UNCLASSIFIED
AD NUMBER
AD830267
NEW LIMITATION CHANGE
TO
Approved for public release, distribution
unlimited
FROM
Distribution authorized to U.S. Gov't.
agencies and their contractors; Critical
Technology; SEP 1963. Other requests shall
be referred to Army Materiel Command,
Washington, DC.
AUTHORITY
USAMC ltr, 14 Jan 1972
THIS PAGE IS UNCLASSIFIED
ENIN
DIGN
ERINT
OO
CHAN
InI
OFATMATEIE
DEIG
ENIERN
a
b
uiaY
t2 ljgn rastL
Sa~
ntr
~
HA:ett
~tl'LoefNDBO
4eetL)Y
tn
Commandl
2 0 315
tt
appr
af'.
AG
4a
't
HEADQUARTERS, U. S. ARMY MATERIEL COMMAND
et
SEPTEMBER 1963
This Document Contains
Missing Page/s That Are
Unavailable In The
Original Document
BEST
AVAILABLE COPY
'0
ENG IEERING DESIGN HANDBOOK SERIES
Tkw Engineering Design Handbook Series is inzended to provide a com21latio= of principles And fundamental data to
in assstng engineers ir the evolution of new designs which will meet tactical and technlcai
supplement experiem
needs while also embod)ing riatlsfacto.y prod.ibillty and mainainability.
ULo.ed . elow are the hlandbooksw.b hale been publishedor submitted forpublicatioh. Handooke wnh prbLicatlon
dates pr'ior to I August 196Z were published as Zrerles Ordnance Corps pamphlets. AMC Circular 31038, 19 .Tuly
1963. redesignated those publications :,a706series AMC pamphIt, (i.e., ORDP 1:0138 was redesignated AIACP 706138); All new, reprinted, or revised nandbooks are being publr.hcd as 706series AMC pamphlets.
General and Miscellaneous Subjects
Title
Number
Elements of Armament Engineering, Part One,
106
Sources of Energy
Eleents of Armament Engineering. Part Two,
107
Ballistics
108
Elements of Armimenl Engineering, Part Three,
Weapon Systems and Components
lie
Experimental Statistics, Section 1, B.sic Concepts and Analysis of Measurement Data
III
Experimental Statistics, Section Z, Analysis of
Enumerative and Classificatory Data
112
Experimental Statiytics, Section 3, Planning and
Analysis of Comparative Experiments
Experimental Statistics, Section 4, Special
113
Topics
114
Experimental Statistics, Section 5, Tables
Maintenance Engineering Guide for Ordnance
134
Design
Inventions, Pztents, and Related Matters
135
Section 1, Theory
136
Servomechaennsm,
Servomechanisms, Section 2, Measurement
137
and Signal Converters
Servomechanisms, Secdn,3, Amplification
138
Servomechanisms, Section 4, Power Elemento
139
and System Design
17G(C)
Armor and Its Applicatim to Vehicles (U)
270
Propellant Actuated Devices
290(C)
WarheadsGenera. (U)
331
355
Compensating Elements (Fire Control Series)
he Automotive Assembly (P.Atomotive Series)
Ammunition and Exploclves Series
Solid Propellants, Part One
175
176(C) Solid Propellants, Part Two (U)
Properties of Explosives of Military Interest,
177
Section 1
1781C)
Properties of Explosives of Military Interest,
Section 2 (U)
21Z0
Fuzes, General and Mechanical
211(C)
212(S)
213(S)
214(S)
215(C)
244
245(C)
246
~acteristics
247(C)
248
249(C)
Fuzes. Proximity, Electrical, Part One (U)
Fuzes, Proximity, Electrical, Part Two (U)
Fuzes, Proximity, Electrical, Part Three (U)
Fuzes, Proximity. Electrical, Part Four (U)
Fuzes, Proximity, Electrical, Part Five (U)
Section 1, Artillery AmmunitionGeneral,
with Table of Contentq Glossary and
Index for Series
Section 2, Design for Terminal Effects (U)
Section 3, Design for Control of Flight Char
Ballistic Missile Series
Number
Z81(SRD)
282
284(C)
Z86
Title
Weapon System Effectiveness (U)
Propulsion and Propellants
Trajectories (U)
Structures
Ballistics Sries
Trajectories, Differential Effects, and Data
140
for Projectiles
Elements of Terminal Ballistics, Part One,
160(5)
Introduction, Hill Mechanisms. and
Vulnerability (U)
Elements of Terminal Ballistics, Part Two,
161(S)
Collection and Analysis of Data Concerni=g Targets (U)
162(5RD) Elements of Terminal Ballistics, Part Three,
Application to Aissile and Space Targets(U)
(Xrriages and Mounts Series
Cradles
341
34,
Rectil Systems
343
Top Carriages
Bottom Car iages
344
Equilibrators
345
Elevating Mechanisms
346
Traversing Mechanisms
347
Materials Handbooks
301
302
303
305
306
3C7
308
309
310
311
Aluminum and Aluminum Alloyc
Copper and Copper Alloys
Magnesium and Magnesium Alloys
Titanium and Titanium Alloys
Adhesives
Gasket Materials (Nonmetallic)
Glass
Plastics
Rubber and RubberLike Materials
Corrosion and Corrosion Protection of Metals
SurfacetoAir Missile Series
Part One, Syttem integration
291
Part Two, Weapon Control
292
293
Part Three, Computers
294(5)
Part Four, Mibsile Armament (U)
295(S)
Part Five, Countermeasures (U)
296
Part Six, Structures and Power Sources
Part Seven, Sample Problem (U)
297(S)
Section 4, Design for Projection (U)
Section 5, Inspection Aspects of Artillery
Ammunition Design
Section 6, Manufacture of Metallic Cornponenta
of Artillery Ammunition (U)
______
*HEADQUARTERS
s
S
L.1,M
LuL
SlILUNITED) STATES
ARMY MATERIEL COMMAND
WASHINGTON 25, D.C.
30 September 1963
AMCP 706286, Structures, forming part of the Ballistic Missile
Series of the Army Materiel Command Engineering Design Handbook
Series, is published for the information and guidance of all concerned.
(AHCRD)
FOR THE COMY14DER:
SELWYN D. SMIT , JR.
Major General, USA
Chief of Staff
')FFICIAL:
R. 0. DAYONC
Colonel,
..
qp
Chief, Adiistrative
DISTRIBUTION:
Office
Special
WOEACE
"Tishandbook has been prepared as one of a series on the design of
ballistic missiles. It presevis selected data on the properties of materials
of construction including the effects of elevated temperatures and other
operational conditions on ballistic missile structures. ParticuJar attention
has been given to the stress analysis of thin shells. in addition the special
design considerations applicable to ballistic missile structures are discussed..
The handbook was prepared for the Office of Ordnance Res':cn,
Ordnance Corp3, U. S. Armyn. The text and illustrations were prepared by
Vitro Laboratories under contract with Duke University, wvth the technical
assistance of Army Ballistic Missile Agency and the Special Projects
Branch of Navy Bureau of Ordnanve.
4!l
STRUCTURES
CONMEtTS
Page
Chapt7.
1
1 . P urpcse and Scope ...................... .......................
2
12. Ballistic Missile Structural Considerations .........
13. Elevated Temperature Effeci ........
.
2 Properties of Materia
..
..
.....
.......
..
.............
21. Mechanical Behavior ...
ior
22. Creep e
23. Heat Transfer ...
................
11
13
.......
24. Selcted Data an Properties of Materials of Construc16
tion 38
25. References and Bibliography ........................................
3 Stress Analysis .. 
..


41

41
31. General Considerations ...............................................
32. Stresses and Deformations in Thin Shells Under Uniform Internal Pressure ............. ..................
...41
33. Disccntinuit Stresses in Thin Sheis
42
34. Buckling of Thin Circular Cylindrical Sheic
44
35. Buckling of Thin Spherical Shells
46
....
46
36. Thermal Stresses ..............................
37. Thermal "Buckling
...
48
38. Creep Effects ..................
5i
39. References and Bibliography............
55
57
4 Design Considerations ...............................
57
.....................
41. General Comments .....................
42. Aerodynamic Influence on External Configuration
..
57
43. Aerodynamic Heating ............................ 58
.....
62
_...........
...............
44. Design for Applied Loads .....
66
......
....... ..........
.......
45. Structural Failure ..............46. Time and Temperature Dependence of Design Criteria 69
70
47. Miscellaneous Factors Affecting Structural Design ......
73
48. References and Bibliography ........................................
11 7'
ndex ......
_ _.
. . .............. 75
STRUCTURES
LST OF ILLUS1
Figure
T1ONS
T;tle
Page
21.
Typical Tensile StressStrain Diagrams for Steel and
Aluminum Alloys
22. DtAtortion of an Element by Normal and Shear Strains
23. Fatigue Leading and Typical StressCycles to Fa~iure
.
(SN) Curve ....
24. Typical Presentation of Creep Dat
.
25.
26.
8
9
11
12
Effect of Temperature on Modulus of Elasticity for Several Typical Materials._ _
Effect
of Temperature on Yield Stress for Sevewal Typical Materials
26
27.
Effect of Temperature on W21ghtStrength Ratio for
Several Typiczl Mterials ....
27
28. Effect of Temperature on CostStren th Ratio for 1everal Typical Materials_
_
23
29. Compressive StressStrain Curves for 2024T3 Aluminum A31o. Clad Sheet Stock at Various Temperatures..
28
210. Compressive StressStrain Curves for 177 PH (TH
950) Stainless Steel Sheet Stock at Various Temperatures
..
...
29
.....
11. Tangent bodulus Curves for 2024T3 Aluminum Alloy
Alcad Sheet in Compression..............
...
29
212. Stress for 0.2" CreepSet of X2020T6 and 2024T6
Aluminum Alloys as a Function of Time and Temperature .....................................
..
...
.....
30
21.. Rupture Stress for 2024T6 Aluminum Alloy as a Function of Time and Temperature ............... 30
214. StressRupture and Creep Characteristics of X2020
Aluminum Alloy at 300F .........................................
31
215. Tension Creep Curves for 2024T4 Aluminum Alloy..
32
216. Creep Rate and Time to Rupture Curves for 3SH18
Aluminum Alloy .................... _.............................
.... 33
217. Master Creep Curves for 2024T3 Aluminum Alloy
Clad Sheet Stock ..............................
........................ . 34
218. Fatigue Properties of 2024T4 Aluminum Alloy (Extruded Bar) at Elevated Temperatures ..... ......
34
219. Thermal Conductivity of Selected Materials .................. 35
220. Specific Heat of Selected Materials................................... 36
221. Mean Coefficient of Thermal Expansion for Several
M aterials.....................................................
.... ......... 37
Qt
Vf
$VflUCTIJRES
IST OF ILLUSTRATIONS (WCmtised).
Title
Page
Figure
31. Theoretical Deformations of Pressure Vessel under
42
Membrane Stresses Alone....
32. Defornations of Thin Circular Cylindrical Shell under
42
Beriding Stresses
33. Determination Of Tangent Modulus and Secant Modur
46
lus from StressStfain
47
34. Bar Heated between Two Insulated Walls
35. Assumed Temperature Distribution in Long Rectan47
gular Flat Plate
36. Ring Frame with Temperature 'Varying Linearly
48
through Depth
37. RingReinforced Circular Cylindrical Shell 49
38.
39.
Simply Supported Plate Subjected to Thermal Stresses
.
.
. .
49
in One Direction ..
Assumed Stress Distribution in Rectangular Flat Plate
50
with Free Edges ...
...
310. Tensile Creep Curve
311. Deflected Column
312. StressStrain Curve at Elevated Temperature ...
41. Effects of Aerodynamic Heating on Model Nose Cone,
Stagnation Temperature 1995OR 42. Effects of Aerodynamic Heating on Model Nose Cone,
Stagnation Temperature 2030'R ................
3, Tank Section, Jupiter Missile ......................
44. Top Section, Redstone Missile .....................................
51
53
54
59
59
4
65
STR UCTURES
LST OF TABLES
Tabie
Title
Page
____I__1
11. Principal Notations
21. Mechanical Properties of Some Typical Alloy Steels.. 18
22. Mechanical Properties of Some Typical Aluminum Al19
Ioys..__
23. Mechanical Properties of Some Typical Alloys of Magnesium, MlyJbeenum, and Niobium
24. Mechanical Pioperties of Titanium and Some of its Typical AllIoIyb
25. Mechanical Propertien of Some Typical Cermets, Metallic
.21
Refractorie and Ceramics
26. Physical Properties of Some Typical Materials 27. Approximate Costs of Structural Materials (1955)______
28. Mechanical Properties of 2024T6 Aluminum Alloy Products at Varicus Temperatures and after Heating ........
41. Weight Breakdown for Redstone A Missile .
vii
20
21
22
23
24
72
U'
I
~A'
(9
'fifE
C)
tapter I~
INTROD UCTICN
14. PURPOSE AND SCOPE
'TABLE 11.
PRlCli(zL O y.kmJ
The information presented in this hand
)

book has been selected to convey ,an appreciation of the problens whicai face the missile
sLructural design 4pecialist and to provide
a basis Vir understanding the considexations
which leal to the final sttuctural coafiguraion. Since a complete o'itline ,ofthe theory
and aractice of tructm'al lesign.and analysis
can(, be included in a single volume, the
more elementary coniderations have been
omitted. It is a.sumed that the reader has
a technical background .and a reasonable
knowledge of the fundamentals of the
strength of materials and structural principes. However, the 'book has been designed
for the use of graduate engineers ixth little
or no direct experience in this field ather
than for the experienced structural design
engineer or structural analyst.
Some of the problems facing the ballistic
missile structual design engineer are outlined in this introductory chapter. In Chapter 2, Properties of Materials, elementary and
basic considerations in the behavior of materials are introduced, mechanical, creep and
thermal properties are discussed, and data
on physical properties are presented. Chapter 3, Stress Analysis, gives formulas of
stress and strain for basic structural elements under selected lods and boundary
conditions. Preliminary stress calculations
can be made by the use of the information
presented in these two chapters. In the final
chapter, Design C(.nsiderations, some of the
problem facing the ballistic missile design
engineer are presented and various possible
ten,.th oflat plate
Crosssectinal area
Radiatiug :area, surface area and
area normal to heat flow in
radiation, convection and conduction modes of heat transfer, respctively
Width offlat plate
Distance to extreme fibers from
neutral plane in a pla'e
Brinell hardness
Constant, creep laws
British thermal unit
Thermal capacity
Distance from neutral axis to extreme fiber
Constant, creeup laws
Temperature, Centigrade scale
Diameter
Internal energy
Young's Modulus
Tangent modulus
Secant modulus
Effective modulus
Temperature, Fahrenheit scale
Compressive yield stress
Shear modulus
Torsional rigidity
Heat transfer coefficient
Beam deptht
Shell or plate thickness
Moment of inertia
Thermal conductivity
a
A
B
Btu
c
C
C
D
E
E,
E,
Em,
F
F1.
Q
GC
h
I
k
Constants, creep laws
solutions are discussed.
k j , k2
The principal notation used in this volume
"sshown in Table 11.
This volume was w.ritsen by S.V.Nardo and N.J.
K
L
Constant
Lngdn.
.,off, Polytechnic Institute of Brooklyn, Brooklyn,
N. Y.. and edited by C. D. Fitz, Vitro Laboratories,
West Orange, N. J.
Bending moment
Constant, creep laws
Numbei of stress cycles
Constint, creep laws
STRUCIURES
TABLE 11. (Continued)
12 BALLISTIC MISSILE STRUCTURAL
CONSIDERATIONS
Critical pressure
End load on column
Rate of cooling
Rate of heat generation in elemental volume
Rate ,.f hett flow
Cornstant, creep laws
Quantity of heat
Q
Muliiplying factor
r
Rockwell hardness
R
Radius
Rc,, R~, R, Stress ratios, interaction formulas
PTemperature, Rankine scale
Room temperature
R.T.
Thickness
t
Time
t~,Critical
time
T
Temperature
u, v, w
Displacements in x, y, z directions
respectively
VVolume
Shear force
wSpecific weight
x, y, z
Space variables
yDsncfrm
euriaiina
The structural designer of a missile or an
airplane h&, much les,, freedom in selecting
the external shape of his design than the designc'r of a bridge or building strufcture. The
form is largely prescribed for him
other
requirements. As a consequence his
by
tzrk is a more difficult one. The designer hast
to work within the space available and inust
ensure that sufficient strength is provided
with the minimum amount of weight. The
conservation of weight is of particuiar importance in ballistic miss"le
111_ ue to the magnitude of the growth factor. This factor is
the additional weight of 'the fuel anid struetural rnaterial requiredi by the addition of
one pound of n)ay!oad to the 3riginal payload.
The most important difference between the
structure of a ballistic missile and mrost other
structures is that the reentry portion of the
tyliss~le structure must be designed to withstand high rates of hcatinir This aerodynarnic lheatng raises tl~e temperature of the
structure and influences adically its loadcarrying capacity. Of course, aerodynamic
healing is present in all supersonic aircraft,
but it is h~i the ballistic missile that it a9ttains
its maximum importance.
For this _7eason a new parameter, namely
4 enperaturL, enters the analysis of the structure. The elevated temperature, in turn, in
GREEK VITTTERS
Thermal diffusivity
a
Coefficient of thermal expansion
Normal strain
e
Emissivity
Half ape;: angle of cone
ly
Constant, creep Jaws
vPoicson's Ratio
Rate of radiant energy emissiop or
absorption
Radius of gyration
p
stress
a'Normal
StpfaiiBoltzmann constant
Criti&. normal stress
Shear st;ess
TlCritical shear stress
must also be considered carefully in the
analysis of missile structures. While 'the
structure of a conve ntional groundbased
subsonic airplane in ,irindcpie lasts for an
almost indefinite per.od, the hot sti uctural
ukuremets of missiles change their shape continuously under load until they become 'o
deformed that they are incapablie of performing their structural duties.
Every structuare that has to perform at
very high temperature has, therefore, a definite lifetime beyond wh~ich it cannot be
utilized. The consideration of this lifetime
muakes the task of designing and analy:,sing
a missile strticture much more complicated
than similar iasks Performed itn connectidn
with more conventional structilres, T~te "one
Constant, creep lav.s
Prefsure
Cceflcient of first harmonic of
thermal stress distribution
external
P
q
INTRODUCTION
shot" nature of the ballistic missile intro.
duces n design philosophy which cannot be
tolerated in other structure..
structures and the thermal stresses of nonuniformly heated structures.
Thermal stresses are of the nature of the
secondary stresses of conventional structural
engineering. They are a consequence of the
deformation. and not of the externally applied loads.
is unlikely
they will
cause
failure
except,It perhaps,
in fatigue.
However,
13. ELEVATED TEMPERATURE EFFECTS
13.1. Effects on Material Properties. The
most immediate effect of elevated temperature s the deterioration of the loadcarrying
failur excelt
e
p
a
fati
e
H er,
he material.
ateial.Man
tets
ave
they
may
cause
unduly
large
plastic
deformacapacity of capaityof
the
Many tests have
tions and, in particular, compressive thermal
been carried out to obtain detailed informastresses might cause buckling.
tion on this effect. The collected data indiThermal buckling may lead to the developcates the deleterious effect elevated temperaof bulges on the surface of the body or
ment
properties
a material's physical
which aftect the rigidity, deformation, ulticontrol surfaces of the missile which coild
interfere seriously with the aerodynamic
mate strength, fatigue and buckling of structhe accuracy of
particular,
performance.
for
that
shows
also
tural members. The data
be ....., ....... For these
m,a,
pathIn
figl
4he
........
the...........
every structural material there exists a maxithe the missile
miigit structural
su
r designer
d
and
reasons
mum temperature beyond which the loadanalyst must be conversant with the calcucarrying capacity is reduced to such an
lation c' thermal stresses and with the
extent that the material cannot be advanA
I
lormuis predicting thermal buzkling.
tageously utilized.13.3. Croep and Creep Buckling. A third
13.2. Thermal Stresses and Thermal
effect of elevated temperatures,
important
of
effect
Buckling. The second ;mportant
little known in conventional
comparatively
of
aerodynamic heating is the development
is creep. This phemachinery,
iterasteesi
teeracture.
risen
in
ahetdrdisujcetoaosantnil
and
structures
in
Even
structure.
internal stresses in the
instance, when
for
observed,
be
can
eometuericnomenon
of uniform
hecase
tensile
constant
a
to
subjected
is
rod
heated
a
cnstructural element, externaltureeric
Under this load the rod elongates conmnaxirnaof different
eomdieetiateonload.
strauctsrora a combination
iusywthie.lhogladfrex
matestraints
tiiuously with time. Although lead, for examrials will induce it ternal stresses. T'hese
pi., exhibits creep at room temperature, with
temperature
called
stresses are sometimes
employed structural materials the
commonly
problem,
practical
every
almost
s!resses. In
p enomenon is practically noneistent at
however, uniform temperatures in a heated
becomes important,
temperature.
unlikely to o room
rather
structural element
strutura are
re
eleent
rtherunliely
wever,
aial ma Creep
even lead to the elimnina.e in
tera to
ma
ic
a
of
I
the
that
are
reasons
occ,.r. Two principal
rticular material from ie in
tion of a
aerodynamic heing itself is nonuniform
structures, when the temperature is raised
over the exterior surface of the structure,
sufficiently.
and that the heat capacity of the structure
first structural effect of creep is the
v e fThe
varies from point to point. In consequence of
deformation created. It is important that the
analyst be
the nonuniform temperature distribution,and
the
material
would
expand
nonuniformly
msiesrcua
einradaaytb
deformathe
when
able to predict the time
y
ia ossibl geoetrcally.Asuthe
weeth
were this possible geometrically. As the nattions will become so large that they substanural 4eformations of the material are incompatible, internal stresses are set up in the
tially change the aerodynamic shape of the
missile or that they interfere with the proper
non uniformly heated st.uctural elements to
parts of the
eemeableof the
functioning
reestablish the continuity of the deformasystem.
as
tions. These stresses are known properly
thermal stresses. However, in this section the
A second effect of creep is the occurrence
term thermal stresses shall be 'used for both
of two entirely new types of structural instathe temperature stresses of uniformly heated
bility. The first is instability in tension. When
STRUCTURES
a straight bar is attached to a rigid structure
at its upper end and is loaded by a weight at
its lower end, the bar elongates continuously
with time. As creep deformations of metals
take place substantially without any change
in the total volume of the material, ;. one
percent increase in the length of the bar
must be accompanied by a one percent decrease in its crosssetiona area. However,
action of these bending moments the column
deforms in creep bending. The increasing.
curvatures in turn lead to increasing lever
arms which again result in increasinir hpnd
ing moments. This is a vicious circle which
can end only in the deveiopment o very large
curvature and compiete collapse of the
column. When the behavior of the materiA is
such that the creep rates increase niore rap
constant load applied to a !:.aller crosssectional area causes an increased stress in
the bar which in turn increases the rate at
which creep deformations take place. The
creep xtrain rate is generally a nonlinear
function of the stress, increasing much more
rapidly than the stress. The consequence is
a very rapid acceleration ,f the creep proce;s
which can end only in a reduction of the
crosssectional area to the value at which the
material will rupture instantaneously under
the applied load. The time at which this rupture takes place is known as the critical tinie.
This characteristic exista for all structural
materials at elevated temperatures and must
be known to the structural designier.
A second instability of importance is
known as creep buckling. This type of buckling differs ini many respects from elastic and
inelastic buckling as these phenomena are
known to the structural engineer. The main
idly than thbnn~r, 
difference is that in creep buckling one cannot utilize the concept of a efriical load.
Buckling takes place under any compressive
load however small if one waits long enough
after the Application of the load. On the other
hand, jL.St as in the case of tension, a new
concept must be introduced, namely that of a
critical ime. An explanation of this phenornenon can be givea without ditlhcuity.
Every practicai column, as distinguished
from the 'dealized ones underlying some
theoretical calculations, is imperfect. This
means that the column's axis is never coinpletely straight and the load is never applied
therefore take the necessary rrecautions in
this regard. Some of these precautions are of
an entirely aerodynamic nature; if the contour.; of the missile are properly shaped, the
heating that the missile will experience can
be minimized. A second precautionary measure is more in the realm of the structural
designer; it consists of provisiorr for thermal protection through shielding the structural materivl with insulation, through cool
ing the structure by means of the evaporation
of a coolant, or. through other means.
In some designs the external metal covering of the missile is used as a heat sink. When
to it completely centrally, Consequently the
column is subjected to bending motmenii. A
bending moment can be calculated as the applied !oad multiplied by its lever arm, or the
distance of the centroid of the crosssection
from the line of load application, Under the
this is the case, information is reqLired azout
the thermal conductivity, the thermal diffusivity, and the specific heat of the material.
The form.r quantities determine how rapidly
heat will 1.e transmitted from the surf;)e of
the missile Vxits interior, while the last item
the time
_
necessary for creep buckling to occur is eways finite (not infinitely long). Consequently we can properly talk of a criticP'1*  e .
Depending upon the material, the 1oda, . 4nd
the temperature, this critical time may be a
few years, a few months, a few days, a few
minutes or even a few seconds. It is therefore
most important that the engineer be conversant with the facts of creep buckling
when designing missile structure.
13.4. Melting and Ablation. Other phenomena which must be treated, if a complete
picture of the problems of ballistic missile
structures is to be give,, are even farther removed from conventional structural analysis.
The heating rates experienced by ballistic
missiles on reentry are so large that the
surface of the material may melt, sublime or
even burn. The designer of the missile must
'
INTRODUCTION
indicates how much heat can be stored in the
material,
Reentry has also been accomplished suecessfully through 1eting ithe surface of the
missile melt and the molten metal be carried
away by the windstream. If sufficient material is available in the structure, the integrity
of the missile may not be impaired by this
melting. However, the rates of melting and
ablation must be calculated, and the designer
must also have information about the sta
_)
I5
bility of the process; in particular, whether
ridges or pits will be formed or whether the
material will be consumed uniformly over the
entire surface. Similarly it has been suggested that a missile miAght be allowed to
burnL during reentry just as in the case of ameteor. Again, a great deal of information
is needed on this problem before the designer
can employ burnifig as a means of maintaining the int.grity of the interior of the
structure.

.
~.,
I
K)
a
'WI
I.
I
II
(_I
~


~
.~
 ~
 

 
I
I
Chapter 2
PROPERTIES OF MATERtALS
!!of
21. MECHANICAL SEHAV*_
In the analysis and design of any structure,
the physical propertiesof available ms
are of primary importance. With relatively
few exceptions, structural engineers in the
past were concerned only with those properties which might be classed as purely mechanical. 'ertain elastic constants, yield and
breaking strengths, and fatigue characteristics might be cited as examples. The relatively recent developments in jet and rocket
propulsive systems, highperformance aircraft, and missiles have greatly increased the
complexity of structural analysis and desig~n
problems,. Designers of 'today's highspeed
in the material properties already mentioned,
Data on the properties of some of the more
pertinent engineering materials used in mise design are collected and presented in
taii.lar and graphical form at the end of this
but also in the thermal, creep and plastic
characteristics. Indeed, it is certain that this
interest will extend to those physical characteristics which were heretofore in the domain
of solidscate physics.
A major reason for this extension of interest ir materiat properties is that ballistic
missiles are subject to high rates of heating.
This heating occurs upon reentry into the
atmosphere and (under certain trajectory
conditions) durinr the ascending phase of
the Bight. Unless the heat is dissipated the
temperature will rise. The loas! carrying capacity of the strecture is affected by the
temperatures
and also by the length
time at an attained,
elevated temperature.
a unit length to obtain the strain. When the
data are presented graphically in the form of
stress as a function of strain, the result is a
s ressstrain curve. Typical curves are shown
ip Figure 21 for steel vnd aluminum alloys.
Before the information derivable from
these curves is discusseul, a brief review of
the concepts of stress and strain will be presented. Stress is the intensity of the force per
unit area acting on a plane. This property is
a vector quantity; the normal component of
the vector is the normal or direct stress, and
the component in the plane, the sfr ar .stress.
In a cylindrical test specimen, the quantitj
obtained
by dividing
the
force
by actthe
crosssectional
area .is
the total
normal
stre.szs
As a genera) rule, it may be said that elevated temperatures have a deleterious effect
on the mechanical properties of materials,
For example, the modulus of elasticity, yield
stress, and the ultimate stress usually decrease with increasing temperature. It will
also be se, n that the rupture stress in
fatigue. or the number of cycles to failure
generally decreases with an increase in
temperature.
ing on a plar.e perpendicular to the axis of
the cylinder With a material that can be
considered homogeneous and isutropic, and
with a cen'.rally applied load, this definition
is reasonzbly consistent and accurate. Inasmuch as this normal stress is the only stress
acting in this simplified case, the stress condition is said to be uniaxial.
The stress picture is accompanivd by a
strain picture with normal and shear strains
21.1. StressStrain Curves. Probably the
mechanical data most familiar to engineers
are stressstraifn curves. This information is
obtained experimentally by testing a prescribed standard specimen of the material in
tension or compression and simultaneously
recording the load and elongation, in thsimplest of thene tests, a cylindrical *pecimen is loaded in tension ;r :atesting machine,
the observed load is divided by the original
s'.ess, and the elongation is measured over
STRUCTURES
developing at each point in the material.
Imagine that parallel planes are passed
through two neighboring points. The normal
strain is the relative displacements of these
points normal to the planes divided by the
Figure 22 illustrates the separate effects
of normal and shear strains on the deformation of a square element of the material. In
part (a) of the figure, the element is subjected only to normai strains. This type of
original distance between the planes. The
disrelativeparalas the
is defined
measured
two points
of the
"placement shear slrain
lel to the planes, divided by the original
deformation would occur, for example, if a
onlya
stress in
normal
.11vie the
that
element, were
anthe
thex direction
stress acting
strain in the y direction accompanies the
distance between the planes. Normal strains
have the effect of distorting a cubical element
of t.e material without altering the orthogonality between tie sides of the element,
whereas shearing trains are a measure of
the change from the original right angles
between the side.
strain in the x direction due to stress in the
x direction. Part (b) of the figure illustrates
the manner in which the element would be
distorted if it were subjected to a condition
of pure shear stresses. In the most general
case there would also be u displacements
(xdirection) of the end points of the element.
]1
iW
/PROPORTIONAL
0 I
MAXIMUM STRESS
YL
YIELD STRESS
IT
STRAIN
ALUMINUM ALLOYS
VEELD STRESS
UJ
'ROPOW'1ONAL
MAXIMUM STRESS
LIMIT
a:
//
,STRAIN
STEELS
FiSu.,e 21. Typical Tensile StjssStrnin Dogrumzl
for Steel and Alum'nuzi Alloys
PROPERTIES OF MATERIALS
~~dx
dx
du
[d
dyj
(a)
NORMAL STRAIN
du
SHE.A
dx
STRAIN
.
=dx
Figure 22. Disfortion of an Elernenst by Nosmal vdSIear Strains
21.2. Propart'onal Limit. Notwithstanding
the relotively simple nature of"the test that
yields the stressstrain curves of Figure 21,
some of the most useful and basic properties
of the material are deduced from it. It will be
noted that both curves exhibit an initial
linear relationship between stress and strain,
The value of the stress corresponding to the
end of this ~rroortina
linear
called the
hmt. relationship
rovied heis tresesye
spring constant or Young's Modulus can be
assumed to be independent of both pesition
and direction within the material. Clearly,
there0vre, this one physical property is of
major importance to the structural designer
when the design criterion is a rigidity
requirement.
2I.4. Y'ield.
Consider a simn!;'e tensile
specield.
im limit
ite to
specimen
loaded Conde
beyond the aelastic
proportionallimit. Provided the stresses remain below thel
,
a s r s
o r p n i g t
on
n F g
a stress corresponding to point B in Fiure 21a. Upon the reo d of the load, the
rial remains elastic, i.e., it assumes its origidecr;ses approximately linearly with
stress
load.
the
of
removal
the
upon
nal shape
line BC because
by deformations
as illustrated
strain
It
Furter,
he
mteril
i
linarlyelasic.i
elastic.
ne
elastic
part of the
canonly
be
linearly
is
Further, the material
might be well to point out that in reality, t
nereaic par t
e
tiosc
n e
material canmateialcanbe
be stressed
the roporsresed above
bovethe~i~orregained. A permanent set is observed when
tionai limit and still remain elastic. However,
the load is completely removed. The unloadfor most engineering materials, the proporing line BC is approxi.: ely parallel to the
straight
linelus.OA
whose
slope is equal
to
tional and elastic limits are frequently
relatively
close
s
O If
f he
e pr
per.
miaterial were
the sltei
Modulus.
Young's
used
o each other and are in fact
interchangeably by engineers.
fectly elastic, it would have unloaded along
the original loading line OAB. For materials
21.3.. Young's Modulus. The slope of the
stressstrain curve, that is the rate of change
of the stress with increasing strain in the
linear region is calied Young's Modulus.
Physically, it may be considered a spring
constant because Young's Moduius is proportional to the force per unit elongation.
Although the stressstrain curve gives the
value of the spring constant in only ornu direction, most engineering material, Are fairly
whose stressstrain curve has the charocteristics shown in part (a) of tne figure, t:p
yield strpss ii defined a$ that value of the
stress c.arre.ipwndirg to an offnet, c,, of 0.002
ur materials whose tressstrain curves are
similar to that shown in part (b) of the figure, the first point of horizontal tangency is
taken as the yield point.
21.5. Ultimate Stress. The maximum value
homogeneous and isotropic, and hence the
of the stress reached in the simple tensile
STRUCTURES
test is called the ultiniate sirfs. Since all
stresses are . sed on the original crossectional area of the specimen, the ultimate
stress corresponds to the maximum load, and
it may be considered as indicaive of the
strength of the material. Relatively few
members in airplane and missile structures,
however, are critical in tension. The ultimate
strems of a material in compression is net
quite as clearcut because of the column
action. Even when the column action is prevented by lateral constraints or by the proportions of the test specimen only some
materials such as wood and stone will exhibit
a definite value of the ultimate stress. For
most metals, an arbitrary criterion must be
used to define this quantity. When the design
criterion is associated with failure or rupture
t
SL
ratio of the lateral strain to the axial strain
in a uniaxially stressed material. As mentioned earlier, a cylindrical tensile test specimen subjected to an axial load produces an
axial strain. This axial strain is always accompanied by a lateral strain smaller in
magnitude, and opposite in sign [Figure
22(a) ]. For most engineering materials
Poisson's Ratio has a value of approximately
0.3.
stress assumes major iUmprtance as a material propeyty.
21.6. Tarigent and Secant Modulus, Modu21.of TaRgetIndsestucatMualsnalyses
lus of Rigid. In some structural analyses
(column buckling, for example), a socalled
tangent modulus is used. This quantity is
simply the slope of the stressstrain curve at
any point. Below the proportional limit, the
tangent modulus is everywhere equal to the
Young's Modulus. Some investigators use the
concept of a secant modulus, which is the
ratio of stress to strain at any point. Young's
Modulus and the secant modulus are also
equal below the proportional limit. Torsion
stressstrain diagrams can also be plotted
from torsion tests on round tubes or round
solid sections. The initial slope of this curve,
analogous to Young's Modulus, yields the
modulus of elasticity in shear, referred to
as the Modulus of Rigidity.
21.8. Fatigue. The physical properties of
materials discussed so far have been based
on static tests in which the load is slowly applied without repetition. It is well known that
some structural elements are subject to vibrations which cause the stresses to fluctuate.
Under these conditions, failure can occur at
a stress level which may be considerably
lower than the ultimate stress previously discussed. This type of failure is known as
fatigue failure.
The fatigue strengtl. of a structural element is probably the least amenable to analysis and hence the most difficult of the various
design strengths to evaluate. Unfortunately,
its determination is not confined !o a knowledge of the nature and magnitude of L%,
fluctuating stresses and some simply physical
properties. it is known that the "atigue
strength of a specimen is affected by size,
local discontinuities, and heat treatment, to
mention a few of the parameters. Investigators in this field have nevertheless made significant strides by their analytical studies
and extensive experimental programs.
The data used by design engineers in determining fatigue strength are based on tests
cf standar.1 specimens (smooth or notched)
subject to controlled fluctuating stresses, in
21.7. Poisson's Ratio. For homogeneous,
isotropic materials, the Modulus of Rigidity
is related to Young's Modulus in the following manner:
(21)
G = E/2( v)
where
E is 'Young's Modulus
G is Modulus of Rigidity
v is Poisson's Ratio
Pqisson's Ratio is the absolute value of the
which the number of cycles to failure is recorded. Part (a) of Figure 23 shows a periodic variation of stress to which a fatiguc
specimen may be subjected and defines the
important stress parameters affecting fatigut
life. In part (b) of The figure, a characteristic
curve is presented which gives results of
fatigue tests on a specimen. The curve is
drawn for a mean stress value of zero, and
relates the stress to the number of cycles to
oh
suctuall nirner, then the ultimate
10
f
a
PROPERTIES OF MATERIALS
4"failure. Curves of
with
this type are commonly
q;aued SN diagrams. The value of the greatest. stress which the material can withstand
for an infinite number of cycles without failurt is called the endurance limnit. This is a
simple and admittedly crude measure of the
fatigue streftgth of the material. If the endurance limit so defined cannot be detected
a reasonable amount of testing, a practical endurance limit is often introduced as
the stress at which 107 load cycles do not
cause failure.
For the case of zero mear, stress, it will be
noted that the mnaximnum ntress is equal to the
stress amplitude. The maximnumr and minimum stresses are equal in value, )pposite in
sign, and correspond to) a case of completely
reversexd loading. When the mean stress is
other than zero, the stress amplitude coreesponding to fatigue decreauses with inereasing mean stress. Trypical curves at room and
elevated temperatures are presented in the
literature.'4
*Superscript numbers refer to the bibliography
at the end of thi chgpter.
OTrEAN
U)
ENDURANCE LI MIT
TIME
NUMBER OF CYCLES TO FAILURE (LOG SCALE)
Figure 23. Fatigue Loading and Typical Strevs. Cycles to Failure (SN) Curve
21.9. Other Mechanical Properties. Of
great importance to the missile designer is
the plastic behavior of materials. Plasticity
is related to the ability of a material to sustamn a permanent deformation without fracture, A perfectly plastic material, for examFle, would require no increase in stress to
produce a continuing plastic deformation
oree [z 7ipld stress was reached; it would be
repr;saiid by a segment of the stress axis
and a horizontal straight line on the stress,.strain diagrini. The importance of plasticity
Ican
be illu.,iialrd by copsidering a structure
which is su~ije, lowaly to a high stress level,
If the material is ductile, plastic flow occurs,
and the load is ridistributed more uniformly.
Neighboring points in the material absorb
the load which is shed at the point of stress
concentration, thus lowering the intensity orA
the maximum stress and lessening the possibility of early failure. A brittle material does
not readily flow plastically and hence there
is the danger of brittle fracture. Quantitative
measures of the ductility of a material are
the percent elongation in 2 inches of a test
specimen and the reduction of area at fracture in a tensi or test.
The hardne,, of a material is a measure of
its ability to iesist permanent surface deformation under applied concentrated lowds.
There are several standard methods of incasuring tis quantity. Values of hardness for
several materials are given in the tabies at
the end of this chapter. Approximate corrlation between the various hardness mcales
are given in t2~e literature." It can be shown
that for many steels thie tensile stftength is
proportioial to the hardness.
22
RPBHVO
22.CE
BAVO
Creep may be defined as the contirwing
deformation tWat occur's in a material subjected t(; a constant stress. At room temperatures, the cre4op of engineering mat!rals is
unimportant, but at eievated temperatuires
STRUC URES
creep effects mnay yield important design
criteria.
Creep data are generally presented in the
form of st:ain as a function of time as shown
~EL
in Figure 24. Specimens are subjected to a
I____________
conslant load, which cozresponds to a con.
stant stress level based on the original cross
STRAIN
JPRIMAR\

sectionl area of the specimen. Thie instantaneous value of the strain, t. corresponds to
the elastic deformation calculable from the
load. original crosssectional area and
Young's Modulus. With the passage of tLime,J
the strain increasts; the strain rate is high
at first and gradually decreases to a constant
value. The portion of this creep curve inI
RGE
__
STAGV
which the creep rate is changing is called the
pi inzary stage or primary e 'eep. The portion
corresponding to a constant ate of strain is
called the seeondary stage, or secondary
creep. In the tertiary stage, the strain rate
suddenly increases until the specimenSAG
ruptures.
Removal of the load during the secondaryj
stage results in the recovery of the elastic
SEODR
STNAE
strain r, and sometimes also of that portion
of the creep strain it,obtained by extrap~iating the straight line of the secondary creep
stage to t = 0. Some solid state physicists
attribute this to a release of internal stresses
which occurred during the loading and creep
processes. it should b,.. noted that the true
physical nature of cr esp has not yet been
esZabhshed with any &gree of finality.
Elongats i rn the specimen that occur
contraction of the specimen in the lateral
direction. This decreases the cros~ssectional
area and increases the value of the true
stress whici' is the loadO divided by the true
crosssectional area. This process acceleratesTIE
TERTIARY STAGE
RtP7URE
rapidly a.2d culmipates in rupture of thetTM
sppcinei when the greatly reduced crosssecticty e,, area can no longer 3upport the applied load. The portion of the creep curve
corresponding to increasing rates of creep is
the tertary stage. it should be mentioned
that met,,ilographic changes in the material
(for instance those caused by annealing during the test) can also lead to increasing rates
of
creep
Figure 24. Typical Presentation of Creep Ocita
PROPERT'SES OF MATERIALS
I!
A socalled f/ray body will reflect some of
th rain nryrcevd ec h misf agra boy wlldepend ulon
siv peer
23. HEAT TRANSSMl
Iiscustomar todstinguish threebasic
paii anoter
cond
(3)
ion hil
th
iectu
absorbed and reflected. It shouid be noted
energy ismntcmtl under
stying ththea
pheomeon
s osered o b c istnt ith
~certain physical laws pustulated by v;%rious
Th paticuar aw hic is
invetigors
tion to being proportional to ihe fourth
pow.?r of the temnperature, is also dependent
onhehycaprptisotetodad
varies with the wave length. It is usual to
applicable can be established by recourse to
experimental methods.onteweleg.ThcrepndgeqWhatever the wrid~ of heat trancfi~, the
miissile designer is bask'.ly interested in
ifUtmperzture distribution ei.her as a function
f sgae or of space nnd time. The temjperaure distribution is required to deterin
heat transfer rates, to make possible the cots,putation of thermal stresses, to deterinine
the change lin material physical properties,
or to cGP vith a problem ;Sinmefting, abla
Lion or burhing.
23.1. Radiatica. The transfer of heat
energy by radiation is probably best exerr
plifled by considering the transfer of heat
rom the sun to the earth. Thermal energy
F
Lrom a hot body, the sun, is converted into
__/ eleetromagnetic energy which is transritted
thrug spceto a cold body, the earth,
wharc iisreconverted into thermal eniergy.
The radiant energy reaching the cold body
can be absorbed, reflected or transmitted
throuf.h the body. Since most e;;gineering
materials are opaque, the tronsmiss~vity of
these materials is zero, hence the cf i.ctivity
and absorptivity only need be consiide.
The concept _f a black body is the limiting
case of a body with 100O/1 absorptivity and
zero reflectivity. For a given absolute temperature, no body emits or absorbs more heat
than a black body. The rate at kvhich the
radiant energy is emitted or absorbed by a
black body is given by the StefanBoltzmann
=rate
' diepdencTe ortespradint poer
tion for 1)for a gray body is written
(2 3)
'cAT,
4",
wherw t is the emissivity of the gry body.
The emissivity foi, some engineering materials 1.gi'ven in the literature.'"
Two final relations should be considered.
The first is that radiant interchange between
5i',res is governxed by the aiove principles
if there is a nonabsorbing medium between
the surfaces. Second, the radiant interchange
also dtepends upon tho.. iverse square of the
distanze between the surfaces and the relative orientation of one surface with respect
to another. If it is assumed that the radiation
fromn a heated surface has the same flux in
all directions, the orientation or "view" becomes purely geometrical in nature. Terv
factors are also given in the literature.'23l.2. Convection. 'The trainsmoission of heat
by tonvection (liffens materiilly from the
transmission (if heat by radiation. In the case
of convetive heat transfer, heat is conveyed
from one point to another by the movement
of a fluid Atream; i.e., ihe transport process
is &ssocidetd w~th the actual movemneat of a
fluid streain. Convective heat transfer may
bentta or forced depending upon whether
the fluid motion occurs by virtue of the nattendency of the he,.ted, lighter elerneit
10111
blower or othct means are used to produce
or abof radia.,L enryem~s
tine
hee
he fluid to Itsha,or pl
T bofy Bt (2 2)ac
so tionby
here that natural convective cooling of a surface area A a0 the b')n dary of a~ solid is
proportionail to (T',, T_) I %%here T,, arnd T..
a ~tefan13olt?mann
ongl th
constant
0.174 x 104 Btu1 ( ir)(Wt)(F')
radiating area, ft"
are the surface ind arnbiont fluid tempera
T =absolute temperature, F
13
4"
STRUCTURES
tures respectively. In practice, a linearized
approximation can usually be taken and tie
analytical expression for the convective cooling of a surface can be written as
where
q = rate of cooling, Btu/hr
h = heat transfer coefficiet,
Btu/(br)(ft )(F)
Strictly speaking, the convective cooling of a
surface is a combination of heat transfer
br th by coivectic n and conduction within the
fluid. The 'eader will note that the above
equation is similar to Fourier's basic heat
conduction Equation (?5). The reason for
its appearance here lies n a concept which
essentially converts the convective cooling
problem to a conduction problem. This concept considers that there is a film adjacent to
the surface throtigh which there is a linear
temperature drop from TW to T,. The effective film thickness is difficult to measure and
is incorporated in the heat transfer coefficient
h which also is a function of the physical
and dys'amical properties of the media
invoived.
23.3. Conduction. Of greatest interest to
the missile structural designer is the tkansmission ;f heat by conduction. This process
is generally interpreted as a simple molealar interchange of kinetic energy from one
molecule to an adjacent molecule by virtue of
elactic impacts. The heat flow is observed to
pass frurn the molecules at high temperature
to those at a relatively low:er temperature.
it has been custo ary ;n Lh azialysis of
the temperature distribution in a structure to
consider convective and radiant heat transfer
as the mneans by which the external air flow
heats the structure, and to disregard these
modes of beat transfer when the flow of heat
in the interior of the structure is sought.
to higher thermal stresses than those basd
on a consideration of all three modes. In a
solid structure the oniy mode f heat transfer
is conduction; in hollow structure, however,
and radiation can significantly
conv
Tmoify ion
the temperature distribution calculated
on the be
basis
of conduction
It might
mentioned
that alone.
a relativly
complete analysis of simultaneous effects of
conduction, convection, and radiation is possible through the use of computers.
The basic law of heat conduction has ts
origin in the fundamental work of the F.ench
mathematician Jean Fourier. Fourier's Law
states that the instantneous rate of heat
flow in a onedimen:lienuO booy is given by
the prduct of the area perpendicular to the
flow, the rate of change of temperature in
the direction of f!c' .nd the thermal
conductivity.
dQ
aT
(
q (25)
kA
t
In t above expresio,
Q =. uqntity of heat, Btu
t =time, hr
q =rate of heat flow, Btu/hr
k =therm co.iductivily, the a.nount of heat
per unit time flowing through a unit
area under a unit temperature gradient,
Btuin.ift2 hr F
A =area normal to heat flow, ft 2
T temperature, F
x =space zoordinate in 'tow direction, ft
The units uised are British Gravitational
units, and the minus sign is arbitrarily
chosen to make Q a positive quantity. In the
formi 2itown, the temperature T is.assumed
to bc a function of space and time (transient
conduction), although it may be a function of
space alone isteady conduction). Taking the
case of steady conduction in one dimension
and one for which A is not a function of x,
Fourier's Equation is
Two reasons can be mentioned in defense of
,k
("r,
this iimplified approach; first, the equations
of conduction have been solved rigorously for
many more cases than those governing the
o'her two modes of heat transfer; and second, the resuits obtained from an analysis
where
L = total path length (x2
T7 = temperature at x2
based on conduction alone always correspond
T, = temperature at x,
T
(2 6)
14
xi)
/
PROPERTIES OF MATERIALS
Cniefor example, a r~rd of unfo01rm
crmsettional area A and length L~. F r a
zontstt vlueof the heat flow q, the tern
assuming k, c. and w eomstant. Utder these
conditions, the most genefal heat conduction
equation redaes to
perature diffarence (Tz  TI) is inversely
13T
Of
eT
q
0
+'
.
a
+
proportional to the Ithermal conductivity k.
Thus amaterial of high thermal conduCtIvity
29
where(29
will allowi s given heat flow at a lower ternperature difference than a material of lower
k_
a=
rmal diffusivity ftl/hr
thermal conductivity. The missile designer
inteestd ineonfrcing
he eatinpu onthe
For systems that do not contain heat sources
snurface of a ntructur3 to the intrirwt
a or sinks, q =0, and Equation (29) becomes
minium tmpeatur
els atthe urfae,
he FoarierEquation. If sources and sinks areI
will choose a mnatearial hazvin~g high thermal
isa
T
present 4at Lkeg temperature iied,t=0
conductivity (assuing all other maicerial
properties are the samt).a
In the transient state of heat conduction,
an Eqain()rdcstPiso'
Equation. Finally, for steady conduction in
the quantity of heat entering and leaving a
systems free of sources and sinks, the wellvolurne element at any instant is not equal.
known Laplace Equation is obtained. This
The increase in internal energy, dE is
linear partialdifferential equation appears
frequently in many branches of science.
(2?7)
dE = c wVa dt
at
Solutions are discussed in the literature.
The thermal diffuiivity is a derived, ratewhere
rial property important to the missile dec thermal capacity of the rnaternal, th;_% signer as a measure of the ability of a
of heat energy the naterial
iraterial to absorb or diffuse heat. As defined
can absorb per unit weight for a unit
by Equation (29) thermal diffusivity is
temperature rise Btu/ib./*F
essentially the ratio of th. thermal conducw = specific weight of the material, lbs,/ft2
tivity to the specific heat. Thue tirse required
V = volume, 11,3
to bring a material to a specified temperature
level varies inversely with the thermal difA missile designer interested in using the
fusivity and directly with the square of the
structaral material as a heat sink, would

.
)amount
material of high thermal capacity,
choose~ a
conducting path length.
in three dimensions c~n be written
upon the material and the temperature rise.
again assuming that other material propertes were equal. The equation of continuity
fr the heat flow thrcugh a volume element
(2 8)
For the single dimension case the ratio of the
unit elongation per unit temperature rise is
called the coefficient of thermnal expansion.
where the effect of the heat input q, Btu/fthr, has been added. In the most general case,
c, w and q are functions of x, y, z and T;
q may also be time dependent. As written,
Equation (28) is valid for an inhomogeneous, but isotropic solid. Moreover, the equation is nonlinear if k, c, or w depend on T.
d
( 0
where
acoefficient of thermal expansiop I F
per unit length
~
Engineers find it more convenient to define a
mean coefficient of thermal expansion as
j
follows:
aTl
T
0
ax (k ax)+ a.(ka
O
+ ew
rk,
23.4, Thermal Expansion. When an engineering material is heated, it expands in all
directions in a manner that is dependent
+ q
OT
oz (I+
(I
by taking q as a function of T only, and
Man caes ar
f pactcalintres
coere
elongation
a

~dT(2
iI'l.
STRUCTURES
where T. is a reference temperature. The
efficient of linear expansion. Coefficients of
uvit elongailM can be easily calculated after
. is computed arid plotted ws a function of
temperature from
e=a=.
1.)
(2 12)
ternfor
a
is
required
elorgation
If the unit
perature rise other than from the reference
temperature, say T, to T., then
: a,..(T..1.)  a,,,(T  '')
(2 13)
thermal expansion for some engineering materials are given in the tables and graphs at
the end of this chapter.
Consider the case of a slender rod of original length 4.,, heated to a temperature T,
from a reference temperature value T., The
increase in the length of the rod due to the
effects of thermal expansion is required.
Taking T,  T,. = T = 100 0F as the ternperature rise, L,, = 1 ft, and assuming L,
for the material ha i value of 10  7,the unit
elongation from (212) is
The ea
of
rod'~)
10~
is
1material
The elongation of the rod is
24. SELECTED DATA ON PROPERTIES OF
MATERIALS OF CONSTRUCTION
Collected in the pages that follow are
selected data on material properties pertinent
to missile design. Data are included on materials that appear to have promise for application in designs of the future. These data
have been gleaned from a number of references which are listed in the bibliography.
One of the principal reasons for collecting
the information has been to give the reader a
quantitative feeling for the order of magnitude of the physical properties involved. in
sore instances, the variations in these physical properties for different alloys of the same
are presented, but this is not con
he
(214)
sidered of paramount interest. In addition to
It might be well to point out the implicit
the objetive mentioned, th curves are set
assumption that the rod is not subject to any
external oastraint. Ifthe rod isconstrained
extrna
~~isrants
I th ro
i coisraied
to remain at
its original
length,
then the
temperature rise would produce normal
up to afford an easy comparison between one
material and another, The frequency with
which~ time and temperature appear as parameters may be noted.
Another objective in the presentation of
the
is tofinds
point
up the
form
presentationdata
which
favor
with
theofmissile
de
AL
L. = 10  2 ft
stress
stresseqire
rodcesanpeqondi
topoprequiredteto produce
an equal and
posite unit elongation. The value of this
(thermal) stress is obtained by multiplying
the unit elongation by Young's Modulus:
tion w
nd s avoryth th m isilces
signer and stress analyst. in many instances
datc
of a particular kind are shown only for
one alloy
of one material. This is particularly
true in the graphical presentation of creep
data, and stressstrain and tangent modulus
curves as functions of temperature.
Tables 21 through 25 cover several alloys
each of steel, aluminum, titanium, magnesium, and a few ceramics. In this compilation
the room temperature mechanical properties
are given for each alloy and for a selected
form and condition. Young's Modulus, yield
and ultimate strengths are presented on the
basis of tensile test data. The yield stress is
based upon a 0.2%. permanent offset condition. The elongation is the percent change in
length measured over 2 inches of a test
specimen.
Table 26 summarizes the physical properties of various materials: density, melting
a = Ee
(215)
Hence if Young's Modulus for t1be rod used
in the above example is 107 psi, the stress
can easily be calculated:
= 101(10  1) = 10,000 psi
Partial restraints at the ends of the rod produce thermal stresses which are between the
Values of zero thermal stress for unrestrained ends and the thermal stress calculated above for fully restrained ends.
Expansion of the rod in the lateral directions may be calculated in a similar manner.
The coefficient of thermal expansion is usually assumed to be independent of direction.
Since most engineering material3 can be considered isotropic, the assumption is fairly
good. The quantity a is also called the co
16
PROPERTIFS OF MATERIALS
tion with the interpretation of the curves
point, specific heat, thermal conductivity and
shown in these Pi/ures. The stressstrain
coefficientof thermal expansion. In this corncurves shown P e obtained at a:constant ternpilatijn, only one or two alloys of steel,
are
perat.me given by the intersection of the
magnesium, aluminum and titanium
curve with the temperature axis. The stresssuch
of
materials
listed, but thi. properties
strain curves are spread in their correct
as copper, tungsten, molybdenum, oxides,
posti*;t on the temperature axis (abscipsa)
carbides and borides are presented because
in order that curves at intermediate ternof the attractiveness of one or more of their
Mean
peratures; m'fy be interpolated. For an.y given
elevated
tcrmperatures.
properties at
stressstrain curve, the strain at any stress
values of the thermal properties are given
level may ;ie obtained by interpolation betogether with the temperature range within
tween constant strain lines or by using the
which they are applicable. For a number of
strain smle (shown only in Figure 29) and
materials, the, thermal properties are plotted
measurinr .from the point of zero stress. An
as a function of temperature and a reference
illustrativ3 example :s given in Figure 29
to the corresponding figure is given.
fo further clarifia;. n.
In Figure,, 25 and 26 the modulus of
ata at various stresses
nio4u
elasticity anSi the yield stress are plotted aangent
J . _niinum alloy in Figure
are shown fot"
are shoTh valuf . nninum as io F
functions of the temperature for a selected
T e value, ,xe 4,owp s Mos of
number of typical materials. As alreAdy menMaulus, of
, b
A
Moulus
at
tioned, the principal purpose is not only to
of
A tabulation
various pronerties
ternper.'n
at
a room
and elevated
mechanical
show the variation of these properties for a
temperatres fir an aluminum alloy is given
material, but: also to indicate the relative
a s
e
Creep aalu iue
.
tr
magnitudes and the behavior between differincluded.
ent materials. It is significant to point out
incle
that these properties are affected by the rate
by
F pres
inh
ree
Th
of heating, soaking time at test temperature,
through 217 are in the forms preferred by
asthe ctrain rate. Most of the mechanical
structural designers and analysts. In Figure
properties shown are for moderate heating
reader should note that the elastic
215,
and strain rate, and are soaked for onehalf
strainthe
is subtracted
from the total strain to
hour or more at the test temperature. Short
time test results, which are not quite as plenobtain the curves. The loglog plot of minitiful, indicate an increase in the modulus of
mum creep rate as a function of stress in
elsticityl,
dyiedstress with an increase
Figure 216 is particularly useful in the
elasticity end yield srswihainese
formulr'~nn of creep laws for engineering
he master
aster creerves
mtr ina
rate
and
with
and
strain
heating
rate
in
the
creep curves
in the
Finally, in
materiai..
time.
soaking
the
in
a decrease
are total
parameters
the
217,
of Figure
strain, and a socalled LarsonMiller parameter which includes both the time and temperature effect. The comments made in connection with Figures 29 and 210 also hold
for this figure except that temperature is
replaced by the LarsonMiller parameter.
Inasmuch as fatigue considerations are
relatively unimportant in missile type structures, only one set of curves is given for an
alloy of aluminum in Figure 218.
The final set of figures, 219 through 221,
show the thermal properties of a wide variety of materials as functions of temperature. To enable a comparison between the
various materials, each figure shows the
variation of a particular thermal property.
The weightstrength and coststrength
ratios of one alloy each of aluminum, magnesium, titanium, steel and nickel are presented in Figures 27 and 28. Note how effectively the curves give an approximate idea
of the maximum useful temperatures of the
various alloys based upon these criteria. Cost
data are listed in Table 27 and strength data
are based on the tensile yield strength given
in Figure 26. The reader is cautioned against
arriving at any broad generalization based
upon criteria of this type.
Stressstrain curves at various temperatures are shown for an aluminum alloy and
a steel alloy in Figures 29 and 210 respectively. A word of caution is made in connec17
I.
SIPRUCTURES
421
co
I.
c:1
to
=t
to
<
1 41
C1
^bai
00
.'
01
.0
'4
COI
0
0
0
00
400
cr00
Go
coto
.4
to
~~
00
4
1
40
m
00
0oMCI0
I*
 W
0,
C4?
40
CI
00t
oo
t E
PROPERTIES OF MATERIALS
)z. 1
ci
oz
10
COO'i~
0
I tt
30
4

4j
4s
a,0.
ej A
V2 a,a.
If
co,
Qp
W
q.
U2
*'1..
.2~
J2 (j
o
t~.
19
M !J
..
;
IN
wo
STRUCTURES
03
v

t
03
co.4
Ir
cIc
03CIA
cq
Lo c9
o
co
0
m
zz
t
uO to 9
~~11
ME
Iq
0
0U
0.
0
o458
to
to
wt toO
P14
%;
00
11'
40
PRO)PERTIES OF MATERIALS
LG
0(br
>
AAI.
1
C>
.2
LCJ
C)
04cl00
04
C:M
C13O
umr
m
CL.C
r Z.
'
.
to
nm in co
U2 1
z2
.2
a A
4W
eq
II
to
0.
LZ
21
at
0r
Ns~
STRUCTURES
~2.
03 04?
a000
00
0~~~~4000o
03
0
E4
E
C)
00
C)
02
C4
c.
0 0
_ __
M co
00
.0
.0t
o
Ir.
C,
E4
03034oM
A.~o
4.,
00
00 V3
1.
0~~
1.~
t
C2C
C4
co
t 0
COi
03 C
co00
1O.
0 J
24
oot
4
04.l
to
WO
020
a2
.
1c
CO~*4
02
'1
cow
0
4
0
o00Z
u>
1.0
m
(102
C4I
PROPERTIES OF MATERIALS
'[I
co
tor4 C3C3o
C91
r,
lu
0**
I
Lu
O0
vrC
023
'000
Lao
;0
ob
/(
STRUCTURES
10
2
0TIz
0 C
Q0C>9
c, C> c>00
o"
00'2I0
C)c
c II 'It t0 0j CO
c> IC
1i
C0'
01
0f
0~0
 DZ
oo.
toCD &C>I
to
*

>
0000 t00
00
C>00
so
OO
..
.4."000
0C
tt
C=,;co
C
00
. 0.O
0
00C0
o
o~o
C>0
IN
24
003
t'00
,0
PROPERTIES OF MATERALS
ALLOYS
MOLYBDENUMBASE
30
I'
INCONFL XI
IQ
I
177PR(THJ050)
28L,,
SIAINLESS STEEL
P112 HARDTYPE 30i
0
0:
IIr
TITANIUM ALLOY'
U.
\i
TIYANIUM
ALUMIN
I"JJ
4
4~~
TITANIUM ALLOY
IN"'\1 '1
AL
75Y
RAUINUM
AIIOAT
2014 T6
1t 0Lz
*MAGNESIUM
400
V 00
1200
1600
2000
TEMPE.RATURE , OF
Figure 25. Effect of Temperature on Modulus of Elasticity for Severa) Typical Materials
25
STRUCT%&'.ES
ThiNES~STEEL 177f'H (TH 1050)
iI
...
,\$INCONEL
MOLYBDENUM
112% T;
STAINLESS STEEL
HARD TYPE,301
,12
0)

i
6C,
ALUMINUM
_ LLOY
4:PAAAM 2020M 486.
,TITANIUM
ALLOY
5% A 2.5% Sn
'
20
T
COMM ANNFALED
TITANIUMl
\ALUMINUM ALLOY 2024T6
\MAGNESIUM ALLOY AZ 31A
o0
400
1200
800
TEMPERATURE, 0 7
1600
2000
Fisq*re 26. Effect of Temperature on Yield Stress for Several Typical Maieri as
26
PROPERTIES OF MATERIALS
... L
KKKWK7
f,
,8
_t
14
.I
MAGNESIUM ALLOY AZ31A
bO
Fr
ALUMINUM ALLOY 2024T6
w
STAINLESS STEEL
12/ HARDTYPE 301
'
i
INCO14EL X
__TITANIUM
25%
ALLOY
Al, 2.5% Sn
0'1600
1200
800
400
TEMPERATURE,
2000
0F
Figure 27. Effect of temperature on WeightStrength Ratio for Several Typical Materials
27
STRUCTURES
sFit
wt
TITANIUM ALLOY
!,
SI
5%A 1, 2.5% Sn
Figure 28. Effect of Toemperaure
4
on CostStrength Ratio for Several
Typical Materials
MAGNES!UM ALLOY AZ31A
IR.ONEL X
0[
A30
oo
Figuic 29. Compressive
ALLOY 2024 T6
UALUMiNUM
STEEL
#A~~
1200
1606
StressStrain .;.,es for
2024T3 Aluminum Atir'
]STAINLESS
Clad Sheet Stock at
Various Temperalure3
2000
TEMPERATURE , F
50
STR~AN RATE0.002/MiN.
I/? HOUR AT TEST TEMPERATURE
.004
0.002 /min strain Woe
.0
. o0*.
0.0018
Emj*t The strain at rf,500 psi and 250OF is0.0018
ISSTRAIN(
28
PROPERTIES OF MATERIALS
'120C
STRAIN RATE O.002/MIN
RATE OF
00
2/min
00
IL
0

00.3
001
200
400
STRESS i0 PSInFigure 21.
Tamresent dlsSf Curves for024
Aluinu
PT3
Allo)SAiclad Stehoot
n mpreso
4).
T M E ATU E Y U GS
OD29S
STRUCTURES
0
0
01
00
0l
cf.,
!0
40
:3
*1
o
5)
0c
0
.0
E
,,
Cl
.0
(n
0
o
I/d
'I5/o
SS3U.
/,I/
o )(0
0
i
o0
0.
I 0
II
.
0j
0
/ 0
NO 00
00
0
C4
00m
!I
i
'
cm
iL00
co
00
ISd F,01
SS'3NIS
30
1
)0
I
PROPERTIES OF MATERIALS
N_
__
__
__
__
__
CRE_
RUTR
o_
_
_
I.
__s
_IN,
30
c3O
SPECIMENS FROM EXTRUSIONS,
ao0
FORGED ROD& PLATE
*10
z0t
0.1
1.0
10
TIMEHOURS
100
1000
Courtesy ALCOA Research Laboratories
Figure 274. StressRupture and Creep Characteristics of X2020 Aluminum Alloy at 300'F
31
STRUCTURES
0
10"
jo
0
00
I
IL
00
dU:
1S
00001
320
PROPERTIES OF MATERIALS
20
2
241
z'a
____
01110
0.000001
e8
100
1000
TIME TO RUPTURE, HOURS
0.01
0.001
0.0001
0.00001
CREEP RATE, IN ./iN./HR.
10000
0.1
216. Creep Rate an~d Time to Rupture Curves for 3SH18 Afuminum Alloy
Ii
STRUCTURES
S120
YIELD STRESS=40000 PSI
DWI
=EPEAURO
0
frg
,IEaOR
it
W
),.
r_
02
moso
wz
TORELR
4034
00)
PROPERTIES OF MATERIALS
3200
t
2800
siOPPER
'V
z
2000
1.600
z0
<.
BERYLLIUM
2oo
'"
SAGNESIUM
ALLOY
MOLYE LYBE!,1
FSIH24
ITALUM ALLOY RC CII.M
400
800
1200
1600
2000
TEMPERATURE
A LF
4 0 0
[
,'
~I~ilL
Figure 219. Thermal Conductivity of Selected Materials
35
2400
STRUCTURESj
kJ
0.80
B.RIUM/
0.30
MAGNESIUM ALLOY
FSI 24
0.30
. .
1
ALUMINUM ALLOY
ALUMINUM OXIDEI
TITANIUM ALLOY
0.20

400
:F CoPPER
0.0 
MOLYBDENUM
1600
1200
800
TEMPERATRE, OF
Figure 2.20. Specific Heat of Selected Materials
36
2000
2400
PROPERTIES OF MATERIALS
114Ai
_
16
..
NUM ALLOY 7075716
J_"COPPER
STAINLESS STEEL
XO""
I/""
"q
INCONEL X
ALLOY STEEL
."
JJ
...
,ALM
"
ALLOY FIH24
'/
14
MAGNESIUM
,_
A'3408630
/"KTiTANIUM
ALLOY

z"'
VA'LUMINUI
J4
MBDENUM
_....,
OWIPE
0=
00
400
800
!200
TEMPERATURE,
1600
2000
2400
OF
Figure 221. Mean Coefficient of Thermal Expansion; for Several Materials
37
ri
STRUCTURES
75. REFERENCES AND BIBLIOGRAPHY
1.o Srenth
Meal Arcrft Eemets,3rd
March 1965 ed.. Air Materiel Command5 Bulletin, Government Printing Office,
Washington., D, C.
2. Engineering Materials Hand.liok, 1st
ed., edited by C. L. Mantell, McGrawHill Book Co., doc., New York, N. Y.,
1958.
3. Metals Hac., 1948 ed., American
Society for Metals, Clevelaned 3, Ohio.
4. Besserer, C. W., twuided Missiles Handbook, D. Van Nostrand Co., Iz.Princeton, N. Y., 19b8.
5. Stainless Steel Handbook, Allegheny
Ludlum Steel Corporation, Pittsburgh
22, Pennsylvania.
6. American Institute of Pkysir6 Handbook, McGrawHill Book Co., Inc., New
York, N. Y., 195?7.
7. Handbook of Chemistry and Physics,
Z9th Edition, Chemical Rubber Publishing Company, Cleveland, Ohio, 1947.
8.Hn0
oko Experimental Stress Analysis, ed. by M. Het~nyi, John Wiley and
Sons, Inc., New York, N. Y., 1950.
9. Alcoa Structural Handbook, Aluminum
Company of America, Pictsburgh, Pa.,
1950.
10. Research a'nd Development Technical
Handbook, Vol. I, Aviation Age, ConoverNast Publication, New York, N. Y.,
195753.
11. Zwikker, C., Physical Propertiesof Solid
ilaterials, Perganion Press, New York,
N. Y., 1954.
12. Frankel, J. P., Principles of tike Properties of Materials, McGrawHill Book
Co., Inc., New Y'ork, N. Y., 1957.
13. Carlson, R. L. and Manrning, G. K.,
"Investigation of Compressive Creep
Properties of Aluminum Columns at
Elevated Tempeeaiures, Part 2, Stability 2"roblems," Wright Air Development
Center, WADC TR52251, May 1954.
14. Dorn, J. E. and Tietz, T. C., "Creep and
StressRt.pture Investig<,:nis on Some
Aluminum, Alloy Sheet Metals," Pro~ceedinf/s of the AnwY.';ian Society for
Test ingi Materis,Vol. 49, P 815, 1949.
15. McAdams, WV. H., Heat Transmi7ssion,
ed., John Wiley and Sons, Inc., New
YrN
. 94
16. Jakob, Max, Heat Transfer, Volume 1,
John Wiley and Sons, Inc., New York
N. Y., 1949.
17. Pohle, P. V., "Hfeat Transmissi~n in the
Structure," Chapter 3 of AGARDoGRAPH entitled Higit Temperature
Effects in Aircraft Structures.
18. Schneider, P. J., Conduction Heat
Transfer, Addison  Wesley Publishing
C o., Inc., C.ambridge, Mass., 1955.
19. Carclaw, H. S., and Jaeger, J. C., Conduction. of Heat in Solids, 1st ed., Oxford, Clarendlon Press, 1947.
20. Evans, Jerry E., Jr., "Thermal Conductivity of 14 Metals and Alloys up to
U100cF,." National Advisory Committee
for Aeronautics, NACA RM E501,7
March 2, 1951.
21. Wisler, P., "MaterialsHeat Sink and
Expanm ;n Problems," General Electric
Company, Missiles and Ordnance Systems Dept... July 26, 1956.
22. Gatewood, B. E., Thermal Stres r3, MeGrawH.ll Book Co., Inc., New York
N. Y., 1957.
23. Manson, S. S., "Behavior of Metals
Under Conditions of Thermal Stress,"
National Advisory Committee for Aeronautics, NACA TR 1170, 1954.
24. Favor, Ronald J., et al., "MaterialsPropertyDesign Criteria for Metals,
Part 5. The Conventional ShortTim~e,
Elevated Temperature Properties of
Selected Stainless Steels and Super
Alloys," Wright Air Development Center
WADC TR 55150, October 1957, ASTIA
Document N;o. AD 142069.
25. Favor, Ronald J., et aL., "MaterialsPropertyDesign Criteria f,)r Metals,
Part 7. The Conventional ShuitTime,
Elevated Temperature 1Proper tiez of
Selectd Ljow and MediumAlloy Steels,"
Wright Air Development Center, WADC
k.
TR 55150, October 1,)Qr57, aSTA XDOcumn!r.t No. AD 14264.
Roel , and Aver;', Chale
11., "Effects of TemperntureTimeSt1 \,ss
'G.Prny
38
~i1
PROPERTIES OF MATERIALS
2Wright
Hnwories on the Mechanical Properties
of Aircraft Structural Metallic Materials, Part I. TemperatureTime Studies
for 2024T3 and 7075T6 Alclad Sheet,"
Air Development Center, TR 56585, September 1957, ASTIA Document
No AD 142007.
27. Schwartzberg, .. R., et al., "The Properties of Titanium Alloys at Elevated
Temperatures," Battelle Memorial Inst tute. Titanium Me~allurgical Laboratory
No. 82,P.September
10, J.1957.
28. Report
Roe, William
and Kattus,
Robert,
Metals Used at High Temperatures,"
Prodzct Engineering,Vol. 25, No. 1, pp.
1356, January 1954.
36. Criner, C. B., and Spuhler, E., "Summary of InformationAlloy X2020,"
Aluminum Company of America, Report
No. 1357AKS7, September 30, 1957.
37. Lemon, R. C., and Sicha, W. E., "New
Aluminum Casting Alloy XA140 for
Elevated Temperature Applications,"
paper presented at 60th Annual Meeting
of
the7, American
1oundrymen's Society,
May
195,.65
"Tensile
l
Properties of Aircraft Structural Metals at Various Rates of Loading After Rapid Heating," Wright Air
Development Center, WADC TR 55199,
September 1957, ASTIA Document No
AD 142003.
29. Shinn, D. A., "Selected Properties of
Structural Materials at Elevated Ternperatures," Advisory Group for Aeronautical Research and Development, Report 104, April 1957.
30. andven, 0. A., "Cermets as Potential
Mlaterials for HighTemperature Service," Advisory Grov;p for Aeronautical
Research and De,e!opment, Report 99,
April 1957.
31. Schwarzkopf, Paul and Kieffer, Richard,
38. Lyle, John P., Jr., "Aluminum Powder
Metallurgy Products," Materials and
Methods, Vol. 43, No. 4, pp. 1061:1,
April 1956.
39. Dix, E. H., Jr., "Aluminum Alloys for
Elevated Temperature Applications,."
paper presented at the ASME Aviation
Conference, Peper No. 56AV& March
2416, 1956..
40. "Symposium on Effect Gf Cyclic Heating
and Stressing on Metals at Elevated
Temperatures," American Society for
Testing Materials 57th Annual Meeting,
Chicago, Illinois, ASTM Special Technical Publication No. 165, June 1954.
Company, New York, N. Y., 1953.
32. Killmar, H. M., and Wroten, W.L.,
"Special Ceramic Materials of Construetion for Rocketry," Ceramic Industry,
Vol. 58, No. 5,pp. 936, May 1956.
33. Ault, Neil N., "Characteristics of Refractory Oxide Coatings Produced by
Flame  Spraying," American Ceramic
Society, J., Vol. 4(/, No. 3, pp. 6974,
March 1957.
cially by Fatigue," Edgar Marburg Lecture, Piroceedings of the American
Society for Testing Materials, Vol. 33,
Part 2, 1933.
42. Howard, H. B., "Merit Indices for Structural Materials," Advisory Group for
Aeronautical Research and Development, L..;poit No. 105, April 1957.
43. Dukes, W. H. and Schmitt, A. (ed.),
"Structural Design for Aerodynamic
34, Westerholm, R. J., et al.. St U:on I"Some Aspects o;f the Development and
Performance of Pure Oxide Coatings,"
Section Il"Appication of ROKIDE:
A Coating Lo Ramjet Engine Combustion
Components," papers presented a'ARSASME Aviation Division Conference.,
Heating." Part JDesign Information;
Part IIAnalytical Studies; (confidential) Part IIIStructural Design
(SECRET), Wright Air Development
Ceater, WADC TR 55305, October 1955.
41. Gough, H. J., "Crystalline Structure in
Refractory Hard Metals, The MacMilhar
Los Angeles , a
35.
Sully,,
A.
H.
"Specia'{l
Relation to Failure of MetalsEspe
March 1956.
/'3.ti
",,
for
39
*1
Al
1!
ii
*1*4
I
II.
40
Chapter 3
STRESS ANALYSIS
*
31. GENERAL CONSIDERATIONS
In the fiftyodd years since the first successful flight of a heavierthanair machine,
the analysis of the stresses in aircraft structures has been developed toa highly perfected
art. Good textbooks on airplane stress analysis are readily available. The purpose of this
chapter is to present those principles and
formulas which are peculiar to missile structural analysis and are not to be found, as a
rule, in books on airplane structures.
In this category belong shell analysis, the
study of thermal stresses, and the investigation of the effects of creep. Shell analysis is
peculiar to ballistic missile design because
most of the structural elements of these
missiles are thin shells, often unreinforced
The customary methods of reinforced thin
;hell analysis, amply di.icussed in textbooks
on aircraft structures, are not applicable to
pressure, and compressive when the opposite
is true.
32.2. Conical Shell. In a conical shell under
a uniform internal pressure the tangential
stress, (,, and the axial stress acting in the
direction of the generator, a,, are
a, = pR/h cos y
(3 3)
at = pRt2h cos y
(3 4)
where R "s the local mean radius of the shell
measured in a plane prpendicular to the
axis of the cone, y is the angle subtended by
any generator with the axis of the cone (the
half apex angle of the cone), and the remaining symbols .re defined as before.
3
3.2.3. Spherical Shell. The normal stress
acting in any direction in the plane tangential to the sphere at the considered point
known as the membrane .Atres.,%, aM,. may be
expressed as
these missiles. Thermal stresses and creep
=pR
arise as
ing; for
tance to
account
32. STRESSES AND DEFORMATIONS IN THIN
SHELLS UNDER UNIFORM INTERNA.
PRESSURE
(3 6)
= E
whpre t is the strain (elongation per unit
length) and E Young's Modulus of elasticity,
the mean radius of tie cylinder inicreases by
the amount
p"
R,
(3 7)
32.1. Circular Cylindrical Shell. The tangential or hoop stress, n. and the stress i
the longitudinalor axial direction, ,, within
circular cylindrical shells under a uniform
internal pressure are
pR
_=
h(3 1)
ax= pR
2h
(3 5)
2h
32.4. Thin Shell. All of the stresses given
heretofore are membrane stresses. These are
stresses distributed uniformly through the
thickness of the wad. Under their action the
shell change. its size slightly. When the material is perfectly slastic and when it follows
Hooke's uniaxial law
a consequence of aerodynamic heatthis reason they are o. great imporveryhighspeed missiles but of little
in conventional airplanes.
where x is Poisson's ratio who,,e value is .',
for steel and aluminum alloys. For the cone
the corresponding quantity is
AR .
P
pR2
A
2 Ehcosy
(38)
and for the sphere
(3 2)
where p is the excess of the internal pressure
in the closed cylindrical shell over the exter
nal pressure, R is the mean radius of the
cylindrical shell, and h is the wall thickness.
The stresses , and 7, are tensixc if the internal pressure is larger than the external
AR,),,
= (1 
(3 9)
41
itt
STRUCTURES
33. DISCONTINUITY STRESSES IN
THIN SHELLS
vie with twoylind
erical
pressure vessel pvided with two hem'spherical
piressure ve.el
heads, and the vesel is filled with a gas at
high pressure, the change in the length of the
radius of the cylinder is more than twice the
change in the length of the radius
,,. of the
sphere in accordance with Equations (37)
and (39) if the wall thicknesses are equal.
Such deformations are obviou.ly impossibhle
when cylinder and hemisphere are welded
together (see Figure 31). For this reason
bending stresses develop in the walls of the
vessel in addition to the membrane stresses
discussed in Section 32. When the material
is perfectly elastic, these bending stresses
vary linearly across the wall and attain their
maximum values on thp 'juter and inner surfaces of the wall. Th 'bending stresses add
up to a bending moment of M., per unit length
of the circumference at the edge of the cylinder of Figure 32 and the shearing stresses
accrmpanying them to a shear force V, per
unit length of the circumference.
Under the action of the edge moment and
the edge shear force the region of the shell
close to the edge deforms. The deformations
of the edges of cylinder and hemisphere are
of such magnitude that the diicontinuity in
the deformations as caused by the membrane
stresses alone is eliminated and cylinder and
hemisphere fit together. The bending stresses
caused by M. and V.. are known as dsconti
When M. and V are applied to the edge of
the cylindrical shell shown in Figure 32, the
thcyidicalhiell csw in Fighe 3. h
by the following equation
2V3(1
VR)
(R
=
Eh
+
The positive directions of M, V and w are as
shown in the figure. The slope of the deforwed shape of the shell is
(dw/dx). =
3(1 
(2MoK + VoR)
(31)
(311) yields a negative slope when
M,, and V., are positive; correspondingly the
radial dispiacement w decreases with increasing x in Figure 32, that is the slope
shown in the figure is negative.
XooX
The value of the constant K in Equations
(310) and (311) is
K = '/3(1  9) V(R/hj
(31?
The bending moment M,, is not necessarily
the largest bending moment in the wall of
the shell; variation in the value of the bend
ing moment is given by the equation
Figure 31. Theoretical Deformations if Pressure
Vessel under Membrane Strestset Alone
M = eKx/
[mcos R + sin Kx
+i
+
V. t
Mo
xR
V,.R
.K
sirlK1c
(313)
The location of the maximum bending moment can be computed from the formula
oX _
R  1 2+VtR
o Kcot
(314)
The connection between the maximum bending stress and the bending moment is
6M/h2
(315)is
The'attenuafion of =the
bending stresses
2MK
Kx
Mo
characterized hy the factor eKx"/ ; this is
unity at x
0 and e I
0.368 when
Figure 32. Doformations of Thin Circular
Cylindrical Shell under Bending S$resses
i,,
42
R/K
(316)
STRESS ANALYSIS
When the attenuation length x., is small as
compared to the undisturbed, uniform portion of the length of the cylinder, the equations given here are valid; if the attenuation
length is larger the equations must be replaced by more complex equations discussed
in the literature.
Experience has shown that satisfactory
Modulus of elasticity
E = 29 x 106 psi
Poisson's ratio
= 0.3
Internal pressure
p = 100 psi
From Equations (31), (32) and Z35) one
obtains
accuracy is obtained for, engineering purposes if the actual shell configurations are
imagined to be replaced by equivalent circular cylindrical shells in the calculation of the
disturbance stresses. In the case of the spher
Hoop stress in cylinder
,, = 20,000 psi
Axial stress in cylinder
Ox
10,000 psi
Membrane stress in hemisphere
ical shell, the equivalent cylindr:al shell has
10,000 psi
the same radius as the spherical shell that it
'.he changes in the length of the radius of
replaces. In the case of the arbitrary rota
cylinder and hemisphere are computed from
tionally symmetric shell which joins the cylinder with a continuous tangent, the radius
of the equivalent shell is
(317)
R, = R(r/R)2
where R is the radius of the cylinder and r
is the radius of the meridian of the rotationally symmetsc shell at the junction with the
cylinder. Equations (310) to (316) retain
their validity for the general rotationally
symmetric shell joining the cylinder with a
Equations (37) and (39)
1.466 X 101 in
R
0.603 X 102 in
The misfit between cylinder and hemisphere
istherefore
AR,,,  AR,phe.. = 0.863 X 107 in
continuous tangent if R is replaced in them
by R, and if x is measured along a meridian,
This is true, however, only if the change in
the valves of the principal curvatures of the
rotationally ,.ymmetric shell is insignificant
over the attenuation length x,t defined in
equivalent cylinder, and a coordinate x, is
introduced in it which is measured to the left
from the juncture between cylinder and
hemispheie. An edge moment M,. and an edge
shear force V., are applied to the cylinder as
shown in Figure 32 and equal and opposite
Equation (316).
V,hen the head of the pressure vessel has
a taagent which changes discontinuously
from the tangent of the cylinder, the equiva..
lent cylinder concept is still valid, but the
forces and displacements must be resolved in
the proper directions.
For illustrative purposes the numerical values
characterizing the pressure vessel of Figure 31
may be taken as
Radius of median surface of cylinder
R,= 25 fn
Radius of med.an 3urface of hemisphere
R,,,h,.r = 2 in
This misfit must be eliminated with the aid
of the bending stresses.
First, the hemisphere is replaced by the
quantities are applied to the cylinder replacing the hemisphere. The conditions must now
be enforced that under the action of M,, and
V. the ;relative displacement of the two shell
wails is 0.863 X 10 in., and that the slope of
the gene.ators remains continuous. The l.tter condi&;:ion is fulfilled if M,, = 0 because
the wall thicknesses are the same in cylinder
and hemisphere. As the real and the equivalent cylinder are eqaal in size, under the
action of the same shear force V,, each one
shows the same numerical value of change
in the length of the radius. Hence Equation
(310) becomes
Length of cylinder
L =100 in
Wall thickness of cylinder and fiemisphere
h 0.125 in
 EK
3(1
hv:)
R
R
(J 18)
43
!I
STRUCTURES
From Equation (312) K =182 and substitation of the numerical values in Equation
when the pressure reaches the following
critical value:
(318) yields
h V7liv
V= 1.99 x 10= x 8.63 X i0=
 17.2 lb per in
With M.,= 0 the condition of a maximum
moment becomes
Kx R
r 4
(319)
Substitution in Equation (313) leads to
M..x
0.322V.RK = 0.442V0
 7.6 in lb per in
The corresponding stress is from Equation
(315)
Ob.=
= 2920 psi
From Equation (319) the iocation of the
maximum bending stress is
x (r, 4) (R/K) = 1.08 in.
p. 0.92E01
R(3Here r7 is Young's Modulus, h the wall th.ckness, R the radius of the median surface of
the cyiinier, and L the length of the cylinder
between s, pports; Poisson's ratio v is taken
ps0.3. Thi equation is valid only if
0.763 ,
7i
1
(323)
The meaning of this restriction should be
understood from a numerical example. When
R = 25 in. ant' h h g in., the inequality
reduces to 0.4321 1. If L = 10 in., the ;efthand member is 4.32; with this value one can
expect a reasonabit approximat.on. The accuracy increases wikh increasing L.
which is very close to the juncture. The at
For very long cyhirders the critical ,res
tenuation length is from Equation (316)
= R,K = 1.37 in.
Hence interference from the ton of the heniisphere and from the other end o.7 the cylir der
is negligible and the results obtained are
accurate enough for engineering purposs.
Superposition of a, and of ah. yields the
maximum normal stress which acts in the
cylinder in the axial direction at a dista ice
of 1.08 in. from the juncture on the outer
surface of the wall
sure is
p,, =E 14(1  2)]1(h,'RJ
fomlI
a .. ,..= 10,000 + 2920 = 12,920 psi
where
Finally a general forniula can be deriied
from Equations (313), (318) and (319'
which is valid when the cylinder and i he
equivalent cylinder have the same radius a id
the same wall thickness:
M.=
0.0465
E
When v' 0.3, this becomes
(, 24)
Equation (,24) is to b. used in preference
'to Equation (322) when it predicts a higher
buckling pressure than Equation (322).
For very short cylinders which do not satisfy inequality (323), the literature should
be consulted. '
34.2. Uniform Axiai Pressure. The buckling stress under the application of a uniorm
axial pressure can be calculated florn the
k = 9(h/R)0. + 0.16(R, L)'.(h R)0 3 (325)
and h, R, and L are as defined in the previous
paragraph. For example, k = 0.24 when
R/h = 500 and R L = 1. For short and com
paratively thickwalled cylinders Equation
h AR%I  ARp,:
1
(325) yields values of k exceeding 0.3; itis
recommended that in such cases k should be
taken as 0.3.
(340)
0.0487Eh 2 ARI  ARph....
R
(321)
4
T34.3.
Pure Bending. It is customary to calmaximum bending stress in a
the
culate
SHELLS
CYLINDRICAL
cylindrical shell from the engineering theory
34.1. Uniform Lateral Pressure. If a uniof bending. If the maximum compressive
form external pressure p is acting over the
stress so found is less than 1.33 times the
circular cylindrical surface of a shell, but Pot
critical axial stress obtained from Equation
over the end sections, the cylinder hackles
(324), the cylinder is riot expected to buckle.
M.,=
414
X
',
STRESS ANALYSIS
34.4. Pure Torsion. The cr'tical shear
'
stress
T,
and circumferential di'qplacement of the
in pure torsion is :
circular median line of ihe wall are prevented
in the end sections.
When the end section of the cylinder is
welded or bolted to another structural element, warping of the plane and rotation of
the generators may be partially or fully prevented. This additional constraint raises the
signifficant
excepi
when
not
the cylinder
is very
is usually
increase
but the
critical stress
shict en the
clinde ve
7, = E(hiL)213.08 + J3._5
(3 27
+ 0.557(L'/hR)"I,1
if Poisson's raiio is taken as 0.3.'.5
For very
long cylinders the fohlowing
formula
applies:
E'(h,;R/ " 2
(32)
(328
34.5. Hydrostatic Pressure. In the hydro=0.25
static case of loading, the cylinder is closed
at its two ends and the closed shell is subjected to a uniform external pressure p.
When the cylinder is very long, it buckles
under the critical pressure corresponding to
uniform lateral pressure; and when it is very
short, under the critical strew of the axial
loading case. In the range between these two
extremes the critical pressure of the hydro
short. In the latter case more accurate values
lIterature2.'
of the bucking stress can be taken from the
d
static loading case is lower than the citical
pressure of either of the simpler loading
cases mentioned. For exact values, see the
literature.'2 An approximate formula can
be given as
E (h/R)y
P " (I,/R)(R.,h)' ,  0.636 (3 29)
24.6. Combined Axial Compression and
Torsion. An empirical formula  can be given
in the form:
R, +iR,< 1
(3 30)
In this interaction formula R,. is the ratio of
the compressive stress of the combined load
stress can b expresed by an addi.ion of k
to the value k .of the factor in Equation
(325). First calculate the nondimensional
pressure p* defined as
p* = (p/E)(R/h)2
(3 31)
The increment Ak is then qiven by
Ak = 1.9 pi if 0 < p* < 0.12 (3 32)
Ak = 0.229 if *> 0.12
Finally, Equation (3 2ti) is replaced by
(k + Ak) E (h/R)
(333)
If the example given under Paragraph 34.2 is
continued ard ar, internal pressure p = 10 psi
is applied to the she)l, one obtains p' = 0.0862
and
ak = 0.164. Hence the critical stress
b
34.8. Stabiization Through Internal Pressure. Thin shells can be partially, or fully,
stabilizedI through application of an internal pressure p. The stabilizing effect in the
case of a circular cylindrical shell subjected
to a uniformly distributed axial compressive
)0.92
ing to the critical compressive stress under
compression alone, and R, :%the ratio of the
0.24 x 29 X i0/500
13,900 psi with
shear stress of the combined !oading to t)l,
critical shear stress of torsion !:'.se. When
the inequality is sat;:.ca, the cylindrical
shell is safe. W en the lefthand member of
the ineo,aty is Lqual to unity, the shell
buck'es. No equilibrium can be maintained
the lefthand member exceeds unity.
4nen
out internal pressure
0,= 0.404 x 29 x 10/500 = 23,450 psi
wi.h internal pressure
Young's Modul~s was assumed to be 29 X 106
s
A limited amount of experimental evidence  indicates that Equation (330) can be
34.?. Different Boundary Conditions. The
equations given above are valid for the case
when the edges af the cylinder are simply
supported. This boundary condition invc'ves
no constraint against rotation of the gem rators of the cylinder over the supports and no
resistance to a warping of the end section of
the cylinder out of its plane; however, radial
used to determine the critical condition of a
circular cylindrical shell subjected simultaneously to compres;ioa, tursion and internal
pressure'. It is only necessary to reinterpret
the meaning of the symbols R, and 9,. R, is
nox the ratio of the applied I coinpressive
stress of the triple loading (compression plus
torsion plus pressure) to the "ritical corn.
o,, =
45
STRUCTURES
pressive
tress in the presence of the pr.
scribed internal pressure p. Similarly, R, is
the ratio of the sitaring stress caused by the
torque of the trifle loading to the critical
shearing stress of torsion in the presence of
'
the prescribed internal pressure p.
A circular cylindrical shell subjected
simultaneously to torsion and internal pressure does not buckle if the following inequalis stisfied: "(ity
CE 5s
where R, is the ratio of the shearing stress
of the combined loading to the critical shearing stress under pu.e torsion while R., is the
ratio of the internal pressure p to the critical
pressure of hydrostatic loading (both pressures are coasiered positive in this context).
Figure 33. Determination of Tangent Modulus
and Secant Modulus from StressStrain Curve
34.9. Inelastic Buckling. When the stresses
at which buckling takes place according to
the formulas given are higher than the limit
UCKLING OF THIN SPHERICAL
SHELLS
The critical compressive stress in a uni
E
of elasticity of the material, the theories
formly compressed thinwalled spherical
underlying the formulas are not valid. They
must be modified in order to arrive at results
useful in engineering applications. As the
theory of inelastic behavior is in lesser agreement with test results than the theory
shell is
= 0.2E(h/R)
(3 35)
The correspondiog value of the critical pressure is
pl
elastic behavior, plastic buckling formulas
must be used with some degree of caution.
In the calculation of buckling stressas
greater than the elastic limit, two moduli
are of importance in addition to Young's
Modulus E. One is the tangent modulus Et
defined as the rate 6f change of the stress
with strain at the particular value of the
36.1. Straight Bar. If a straight bar were
heated from a uniform initial temperature
T, to a uniform final temperature T, it would
expand and its initial length L would increase
by the amount
stress; and the other the secant modulus E.
which is the ratio of the particular value of
the stress to the c,,)rresponding value of the
(3 37)
AL = a(T  T,)L
expanof
thermal
a is the coefficient
whereusually
r, asured in inchies per inch per
sion,
strain (see Figure 33). Both E, and E. are
equal to E when the stressstrain relation is
represented by Hooke's straight line. In the
case of plastic bucliing, E in the formulas
s ome combination of E,
must be replaced
degree Fahrenheit.
If, however, the bar is held between insulated walls as shown in Figure 34, the insuthis
'expansion
prevent
wallsup
lated
magnitude
of a completely
stresses
thermal
setting
by
and E4 ; in a first approximation one may use
E. inste.,d of E. For more accurate results
recourse should be had to the literature."."
a=  E(T  T)
(3 38)
where E is Young's Modulus and the negative sign indicates that the thermal stress is
46
41
(3
36. THERMAL STRESSES
STRESS ANALYSIS
compressive if the final temperature T is
higher than
oe th at emperature T. Itis
easy to prove that the shortening of the bar
in consequence of this thermal stress is equal
to the elongation caused by the rise in temperature as given by Equation (337).
L
L
/

d,
i/
Figure 34. Bar Heated between Two
Inulatd Walls.r
where the average rise in temperature is
b/2
(T + T,). = (1/b)/b/2  T)dy (340)
/2
'This stress is a normal stress parallel to the
x direction. At the free ends x = _L/2 the
normal stress is zero and significant deviations from the values given by Equation
(339) exist in regions of the plate extending
from xL/2
to x (L/2) + b and
from x
(L/2)  b to x = L/2.
When the temperature distribution is not
symmetric to the x axis and the thickness h
of the plate varies with y, the themal stress
in regions distant from the free edges
L/2 is given by

where
3S.2. Flat Plate. If a long rectangular flat
plate ,)f constant thickness h is heated from
a uniform initial temperature T, to a final
variable
temperature
is
symmetric
with respect
T, whose
to the distribution
x axis (see
.5
(T  Ti). ,
aE[(T
T.)
Cb =yd
(31/42Tb)J
"
and A is the
ctn aeotep e
formulas y must be measured from
In thesh
the ceftroidal axis of the section; be and b
are the distances of the lower and upper extreme fibers from the centroidal axis; and I
isthe moment of inertia of the section with
resuect to the neutral axis:
b.
I__y'dA

(T
T,),]
(T  Tj)hdy
(342a)
by

(I/A)f
,
Figure 35), and if rigid end walls prevent
all expansion, the thermal stress distribution
is still given by Equation (338). In this case,
T. as well as the thermal stress, is a function
of the coordinate y while it is independent of
the coordinate x. In the absence of rigid walls
such normal stresses cannot exist at the free
ends, x = L/2, the thermal stress is given
a=
aEj(T  T)  (T  Ti).,  Cy
(341)
(339)
hy2dy
b,
(343)
Ti
_______________0
L
r
Figure 3S. Assumed Temperature Distribution in Long Rectangular Flat Platef
47
'p
h]
"
STRUCTURES
36.3. Ring Frame. if a ring frame is
heated uniformly, no thermal stresses arise
in it provided it is made of a single material
and is free to expand. if, however, the ternperature of the outer surface of the ring exceeds by _%T that of the inner surface, as
indicated in Figure 36, the thermal change
(T TI)., which varies with y. The first
term causes no thermal stresses and ihe
maximum stress cauced by the second is
given by Equati3n (344) with the stress,
distributed linearly across the section. The
third term, (T  T).,,, can be substituted in
Equation (338) in order to obtain the additional thermal stress.
3f.A. Circular Cylindrical Shell. The thinwalled circular cylindrical shell of Figure 37
T, AT
is uniformly heated to a temperature exceeding by AT the uniform temperature of the
ring frame. If the wall thickness of the shell
is h and the crosssectional area of the ring
frame is A, the hoop compressive stress in
the shell at the location of the ring is
. .= raET
(345)
and the maximum axial bending stress is
A. A
H
c,, = 1.82rfET
(346)
Iprovided
that Poisson's ratio is 0.3. The
[,
SECTION AA
value of the multiplying factor is
r = 1/[1 + (2,'K)(Rh,A)j
(347)
with
K = 1.29.v/R/i
(348)
It can be seen from these equations that
the value of r is always between 0 and 1.
Figure 36. Ring Frame with Temperature
When the area of the ring is very small, r is
Varying Linearly Through Depth
very small, and the small amount of constraint provided by the ring does not give
in the curvature is restrained by the contirise to large thermal stresses. When A is
nuity of the structure. When the temperature
very large, r is not much smaller than unity,
increase from the initial uniform temperaand the hoop compressive stress corresponds
to almost perfect constraint.
ture is a linear function of the depth of the
section, but is uniferm along the circumference and at the same time the ring frame has
3.7. THERMAL BUCKLING
two axes of symmetry
asbndig
shown in
the figure,
theheral
axium
sres
!s37.1.
Columns. The column does not know
th
awhether
the compressive stress in it was
I,.
(1 2)aEAT
(3 44)
In the derivation of this formula the additional assumptions were made that the cross
section of the ring was doubly symmetric
and that one axis of symmetry was in the
plane of the ring.
caused by restrained thermal expansion or
by an applied load. Consequently the customary buckling formulas can be used wth
confidence as soon as the compressive thermal stress is known. In particular, one may
write for the critical stress of a simply sup
When the temperature vwries arbitrarily I ported column
=r2 i E,(L,'p)(349)
with y (through the section), but is constant
around the circumference of the ring, the inwhere Et is the tangent moclulus (sue Figure
crease in temperature can be represented as
33), L is the length of the column, and P is
the sum of an average increase, a linear
the radius of gyration of its cross section.
variation with y, and an additional term
The ratio L/p is known as the slenderness
48
STRESS ANALYSIS
A.....
.1
"~
'
A'
Figure 37. RingReinforced Circular Cylindriced Shelf
SI0
X1
Figure 38. Simply Supporled Pla',e Subjected to Thermal Stresses in One Direction
ratio. When the stres does not exceed the
elastic limit of the material at the operating
temperature, E, becomes E.
I
VI
buckling taks place at the following critical
value of the average st.ress:
37.2. Simply Supported P:ates. In the case
of plates the customary buckling formulas
aao plae
n
theid.
cust
buckling formuls
again are valid. Thus the buckling stress of
a simply supported
rectangular
of constan
s i th(se
ckneFi ure plate
8)Equation
stant thickness is (see Figure 3
The following
meaning of this
(351)
(y
o and y =
acin i 3.6E(h,
h
b)2
(.q 50)
where a
is a uniform compressive stress
acting in the x direction.
The complication arising from the thermal
nature of the stiesser, is that the distribution
is usuaily not uniform. For a :igorous evaluation of the conditions of buckling, reference
should then be made to the literature. When
the distribution can be approximated by the
formula
or= ao1 + p cos (2iry/b))
(351)
(y
b 2) it is zero. Equation (352) yields
in this case a......  2a.,.. Hece the shifting
of the compression toward the supported
edges and the consequent reliel of the unsupported niiddle portion increases the critical value of the average stress by a factor
of 2. On the other hand, p = I increases
the compressive stress in the middle to 2 a..
and decreases it along the edges; the result
is a decrease of the critical value of the
average stress by a factor of 2,3.
0,0..
49
(p 2)1
(3 52)
e. amples will illustrate the
formula: When p = , from
thp stiess at the two edges
b) is 2(%,., while in the middle
I'
7
STRUCTURES
37.3. Rectangular Plates with Free Edges.
The control surfaces of ms;.siles can often be
considered as flat plates of constant or varable thickness with three edges completely
free and the fourth constrained elastically
(see Figure 39). If the atresses a acting in
the xdirection atid caused by nonuniform
rises in temperature are distributed as shown
in the figure, the control surface is likely tob
buckle in a torsional mode when the stress
intensity becomes sufficiently high." The
stress can be given as
,  =of(yl
(3 53)
where a. is a conttant while f(y) defines the
sh;.pe of the stress distribution curve. Tne
stress is considered positive .%hen it is coinpressive. Buckling takes place when
(354)
= GC I
where GC is 'he St. Venant torsional rigidity
of The plate and
b
(3 55)
I  hyf(y)dy
Thin integral can always be evalua;teJ graphically or numerically without difficulty. When
the plate thickness is constant,
= (2 3)h~b( (3 56)
GC. .
and G is the shear modulus of the material.
1~PPI
COMPIRESSION
TENS10 O
b
cYf(y)
Figure 39. Assumed Stress Distibulion in Reciangular Flat Plafe with Free Edges
37A. Thin Circular Cylindrical Shells. In
order to simplify the problem of thin circular
cylindrical shells one may consider separately
thermal
three cases of temperature and
temperature
tures become very high.
stress distribution. First, the
The situation is different when the ternvaries only in the circumferential
perature
direction and remains constant through the
can vary through the wall thickness, but remain constant along the axis of the cylinder
and around the circumference. In this case
thermal buckling does not occur.
independent of
When the temperature i..
the circumferential cocrdinate and is constant through the wall thickness but varies
with the axial coordinate, the resulting hoop
compression can, in principle, cause buckling,
Investigations have shown," ' however,
that under the practical conditions'of ballistic missile design the danger of elastic buckling isvery remote and that at most some
localized inelastic dimpling may occur near
restraining ring frames when the tempera
wall thickness and with the axial coordi.
nate.1" A typical situation of this kind arises
when a circular cylindrical shell is stiffened
by means of a number of internally arranged
stringers whic are parallel to the axis of the
cylinder. if aerodynamic heating raises the
skin tempera're vry rapidly and the heat
flow to the stringers is relatively slow, the
stringers are subjected to tensile stresses
and the shell wall to compressive thermal
stresses in the axial direction. Under these
stresses the shell wall backles when the average intensity of the stress reaches approximately the critical value given in Eqiation
(325).
50
,/N
ST RESS ANALYSIS
3.8. CROff EMFECTS
When the upper end of a vertically arr:nged tensile te.4 specimen i4 attached to a
rigid frame, and a weight is suspended from
its lower end, the observer can measure an
instantaneous elastic, and possibly a plastic
elongation at the noment the load is applied.
If the temperature of the specimen is suffi.
ciently high, the instantaneous elongation is
followed by furiher elongations which increase with time even though the load remains constant. These elongations are desigPated creep deformations.
The physical laws governing creep have
not yet been explored sufficiently to explain
completely the reasons for creep and to predict accurately the creep deformations under
prescribed conditions of loading. Moreover,
creep is influenced not only by the chemicai
co.position of the material. but also by heat
treatment and cold work. For this reason
care should be taken in the application of
creep formulas; they may be valid for a
material before, but not after the manufacturing process.
Figure 310 presents the usual shape of
*he creep curve as obtained in a tensile test
carried out under a constant load Ft constant
temperature. rhe instantaneous deformation
is followed by a period of creep at decreasing
rates of strain, the socalled primary, or
transient, stage of creep. After a while the
time rate of strain becomes constant; the corresponding straight line in the diagram represents the secondary, or steady, stage of
creep. Finally the creep rate increases again
in the tertiary range until the specimen fails
by fracture.
ISTRAIN
)'
SECONDARY
INSTANTANEOUSAR
Figure 310. Tensile Creep Curve
I
V
h
T RY
TIME
C and n are conIn these formulas k, I,,
38.!. Secondary Creep. The greatest
stants. Of these, k has the dimension in. per
amount of information is available today on
in. per unit time, the dimension of ). is that
secondary creep. For a given material the
of a stress, the dimension of C is the reciprothe
on
only
depends
d,
'dt
rate
creep strain
cal of a stress, and n is a number. Results of
stress and tLa temperature (tis time). On
creep tests carried out witl a material in a
the basis of test results it appears possible to
given condition of heat treatment and cold
represent the creep rate as the product of a
work indicate which one of the three forms
function f(a) of the stress a and a function
ismost suitable for an analytic ren:sentaF(T) of the temperature T:
constants
tion. When the form is selected, thepossible
(357)
de "dt = f(()F(T)
fit
can be chosen to obtain the best
)((5
d t
of
f(.
ally
accepted
forms
The
most
genm,
points.
test
the
with
curve
empirical
the
of
The:
The form of the function of the temperature
Sare:
f
( .
).
(3 58)
is usually given as
f() =
(3
59)
F(T) =eu T
(3 60) 56
i h C~a
sn
f.(s) 
r.,o
kSTRUCTURES
where the constant Btu b.a. .,e dimension
of a t..mperature.
Although the.;e formulas represent the resuits of constantstress tests, tbey have been
generally used under variable stress conditions also. They are more likely to gve reliable results during loading than during unloading. The values of n in ballistic missile
applications usually range between 3 and 15.
crocp; i.Ns is characterized by =
in
Equations (353) and (3r62). L% . ar creep
is occasiormly observed with high polymers
but not with metals.
Since the material, in general, is subject
to elastic deformations even though crrep
deformations are also present, the Fress
distribution caused by a prescribed loding
depends on the interaction betwean elasticity
and creep, and it changes with time. At the
moment when the loads are applied, the
lna isi iti ui np r al.W t n
creasing tic thi changes over into oe that
is governed by the creep law alone.
This final distribution is the some whether
the creep phenoiieron is re.presented by
Equation (358) ;cr secondary creep or
Equation (362) for primary creep as long
as the value of. n ir. the same. Its analysis
can be carried out with the aid of analogues
and minimal principles derived in the literature" "As an example, the results of the analysis
of
stresssubjected
distribution
a solid rectangularthebeam
to ainconstant
bending
38.2.of Primary Creep. During the primary
~~stage
creep the creep r~te changes with
time. Results of tests carried out at constant
tress and constant tPemperaturie can usually
be represented as
o
=k4 ) '
(&62)
However, this equation is unsuitable under
conditions of variable stress. It can be
brought into a form in which time does not
appear explicitly; this is necessary because
the presence of t in a formula representing
creep strain leads to logical inconsistencies
undr variable stress conditiono. Discussion
of
this can be found in the literature.'" This
assumption
on which the transformation is
Abased is that the creep rate at constant ternperature Js only a function of the instantaneous values of stress and strain but not of the
history of loading. This assumption implies
the ei istence of a mechnical equa'ion of state.
The recommended formulation is
N
s.
moment may be presented. If h is the depth
of the beam and y the distance measured
frcrn the neutral axis of the cross section, the
distribution of the normal stresses (bending
stresses) is characterized by the equation
(Y4 X)',
.
".
i

(2ylh);"'
(364)
where ,,,,,.
is the stress in the extreme fiber
and n isthe exponent in the creep law. The
maximum stress in turn is given by the
formula
(365)
2n+1 Mh
n
61
where M is the applied moment and I the
moment of inertia of the cross section.
It the case of linear creep or linear elasticity
o/c,
= p",,(363)
,herethe constants have the same values as
in Equation (362).
As the value of p is usually between 2 and
3, the strain rate decreases with increasing
strain, that is with increasing time. For this
reason materials whose creep behavior can
be expressed by Equation (363) are often
referred to as strainhardening materials.
Again, tivs creep law is more reliable in
applications in which the load hicreases than
in which it decreases.
38.3. Stress Distribution Accompanying
Creep. The presence of creep 'hange the
stress distribution in a structure from the
one calculated from the linear theory of elasticity. The only exception to this statement
is the stress (li:itribution accompanying linear
2y h
(366)
S...=Mh, 21
(367)
The stress distribution is linear only when
creep deformations are absent, or when they
are governed by a linear law as is the case
with some high polymers. With metals the
final stress distribution is nonlinear since
the creep law is nonln,ar.
,
52
STE~ANALYSIS
~.
pre:?nce of creep to that caused by p~urely
elastice (linear) deformations depends only
on the value of n:V
R.
(38
2n + 1
a iew typical values fol~ow:
n 1
3
7
1 0.78073;0.(1714
91L
0.704~ 0.6I
39.4. Failur2, A tensile speckmen s:ubjected to a con.tant load fails by necking/
down and fracturing through the smallest
C35section. Wrhen the material is metallographicaily s~tble and when its creep deJ
formations follow the law of Bquaticn
(358), the critical time, that is the time
elapsed between load application and fracture, can be given a,,:Fgr
t,, = I ne.
(1369)
wherv i., is the creep rate at the moment of~
load application
i,,= ksv Xy,'
(0, 70)
with k,, the xerimentaiy determined con) stant at th~e proe epeaue
k,= k, e 111.T
(37~
When ~~
~ smealgapia
~ ~~ t71aera
G
Wentemierial las m~~e tyrpial
(3 72)
', =const.
=;
untale
uch
exerient,.
frmdeive
are fte
are ftenderved
romexpeimets.
uch
recent information on creep rupture can be
found in R~eference 24.
With a machine or .r,?hanisnr which binds
and becomes inoperative if a maximum permissible dleformation is exceeded, the analyst
4;
must calculate the time at which the corres.
pondng
stainis
ree eaced.An pprxiponingcrep
tran i rechd. n aproimate evaluation of this time is facilitated by
isochronous stressstrain curves; they are a
family of curves in the stressstrain plane
with t = const. as the parameter. The creep
behvio atdiferet
empratrescanbe
}
~correlated by means of parameters discussed
int
38,5. Creep Buckling. ,When a column
4'(see Figure 311) is subjected to two equal
&and
opposite compressive loads, P, small besiding moments M z Py, arise in its cross se,
(
tions because of tL~e unavoidable deviation,
li
1/
 .DsetdClm
. fte}trln
o h ounfo,
h
o h etrl'e
fteclm
ri
h
line of load application. The small leve'
arms y., upon which the load "' is atn ~
equally be cau.,ed by inaccuracies in the manufacture (if the column and in its centering
in the loading machine. Ur',er these bending
moments the small initial curvatire of the
Column increases in conse~quence c.J bending
creep. The increased curvature, in turn,
leads to increased bending momtnts because
of the increases in the lever arm , y. As creep
in metals is nonlinear, the Jncre.sps in curvatr
n
elcintk
lc eysol
tit the beginning, but they are accelerated in
atraildeecinak
aevrysoy
the later phases, of the creep column testI
unercsttlod
Probably the mo. ;nteresting fact about
the
buckling of a cojumn in the istae
presence
ofcepithtwlehecum
i t nta
lgtycre
tti
e
coeustbew
nthcrppoesicurvature ufciety.oThis increaes intbl
be shown
with the aid of the stressri
dagmofte
aeraesblhd
theeece2
temperature oftecolumn test
Figure (3 '2)..
At the omen, of load applicaldon the
deviation
is so small that virtually the
same stres.s (;,, prevails throughout the cross
section. But as the curvature increases iii
STRUCTURES
consequence of the creep deformations, the
convex side of the middle of the column will
be under the action of a compressive stress
a, considerably smaller than the compressive
stress a2 on the concave side of the'middle of
When the column is an idealized I sectio
in which all the material is concentrated in
two flanges a tiistance h apart, the effective
flexural rigiditv csn be calculated from tLe
values of . 'M
 d: ; with the Wi of the
the column. When the bending moment is
increased slightly, the stress on the concave
side inc.eases along the tangent to the curve.
The slope of this tangent, namely tan a_, is
the tangent modulus E, as discus'od under
paragraph 34.9, On the other band, on the
convex side the stress decreases along a line
parallel to the initia] straight portion of the
stressstrain curve because onily the elastic
part of the deforinaons can be regained.
The slope of the initial straight line, tan ,
is Young's Modulus E.
formula
0'radius
2EE.
(T
EE
(373)
For'other sections Equation (3.72) represents a satisfactory approximation.
The buckling load of the column is
_
L)
and the critical stress can be given as
L
(75)
where L is the length of the column, p, the
Eqf is simply
of gyration of its cross section, and
E2.t= (EI)a./I = 2EE,/(E + E,)
a2
The value of Eef decreases as the deflection
V
/
o
of the column ircreases. Thus the critical
load P,, also decreasei until, zt the critical
value of the deflection, it becomes equal to
the applied load P. At that moment the
column collapses.
the critical deflection wu?/
be ca!culated from the instantaneous stresscurve of the material established at
the proper temperature and after the proper
time of exposure. The critical time is then
simply the time necessary for'the creep deformations to increase to the critical value.
For the calculation of the critical time, data
on creep are also needed. Since all this information is not yet available for the materials
wConsequently,
#
ostrain
EO
of construction, no closed formulas can be
STR
IN
guro J.12. Sfiesstroin Cur
Temperaturo
a Elevated
given for the critical time of columns.
Efforts have also been made to calculate
the critical tLne of initially perfect columns.
The results obtained, however, can,ot be
logically corrtAated t 'the practical buckling
times observed in experiment. :"
(54l
454
STRESS ANALYSIS
3.9. RrEiZEFCESAND BBLOGRAPHY
Y.., 1936.
Bo
.,NwYr,
ElasticStability Analysis jfor Thin CyShells," National Advisory
*lindrical
Committee for Aeronautics, NACA ReunoDnai
port No. 874, 1947.
nDymk
3Fliigge, Wilhelm, Sai
der Schalen, Julius Springer, Berlin,
4Cylindrical
~Hydrostatic
*852,
Shell,."
ational Avisory Commrittee
the Plastic Stability ofPlate and
Sciences, Vol. 16, No. 9, pp. 5294J, Septembir 1949.
12. Gera~rd, George, "Plastic Stability Theof Thin Shells," Journal of the AeronutclSciences, Vo.24, No. 4,pp. 29
74, April 1957.I
1934.
4. Fung. Y. C. and Schier, E. E., "Buckling of ThinWal!ed Circular Cyliniders
under Axial Compression and Jiiternal
Pressure," JauWn of the Aeronawtical
Scienes, Vol. 24, No. 5, pp. 35156, May
13. Hoff, N. J., "Approximate Analysis of
the Redliction in Torsional Rigidity andp
of the I1orsional Buckling of Solid Wings
under Therma! Stresses," Journa~l of the
Ae onuuticalScience3, Vol. 23, No. 6, p.
FiX June 1S563.
L.. H.Rtbltyo.Tiale
i5.
ochnWle
5 Donneffl, L.H,"tblt
Tubes under Torsion," National Advisory
Commnittee for AerQnautics, NACA Report No. 479, 1933.
6. Bijilaard, P. P., "Buckling Stress of Thin
Clamped Shella Subject to
pressure," Journal of the
Aeronautbo'Sciences, Vol. 21, No. 12, p.
December 1954.
'7. isatdori, S. B., et a, "Critical Combina..
N. J., 'Piickling of Thin Cylindrial Shell Under Hoop Stresses Varying
7oitrizal of Appi,d
in AxaI Directie
Mechanics, Vol. k4, No. 3, pp. 40512.
September 1957.
15. Eoff, N. J., "Tihermnal Buckling of Supersonic Wing Panels," Journal of rU Aeronautical Sciences, Vol. 23, No. 11, pp.
101928, November 1956.p
16. Abir, David, "Thermal Buckling of Circular Cylndrical Sheil Under Circum
for ThinWalled Cylinders,"
National
Advisory Committee for Aeronautics,
NACA TN 1345, June 1947.
8. Harris, Leonard A., et 0l, "The Effect
of Internal Pressure on the Buckling
Stress of ThinWalled Circular Cyikers
Under Combined Axial Compression
Torsion," Journal of the Aeronavlicd,.
Sciences, Vol. 25, No. 2, pp. 1423, F'ebruary 1958.
9. Crate, H., et al, "The Effect of Internal
Pressure on the Buckling Stress of ThinWalled Circular Cylinders under Tor
sion," National Advisory Committee for
Aeronautics, NACA APR 14E27, May
1944.
10. Gerard, George and Becker, Herbert,
"Handbook of Structural Stability, Part
III, Buckling of Curved Plates and
',"
Dimsertation, Polytechnic
Institute of
Brooklyn, Brooklyn, N. Y., June 1958.
17. Hoff, N. J., "Ifigh Temperature Structures," Stanford tUnivernnity, Stanford.
California, SUDAER No. 79, June 1958;
prese ited at the 3rd Symposium on HighSpeed Aerodynamics and Structures, San
Diego, California, March 2527, 1958.
18. Hoff, 1'I. J., "Approximate Aialysis of
Struct kres in th, . Presence of Moderately
Large Creep Deformatiens," Qalarterly
of Ap )lief& M ~thernatic3, Vol. 12, No. 1,
p. 49 April 19,"4.
19. Warg, Alexander
J.
and Prager,
Wil'iam, "Thermal and Creep Effects in
WorkHardening Ela.dicPlastic Solid,"
Journal of the Aeronaidical Sciences,
Vol. 21, No. 5, pp. 3434, 360, M'ay 1954.
20. Hoff, N, J., "Stresis Distributioi, in the
STRUCTURES
Presenca of steady Creep," P'roceedings
of the Coiference on ffighSpeed Aeronautics, Polytechnic Institute of Brooka, ""rooklyn, N. Y., 1955.
21. 7.*off, if. J,, "Effets thermiques dans le
ca,'cu) (!e la r~sistence dea structure~s
24. Heimerl, George J., "TimeTemperatre
Parameters and an Application to Rupture and Creep of Aluminum Alloys,"
National Advisory Committe,. for Aeronautics, NACA TN 3195, June 1954.
25. Fraeijs de Veubeice, B., "Creep Buok
d'avions et d'engins," Advisory Group
for Aerrnatztical Research ak:d Developmernt, AGARD Report No. 52, Paris, January 1956.
22. Sanders, J. Lyell, Jr., et al, "A Variational Theorem for Creep with Lpplications to Plates and Columns," National
Advisor Committee for Aeronautics,
NACA TN 4003, May 1957.
23. Hoff, N. J., "The Necking and the Rupture of Rods Subjected to Constait Tensile Loads," Journalof Applied Mecha?;ic.,, Vol. 20, No. 1, pp. 1058, March 1953.
ling," Chapter 13 of High Tertera,. :v
Effects in Aircraft Scructures, edited b.
N. J. Hoff, published for AGARD by Pergamon Press, London, England, 1958.
26. Chapnan, J. C., et al, "A Theoretical
and Experimental Investigation of Creep
'Buckling," Polytechnic Institute of
Brook!yn, Brooklyn, N. Y., P1BAL Report No. 406, October 1957.
27. Hoff, N. J., "A Survey of the Theories of
Creep Buckli;.g," Stanford University,
Stanford, California, SUDAER No. 80,
1958.
JI
._
41
24
Chapter 4
DESIGN CONSIDERATIONS
41. GENERAL COMMENTS
The design of a complete weapon system
is a rather formidable task requiring 'he
services of scientists and engineers in a 'variety of fields and a high degre of coordination of effort.' While attentior in this chapter is focused on structural design consideratiors, it will be seen that these c )nsiderations
are intimatel connected with the guida!.ce,
control, propulsion and other aspects of thu
entire missile design.
The structural design prolem can be visualized in a manner similar to closed loop
systems for the guidance of aircraft and missiles. One starts, for example, with a study
and establishment of the envronmental conditions, goes cn toa consideration of the mission of the missile and then to a :etup of the
design criteria. At this stage, the choice of
a propulsive system may be made. The structural analysis and design is undertaken, followed by laboratory tests of components or
assemblies. The ainal stage involves the flight
tests of the vehicle with instrumentation to
check the environment assumed at the outsa
Thus one arrives at the start of the .
and refinements are made by going through
the loop as many times as necessary.
It may be well to distinguish between
structural analysis and structural design. In
structural analysis, the structure, loads and
boundary conditions are given, and the
analyst is required to find the stresses and
deformations. In a design problem, however,
a set of loads are to be transmitted o'. er given
distances, subject, for example, to a weight
and stiffness specification. The desigver mus.
and analysis portion of the closed loop outlined above. Some of the inost important
aspects of the environment, mission, or design criteria of the missile are considered,
one at a time. and their effects on the design
are noted. While the treatment is qualitative
in nature, the discussion will point up the
manner in which the selection of materials,
structural configuration, and crosssectional
areas are affected by the environment, mission and design criteria.
The radically different environments and
mibsions of the many varieties of missiles
precludes the possibility of establishing one
set of desi.4n criteria. A particularly good
example is the importance of the thermal
environment, especially it the case of long
rarge ballistic mL;ziles. For these missiles,
high speeds are obtained ' high altitudes
and a whole array of pro .em., arise when
the missile enters the lower, denser atmosphere at the end of its trajectory. These
are usually called re.entry problems and involve all aspects of the structural design.
De..ign criteria must be intelligently set up
for each missilc desigii from a carefA! study
of all the information available on the envir.inment, mission, requixement., of rigidity,
deformation, weight and balanc,, and others.
42. AERODYNAMIC 1NFLUENC! ON
ERODNAM. ONFURATION
figuration, and hardware crosssection which
he feels will meet the specifications. The
structure, under the given loads, is now
checked against the requirements, and one
is back to analysis once more
The accent in this chapter is on the design
tions on the overall size and shape of the
structure, and therefore on the capacity to
transmit applied loads. If the restrictions
are very severe, it may be necessary to seek
better materials or more eficient st;uctural
configurations. A typical example is the wing
One of the major problems confronting
do.igner.; of high speEd aircraft Pnd missile.s
is the roic that aerodynamics plays in fixing
the external confguration of the missile.
From the structural designer's standpoint,
aerodynamic consideration, impose restrie
tentatively choose a material, structural cor,
57
*1F
STRUCTURES
''
of a high speed aircraft or the contro, vanes
of a rnissilp. The wave drag is proportioaial
the square of the wing thickness, hence the
acrodynarnicist would like to have as thin a
wing as possible. The ability of the structure withi the wing to carry certain loads,
however, increases approximately as the
square of the wing thickness. In general,
such cases of conflicting inteiELs are settled
by a compromise; this is almost always true
in design problems. Should he aerodynamic
requirement be very severe ard have high
priority, decreasing the wing thickness may
lead to a prohinitively high weight for the
wing design. Jn this eve'., the structural
designermay have to"study lhe possibility of
a more efficient structural configuration,
stronger alloys of conventional materials, or
even the use of a neiv material,
in The example ci ed has been over implified
inorder to illustrate the maner in which the
interests of one group affect those of anc.her.
Onr could go fuether with Che above exas ile
cnd show the effect that an extremely .'in
wing may have on other aspects of the dcli:,,.
The thin wing may introduce problems in
connection with the housing of certain cornponents. Generally speaking, changes in one
design feature will affect many ocher aspects
of the design; the importance of making intePigent compromises cannot be overemphasized in the success of an aircraft or mis.sile design.
4.3. AERODYNAMIC HEATING
Probably the most serious problems aris
lyn. The model showr in I'igure 41 was
made of cast aluminum alloy and had a wall
thickness f 9.5 inch. After 50 seconds of
ieating at a tunnel stagnetio: presstre bf
16) psia an.. a stagnation tpnperature of
1995 "R,a moderate amount of melting occurred at the stz.gnatior region. Of interest is
the symmetric manner in which the molten
rietal recrystallized after being carried
downstream to the relatively cooler surface
of the model. The model shown in Figgre
42 was tested for 62 seconds. The model and
test conditions were identical to those dpscribed above except that the stagnation ternperati:e was 203 0 R. The melted portion of
the nose cone was observed to have disintegrated into relatively,large particles.
Structural design for aerodynamic heating
is undobutedly one of the major problem
areas for the missile designer and analyst
Under the influence of high heating rates,
and the resulting high temperatures, mater'als which had performed satisfactorily in
the paet may be wholly inadequate, In the
most critical regions (nose cone, control surfaces), several solutions may be considered
by the designer to cope with the adverse
effects of elevated temperatures on material
properties and structural behavior of missile
components. Even in regions subject tc moderate conditions of thetmal environment, the
structural design problems are iormidable.
Structural design configurations for aerody.
namic heating may be conveniently classified
as (1) unprotected, (2) inbulated, (a)
coolcd, or a gombination of these.
ing from the aerodynamics of flight that
43.1. Unprotected Structures, Unprotected
affect the missile structural designer are
tho&e associated with aerodynamic heating.
Depending on the speed, character of the
boundary layer, and other factors, the heat
input over crtical portions of the s~wface of
a missile can attain values of the order of
structures may be defined as structvres that
do not employ any fluid coolant or special
insulators to deal with the thermal input
from the boundary layer air. They present
unusually difficult design problems inasmuch
as the designer is concerned directly w'*th
10" Btu/ft"/hr. The qualitative and quantitative aspects of thi3 phenomenoai are coyered in another volume of this series,
The effects of aerodynamic heating on nose
cone models are shown in Figores 41 and
42. These tests were run in the hypersonic
tunnel at the Polytechnic Institute of Brook
material properties and structural behavior
at eleva ed temperatures, aeroelastic, and
propellant task problems. In addition, the
designer must seriously consider transfer of
heat to the interior of the missile because
there may be temperature limitations on the
proper functioning of electronic and other
!o
58
DESIGN CONSIDERATIONS
.iue41]. Effects of Aerodynamic Healing an Model Nose Cone, Stagnat~an
Temperature I 995*R
Figure 42. Effects of Aerodynamic Heating on Modol Nose Cone, Stagnation
Temperature 2030"R
59
LJ
STRUCTURES
L,
eqaipi ant. The unprotected design should
that minii.:v.s the thermal stresses caused
of course be compared with insuleted and
by uneven distribution of material and varicooled designs in order tr arrive at an opiiations in the heat transfer coefficient. For
mum of r:;,fig4:ation.
winglike structures, trussed and corrugated
Unprotected structures ha~e been classiwebs and ribr have been investigated. Confled  into three categories depending on the
sideration has also been give, to sandwich
manner N which the effects of aerodynamic
construction of the surface to surfate and
heating are handied: (a) heatsustainlng,
individual surface types, and the ze of shal(b) beat sink, and (c) thermal stre.s reduclow, closely spaced s.ringers. In ech of the
ing types. On the heat susaining type of
configurations mentioned thb. basic idea is to
structure t.e most complex problems arise
avoid concentrations of material which give
from the effects of th3rrmal expansion which
rise to temperature differences, and hence
cause structural growth, structural distorthermal stresses, in a structural component.
tion, and theri.ai stresses. Structural growth,
The ::losely spaced shallow stringer configfor example, q:ves rise to detail design prob.
them down into a numIser of smaller heat
lemni in the Ltachment of internal equipment
sinks. The shallow depth keeps the average
to the missile body and of surface control
stringer temperature close to the skin tmwhich compensates for growth of the struc.
perature, thus reducing the thermal stresses.
ture. Structural distortions are unavoidable
Introducing a flexible connection or support
in the unprotected type of strnture. Even
between components avoids a build up of inif equilibrim temperatures are attained, temteral stresses due to constraints.
peratures are not uniform over the surface
of the structure due to variations in heat
43.2. Insulated Structures The insukto
structure introduces the conccpt of one structransfe" coefficients and local concentrations
ture to absorb the primary loads and a second
of material which serve as heat sinks. As
structure to cope with the thermal input.
mentioned in the early part of this chapter,
Consider, for example, a corntional strucvariations in temperature from an average
tura configuration of a material which may
temperature give rise to thermal stresses and
attention must be given to the possibility of
iais A fair oinsulatg mater
stress concentrations or thermal bitckling.
be
In the h ?at s ink ap pr o a ch , ex tr, n at~ ' a l
k eapplied
p t e to
e mther surface
a u e o ofththel structure
ad c ry to
n
is added to the external surface of the miskeep the temperature of the load carrying
sile. Tbis addition of material has the effect
posite type of structure, however, requires
of reducing temperatures, increasing mateeite
reior
stc
h
r re
rf
rial strength and reducing structural growth
the insulating material in order to compare
and distortions. The thermal stresses, howfavorably with the double wall solution.
ever, di.sy not be changed significantly. This
Moreover, the insulatitig material has strinsolution to the aerodynamic heating problem
soltio
adiabatic wall tmpera e gent mechanical and thermal requirements.
whentotheaerdynmicheaingprolem
is most efficient
It should be erosion and beat resistant, and
tures and flight times are such that additional
characteradhesive and
good ceramics
possess some
mustAlthough
it
strucalloy of
a lightheats
added to
canisbebecause
material
refracistics.
the
specific
ture. This
light alloys are considerably higher than the
specific heats of alloys of steel and titanium.
For very short exposure times, the mechanical properties of this configuration are far
onehalf 2hour
superior
to thosetimes
shown
mayinbe Chapter
for
wh ich exposure
c eo
ue
or more.
tories appear to have some of these esirable
properties, it seems that the necessary requirements are too severe to be met by a
single material. Many refractory materials
shock reare brittle and have low thermal the
in h oxide
sistance. Although ceramics
ssac.Atog
eaisi
xd
family have a lower thermal conductivity
than those in the silicide, boride or carbide
The stress redtcing type of structure A.an
Otempt to evolve a structural configuration
groups, a 0.10 inch coating of aluminum
oxide would have little insulating value in
60
4~
"
"
Hi
DESIGN CONSIDERATIONS
heat sink is quite different from the heat sink
uprotected type structure. In this latter
case, the mass of the unprotected primary
structure was simply increased in order to
enlarge its heat absorption capacity. The
heat sink approach has received a great deal
of attention bItcause of its simplicity. Copper
appears to be a suitable cover material for
heat sink purposes primarily because of its
excellent thermal diffusiviry. The poor high
temperature mechanical characteristics of
copper have not affected its use for this purpose.
Should the thermal input conditions be exceptionally severe, the designee may still
utilize the heat sink appr.ah, except that
the heat sink material is allowed to melt.
The molten material is then ablated or carried away by the windstream. This approach
has been used successfully in several reentry
bodies. One of the major problems in the
application of this solution is the instability
of the ablation process. Any irregularity
or pit in the surface may increase the heat
transfer locally, melting material at a greater
rate and precipitating a very rapid and
wholesale ablation of the material in the entire area.
Sublimation cooling is another interesting
and promising approach to the aerodynanic
heating problem for very high thermal inputs
.nwhich the thermal shield material itself is
sublimated at the surface, The sublimation
cooling solution is here classified as an insulated structure inasmuch as there is a "orimary structure which is being protected by
the outer cover of material which is sublimating In common with the nonmelting and
melting heat sink discussed in the preceding
paragraphs, the sublimation cooling technique is basically one in which a thermal
shield protects a primary loadcarrying
structure. Materials which show promise as
outer covers, are poor thermal conductors,
and it appears that there is no necessity to
interpose insulation between the primary
structure and the outer cover.
The principal beneficial effects of sublimation cooling are in the reduction of the heat
transfer coefficient due to introduction of
egions of moderate to high thermal inuts.
Developmenta! work in this directicn is
nevertheless worthwhile bemu.L of the sirplicity and low cost possibilities
A more promising and feasible insulated
structure is a double wall construction. The
outer shell is fabricated from a heat rEsistant
material and is relatively thin and light,
The panels of the outer shell esse:atia!ly carry
only the local normal loads, and therefore
need be designed to withstand these. loads in
only at the elevated temperature. The inner
wall or shell is the primary loae carrying
structure and may be cunventionai n design
and material. The reason for this is that the
inner structure will not be sub.*ect to high
thermal inputs and the temperature rise is
kept low enough that material properties are
not significantly affected. Interposed between
the walls in a suitable light and effective insulator. Fo relatively close sracing between
effectively as
the walls, air aoi. mnzht
the insulating medium. If the outer panels
are designed in such a manner that each individual panel can expand or contract freely,
it offers the advantage o! complete absence
of thermal stresses. The cuter phnels can be
allowed to float by attaching them to the primary structure with thin strips or clips
which allow expansion of the panel in its own
plane. In additirn to tie advantage of complete absence of thernid stresses, other features of this design .nclude its feasibility
from the standpoint (if developmental time,
heat protection for in.ternal equipment, and
freedom from panel flutter.
Some ol the most interesting insulated
structure solutions co the aerodynamic heating problem have been proposed for missiles
of long range reentering the atmosphere at
high speeds. One of these solutions is the
nonmelting heat sink. In this application,
the primary or loadcarrying structure is
covered with an engineering material in
sufficient quantity that it can absoib the
thermal input without melting. Insulation is
interposed be':eewmn the outer heatabsorbing
material and the inner lcadcarrying structure in orde. to keep the lcadcarrying structure relatively cool. It is te be noted that this
61
fQ
STRUCTUPES
fluid material into the boundary layer, and
in the large heat required 1or sublimation.
It is also reported that uneven heating or
local hot spots are absent Moreover there is
a builtin stability in the system; local high
heating increases the sublimation rate at that
location, thus affecting a reduction in heat
transfer. A possible defect in this approach
exists in connection with the use of low heat
of sublimation materials which may blow so
much material that the characteristics of the
boundary layer could be altered. This could
lead to large scale turbulence phenomena
with attendant higher heating rates. It has
also been reported that burning accompanyinkf sublimatio d,o rnot seem t;.
be a serious
matter.
Another method for protecting the structure consists of the application of surface
material must be porous in order to allow
the passage of the fluid through its thiclnes.
The structural material may be manufactured by powder meta.uy technq
constructed by compressing layers of screens
to the desired thickness. The fluid may be
either a gas or a liquid, but probably a liquid
coolant would be preferable in order to take
advantage of the large heat absorption asaociated with the phtse change and the decreased container requirements. There are a
number of interesting features associated
with this method of cooling a structure. it
appears that blowing through the boundary
layer reduces the magnitude of the heat
tA
nrsfer ccefficiert which is. of course, desirable. There is also a builtin stability in
the system inasmuch as a local hot spot will
evaporate more fluid at that point. 'Under
coatings which maximize the radiation of
heat Materials such as cerium having high
radiant emissivity characteristics may be
utilized for this purpose.
these conditions, more heat is absorbed at the
location in question anzd the net effect is a
tendency to make the thermal inputs more
uniforra over the surface of the missile.
Structures employing a combination of cooling and insulation techniques are also feasible,
43.3. Cooled Structures. Under certain conditions of the missile mission and expected
thermal environment, a cooled structure may
be necessary or desirable. It appears desirable to consider direct cooling of the primary
structure for small missiles where space is
not available for insulation or for very short
flight times.
Cooling of the primary structure may be
accomplished by an internal cooling system
or by porous coolirPg. In the internal cooling
system, a fluid is circulated through the
structure to effect a transfer of heat from
the structural material to the fluid. The
coolant may be circulated in channels which
are an integral part of the structure or otherwise in separate ducts which are in intimate
contact with the structure. For long range
missiles or where there is a reentry problem,
it appears that the high heat capacity of the
low atomic number metals is an attractive
property. These metals would be used in the
molten state. Some of the undesirdable characteristics mentioned have been their poor
thermal conductivities and difficulties of fabrication and handling.
In a porous cooling solution, the structural
k'
44. DESIGN FOR APPLIED LOADS
Design for the transmission of the applied
loads requires a careful study of critical loading conditions, material properties and structural configurations. In order to enable the
structural designer to predict required hardware crosssectional areas, there must also be
available a rational set of structural design
criteria. Where there exists a long history
of experience with a particular type of structure and its environment, adequate design
criteria have been established. Unfortunately, this is not the case in the missile field.
The experience accumulated thus far is relatively limited, and a great deal of complication has been introduced by the dependence
of the design criteria on both temperature
and time.
While details of the considerations involved in obtaining design loads, factors of
safety, and allowable stresses and deformations will not be discussed, the more carefully these items are selected, the more effi
62
DESIGN CONSIDERATIONS
missiles until such time as separate design
criteria are established for missiles or if the
procuring agency dos not specify safety fLctoiz in a specific missile development. ]a,tors cf safrt y are applied to limit loads to
take into a ccount neertainties in coniectior.
with the establishment tif the limit !oadi.,
cient will the design be from the structural
standpoint.
Stress analysts fre44.1. Types
quent1
spaes ofofLoads.
primrk
s.
nd second
ofc
loads. Primary loads are the maj,,r loads
that must be resisted and transmitted b, tntL
sructure. Consider, for example, that under

a particular flight wzondition, certain loads
are acting on the missile structure. If we cut
off tne tail portion and examine the equilibrium of this portion as a free rigid body,
then three forces and three moments will be
required at the section of the cut in order o
maintain equilibrium. These may be considered as the primary loads being transmitted
Sat the section,
F.
Secondary loads are selfequilibrating.
They are a consequence of structural defermations rafnez than externally applied loads.
Thma ratheere
an exnalpledseoad,
second
Thrmlstresses Areanhexampuh
ary stresses. Although secondary stre ens
Scan be disregarded in some branches of engineering, they cannot be overlooked in missile structural design.
The principal loads on the missile structure areaerodynamic, inertial, and thermal.
Therearealso important loads which are imposed on the structure that do not fall precisely in the above categories, but which are
tnevertheless important in determining the
structural design. These include loads inpropulsive system,
by thepressurization,
duced
vibraslosh,induced
fuel those
by internal
essurizatontfue s
, viraboyintern lo
tory loads, and loads encountered in transportation, handling and launching.
44.2 Limit and Ultimate Loads; Factor
of Safety. The loads that are actually anticipated during the lifetime of a missile are
called limit load8. These loads may be calculated by specialists in various fields and
the information made available to the structural analyst and designer. It is possible that
data in the form of tables and curves can be
made available to the designer for preliminary design purposes. Limit loads are multiplied by a factor of s8fety to 'jbtain ultimate
loads. In the design of aircraft, factors of
safety are specified; these can be applied to
properties
a.lithods
of materials, and
stress analyss.
of
tural DeinAsue tha tydiner s
ural Design. Assume thathe designer has
tentatively chosen a thinwalled circular .ylinder as a structural configuration; this
choice based upon his experience and information available on the expected environment
and mission of the vehicle. It may be convenient to consider that aside from the nose
cone and control vanes, the circular cylinder
comprises the entire body of the missile.One
can now confine himself to the applied loads,
and inquire into the efficiency of the structural configuration, and into the effect that
the magnitude of the loads and design criteria have on material selection and required
on areas.
c igure 43 i a photograph of he tni section of the Jupiter missile illustr.tive of
_ructhe thiwalled circular cylinder as a
tural configuraion. Note the extremely large
ratio of radius to wall thickness and the circumferential reinforcement rings, The photos The pto
re 44so
,ph in Figure 44 shows the top section of
ggraphi
the Redstone being inspected prior to movement to the assembly area where internal
equipment is installed. In the background is
the nose cone in which the warhead is installed.
Treating the circular cylinder as a beam,
the cylinder is required to transmit three
f'irce and three moment components at any
crosssection. For simplicity, the two force
components in the plane of the crosssection
can be obtained Into a single shear force, and
the two moment components in the plane of
the crosssection into a single bending moment. The remaining foree component is an
oxial force, the rtmaining moment compcnent is called the tor 1 e.

6
63
5TRUCTURES
ff
Vovo0
~,II
r
DES!GN CO?'JSIDERAVONS
K'
LI)
I
I
J.
4,'
***
?*/
2'
N
K
&
~
1/
'~
St~
~'.~
)
65
2~~
0l
STIZ4CW1RF
Taking the shear force in the plane of Ube
crosssection passing through the center of
the circle, this force is tranbrii tted by means
f shear stresses which are symmetrically
dispose about the line of action of the resultant force. The variation is sinusoidal;
varying from zero at the point of intersection
of the line of action of Vhe force and the cy.lind,", to h maximum value at 90 degrees, and
back
,zezo at lMl0 degrees. The maximum
shear stress is twice the ntverage value obLa:ed by dividing theforce by the crossseetior;al area. The axial force gives rise to a
uniform nomal stress distribution whoae
value is given by the axial force divided by
sisted by hoop or membrane stresses. Curved
twodimensional elements are rturally effieient in this resect. A flat twodimensional
eleimen, or plate, can resist normal pressures
only by means of its bending resistance. This
generally gives rise to fairly high stresses.
On the 9ther hand, curved twodimensional
elements are able to supply tension force components opposite to the direction of the preasure vector and therefore ar.used wherever
pos ble to withstand high normal pressures.
()
4.5. STRUCTURAL FAILURE
In analyzing 'he integrity of a design, thb
stress analyst considers yield and ultimate
the crosssectional azea.
Turning now to t~e noments which are to
be transmitted,
it c~n
that the
be ranmitedit
an bee sown
zowntha
th thinthnwalled circular cylinder is about one and onehalf as efficient as a solid rectangular section
Comparing these cxosssections on the basis
of equal areas and equal maximum normal
stresses, the criterion is simply the sectu
..
wdu!,us, which is the raio of the moment of
rtrength, stiffness, fatigue strength, and
buckling Assume that the designer has tentatively
chosen a structua! configration and
tatel
o
atrcua conution
the material or materla of construction.
Also assume that teitative crosssectional
areas have been established. The primary
object then is to study the manner in which
the design criteria affect the crosssectional
inertia to the d;stance from the neutral axis
to the extreme fibers. The relative efficiencies
of the Ibeam, circular cylinder and solid raectangc!i section are 1, '/,w and i respec
arEi Atastructme
very es'Jj
stage)ein
design
a
tentative
must
laitheout,
or it
would o be possible to cYculate preliminary
tively. The thinwalled circular cylinder has
a moment oZ iaertia which isconstant about
any diameter. Thus this shape can abs9rb a
bending nomegit acting in any direction with
equal facility.
The remaining moment component, the
torque, is most efficiently transmitted by the
thinw led circular section. In a circular
section the twisting rigidity (torque per unit
angle of twist) is proportional to its polar
momert of inertia. A tubular crosssection
has a higher polar moment than a solid circular section of the same area. The torque
is resisted by a uniform distributon of shear
stresses, and the moments by a distribution
of normal stresses given by the wellknown
'tresses, ix.e {ia loads and other quantities.
45.1o Yield and Ultimate Strength. The
criteria established in analyzing the yield
and ultimate strength of a structure are essentisly the following. At limit load, the
structure should not suffer more than a certain, maximum permanent deformation. A
0.2% permanent set is ustally specified. The
integrity of the structure should be maintaied, i.e., no collapse or failure up to the
ultimate load. This means that the stress
analyst, in checking the yield strength of the
structure, must perform stress calculations
for the limit loads under a variety of environmental conditions and ascertain that the
stresses nowhere exceed the yield stress of
MC/I formula.
It is to be noted that considerations of
bukling, concenttated loads, edge effects and
others have been disregarded.
Normal pressures acting over the surface
of the cylindrical shell are conveniently re
the material of construction. In order to simplify the calculations, the analyst generally
investigate. the most critical conditions and
the most critical portions of the structure.
If the maximum stresses are below the yield
stress of the material, then the structure has
aerodynamic loads, thermal inputs, thermal
66
DESIGN CONSIDERATIONS
sumcient yield strength. To check the vltimate strength, the analyst multiplies the
limit loads by a factoi of safety and goes
through the stress calculation. The maximum
str&sses must c=erywhere be less than the
ultimate stress of %he material
Should the yield stress of the material be
exceeded under limit loads, or the ultimate
stress exceeded under ulfiPate loads, then
the structure is obviously too wPak and the
design must be strengthened. One of the
most straightforward measures that can he
taken by the designer is to increase the crosssectional arsa suffici nly that the maximum
stresses are reduced below the yield or ultimate strength criterion, depending upon
which one is critical. In the case of tic thinwalled circular cylinder, the designer iamy
increase the wall thickness. This solution, of
course, adds weight to the structure which
may well be intolerable. The designer may
then consider another alloy of the same
material with higher yield and ult;mate
strengths. This may give a satisfactory soluion, prcvided that no trouble is anticipated
due to increased cust. manufacturing difficulties, the possibility that the material is toe
brittle, and other considerations. The designer may also investigate the desirability
of altering the structuralconfiguration or the
Stictly speaking, the stiffness criterion is
not directly related to the strength of a structure. A stiffness criterion on a structurc ,or
structural component may be establIshed because too much flexibility of the structure
mi.r Inerfere with the proper operation of
the control system, er may seriously affect
tte stability and control characteristics of
the mixle. Consider the design of control
Unkage systems and their supports. it is
obvious that if the supports were attached to
a very flexible structure, the supports would
fall out of line under the appEed loads and
could :ender the control system inoperable.
The same would be true if the control shaft
itself were to deflect transversely between
supports. To insure the proper o eration of
such control syctems, the design is governed
by a stiffness criterion
A number of other examples may be cited.
It is ixrportant to consider the relative deformation between the sensing instrument, say
in a stabilization system, and the control surface. If too much twisting or bending of the
structure occurs between tLe location of the
sensing equipment ard the control surface,
the desired stabilization may not be achieved.
It is also possible to adversely affect the stability derivatives of a missile if there is excessive deformation in bending, or control
__e of another material
efctiveness may be impaired if a control
A measure of the stiffness of a structure
is the product of Young's Modulus of the
surface does not have suif ;ient torsional stiffAnes. The important point that mustbe borne
in mind is that rigidity requirements are important conisiderat.ans in the design of a mis
material and the geometry of the crosssection, namely the moment of inertia. This product, EI, is called the bending rigidity. Torsional rigidity, GC, is also the product of a
material elastic constant and the crosssection. G is the Modulus of Rigidity and C is
a property only of the crosssection. For
solid or tubular crosssections, C is the polar
moment of ineria of the section. The stiffnecs of a structural element is also affected
by.the magnitude and type of load to which
it is subjected. Compressive loads tend to
decrease the effective stiffness of a structural
element, whereas tensile loads have the opposite effect.
sile and some of its components.
There are instances in which a rigidity
requirement is imposed in order to safeguard
the integrity of a structure. Under some
flight conditions, the aerodynamic fcrces may
vary in resonance with the natural frequency
of the structure which in turn is a function
of the structures stiffness and mass distribution. This resonance can lead to catastrophic
failure. Interaction: between aerodynamic,
elastic, and inertia forces are called aeroelastic ,.flects. The speed at which these
forces resonate is called the flutter speed and
the phenomenon is called flutter. The components of a missile that are most apt to
.21RU ICTU
45L
. TI1ES$
67
STRUCTURES
flutter are the control surfaces. Individual
panels of a structure may also flutter under
the proper conditions.
The above considerations are very important and the aeroalastician must calculate
the flutter speed to insure that it is well above
the speeds expected to be encountered. The
analytical compatations cannot be performed
until the design has progressed sufficiently to
enable the analyst to determine the elastic
resistace and mass distribution of the structur& Aerodynamis and structural damping
are usually included in the analytical formulation of the problem. Once the missile has
been constructed, an experimental chec1& can
be made of the structural stiffness by
statically loading the member or missile and
observing the deflections a.d slopes, or by
forced vibration tehnique bv which the
stiffness can be deduced from the amplitude
and frequency response of the structure.
it may also be worthwhile to mention that
liquid sloshing and liquid impact against
bulkheads P.t thrust cutoff may affect rigidity requdrements of certain structul cornponents.
additional oscuflating loads on the fin s
po rtpoints and thereby creating the possibility of fatigue filure.
The problems associated with vibratory
loads are more in the nature of their effect
on the performance of the missile rather than
with the strength or integrity of the structural cpmponent. A vibrating panel could
conceivably affect the aerodynamic behavior
of the missile. At times the problem is assc.
ciated with the proper functioning of electronic equipment or with .acoustical transmissiom. There are also ,other problems
created by the vibratory loads, some of which
may be of greater concern to some designers
than those mentioned above. Having noted
that fatigue is not the important consideration, but rather one of interaction with other
phaass of the missile design. the question
raised is in connection with the step3 that
the structural designer can take to alleviate
the undesirable conditions.
One obvious method of approach is toeliminate the vibratory forces at their source.
This cannot always be accomplished. However, if the source of the virbatory loads can
453 Fatigue. The design of a structure
for vibratory loadv is a source of wr.jor concern for all missile structural designers, A
vibratory load on v,eructural elemet immediately gives r!K to the possibility of
ure by fatigue. For flight vehicles that are
designed to last (in principle) an indifinitely
long time, the possibility of fatigue failure
is important and always inv.stigated. For
ballistic missiles, however, the relatively
short life duration usually precluleb 'he pQssibility of fatigue failure. Fatigue faiiure
should be investigated in a missile design,
but it is generally not important as a design
consideration.
An example under which a ballistic missile
may be subject to fatigue failure can be
cited. A ballistic missile is erected on its
launching pad in such a manner that the
stand load is introduced at the tips of the
fins. The missile may remain in this position
for a relatively long time. At a particular
speed, lateral winds passing over the fins
way cause the fins to resonate, introducing
be traled to theengine, or if vibrations occur
in transit aid storage, a shock mounting
technique may be effective. In other cases,
the designer must seek to increase the 3tiffness of the strciure or portion of the stru 2:e which is involved. Inasmuch as the
structural stiffness is proportional to Ahe
rroduct of Young's Modulus (or the mo'duius
of rigidity) and the geometry of the cro.ssection, the designer can either choose a nev
material with a higher Young's Modulus, or
he can change the crosssection. Changing
the crosssection to increase the stiffness can
be achieved by altering the configuration
without increasing the area, or by increasing
the area.
45.4. Buckling. She. r etresses and normal
compressive stresves v_ng on thinwalled
elements give risc to tl possibility of buckling. Fortunately, buckling of a curved twodimensional thinwall element requires a
stress which is much higher than the buckling stress of a flat elemeat. The buckling
strese of a square fiat plate simply supported
68
A
DESIGN CONSIDERATIONS
on all four sides and un.tormly .ompressed in
the case, hower, if the bar in iuest" )n is
one direction can easily be calculated for
elastic buckling. If the plate were to i~
rolled into a cylinder, the critical stress due
to uniform axial compression would be
about 200 times higher basd on an original
square plate of 12 inch sides and a thickness
of J.032 inch.
Failure by bickling is at times the decisive criterion in the selection of a material
of construction and in the determination of
crosssection areas. As an example of how
buckling considerations can affect the choice
of a materipl, consider a fiat plate subject to
edge compression in one direction. The buckling stress of the plate varies with the square
of the thickness. A comparison of the buckling stress is desired between two plates; ona
but one bar of a redundant framework, because the other bars in the framework can
absorb additional loads. A flat or curved twodimensional element in edgew.e compression in one direction will buckle a, a relatively Jow load, especially if the thickness is
small. This does not correspond to the maximum load or failure if the unloaded sides are
supported. The material near the supports
is constrained against transverse deflection
and can absorb loads beyond those corresponding to the bu'kling load.
of aluminum and one (.f steel. The plates are
previotsly disussed are applabe not only
to have the same length and width, and the
sam,. weight. Since alurninum is about 1/,3
lighter than gteei, the thickness of the aluminum plate can be three tihnes thicker for
the same weight. Inasmuch as the buckling
Atress varies with the square of the thickness, the aluminum plate .Aill have a buckling
strtss that is approximately nine times that
ol the comparable steel plate. This is one
of the strong reasons that aluminum alloys
are selected over steels in the design of most
aircraft and missiles that do not operate at
temperature in excess of say 400CF.
if buckling of a structural component resuits in the collapse or.failure of the entire
structure, it is clear that the buckling stress
of the structural component would be the
maximum allowable stress. In such Ln event,
the structural designer must increase the
crosssectional area of the componeat until
the value of the stiess falls below the maximum allowable. It might be proper to point
out, however, that buckling of a structural
component does not necessarily correspond to
failure of the entire structure. Consider, for
example, a structure composed of a single
bar in compression. The buckling load and
maximum load or s11 onedimensional element are identical fo k.': practical purposes,
Buckling wculd correpord to failure o both
the elemert and the structure. This is not
to missiles but also to other types of flight
vehicles and to many types of ground babed
structures. What problems in this area then
are peculiar to the design of missiles? The
answer is that if one again goes back to
considerations of aerodynamic heating and
finds that the thermal inputs are significant,
then design criteria established in connection
with the strength and stiffness requirements
mupt be reexamined. Reference 4 contalps
a more detailed discussion of the strength,
stiffness and fatigue requirements for both
cold and elevated temperature structures and
proposes revised design criteria for struc.tures at elevated temperatures.
Briefly, the complicating factor is the dependence of the design criteria, and even the
stress analysis itself, on temperature and
time parameters. Investigation of the yield
strength of a heated structure may serve as
an illustration. At room temperature, the
yield strength of a tructure, in the simplest
case, is a matter of keeping the stresses
below the yield' stress of the material used
in the construction of the structure. Room
temperature mechanical data of this sort is
abundant and easily available. At elevated
temperatures, however, the mechanical properties of the raterial may undergo significent
changes. Inasmuch as these data are also
available, it would seem logical to use the
46. 1319E AND TMPERATURE DVENDENCE
OF MIGN CRITEIA
Most of the mechanical deaign criteria
69
STRUCTURES
yield tres at an appropriate design temperacause the differential pressure wil! cause
high hoop stresses in the ,hot skin of the
ture as a yield strength criterion. While
comparvment, significant creeping can occur
this procedure may appear logical and exi: a period of time corresponding to a few
pedient, the time parameter cannot be
minutes.
ignored.
Similar remarks can be made in connecA ballitc missile may be subject to significant thermal inputs and temperdtures for tion" with temperaturtideload eIfects on
other structuralstre'agth equiremnimt. The
relatively short time periods, i.e., of the
immediate objecive of the :preceding disorder of several minutes. Thf temperaturecussion was to show the significnt effects
time nistory is generally transient during
that the thermal environment ad in the
this period. Even if u design temper.ture
tia the
esign
str al design
tiate
could be established for this period, the ultimate
structural
criteia that were
elevated temperature mechanical data at the
r.1,
same temperature may not be strictly app pl established.
tis importnttpotit
outthat the detercable because these data are usually obtained
rIination of the loads on a nissile. ac l:
establishment of th. associated deign ciiafter the specimens have been soaked at the
it
test temperature 'for 1 hour or more. It
conditions
teria are not confined t o diz.
the
leaves
'ftnr
issthe
occur
which
temperature
is well known that'the elevated
r t Maso
launching,platform. The
properties of a material are .affected by the
which iy be
obtaipirormationonmh1
tinie element as well es by .other time deimposed on the iniissile from the time of
pendent factors such as th, strain rate.
fibication to the time of launching and
Mechanical data for very short times are
establih appropriate :eoigncriterand
xelatively scarce. Hence, a ,missile designer
l
f
,ar important and not to be
flight
loads
may have no recourse but to use the data
that is available with the risk of being
overlooked. They may be critical in the
design of a structural detail or even in a
overconservative,
major :portion of the structure. Freflight
A final consideration is 'the possibiiit of
loads may ha,e to be investigated during
the occurence of creep of the structural ,maproduction, handling,.tran3portation, or while
terial. The yield ntrength criterion in this
case could be taken as the stress required to
poised in the launcher. A ,particularly intereating
is that requires
of a missile
of thininpositive
design which
walled example
a
also
is
stress
of
this
value
The
perature.
ternal pressure to stabilize the skin against
functiun of the time; the longer the time
the
required to produtce the C.27, creepset,
buckling, even while it is in production!
higher the allowable stress value (see Figure
MISCELLANEOVS ACTORS AFFECTING
4
at
requirements
shorttime
212). For the
DESIGN
STRUCTURAL
missiles,
for
ballistic
and
temperature
load
a
maior
is
not
creep
that
it
appears
however,
A number s miscellaneous considEiations
design problem. Nevertheless, there are
in one way or another the final missile
affect
may
creep
that
feel
who
designers
some
structsra: design. The considczaT.is should
become a major design problem. For examnot be assumed secondary or unimportant
ple, a missile compartment must ma.ntain
solely on the basis of space allocation in this
bealtitudes
high
at
pressure
atmospheric
discussion. Maintenance considerations, for
cauce some of the components within the
example, are not separately discussed, yet a
compartment would not function at lower
poor design would result if the designer
pressures. The skin of this compartment is
neglected t make provision for replacement,
the
during
temperature
high
a
to
heated
rep,.t, adjistment, access, etc., of various
ascending phase of the uight due to aeroLomponents of the missile. Indeed, some dedynamic heating and cools slowly while
signers can point to one or more of the items
coasting outside the denser at.nosphere. Be
70
CA
i
v
Ais
___A
'J
DESIGN CONSrOERATIONS
to b . mentioned as the consideration that
pervaded an entire design. Most of the following factors entering and influencing the
missile structural design are not peculiar to
the missile field, but rather to the general
field of design of flight structures. Thi is
While material A may be superior in a particular ap ,icatioL material B may nevertheless be called .ut fer the materfid of
constrvcten becausa of lower cost. A structural component can be made lighter and
stronger by choosingsolution Aoversolution
the principal reason that little more than a
lsting of 'the various considertbins is
B. Aoain, solution B may be incorporated
intot. design if manufacturing costs of the
othera re excessive. Numerous other examples c n be cited i,which one notesthat both
iajor and minor sarutnral designs are involved. This subject is discuss d in greater
detgilinanother volume of this series.
47.3. Weight Considerations. Up to this
point, only brief mention has been made of
the importance of weight control in the design Of a missile. The important advansinoamiil.Temprntdvarerealized, in a weightconscious
design are apparent when performance
characteristics are ntuwied. Each pound of
weight saved is reflected either in an increase
in the flight perfonrince of the mlissile or,
for the same weight, a, increase in the payload is maie possibe. The rnisiie structural
designer is always weightconscious not
w'thstanding the fact that the structural
weight may be only of the orderof 5 percent
of the missile takeoff gross weight.
presented.
th47;1. Time chedule. As a firt item,
as aposis takenstructural
sch,dule
of atime
sileeffect
factor
influencing
a risile
dtsign. Thisfactor is purposely taken atthe
autset because one normally might not conceive of the possibility of a development,
production or defivery schedule profoundly
affectingthe final product. It is nevertheless
if a strict deadline faces a missile
trueethat
htagesthat
designer, thefinal design may be quite differentfrom the designenvwaged had moretime
been available. One n'.;ht imagine that a
missile designer is intrigued by the attractive properties possessed by a particulkr
materiai and seriously contemplates its use
in the design. One or two important physical
properties of this material may be unknown
or uncertain, and the designer feels that an
investigation of these properties is warranted
before authorizing its use in the design. In
the face of a tight time schedule and the
uncertain length of time required to estblish the acceptability of the new material,
he may be forced to stand by a material with
which he has already had experience, but
which might have effectively been replaced
by the new material.
47.2. Cost Consideratirus. In .11 structural design of the pzimary or .. ondary
,tructure, the designer frequently har several
possible solutions to the problem orn hand.
The choice of the best solution involves a
careful evaluation of the advantages and disadvantages of each possible solution from
the standponit of pei formance, ease of manufact,re, weight, rutiabilit3 and many other
factors. Whe. a design is exceptionally cost
conscious, it is reflected in assigning a large
figure of merit to cost considerations in evaluating several possible structural solutions.
Weightcoisciousness is reflected in the
structural design in a number of ways. In the
selection of the material of construction, it
means that the weightstrength characteristics must be given serious consideration.
Unfortunately, no simple criterion exists for
the weightstrength characteristic of a material. If Young's Modulus is chosen a. a
strength criterion, it is noted that most cornmon metals of construction have approximately the same ratio of dens! ',y o Young's
Modulus. Molybdenum and t,i _en, however, are notable exceptions. The ratio of
density to the yield stress of a material is
generally taken as the best indication of its
weightstrength ratio. One of the principal
reason:s that B single simple weightstrength
riterion is not decisive, lies in the fact that
the true weightstrength property of a
structural element is dependent on its configuration and the type of load that it is re
71
STRUCTURES
quired to transmiit. Further complication is
plained by the relative infancy an ail
Introduced by the variation of ally weightchanging characteristic of the missile destrength criterion with temp~erature.
sign fild. In addition. ready access to weight
The effort~s of structural designers to save
data has been curtailed by security considweight also manifests itself in '1selection
erations. H~owever, a coarse weight breakA
of the mo.t efficient structural configuration
down for the Redstonp A missile, wha'ch has
consistent with the physical properties of
a range of 200250 miles, is available. This
materials that are available. The remnforced
information is presentee ir Table 41.
monocoque construction is a r.om~worthy
Depending upon the range. mission, and
example. In weig'lteonscious designs, facoflier parameters, the gross weight of mistors, of safety are kept as !ow as possible andi
siles will vary widely. The portion of the
the most powerful methods ot stressanalysis
total weight chaigeable to payload, fuel,
methods are used. Needless to say, the
engine, structure and so forth will also vary
structural designer carries his n~eightsaving
for different missiles. Hlence, no general
philosophy down to the most minute strucconclusions should be drawn based on the
tural detail,
weight data given for thle Redeto.ne.
It is not to be implied, howerer, that
missile structural designers consider weightle[
4.1
sa:ir, if paramount importance. Changes
WEIGHT VIREAKDOWN FOR
in one design feature invariably affect other
characteristics of tha missile system. Figures
REWSTONE A 44ISSILE
of merit are assigned as: inteiligeniy and
rationally as possible, and tcompropises trustAprime
be made. In rziaking dec sions. an whether
Wyeight
to change weigh$4 or fixing weight objectives,
(lbs)
~
many factor' r.xsL be considered by the
Tpni
226
designer. So~ne of these factoi s have dlready
Booster'Unit
6,757
been mentioned. such as cost, e~Teet on schedEngineo
1,50)
ule, manufacturing ease, and ffect on perPayload
6,400
formance. There are other considerations
Totaln
that can be mentioned such as questions
Empty
19,053
re6tug to the importance of the design
FVl
C2,1
f' acoon of the structural element, structural
Length
820 inches
integrity', safety and reliability, effects on
Diameter 70 inches
bandli,ig and launching equipmert. 'Conplete weight optimization, in short, is expen47.A. Reliability. The ouestici of reliability, treated in greater detail in another
sive, time consuming, and gencrally not
warranted. One may consider having reached
volume of this series, may enter into and
a point of diminishing return in trying to
affect the choice of the material of consirucsay a 90 percent weightoptimized
t~oii, structural configuration, and crossrnissfle, ip view of the excessihc cost and
seetional area. A simplified examnple can be
time requirements to achieve the additional
Laken to show the effect that relitUity cor10 peent weij,,t optimization.
siderations may have on either tne seteetion
of m'aterial or the crosssection area. Ever.
.In the aipl_416 design field, the wealth of
wheni a strL~ctura! material having wellweight dat'a accawt'&d ed over the years has
rrau&, it p *.ible to obtnin p:eI~minary strueestablished mechanical p Dperties iAs en'tural weight estimates by Aatistical ard
ployed in a design, it i.3 necessary to estabanalytical me~'iods. Tinfortuo ately, the data
lish the u.iiformity of the material pr,)perties
available and applicable to mi.'les is relaby staiA'sticni testing of various I),tch and
tively meager. This may ;.P partially exloL. Consider', on the ctber
I~
use
trefine
DESIGN CONSIDIERATIONS
of a material of inconsistent physical properties. The scatter of tests about the sample
mean or median of a strength characteristic
would dictate a lower design allowable, and
hence a larger crosssectional area for a
given applied stress. In a preliminary design
evaluation this may well lead to the structural designei" to choose a material of more
missiles than it is in the design of manned
flight vehicles. In ballistic missile design the
physical safety of flight personnel need not
be considered. Thus the structural designer
can keep safety factors at a minimum, especiaily for wellestablished design conditions
and for secondary structures, when he can
choose a structural material without the
consistent physical properties.
necessity of considering the potential loss of
life. The possibility of structural failure
near the launching platform, however, does
47.5. Safety. The philosophy of design
for safety is somewhat different in ballistic
involve physical safety of personnel, and
cannot be overlooked.
40. RFERENCES AND 21BL1OGRAPHW
November 1957.
1. Ramo, S., "ICBM: Giant Step into Space,"
Astronautics, Vol. 2, No. 1, pp. 3441,
August 1957.
2. Dukes, W. H., and Schwitt, A. (ed.),
"Structural Design for Aerodynamic
Heating." Part IDesign Information
(Confidential).
Part 11Analytical
Studies (Confirential). Part IIIStruetural Design (Sezret), Wright Air Development Center, WADC TR 55305,
October 1955.
3. Strength of Aircraft Metal Elements,
March 1955 ei., Air Materiel Command5 Builletin, Government Printing Office,
Washiagton, D. C.
4. Goldin, Robert, "Thermal Creep Design
Criteria," paper presented to Institute
of Aeronavuical Science National Summer Mee,ng, Los Angeles, California,
Reprint No. '730, July 1957.
5. Batdori, Samuel B., "Structural Problems in Hypersonic Flight," Jet Propulsion, Vy5. 27,, No. II, pp. 11571161,
6. Bonney, E. A., et al, Aerodynamics, ProIntivion, Structures, Vol. 2, Principles of
Guided Missile Design Series, D. Van
Nostrand Co., Inc., Princeton, N.J., 1956.
7. Cox, Donald D, "Selecting Structural
Materials for Supersonic Flight," Aeronautical Engineering Review, Vol. 17,
No. 1, pp. 2831, January 1958.
8. Hoff, N. J., "The Structural Effects of
Aerodynamic Heating," Proceelings of
the Third General Assembly of the Advisory Group for Aeronautical Research
and Development, NATO, London, England, p. 47, 1953.
9. Hod, N. J., "Structural Problems of
Future Aircraft," Proceedings of the
Third AngloAmerican
Aeronautical
Conference, The Royal Aeronautical
Society, Loadon, England, p. 77, 1951.
10. Hoff, N. J., "The Thermal BarrierStructures," Transactions of the Ameica? Society of Mechanical Engineers,
Vol. 77, No. 5, p. 759, July 1955.
7
ii
"
73
STRUCTUPES
INDEX
A
AbIati . 4
Aerodynamic heating, 2,41, 58
Aeroehstic effects, 67
B
Bending moments, 4
Bending rigidity, 67
Brittle fracture, 11
Buckliag, 3, 44, 46, 48, E3, 68
Cefficient of thermal expansion, 15. 17
Conduction, 13
Convection, 13
Cost, 71
creep, 3
behavior, 11
buckling, 3, 53
failure, 53
Sprimary, 12, 52
seco dary, 12,51
stress dictribution, 52
Critical load, 4
Critical time; 4
Design considerations, 57
aerodynamic heating, 58
buckling,
cost, 71 68
external configuration, 57
fatigue, 68
loads, 63
miscellaneous, 71
reliability, 72
safety, 63, 72
structural failure, 66
structural stiffness, 67
thinwalled circular cylinder structures, 63
time schedule, 71
timetemperature dependence, 69
ultimate strength, 66
weight, 71
Direct stress, 7
Discontinuity stresses, 42
Ductility, 11
Elasticity, 9
Elevated temperature effects, 3, 7, 69
Endurance limit, 11
External configuration, 57
F
Factor of safety, 63
Fatigue, 68
failure, 10
strength, 10
Flutter, 67
Fourier's equation, 14
Fouriers law, 14
Growth factor, 2
Hardnes, 11
Ht sink, 4, 15, 61
Heal transfer, 13
conduction, 14
convection, 13
radiation, 13
Hooke's uniaxial law, 41
Hoop stress, 41
Hydrostatic pressure, 45
L
Laplace equation, 15
LarsonMiller parameter, 17
Limit loads, 63
Loadcarrying capacity, 3
M
Materials, 7
creep behavior, 11
ductility, 11
elasticity, 9
hardness, 11
loadcarrying capacity, 3
mechanical behavior. 7
melting pcint, h
plasticity, 11
75
*U. S. GOVERNMENT PRINTING OFFICE
1907 0  255012 (6041A)
[I
*1.
ri
STRUCTURES
Mechanical behavior, 7
Melting point, 16
Modulus of elasticity, 17
Modulus of rigidity, 10
N
Normal strain, 8
Normal stress, 7
P
Plasticity, 11
Poisswn's equation, 15
Poisson's ratio, 10
Primary creep, 12, 52
Proportional limit, 9
Radiation, 13
Radial displacement, 42
Rerei~ry problems, 2,4
Reliability, 72
S
Safety, 72
Secant modulus, 10
Secondary crerp, 12. 51
Section modulus, 66
Shearstrain, g
Shearstress, 7
Shel analysis, 41
Specific heat, 4
.tet'anBoltzmann law, 13
Stress,?7
analysis,
direct,'l 41
disontinuity, 42
hoop, 41
normal, 7
shear, 7
tangential, 41
ultimate, 9
uniaxial, 7
yiel, 17
Siressstrain curves, 7
Structural failure, 66
Structural Oiffness, 67
T
T!
Tangent modulus, 10
Tangential stress, 41
Temperature distribution, 13
Temperature stresses, 3
Tensile strength, 11
Thermal buckling, 3
columns, 48
rectangular plates with free edges, 50
simply supported plates, 49
thin circular cylindrical shells. 50
Thermal conductivity, 4, 17
Thermal diffusivity, 4, 15
Thermal expansion, 15
Thermal
protection, 4, 58
cooled structures,
62
insulated structures, 60
unprotected structures, 58
Thermal stresses, 3, 41,43
circular cylindrical shell, 48
flat plate, 47
ring frame, 48
straight bar, 46
Thinwalled circular cylindrical structures, 63
lime shedle, 71
Timetemperature dependence, 60
Torsional rigidity, 67
Twisting rizidity, 66
0 1
U
Ultimate load, 63
Ultimate strengi, 6(6
Ultimat !ress, 9
Unixial stress, 7
W
Weight, '
Y
YWed poiYt, 9
Yield strength, 61, 59
Yield s .re,17
Younes Modulua,
12, 16, i7, 41, 71
t
GPO C65365
01
Lebih dari sekadar dokumen.
Temukan segala yang ditawarkan Scribd, termasuk buku dan buku audio dari penerbitpenerbit terkemuka.
Batalkan kapan saja.