Integrated Flight Dynamic Modeling of Flexible Aircraft with

Inertial Force-Propulsion-Aeroelastic Coupling

Nhan Nguyen

NASA Ames Research Center

Moffett Field, CA 94035
This paper presents an integrated ight dynamic modeling method for exible aircraft that captures cou-
pled physics effects due to inertial forces, aeroelasticity, and propulsive forces that are normally present in
ight. The present approach formulates the coupled ight dynamics using a structural dynamic modeling
method that describes the elasticity of a exible, twisted, swept wing using an equivalent beam-rod model. The
structural dynamic model allows for three types of wing elastic motion: apwise bending, chordwise bending,
and torsion. Inertial force coupling with the wing elasticity is formulated to account for aircraft acceleration.
The structural deections create an effective aeroelastic angle of attack that affects the rigid-body motion of
exible aircraft. The aeroelastic effect contributes to aerodynamic damping forces that can inuence aerody-
namic stability. For wing-mounted engines, wing exibility can cause the propulsive forces and moments to
couple with the wing elastic motion. The integrated ight dynamics for a exible aircraft are formulated by
including generalized coordinate variables associated with the aeroelastic-propulsive forces and moments in
the standard state-space form for six degree-of-freedom ight dynamics.
I. Introduction
Modern aircraft are increasingly designed to be highly maneuverable in order to achieve high-performance mission
objectives. Toward this goal, aircraft designers have been adopting light-weight, exible, high aspect ratio wings in
modern aircraft. Aircraft design concepts that take advantage of wing exibility to increase maneuverability have been
investigated. By twisting a wing structure, an aerodynamic moment can be generated to enable an aircraft to execute a
maneuver in place of the use of traditional control surfaces. For example, a rolling moment can be induced by twisting
the left and right wings in the opposite direction. Similarly, a pitching moment can be generated by twisting both
wings in the same direction. Wing twisting or warping for ight control is not a new concept and was used in the
Wright Flyer in the 1903. The U.S. Air Force conducted the Active Flexible Wing programin the 1980s and 1990s to
explore potential use of leading edge slats and trailing edge aps to increase control effectiveness of F-16 aircraft for
high speed maneuvers.
1
In the recent years, the Active Aeroelastic Wing research program also investigated a similar
technology to induce wing twist in order to improve roll maneuverability of F/A-18 aircraft.
2
Structural deections of lifting surfaces interact with aerodynamic forces to create aeroelastic coupling that can
affect aircraft performance. Understanding these effects can improve the prediction of aircraft ight dynamics and can
provide insight into how to design a ight control system that can reduce aeroelastic interactions with a rigid-body
ight controller. Generally, high aspect ratio lifting surfaces undergo a greater degree of structural deections than
low aspect ratio lifting surfaces. In general, a wing section possesses a lower stiffness than a horizontal stabilizer or a
vertical stabilizer. As a result, its natural frequency is normally present inside a ight control frequency bandwidth that
potentially can result in ight control interactions. For example, when a pilot commands a roll maneuver, the aileron
deections can cause one or more elastic modes of the wings to excite. The wing elastic modes can result in changes
to the intended aerodynamics of the wings, thereby potentially causing undesired aircraft responses. Aeroservoelastic
ltering is a traditional method for suppressing elastic modes, but this usually comes at an expense in terms of reducing
the phase margin in a ight control system.
3
If the phase margin is signicantly reduced, aircraft responses may
become more sluggish to pilot commands. Consequently, with a phase lag in the control inputs, potential pilot-

Research Scientist, Intelligent Systems Division, Mail Stop 269-1, AIAA Senior Member
1 of 20
American Institute of Aeronautics and Astronautics
induced oscillations (PIOs) can occur. Numerous studies have been made to increase the understanding of the role of
aeroservoelasticity in the design of ight control systems.
36, 15
Aeroelasticity remains a fertile eld of research in spite of many early contributions to this eld dated back to the
1930s. In the early days of National Advisory Committee for Aeronautics (NACA), numerous studies on aeroelas-
ticity by investigators, such as Theodorsen and Garrick, had made signicant advances in this eld.
711
Theodorsens
theory of unsteady aerodynamics still remains an essential tool for aeroelastic analysis. Recently, interests in coupled
ight-structural dynamics have seen a renewal. Due to exibility of modern aircraft structures, ight dynamic models
of rigid-body aircraft have limitations and cannot accurately predict behaviors of exible aircraft when elastic modes
participate in the rigid-body motion. Recently, some investigators have investigated theoretical approaches to devel-
oping integrated ight dynamics with aeroelasticity. Shearer develops an integrated ight dynamic model for a rep-
resentative High Altitude Long Endurance (HALE) vehicle.
13
Meirovitch and Tuzcu also develop another integrated
approach to ight dynamics of exible aircraft.
14, 15
In both of these approaches, detail inertial-aeroelastic-propulsive
force coupling in the governing structural dynamic partial differential equations were not made available.
The purpose of this study is to produce a modeling capability for integrated ight dynamics of exible aircraft
that can better predict some of the complex behaviors in ight due to multi-physics coupling. Some of the important
features in the present method are: inertial force coupling due to aircraft rigid-body acceleration, bending-torsion
coupling due to wing pre-wist, aeroelastic-propulsive force coupling due to engine mounting on exible wing struc-
tures, and lastly an extension of Theodorsens theory to include chordwise bending in the aeroelastic angle of attack.
Through the use of generalized coordinates that represent elastic deections, the standard ight dynamic equations for
six degree-of-freedommotion are modied to include effects of aeroelasticity and propulsive forces on exible wings.
II. Reference Frames
Fig. 1 - Aircraft Reference Frames
Figure 1 illustrates three orthogonal views of a typical aircraft. Several reference frames are introduced to facilitate
the rigid-body dynamic and structural dynamic analysis of the lifting surfaces. For example, the aircraft inertial
reference frame A is dened by unit vectors a
1
, a
2
, and a
3
xed to the non-rotating earth. The aircraft body-xed
reference frame B is dened by unit vectors b
1
, b
2
, and b
3
. The reference frames A and B are related by three
successive rotations: 1) the rst rotation about a
3
by the heading angle that results in an intermediate reference
frame A

dened by unit vectors a

1
, a

2
, and a

3
(not shown), 2) the second rotation about a

2
by the pitch angle
that results in an intermediate reference frame B

dened by unit vectors b

1
, b

2
, and b

3
(not shown), and 3) the third
2 of 20
American Institute of Aeronautics and Astronautics
rotation about b

1
by the bank angle that results in the reference frame B. This relationship can be expressed as

a
1
a
2
a
3

cos sin 0
sin cos 0
0 0 1

cos 0 sin
0 1 0
sin 0 cos

1 0 0
0 cos sin
0 sin cos

b
1
b
2
b
3

coscos sincos +cossin sin sinsin +cossin cos

sincos coscos +sinsin sin cossin +sinsin cos
sin cos sin cos cos

b
1
b
2
b
3

(1)
The left wing elastic reference frame D is dened by unit vectors d
1
, d
2
, and d
3
. The reference frames B and
D are related by three successive rotations: 1) the rst rotation about b
3
by the elastic axis sweep angle
3
2
that
results in an intermediate reference frame B

dened by unit vectors b

1
, b

2
, and b

3
(not shown), 2) the second rotation
about negative b

2
by the elastic axis dihedral angle that results in an intermediate reference frame D

dened by unit
vectors d

1
, d

2
, and d

3
(not shown), and 3) the third rotation about d

1
by an angle that results in the reference frame
D. This relationship can be expressed as

b
1
b
2
b
3

sin cos 0
cos sin 0
0 0 1

cos 0 sin
0 1 0
sin 0 cos

1 0 0
0 1 0
0 0 1

d
1
d
2
d
3

sincos cos sinsin

coscos sin cossin
sin 0 cos

d
1
d
2
d
3

(2)
Generally, the effect of the dihedral angle can be signicant. A full analysis with the dihedral angle can be
performed but can also result in a very complex analytical formulation. Thus, to simplify the analysis, the dihedral
effect is assumed to be negligible in this study. The right wing reference frame C can be established in a similar
manner. In the analysis, the aeroelastic effects on the fuselage, horizontal stabilizers, and vertical stabilizer are not
considered, but the analytical method can be formulated for analyzing these lifting surfaces if necessary. In general, a
whole aircraft analysis approach should be conducted to provide a comprehensive assessment of the effect of exibility
on aircraft stability. Such an analysis is feasible in the current framework as described in this study.
III. Elastic Analysis
In the subsequent analysis, the combined motion of the left wing is considered. The wing has a varying pre-
twist angle (x) common in many aircraft. Typically, the wing pre-twist angle varies from being nose-up at the wing
root to nose-down at the wing tip. The nose-down pre-twist at the wing tip is designed to delay stall onsets. Under
aerodynamic forces and moments, wing structural deections introduce strains in the wing structure. For high aspect
ratio wings, an equivalent beam approach can be used to analyze structural deections with a reasonable accuracy.
The equivalent beam approach is a typical formulation in many aeroelasticity studies.
12, 16
Experimental validation
can show that equivalent beam approach is accurate for an aspect ratio as low as 3:1. The internal structure of a wing
typically comprises a complex arrangement of load carrying spars and wing boxes. Nonetheless, the elastic behavior
of a wing can be captured by the use of equivalent stiffness properties. These properties can be derived from structural
certication testing that yields information about wing deection as a function of loading.
Consider an airfoil section on the left wing as shown in Fig. 2 undergoing bending and twist deections.
3 of 20
American Institute of Aeronautics and Astronautics
Fig. 2 - Left Wing Reference Frame
Let (x, y, z) be the coordinates of a point Q on the airfoil. Then

y
z

cos sin
sin cos

(3)
where and are local airfoil coordinates, and is the wing section pre-twist angle, positive nose-down.
The axial or extensional deection of a wing is generally very small and therefore can usually be neglected. Let V
and W be chordwise and apwise bending displacements of point Q, respectively, and let be a torsional twist angle
about the x-axis, positive nose-down. Then, the rotation angle due to the structural deformation can be expressed as
(x, t) =d
1
W
x
d
2
+V
x
d
3
(4)
where the subscripts x and t denote the partial derivatives of V, W, and .
Let (x
1
, y
1
, z
1
) be the coordinates of point Q on the airfoil in the reference frame D. Then the coordinates (x
1
, y
1
, z
1
)
are computed using the small angle approximation as
11

x
1
(x, t)
y
1
(x, t)
z
1
(x, t)

x
y +V
z +W

(yd
2
+zd
3
).d
1
(yd
2
+zd
3
).d
2
(yd
2
+zd
3
).d
3

x yV
x
zW
x
y +V z
z +W +y

(5)
Differentiating x
1
, y
1
, and z
1
with respect to x yields

x
1,x
y
1,x
z
1,x

1 yV
xx
+z

V
x
zW
xx
y

W
x
z

+V
x
z
x
y

+W
x
+y
x
z

(6)
Neglecting the transverse shear effect, the longitudinal strain is computed as
=
ds
1
ds
ds
=
s
1,x
s
x
1 (7)
where
s
x
=

1 +y
2
x
+z
2
x
=

1 +(y
2
+z
2
)

2
(8)
s
1,x
=

x
2
1,x
+y
2
1,x
+z
2
1,x
=

1 +(y
2
+z
2
)

2
2yV
xx
2zW
xx
+2(y
2
+z
2
)

x
(9)
Using the Taylor series expansion, it can be shown that for small a and b

1 +a

1 +b

1 +b

a
2

b
2
1 +
b
2
=
a b
2 +b
(10)
4 of 20
American Institute of Aeronautics and Astronautics
Therefore, the longitudinal strain is expressed as
=
s
1,x
s
x
1
yV
xx
zW
xx
+

y
2
+z
2

x
1 +
1
2
(y
2
+z
2
)

2
(11)
For a small wing twist angle , the longitudinal strain is further simplied as
=yV
xx
zW
xx
+

y
2
+z
2

x
(12)
The moments are now computed as

M
x
M
y
M
z

GJ
x
0
0

y
2
+z
2

+
x

z
y

dydz = E

GJ
E
+B
1

2
B
2

B
3

B
2

I
yy
I
yz
B
3

I
yz
I
zz

x
W
xx
V
xx

(13)
where the area moments of inertia and the bending-torsion coupling constants B
1
, B
2
, and B
3
are dened as

I
xx
0 0
0 I
yy
I
yz
0 I
yz
I
zz

y
2
+z
2
0 0
0 z
2
yz
0 yz y
2

dydz (14)

B
1
B
2
B
3

y
2
+z
2

y
2
+z
2
z
y

dydz (15)
The strain analysis shows that for a pre-twisted wing the exural deections V and W are coupled to the torsional
deection via the slope of the wing pre-twist angle. This coupling is usually not present in the Euler-Bernoulli beam
formulation, and can be signicant if the term

is dominant as in highly twisted wings such as turbomachinery rotor

blades.
IV. Inertial Coupling
During a high-g maneuver, rigid-body acceleration of an aircraft due to the Coriolis effect can generate inertial
forces on lifting surfaces. For highly exible wings, the inertial forces can couple with the structural deection.
In particular, if an aircraft experiences a very rapid roll rate, the inertial forces created by the roll acceleration can
cause natural frequencies to change due to the stiffness contribution of the roll acceleration. This is a well-known
phenomenon for rotary wing structures, whereby the rotation contributes to the apparent structural stiffness which is
known as rotational stiffening. Generally, the roll motion of an aircraft has a greater rotational stiffening effect than
the pitch and yaw motions due to a larger moment arm, which is the wing span.
For the analysis, only the roll motion of an aircraft is considered. The kinematic relationship is to be developed
to establish the velocity and acceleration due to rigid-body aircraft motion at a reference point on the wing. The
velocity establishes aeroelastic angle of attack that is used to develop aeroelastic forces and moments. Similarly, the
acceleration establishes the inertial force coupling with the wing elasticity.
A. Kinematics
Working in the left wing reference frame D, we dene points O, P, and Q as the center of gravity of the aircraft, the
origin of the reference frame D of the left wing, and the reference point of a mass particle on the airfoil section as
shown in Fig. 2. Then the position vectors from O to P and from P to Q are dened as
r
OP
= r
OP
x
b
1
+r
OP
y
b
2
+r
OP
z
b
3
= x
P
d
1
+y
P
d
2
+z
P
d
3
(16)
r
PQ
= x
1
(x, t)d
1
+y
1
(x, t)d
2
+z
1
(x, t) d
3
(17)
5 of 20
American Institute of Aeronautics and Astronautics
where

x
P
y
P
z
P

sin cos 0
cos sin 0
0 0 1

r
OP
x
r
OP
y
r
OP
z

(18)
Using Kanes dynamical notations, the velocity of the center of gravity, point O, and the angular velocity of the
aircraft are dened as
A
v
O
= ub
1
+wb
3
(19)
A

B
= pb
1
=
x
d
1
+
y
d
2
(20)
where u and w are the axial and vertical velocity components at point O , respectively; p is the roll rate; and
x
and
y
are the angular rates in the reference frame D

sin cos
cos sin

p
q = 0

(21)
The acceleration of the center of gravity and angular acceleration of the aircraft are computed as
A
a
O
=
B
d
A
v
O
dt
+
A

A
v
O
= ub
1
pwb
2
+ wb
3
= a
O
x
d
1
+a
O
y
d
2
+a
O
z
d
3
(22)
A

B
=
B
d
A

B
dt
= pb
1
=
x
d
1
+
y
d
2
(23)
where a
O
x
, a
O
y
, and a
O
z
are the acceleration components at point O in the reference frame D

a
O
x
a
O
y
a
O
z

sin cos 0
cos sin 0
0 0 1

u
pw
w

(24)
The velocity of point Q is contributed by the structural deection and is computed as
A
v
Q
=
A
v
O
+
A

r
OP
+r
PQ

+
D
r
PQ
t
= v
Q
x
d
1
+v
Q
y
d
2
+v
Q
z
d
3
(25)
where v
Q
x
, v
Q
y
, and v
Q
z
are the velocity components in the left wing reference frame D

v
Q
x
v
Q
y
v
Q
z

usin+
y
(z
P
+z
1
) +x
1,t
ucos
x
(z
P
+z
1
) +y
1,t
w+
x
(y
P
+y
1
)
y
(x
P
+x
1
) +z
1,t

(26)
The acceleration of point Q is also contributed by the structural deection and is computed as
A
a
Q
=
A
a
O
+
A

r
OP
+r
PQ

+
A

A
v
Q
+
D

2
r
PQ
t
2
+
A

D
r
PQ
t
= a
Q
x
d
1
+a
Q
y
d
2
+a
Q
z
d
3
(27)
where a
Q
x
, a
Q
y
, and a
Q
z
are the acceleration components in the reference frame D

a
Q
x
a
Q
y
a
Q
z

a
O
x
+
y
(z
P
+z)
y
w+
x

y
(y
P
+y)
2
y
(x
P
+x)
a
O
y

x
(z
P
+z) +
x
w
2
x
(y
P
+y) +
x

y
(x
P
+x)
a
O
z
+
x
(y
P
+y)
y
(x
P
+x)

2
x
+
2
y

(z
P
+z)

a
x
a
y
a
z

(28)
and a
x
, a
y
, and a
z
are the structural-deection induced acceleration components in the reference frame D

a
x
a
y
a
z

y
(W +y) +
x

y
(V z)
2
y
(yV
x
zW
x
) yV
xtt
zW
xtt
+2
y
(W
t
+y
t
)

x
(W +y)
2
x
(V z) +
x

y
(yV
x
zW
x
) +V
tt
z
tt
2
x
(W
t
+y
t
)

x
(V z)
y
(yV
x
zW
x
)

2
x
+
2
y

(W +y) +W
tt
+y
tt
+2
x
(V
t
z
t
) 2
y
(yV
xt
zW
xt
)

(29)
It is observed that angular rates and accelerations are coupled with the structural deection to give rise to additional
acceleration components on a wing structure.
6 of 20
American Institute of Aeronautics and Astronautics
B. Inertial Forces and Moments
For simplicity, we assume that the neutral axis coincides with the elastic axis. This is a reasonable approximation
and good results have been obtained with this assumption.When the neutral axis is offset from the elastic axis, the
expressions for inertial forces and moments can be quite complex. The offset between the neutral axis and elastic axis
is evaluated as

e
y
e
z

=
1
A

y
z

dydz =

0
0

(30)
The inertial forces at the elastic center due to the structural deection per unit length are then computed as

f
i
x
f
i
y
f
i
z

a
Q
x
a
Q
y
a
Q
z

dydz =A

x
+
y
W +
x

y
V +2
y
W
t
a

y

x
W
2
x
V +V
tt
2
x
W
t
a

z
+
x
V

2
x
+
2
y

W +W
tt
+2
x
V
t

(31)
where a

x
, a

y
, and a

z
are the rigid-body acceleration components at the elastic center

x
a

y
a

a
O
x
+
y
z
P

y
w+
x

y
y
P

2
y
(x
P
+x)
a
O
y

x
z
P
+
x
w
2
x
y
P
+
x

y
(x
P
+x)
a
O
z
+
x
y
P

y
(x
P
+x)

2
x
+
2
y

z
P

(32)
The inertial moments due to the structural deection per unit length are now computed as

m
i
x
m
i
y
m
i
z

za
Q
y
ya
Q
z
za
Q
x
ya
Q
x

dydz
=

(
x

y
I
yz

y
I
zz
)V
x
+(
y
I
yz

y
I
yy
)W
x
+

2
x
I
xx
+
2
y
I
zz

I
xx

tt
+2
y
(I
zz
V
xt
+I
yz
W
xt
)
(
y
I
yz
+
x

y
I
yy
)
2
y
(I
yz
V
x
I
yy
W
x
) +I
yz
V
xtt
I
yy
W
xtt
2
y
I
yz

t
(
y
I
zz
+
x

y
I
yz
)+
2
y
(I
zz
V
x
I
yz
W
x
) I
zz
V
xtt
+I
yz
W
xtt
+2
y
I
zz

(33)
V. Aeroelastic Analysis
The relative velocity of the air approaching a wing section includes the contribution from the wing structural
deection that results in changes in the local angle of attack. Since aerodynamic forces and moments are dependent
on the local angle of attack, the wing structural deection will generate additional elastic forces and moments. The
local angle of attack depends on the relative approaching air velocity as well as the rotation angle from Eq. (4).
The relative air velocity in turn also depends on a structural-deection induced velocity. The oscillation of the airfoil
results in an unsteady circulation. Kussner showed that the velocity at the 3/4-chord point determines the circulation
force on an oscillating airfoil.
17
This is in contrast to the lift due to circulation that acts at the 1/4-chord point in a
steady motion.
Fig. 3 - Airfoil Coordinates
Referring to Fig. 3, we adopt the convention used in Theodorsens theory of unsteady aerodynamics whereby the
chord of an airfoil has a length 2b and the elastic center is located at a distance ab from the mid-chord, where the
7 of 20
American Institute of Aeronautics and Astronautics
parameter a is between -1 and 1 and a is negative when the elastic center is forward of the mid-chord, which is usually
the case for a conventional airfoil
16, 17
. The local velocity is computed by evaluating the velocity of point Q in Eq.
(25) with the coordinates at the 3/4-chord point y
C
=

1
2
a

bcos and z
C
=

1
2
a

bsin in the reference frame D

v
C
x
v
C
y
v
C
z

usin+
y
(z
P
+z
C
+W +y
C
) y
C
V
x
z
C
W
x
ucos
x
(z
P
+z
C
+W +y
C
) +V
t
z
C

t
w+
x
(y
P
+y
C
+V z
C
)
y
(x
P
+x y
C
V
x
z
C
W
x
) +W
t
+y
C

(34)
In order to compute the aeroelastic forces and moments, the velocity must be transformed from the reference
frame D to the airfoil local coordinate reference frame dened by (, , ) (see Fig. 2). Then the transformation can
be performed using three successive rotation matrix multiplication operations as

v
C

v
C

v
C

1 0 0
0 1 +
0 1

1 0 W
x
0 1 0
W
x
0 1

1 V
x
0
V
x
1 0
0 0 1

v
C
x
v
C
y
v
C
z

(35)
where v

, v

, and v

are rigid-body velocity components given by

usin+
y
(z
P
+z
C
)
ucosw
x
(z
P
+z
C
)
w+ucos+
x
(y
P
+y
C
)
y
(x
P
+x)

(36)
and v

, v

, and v

are the induced velocity components due to the structural deection given by

V
x
v

+W
x
v

+
y
(W +y
C
) y
C
V
x
z
C
W
x
V
x
v

+v

x
(W +y
C
) +V
t
z
C

t
W
x
v

+
x
(V z
C
)
y
(y
C
V
x
z
C
W
x
) +W
t
+y
C

(37)
A. Aeroelastic Angle of Attack
Referring to Fig. 4, the local aeroelastic angle of attack on the airfoil section is due to the velocity components v
C

and
v
C

and is computed as

c
=
v
C

v
C

=
v

+
v

v
2

(38)
Fig. 4 - Aeroelastic Angle of Attack
Let

be a local rigid-body angle of attack. Then

is computed as

=
v

=
w+ucos+
x
(y
P
+y
C
)
y
(x
P
+x)
ucosw
x
(z
P
+z
C
)


cos

x
(y
P
+y
C
)
y
(x
P
+x)
ucos

x
(z
P
+z
C
)
ucos
2

(39)
where we recognize that is the aircraft angle of attack which is dened as
=
w
u
(40)
8 of 20
American Institute of Aeronautics and Astronautics
Let
e
be the local elastic angle of attack. Then
e
is computed as

e
=
v

v
2

=
v

W
x
v

+
x
(V z
C
)
y
(y
C
V
x
z
C
W
x
) +W
t
+y
C

v
2

V
x
v

+v

x
(W +y
C
) +V
t
z
C

v
2

(41)
Upon evaluation, we get

V
x

cos

W
x

tan

y
(z
P
+z
C
)
ucos

x
(V z
C
)
y
(y
C
V
x
z
C
W
x
) +W
t
+(y
C
+

z
C
)
t
+

x
(W +y
C
)

V
t
ucos
(42)
Then, the local aeroelastic angle of attack is expressed as

c
=

+
e
(43)
The terms V
t
, W
t
, and
t
contribute to aerodynamic damping forces which can be signicant in aeroelastic analysis.
If p = 0 and is small, then the elastic angle of attack is approximately equal to

e
=V
x

cos

tanW
x
tan
W
t
+

1
2
a

b
t

V
t
ucos
(44)
which agrees with the well-known Theodorsens result for a straight wing with = 0.
17
B. Aeroelastic Forces and Pitching Moment
In unsteady aerodynamics, the lift force is comprised of noncirculatory and circulatory components. The non circu-
latory component is due to the apparent mass and inertia effects which are generated when the wing has a non-zero
acceleration. This acceleration causes the surrounding air which has a nite mass to generate inertial forces that oppose
the acceleration. The circulatory component is more important for wing sections. This is due to the vortical strength
of circulation that generates lift. In unsteady aerodynamics, the vortices are shed in the ow in a complex fashion.
Unsteady thin-airfoil theories developed by Theodorsen and Peters can be used to estimate the effect of unsteady shed
vorticity. The former theory is appropriate for classical utter analysis while the latter is a nite-state theory cast in
the time-domain and is appropriate for time-domain analysis.
18
The total aeroelastic forces and pitch momenting at the elastic center are the sum of the noncirculatory and circu-
latory components and are expressed as

f
a
y
f
a
z
m
a
x

f
n
y
f
n
z
m
n
x

f
c
y
f
c
z
m
c
x

(45)
where the superscripts a, n, and c denote aeroelastic, noncirculatory, and circulatory, respectively.
1. Noncirculatory Forces and Pitching Moment
The noncirculatory unsteady forces are due to the apparent air mass effect which is based on the acceleration of the air
mass enclosed by a circular cylinder whose diameter is the airfoil chord. These noncirculatory forces are computed
from the acceleration in the reference frame D acting at the mid-chord as

f
n
y
f
n
z

b
2

a
y
( y, z)
a
z
( y, z)

b
2

x
(W + y)
2
x
(V z) +
x

y
( yV
x
zW
x
) +V
tt
z
tt
2
x
(W
t
+ y
t
)

x
(V z)
y
( yV
x
zW
x
)

2
x
+
2
y

(W + y) +W
tt
+ y
tt
+2
x
(V
t
z
t
) 2
y
( yV
xt
zW
xt
)

(46)
9 of 20
American Institute of Aeronautics and Astronautics
where y =abcos, z =absin, and

is the air density.

In addition, the structural rotation of the wing induces a downwash component
v

ucos
e
(47)
where u is the free-stream air speed, which generates a noncirculatory force acting at the 3/4-chord point equal to
f
n

b
2
v

b
2
ucos

e
t
(48)
The angular acceleration of the wing twist also induces a nose-down pitching moment at the mid-chord
m
n
x
=
1
8

b
4

e
t
2
(49)
Hence, the total noncirculatory forces and pitching moment at the elastic center are computed as

f
n
y
f
n
z
m
n
x

f
n
y
f
n

sin
f
n
z
+ f
n

cos
y f
n
z
z f
n
y
+

1
2
a

b f
n

+m
n
x

(50)
If p = 0 and is small, the noncirculatory forces and pitching moment are approximately equal to

f
n
y
f
n
z
m
n
x

b
2

V
tt
+ab
tt
+ucos

e
t
W
tt
ab
tt
ucos

e
t
ab(W
tt
ab
tt
V
tt
)

1
2
a

bucos

e
t

1
8
b
2
2

e
t
2

(51)
Equation (51) is in agreement with the well-established results in aeroelasticity.
17
2. Circulatory Aeroelastic Forces and Pitching Moment
The circulatory lift, drag, and pitching moment based on the Theodorsens theory are given by
d
dx

L
D
M

bu
2
cos
2

C
L
C
D
2bC
m

+C(k)
e

C
L,
C
D,
0

(52)
The 2-D section lift, drag, and pitching moment coefcients C
L
, C
D
, and C
m
due to the rigid-body angle of attack
are dened as

C
L
C
D
C
m

C
L,0
+C
L,

+C
L,
a

a
C
D,0
+C
D,

+C
D,
a

a
+KC
2
L
C
m,AC
+C
m,
a

(53)
where
a
is the aileron deection, K is the induced drag constant, and C
m,AC
is the quarter-chord pitching moment
coefcient which is positive nose up and independent of the angle of attack.
The functionC(k) is called the Theodorsens function which is a complex-valued function of the reduced frequency
parameter k
k =
b
ucos
(54)
where is the utter or aeroelastic mode frequency.
C(k) can also be expressed in terms of Hankel functions of the second kind H
(2)
n
(k) as
C(k) =
H
(2)
1
(k)
H
(2)
1
(k) +iH
(2)
0
(k)
= F (k) iG(k) (55)
where F (k) > 0 and G(k) > 0.
When k = 0, the airfoil motion is steady and C(k) is real and unity. As k increases, there is a phase lag introduced
as the magnitude of G(k) increases as shown in Figs. 4 and 5. The limiting values of F (k) and G(k) are 1/2 and 0 as
k .
17
10 of 20
American Institute of Aeronautics and Astronautics
0 0.5 1 1.5 2 2.5 3 3.5 4
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
k
F
(
k
)

o
r

G
(
k
)
F(k)
G(k)
Fig. 5 - Theodorsens Function
0 0.5 1 1.5 2 2.5 3 3.5 4
16
14
12
10
8
6
4
2
0
k

t
a
n

1
[
G
(
k
)
/
F
(
k
)
]
Fig. 6 - Theodorsens Function Phase lag
The aerodynamic forces and moment on a 2-D wing section acting at the elastic center in the reference frame D
are determined by

f
c
y
f
c
z
m
c
x

=
d
dx

Lsin(
c
+) +Dcos(
c
+)
Lcos (
c
+) +Dsin(
c
+)
M

1
2
+a

b[Lcos
c
+Dsin
c
]

(56)
as illustrated in Fig. 7.
Fig. 7 - Airfoil Forces and Moment
11 of 20
American Institute of Aeronautics and Astronautics
For a small elastic angle of attack, the circulatory unsteady aerodynamic forces and moment are approximated as

f
c
y
f
c
z
m
c
x

bu
2
cos
2

C
y
C
z
2bC
x

+C(k)
e

C
y,
C
z,
2bC
x,

(57)
where C
y
, C
z
, and C
x
are the force and moment coefcients due to rigid-body aerodynamics

C
y
C
z
C
x

C
L
sin(

+) +C
D
cos(

+)
C
L
cos(

+) +C
D
sin(

+)
C
m

1
4
+
a
2

[C
L
cos

+C
D
sin

(58)
and C
y,
, C
z,
, and C
x,
are the derivatives of the force and moment coefcients with respect to the angle of attack

C
y,
C
z,
C
x,

(C
L,
+C
D
)sin(

+) +(C
D,
C
L
)cos(

+)
(C
L,
+C
D
)cos(

+) +(C
D,
C
L
)sin(

+)

1
4
+
a
2

L,
+c

cos

+(C
D,
C
L
)sin

(59)
C. Aerodynamic Damping
Dynamic stability of a system is dependent on dissipative forces acting on it. The dissipative forces contribute posi-
tively to damping of the system. Dynamically stable systems require positive damping. Aeroelastic forces give rise to
the aerodynamic damping mechanism that inuences aerodynamic stability of an oscillating wing or a ight vehicle.
Positive aerodynamic damping results in aerodynamically stable operation. Conversely, negative aerodynamic damp-
ing causes utters and self-excited vibrations. Flutter boundaries are dened by air speed at which the aerodynamic
damping crosses from a positive value to a negative value.
Consider the case when p = 0 and = 0, the damping forces are obtained from the noncirculatory and circulatory
components of the aeroelastic forces as

f
d
y
f
d
z
m
d
x

b
2
ucos

V
xt
tan
cos
W
xt
tan
t

1
2
a a
2

bu
2
cos
2
C(k)
e

C
y,
C
z,
2bC
x,

(60)
For a harmonic motion, the structural deections may be expressed as

W
V

(x)

W
(x)

V
(x)

e
it
(61)
where

,
W
, and
V
are the elastic mode shapes of the torsion, apwise bending, and chordwise bending modes,
respectively.
Then, the elastic angle of attack may be expressed as

e
= (
r
+i
i
)e
it
(62)
where
r
and
i
are the real and imaginary parts of
e

r
=

V
tan
cos

W
tan

(63)

i
=

W
+

1
2
a

V
ucos

(64)
where the prime denotes derivative with respect to x.
12 of 20
American Institute of Aeronautics and Astronautics
The damping forces and pitching moment are then obtained as the imaginary part of the circulatory aeroelastic
forces and pitching moment

f
d
y
f
d
z
m
d
x

b
2
ucos
r
ie
it

1
2
a a
2

bu
2
cos
2
[G(k)
r
+F (k)
i
] ie
it

C
y,
C
z,
2bC
x,

(65)
which can also be written as

f
d
y
f
d
z
m
d
x

b
2
ucos

V
xt
tan
cos
W
xt
tan
t

1
2
a a
2

G(k)
k

C
y,
C
z,
2bC
x,

bucosF (k)

W
t

1
2
a

b
t
+

V
t

C
y,
C
z,
2bC
x,

(66)
It should be noted that the sign of the aerodynamic damping is inuenced by the sign of C
L,
+C
D
. While C
D
is always positive, C
L,
can be negative when the airfoil is stalled. Therefore, it is possible that the aerodynamic
damping can become negative. If the overall damping which includes structural damping that inherently exists in
the wing structure transitions from a positive value to a negative value, the ensuing motion will be aerodynamically
unstable due to positive work inputs to the wing by the air. For example, the generalized damping coefcient for the
apwise bending, by neglecting the term C
D,
C
L
and the contributions from chordwise bending and torsion, may
be estimated as

ucosF (k)

b(C
L,
+C
D
)cos

(x)[
W
(x)]
2
dx
2

A[
W
(x)]
2
dx
(67)
It is obvious that if C
L
becomes negative then if it possible for
W
to be negative. The aerodynamic damping for
the torsion is more complex. In general, the aerodynamic damping for the combined bending-torsion motion must be
analyzed by matrix analysis
VI. Gravity and Propulsive Forces and Moments
The gravity and propulsive forces are signicant contributing active forces acting on the aircraft airframe. These
forces can inuence structural deections of a wing. For high-aspect ratio, exible wings, gravity can signicantly
offset wing bending deections at low airspeed. Similarly, the engine thrust force can also affect twist and bending
deections of exible wings.
A. Propulsive Forces and Moments
To deal with the propulsive force, we assume that the left engine produces a thrust vector aligned with the b
1
-direction
at the thrust center E in the rigid-body aircraft reference frame B. Point E is assumed to be located relative to point P
at x = x
E
, y = y
E
, and z = z
E
forward and below the elastic center of the wing section in the reference frame D. Since
the engine is mounted on a exible wing, then the thrust center E is dependent on the wing structural deections.
Since the engine thrust is a concentrated force, it can be formulated as a distributed force using the Dirac delta
function which is dened as
(x x
0
) =

, x = x
0
0 , x = x
0
(68)
such that for an arbitrary function g(x), then

g(x) (x x
0
)dx = g(x
0
) (69)
13 of 20
American Institute of Aeronautics and Astronautics
Then, the distributed forces in the reference frame D due to the small structural rotation angle are computed
using three successive rotation matrix multiplication operations as

f
e
x
f
e
y
f
e
z

(x x
E
)T

1 0 0
0 1
0 1

1 0 W
x
0 1 0
W
x
0 1

1 V
x
0
V
x
1 0
0 0 1

sin
cos
0

(x x
E
) T

sin+V
x
cos
cosV
x
sin
W
x
sincos

(70)
where T is the engine thrust.
The distributed thrust forces are transformed back into the aircraft reference frame B as

f
e
x
f
e
y
f
e
z

(x x
E
) T

sin cos 0
cos sin 0
0 0 1

sin+V
x
cos
cosV
x
sin
W
x
sincos

(x x
E
) T

1
V
x
W
x
sin+cos

(71)
The left engine thrust force can now be obtained by integration as
T
L
=

(x x
E
)T [b
1
V
x
b
2
(W
x
sin+cos)b
3
] dx = T [b
1
V
x
b
2
+(W
x
sin+cos)b
3
]
x=x
E
(72)
It can be observed that the structural deections at the thrust center generate additional thrust force components
in the b
2
- and b
3
-directions. Thus, the coupled effect between the propulsive force and structural deections can be
signicant if the wing is highly exible and the engine thrust is large.
The distributed moments due to the propulsive force in the reference frame D are computed as

m
e
x
m
e
y
m
e
z

y
E
f
e
z
z
E
f
e
y
z
E
f
e
x
y
E
f
e
x

= (x x
E
) T

(z
E
y
E
W
x
) sin+(z
E
V
x
y
E
)cos
z
E
(sinV
x
cos)
y
E
(W
x
sin+cos)

(73)
B. Gravity Forces and Moments
The gravity can exert signicant forces and moments on a wing. The gravity forces include the weight of the wing
including fuel, and the engine weight. The distributed gravity forces can be expressed in the reference frame D as

f
g
x
f
g
y
f
g
z

[Ag + (x x
E
) m
E
g]

sin sincos sin cos

sin cos+cos sin sin
cos cos

(74)
where is the mass density of the wing including fuel, A is the cross sectional area, and m
E
is the engine mass.
Assuming that the center of gravity of the wing and fuel coincides with the elastic axis, and the center of gravity
of the engine coincides with the thrust center, then the distributed moments due to gravity in the reference frame D are
computed as

m
g
x
m
g
y
m
g
z

= (x x
E
) m
E
g

y
E
cos cos z
E
(cos+cos sin sin)
z
E
(sin sincos sin cos)
y
E
(sin sincos sin cos)

(75)
VII. Coupled Structural Dynamic Equations
The equilibrium conditions describe the force and moment balance of all forces and moments acting on a wing.
The resulting force and moment equilibrium conditions are given by

F
x
F
y
F
z

f
x
f
y
f
z

0
0
0

(76)
14 of 20
American Institute of Aeronautics and Astronautics

M
x
M
y
M
z

m
x
m
y
m
z

F
z
V
x
F
y
W
x
F
x
W
x
+F
z
F
x
V
x
+F
y

0
0
0

(77)
where f
x
, f
y
, f
z
, m
x
, m
y
, and m
z
are distributed forces and moments due to inertial, aeroelastic, gravity, and propulsive
effects
f
(x,y,z)
= f
i
(x,y,z)
+ f
a
(x,y,z)
+ f
g
(x,y,z)
+ f
e
(x,y,z)
(78)
m
(x,y,z)
= m
i
(x,y,z)
+m
a
(x,y,z)
+m
g
(x,y,z)
+m
e
(x,y,z)
(79)
The shear forces F
y
and F
z
can be eliminated by solving the last two equations in Eq. (77). Neglecting nonlinear
terms, the equilibrium conditions can be written as
M
x
x
+m
x
m

y
V
x
+m

z
W
x
= 0 (80)

2
M
y
x
2
+
m
y
x
f
z
+

x

W
x

x
0
f

x
d

= 0 (81)

2
M
z
x
2
+
m
z
x
f
y
+

x

V
x

x
0
f

x
d

= 0 (82)
where the superscript * denotes rigid-body forces and moments, and is a dummy variable that replaces x.
The resulting equations are three structural dynamic partial differential equations that relate the apwise bending,
chordwise bending, and torsion with the coupled effects under consideration. The integral term involving the rigid-
body inertial force f

x
can be important if the inertial acceleration of the aircraft is signicant. This term gives rise to
the rotational stiffening effect when the angular speed of a wing structure is large.
A. Example
Consider a ight vehicle with non-twisted, unswept wings = 0, = 0, on a horizontal ight, = 0, making a
constant roll rate p maneuver. The rigid-body and elastic angles of attack from Eqs. (39) and (42) reduce to

= +
p(x
P
+x)
u
(83)

e
=W
x
pz
P
u

W
t
+

1
2
a

b
t
ucos
(84)
Neglecting the chordwise bending motion and the damping forces, then the structural dynamic equations for ap-
wise bending and torsion are
(GJ
x
)
x
p
2
I
zz
+I
xx

tt
+

b
3

3
8
a +2a
2

b
tt
+

1
2
2a

W
tt

b
2
u
2
cos
2
F (k)C
z,

1
2
+a

W
x
pz
P
u
+

b
3
G(k)
k
C
z,

1
2
+a

W
tt
+

1
2
a

b
tt

bu
2
cos
2
2bC
x
(85)
(EI
yy
W
xx
)
xx
+A

W
tt
p
2
W

p
2
I
yy
W
x
I
yy
W
xtt
+W
x

x
0
Aa

x
d

x
+

b
2

1
2
2a

b
tt
+2W
tt

bu
2
cos
2
F (k)C
z,

W
x
pz
P
u
+

b
2
C
z,
G(k)
k

W
tt
+

1
2
a

b
tt

=Aa

z
+

bu
2
cos
2
C
z
Agcos (86)
15 of 20
American Institute of Aeronautics and Astronautics
For a quasi-steady-state motion when the wing responds statically to the inertial, aeroelastic, and propulsive forces,
by setting all the partial time derivative terms to zero in the above equations, one obtains
(GJ
x
)
x
p
2
I
zz

b
2
u
2
cos
2
C
z,

1
2
+a

W
x
pz
P
u
+

bu
2
cos
2
2bC
x
(87)
(EI
yy
W
xx
)
xx
Ap
2
W +

p
2
I
yy
W
x
+W
x

x
0
Aa

x
d

x
+

bu
2
cos
2
C
z,

W
x
pz
P
u
+

=A

w+ p
2
z
p

bu
2
cos
2
C
z
Agcos (88)
B. Solution Methods
Structural dynamic problems are generally solved by the nite-element analysis (FEA) method. The FEA method
formulates a structural dynamic problem described by a system of partial differential equations as a matrix equation
in terms of the mass matrix, stiffness matrix, and force vector in the form of
19

M
i
+M
n

G(k) H +(K
s
+K
c
)

U = F (89)
where M
i
is an inertial mass matrix, M
n
is an apparent mass matrix due to noncirculatory forces, K
s
is a structural
stiffness matrix, K
e
is an aeroelastic stiffness matrix due to circulatory forces, F is a force vector, H is a matrix due
to the phase lag resulting from vortex shedding, and U is a solution vector. Due to the aeroelasticity, the matrices
H and K
c
are non-symmetric, but they can be decomposed into symmetric and skew-symmetric matrices. Only the
symmetric matrices are retained for the solution.
16
Equation (89) is a generalized nonlinear eigenvalue problem due to the presence of G(k) which is a nonlinear
function of . Thus, the eigenvalue solution is an iterative process by rst guessing for k, and then solving an
generalized quadratic eigenvalue problem for , which in turn is used to update k. An alternate approach is to simply
ignore G(k) since G(k) is a small value. The eigenvalue problem then reverts to the standard form which can easily
be solved.
There are two types of problems: 1) static aeroelasticity and 2) dynamic aeroelasticity. The static aeroelasticity
describes physical effects that do not involve dynamic responses of a wing structure such as divergence and control
reversal. The wing-deected shape can inuence the aerodynamics of a vehicle. The static deection can be large
if the wing structure is highly exible. The static problem can be formulated as a coupled uid-structure interaction
problem. The vehicle is modeled by computational uid dynamics (CFD) method for aerodynamic calculations of
coefcients and derivatives. The results are used as inputs to the FEA model to compute the vehicle deected shape.
This shape is then used as the new input to the CFD model and the process is repeated until the solution converges.
The dynamic aeroelasticity describes utter behaviors and transient responses of a wing structure that is subject
to wind gusts or instantaneous control surface deections. Dynamic responses of the wing structure can affect the
overall vehicle control and stability. The solution of the dynamic problem can be solved by implementing the modal
decomposition method on the eigenvalue solution of the FEA. Elastic modes can be described by a set of uncoupled,
scalar second-order differential equations that can be readily analyzed.
For a symmetric aircraft conguration for which both wings are identical, two types of elastic modes are present.
Symmetric modes are those for which the structural deections of both wings are in the same sense. Anti-symmetric
modes are those that exhibit structural deections in an opposite sense between the left and right wings. Figure 8
illustrates symmetric and anti-symmetric modes.
16 of 20
American Institute of Aeronautics and Astronautics
Fig. 8 - Symmetric and Anti-Symmetric Modes
Because of the symmetry that exists at the fuselage centerline, only one wing can be analyzed with appropriate
boundary conditions.
20
The fuselage and tails contribute to the elastic modes as a concentrated mass, half of which is
located at each of the wing roots. Then for symmetric modes, the boundary conditions at the wing roots must match
the bending displacement slopes for both left and right wings, and are given by

W
x
(0, t)
V
x
(0, t)

0
0

(90)
The torsion boundary condition for symmetric modes is null which corresponds to a free-free boundary condition.
For anti-symmetric modes, the boundary conditions at the wing roots must be zero

(0, t)
W (0, t)
V (0, t)

0
0
0

(91)
To be more precise, the entire aircraft structure including wings, fuselage, and tails can be discretized and solved
by the FEA. In particular, the whole-aircraft analysis becomes necessary if the aircraft conguration is asymmetric.
The structural deections are obtained from the FEA solution and can be expressed as

(x, t)
W (x, t)
V (x, t)

(x)

W (x)

V (x)

+
m

j=1
q
j
(t)

, j
(x)

W, j
(x)

V, j
(x)

+
n

j=1
r
j
(t)

, j
(x)

W, j
(x)

V, j
(x)

(92)
where the overbar symbol denotes static solutions,
(,W,V), j
and
(,W,V), j
are normalized eigenvectors, q
j
and r
j
are generalized coordinates for the j th symmetric and anti-symmetric modes that solve a set of uncoupled scalar
differential equations
m
gen
j
q
j
+c
gen
j
q
j
+k
gen
j
q
j
= g
gen
j
( u, w, u, w, p, p,
a
) , j = 1, . . . , m (93)
m
gen
j
r
j
+c
gen
j
r
j
+k
gen
j
r
j
= h
gen
j
( u, w, u, w, p, p,
a
), j = 1, . . . , n (94)
The scalar quantities m
gen
, c
gen
, and k
gen
are called generalized mass, damping, and stiffness matrices, respectively,
and g
gen
and h
gen
are called the generalized forces. They can be computed using the standard procedure in elementary
vibration theory.
17 of 20
American Institute of Aeronautics and Astronautics
VIII. Flight Dynamics of Flexible Aircraft
A exible aircraft has various elastic modes that can participate in the motion to affect its ight characteristics.
Wing elastic modes constitute signicant structural dynamics of exible aircraft. In addition, fuselage bending modes
are also known to affect pitch characteristics. There are other elastic modes such as those due to horizontal stabilizers
and vertical stabilizer. In totality, all these components contribute to ight characteristics and should be included in
the equations of motion. The coupled ight-structural dynamics can be quite complicated when all elastic modes are
accounted for. In this study, we will only focus on the coupling of wing elastic modes and rigid-body ight dynamics.
A. Aeroelastic Forces and Moments in Aircraft Reference Frame
We dene the elastic angle of attack as a function of the structural deections as

e
(,W
x
,V
x
) =

V
x

cos

W
x

tan

y
(z
P
+z
C
)
ucos

(95)
Wing structural deections affect the lift characteristics of an aircraft. Assuming both the left and right wings are
of the same geometry, then the elastic angle of attack contributes to the static and dynamic forces and the moments in
the aircraft reference frame B as

X (u, w, p, , , ,
a
)

Z(u, w, p, , , ,
a
)

M(u, w, p, , , ,
a
)

= 2

L
0

bu
2
cos
2

,

W
x
,

V
x

C
y,
cos
C
z,
2bC
x,
cos

dx (96)

X
S
j
(u, w, p, , , ,
a
)
Z
S
j
(u, w, p)
M
S
j
(u, w, p)

= 2F (k)

L
0

bu
2
cos
2

, j
,
W, j
,
V, j

C
y,
cos
C
z,
2bC
x,
cos

dx (97)

Y
A
j
(u, w, p, , , ,
a
)
L
A
j
(u, w, p)

= 2F (k)

L
0

bu
2
cos
2

, j
,
W, j
,
V, j

C
y,
sin
2bC
x,
sin

dx (98)
where X, Y, and Z are the aircraft axial, side, and normal forces, L, M, and N are the aircraft rolling, pitching, and
yawing moments, the overbar symbol denotes static forces and moments, and the superscript and denote dynamic
forces and moments corresponding to symmetric and anti-symmetric modes, respectively.
It can be seen that symmetric modes affects forces and moments in the longitudinal direction. On the other hand,
anti-symmetric modes affects the lateral motion of the aircraft. The aeroelastic forces and moments contribute to
the ight dynamics of an aircraft in such a way that elastic deections can adversely affect the rigid-body aircraft
responses and can result in elastic mode interactions with a ight control system. These interactions necessitate the
use of aeroelastic mode lters in the ight control design in order to attenuate structural dynamic responses of exible
aircraft lifting surfaces.
B. Propulsive Forces and Moments in Aircraft Reference Frame
In addition, the elastic modes also affect the propulsive forces and moments generated by wing-mounted engines. The
propulsive force for the left engine is given by Eq. (72). Assuming a twin-engine aircraft conguration, the propulsive
force for the right engine is given by
T
R
= T [b
1
V
x
b
2
+(W
x
sincos) b
3
]
x=x
E
(99)
where V, W, and are dened according to the right wing reference frame C such that V and are in the opposite
sense to those deections in the reference frame D, i.e., V is positive towards the leading edge and is positive nose
up.
The total static and dynamic propulsive forces are then computed to be
X
T
= 2T (100)
18 of 20
American Institute of Aeronautics and Astronautics

Y
TA
j
Z
TS
j

= 2T

V, j

W, j
sin+
, j
cos

x=x
E
(101)
where the superscript TS and TA denote thrust forces due to symmetric and anti-symmetric modes, respectively. Thus,
symmetric modes create a normal force and anti-symmetric modes create a side force due to the combined thrust of
the two engines.
The propulsive moment is computed as
M =

r
E
x
b
1
r
E
y
b
2
+r
E
z
b
3

T
L
+

r
E
x
b
1
+r
E
y
b
2
+r
E
z
b
3

T
R
(102)
where

r
E
x
, r
E
y
, r
E
z

are the coordinates of the right engine thrust center relative to the aircraft center of gravity in the
reference frame B.
Upon evaluation, this yields the total static and dynamic propulsive moments
M
T
= 2Tr
E
z
(103)

L
TA
j
M
TS
j
N
TA
j

= 2T

r
E
z

V, j
r
E
y

W, j
sin+
, j
cos

r
E
x

W, j
sin+
, j
cos

r
E
x

V, j

x=x
E
(104)
It can be seen that symmetric modes result in an additional pitching moment, whereas anti-symmetric modes create
both rolling and yawing moments.
C. Equations of Motion
The ight dynamic equations with elastic mode and propulsive force interactions can now be written as
m( urv +qw+gsin) = X
T
+

C
L
qSsin

C
D
qScos cos +

X +
m

j=1
X
S
j
q
j
(105)
m( v +ru pwgcossin) =

C
Y
qS

C
D
qSsin +
n

j=1

Y
TA
j
+Y
A
j

r
j
(106)
m( wqu + pv gcoscos) =

C
L
qScos

C
D
qSsin cos +

Z+
m

j=1

Z
TS
j
+Z
S
j

q
j
(107)
L =

C
l
qS

b+
n

j=1

L
TA
j
+L
A
j

r
j
(108)
M =

C
m
qS c +M
T
+

M+
m

j=1

M
TS
j
+M
S
j

q
j
(109)
N =

C
n
qS

b+
n

j=1
N
TA
j
r
j
(110)
where the coefcients with the overbar are for the aircraft, q is the dynamic pressure, c isthe mean aerodynamic chord,

b is the wing span, S is the reference wing area, and is the angle of sideslip.
These equations constitute six degrees of freedom dynamics that are coupled with structural dynamics equations
fromEqs. (93) and (94), which are obtained fromthe FEA. In analyzing the elastic modes, only the rst few signicant
modes whose natural frequencies are within a ight control frequency bandwidth are usually considered. These elastic
modes can become excited by the ight control surface deections. Examining Eqs. (105) to (110) reveals that
symmetric modes only affects longitudinal dynamics of aircraft. On the other hand, anti-symmetric modes play a role
in both lateral and directional dynamics.
19 of 20
American Institute of Aeronautics and Astronautics
IX. Conclusions
This paper has presented an integrated ight dynamic modeling method for exible aircraft. The method combines
structural dynamics of an equivalent beam model of a exible wing with rigid-body ight dynamics that accounts for
coupled effects due to aeroelasticity, inertial forces, and propulsive forces. A formulation of aeroelastic angle of
attack for the combined chordwise bending, apwise bending, and torsion is developed that extends Theodorsens
theoretical result. The structural dynamic equations can be solved using the nite-element method to determine static
and dynamic structural deections as functions of aircraft states. The elastic modes are decomposed into symmetric
and anti-symmetric modes with associated generalized coordinates. The standard ight dynamic equations for six
degree-of-freedom motion then includes the generalized coordinates as additional state variables. These equations
become coupled with a set of uncoupled second-order differential equations in terms of the generalized coordinates
that describe the elastic responses of the wing structure. These equations must be solved simultaneously to obtain a
solution that describes the combined motion of exible aircraft.
References
1
Pendleton, E., Lee, M., and Wasserman, L., Application of Active Flexible Wing Technology to the Agile Falcon, AIAA Journal of
Aircraft, Vol. 29, No. 3, May-June 1992.Pendleton, E., Lee, M., and Wasserman, L., Application of Active Flexible Wing Technology to the Agile
Falcon, AIAA Journal of Aircraft, Vol. 29, No. 3, May-June 1992.
2
Wilson, J.R., Active Aeroelastic Wing: a New/Old Twist on Flight, Aerospace America, September 2002.
3
Pratt, R.W., Taylor, R., and Caldwell, B.D., Aeroservoelasticity: Key Issues Affecting the Design of Flight Control Systems, IEEE Control
Conference Publication No. 389, March 1994.
4
Gupta, K.K., Brenner, M.J., and Voelker, L.S., Development of an Integrated Aeroservoelastic Analysis Program and Correlation with Test
Data, NASA Technical Paper 3120, May 1991.
5
Cheng, P.Y. and Hirner, T.J., Aircraft Aeroservoelastic Compensation Using Constrained Optimization, AIAA Structures, Structural Dy-
namics, and Materials Conference, AIAA-1992-2399, April 1992.
6
Brenner, M.J., Aeroservoelastic Modeling and Validation of a Thrust-Vectoring F/A-18 Aircraft, NASA Technical Paper 3647, September
1996.
7
Strass, H.K., and Stephens, E.W., An Engineering Method for Determination of Aeroelastic Effects upon the Rolling Effectiveness of
Ailerons on Swept Wings, NACA Research Memorandum L53H14, November 1953.
8
Theodorsen, T. and Garrick, I.E., Mechanism of Flutter - a Theoretical and Experimental Investigation of the Flutter Problem, NACA
Report 685, 1940.
9
Theodorsen, T, General Theory of Aerodynamic Instability and the Mechanism of Flutter, NACA Report 496, 1935.
10
Garrick, I.E., Bending-Torsion Flutter Calculations Modied by Subsonic Compressibility Corrections, NACA Report 836, 1946.
11
Houbolt, J.C. and Brooks, G.W., Differential Equations of Motion for Combined Flapwise Bending, Chordwise Bending, and Torsion of
Twisted Nonuniform Rotor Blades, NACA Technical Note 3905, February 1957.
12
Lee, U., Equivalent Continuum Beam-Rod Models of Aircraft Wing Structures for Aeroelastic Analysis, AIAA Structures, Structural
Dynamics, and Materials Conference, AIAA-1994-1695, April 1995.
13
Shearer, C.M. and Cesnik, C.E.S., Nonlinear Flight Dynamics of Very Flexible Aircraft, AIAA Atmospheric Flight Mechanics Conference,
AIAA-2005-5805, August 2005.
14
Meirovitch, L. and Tuzcu, I, Integrated Approach to Flight Dynamics and Aeroservoelasticity of Whole Flexible Aircraft - Part I: System
Modeling, AIAA Guidance, Navigation, and Control Conference, AIAA-2002-4747, August 2002.
15
Meirovitch, L. and Tuzcu, I, Integrated Approach to Flight Dynamics and Aeroservoelasticity of Whole Flexible Aircraft - Part II: Control
Design, AIAA Guidance, Navigation, and Control Conference, AIAA-2002-5055, August 2002.
16
Bisplinghoff, R.L and Ashley, H., Principles of Aeroelasticity, Dover Publications, Inc., New York, 1975.
17
Abramson, N., An Introduction to the Dynamics of Airplanes, Dover Publications, Inc., New York, 1971.
18
Hodges, D.H. and Pierce, G.A., Introduction to Structural Dynamics and Aeroelasticity, Cambridge University Press, 2002.
19
Hughes, T.J.R., The Finite Element Method: Linear Static and Dynamic Finite Element Analysis, Dover Publications, Inc., 2000.
20
Przemieniecki, J.S., Theory of Matrix Structural Analysis, Dover Publications, Inc., New York, 1985.
20 of 20
American Institute of Aeronautics and Astronautics