Aircraft Flight Dynamics, Stability, and Control

Instructor:

Prof. Dimitri Mavris

Boeing Endowed Regents Professor in

Advanced Aerospace Systems Analysis

Special Thanks:

Dr. Evan D. Harrison

 Contents
• Fundamental Concepts and Reference Frames
• Aircraft Equations of Motion
• Longitudinal Stability
– Flying Wing, Wing-Tail, and Canard-Wing Configurations
– Additional factors affecting longitudinal stability
• Longitudinal Control and Maneuverability
• Lateral/Directional Stability, Control, and Maneuverability
• Stability in Steady Flight
– Longitudinal Modes
• Phugoid and Short Period Modes
• Open Loop Response to Control Inputs
– Lateral Modes
• Additional Topics

2
 Uneven Distribution of Knowledge Effects
Design Freedom 100% 100% 100%

1 1 1 1. Aerodynamics

2 2 2 2. Propulsion

3 3 3 3. Structures

4 4 4 4. Controls

5 5 5 5. Manufacturing

6 6 6 6. Supportability

7 7 7 7. Cost

Knowledge Conceptual Preliminary Detailed

about design Time into the Design Process

3
 Contents
• Fundamental Concepts and Reference Frames
• Aircraft Equations of Motion
• Longitudinal Stability
– Flying Wing, Wing-Tail, and Canard-Wing Configurations
– Additional factors affecting longitudinal stability
• Longitudinal Control and Maneuverability
• Lateral/Directional Stability, Control, and Maneuverability
• Stability in Steady Flight
– Longitudinal Modes
• Phugoid and Short Period Modes
• Open Loop Response to Control Inputs
– Lateral Modes
• Additional Topics

4
 Atmospheric Flight Mechanics
Atmospheric Flight Mechanics
encompasses three major disciplines

Performance Flight Dynamics Aeroelasticity

• Determination of • Analysis of motion due to • Analysis of phenomena
performance characteristics internal or external associated with
disturbances interactions between
• Range, endurance, rate of inertial, elastic, and
climb, takeoff & landing • Determination of stability aerodynamic forces
distances, fuel burn, flight and control capabilities
path optimization, etc. • Deals with both static and
• Requires complete dynamic aeroelastic
• Point mass assumption, with equations of motion (6- phenomena
focus on 𝐿, 𝐷, 𝑇, 𝑊 forces DoF for rigid body)

Nelson, R.C., “Flight Stability and Automatic

5 Control”, 2nd Ed., McGraw Hill, 1997
 Stability & Control (or Versus?)
• Stability is defined as the aircraft’s tendency to
return to its equilibrium position after it has been
disturbed either by pilot action or atmospheric
phenomena

• Static stability is the initial tendency to return to

original equilibrium state
– Statically stable, statically unstable, neutrally stable Non-oscillatory

• Dynamic stability deals with the time history of the Neutral

motion of the vehicle after it is disturbed from the Time

equilibrium point
– Non-oscillatory motion: subsidence, divergence, neutral

Nelson, R.C., “Flight Stability and Automatic

6 Control”, 2nd Ed., McGraw Hill, 1997
 Stability & Control (or Versus?)
• Dynamic stability often manifests as oscillatory motion
– Damped, undamped, divergent
Damped oscillation Undamped oscillation Divergent oscillation

Time Time Time

• A vehicle can be statically stable but dynamically unstable

– Static stability does not guarantee dynamic stability
• However, for a vehicle to be dynamically stable, it must also be statically stable
• Control is defined as the ability of the aircraft to be maneuvered as desired from
one flight condition to another

7
 Stability: Pendulum Example

Pin with Pin without

friction friction

gravity

Statically Stable Statically Stable Statically Unstable

Dynamically Stable Dynamically Neutral

8
Stability: Mass-Spring-Damper
• Behavior can be simply
described with Newton’s 2nd
law
– 2nd order ODE
• System is autonomous, linear,
and time-invariant
• Is this system statically and
dynamically stable?
– How does the stability change
as the system parameters
change?

9
 Eigenvalues and Eigenvectors
• Consider the eigenproblem for the system
𝑿ሶ = 𝐴 𝑿
• The eigenvalues 𝜆 = {𝜆1 , … , 𝜆𝑛 } are the roots of the characteristic equation
𝐝𝐞𝐭 𝝀 𝑰 − 𝑨 = 𝟎

• The eigenvector 𝒗𝒊 for each eigenvalue 𝝀𝒊 is found by solving the equation

𝑨𝒗𝒊 = 𝝀𝒊 𝒗𝒊
• Eigenvectors are typically normalized to a magnitude of unity
• Eigenvalues and eigenvectors can be real or complex, with implications on dynamic
stability

10
 An Example Eigenproblem

Consider a simple 2 x 2 system Find eigenvector for 𝝀𝟏 = 𝟏

2 1 2 1 𝑎1 𝑎1
𝐴= 𝐴𝑣1 = 𝜆1 𝑣1 → 𝑏 =1 𝑏
1 2 1 2 1 1
0.707
𝑎1 + 𝑏1 = 0 → 𝑣1 =
−0.707

Find eigenvalues of this system Find eigenvector for 𝝀𝟐 = 𝟑

2−𝜆 1 2 1 𝑎2 𝑎2
(𝐴 − 𝜆𝐼) = 𝐴𝑣2 = 𝜆2 𝑣2 → =3 𝑏
1 2−𝜆 1 2 𝑏2 2
det(𝐴 − 𝜆𝐼) = 2 − 𝜆 2 − 1 = 0 0.707
𝑎2 − 𝑏2 = 0 → 𝑣2 =
0.707
𝜆1 = 1, 𝜆2 = 3

11
 Eigenvalues of the MSD System
• Consider the eigenproblem for the system
0 1
𝑥ሶ
= 𝑘 𝑏 𝑥
𝑥ሷ − − 𝑥ሶ
𝑚 𝑚
• Construct the characteristic equation:
𝜆 −1 𝑏 𝑘
𝑘 𝑏 2
det =0 →𝜆 + 𝜆+ =0
𝜆− 𝑚 𝑚
𝑚 𝑚
2
𝑏 𝑏 𝑘
− ± −4
𝑚 𝑚 𝑚
𝜆=
2
• In general the values of 𝜆 will be complex numbers of the form
𝝀 = −𝝈 ± 𝒋𝝎𝒅

12
 Interpretation of Eigenvalues
• Consider a general complex conjugate eigenvalue pair:
𝝀 = −𝝈 ± 𝒋 𝝎𝒅 𝝈: damping rate 𝝎𝒅 : damped natural frequency

• Purely real eigenvalue (𝜔𝑑 = 0):

- non-oscillatory mode, damping = 𝜎
– 𝑹𝒆 𝝀 = 𝟎 → 𝜎 = 0 : rigid-body mode, describing rigid-body displacement
– 𝑹𝒆 𝝀 < 𝟎 → 𝜎 > 0 : convergent mode, exponential return to equilibrium
– 𝑹𝒆 𝝀 > 𝟎 → 𝜎 < 0 : divergent mode, exponential deviation from equilibrium

13
 Interpretation of Eigenvalues
• Consider a general complex conjugate eigenvalue pair:
𝝀 = −𝝈 ± 𝒋 𝝎𝒅 𝝈: damping rate 𝝎𝒅 : damped natural frequency

• Complex pair of eigenvalues (𝜔𝑑 ≠ 0):

- oscillatory mode, damping = 𝜎, damped natural frequency = 𝜔𝑑
– 𝑹𝒆 𝝀 = 𝟎 → 𝜎 = 0 : undamped oscillatory mode
– 𝑹𝒆 𝝀 < 𝟎 → 𝜎 > 0 : damped oscillatory mode
– 𝑹𝒆 𝝀 > 𝟎 → 𝜎 < 0 : divergent oscillatory mode

14
 Interpretation of Eigenvalues
𝝀 = −𝝈 ± 𝒋 𝝎𝒅 𝝈: damping rate 𝝎𝒅 : damped natural frequency
Imaginary axis Undamped natural frequency
𝜆 = −𝜎 + 𝑗 𝜔𝑑
𝜔𝑛 = 𝜎 2 + 𝜔𝑑2
𝜔𝑛 𝜔𝑑 = 𝜔𝑛 1 − 𝜁 2 Damping ratio
cos −1 (𝜁) = 𝜃 𝜎
𝜁=
Real axis
𝜎 2 + 𝜔𝑑2
𝜎 = 𝜁𝜔𝑛
Left-half plane Right-half plane Graphical analysis of how the roots
Stable Unstable (eigenvalues) of a dynamical system change
(convergent) (divergent) with a change in gain is called root locus
analysis.

This type of graphical representation is

called a root locus plot.

15
 Stability of the MSD
0 1
𝑥ሶ
= 𝑘 𝑏 𝑥
𝑥ሷ − − 𝑥ሶ
𝑚 𝑚

2
𝑏 𝑏 𝑘
− 𝑚 ± −4 𝑚
𝑚
𝝀=
2

𝝀 = −𝝈 ± 𝒋𝝎𝒅

• In general for positive values of k

and b of similar magnitude, the
system is statically stable and
dynamically stable with damped 𝑚 = 1, 𝑘 = 1, 𝑏 = 0.5
oscillations

16
 Stability of the MSD
• The behavior of the system is highly dependent upon the value of the system
parameters

𝑘 = 0.5, 𝑏 = 0.5 𝑘 = 1, 𝑏 = 0.5 𝑘 = 1.5, 𝑏 = 0.5

𝑘 = 1, 𝑏 = −0.5
𝑘 = 1, 𝑏 = 0

17
 Reference Frames
• A reference frame is an orthonormal set of axes attached to an origin, which may be
stationary or moving

• Positions, translation and angular velocities, forces, moments, etc. are vector
quantities. A reference frame is needed to represent them with numbers
• An inertial frame of reference is one which is stationary with respect to the distant
stars. For such frames, Newton’s 2nd Law (𝐹ത = 𝑚 𝑎ത𝑐 ) and Euler’s Law (𝑀 ഥ𝑐 = 𝐻 ഥሶ 𝑐 )
can be applied in these stated forms

• In many cases in aircraft dynamics, it is more convenient (or necessary) to use

reference frames whose origins (and axes) are fixed relative to the vehicle. These
reference frames are not inertial
• As a result, the equations of motion developed using such reference frames have
additional terms that account for the movement of the reference frame

18
 Earth-Fixed Reference Frames
Earth-Surface Earth-Fixed (ESEF) Earth-Center Earth-Fixed (ECEF)
• Origin: Point on Earth’s surface, • Origin: Center of Earth
preferably near vehicle
• X-axis: direction fixed by
• Z-axis: Directed vertically down reference point on equator

• X-axis: North, Y-axis: East • Z-axis: along Earth’s rotation

axis, pointing North
• XY plane: local horizontal
plane

Etkin, B., “Dynamics of Atmospheric Flight”,

19 John Wiley & Sons Inc., 1972
 Wind Axes and Body-Fixed Axes
Wind Axes Body-fixed Axes
• Origin: Usually C.G. • Origin: Usually C.G.
• X-axis: along velocity vector relative to atmosphere • X-axis: longitudinal direction
(multiple possibilities)
• Z-axis: lies on plane of symmetry (if present)
• Y-axis: rightward

The angle of attack and angle of sideslip rotate the

wind axes into the body-fixed axes or vice versa
𝑤 𝑣
𝑉∞ = 𝑢2 + 𝑣 2 + 𝑤 2 , 𝛼 = tan−1 , 𝛽 = sin−1
𝑢 𝑉∞
Etkin, B., “Dynamics of Atmospheric Flight”,
20 John Wiley & Sons Inc., 1972
 Stability Axes and Principal Axes
Stability Axes Principal Axes
• Special case of body-fixed axes used to • Body-fixed axes have the advantage
study disturbances from reference that components of the inertia tensor do
condition not change with change in orientation
• For symmetric flight condition, (attitude)
coincides with wind axes in reference • Principal axes have a further advantage
condition, departs from it and moves that the inertia tensor is a diagonal
with the body during disturbances matrix
• For asymmetric flight condition, x-axis – Only moments of inertia (diagonal
lies on projection of 𝑉∞ on plane of terms) are present
symmetry, and z-axis also lies on plane – All products of inertia (off-
of symmetry diagonal) terms are zero

∗
𝐼𝑥𝑥 0 0
∗
𝐼𝑐∗ = 0 𝐼𝑦𝑦 0
0 0 ∗
𝐼𝑧𝑧

21
 Rotation Between Reference Frames
• Commonly convenient to work in multiple From vehicle carried frame to body fixed frame
reference frames 𝑥𝐵 = 𝐿𝐵𝑉 𝑥𝑉
– e.g: Wind axis convenient for aerodynamic
analysis while body axis convenient for 𝐿𝐵𝑉 = 𝐿1 𝜙 ∙ 𝐿2 𝜃 ∙ 𝐿3 𝜓
kinematic analysis
cos 𝜃 cos 𝜓 cos 𝜃 sin 𝜓 − sin 𝜃
• Can express vectors generated in one sin 𝜙 sin 𝜃 cos 𝜓 sin 𝜙 sin 𝜃 sin 𝜓
reference frame in a different frame sin 𝜙 cos 𝜃
𝐿𝐵𝑉 = − cos 𝜙 sin 𝜓 + cos 𝜙 cos 𝜓
through rotation matrix cos 𝜙 sin 𝜃 cos 𝜓 cos 𝜙 sin 𝜃 sin 𝜓
cos 𝜙 cos 𝜃
+ sin 𝜙 sin 𝜓 − sin 𝜙 cos 𝜓
1 0 0
𝐿1 𝑋1 = 0 cos 𝑋1 sin 𝑋1 From wind axes to body-fixed axes
0 − sin 𝑋1 cos 𝑋1
𝑥𝐵 = 𝐿𝐵𝑊 𝑥𝑊
cos 𝑋2 0 − sin 𝑋2
𝐿2 𝑋2 = 0 1 0 𝐿𝐵𝑊 = 𝐿2 𝛼𝑥 ∙ 𝐿3 𝛽
sin 𝑋2 0 cos 𝑋2
cos 𝛼𝑥 cos 𝛽 − cos 𝛼𝑥 sin 𝛽 − sin 𝛼𝑥
1 0 0 𝐿𝐵𝑊 = sin 𝛽 cos 𝛽 0
𝐿3 𝑋1 = 0 cos 𝑋1 sin 𝑋1 sin 𝛼𝑥 cos 𝛽 − sin 𝛼𝑥 sin 𝛽 cos 𝛼𝑥
0 − sin 𝑋1 cos 𝑋1
Etkin, B., “Dynamics of Atmospheric Flight”,
22 John Wiley & Sons Inc., 1972
 Contents
• Fundamental Concepts and Reference Frames
• Aircraft Equations of Motion
• Longitudinal Stability
– Flying Wing, Wing-Tail, and Canard-Wing Configurations
– Additional factors affecting longitudinal stability
• Longitudinal Control and Maneuverability
• Lateral/Directional Stability, Control, and Maneuverability
• Stability in Steady Flight
– Longitudinal Modes
• Phugoid and Short Period Modes
• Open Loop Response to Control Inputs
– Lateral Modes
• Additional Topics

23
 Transitioning to Rigid Body Motion
• Early performance analysis can
typically be conducted using a point-
mass assumption
– Total forces are assumed to act at some
reference point

• S&C analysis typically involves

distributed forces and moments,
which prevents the use of the point-
mass assumption

• Will also consider the full set of rigid-

body degrees of freedom
– Results in more vehicle states and
corresponding equations

24
 Aerodynamic Forces on a Body
ഥ is the integrated effect of the pressure and shear
The resultant aerodynamic reaction 𝑹
forces acting on the body moving within the fluid
Pressure Shear Stress
𝑝 = 𝑝(𝑠)
𝜏=𝜏 𝑠
𝑠
𝑠

acts perpendicular to the surface acts parallel to the surface

ഥ
𝑹
𝑳 Lift (𝑳), defined as
perpendicular to 𝑉∞
ෝ
𝒏
෡
𝒌 Drag (𝑫), defined
as parallel to 𝑉∞
𝑑𝑠
𝑫
Integrate 𝑝 𝑠 and 𝜏 𝑠 around the surface 𝑽∞ 𝒂 C.P.

of the body to get the total force:

These aerodynamic forces act at the center of
ഥ = ∫ ∫ 𝑝 𝑠 𝑛ො 𝑑𝑠 + ∫ ∫ 𝜏 𝑠 𝑘෠ 𝑑𝑠
𝑹 pressure (the centroid of the distributed load)
 Center of Pressure and Aerodynamic Center
𝑳 𝑳

𝑴𝒂𝒄
𝑫 𝑫
𝑽∞ 𝒂 C.P. 𝑽∞ 𝒂 A.C.

The aerodynamic forces act at the center of 𝐿 and 𝐷 can be transferred to a different point,
pressure (the centroid of the distributed load) with a moment added to maintain equivalency

The net aerodynamic moment taken about this A point called the aerodynamic center (A.C.)
point is zero. However, the point moves with exists on the airfoil such that the moment 𝑀𝑎𝑐
change in angle of attack 𝛼 about it is invariant with angle of attack 𝛼

This makes it inconvenient to use the C.P. in The location of the A.C. can be found
flight dynamics analyses, since its movement experimentally. It is near 25%-chord for subsonic
would have to be tracked and near 45%-chord for supersonic flow
 Kinematics for 6-DoF Rigid Body Aircraft Motion
• The translational velocity can be expressed either in body-
fixed axes as 𝑉ത𝑏 = {𝑢, 𝑣, 𝑤} or in wind axes as 𝑉ത𝑤 =
{𝑉∞ , 0,0}
• The angular velocity can also be expressed in either of
these two axes: 𝜔ഥ = {𝑝, 𝑞, 𝑟} or 𝜔ഥ𝑤 = {𝑝𝑤 , 𝑞𝑤 , 𝑟𝑤 }

• The position of the aircraft (𝑥𝑓 , 𝑦𝑓 , 𝑧𝑓 ) is expressed with

respect to an axis system fixed to the ground
• The attitude of the aircraft is expressed through the Euler
angles (𝜙, 𝜃, 𝜓) or (𝜙𝑤 , 𝜃𝑤 , 𝜓𝑤 ) that rotate the earth-fixed
6 x 1st order ODEs
axes to the body or wind axes
𝑥ሶ𝑓 𝑐𝜃 𝑐𝜓 𝑠𝜙 𝑠𝜃 𝑐𝜓 − 𝑐𝜙 𝑠𝜓 𝑐𝜙 𝑠𝜃 𝑐𝜓 + 𝑠𝜙 𝑠𝜓 𝑢
𝑦ሶ𝑓 = 𝑐𝜃 𝑠𝜓 𝑠𝜙 𝑠𝜃 𝑠𝜓 + 𝑐𝜙 𝑐𝜓 𝑐𝜙 𝑠𝜃 𝑠𝜓 − 𝑠𝜙 𝑐𝜓 𝑣
• Kinematic relationships allow the −𝑠𝜃 𝑠𝜙 𝑐𝜃 𝑐𝜙 𝑐𝜃 𝑤
𝑧𝑓ሶ
velocities and rates to be converted to the
𝜙ሶ 1 𝑠𝜙 𝑠𝜃 /𝑐𝜃 𝑐𝜙 𝑠𝜃 /𝑐𝜃 𝑝
time rates of position and attitude variables
𝜃ሶ = 0 𝑐𝜙 −𝑠𝜙 𝑞
• Integrating these allows the aircraft motion to be “tracked” 𝑟
𝜓ሶ 0 𝑠𝜙 /𝑐𝜃 𝑐𝜙 /𝑐𝜃

27
 The Force Equations
• Newton’s 2nd Law, written in vector form for an
accelerating frame of reference:

𝐹ത𝑎𝑒𝑟𝑜 + 𝐹ത𝑝𝑟𝑜𝑝 + 𝐹ത𝑔𝑟𝑎𝑣 = 𝑚 𝑉തሶ + 𝜔

෥ 𝑉ത

• The gravitational force can be expressed in terms of the Euler angles 𝜙, 𝜃, 𝜓 (if wind
axes are used, the same relationships are valid with 𝜙𝑤 , 𝜃𝑤 , 𝜓𝑤

𝐹ത𝑔𝑟𝑎𝑣 = 𝑚𝑔 {− sin 𝜃 , sin 𝜙 cos 𝜃 , cos 𝜙 cos 𝜃}

28
 The Force Equations
• If body-fixed axes are chosen, then 𝑉ത = {𝑢, 𝑣, 𝑤}, 𝐹ത𝑎𝑒𝑟𝑜 = {𝑋, 𝑌, 𝑍}, 𝐹ത𝑝𝑟𝑜𝑝 =
{𝑇𝑥,𝑏 , 𝑇𝑦,𝑏 , 𝑇𝑧,𝑏 }

𝑢ሶ 𝑋 + 𝑇𝑥,𝑏 − 𝑚𝑔 sin 𝜃 + 𝑟𝑣 − 𝑞𝑤
1
𝑣ሶ = 𝑌 + 𝑇𝑦,𝑏 + 𝑚𝑔 sin 𝜙 cos 𝜃 + 𝑝𝑤 − 𝑟𝑢 3 x 1st order ODEs
𝑚
𝑤ሶ 𝑍 + 𝑇𝑧,𝑏 + 𝑚𝑔 cos 𝜙 cos 𝜃 + 𝑞𝑢 − 𝑝𝑣

• If wind axes are chosen, then 𝑉ത = {𝑉∞ , 0, 0}, 𝐹ത𝑎𝑒𝑟𝑜 = −{𝐷, 𝐶, 𝐿}, 𝐹ത𝑝𝑟𝑜𝑝 =
{𝑇𝑥,𝑤 , 𝑇𝑦,𝑤 , 𝑇𝑧,𝑤 }

𝑚𝑉ሶ = 𝑇𝑥,𝑤 − 𝐷 − 𝑚𝑔 sin 𝜃𝑤

𝑚𝑉𝑟𝑤 = 𝑇𝑦,𝑤 − 𝐶 + 𝑚𝑔 cos 𝜃𝑤 sin 𝜙𝑤
−𝑚𝑉𝑞𝑤 = 𝑇𝑧,𝑤 − 𝐿 + 𝑚𝑔 cos 𝜃𝑤 cos 𝜙𝑤

29
 The Moment Equations
• The sum of externally applied moments equals the rate of change of angular
momentum
– Angular momentum of the vehicle, expressed in body-fixed axes
ഥ + ∑ℎത 𝑟
ഥ𝐶 = 𝐼𝐶 𝜔
𝐻
– ∑ℎത 𝑟 is the net angular momentum contribution of rotating internal systems, e.g.
propellers, jet engines, etc.

• Taking a time derivative, and remembering that the body-fixed axes are not inertial,
the moment equations are obtained as

Net moments 𝑀𝑥,𝑏

due to both ഥሶ + 𝐼𝐶ሶ 𝜔
𝑀𝑦,𝑏 = 𝐼𝐶 𝜔 ഥ + 𝜔𝐼 ഥ + ∑ℎതሶ 𝑟 + 𝜔∑
෥ 𝐶𝜔 ෥ ℎത 𝑟
aero and thrust 𝑀𝑧,𝑏

30
 The Moment Equations
𝑀𝑥,𝑏
ഥሶ + 𝐼𝐶ሶ 𝜔
𝑀𝑦,𝑏 = 𝐼𝐶 𝜔 ഥ + 𝜔𝐼 ഥ + ∑ℎതሶ 𝑟 + 𝜔∑
෥ 𝐶𝜔 ෥ ℎത 𝑟
𝑀𝑧,𝑏
• For the restricted case of a rigid body (𝐼𝐶ሶ = 0) with no rotating internal systems
(∑ℎത 𝑟 = 0ത ), and with an x-z plane of symmetry (𝐼𝑥𝑦 = 𝐼𝑦𝑧 = 0), the moment
equations simplify to:

−1 𝑀𝑥,𝑏 + 𝐼𝑦𝑦 − 𝐼𝑧𝑧 𝑞𝑟 + 𝐼𝑥𝑧 𝑝𝑞

𝑝ሶ 𝐼𝑥𝑥 0 −𝐼𝑥𝑧
𝑞ሶ = 0 𝐼𝑦𝑦 0 𝑀𝑦,𝑏 + 𝐼𝑧𝑧 − 𝐼𝑥𝑥 𝑝𝑟 + 𝐼𝑥𝑧 (𝑟 2 − 𝑝2 ) 3 x 1st order ODEs
𝑟ሶ −𝐼𝑥𝑧 0 𝐼𝑧𝑧 𝑀𝑧,𝑏 + 𝐼𝑥𝑥 − 𝐼𝑦𝑦 𝑝𝑞 − 𝐼𝑥𝑧 𝑞𝑟

31
 The Rigid Body 6-DoF Equations of Motion
• The system of 12, coupled, nonlinear 1st order differential equations does not (in general)
have analytical solutions, but can be integrated numerically to yield time histories of:

𝑢ሶ 𝑋 + 𝑇𝑥,𝑏 − 𝑚𝑔 sin 𝜃 + 𝑟𝑣 − 𝑞𝑤
Translational
1
𝑣ሶ = 𝑌 + 𝑇𝑦,𝑏 + 𝑚𝑔 sin 𝜙 cos 𝜃 + 𝑝𝑤 − 𝑟𝑢
velocity 𝑚
𝑤ሶ 𝑍 + 𝑇𝑧,𝑏 + 𝑚𝑔 cos 𝜙 cos 𝜃 + 𝑞𝑢 − 𝑝𝑣
−1 𝑀𝑥,𝑏 + 𝐼𝑦𝑦 − 𝐼𝑧𝑧 𝑞𝑟 + 𝐼𝑥𝑧 𝑝𝑞
𝑝ሶ 𝐼𝑥𝑥 0 −𝐼𝑥𝑧
Rotational
velocity 𝑞ሶ = 0 𝐼𝑦𝑦 0 𝑀𝑦,𝑏 + 𝐼𝑧𝑧 − 𝐼𝑥𝑥 𝑝𝑟 + 𝐼𝑥𝑧 (𝑟 2 − 𝑝2 )
𝑟ሶ −𝐼𝑥𝑧 0 𝐼𝑧𝑧 𝑀𝑧,𝑏 + 𝐼𝑥𝑥 − 𝐼𝑦𝑦 𝑝𝑞 − 𝐼𝑥𝑧 𝑞𝑟
𝑥ሶ𝑓 𝑐𝜃 𝑐𝜓 𝑠𝜙 𝑠𝜃 𝑐𝜓 − 𝑐𝜙 𝑠𝜓 𝑐𝜙 𝑠𝜃 𝑐𝜓 + 𝑠𝜙 𝑠𝜓 𝑢
Position
vector 𝑦ሶ𝑓 = 𝑐𝜃 𝑠𝜓 𝑠𝜙 𝑠𝜃 𝑠𝜓 + 𝑐𝜙 𝑐𝜓 𝑐𝜙 𝑠𝜃 𝑠𝜓 − 𝑠𝜙 𝑐𝜓 𝑣
𝑧𝑓ሶ −𝑠𝜃 𝑠𝜙 𝑐𝜃 𝑐𝜙 𝑐𝜃 𝑤

𝜙ሶ 1 𝑠𝜙 𝑠𝜃 /𝑐𝜃 𝑐𝜙 𝑠𝜃 /𝑐𝜃 𝑝 The dependency of aircraft dynamics on design

Euler angles
(attitude) 𝜃ሶ = 0 𝑐𝜙 −𝑠𝜙 𝑞 and operating parameters is not easily
𝑟 discernable from numerical solutions.
𝜓ሶ 0 𝑠𝜙 /𝑐𝜃 𝑐𝜙 /𝑐𝜃
To obtain some closed-form analytical
32
solutions, these equations have to be linearized
 Linearization – The Small Disturbance Theory
• Linearization using small disturbance theory requires that the aircraft motion be
confined to small deviations about some steady equilibrium condition

• Each variable in the E.o.M. is written as its equilibrium reference state value plus a
small disturbance from that state: 𝑋 = 𝑋0 + Δ𝑋
– This applies to (1) kinematic variables, (2) forces, and (3) moments

• The forces and moments, which are functions of kinematic variables and control
deflections are approximated using partial derivatives evaluated at the reference
flight condition
𝜕𝑀𝑥,𝑏 𝜕𝑀𝑥,𝑏 𝜕𝑀𝑥,𝑏 Δ𝑢 𝜕𝑀𝑥,𝑏 𝜕𝑀𝑥,𝑏 𝜕𝑀𝑥,𝑏 Δ𝑝
Δ𝑀𝑥,𝑏 = Δ𝑣 + Δ𝑞
𝜕𝑢 𝜕𝑣 𝜕𝑤 𝜕𝑝 𝜕𝑞 𝜕𝑟
Δ𝑤 Δ𝑟
Δ𝑢ሶ Δ𝛿𝑎
𝜕𝑀𝑥,𝑏 𝜕𝑀𝑥,𝑏 𝜕𝑀𝑥,𝑏 𝜕𝑀𝑥,𝑏 𝜕𝑀𝑥,𝑏 𝜕𝑀𝑥,𝑏
+ Δ𝑣ሶ + Δ𝛿𝑒
𝜕𝑢ሶ 𝜕 𝑣ሶ 𝜕𝑤ሶ 𝜕𝛿𝑎 𝜕𝛿𝑒 𝜕𝛿𝑟
Δ𝑤ሶ Δ𝛿𝑟

33
 Small Disturbance Theory Example
• Start with one of the nonlinear equations
1
𝑢ሶ = (𝑋 + 𝑇𝑥,𝑏 − 𝑚𝑔 sin 𝜃 + 𝑟𝑣 − 𝑞𝑤)
𝑚

• Apply perturbated state relationships

1 𝑋 + 𝑇𝑥,𝑏 − 𝑚𝑔 sin 𝜃0 + Δ𝜃
𝑢ሶ 0 + Δ𝑢ሶ =
𝑚 + 𝑟0 + Δ𝑟 𝑣0 + Δ𝑣 − 𝑞0 + Δ𝑞 𝑤0 + Δ𝑤

• Recall that these are small deviations about some steady equilibrium condition
– Typically choose symmetric rectilinear flight, but this is not a restrictive
decision

34
 Small Disturbance Theory Example
• At the reference condition we see
1 1
𝑢ሶ 0 = 0 = 𝑋 + 𝑇𝑥,𝑏,0 − 𝑚𝑔 sin 𝜃0 = 𝐹 − 𝑚𝑔 sin 𝜃0
𝑚 0 𝑚 𝑥,𝑏,0

• The small perturbation case also can be simplified

1
Δ𝑢ሶ = 𝐹𝑥,𝑏 − 𝑚𝑔 sin 𝜃0 + Δ𝜃 + Δ𝑟Δ𝑣 − Δ𝑞Δ𝑤
𝑚

• Now we can simplify this further considering the small angle approximation and that
the product of small values is negligible
1
Δ𝑢ሶ = 𝐹 − 𝑚𝑔 sin 𝜃0 + 𝑚𝑔 cos 𝜃0 Δ𝜃
𝑚 𝑥,𝑏

35
 Small Disturbance Theory Example
• Next a first-order expansion of the force
𝜕𝐹𝑥,𝑏 𝜕𝐹𝑥,𝑏 𝜕𝐹𝑥,𝑏 𝜕𝐹𝑥,𝑏 𝜕𝐹𝑥,𝑏 𝜕𝐹𝑥,𝑏
𝐹𝑥,𝑏 = 𝐹𝑥,𝑏,0 + Δ𝑢ሶ + Δ𝑤ሶ + Δ𝑢 + Δ𝑤 + Δq+ Δ𝛿𝑒
𝜕𝑢ሶ 𝜕𝑤ሶ 𝜕𝑢 𝜕𝑤 𝜕𝑞 𝜕𝛿𝑒
• Substituting in to the equation of motion and rearranging
𝜕𝐹𝑥,𝑏 𝜕𝐹𝑥,𝑏
m− Δ𝑢ሶ − Δ𝑤ሶ
𝜕𝑢ሶ 𝜕 𝑤ሶ
𝜕𝐹𝑥,𝑏 𝜕𝐹𝑥,𝑏 𝜕𝐹𝑥,𝑏
= 𝐹𝑥,𝑏,0 − 𝑚𝑔 sin 𝜃0 + Δ𝑢 + Δ𝑤 + Δq
𝜕𝑢 𝜕𝑤 𝜕𝑞
𝜕𝐹𝑥,𝑏
+ 𝑚𝑔 cos 𝜃0 Δ𝜃+ Δ𝛿𝑒
𝜕𝛿𝑒
• This is now linear equation that we can rewrite in matrix format
Δ𝑢
W 𝜕𝐹𝑥,𝑏 𝜕𝐹𝑥,𝑏 Δ𝑢ሶ 𝜕𝐹𝑥,𝑏 𝜕𝐹𝑥,𝑏 𝜕𝐹𝑥,𝑏 Δ𝑤 𝜕𝐹𝑥,𝑏
− − = W cos 𝜃0 + Δ𝛿𝑒
g 𝜕𝑢ሶ 𝜕𝑤ሶ Δ𝑤ሶ 𝜕𝑢 𝜕𝑤 𝜕𝑞 Δ𝑞 𝜕𝛿𝑒
Δ𝜃

36
 Linearized Longitudinal and Lateral Equations
• Introducing the derivatives into the E.o.M., applying small angle approximations,
and neglecting second order terms, we get
– The linearized longitudinal equations
𝑊/𝑔 − 𝐹𝑥,𝑏𝑢ሶ −𝐹𝑥,𝑏𝑤ሶ 0 0 0 0 Δ𝑢ሶ 𝐹𝑥,𝑏𝑢 𝐹𝑥,𝑏𝑤 𝐹𝑥,𝑏𝑞 0 0 −𝑊 cos 𝜃0 Δ𝑢 𝐹𝑥,𝑏𝛿𝑒
−𝐹𝑧,𝑏𝑢ሶ 𝑊/𝑔 − 𝐹𝑧,𝑏𝑤ሶ 0 0 0 0 Δ𝑤ሶ 𝐹𝑧,𝑏𝑢 𝐹𝑧,𝑏𝑤 𝐹𝑧,𝑏𝑞 + 𝑉0 𝑊/𝑔 0 0 −𝑊 sin 𝜃0 Δ𝑤 𝐹𝑧,𝑏𝛿𝑒
𝐼𝑦𝑦 0 0 0 Δ𝑞ሶ 0 0 0 Δ𝑞
= 𝑀𝑦,𝑏𝑢
−𝑀𝑦,𝑏𝑢ሶ −𝑀𝑦,𝑏𝑤ሶ 𝑀𝑦,𝑏𝑤 𝑀𝑦,𝑏𝑞 𝑀𝑦,𝑏𝛿𝑒 Δ𝛿
0 1 0 0 Δ𝑥ሶ𝑓
cos 𝜃0 sin 𝜃0 0 0 −𝑉0 sin 𝜃0 Δ𝑥𝑓 + 𝑒
0 0 0 0
0 0 0 0 1 0 Δ𝑧𝑓ሶ − sin 𝜃0 cos 𝜃0 0 0 −𝑉0 cos 𝜃0 Δ𝑧𝑓
0 0
0 0 0 0 0 1 Δ𝜃ሶ 0 0 1 0 0 1 Δ𝜃 0

– The linearized lateral equations

𝑊/𝑔 0 0 0 0 0 Δ𝑣ሶ 𝐹𝑦,𝑏𝑣 𝐹𝑦,𝑏𝑝 𝐹𝑦,𝑏𝑟 − 𝑉0 𝑊/𝑔 0 𝑊 cos 𝜃0 0 Δ𝑣 𝐹𝑦,𝑏𝛿𝑎 𝐹𝑦,𝑏𝛿𝑟
0 𝐼𝑥𝑥 −𝐼𝑥𝑧 0 0 0 Δ𝑝ሶ 𝑀𝑥,𝑏𝑣 𝑀𝑥,𝑏𝑝 𝑀𝑥,𝑏𝑟 0 0 0 Δ𝑝 𝑀𝑥,𝑏𝛿𝑎 𝑀𝑥,𝑏𝛿𝑟
0 −𝐼𝑥𝑧 𝐼𝑧𝑧 0 0 0 Δ𝑟ሶ 𝑀𝑧,𝑏𝑟 0 0 0 Δ𝑟 Δ𝛿𝑎
Δ𝑦ሶ𝑓 = 𝑀𝑧,𝑏𝑣 𝑀𝑧,𝑏𝑝
Δ𝑦𝑓 +
𝑀𝑧,𝑏𝛿𝑎 𝑀𝑧,𝑏𝛿𝑟
0 0 0 1 0 0 1 0 0 0 0 𝑉0 cos 𝜃0 Δ𝛿𝑟
0 0
0 0 0 0 1 0 Δ𝜙ሶ 0 1 tan 𝜃0 0 0 0 Δ𝜙 0 0
0 0 0 0 0 1 Δ𝜓ሶ 0 0 sec 𝜃0 0 0 0 Δ𝜓 0 0

37
 Linearized Equations of Motion
𝑊/𝑔 − 𝐹𝑥,𝑏𝑢ሶ −𝐹𝑥,𝑏𝑤ሶ 0 0 0 0 Δ𝑢ሶ 𝐹𝑥,𝑏𝑢 𝐹𝑥,𝑏𝑤 𝐹𝑥,𝑏𝑞 0 0 −𝑊 cos 𝜃0 Δ𝑢 𝐹𝑥,𝑏𝛿𝑒
−𝐹𝑧,𝑏𝑢ሶ 𝑊/𝑔 − 𝐹𝑧,𝑏𝑤ሶ 0 0 0 0 Δ𝑤ሶ 𝐹𝑧,𝑏𝑢 𝐹𝑧,𝑏𝑤 𝐹𝑧,𝑏𝑞 + 𝑉0 𝑊/𝑔 0 0 −𝑊 sin 𝜃0 Δ𝑤 𝐹𝑧,𝑏𝛿𝑒
𝐼𝑦𝑦 0 0 0 Δ𝑞ሶ 0 0 0 Δ𝑞
−𝑀𝑦,𝑏𝑢ሶ −𝑀𝑦,𝑏𝑤ሶ
Δ𝑥ሶ𝑓 = 𝑀𝑦,𝑏𝑢 𝑀𝑦,𝑏𝑤 𝑀𝑦,𝑏𝑞
Δ𝑥𝑓 +
𝑀𝑦,𝑏𝛿𝑒 Δ𝛿
𝑒
0 0 0 1 0 0 cos 𝜃0 sin 𝜃0 0 0 0 −𝑉0 sin 𝜃0 0
0 0 0 0 1 0 Δ𝑧𝑓ሶ − sin 𝜃0 cos 𝜃0 0 0 −𝑉0 cos 𝜃0 Δ𝑧𝑓
0 0
0 0 0 0 0 1 Δ𝜃ሶ 0 0 1 0 0 1 Δ𝜃 0

𝑊/𝑔 0 0 0 0 0 Δ𝑣ሶ 𝐹𝑦,𝑏𝑣 𝐹𝑦,𝑏𝑝 𝐹𝑦,𝑏𝑟 − 𝑉0 𝑊/𝑔 0 𝑊 cos 𝜃0 0 Δ𝑣 𝐹𝑦,𝑏𝛿𝑎 𝐹𝑦,𝑏𝛿

𝑟
0 𝐼𝑥𝑥 −𝐼𝑥𝑧 0 0 0 Δ𝑝ሶ 𝑀𝑥,𝑏𝑣 𝑀𝑥,𝑏𝑝 𝑀𝑥,𝑏𝑟 0 0 0 Δ𝑝 𝑀𝑥,𝑏𝛿𝑎 𝑀𝑥,𝑏𝛿𝑟
0 −𝐼𝑥𝑧 𝐼𝑧𝑧 0 0 0 Δ𝑟ሶ 𝑀𝑧,𝑏𝑟 0 0 0 Δ𝑟 Δ𝛿𝑎
Δ𝑦ሶ𝑓 = 𝑀𝑧,𝑏𝑣 𝑀𝑧,𝑏𝑝
Δ𝑦𝑓 +
𝑀𝑧,𝑏𝛿𝑎 𝑀𝑧,𝑏𝛿𝑟
0 0 0 1 0 0 1 0 0 0 0 𝑉0 cos 𝜃0 Δ𝛿𝑟
0 0
0 0 0 0 1 0 Δ𝜙ሶ 0 1 tan 𝜃0 0 0 0 Δ𝜙 0 0
0 0 0 0 0 1 Δ𝜓ሶ 0 0 sec 𝜃0 0 0 0 Δ𝜓 0 0

• The linearized 6-DoF equations are considerably simpler than the nonlinear 6-DoF
ones
– Valid iff aircraft motion is confined to small deviations about equilibrium
condition
– Valid for most commonly encountered flying conditions
– Not applicable to spinning motion, stall recovery, or large-amplitude rapid
maneuvers

38
 Stability & Control Derivatives
• Traditional flight dynamics discussions will reference a vehicle’s stability
derivatives

• Stability (and control) derivatives are the partial derivatives evaluated about a
reference condition
– Often expressed with respect to nondimensional form of forces or moments

𝜕MX 𝜕𝐶𝑀
MX,u = 𝐶𝑀𝑢 =
𝜕𝑢 𝜕𝑢

39
Some Key Stability & Control Derivatives
Lift Force Side Force Roll Mom. Pitch Mom. Yaw Mom.
(𝐶𝐿 ) (𝐶𝑌 ) (𝐶𝑙 ) (𝐶𝑚 ) (𝐶𝑛 )
AOA (𝛼) 𝑪 𝑳𝜶 𝑪𝒎𝜶
෠ሶ
AOA rate (𝛼) 𝑪𝑳𝜶෡ሶ 𝑪𝒎𝜶෡ሶ

Sideslip (𝛽) 𝑪 𝒀𝜷 𝑪𝒍𝜷 𝑪𝒏𝜷

Roll rate (𝑝)Ƹ 𝑪𝒀𝒑ෝ 𝑪𝒍 𝒑ෝ 𝑪𝒏𝒑ෝ

Pitch rate (𝑞)

ො 𝑪𝑳𝒒ෝ 𝑪𝒎𝒒ෝ

Yaw rate (𝑟)Ƹ 𝑪𝒀𝒓ො 𝑪𝒍𝒓ො 𝑪𝒏𝒓ො

Aileron (𝛿𝑎 ) 𝑪𝒍𝜹𝒂 𝑪𝒏𝜹𝒂

Elevator (𝛿𝑒 ) 𝑪𝑳𝜹𝒆 𝑪𝒎𝜹𝒆

Rudder (𝛿𝑟 ) 𝑪𝒀𝜹𝒓 𝑪𝒍𝜹𝒓 𝑪𝒏𝜹𝒓

1st or 2nd order “stiffness” Rate damping derivatives

Control powers Roll/yaw axis coupling

40
 State-Space Representation of Aircraft Dynamics
• Note that the linearized equations are of the form:
Δ𝑿ሶ = 𝐴 Δ𝑿 + 𝐵 Δ𝑼 Δ𝑋ሶ long = Along Δ𝑋long + Blong Δ𝑈𝑙𝑜𝑛𝑔
• This is called the state-space representation of a Δ𝑋ሶ lat = Alat Δ𝑋lat + Blat Δ𝑈𝑙𝑎𝑡
dynamical system, in which
𝒏 Δ𝑢 Δ𝑣
– 𝑿 ∈ 𝑹 : state vector Δ𝑝
Δ𝑤
– 𝑼 ∈ 𝑹𝒑 : control or input vector Δ𝑞 Δ𝑟
Δ𝑋long = Δ𝑥 Δ𝑋lat = Δ𝑦
– 𝑨 ∈ 𝑹𝒏 x 𝒏 : state or system matrix 𝑓 𝑓
– 𝑩 ∈ 𝑹𝒏 x 𝒑 : control or input matrix Δ𝑧𝑓 Δ𝜙
Δ𝜃 Δ𝜓

• To assess the dynamic stability of the uncontrolled (or

controls-fixed) system, we set Δ𝑼 = 𝟎 ഥ, and look at the Δ𝑈𝑙𝑜𝑛𝑔 = Δ𝛿𝑒
stability of the system Δ𝑿ሶ = 𝐴 Δ𝑿 Δ𝑈𝑙𝑎𝑡 =
Δ𝛿𝑎
Δ𝛿𝑟
• This can be done by solving the eigenproblem, which
yields the system’s eigenvalues and corresponding
eigenvectors

41
 Contents
• Fundamental Concepts and Reference Frames
• Aircraft Equations of Motion
• Longitudinal Stability
– Flying Wing, Wing-Tail, and Canard-Wing Configurations
– Additional factors affecting longitudinal stability
• Longitudinal Control and Maneuverability
• Lateral/Directional Stability, Control, and Maneuverability
• Stability in Steady Flight
– Longitudinal Modes
• Phugoid and Short Period Modes
• Open Loop Response to Control Inputs
– Lateral Modes
• Additional Topics

42
 Trim
• When the controls are set so that the resultant forces and the moments about the
center of gravity are all zero, the aircraft is said to be in trim, which simply means
static equilibrium.
∑𝐹𝑥 = ∑𝐹𝑦 = ∑𝐹𝑧 = 0, ∑𝐿𝑐𝑔 = ∑𝑀𝑐𝑔 = ∑𝑁𝑐𝑔 = 0

• The ease with which the vehicle can be trimmed is an important aspect of the
aircraft characteristics known as handling qualities
– A difficult-to-trim airplane will be tiresome or even dangerous to fly
– Such an aircraft will never be popular among pilots

43
 Trim
• For an aircraft to be statically stable in rotation, disturbances in roll, pitch, or yaw
axes must result in generation of restoring moments tending to return the aircraft to
the original equilibrium (trimmed) attitude

• Note: Simply being able to trim at a flight condition does not imply that the vehicle
is dynamically (or even statically) stable

44
 Pitch Trim & Stability of a Flying Wing
• Equilibrium of forces: 𝑇 = 𝐷, 𝐿 = 𝑊 ℓ𝒘 𝑳

𝑴𝒂𝒄
• Summation of pitching moments 𝑽∞ 𝑻 𝑫
about CG must also be zero:
𝑾
𝑀 = 𝑀𝑎𝑐 − ℓ𝑤 𝐿 = 0
ℓ𝑤
→ 𝐶𝑚 = 𝐶𝑚𝑎𝑐 − 𝐶𝐿 = 0
𝑐ҧ

𝑊
• For a given speed, altitude, and weight, 𝐶𝐿 = 1 >0
2
𝜌𝑉 2 𝑆
𝑐ҧ
• Solving for the moment arm, ℓ𝑤 = 𝐶𝑚𝑎𝑐
𝐶𝐿

• The above is the requirement for static equilibrium - we also need static stability

45
 Pitch Trim & Stability of a Flying Wing
• For static stability, a restoring pitching moment must develop in response to a
disturbance in the pitch axis, i.e., an angle of attack disturbance Δ𝛼
– Restoring means, if Δ𝛼 > 0, then Δ𝐶𝑚 < 0. And if Δ𝛼 < 0, then Δ𝐶𝑚 > 0
𝜕𝐶𝑚
– Thus, the general pitch static stability criterion for any aircraft: 𝐶𝑚𝛼 = <0
𝜕𝛼

ℓ𝑤 𝜕𝐶𝑚
• Recall: 𝐶𝑚 = 𝐶𝑚𝑎𝑐 − 𝐶 .Differentiating this w.r.t. 𝛼 and imposing <0
𝑐ҧ 𝐿 𝜕𝛼
𝜕𝐶𝑚 𝜕𝐶𝑚𝑎𝑐 ℓ𝑤 𝜕𝐶𝐿
= − <0
𝜕𝛼 𝜕𝛼 𝑐ҧ 𝜕𝛼

𝜕𝐶𝑚𝑎𝑐 𝑐ҧ
• From the definition of aerodynamic center, = 0. And recall, ℓ𝑤 = 𝐶𝑚𝑎𝑐
𝜕𝛼 𝐶𝐿
• Using these, the pitch static stability criterion simplifies to the requirement that:
𝐶𝑚𝑎𝑐 𝜕𝐶𝐿
− < 0 → 𝐶𝑚𝑎𝑐 > 0
𝐶𝐿 𝜕𝛼

46
 Pitch Trim & Stability of a Flying Wing
Criterion → Static equilibrium Static stability
𝑐ҧ
Requirement → ℓ𝑤 = 𝐶𝑚𝑎𝑐 𝐶𝑚𝑎𝑐 > 0
𝐶𝐿

Symmetric
(𝑪𝒎𝒂𝒄 = 𝟎) ✓ ℓ𝑤 = 0
(place CG at AC)

Positive camber
(𝑪𝒎𝒂𝒄 < 𝟎)
✓ ℓ𝑤 < 0
(place CG behind AC) 
Negative camber
(𝑪𝒎𝒂𝒄 > 𝟎) ✓ ℓ𝑤 > 0
(place CG ahead of AC)
✓
• Negative camber will get the job done, but is quite inefficient
– Low 𝐶𝐿,𝑚𝑎𝑥 and lift-to-drag ratio

47
 Pitch Trim & Stability of a Flying Wing
Criterion → Static equilibrium Static stability
𝑐ҧ
Requirement → ℓ𝑤 = 𝐶𝑚𝑎𝑐 𝐶𝑚𝑎𝑐 > 0
𝐶𝐿

Reflexed
Trailing Edge ✓ ℓ𝑤 > 0
(place CG ahead of AC)
✓
(𝑪𝒎𝒂𝒄 > 𝟎)

Positive camber over most of chord Negative camber near

(good aerodynamic efficiency) trailing edge
(to satisfy 𝑪𝒎𝒂𝒄 > 𝟎 )

While static stability is achievable, pure flying wings usually require some form of
stability augmentation to obtain acceptable handling qualities

48
 Pitch Stability of Combined Lifting Surfaces
• The majority of fixed-wing aircraft use multiple lifting surfaces
– The primary (main) lifting surface is commonly called the wing

• Additional lifting surface(s) contribute to pitch stability and also allow pitch control
– A tail, or tail-plane (a.k.a. stabilizer), located aft of the main wing
– A canard, or fore-plane, located ahead of the main wing
• Thus, wing-tail, canard-wing, and canard-wing-tail configurations are possible

F-16 Fighting Falcon (wing-tail) JAS 39 Gripen (canard-wing) Sukhoi Su-30 MKI (canard-wing-tail)
https://s-media-cache-
https://c1.staticflickr.com/3/2906/13873859334_7106cff425_b.jpg https://en.wikipedia.org/wiki/Canard_(aeronautics) ak0.pinimg.com/564x/93/d6/01/93d6012be76c0d0723e6116ad9a42f86.jpg

49
 Wing & Aft Tail Combination
• A simplified analysis will allow the major
contributions to trim and static stability to
be analyzed without undue complexity
• More general case will be then be shown
𝜸 𝑽∞ ℓ𝒘
𝑻 𝑳𝒘
𝒊𝒘 ℓ𝒕
𝑽∞
𝑫𝒘
𝝐𝒅
𝑾 𝑴𝒂𝒄,𝒘 𝑽𝒚
𝒊𝒕
𝑳𝒕
𝑴𝒂𝒄,𝒕
• The assumptions of this simplified analysis are:
𝑫𝒕
– Fuselage reference line (FRL) taken to be in
direction of flight, passing through CG 𝜹𝒆
– The thrust line and the aerodynamic centers (AC)
of both wing and tail lie on FRL
50
 Simplified Pitch Equilibrium Analysis
𝑇 = 𝐷𝑤 + 𝐷𝑡 cos 𝜖𝑑 + 𝐿𝑡 sin 𝜖𝑑 + 𝑊 sin 𝛾
𝐿𝑤 + 𝐿𝑡 cos 𝜖𝑑 − 𝐷𝑡 sin 𝜖𝑑 = 𝑊 cos 𝛾
𝑀𝑐𝑔 = 𝑀𝑎𝑐,𝑤 + 𝑀𝑎𝑐,𝑡 − ℓ𝑤 𝐿𝑤 − ℓ𝑡 𝐿𝑡 cos 𝜖𝑑 + ℓ𝑡 𝐷𝑡 sin 𝜖𝑑 = 0

• If the downwash angle is small, then cos 𝜖𝑑 ≈ 1, 𝐷𝑡 sin 𝜖𝑑 ≪ 𝐿𝑡

1 2
𝐿𝑤 𝜌∞ 𝑉𝑡 𝑆𝑡 𝐿𝑡 𝑊 cos 𝛾 𝑆𝑡
+ 2 =
1 1 𝑆 1 1 𝐶𝐿,𝑤 + 𝜂 𝑡 𝐶𝐿,𝑡 = 𝐶𝑊 cos 𝛾
2 2 𝑤 2 2 𝑆
2 𝜌∞ 𝑉∞ 𝑆𝑤 2 𝜌∞ 𝑉∞ 2 𝜌∞ 𝑉𝑡 𝑆𝑡 2 𝜌∞ 𝑉∞ 𝑆𝑤
𝑤

Tail efficiency, 𝜂𝑡 ∈ 0.8, 1.2

𝑀𝑐𝑔 𝑀𝑎𝑐,𝑤 𝑆𝑡 𝑐𝑡ҧ 𝑀𝑎𝑐,𝑡 ℓ𝑤 𝐿𝑤 𝑆𝑡 ℓ𝑡 𝐿𝑡
= + 𝜂𝑡 − − 𝜂𝑡 =0
1 1 𝑆𝑤 𝑐𝑤 ҧ 1 𝜌 𝑉 2 𝑆 𝑐ҧ ҧ 1 𝜌 𝑉2𝑆
𝑐𝑤 𝑆𝑤 𝑐𝑤
ҧ 1
𝜌∞ 𝑉∞2 𝑆𝑤 𝑐𝑤
ҧ 𝜌∞ 𝑉∞2 𝑆𝑤 𝑐𝑤
ҧ 2
𝜌 𝑉 𝑆
2 2 2 ∞ 𝑡 𝑡 𝑡 2 ∞ ∞ 𝑤 2 ∞ 𝑡 𝑡

𝑆𝑡 𝑐𝑡ҧ ℓ𝑤 𝑆𝑡 ℓ𝑡
𝐶𝑚,𝑐𝑔 = 𝐶𝑚,𝑎𝑐,𝑤 + 𝜂𝑡 𝐶𝑚,𝑎𝑐,𝑡 − 𝐶𝐿,𝑤 − 𝜂𝑡 𝐶𝐿,𝑡 = 0
𝑆𝑤 𝑐𝑤 ҧ 𝑐𝑤
ҧ 𝑆𝑤 𝑐𝑤
ҧ

51
 Tail Lift: Positive, Negative, or Zero?
• Solving the previous two relationships for wing and tail lift yields:
𝑆𝑡 𝑆𝑤 𝑐𝑤
ҧ 𝑆𝑡 𝑐𝑡ҧ ℓ𝑤
𝐶𝐿,𝑤 = 𝐶𝑊 cos 𝛾 − 𝜂𝑡 𝐶𝐿,𝑡 𝐶𝐿,𝑡 = 𝐶𝑚,𝑎𝑐,𝑤 + 𝜂𝑡 𝐶𝑚,𝑎𝑐,𝑡 − 𝐶𝐿,𝑊
𝑆𝑤 𝑆𝑡 ℓ𝑡 − ℓ𝑤 𝜂𝑡 𝑆𝑤 𝑐𝑤 ҧ 𝑐𝑤
ҧ

• If tail carries negative lift (𝐶𝐿,𝑡 < 0), the wing lift has to exceed the weight of the
aircraft (𝐶𝐿,𝑤 > 𝐶𝑊 ), which is clearly inefficient

• A tail carrying positive lift (𝐶𝐿,𝑡 > 0) is more efficient, but still not as efficient as if
the same amount of lift was carried by the wing
– Wing AR > tail AR, which means tail sees more induced drag than wing for
same lift

52
 Tail Lift: Positive, Negative, or Zero?
• So for aerodynamic efficiency alone (and not trim or stability), a good preliminary
design choice is to trim with no elevator deflection with zero tail load (𝐶𝐿,𝑡 = 0) at a
design weight and flight condition (i.e. 𝐶𝐿,𝑤 = 𝐶𝑊 )

• This is achieved by positioning the wing A.C. relative to the C.G. such that

𝑆𝑡 𝑐𝑡ҧ ℓ𝑤
𝐶𝑚,𝑎𝑐,𝑤 + 𝜂𝑡 𝐶𝑚,𝑎𝑐,𝑡 − 𝐶𝑊 = 0
𝑆𝑤 𝑐𝑤 ҧ 𝑐𝑤
ҧ

𝑐𝑤
ҧ 𝑆𝑡 𝑐𝑡ҧ
ℓ𝑤 = 𝐶𝑚,𝑤 + 𝜂 𝐶
𝐶𝑊 𝑆𝑤 𝑐𝑤 ҧ 𝑡 𝑚,𝑎𝑐,𝑡

53
 Wing and Tail Mounting Angles
Wing Tail
• At trim, under the assumed conditions, • The downwash seen by the tail is
1 𝜕𝜖
wing lift: 𝐶𝐿,𝑤 = 𝐶𝑊 = 𝑊/ 2 𝜌∞ 𝑉∞2 𝑆𝑤 expressed as 𝜖 = 𝜖0 + 𝜕𝛼 𝛼

• Referring to the wing’s lift curve, • The tail AOA is given by:
𝐶𝐿,𝑤 = 𝐶𝐿𝛼,𝑤 𝛼 + 𝑖𝑤 − 𝛼𝐿0,𝑤 𝜕𝜖
𝛼𝑡 = 𝛼 + 𝑖𝑡 − 𝜖 = 𝛼 1 − + 𝑖 𝑡 − 𝜖0
𝜕𝛼
• The angle of attack 𝛼 is measured relative • If the tail is symmetric, then for 𝐶𝐿,𝑡 = 0,
to the fuselage reference line (F.R.L.), and we require that 𝛼𝑡 = 0
we may want this to be aligned with the
freestream (𝛼 = 0) at the design condition • Solving the above, the tail mounting angle
• This allows us to determine the wing is then numerically equal to the downwash
mounting angle as at zero angle of attack of the F.R.L.
1 𝑖 𝑡 = 𝜖0
𝑖𝑤 = 𝛼𝐿0,𝑤 + 𝑊/ 𝜌∞ 𝑉∞2 𝑆𝑤 𝐶𝐿𝛼 ,𝑤
2

54
 Elevator Angle to Trim
• For flight conditions other than the design condition, the airplane will have to be
trimmed with elevator deflection 𝛿𝑒 ≠ 0
• In this case, the tail lift (assuming a symmetrical airfoil) is given by
𝜕𝜖
𝐶𝐿,𝑡 = 𝐶𝐿𝛼,𝑡 1 − 𝛼 + 𝑖𝑡 − 𝜖0 + 𝜖𝑒 𝛿𝑒
𝜕𝛼
The new quantity 𝜖𝑒 is called the elevator effectiveness. Can be thought of as
“change of effective tail AOA per unit elevator deflection”

• Further, elevator deflection causes a moment about the tail A.C., which for small
elevator deflections is proportional to the deflection: 𝐶𝑚,𝑎𝑐,𝑡 = 𝐶𝑚,𝑎𝑐,𝑡𝛿𝑒 𝛿𝑒
• These relationships can be substituted into the original lift and pitching moment
relationships, which can then be solved for the following two quantities
– Angle of attack at trim, 𝛼𝑡𝑟𝑖𝑚
– Elevator deflection to achieve trim, 𝛿𝑒,𝑡𝑟𝑖𝑚

55
 Simplified Analysis of Pitch Stability
• After a lot of algebra and grouping of terms, the following matrix equations are
obtained: 𝐶𝐿 𝐶𝐿 𝛼 𝐶 −𝐶
𝛼 𝛿𝑒 𝑡𝑟𝑖𝑚 𝐿 𝐿0
𝛿𝑒,𝑡𝑟𝑖𝑚 =
𝐶𝑚 𝛼 𝐶𝑚𝛿𝑒 −𝐶𝑚0

𝑆𝑡 𝜕𝜖
𝐶𝐿𝛼 = 𝐶𝐿𝛼,𝑤 + 𝜂 𝐶 1− Effective lift curve slope of wing + tail
𝑆𝑤 𝑡 𝐿𝛼,𝑡 𝜕𝛼
𝑆𝑡
𝐶𝐿𝛿𝑒 = 𝜂 𝐶 𝜖 Elevator lift effectiveness Lift coefficient in trimmed,
𝑆𝑤 𝑡 𝐿𝛼,𝑡 𝑒 steady-level flight
ℓ𝑤 𝑆𝑡 ℓ𝑡 𝜕𝜖 𝑊
𝐶𝑚𝛼 = − 𝐶𝐿𝛼,𝑤 − 𝜂𝐶 1− Pitch stiffness 𝐶𝐿 =
𝑐𝑤ҧ 𝑆𝑤 𝑐𝑤ҧ 𝑡 𝐿𝛼,𝑡 𝜕𝛼 1 2𝑆
𝜌∞ 𝑉 ∞ 𝑤
𝑆𝑡 𝑐𝑡ҧ 𝑆𝑡 ℓ𝑡 2
𝐶𝑚𝛿𝑒 = 𝜂𝑡 𝐶𝑚,𝑎𝑐,𝑡𝛿𝑒 − 𝜂𝐶 𝜖 Elevator control power
𝑆𝑤 𝑐𝑤 ҧ 𝑆𝑤 𝑐𝑤ҧ 𝑡 𝐿𝛼,𝑡 𝑒
𝑆𝑡
𝐶𝐿0 = 𝐶𝐿𝛼,𝑤 𝑖𝑤 − 𝛼𝐿0,𝑤 + 𝜂 𝐶 𝑖 − 𝜖0 Total lift (wing + tail) at zero AOA
𝑆𝑤 𝑡 𝐿𝛼,𝑡 𝑡
ℓ𝑤 𝑆𝑡 ℓ𝑡
𝐶𝑚0 = 𝐶𝑚,𝑎𝑐,𝑤 − 𝐶𝐿𝛼,𝑤 𝑖𝑤 − 𝛼𝐿0,𝑤 − 𝜂𝑡 𝐶𝐿𝛼,𝑡 (𝑖𝑡 − 𝜖0 ) Total pitching moment (wing
𝑐𝑤
ҧ 𝑆𝑤 𝑐𝑤ҧ + tail) at zero AOA

56
 Static Stability in Pitch
𝜕𝐶𝑚
• Recall the general pitch static stability criterion: <0
𝜕𝛼

• Recall also the equation for pitching moment, expressed in terms of wing and
(symmetric) tail lift curve slopes, incidence angles, and tail downwash:
𝑆𝑡 𝑐𝑡ҧ ℓ𝑤
𝐶𝑚 = 𝐶𝑚,𝑎𝑐,𝑤 + 𝜂𝐶 𝛿 − 𝐶 𝛼 + 𝑖𝑤 − 𝛼𝐿0,𝑤
𝑆𝑤 𝑐𝑤 ҧ 𝑡 𝑚,𝑡𝛿𝑒 𝑒 𝑐𝑤 ҧ 𝐿𝛼,𝑤
𝑆𝑡 ℓ𝑡 𝜕𝜖
− 𝜂𝐶 1− 𝛼 + 𝑖𝑡 − 𝜖𝑑,0 + 𝜖𝑒 𝛿𝑒
𝑆𝑤 𝑐𝑤 ҧ 𝑡 𝐿𝛼,𝑡 𝜕𝛼

• Taking the derivative with respect to AOA gives the static stability criterion:
𝜕𝐶𝑚 ℓ𝑤 𝑆𝑡 ℓ𝑡 𝜕𝜖
=− 𝐶 − 𝜂𝐶 1− <0
𝜕𝛼 ҧ 𝐿𝛼,𝑤 𝑆𝑤 𝑐𝑤
𝑐𝑤 ҧ 𝑡 𝐿𝛼,𝑡 𝜕𝛼
• Now, we can analyze how to ensure static stability of the wing and aft tail
combination by studying the contributions of both wing and aft tail

57
 Static Stability in Pitch
• The static stability criterion for the wing and aft tail configuration:

𝜕𝐶𝑚 ℓ𝑤 𝑆𝑡 ℓ𝑡 𝜕𝜖
Cm𝛼 = =− 𝐶𝐿𝛼,𝑤 − 𝜂𝑡 𝐶𝐿𝛼,𝑡 1 − <0
𝜕𝛼 𝑐𝑤
ҧ 𝑆𝑤 𝑐𝑤
ҧ 𝜕𝛼
Wing contribution Aft tail contribution

𝜕𝜖
• The downwash gradient < 1, and so the aft tail is always stabilizing. The
𝜕𝛼
stabilizing influence of the tail is proportional to tail volume 𝑆𝑡 ℓ𝑡
𝑆𝑡 ℓ𝑡
– Often expressed as horizontal tail volume ratio 𝑉ℎ = ҧ
𝑆𝑤 𝑐𝑤

58
 Static Stability in Pitch
𝜕𝐶𝑚 ℓ𝑤 𝑆𝑡 ℓ𝑡 𝜕𝜖
Cm𝛼 = =− 𝐶𝐿𝛼,𝑤 − 𝜂𝑡 𝐶𝐿𝛼,𝑡 1 − <0
𝜕𝛼 𝑐𝑤
ҧ 𝑆𝑤 𝑐𝑤
ҧ 𝜕𝛼

• The wing’s contribution depends on the sign of the wing’s moment arm ℓ𝑤
– If ℓ𝑤 > 0 (CG forward of wing AC), the wing is stabilizing. However, the tail
will carry negative lift for all operating conditions
– If ℓ𝑤 < 0 (CG aft of wing AC), the wing is de-stabilizing. In that case, the tail
volume ratio 𝑉ℎ has to be large enough to ensure that 𝐶𝑚𝛼 < 0

• Important: ℓ𝑤 , ℓ𝑡 , 𝑉ℎ are all affected by CG location. CG location has a dramatic

impact on pitch stability

59
 Stick-Fixed Neutral Point & Static Margin
• Just like the pitching moment about the aerodynamic center of an airfoil is invariant
to AOA, a similar point exists for the entire airplane. It is called stick-fixed neutral
point
– If the CG is located at this point, the airplane is neutrally stable (𝐶𝑚𝛼 = 0)

• We can find this point for the wing-tail combination by solving for the CG location
that gives 𝐶𝑚𝛼 = 0
Numerator: Pitch stiffness of
wing + tail combination, 𝐶𝑚𝛼
ℓ𝑤 𝑆ℓ 𝜕𝜖
ℓ𝑛𝑝 𝐶𝐿𝛼,𝑤 + 𝑡 𝑡 𝜂𝑡 𝐶𝐿𝛼,𝑡 1 −
𝑐ҧ 𝑆𝑤 𝑐𝑤
ҧ 𝜕𝛼
=− 𝑤
𝑐𝑤
ҧ 𝑆 𝜕𝜖
𝐶𝐿𝛼,𝑤 + 𝑡 𝜂𝑡 𝐶𝐿𝛼,𝑡 1 −
𝑆𝑤 𝜕𝛼
Denominator: Lift curve slope
of wing + tail combination, 𝐶𝐿𝛼

60
 Stick-Fixed Neutral Point & Static Margin
ℓ𝑤 𝑆ℓ 𝜕𝜖
ℓ𝑛𝑝 𝐶𝐿𝛼,𝑤 + 𝑡 𝑡 𝜂𝑡 𝐶𝐿𝛼,𝑡 1 −
𝑐ҧ 𝑆𝑤 𝑐𝑤
ҧ 𝜕𝛼
=− 𝑤
𝑐𝑤
ҧ 𝑆 𝜕𝜖
𝐶𝐿𝛼,𝑤 + 𝑡 𝜂𝑡 𝐶𝐿𝛼,𝑡 1 −
𝑆𝑤 𝜕𝛼

ℓ𝑛𝑝
• This yields a very important relationship: 𝐶𝑚𝛼 = − ҧ
𝐶𝐿𝛼
𝑐𝑤
ℓ𝑛𝑝
– The quantity ҧ
is called the stick-fixed static margin: the distance between
𝑐𝑤
CG and stick-fixed neutral point, expressed as fraction of MAC of main wing

• Static stability requires a positive stick-fixed static margin, which requires that
ℓ𝑛𝑝 > 0 , i.e., CG must be located forward of stick-fixed neutral point

61
Effect of Airspeed and CG Location on Trim

𝜕𝜖
𝑊 = 2700 𝑙𝑏, 𝑆𝑤 = 180 𝑓𝑡 2 , 𝑆𝑡 = 36 𝑓𝑡 2 , 𝑐𝑤
ҧ = 5.45 𝑓𝑡, 𝑐𝑡ҧ = 3 𝑓𝑡, 𝐶𝐿𝛼,𝑤 = 4.44, 𝐶𝐿𝛼,𝑡 = 3.97, 𝐶𝑚,𝑎𝑐,𝑡,𝛿𝑒 = −0.55, 𝜂𝑡 = 1.0, = 0.44,
𝜕𝛼
𝜖𝑑,0 = 0.04041, 𝜖𝑒 = 0.60, 𝛼𝐿0,𝑤 = −2.2∘ , 𝐶𝑚,𝑎𝑐,𝑤 = −0.053

62
 Wing & Canard Combination

𝜸 𝑽∞ ℓ𝒄
ℓ𝒘
𝑻 𝑳𝒄
𝝐𝒖 𝑽𝒚 𝒊𝒄
𝑽∞ 𝑫𝒄
𝑴𝒂𝒄,𝒄 𝑳𝒘
𝑾
𝒊𝒘
𝜹𝒄 𝑫𝒘
𝑴𝒂𝒄,𝒘

• The assumptions of this simplified analysis are:

– Fuselage reference line (FRL) taken to be in
direction of flight, passing through CG
– The thrust line and the aerodynamic centers (AC) of
both wing and tail lie on FRL

63
Simplified Pitch Equilibrium (Trim) Analysis
Support weight through lift:
𝑆𝑐 𝑊 cos 𝛾
𝐶𝐿𝑤 + 𝐶𝐿𝑐 = = 𝐶𝑊 cos 𝛾
𝑆𝑤 1 2
𝜌 𝑉 𝑆
2 ∞ ∞ 𝑤
Balance out (trim) pitching moments:
𝑆𝑐 𝑐𝑐ҧ ℓ𝑤
𝐶𝑚 = 𝐶𝑚,𝑎𝑐,𝑤 + 𝐶𝑚,𝑎𝑐,𝑐 − 𝐶
𝑆𝑤 𝑐𝑤 ҧ ҧ 𝐿,𝑤
𝑐𝑤
𝑆𝑐 ℓ𝑐
− 𝐶 =0
ҧ 𝐿,𝑐
𝑆𝑤 𝑐𝑤

𝑐𝑤
ҧ 𝑆𝑐 𝑐𝑐ҧ ℓ𝑐 For the wing to carry positive lift
𝐶𝐿,𝑤 = 𝐶𝑚,𝑎𝑐,𝑤 + 𝐶𝑚,𝑎𝑐,𝑐 − 𝐶𝑊 cos 𝛾
ℓ𝑤 − ℓ𝑐 𝑆𝑤 𝑐𝑤 ҧ 𝑐𝑤
ҧ (𝐶𝐿,𝑤 > 0), the canard must be
located ahead of the CG (ℓ𝑐 < 0)
> 𝟎 ≤ 𝟎 ≤ 𝟎 >𝟎
Unlike aft-tail configurations, in a
𝑐𝑐ҧ 𝑆𝑤 ℓ𝑤 𝑆𝑤 𝑐𝑤
ҧ canard-wing configuration, the canard
𝐶𝐿,𝑐 = 𝐶𝑊 cos 𝛾 − 𝐶 − 𝐶𝑚,𝑎𝑐,𝑐
ℓ𝑤 − ℓ𝑐 𝑆𝑐 𝑐𝑐ҧ 𝑆𝑐 𝑐𝑐ҧ 𝑚,𝑎𝑐,𝑤 always carries positive lift (𝐶𝐿,𝑐 > 0)
- For aero efficiency, canards are
> 𝟎 >𝟎 ≤ 𝟎 ≤ 𝟎 designed with high aspect ratio

64
 Simplified Pitch Stability Analysis
• Recall the pitching moment of the canard-wing configuration
𝑆𝑐 𝑐𝑐ҧ ℓ𝑤 𝑆𝑐 ℓ𝑐
𝐶𝑚 = 𝐶𝑚,𝑎𝑐,𝑤 + 𝐶𝑚,𝑎𝑐,𝑐 − 𝐶𝐿,𝑤 − 𝐶 𝐶𝐿,𝑐 = 𝐶𝐿𝛼,𝑐 (𝛼 + 𝑖𝑐 − 𝛼𝐿0,𝑐 + 𝜖𝑢 + 𝜖𝑐 𝛿𝑐 )
𝑆𝑤 𝑐𝑤 ҧ 𝑐𝑤
ҧ ҧ 𝐿,𝑐
𝑆𝑤 𝑐𝑤

𝐶𝐿,𝑤 = 𝐶𝐿𝛼,𝑤 (𝛼 + 𝑖𝑤 − 𝛼𝐿0,𝑤 )

𝜕𝐶
• Differentiating w.r.t. 𝛼 and enforcing the static pitch stability criterion 𝐶𝑚𝛼 = 𝑚 <
𝜕𝛼
0 yields the condition for static stability in pitch for the canard-wing configuration:

𝜕𝐶𝑚 ℓ𝑤 𝑆𝑐 ℓ𝑐 𝜕𝜖𝑢
= − 𝐶 − 𝐶 1+ <0
𝜕𝛼 ҧ 𝐿𝛼,𝑤 𝑆𝑤 𝑐𝑤
𝑐𝑤 ҧ 𝐿𝛼,𝑐 𝜕𝛼
Since CG is ahead of AC Since CG is aft of AC of
of wing (ℓ𝑤 > 0), the canard (ℓ𝑐 < 0), the canard is
wing is always stabilizing always de-stabilizing

• Thus, it is the main wing (and not the canard) that provides static stability in pitch
for a canard-wing configuration
65
 Understanding how a Canard is Different
𝑆𝑐 (−ℓ𝑐 )
• Similar to an aft tail, it is possible to define a canard volume ratio, 𝑉𝑐 = ҧ
𝑆𝑤 𝑐𝑤

ℓ𝑤 𝐶𝐿𝑤 𝐶𝑚,𝑎𝑐,𝑤 𝑆𝑐 𝑐𝑐ҧ 𝐶𝑚,𝑎𝑐,𝑐

• The trim requirement expressed in terms of 𝑉𝑐 : 𝑉𝑐 = ҧ 𝐶𝐿 𝑐
− − ҧ
𝑐𝑤 𝐶𝐿 𝑐 𝑆𝑤 𝑐𝑤 𝐶𝐿 𝑐

ℓ𝑤 𝐶𝐿𝛼,𝑤
• The 𝐶𝑚𝛼 < 0 requirement in terms of 𝑉𝑐 : 𝑉𝑐 < ҧ 𝐶𝐿𝛼,𝑐 1+𝜕𝜖𝑢 /𝜕𝛼
𝑐𝑤

• Canard sizing is a tradeoff between pitch control and pitch stability

𝐶𝐿,𝑤 𝐶𝐿𝛼,𝑤 ҧ 𝐶𝑚,𝑎𝑐,𝑤
𝑐𝑤 𝑆𝑐 𝑐𝑐ҧ 𝐶𝑚,𝑎𝑐,𝑐
• Combining the above relationships: < + +
𝐶𝐿,𝑐 𝐶𝐿𝛼,𝑐 (1+𝜕𝜖𝑢 /𝜕𝛼) ℓ𝑤 𝐶𝐿,𝑐 𝑆𝑤 ℓ𝑤 𝐶𝐿,𝑐

typically < 𝟏 typically < 𝟎

• Therefore, 𝐶𝐿𝑐 > 𝐶𝐿,𝑤 , which means that the canard will stall before the main wing
– It is very difficult to stall the main wing through a gradual maneuver

66
 Vertical Offset of Wing, Tail, and Thrust

ℓ𝒘 𝑽∞
𝑻 𝝐𝒅
ℓ𝒕 𝑽𝒚 𝑳𝒕
𝑳𝒘 𝒊𝒕
𝒉𝑻 𝑴𝒂𝒄,𝒕
𝜽 𝜶 𝒊𝒘 𝑴𝒂𝒄,𝒘
𝑽∞
𝜸 𝑫𝒕
𝜽=𝜸+𝜶 𝑫𝒘
𝜽 𝜹𝒆
𝒉𝒘
𝑾

• Wing, tail, and thrust line now have 𝒉𝒕

vertically offset from the FRL
• The freestream velocity is no longer
parallel to the FRL

67
 Vertical Offset of Wing, Tail, and Thrust
𝑆𝑡
𝐶𝑇 = 𝐶𝑊 sin 𝜃 + 𝐶𝐷,𝑤 cos 𝛼 − 𝐶𝐿,𝑤 sin 𝛼 + 𝜂 𝐶 cos(𝛼 − 𝜖𝑑 ) − 𝐶𝐿,𝑡 sin(𝛼 − 𝜖𝑑 )
Force balance in the 𝑆𝑤 𝑡 𝐷,𝑡
x and z directions 𝑆𝑡
𝐶𝑊 cos 𝜃 = 𝐶𝐿,𝑤 cos 𝛼 + 𝐶𝐷,𝑤 sin 𝛼 + 𝜂 𝐶 cos(𝛼 − 𝜖𝑑 ) + 𝐶𝐷,𝑡 sin(𝛼 − 𝜖𝑑 )
𝑆𝑤 𝑡 𝐿,𝑡

ℓ𝑤 ℎ𝑤 ℎ𝑤 ℓ𝑤
Pitching moment balance. 𝐶𝑚 = 𝐶𝑚,𝑎𝑐,𝑤 − cos 𝛼 + sin 𝛼 𝐶𝐿,𝑤 + cos 𝛼 − sin 𝛼 𝐶𝐷,𝑤
𝑐𝑤
ҧ 𝑐𝑤 ҧ 𝑐𝑤ҧ 𝑐𝑤 ҧ
Note 1: there is a moment 𝑆𝑡 𝑐𝑡ҧ 𝑆𝑡 ℓ𝑡 𝑆𝑡 ℎ𝑡
due to the thrust axis offset + 𝜂𝑡 𝐶𝑚,𝑎𝑐,𝑡 − cos 𝛼 − 𝜖𝑑 + sin 𝛼 − 𝜖𝑑 𝜂𝑡 𝐶𝐿,𝑡
𝑆𝑤 𝑐𝑤 ҧ 𝑆𝑤 𝑐𝑤ҧ 𝑆𝑤 𝑐𝑤ҧ
Note 2: 𝐶𝑚𝛼 = 𝜕𝐶𝑚 /𝜕𝛼 𝑆𝑡 ℎ𝑡 𝑆𝑡 ℓ𝑡 ℎ𝑇
+ cos(𝛼 − 𝜖𝑑 ) − sin(𝛼 − 𝜖𝑑 ) 𝜂𝑡 𝐶𝐷,ℎ − 𝐶𝑇 = 0
will now be a function of 𝛼 𝑆𝑤 𝑐𝑤 ҧ 𝑆𝑤 𝑐𝑤 ҧ 𝑐𝑤ҧ

𝐶𝐿,𝑤 = 𝐶𝐿𝛼 ,𝑤 (𝛼 + 𝑖𝑤 − 𝛼𝐿0,𝑤 ) 𝐶𝐿,𝑡 = 𝐶𝐿𝛼,𝑡 (𝛼 + 𝑖𝑡 − 𝛼𝐿0,𝑡 − 𝜖𝑑 + 𝜖𝑒 𝛿𝑒 )

In addition to lift and pitching
moment about AC, we also 2 2
𝐶𝐷,𝑤 = 𝐶𝐷0,𝑤 + 𝐾1,𝑤 𝐶𝐿,𝑤 + 𝐾2,𝑤 𝐶𝐿,𝑤 𝐶𝐷,𝑡 = 𝐶𝐷0,𝑡 + 𝐾1,𝑡 𝐶𝐿,𝑡 + 𝐾2,𝑡 𝐶𝐿,𝑡
have to build up the drag on
the wing and tail 𝐶𝑚,𝑎𝑐,𝑡 = 𝐶𝑚,𝑎𝑐,𝑡𝛿𝑒 𝛿𝑒

This set of nonlinear equations has to be solved numerically to determine the trim point, and also to
determine the static stability at the trim point (which may be different at different trim points)

68
 Contribution of Fuselage and Nacelles
• Contribution of the fuselage and nacelles to static stability is typically destabilizing (often, the
effect can be quite significant)

• Aero forces and moments on the fuselage are complex and affected by interactions with wing
and tail. Computer simulations and/or wind tunnel tests are typically required

• However, for a first approximation in early design, the effect of the fuselage and nacelles on
airplane lift is usually neglected. The effect of moment may be assessed using an
experimental correlation (Hoak , 1960):
3
𝑆𝑓 ℓ𝑓 𝑑𝑓 2
Δ𝐶𝑚 𝑓𝑢𝑠 = −2 1 − 1.76 𝛼𝑓
𝑆𝑤 𝑐𝑤
ҧ 𝑐𝑓

• 𝑆𝑓 : max fuselage cross-sectional area, 𝑐𝑓 : length of fuselage, 𝑑𝑓 : diameter of circle of area 𝑆𝑓 ,

ℓ𝑓 : distance that fuselage center of pressure is aft of aircraft CG (typically, ℓ𝑓 < 0), 𝛼𝑓 :
AOA of fuselage minimum drag axis

Hoak, D.E., “USAF Stability and Control

69 Datcom”, AFWAL-TR-83-3048, 1960
 Contribution of Running Propellers
• Affect both trim and static stability due to (1) direct effects: due to forces and moments on prop and (2)
Indirect effects: due to complex prop slipstream interactions
• Considering the direct effects only: ℎ𝑝 2 𝑑𝑝2 ℓ2𝑝 𝐶𝑁,𝑝𝛼
Δ𝐶𝑚 𝑝 = − 𝐶𝑇 − 𝛼𝑝
𝑐𝑤ҧ 𝑆𝑤 𝑐𝑤ҧ 𝐽2
• Prop incidence angle depends on downwash 𝜖𝑑,𝑝 and also 𝑵𝒑
thrust axis inclination 𝛼0𝑝 : 𝛼𝑝 = 𝛼 − 𝜖𝑑,𝑝 + 𝛼0𝑝
−ℓ𝒑
• The net impact can be represented as 𝑻
Δ𝐶𝑚 𝑝 = Δ𝐶𝑚0
𝑝
+ Δ𝐶𝑚𝛼
𝑝
𝛼 𝜶𝒑
𝑽∞
• The first term affects trim:
ℎ𝑝 2 𝑑𝑝2 ℓ𝑝 𝐶𝑁,𝑝𝛼 𝒉𝒑
Δ𝐶𝑚0 = − 𝐶𝑇 − (𝛼0𝑝 − 𝜖𝑑,0𝑝 )
𝑝 𝑐𝑤
ҧ 𝑆𝑤 𝑐𝑤ҧ 𝐽2
• The second term is the prop’s effect on static stability 𝑇
𝐶𝑇 =
2 𝑑𝑝2 ℓ𝑝 𝐶𝑁,𝑝𝛼 1
Δ𝐶𝑚𝛼 =− 1 − 𝜕𝜖𝑑,𝑝 /𝜕𝛼 𝜌∞ 𝑉∞2 𝑆𝑤
𝑝 𝑆𝑤 𝑐𝑤ҧ 𝐽2 2
𝐶𝑁,𝑝𝛼
• Tractor prop: destabilizing [ℓ𝑝 < 0 → Δ𝐶𝑚𝛼 𝑝
> 0] has to be read off a chart
𝐽2
• Pusher prop: stabilizing [ℓ𝑝 > 0 → Δ𝐶𝑚𝛼 𝑝
< 0] 𝑑𝑝 : prop diameter
Image: http://images.fineartamerica.com/images- Phillips, W.F., “Mechanics of Flight”, John
medium-large-5/2-old-crow-p-51-mustang-white- Wiley & Sons Inc., 2004
background-craig-tinder.jpg 70
 Contribution of Running Jet Engines
• Similar to a propeller, running jet engines also affect both trim and static stability in pitch
through direct and indirect effects.
– Direct effects: Due to moments from thrust offset and normal force
– Indirect effects: Due to flow entrainment by the jet, interaction of jet exhaust with other
airplane surfaces (neglected in early design)

• In a very similar manner to a propeller, the net pitching moment impact due to a jet can be
represented as Δ𝐶 = Δ𝐶 + Δ𝐶 𝛼
𝑚 𝑗 𝑚0 𝑗 𝑚𝛼 𝑗

𝜼𝒑,𝒊 : ideal propulsive

ℎ𝑗 𝜂𝑝,𝑖 ℓ𝑗 efficiency of jet engine
Δ𝐶𝑚0 = −𝐶𝑇 + 𝛼0𝑗 − 𝜖𝑑,0𝑗
𝑗 𝑐𝑤
ҧ 2(1 − 𝜂𝑝,𝑖 ) 𝑐𝑤
ҧ
Geometric parameters are
identical to those shown
𝜂𝑝,𝑖 ℓ𝑗 for the propeller airplane,
Δ𝐶𝑚𝛼 = −𝐶𝑇 1 − 𝜕𝜖𝑑,𝑗 /𝜕𝛼
𝑝 2 1 − 𝜂𝑝,𝑖 𝑐𝑤
ҧ with subscript ‘j’ for jet,
instead of ‘p’ for propeller

Phillips, W.F., “Mechanics of Flight”, John

71 Wiley & Sons Inc., 2004
 Contents
• Fundamental Concepts and Reference Frames
• Aircraft Equations of Motion
• Longitudinal Stability
– Flying Wing, Wing-Tail, and Canard-Wing Configurations
– Additional factors affecting longitudinal stability
• Longitudinal Control and Maneuverability
• Lateral/Directional Stability, Control, and Maneuverability
• Stability in Steady Flight
– Longitudinal Modes
• Phugoid and Short Period Modes
• Open Loop Response to Control Inputs
– Lateral Modes
• Additional Topics

72
Some Key Stability & Control Derivatives
Lift Force Side Force Roll Mom. Pitch Mom. Yaw Mom.
(𝐶𝐿 ) (𝐶𝑌 ) (𝐶𝑙 ) (𝐶𝑚 ) (𝐶𝑛 )
AOA (𝛼) 𝑪 𝑳𝜶 𝑪𝒎𝜶
෠ሶ
AOA rate (𝛼) 𝑪𝑳𝜶෡ሶ 𝑪𝒎𝜶෡ሶ

Sideslip (𝛽) 𝑪 𝒀𝜷 𝑪𝒍𝜷 𝑪𝒏𝜷

Roll rate (𝑝)Ƹ 𝑪𝒀𝒑ෝ 𝑪𝒍 𝒑ෝ 𝑪𝒏𝒑ෝ

Pitch rate (𝑞)

ො 𝑪𝑳𝒒ෝ 𝑪𝒎𝒒ෝ

Yaw rate (𝑟)Ƹ 𝑪𝒀𝒓ො 𝑪𝒍𝒓ො 𝑪𝒏𝒓ො

Aileron (𝛿𝑎 ) 𝑪𝒍𝜹𝒂 𝑪𝒏𝜹𝒂

Elevator (𝛿𝑒 ) 𝑪𝑳𝜹𝒆 𝑪𝒎𝜹𝒆

Rudder (𝛿𝑟 ) 𝑪𝒀𝜹𝒓 𝑪𝒍𝜹𝒓 𝑪𝒏𝜹𝒓

1st or 2nd order “stiffness” Rate damping derivatives

Control powers Roll/yaw axis coupling

73
 Effect of Pitch Rate on Lift and Pitching Moment
• Consider an airplane in a pull-up maneuver. 𝚫𝜶(.) 𝑽∞
𝒒
ℓ.
ҧ
𝑐𝑤 𝑽∞
The pitch rate 𝑞 is non-dimensionalized as 𝑞ො = 𝑞 ℓ. ℓ.
2𝑉∞
𝚫𝜶(.) = tan−𝟏 𝒒 ≈𝒒
𝑽∞ 𝑽∞
• The pitch rate changes the wing and tail incidence angles
𝑞ℓ𝑤 2ℓ𝑤 𝑞ℓ𝑡 2ℓ𝑡
– For wing: Δ𝛼𝑤 = = ҧ
𝑞ො and for tail: Δ𝛼𝑡 = = ҧ
𝑞ො
𝑉∞ 𝑐𝑤 𝑉∞ 𝑐𝑤

𝑆𝑡
• Change in airplane 𝐶𝐿 due to the pitch rate: Δ𝐶𝐿 𝑞 = 𝐶𝐿𝛼,𝑤 Δ𝛼𝑤 + 𝜂 CL𝛼,𝑡 Δ𝛼𝑡
𝑆𝑤 𝑡
2ℓ𝑤 𝑆𝑡 ℓ𝑡
Δ𝐶𝐿 𝑞 = 𝐶𝐿𝛼,𝑤 + 2 𝜂C 𝑞ො = 𝑪𝑳𝒒ෝ 𝑞ො
𝑐𝑤
ҧ ҧ 𝑡 L𝛼,𝑡
𝑆𝑤 𝑐𝑤
ℓ ℓ 𝑆𝑡
• Change in airplane 𝐶𝑚 due to pitch rate: Δ𝐶𝑚 𝑞 = − 𝑐𝑤ҧ 𝐶𝐿𝛼,𝑤 Δ𝛼𝑤 − 𝑐ҧ 𝑡 𝜂𝑡 CL𝛼,𝑡 Δ𝛼𝑡
𝑤 𝑤 𝑆𝑤
2ℓ2𝑤 𝑆𝑡 ℓ2𝑡
Δ𝐶𝑚 𝑞 = − 2 𝐶𝐿𝛼,𝑤 − 2 𝜂C 𝑞ො = 𝑪𝒎𝒒ෝ 𝑞ො
𝑐𝑤
ҧ 𝑆𝑤 𝑐𝑤ҧ2 𝑡 L𝛼,𝑡

Phillips, W.F., “Mechanics of Flight”, John

74 Wiley & Sons Inc., 2004
 Effect of Pitch Rate on Lift and Pitching Moment

2ℓ𝑤 𝑆𝑡 ℓ𝑡
Δ𝐶𝐿 𝑞= 𝐶𝐿𝛼,𝑤 + 2 𝜂𝑡 CL𝛼,𝑡 𝑞ො = 𝑪𝑳𝒒ෝ 𝑞ො
𝑐𝑤
ҧ 𝑆𝑤 𝑐𝑤
ҧ
2ℓ2𝑤 𝑆𝑡 ℓ2𝑡
Δ𝐶𝑚 𝑞 = − 2 𝐶𝐿𝛼,𝑤 − 2 𝜂C 𝑞ො = 𝑪𝒎𝒒ෝ 𝑞ො
𝑐𝑤
ҧ 𝑆𝑤 𝑐𝑤ҧ2 𝑡 L𝛼,𝑡

• The coefficient𝑪𝑳𝒒ෝ represents change in lift due to pitch rate

• The coefficient 𝑪𝒎𝒒ෝ represents change in pitching moment due to pitch rate and is
called the pitch damping coefficient (derivative).

• They have very important effects on aircraft dynamic response in pitch

Phillips, W.F., “Mechanics of Flight”, John

75 Wiley & Sons Inc., 2004
 Elevator Angle Per g
Lift and pitching moment equations:
• Recall that for the pull-up maneuver:
𝑛−1 𝑔 𝑔𝑐𝑤ҧ ෝ + 𝑪𝑳𝜹𝒆 𝜹𝒆
𝑪𝑳 = 𝑪𝑳𝟎 + 𝑪𝑳𝜶 𝜶 + 𝑪𝑳𝒒ෝ 𝒒
– Pitch rate: 𝑞 = 𝑉∞
→ 𝑞ො = 𝑛 − 1 2 𝑉∞ 2
ෝ + 𝑪𝒎𝜹𝒆 𝜹𝒆
𝑪𝒎 = 𝑪𝒎𝟎 + 𝑪𝒎𝜶 𝜶 + 𝑪𝒎𝒒ෝ 𝒒
– Also: 𝐿 = 𝑛 𝑊 → 𝐶𝐿 = 𝑛 𝐶𝑊
• Using these to solve the lift and pitching moment equations for the elevator
deflection:
𝑔𝑐𝑤ҧ
𝑛𝐶𝑊 − 𝐶𝐿,0 𝐶𝑚𝛼 + 𝐶𝐿𝛼 𝐶𝑚0 + 𝐶𝐿𝛼 𝐶𝑚𝑞ෝ − 𝐶𝐿𝑞ෝ 𝐶𝑚𝛼 (𝑛 − 1)
2 𝑉∞2
𝛿𝑒 = −
𝐶𝐿𝛼 𝐶𝑚𝛿𝑒 − 𝐶𝐿𝛿𝑒 𝐶𝑚𝛼

• Differentiating this w.r.t. load factor gives elevator angle per g:

𝑔𝑐𝑤ҧ
𝐶𝑊 𝐶𝑚𝛼 + (𝐶𝐿𝛼 𝐶𝑚𝑞ෝ − 𝐶𝐿𝑞ෝ 𝐶𝑚𝛼 )
𝜕𝛿𝑒 2 𝑉∞2
=−
𝜕𝑛 𝐶𝐿𝛼 𝐶𝑚𝛿𝑒 − 𝐶𝐿𝛿𝑒 𝐶𝑚𝛼

Phillips, W.F., “Mechanics of Flight”, John

76 Wiley & Sons Inc., 2004
 Elevator Angle Per g
• Elevator angle per g:
𝑔𝑐𝑤ҧ
𝐶𝑊 𝐶𝑚𝛼 + (𝐶𝐿𝛼 𝐶𝑚𝑞ෝ − 𝐶𝐿𝑞ෝ 𝐶𝑚𝛼 )
𝜕𝛿𝑒 2 𝑉∞2
=−
𝜕𝑛 𝐶𝐿𝛼 𝐶𝑚𝛿𝑒 − 𝐶𝐿𝛿𝑒 𝐶𝑚𝛼
• This provides a measure of airplane maneuverability
– Smaller magnitude implies more maneuverable aircraft (less elevator needed per g)
– But if it is too small, the airplane will become difficult to fly

What is the impact of:

- Airspeed?
- Weight?
- CG position?
- pitch damping?
- elevator control power?

Phillips, W.F., “Mechanics of Flight”, John

77 Wiley & Sons Inc., 2004
 The Stick-Fixed Maneuver Point
ℓ𝑛𝑝
• Using the results 𝐶𝑚𝛼 = − ҧ
𝐶𝐿𝛼 and ℓ𝑛𝑝 = 𝑥𝑛𝑝 − 𝑥𝑐𝑔 , in the 𝜕𝛿𝑒 /𝜕𝑛 relationship:
𝑐𝑤
𝑔𝑐𝑤ҧ 𝑔𝑐𝑤 ҧ
𝐶𝑊 − 𝐶𝐿𝑞ෝ 𝐶𝑚𝑞ෝ
𝜕𝛿𝑒 2𝑉∞2 2𝑉∞2 𝑥𝑛𝑝 𝑥𝑐𝑔
=− 𝑥𝑛𝑝 − 𝑥𝑐𝑔 − −
𝜕𝑛 𝑔𝑐𝑤
ҧ 𝑐𝑤ҧ 𝑐𝑤
ҧ
𝐶 𝑚 𝛿𝑒 + 𝐶𝐿𝛿𝑒 𝐶𝑊 − 𝐶𝐿𝑞ෝ 2
𝑐𝑤 ҧ 2𝑉∞

𝜕𝛿𝑒
• There is a CG location for which the red term in the above becomes zero, giving =0
𝜕𝑛

𝜕𝛿𝑒
• The CG location that gives = 0 is called the stick-fixed maneuver point
𝜕𝑛
• The distance between CG and SFMP is called the stick-fixed maneuver margin

Phillips, W.F., “Mechanics of Flight”, John

78 Wiley & Sons Inc., 2004
 The Stick-Fixed Maneuver Point
𝑔𝑐𝑤ҧ 𝑔𝑐𝑤 ҧ
𝐶𝑊 − 𝐶𝐿𝑞ෝ 𝐶𝑚𝑞ෝ
𝜕𝛿𝑒 2𝑉∞2 2𝑉∞2 𝑥𝑛𝑝 𝑥𝑐𝑔
=− 𝑥𝑛𝑝 − 𝑥𝑐𝑔 − −
𝜕𝑛 𝑔𝑐𝑤
ҧ 𝑐𝑤ҧ 𝑐𝑤
ҧ
𝐶 𝑚𝛿𝑒 + 𝐶𝐿𝛿𝑒 𝐶𝑊 − 𝐶𝐿𝑞ෝ 2
𝑐𝑤 ҧ 2𝑉∞

𝜕𝛿𝑒
• The stick-fixed MP = 0 is always aft of the stick-fixed neutral point 𝐶𝑚 𝛼 = 0
𝜕𝑛
– Stick-fixed maneuver margin is always greater than stick-fixed static margin

• Note: A statically unstable airplane (stick-fixed SM < 0) can be reliably flown by a human
pilot
– Requires a Stability Augmentation System (SAS), i.e., computer control
• Warning! Even a computerized SAS cannot control an airplane with negative stick-fixed
maneuver margin
– The CG must always be kept forward of the SFMP in all airplanes

Phillips, W.F., “Mechanics of Flight”, John

79 Wiley & Sons Inc., 2004
 Control Surface Hinge Moments and Trim Tabs
• Consider the case of the elevator (but the
concept applies to all hinged controls)
• To trim the aircraft at a given flight condition
requires a certain elevator deflection
– At all but one deflection, airflow will cause
a hinge moment about the hinge-line
– In a reversible flight control system, the pilot will have to apply a force to the control column to
offset this hinge moment. This will be quite… fatiguing
• A trim tab is used to offset (zero out) the hinge moment, and “trim” the control. Thus, the
elevator “trims” the airplane. The elevator tab “trims” out the elevator control force
– All of this can be studied using hinge moment coefficients
𝐶𝐻,𝑒 = 𝐶𝐻,𝑒0 + 𝐶𝐻,𝑒𝛼 𝛼𝑡 + 𝐶𝐻,𝑒𝛿𝑒 𝛿𝑒 + 𝐶𝐻,𝑒𝛿 𝛿𝑡𝑡
𝑡 𝑡𝑡

• “Letting go” of the yoke/stick (“stick-free”) results in the elevator

“floating” to a position 𝛿𝑒,𝑓𝑟𝑒𝑒 where 𝐶𝐻,𝑒 = 0
Trim
• “Trimming out” the elevator control force means setting 𝛿𝑡𝑡 such
that 𝐶𝐻,𝑒 = 0, at the elevator angle that you need for trim.
wheel
The trim tab allows 𝛿𝑒,𝑡𝑟𝑖𝑚 = 𝛿𝑒,𝑓𝑟𝑒𝑒 → 𝐶𝐻,𝑒 = 0
Image: https://upload.wikimedia.org/wikipedia/commons/1/10/Cessna-172-trim-control.jpg

80
 How Heavy Are the Controls? Control Force Gradient
𝐻𝑒
• Consider a simplified relationship between hinge moment and control force: 𝐹𝑝 = ℓ𝑔

– ℓ𝑔 represents the effective gearing of the elevator control system

• The control force (that the pilot must apply) can be expressed as (skipping derivation)

𝑆𝑒 𝑐𝑒ҧ 𝑉∞2 𝜌∞ 𝑔 𝑆𝑤 𝑐𝑤
ҧ
𝐹𝑝 = 𝜂𝑡 𝐶1 𝑛 − 2 𝑊 + 𝐶2 𝑛 − 1
𝑆𝑤 ℓ𝑔 𝑉𝑡𝑟𝑖𝑚 4

• Note that at trimmed steady-level flight, 𝑉∞ = 𝑉𝑡𝑟𝑖𝑚 , 𝑛 = 1 → 𝐹𝑝 = 0

• Differentiating 𝐹𝑝 w.r.t. 𝑉∞ with 𝑉 = 𝑉𝑡𝑟𝑖𝑚 , 𝑛 = constant gives the control force gradient

𝜕𝐹𝑝 𝑆𝑒 𝑐𝑒ҧ 𝑊
= −2 𝜂𝑡 𝐶
𝜕𝑉∞ 𝑉=𝑉𝑡𝑟𝑖𝑚 𝑆𝑤 ℓ𝑔 1 𝑉𝑡𝑟𝑖𝑚
𝑛= const

Phillips, W.F., “Mechanics of Flight”, John

81 Wiley & Sons Inc., 2004
 How Heavy Are the Controls? Control Force Gradient

𝜕𝐹𝑝 𝑆𝑒 𝑐𝑒ҧ 𝑊
= −2 𝜂𝑡 𝐶1
𝜕𝑉∞ 𝑉=𝑉𝑡𝑟𝑖𝑚 𝑆𝑤 ℓ𝑔 𝑉𝑡𝑟𝑖𝑚
𝑛= const

• Note from the above that controls become “heavier” at higher gross weight 𝑊 and lower trim
speeds 𝑉∞ = 𝑉𝑡𝑟𝑖𝑚
– They also increase with 𝑐𝑒ҧ , which is ∝ airplane size

𝝏𝑭𝒑
• There is a certain CG location at which the control force gradient = 𝟎. This is called
𝝏𝑽∞
the stick-free neutral point
– It is the neutral point if no control force were exerted on a hypothetical frictionless,
massless flight control system

Phillips, W.F., “Mechanics of Flight”, John

82 Wiley & Sons Inc., 2004
 How Heavy Are the Controls? Control Force Per g
• The control force (that the pilot must apply) can be expressed as (skipping derivation)
𝑆𝑒 𝑐𝑒ҧ 𝑉∞2 𝜌∞ 𝑔 𝑆𝑤 𝑐𝑤ҧ
𝐹𝑝 = 𝜂𝑡 𝐶1 𝑛 − 2 𝑊 + 𝐶2 𝑛 − 1
𝑆𝑤 ℓ𝑔 𝑉𝑡𝑟𝑖𝑚 4

• Taking a derivative w.r.t. load factor at constant airspeed gives the longitudinal control force
per g:
𝜕𝐹𝑝 𝑆𝑒 𝑐𝑒ҧ 𝜌∞ 𝑔 𝑆𝑤 𝑐𝑤ҧ
= 𝜂𝑡 𝐶1 𝑊 + 𝐶2
𝜕𝑛 𝑉=𝑐𝑜𝑛𝑠𝑡 𝑆𝑤 ℓ𝑔 4
• Note that it increases in magnitude with increasing gross weight and decreasing altitude
𝝏𝑭𝒑
• There is a CG location where the control force per g becomes zero, = 𝟎. This is called
𝝏𝒏
the stick-free maneuver point 𝜕𝜖𝑑
𝐶𝑚 𝛿 𝐶𝐻,𝑒𝛼 1− − 𝐶𝑚𝛼 𝐶𝐻,𝑒𝛿𝑒
Constants 𝐶1 , 𝐶2 depend on aircraft stability and 𝑒 𝑡 𝜕𝛼
𝐶1 =
control derivatives as well as hinge moment 𝐶𝐿𝛼 𝐶𝑚𝛿𝑒 − 𝐶𝐿𝛿𝑒 𝐶𝑚𝛼
derivatives 𝜕𝜖
2ℓℎ 𝐶𝐻,𝑒𝛼 𝐶𝐿𝛿𝑒 𝐶𝑚𝑞ෝ − 𝐶𝐿𝑞ෝ 𝐶𝑚𝛿𝑒 𝐶𝐻,𝑒𝛼 1 − 𝑑 − 𝐶𝐿𝛼 𝐶𝑚𝑞ෝ − 𝐶𝐿 qෝ 𝐶𝑚𝛼 𝐶𝐻,𝑒𝛿𝑒
𝑡 𝑡 𝜕𝛼
𝐶2 = +
𝑐𝑤
ҧ 𝐶𝐿𝛼 𝐶𝑚𝛿𝑒 − 𝐶𝐿𝛿𝑒 𝐶𝑚𝛼

83
 Summary – Neutral & Maneuver Points
Sign convention: “Pull” force Stick-free Stick-fixed Stick-free Stick-fixed
and “down elevator” are positive
Neutral Neutral Maneuver Maneuver
If CG lies ahead of all neutral and Point Point Point Point
maneuver points: (Zero control (Zero pitch (Zero control (Zero elevator
Increasing 𝑛 requires 𝜕𝛿𝑒 < 0 force gradient) stiffness) force per g) angle per g)
“up” elevator 𝜕𝑛
𝐶𝑚𝛼 = 0 𝜕𝐹𝑝 𝜕𝛿𝑒
Increasing 𝑛 requires 𝜕𝐹𝑝 > 0 𝜕𝐹𝑝
=0 =0
“pull” force 𝜕𝑛 =0 𝜕𝛿𝑒 𝜕𝑛 𝜕𝑛
𝜕𝑉∞ =0
𝜕𝛿𝑒 𝜕𝑉∞
Increasing 𝑉∞ requires >0
“down” elevator 𝜕𝑉∞

Increasing 𝑉∞ requires 𝜕𝐹𝑝

<0
“push” force 𝜕𝑉∞
Aft

Statically Uncon-
Stable Statically trollable
Fwd Unstable

84
 Ground Effect
• Flow around the aircraft is significantly modified when it flies in close proximity to
the ground (typically within one wingspan) – this is referred to as “ground effect”
• The proximity to the ground reduces the downwash for wing and other components
– Increases lift curve slope, reduces induced drag
• Ground effect typically produces a substantial rearward shift in the neutral point
for a conventional airplane
– Increase is wing and tail lift curve slopes tend to move the N.P. forward
– However, the dominant effect is from reduction in aft tail downwash, which
increases the stabilizing effect of an aft tail, moving the N.P. aft
• Thus, ground effect increases the amount of up-elevator needed to trim at low
airspeeds
– Also increases the control force gradient, making reversible controls feel heavier
• The combination of (1) low airspeed, (2) forward C.G., and (3) ground effect under
approach/landing conditions can be a critical constraint on elevator sizing

Phillips, W.F., “Mechanics of Flight”, John

85 Wiley & Sons Inc., 2004
 Effect of Deployed Flaps and Landing Gear
• Deploying landing gear generates nose-down pitching moment, as it is below the CG
• Deploying flaps (1) increases max lift coefficient, (2) reduces zero-lift AOA, (3) causes
additional nose-down pitching moment about wing AC, and (4) increases wing drag

• Recall the solution for trim angle of attack 𝛼𝑡𝑟𝑖𝑚 and elevator angle 𝛿𝑒,𝑡𝑟𝑖𝑚 for flight at
weight coefficient 𝐶𝑊 . The changes in these quantities due to flaps + gear deployment:
𝐶𝑊 − 𝐶𝐿0 𝐶𝑚𝛿𝑒 + 𝐶𝐿𝛿𝑒 𝐶𝑚0 −𝐶𝑚𝛿𝑒 Δ𝐶𝐿0 + 𝐶𝐿𝛿𝑒 Δ𝐶𝑚0
𝑓𝑔 𝑓𝑔 Note: denominators
𝛼𝑡𝑟𝑖𝑚 = Δ𝛼𝑡𝑟𝑖𝑚 𝑓𝑔 =
𝐶𝐿𝛼 𝐶𝑚𝛿𝑒 + 𝐶𝐿𝛿𝑒 𝑙𝑛𝑝 /𝑐𝑤
ҧ 𝐶𝐿𝛼 𝐶𝑚𝛿𝑒 + 𝐶𝐿𝛿𝑒 𝑙𝑛𝑝 /𝑐𝑤
ҧ are negative
𝑙𝑛𝑝 𝐶𝑚𝛿𝑒 < 0, 𝐶𝐿𝛿𝑒 > 0
𝐶𝑊 − 𝐶𝐿0 𝑙𝑛𝑝 /𝑐𝑤
ҧ − 𝐶𝑚0 − Δ𝐶𝐿0 − Δ𝐶𝑚0
𝛿𝑒,𝑡𝑟𝑖𝑚 = 𝑐𝑤ҧ 𝑓𝑔 𝑓𝑔
𝛿𝑒,𝑡𝑟𝑖𝑚 = |𝐶𝑚𝛿𝑒 > 𝐶𝐿𝛿𝑒 |
𝐶𝑚𝛿𝑒 + 𝐶𝐿𝛿𝑒 𝑙𝑛𝑝 /𝑐𝑤
ҧ 𝑓𝑔
𝐶𝑚𝛿𝑒 + 𝐶𝐿𝛿𝑒 𝑙𝑛𝑝 /𝑐𝑤
ҧ

• When flaps + gear are deployed, Δ𝐶𝐿0 > 0 and Δ𝐶𝑚0 < 0. The net effect:
𝑓𝑔 𝑓𝑔
– Angle of attack is significantly reduced in landing configuration (good: better visibility)
– For low static margin, more up-elevator. For high static margin, some down-elevator
Phillips, W.F., “Mechanics of Flight”, John
86 Wiley & Sons Inc., 2004
 Elevator Sizing
• The elevator must be sized to have sufficient authority to trim the airplane to achieve 𝐶𝐿𝑚𝑎𝑥 at
forward CG and in ground effect with gear extended
– For both flaps extended and flaps retracted (either may occur during landing)
– For a sized elevator, this constraint may be used to specify forward CG limit
– If forward CG limit has been otherwise established, this constraint can be used to size the
elevator
– Elevators are typically a constant fraction of the
stabilizer chord, extending up to 95% or to the tip
of the stabilizer (Raymer)
– Flutter tendencies should be minimized by using
mass balancing and aerodynamic balancing (the
latter also reduces required control forces)
• For transonic and supersonic airplanes, the a) Supersonic usually all-moving only.
entire horizontal stabilizer rotates b) Often all-moving plus elevator.

– If this is done to aid in trim, it is called a trimmable horizontal stabilizer

– If this is done to aid in control as well, it is often called a “stabilator”
Raymer, D., “Aircraft Design: A Conceptual
87 Approach”, 4th Ed. AIAA Education Series
Let’s Check Out Some Empennages
“Horn” type Trimmable Horizontal
aerodynamic Stabilizer on
balance commercial jet
transport. This allows
the aircraft to be
trimmed with the
elevator faired (neutral).
The elevators provide
primary pitch control
Elevator
F-16 aircraft with stabilators.
Hinge-line
There is no separate elevator.
The stabilator provides pitch
GA aircraft elevator with notch stability and pitch control
or “horn”-type aerodynamic
balance. Located forward of the
hinge-line, the horn deflects into
the flow in the opposite direction
of the elevator, and cancels out
some of the hinge moment
http://cdn.c.photoshelter.com/img-get2/I0000XGiIYk_2JrY/fit=1000x750/MG-8445.jpg
https://i.stack.imgur.com/9RQY4.jpg
https://en.wikipedia.org/wiki/Stabilator 88
 Contents
• Fundamental Concepts and Reference Frames
• Aircraft Equations of Motion
• Longitudinal Stability
– Flying Wing, Wing-Tail, and Canard-Wing Configurations
– Additional factors affecting longitudinal stability
• Longitudinal Control and Maneuverability
• Lateral/Directional Stability, Control, and Maneuverability
• Stability in Steady Flight
– Longitudinal Modes
• Phugoid and Short Period Modes
• Open Loop Response to Control Inputs
– Lateral Modes
• Additional Topics

89
 Yaw Stiffness (Directional / Weathervane Stability)
• Analogous to pitch stiffness, a directionally
stable aircraft should develop a moment that
tends to reduce any sideslip disturbance.

• The criterion for positive yaw stiffness (static

stability in yaw) is:
𝜕𝐶
𝐶𝑛𝛽 = 𝑛 > 0
𝜕𝛽

• The primary stabilizing contribution is from

an aft tail
• Contributions from fuselage and propeller are
typically destabilizing
• Wing contribution is not significant unless
the wing is highly swept

90
 Effect of Vertical Tail on Side Force
• Some immediate analogies with the longitudinal case:
𝑆𝑣 ℓ𝑣
– Vertical tail volume ratio: 𝑉𝑣 = , sidewash 𝜎 and
𝑆𝑤 𝑏𝑤
𝜕𝜎
sidewash gradient , vertical tail efficiency 𝜂𝑣
𝜕𝛽

• Contribution of vertical tail to side force:

𝑆𝑣 𝜕𝜎
Δ𝐶𝑌 𝑣 = −𝜂𝑣 𝐶 1− 𝛽 − 𝜖𝑠,0 + 𝑖𝑣 − 𝜖𝑟 𝛿𝑟 ℓ𝒗
𝑆𝑤 𝐿𝛼,𝑣 𝜕𝛽

• Taking a partial derivative w.r.t. sideslip 𝛽

𝑆𝑣 𝜕𝜎
𝑪𝒀𝜷 = −𝜂𝑣 𝐶 1−
𝑣 𝑆𝑤 𝐿𝛼,𝑣 𝜕𝛽
• Taking a partial derivative w.r.t. rudder angle 𝛿𝑟
𝑆𝑣
𝑪𝒀𝜹𝒓 = 𝜂𝑣 𝐶 𝜖
𝑆𝑤 𝐿𝛼,𝑣 𝑟

91
 Effect of Vertical Tail on Yawing Moment
• Considering the vertical tail moment arm ℓ𝑣 , the yawing moment contribution can
be summed up as:
Δ𝐶𝑛 𝑣
𝑆𝑣 ℓ𝑣 𝜕𝜎 𝑐𝑣ҧ
= 𝜂𝑣 𝐶𝐿𝛼,𝑣 1 − 𝛽 − 𝐶𝐿𝛼,𝑣 (𝜖𝑠,0 + 𝑖𝑣 ) − 𝜖𝑟 𝐶𝐿𝛼,𝑣 − 𝐶𝑚,𝑎𝑐,𝑣𝛿𝑟 𝛿𝑟
𝑆𝑤 𝑏𝑤 𝜕𝛽 ℓ𝑣

• Taking a partial derivative w.r.t. sideslip 𝛽

𝑆𝑣 ℓ𝑣 𝜕𝜎
𝑪 𝒏𝜷 = 𝜂 𝑣 𝐶𝐿𝛼,𝑣 1 −
𝑣 𝑆𝑤 𝑏𝑤 𝜕𝛽

• Taking a partial derivative w.r.t. rudder angle 𝛿𝑟

𝑆𝑣 ℓ𝑣 𝑐𝑣ҧ
𝑪𝒏𝜹𝒓 = −𝜂𝑣 𝐶 𝜖 − 𝐶
𝑆𝑤 𝑏𝑤 𝐿𝛼,𝑣 𝑟 ℓ𝑣 𝑚,𝑎𝑐,𝑣𝛿𝑟

92
 Effect of Fuselage + Nacelles
& Direct Effect of Propellers
• Fuselage + nacelles: Analogous to the preliminary estimate for contribution to pitch stability,
except with sideslip angle 𝛽 replacing angle of attack 𝛼
3 3
𝑆𝑓 ℓ𝑓 𝑑𝑓 2 𝑆𝑓 ℓ𝑓 𝑑𝑓 2
Δ𝐶𝑛 𝑓𝑢𝑠 =2 1 − 1.76 𝛽𝑓 𝐶𝑛𝛽 =2 1 − 1.76
𝑆𝑤 𝑏𝑤 𝑐𝑓 𝑓𝑢𝑠 𝑆𝑤 𝑏𝑤 𝑐𝑓

• Direct effect of propellers: Also analogous to the approach for pitch stability
Δ𝐶𝑛 𝑝 = Δ𝐶𝑛0 + Δ𝐶𝑛𝛼 𝛼 + Δ𝐶𝑛𝛽 𝛽
𝑝 𝑝 𝑝
The interaction of the
𝑦𝑏𝑝 2 𝑑𝑝3 𝐶𝑛,𝑝𝛼 𝐶𝑛,𝑝𝛼 < 0 for clockwise- rotating propeller
Δ𝐶𝑛0 =− 𝐶 + (𝛼0𝑝 − 𝜖𝑑,0𝑝 )
𝑝 𝑏𝑤 𝑇,𝑝 𝑆𝑤 𝑏𝑤 𝐽2 turning propellers. Thus slipstream with vertical
these create a yawing tail in particular also
moment to the left at generates yawing
2 𝑑𝑝3 𝐶𝑛,𝑝𝛼
Δ𝐶𝑛𝛼 = 1 − 𝜕𝜖𝑑,𝑝 /𝜕𝛼 positive AOA. This is moments (leftward, for
𝑝 𝑆𝑤 𝑐𝑤
ҧ 𝐽2
called “p-factor” clockwise prop)

2 𝑑𝑝2 ℓ𝑝 𝐶𝑁,𝑝𝛼 As before, tractor propellers (ℓ𝑝 < 0) are

Δ𝐶𝑛𝛽 = 1 − 𝜕𝜖𝑠,𝑝 /𝜕𝛽
𝑝 𝑆𝑤 𝑐𝑤ҧ 𝐽2 destabilizing in yaw, while pusher propellers
(ℓ𝑝 > 0) are stabilizing
Hoak, D.E., “USAF Stability and Control
93 Datcom”, AFWAL-TR-83-3048, 1960
 Turning Tendencies due to Rotating Propeller
• This discussion applies to propellers turning clockwise when viewed from behind.
The effects will reverse from counter-clockwise turning propellers

• Overall, a clockwise propeller has a left turning tendency, due to 4 causes:

– Torque
– P-Factor
– Gyroscopic moment
– Slipstream interaction

94
 Rudder Sizing and Control Power (𝐶𝑛𝛿𝑟 )
• The rudder has to be sized to permit aircraft
control for a variety of constraining conditions:
– Coordinated turns: against adverse yaw (most
significant when flying at low airspeed)
– Crosswind landings: rudder must be powerful enough to overcome the natural
yaw stiffness of the aircraft and allow alignment with runway
– Asymmetric power conditions: Must permit positive directional control after
failure of the critical engine at minimum controllable airspeed
– Spin recovery: The rudder must be powerful enough to oppose and stop the
yaw rotation in a spin and permit recovery from the spin

𝑆𝑣 ℓ𝑣 𝑐𝑣ҧ 𝑐𝑣ҧ
𝑪𝒏𝜹𝒓 = −𝜂𝑣 𝐶 𝜖 − 𝐶 = −𝜂𝑣 𝑉𝑣 𝐶𝐿𝛼,𝑣 𝜖𝑟 − 𝐶𝑚,𝑎𝑐,𝑣𝛿𝑟
𝑆𝑤 𝑏𝑤 𝐿𝛼,𝑣 𝑟 ℓ𝑣 𝑚,𝑎𝑐,𝑣𝛿𝑟 ℓ𝑣

95
 Roll Stiffness (Dihedral Effect, 𝐶𝑙𝛽 )
• Unlike pitch and yaw, there is no 1st order roll stiffness
• Airplanes still have an inherent tendency to fly with wings
level due to the presence of wing dihedral
• When there is a roll disturbance, a component of weight in
the direction of the roll causes a sideslip to develop in that
direction (i.e., if 𝜙 > 0, then 𝛽 > 0)
• The presence of dihedral causes the “windward” wing to see
a higher angle of attack. This develops a restoring moment that 2Γ
𝐶𝑙𝛽 =− 𝜅 𝐶
reduces the roll (𝜙 > 0 → 𝛽 > 0 → 𝐶𝑙 < 0) Γw 3𝜋 𝑙 𝐿𝛼,𝑤
Γ: dihedral (rad)
𝜅𝑙 : chart look-up
• The derivative 𝐶𝑙𝛽 < 0 is used as a measure of the dihedral effect. The mounting
position of the wing relative to fuselage (low, mid, or high) also has an impact, with
high wing having the strongest roll stiffness and low wing the least
• Excessive roll stiffness may be dealt with using some negative dihedral, called
anhedral
96
 Roll Damping (𝐶𝑙𝑝ෝ )
• For most airplanes, the rolling motion is heavily
damped. The primary influence is from the wing

• As the aircraft rolls, the roll rate causes the down-

moving wing to see a higher angle of attack, and
thus more lift.

• This creates an opposing rolling moment

proportional to roll rate (i.e., a damping moment)

• This is captured through the roll damping

derivative, 𝐂𝐥𝐩ෝ
• Note that at very high roll rates, the down-moving
wing may stall. What then?

Phillips, W.F., “Mechanics of Flight”, John

97 Wiley & Sons Inc., 2004
 Aileron Sizing and Control Power (𝐶𝑙𝛿𝑎 )
Balance between rolling
• The aileron control power 𝐶𝑙𝛿𝑎 denotes the rolling moments due to maximum
aileron input and roll damping
moment authority of the ailerons. Ailerons are typically
𝑝 𝑏 𝐶𝑙 = 𝐶𝑙𝛿𝑎 Δ𝛿𝑎 + 𝐶𝑙𝑝ෝ 𝑝Ƹ𝑠𝑠 = 0
sized based on a steady-state roll rate criterion 𝑠𝑠 𝑤
2𝑉0 𝑝𝑠𝑠 𝑏𝑤
𝐶𝑙𝛿𝑎 Δ𝛿𝑎 + 𝐶𝑙𝑝ෝ =0
2𝑉∞
𝑝𝑠𝑠 𝑏𝑤 𝐶𝑙𝛿𝑎 Δ𝛿𝑎
• An undesirable side-effect of aileron deflection is =−
2𝑉∞ 𝐶𝑙𝑝ෝ
adverse yaw, caused by the down-deflected aileron
generating more drag Low to medium
maneuverability aircraft, e.g.,
• This causes a yawing moment in the opposite direction cargo & transport airplanes:
as the rolling moment 𝑝𝑠𝑠 𝑏𝑤
≥ 0.07
– Captured using the derivative 𝐂𝐧𝛅𝐚 2𝑉∞
High maneuverability aircraft, e.g.,
fighters and aerobatic airplanes
𝑝𝑠𝑠 𝑏𝑤
≥ 0.09
2𝑉∞

98
 Aileron Sizing and Control Power (𝐶𝑙𝛿𝑎 )
• Adverse yaw can be reduced or eliminated using:
– Differential aileron deflection: Deflecting
the up-deflected aileron more

– Frise ailerons: where the leading edge of the

up-deflected aileron projects into the flow to
create drag
https://www.aviationsafetymagazine.com/features/two-
aileron-types/

– Differential spoiler deflections: deploying

spoilers only on the down-moving wing. This
creates rolling and yawing moments in the
same direction

https://www.airlineratings.com/did-you-know/what-
are-spoilers-and-what-are-they-used-for/

99
A Look at Lateral Control Surfaces

Flaperon Flaps

Rudder

Aileron

100
 Aeroelastic Effects
• Common flexing occurs in the fuselage bending longitudinally and spanwise, as well
as torsional bending of the wing
• Flexing of a swept wing will cause the lift-curve slope to be reduced
– Negatively effects stability
– Typical swept-wing transport at high subsonic speed can see a wing lift-curve
slope reduction of about 20%, tail pitching moment reduction of 30%, 50% less
elevator effectiveness, and the wing aerodynamic center moved 10% forward
• Aileron effectiveness can be reduced to over 100% due to ailerons twisting the
wing in the opposite direction of aileron travel (known as aileron reversal)

101
 Aircraft Performing a Steady Turn
Longitudinal
• Build up the
Lift 𝐶𝐿 = 𝐶𝐿0 + 𝐶𝐿𝛼 𝛼 + 𝐶𝐿𝑞ෝ 𝑞ො + 𝐶𝐿𝛿𝑒 𝛿𝑒
longitudinal and
lateral forces and Pitching moment 𝐶𝑚 = 𝐶𝑚0 + 𝐶𝑚𝛼 𝛼 + 𝐶𝑚𝛽 𝛽 + 𝐶𝑚𝑞ෝ 𝑞ො + 𝐶𝑚𝛿 𝛿𝑒
𝑒
moments
Lateral
• Assumption: Side force 𝐶𝑌 = 𝐶𝑌𝛽 𝛽 + 𝐶𝑌𝑝ෝ 𝑝Ƹ + 𝐶𝑌ෝ𝑟 𝑟Ƹ + 𝐶𝑌𝛿𝑟 𝛿𝑟
Excess power
equation is Rolling moment 𝐶𝑙 = 𝐶𝑙𝛽 𝛽 + 𝐶𝑙𝑝ෝ 𝑝Ƹ + 𝐶𝑙ෝ𝑟 𝑟Ƹ + 𝐶𝑙𝛿𝑎 𝛿𝑎 + 𝐶𝑙𝛿𝑟 𝛿𝑟
satisfied implicitly
Yawing moment 𝐶𝑛 = 𝐶𝑛𝛼 𝛼 + 𝐶𝑛𝛽 𝛽 + 𝐶𝑛𝑝ෝ 𝑝Ƹ + 𝐶𝑛ෝ𝑟 𝑟Ƹ + 𝐶𝑛𝛿𝑎 𝛿𝑎 + 𝐶𝑛𝛿𝑟 𝛿𝑟

𝑨 𝑿 𝑩 𝑪 𝑫
• Collect
𝐶𝐿𝛼 0 0 𝐶𝐿𝛿𝑒 0 𝐶𝐿 − 𝐶𝐿0 0 𝐶𝐿𝑞ෝ 0
relationships 𝜶
0 𝐶𝑌𝛽 0 0 𝐶𝑌𝛿𝑟 𝜷 𝐶𝑌 𝐶𝑌𝑝ෝ 0 𝐶𝑌ෝ𝑟 𝑝Ƹ
into matrix
0 𝐶𝑙𝛽 𝐶𝑙𝛿𝑎 0 𝐶𝑙𝛿𝑟 𝜹𝒂 = 𝐶𝑙 − 𝐶𝑙𝑝ෝ 0 𝐶𝑙ෝ𝑟 𝑞ො
equation
𝐶𝑚𝛼 𝐶𝑚𝛽 0 𝐶𝑚𝛿𝑒 0 𝜹𝒆 𝐶𝑚 − 𝐶𝑚0 0 𝐶𝑚𝑞ෝ 0 𝑟Ƹ
• We want to 𝜹𝒓
𝐶𝑛𝛼 𝐶𝑛𝛽 𝐶𝑛𝛿𝑎 0 𝐶𝑛𝛿𝑟 𝐶𝑛 𝐶𝑛𝑝ෝ 0 𝐶𝑛ෝ𝑟
solve for 𝑿
𝐴𝑿 = 𝐵 − 𝐶𝐷 → 𝑿 = 𝐴−1 𝐵 + 𝐴−1 𝐶𝐷
Phillips, W.F., “Mechanics of Flight”, John
102 Wiley & Sons Inc., 2004
 Aircraft Performing a Steady Turn
• For aircraft in steady, climbing, coordinated turn: 𝐶𝐿 = 𝑛 𝐶𝑊 cos 𝛾 = 𝐶𝑊 cos 𝛾 / cos 𝜙
• For a coordinated turn, the side force should be zero: 𝐶𝑌 = 0
• There are no unbalanced rolling, pitching, or yawing moments.
Therefore, 𝐶𝑙 = 𝐶𝑚 = 𝐶𝑛 = 0
• The turn rate is related to bank angle and speed as: Ω = 𝑔 tan 𝜙/𝑉∞
– If this is resolved in the body-fixed axes, we get the body-fixed angular velocities
𝑝 = −Ω sin 𝛾, 𝑞 = Ω cos 𝛾 sin 𝜙 , 𝑟 = Ω cos 𝛾 cos 𝜙
– Remember, these need to be converted to non-dimensional angular rates
𝑝𝑏𝑤 𝑞𝑐𝑤ҧ 𝑟𝑏𝑤
𝑝Ƹ = , 𝑞ො = , 𝑟Ƹ =
2𝑉∞ 2𝑉∞ 2𝑉∞

• We can now solve for 𝑋 = {𝛼, 𝛽, 𝛿𝑎 , 𝛿𝑒 , 𝛿𝑟 } over a range of bank angles 𝜙 ∈ [−60∘ , +60∘ ] to study the
steady turn performance. For simplicity, we shall do this for 𝛾 = 0
𝐶𝐿𝛼 0 0 𝐶𝐿𝛿𝑒 0 cos 𝛾 0 𝐶𝐿𝑞ෝ 0
𝜶 𝐶𝑊 − 𝐶𝐿0
0 𝐶𝑌𝛽 0 0 𝐶𝑌𝛿𝑟 𝜷 cos 𝜙 𝐶𝑌𝑝ෝ 0 𝐶𝑌𝑟ො −𝑏𝑤 sin 𝛾
𝐶𝑙𝛽 0 𝑔 tan 𝜙
0 𝐶𝑙𝛿𝑎 0 𝐶𝑙𝛿𝑟 𝜹𝒂 = − 𝐶𝑙𝑝ෝ 0 𝐶𝑙𝑟ො 𝑐𝑤
ҧ cos 𝛾 sin 𝜙
0 2𝑉∞2
𝐶𝑚𝛼 𝐶𝑚𝛽 0 𝐶𝑚𝛿𝑒 0 𝜹𝒆 0 𝐶𝑚𝑞ෝ 0 𝑏𝑤 cos 𝛾 cos 𝜙
−𝐶𝑚0
𝐶𝑛𝛼 𝐶𝑛𝛽 𝐶𝑛𝛿𝑎 𝐶𝑛𝛿𝑟 𝜹𝒓 𝐶𝑛𝑝ෝ 𝐶𝑛𝑟ො
0 0 0
Phillips, W.F., “Mechanics of Flight”, John
103 Wiley & Sons Inc., 2004
 Aircraft Performing a Steady Turn
We are already familiar AOA must increase with A coordinated turn
with the load factor vs. bank angle to generate requires 𝐶𝑌 = 0,
bank angle relationship the additional lift needed not zero sideslip

𝐶𝑙𝛽 < 0
𝐶𝑙𝑝ෝ < 0
𝐶𝑙𝑝 < 0
𝐶𝑙𝛿𝑎 < 0
𝐶𝑙ෝ𝑟 > 0
𝐶𝑙𝛿𝑟 > 0 “Hold bank with aileron”: “Bank-and-yank”: higher “Keep the ball centered”:
input direction depends on bank angles need more This ensures 𝐶𝑌 = 0, and
rolling moment balance up-elevator (“yank”) to requires rudder input
achieve the higher AOA “into” the turn
1
𝛿𝑎 = − (𝐶 𝛽 + 𝐶𝑙𝑝ෝ 𝑝Ƹ + 𝐶𝑙ෝ𝑟 𝑟Ƹ + 𝐶𝑙𝛿𝑟 𝛿𝑟 ) Phillips, W.F., “Mechanics of Flight”, John
𝐶𝑙𝛿𝑎 𝑙𝛽 104 Wiley & Sons Inc., 2004 (Example 6.7.1)
 Contents
• Fundamental Concepts and Reference Frames
• Aircraft Equations of Motion
• Longitudinal Stability
– Flying Wing, Wing-Tail, and Canard-Wing Configurations
– Additional factors affecting longitudinal stability
• Longitudinal Control and Maneuverability
• Lateral/Directional Stability, Control, and Maneuverability
• Stability in Steady Flight
– Longitudinal Modes
• Phugoid and Short Period Modes
• Open Loop Response to Control Inputs
– Lateral Modes
• Additional Topics

105
 State-Space Representation of Aircraft Dynamics
• Note that the linearized equations are of the form:
Δ𝑿ሶ = 𝐴 Δ𝑿 + 𝐵 Δ𝑼 Δ𝑋ሶ long = Along Δ𝑋long + Blong Δ𝑈𝑙𝑜𝑛𝑔
• This is called the state-space representation of a Δ𝑋ሶ lat = Alat Δ𝑋lat + Blat Δ𝑈𝑙𝑎𝑡
dynamical system, in which
𝒏 Δ𝑢 Δ𝑣
– 𝑿 ∈ 𝑹 : state vector Δ𝑝
Δ𝑤
– 𝑼 ∈ 𝑹𝒑 : control or input vector Δ𝑞 Δ𝑟
Δ𝑋long = Δ𝑥 Δ𝑋lat = Δ𝑦
– 𝑨 ∈ 𝑹𝒏 x 𝒏 : state or system matrix 𝑓 𝑓
– 𝑩 ∈ 𝑹𝒏 x 𝒑 : control or input matrix Δ𝑧𝑓 Δ𝜙
Δ𝜃 Δ𝜓

• To assess the dynamic stability of the uncontrolled (or

controls-fixed) system, we set Δ𝑼 = 𝟎 ഥ, and look at the Δ𝑈𝑙𝑜𝑛𝑔 = Δ𝛿𝑒
stability of the system Δ𝑿ሶ = 𝐴 Δ𝑿 Δ𝑈𝑙𝑎𝑡 =
Δ𝛿𝑎
Δ𝛿𝑟
• This can be done by solving the eigenproblem, which
yields the system’s eigenvalues and corresponding
eigenvectors

106
 Eigenvalues and Eigenvectors
Consider a simple 2 x 2 system
• Consider the eigenproblem for the system
2 1
Δ𝑿ሶ = 𝐴 Δ𝑿 𝐴=
1 2
Find eigenvalues of this system
• The eigenvalues 𝜆 = {𝜆1 , … , 𝜆𝑛 } are the roots of the (𝐴 − 𝜆𝐼) =
2−𝜆 1
characteristic equation: 1 2−𝜆
2
𝐝𝐞𝐭 𝑨 − 𝝀 𝑰 = 𝟎 det(𝐴 − 𝜆𝐼) = 2 − 𝜆 −1=0
𝜆1 = 1, 𝜆2 = 3
• The eigenvector 𝒗𝒊 for each eigenvalue 𝝀𝒊 is found by Find eigenvector for 𝝀𝟏 = 𝟏
solving the equation 2 1 𝑎1 𝑎1
𝑨𝒗𝒊 = 𝝀𝒊 𝒗 𝐴𝑣1 = 𝜆1 𝑣1 → = 1
1 2 𝑏1 𝑏1
0.707
𝑎1 + 𝑏1 = 0 → 𝑣1 =
• Eigenvectors are typically normalized to a magnitude −0.707
of unity Find eigenvector for 𝝀𝟐 = 𝟑
2 1 𝑎2 𝑎2
𝐴𝑣2 = 𝜆2 𝑣2 → = 3
1 2 𝑏2 𝑏2
• Eigenvalues and eigenvectors can be real or complex,
with implications on dynamic stability 0.707
𝑎2 − 𝑏2 = 0 → 𝑣2 =
0.707

107
 Interpretation of Eigenvalues
• Consider a general complex conjugate eigenvalue pair:

𝝀 = −𝝈 ± 𝒋 𝝎𝒅 𝝈: damping rate 𝝎𝒅 : damped natural frequency

• Purely real eigenvalue (𝜔𝑑 = 0):

- non-oscillatory mode, damping = 𝜎
– 𝑹𝒆 𝝀 = 𝟎 → 𝜎 = 0 : rigid-body mode, describing rigid-body displacement
– 𝑹𝒆 𝝀 < 𝟎 → 𝜎 > 0 : convergent mode, exponential return to equilibrium
– 𝑹𝒆 𝝀 > 𝟎 → 𝜎 < 0 : divergent mode, exponential deviation from equilibrium

• Complex pair of eigenvalues (𝜔𝑑 ≠ 0):

- oscillatory mode, damping = 𝜎, damped natural frequency = 𝜔𝑑
– 𝑹𝒆 𝝀 = 𝟎 → 𝜎 = 0 : undamped oscillatory mode
– 𝑹𝒆 𝝀 < 𝟎 → 𝜎 > 0 : damped oscillatory mode
– 𝑹𝒆 𝝀 > 𝟎 → 𝜎 < 0 : divergent oscillatory mode

108
 Interpretation of Eigenvalues
𝝀 = −𝝈 ± 𝒋 𝝎𝒅 𝝈: damping rate 𝝎𝒅 : damped natural frequency

Imaginary axis
𝜆 = −𝜎 + 𝑗 𝜔𝑑
Undamped natural frequency
𝜔𝑛 𝜔𝑑 = 𝜔𝑛 1 − 𝜁 2 𝜔𝑛 = 𝜎 2 + 𝜔𝑑2
cos −1 (𝜁) = 𝜃
Real axis Damping ratio
𝜎
𝜎 = 𝜁𝜔𝑛 𝜁=
Left-half plane Right-half plane 𝜎 2 + 𝜔𝑑2
Stable Unstable
(convergent) (divergent)

109
 Longitudinal Modes – Eigenproblem Solution
6 states, therefore 6 eigenvalues
Another complex
Δ𝑢
conjugate pair of Δ𝑤
eigenvalues represents Imaginary Δ𝑞
axis (𝜔𝑑 ) Δ𝑋long = Δ𝑥
a low frequency (long 𝑓
period), lightly Δ𝑧𝑓
damped mode called Δ𝜃
the phugoid mode

Real axis (𝜎)

The remaining two eigenvalues are

always zero, and represent rigid
body displacement modes. These
One complex conjugate pair of just indicate that one can fly at
eigenvalues represents a high frequency different latitudes and altitudes
(short period), highly damped mode
called the short period mode (not drawn to scale)

110
Longitudinal Modes: Phugoid & Short Period
Phugoid (Long Period) Mode Short Period Mode
• Aircraft exchanges translational KE & PE • Aircraft exchanges rotational KE & PE
• Speed leads the pitch angle (or FPA) by • There is negligible speed variation
about 90 deg in phase • Rapid AoA and pitch variations which
• Very little AoA variation lead variation in vertical displacement
• Long period, lightly damped. Damping (altitude) by about 90 deg
reduces with increasing L/D. Damping • If high frequency and heavily damped,
can even be negative at certain airspeeds airplane responds rapidly to elevator
• Light phugoid damping objectionable in input without undesirable overshoot
IFR flight, but usually not in VFR flight

111
 Short Period Mode Approximation
• Damped natural frequency (𝜔𝑑,𝑠𝑝 ): 𝜌𝑆𝑤 𝑐𝑤ҧ3
𝑅𝑚𝛼 = 𝐶
2𝑉0 𝑅𝑧𝛼 + 𝑅𝑚𝑞 + 𝑅𝑚𝛼෡ሶ 2 8 𝐼𝑦𝑦 𝑚𝛼
𝜔𝑑,𝑠𝑝 = 𝑅𝑧𝛼 𝑅𝑚𝑞ෝ − 𝑅𝑚𝛼 −
𝑐𝑤
ҧ 2 𝜌𝑆𝑤 𝑐𝑤ҧ3
𝑅𝑚𝑞ෝ = 𝐶
8 𝐼𝑦𝑦 𝑚𝑞ෝ
𝜌𝑆𝑤 𝑐𝑤
ҧ
• Damping rate (𝜎𝑠𝑝 ): 𝑅𝑧𝛼 = (−𝐶𝐿𝛼 − 𝐶𝐷 )
𝑊
𝑉0 4
𝑔
𝜎𝑠𝑝 = − 𝑅𝑧𝛼 + 𝑅𝑚𝑞ෝ + 𝑅𝑚𝛼෡ሶ
𝑐𝑤
ҧ 𝜌𝑆𝑤 𝑐𝑤ҧ3
𝑅𝑚𝛼෡ሶ = 𝐶
8 𝐼𝑦𝑦 𝑚𝛼෡ሶ

• Increasing the pitch damping derivative 𝐶𝑚𝑞ෝ and AOA-rate damping derivative 𝐶𝑚𝛼෡ሶ will increase short
period damping

• Increasing pitch stiffness |𝐶𝑚𝛼 | (static stability) will increase short period frequency

• Note that altitude effects enter some of the coefficients through density (𝜌). Damping is higher in the
denser lower atmosphere. The frequency does not vary much with altitude

Nelson, R.C., “Flight Stability and Automatic Phillips, W.F., “Mechanics of Flight”, John
Control”, 2nd Ed., McGraw Hill, 1997 112 Wiley & Sons Inc., 2004
 Phugoid Mode Approximation
• Approximate phugoid damped natural frequency (𝜔𝑑,𝑝 ): 𝑅𝑚𝛼
𝑅𝑠 =
2 𝑅𝑚𝛼 − 𝑅𝑧𝛼 𝑅𝑚𝑞ෝ
2𝑉0 𝑅𝑧𝜇 4𝑅𝑔𝑥 𝑅𝑥𝜇
𝜔𝑑,𝑝 = − 𝑅𝑠 − + 𝑅𝑑 𝑅𝑥𝛼 𝑅𝑚𝑞ෝ
𝑐𝑤
ҧ 2 𝑅𝑧𝜇 𝑅𝑧𝜇 𝑅𝑑 =
𝑅𝑚𝛼 − 𝑅𝑧𝛼 𝑅𝑚𝑞ෝ
𝑅𝑥𝛼 𝑅𝑚𝑞ෝ
• Approximate phugoid damping rate (𝜎𝑝 ): 𝑅𝑑 =
𝑅𝑚𝛼 − 𝑅𝑧𝛼 𝑅𝑚𝑞ෝ
2𝑉0 𝑅𝑧𝜇 𝑅𝑥𝜇
𝜎𝑝 = − + 𝑅𝑑 − 𝑅𝑝 𝑅𝑧 𝛼 + 𝑅𝑚𝑞ෝ
𝑐𝑤ҧ 2 𝑅𝑧𝜇 𝑅𝑝 = 𝑅𝑔𝑥 𝑅𝑠
𝑅𝑚𝛼 − 𝑅𝑧𝛼 𝑅𝑚𝑞ෝ
𝑔𝑐𝑤ҧ
• Phugoid damping can be increased by 𝑅𝑔𝑥 =
2 𝑉02
reducing the lift-to-drag ratio, but this is
𝜌𝑆𝑤 𝑐𝑤
ҧ 𝑇𝑉 cos 𝛼 𝑇0
clearly inefficient. An alternative (if phugoid) 𝑅𝑥𝜇 = −2𝐶𝐷 − 𝑀 𝐶𝐷𝑀 +
4𝑊/𝑔 1
damping is unacceptable is to use a phugoid 𝜌𝑉 𝑆
2 0 𝑤
suppression system (example of SAS)
𝜌𝑆𝑤 𝑐𝑤
ҧ 𝑇𝑉 sin(𝛼 𝑇0 )
𝑅𝑧𝜇 = −2𝐶𝐿 − 𝑀 𝐶𝐿𝑀 −
4𝑊/𝑔 1
• Damping and frequency are higher at denser, 𝜌𝑉 𝑆
2 0 𝑤
lower altitudes and lower at higher altitudes

Nelson, R.C., “Flight Stability and Automatic Phillips, W.F., “Mechanics of Flight”, John
Control”, 2nd Ed., McGraw Hill, 1997 113 Wiley & Sons Inc., 2004
Effect of CG Position and Tail Area on Longitudinal Modes
As the CG moves aft, the As tail area increases, static
frequency of both longitudinal stiffness increases, which
modes reduces (longer period) increases frequency of the modes
Imaginary Imaginary
axis (𝜔𝑑 ) axis (𝜔𝑑 )

Real axis (𝜎) Real axis (𝜎)

As the CG moves further As tail area increases, pitch and

rearward, the modes become AOA-rate damping derivatives also
aperiodic, and ultimately unstable increase in magnitude, which
increases short period mode damping
Nelson, R.C., “Flight Stability and Automatic
Control”, 2nd Ed., McGraw Hill, 1997 114
 Longitudinal Step Response
• Longitudinal motion
in response to a step
elevator input

• Short period motion

in initial response
• Phugoid motion
present for majority
of motion

115
 Longitudinal Step Response – Short Period
• Same step elevator
input but utilizing short
period approximation
– Constant forward
velocity and pitch angle

116
 Longitudinal Perturbation Response
• This example is initialized
with a small perturbation
in forward velocity

• What would this motion

look like if you were in a
wingman aircraft?
– Assume both aircraft are
trimmed to the same
starting velocity, altitude,
etc.

117
Longitudinal Perturbation – Relative Location
Note the scales here – lots of motion forward and back
with small motion up and down

118
 Lateral Modes – Eigenproblem Solution
A slowly 6 states, therefore 6 eigenvalues
convergent or Δ𝑣
divergent Imaginary Δ𝑝
motion called axis (𝜔𝑑 ) Δ𝑟
A lightly damped Δ𝑋lat = Δ𝑦
𝑓
the spiral mode
oscillatory motion with Δ𝜙
low frequency called Δ𝜓
the Dutch roll mode

Real axis (𝜎)

A highly convergent The remaining two eigenvalues are

non-oscillatory mode always zero, and represent rigid
called the roll mode body displacement modes. These
just indicate that one can fly at
different longitudes and headings

(not drawn to scale)

119
 The Roll Mode
• The roll mode is very heavily damped and is of no concern to pilots/passengers

• Because it is overdamped and highly convergent, transients in rolling motion are quickly damped out and
a steady rolling rate is quickly approached
– As a result, pilots tend to associate aileron input with the steady state roll rate (i.e., larger aileron
input yields higher steady state roll rate)

• If roll damping is insufficient, pilots will feel that they are commanding roll acceleration instead of roll
rate, and will give a poor handling quality rating

𝜌𝑆𝑤 𝑏𝑤2𝑉
0
• Approximate roll mode damping rate, 𝜎𝑟 = − 𝐶𝑙𝑝ෝ
4 𝐼𝑥𝑥
– Clearly, it is driven by the magnitude
of the roll damping derivative
– Roll damping reduces with reducing speed

120
 The Spiral Mode
• The spiral mode is usually convergent (damping reduces at lower speeds). If divergent, it is so slowly
divergent that the pilot usually corrects for it without any problem

• If an airplane with an unstable spiral mode is disturbed laterally, it will begin a slow spiral in the
direction of the disturbance.
– As the spiral deviation becomes large, longitudinal motion will also be induced and an ever-
tightening high-speed spiral dive may result
– This can be a problem if inadvertently entering clouds while VFR or suffering failure of attitude
indicating instruments
𝑔 𝐶𝑙𝛽 𝐶𝑛ෝ −𝐶𝑙ෝ 𝐶𝑛𝛽
𝑟 𝑟
• Approximate spiral mode damping rate: 𝜎𝑠 =
𝑉0 𝐶𝑙𝛽 𝐶𝑛 ෝ −𝐶𝑙 ෝ 𝐶𝑛𝛽
𝑝 𝑝

• Root locus analysis reveals that increasing the dihedral

effect (𝐶𝑙𝛽 more negative) makes the spiral mode more
stable. Increasing directional stability (𝐶𝑛𝛽 more positive)
makes the spiral mode less stable.

• However, the Dutch roll mode behaves in the exact opposite manner!

Nelson, R.C., “Flight Stability and Automatic Phillips, W.F., “Mechanics of Flight”, John
Control”, 2nd Ed., McGraw Hill, 1997 121 Wiley & Sons Inc., 2004
 The Dutch Roll Mode
• Damped oscillatory motion with combination of sideslip, roll, and yaw. Dutch Roll Simulation
– When amplitude is large, longitudinal motion is also induced (courtesy Prof. Eric
Johnson)
– Damping reduces as lower speeds

• Period for this motion is of the order of a few seconds. Thus, it can be very annoying or uncomfortable
for pilots and passengers if damping is insufficient

• Approximate damped natural frequency and damping rate for Dutch roll mode:
2
2𝑉0 𝑅𝑦𝛽 + 𝑅𝑛𝑟ො
𝜔𝑑,𝑑𝑟 = 1 − 𝑅𝑦rො 𝑅𝑛𝛽 + 𝑅𝑦𝛽 𝑅𝑛𝑟ො + 𝑅𝐷𝑠 −
𝑏𝑤 2
𝑉0
𝜎𝑑𝑟 = − 𝑅 + 𝑅𝑛𝑟ො − 𝑅𝐷𝑐 + 𝑅𝐷𝑝
𝑏𝑤 𝑦𝛽
• Dutch roll mode frequency increases if directional stability (𝐶𝑛𝛽 ) is increased.

• Increasing yaw damping (|𝐶𝑛ෝ𝑟 |) will result in better Dutch roll damping. On the other hand, increasing
dihedral effect (|𝐶𝑙𝛽 |) makes the Dutch roll less stable. These are opposite to what is observed for the
spiral mode

Nelson, R.C., “Flight Stability and Automatic Phillips, W.F., “Mechanics of Flight”, John
Control”, 2nd Ed., McGraw Hill, 1997 122 Wiley & Sons Inc., 2004
Spiral Mode vs. Dutch Roll Mode - Tradeoff
• Designing good Dutch roll characteristics With increasing Dutch Spiral
while maintaining acceptable roll, yaw, and magnitudes of Roll Mode Mode
spiral stability is challenging these derivatives… Damping Damping
• The only way to increase both Dutch roll and Roll stability |𝐶𝑙𝛽 | Decreases Increases
spiral damping is to increase yaw damping
(higher |𝐶𝑛ෝ𝑟 |) Roll damping |𝐶𝑙𝑝ෝ | Increases Decreases
– This is difficult to achieve using Yaw stability |𝐶𝑛𝛽 | Increases Decreases
geometric design changes alone
Yaw damping |𝐶𝑛ෝ𝑟 | Increases Increases

𝑆𝑣 ℓ𝑣 𝜕𝜎
𝑪𝒏𝜷 = 𝜂𝑣 𝐶𝐿𝛼,𝑣 1 −
𝑣 𝑆𝑤 𝑏𝑤 𝜕𝛽
Note: Attempting to
change the vertical tail ℓ𝑣
𝑆𝑣 ℎ𝑣 𝜕𝜎
𝑪𝒍𝜷 = −𝜂𝑣 𝐶𝐿𝛼,𝑣 1 − area to increase yaw
𝑣 𝑆𝑤 𝑏𝑤 𝜕𝛽 𝑆𝑣
damping (|𝐶𝑛𝑟ො |) will
ℎ𝑣
also result in changes
to the other derivatives
2 𝑆𝑣 ℓ2𝑣 𝜕𝜎
𝑪𝑛𝑟ො = −𝜂𝑣 𝐶
2 𝐿𝛼,𝑣
1 −
𝑣 𝑆𝑤 𝑏𝑤 𝜕𝛽

Nelson, R.C., “Flight Stability and Automatic Phillips, W.F., “Mechanics of Flight”, John
Control”, 2nd Ed., McGraw Hill, 1997 123 Wiley & Sons Inc., 2004
Let’s Have a Look!

124
 Contents
• Fundamental Concepts and Reference Frames
• Aircraft Equations of Motion
• Longitudinal Stability
– Flying Wing, Wing-Tail, and Canard-Wing Configurations
– Additional factors affecting longitudinal stability
• Longitudinal Control and Maneuverability
• Lateral/Directional Stability, Control, and Maneuverability
• Stability in Steady Flight
– Longitudinal Modes
• Phugoid and Short Period Modes
• Open Loop Response to Control Inputs
– Lateral Modes
• Additional Topics

125
 Beyond the Basics
• With an understanding of the basic notions of aircraft stability & control one may
begin to entertain more advanced considerations
– Design and integration of physical control systems
– Vehicle Management
– Vehicle Handling Qualities
– Stability Augmentation Systems
– Relaxed Stability
– Control-Configured Vehicle Designs

• Most of these topics are natural extension of the fundamental dynamics we have
covered already
– Generally requires supporting analysis to discern underlying vehicle dynamics
– Vehicle dynamic behavior can then be exploited or corrected

126
Control System Development

127
The Vehicle Management System (VMS)

128
VMS Actuation

129
 Open-Loop vs. Closed-Loop
• Earlier we discussed the longitudinal and Open-Loop System
lateral dynamic properties of aircraft
– “How does the aircraft respond to various Input State
Aircraft
inputs?”
• The “feel” of an aircraft (i.e. it’s open-loop
characteristics) are typically assessed using
Handling Qualities Ratings
• When the state of the vehicle is used to
Closed-Loop System
help determine the input (feedback
control) we “close the loop” on our system Command Input State
• Closed-loop system allows for: Aircraft
– Steady-state error reduction
– Improve open-loop characteristics
– Augment pilot capabilities
– Provide automatic control (e.g. autopilot) Controller

130
Cooper Harper Handling Quality Ratings

131
Aircraft and Flight Phase Classification

132
Longitudinal Mode Requirements from HQ Perspective
Short Period Mode

Phugoid (Long Period Mode)

Phillips, W.F., “Mechanics of Flight”, John

133 Wiley & Sons Inc., 2004
Lateral Mode Requirements from HQ Perspective
Roll Mode Spiral Mode

Long Period Mode

Dutch Roll Mode

For some modern aircraft with low wing area, the roll
damping may be quite low. Then, it is occasionally
possible for the usual roll and spiral modes to combine to
form a complex pair. In this case, the aircraft will exhibit
a long-period lateral oscillation, called the lateral
phugoid or the roll-spiral mode. It is not desirable, but
not objectionable either if there is sufficient damping

Phillips, W.F., “Mechanics of Flight”, John

134 Wiley & Sons Inc., 2004
 Stability Augmentation Systems (SAS)
• Recall the state-space representation of the
aircraft dynamical system:
𝑿ሶ = 𝐴 𝑿 + 𝐵 𝑼
• Suppose the modes (eigenvalues) of the uncontrolled system (𝑨) are either unstable
or undesirable (e.g. insufficient damping)
• Split the control 𝑼 into a parts supplied by pilot 𝑼𝒑 and the SAS 𝑼𝒔𝒂𝒔
𝑼 = 𝑼𝒑 + 𝑼𝒔𝒂𝒔
• Design 𝑼𝒔𝒂𝒔 to be a function of the feedback of aircraft states: 𝑼𝒔𝒂𝒔 = −𝑲𝑿
• The augmented system is now given by
𝑿ሶ = 𝐴 − 𝐵𝐾 𝑿 + 𝐵𝑼𝒑
• By tuning the elements (called gains) within matrix 𝑲, it is possible to
– Stabilize an otherwise unstable aircraft
– Modify an unacceptable mode to have acceptable characteristics
– Change the entire “feel” of the vehicle

135
 Relaxed Stability
• Most aircraft are designed to be
statically stable while
maintaining adequate
performance

• There are some classes of

aircraft where reduced static
stability, or even instability, is
accepted in order to gain other
system benefits

136
 Potential Benefits of Relaxed Stability
• Increased maneuverability is the most
common motivation for relaxed
stability
– As static margin is reduced the
vehicle’s pitch stiffness is also reduced
– Reduced stiffness allows for more
responsive vehicle maneuverability

• For more traditional vehicles without

large maneuverability demands,
relaxed stability can still be of interest
– Reduce the size of control surfaces
– Reduction of total drag at trim conditions
Perez, Ruben E., Hugh HT Liu, and Kamran Behdinan.
"Relaxed static stability aircraft design via longitudinal control-
configured multi-disciplinary design optimization
methodology." Canadian Aeronautics and Space Journal 52.1
(2006): 1-14.

137
 Control-Configured Vehicle Design
• Traditionally a vehicle’s controllers are designed in the
preliminary and detailed design stages
– Controllers are designed to augment the stability or performance of
a fixed design

• This traditional approach carries some inherent risk,

particularly for vehicles that require SAS to overcome
relaxed stability

138
 Control-Configured Vehicle Design
• Alternative methods to integrate control design considerations
alongside vehicle design are generally known as Control-
Configured Vehicle (CCV) Design

• General approaches balance vehicle characteristics with stability

augmentation to meet desired vehicle handling or
maneuverability requirements
– Requires greater level of detail for design knowledge (aerodynamics, mass
properties, etc.) in order to estimate S&C properties

• Some control considerations pertaining to vehicle details (e.g.

actuator dynamics) will be needed to refine controllers in later
design stages
139