SUMMARY

In the autumn of 1902 the Wright brothers spent just over eight weeks
at their test site in the Kill Devil Hills near Kitty Hawk, North
Carolina, testing their third Glider design. During the trial period they
implemented an inter-linked roll-yaw control system. Together with
the forward canard surface, this gave them control over vertical and
horizontal components of the flight path. They were also able to hone
and perfect their piloting skills. In just two days in the final week, they
made about 250 glides. The success of the trials instilled the confi-
dence in the Wright brothers to proceed rapidly to the construction of
a powered aircraft. Within a month of returning to Dayton, they were
writing to engine manufacturers with their specification an engine
that would develop eight to nine brake horse power, weigh no more
than 180lb and be free from vibration; they would not find a suitable
powerplant and had to design and build their own. The invention of
the powered aeroplane in 1903 somewhat overshadows the earlier crit-
ical flight control developments, but the birth of flight control in 1902
opened the way for aviation to flourish. With the aid of modern flight
science techniques wind-tunnel testing, computational flight dynam-
ics and piloted simulation, this paper examines the technology of the
Wrights 1902 glider. The research forms a part of the Liverpool
Wright Project, aiming to bring to life the Wright brothers achieve-
ments in this centenary period. Wilbur and Orville Wright are recog-
nised by many as the first aeronautical engineers and test pilots. In so
many ways they set standards that todays engineers and organisations
benefit from. Their work in the period 1901 to 1902 reflects their
genius and the paper reviews this work in detail, examining the design,
aerodynamic characteristics and flying qualities of the aircraft that first
featured a practical three-axis control system.
NOMENCLATURE
A system matrix
AR aspect ratio (span
2
/wing area)
b wing span
B control matrix
c wing chord
C
D
drag coefficient
C
l
rolling moment coefficient
C
L
lift coefficient
C
M
pitching moment coefficient
C
n
yawing moment coefficient
g gravitational acceleration
H
n
static margin
I
xx
, I
yy
, I
zz
moments of inertia about body axes
k (Smeaton) coefficient of air pressure
K

, K

, K

pilot gains
K
p
pilot transfer function between error and control
p, q, r perturbation angular velocities about body axes
Re Reynolds number (Vc/)
S wing area
T
1
, T
2
time constants of closed-loop zeros
T
D
time to double amplitude
T
s
spiral mode time constant
T
SP1
, T
SP2
time constants of short period mode
u control vector
U
e
trim (equilibrium) speed component along x axis
u,v,w perturbation velocities along body axes
V flight velocity
X
u
, M
w
aerodynamic derivatives
THE AERONAUTICAL JOURNAL DECEMBER 2003 697
Paper No. 2854. Manuscript received 22 August 2003 accepted 29 September 2003.
The birth of flight control:
An engineering analysis of the Wright brothers
1902 glider
G. D. Padfield
Professor of Aerospace Engineering
B. Lawrence
Research student
Department of Engineering
The University of Liverpool
Liverpool, UK
x state vector
Y
c
transfer function between canard and pitch attitude
angle of incidence (attack)
angle of sideslip

c
canard angle (positive pitch up)

w
warp angle (positive roll left) = incidence right tip
incidence left tip

r
rudder angle (positive nose right)

p
,
p
damping ratio and frequency of phugoid

d
,
d
damping ratio and frequency of Dutch roll
coefficient of viscosity
air density
, , Euler attitudes

e
,
d
pitch attitude error and demand
eigenvalue
rate of change of rolling moment coefficient with warp angle
rate of change of yawing moment coefficient with rudder
angle
1.0 INTRODUCTION
The Wright brothers had intended their flights on 17 December 1903
to be a private matter and had no intention of making a detailed
public statement. However, since, as Wilbur wrote, the contents of
a private telegram, announcing to our folks at home the success of
our trials, was dishonestly communicated to the newspapers, they
felt the need to put the record straight
(1)
. Wilburs statement to the
Associated Press, written on 5 January 1904, ended thus,
...... we were determined, before returning home, to know
whether the machine possessed sufficient power to fly, sufficient
strength to withstand the shocks of landings, and sufficient capac-
ity of control to make flight safe in boisterous winds, as well as in
calm air. When these points had been definitely established, we at
once packed our goods and returned home, knowing that the age
of the flying machine had come at last.
In this brief and reluctant communication to the world, Wilbur
revealed the purposeful intentions of their aeronautical endeavours,
emphasising performance, structural strength and control as the com-
bined objectives of the worlds first successful flight test of a
powered aeroplane. The following two years were to prove that the
Wrights had much more development work to do to arrive at a prac-
tical aeroplane

, but the fundamental challenge of flight control had

been solved through research and development over the two years
leading up to December 1903. The critical breakthrough came in the
autumn of 1902.
On the night of 23 October 1902, Orville Wright wrote to his sister
Katharine from their camp at the Kill Devil Hills near Kitty Hawk,
North Carolina
(1)
.
The past five days have been the most satisfactory for gliding
that we have had. In two days we made over 250 glides, or more
than we had made all together up to the time Lorin left. We have
gained considerable proficiency in the handling of the machine
now, so that we are able to take it out in any kind of weather. Day
before yesterday we had a wind of 16 meters per second or about 30
miles an hour, and glided in it without any trouble. That was the
highest wind a gliding machine was ever in, so that we now hold all
the records! The largest machine we handled in any kind of
weather, made the longest distance glide, the longest time in the air,
the smallest angle of descent, and the highest wind!!!
The machine that Orville referred to was the Wright brothers 1902
glider, featuring, for the first time, three-axis control over the motion
of the aircraft. The Wrights had flown between 700 and 1,000 flights
in this aircraft during the autumn of 1902. It is testament to the genius
of the Wrights that they had accomplished this achievement working
part time in about three years.
Figure 1 shows one of Orvilles first flights with the three-axis
control operational on the 1902 glider. He and Wilbur had thought
together to arrive at this design innovation. The aircraft weighed
119lbf (53kgf) empty, 259lbf (115kgf) with Wilbur on board and
264lbf (118kgf) with Orville on board. According to Orvilles diary
(1)
,
the longest distance of 622ft was flown by Wilbur in 26 seconds with
a glide angle of just over 8 deg. The best glide angle was quoted as
about 6 deg
(1)
(see Ref. 1, p 266 for flights on 30 September) when the
wind was 5ms
-1
and speed over ground averaged 23ft/sec

. The
longest and highest duration glides were made on the north slope of
the big Kill Devil Hill (max slope 95, Ref. 1 p 259) on 23 October
with a glide slope of between 8 and 9. Over the few weeks at the end
of September and into October, the brothers developed and perfected
their flying skills. To quote their friend and aeronautical colleague,
Octave Chanute
(1)
,
The two brothers glided alternately and they soon obtained
almost complete mastery over the inconsistencies of the wind. They
met the wind gusts and steered as they willed. They did not venture
to sweep much more than one quarter circle, so as not to lose the
advantage of a headwind, but they constantly improved in the
control of the machine and in learning the arts of the birds. Some
800 glides were made.
The autumn flight test campaign of 1902 was very successful for the
Wrights and one is left with the thought that their solution to three-axis
control, linking roll and yaw control to mitigate the powerful adverse
yaw effects, was one of the critical breakthroughs in the history of
aviation and aeronautical engineering. Peter Jakab sums this up in his
book on the Wrights process of invention
(3)
; If the Wright brothers
are to be cited as the inventors of the airplane based on having
resolved all the fundamental problems of mechanical flight then it is
not necessary to look beyond the 1902 glider what was innovative
about the (1903) Flyer was present in the earlier 1902 glider..
This paper records a study of this aircraft and its development, set in
the context of aeronautical knowledge at the turn of the 20th century
and reflected by the synthesis and analysis conducted in the Liverpool
Wright project. Section 2 presents a resume of the Wrights work prior
to the 1902 tests and Section 3 goes on to interpret the research and
development process undertaken during the winter of 1901-2. Section
4 describes the 1902 flights tests and how the three-axis control
698 THE AERONAUTICAL JOURNAL DECEMBER 2003

l
w
C

n
r
C

Figure 1. The Wright brothers glider on 10 October 1902, Orville at the

controls, Wilbur at the starboard wing, Dan Tate to port.

On 16 Oct 1905 the Wrights flew their Flyer No 3 for the last time until May 1908; two
weeks before, Orville had flown for 39km in 38 minutes. In the first two weeks of
October 1905, the Wrights (Ref. 1, p 517) ..did more flying than in all our previous
flights of three years put together. The first powered aircraft flight in Europe (by
Santos Dumont) would not take place until the autumn of 1906 (Ref. 2).

The Wrights regularly mixed their units as a result of both US and European influences.
solution evolved. In Section 5, the Wrights invention is re-visited
with modern engineering tools including wind tunnel testing, com-
putational aerodynamics, flight dynamics analysis and piloted simu-
lation. Here, some of the fundamental aspects of the Wrights
approach the use of the canard, wing warping and the warp-rudder
interlink, wing anhedral and the mastery of control over stability
are explored to reveal the unique nature of their genius. The paper is
rounded off with a very brief reflection on the consequent, post-
1902, activities of the Wrights in Section 6 and some concluding
remarks in Section 7.
2.0 THE WRIGHT BROTHERS AND THEIR
PRE-1902 WORK
2.1 Background
Wilbur was born in 1867 and Orville in 1871, the same year the
family moved into a new house, No 7 Hawthorn Street, in Dayton,
Ohio. They had two elder brothers Reuchlin and Lorin, and a
younger sister, Katharine. A clear picture of the kind of life the
Wright brothers experienced during their upbringing is portrayed in
Tom Crouchs thesis on a life of Wilbur and Orville Wright The
Bishops Boys
(4)
. Their father, Milton Wright, a minister in the
Church of the United Brethren in Christ, established a corporate
family identity. Crouch describes Milton as rather uncompromising
and strong principled inherited from his father and passed to his
children. Milton considered that the world was not a friendly place
for honest men and women, temptations beckoned, unscrupulous
persons lay in wait; friends would fall away in times of trial and
ultimately the strength of family bonds offered the only real support
one could hope for in life. This austerity seems to have been bal-
anced by the caring attention of their mother, Susan Catherine
(Koerner) Wright. Crouch notes that while no diary survived to show
her side of the story, she played a key role in her childrens fascina-
tion for engineering. Her children remember her as having consid-
erable mechanical aptitude, having spent time in her fathers
carriage shop as a young girl. She designed and built simple house-
hold appliances for herself and made toys for her children. When the
boys wanted mechanical advice and assistance, they came from their
mother. Milton was one of those men who had difficulty driving a
nail straight. Crouch makes the keen observation that, Wilbur and
Orville had their mother to thank for their extraordinary ability to
visualise the operation of mechanisms that had yet to be
constructed.
Tom Crouch provides us with vivid images of the two inventors;
to paraphrase Wilbur was an outgoing person and a gifted speaker,
never rattled in thought or temper; cool, aloof and controlled. He had
struggled to overcome fits of depression in his young manhood,
developing considerable self confidence in the process. He drew
friends from his older brothers circles. Orville, on the other hand,
was especially close to his sister Katharine, almost her protector. He
was impulsive and excitable, an enthusiast and optimist, on fire with
new inventions. While the airplane was Wilburs idea, Orville sup-
plied the drive to continue in difficult times. With the family, Orville
was something of a tease and practical joker, but with strangers he
was painfully shy. He outlived Wilbur by 36 years but refused to
speak in public.
When Wilbur was eleven and Orville seven years old, Milton
bought them a toy helicopter, designed by the Frenchman Alphonse
Pnaud, who himself had based his design on one of Leonardo da
Vincis aircraft concepts. This made a big impression and the broth-
ers went on to construct several models from their own designs. We
see here the seed of their interest in aviation and it was to be nur-
tured over the next 20 years in a variety of ways their fascination
with bird flight, following the work of the German engineer and
PADFIELD AND LAWRENCE THE BIRTH OF FLIGHT CONTROL AN ENGINEERING ANALYSIS OF THE WRIGHT BROTHERS 1902 GLIDER 699
Figure 2. Wilbur and Orville Wright.
Figure 3. Sketch of the Wrights 1899 Kite.
Wilbur Wright
1867-1912
Orville Wright
1867-1948
glider pilot, Lilienthal, and the confidence in their own engineering
knowledge and skills gained from designing and making their own
bicycles the Van Cleve, the St Clair and the Wright Special
during the late 1890s. The story of the Wright brothers is a fascinat-
ing one, leaving the thought that the invention of the aeroplane was
the result of a combination of several critical success factors and
brought to us by the first aeronautical engineers who addressed
structural strength, performance and control with equal attention.
They took what we now describe as a systems approach but they
also realised that control was the most critical aspect.
2.2 1899-1900: The quest begins
By the time Wilbur was 32 he was ready to begin a serious study of
aviation. On 30 May 1899, he wrote to the Smithsonian Institution in
Washington DC
(1)
.
I have been interested in the problem of mechanical and
human flight ever since as a boy I constructed a number of bats of
various sizes after the style of Cayleys and Pnauds machines.
My observations since have only convinced me more firmly that
human flight is possible and practicable. It is only a question of
knowledge and skill just as in acrobatic feats. Birds are the most
perfectly trained gymnasts in the world and are specially well
fitted for their work, and it may be that man will never equal them,
but no one who has watched a bird chasing an insect or another
bird can doubt that feats are performed which require three or
four times the effort required in ordinary flight. I believe that
simple flight at least is possible to man and that the experiments
and investigations of a large number of independent workers will
result in the accumulation of information and knowledge and skill
which will finally lead to accomplished flight. I am about to
begin a systematic study of the subject in preparation for practical
work to which I expect to devote what time I can spare from my
regular business. I wish to obtain such papers as the Smithsonian
Institution has published on this subject, and if possible a list of
other works in print in the English Language. I am an enthusiast,
but not a crank in the sense that I have some pet theories as to the
proper construction of a flying machine. I wish to avail myself of
all that is already known and then if possible add my mite to help
on the future worker who will attain final success.
Wilburs request was granted and the Smithsonian sent him a list of
their publications, which included works by Chanute, Langley and
Means, in addition to a set of Smithsonian pamphlets including ones
by Lilienthal and Langley. In the summer of 1899 Wilbur con-
structed his first kite with a span of 5ft. He was convinced that the
method of control adopted by birds, by shifting their centre of pres-
sure, was more efficient than by shifting the centre of gravity, a
method used in several glider designs during the 1890s e.g. those by
Lilienthal, Pilcher and Chanute. Wilburs control mechanism con-
sisted of wing-warping for lateral control, whereby the outer sections
of the wings of the biplane were twisted by rotating the control
levers in opposite directions; longitudinal control was achieved
through a symmetric rotation of the two levers, shifting the two sur-
faces longitudinally relative to one another, and simultaneously
rotating the canard surface (Fig. 3).
Confident in his method of control, Wilbur proceeded to design a
man-carrying glider, based on Lilienthals data, with a wing area a
little over 150ft
2
. This would be flown tethered, and, according to
the data Wilbur had access to, would support the weight of a man in
a wind of about 16mph, to give the operator practice in control, with
minimum effort. Wilbur needed a location where such steady winds
blew daily during the autumn and winter months, the period his busi-
ness commitments allowed him to pursue his new hobby. He wrote
to the US Weather Bureau in December 1899 and received informa-
tion on the locations where such steady winds could be found.
Eventually, Wilbur chose the place with the 6th highest average
wind in the US, Kitty Hawk in North Carolina, as the location where
the brothers would conduct their flying trials over the next four
years. Eager to set up a dialogue with the author of the book
Progress in Flying Machines, Wilbur wrote to Octave Chanute on 17
May 1900. This would be the first of a great many communications
between the two friends until Chanute died in 1910. The letters,
making up a large part of Ref. 1, provide us with a trail of the work
of Wilbur and Orville Wright, unfolding the process of discovery
and invention and documenting much of the engineering data col-
lected by the Wrights over the period 1900-1905.
In his first letter, Wilbur shared with Chanute his plans to build a
flying machine;
My general ideas on the subject are similar to those held by
most practical experimenters, to wit: that what is chiefly needed is
skill rather than machinery. The flight of the buzzard and similar
sailors is a convincing demonstration of the value of skill, and the
partial needlessness of motors. It is possible to fly without motors,
but not without knowledge and skill. This I conceive to be fortu-
nate, for man, by reason of his greater intellect, can more reason-
ably hope to equal birds in knowledge, than to equal nature in the
perfection of her machinery. My observation of the flight of buz-
zards leads me to believe that they regain their lateral balance,
when partly overturned by a gust of wind, by torsion of the tips of
the wings. If the rear edge of the right wing tip is twisted upward
and the left downward, the bird becomes an animated windmill
and instantly begins to turn, a line from its head to its tail being
the axis.
Wilbur was confiding in Chanute to
learn to what extent similar plans have been tested and found
to be failures, and also to obtain such suggestions as your great
knowledge and experience might enable you to give me. I make
no secret of my plans for the reason that I believe no financial
profit will accrue to the inventor of the first flying machine, and
that only those who are willing to give as well as to receive
suggestions can hope to link their names with the honor of its
discovery.
Wilbur would be more cautious in sharing his ideas in years to come,
but for the crucial few years ahead, he was open in his correspondence
and the encouragement and curiosity displayed by Chanute was clearly
a spur to Wilburs progress during the development of the flying
machine up to 1905.
Wilbur went on to describe his wing warping concept and his plan to
erect a tower for tethering his new glider design, thereby enabling
soaring flight. Wilbur arrived in Kitty Hawk on 11 Sept 1900
(Orville, who was now very much part of the team, arrived on 28
September). The 1900 glider (Fig. 4), with its wing area of about
165ft
2
, aspect ratio 34 (per surface) and 1/23 camber, was assembled
and tested as a kite in early October. From the entries in Wilburs note-
book we can get an understanding of the kind of data he was gathering
from these tests
(1)

the proportion of total drift

to lift would have been less at thirty

five miles per hour than at twenty because the drift of surfaces due to
the smaller angle would have decreased faster than the resistance of
framing increased.
Wilbur was acknowledging the difference between pressure drag
and skin friction drag; he had also discovered that a wings efficiency
improves with speed; later he would realise that the drift to lift ratio had
a minimum value. At small angles of incidence threads projecting
from the under side of the surfaces were turned forward. This
reversed flow effect contributed to the large movement of the centre of
pressure on the cambered section.
700 THE AERONAUTICAL JOURNAL DECEMBER 2003
(a) perspective sketch.
Figure 4. The Wrights 1900 Glider.
(b) kiting the 1901 Glider.

used to mean drag.

With the wind blowing up a slope of one in twelve, the
resistance of the machine at 50lbs was only three or four pounds, the
wind blowing over 20 miles. On the face of a steep hill (one-in-three)
the machine soared in a wind of about ten miles. When a rope is used
to prevent drift the center of gravity must coincide with both center of
lift and center of drift, and the rope must be attached to this point.
Wilbur eventually managed to glide a distance of several hundred
feet with the warp control tied off and flying very close to the surface
of a one-in-six hill, landing at nearly 30mph. Referring to the pitch,
fore-and-aft control, Wilbur informed Chanute from Dayton on 16
November that. The ease with which it was accomplished was a
matter of great astonishment to us.we never found it necessary to
shift the body. Wilburs excitement with the partial success of the first
trials was shared by Chanute

and the latter was keen to report the

progress in a paper he was writing. He asked permission to do this and
Wilbur reluctantly agreed but stated that ..for the present (we)
would not wish any publication in detail of the methods of operation or
construction. The methods of operation were the warp and canard
control concepts and Wilbur Wright knew that his design was innova-
tive and unique. He had spent about two minutes airborne but was con-
vinced that the problem was one of performance and that more wing
area was essentially all that was required.
2.3 1901: Second step into the unknown
The Wright brothers arrived in Kitty Hawk on 11 July with a new
glider. Their 1901 glider had been designed with ample margin to fly in
winds less than 15mph. With a 22ft wing span and a chord of 7ft, the
total lifting wing area was approximately 300ft
2
(Fig. 5). They also
increased the camber to one-in-twelve, the value used by Lilienthal on
his circular arc wing sections. Pitch control was achieved through
varying the camber of the canard by pushing down on a pitch bar; roll
control through wing warp was activated through foot pedal controls.
The 1901 trial period lasted about six weeks, although the first test
flight did not take place until 27 July and they left for Dayton on 20
August. They were joined for part of this period by Chanute. While
they had some success, the overwhelming feeling was of failure. Later,
when reflecting on this period, Wilbur would state
(4)
,
..we doubted that we would ever resume our experiments.
Although we had broken the record for distance in gliding, yet when
we looked at the time and money which we had expended, and con-
sidered the progress made and the distance yet to go, we considered
our experiments a failure. At this time I made the prediction that men
would sometime fly, but that it would not be within our lifetime.
The problems fell into three categories and we look to Wilburs
diary to get his impressions;
Performance (29 July 1901)
Afternoon spent in kite tests. Found lift of machine much less than
Lilienthal tables would indicate, reaching only about a third as much.
Found that machine at 100lbs would not glide at 3 or 4 on wind of
less than about 23 to 25 miles per hour.
(1)
During a glide on 8 August,
the measurements indicated that 10deg of incidence were required at
24mph. With this level of performance there was little chance of
achieving soaring flight, although Fig. 6 shows a classic picture of
Wilbur almost soaring.
Longitudinal Control (30 July 1901)
The control of our machine is not as good as last year. We
attribute this to the fact that the travel of the center of pressure
toward the front edge is slower than in the machine of last year
PADFIELD AND LAWRENCE THE BIRTH OF FLIGHT CONTROL AN ENGINEERING ANALYSIS OF THE WRIGHT BROTHERS 1902 GLIDER 701
Figure 5. The Wrights 1901 Glider.
(b) kiting the 1901 Glider.
(a) perspective sketch.

Figure 6. Wilbur being launched and almost soaring in 1901.

Crouch (Ref. 3) notes that Chanute may never have properly understood Wilbur's
control mechanism, being rather fixed in the view that turning an aircraft was accom-
plished by yawing, much like a boat.
which had less fore-and-aft curvature. We even think that the
direction of travel reverses at an angle within those used in gliding
so that the forward plane, and the variation between centers of pres-
sure and gravity, instead of counteracting each other, at times act
together to disturb the equilibrium.
The use of the 1/12 camber had created such a large nose down
pitching moment that the pilot had to move back to avoid nose diving
into the sand. Wilbur later described the first test flight in his first
public Lecture entitled Some Aeronautical Experiments, in September
1901. Referring to himself as pilot, he stated that,
He kept moving further and further back with each trial, till
finally he occupied a position nearly a foot back of that at which we
had expected to find the center of pressure..The machine then
sailed off and made an undulating flight of a little more than 300ft.
To the onlookers this flight seemed very successful, but to the opera-
tor it was known that the full power of the rudder

had been required

to keep the machine from either running into the ground or rising so
high as to lose all headway. In the 1900 machine, one fourth as much
action had been sufficient to give much better control.
Wilbur was experiencing the consequences of flying a pitch-unstable air-
craft. The center of gravity was behind the neutral point, the aerodynamic
centre of the whole aircraft. The large nose down pitching moment on the
cambered wing section meant that the center of pressure (cp) was a long
way aft at very small angles of incidence, moving forward as the inci-
dence increased and moving aft again at high incidence. The reversing of
the movement of the cp caused major problems for the Wrights. During
the 1901 flight tests, they would reduce the camber to about 1/20 by
trussing down the rib in the center. Even with this arrangement the air-
craft was likely to have been, at best, marginally stable, requiring the full
concentration of the pilot to stay safely airborne.
Lateral control (22 August 1901)
we proved that our machine does not turn towards the lowest
wing under all circumstances, a very unlooked for result and one
which completely upsets our theories as to the causes which
produce the turning to right or left
The problem is illustrated in the sequence in Fig. 7.
The pilot applies warp to roll and turn the aircraft to starboard, (a).
The increased lift on the port wing induces an increase in drag leading
to an adverse yawing motion, dragging the port wing back (b), and
sometimes resulting in the aircraft turning to port rather than star-
board.
The pitch, roll and yaw problems were inherent in the unstable con-
figuration selected by the Wrights but they seemed to know instinc-
tively that control was more important than stability and that, at the
low speeds they wanted to fly, the increased control power conferred
by the canard was quite critical for manoeuvrability. The Wrights
placed a great deal of emphasis on control while at the same time, not
wholly appreciating the role of the aerodynamic moments in stabil-
ity
(5)
. This is entirely understandable considering the mathematical
complexities involved in quantifying the dynamic behaviour of air-
craft; these were being worked on at the time by the British mathe-
matician Bryan, and would be published for the first time two years
later
(6)
.
Despite their obvious despondency as a result of the performance
and control problems, the Wright brothers had some positive aspects
to reflect on. Their aircraft, the largest ever flown, was very strong and
had withstood many heavy landings. In higher winds they had glided
for more than 300ft. Their confidence in the canard first design was
strongly reinforced. To quote from Wilburs Lecture again,
In one glide the machine rose higher and higher till it lost all
headway. This was the position from which Lilienthal had always
found difficulty to extricate himself, as his machine then, in spite of
his greatest exertions, manifested a tendency to dive downward
almost vertically and strike the ground head on with a frightful
velocity. In this case a warning cry from the ground caused the
operator to turn the rudder

to its full extent and also to move his

body slightly forward. The machine then settled slowly to the
ground, maintaining its horizontal position almost perfectly, and
landed without any injury at all. This was very encouraging as it
showed that one of the very greatest dangers in machines with hori-
zontal tails had been overcome by the use of a front rudder.
They had experienced this nose dive tendency at stall with their
1900 glider flown with a tail, but their 1901 glider tended to stall flat.
In a letter to Chanute on 29 August, Wilbur refers to the muddled
state of affairs left over from the 1901 trials. On the same day he per-
formed extensive analysis on the measurements of glide angles and
wind speeds, estimating values for drift and lift. He deduced that it
would not be possible to glide permanently at an angle less than 8 in
still air. He calculated that, based on Lilienthals tables, a glide angle
of 4 should be possible at 17mph. Also on 29 August, Wilbur
received a letter from Chanute inviting him to present a paper on their
gliding experiments to the Society of Western Engineers in Chicago in
three weeks time. His sister Katharine ..nagged him into going. He
will get acquainted with some scientific men and it may do him a lot of
good. Wilburs paper, published in full in Ref. 1, contained about
7,000 words of text, figures and photographs and was a classic
research paper on applied aeronautics. It is easy to imagine that the
failings of the 1901 glider and the reflections Wilbur made in writing
his Some Aeronautical Experiments paper were spurs to the next and
critical stage of development.
702 THE AERONAUTICAL JOURNAL DECEMBER 2003
(a) pilot applies warp to roll to starboard.
(b) adverse yaw turns aircraft to port.
Figure 7. Sequence showing effects of warp-induced adverse yaw.

The Wrights often used the term rudder when referring to the canard control surface.

In Ref. 4, Culick observes that ...the backward state of the general theory and under-
standing of flight mechanics hindered them and in fact caused them considerable diffi-
culties.

Again, the Wrights are referring to the canard pitch control surface.
3.0 1901-1902: INTERPRETING THE
WRIGHTS PROCESS OF INVENTION
In Ref. 1, the account of activities during the period between
October 1901 and February 1902 is documented in about 100 pages
of text, numerous table and figures, and is arguably the richest
record of progress in the entire reference. Here we see the Wright
brothers working together towards a greater understanding of the
performance problems they had experienced. They had made exten-
sive use of Lilienthals data tables in their design and were perplexed
by the results, losing confidence in the process. In a letter to Chanute
written on 6 October 1901, Wilbur writes, I am now absolutely
certain that Lilienthals table is very seriously in error Less than
two weeks later, after painstaking analysis of available data Wilbur
wrote to Chanute again, describing a new balance they would be
using to measure lift and drift; Wilbur adds, It would appear that
Lilienthal is very much nearer the truth than we have heretofore
been disposed to think In this same month, Chanute would send
Wilbur a translation copy of Lilienthals book
(7)
, which would serve
to inform him even further of the errors that both he and Lilienthal
had made; Wilburs letter to Chanute on 2 November
(1)
summarising
his analysis of errors reads like a classic researchers logbook. One
of the major discoveries made by Wilbur in the analysis of existing
data was the error in the, so-called, Smeaton coefficient in the
expression that Wilbur believed Lilienthal had used for computing
lift,
Lift = k S V
2
C
L
k is the Smeaton coefficient of air pressure
(8)
, S the wing area, V the
velocity and C
L
the lift coefficient. Wilbur assumed that Lilienthal
had used Smeatons value of 0005 for k. In fact, Lilienthal had mea-
sured C
L
directly and had no need to use the Smeaton coefficient, so
Wilburs back-calculation to estimate the required wing area and
resulting glide angles were based on an erroneous assumption (Ref.
8 gives a detailed explanation of how the interpretation error came to
pass). Wilbur had reported his suspicion that the Smeaton coefficient
was too high in his letter to Chanute on 26 Sept, Professor
Langley and also the Weather Bureau officials found that the correct
coefficient of pressure was only about 00032, instead of Smeatons
0005. Wilbur had unearthed one of the reasons for the perfor-
mance shortfall in the 1901 glider, but he was determined to estab-
lish the optimum wing design.
With the purpose of creating a detailed aerodynamic database that
would serve their design goals, the Wright brothers built a wind
tunnel and two force balance systems that would provide, effectively,
direct measurements of the lift coefficient and the ratio of drift to lift.
The operation of this balance is very well described by Jakab in Ref.
3. They made measurements to examine the effect of a wide range of
design parameters aspect ratio, camber, point of maximum camber,
biplane surface spacing and others on the lifting capacity and effi-
ciency of the surfaces. Wilbur wrote to Chanute on 15 December,
continuing the information exchange with a summary of the main
results to date, but also indicating that the experimenting had now to
stop to give way for bicycle making. In his letter Wilbur described
laws for superimposing multiple surfaces, re-shaping single surfaces
and revealed the major effect of aspect ratio on the lift curve slope.
He emphasised the benefits of their systematic approach in this
context,
It has been a great advantage we think to make a systematic
measurement of several typical series of surfaces rather than to
work blindly on all sorts of shapes, as a study of the series plates
quickly discloses the general principles which govern lift and tan-
gential and thus renders the search for the best shapes much
easier. A mere glance shows that while increasing the ratio of
breadth to length does not increase the maximum lift to any great
extent it does cause the maximum to be reached at a smaller
angle; and the wider the spread from tip to tip the smaller the
angle at which large lifts can be obtained.
Wilburs breadth was what we now call the wing span while his
length was our chord. In this paragraph, he emphasised the word
best, acknowledging that the physics of aerodynamics clearly did
favour particular arrangements.
The Wrights tested more than 150 surfaces during the last quarter
of 1901. Figure 8, redrawn from Ref. 1, shows eight planforms and
corresponding cambers.
The lift coefficient and drag/lift ration for many of these surfaces
are tabulated in Ref. 1, along with the raw measurements from which
these were derived. Figure 9 shows the effect of aspect ratio (AR) on
lift coefficient and drag/lift ratio, for the parabolic section with 1/20
camber (Wright designated surfaces nos 12 (AR = 6 ), 19 (AR = 4)
and 17 (AR = 1)). All had an area of 6in
2
. The aspect ratio of the
wing of the 1901 glider had been about 314 (per surface) and the
Wrights could now see clearly the efficiency penalty arising from
their choice of this configuration relative to, say, using a wing with
an AR of 6. Over the range of typical incidences used in gliding, a
wing with an aspect ratio of 35 might only be able to lift 75% as
much as a wing with an aspect ratio of 6.
Moreover, the comparison of drag/lift ratio shown in Fig. 9 must
have been a startling revelation to the Wrights. This was a direct
measure of the glide angle at different angles of incidence, or the
efficiency of the lifting surface. The Wright brothers could see for
the first time that a minimum occurred, and that the AR = 6 surface
was twice as efficient as the AR = 1 surface. They could also see that
the minimum for the AR = 6 surface was around 01 and occurred at
an angle of incidence of about 6. If this worked out in practice they
PADFIELD AND LAWRENCE THE BIRTH OF FLIGHT CONTROL AN ENGINEERING ANALYSIS OF THE WRIGHT BROTHERS 1902 GLIDER 703
Figure 8. Sample of surfaces tested with three aspect ratios and three
cambers.
Figure 9. Lift coefficient and drag/lift ratio of surfaces with different ARs.
. . . (1)
would be able to glide flat down a 6 slope! For the 1901 design,
they had used a value of the pressure coefficient which was 50% too
high and they could now see that by changing the aspect ratio from
31 to perhaps 7, they could compensate for this error and design a
new glider with essentially the same wing area as in their 1901
machine. The aspect ratio effect came as a surprise to the Wrights,
even though Lilienthals machine clearly had a higher ratio and they
had used Lilienthals tables apparently oblivious to this important
design parameter. In Ref. 7, Lilienthal recommends particular
dimensions; to quote from his list of fundamental construction
points;
9th It will be a matter of experiment to determine whether
the broad shape of wing with resolved pinions, such as we see in
birds of prey, or the long pointed shape of wing of the sea birds,
are preferable.
10th In the former case the dimensions of the wing would be
8m span, and 16m greatest width (AR = 5).
11th When employing the slender wing shape, the correspond-
ing dimensions would be 11m span and 14m greatest width
(AR=79).
Wilbur had written in his letter to Chanute on 24 November that,
It is very evident from these measurements that a table based
on one aspect and profile is worthless for a surface of different
aspect and curvature. This no doubt explains why we have had so
much trouble figuring all our machines from Lilienthals table.
In his book, A History of Aerodynamics
(8)
, John Anderson pro-
vides a comprehensive analysis of the causes of the under-perfor-
mance in the Wrights 1901 glider, compared with their expectations
a combination of using a lower aspect ratio than Lilienthal, using a
parabolic section rather than a circular arc and using a pressure coef-
ficient of 0005, meant that the Wrights lift calculations were in
error by a factor of 036, correlating very closely with Wilburs
assertion recorded in his diary on 29 July 1901, that they, ..Found
lift of machine much less than Lilienthal tables would indicate,
reaching only about 1/3rd as much.
The wind tunnel experiments conducted during the months of
October to December 1901 had provided the Wrights with the
answers to the performance questions arising from their 1901 flight
trials. For Wilbur and Orville Wright, the control and stability prob-
lems were intractable to solution at model scale or by analysis, and
they would have to wait until the next flight tests to address these.
On 20 August 1902, Katharine Wright, in a letter to her father,
writes,
The flying machine is in process of making now. Will spins
the sewing machine around by the hour while Orv squats around
marking the places to sew. There is no place in the house to live
but Ill be lonesome enough by this time next week and wish that I
could have some of their racket around.
By this time next week, Wilbur and Orville had left for Kitty
Hawk, arriving on Thursday 28 August to undertake their third
flight trial.
4.0 OCTOBER 1902: THE BIRTH OF
THREE-AXIS FLIGHT CONTROL
The aircraft the Wrights made over the summer is illustrated in Figs
10 and 11.
The aircraft featured several design changes compared with 1901.
The aspect ratio was 64, the wing area just over 300ft
2
, the wing
camber reduced to 1/24, the canard had an increased area and aspect
ratio and the aircraft was fitted with a fixed two-surface vertical tail,
located 35ft aft of the wing trailing edge. A small amount of wing
anhedral was built into the new aircraft. The vertical tail was
704 THE AERONAUTICAL JOURNAL DECEMBER 2003
(a) perspective sketch. (b) Kiting the 1902 Glider.
Figure 10. The Wrights 1902 Glider.
Figure 11. Wilbur in Flight with the 1902 Glider on the north slope of
the Big Kill Devil Hill, 2 October 1902.
included to compensate for the adverse yaw caused by wing warp,
but they may also have anticipated the adverse effects of anhedral
with this design feature. It is interesting to note that the wing camber
and aspect ratio of the new design was not one that they had tested in
their wind tunnel during the winter. The Wrights also changed the
control mechanism for warp and canard pitch. Wing warping was
effected by the operator moving his hips in a cradle in the direction
of turn and the canard was now operated by a twist bar twisting up
for increased pitch.
Wilbur recorded his satisfaction with the first kite tests (see Fig.
10) in a letter, written on 16 September, to his aeronautical col-
league, George Spratt, who had joined the 1901 trials for a period.
We had it out making some tests of its efficiency today and are
very much pleased with the results of our measurements. .In a
test for soaring as a kite the cords stood vertical or a little to the
front on a hill having a slope of only 75, (see Fig. 10(b)). This is
an immense improvement over our last years machine which
would soar only when the slope was 15 to 20, as you will
remember.
They began testing on 19 Sept and made about 50 glides during
the first two days. Wilbur crashed during one glide when he inadver-
tently increased the canard pitch when the glider began rolling due to
a port gust. In a letter to Chanute, written on a rainy Sunday, 21
Sept, Wilbur invited Chanute to join the trial and noted that, I made
a glide of 140ft at an angle of 7 10 on a straight slope whose great-
est inclination was 7 30, exactly facing the wind, at a speed of
about 10 miles over ground, and wind of nine miles, a total of not
over 19 miles. Their prediction that they would be able to glide at 7
appeared to be correct and they could now look forward to many
hours learning to fly. However, during the following Tuesday, when
they made about 75 glides, Orville, on his first free flight, had a
serious crash, surviving ..without a bruise or a scratch. The
problem stemmed from the strong coupling between longitudinal
and lateral motions (incidence and sideslip) when the aircraft began
to yaw, although Wilbur and Orville could not agree on the exact
nature of the problem
(1)
.
After the machine was repaired, they continued testing for another
week but Orville remained uneasy about the control problems they
were continuing to experience. Later, Wilbur would recall the nature
of the outstanding problem in his second lecture to the Society of
Western Engineers on 23 June 1903
(1)
. To quote, It had been
noticed during the day that when a side gust struck the machine its
effect was at first partly counteracted by the vertical tail, but after a
time, when the machine had acquired a lateral motion, the tail made
matters worse instead of better. The vertical tail was suppressing
the adverse yaw but then making things worse. The expression
necessity is the mother of invention comes to mind when making
sense of the Wright brothers next move. On the morning of 3
October, Orville shared with Wilbur his idea of giving the pilot sepa-
rate control of the vertical tail so that the compensation could be
reversed as required. Wilbur was concerned about giving the opera-
tor too much to do and proposed that the movement of the vertical
tail be inter-linked with the wing warp so that when the right roll
was commanded, the tail was displaced into the turn giving a more
coordinated motion and reducing the build up of sideslip. The
problem and solution are illustrated in the sequence in Fig. 12.
Following a side gust, the anhedral wing causes the aircraft to roll
into the gust (a much stronger effect than on the flat winged 1901
machine). Figure 12(a) shows how the fixed tail worked as designed
during a transient disturbance, leading to a proverse, restoring,
yawing moment, following the pilots corrective warp input.
However, if the sideslip builds up, as with a sustained gust or change
of wind direction, then the fixed tail contributed an adverse yaw, due
to the weathercock effect, making the situation worse (Fig. 12(b)).
The operation of the warp-rudder interconnect caused an additional
proverse yaw even when the aircraft slipped out of the direction of
the turn (in Fig. 12(c), the turn is to starboard). The roll-yaw inter-
link effectively assisted turns or corrective rolls. The three-axis
PADFIELD AND LAWRENCE THE BIRTH OF FLIGHT CONTROL AN ENGINEERING ANALYSIS OF THE WRIGHT BROTHERS 1902 GLIDER 705
(a) corrective roll and yaw following side gust.
(c) interlinked wing-warp and rudder suppressing adverse yaw.
(b) prolonged gust leads to increased adverse yaw.
Figure 12. The 1902 glider, (a) vertical tail counteracting wing adverse
yaw, (b) vertical tail giving adverse yaw in sustained lateral gust, (c)
interlinked warp-rudder counteracting effects of sustained gust.
control system would later feature as the basis of the Wright
Brothers patent, first applied for in 1903 and finally granted in the
United States in 1906

.
On 4 October, Orville noted in his diary, we began making
new vertical rudder, one that is operated at same time as end tips.
They began testing the new aircraft on 8 October. Later, in his
second paper to the Society of Western Engineers on 24 June 1903,
Wilbur would record, With this improvement our serious troubles
ended and thereafter we devoted ourselves to the work of gaining
skill by continued practice. When properly applied the means of
control proved to possess a mastery over the forces tending to
disturb the equilibrium. Since balancing was effected by adjust-
ments of the surfaces, instead of by movements of weights, the con-
trolling forces increase in power in the same ratio as the
disturbing forces, when the machine was suddenly struck by a wind
gust. Figure 13 shows four of the collection of photographs taken
during the last two weeks of the 1902 flight trials. In the first
picture, the faint image on the hillside behind the glider may well
be the Chanute-Herring multi-wing machine, with its top surface
removed. It is known that this aircraft was tested in this configura-
tion during this period
(1)
.
During the trial period the Wrights and their colleagues made
measurements of the performance of the glider. They recorded
glide distance and time, the glide angle, the wind speed and ground
speed. From these data, tabulated in Ref. 1, the airspeed can be
estimated. Knowing the weight of the aircraft (115kg with Wilbur
on board and 118kg with Orville on board
(1)
), and assuming a con-
stant sea level density, the lift coefficient can be computed. Figure
14 shows the collected data points plotted on a chart of lift/drag
ratio vs lift coefficient, along with the best (least squares) fit for a
quadratic variation. The scatter in the data is so large, and hence
correlation so weak, that any estimate of an average performance is
likely to be significantly in error. The problem stems from the
single point nature of the data whereas in reality the glide slope
and airspeed almost certainly varied significantly during the glides.
In correspondence between Wilbur Wright and Octave Chanute
during July 1903, their differing opinions on the value of the glide
measurements would aggravate the developing tension between the
two colleagues.
Chanute was confused by the measurements and implied in a
letter, written on 12 July that the Wrights may have over-estimated
the performance of their glider. Wilbur replied robustly, in a letter
dated 14 July, requesting that If you will furnish us data of the
speed at starting, rate of acceleration, maximum speed, rate of
retardation, speed at landing, and a diagram of the path and undu-
lations of the machine, with a mark to indicate the points at which
observations of angle were taken, I think it would be possible to
compute one of these glides though it might require some time.
Wilbur had clearly tried to make the same computations as shown
in Fig. 14 himself and the tone of his response hardly concealed his
frustration with Chanute. Later in the same letter, Wilbur made the
point firmly, The data of ordinary glides are, in my opinion,
almost worthless for purposes of computation unless a possible
error of 50%, or sometimes more, is no serious objection. During
his lecture on Experiments and Observations in Soaring Flight,
given to the Western Society of Engineers on 24 June 1903 and
later published in their journal in December 1903, Wilbur stated
that Observations were almost constantly being made for the pur-
poses of determining the amount and direction of the pressures
upon the sustaining wings; the minimum speed required for
support; the speed and angle of incidence at which the horizontal
resistance became least; and the minimum angle of descent at
which it was possible to glide. To determine any of these points
with exactness was found to be very difficult indeed. With a
total weight of 250lb, Wilbur estimated that a horizontal resistance
706 THE AERONAUTICAL JOURNAL DECEMBER 2003
Figure 13. Wilbur Wright flying the modified 1902 Glider; from the top
10 Oct, 10 Oct, 24 Oct and 10 Oct.

In Ref. 3, Crouch notes that the first application, filed on 23 March 1903, was rejected
on the grounds that it was incapable of performing its intended function!
of the machine was 30lb and remained nearly constant at speeds
between 18 and 25mph. These speeds correspond to a lift/drag of
approximately 833 and C
L
values between 075 and 05, corre-
sponding very roughly to the maximum portion of the fitted curve
in Fig. 14. In his lecture, Wilbur actually remarks on the flights
made on 24 October declaring the point shown on Fig. 14 to be a
good estimate of the glider performance; coincidently, it lies on the
line of best fit.
From Orvilles diary dated Tuesday 21 October 1902, the centre
of gravity of the machine was approximately 18 from front
edge. The centre of gravity of the aircraft plus pilot was probably
close to this point. With its 5ft chord, the aerodynamic centre of
the main wing was therefore ahead of the cg and the aerodynamic
centre of the whole aircraft, the neutral point, was even further in
front. The Wrights had designed an unstable aircraft, but just how
unstable how much negative static margin was present and how
much dynamic instability requires different aerodynamic mea-
surements and analyses than the Wrights were able to handle or
contemplate. Later in this paper, the critical success factors of the
Wright brothers work are reviewed, and the notion that Wilbur
and Orville perceived that control was more important than stabil-
ity is further discussed. As Culick notes in Ref. 5, Only the Wright
brothers recognised that the great problem of control still
remained to be solved.They faced and effectively solved, to the
extent they required, problems of stability and control about all
three axes. Their accomplishments are all the more impressive
when put in the context of the continuing development of aircraft
flight mechanics post 1903; as pointed out by W.H. Phillips in Ref.
9, The entire period from the Wright brothers to 1935, is charac-
terised by a lack of understanding of the relation between stability
theory and flying qualities.
Wilbur and Orville left Kitty Hawk for Dayton on 28 October
1902, just a few days after Orville had shared his euphoria with his
sister, The past five days have been the most satisfactory for
gliding that we have had we are able to take it out in any kind of
weather. The flying practice they had gained would stand them in
good stead and the success of the 1902 glider, from both perfor-
mance and control standpoints, would define a central thread of the
Wright brothers progress for some years to come. To gain a better
understanding of the capabilities and shortcomings of this aircraft,
in the following section the 1902 glider design is appraised using
current day engineering methods including wind-tunnel testing,
simulation modelling and the syntheses and analyses of aircraft
flying qualities.
5.0 MODELLING AND SIMULATION OF THE
WRIGHT GLIDERS
From a flight mechanics perspective, the 1902 glider was rich in
characteristics that have continued to feature, in one form or another,
in aircraft throughout the 100 years of powered flight adverse yaw,
deep stall, unstable pitch motion etc. The aims of the Liverpool
Wright Project are to synthesise and analyse the flight characteristics
of the Wright aircraft from 1900 to 1905 using modern engineering
methods. The primary motivation behind the project is to create
high-fidelity simulations of the Wright aircraft and to make formal
handling qualities assessments of the aircraft including test pilot
evaluations. The process of discovery experienced by the Wright
brothers can be powerfully brought to life through the piloted tests
and the important threads in the design evolution highlighted. The
three elements of the engineering analysis of the Wright Glider are
wind tunnel testing, simulation model development and handling
qualities analysis activities under these three headings applicable
to the 1902 Glider will be summarised in the remainder of the paper.
A full description of the project will be given in the second authors
PhD thesis, due to be published in 2004. Some of the results for the
1902 Glider can also be found in Ref. 10.
5.1 Wind-tunnel testing
The objectives of the tests were to measure force and moment coeffi-
cients over ranges of angle-of-attack, (10 to +24) and sideslip
(30), combined with effects of control surface deflections, to
provide the steady aerodynamic data for the simulation model.
Three Glider models have been constructed, two 1901 gliders at
1/5th scale (44ft span) and one 1902 glider at 1/8th scale (401ft
span). The 1901 glider models featured the original 1/12 camber and
the modified 1/19 camber. Both models featured a variable camber
canard and wing-warping. The 1902 glider model featured wing
warping, an adjustable canard and a single, adjustable vertical tail.
Figure 15 shows the 1902 model mounted in the Manchester
University Goldstein tunnel, where the tests were conducted. The
tunnel is closed return, runs at atmospheric pressure and has a
working section of 9 73ft.
The models were mounted on a T strut connected to an overhead
six degree-of-freedom force and moment balance. The models were
thus tested inverted with the front supported by a vertical nose-wire.
This configuration gave the minimum of interference from the strut
mount with the aerodynamic surfaces and left no attachments along
the wings an important feature to allow for the wing warping.
PADFIELD AND LAWRENCE THE BIRTH OF FLIGHT CONTROL AN ENGINEERING ANALYSIS OF THE WRIGHT BROTHERS 1902 GLIDER 707
Figure 14. Lift/Drag vs C
L
for the 1902 Glider data (from Ref. 1 tables). Figure 15. 1/8th scale 1902 glider model.
The conventions for the various control deflections are as follows:
a) Canard positive leading edge up (pitch up control), +
c
b) Warp positive right wing tip increased incidence, i.e. control
input positive, +
w
gives a roll to the left (i
right
i
left
).
c) Rudder control positive trailing edge to starboard (nose to right),
r
.
The reference datum for the angle of incidence was the chord line
from the leading edge to trailing edge of the wing.
The 1901 glider was tested at speeds between 14-15ms
-1
, giving a
Reynolds number (based on the wing chord, c = 0426m) of Re =
400,000 430,000. This compares with a full scale Reynolds
number at 20 30mph of 128 10
6
21 10
6
(c = 21336m). The
1902 glider was tested at speeds of 175-215ms
-1
giving a Re =
223,000 275,000, while full-scale at 20 30mph gives Re =
910,000 137 10
6
. The 1902 tests were thus conducted at a
quarter of full-scale Re values. Nevertheless, it is considered that
the results are relatively insensitive to Reynolds number over the
range between full scale and model tests. Several sources report the
performance of thin cambered airfoils as fairly insensitive to
Reynolds numbers (e.g. Ref. 11). The primary Reynolds number
effect is a reduction in C
Lmax
of around 10% when reducing the Re
1 10
6
to Re 1 10
5
. The remaining components of the aircraft
such as the struts, wires and attachments are made up from standard
shapes (cylinders, squares) and have known Reynolds number prop-
erties and are either insensitive to the Reynolds number difference
between full-scale and model, or experience a large degree of flow
separation.
Longitudinal Aerodynamics; Despite the foregoing differences,
Fig. 16 shows remarkable agreement between the lift coefficient
results from the current tests and those documented by the Wrights
in Ref. 1 for single aerofoil No 12. Also shown on Fig. 17 are results
for the biplane combination of aerofoils Nos 22 and 23, which were
very similar to the No 12 aerofoil. The Wright wind tunnel was run
at about 25mph (112ms
-1
)
(1)
. The biplane model (aerofoils 22 and 23
superimposed) had a chord of 066in (0017m) giving a Re of only
about 13,000. The thin sections used on both model and full scale
reduce the flow differences due to Re effects but as pointed out by
Anderson in Ref. 8, at low Res, below 100,000, the lift curve slope
reduces to about 70% of full scale, as suggested in Fig. 16. Wilbur
had noticed this loss of efficiency with superimposed surfaces. In a
letter to Chanute on 15 Dec 1901, he points out that
(1)
that his data
...will tend to establish a general law that the lift and tangentials of
a set of superimposed or following surfaces spaced about their
length apart are approximately equal to those of a single surface of
similar profile or curvature having a breadth equal to that of one
surface and a length equal to the sum of their lengths. Wilbur refers
to the chord as the length and the span as the breadth and is noting
an equivalence between superposition and aspect ratio, as far as lift
curve slope is concerned.
The Lawrence data in Fig. 16 shows a zero-lift angle-of-attack of
about 25 deg and a lift curve slope of about 38/rad. C
L
peaks at
about 085 and remains at this level through to the measurement
range limit of 25. Flow visualisation reveals a large leading-edge
separation bubble over the upper surface at high angles-of-attack,
preserving the low pressure and inhibiting the lift loss at stall. The
drag increases however and Fig. 17 shows comparisons of the glider
efficiency the ratio of lift to drag, plotted against the lift coeffi-
cient. In Fig. 17, the drag coefficients for the struts, wires and sup-
porting structure have been corrected for known Re effects.
Included on Fig. 17 is the trend line from the Wrights flight tests
and the 24 Oct data point referred to by Wilbur in his 1903 lecture to
the Society of Western Engineers. The Wrights would have realised
from their flight tests that the performance was not as good as the
wind tunnel data suggested. Nevertheless they had improved signifi-
cantly on their 1901 glider as illustrated in Fig. 18 where the lift/drag
ratio of the two glider models are compared to show a 60% increase
in maximum efficiency.
The pitching moment characteristics are shown in Fig. 19 (as
function of incidence for three canard settings) and 20 (as function
708 THE AERONAUTICAL JOURNAL DECEMBER 2003
Figure 16. Lift coefficient vs Incidence.
Figure 17. Lift/Drag ratio of 1902 Glider as a function of C
L
.
Figure 18. Comparison of the Lift/Drag ratio of 1901 () and 1902
(L)Gliders.
Figure 19. Pitching moment coefficient about centre of gravity
(035c) as a function of incidence.
of C
L
for three cg locations). Strong non-linearity is clearly visible
with a reversal of the overall slope at higher incidences (for >
10deg for zero canard). In Fig. 19 the canard can be seen to give a
positive, albeit weak, increase in moment up to about +6 with
+10
c
. Any further increase in a results in total loss of nose up
control power. The same is seen for a 10
c
where at 5 there a
loss of nose down control. In both cases the canard is stalling at inci-
dences of around 15. The slope of the curve at low in Fig. 20 is
equivalent to the static margin, H
n
. A value of 16% is measured for
the working range of C
L
between 02 and 06. This value corresponds
to a cg location of 035c, and is probably more unstable than when
the Wrights were flying. Wilbur describes the action of the 1902
glider, The action of the machine is almost perfect, suggesting
an easily controlled vehicle. Shown in Fig. 20 is the effect of shifting
the cg in the 1902 glider. The cg of the 1902 glider without pilot is
quoted as being approximately 18 inches from the wing leading
edge
(1)
. This is at 30% chord, and when the pilot is included this is
likely to shift slightly forward. Taking the gliders weight of 116lb
and the pilots (Orville) weight of 140lb the resultant cg is calculated
to be approximately at the 24% chord position. At this cg position,
the static margin is slightly unstable, H
n
= 6.6%, for C
L
between 02
and 06. The shifting centre of pressure with incidence was noted by
the Wrights and reflects a corresponding variation in stability.
Another major feature of the Figs 19 and 20 is the reversal of the
slope, indicating a large change in the pitching moment, at high inci-
dence. This effect is due to separated flow near the leading edge of
the airfoil. At high incidence, the wing continues to create lift, but a
change in the pressure distribution moves the centre of pressure aft
causing a large nose-down pitching moment. Figures 21 and 22
show two flow visualised images. The wing in Fig. 21 is at an angle-
of-attack where the flow is smooth and attached to both surfaces,
representing the range 2 8. Figure 22 shows the model at high
a when the flow separates and a re-circulation bubble on the upper
surface forms. The consequences are a low but sustained C
Lmax
and a
large C
M
break. Both of these characteristics are typical of very thin,
cambered airfoils. The Wrights also detected this phenomenon when
testing a single surface of their 1902 glider Found that surface
tended to duck at large angle, but on increase of wind and decrease
of angle of incidence centre of pressure seemed to move forward,
and pitching ceased.
(1)
, The consequence of this effect is that if the
pilot were to attain a high flight condition, by pitching nose-up and
losing airspeed, there would a restoring nose-down pitching
moment. Also, Fig. 19 displays a second crossing of the x-axis (C
M
=
0) at a high incidence of around 15-16. This zero moment trim
point, combined with the C
Lmax
which is preserved up to high angles-
of-attack, represents a flight condition that the Wrights would often
find themselves in, where the aircraft would lose airspeed and
descend in a flat deep-stall. Many of the longitudinal problems
with the 1901 and 1902 Gliders can be explained by examination of
the detailed features in Fig. 19; such information was, of course, not
available to Wilbur and Orville.
Lateral-Directional Aerodynamics; Figs 23-26 illustrate the lateral-
directional aerodynamic characteristics of the 1902 glider wind-
tunnel model. Figure 23 shows the yawing moment coefficient with
sideslip at three rudder settings; the aircraft is directionally stable.
The control sensitivity is approximately constant up to the
stall angle of the tail surface, whereupon it falls off rapidly. The
rolling moment is shown in Fig. 24; the anhedral angle of the wings
gives a positive value of .
The effect of wing-warping is illustrated in Figs 25 and 26. Figure
25 shows the rolling moment coefficient plotted against the angle of
attack with curves for warp angles,
w
of +14, 14, and 0. The warp
control is effective at creating rolling moments at angles of inci-
dence between 5 and +10. Beyond these angles the roll control
sensitivity falls off, and by = 15, 0. Figure 26 presents the
yawing moments created when the warp control is operated (without
interlinked rudder). Large adverse yaw moments are generated,
PADFIELD AND LAWRENCE THE BIRTH OF FLIGHT CONTROL AN ENGINEERING ANALYSIS OF THE WRIGHT BROTHERS 1902 GLIDER 709
Figure 20. Pitching moment coefficient about centre of gravity for
different cg locations.
Figure 22. Smoke flow visualisation 20.
Figure 21. Smoke flow visualisation 6.
n
r
C

l
C

l
w
C

710 THE AERONAUTICAL JOURNAL DECEMBER 2003

Figure 23. Yawing moment coefficient as a function of sideslip.
Figure 24. Rolling moment coefficient as a function of sideslip.
Figure 26. Yawing moment coefficient as a function of incidence.
Figure 27. Canard angle to trim as a function of speed.
Figure 25. Rolling moment coefficient as a function of incidence. Figure 28. Drag and power as a function of speed.
becoming increasingly worse with increasing incidence. For the
range of 5 to +10, the adverse yawing moments are less severe for
the maximum warp of 14. Comparing the yawing moments avail-
able from the rudder and the warp controls, it can be seen that a C
n
of
0004 for 10
r
is achieved, and the maximum C
n
of 0005
for 14
w
is not achieved until = 10. This indicates that the warp-
to-rudder interlink system devised by the Wrights was very effective
at cancelling the adverse yaw moments. In Ref. 12, Jex and Culick
refer to a gearing ratio of
r
= 125
w
. Comparing the control
derivatives of = 00004deg
-1
and = 000036 deg
-1
(worst case below = 10), it can be seen that small proverse
yawing moments are created by hip cradle motion.
5.2 Simulation model construction
The Wright aircraft simulation models are created using the
FLIGHTLAB modelling and simulation software environment
(13)
.
The software was developed to address the particular demands of
high-fidelity simulation of rotorcraft and uses a multi-body dynam-
ics approach to model flight vehicles. It provides a range of tools
using a modular approach to assist in rapid generation and analysis
of complex, non-linear models. FLIGHTLAB has a number of high-
level graphical user interfaces to aid the generation of models
GSCOPE, FLIGHTLAB model editor (FLME) and XAnalysis.
GSCOPE and FLME enable the construction of models through the
manipulation of icons or model trees. The process builds scripts
(code) using FLIGHTLABs interpretative language SCOPE. Not all
the features required to model the Wright aircraft were available in
the component library. Consequently, for the Wright models, a
library of novel user-defined components and model scripts were
developed to implement the features required. XAnalysis offers a
suite of functions that enable the user to conduct control system syn-
thesis and overall vehicle analyses on the simulation model. Within
XAnalysis, the user has the ability to set the desired test conditions,
trim the aircraft and perform dynamic response, stability, perfor-
mance and handling qualities analyses. Finally the real-time operat-
ing system, Pilotstation, links with FLIGHTLAB to enable piloted
simulation.
The wind tunnel tests provided a new set of data describing the
(static) aerodynamic characteristics of the aircraft as a whole. In
order to incorporate this data a multiple lookup table, super-compo-
nent, was developed. This component calculates the total aerody-
namic loads for the canard, wings, vertical surfaces, airframe and
pilot in one system. The aero-tables are generally three dimensional,
with the coefficients a function of , , and
control
. The super com-
ponent is then attached to the aircraft model at a predefined airload
reference point. The structure of the aircraft is modelled as a rigid
body with a mass placed at the aircraft centre-of-gravity and
moments of inertia distributed about the body axes. No published
inertial data were available for the 1902 glider hence a spreadsheet
method was developed, based on Ref. 14, to estimate the moments
of inertia. A simple landing skid model based on the standard
FLIGHTLAB two-strut undercarriage model was included to enable
the glider to be landed. The spring, damping and friction coefficients
were selected to emulate the behaviour of landing skids on a sandy
surface.
Aerodynamic damping effects are included in derivative form for
pitch, roll and yaw based on simple linear, two-dimensional
theory
(15)
.
5.3 Handling qualities analysis
Handling Qualities analysis addresses the requirements for trim, sta-
bility and response and the consequent pilots impressions of an air-
crafts suitability for its role, defined by a collection of mission task
elements. Using the FLIGHTLAB trim analysis tool, the steady per-
formance characteristics of the 1902 glider can be analysed. Figure 27
shows the canard deflection angle required for trim across a speed
range for three locations of the centre of gravity. It can be seen that as
the cg moves forwards from the aft position (035c), the slope
changes from positive to slightly negative at a speed of 24kt. The
minimum speed that the model would trim was about 16kt; below
that there was insufficient control power to trim. For the unstable
configurations, aft stick was required to trim as the speed increased.
Figure 28 shows the variations in drag and power (drag speed) with
forward speed. The minimum drag speed occurs at about 20kt and the
minimum power required for flight is 18 horsepower at a speed of
about 185kt. The speed at minimum drag corresponds to the
minimum glide angle (maximum C
L
/C
D
), whereas the minimum
power speed represents the condition for the minimum sink rate.
The most common flight speed that Wrights glided at was approx-
imately 24kt, and this case has been selected for further analysis
using open and closed loop stability analysis. From the non-linear
FLIGHTLAB model a linear model can be derived, valid for small
perturbations from the trim flight condition. Table 1 gives the con-
figuration data for the trim case.
The system and control matrices for the aircraft at the 24kt flight
condition are given below in semi-normalised form (e.g. X
u
(1/sec),
M
w
(rad/ft.sec), M
c
(1/sec
2
)). The perturbation velocities along the
body-fixed X, Y and Z axes are u, v and w; the angular rates are p, q
and r and the Euler angles are , and . The system equations are
in the general form;
Longitudinal A matrix ; x = [u w q ]
T
, cg @ 035c
Longitudinal A matrix, cg @ 024c
Note the positive M
w
derivative (element 3,2), indicating the static
instability of the aircraft (particularly strong in the aft cg configura-
tion) and the relatively large heave and pitch dampings (Z
w
and M
q
).
Longitudinal B matrix, u = [
c
], for cg at .35c and .24c respectively.
PADFIELD AND LAWRENCE THE BIRTH OF FLIGHT CONTROL AN ENGINEERING ANALYSIS OF THE WRIGHT BROTHERS 1902 GLIDER 711
n
r
C

n
w
C

. . . (2)
+ x Ax Bu
0 2158 0 7225 2 7944 32 0204
1 0274 8 1751 44 9362 3 3716
0 0643 0 9543 3 5995 0
0 0 1 0
, ]
, ]

, ]

, ]
, ]
]
Long
A
0 1742 0 6877 2 5623 32 0961
1 1065 7 6488 43 0690 2 6753
0 0074 0 1191 3 2926 0
0 0 1 0
, ]
, ]

, ]

, ]
, ]
]
Long
A
9 6691
18 2279
18 5695
0
, ]
, ]

, ]

, ]
, ]
]
Long
B
2 1548
2 5659
=
12 6496
0
, ]
, ]

, ]
, ]
, ]
]
Long
B
,
Table 1
Configuration data for 1902 Glider analysis
Trim speed 24kt (405ms
-1
)
Weight 252lbf
I
XX
= 228 slugs-ft
2
I
YY
= 48 slugs-ft
2
I
ZZ
= 229 slugs-ft
2
Span b = 32ft
Chord c = 5ft
Total wing area = 305ft
2
Lateral A matrix, x = [v p r ]
T
cg @ 035c
Lateral-Directional B matrix, u = [
wr
]; u=[
w

r
]
T
,
Note the high value of roll damping L
p
, the adverse yaw effects
(N
p
and N
w
), negative dihedral (positive L
v
) and relatively weak
weathercock stability, N
v
.
For the longitudinal motions, the B matrix represents the effects of
the canard control,
c
. For lateral-directional motions, B
wr
repre-
sents the control with the warp-rudder interconnect system, while
B
lat
breaks down the separate effects of warp,
w
, and rudder,
r
.
The eigenvalues for the longitudinal and lateral-directional system
matrices are:
Longitudinal eigenvalues (cg @ 035c)
127758
19558
05852 + 12865i
05852 12865i
Longitudinal eigenvalues (cg @ 024c)
85371
20827
08833
03875
Lateral-directional eigenvalues (cg @ 035c)
143685
09226 + 14639i
09226 14639i
01168
Both longitudinal and lateral-directional motions are unstable. For
the aft cg case, the pitch mode with = 19558/sec, is particularly
unstable, with a time to double amplitude T
D
= 035 seconds. The
spiral mode, = 01168/sec, has a T
D
= 59 seconds. In the following
discussion, the flying qualities of the 1902 Glider are referred to
modern day criteria set in standards such as Refs 16 and 17, and
referred to flying qualities Levels as originally defined by Cooper
and Harper
(18)
.
Longitudinal Dynamics
Referring to the short period criteria within MIL-F-8785C
(16)
, with
the cg at 035c the dynamics lie outside the Level three boundary.
Moving the cg forward to 024c reduces the instability ( = 03875,
T
D
= 179s) but the aircraft remains unstable. This confers Level 3
flying qualities for the aircraft. The source of the problem for the
Wrights was, of course, the positive pitching moment change with
incidence, already referred to in Fig. 19, and the requirement for aft
stick to trim as the speed increased in Fig. 27. The derivative M
w
at
both cg locations is positive, reflecting the negative static margin but
it is interesting to examine the manoeuvre margin and the short
period approximation for these cases. The approximation for the
short period eigenvalues ()

, assuming no speed changes during this

rapid incidence adjustment mode, is given by;
The term is proportional to the manoeuvre margin
and is negative (-134) with the cg @ 035 and positive (+20) with the cg
@ 024. A comparison of the exact and approximate roots is given in
Table 2. The large stable root is captured well in both cases but the
unstable mode is only captured in the aft cg configuration and then with
a nearly 50% error. The forward speed change in this mode is clearly
significant. In the forward cg configuration, the approximation picks up
the stable pitch mode; as the cg is moved further forward these two real
roots combine to form the true short period mode and the approximation
given in Table 2 then captures both frequency and damping very well.
The approximation for the phugiod mode, assuming a motion
made up of essentially vertical and horizontal motions
(19)
is given by
the expression;
A comparison of exact and approximate eigenvalues is given in
Table 3. Although an oscillation is predicted for the aft cg configura-
tion, the frequency and damping are more than 50% in error. For the
forward cg configuration the pitch instability is now predicted,
although with a 30% error. The approximations given by Tables 2
and 3 rely on a weak coupling between the motions and a wide sepa-
ration between the modes. For conventional aircraft they tend to give
good approximations and provide insight into the derivatives that
contribute to stability and instability. The 1902 glider was unconven-
tional and the success of the approximations is much more limited.
An unstable aircraft is not necessarily un-flyable of course. Acting
as sensor and actuator elements in the feedback loop, a pilot can
modify the closed loop dynamics in a wide variety of ways. Figure
29 shows a schematic describing the aircraft-pilot system. The
pilots action is approximated as a proportional controller (in a
712 THE AERONAUTICAL JOURNAL DECEMBER 2003
. . . (3)
( )
2
( ) 0
w q w q w q e
Z M Z M M Z U + + +
Table 3
Comparison of Exact and
Approximate Phugoid eigenvalues
cg @ 035c cg @ 024c
Exact Approx Exact Approx
Real 05852 02762 08833 06346
Imag. 12865i 19706i 0i 0i
Real 03875 05052
Imag. 0i 0i
Table 2
Comparison of Exact and Approximate
Short Period eigenvalues
cg @ 035c cg @ 024c
Exact Approx Exact Approx

1
127588 -128239 85371 86129

2
19558 10493 20827 23285

Note that with real eigenvalues there is strictly no short period mode but the approxi-
mation can still work for aperiodic modes when the constant speed assumption holds.
0 3460 1 8597 40 4611 32 0208
0 0441 14 7847 3 2618 0
0 0625 1 6653 0 9663 0
0 1 0 1053 0
, ]
, ]

, ]

, ]
, ]

]
Lat
A
4 470
13 8699
2 2536
0

, ]
, ]

, ]

, ]
, ]
]
wr
B
0 4 5321
13 8699 0
0 4796 2 0197
0 0
, ]
, ]

, ]

, ]
, ]
]
Lat
B
( )
w q w q e
Z M M Z U +
( ) ( ) ( ) ( )( )
( )
( ) ( )
( )
2
gCos /
gCos
0
u
w e e u q u q e q e w u w u
q w w q e
w u q u q e
e
u
e q w w q e
X
X U Z M M Z U X W Z M M Z
M Z M Z U
Z Z M M Z U
Z
U M Z M Z U
+
+


+ + +


+



+



+


. . . (4)
PADFIELD AND LAWRENCE THE BIRTH OF FLIGHT CONTROL - AN ENGINEERING ANALYSIS OF THE WRIGHT BROTHERS' 1902 GLIDER 713
Figure 29. Closed-loop system with pilot model.
Figure 30. Time response comparison of open and
closed loop control.
Figure 31. Longitudinal root loci for varying K

; cg @ 035c. Figure 34. Lateral-directional root loci for varying K

.
Figure 32. Longitudinal root loci for varying K

; cg @ 024c.
Figure 33. Root loci showing effect of cg location
on the longitudinal modes.
similar manner to the Jex and Culick studies of the Wright 1903
Flyer in Refs 12 and 20). Using airframe and outside world refer-
ences such as the canard and horizon, the pilot can deduce the error
between the desired pitch attitude and the current sensed attitude and
enter a proportional amount of control deflection to reduce the error.
This classical pilot model can be used to predict the dynamics of the
aircraft when under the simplest form of pilot control with constant
gain K
p
= K

. It also provides insight into how the Wright brothers

initially experienced the flying characteristics of their Glider.
In Fig. 29, the open loop transfer function is shown as the ratio
of the finite zeros polynomial and the open loop transfer function
containing the short period and phugoid modes.
Figure 30 shows a comparison of the response of the 1902 glider
under open loop and closed loop control when disturbed by a 05
second control pulse. The unstable mode causes the open loop aircraft
to diverge rapidly in pitch, reaching = 20 after only 25 seconds. In
the closed loop response, the pilot gain, K

was set to four(deg

canard/deg pitch attitude error). The closed loop response is stable
with the aircraft motion damping out after about three oscillations.
The effect of closing the loop can also be investigated through the
root loci for the system, showing how the various modes change as
the gain is increased. Figs 31 and 32 show the longitudinal root loci
for the two different cg locations.
In Fig. 31 the application of pilot feedback is seen to stabilize the
unstable mode at high gain. With a gain of K

= 1, the pitch mode is

still unstable and the oscillatory phugoid mode has been driven to a
condition of neutral stability. This loss of stability in the oscillatory
mode is a consequence of the condition that pitch attitude feedback
cannot change the overall damping of the system hence if in one
mode the damping increases, in another the damping must decrease.
If the gain is further increased the unstable aperiodic mode becomes
marginally stable and the oscillatory mode moves away from the
right hand side but with increased frequency.
As discussed earlier the cg location was probably nearer 024c
when the Wrights flew their 1902 Glider, and the root loci for this
condition is shown in Fig. 32. In the open loop configuration, we see
four aperiodic modes. When closed loop control is applied, the diver-
gent mode is stabilised with a gain of K

<1. Two of the aperiodic

modes then join to form an oscillatory mode similar to the cg = 035c
case.
The effect of the movement of the centre of gravity is illustrated
in Fig. 33, showing how, as the cg is moved forward, the unstable
mode moves from the right hand side to the left meeting another
pole that was formed by the splitting of the quasi phugoid mode.
This new mode forms a low frequency, weakly damped mode the
classical phugoid. The remaining pole splits and travels left to form
another oscillatory mode of higher frequency and high damping
the classical short period mode.
The moving of the cg forward clearly makes the 1902 glider more
stable so a question that arises is why did the Wrights persist with
the unstable aft cg? Part of the reason was that the configuration of
the aircraft forced the cg to be aft and the pilot could only practically
get his body so far forward within the aircraft. Secondly, if the cg is
too far forwards then the canard has to have a significant upload to
trim; for the 02c cg location,
c
= 10 at 24kt. With the estimated
canard stall angle of 12-15 there is little control margin left for
pitch up control, and at the lower speeds the pilot would not have
adequate control power to achieve a trim.
In summary, the 1902 glider was likely to have been marginally
unstable as flown by Wilbur and Orvillle, with ample control power,
pitch up and down. However, the conditions flown in would have
been gusty, resulting in the pilot needing to make continuous control
corrections to keep a steady flight path. In these conditions the air-
craft would undulate in a lightly damped oscillation. Furthermore,
the frequencies (3-7 rad/s, 047-111Hz) of the closed-loop oscilla-
tory mode, corresponding to proportional gains of between 1 and 4
deg/deg, indicate a susceptibility to PIOs (Pilot Induced
Oscillations), unless a degree of anticipation could be applied. The
skill to apply this anticipation would come with practice and the
Wright brothers had plenty of opportunity to develop such skills
during the autumn of 1902.
Lateral directional dynamics
The 1902 glider also featured the classical set of lateral-directional
modes an unstable spiral mode, (T
D
= 59 secs) a moderately
damped Dutch roll oscillation and heavily damped roll subsidence.
At 24kt, the 1902 glider roll mode
r
=-143685 has a time constant,
T
r
= 007 seconds, well within the Level 1 HQ boundary (and match-
ing the classical approximation, L
p
= 1479 well). The Dutch roll is
a coupled yaw/sideslip/roll oscillation, with a mix determined by the
characteristics of the rolling and yawing moments with sideslip and
body rates. For the 1902 glider, the Dutch roll mode is stable ( =
092 146i). MIL-F-8785C requires for Level 1 that
d
> 019,
d
> 1.0 and
d

d
>035. Comparing the 1902 glider values, these are

d
= 054,
d
= 173 and
d

d
= 092. Hence Level 1 handling qual-
ities are met for this mode. A good approximation to the Dutch roll
mode can be derived by assuming the motion is a coupled yaw-roll
motion with little sideways motion (sideslip is derived only from
714 THE AERONAUTICAL JOURNAL DECEMBER 2003
Figure 35. Lateral-directional root loci for roll control with warp-rudder
interlink disconnected.
Figure 36. Time response roll command.
c
Y

yaw and roll Ref. 21). Then we can write;

dapprox
= 0973 (exact = 092)

dapprox
= 1523 (exact = 173)
The cross damping derivatives N
p
and L
r
determine the ratios of roll
and yaw in the Dutch roll. In the 1902 Glider, both are strong effects,
N
p
being even stronger than the damping N
r
and, of course, adverse.
A certain degree of spiral instability can be accepted. For Level 3
flying qualities, the spiral mode time constant, T
s
must be greater
than 72 secs. For the 1902 glider this parameter (defined T
s
= 1/s)
is calculated to be 86 seconds, just inside the Level 3 boundary but
outside Level 2, defined at T
s
> 115s. The unstable behaviour is
partly a result of the wing anhedral. It was known to aviators at the
time that a dihedral angle gave stability and this was favoured by
model aircraft makers for obvious reasons, but the Wrights had
deliberately designed their aircraft with anhedral. They were uncom-
fortable with the roll response induced by side gusts with dihedral,
which tended to lift the wing on the side of the gust and drive the
opposite wingtip toward the ground. The unstable response tended to
cancel the sideslip induced by the gust by rolling the aircraft into the
wind. The pilot would apply warp to lift a dropped right wing. The
warp induced yaw would pull the starboard wing back, inducing left
sideslip which would further decrease the roll angle. The classical
approximation to the damping of the spiral mode is given by the
expression;
where
giving,
s
(approx) = +024, compared with +012 for the exact
value, a rather poor comparison. It can be shown that the
approximation is dominated by the L
r
effect;
which itself causes the pilot to hold significant out-of-turn warp
control during a turn.
The lateral problems had not completely disappeared with the addi-
tion of anhedral, and, as described in the earlier section, the Wrights
linked the wing warp to the vertical tail. This vertical tail gave the
1902 glider a degree of directional stability which, after being dis-
turbed by a short duration side gust, stabilised the heading of the air-
craft by creating a restoring, proverse, yawing moment. In a prolonged
gust or a change of wind however, the aircraft would roll and yaw
toward the gust because of the directional stability provided by the
vertical tail. The pilot, trying to maintain heading towards some fixed
point, would naturally try to keep the wings level, warping the wings
to correct. The warp produced an adverse yaw bringing the low
wingtip further around toward the wind and causing a yaw rate
induced roll (positive L
r
) further increasing the roll angle. This behav-
iour was dubbed well digging by the Wright brothers one wingtip
would drive into the sand and dig in with the aircraft corkscrewing
around this point.
Well-digging was the result of a rather complex situation fea-
turing closed loop control by the pilot and the rapid loss of air-
speed with the subsequent angle of incidence changes as the
aircraft acquired large sideslip velocities. As discussed earlier
in the paper, the Wrights struggled to understand what was hap-
pening. Orville notes in his diary that he and Wilbur disagreed
about the source of the problem
(1)
I now became thoroughly
convinced that the trouble in the fore and aft control was a
result of one wing getting higher than the other, Will main-
tained the trouble was just the reverse, and that the lateral
tipping of the machine was the result of the loss of forward
motion... Figures 25 and 26 show us that the source of the
problem was the loss of roll control combined with the signifi-
cant increase in yaw response to warp at high incidence. So, in a
sense, both Orville and Wilbur were correct and they found the
solution together. The fix was to replace the double vertical
surface with a controllable single surface tail that moved with
the wing-warp. The effect of this was twofold. The first was
that the smaller tail area reduced the directional stability, which
reduced the tendency of the aircraft to swing into the wind, and
secondly, when the wing was warped to counter the into-wind
roll, the rudder created a proverse yawing moment cancelling
the warp adverse yaw (see Fig. 12). Unfortunately, the large
rolling moment due to yaw rate meant that during a turn, as
already mentioned, the pilot had to hold out-of-turn lateral
control, which in turn gave an out-of-turn rudder input; the air-
craft then side-slipped into the turn and the Wrights could never
fly properly co-ordinated turns.
The analysis carried out for the longitudinal motion can also be
applied to the lateral axis, either feeding back the roll attitude, ,
or the heading angle, to the wing-warping control. Of course,
the 1902 glider had no independent yaw control, the only yaw
control coming via the warp-rudder interconnect system. The
two-axis control without interconnect system is also investigated.
Figure 34 shows the root loci for bank angle control with the
interlinked warp and rudder control,
wr
. The effect of increasing
the gain is to stabilise the spiral mode. The gain increase has little
effect on the Dutch Roll mode; the numerator in the roll-warp
transfer function features a second order zero with similar
dynamics to the open loop pole. Increasing the gain to K

= 4, a
second damped oscillatory (combined roll-spiral) mode forms.
This insensitivity of the Dutch roll to warp control can be consid-
ered as a design feature of the aircraft, although the Wrights
never interpreted it in quite this way of course. Another positive
feature of the anhredral wing is that the effective directional sta-
bility in closed-loop control
(19)
is increased when
adverse yaw is present.
Figure 35 shows the roll loci with warp-to-rudder interconnect
disconnected. There is little difference compared with Fig. 34.
Time responses for the feedback gain of K

= 1 are presented in
Fig. 36. Results for the interconnect system operating and discon-
nected are compared. In contrast with the stability results in the
root loci, the time responses show significant differences in yaw
and sideslip response. The response with interconnect-off
exhibits an untidy turn entry with large amounts of adverse yaw
and sideslip, whereas with interconnect-on the turn entry is
smooth and more controlled.
The previous analyses have shown that the 1902 glider was
unstable in pitch and roll-yaw. For the aft cg configuration, the
pitch stability is predicted to be outside of the Level three bound-
ary, based on contemporary flying qualities theory. For the
Lateral axis, Level one criteria are met for the roll subsidence and
Dutch roll modes. The spiral mode is unstable, falling into the
Level three range. The rudder-to-warp interconnect has no effect
on open-loop stability of course, and only a minor effect on
closed loop stability, but is seen to improve the turn entry charac-
teristics when using the simple warp-to-roll pilot model.
PADFIELD AND LAWRENCE THE BIRTH OF FLIGHT CONTROL - AN ENGINEERING ANALYSIS OF THE WRIGHT BROTHERS' 1902 GLIDER 715
. . . (5)
2
1
v r
r v d
p
d d
d r
p
L L
N Y
V L
L
L V
j \ , ]
+ +
, ] , (
, ] ( ,
]

j \

, (
( ,
. . . (6)
( )
2
1
v d v
d
d r
p
VN L
L
L V
+

j \

, (
( ,
. . . (7)
p
d
p
g N V
L

. . . (8)
g
v r v r
s
p v s v
L N N L
L VN L
j \

, (
+ ( ,
. . . (9)
s d

,
v r v r v s v
N L L N VN L >> >>
veffective v v
N
N N L
L

5.4 Exploratory piloted simulation tests
The dynamic analysis presented in the previous sections has shed
some light on the flying characteristics of the 1902 glider, explaining
many of the problems experienced by the Wrights. However, the
assumptions concerning linear aerodynamics, decoupled longitudinal
and lateral control and the simple pilot model, preclude one from
answering some important questions how did the lateral and longi-
tudinal axes couple? What is the manoeuvre flight envelope and gen-
erally how well can a pilot fly the 1902 glider? How prone is the
aircraft to pilot-induced-oscillations? To answer these questions,
piloted trials using the full non-linear FLIGHTLAB simulation
model have been carried out at the University of Liverpools flight
simulation laboratory, using a single seat cockpit on a six-axes
motion system. The project test pilot, experienced in flying vintage
aircraft, flew a set of mission task elements (MTEs) in exploratory
trials. MTEs are manoeuvres that are considered critical to the
mission performance of an aircraft and are selected to qualify the
handling qualities throughout the operational flight envelope

. MTEs
are derived as tasks from a typical mission. The 1902 glider was an
aircraft developed as a test bed to refine the flight control system and
allow the Wrights to gain flying experience. Hence, the MTEs could
be relatively simple (see Refs 10 and 22 for more details).
To enable sustained flight, the glider was operated in a continuous
thermal of strengths varying between 450-600ft/min, allowing the
aircraft to be flown straight and level and also to climb. The initial
flight by the test pilot highlighted many of the characteristics dis-
cussed in the previous section. The pilot noted that the aircraft,
although pitch unstable, became stable at high angle of attack. He
also commented on the spiral instability, the adverse warp-yaw and
found particularly control difficult if the sideslip were allowed to
build up beyond about 10. The piloted tests also confirmed many of
the predictions of the closed-loop analysis, including the undulating
oscillations (weakly damped pitch/heave mode) typical of many of
the Wright canard aircraft. Lateral control was effective and as long
as bank angles were kept below about 15, successful turns could be
made. However, if the airspeed were allowed to reduce below about
20kts (minimum drag speed) then the lateral control became sluggish
and the aircraft would often sidle off, building up high, and eventu-
ally uncontrollable, sideslip angles. The handling qualities ratings
awarded for the different (simple) tasks were in the 4-7 range, sur-
prising good considering the aircraft being studied. Undoubtedly, the
aircraft required skill to fly but offered unprecedented levels of
control and manoeuvrability for the time. Results from these trials
will be reported in detail in future publications and initial findings
are discussed in Refs 10 and 22.
6.0 BRIEF REVIEW OF THE WRIGHT
BROTHERS WORK POST-1902
This paper concerns the birth of flight control as conceived by the
Wright brothers in the design of their 1902 Glider. The story would
be incomplete, however, without a brief mention of how this was
taken forward in the powered aircraft designs. In his lecture to the
Society of Western Engineers on 24 June 1903, Wilbur remarked
that .. at the slower speed (18mph), 166lb were sustained for each
horsepower consumed; at the higher speed (25mph), 125 pounds per
horsepower. Discussing power requirements, the Wrights were well
advanced on the next stage of their aeronautical progress at this time.
Late in 1902, after a lengthy dialogue on the data from the 1902
tests, Chanute enquired of Wilbur, How far do you think of carrying
on aeronautical work? Wilbur relied, It is our intention next year
to build a machine much larger and about twice as heavy as our
present machine. With it we will work out problems relating to
716 THE AERONAUTICAL JOURNAL DECEMBER 2003
Figure 37. Testing the two-surface rudder on the 1902 glider in
October 1903.
Figure 38. The Wright Flyer crashes on take-off Wilbur at the
controls on 14 December 1903.
Figure 39. The First Flight of the Wright Flyer Orville at the controls
on 17 December 1903.

MTEs were originally introduced in the mid-1980s as part of the helicopter handling
qualities engineering discipline
(21)
but have since seen more general use by the fixed-
wing community.
starting and handling heavy weight machines, and if we find it under
satisfactory control in flight, we will proceed to mount a motor.
Their thinking quickly developed over the next few months and,
with no suitable powerplant to be found on the market, they pro-
ceeded to design and build their own complete drive system
engine, transmission and propellers. The Wrights designed and con-
structed the components of their 4th piloted flying machine, the
Flyer (see Ref. 23 for a comprehensive technical review of this air-
craft), and they journeyed to Kitty Hawk in September 1903 to
assemble and test the aircraft. They spent the first month further
developing their flying skills and refining the control system on the
1902 Glider; Orville recorded in his diary on 26 October, We suc-
ceeded in breaking our former record six times out of about 20
attempts
(1)
(see Fig. 37).
Their first attempts to run the new engine mounted on the Flyer in
early November failed due to excessive vibration. Then, on two
occasions, propeller shaft failures delayed their progress further. The
Wrights were very concerned that the weight had increased to over
700lb the thrust required might be 100lb, and we got to doubting
whether the engine would have the power, using the gears we have,
to give the necessary thrust
(1)
. Finally the machine was ready for its
first flight on 14 December. With Wilbur at the controls, the
machine turned up in front and rose to a height of about 15 feet from
ground After losing most of its headway it gradually sank to
ground turned up at an angle of probably 20 incidence.Time of
flight from end of track was 3 seconds, for a distance of 105ft.
(1)
.
On its first flight, The Flyer had crashed on take off (Fig. 38) and
Wilbur stated in a letter to his father on the same day,
the real trouble was an error in judgement, in turning up too
suddenly after leaving the track, and as the machine had barely
speed enough for support already, this slowed in down so much
that before I could correct the error, the machine began to come
down, though turned up at a big angle. Characteristically,
Wilbur was first to admit his mistake and it took two days to repair
the damage. They took the machine out again on the morning of
17 December.
Figure 39 captures take-off on the first of four flights made on 17
December Orville flew for about 12secs, covering a ground distance
of 120ft. On the fourth flight Wilbur flew for 59 seconds covering a
distance of 852ft
(1)
.
The four flights on 17 December 1903 gave the brothers a very
brief glimpse of the future. It would take another nine months before
the Wrights were able to fly a circuit of the Huffman field near
Dayton, Ohio and a further twelve months before they were satisfied
that they understood the handling qualities sufficiently well to fly
safely at altitude. The autumn of 1905 in many ways mirrored the
autumn of 1902, with the Wrights flying extensively for two weeks to
master the flying characteristics of their Flyer No 3. To quote from
Wilbur Wrights summary of the 1905 experiments, written in 1912.
The trouble was really due to the fact that in circling, the
machine has to carry the load resulting from centrifugal force, in
addition to its own weight, since the actual pressure that the air
must sustain is that due to the resultant of the two forces. The
machine in question had but a slight surplus of power above what
was required for straight flight, and as the additional load, caused
by circling, increased rapidly as the circle became smaller, a limit
was finally reached beyond which the machine was no longer able
to maintain sufficient speed to sustain itself in the air. And as the
lifting effect of the inner wing, owing to its reduced speed, coun-
terbalanced a large part of the increased lift resulting from the
greater angle of incidence on that wing, the response to lateral
control was so slow that the machine sank to the ground, usually
before it had been brought back to the level again. In other words
the machine was in what has come to be known as a stalled condi-
tion. When we had discovered the real nature of the trouble, and
knew that it could always be remedied by tilting the machine
forward a little, so that its flying speed would be restored, we felt
that we were ready to place flying machines on the market.
Wilbur, perhaps unknowingly, was also highlighting that a pitch
unstable aircraft required forward stick in a turn to counter the neg-
ative manoeuvre margin. From autumn 1902 to autumn 1905 the
Wrights would work hard to design a powered machine with satis-
factory flying qualities, but without design criteria or supporting
analytical tools their progress would be limited. Wilbur also
believed that The remedy for the difficulty lies in more skilful
operation of the aeroplanes and one wonders whether he consid-
ered that instability was the natural condition of flying, much as it
was for bicycle riding. However, he was not alone and it would be
10 years after the first powered flight that the first aircraft with sat-
isfactory all-round inherent stability flew (the Farnborough BE2 in
1913, a Green-de Havilland design Ref. 2).
The Wrights returned to Kitty Hawk in May 1908 to carry out
the final design refinements to the Flyer No 3 and prepare for
their aerial demonstrations in the US and Europe. The pilot now
sat up in the machine, and had independent control of pitch, roll
and yaw. Flights with a passenger on board were also being made
and the time of Wilburs last flight at the Kill Devil Hills (on 14
May 1908) was seven minutes 29 seconds, covering a distance of
over 8km; the flight ended in a crash! In the months to come,
Wilbur would amaze the world with his feats of aerial navigation
in the skies over France. He wrote to Orville from Le Mans on 15
August
(1)
, The newspapers and the French aviators nearly went
wild with excitement. Blriot and Delagrange were so excited
they could scarcely speak, and Kapferer could only gasp and
could not talk at all. Bleriot was quoted as saying, I consider
that for us in France and everywhere a new era in mechanical
flight has commencedit is marvelous. What the Wright broth-
ers had achieved three years earlier far exceeded the progress in
Europe since the first powered flight there in 1906 the
Europeans could only fly in straight lines! With Wilbur demon-
strating what could be done, after 1908 the rate of progress in
aviation was quite startling; progress which, in many ways,
would leave the Wright brothers behind with their legacy as the
first aeronautical engineers and first test pilots
(24)
. Perhaps they
had reached their horizon?
7.0 CONCLUDING REMARKS ON THE
WRIGHT CRITICAL SUCCESS FACTORS
October 1902 was a watershed in the history of aviation. The use of
three-axis flight control paved the way for the invention and devel-
opment of powered flight to proceed. Flight control issues would
continue to feature large in the Wright brothers aircraft for the next
five years but the basic design concept would remain the same
forward canard for pitch/vertical flight path, wing warp for roll and
rudder for yaw.
Looking back to the period leading up to the autumn of 1902, it is
possible to suggest the critical success factors for the birth of flight
control
1) The Wright brothers were very capable engineers, very good at
conceptual design and what is described as visual thinking, and
particularly good at being able to stay focused on goals,
2) They had developed a high proficiency in the practical skills of
manufacturing and a knowledge in the use of materials,
3) They understood the laws of physics and aspects of the mathe-
matical analysis of mechanical systems,
4) They were powerfully motivated to learn how to fly, and made
thousands of gliding flights to develop performance, flight
control concepts and piloting skills,
5) They designed, built and tested exhaustively (aerofoil aerody-
namics, structures, control systems, engines, propellers etc.),
6) They realised that control and manoeuvrability were more impor-
tant than stability,
PADFIELD AND LAWRENCE THE BIRTH OF FLIGHT CONTROL - AN ENGINEERING ANALYSIS OF THE WRIGHT BROTHERS' 1902 GLIDER 717
7) They took what is nowadays referred to as a systems approach to
their work understanding that an aeroplane is a technological
system of discrete elements, including the pilot, all of which had to be
successfully designed and integrated to achieve success. Within this
context their continuity of design was also important, solving prob-
lems in an evolutionary fashion,
8) They were a team, sharing responsibilities, sharing a common vision.
Other success factors can be mentioned, including the Wrights choice
of test site, their willingness and (financial) ability to undertake annual
campaigns, but from the above list, factor number six was particularly
important. The aft cg and canard control conferred the 1902 Glider with
greater control power than an equivalent tailed aircraft, crucial at the low
speeds flown by the Wrights; the manoeuvrability was further enhanced
by the negative static margin (and for the furthest aft case, negative
manoeuvre margin). The canard also provided a powerful attitude refer-
ence and protection during heavy landings or crashes. Of critical impor-
tance to the Wrights, the wing design gave safe stall characteristics. The
wing anhedral increased the effective roll control power and closed-loop
directional stability and the warp-rudder control interlink assisted turn
entry.
All of the success factors listed above are considered important, and it
is hard to imagine the invention of powered flight being achieved quite
so rapidly, or without more serious mishaps, in the early 1900s without
the Wright Brothers and their pursuit of excellence. In the years to come,
many chief designers would be noted for similar attributes. Nowadays,
however, although University degree and Industry training programmes
encourage innovation, and attempt to instill a broad appreciation of the
technologies involved and an ethos of engineering excellence, few prac-
ticing engineers have the opportunity to work across such a broad front
and put ideas into practice quite so readily as the Wrights did. Perhaps
this is a shortcoming of our current systems and the centenary of
powered flight provides a good opportunity to reflect on this. For the
Wrights, almost at every turn, necessity was the mother of invention.
The desire to fly and ultimately to invent the practical aeroplane was the
spur to their relentless pursuit of the mastery of flight control. Perhaps
one of the challenges for 21st century aerospace engineering is to under-
stand how to enable innovation to flourish amidst so much advanced
technology.
Wilbur and Orville can also teach us something about patience. They
first applied to patent their three-axis flight control system in March
1903. It was rejected by the US Patent Office, who claimed the descrip-
tion was vague and indefinite and incapable of performing its
intended function
(4)
. The patent was eventually granted in the US in
May 1906, two years after Great Britain and France had approved the
application.
This paper has presented the story of the birth of flight control as fea-
tured in the Wrights 1902 Glider, and appraised in the Liverpool Wright
Project. High-fidelity simulations of the 1901 and 1902 gliders and the
1903 Flyer are currently operational on the motion simulator at Liverpool
and work is underway to create a simulation of the 1905 design, the first
practical aircraft. This ongoing research is supported by a Royal
Aeronautical Society Centenary Award. Future publications will report
the continuing progress.
ACKNOWLEDGEMENTS
The research reported in this paper was partly funded by an UK EPSRC
Doctoral Training award for the second author. Financial support from
The Friends of The University of Liverpool, enabling the wind-tunnel
testing to be conducted, is greatly appreciated and acknowledged. The
authors would like to thank Peter Jakab and Tom Crouch, Aeronautics
Curators at the National Air and Space Museum, Washington DC, for
fruitful discussions and for the opportunity to work with them in this
Centenary period. The authors are also particularly grateful to Professor
Norman Wood at the University of Manchester and the team at Flow
Science for the use of the Goldstein wind-tunnel facilities. The project
Test Pilot, Roger Dodge Bailey from Cranfield University/Shuttleworth
718 THE AERONAUTICAL JOURNAL DECEMBER 2003
Collection has worked with the Liverpool team during the project; his
insight into handling qualities from the pilots perspective has proved
invaluable. Most of the photographs taken during the Wrights test cam-
paigns and presented in this paper are the property of the US Library of
Congress who are gratefully acknowledged. Figure 10(b) and Figs 13(a),
13(b) and 13(d) are the property of Wright State University who are also
acknowledged. The Sketches in Figs 3, 4 and 10 were taken from Ref 25.
REFERENCES
1. MCFARLAND, M.W. (Ed), The Papers of Wilbur and Orville Wright, includ-
ing the Chanute-Wright Letters and other papers of Octave Chanute, 1953,
McGraw Hill, New York.
2. GIBBS-SMITH, C.H. The Aeroplane; An historical survey of its origins and
development, 1960, Her Majestys Stationery Office, London.
3. JAKAB, P.L. Visions of a Flying Machine, The Wright Brothers and the
Process of Invention, 1990, Smithsonian Institution Press.
4. CROUCH, T.D. The Bishops Boys, A Life of Wilbur and Orville Wright,
1989, W.W. Norton, New York.
5. CULICK, F.E.C. What the Wright brothers did and did not understand about
flight mechanics in modern terms, AIAA-2001-3385, 37th
AIAA/ASME/SAE/ASEE Joint Propulsion Conference and Exhibit, July
2001 Salt Lake City, Utah.
6. BRYAN, G.H. and WILLIAMS, W.E. The longitudinal stability of gliders, June
1903, Proc Royal Society of London.
7. LILIENTHAL, O. Der Vogelflug als Grundlage der Fliegekunst. Berlin: R.
Gaertners Verlagsbuchhandlung (Birdflight as the Basis of Aviation,
American Aeronautical Archives, Markowski International Publishers,
2001).
8. ANDERSON, J.D. A History of Aerodynamics and its Impact on Flying
Machines, 1998, Cambridge University Press.
9. PHILLIPS, W.H. Flying qualities from early airplanes to the space shuttle,
Dryden Lectureship in Research, AIAA 26th Aerospace Sciences Meeting,
Reno Jan 1988, AIAA J Guidance and Control, July-August 1989, 12, (4).
10. LAWRENCE, B. and PADFIELD, G.D. A handling qualities analysis of the
Wright brothers 1902 Glider, AIAA Atmospheric Flight Mechanics
Conference, Austin, Texas, 11-14 August, 2003.
11. MCCORMICK, B.W. Aerodynamics, Aeronautics, and Flight Mechanics,
1995, 2nd ed, John Wiley and Sons, New York, 1995.
12. JEX, H.R. and CULICK, F.E.C. Flight control dynamics of the 1903 Wright
Flyer, 1985, AIAA Paper, p 534-548.
13. PADFIELD, G.D and WHITE, M.D. Flight simulation in academia;
HELIFLIGHT in its first year of operation, Aeronaut J, October 2003, 107,
(1076), (also in The challenge of realistic rotorcraft simulation, RAeS
Conference, London, Nov 2001).
14. KOCHERSBERGER, K., SANDUSKY R., HYDE K., ASH R., BRITCHER C. and
LANDMAN, D. An evaluation of the Wright 1901 Glider using full scale wind
tunnel data, AIAA Aerospace Sciences Meeting & Exhibit, 40th, Reno, NV,
2002.
15. ETKIN, B. Dynamics of Atmospheric Flight, 1972, John Wiley and Sons,
New York.
16. Anon MIL-F-8785C, Military Specification Flying Qualities of Piloted
Airplanes. U.S. Department of Defense, 1980.
17. Anon MIL-HDBK-1797, Flying Qualities of Piloted Aircraft, US
Department of Defense, 1997.
18. COOPER, G. E., HARPER, R.P. The use of pilot rating in the evaluation of air-
craft handling qualities and pilot evaluation, NASA TN D-5153, April 1969
19. MILNE, R.D. and PADFIELD, G.D. The Strongly controlled aircraft, Aeronaut
Q, May 1971, XXII, pp 146-168.
20. CULICK, F.E.C. and JEX, H.R. Aerodynamics, stability and control of the
1903 Wright Flyer, The Wright Flyer, an engineering perspective, National
Air and Space Museum, Washington DC, 1987.
21. PADFIELD, G.D. Helicopter Flight Dynamics, 1996, Blackwell Science,
Oxford.
22. LAWRENCE, B. and PADFIELD, G.D. Flight testing simulations of the Wright
1902 Glider and 1903 Flyer, 34th Annual International Symposium of the
SFTE, Portsmouth, Va, September 2003.
23. WOLKO, H.S. (Ed) The Wright Flyer; an engineering perspective, 1987,
Smithsonian Institution.
24. CULICK, F.E.C. Wright brothers: first aeronautical engineers and test pilots,
Sept 2001, 45th Annual Symposium of the Society of Experimental Test
Pilots, Los Angeles, California.
25. CULICK, F.E.C., The origins of the first powered, man-carrying airplane,
December 1979, Scientific American.