Static and Dynamic Aeroelasticity

Terrence A. Weisshaar
School of Aeronautics and Astronautics, Purdue University, West Lafayette, IN, USA

aerodynamic forces, structural deformation (elasticity), and

1 Purpose and Scope 1 motion (dynamics) of aerodynamic and hydrodynamic lifting
2 Aircraft Structural Dynamic Response, surfaces. These highly interactive, interdisciplinary activi-
Stability, and Flutter 3 ties are summarized in Figure 1 with interactions among
3 The Origins of Static and Dynamic Aeroelastic four major technical areas involved in aeroelastic phenomena
Problems 4 (Friedmann, 1999). The term aeroelasticity was coined in the
early 1930s by Alfred Pugsley and Harold Roxbee Cox, two
4 Compressibility: Mach Number Effects 7
distinguished engineers at the Royal Aircraft Establishment
5 Control Effectiveness 7 (RAE).
6 Flutter and Buffet Appear 8 Aeroelastic interactions determine airplane loads and per-
7 Swept Wing Aeroelasticity 9 formance in four primary areas: (i) wing and tail surface
8 High-Speed Flutter Problems 10 lift redistribution that changes external loads; (ii) stability
9 Flutter Design Solutions 10 derivatives, including lift effectiveness, that changes flight
10 Controlling and Exploiting Aeroelasticity:
Aeroelastic Tailoring and Aeroservoelasticity 11
11 Static Stability or Flight Dynamic Stability:
Divergence or Flutter? 12
12 Summary 12
References 13
Textbooks and Further Reading 13

1 PURPOSE AND SCOPE

The purpose of this chapter is to describe aeroelastic effects

on fixed wing air vehicles. Rotating systems such as pro-
pellers, helicopter blades, or gas turbines are not addressed.
Aeroelasticity is a design activity concerned with the conse-
quences created and tradeoffs required by interactions among

Figure 1. Aeroelastic interactions among major technical areas

create new problems and opportunities.

Encyclopedia of Aerospace Engineering, Online © 2010 John Wiley & Sons, Ltd.
This article is © 2010 John Wiley & Sons, Ltd.
This article was published in the Encyclopedia of Aerospace Engineering in 2010 by John Wiley & Sons, Ltd.
DOI: 10.1002/9780470686652.eae149
 2 Aeroelasticity and Aeroservoelasticity

2.0

Unswept wing
divergence
1.5
Unswept
wing

LIft effectiveness
1.0 15º
sweep

0.5

30º
sweep
0.0
0 50 100 150 200 250 300 350
Figure 2. Boeing 767 static structural proof test to wing failure
Dynamic pressure (psf)
showing extreme strength with large deformation.
Figure 4. Lift effectiveness, showing increased lift per unit angle
static and dynamic control features such as aircraft trim and of attack associated with unswept wings.
dynamic response (Abzug and Larabee, 2002); (iii) control
effectiveness, including aileron reversal, that limits maneu- Unswept wing torsional deformation causes the loading
verability; and (iv) aircraft structural dynamic response and to shift outboard, but wing bending produces no change in
stability, in particular buffeting and flutter. local angle of attack. The aerodynamic load on an unswept
Highly efficient wing and tail surface structural design per- wing resembles the curve marked “flexible wing” in Figure 3,
mits substantial structural bending and twisting deformation whereas the unswept flexible wing lift distribution resembles
during flight, as indicated in the picture of the Boeing 767 the curve marked “rigid wing.”
ground test shown in Figure 2. Wing twist produces changes Lift effectiveness is defined as the ratio of the total lift
in local angle of attack on swept and unswept surfaces. Swept- on a flexible surface divided by the lift on the identical,
back wing bending reduces local angle of attack, whereas but rigid, surface at identical angles of attack. Lift effec-
sweptforward wing bending increases local angle of attack. tiveness changes with airspeed. Figure 4 shows three plots
Figure 3 shows two different aerodynamic load distribu- of typical wing lift effectiveness as a function of airspeed.
tions along a hypothetical, 35◦ sweptback wing, one for a Sweptback wings are lift ineffective (lift effectiveness less
rigid wing, the other for an identical but flexible wing. The than one) because of lifting surface bending. Conversely, flex-
total lift, shown as the areas under each of the two curves, ible unswept lifting surfaces generate more lift than similar
is equal; the aircraft angle of attack for the flexible surface rigid surfaces.
must be larger for the rigid surface. Control effectiveness is the ability of a control surface
such as an aileron or a rudder to produce aerodynamic forces
and moments to control airplane orientation and maneuver
along a flight path. Asymmetrical aileron rotation produces
rolling acceleration and roll rate. Aircraft rolling motion cre-
ates damping in roll that opposes rolling motion. The ability
to create a terminal or steady-state roll rate is the primary
measure of aileron effectiveness.
Consider Figure 5. Without wing torsional flexibility, the
terminal roll rate is a linear function of airspeed. Rotating
the aileron downward produces lift, but also twists the wing
surface nose-down, reducing the local wing angle of attack as
well as the lift force and moment. The size of the nose-down
twisting moment and nose-down twist depends upon the:
Figure 3. An example flexible 35◦ sweptback wing lift redistribu- (i) size of the control surface; (ii) amount of aileron deflec-
tion caused by aeroelastic effects at cruise. tion; (iii) structural stiffness; and, (iv) dynamic pressure, q.

Encyclopedia of Aerospace Engineering, Online © 2010 John Wiley & Sons, Ltd.
This article is © 2010 John Wiley & Sons, Ltd.
This article was published in the Encyclopedia of Aerospace Engineering in 2010 by John Wiley & Sons, Ltd.
DOI: 10.1002/9780470686652.eae149
 Static and Dynamic Aeroelasticity 3

Figure 5. Control effectiveness declines with increasing airspeed.

Dynamic pressure is one half the product of air density times Figure 6. Wing vibration mode shapes.
airspeed squared.
Because aileron induced twisting reduces the lift force,
the rolling acceleration and the steady state roll rate will be Analysis and test are used to compute and then measure
less than the rigid wing roll rate. As indicated in Figure 5, aircraft natural frequencies and the mode shapes associated
the roll rate is maximum at a certain flight speed and then with these dynamic pressure independent natural frequencies.
declines rapidly as airspeed increases. At a special airspeed, Natural frequencies are identified during ground resonance
called reversal speed, the ailerons will not generate a rolling tests. The vibration shapes at these natural frequencies are
moment even though there is substantial wing surface dis- the mode shapes, unique to each natural frequency. A typical
tortion and substantial aileron rotation. At speeds above the result is shown in Figure 6.
reversal speed, the aileron produces a roll rate, but in the Flutter is a self-excited structural dynamic instability that
opposite direction to that intended. occurs when an oscillating wing or tail surface vibrates in
such a way as to extract energy from the airstream; the
amplitude of vibratory motion increases exponentially. Flut-
2 AIRCRAFT STRUCTURAL DYNAMIC ter belongs to a special class of mechanics problems called
RESPONSE, STABILITY, AND FLUTTER nonconservative problems. The flutter mechanism depends
on flying at the right speed and altitude to allow two or
Unsteady dynamic loads such as those created by abrupt con- more aircraft vibration modes to interact or couple together.
trol movement, gusts, or buffet may produce moderate or In rare instances flutter occurs as the result of the loss of
even severe transient structural dynamic response. Buffet is aerodynamic damping. Flutter can be loosely categorized
an unsteady aerodynamic loading created by aerodynamic into at least five different areas: (i) classical – wing bending
wakes shed from wings, nacelles, and fuselage pods dur- and torsion; (ii) control surface – surface rotation and wing
ing operation at high angles of attack. Buffet can create bending; (iii) empennage – fuselage torsion and tail torsion;
severe dynamic loads and stresses, particularly on vertical (iv) stall – wing torsion; and (v) body freedom – wing bending
stabilizers. and fuselage pitch.
Aircraft and missile resonant natural frequencies depend Figure 7 shows time histories of three possible types of
on stiffness and mass distribution. Lifting surface structural dynamic responses at a point on a wing surface responding
flexibility depends on the amount of displacement created to a disturbance during flight at different airspeeds (Ricketts,
by forces and moments. If large deflections are created by 1983). Disturbances decay with time at an airspeed corre-
small forces then the surface is very flexible. Stiffness is sponding to point A where the resonant natural frequencies
the inverse of flexibility. Because the source of aeroelas- are well separated. At Point B a disturbance produces har-
tic stiffness or flexibility results from the structure and the monic oscillatory motion with fixed amplitude. The two
aerodynamic displacement-dependent forces, aircraft flexi- natural frequencies have changed and are very close to each
bility and stiffness, as well as resonant frequencies and mode other. At point C two natural frequencies are extremely close
shapes, change with airspeed or altitude. together. Any disturbance grows rapidly. This is flutter.

Encyclopedia of Aerospace Engineering, Online © 2010 John Wiley & Sons, Ltd.
This article is © 2010 John Wiley & Sons, Ltd.
This article was published in the Encyclopedia of Aerospace Engineering in 2010 by John Wiley & Sons, Ltd.
DOI: 10.1002/9780470686652.eae149
 4 Aeroelasticity and Aeroservoelasticity

As the wing plunges downward it also rotates downward

to place the lift force in phase with the downward motion.
At the beginning of the downward motion the lift is zero
and reaches its maximum downward value when the plunge
displacement is zero.
The aerodynamic forces behave like a sine wave, whereas
the displacement is a cosine wave. These two functions are
90◦ out-of-phase. As a result, the integral of the product of
the lift force times the plunge velocity is positive. Energy is
extracted from the flow and fed into the system.

3 THE ORIGINS OF STATIC AND

DYNAMIC AEROELASTIC PROBLEMS
Figure 7. Aircraft vibration modes couple together to allow energy
to be absorbed from the airstream. Two factors drive aviation development: the quest for speed
and the competition for new air vehicle military and com-
The proximity of the two natural frequencies indicates
mercial applications. These factors trigger the appearance of
that the two modes of motion are interacting in a way such
new aircraft shapes, new materials, and applications of other
that the lifting surface is naturally extracting energy from the
new technologies such as avionics. This in turn has created
airstream and feeding it into wing vibration.
and continues to create new challenges for aeroelasticity.
Flutter is not forced resonant response requiring a fixed
Static aeroelastic phenomena appeared at the very begin-
frequency harmonic force input at a system natural frequency.
ning of powered flight. Samuel Langley’s Aerodrome failed
The airstream causing flutter is steady and nonoscillatory
until the system is disturbed. Without internal structural
damping, resonance response amplitude grows linearly with
time, whereas the flutter dynamic response has an exponen-
tial increase until the structure is destroyed or some nonlinear
mechanism comes into play.
The flutter instability mechanism is illustrated in Figure 8,
which shows a wing section undergoing torsional and bend-
ing motion. The torsion mode creates harmonic aerodynamic
forces, whereas the bending or “plunge” mode supplies the
displacements for the aerodynamic forces to do work.

Figure 8. Airspeed dependent phasing between two distinct types Figure 9. Langley Aerodrome aeroelastic launch failure December
of oscillatory motion at nearly the same frequency is required to 1903 showing forward wing collapse. Reproduced with permission
create flutter dynamic instability. from Phil Callihan.
Encyclopedia of Aerospace Engineering, Online © 2010 John Wiley & Sons, Ltd.
This article is © 2010 John Wiley & Sons, Ltd.
This article was published in the Encyclopedia of Aerospace Engineering in 2010 by John Wiley & Sons, Ltd.
DOI: 10.1002/9780470686652.eae149
 Static and Dynamic Aeroelasticity 5

Figure 10. 1901 Wright Brothers wing warping kite.

twice, including once several days before the Wright

Brother’s first flight in 1903 (Figure 9). Collar (1978)
attributes this failure to a static structural instability called
wing divergence. Garrick and Reed (1981) disagree with this
conclusion, citing recent evidence that rules out static insta-
bility, but identify excessive flexibility that led to an overload
situation.
In 1901 the Wright Brothers used a tethered kite, sketched
in Figure 10, to demonstrate wing warping – controlled, anti-
symmetrical biplane wing twisting – to create active roll
control. Wing warping depends on building torsionally flexi- Figure 12. (a) (b) The typical section aeroelastic idealization.
ble wing surfaces easily distorted by the pilot. Wing warping
surfaces must be very flexible; they are also easily distorted
by the airstream to produce unintended airloads. This concept its rear spar created a tendency of the leading edge to twist,
was used on the Wright Flyer in 1903. Most early monoplanes producing excessive lift forces during dive pullout. Design-
were controlled by wing warping. These included the Bleriot ers and operators struggled to understand the source of these
XI, shown in Figure 11, the British Bristol Prier Monoplane, problems.
and, as late as 1915, the Fokker Eindecker. The most often The Bleriot XI and the Bristol Prier had fatal accidents
cited early aeroelastically induced wing failure occurred on around 1912. Bleriot strengthened the Bleriot XI guy wires
the Fokker D-VIII monoplane (Fokker, 1931). Strengthening and increased the main wing spar size, but wing failures
still occurred. On 12 September 1912 the British War Office
placed a ban on Royal Flying Corps pilots flying mono-
planes after a series of accidents involving the Bristol Prier
monoplane (Baker, 1994). Some witnesses to these crashes
reported in-flight explosions. The French issued their own
monoplane flight ban soon after. These temporary bans cre-
ated an anti-monoplane bias lasting through World War I.
Designers did not understand that aeroelastic problems are
stiffness related, not stress related.
Figure 12 shows the typical section aeroelastic idealization
used to illustrate static aeroelastic phenomena. The model
has two degrees of freedom called pitch and plunge and is
supported by two linear elastic springs located at the shear
center. The two springs, with stiffnesses Kh and KT , simulate
the stiffness of wing spars and skin. The displacement h,
called “plunge” represents wing surface bending displace-
ment, measured at the wing section shear center, whereas θ
represents wing structural torsion. The shear center or center
of twist is a point on the wing section where the application of
Figure 11. The wire-braced Bleriot XI with a wing warping control a concentrated force creates plunge/bending deflection, but
system first flew between England and France in 1909. no twist.
Encyclopedia of Aerospace Engineering, Online © 2010 John Wiley & Sons, Ltd.
This article is © 2010 John Wiley & Sons, Ltd.
This article was published in the Encyclopedia of Aerospace Engineering in 2010 by John Wiley & Sons, Ltd.
DOI: 10.1002/9780470686652.eae149
 6 Aeroelasticity and Aeroservoelasticity

The typical section lift vector and pitching moment act offset between the shear center and wing aerodynamic center
at the aerodynamic center. Thin airfoil theory for incom- excludes wing divergence as a mode of failure but not aeroe-
pressible flow predicts the aerodynamic center to be at the lastic overloading caused by high camber and low torsional
1/4 chord position. At the aerodynamic center the aerody- stiffness.
namic pitching moment depends on airfoil camber but does The typical section effective lift curve slope CLα is modi-
not change with angle of attack. The lift and moment are fied by aeroelasticity to be:
expressed in coefficient form as:
CLα CLα
CLflex = = (5)
L = qSCLα (α0 + θ) (1) α
1− qSeCLα
KT
1 − q̄
MAC = qScCMAC (2)
Because e is usually positive, the typical section develops
with wing chord dimension, c, planform area S, airspeed V , more lift than a similar rigid surface at a given angle of attack.
dynamic pressure q = 21 ρV 2 and initial angle of attack, αo . For large values of q̄ wing lift is sensitive to small changes
The CMAC term is an aerodynamic coefficient that depends in angle of attack.
on wing camber. For positive camber (curvature downward) Aircraft load factor, n, is defined as n = L/W = lift/weight.
CMAC is negative. For the typical section model the change in the load factor
Aerodynamic lift and moment twist the airfoil an amount with respect to the aircraft angle of attack is
θ, creating additional lift because the aerodynamic center is
offset from the shear center by the distance e. The structural ∂n ∂L qCL
= W = W  α (6)
twist is: ∂αo ∂αo S
(1 − q̄)
qSeCLα αo

KT qScCMAC 1 Braced-wing monoplanes, such as the Bleriot XI, had
θ= + small wing torsional stiffness and operated at large values
1 − qSeCKT
Lα KT 1 − qSeC
KT
Lα

      of q̄. Small changes in the angle of attack, either by pilot

q̄ cCMAC q̄ input or by atmospheric gusts, created large increases in wing
= αo + (3)
1 − q̄ eCLα 1 − q̄ external force and led to the documented failures.
When two typical section wings, each with area S, pro-
The typical section lift is: vide total lift, L, to support an airplane with weight, W, the
required wing angle of attack is:
 
qSCLα qSCLα qScCMAC
L= αo +  
1− qSeCLα
1− qSeCLα
KT 1W 1 − q̄
KT KT αo = − αM (7)
2 S qCLα
= qSCLflex
α
αo + flex
qScCMAC (4)
The wing section twist angles, θ, are both:
The nondimensional aeroelastic parameter q̄ = qSeCKT
Lα
rep-  
resents the ratio of aerodynamic overturning moment divided 1 eW
θ= + qScCMAC (8)
by structural restoring moment. When qSeC KT
Lα
= 1 we have KT 2
wing divergence. An initial angle of attack, no matter how
small, creates an infinitely large lift and twist. Wing diver- No aeroelastic parameter appears in this fixed lift equa-
gence is a static structural instability. tion. If camber is zero, then the twist angle is constant at all
The typical section model results in equations (3) and (4) flight speeds; the airplane angle of attack decreases to zero
show that flexible surfaces with large camber, such as the as the flight q approaches the divergence condition q̄ = 1.
Langley Aerodrome wing and tail, develop large nose-down However, sensitivity to changes in angle of attack remains.
twist at low speeds because the second terms in equations (3) The divergence instability still exists.
and (4) are negative. With large camber the second term in In 1926, Hans Reissner published a landmark paper
equation (4) dominates even if the aerodynamic center/shear “Neuer Probleme aus der Flugzeugstatik” (New Static
center offset e is close to zero. Structural tests (Garrick and Structural Problems of Wings) detailing static aeroelastic
Reed, 1981) performed on the original Langley Aerodrome phenomenon such as lift effectiveness and wing divergence.
in 1980 found the front wing shear center to be very close Similar papers soon appeared in other countries; static aeroe-
to the wing aerodynamic center. This lack of a measurable lastic problems began to be understood and controlled.

Encyclopedia of Aerospace Engineering, Online © 2010 John Wiley & Sons, Ltd.
This article is © 2010 John Wiley & Sons, Ltd.
This article was published in the Encyclopedia of Aerospace Engineering in 2010 by John Wiley & Sons, Ltd.
DOI: 10.1002/9780470686652.eae149
 Static and Dynamic Aeroelasticity 7

4 COMPRESSIBILITY: MACH NUMBER computed when the Mach number effects are ignored. Com-
EFFECTS pressibility effects are particularly important at high altitude.

Wing divergence dynamic pressure depends on the Mach

number. Mach number, M, is the ratio of airspeed divided 5 CONTROL EFFECTIVENESS
by the local speed of sound and measures the importance of
flow compressibility. Dynamic pressure is written in terms of Aeroelastic flight control problems began in the 1920s. In
Mach number as: 1927 the British Bristol Bagshot fighter prototype encoun-
tered aileron reversal. Figure 14 shows the typical section
1 2 1 with a control surface added. An angle of attack αo and a
q= ρV = ρaM 2 (9)
2 2 control surface rotation δo produce lift, L, and a pitching
moment at the aerodynamic center, MAC , expressed as:
The symbol a represents the speed of sound and depends
on altitude; this relationship is called the atmosphere line. L = qSCLα (αo + θ) + qSCLδ δo (12)
The Prandtl–Glauert transformation provides an estimate
of the lift curve slope for compressible flow written as:
MAC = qScCMAC + qScCMδ δo (13)
CLα0
CLα = √ (10) where
1 − M2
∂CL
As a result, = C Lδ (14)
∂δ

KT 1 − MD2 1 and
qD = = ρaMD2 (11)
SeCLα0 2 ∂CMAC
CMδ = (15)
∂δo
The solution to this equation is called the match point. The
Mach number used to compute the aerodynamic features of a negative number. The twist, θ, is:
the typical section matches the local Mach number used to
compute the dynamic pressure. 
Figure 13 plots divergence speed as a function of altitude qSeCLα αo + c CMAC
e CLα
+ qSeCLδ + ce CMδ δo
θ=
for a typical section that will diverge at 250 lb ft−2 in incom- KT − qSeCLα
pressible flow. This is compared to the divergence speed
(16)

The lift only resulting from δo (with αo = 0) is

CLα  
1 + q KScT CLδ
CM δ
L = qsCLδ δo (17)
1 − q̄

Figure 13. Altitude vs. divergence speed showing compressibility

effects (qD0 = 250 lb ft−2 ). Figure 14. Typical section with control surface.

Encyclopedia of Aerospace Engineering, Online © 2010 John Wiley & Sons, Ltd.
This article is © 2010 John Wiley & Sons, Ltd.
This article was published in the Encyclopedia of Aerospace Engineering in 2010 by John Wiley & Sons, Ltd.
DOI: 10.1002/9780470686652.eae149
 8 Aeroelasticity and Aeroservoelasticity

with

CLδ ∂CL ∂α ∂α
= × = (18)
CL α ∂δ ∂CL ∂δ

defined as the apparent change in section angle of attack

resulting from a unit control deflection, δo . This term is a
function of the ratio of the control surface chord to wing
chord ratio. The lift expression is also written as: Figure 15. Interactions between ailerons and wing bending motions
at high speed produced the first catastrophic flutter events in the

 1920s (Cook, 1991).
1 − (q/qR )
L = Lr (19)
1 − (q/qD )

where 1915 when the British Handley bi-plane bomber experienced

severe tail flutter.
−KT CLδ −KT ∂α After World War I airspeeds increased; low drag, semi-
Lr = qSCLδ δo qR = =
ScCMδ CLα ScCMδ ∂δo monocoque, monoplane designs appeared and with them
wing-aileron flutter. As indicated in Figure 15, at the onset of
KT
qD = (20) wing-aileron flutter, phasing between the aileron motion and
SeCLα the wing motion is such that lift from the aileron motion
inputs energy into the wing (Cook, 1991). The result is
The term qR is called the aileron reversal dynamic pressure undamped, exponentially divergent oscillations. The flutter
and is a positive number since CMδ is a negative number. speed depends on the aileron/wing connection stiffness, wing
Control effectiveness for the typical section is defined as: bending frequency and the location of the aileron center on

 gravity with respect to the aileron/wing hinge.
L 1 − (q/qR ) Wing/aileron flutter and its counterpart, rudder/vertical tail
= (21)
Lr 1 − (q/qD ) flutter, were common in the 1920s. The British Gloster-Grebe
biplane fighter encountered wing/aileron flutter in 1923. In
If the flight dynamic pressure, q, is less than the divergence the Netherlands, the Van Berkel W.B. seaplane experienced
dynamic pressure, then, as dynamic pressure approaches qR , violent wing oscillations in flight.
L/Lr decreases; it becomes zero when q = qR . At this speed In Germany, Von Bambauer and Koning (1923) identi-
the control surface deflection produces no lift because lift fied the cause of and solutions to wing/aileron flutter by
created by nose-down twist cancels control deflection lift. conducting experimental and theoretical investigations of
Control reversal is not a structural instability. Near the wing/aileron dynamic interaction. They were the first to
divergence speed a very small input αo produces a very large understand that flutter can be prevented by decoupling modes
structural response θ. At reversal a large input δo produces no of motion. Their decoupling was the result of attaching small
output (L/Lr ). At reversal the twist angle θ may be large or weights ahead of the aileron hinge line; this “mass balanc-
small. Above the reversal speed, a control input δo produces ing” solution is still used. In 1925 the last wooden British
lift force opposite to that intended, but the structure is still fighter, the Gloster Gamecock, encountered rudder/vertical
stable. tail flutter. This was also eliminated by mass balancing the
rudder.
Accurately predicting flutter speed depends strongly on
6 FLUTTER AND BUFFET APPEAR unsteady airload calculations. German researchers took
the lead developing unsteady aerodynamic theories that
The term flutter first appeared in 1924 in a published doc- addressed flutter problems (Herschel, Prem and Madelung,
ument by R.T. Glazebrook in the Yearbook of the British 2004). In 1922 Walter Birnbaum provided expressions for
Aeronautical Committee. Flutter is a dynamic, oscilla- the oscillatory aerodynamic loading on a two-dimensional
tory structural instability enabled by interactions between flat plate (much like the typical section) undergoing low-
unsteady aerodynamic forces and moments created by vibra- frequency oscillatory motion in an airstream. Two years
tory motion of lifting surfaces and the vehicles to which these later Herbert Wagner developed the first transient solu-
surfaces are attached. Empennage/tail flutter appeared first in tion, the Wagner function, describing lift development on a

Encyclopedia of Aerospace Engineering, Online © 2010 John Wiley & Sons, Ltd.
This article is © 2010 John Wiley & Sons, Ltd.
This article was published in the Encyclopedia of Aerospace Engineering in 2010 by John Wiley & Sons, Ltd.
DOI: 10.1002/9780470686652.eae149
 Static and Dynamic Aeroelasticity 9

two-dimensional flat plate airfoil when it was given an instan-

taneous increase in its angle of attack. By the end of the 1920s
the fundamentals of flutter were clearly understood (Frazer
and Duncan, 1920).
In the United States, Theodorsen’s work on unsteady aero-
dynamics added further fidelity to aeroelastic analysis of
the typical section (Theodorsen, 1935). A key parameter in
these calculations is the reduced frequency of oscillation,
defined as k = ωbV
. The parameter b is the wing section semi-
chord, whereas V is the airspeed, and ω is the frequency of
oscillation. Flow unsteadiness is very important for calcula-
tions for systems with reduced frequencies beginning about
k = ωbV
≈ 0.04. Smilg and Wasserman (1942) outlined a flut-
ter analysis procedure called the V-g method. This method
was a standard computational technique for several decades.

7 SWEPT WING AEROELASTICITY

In 1935 Dr Adolph Busemann first proposed sweeping wings

to delay the onset of wave drag resulting from high speed Figure 17. Swept wing bending flexibility produces washout that
reduces local angle of attack at wing sections such as A-A.
flow compressibility. High-speed swept wing aircraft designs
appeared in Germany late in World War II. There are three
reasons to sweep a wing forward or backward: (i) to improve led to radically new ways of analyzing aeroelastic behavior
longitudinal stability by reducing the distance between the (Bhatia, 2003).
aircraft center of gravity and the wing aerodynamic cen- As indicated in Figure 17, sweptback wing bending dis-
ter; (ii) to provide longitudinal and directional stability for placements reduce wing section angles of attack, leading
tailless (flying) wings; and (iii) to delay transonic drag rise to three static aeroelastic problems: (i) flexible sweptback
(compressibility). wings are lift ineffective because wing bending reduces the
Although not the first swept wing aircraft, the American total wing lift for a given wing angle of attack; (ii) bending
B-47 jet bomber shown in Figure 16, was the first to encounter deformation moves the wing center of pressure inboard and
and address high-speed swept wing aircraft aeroelastic issues forward; and, (iii) bending displacement reduces sweptback
ranging from aerodynamic load redistribution to flutter. The wing aileron effectiveness as well as sweptback tail/rudder
B-47s high aspect ratio swept wings and speed of 607 mph effectiveness so much that ailerons are often replaced by
spoilers. On the positive side, local angle of attack reduc-
tion created by swept wing bending counters the increased
angle of attack created by torsion. This cancellation makes
wing divergence unlikely if wings are swept back >10–15◦ ,
but more likely if the wings are swept forward.
The increased complexity of high-speed swept wing jet
aircraft construction challenged engineers. Swept wing static
aeroelasticity studies appeared in the late 1940s and early
1950s (Diederich and Budiansky, 1948; Pai and Sears, 1949;
Diederich and Foss, 1952). Static aeroelastic methods for
the complete airplane were developed at the Boeing Com-
pany about 1950 and published in an NACA report (Gray and
Schenk, 1953). Design, testing and certification created new
demands for analytical fidelity. These demands led Boeing
engineers to develop a new aero/structural analysis method
Figure 16. The Boeing B-47, first flown 17 December 1947, was called the finite element method. The paper by Turner et al.
the first aircraft to address swept wing aeroelasticity. (1956) coupled with the emergence of high-speed computers,

Encyclopedia of Aerospace Engineering, Online © 2010 John Wiley & Sons, Ltd.
This article is © 2010 John Wiley & Sons, Ltd.
This article was published in the Encyclopedia of Aerospace Engineering in 2010 by John Wiley & Sons, Ltd.
DOI: 10.1002/9780470686652.eae149
 10 Aeroelasticity and Aeroservoelasticity

laid the groundwork for a new powerful method of structural (ii) its onset usually leads to limit cycle oscillations that cre-
analysis that has since been embraced by the entire structural ate a severe, sustained, acoustic environment and lead to high
engineering community. cycle fatigue.
High-speed flutter instabilities include single degree of
freedom flutter (Lambourne, 1968). Unlike classical flutter
8 HIGH-SPEED FLUTTER PROBLEMS requiring modal interaction, single degree of freedom flut-
ter involves instability mechanisms such as loss of damping,
In 1947 the Bell X-1 broke the sound barrier; soon many flow separation, pure pitching oscillations, or pure bend-
fighters and a few bombers were flying at supersonic speeds. ing oscillations. Single degree of freedom pitch oscillations
In 1958 the X-15 flew at hypersonic speeds at the edge of the require oscillation at special reduced frequencies and pitch
Earth’s atmosphere. By 1982 the first Space Shuttle showed axis locations.
that it could operate at speeds ranging from subsonic to hyper- A common sustained, limit cycle flutter with one degree
sonic. of freedom is control surface buzz. This instability requires
During the post-World War II era, airplane size also grew, shock wave–boundary layer interaction to trigger separated
as did the variety of different configurations, including fly- flow with periodic shock wave reattachment. This instability
ing wings. A new phenomenon called Body Freedom Flutter is highly nonlinear with the onset of a limit cycle oscillation
appeared. Body freedom flutter occurs as the aircraft short dependent upon unsteady aerodynamic phenomena.
period mode, a basic flight mechanics parameter, increases
with airspeed and comes in close proximity to a wing vibra-
tion mode. The modal coupling between aircraft pitching 9 FLUTTER DESIGN SOLUTIONS
motion and wing structural vibration strongly affects the
dynamics of the aircraft. This problem appeared on the RB-57 The key to preventing classical flutter events is to eliminate or
reconnaissance aircraft, a version of the B-57 with extended, delay the modal interaction shown in Figure 19. Flutter prob-
high aspect ratio wings and a short fuselage. It is also a pos- lems are ameliorated by combinations of: (i) mass distribution
sible mode of instability on flying wings and forward swept changes (adding or removing mass at specific lifting sur-
wing designs. face locations); (ii) structural stiffness changes (thickening
High speed flight created new aeroelastic challenges for wing skin or other structural members); or (iii) aerodynamic
launch vehicle structures. Supersonic panel flutter, depicted surface changes (changing wing span, chord dimensions or
in Figure 18, occurs because of modal coupling between the reshaping planiform areas).
first two bending modes of a platelike surface. Early models Figure 20 shows that the location of the wing section
of the German V-2 rocket were rumored to have been lost center of gravity changes flutter speed. Placing the center
because of flutter of panels near the rocket nose. Panel flutter of gravity near the 1/4 chord significantly increases flut-
plagued X-15 tail surface panels, creating loads and acoustic ter speed. Center of gravity position also depends on fuel
problems. The Saturn V launch vehicle also suffered from
this problem.
Panel flutter differs from other wing flutter because;
(i) only one side of the platelike structure is exposed to high
speed flow whereas the other is in contact with dead air; and

U
y=f(x) e-iωt

L
x=o x=w

Figure 18. Panel flutter occurs at supersonic speeds and is charac- Figure 19. Delaying or eliminating modal coupling as airspeed
terized by bending vibration modal coupling. increases is the key to preventing classical flutter.

Encyclopedia of Aerospace Engineering, Online © 2010 John Wiley & Sons, Ltd.
This article is © 2010 John Wiley & Sons, Ltd.
This article was published in the Encyclopedia of Aerospace Engineering in 2010 by John Wiley & Sons, Ltd.
DOI: 10.1002/9780470686652.eae149
 Static and Dynamic Aeroelasticity 11

Figure 20. Wing cross-sectional center of gravity location is an

important factor in determining flutter speed. Forward c.g. locations
are better. Figure 21. The DARPA X-29 research aircraft promoted aeroelas-
tically tailored advanced composite materials to control undesirable
tank placement, under-wing missile or bomb placement, and static aeroelastic effects.
engine placement. Re-positioning engines, including length-
ening or shortening attachment pylons, affects flutter speed. aligned along the wing swept axis leads to deformation called
Flutter speed increase may require mass-balancing, adding wash-in and increased airloads. This reduces the wing diver-
nonstructural mass, such as lead weights, to critical places on gence airspeed compared to an unswept wing. Substantial
a wing, an engine, an aileron or an elevator. added structural stiffness (and weight) is required to provide
Many fighter aircraft carry a wide variety of externally aeroelastic stability.
mounted, under-wing armaments or “stores.” Operational Re-orienting laminate fibers changes bend/twist coupling.
requirements may create a large number of possible store As the wing bends upward it twists in the nose-down direc-
combinations. This, in turn, may result in missions where the tion, creating wash-out. This reduces the local airloads and
flutter speed is less than possible operational airspeeds. increases wing divergence speed without extra weight. Fig-
ure 23 summarizes the effects of laminate fiber orientation
on sweptback wing aeroelastic phenomena (Shirk, Hertz and
10 CONTROLLING AND EXPLOITING Weisshaar, 1986).
Aeroelastic phenomena such as aileron reversal have been
AEROELASTICITY: AEROELASTIC
exploited to create actively controlled, light-weight struc-
TAILORING AND tures. The X-53 Active Aeroelastic Wing (AAW) test aircraft
AEROSERVOELASTICITY

With the advent of modern advanced composite materials,

passive aeroelastic control, known as aeroelastic tailor-
ing, has come to include structural stiffness design to
control deformation mode coupling. Strong directional stiff-
ness exhibited by advanced composite materials such as
graphite/epoxy provides the ability to decouple or couple
aeroelastic deformations. Aeroelastic tailoring through use
of composite laminate design is used to increase lift effec-
tiveness, control effectiveness, and flutter speed.
In the early 1970s aeroelastic tailoring was applied to
sweptforward wing designs with inherently low wing diver-
gence speeds. The DARPA X-29 research aircraft shown in
Figure 21 was the result of these efforts.
Aeroelastic tailoring introduces structural bending/torsion
elastic coupling by offsetting the direction of laminate fibers, Figure 22. Sweptforward wing aeroelastic tailoring increases wing
as indicated in Figure 22. A sweptforward wing with the fibers divergence speed.

Encyclopedia of Aerospace Engineering, Online © 2010 John Wiley & Sons, Ltd.
This article is © 2010 John Wiley & Sons, Ltd.
This article was published in the Encyclopedia of Aerospace Engineering in 2010 by John Wiley & Sons, Ltd.
DOI: 10.1002/9780470686652.eae149
 12 Aeroelasticity and Aeroservoelasticity

Figure 25. Forward swept wing body-freedom flutter replaces

static divergence as an instability mode.

11 STATIC STABILITY OR FLIGHT

DYNAMIC STABILITY: DIVERGENCE
OR FLUTTER?

Wing static divergence is a theoretical phenomenon; it has

Figure 23. Aeroelastic tailoring applied to sweptback wing aeroe- been called “flutter at zero frequency.” Divergence of lift-
lastic performance, showing preferred directions for improving ing surfaces is unlikely to appear in flight because, unlike
aeroelastic performance.
a wing rigidly mounted in a wind tunnel, freely flying air-
craft respond to large forces developed when an aerodynamic
surface deforms. This is particularly true of flying wings,
shown in Figure 24 uses a combination of active and passive sweptforward wings, joined wings, and oblique wings.
aeroelastic control to produce a highly flexible, lightweight When the airplane is treated as a dynamic system, the
control system (Pendleton et al., 2007). The X-53 wing has a wing divergence mode appears as a low frequency wing-body
relatively low reversal speed; it uses ailerons to create wing vibration mode. This mode displays a classical modal inter-
distortion much like wing warping to roll the aircraft. action with other flight mechanics modes, in particular the
Aeroservoelasticity uses interactive, active flight control to short period mode, to create body freedom flutter as shown
modify aeroelastic dynamic response and stability. In the past in Figure 25.
few decades, aircraft active flight control has brought flight As a result, most aeroelastic analyses are conducted with
mechanics much closer to aeroelasticity than it has been in the models that include rigid-body degrees of freedom. These
past. Until a few decades ago, except in unusual cases, aeroe- mathematical models are far more complex than those first
lasticians isolated lifting surface aeroelastic response from imagined decades ago when the first flutter analysts began
vehicle response. Aeroservoelasticity began by improving their quest for understanding and control. In addition, this
XB-70 supersonic bomber ride quality. Later active flutter approach has moved aeroelasticity from a specialty area into
suppression using actively controlled ailerons was demon- the mainstream of aircraft structures that recognizes the need
strated on a B-52 test aircraft (Roger, Hodges and Felt, 1975). to treat the system as a truly integrated flying system.

12 SUMMARY

The strong interactions among various aeroelastic con-

stituencies, such as structures, flight mechanics and control,
unsteady aerodynamics, and vehicle performance has cre-
ated unusual problems for aircraft development during the
past century. Creative, skilled aeroelasticians have risen to
the occasion and solved these problems. Today, the wealth of
comprehensive analytical methods available to engineering
analysts allows developers to examine aeroelastic phenomena
before they produce unpleasant surprises. However, aeroe-
lasticity continues to be a vibrant discipline required when
new and unusual aircraft configurations are designed. Future
vehicles such as those with rapidly changing mass prop-
Figure 24. X-53 Active flexible wing aircraft. erties, trans-atmospheric aircraft or vehicles with rapidly

Encyclopedia of Aerospace Engineering, Online © 2010 John Wiley & Sons, Ltd.
This article is © 2010 John Wiley & Sons, Ltd.
This article was published in the Encyclopedia of Aerospace Engineering in 2010 by John Wiley & Sons, Ltd.
DOI: 10.1002/9780470686652.eae149
 Static and Dynamic Aeroelasticity 13

changing shape, such as morphing aircraft may hold inter- Roger, K.L., Hodges, G.E. and Felt, L. (1975) Active flutter
active surprises. This statement is particularly applicable to suppression: a flight test demonstration. J. Aircraft, 12, 551–
new generations of unmanned vehicles, micro air vehicles 556.
and hypersonic vehicles. Each new flight era brings with it Shirk, M.H., Hertz, T.J. and Weisshaar, T.A. (1986) A survey of
aeroelastic tailoring, theory, practice, promise. J. Aircraft, 23(1),
the prejudices of the past and the challenges for the future.
6–18.
So too it will be with aeroelasticity.
Smilg, B. and Wasserman, L.S. (1942) Application of three-
dimensional flutter theory to aircraft structures. Army Air Force
Technical Report 4798.
REFERENCES Theodorsen, T. (1935) General Theory of Aerodynamic Instability
and the Mechanism of Flutter, NACA, Washington, DC.
Abzug, M.J. and Larabee, E.E. (2002) Airplane Stability and Con- Turner, M.J., Clough, R.W., Martin, H.C. and Topp, L.J. (1956)
trol: a History of the Technologies That Made Aviation Possible, Stiffness and deflection analysis of complex structures. J. Aero-
Cambridge University Press, Cambridge. naut. Sci., 23, 805.
Baker, D. (1994) Flight and Flying: A Chronology, Facts On File, Vogel-Prandtl, J. (2004) Zeitschrift fur Flugtechnik und Motor
New York. Luftschiffahrt, 17(7), 154–169.
Bhatia, K.G. (2003) Airplane aeroelasticity: practice and potential. Von Bambauer, A.G. and Koning, C. (1923) On the Stability of
J. Aircraft, 40(6), 1010–1018. Oscillations of an Airplane Wing, International Air Congress,
Collar, A.R. (1978) The first fifty years of aeroelasticity. Aerospace, London, re-published as NACA-TM-223.
5(2), 12–20.
Cook, W.H. (1991) The Road to the 707, The Inside Story of Design-
ing the 707, TYC Publishing, Bellevue, WA.
Diederich, R.W. and Budiansky, B. (1948) Divergence of Swept TEXTBOOKS AND FURTHER READING
Wings, NACA, Washington, DC.
Diederich, F.W. and Foss, K.A. (1952) Charts and Approximate Classic aeroelasticity textbooks describe aeroelastic phenomena in
Formulas for the Estimation of Aeroelastic Effects on the Loading all flight speed regimes and provide methods to analyze them.
of Swept and Unswept Wings, NACA, Washington, DC. These methods have been largely supplanted by modern compu-
tational developments, but textbooks remain an excellent source
Fokker, A.H.G., (1931) The Flying Dutchman, Henry Holt and Co.,
of fundamental explanations, classic examples, and valuable, basic
New York.
computational techniques that can be used for first estimates of
Frazer, R.A. and Duncan, W.J. (1920) The Flutter of Airplane aeroelastic effects. The following references are recommended for
Wings. Reports and Memoranda 1155, Aeronautical Research further study.
Committee.
Bisplinghoff, L., Ashley, H. and Halfman, R.L. (1996) Aeroelas-
Friedmann, P.P. (1999) The renaissance of aeroelasticity and its ticity, Courier Dover Publications, New York. (The timeless
future. J. Aircraft, 36(1), 105–121. classical textbook for aeroelasticians by three prominent contrib-
Garrick, I.E. and Reed, W.H. (1981) Historical development of utors to aeroelasticity research Filled with historical background
aircraft flutter. J. Aircraft, 18(11), 897–912. references and a wealth of examples.)
Gray, W.L. and Schenk, K.M. (1953) A Method for Calculating the Bisplinghoff, L. and Ashley, H. (2002) Principles of Aeroelasticity,
Subsonic Steady-State Loading on an Airplane with a Wing of Courier Dover Publications, New York.
Arbitrary Plan Form and Stiffness, NACA, Washington, DC. Dowell, F., Crawley, E.F., Curtiss, H.C. Jr., Peters, D.A., Scanlan,
Herschel, E..-H., Prem, H. and Madelung, G. (2004) Aeronauti- R.H. and Sisto, F. (1995) A Modern Course in Aeroelasticity,
cal Research in Germany: from Lilienthal until Today, Springer, Kluwer Academic, London. (A valuable modern reference book
New York. with broad perspective.)
Lambourne, N.C. (1968) Flutter in one degree of freedom, Fung, C. (2002) An Introduction to the Theory of Aeroelasticity,
In AGARD Manual on Aeroelasticity, Advisory Group for Courier Dover Publications, New York. (One of the two timeless
Aerospace Research and Development, Neuilly-sur-Seine, classical textbooks on aeroelasticity created by a master writer
France. and teacher.)
Pai, S.I. and Sears, W.R. (1949) Some aeroelastic properties of swept Hodges, H. and Pierce, G.A. (2002) Introduction to Structural
wings. J. Aeronaut. Sci., 16(2), 105–116. Dynamics and Aeroelasticity, Cambridge University Press,
Pendleton, E., Flick, P., Voracek, D., Reichenbach, E., Griffin, K. Cambridge.
and Paul, D. (2007) The X-53, a summary of the active aeroelastic Scanlan, H. and Rosenbaum, R. (1968) Introduction to the
wing flight research program. AIAA Paper 2007-1855. Study of Aircraft Vibration and Flutter, Dover Publications,
Ricketts, R.H. (1983) Structural testing for static failure, flutter and New York.
other scary things, NASA Technical Memorandum 84606. NASA Wright, R. and Cooper, J.E. (2008) Introduction to Aircraft Aeroe-
Langley Research Center, Hampton, VA. lasticity and Loads, Wiley, New York.

Encyclopedia of Aerospace Engineering, Online © 2010 John Wiley & Sons, Ltd.
This article is © 2010 John Wiley & Sons, Ltd.
This article was published in the Encyclopedia of Aerospace Engineering in 2010 by John Wiley & Sons, Ltd.
DOI: 10.1002/9780470686652.eae149
Encyclopedia of Aerospace Engineering, Online © 2010 John Wiley & Sons, Ltd.
This article is © 2010 John Wiley & Sons, Ltd.
This article was published in the Encyclopedia of Aerospace Engineering in 2010 by John Wiley & Sons, Ltd.
DOI: 10.1002/9780470686652.eae149