Some Notes On Aircraft Stability and Control: Michael Carley

Some notes on aircraft stability and control
Michael Carley, m.j.carley@bath.ac.uk

A lonely impulse of delight
Drove to this tumult in the clouds; . . .
Contents
Contents i
List of Figures iii
List of Tables iv
I Static stability 1
1 How aircraft y 3
1.1 Equilibrium and stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.2 Functions of aircraft controls . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3 Forces and moments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.4 Effects of symmetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.5 Trim and stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.6 Aerodynamic centre and neutral point . . . . . . . . . . . . . . . . . . . . . 7
1.7 Denitions of static and c.g. margins . . . . . . . . . . . . . . . . . . . . . . 9
1.8 Basic aerofoil and control characteristics . . . . . . . . . . . . . . . . . . . . 9
Aerofoils and wings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Control forces and moments . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Control hinge moments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2 Longitudinal static stability 15
2.1 Elementary longitudinal stability . . . . . . . . . . . . . . . . . . . . . . . . 15
2.2 Effect of downwash on tailplanes . . . . . . . . . . . . . . . . . . . . . . . . 16
2.3 Stick xed stability and c.g. margins . . . . . . . . . . . . . . . . . . . . . . 19
2.4 Stick free stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.5 Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3 Flight testing 23
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.2 K
n
elevator angle to trim . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.3 What does the pilot feel? . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.4 K
n
tab angle to trim . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
i
4 Tailless aircraft 29
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.2 Stick xed stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4.3 Static margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
5 Stick forces 33
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
5.2 Analysis to calculate stick forces . . . . . . . . . . . . . . . . . . . . . . . . 33
5.3 More ight testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
5.4 Modication of stick forces . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
6 Manoeuvre stability 39
6.1 Analysis of a steady pullout . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
6.2 Stick xed manoeuvre point . . . . . . . . . . . . . . . . . . . . . . . . . . 42
6.3 Stick xed manoeuvre stability . . . . . . . . . . . . . . . . . . . . . . . . . 43
6.4 Stick free manoeuvre stability . . . . . . . . . . . . . . . . . . . . . . . . . . 43
6.5 Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
6.6 Tailless aircraft . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
6.7 Tailless aircraft manoeuvre point . . . . . . . . . . . . . . . . . . . . . . . . 46
6.8 Tailless aircraft manoeuvre margins . . . . . . . . . . . . . . . . . . . . . . 47
6.9 Relationships between static and manoeuvre margins . . . . . . . . . . . . . 47
Conventional aircraft . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Tailless aircraft . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
6.10 Modication of stick free neutral and manoeuvre points . . . . . . . . . . . . 48
7 Compressibility effects 51
7.1 High speed effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
II Dynamic stability 55
8 Dynamic behaviour of aircraft 57
8.1 Axes and notation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
8.2 Aerodynamic derivatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
8.3 Longitudinal symmetric motion . . . . . . . . . . . . . . . . . . . . . . . . . 61
9 Normal modes of aircraft 63
Phugoid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Short period oscillation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
9.1 Lateral motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Dutch roll . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Spiral mode and roll subsidence . . . . . . . . . . . . . . . . . . . . . . . . 66
9.2 Dihedral effect and weathercock stability . . . . . . . . . . . . . . . . . . . . 67
Dihedral effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
ii
iii
Weathercock stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
List of Figures
1.1 Phases of ight . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Denitions of stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3 Axes and sign conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.4 Sign conventions for longitudinal stability . . . . . . . . . . . . . . . . . . . . . 6
1.5 Trim and stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.6 Centre of pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.7 Incremental loads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.8 c.g and aerodynamic centre relationships . . . . . . . . . . . . . . . . . . . . . . 8
1.9 Lift curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.10 Loading due to control deection . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.11 Measurement of control/tab deections . . . . . . . . . . . . . . . . . . . . . . . 11
1.12 Pressure distributions due to deections, a
1
> a
2
> a
3
. . . . . . . . . . . . . . . 11
1.13 Measurement of control surface area . . . . . . . . . . . . . . . . . . . . . . . . 12
1.14 Control hinge moments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.1 Stick xed stability (conventional aircraft) . . . . . . . . . . . . . . . . . . . . . 15
2.2 Trailing vortices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.3 Effect of downwash on tailplane . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.4 Stick free elevator conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.1 Elevator angle to trim at varying lift coefcients . . . . . . . . . . . . . . . . . . 24
3.2 Measurement of neutral point location . . . . . . . . . . . . . . . . . . . . . . . 25
3.3 What the pilot experiences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.4 Tab angle to trim at varying lift coefcients . . . . . . . . . . . . . . . . . . . . 27
3.5 Measurement of stick free neutral point location . . . . . . . . . . . . . . . . . . 28
4.1 Canard conguration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.2 Control surfaces for tailless aircraft . . . . . . . . . . . . . . . . . . . . . . . . . 30
4.3 Representation of tailless aircraft . . . . . . . . . . . . . . . . . . . . . . . . . . 30
5.1 Aerodynamic assistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
6.1 Manoeuvre conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
6.2 Manoeuvre conditions for a tailless aircraft . . . . . . . . . . . . . . . . . . . . . 45
6.3 Aerodynamic forces during pitching motion . . . . . . . . . . . . . . . . . . . . 46
6.4 Modication of neutral and manoeuvre points . . . . . . . . . . . . . . . . . . . 48
6.5 Effects of positive springs and bob-weights . . . . . . . . . . . . . . . . . . . . 49
6.6 Modication of stick force per g . . . . . . . . . . . . . . . . . . . . . . . . . . 49
7.1 Compressibility effects on lift curve slope and aerodynamic centre . . . . . . . . 52
7.2 Compressibility effects on zero lift pitching moment and zero lift angle . . . . . . 52
7.3 Aeroelastic effects on lift curve slope . . . . . . . . . . . . . . . . . . . . . . . . 53
7.4 Aeroelastic effects on tailplane and elevator . . . . . . . . . . . . . . . . . . . . 53
7.5 Variation of downwash with Mach number . . . . . . . . . . . . . . . . . . . . . 54
7.6 Variation of pitching moment with Mach number . . . . . . . . . . . . . . . . . 54
7.7 Variation of stick forces with Mach number . . . . . . . . . . . . . . . . . . . . 54
8.1 Notation for analysis of dynamic stability . . . . . . . . . . . . . . . . . . . . . 58
9.1 Phugoid oscillation trajectory . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
9.2 Short period oscillation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
9.3 Rolling subsidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
9.4 Stability of the lateral modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
9.5 dihedral effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
9.6 Wing sweep effects on L
v
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
9.7 Wing-fuselage interference effects on L
v
. . . . . . . . . . . . . . . . . . . . . . 69
9.8 Use of twin ns at high speed . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
9.9 Intake effects on N
v
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
List of Tables
5.1 Maximum control forces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
iv
Part I
Static stability
1
Chapter 1
How aircraft y
Aircraft y by generating enough lift to reach or exceed their weight. They usually do this by
holding a wing at a certain angle of attack, or incidence. Longitudinal control is the study of
how to set, maintain and change that angle of attack; stability is the study of how the incidence
behaves when it is perturbed slightly.
The easiest way to think about this is to look at the phases of aircraft ight, Figure 1.1.
When an aircraft takes off, its speed is quite lowand it must generate lift greater than its weight
in order to leave the ground. In cruise, the aircraft operates at a constant speed and constant
lift. Finally, when the aircraft lands, it needs to reduce its speed without losing too much lift.
In each case, the issue for control of the aircraft is how it can maintain its incidence at
Take-off: the incidence increases to generate more lift at low speed.
Cruise: the aircraft ies at constant incidence and speed
Landing: the aircraft increases incidence to allow it to slow down.
Figure 1.1: Phases of ight
3
4 CHAPTER 1. HOW AIRCRAFT FLY
a given speed. When it takes off or lands, it does so by rotatingraising or lowering its
nosein order to change the incidence of the wing, altering the relationship between speed
and lift. Aircraft control is the study of how a pilot can x the relationship between speed and
incidence.
When an aircraft cruises, it is desirable that it do so at constant speed and incidence, so
the controls are at a xed setting. Aircraft stability is the study of how an aircraft responds to
small disturbances in ight and how it can be designed so that it remains at a xed incidence
and speed.
In each case, the basic question is how to generate a moment on the aircraft so that it
rotates and changes the wing incidence or so that the net moment is zero and the aircraft ies
at constant incidence. This is done via the aircraft controls. First, we need to clarify what we
mean by some basic terms.
1.1 Equilibrium and stability
The requirements of an aircraft control is system are that it be able to bring the aircraft into
some required equilibrium and that it be able to maintain that equilibrium stably.
A system is in equilibrium if the sum of all the forces and moments acting on it is zero.
A system is statically stable if, when disturbed from equilibrium, it tends to return to the
equilibrium position.
A system is dynamically stable if, given enough time, it actually returns to the equilibrium
position.
The distinction between the two cases is shown in gure 1.2.
R
e
s
p
o
n
s
e
Time
Neutral stability
Statically and dynamically stable
Statically stable and dynamically unstable
Statically unstable
Figure 1.2: Denitions of stability
1.2 Functions of aircraft controls
The elevator provides pitch control. It is used to provide pitching moment equilibrium and
enables adjustment of incidence.
1.3. FORCES AND MOMENTS 5
Ailerons give roll control. They are used to bank the aircraft to initiate turns and to oppose
disturbances in roll due to crosswinds or gusts.
The rudder gives yaw control. Its most important functions are in balancing engine-failed
yawing moments, turn co-ordination, spin recovery and crosswind landing.
The three controls are used to enable moment equilibrium in all three axes simultaneously.
Controls may be operated mechanically by the pilot, or there might be aerodynamic assis-
tance, power assistance or full power operation. Full power operation can be mechanically or
electronically signalled (y by wire).
The sign conventions for the controls and motions are shown in gure 1.3.
Positive up
Positive down
Positive left
Positive down
z
Yaw
x
Roll
y
Pitch
Figure 1.3: Axes and sign conventions
1.3 Forces and moments
The forces and moments on an aircraft arise from two sources: the mass of the aircraft and its
aerodynamics.
The mass and mass distribution give rise to inertial forces (masslinear acceleration,
moment of inertiaangular acceleration) and gravitational forces (massacceleration due to
gravity).
The aerodynamic contributions are:
1. static forces and moments due to linear velocities;
2. damping forces and moments due to angular velocities;
3. control forces and moments produced by rudder, ailerons, elevators etc.
1.4 Effects of symmetry
Since most aircraft are laterally symmetric (i.e. the left-hand side is a mirror image of the
right-hand side) it follows that in forward ight with wings level and with no roll or yaw the
resultant of the aerodynamic forces must lie in the plane of symmetry. Hence, in straight ight
any symmetric disturbance will result in only horizontal and vertical motion of the centre of
gravity (c.g.) and pitching about the c.g. This is longitudinal symmetric motion.
1.5 Trim and stability
The sign conventions for examining equilibrium and stability in free ight are shown in g-
ure 1.4.
Horizontal
F
lig
h
t
d
ir
e
c
t
io
n
Z
ero
lift
lin
e
W
L
D
T
Figure 1.4: Sign conventions for longitudinal stability
For equilibrium there must be no net forces or moments acting on the aircraft. Hence,
resolving parallel and perpendicular to the aircraft at equilibrium:
Perpendicular:L W cos = 0, (1.1)
Parallel:T D W sin = 0, (1.2)
Moments about the c.g.:M
cg
= 0. (1.3)
M
cg
cannot come from the weight of the aircraft since, by denition, the gravitational
moments about the c.g. are zero. Therefore, the aerodynamic moments coming from the
wing/fuselage and tail must all be in balance. If M
cg
= 0 the aircraft is said to be trimmed.
For stability all that is necessary is that if, say, the aircraft pitches nose up then the resulting
incremental aerodynamic moment about the c.g., M
cg
, should be negative in order to push
the nose down again, i.e. M
cg
/ must be negative.
It is quite possible to achieve trim without stability. It is also possible to achieve stability
without being able to trim at a useful incidence. These states are shown in gure 1.5.
1.6. AERODYNAMIC CENTRE AND NEUTRAL POINT 7
C
M
c
g
Neutrally stable
cannot trim
C
M
/ > 0
C
M
/ < 0
C
M
/ < 0
C
L
< 0
Figure 1.5: Trim and stability
1.6 Aerodynamic centre and neutral point
The aerodynamic forces and moments acting on an aircraft of a given shape are independent
of the position of the centre of gravity. Therefore, the aerodynamic loads can be considered
separately from the gravitational loads.
The loads around an aircraft in trim conditions are sketched in Figure 1.6. The aerody-
namic loads can be considered to act through a point at which the aerodynamic moments are
zero. This location is known as the centre of pressure. However, as the incidence of the air-
craft varies the location of the centre of pressure changes, it is therefore not very useful for
calculations or consideration of stability.
L
Figure 1.6: Centre of pressure
A more useful concept is that of aerodynamic centres. The aerodynamic centre is generally
dened as the location at which the pitching moment is unaffected by the incidence (this as-
sumes linear aerodynamics). One way of thinking of the aerodynamic centre is by considering
the incremental aerodynamic loads. These are sketched in gure 1.7.
When the incidence is increased slightly the incremental aerodynamic load distributions
can be considered to act through a certain pointtheir combined centre of pressure. This
point is the aerodynamic centre of the whole aircraft, and is the point where dM/d = 0.
L
Figure 1.7: Incremental loads
The relationship between the aerodynamic centre of the whole aircraft and the location of
the centre of gravity of the aircraft denes its static stability. There are three possible cases,
shown in Figure 1.8.
W
L
M
0
W
L
M
0
W
L
M
0
Figure 1.8: c.g and aerodynamic centre relationships
1. If the c.g. is forward of the aerodynamic centre, dM
cg
/d will be negative and the
aircraft will therefore be statically stable.
2. If the c.g. is aft of the aerodynamic centre, dM
cg
/d will be positive and the aircraft
will therefore be statically unstable.
3. If the c.g. is at the aerodynamic centre, dM
cg
/d will be zero and the aircraft will
therefore be neutrally stable.
The neutral point of an aircraft is dened as the position of the centre of gravity for which
the static stability is neutral. A neutral point is a purely aerodynamic property of a particular
conguration (neglecting distortion and aeroelastic effects). For a wing alone dM/d = 0 is
the condition dening the aerodynamic centre (i.e. for a wing alone the neutral point is the
aerodynamic centre. Similarly, the neutral point of a complete aircraft may be regarded as the
aerodynamic centre of the complete aircraft.
The term aerodynamic centre is usually reserved for aerofoils only, and neutral point
is applied to complete aircraft.
1.7. DEFINITIONS OF STATIC AND C.G. MARGINS 9
1.7 Denitions of static and c.g. margins
The gradient of the C
M
vs graph obviously indicates the degree of static stability (see g-
ure 1.5), but the parameters used to communicate the degree of stability and their precise
denitions are the static margin:
K
n
=
dC
Mcg
dC
R
(1.4)
where C
R
= (C
2
L
+C
2
D
)
1/2
, the resultant force.
The c.g. margin, H
n
, is the distance of the neutral point aft of the c.g. (measured in mean
chord, c). It can easily be shown that:
H
n
=
dC
M
dC
L
= h
n
h (1.5)
where hc is the distance of the c.g. aft of a reference point and h
n
c is the distance of the
neutral point aft of a reference point.
At low speeds and for small ight path inclinations C
R
C
L
and C
L
, C
M
, C
R
are not
inuenced by Mach number or aeroelastic effects. Hence H
n
K
n
under these conditions.
1.8 Basic aerofoil and control characteristics
Aerofoils and wings
To enable comparisons to be made between different ight conditions, and between wind-
tunnel testing and full scale ight the lift, drag and pitching moments are generally non-
dimensionalised:
Lift coefcient C
L
=
L
V
2
S/2
(1.6)
Drag coefcient C
D
=
D
V
2
S/2
(1.7)
Moment coefcient C
M
=
M
V
2
Sc/2
(1.8)
The root chord length, c
0
, is often used as the characteristic length for tailless aircraft.
Note that for each of these coefcients, the subscript is upper case. This indicates that the
coefcients relate to three dimensional planforms.
A typical lift curve for a wing is shown in gure 1.9.
For convenience, in stability work we usually use incidence, , measured from the zero
lift line. This is different to most other aerodynamic work, and should be checked for each
case. All of the stability analysis in this course will examine the linear region of the lift curve,
and hence would need to be signicantly modied for high angle of attack regimes and/or
manoeuvring combat aircraft.
C
L
Figure 1.9: Lift curve
Control forces and moments
If the geometric incidence of a wing/tailplane is held constant the effect of a positive deec-
tion of a trailing edge control is to generate extra lift along with a pitching moment. This
pitching moment is generally in the nose down sense due to the form of the additional loading
distribution (see gure 1.10).
x/c
C
p
Figure 1.10: Loading due to control deection
A positive control deection therefore gives a positive contribution to lift and a negative
pitching moment. Similarly, if we have a tab at the trailing edge of the control it will gen-
erate lift together with a negative pitching moment. The deections of controls and tabs are
generally given the symbols and and they are measured as shown in gure 1.11.
The contributions of control surfaces and tabs to lift and pitching moment are generally
assumed to be linear. To simplify future derivations and expressions the following denitions
are used. These refer to the lift generated due to tailplane incidence,
T
, elevator deection,
, and tab deection, :
a
1
=
C
L
T
T
, a
2
=
C
L
T
, a
3
=
C
L
T
. (1.9)
The pressure distributions associated with each of these deections are shown in gure 1.12.
1.8. BASIC AEROFOIL AND CONTROL CHARACTERISTICS 11
Figure 1.11: Measurement of control/tab deections

a
1
a
2
a
3
C
p
x/c
C
p
x/c
C
p
x/c
Figure 1.12: Pressure distributions due to deections, a

1
> a
2
> a
3
.
It is assumed that the lift and pitching moments can also be added together. Hence
C
L
T
= a
1
T
+a
2
+a
3
.
Similarly,
C
Mac
= C
M
0
+
C
M
0
+
C
M
0
.
Control hinge moments
To enable controls to be designed such that the forces that the pilot must exert can be estimated
the moments acting on the control itself (rather than on the whole aircraft) need to be calcu-
lated. Similarly to the forces and moments acting on the aircraft described in the preceding
section the contributions from tailplane incidence, elevator and tab deection are considered
separately and summed to give the total control hinge moment. The hinge moment coefcient,
C
H
, is dened as:
C
H
=
Hinge moment
V
2
S
/2
(1.10)
Where S
is the control surface area and c
is the control surface mean chord. Both of these

values are measured aft of the hinge line, as shown in gure 1.13.
Hinge line
Aerodynamic balance
S
Figure 1.13: Measurement of control surface area

Again, to simplify future derivations the following denitions are used to represent the
control hinge moments due to tailplane incidence, elevator and tab deection respectively:
b
1
=
C
H
T
, b
2
=
C
H
, b
3
=
C
H
, (1.11)
For non-symmetric tailplane cross-sections there is usually a hinge moment when all other
deections are zero. This is given the symbol b
0
. The pressure distributions associated with
each of these terms are shown in gure 1.14.
The hinge moments acting on a given control can therefore be dened as:
C
H
= b
0
+b
1
T
+b
2
+b
3
. (1.12)
The characteristics of any given control will depend on a number of factors including the
relative size of the control (c
f
/c), the hinge position and the shape of the balance portion,
as well as the aerofoil cross-section. These factors will be discussed when stick forces for
manoeuvres are examined.
1.8. BASIC AEROFOIL AND CONTROL CHARACTERISTICS 13
b
1
b
2
b
3
H
i
n
g
e
l
i
n
e
C
p
x/c
H
i
n
g
e
l
i
n
e
C
p
x/c
H
i
n
g
e
l
i
n
e
C
p
x/c
Figure 1.14: Control hinge moments

Chapter 2
Longitudinal static stability
2.1 Elementary longitudinal stability
Stick xed stability is concerned with calculating the trim angles and stability of an aircraft
with the control surfaces held at a constant location. Hence, no account is made of the control
forces that the pilot (or power assist etc.) must provide. This simplies the analysis such that
only the forces and moments acting on the whole aircraft need to be in equilibrium for a trim
condition. The simplest way to consider the trim and stability of a conventional aircraft is
shown in gure 2.1.
Datum
W
h c
L
WBN
h
0
c
M
0
L
T
l
T
z
T
D
z
D
Figure 2.1: Stick xed stability (conventional aircraft)
It is assumed that the aerodynamics are linear, that the angles and
T
are small and that
there is no wake effect on the tailplane (2.2). Note that the total lift of the aircraft has now
been divided into a Wing/Body/Nacelle (WBN) component and a Tailplane (T) component.
Taking moments about the c.g.:
M
cg
= M
0
L
WBN
(h
0
h)c L
T
((h
0
h)c +l) Tz
T
+Dz
D
= M
0
(h
0
h)c(L
WBN
+L
T
) L
T
l Tz
T
+Dz
D
= M
0
(h
0
h)cL L
T
l Tz
T
+Dz
D
.
15
16 CHAPTER 2. LONGITUDINAL STATIC STABILITY
If we assume that Tz
T
and Dz
D
are small, which is reasonable since lift is normally much
larger than drag (and hence thrust):
M
cg
= M
0
(h
0
h)cL L
T
l.
As is usual for aerospace calculations, non-dimensional parameters are used. C
L
, C
D
and
C
M
have been dened previously (1.8). However, a new denition is required for the lift
coefcient of the tailplane, C
L
T
:
C
L
T
=
L
T
V
2
S
T
/2
(2.1)
where S
T
is the area of the tailplane.
Hence, if the pitching moment M
cg
is divided through by S(V
2
/2) we get:
C
Mcg
= C
M
0
(h
0
h)C
L
C
L
T
S
T
S
l
c
.
If we dene the tail volume coefcient, V , as:
V =
S
T
S
l
c
(2.2)
C
Mcg
= C
M
0
(h
0
h)C
L
V C
L
T
. (2.3)
This is an extremely important equation and is at the root of all stick-xed stability calcu-
lations. If youre stuck when attempting a question that asks, What is the lift required at the
tailplane for trim. or, Calculate the elevator angle required for trim. this is always a good
starting point!
Remember, for trim M
cg
= 0 hence C
Mcg
= 0.
2.2 Effect of downwash on tailplanes
Downwash is present at the tailplane of a conventional aircraft due to the trailing vortices
generated by the wing, as shown in gure 2.2.
This has a signicant effect on the lift generated by the tailplane, since it alters the effective
angle of attack as shown in gure 2.3.
Hence:
T
= +
T
where
T
is the tailplane setting relative to the wing zero lift line.
For an untwisted wing, the downwash velocity (and hence the downwash angle, ) at the
tailplane is proportional to the lift generated by the wing. Therefore, since we are only in-
terested in the linear region of the lift curve, the downwash angle is also proportional to .
Hence:
=
d
d
+
0
.
2.2. EFFECT OF DOWNWASH ON TAILPLANES 17
Figure 2.2: Trailing vortices
Z
L
L
t
a
i
l
p
l
a
n
e
Z
L
L
W
B
N
T
Free stream
R
e
s
u
lta
n
t
o
w
Figure 2.3: Effect of downwash on tailplane
0
is only present for a wing where the zero lift angle of attack varies along its length (i.e. a
wing with a varying cross-section/camber along its length or with twist). Putting these two
equations together results in:
T
= +
T
0
+
d
d
_
=
_
1
d
d
_
+ (
T
0
).
By denition (1.8):
C
L
T
= a
1
T
+a
2
+a
3
.
So,
C
L
T
= a
1
( +
T
) +a
2
+a
3
and therefore:
C
L
T
= a
1
_
1
d
d
_
+a
1
(
T
0
) +a
2
+a
3
.
To be able to solve this equation, and calculate the lift acting on the tailplane, we require a
relationship between this and C
L
.
Since we are looking at the linear region of aerodynamics and we have dened that there
is zero lift at zero incidence (gure 1.9), we can therefore say that:
C
L
= a
and hence
= C
L
/a
where a is the overall lift curve slope of the aircraft. We can substitute this into the previous
equation to get:
C
L
T
=
a
1
a
_
1
d
d
_
C
L
+a
1
(
T
0
) +a
2
+a
3
. (2.4)
Equation 2.3 made no assumptions about where the lift at the tailplane came from, or whether
there was a downwash effect etc. Therefore, this equation can be substituted into Equation 2.3,
resulting in the rather torturous, but extremely useful equation below:
C
M
= C
M
0
(h
0
h)C
L
V
_
a
1
a
_
1
d
d
_
C
L
+a
1
(
T
0
) +a
2
+a
3
_
.
This equation is so useful because it allows us to calculate the elevator angles needed for trim
at given ight conditions. The elevator angle to trim is given the symbol, . This is often a
design criterion for aircraft. For example, the c.g. range available for an aircraft is generally
limited by the elevator angle required to trim on approach to a runway at low speed with aps
2.3. STICK FIXED STABILITY AND C.G. MARGINS 19
and undercarriage extended. To calculate the elevator angle to trim, , the previous equation is
rearranged to make the subject and, since we are calculating the conditions for trim, C
M
=0.
This results in:
=
1
V a
2
_
C
M
0
(h
0
h)C
L
V
_
a
1
a
_
1
d
d
_
C
L
+a
1
(
T
0
) +a
3
__
.
This equation is not really worth remembering, since it is easily derived by rearranging Equa-
tion 2.3.
2.3 Stick xed stability and c.g. margins
By denition,
K
n
H
n
=
dC
M
dC
L
.
Hence, to examine stability we need to differentiate equation 2.3 with respect to C
L
:
dC
M
dC
L
= (h
0
h) +V
a
1
a
_
1
d
d
_
By denition, the neutral point of the aircraft is the c.g. position giving zero dC
M
/dC
L
.
Hence,
h
n
= h
0
+V
a
1
a
_
1
d
d
_
.
h
0
is the neutral point of the wings, body and nacelle of the aircraft. Therefore, the addition
of a tail has moved the neutral point of the aircraft aft by an amount V (a
1
/a)(1 d/d).
2.4 Stick free stability
So far, we have examined the elevator angles required to trim, and the effect of the tailplane
on the overall longitudinal stability of the aircraft. However, the amount of force that the pilot
(or hydraulic system etc.) needs to exert has been ignored. Clearly, for an aircraft to be as
effective as possible these factors must be considered.
The rst step in this analysis is to examine the stick free stability of the aircraft. The
phrase stick free indicates that the pilot has released the controls. Therefore, the elevator
may move to any location. In the equilibrium condition, the controls will settle in a location
where the hinge moments are zero (ignoring frictional effects).
Note that stick-free conditions only affect the elevator. Typically the tailplane,
T
, and
the tab, , are locked into position and will not move if the pilot releases the stick.
2.5 Analysis
The basic hinge moment equation is:
C
H
= b
0
+b
1
T
+b
2
+b
3
.
If the elevator is free (stick free), then C
H
=0. We can therefore rearrange to calculate
the elevator angle, , for the stick free case.
=
b
0
+b
1
T
+b
3
b
2
.
From our earlier analysis (and from the equation sheet),
C
L
T
=
a
1
a
_
1
d
d
_
C
L
+a
1
(
T
0
) +a
2
+a
3
.
So, substituting for results in:
C
L
T
=
a
1
a
_
1
d
d
_
C
L
+a
1
(
T
0
)
a
2
b
2
(b
0
+b
1
T
+b
3
) +a
3
.
For a given ight condition we know everything on the right hand side of this expression
except
T
. Fortunately, weve already worked out an expression for (2.2, page 18)
T
=
C
L
a
_
1
d
d
_
+ (
T
0
).
Collecting terms and rearranging results in:
C
L
T
=
_
a
1
a
2
b
1
b
2
__
1
d
d
_
C
L
a
+
_
a
1
a
2
b
1
b
2
_
(
T
0
)
+
_
a
3
a
2
b
3
b
2
_

a
2
b
0
b
2
.
This is obviously quite a complex expression. However, we can make it more straightfor-
ward by dening two new variables, a
1
and a
3
where:
a
1
= a
1
_
1
a
2
b
1
a
1
b
2
_
,
a
3
= a
3
_
1
a
2
b
1
a
3
b
2
_
(both of these expressions are given on the equation sheet).
If we substitute these expressions back into the equation for C
L
T
we get:
C
L
T
= a
1
_
1
d
d
_
C
L
a
+a
1
(
T
0
) +a
3

a
2
b
0
b
2
.
2.6. STICK FREE STABILITY 21
We now have an expression that allows us to calculate the lift generated by the tailplane if
the stick is released and the elevator stabilises such that the elevator hinge moments are zero.
We already have equation 2.3 which enables us to calculate the lift required at the tailplane.
Therefore, since:
C
M
= C
M
0
(h
0
h)C
L
V C
L
T
,
we can substitute for C
L
T
to get:
C
M
= C
M
0
(h
0
h)C
L
V
_
a
1
a
_
1
d
d
_
C
L
+a
1
(
T
0
) +a
3

a
2
b
0
b
2
_
.
This expression allows us to calculate the tab angle required to trim with zero stick force, .
Remember, for trim C
M
=0.
As previously, to examine stick free stability we differentiate with respect to C
L
.
dC
M
dC
L
= (h
0
h) V
a
1
a
_
1
d
d
_
.
The static margin stick free,
K
n
=
dC
M
dC
L
= (h
0
h) +V
a
1
a
_
1
d
d
_
.
Since the static margin (stick xed), K
n
, is:
K
n
= (h
0
h) +V
a
1
a
_
1
d
d
_
the two static margins will be the same if:
a
1
= a
1
= a
1
_
1
a
2
b
1
a
1
b
2
_
that is, if
a
2
b
1
a
1
b
2
= 0.
a
1
and a
2
are both positive, since increasing either the incidence of the tailplane or the elevator
deection increases the lift generated by the tailplane. b
2
must be negative for correct feel
of elevator (i.e. when the stick is pulled, it pushes back at you). b
1
can be made zero by using
aerodynamic balances or by moving the hinge line of the elevator (which also affects b
2
).
There are 3 possible cases for the stick free case. These are shown in gure 2.4.
a
2
b
1
/a
1
b
2
> 0 a
2
b
1
/a
1
b
2
< 0 a
2
b
1
/a
1
b
2
= 0
Figure 2.4: Stick free elevator conditions
1. If a
2
b
1
/a
1
b
2
> 0 (i.e. b
1
is negative), the aircraft becomes less stable in the stick free
case. This is known as having a convergent elevator.
2. If a
2
b
1
/a
1
b
2
< 0 (i.e. b
1
is positive), the aircraft becomes more stable on freeing the
stick. This is known as having a divergent elevator.
3. If a
2
b
1
/a
1
b
2
= 0 (i.e. b
1
= 0), the aircraft is equally stable in in both stick- xed and
stick-free cases. This is known as having a null elevator.
The neutral point stick free h
n
and static and c.g. margins stick free H
n
are precisely
analogous to the stick xed cases.
Chapter 3
Flight testing
3.1 Introduction
So far, we have used analytical techniques to examine the static stability of aircraft, and it is
possible to use semi-empirical techniques to estimate parameters such as a
1
, a
2
, b
1
, b
2
(see
supplementary sheet). However, when we have designed an aircraft, how do we check that its
stability characteristics are as we have calculated?
For most loadings or ight conditions our estimates will be adequate, but we need to
know what the allowable loads and c.g. range of the aircraft are. If our estimates are too
conservative, we lose potential revenue due to limitations on how the aircraft can be loaded.
Conversely, if we have estimated that the aircraft is more stable than it actually is, then the
aircraft may at some time be loaded in such a way that it is unstable or unsafe.
It is therefore necessary to be able to measure the stability of an aircraft in ight. This
forms one part of the ight test course that will be undertaken in Semester 2, during the Group
Design and Business Project.
3.2 K
n
elevator angle to trim
In trim conditions, the pitching moment acting on the aircraft, and hence C
M
, are zero. We
have already shown that (on equation sheet):
C
M
= C
M
0
(h
0
h)C
L
V C
L
T
and,
C
L
T
=
a
1
a
_
1
d
d
_
C
L
+a
1
(
T
0
) +a
2
+a
3
.
Hence,
C
M
= 0 = C
M
0
(h
0
h)C
L
V
_
a
1
a
_
1
d
d
_
C
L
+a
1
(
T
0
) + a
2
+a
3
_
.
23
24 CHAPTER 3. FLIGHT TESTING
We have previously shown (2.3) that this can be differentiated to examine the stick-xed
stability of an aircraft such that:
K
n
H
n
=
C
M
C
L
= (h
0
h) +V
a
1
a
_
1
d
d
_
.
We have also shown (Example Question 2.3) that the pitching moment equations can be rear-
ranged to get the elevator angle to trim, :
=
1
V a
2
_
C
M
0
(h
0
h)C
L
V
_
a
1
a
_
1
d
d
_
C
L
+a
1
(
T
0
) +a
3
__
and that there is a relationship between and K
n
because:
d
dC
L
=
1
V a
2
_
(h
0
h) + V
a
1
a
_
1
d
d
__
=
K
n
V a
2
.
But what do these relationships look like? is plotted against C
L
in gure 3.1 and the
relationship between d/dC
L
and the location of the centre of gravity is shown in gure 3.2.
C
L

h
1
h
2
h
3
c.g. forward
Figure 3.1: Elevator angle to trim at varying lift coefcients
Therefore, if the aircraft is own straight, level and trimmed, and the elevator deection is
measured, a point on one of the lines on Figure 3.1 can be plotted. If the incidence and speed
of the aircraft are then changed such that the aircraft is trimmed at a different lift coefcient
another point at the same centre of gravity location can be measured. This can be repeated
further to measure more points for the same c.g. location.
If the aircraft is now landed, the location of the centre of gravity can be moved by a
number of means; the fuel might be added to a different fuel tank, passengers might be loaded
to different parts of the aircraft or luggage/freight might be loaded to different areas of the
aircraft. The aircraft can now be trimmed at a number of speeds for this centre of gravity
3.3. WHAT DOES THE PILOT FEEL? 25
h
d /dC
L
h
1
h
2
h
3
Figure 3.2: Measurement of neutral point location
location, thus resulting in a second line (of different gradient) on the C
L
vs graph. If the
gradients of these lines (i.e. d/dC
L
) are now plotted against the c.g. location, as in gure 3.2,
the stick xed neutral point location can be found since it is the c.g. location where d/dC
L
,
and hence K
n
, is zero.
3.3 What does the pilot feel?
Something that is worth noting is that this is not what the pilot will experience. Although the
pilot will have a good feel for the elevator deection (it will generally be directly proportional
to the distance he/she has pushed/pulled the stick) the concept of a lift coefcient is somewhat
contrived. The pilot might, however, note how the elevator angle varies with ight speed.
Lets have a look at what the pilot will experience, then.
The variation of elevator angle with speed can easily be found from the derivation that we
have already done, since:
d
dV
=
d
dC
L
dC
L
dV
.
By denition,
C
L
=
L
V
2
S/2
.
Hence,
dC
L
dV
=
2L
V
3
S/2
=
2C
L
V
.
Therefore,
d
dV
=
2C
L
V
d
dC
L
=
2C
L
V
K
n
V a
2
.
which is sketched in gure 3.3.
V

c.g. moving forward
Figure 3.3: What the pilot experiences
3.4 K
n
tab angle to trim
As normal, we start with the standard pitching moment equation and we are looking at the
case where we are trimmed, hence:
C
M
= C
M
0
(h
0
h)C
L
V C
L
T
= 0
and
C
L
T
=
a
1
a
_
1
d
d
_
C
L
+a
1
(
T
0
) +a
2
+a
3
.
From 2.5:
=
b
0
+b
1
T
+b
3
b
2
and C
L
T
=
a
1
a
_
1
d
d
_
C
L
+a
1
(
T
0
) +a
3

a
2
b
0
b
2
Hence
C
M
= C
M
0
(h
0
h)C
L
V
_
a
1
a
_
1
d
d
_
C
L
+a
1
(
T
0
)
a
2
b
0
b
2
_
If we rearrange to nd the tab angle to trim, , where C
M
=0 we get:
=
1
V a
3
_
C
M
0
(h
0
h)C
L
V
_
a
1
a
_
1
d
d
_
C
L
+a
1
(
T
0
) + a
3

a
2
b
0
b
2
__
3.4. K
N
TAB ANGLE TO TRIM 27
Differentiating with respect to C
L
results in:
d
dC
L
=
1
V a
3
_
h
0
h +V
a
1
a
_
1
d
d
__
.
But from 2.6:
K
n
= (h
0
h) +V
a
1
a
_
1
d
d
_
.
Hence:
d
dC
L
=
K
n
V a
3
.
Therefore, there is a relationship between the stick free static margin, K
n
, and the tab
angle to trim, , in exactly the same way as there is a relationship between the stick xed static
margin and the elevator angle to trim. So, we can use a very similar technique to measure the
stick free neutral point, h
n
, as we would to measure the stick xed manoeuvre point, h
n
.
Sketches of the variation of with lift coefcient and the variation of d/dC
L
with c.g.
location are shown in gure 3.4 and gure 3.5 respectively.
C
L
h
1
h
2
h
3
Figure 3.4: Tab angle to trim at varying lift coefcients
The ight tests to measure the stick free static margin are therefore undertaken in exactly
the same way as those to measure the stick xed static margin, except the tab angle to trim
with zero stick force is measured instead of the elevator angle.
h
d
/dC
L
h
1
h
2
h
3
Figure 3.5: Measurement of stick free neutral point location
Chapter 4
Tailless aircraft
4.1 Introduction
So far, we have only examined conventional aircraftthose with a tailplane aft of the main
wing. This analysis can easily be adapted to examine canard congured aircraft, where
the balancing lifting surface is known as a foreplane and is positioned forward of the main
wing. For such congurations, the basic equations are the same, but the distance, l, from
the aerodynamic centre of the wing/body/nacelle to the foreplane is negative, as shown in
gure 4.1.
L
F
L
WBN
l
M
0
W
h c
h
0
c
Figure 4.1: Canard conguration
The tailless conguration has no lifting surface other than the main wing. An obvious
example is Concorde. A number of military aircraft are tailless, such as the Avro Vulcan,
although in recent years close coupled canard-deltas have become popular. Examples of
these aircraft are Euroghter Typhoon and the Saab Gripen. The control surfaces used for
tailless aircraft are shown in gure 4.2.
29
30 CHAPTER 4. TAILLESS AIRCRAFT
Elevon
Rudder
Figure 4.2: Control surfaces for tailless aircraft: the elevons operate together for pitch control
and differentially for roll
4.2 Stick xed stability
There are two key differences between the analysis of conventional and tailless aircraft. The
rst, and most obvious, is that there is no tailplane and thus no terms relating to tailplane lift.
Secondly, in addition to deections of the elevons causing a change in lift, they also generate
a large pitching moment. This can cause complications, especially on landing.
The representation of a tailless aircraft used for examining static stability is given in g-
ure 4.3
L
WBN
M
0
W
h c
h
0
c
Figure 4.3: Representation of tailless aircraft
4.3. STATIC MARGIN 31
The total lift generated by a tailless aircraft is broken down in the same way as for the
tailplane of a conventional aircraft so that:
C
L
= a
1
+a
2
.
Tailless aircraft very rarely have tabs, since the controls are almost always powered, hence the
reduction of stick forces by using a tab is rarely necessary. The analysis of tailless aircraft is
straightforward. If we take moments about the centre of gravity:
M
cg
= M
0
+
M
0
(h
0
h)c
0
L
Non-dimensionalizing results in:
C
M
= C
M
0
+
C
M
0
(h
0
h)C
L
.
As for conventional aircraft, this expression can be rearranged to calculate the elevon angle
required to trim.
4.3 Static margin
As for conventional aircraft, we differentiate the pitching moment equation with respect to C
L
to calculate the static margin:
dC
M
dC
L
= (h
0
h).
Therefore,
K
n
= h
0
h.
The stick free case for tailless aircraft is even more straightforwardsince in practice there
isnt a stick free case!
Due to the size of the elevons, and the size of typical aircraft, tailless aircraft almost always
have powered controls, which do not have a stick free condition because they are held in
place by hydraulic systems. Therefore, within the scope of this course it is not necessary to
consider the stick free case.
Chapter 5
Stick forces
5.1 Introduction
All the analysis that we have undertaken to date has ignored the stick force required to trim
an aircraft. Obviously, this is an important criterion since it will directly affect whether a pilot
can hold the aircraft in trim for any length of time. Also, when we consider the stick forces
required for manoeuvring, later in the course, we need to ensure that the forces required are
realistic. But what is a reasonable force? Table 5.1 indicates the maximumforces, in Newtons,
that can be applied to the controls for different lengths of time.
Aileron Elevator Rudder
Stick Wheel Stick Wheel (Push)
Maximum all-out effort 2 hands 400 530 800 980 1780 N
Maximum permissible effort 2 hands 360 440 440 890 N
1 hand 220 220 310 310 N
Maximum comfortable effort 2 hands 130 180 270 N
1 hand 90 90 130 130 N
Largest full travel 254 508 230 230 126 mm
Table 5.1: Maximum control forces
It is worth noting that the elevator push/pull force is always higher than the maximum
aileron force and that the maximum rudder pedal force is higher than both of the stick forces.
This information is used to harmonise the controls. The controls are said to be harmonised
if the aileron, elevator and rudder force ratios are 1:2:4 to achieve equal effect, dened as
response in angular rate (i.e. the rudder pedal force required for a 10
/s yaw is double the

stick force for a 10
/s pitch, which in turn is double the force for a 10
/s roll).
5.2 Analysis to calculate stick forces
The stick force to trim, generally given the symbol P
e
, is equal to the hinge moment multiplied
by the gearing ratio (m
e
, between the angular movement of the elevator and the distance the
33
34 CHAPTER 5. STICK FORCES
stick moves). Hence:
P
e
= m
e
V
2
2
S
C
H
We therefore need to nd the hinge moment required to trim, C
H
. We have already found
a way to calculate the tab angle required to trim for zero stick force (2.5). Calculating the
stick force required at different tab angles is very similar.
We start, as for the zero stick force (stick free) case, with the denition of the hinge mo-
ment, C
H
:
C
H
= b
0
+b
1
T
+b
2
+b
3
.
For the stick free case, C
H
=0, which is not now true. However, we can still rearrange to
nd the elevator angle as a function of the hinge moments:
=
C
H
b
0
b
1
T
b
3
b
2
.
We also know, from the equation sheet, that:
C
L
T
= a
1
_
1
d
d
_
C
L
a
+a
1
(
T
0
) +a
2
+a
3
.
If we substitute into this equation we get, after substituting for
T
and using the denitions
of a
1
and a
3
:
C
L
T
= a
1
_
1
d
d
_
C
L
a
+a
1
(
T
0
) +a
3
+
a
2
b
2
(C
H
b
0
).
This is the more general case of an equation that we have already derived for C
H
=0 (i.e.
the stick free case). We know an equation that links the lift at the tailplane to the pitching
moments:
C
M
= C
M
0
(h
0
h)C
L
V C
L
T
.
Substituting for C
L
T
in this equation results in:
C
M
= C
M
0
(h
0
h)C
L
V
_
a
1
a
_
1
d
d
_
C
L
+a
1
(
T
0
) +a
3
+
a
2
b
2
(C
H
b
0
)
_
.
This can easily be re-arranged to nd the tab angle to trim with zero hinge moments, :
V a
3
= C
M
0
(h
0
h)C
L
V
_
a
1
a
_
1
d
d
_
C
L
+a
1
(
T
0
)
a
2
b
0
b
2
_
(5.1)
Alternatively, we can rearrange to get the hinge moments required to trim:
V
a
2
C
H
b
2
= C
M
0
(h
0
h)C
L
V
_
a
1
a
_
1
d
d
_
C
L
+a
1
(
T
0
) +a
3

a
2
b
0
b
2
_
(5.2)
5.3. MORE FLIGHT TESTING 35
We could use this expression to calculate the elevator hinge moments for an arbitrary
trimmed ight condition, but there is actually a shortcut to doing this. If we subtract equa-
tion 5.2 from equation 5.1 most of the terms cancel and we get:
V
_
a
3

a
2
b
2
C
H
_
= V a
3
which can be rearranged to get:

C
H
=
b
2
a
2
a
3
( )
and hence:
P
e
= m
e
V
2
2
S
b
2
a
2
a
3
( ).
Therefore, to calculate the stick force required to trim simply calculate the tab angle to trim
stick-free and use this equation.
5.3 More ight testing
It is theoretically possible to use the stick forces to calculate the stick free neutral point. We
will only go through this briey here, since in practice this method is problematic due to
friction between the stick and the elevator.
We have already shown that:
C
H
=
b
2
a
2
a
3
( ).
Differentiating with respect to lift coefcient, if the tab angle is held constant, results in:
C
H
C
L
=
b
2
a
2
a
3
C
L
but from 3.4:
d
dC
L
=
K
n
V
1
a
3
.
Hence,
dC
H
dC
L
=
b
2
K
n
V a
2
Therefore, by measuring the hinge moments (stick forces) at different ight conditions and
repeating for different c.g. positions (in a similar way to that used for tab angle and elevator
angle measurements of the neutral points) we can measure the location of the stick-free neutral
point.
36 CHAPTER 5. STICK FORCES
5.4 Modication of stick forces
Three methods are available to decrease the pilots effort, bearing in mind that the pilot must
retain some feel of the control forces:
Gearing between stick and control surface. This has limited application since the gearing is
restricted by stick travel limits.
Power assistance, either by load sharing or by using full power assistance with articial
feel.
Aerodynamic assistance can be used, by using horn balances, by moving the hinge line of
the elevator or by using balance/anti balance tabs (see gure 5.1).
Hinge line
Horn balance
S
a: horn balance b: hinge location

c: geared tab d: anti-balance tab
Figure 5.1: Aerodynamic assistance
What are we trying to do when we use aerodynamic balancing? Effectively, we are trying
to reduce the control force (hinge moment) required for a given elevator deection (i.e. we are
trying to reduce dP
e
/d):
P
e
= m
e
V
2
2
S
C
H
,
but,
C
H
= b
0
+b
1
T
+b
2
+b
3
,
5.4. MODIFICATION OF STICK FORCES 37
so that
dP
e
d
= m
e
V
2
2
S
b
2
.
Therefore, to achieve aerodynamic balance we are trying to reduce b
2
. However, dP
e
/d
and hence b
2
, must be negative for the correct feel of the controls. Therefore, reduction of b
2
enables us to reduce the control forces. This is useful at high speeds. However, at low speeds
this solution may give inadequate feel for the pilot. Therefore, aerodynamic assistance can
also be limited.
Chapter 6
Manoeuvre stability
We have now completed our analysis of straight and level static stability. The next step is
to examine longitudinally symmetric manoeuvres (i.e. manoeuvres that affect the left and
right hand side of the aircraft equally). The most straightforward example of this is a steady
pullout at constant velocity. Somewhat surprisingly, the elevator angle required for pitch
trim in a steady banking turn can also be calculated in the same way. This is because the
radius of a typical banked turn is very large. Hence, the asymmetry in the ow is small once
the turn has been initiated.
6.1 Analysis of a steady pullout
Consider two conditions, shown in gure 6.1:
1. The aircraft is in steady, level ight at speed V .
2. The aircraft is in a steady pullout at speed V .
In the steady pullout the aircraft has a radial (centripetal) acceleration V
2
/r = ng. The
difference in lift between (1) and (2) is nmg = nW. Hence:
C
L
= nC
L
=
nW
V
2
S/2
where C
L
is the lift coefcient in the straight and level case.
The other key difference between the two cases is that in the steady pullout the aircraft
has an angular (pitch) velocity as well as a linear velocity. This angular velocity can easily be
found by considering the amount of time that the aircraft would take to complete a full circle
at constant speed. If the aircraft completed a full circle it would pitch though 2 radians and
would therefore cover a distance of 2r, where r is the radius of the circle. The time taken, t,
at speed V would be:
t =
2r
V
39
40 CHAPTER 6. MANOEUVRE STABILITY
V
C
L1
, L = W
W = mg
C
L2
= (1 +n)C
L1
, L = (1 +n)W
W = m(1 +n)g
r
1: Steady, level ight 2: Steady pullout
Figure 6.1: Manoeuvre conditions
but by considering the centripetal acceleration we also know that:
r =
V
2
ng
.
Hence,
t =
2V
ng
.
The aircraft has pitched through a total angle of 2 radians in this time. Therefore the pitch
rate, q, is:
q =
2
2V/ng
=
ng
V
.
But why is this pitch rate, q, important? The pitch rate will cause the tail of the aircraft to
move down relative to the incoming air velocity. This causes the incidence at the tailplane to
increase by an amount:
T
=
ql
T
V
where l
T
is the tail arm measured from the centre of gravity. We already have an expression
for q, so we get:
T
=
ngl
T
V
2
.
Unfortunately, this expression has a V
2
term, and hence will vary with the ight conditions.
We can get rid of this awkward term by applying the denition of the lift coefcient, C
L
, for
6.1. ANALYSIS OF A STEADY PULLOUT 41
the straight and level case:
C
L
=
W
V
2
S/2
.
Hence
V
2
=
W
SC
L
/2
.
Substituting this back into the expression for
T
results in:
T
=
gSl
T
W
nC
L
2
or
T
=
nC
L
2
1
.
where
1
= W/gSl
T
and is known as the longitudinal relative density.
The change in incidence of the tailplane causes the lift coefcient of the tailplane to alter
by an amount a
1
T
.Therefore, remembering that the lift coefcient of the aircraft in the
steady pullout is (1 +n)C
L
:
C
L
T
=
a
1
a
_
1
d
d
_
(1 +n)C
L
+ a
1
(
T
0
) +a
2
+a
3
+a
1
T
.
Hence,
C
L
T
=
a
1
a
_
1
d
d
_
(1 +n)C
L
+a
1
(
T
0
) +a
2
+a
3
+ a
1
nC
L
2
1
.
The basic pitching moment equation is still valid, since it makes no assumptions about the
source of the lift and momentsit is simply the result of non-dimensionalising a free body
diagram. Therefore, this revised expression for C
L
T
can be substituted. Again, remembering
that the lift coefcient in the steady pullout is (1 +n)C
L
:
C
M
= C
M
0
(h
0
h)(1 +n)C
L
V
_
a
1
a
_
1
d
d
_
(1 +n)C
L
+a
1
(
T
0
) +a
2
+a
3
+a
1
nC
L
2
1
_
.
For the straight and level ight of the aircraft, in trim, we have previously derived the
equation:
C
M
= 0 = C
M
0
(h
0
h)C
L
V
_
a
1
a
_
1
d
d
_
C
L
+a
1
(
T
0
) + a
2
+a
3
_
.
(6.1)
In trim, the pitching moments acting on the manoeuvring aircraft will be zero if the aircraft
is undertaking a steady manoeuvre. If we now look at the expression for trim in a steady
pullout, and look at the change in elevator angle required for trim, such that the elevator angle
is now + we get:
0 = C
M
0
(h
0
h)(1 +n)C
L
V
_
a
1
a
_
1
d
d
_
(1 +n)C
L
+a
1
(
T
0
) +a
2
( +) +a
3
+a
1
nC
L
2
1
_
. (6.2)
Equations 6.1 and 6.2 are very similar. We can therefore perform a similar mathematical
trick to the one we used to get the stick forces to trim: we subtract one equation from the other.
If we subtract equation 6.1 from equation 6.2 we get:
0 = (h
0
h)nC
L
V
_
a
1
a
_
1
d
d
_
nC
L
+a
1
nC
L
2
1
+a
2
_
. (6.3)
This can be rearranged to get the elevator deection/g required for a steady pullout:
n
=
C
L
V a
2
_
(h
0
h) +V
_
a
1
a
_
1
d
d
_
+
a
1
2
1
__
.
This must always be negative, otherwise the pilot would pitch nose-down on pulling back
on the stick.
6.2 Stick xed manoeuvre point
When /n = 0 the c.g. is at the stick xed manoeuvre point. Hence, at h = h
m
:
0 =
C
L
V a
2
_
(h
0
h
m
) +V
_
a
1
a
_
1
d
d
_
+
a
1
2
1
__
,
h
m
= h
0
+V
_
a
1
a
_
1
d
d
_
+
a
1
2
1
_
.
This should be compared with the neutral point location stick xed, h
n
, which we have previ-
ously shown to be:
h
n
= h
0
+V
a
1
a
_
1
d
d
_
.
Therefore, the stick xed manoeuvre point is a distance a
1
c/2
1
aft of the stick xed
neutral point. It is worth noting that the location of the stick xed manoeuvre point varies
with altitude, since
1
is a function of the air density as well as of geometry.
6.3. STICK FIXED MANOEUVRE STABILITY 43
6.3 Stick xed manoeuvre stability
The stick xed manoeuvre margin, H
m
, is dened by:
H
m
= h
m
h.
We showed in 6.1 that:
n
=
C
L
V a
2
_
(h
0
h) +V
_
a
1
a
_
1
d
d
_
+
a
1
2
1
__
.
Hence,
h = h
0
+
V a
2
C
L
n
+V
_
a
1
a
_
1
d
d
_
+
a
1
2
1
_
Therefore,
H
m
=
V a
2
C
L
n
.
This result is important because it demonstrates that there is a relationship between the stick
xed manoeuvre margin and the elevator angle to trim. There is, seemingly, a discrepancy
between the fact that h
m
moves with changing altitude and the above expression. How can
this discrepancy be explained?
6.4 Stick free manoeuvre stability
Stick xed analysis has enabled us to calculate the elevator angles required to trimthe aircraft
in a steady pullout/bank, but tells us nothing about the stick forces required for the manoeuvres
(just as stick xed analysis told us nothing of the stick forces for straight and level ight).
Stick free manoeuvre stability analysis will allow us to calculate these stick forces.
6.5 Analysis
In 5.2, we derived an expression that allowed us to calculate the hinge moments for trim in
straight and level ight:
C
M
= 0 = C
M
0
(h
0
h)C
L
V
_
a
1
a
_
1
d
d
_
C
L
+a
1
(
T
0
) +a
3
+
a
2
b
2
(C
H
b
0
)
_
.
The same process can be undertaken to nd the hinge moments for trim in a steady pullout,
C
H
+ C
H
. For the pullout, assuming that the tab is not used:
C
M
= 0 = C
M
0
(h
0
h)(1 +n)C
L
V
_
a
1
a
_
1
d
d
_
(1 +n)C
L
+a
1
(
T
0
) +a
3
+a
1
nC
L
2
1
+
a
2
b
2
(C
H
+C
H
b
0
)
_
.
where C
L
is, again, the lift coefcient in straight and level ight.
Equating these two expressions and cancelling the identical terms results in:
0 = (h
0
h)nC
L
V
_
a
1
a
_
1
d
d
_
nC
L
+a
1
nC
L
2
1
+
a
2
C
H
b
2
_
.
Hence,
V a
2
b
2
C
L
C
H
n
= (h
0
h) V
_
a
1
a
_
1
d
d
_
+
a
1
2
1
_
.
The stick free manoeuvre point, h
m
, is dened as the c.g. position that gives C
H
/n = 0.
Therefore,
h
m
= h
0
+V
_
a
1
a
_
1
d
d
_
+
a
1
2
1
_
.
The stick free manoeuvre margin, H
m
, is dened as:
H
m
= h
m
h
which can easily be shown to be:
H
m
=
V a
2
b
2
C
L
C
H
n
.
The stick force per g is calculated from the hinge moment per g, in exactly the same way
as for straight and level ight:
P
e
n
= m
e
V
2
2
S
C
H
n
.
For handling safety the stick force required to pull high g should be appreciable to avoid
accidentally exceeding the structural limitations of the aircraft. A typical value for a non-
aerobatic aircraft is usually of the order of 20 N/g.
6.6 Tailless aircraft
The analysis for tailless aircraft is very similar to that for conventional aircraft. The ight
conditions, as for conventional aircraft, are shown in gure 6.2.
For a tailless aircraft in steady trimmed ight we have already derived the equation (4.2):
C
M
= 0 = C
M
0
+
C
M
0
(h
0
h)C
L
.
When the aircraft is in a steady pullout with radial acceleration ng and with pitch rate q we
can write:
C
M
= 0 = C
M
0
+
C
M
0
( + ) (h
0
h)(1 +n)C
L
+
C
M
q
q.
6.6. TAILLESS AIRCRAFT 45
V
C
L1
, L = W
W = mg
V
C
L2
= (1 +n)C
L1
, L = (1 +n)W
W = m(1 +n)g
r
q = ng/V
1: Steady, level ight 2: Steady Pullout
Figure 6.2: Manoeuvre conditions for a tailless aircraft
Again, we can use the trick of subtracting one equation from the other to get:
=
1
C
M
0
/
_
(h
0
h)nC
L
C
M
q
q
_
.
C
M
/q is what is known as an aerodynamic derivative. There are a large number of
aerodynamic derivatives that can be dened for any aircraft, and they enable us to calculate
the aerodynamic behaviour of the aircraft. We will encounter more aerodynamic derivatives
when we examine the dynamic stability of aircraft. There is a standard non-dimensionalised
form for each of these parameters. The non-dimensional form of C
M
/q is given the symbol
m
q
, and is dened as:
m
q
=
1
V Sc
2
0
M
q
.
Hence,
C
M
q
=
1
V
2
Sc
0
/2
M
q
=
2c
0
V
m
q
and we know that
q =
ng
V
.
Therefore,
=
1
C
M
0
/
_
(h
0
h)nC
L
2c
0
V
m
q
ng
V
_
.
As for the conventional aircraft, we have an expression that includes the ight velocity.
Again, we can remove this by using the denition of the lift coefcient and rearranging such
that:
V
2
=
W
SC
L
/2
.
Making this substitution results in:
=
1
C
M
0
/
_
(h
0
h)nC
L
m
q
nC
L
gc
0
S
W
_
.
The longitudinal relative density,
1
, for a tailless aircraft is dened as:
1
=
W
gSc
0
Using this denition and rearranging results in:
n
=
1
C
M
0
/
_
(h
0
h)
m
q
1
_
C
L
.
This expression can be used to calculate the elevator deections required to undertake ma-
noeuvres.
6.7 Tailless aircraft manoeuvre point
As for the conventional aircraft, the manoeuvre point is dened by the c.g. location that results
in /n = 0. Therefore,
h
m
= h
0
m
q
1
.
The resulting aerodynamic forces due to a positive pitch rate are shown in gure 6.3.
Forces oppose motion
Figure 6.3: Aerodynamic forces during pitching motion
These forces all oppose the motion of the aircraft. Hence m
q
is always negative. This
means that the manoeuvre point for a tailless aircraft is always aft of the neutral point for the
aircraft (which is at h = h
0
). The damping in pitch has therefore increased the stability of the
aircraft.
6.8. TAILLESS AIRCRAFT MANOEUVRE MARGINS 47
6.8 Tailless aircraft manoeuvre margins
The manoeuvre margin for a tailless aircraft, H
m
, is dened identically to that for a conven-
tional aircraft:
H
m
= h
m
h.
Hence
H
m
= (h
0
h)
m
q
1
= K
n
m
q
1
.
The elevon angle per g can therefore be written as:
n
=
H
m
C
L
C
M
0
/
.
The elevon angle per g is therefore directly proportional to the manoeuvre margin.
6.9 Relationships between static and manoeuvre margins
Conventional aircraft
We have shown that the static margins, stick xed and stick free, for conventional aircraft are:
K
n
= (h
0
h) +V
a
1
a
_
1
d
d
_
K
n
= (h
0
h) +V
a
1
a
_
1
d
d
_
.
Also, the manoeuvre margins for conventional aircraft are:
H
m
= (h
0
h) +V
_
a
1
a
_
1
d
d
_
+
a
1
2
1
_
H
m
= (h
0
h) +V
_
a
1
a
_
1
d
d
_
+
a
1
2
1
_
.
Therefore,
H
m
= K
n
+
V a
1
2
1
H
m
= K
n
+
V a
1
2
1
.
The manoeuvre points of conventional aircraft are aft of the respective neutral points. This
is due to the stabilising inuence of additional lift at the tailplane due to the pitch rate. Note
that since
1
is a function of the air density the manoeuvre margin decreases with increasing
altitude.
Tailless aircraft
We have shown that the static margin for tailless aircraft is:
K
n
= h
0
h
and that the manoeuvre margin is:
H
m
= (h
0
h)
m
q
1
.
Therefore,
H
m
= K
n
m
q
1
.
As for conventional aircraft, a tailless aircraft is more stable when manoeuvring due to
the stabilising effect of the pitch damping term m
q
(remember, m
q
is negative). Again, the
manoeuvre margin is reduced at high altitudes due to the presence of a density term in
1
.
6.10 Modication of stick free neutral and manoeuvre
points
Two common ways of modifying the stick free neutral and manoeuvre points are shown in
gure 6.4.
Bob weight
a: Spring b: Bob weight
Figure 6.4: Modication of neutral and manoeuvre points
A spring or bob-weight is dened as positive if it exerts a moment that would cause a
positive deection of the elevator.
These two additions have no effect on the stick-xed neutral or manoeuvre points since the
calculation of these locations does not require the consideration of hinge moments. However,
it can be shown that a positive spring moves the stick free neutral point aft but has no effect
on the stick free manoeuvre point. In contrast, a positive bob-weight results moves both the
stick free neutral point and the stick free manoeuvre point of an aircraft aft. These effects are
shown in gure 6.5.
By combining positive and negative springs and bob-weights it is possible to move the two
stick free points independently of each other. Since the stick forces are directly proportional
6.10. MODIFICATION OF STICK FREE NEUTRAL AND MANOEUVRE POINTS 49
N M
Spring
Bob-weight
N
Figure 6.5: Effects of positive springs and bob-weights

to the stick free static margin and the stick free manoeuvre margin this enables the stick forces
to be modied by a simple mechanical addition to the system.
For example, an aircraft might have suitable levels of stick free static stability but insuf-
cient manoeuvre margins. This results in a stable aircraft with good feel for the pilot and
suitable stick loads for trim, but the low stick force per g resulting from the low manoeuvre
margin might cause a risk of inadvertently overstressing the aircraft. The addition of a nega-
tive spring together with a positive bob-weight would solve this problem since the stick free
static margin would be unchanged but the stick free manoeuvre margin would increase. This
is shown in gure 6.6.
Positive bob-weight
Negative spring
N
Figure 6.6: Modication of stick force per g

Chapter 7
Compressibility effects
Everything that we have examined so far has assumed:
linear aerodynamics;
incompressible ow;
rigid aircraft.
In reality, of course, none of these assumptions will be valid at all ight conditions. At higher
angles of attack the aerodynamics become non-linear (e.g. C
L
not proportional to ) and as
the aircraft ies faster other effects become important. In this section we will briey outline
the major changes that occur at high speeds, and the effect that this has on the control of the
aircraft.
7.1 High speed effects
Changes from the low speed case arise primarily from Mach number (compressibility) and
distortion (aeroelastic) effects. The dominant effects of increasing Mach number come from:
1. change of lift curve slope with Mach number;
2. movement of aerodynamic centre rearwards, moving from quarter-chord at low speed
to mid-chord at supersonic Mach numbers.
These effects are shown in gure 7.1.
Combined with these are the effects of Mach number on zero-lift pitching moment and
zero-lift angle.
The effects of aeroelasticity are to reduce lift curve slopes with increasing V
2
/2, dynamic
pressure. The loads acting on an aircraft are proportional to the dynamic pressure, if the lift
and drag coefcients are constant. The deections are, similarly, proportional to the forces.
Hence, all aeroelastic effects are dependent on the dynamic pressure. This results in changes
in the aeroelastic response of the aircraft at different altitudes, since the ambient air density
51
52 CHAPTER 7. COMPRESSIBILITY EFFECTS
M
a
1.0 M
h
1.0
c/4
c/2
Figure 7.1: Compressibility effects on lift curve slope and aerodynamic centre
M
C
M0
1.0 M
0
1.0
Figure 7.2: Compressibility effects on zero lift pitching moment and zero lift angle
varies with altitude. If we superimpose altitude effects onto the variation of lift curve slope
with Mach number for a typical (aft) swept wing aircraft, we get a result as in gure 7.3.
There are similar effects on the effectiveness of the tailplane and the elevator, as shown in
gure 7.4.
The downwash at the tail typically varies as shown in gure 7.5.
It decreases to zero at high supersonic speeds, since any downwash is conned to the
volume of air inuenced by the wing.
The usual result of the combined effects is that stability reduces as the Mach number nears
unity and then increases, sometimes rapidly, to a higher value at supersonic speeds. The
variation of C
M
for a typical aircraft is shown in gure 7.6.
To counteract the nose down pitching moment that often occurs on swept wing aircraft
(subsonic jet transportsBoeing 707, 747, etc.) an up-elevator or stabilisation input is pro-
vided by a Mach number sensing system. This is known as Mach trim. If the nose down
moment were allowed to take effect the stick force gradient would be reversed, and there is
also a danger that the maximum allowable speed of the aircraft due to structural limits would
be exceeded. The stick forces for such an aircraft are shown in gure 7.7.
7.1. HIGH SPEED EFFECTS 53
M
a
1.0
Rigid aircraft
Sea level
Decreasing altitude
Figure 7.3: Aeroelastic effects on lift curve slope
M
V a
1
1.0
Rigid aircraft
Sea level
Decreasing altitude
M
V a
2
1.0
Rigid aircraft
Sea level
Decreasing altitude
Figure 7.4: Aeroelastic effects on tailplane and elevator
54 CHAPTER 7. COMPRESSIBILITY EFFECTS
M
/
1.0
Figure 7.5: Variation of downwash with Mach number
M
C
M
1.0
Mach tuck
Figure 7.6: Variation of pitching moment with Mach number
M
Push
Pull
1.0
Uncorrected stick force
Mach trim input
Figure 7.7: Variation of stick forces with Mach number
Part II
Dynamic stability
55
Chapter 8
Dynamic behaviour of aircraft
In the rst part of the course, we examined the static stability of aircraft, which means that we
have considered whether an aircraft tends to return to its equilibrium position after a pertur-
bation, 1.1. We are now going to analyze the dynamic stability of aircraft and see how they
respond over time to perturbations in ight.
8.1 Axes and notation
The axes and notation for the analysis of dynamic stability of an aircraft are given in Figure 8.1
and follow a logical order. Once the x, y and z-axes are dened we then have, for example,
L, M and Nthe rolling moment about the x-axis, the pitching moment about the y-axis and
the yawing moment about the z-axis respectively.
The axis system uses what are known as body axes. This axis system is not locked in
position in space, but moves with the aircraft. The origin of the axis system is at the centre of
gravity of the aircraft, since all rotations take place about the c.g.
A rigid aircraft has six degrees of freedom. To simplify the equations used when perform-
ing analysis of the dynamic modes of an aircraft, these degrees of freedom are expressed as
perturbation quantities in relation to steady straight ight (i.e. velocity perturbations u, v and
w and rotational velocities p =

, q =

and r =

).
8.2 Aerodynamic derivatives
We now have a co-ordinate system that allows us to dene any perturbation of the aircraft
from straight and level ight. To continue, we need to nd out what forces are acting on the
aircraft for a given perturbation
1
.
1
The analysis which follows is taken from MILNE-THOMSON, L. M., Theoretical aerodynamics, MacMillan
and Company, 1966.
57
58 CHAPTER 8. DYNAMIC BEHAVIOUR OF AIRCRAFT
z
x
y
Axis Perturbation Mean Perturbation Rotation Angular Moment Moment
force velocity velocity angle velocity of inertia
x X U u p A L
y Y V v q B M
z Z W w r C N
Figure 8.1: Notation for analysis of dynamic stability
If we assume that the effects of each perturbation are linear (true for small perturbations),
then:
M =
M
u
u +
M
v
v +
M
w
w +
M
p
p +
M
q
q +
M
r
r.
The partial derivatives in this expression are known as aerodynamic derivatives or stability
derivatives. We have already met the derivative M/q, often known as pitch damping, in
our analysis of control deections for tailless aircraft. Therefore, if we know the aerodynamic
derivatives and the perturbations, we can calculate all of the forces and moments acting on
the aircraft (6 equations). If, in addition, we know the mass of the aircraft and its inertia in
roll, pitch and yaw (A, B, C) we can calculate the acceleration of the aircraft, and hence its
dynamic response.
The forces on the aircraft are the aerodynamic force F and mg:
F = Xi +Y j +Zk, (8.1)
mg = mg
1
i +mg
2
j +mg
3
k, (8.2)
8.2. AERODYNAMIC DERIVATIVES 59
where the components of g are needed because the reference frame is xed to the aircraft and
is not necessarily horizontal. The other quantitites we need for the aircraft are:
v = ui +vj +wk, velocity
= pi +qj +rk, angular velocity
h = h
1
i +h
2
j +h
3
k, angular momentum.
The equations of motion in translation and rotation are then:
d
dt
(mv) = m v + (mv) = F +mg, (8.3a)
dh
dt
=

h + h = L, (8.3b)
where the boxed terms are required because the frame of reference is rotating. The applied
moment L is
L = Li +Mj +Nk.
Equations 8.3 are the general equations of motion for an aircraft and could, in principle, be
used to calculate the motion given enough information about the aerodynamics and mass dis-
tribution. We, however, want to know if the aircraft is dynamically stable, so we need to make
some approximations to see how the aircraft behaves when perturbed from steady ight.
In steady ight, we write:
v = V, = 0, F +mg = 0,
and add the small perturbation quantities so that:
V = V
1
+u;
V
1
= Ui,
u = ui +vj +wk;
= pi +qj +rk.
For a small rotation ,
= i +j +k,
=

i +

j +

k.
Similarly, the perturbation forces are:
F +F, m(g +g)
and it can be shown that
g + g = 0.
Inserting these assumptions in Equations 8.3 yields the equations of motion under small per-
turbations:
m u +m( V
1
+ g) = F, (8.4a)
h = L. (8.4b)
We can now simplify the system by making certain (reasonable) assumptions. First, we as-
sume that the forces and moments depend only on velocities and not on accelerations, with
the exception of the dependence of pitching moment on w, the downwash velocity. Then:
F = Xi +Y j +Zk,
L = Li +Mj +Nk,
and, for example,
X =
X
u
u +
X
v
v +
X
w
w +
X
p
p +
X
q
q +
X
r
r,
M =
M
u
u +
M
v
v +
M
w
w +
M
w
w +
M
p
p +
M
q
q +
M
r
r.
Secondly, we are assuming that the aircraft is symmetric so that a symmetric perturbation can
only cause a symmetric response. This means that a pitch disturbance, for example, cannot
cause a response in yaw or roll. Also, the symmetric response to an asymmetric input has to
be symmetric: if the aircraft rolls at a given rate, the pitch response must be the same whether
it rolls in a positive or negative sense. The rst of these statements implies that
Y
u
,
Y
w
,
Y
q
;
L
u
,
L
w
,
L
q
;
N
u
,
N
w
,
N
q
are all zero. From the second requirement, we can say that
X
p
,
X
q
,
X
r
;
Z
p
,
Z
q
,
Z
r
;
M
p
,
M
q
,
M
r
are likewise all zero.
Eliminating zero terms, we can write:
F = (
X
u
u +
X
w
w +
X
q
)i + (
Y
v
v +
Y
p
+
Y
r
)j
+ (
Z
u
u +
Z
w
w +
Z
q
)k,
L = (
L
p
+
L
r
+
L
v
v)i + (
M
q
+
M
u
u +
M
w
w +
M
w
w)j
+ (
N
p
+
N
r
+
N
v
v)k.
We need one more assumption about the aircraft, which is that there is no inertial coupling
between yaw and roll. This means that the only moments of inertia we need consider are A,
B and C, the moments of inertia about the coordinate axes.
8.3. LONGITUDINAL SYMMETRIC MOTION 61
Now, assuming disturbed horizontal ight and expanding the cross products in Equa-
tions 8.4 yields the equations of motion for each translational and rotational component:
m u =
X
u
u +
X
w
w +
X
q
q mg, (8.5a)
m( w Uq) =
Z
u
u +
Z
w
w +
Z
q
, (8.5b)
B q =
M
q
q +
M
u
u +
M
w
w +
M
w
w. (8.5c)
and
m( v +Ur) =
Y
v
v +
Y
p
p +
Y
r
r +mg, (8.6a)
A p =
L
p
p +
L
r
r +
L
v
v, (8.6b)
C r =
N
p
p +
N
r
r +
N
v
v. (8.6c)
The rst of these sets of equations covers symmetric motion, e.g. pitch oscillations, while the
second covers lateral motion, such as yaw and roll. An important point to note is that these
equations are uncoupled, longitudinal motion does not affect lateral and vice versa.
8.3 Longitudinal symmetric motion
The important information about the dynamic response of a systemis the set of modes in which
it oscillates
2
. These can be found by the usual method of inserting an assumed solution into
the differential equations and nding combinations of parameters which satisfy the system.
The most convenient form of solution is:
u = u
0
e
t
, v = v
0
e
t
, =
0
e
t
.
Inserting these assumptions into Equation 8.5a, for example, yields:
mu
0
e
t
=
X
u
u
0
e
t
+
X
w
w
0
e
t
+
X
q

0
e
t
mg
0
e
t
.
Now, we can divide through by exp t and, as always, non-dimensionalize the parameters, to
give the non-dimensional equations of motion:
( x
u
)u
x
w
w
_
x
q
C
L
2
_
0
= 0, (8.7a)
z
u
u
+ ( z
w
)w
_
1 +
z
q
c
_
0
= 0, (8.7b)
_
m
w

c
+m
w
_
w
+
(b m
q
)
c
= 0. (8.7c)
2
The following analysis, with different notation, is based on GRAHAM, W., Asymptotic analysis of the
classical aircraft stability equations, Aeronautical Journal, February 1999, pp95103.
The non-dimensional parameters are:
x
u
=
X
u
US
, x
w
=
X
w
US
, z
u
=
Z
u
US
, z
w
=
Z
w
US
,
x
q
=
X
q
USc
, z
q
=
Z
q
USc
, m
u
=
M
u
USc
, m
w
=
M
w
USc
,
m
q
=
M
q
USc
2
,
m
w
=
M
w
Sc
,
b =
B
mc
2
and
=
m
US
,
c
=
m
Sc
.
Chapter 9
Normal modes of aircraft
Phugoid
The rst approximate solution we consider is a low frequency oscillation. We state without
proof that there is a solution with and u
/ of order one and w
/
0
of order 1/
c
. This
means that, in this case, the vertical motion is negligible or, equivalently, the incidence is
almost constant. We can rewrite Equations 8.7 in matrix form, with the negligible terms in
each equation removed:
_
_
x
u
0 C
L
/2
z
u
0
0 m
w
(b m
q
)/
c
_
_
_
_
u
0
_
_
=
_
_
0
0
0
_
_
.
This equation can only have a non-trivial solution if the determinant of the matrix is zero:
2
x
u

z
u
2
C
L
= 0.
Solving for gives:
=
(x
u
)
2
j
ph
_
1
_
x
u
2
ph
_
2
_
1/2
,
which species oscillatory motion with:
ph
=
_
z
u
C
L
2
_
1/2
, natural frequency, (9.1a)
c
ph
=
x
u
2
ph
, damping. (9.1b)
This solution denes the phugoid mode, which is a lightly damped long period oscillation.
The incidence is almost constant and the aircraft varies altitude at constant energy, trading
potential for kinetic and back again, Figure 9.1.
An important point to note is that the damping is proportional to (z
u
), the rate of change
of vertical force with small changes in horizontal speed. Remembering that the z axis points
vertically down, we can see that z
u
< 0 and the damping is positive. Although it is not proven
on the basis of these results, a statically stable aircraft always has a stable phugoid.
63
64 CHAPTER 9. NORMAL MODES OF AIRCRAFT
h
max
, V
min
h
min
, V
max
Figure 9.1: Phugoid oscillation trajectory
Short period oscillation
The second solution for longitudinal oscillation is for the case where is of order
1/2
c
, u
0
/
0
is of order
1/2
c
and w
0
/
0
is of order one. In this case, the approximation to equations 8.7 is:
_
_
x
u
x
w

_
x
q
C
L
2
_
0 z
w

0
_
m
w
c
+m
w
_
(b m
q
)
c
_
_
_
_
u
0
_
_
=
_
_
0
0
0
_
_
.
Again, we nd the natural frequency by requiring that the determinant of the matrix be zero:
( x
u
)
_
_
z
w
+
m
q
+m
w
b
_
+
z
w
m
q
m
w
c
b
_
= 0,
which, on solving the quadratic, gives a result for the non-dimensional natural frequency and
damping:
spo
=

c
(m
w
) +m
q
z
w
b
, natural frequency, (9.2a)
c
spo
=
1
2
spo
_
z
w
+
m
q
+m
w
b
_
, damping. (9.2b)
This is the short period oscillation and is a heavily damped mode with period typically of a
few seconds. The aircraft pitches rapidly about its centre of gravity which continues to y
at almost constant speed in a straight line. The periodic time is typically a few seconds, but
must not be less than about 1.25 seconds, otherwise there is a risk of Pilot Induced Oscillation
(PIO).
V
Figure 9.2: Short period oscillation
The frequency is proportional to K
1/2
n
, and increases with dynamic pressure, V
2
/2.
Therefore the aircraft will have the highest frequency SPO, and hence the shortest time period,
at high speed with the centre of gravity in the furthest forward position. The SPO is always
stable for a statically stable aircraft.
9.1. LATERAL MOTION 65
9.1 Lateral motion
In the case of lateral motion, we again need to insert the assumed form for the solution:
v = v
0
e
t
, =
0
e
t
, r = r
0
e
t
,
and non-dimensionalize quantities in equations 8.6, which we do in the same way as before
except that our reference length is now s, the wingspan:
( y
v
)v
_
y
p
s
+
C
L
2
_
0
+
_
1
y
r
s
_
r
= 0, (9.3a)
l
v
v
+ (a l
p
)

l
r
s
r
= 0, (9.3b)
n
v
v
n
p
0
+
c n
r
s
r
= 0. (9.3c)
The non-dimensional quantities are:
y
v
=
Y
v
US
,
y
p
=
Y
p
USs
, y
r
=
Y
r
USs
, l
v
=
L
v
USs
, n
v
=
N
v
USs
,
l
p
=
L
p
USs
2
, l
r
=
L
r
USs
2
, n
p
=
N
p
USs
2
, n
r
=
N
r
USs
2
,
a =
A
ms
2
, c =
C
ms
2
,
v
=
v
0
U
, r
=
mr
0
US
,
s
=
m
Ss
.
Dutch roll
The rst lateral mode we consider is Dutch roll which has oscillations of roughly equal mag-
nitude in pitch, yaw and roll. In this case, equations 9.3 reduce to:
_
_
0 1
l
v
a
2
/
s
0
n
v
0 c/
s
_
_
_
_
v
0
r
_
_
=
_
_
0
0
0
_
_
.
As before the determinant of the matrix must be zero for a non-trivial solution to exist:
2
(c
2
+
s
n
v
) = 0,
and the frequency of the oscillation is, on the approximations we are using:
dr
=
_
s
n
v
c
_
1/2
. (9.4)
In Dutch roll, yawing oscillation (analogous to the longitudinal SPO) causes alternating sideslip.
This in turn causes a rolling oscillation via L
v
v. The periodic time is typically a few seconds,
but as for the SPO it should not have a period of less than 1.25 seconds due to PIO.
Dutch roll is not permitted to be divergent. Divergent Dutch roll can be xed by a yaw
damper on the rudder which damps the yawing oscillation, and hence the roll response as well.
Spiral mode and roll subsidence
There are two further solutions to the dynamic equations which have small values of . These
are dominated by yaw and roll with weak sideslip and the corresponding approximations to
equations 9.3 are:
_
_
0 C
L
/2 1
l
v
(a l
p
)/
s
l
r
/
s
n
v
n
p
/
s
(c n
r
)/
s
_
_
_
_
v
0
r
_
_
=
_
_
0
0
0
_
_
.
The requirement for non-trivial solution is then that:
an
v
2
+ [l
v
(n
p
cC
L
/2) l
p
n
v
] + (l
v
n
r
l
r
n
v
)C
L
/2 = 0.
The two roots of this equation can be approximated as:
rs
=
(l
p
)n
v
+ (l
v
)[cC
L
/2 + (n
p
)
an
v
(9.5)
and
sm
=
C
L
2
l
v
n
r
l
r
n
v
(l
p
)n
v
+ (l
v
)[cC
L
/2 + (n
p
)]
. (9.6)
Note that both of these roots are real and so they do not describe oscillations. The rst,
rs
, de-
scribes rolling subsidence which is a pure rolling motion that is generally heavily damped, and
is therefore generally stable. The damping is primarily from the wings, where the incidence
along the wing is changed due to the roll-rate, as shown in gure 9.3.
Roll rate p
Loading
Rolling moment L
p
p < 0
Figure 9.3: Rolling subsidence
This roll-rate results in a rolling moment L
p
p. Therefore, if L
p
is negative the rolling
subsidence mode is stable. This is generally the case. However, if L
p
becomes negative,
9.2. DIHEDRAL EFFECT AND WEATHERCOCK STABILITY 67
generally due to non-linearities in the lift curve slopes at high roll rates, auto rotational rolling
can occur. This is what happens when an aircraft spins.
The second root
sm
, which is much smaller than
rs
, corresponds to the spiral mode of
the aircraft. This is a combined yaw and roll motion which is allowed to be unstable (i.e.
negatively damped) as long as it does not double amplitude in less than twenty seconds, so
that it can be controlled out.
The dynamics of the spiral mode are that if the aircraft rolls slightly, it will start to sideslip,
and the n then tries to turn the aircraft into the relative wind due to a yawing moment N
v
v.
However, the rolling moment due to sideslip L
v
v tries to roll the wings back level. Depending
on which of the effects wins, the aircraft will be spirally unstable or stable, as can be seen
from the numerator of equation 9.6.
9.2 Dihedral effect and weathercock stability
The aerodynamic derivatives L
v
and N
v
dene whether an aircraft is stable or unstable in
rolling subsidence and Dutch roll. L
v
and N
v
are known as the dihedral effect and weath-
ercock stability respectively. The effect of the two aerodynamic derivatives on the lateral
stability of the aircraft is shown in gure 9.4.
L
v
N
v
Increasing altitude
Unstable spiral mode
Unstable Dutch roll
All lateral modes stable
Figure 9.4: Stability of the lateral modes
Dihedral effect
L
v
is known as the dihedral effect since the majority of the rolling moment due to sideslip
comes from dihedral (on an aircraft with unswept wings), as shown in gure 9.5. Positive
dihedral combined with positive sideslip results in a negative rolling moment (and hence neg-
ative L
v
).
Relative wind
Positive sideslip
L
v
< 0
Figure 9.5: dihedral effect
Wing sweep has a large, negative, effect on L
v
due to reduced or increased effective sweep
for positive sideslip. this is shown in gure 9.6.
Reduced eective sweep:
increased lift
Increased eective sweep:
reduced lift
Positive sideslip
L
v
< 0
Figure 9.6: Wing sweep effects on L
v
Wing/fuselage interference effects give contributions to L
v
due to changes in wing effec-
tive incidence near the root. These contributions are negative for high mounted wings and
positive for low mounted wings, as shown in gure 9.7.
A reasonable level of L
v
may be achieved by using anhedral with swept and high mounted
wings (e.g. Harrier). Ground clearance issues may limit anhedral on low wing aircraft, result-
ing in an unstable Dutch roll mode.
Weathercock stability
The aerodynamic derivative N
v
is known as weathercock stability since it is, effectively, the
ability of an aircraft to turn into the wind. It is produced mainly by the sideways lift-force
9.2. DIHEDRAL EFFECT AND WEATHERCOCK STABILITY 69
L
v
< 0
L
v
> 0
High wing Low wing
Figure 9.7: Wing-fuselage interference effects on L
v
of the n in sideslip, and should always be negative. However, as shown in 9.2, if N
v
is too
large the aircraft may be spirally unstable.
The n contribution to N
v
generally reduces with increasing Mach number, since the ns
lift curve slope is reducing. Therefore an aircraft with a large n may be spirally stable at high
speeds but unstable at low speeds. This can be solved by using paired ns close together. At
low speeds their mutual interference reduces their effectiveness, while at supersonic speeds
this interference is progressively removed, increasing their effectiveness to combat the de-
creasing lift curve slope. This is shown in gure 9.8.
M
n a
1
1.0
Mach cone
Figure 9.8: Use of twin ns at high speed
For VSTOL aircraft (e.g.Harrier) engine air intake mass ow may give a negative contri-
bution to N
v
making the aircraft directionally unstable in the hover and at low forward speeds.
This is because the air undergoes a change in direction and hence a momentum change to go
down the intake, giving a sideforce acting ahead of the aircraft c.g., as shown in gure 9.9.
Intake ow
S
id
e
fo
r
c
e
N
v
< 0
Flight speed
N
v
Fin
In
ta
k
e
o
w
Total
Directionally
unstable
Figure 9.9: Intake effects on N
v

Some Notes On Aircraft Stability and Control: Michael Carley

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Some Notes On Aircraft Stability and Control: Michael Carley

Uploaded by

Copyright:

Available Formats

Some notes on aircraft stability and control

Michael Carley, m.j.carley@bath.ac.uk

Figure 1.11: Measurement of control/tab deections

Figure 1.12: Pressure distributions due to deections, a

is the control surface area and c

is the control surface mean chord. Both of these

Figure 1.13: Measurement of control surface area

Figure 1.14: Control hinge moments

/s yaw is double the

/s pitch, which in turn is double the force for a 10

which can be rearranged to get:

a: horn balance b: hinge location

Figure 6.5: Effects of positive springs and bob-weights

Figure 6.6: Modication of stick force per g

/ of order one and w

You might also like