l”; -O‘. ' W. h- ' :‘511.(:A‘. ' lair.

 

mum of THE nmmmm Bf THE AIBPlANE

THESIS FOR DEGREE OF M. E. .

WlLLIAM EDWARD SAVAGE

AAAA ~ 2%.:t7flf-‘z’: I — _ o.

 

 

“'H HQfit

.r 3“ ya.
K .‘ J A I ~ \ ‘ .
’ ) P . I. 7 ' "" " '\. 3i... ‘y'\;’ 9‘; '.
0....L/ Lia § . .

WWWWWW

“\‘fl \ \\

3 1293 01074 0284

U)

PLACE IN RETURN Box
to remove this checkout from your record.
To AVOID FINES return on or before date due.

DATE DUE I DATE DUE am we

T *2, 37‘
"d i . r‘
N.‘ ' 3r

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

me m.m‘"

 

   

 

mw\\CLqM‘A.'Q~ flu“

(h:

9*

.9‘1.

V ('1

HISTORY OF THE DEVELOPMENT OF THE AIRELfiflE

By

William Edward Savage

w

Submitted in partial fulfillment of the Requirements for the
Degree of Mechanical Engineer
From

MICHIGAN STATE COLLEGE

Date: May 16, 1937.

 

TH F'QIQ

§l§TORY OF THE DEVELOPMENT OF THE AIRPLANE

 

In writing this history of the development of the airplane,
it is the aim to present only the events and changes that have be-
come the structure landmarks between the old stick—and—wire air—
planes and our modern stressed-skin designs. Much has been written
on the airplane and related subjects from the layman's point of
view. In this thesis the material will be presented in such a man—
ner as to be of interest, it is hOped, to the professional engineer,
to the engineering student, and to others concerned with the develop-
ment of the machine that has reduced distance in terms of time and
brought formerly remote and isolated.parts of the earth, where other
means of transportation have been impractical, if not impossible,
into the path of world travel and commerce. For mountains and
oceans are no obstacle to these man-made birds.

Records show that the first flight in a machine whose sup-
port depended on velocity and on no other means of buoyancy was
made in the United States, on December 17, 1903. From this time
until the outbreak of the World War in.&ugust, 191M, the develop-
ment and advancement of the airplane was slowz'because it was men
with small capital, small shops, and in some cases, no shops at
all, who had the desire to look down on this earth from a flying
machine and to try their luck in exhibition work at various county
and state fairs. The construction of the airplane during this
period was possible only because of the materials used, which con-

sisted of wood, piano wire, a few low-grade steel fittings for

- 1 -
"1 “‘91.: .‘l

the supporting structure, and linen or a good grade of cotton for
the covering. The use of these materials required few tools,
making the building and repairing of the airplane in use during
these years a comparatively simple matter. This was desirable,
because repairs often had to be made quickly and on the landing
field where the exhibition was being given. The biplane was

the type most used by these flyers, which, with its low wing
loading and low power loading, was a happy choice for this type
of work. For, although these airplanes were not built to severe
requirements, this low loading, tOgether with the lack of either
streamlining or cleanness of design, produced a low-speed air-
plane, whose strength requirements were adequately met by the
type of construction then known. This and careful handling in
the air are the only reasons why more of these airplanes did not
fail in flight. With the earnings from exhibition flying at
fairs, the pilot, who was his own mechanic in many cases, would
prepare during the off season for the next year's work. Either
he made changes in his existing airplane or he built a new one
incorporating the ideas obtained from the period of flying.

When an exhibition was over, the airplane was not flown to
another, but was disassembled, boxed, and shipped by rail.
Shipping airplanes by rail was practiced.until after the sum-
mer of lBlh, when, owing to the war, exhibition work was dis-
continued.

Prior to the entry of the United States into the World

War. there were a few aircraft companies, small in size, who
contracted to produce airplanes for foreign countries. The
work was done in.American factories, but was under the super-
vision of representatives of the country buying the airplanes.
At that time none of these countries knew much about require-
ments for military airplanes; but the United States knew even
less. However, by the time this country entered the conflict,
considerable information had been collected by the Allies, which
they passed on to American government agencies. In the summer
of 1917 the.Airp1ane and Engine Sections of the U.S. Signal
Corps were created in Washington, and in November of that year,
they were moved to McCook Field. Dayton, Ohio, where they
finally developed into the present Materiel Division of the

U. S. Army'Air Corps. This was the real beginning of a splen-
did work, but owing to the pressure that was necessary at this
time, items of prime engineering importance were neglected.

Realizing that there was little basic design information
which could be used, it was decided to conduct experiments in
order to obtain the necessary information. Therefore, the
Structures Branch of the Research.Department was created.

It was the duty of this newly formed Branch to determine
and approve the strength characteristics of all military air-
planes. Little was known about calculating loads in the members
of such flimsy structures: so the best method was to actually

load the structural assembly to simulate air loads. The loads

to be applied were calculated from the meager wind tunnel test
data that existed at the time and from the acceleration values
that might be expected in maneuvering the airplane. By means
of this work, as well as through an increasing number of wind
tunnel tests, new rules and methods for calculating the strength
of structural members were formulated. The wind.tunnel work
consisted in placing in the tunnel small models of airfoils or
airplanes mounted on a weighing device, then in forcing the air
past the model while at the same time, weighing the forces
caused by the air in passing over the model. The most common
forces measured were lift, drag, and moments for the airfoil
or airplane model; at all angles of attack within the flying
range of the airplane. Model after’model was tested in the
wind tunnel, and the results used in the design of future air-
planes. With these data the engineer's best effort was put
forth to design a structure that would satisfy the strength
requirements and yet be light in weight. This procedure was
the best possible at that time. However, the static testing
of complete airplanes was continued, each test furnishing a
lesson and indicating where improvements could be made. It
became customary for manufacturers building airplanes to
specifications to furnish three full-scale models of a design.
One of these models was used for the purpose of conducting
flight tests, another for destructive structural tests to de-

termine strength characteristics, and the third was stowed

for use as a working model, if needed. This procedure continued
until the engineering departments of the different manufacturing
companies could be improved and become well organized, which re-
quired no little time. Much of the improvement came about through
the work of the different colleges-and universities, who con-
tinually sent out men better qualified in aeronautical engineer-
ing work. Many of these institutions became interested enough to
build and Operate their own wind tunnels in order to obtain in-
formation and to check tests that had been made in other wind
tunnels. In addition, structural problems were studied from the
reports of actual structural tests. As a result of the interest
shown by the different institutions of learning, a marked im-
provement took place in the engineering work submitted as early
as 192”, when the proposed designs submitted to the Air Corps in-
cluded all the calculations made and the methods used.

It was also around l92h when the Post Office Department
began letting contracts to private owners of airplanes for the
carrying of the air mail, which until then had been carried in
remodeled wartime airplanes Operated by the Post Office Depart-
ment itself. Manufacturers were thus rewarded for their efforts
in producing airplanes that would carry more pay load at a greater
speed.

As early as 1922, structural tests were conducted on all-
metal, internally-braced monoplanes manufactured by Junker in

Germany. The type of structural design incorporated in this airplane

was new to this country, as was also the kind of metal used. The
Junker wing structure was built with duralumin tubing and steel
fittings to form the spare, and corrugated sheet duralumin to form
the cover. Thus, the cover acted as a structural member of the
wing, and resisted drag loads imposed upon it. As this was the
first wing to be tested in the United States in which the cover—
ing was used to resist and transmit load, a new structural term,
"stressedvskin" wing, was coined; this is the type of construction
in common use today. Without a doubt, this construction is the
result of the early work of Dr. Hugo Junker and Herr Dornier, of
Germany, and Messieurs DeBoysson and.Dewoitine, of France, all of
whom have patents on some modified type of structure in which the
load is transmitted by a reinforced skin. The work of these men
dates back to 1915. However, because of the thinness of the
sheet, difficulties were encountered in its application and
manufacture. These difficulties were the result of the light
wing loading in use at that time, which with the stressed skin
type of construction, produced such a heavy airplane that it was
several years before the type was generally adapted. When air-
planes with greater wing loading began to appear, which was in
1927, a study was made by the Materiel Division of all types

of metal construction that would have a high strength-weight
ratio. From the results of this study it appeared that the
logical way to resist a distributed load, such as an air load

on the wing, was to keep the stresses distributed, which thought

is sound; for, when a distributed load is concentrated, a pre-
mium in structural weight must be paid. With this in mind, a
wing having two shear webs some distance apart and corrugated
thin sheet between these webs, which formed a box, was built
and tested. This wing had a span of 55 feet and an area of

H23 square feet. The corrugations of the material, running
spanwise and forming the flanges of the cantilever beam, re-
sisted the moment, while the two webs resisted the shear. The
tests conducted on this wing, during October and November, 1930,
furnished some interesting and valuable data even though the wing
was similar in construction to those patented in 1915. The
principal lesson learned from these tests was how to distribute
the material to produce the radius of gyration that prevents
buckling while the wing is under load. This is a problem in
itself, and the only secret is to keep the structural material
where it is placed. This is not a difficult thing to do when
one has done it many times. If structural members are removed,
the stress in them is removed also; for this reason one of the
two shear members has been eliminated in the box stressed-skin
wing, leaving a wing with a single-shear web (which must be
properly located) and with a reinforced nose section to resist
the moments due to torque and bending. Several wings having a
single-shear web have been built with.much success. In using
but one shear member, the thickness must be increased, which

tends to reduce the possibility of failure due to fatigue caused

by vibration, but it must be remembered that the single-shear web
should.be used in wings with airfoils having low moment coeffi—
cients or a constant center of pressure.

Stressed-skin construction has been used in fuselage or
body structures much longer than in wing structures. During the
war, when much wood was used, many fuselages were of this type of
construction: but designing and constructing a stressed-skin
fuselage is not as difficult as designing and constructing a wing
of this type, on account of the depth and size and the much lower
load it must support in comparison with the load the wing must
support. If the sizes that structural design dictates were used
in the fuselage, it would be undesirable in appearance and difficult
to maintain. The design principles are the same, but it is much
easier to design and'build a fuselage for a reasonable weight; that
is, provided it is not desired to have too many Openings, which tend
to lower the natural frequency of vibration in torsion and.bending
regardless of how reinforcing material is used around them. Vertical
and horizontal tail surfaces may be of the same type of construction
as the wing.

Several earlier attempts were made by the Air Corps to use
all-metal construction in airplane design. Late in the year of 1992
tests were conducted on the first all-metal monoplane designed and
built in the United States. This airplane was a corps observation
type known as the 00-1. During the same year the Gallaudet day

bomber was built by the Galladet Aircraft Corporation of East

Greenwich, Rhode Island, who used metal construction with fabric
covered wings and tail surfaces. In both of these airplanes, much
use was made of duralumin and highly heat-treated steels. The
structure of the 00-1 was built of duralumin corrugated sheet,
duralumin tubing, and steel tubing and fittings. The fuselage

of this airplane was designed similarly to that of the Junker
J.L.6, whereas the main structural members of the wing were
definite members forming a space-frame structure. The duralumin
cover of the fuselage, which had its corrugations running length-
wise, supported at intervals by bulkheads, resisted the shears
and moments under the different loadings. These bulkheads
stabilized the covering against buckling, thus permitting the
cover to be stressed more highly before failure. The wing was
constructed with steel and duralumin spare and a corrugated
duralumin cover. The spars resisted the beam loads, whereas
both the spare and cover resisted the torsional and drag loads.
The structural design of this airplane followed the same thought
as the older type of definite struchure except that the fuselage
was of the stressed—skin type and the wing was metal covered to
resist shear due to drag loads and shear due to torque.

The Gallaudet day bombardment airplane was the first air-
plane to have its fuselage streamline in shape and to have an
elliptical cross section. This was an advanced step in fuselage
construction, and in order to build it, the services of silversmiths

and coppersmiths were used. The internallyAbraced,monoplane wing

of this airplane was constructed of two truss-type spars having
heat-treated shear and chord members; but the spars were stabilized
by the usual drag wires, bulkheads, and ribs, which also resisted
the drag and torsional loads. This airplane, although an outstand-
ing achievement in construction and an advanced thought in design,
never took the air, because of its excessive weight, most of which
was built into the fuselage. The airplane was redesigned with
fabric covering for the space-frame structure. When the redesigned
airplane arrived, it showed a tendency during flight tests to roll
to the right, even though the engine torque should have given it a
tendency to roll to the left. .Assuming that something was wrong
with the structure, it was decided to conduct proof tests on the
internally-braced wing; these were completed during the months of
April and May, 192h, From the results of these tests it was
found that the aileron, or lateral control system, was faulty and
that a slightly greater angle of incidence was built into the left
wing panel than was built into the right panel. This, without a
doubt, caused the tendency of the airplane to roll to the right.
On the first flight, when the rolling:tendency was soon discovered,
it was not certain whether the airplane could be landed safely.
After the airplane was made to fly pr0perly, it developed resonance
vibration, or flutter, or both. The airplane was landed by skillful
piloting in a badly damaged condition. Soon after this accident,
further development of this type airplane was abandoned.
Discontinuing work on this type airplane was a mistake,

because it was not long before a similarly constructed airplane was

_ 10 _

built by one of the aircraft companies and successfully flown in
carrying the air mail over one of the air transport lines. This

new all-metal monoplane was the forerunner of the present airline
passenger and mail planes. Of course, airplanes have been improved
greatly since this one was built, and much has been done to in-
crease their size. The possibilities of large airplanes were con-
sidered as far back as 1921, and as a result, the first large outer
wing panel was tested in March 1922. This wing panel was internally
braced and had a span of HO feet; the total span of the airplane was
97 feet h inches with a root chord of 26 feet, a tip chord of 13
feet, and a total of 2082 square feet lifting surface in the complete
wing. The design, construction, and tests of this outer wing panel
constituted a splendid experiment, but had such construction been
put into service, it would have required a trained engineer to be

one of the crew of each airplane in order to keep the proper amount
of tension in the tie rods or tension members. The complete airplane
for which this outer panel was designed was never built.

For all the thought given to the large airplane at this time,
the small airplane was not neglected. From the close of the war and
until 1925, encouragement was given by the.Air Corps to the design
and building of racing airplanes. The racing airplane, usually small
and fast, furnished useful information because of the absence of
equipment, in checking both the engineering data and the accuracy of
the method used to obtain these data. ,As the result of the engineering

data obtained in the flight of these small, fast, highly maneuver-

-11..

able airplanes, the strength requirements for the pursuit type

were increased nearly 50 per cent. This change in the requirements

of the pursuit airplane was made in 1923; there has been little
change necessary in these requirements since. However, many new
requirements have been added, and a change has been made in the
presentation of the design requirements for aircraft. The first
pursuit airplane to be tested under these new requirements was the
small biplane known as the PW-9, weighing 2910 pounds. This air-
plane was designed and built during the summer of l923, and was
subjected to the usual structural tests in November and December

of the same year. It was far in advance of its time, and many of

its features have been incorporated in later designs. Contemporaneous
with the arrival of this airplane at McCook Field, in Dayton, Ohio,
was that of another pursuit practically the same in size and known

as the PW-s. This airplane was also a biplane, but it was unlike

the PW-9 in that it had two bays in its lift truss. During flight
tests, while these two airplanes were diving down as straight as possible,
the propeller on the PW-S failed owing to the increase in its rotational
speed. The failure was a lucky one, because it occurred at the hub,
allowing both halves of the blade to fly off at the same time. One

of the halves landed in Northwest Dayton, while the other came to
earth in North Dayton, a distance of approximately two miles apart.
This was a propitious occurrence, because much was learned, with no
one hurt even though one-half of the propeller landed in a back yard

where children were at play. Some of the prOpeller hub bolts fasten-

- 12 -

ing the hub to the engine crankshaft flange tore pieces out
of some interplane struts, but the struts were not sufficiently
weakened to cause a collapse of the entire wing structure.
These two airplanes had the usual wood, wire-braced wings,
and had welded-steel tubular, wireabraced fuselages. These
airplanes gave way to the P-12 series pursuit airplanes, the
first of which was designed in 1929 and tested for structural
properties in January, 1930, just prior to the production status
of the type. The P—l2 airplane resulted in one of the largest
production orders for airplanes since the close of the World
War. No one was ever killed in this type on account of failure
in a structural member. During the production period, the
duralumin space-frame fuselage was changed to a stressed—skin
design. This P-12 stressed skin type pursuit airplane, in which
much wood was used in the wing cellule, was followed.by one
constructed entirely of metal and known as the P-26 airplane.
All the structural members with the exception of a few steel
fittings were of duralumin. One error made, which.was no fault
of the designer, was the use of external wire bracing in the
lift truss. Many of these airplanes, after the addition of
auxiliary lifting surfaces to the original design, are still
giving good service.

While the pursuit airplane was having so much thought
given to its design and performance, development of the bombard-

ment type lagged behind, until 1929, when the 3—9 airplane,

_ 13 -

entirely of metal, was designed and put under construction.
The fuselage of the 3-9 was so designed that the loads were
resisted not only by the skin, but by four longerons, or heavy
structural members, spaced as far apart as possible in the
elliptical cross section. The wing was of the two-spar con-
struction, with ribs and bulkheads to support the thin metal
cover. The two spars resisted the shear and moments due to beam
load, while both the spars and the thin metal cover resisted the
shear due to the torsional moment. The two engines were mounted
in nacelles on the wing and thus balanced some of the shear and
moments due to beam load while adding to the torsional load to
be resisted by the wing. Although only seven of these B-9 bombard-
ment airplanes were constructed, much was learned from their use.
Early in 1932 the stressed-skin type wing, which had been
in various stages of development prior to this time, was used in
a biplace pursuit monoplane. The stressed-skin wing for this air-
plane was constructed of heavy flat sheet reinforced with stringer
material having various cross-section shapes, which replaced the
corrugated material and thin flat sheet previously used. This
cantilever wing had a hu—foot span and a 297-square foot area.
The air loading was 17.5 pounds per square foot; the design loading
210 pounds per square foot. The weight, including the reinforce-
ment around the opening for the retractable landing gear, was
2.9% pounds per square foot. The use of heavier flat sheet with

stringer reinforcing produces a smoother surface, which is also

- 1h -

better aerodynamically under.static and dynamic loading than
can be expected with the corrugated reinforcing and thin flat
sheet.

The resisting of load by means of a reinforced skin is
not so problematical in the design of the fuselage and engine
nacelles on account of their depth and comparative low loading.
In the design of a wing, where high dynamic loading is to be
resisted, the problem becomes more involved in the stressed-skin
construction, particularly when many openings are necessary,
such as those for the retraction of the landing gear, for the
installation of various items, and for inspection purposes.

The first airplane to be submitted in which a smooth
reinforced skin was incorporated in the design was the C-l9,

a single-engine transport or cargo airplane delivered in June,
1931. The methods of design for the corrugated construction
and the smooth reinforced skin are practically the same, there
being very little difference in the strength-weight ratio of
the two types of construction. In the manufacturing process,
more riveting is necessary with the reinforced flat sheet than
with the corrugated and flat sheet. However, in the latter
construction much difficulty is experienced in the spring back
of the corrugations, both in the shearing and fitting oper-
ations. The excessive amount of riveting may be eliminated by

electric spot~welding.

I
H
\J‘I
I

Duralumin or aluminum alloy, which contains aluminum
and a varying amount of copper, tOgether with small amounts
of manganese, magnesium, and small amounts of certain impurities,
is subject to inter-crystaline corrosion due to electrolytic
action; this action is accelerated in sea water and sea air.
However, even with this bad characteristic, it is suitable for
aircraft construction, for, with an organic protective coating
or with a coating of commercially pure aluminum rolled integrally
with the alloyed metal sheet, excessive corrosion is retarded
even though the c0pper and aluminum are of opposite polarities.
Owing to the action of this metal in and around sea water, there
was designed and built a corrosion-resistant steel wing for an
amphibian airplane. With.this airplane operating in and out of,
or near, the sea, it was thought that it would be subjected to
the most severe service conditions. When the use of this material
for aircraft was first proposed, many considered it infeasible, mean—
ing that it could not be done for a reasonable weight. Well, it
proved practicable, but was delayed three years because many did
not understand the problem. Since this material is denser than
the aluminum alloy commonly used in aircraft construction, much
more care must be exercised in its support and distribution and
in the method of fabrication, in order to obtain a reasonable
strength-weight ratio. The tensile strength of corrosion-
resistant steel may be increased to more than 200,000 pounds

per square inch by "cold work", not by heat-treatment. In the

"cold worked" state it is impractical to drill or ream holes

in it. Hence, much use has been made of electric spot—welding

in making Joints. This method of making joints is not new,

for it was used when the automotive industry made fenders from
many pieces. The automotive industry achieved satisfactory
spot—welds, but when the welds were subjected to vibration, the
metal adjacent to the weld would fail. The cause of these
failures was thought to be a lack of proper timing and current;
therefore, much time was spent in a study of timing devices and
the proper current to use. This work had to be done before any
attempt could be made toward a design. .After many spot-welded
sample Joints were made and tested, reasonable allowable shear-
ing and fatigue values of the welds were established. Upon
completion of this work a wing having a 60-foot span and an

area of 5h0 square feet was designed, built, and tested. The
static test of this wing indicated that it had greater strength
than the wood wing of like dimensions which it replaced and that
it was somewhat lighter. This wing was subjected to a vibration '
test in the laboratory in order to test the spot-welds for fatigue
strength. After 120 hours of mechanically vibrating the wing,
there were no failures due to fatigue. At present, the destructive
forces due to vibration under service conditions have not been
investigated, but it is not thought that the amplitude of
vibration under service conditions will be any greater than

that to which the wing was subjected in the laboratory test.

- 17 -

A number of these wings are now under construction for service
test purposes.

An investigation and study of the use of electric
spot-welded aluminum alloys has been made. This problem is
much.more difficult than the electric spot-welding of steel
because of the character of the metal. Although this study
has been in progress five years or more, to date no electric
spot-welding in main structural members of aluminum alloy has
been given a service test. This method of making a mechanical
joint in.aluminum alloy will be perfected, but not in a short
space of time or without disappointments. The use of electric
spot-welding in airplane structures reduces the cost of build-
ing and eliminates rivet heads on the exposed surfaces of thin
sheet, where countersinking for the use of countersunk head
rivets is impractical and costly.

Much investigation has been done on the design of a
metal wing in which the forces are resisted by a single-shear
web and thin sheet, reinforced covering. This structure acts
as a unit in resisting the shear in the beam direction; the
moment due to shear in the beam direction; and the shear due
to torque. Because of the less indeterminate features in
single-shear web construction, it is superior to and more
efficient than the multiple-shear web construction. The
former type of construction was first used in an airplane

four years ago, and is giving good service.

- 13 -

In the past the design of an airplane structure was
based on tests of small models, conducted in the wind tunnel
to obtain the forces such as lift, drag, and moments at dif-
ferent angles of attack. The most important of these forces
for structural design purposes is the force due to moments.

The scale of such models is often such that the results ob—
tained are not sufficiently accurate to be used with satis-
faction for structural design. Recently, the scale of models
has been increased, with the result that more accurate measure-
ments are obtained in a larger wind tunnel.

In the design and construction of large airplanes, it
will be necessary to rely more and more on computations for the
design of structural members, because of the smaller number of
large airplanes to be built and the cost involved. It is de-
sirable to test to destruction an airplane of each type selected
for service, in order to prove whether the engineering work has
been done properly and in accordance with the assumptions made,
the quantity to be obtained may not always warrant such tests.
This procedure will be appreciated more if it is considered that
the design load is but 50 per cent greater than the expected load;
in many cases, according to records obtained, the expected load
has been exceeded. It is interesting to compare this increase
of only 50 per cent in the design load over the expected load
with the 300 per cent used in the design of other engineering

projects.

_ 19 -

To replace the destruction tests of large expensive
airplanes, structural proof tests will be made in flight. This
will be satisfactory so long as the proof test is not carried
too far. It must be understood that any airplane can be dis—
integrated in the air by control forces. Furthermore, the de-
sign load for the large airplane is much lower than it is for
the small airplane; therefore, any over control would be more
effective on the design strength of the large airplane than it
would be on the smaller airplane. Proof tests and load distri-
bution tests should.be carried on under the direction of an
engineer with much experience. many large airplanes are now
being developed in the United States, in each of which a
modification of some type of structure herein discussed is
exemplified. The airplane will not be increased further in
size when the wing weight becomes such that it is no longer
economical to build larger airplanes.

Considering the efficiency of present lift sections,
or airfoils, a 30-pound per square foot wing loading is about
as high as is desirable if a reasonable landing speed, even
with the present auxiliary lifting devices, is to be maintained.
Again, considering the limit of loading to be 30 pounds per
square foot, it is believed that, with the present commercially
available construction material and with the efficiency of the
present airfoils and auxiliary lifting devices, a wing weight of

6 pounds per square foot is the limit for dead weight. From

data available on the weight of large airplane wing structures,
it appears that a weight of 6 pounds per square foot will be re-
quired in a wing having a span of 250 feet. It must be remembered
that the ratio of wing weight to span does not increase as the
first power or straight line function, but is parabolic.

In the design of large airplane wings, advantage is
taken of the distribution of the weights of fuel, oil, engines,
and cargo, in order to reduce the shear and moment along the
span, since the weights of these items act in Opposition to the
air loads. When a load of 25 tons of fuel is carried, such an
item is not to be overlooked in counteracting air loads if the
wing;weight is to be kept to a minimum. Of course, the spanwise
distribution of these weights can not be carried too far, because
increasing the mass moment of inertia of the airplane about the
vertical and longitudinal axes would result in unfavorable, if
not dangerous, handling characteristics. When the distribution
of these weights has been taken into consideration in the design,
any change in their distribution while the airplane is in service
will lead to difficulty and perhaps to failure of the structure.

The design requirements for each particular type of
airplane, besides being based on the expected usage, should be
based on records, collected over a long period of time, show—
ing the accelerations to which similar airplanes have been sub-
jected. Such records are obtained by using a velocity-acceleration

recorder, the instrument commonly called a V-G recorder. This

- 21 -

instrument is a development of the National Advisory Committee

for Aeronautics and is designed with a system of links connected

to an operating a stylus that draws a diagram on a smoked glass.

.An accelerometer mass damped by a spring in either direction
produces the vertical motion of the stylus, whereas the airspeed
diaphragm, which is Operated.by the pressure in the pitot tube,
furnishes the forces for the horizontal motion of the stylus.

The V-G recorder is a valuable instrument for collecting data

for research, but it is of no value as a guide in flying the air-
plane in which it is mounted because it is non-visual and is usually
located at the center of gravity of the airplane. In many instal-
lations it is located out of reach of the pilot. The many V—G
records collected during a period of several years show that the
design requirements used prior to Obtaining these records were not
far wrong. Although the present military airplane has sufficient
strength, as shown by these records, it is an easy matter to effect
a complete collapse of the present-day airplane by rough handling

of the controls. In fact, because of the different physical ef-
fects on the human body, it is very easy to misjudge the deceleration
to which an airplane is being subjected. In one experiment in which
the airplane was to be subjected to a deceleration of 8 g., it

was by accident decelerated to 10.5 g. The overhcontrolling of

this airplane shows how easy it is for the pilot to make an error

in Judgment because of the physical effects of accelerated forces

_ 22 -

upon him. It also indicates the urgent necessity for a visual
type accelerometer to be mounted in the airplane so that the

pilot will not need to depend upon his physical reactions. These
instruments are more important in the larger airplanes, because of
the lower load factors to which they are designed. Moreover, in
the larger airplane there are more lives and a greater investment
to protect.

With the increase in speed of the airplane came the annoy-
ing troubles of flutter and buffeting, which cause great concern.
Flutter and buffeting are two subjects that are of first importance
in the design of aircraft. Flutter may be defined as a force re-
sulting from some mechanical or aerodynamic force, or from a com-
bination of both these forces, and matching the natural frequencies
of the different assemblies or parts of the airplane. Some of the
common causes of flutter are rough engines; dynamically unbalanced
propellers; and loose vertical, lateral, and horizontal control
systems. In order for the designer to avoid flutter, the natural
frequency of the different assemblies or parts of the airplane
should be widely separated, the movable control surfaces should
be dynamically balanced, and initial loads should be maintained
in the control system. The maintenance of initial load in the
control system will prevent any looseness in the system that
might be effected by deformation of the structural members, which

is caused by load and temperature changes.

_ 23 _

Buffeting may be defined as a force caused by impulses
set in motion by a disturbed airflow whose frequency matches the
frequency of that assembly or part of the airplane in its path.
The parts of the airplane usually affected by buffeting are the
wing, tail surfaces, and, in some cases, the fuselage.

In order to avoid buffeting, the airflow should leave
the surface undisturbed, and this may be accomplished by
eliminating abrupt changes in section, unfaired projections,
and aerodynamically unstable airfoils.

Forces caused by either flutter or buffeting are destructive
if allowed to continue. Since the frequency of an engine or a
prOpeller, or a combination of the two, as well as the disturbed
airflow, will change with airplane speed, the engine speed should
be reduced in order not to create destructive forces should flutter
or buffeting occur and should it be rec0gnized in sufficient time.

During the year of 1920 experiments were first undertaken
to determine the possibility of supercharging an airplane cabin in
order to maintain low altitude temperature and pressure conditions
during flights at high altitudes. By 1922 sufficient flights had
been made to establish the fact that airplanes carrying passengers
and crew under pressure were entirely feasible, provided the
mechanical details could be preperly designed to operate under
all conditions encountered during such flights. When these

facts were established, the work on supercharged cabins was dis-

- 2h _

continued because of more urgent work. During the summer of

1935 it was decided to make an extensive study of cabin super-
charging for high-altitude flying. When flying above 12,000

feet without supercharging, the occupants of the airplane should
take oxygen into the lungs through a hose connected to an oxygen
bottle and the mouth. To supply the lungs with oxygen in low
pressure air by this method tends to lower both the physical

and mental efficiency of the occupants. For this reason it was
desired to undertake further research work on the supercharging
of cabins. Another reason, directly connected with high altitude
flying on the commercial airlines, was added when it was reported
that a number of deaths of passengers had occurred in flight.

It is not known whether all these deaths were caused from the
lack of pressure. In order to avoid any accidents in super-
charged cabin airplanes such as have occurred in foreign countries,
many problems had to be considered. Most of these problems had
already been investigated, but not with the thought of applying
them to such an airplane. Some of these problems are: the ef-
fect on the structure of stresses due to pressure change, temperature
change, flying stresses, bullet punctures, vibration, explosions,
ngging or frosting of windows, frosting of regulating and safety
valves, and failure of the exhaust-driven supercharger supplying
air to the cabin. Many of these problems had not only to be

studied, but experiments had to be conducted at both room

_ 25 _

temperature and room pressure, at higher pressures, and in
cold chambers to check these details as they apply to the super-
charged cabin airplane. The physiological investigations carried
on were the most interesting, because little work had been done
and much less had been written on this work. Some of these in-
vestigations were: the air required per passenger; pressure,
temperature, and humidity control; ventilation; rate of discharge
of air permissible; and oxygen spray in an emergency.

From experiments it was found that air contains the same
amount of oxygen as far away from the earth as man has traveled
as it does at the earth's surface. At sea level under standard
conditions, oxygen enters the lungs of a man at a partial pressure
of 159 mm. Hg., and the effect of the partial pressure of oxygen
depends on the individual and the altitude.~ If sufficient altitude
is reached there will be produced in any individual oxygen-want
‘(anoxemia), unconsciousness, and finally death. If an altitude
of 67,500 feet, where the pressure is reduced to M7 mm. Hg., could
be reached without a supercharged cabin, the fluids of the body
would begin to boil and, of course, the results would be fatal.
In the supercharged cabin there will be gases, dangerous and
annoying, to be either eliminated or controlled. The air inlet
to the cabin supercharger, which is driven by engine exhaust,
will be located far from the engine to avoid any carbon monoxide

entering the cabin, and.there will be no exhaust heater used. A

carbon monoxide indicating instrument, similar to those used in
mines, will be installed to determine the quantity of this gas at
all times in case of a leak. This gas should never exceed the
maximum allowable concentration of .005 per cent. Exhaust heaters
will not be needed, because the heat of compression due to super-
charging will be sufficient with the proper insulation of the cabin.
The carbon dioxide may be eliminated from the cabin by ventilation
or by chemical absorption, or by both of these methods.

A factor of importance is the sudden or instantaneous drOp
in pressure caused by‘a structural seam failure, the blowing out
of a window, or the Opening of the air—tight door. From such
failures there is evidence that brain hemorrhages have resulted,
and probably there would be marked disturbances of the lungs,
intestines, and heart.

A multi-engined airplane with a supercharged cabin will
be completed in April, 1937. It will be officially known as the
XC-35 airplane, but unofficially, as the flying laboratory. When
experiments are conducted with this airplane, there will be an
increased demand for human guineaepigs.

At the close of the War most people thought there would
be no further use for the airplane. These peOple had good reasons
for such thoughts after completing a study of the airplane at that
time.

The efficiency of American-built airplanes has steadily

- 27 _

increased, making them desirable for purchase by foreign countries.
Today'these countries would purchase, if such were possible, more
commercial transport airplanes and fighting airplanes than this
country is capable of building with its present facilities. The
reason for the desire to obtain American-built airplanes is that
they are second to none in performance, efficiency, and adapt-
ability to production.

Some estimation of the advancement in the develOpment of
the airplane may be made when considering the high speed of 127
miles per hour for the pursuit airplane at the close of the War
as compared to more than 300 miles per hour for this type airplane
today. In making such a comparison, it must not be forgotten that
the power required varies approximately as the cube of the speed,
whereas the required strength varies as the square of the speed.
Not only has the speed been increased, but the flying range has been
increased also. This increase in performance came about as a result
of research and develOpment work, from which better design conditions
and requirements were formulated.

Much credit should be given to the men who set such an ideal
standard many years ago and whose work has resulted in this vast im-

provement in airplane design.

(NOTE: This thesis or any part of it shall not be reprinted without
the permission of the Chief of the Air Corps, United States War Depart-
ment, and the.Author.)

- 23 _

1 u

EL“

Z 9
I
JR;

an.»

. , 3, ,.
has“;

 

3264

arre

109103

. ‘. J ..

'h".

 

HICHIGQN STQTE UNIV.

LIBRQRIES

3129301 074028 84