aircraft that wish to operate at high angles of attack tend to generate and utilise vortex lift
– This can be from areas other than the wing, e.g. leading-edge extensions (LEX)
• N.B. Fin Buffet

Aerodynamics 2009/2010 Page 244


Swept wings (23)
• Vortex lift
– The limiting factor is a phenomenon called vortex burst
– Vortex bursting is a phenomenon in which the structured character of the
vortex is destroyed resulting in a loss of most of the vortex lift

– It occurs due to
adverse pressure
gradients acting
on the vortex
– When the vortex
burst occurs on
the wing (as
opposed to
downstream of the
wing) the lift drops
substantially.

Aerodynamics 2009/2010 Page 245


Swept wings (24)
• Vortex flap
– The concept of the vortex flap was to reposition the leading-edge vortex
system which normally develops over a delta (or high-sweep) wing at high
angles of attack onto a forward facing flap surface
– This results in a reduction of induced drag due to a thrust component
derived from the low pressure on the flap.

Aerodynamics 2009/2010 Page 246


Swept wings (25)
• Vortex flap
– NASA Langley conducted flight tests on a modified NF-106B Delta
Dagger

Aerodynamics 2009/2010 Page 247


Swept wings (26)
• Vortex flap
– NASA Langley conducted flight tests on a modified NF-106B Delta
Dagger

α =9°, 30° vortex flap

Aerodynamics 2009/2010 Page 248


Swept wings (27)
• Vortex flap

α =13°, 30° vortex flap

Aerodynamics 2009/2010 Page 249


Swept wings (28)
• Vortex flap

α =13°, 40° vortex flap

Aerodynamics 2009/2010 Page 250


Swept wings (29)
• Vortex flap
– The flow visualisation shows that
the flowfield does not behave as
anticipated

– The vortex flap generates lots of


weak vortices rather than a single
strong one

– The next vortex along the span is


triggered by the secondary vortex
of the previous one

Aerodynamics 2009/2010 Page 251


Swept wings (30)
• Vortex flap
– Never really adopted
operationally
– Typhoon uses a drooped nose
device not dissimilar to the A380

Aerodynamics 2009/2010 Page 252


Aerodynamics 2009/2010 Page 253