TABLE OF CONTENTS

Chapter No.

Al The Work of the Aerospace Structures Engineer.

.. ~ ... STATICALLY DETERMINATE STRUCTURES

(Loads. Reactions, Stresses. Shears. Bending Moments, Deflections>

A2 Equilibrium of Force Systems. Truss Structures. Externally Braced Wings. Landing Gear.
A3 Properties of Sections - Centroids. Moments of Inertia, etc.
j A4 General Loads on Aircraft.
A5 Beams - Shear and Moments. Beam - Column Moments.
A6 Torsion - Stresses and Deflections.
! ;, A7 Deflections of Structures. Castigliano's Theorem. Virtual Work. Matrix Methods.

r j,

THEORY AND METHODS FOR SOLVING STATICALLY

INDETERMINATE STRUCTURES

AS Statically Indeterminate Structures. Theorem of Least Work. Virtual Work. Matrix Methods.
A9 Bending Moments in Frames and Rings by Elastic Center Method.
A1Q Column Analogy Method.
All Continuous Structures' Moment Distribution Method.
A12 Slope Deflection Method.

BEAM BENDING AND SHEAR STRESSES.

MEMBRANE STRESSES. COLUMN AND PLATE INSTABILITY.

A13 Bending Stresses.

A14 Bending Shear Stresses - Solid and Open Sections- Shear Center.
A15 Shear Flow in Closed Thin-Walled Sections.
A16 Membrane Stresses in Pressure Vessels.
A17 Bending of Plates.
A18 Theory of the Instability of Columns and Thin Sheets.

INTRODUCTION TO PRACTICAL AIRCRAFT STRESS ANAL YSIS

A19 Introduction to Wing Stress Analysis by Modified Beam Theory.

A20 Introduction to Fuselage Stress Analysis by Modified Beam Theory.
A21 Loads and Stresses on Ribs and Frames.
A22 Analysis of Special Wing Problems. Cutouts. Shear Lag. Swept Wing.
A23 Analysis by the "Method of Displacements".

THEORY OF ELASTICITY AND THERMOELASTICITY

A24 The 3-Dimensional Equations of Thermoelasticity.

A25 The 2-Dimensional Equations of Elasticity and Thermoelasticity.
A26 Selected Problems in Elasticity and Thermoelasticity.

;...·,..... l_~ .... •

.j 1-3;
 TABLE OF CONTENTS Continued

Chapter No.

FLIGHT VEHICLE MATERIALS AND THEIR PROPERTIES

81 Basic Principles and Definitions.

82 Mechanical and Physical Properties of Metallic Materials for Flight Vehicle Structures.

STRENGTH OF STRUCTURAL ELEMENTS AND COMPOSITE STRUCTURES

Cl Combined Stresses. Theory of Yield and Ultimate Failure.

C2 Strength of Columns with Stable Cross-Sections.
C3 Yield and Ultimate Strength in Bending.
C4 Strength and Design of Round. Streamline, Oval and Square Tubing in Tension, Compression. Bending,
Torsion and Combined Loadings.
CS Buckling Strength of Flat Sheet in Compression, Shear, Bending and Under Combined Stress Systems.
C6 Local Buckling Stress for Composite Shapes.
C7 Crippling Strength of Composite Shapes and Sheet-Stiffener Panels in Compression. Column Strength.
C8 Buckling Strength of Monocoque Cylinders.
C9 Buckling Strength of Curved Sheet Panels and Spherical Plates. Ultimate Strength of
Stiffened Curved Sheet Structures.
C10 Design of Metal Beams. Web Shear Resistant (Non-Buckling) Type.
Part 1. Flat Sheet Web with Vertical Stiffeners. Part 2. Other Types of Non-Buckling Webs.
ell Diagonal Semi-Tension Field Design.
Part 1. Beams with Flat Webs. Part 2. Curved Web Systems.
C12 Sandwich Construction and Design.
C13 Fatigue.

CONNECTIONS AND DESIGN DETAILS

01 Fittings and Connections. Bolted and Riveted.

02 Welded Connections.
03 Some Important Details in Structural Design.

Appendix A Elementary Arithmetical Rules of Matrices.

 INDEX
Accelerated MoUon of Buckling of Stiffened Flat Deflections by Moment Areas. A7.30
Rigid Airplane. A4. 8 Sheets under Longitudinal Deflections for Thermal
Aircraft Bolts . D1.2 Compression C6.4 Strains . A7.17
Aircraft Nuts D1. 2 Buckling under Bending Loads CS.6 Deflections by Virtual Work A7.9
Aircrait Wing Sections - Buckling under Shear Loads. CS.6 Delta Wing Example Problem. A23.2
Type, A19. I Buckling under Transverse Design for Compression C4.2
Aircraft Wing Structure - Shear C8.14 Design Conditions and Deaign
Truss Type. A2.14 Weights .. A5.12
Air Forces on Wing . A4.4 Carry Over Factor All. 4 Design Flight Requirements
Allowable Stresses (and casngnaao'e Theorem A7.5 for Airplane . A4.6
Interactions) . Cll.36 Centroids - Center of Gravity. A3.1 Design Loads .• A4.1
Analysis of Frame with Cladding Reduction Factors. C5.5 Design for Tension . . C4. 1
Pinned Suppor-ts . A9.l6 Column Analogy Method. AlO.l Differential Equation of
Angle Method C7.1 Column Curves - Non- Deflection Surface. ALB. 12
Application of Matrix Methods Dimens tonal . C2.2 Dtsconttnutttes AZD.15
to Various Structures A7.23 Column Curves - Solution C2.l3 Distribution of Loads to
Applied Load A4.1 Column End Restraint. C2.1 Sheet Panels . A2l.2
Axis of Symmetry. A9.4 Column Formulas . C4.2 Ductility. . . B1. 5
Column Strength. C7.21 Dummy Unit Loads A8.6
Beaded Webs CIO.16 Column Strength with Known Dynamic Effect of Air Forces. M.13
Beam Design - Special Cases. D3.10 End Restraining Moment C2.16
Beam Ftxed End Moments by Combined Axial and Trans- Effect of AxIal Load on
Method of Area Moments A7.32 verse Loads - General Moment Distribution. All. 22
Beam Rivet Des Ign ClO. B Action A5.21 Effective Sheet Widths C7.l0
Beam Shear and Bending COmbined Bending and Elastic Buckling Strength of
Moment AS.I Compression. C4.22 Flat Sheet in Compression. C5.1
Beams - Forces at a Section A5.7 Combined Bending and Elasttc - Inelastic Action. Bt. 5
Beams Moment Diagrams.
> AS. 6 Flexural Shear. . • C3.l0 Elastic Latera! SUpPOrt
Beams with Non-Parallel Combined Bending and Columns. C2.l7
Flanges en, 9 Tension C4.23 Elastic Stability of Column .'\17.2
Beams - Shear and Moment Combined Bending and Elastic Strain Energy . ci, 6
Diagrams AS.2 Tension or Compression of Elasticity and Thermo-
aeams Statically Deter-
> Thin Plates A18.l7 elasticity - One-Dimensional
minate & Indeterminate . A5.1 Combined Bending & Torsion. C4.23 Problems .•.. A26.l
Bending and Compression Combined Stress Equations . ci. 2 Elasticity and Thermo-
of Columns AlB. 1 Compatability Equations. A24.7 elasticity - Two-Dimensional
Bending Moments - Elastic Complex Bending' - Equations A25.1
Center Method. A9. 1 Symmetrical Section. C3.9 Electric Arc Welding . . • D2.2
Bending of Rectangular Compressive Buckling Stress End Bay Effects. Cll.23
Plates Ala. 13 for Flanged Elements . C6.1 End Moments for Continuous
Bending Strength - Basic COnical Shells - Buckling Frameworks •. All. 10
Approach. <:3.1 Strength CB.22 Equations of Static
Bending Strength - Example Constant Shear Flow Webs A14.l0 Equilibrium A2.l
Problems C3.4 Constant Shear Flow Webs ~ Equilibrium Equations A24.2
Bendtng Strength of Round Single Cell - 2 Flange Beam. A15.3
Tube' cr. 15 Constant Shear Flow Webs - Failure of Columns by
Bending Strength - Solid Single Cell ~ 3 Flange Beam. A15.5 Compression. A18.4
Round Bar. C3.1 Continuous Structures - Failure Modes in Curved
Bending Stresses A13.1 Curved Members All. 31 Honeycomb Panels. . . C12.20
Bending Stresses - Curved Continuous Structures - FaUure of Structures 81. 1
Beams A13. 15 Variable Moment of Inertia All. 15 Fatigue AnalySiS - Statistical
Bending Stresses - Elastic COre Shear C12.26 Distri.bution ... . C13.4
Range A13. 13 Correction for Cladding. C7.4 Fatigue and Fail-Safe Design . C13.8
Bending Stresses - Non- Corrugated Core Sandwich Fatigue of Materials B1. 14
hcmcgenecus Sections. A13.11 Failure Modes. e12.27 Fatig-le S-N Curves . C13.13
Bending Stresses About Cozzone Procedure C3.2 Fillers. 03.5
pr-mcipal Axes . A13.2 Creep of Ma.terials 31. 8 Fitting Design . 01. ~
Bending of Thin Plates AlS. 10 Creep Pattern . B1. 12 Fixed End Moments All. :3
Bolt Bending Strength. D1.9 Crippling Stresses Fjxed End Moments Due to
Bolt & Lug Strength AnalysiS Calculations . C7.7 Support Deflections All. 9
Methods D1.5 Critical Shear Stress Cll.16 Fixi.ty cceutcterus. C2. 1
Eolt Shear-, Tension & Crystallization Theory C13.1 Flange Design . CIO.l
Ber:ding Str-engths D1.3 Cumulative Damage Theory. C13.3 Flan~e Design Stresses . ClG.2
Boundary Conditions A24.8 Curved Beams A5. a Flange Discontinuities. ClO.7
Box Beams Analysis A22.5 Curved Sheet Panels - Flange Loads Cll. a
Brazing D2.4 Buckling Stress C9.1 Flange Strength (Crippling) . CIO.4
Buckling Coefficient CS.l Curved Web Systems Cll.29 Flat Sheet Web with Vertical
Buckling of Flat Panels wiui Cut-OUts in Webs or Skin Stiffeners CIO.l
ntsetmuer Feces C12.25 Panels . D3.7 Flexural Shear Flow
Buckling of Flat Sheets under Distribution AlS.24
Combined Loads. CS.6 Deflection Limitations in Flexural Shear Flow -
Buckli.ng or Rectangular Plate Analyses. . . A17.4 Symmetr-ical Beam Section A14.5
Plates A18.20 Deflections by Elaanc Weights .\7.27 Flexural Shear Stress. A14.1

Static Tension Stress-
Strain Diagram . . . Bl.2
Statically Determinate
Coplanar Structures and
Loadings . A2.7
Statically Determinate and
Indeterminate Structures A2.4
Statically Indeterminate
Frames - Jomt Rotation Al2.7
Statically Indeterminate
Problem . . • • . • . . . AB.l
Stepped Column - Strength C2.14
Stiliened Cylindrical
Structures - illUmate
3trength . • . • . • . C9.S
Stiffness &I Carry-over
Factors lor CUrved Members All. 30
Stiffness Factor. • . . All. 4
Strain - Displacement
Relations • . . . . . A24.5
Strain Energy . . . . . A7.l
Strain Energy of Plates Due
to Edge Compression and
Bending A18.19
Strain Energy In Pure Bending
INDEX· Continued
ToUlgent Modulus Bl.5 Trusses With Multiple
Tangent-Modulus Theory A1B.a Redundancy . . . • A8.11
Taxi Loads . . • . • • • C13.33 Trusses With Single
Tension CHps . • . . . . D3.2 Redundancy . . . . AS. '7
Tension-Field Beam Action. en. 1 Tubing Design Facts C4.5
'renetcn- Field Beam Formulas C11. 2 Two-Dimensional Problems. A.26. 5
Theorem of Castiglta.no . . • A7.5 Two-Cell Multiple Flange
Theorem of Complementary Beam. • One Axis ot
Energy. . • • • • . . . • • . • A7.5 Symmetry • . • . . • . • • · Al5.11
Theorem of Least Work •.• A8.2 Type of Wing Ribs. • • . • • A2l.t
Theorems of Virtual Work and
Minimum Potential Energy A7.5 tntimate Strength in Combined
Thermal Deflections by Bending &: F1exural Shear • • C4.25
MatriX Methods • . • • . AB.39 tnUmate Strength in Combined
Thermal Stresses . • . • • A8.14 Compression, Bending,
Thermal Stresses . . . • • AB.33 Flexural Shear &: Torsion. C4.26
Thermoelasttcity - Three- Ultimate Strength in Combined
Dimensional Equations. • A24.1 Compression, Bending &I
Thin Walled Shells A16.5 Torsion • . . • • • • • • . • Ct. 24
Three Cell - Multiple Flange illtimate Strength in Combined
Beam. - Symmetrical about Tension. Torsion and
One Axis . . • • • • . . • • AIS. 15 Internal Pressure p in psi. C4.26
Three Flange - Single Cell Uniform Stress Condition. • Cl.I
Wing • • • • • • • • • • • • Al9.5 Unit Analysis lor Fuselage
Torsion - Circular Sections. A6.l Shears and Moments. . • AS. IS"

R_ .
of Plates. • • • . . . . . . . AlB. 12 Torsion - Effect of End UnsymmetriCal Frame • . . A9.2
Streamline Tubing - Strength. C4. 12 Restraint . • • . . . . • A8.16 Unsymmetrical Frames or
Strength Checking and Torsion - Non-circular AIO.4
Design - Problems • . . . C4.5 Sections • . . • . • • • A6.3 Unsymmetrical Frames using
Stren~··_ ".: Round Tubes Torsion Open sections AS. 4 PrinCi~ Axes. • • . •. M.tS
_ ..nder Combined Loadings . C4.22 Torsion of Thin-Wailed .'Tnsymmeirical Structures A9.13
sn-ess Analysis Formulas Cll. 15 Cylinder having Closed Type .' :''"it
" , ':-·"'·A
Stress Analysis of Thin Skin - StiHeners . . . . . . . . . • AS. IS \t, . 'Jy - Load Factor
Multiple stringer Cantilever Torsion Thin Walled Sections. A6.S .....
" • • • .("r- A4.7 •
Wing • . . . . . . . . . . • • A19.10 Torsional Moments - Beams • A5.9
Stress Concentration Factors. C13.10 Torsional Modulus of Rupture. C4.1'7 Wagner Equations .. Cl1.4
Stress Distribution & Angle Torsional Shear Flow in Web Bending &: Shear Stresses CtO.5
of Twist for 2-Cell Thin 4
Multiple Cell Beams by Web Design • . • . • . • • • Cll.18
Wall Closed Section . . A.6.7 Method of Successive Web Splices • • . • . . • • . CIO. 10
Stress-Strain Curve • . • . • B1.7 Corrections . • • • . . A6.10 Web Strength. Stable Webs. CtO.5
Stress-Strain Relations .•• A24.6 Torsional Shear Stresses in Webs with Round Lightening.
Stresses around Panel Cutout. A22.1 Multtple 4Cell
Thin-Wall Holes • • • • • . . • • ClO.17
Stresses in Uprights Cll.17 Closed Section - Distribution AS.7 Wing Analysis Problems A19.2
Stringer Systems in Diagonal Torsional Strength of Round Wtng Arrangements. • • Al9.1
Tension . . . • . • . • • • . Cll.32 Tubes •.•.•..•.•• C4.17 Wing Effective Sectton A19.12
Structural Design Philosophy. ci, 6 Torsional Stresses 1n Wing Internal Stresses • A23.14
Structural Fittings • • • • . . A2.2 Muitiple Cell Thin-Walled
4
Wing Shear and Bending
Structural Skin Panel Details. D3.12 Tubes ••....•... A6.S AnaLysis • • • . . . • • • A19.14
Structures with Curved Transmission of Power by Wing Shear and Bending
Members • . • . • . . . All. 29 Cylindrical Shaft. . . . • A6.2 Moments • . • • . • . . A5.9
Successive Approximation TriaXial Stresses . • • • . Cl.5 Wing - Sbear Lag . • • • A19.25
Method for Multiple Cell Truss Deflection by Method Wing Shears and Moments AS. 10
Beams . . . . . • . . AlS.24 of Elastic Weights • A7.33 Wing Stiffness Matrix.. • • A23.11
Symbols for Reacting Truss Structures A2.9 Wing Strength ReqUirements A19.5
Fitting Units . . . . • A2.3 Trusses with Double Wing Stress Analyl'lis Methods .'1.19.5
Symmetrica sections - Redundancy. • . . AS. 10 Wing - Ultimate Strength . A19.11
External Shear Loads A14.2 Work of Structures Group. Al.2

Y Stiffened Sheet Panels C7.2C

 CHAPTER Al
THE WORK OF THE
AEROSPACE STRUCTURES ENGINEER
AI. 1 Introduction.
The first controllable human flight in a
heavier than air machine was made by Orville
Wright on December 17, 1903, at Kitty Hawk,
North Carolina. It covered a distance ot 120
feet and the duration of flight was twenty
seconds. Today, this initial flight appears
very unimpressive, but it comes into its true
perspective of Lmportance when we realize that
mankind for centuries has dreamed about dOing
or tried to do what the ~rlght Brothers
a:campllshed in 1903.
The tremendous progress accompliShed in the
first 50 years of aviation history, with most
of i t occurring in the last 25 years, 1s almost
unbelievable, but without doubt, the progress
in the second 50 year periOd will still be more
~~believable and fantastic. As this is written
in 1964, jet airline transportation at 600 MPH
is well established and several types of
military aircraft have speeds in the 1200 to
2000 ~ range. ?reliminarJ designs of a
supersonic airliner with Mach 3 speed have been
ccmpleted ~~d the government is on the verge of
sponsoring the development of such a flight
vehicle, thus supersonic air transportation
should become co~on in the early 1970's. The
rapid progress i~ ~!ssile design has ushered
in the Space Age. Already many space vehicles
have been flown in search of new knowledge
which is needed before successful exploration
of space such as landings on several planets
can take place. Unfortunately. the rapid
development of the missile and rocket power
has given mankind a flight vehicle when combined
with the nuclear bomb, the awesome potential to
quickly destroy vast regions of the earth.
TMhile no person at ~resent ~~ows where or what
space exploration will lead to, relative to
benefits to ~nkind, we do know that the next
great aviation expanSion besides supersonic
airline transportation will be the full develop-
~ent and use of vertical take-off and landing
aircraft. Thus persons who will be living
through the second half century of aviation
progress will no doubt witness even more
fantastic progress than oceurred in the first
50 years of aviation history.
A!. 2 General Organization of an Aircraft Company
Engineering Dfvtetcn,
The ~odern commercial airliner, militarJ
airplane, missile and space vehicle is a highly
Al.I
scientific machine and the combined knowledge
and experience of hundreds of engineers and
scientists working in close cooperation is
necessary to insure a successrul product. Thus
the engineering division of an aerospace company
consists of many groups of specialists whose
specialized training covers all fields of
engineering education such as PhySics, Chemical
and Metallurgical, MeChanical, Electrical and,
of course, Aeronautical ~~lneering.
It so happens that practically all the
aerospace companies publiSh extensive pamphlets
or brochures explaining the organization of the
engineering division and the duties and
responSibilities of the many sections and groups
and illustrating the tremendous laboratory and
test facilities which the aerospace industry
possesses. It is highly recommended that the
student read ~~d study these tree publications
in order to obtain an early general under-
standing on how the ~odern flIght vehicle is
conceived, deSigned and then prOduced.
In general, the engineering department of
an aerospace company can be broken down into sax
large rather distinct sections, which in turn
are further divided into specialized groups,
which in turn are further divided into smaller
working groups of engineers. To illustrate, the
six sections will be listed together with some
at the various groups. ThiS is not a complete
list. but it should give an idea or the broad
engineering set-up that is necessarJ.
I. Preliminary Design Section.
II. TecrJlical _~alysis Section.
(1) Aerodynamics Group
(2) Structures Group
(3) ~eight and 3alance Control Group
(4) Power Plant Analysis Group
(5) Materials and Processes Group
(5; Centrols AnalYSiS Group
III. Component DeSign Section.
(1) Structural DeSign Group
(~lng. Body and Control Surfaces)
(2) Systems Design Group
(All mechanical, hydraulic, electrical
and ther.nal installations)
IV. Laborato~J Tests Section.
Al.2 THE WORK OF THE AEROSPACE STRUCTURES
(1) Wind Tunnel and Fluid Mechanic5 ~est
Labs.
(2) Structural Test Labs.
(3) Propulsi~n Test Labs.
(4) Electronics ~est Labs.
(5) Electro-Mechanical Test Labs.
(6) Weapons and Controls Test Labs.
(7) Ar~log and Digital Computer Labs.
v. Flight Test Section.
VI. Engineering Field Service Section.
SinCe this textbook deals with the subject
of structures, it seems appropriate to discuss
in some detail the work of the Structures Group.
For the detailed discussion of the other grou~s,
the student should refer to the various air-
craft company publications.
At. 3 The Work of the Structures Group
The structures group, relative to number of
engineers, is one of the largest of the ~any
groups ot engineers trat make up Section II,
the technical analySis section. The structures
group is primarily responsible for the
structural integrity (safety) ot the airplane.
safety may depend on sufficient strength or
sufficient rigidity. This structural integrity
must be accompanied with lightest pOSSible
weight, because any excess weight has detri-
mental etfect upon the perfo~ce of aircraft.
For example, in a large, long range missile,
one pound of '~ecessary structural weight may
add mora than 200 Ibs. to the overall weight of
the missile.
The structures group is usually divided
'into sUb-groups as tollows:-
(1) Applied Loads Calculation Group
(2) Stress AnalySiS an~ Strength Group
(3) Dynamics AnalYSiS Group
(4) Special Projects and Research Group
THE '"ORK OF THE APPLIED LOADS GROUP
Before any part ot the structure can be
finally proportioned relative to strength or
rigidity, the true external loads on the air-
craft must be determined. Since critical loads
came tTom many sources, the Loads Group must
analyze loads fram aerOdynamiC forces, as well
as those forces from power plants, aircraft
inertia; control system actuators; launching,
landing and recovery gear; a~ent, etc. The
etrects of the aerOdynamic forces are initially
calcUlated on the assumption that the airplane
structure 1s a rigid bOdy. Atts: the aircraft
structure is Obtained, its true rigidity can
be used to obtain dynamic effects. Results of
wind t~~el model tests are usually necessary
in the application of aerodynamic principles to
load and pressure analYSiS. • or danger of failure and therefore the dynamic
ENGINEER
The final results of t.he work of this
group are formal reports glv~n~ complete a~plied
load design criteria, with ~ny graphs ~nd swu-
mary tables. The final results ~y 61v8 com-
plete shear, moment and no~l forc~s =e~er=~d
to a convenient set of :CY2 axes for major air-
c ra.r t units such as the Wing, rus eIage , e t c .
THE WORK OF STRESS ANALYSIS ~\m S~R~~GTH GROUP
Essentially the primary job of :he stress
group is to help specify or deter.nine the kind
of material to use and the :h:c~~ess, size and
cross-sectional shape Jt every struct~l ~eQ­
ber or unit on the airplane or ~issile, and
also to assist in the deSign of all jOints and
connections for such ~embers. safety with ~ight
weight are the paramount str~ctural jesl~ re-
quirements. ~he stress group ~ust consta~tly
work closely with the Structural DeSign Sect:Gn
in order to evolve the best structural over-all
arrangement. Such factors as ~ower ~lants,
bUilt in fuel tanks, landing gear retracting
wells, and other large cut-outs can d~ctate the
type of wing structure, as for example, a two
spar single cell wing, or a multiple spar
~ultiple cell wing.
To expedite the initial struct~r~l ~esign
studies, the stress group ~ust s~~ply initial
structural sizes based on approximate loads.
The fi~l results of the work by the stress
group are recorded in elaborate reports which
show how the stresses were calculated and hew
the reqUired member sizes were obtained to carry
these stresses efficiently. The r:nal size of
a member may be dictated by one or more rae tors
such as elastic action, tne Ias t t c action, ele-
vated temperatures, fatigue, etc. To insure
the accuracy of theoretical calculations, the
stresS group must have the assistance of the
structures test laboratory in order to obtain
information on which to base allowable design
stresses.
THE WORK OF THE DYNAMICS A~LYSIS GROUP
The Dyna~ics AnalysiS Grou) has rapidly
expanded in recent years ~elative to number of
engineers required because supersonic airplanes,
missiles and vertical riSing ai~craft have pre-
sented many new and complex problems in the
general field of dynamics. In some airc~tt
companies the dynamiCS group 1s set up as a
separate group outside the Structures Group.
7he engineers in the dynamiCS group are
responSible for the investigation ot Vibration
and shOCk, aircraft flutter and the establish-
~ent of desig~ requirements or c~2nges for its
control or correction. Aircraft contain dozens
of mechanical installations. Vicration of ~~y
part of these installations or systems ~y be
of such character as to cause faulty operation
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A1.3

characteristics must be changed or modified in
order to insure reliable and safe operation.
The major structural units of aircraft such gresses. For example, in the r.t r-uc tur-ea Group,
as the wing and fuselage are not rigid bodies.
7hus when a Sharp air gust strikes a fleXible
wing in high speed flight, we have a dynamic
load situation and the wing Nill vibrate. The
dynamicist must determine whether this vibration structure; (3) what type of body str~cture is
1s serious relative to induced stresses on the
wir~ structure. The dynamics group is also
responsible for the determination of the
stability and performance of miSSile and flight
vehicle guidar.ce and control systems. The
dynamics group must work constantly with the
various test laboratories in order to obtain
reliable values of certain factors that are
necessary in many theoretical calculations.
THE ',jaRK OF THE SPEC IAL PROJECTS GROUP
In general, all the various technical
groups have a speCial SUb-group which are work-
ing on deSign problems that Nill be encountered
in the near 1r distant future as aviation pro-
this sub-group might be studying such problems
as: (1) how to calculate the thermal stresses
in the wing structure at super-sonic speedS;
(2) how to stress analyze a new type of wing
best for future space travel and what kind of
materials will be needed, etc.
Chart 1 illustrates in general a typical
~ke-up of the Structures Section of a large
aerospace company. Chart 2 lists the many
items which the structures engineer must be
concerned with in insuring the structural
integrity of the flight vehicle. Both Charts
1 and 2 are from Chance-Vought Structures
DeSign Manual and are reproduced with their
permi saton.

~EIlOE'-'\STlc:'···4
U..f ,.

~'IlOE'-'\STIC ,..- --l

i I STWCTVI~S
cU"",
, I

11:'
' ""
lAKlIlATo'" I
~~u,

I. "'",ru,,,
'liSI U,,"T
1

J i
, -.c...."I.IC
'NO "Owu
' ..... 1 ""ST
: UN"

, .....c:.. 'H~
COMPUIATICH
~'''G'' GIlOU'
~~,

Chart 1. Structures Section Organization

Chance- Vought Corp.

-:3
e.13W m
= r ETT? 5ZWF

AI. 4 THE WORK OF THE AEROSPACE STRUCTURES ENGINEER

THE LINKS TO STRUCTURAL INTEGRITY

••••• ARE NO BETTER THAN THE WEAKEST LINK

MATERIALS OF
CONSTRUCTION
FA.STENERS
WELDING
BONDING
?LATE~ND BAR
FORGINGS
STIFFNESS CASTINGS
CRITERIA €XTRUSIONS STRESS
SHEET METAL
FLUTTER SANDWICH
ANALYSIS
CONTROL SYSTEM STABILITY ?tASTIC l,OMINAIT SKIN PANElS
BEARINGS BEAM ANALYSIS
PANEL fLIJTTER-SKIN CONTOURS STRAIN COMPATIBILITY
CONTROL SYSTEM DEfLECTIONS STRAIN CONCENTRATION
JOINT ANALYSIS
THERMAL EFFECTS BEARI~G ANALYSIS
MtCHANICAL VIBRATIONS BULKHEAD ANALYSI S
fITTING ANALYSIS
ROLL POW£R-O IVERGENCE i1'ERMAL STRESS
AEROtWtAMIC CENTER SHIFT ,'oIECHAN1CAL COMPONENTS
D:PERIMENTAL STRESS ANALYSIS
DYNAMIC RESPONSE

LOADS AND
ENVIROMENT MATERIALS AND
FLIGHT LOAD CRITERIA
GROUND LOAD eRITER IA QUALITY CONTROL
FLIGHT LOAD DY~M'CS DUCTILITY
LAUNCHING DYNAMICS
LANDING DYNAMICS STRESS-STRAIN
DYNAMIC RESPONSE HOMOGENEOUS MATERIAL
aECOVERY DYNAMICS
,UGHT LOAD DISTRIBUTIONS RESIDUAL STRESS
INERTIAL LOAD DISTRIBUTIONS COMPONENT ALLOWABLES
HEAT TREAT CONTROL
Fl..D:IBILITY EFFECTS YIHDING
GROUND LOAD DISTRIBUTIONS
ANALYSIS STRESS CORROSION
FRACTURE
REPEATEO LOAD SPECTRUMS UNIT SOLUTIONS STABILITY AT TE~P£RATURE
TEMPERATURE DISTRIBUTIONS °ATlGUE
INOffiRMINATE STRUCTURES SPECIFICATION CONFORMANCE
LOADS FROM TllERMAL WEAR, BRINELLING
DEFORMATIONS WING ANALYSIS BLUE PRINT CONFORMANCE
PRESSURES-IMPACT CREEP
TAIL ,l,NAlYS1S
DEFliCTIONS
FUSElJ.GE SHElL A""lYSIS
Tl'ERMAl O"FECTS
THERMAL ANALYSIS
STIFFNESS
DfFlICTION ANALYSIS
COMBINED LOADINGS
STIFFNESS
3UCKlING

Chart 2
From Chance- Vought Structures Destgn Manual

EQUILIBRIUM OF FORCE SYSTEMS. TRUSS STRUCTURES
A2.1 Introduction. The equations of static
equilibrium must constantly be used by the
stress analyst and structural designer 1~ ob-
taining unknown forces and reactions or unkno~n
internal stresses. They are necessary whether
the structure ..or machine be s tmp.Le or complex.
The ability to apply these equations 1s no
doubt best developed by solving ~ny problems.
This Chapter l~ltlates the application of these
important phySical laws to tt.e force and stress
ar~lysls of structures. It 1s assumed that a
student has completed the usual college course
in engineering mecrznlcs called statics.
•
A2.2 Equations of Static Equilibrium.
To completely Cefl~e a force, we must know
its ~agnltude, direction and ~olnt of ann1ica-
tion. These facts regaTding the ~arce are
generally refer~ed to as the characteristics of
the ~orce. So~etimes the more ~eneral te~ of
line Of act~on or lecation is used as a force
cr~racteristic in Place of paint of application
designation.
A force acting in space is completely
defir-ed :: we %now its components in three
directions and its ~oments about 3 axes, as for
example FX J F~, Fz ~nd ~x, tly and Xz • ~or
equilibrium o~ a force system there can be no
resultant force and thus the equations of
equllibri'4n are obtained ~y equating the force
and moment cCill~onents to zero. The equations
~f static equilibrium for the various types of
force systems 'Nill now be suear-tzeo .
EQ.UILI3RIUi1 SO.UATICNS FOR GSNERAL
SPACE: urCN-COPLANAR) ?ORCE SYSTE?:
l:F x= 0 ~ ... = 0
ZF y = 0 m, = a } - - - - -(2.1)
ZF z = 0 m, = 0
rhus :or a general space ~orce system,
there are 6 equations of static eqUilibrium
available. T~ree of these and ~o ~or8 can be
force equations. It is or t en acr-e convenient
to ~ake the moment axes, 1, 2 ~nd 3. as any set
ot X, y and z axes. All 6 equat10ns could be
no~ent equaticns about 6 ciffere~t axes. ~~e
force equations are written for 3 ~utually
~er?endiC~lar ~es ~nd need not be t~e x, y
and z axes.
sq,UILI3RI'L11 OF' SPACE ::CNCJEP,S::T ropes
Co~curre~~ 7;e~ns that all forces of the
CHAPTER A2
force system pass through a cammon point. The
reSUltant, if any, must therefore be a force
and not a moment and thus only 3 equations are
necessary to completely define the condition
that the resultant must be zero. The equat10ns
of equilibrium available are therefore:-
l:F x =a m, = o
ZFy = a or m, = o } - - - - -(2.2)
l:Fz =0 m, = o
A combir~tion of force and moment equations
to make a total of not more than 3 can be used.
For the moment equations, axes through the point
of concurrency cannot be used since all forces
of ~he system pass through this point. The
moment axes need not be the same direction as
the directirns used in the force equations but
of course. they could be.
EQUILIBRIUM OF S?AC::; ?J..RALLEL FORCE SYSTlli
In a parallel force system the direction of
all forces is known, but the magnitUde and
location of each is unknown. Thus to determine
~gnitude, one equation Is required and for
location twa equations are necessary since the
force is not confined to one plane. In general
the 3 equations commonly used to make the re-
sultant zero for this type of ~orce system are
one force equation and two moment equatiOns.
For example, for a space parallel force system
acting in the y direction, the equations of
equilibrium would be:
ZFy = 0, ZI1x ::: 0, n1z:= 0 - -(2.3)
SG.UILIBRIUM OF GE:N""'..2.AL CO-PUu'JAR ~ORCE SYSTEM
In this type ot farce system all forces lie
in one plane ~d it ta~es only 3 equations to
deter.Aine the magnitUde, direction and location
or the resultant of such a force system. Either
torce or moment equations C3n be ~sed. except
that a ~aximum of 2 :orce equa~lons can be used.
For example, for a force system acting in the
xy plane, the follcwiig co~blnaticn of equili-
bri~ equations could be used.
ZFx =0 ZFx =0 ZF y • a zr1z .. = a
ZF y =0 or L:!":z..= 0 or ZMz ... :: 0 or 1:1z II = 0 2.4
z:1z = 0 Z!"!Zil= 0 z:M: Z ll = 0 mz" = a
SYS':'~
(':'he subsc=-lpts 1, 2 and 3 refer to different
locations for z axes or moment center-s . )
A2.1
.AC
A2.2 EQUILIBRIUM OF FORCE SYSTEMS.
Since all forces lie in the ss~e plane and
also pass thr8ugh a c~~,on ~oint, the ~a~~itude
and direction of the r~s~ltant of this type of
force system is unknown ~ut the location ~s
£~own sin~e the ;oint of conc'~iency is on the
lin8 of action cf the resultant. ~hus only two
equations 0: equilibri~ are necessary to define
the resultant and ~ke it zero. T~e combin-
ations available are,
: 0 or ZFx = a or ZFy = a or ZMz~
l:F x
ZFy : a =0 ZI'1z = O ZI'1z a
(The z axis or moment center locations must be
other than through the paint of concurrency)
EQUILIBRIUM: OF CO-PLANAR P.4RALLt:L FCRCE
Since the direction of all forces in this
type or rorcs system 1s known and since the
forces all lie 1n the same plane, it only takes
2 equations to define the magnitude and location
of the resultant of such a force system. Hence,
there are only 2 equations of equil1bri~ avail-
able for this type of force system, namely, a
force and :noment equation or two moment
equations. For example, for forces parallel to
y axis and located in the xy plane the equili-
brium equations available would be: -
l:l1z. : a
l:F y : 0
n1 z =0
or
4I'1z Ii =a } -----2.6
(The moment centers 1 and 2 cannot be on the
same y axis)
EQUILIBRITJr1 OF COLINEAR FORCE SYSTEM
A collnear torce system 1s one where all
forces act along the same line or in other
words, the direction and location of the forces
is known but their nagnt tudes are unknown, thus
only magnitude needs to be found to define the
resultant of a collnear force system. Thus
only one equation ot equilibrium is a'ffiilable,
namely
ZF =a or ZI1 1 = O
where moment center 1 is not on the line at
action of the force system
A2.3 Structural Fitting Units for Establishing the Force
Characteristics of Direction and Point of Application.
To completely define a force in space re-
qUires 5 equations and 3 equations if t~e force
is limited to one ?lane. In ~eneral a structure
is loaded by ~own forces ar.d these !orces are
transferred tPIough the struct~re in some
~nner of internal stress distribution and then
TRUSS STRUCTURES.
reacted 'Jy other ext er-ne ; -or-ces , ccmnorLy
referred to as reactions which hold the j:,cwn
forces on ~he str~cture in equiliJri~ll. Sl~ce
the static equa~ions of equil:bri:2n ~vai:~~:e
for the various t~?es ot force syste~ are
linited, the str~c:ural engineer resorts tc the
use of fitting units whic~ establ:sh ths
direction 8f an ~n}~i8~TI fc~ce C~ ~~s ~o:nt Jf
ap~11c8:~on or both, t~~s decreasins th3 ~~ber
of UIL~owns t8 be determi~ed. ~~e ~l~ures
which follow illustrate t~e ty;e of ~itt1~g
units employed or o:her gene~~l Ieth~ds ~Jr
=0 } 2.5 establiShing the ~orcs c~aracterist1cs of
dirsction and pOint of application.
= O
Ball and Socket Fitting
SYS'I'~
For any space or coplanar f8rce suc~ as ?
and Q acting on the bar, the line o~ action of
such forces must act t..u-cucn the cent.cr- 0': the
ball ~! rotation of the bar is prevented. ~hus
a ball and socket joint can be used to establ:sh
or control the di~ection and line action of a
force applied to a struct~re through :~i~ :j~e
of fitting. Since the ~oint has ~o rctaticTzl
reSistance, no couples in any plane can be
applied to it.
~'l
-
II p
For any force such as P and Q acting i~ the
xy plane, the line of action of such a ~~rce
~ust pass through the ,in center since t~e
fitting unit cannot resist a ~cme~t about a z
axlo through the pin center. Therefore, fJr
forces acting in the xy plane, th~ direct:cn
~~d line of action are established by th8 pin
joint as illustrated in the f:gure. Sir.ce a
Single pin fitting can resist :noments atou: axes
----2.7 perpendicular to the ~in axis, the cirectlon and
line of action of out of ,lane :orces is ~tere­
fore not established by s tng l e pfn ~it-:lng umt.s .
A ,B
-<@o========:j):Jl_
If a bar AB has si~Gle pin f~t~ings at
eac~ end, then any :orce P lying in the xy
plane and anplied to end B ~ust have a direction
and line of action co tnc tc t ng ....ith a line jo tn-.
i~g the pin centers at end ~it~1ngs A ar.d 3,
since the :lttings cannot resist a ~oment about
the 3 axis.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
Double Pin - Universal Joint Fittings
z Q
L. § /-/11
Fh" liillirn
y Cables - Tie Rods
L B
Since single pin fitting units can resist
applied moments about axes normal to the pin
axiS, a double_pin joint as illustrated above
is otten used. T~is fitting unit cannot resist
~oments about y or z axes and thus applied
forces such as P and Q ~ust have a line of
action and direction such as to pass through
the center of the fitting unit as illustrated
in the figure. The fitting unit can,. however,
resist a moment about the x axis or in other
woras, a universal type of fitting unit can
resist a torsional ~oment.
Rollers
-~~-
t t
In order to pe~it structures to move at
support paints, a fitting unit involving the
idea of rollers is often used. For example,
the truss in the figure above is supported by
a pin :'itting at (A) which is rur-tner- attached
to a fitting portion that prevents any hori-
zontal movement of truss at end (A), however,
the other end (8) is supported by a nest of
rollers which provide no hor1zontal resistance
to a horizontal movement of the truss at end (8).
The rollers fix the direction of the reaction
at (B) as ,erpend~cular to the roller bed.
Since t~e fitti~g uni~ is Jol~ed to the truss
jOint by a pin, the ,oint of application of the
reaction is also known. hence only one :orce
characteristic, na~ely magnitude, 1s unknown
for a roller-pin type of fitting. ?or the
fitting unit at (A), ,oint of application of the
reaction to the truss 1s kncvrn because of the
pin, but di~ectlon and magnitUde are unknown.
Lubricated Slot or Double Roller
Unit.
",4",
Lubricated Slot ! ~
b'RI=Jn-'
Double RoUer
A2.3
Another general fitting type that is used
to establish the direction of a force or reaction
is illustrated in the figure at the bottom of the
P
first column. Any reacting force at joint (A)
~ust be horizontal since the support at (A) is
so designed to provide no vertical resistance.
~P
Since a cable or tie rod has negligible
bending resistance, the reaction at joint B on
the crane structure' tram the cable must be
colinear with the cable axiS, hence the cable
establishes the force characteristics of direc-
tion and point of application of the reaction
on the truss at point B.
A2.4 Symbols for Reacting Fitting Units as Used in
Problem Solution.
In solVing a str~cture for reactions,
member stresses, etc., one must know what foroe
characteristics are unknown and it is cammon
practice to use simple symbols to indicate. what
fitting support or attachment units are to be
used or are assumed to be used in the tinal
design. The following sketch symbols are com-
~onlY used for coplanar tares systems.
Pin
ioi
A small circle at the end of a member or on
a triangle represents a single pin connection
and fixes the point of application of forces
acting between this unit and a connecting member
or structure.
b..........Kni.fe Edge
Rx
b t
tRy
The above graphical symbols represent a
reaction in which translation of the attach-
ment point (b) is prevented but rotation of the
Ty~e of Fitting attached structure about (b) can take place.
Thus the reaction is lxnknown in direction and
magnitude but the point of applIcation is known,
~
namely through point (b). Instead of using
I direction as an unknown, it is more convenient
',~
to replace the resultant reaction by two com-
ponents at right angles to each other as indi-
cated in the sketches.
/1.
~,
A2.4 EQUILIBRIUM OF FORCE SYSTEMS.
l~eEdge ~ can be found by the equations o~ static eqUili-
!
Rollers FA Roller.
I I
The above fitting units using rollers fix
the direction of the reaction as normal to the
roller bed since the fitting unit cannot resist
a horizontal force through point (b). Hence
the direction and point of application of t~e
reaction are established and only magnitude is
unknown.
fixed
(bJM.,
rrrrrt '" R X - I I I \ \ \ \
/ Illy
The graphical symbol above is used to
represent a rigid support which Is attached
rigidly to a connecting structure. The re-
action is campletely unknown since all 3 force
characteristics are unknown, namely, magnt 'tude ,
direction and point of application. It Is con-
venient to replace the reaction R by two force
components referred to some paint (b) plus the
unknown moment M which the resultant reaction R
caused about point (b) as indicated in the
above sketch. This discussion applies to a
coplanar structure with all forces in the same
plane. For a space structure the reaction
would have 3 further unsnowns , name Ly, Rz, I1x
and lly.
A2. 5 Statically DetermLnate and Statically Indeterminate
Structures.
A statically determinate structure is one
in which all external reactions and internal
stresses tor a given load system can be found
by use at the equations of static eqUilibrium
and a statically indeter.nlnate structure is
one in which all reactions and internal stresses
cannot be round by using only the equations of
equilibrium..
A statically dete~inate structure is one
that has just enough external reactions, or
just enough internal members to make the
structure stable under a load system and if one
reaction or member is removed, the structure 1s
reduced to a linkage or a mechanism and is
therefore not further capable or reSisting the
load system. If the st~ucture has more ex-
ternal reactions or internal members than is
necessary for stability ot the structure under
a given load system it is statically ir.deter-
TRUSS STRUCTURES.
minate, and the degree of redundancy dapends on
the number of ~~nowns beyond that ~umber which
brium. A structure can be statically indeter-
minate with respect to exter~al reac~ions alone
or to In:ernal stresses alene or to 60th.
The additional eq~ations :hat are needed
to solve a statically indeternlnate structure
are obtained by conSidering the distortion of
the structure. This means that the size of all
members, the ~~terial ~rom which members are
~de must be known since dist~rtions ~ust be
calculated. In a statically cete~lnate
structure this information on sizes and ~terial
1s not required but only the configuration of
the structure as a whole. Thus deSign analYSiS
for statically deter,ninate structure is stra~ght
forward whereas a general t~ial and error pro-
cedure is required for design analYSiS of
statically indete~inate structures.
A2. 6 Examples of Statically Determinate and Statically
Indeterminate Structures.
The first step in analyzl~g a stTuc~ure 1s
to determine whether the str~cture as presented
is statically det.ermtnat e . It so, the reactions
and internal stresses can Qe found without know-
ing sizes of members or kind of mater1al. If
not statically determinate, the elastic :heory
~ust be applied to obtain additional equa~ion5.
The elastic theory is treated in considerable
detall in Chapters A7 to Al2 inclusive.
To help the student become f~illar with
the problem of determining whether a structure
is statically determinate, several ex~ple
problems will be presented.
~ample Problem 1.
w '" 10 lb. lin.
t
Fig. A2.1
In the st~~cture shown in Fig. 2.1, the
~~Qwn forces or loads are the distri8uted loads
of 10 lb. per inch on member ABD. The reactions
at points A and C are unknown. The reaction at
C has only one ~~own Characteristic, namely,
magnitude because the point ·of application of He
is tnrough the cin center a: C and the directicn
of Rc must be parallel to line SB because there
is a pi~ at the other end B of ~ember CE. At
point A the reaction 1s unknown in 11rection
and ~gnitude but the point of application ~ust
be through the pin center at A. Thus there are
2 unknowns at A and one ~~o~vn at C or a total
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
of 3. with 3 equations of equilibrium avail-
able for a co?lanar force system the structure
1s statically determinate. Instead of uSing an
angle as an '~nown at A to find the direction
of the reaction, it is usually more convenient
to re~lace the reaction by components at right
angles to each other as HA and VA in the figure
and thus the 3 unknowns for the structure are 3
magnf tudes ,
E:x:ample Problem 2.
P available and thus the first conclusion might
-Pin
Fig. A.2.3 with the 3 equations ot equilibrium. Thus
Fig. A2. 2
~lg. 2.2 shows a structural frame carrying
a known load system P. Due to the pins at
redction points A and B the paint or application l~OwnS listed above. The structure is there-
is known tor each reaction, however, the magni-
tude and direction of each is unknown making a
total of 4 unknowns With only 3 equations of
equilibrium available for a coplanar force
system. At first we might conclude that the
structure is statically indeterminate but we
must realize this structure has an internal pin
at C which means the bending moment at this
point 1s zero since the pin has no resistance
to -ceat ton. If the entire structure is in
equilibrium, then each part must likewise be
in equilibrium and we can cut out any portion
as a f~ee bOdy and apply the equilibrium
equations. Fig. 2.3 shows a free bOdY of the
frame to left of pin at C. Taking moments
about C and equating to zero gives us a fourth
equation to use in deter.nining the 4 unknowns,
HA, VA' VB and Hg. The moment equation about C whether the structure is statically deter.nir~te.
does ~ot include the unknowns Ve and He since
they have no ~oment about C because of zero
ams. As in exa~ple problem 1, the reactions
at A and B have been r-ep Iac ed by H and V com-
ponents ins:ead of ~sing an angle (direction)
as an unknown characteristic. The struct~e is
statica:ly dete~inate.
Exa'Ilple Problem 3.
p p p P
(1)~ .~
t E I
~In \~S::
-,
,He fRB .J.VC
Fig. A2.4
~ig. 2.4 shows a straight ~ember 1-2 carrying a
known load system ? and supported by 5 struts
A2.5
attached to reaction paints ABeD.
At reaction points A, E and D, the react.on
is known in direction and pOint of application
but the magnitude is ~~own as indicated by the
vector at each support. At point C, the re-
action is unknown in direction because 2 struts
enter jOint C. Magnitude is also unknown but
paint of application is known since the reaction
must pass through C. Thus we have 5 unknowns,
namely, Re, Ra, RO, Vc and He. For a copjanar
force system we have 3 equilibrium equations
be that we have a statically indeterminate
structure to (5-3) = 2 degrees redundant. How-
ever, observation of the structure shows two
internal pins at points E and F which means
that the bending moment at these two points is
zero, thus giVing us 2 more equations to use
drawing tree bodies of the structure to lett at
pin E and to right or pin F and equating moments
about each pin to zero we obtain 2 equations
which do not includ& unknowns other than the 5
tore statically determinate.
Example Problem 4.
):0
.. c
1;A
0
""llC'fMc "'-Rn
Vc
Fig. A2.5 Fig. A2.6
Fig. 2.5 shows a beam AB which carries a
super-structure CED which in turn is SUbjected
to the known loads P and Q. The question is
The external unknown reactions for the entire
structure are at points A and B. At A due to
the roller type of action, magnitude is the only
unknown characteristic ot the reaction since
direction and point ot application are known.
At B, ~gnitude and direction are unknown but
paint of application 1s known, hence we have 3
unknowns, namely, RA, VB and Hg, and with 3
equations ot equilibrium available we can find
these reactions and therefore the structure is
statically dete~inate with respect to external
reactions. we now investigate to see it the
internal stresses can be found by statics arter
haVing found the external reactions. Obviously,
the internal stresses will be affected by the
internal reactions at C and 0, so we draw a free
body of the super-st~cture as illustrated in
?ig. 2.6 and consider the internal forces that
eXisted at e and 0 as exterr~l reactions. In
the actual structure the members are rigidly
attached together at point C such as a Nelded or
J 1 'L.-
I
 ' _
A2.6 EQUILIBRIUM OF FORCE SYSTEMS.
~ultiple bolt connection. This ~eans tha: all
three force or ~eaction charact9ristics, na~ely,
magnitude, direction and point of a?pllcation
are unknown, or in o~her words, 3 ~owns
exist at C. For convenience we will represent
these unknowns by three components as sho~m in
Fig. 2.6, namely, He, Vc and Me. At jo Int D in
Fig. 2.6, the only unknown regarding the re-
action is RD a magnitude, since the pin at each
end of the nember DE establishes the direction
and point of applIcation at the reaction RD.
Hence we have 4 unknowns and only 3 equations
of equilibrium for the structure in Fig. 2.6,
thus the structure is statically indeterninate
with respect to all of the internal stresses.
The student should observe that internal
stresses between paints AC, 80 and FE are
statically determinate, and thus the statically
indeterminate portion is the structural tri-
angle CEDC.
Example Problem 5
Figs. 2.7, 2.8 and 2.9 show the same
structure carrying the same known load system
P but with different support conditions at
points A and B. The question is whether each
structure is statically indeterminate and if
so, to what degree, that is, what number of
unknowns beyond the equations of statics avail-
able. Since we have a coplar~r force system,
only 3 equations at statics are available for
equilibrium of the structure as a whole.
In the structure in Fig. 2.7, the ~eaction
at A and also at B is unknown in ~gnitude and
direction but point of application is ~~own,
hence 4 urJcnowns and with only 3 equations of
statics available, makes the structure
statically indeterminate to the first degree.
In Fig. 2.8, the reaction at A is a rigid one,
thus all 3 characteristics of magnitude,
direction and point of application of the re-
action are unknown. At point B, due to pin
only 2 unknowns, na~ely, ~gnitude and di-
rection, thus making a total of 5 unknowns
with only 3 equations of statics available or
the structure is statically indeterminate to
the second degree. In the structure of Fig.
2.9, both supports at A and B are rigid thus
all 3 force characteristics are unknown at each
support or a total of 6 unknowns which makes
the structure statically indete~inate to the
third degree.
'\{4x5(
TRUSS STRUCTURES.
Example Problem 5
Fig. A2. 11 Fig. A2. 12
Fig. 2.10 shows a 2 ~a7 ~~~ss supported at
points A and B and carrying a known load syst&~
P, Q. All members of the truss are conr.ectec
at their ends oy a common ~in at each jCi~t.
The reactions at A and a are applied trIough
fittings as indicated. The question is wt-ether
the structure is statically dete~inate.
Relative to external react~or.s at A and B the
structure is statically det8~inate ~eca~se the
type of support produces only one unkncwn at A
and two unknowns at B, name Iy , VA, 'IS and HS as
shown in Fig. 2.10 and we have 3 equations of
static equilibrium available.
~e now investigate to see ~~ we can ~ind
the internal ~ember stresses after :iaving :ow~d
the values of the reactions at A and 8. Suppose
we cut out joint B as indicated 8Y section 1-1
in Fig. 2.10 and draw a free body as sho'#TI in
Fig. 2.11. Since the ~embers of the truss ~ave
pins at each end, the loads in theSe ~enbers
~ust be axial, thus direc~ion and 11~e of action
is known and only magnitUde is '~~~~own. In
rig. 2.11 HE and ~~ are known but AS, CE, ~~d
DB are ~~own in magnitUde hence we have 3 un-
knowns but only 2 equations of equilibrium for
a coplanar concurrent force syste~. If we cut
through the truss in Fig. 2.10 by the section
2-2 and draw a free body of the lcwer'portion
as shown in Fig. 2.12, we ~ave 4 unknowns,
namely, the axial loads in CA, DA, C8, DB but
only 3 equations of equilibrl~ available for
a coplanar ferce sys~em.
Suppose we were able to find the stresses
in CA, DA, CB, DB in some ~nr.er, and we would
now ?Toceed to joint D and treat it as a tree
body or cut through the upper panel along
section 4-4 and use the lower portien as a free
body. The same reasoning as used above would
show us we have one mor-e unknown than the number-
of equilibrium equations available and thus
we have the truss statically lncete~1~t8 to
the second deg~ee relative to internal member
stresses.
Physically, the structure has t~o ~ore
members than is necessary for the stability of
the structure under load, as we cou:d leave
out one diagc~l member in each tr~ss panel and
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
the strJcture Nould be still stable and all
~ember axial stresses could be found by the
equations of static equilibrium without regard
to their size of cross-section or the kind at
material. Adding the second diagonal member
in each panel would necessitate knOWing the
size of all truss members and the kind of
material used before member stresSes could be
found, as the additional equations needed must
come from a consideration involving distortion
of the truss. Assume for exa~ple, that one
diagonal in the upper panel was lett out. We
would then be able to find the stresses in the
~embers of the_upper ?anel by statics but the
lower panel Would still be statically inde-
terminate to 1 degree because or the double
diagonal system and thus one additional equation ternal single pin connections at points a and 0'.
is necessary and would involve a consideration
of truss distortion. (The solution ot static-
ally indeterminate trusses is covered in
Chapter A.B.)
A2.7 Example Problem Solutions of Statically Determinate
Coplanar Structures and Coplanar Loadings.
Although a student has taken a course in
sta:ics ~efore taking a beginning course in
aircraft structures, it is felt that a limited
~eview of )roblems involving the application
of the equations of static equilibrium is quite
j~stlfied, part~cularly 1: the prOblems are
posSibly somewhat more difficult than ~ost or
~he problems in the usual begir~ing course in
stat~cs. Since one ~ust use the equations or
static equilibrium as ?art ot the necessary
equations in solving statically indeterminate
st~uctures and since statically indeterminate
structures are covered in rather complete detail load on the spar bas been replaced by the re-
Ln other c napt ers at :::115 book, only limited
s?ace will be given to ~roblems involVing
statics in this chapter.
~a~ple Problem 8.
Fig. A2.14 shows a much SI~plitied wing
structure, cor~lsting of a wing spar supported
'oy lift and cabane struts wmch tie the wing
spar to the fuselage structure. The distributed figure.
air load on the wing spar is unsymmetrical about
the center line of the air!r~e. The wing spar
is ~ade in t~ree units, readily disassembled by graphically by an arrow head on ~he end of a
using ern fittings at points 0 and 0 1 • All
su~porting wing struts have Single pin fitting
~l~ts a: eac~ end. The problem is to deter-
201ll/in. 401ll/in.
..I...,....,--r;i t • j " " ' . i
~Oi/in.
Ill'
l5#/in.
t f ! ,--,
-r i- 45"~ 82" ---iZO'!_60" ---! Hinge
o' , ';0
A 8' B 0" A PIn
Lilt Caban , -+-
Struts truts C~ c 36"
Fuselage -' "30":.......i..-
E' E Fig. A2.l4
'i 8ym. 'i
A2.7
~ine the axial loads in the members and the re-
actions on the spar.
Solution: The first thing to decide is whether
the structure is statically determinate. From
the figure it is observed that the wing spar is
supported by five struts. Due to the pins at
each end of all struts, we r~ve five unknowns,
~~elYJ the magnitude of the load in each strut.
DirectIon and location at each strut load is
known because of the pin at each end at the
struts. We have 3 equations of equilibrium for
the wing spar as a single unit support'ed by the
5 struts, thus two ~ore equations are necessary
it the 5 unknown strut loads are to be found.
It is noticed that the wing spar includes 2 in-
This establishes the fact that the moment of all
forces located to one side of the pin must be
equal to zero since the single pin fitting can-
not resist a moment. Thus we obtain two addi-
tional equations because of the ~NO internal pin
fittings and thus we have 5 equations to :ind 5
unknowns.
Fig. 2.15 shows a tree body of the wing
spar to the right of hinge fitting at O.
1013",(30+ 15)45
I 20~ 2
Fig. A2.15
In order to take moments, the distributed
sultant load on each spar portion, namely, the
total load on the portion acting through the
centroid of the distributed load system. The
strut .r-eact.t on EA at A tas been shown in phantom
as it is ~ore convenient to deal with its com-
ponents YA and XA' The reaction at 0 15 un-
known tn ~gnltude and direction and for con-
venience we will deal with its components xo
and YO. The sense assumed is indicated on the
The sense of a force Is represented
vector. The correct sense is obtained from the
solution ot the equations of equilibrium since,
a force or ~oment must be given a plus or minus
Sign in writing the equa~ions. Since the sense
of a force or moment is unknown, 1t is assumed,
and if the algebraic solution of the equilibrium
equations gives a ~lus value to the magnitude
then the true sense is as ass~ed, and oppOSite
to that assumed it the solution gives a minus
sign. If the unknown forces are axial loads in
~embers it is camoan practice to call tensile
stress plus and compressive stress minus, thus
if we ass~.e the sense of an unknown axial ~oad
~s tenSion. the solution of the equilibr!um
A2.8 EQUILIBRIUM OF FORCE SYSTEMS. TRUSS STRUCTURES.

equations will give a plus value for the magni- To find strut load g'C I ~ake m~ents about
tude of the unknown if the true stress is point C.
tension and a ~inus sign will indicate the
assumed tension stresses should be reversed or L~ ~ 1325 x 65 + 2000 x 40 + (58eO - 44JO)
compression, thus giving a consistency of signs. 30 - 1500 x 10 - 993 x 35 - 30 (3'8')
30/33.6 =0
To find the unknown YA we take ~cments
about point 0 and equate to zero for equilibrium whence, BtCI = 6000 lb. with sense as sho~TI.

- 2460 x 41 - 1013 x 102 • 82YA =0 To find load in member BIC use equation
ZFy '::: 0 =. 1325 + 2000 + 1500 + 993 - 6000
Hence YA = 204000/82 = 2480 lb.. The plus sign
means that the sense as assumed in the figure (30/33.6) - 2720 (30/33.6) - B'e (30/54)
1s correct. By geometry XA = 2480 x 117/66 = =0
4400 lb. and the load in strut ~ equals
v4400 3 + 2480 3 = 5050 lb. tension or as whence, BIC =-
3535 lb. The ~inus sign
assumed in the :lgure. means it acts opposite to that sho'Nn in fi~~e
or is compression instead of tension.
To find Xo we use the equilibrium equation
ZFX = 0 = Xc - 4400 = 0, whence Xc = 4400 lb. The reactions on the spar can now be
deter.nined and shears, ber.ding moments and
To find YO we use, axial loads on the spar could be round. The
numerical results should ~e checked for eqUili-
ZFy = 0 = 2460 + 1013 - 2480 - YO = 0, whence brium or the spar as a whole by taking moments
YO = 993 lb. ot all forces about a dif:erent ~oment center
to see i t the result "is zero.
To check our results for eqUilibrium we
will take moments at all torces about A to see Example Problem 9.
it they equal zero.
l:11A = 2460 x 41 - 1013 x 20 - 993 x 82 = 0 check
On the spar portion OIA' J the reactions -r-r-'
are obViOUSly equal to 40/30 times those :ound
for portion OA since the external loading is 40
I
strut i ~ 12 12"
as compared to 30.
i\~ I ~ Brace Strut I

Hence AIEl. 6750, XOI = 5880, YO' = 1325 24lJ~ \, E' ---+--
i
12 ~ I
Fig. 2.16 shows a tree body ot the center 12"
spar portion with'the reactions at 0 and at as
found preViously. The unknown loads in the
struts have been assumed tens10n as shown by
the arrows.
Fig.A2.17 I
F--

Shock
Strut
+
12"
1500",50x30
A -L
r 300

" 30
'':In,"
15
, " Fig. 2.17 shows a Simplified airplane
C"9----~ landing gear unit with all members and loads
I Flg.A2.16 confined to one plane. The brace struts ~e
I I pinned at each end and the support at C is of
L __ J the roller type, thus no vertical reaction can
be produced by the support fitting at point C.
To find the load in strut Be take moments The member at C can rotate on the roller but
about s' horizontal movement is ~revented. A knovm load
ot 10,000 lb. 1s applied to axle ~nit at A. The
!MB' = 1325 x 20 - 2000 x 5 - 1500 x 55 problem is to f1nd the load in the brace struts
and the reaction at C.
- 993 x 80 • 60 (BC) 30/33.6 = a
Solution:
Whence, Be • 2720 lb. with sense as Due to the Single pin fitting ~t each end
assumed. of the brace struts, the ~eact1ons ~t a and D
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A2.9

are collnear with the strut axis, thus direction the aluminum alloy tubular truss. Trussed type
and point of application are known for reaction beams composed of closed and open type sections
RB and RD leaving only the magnitude of each as are also frequently used in Wing beam construc-
unknown. The roller type fitting at C fixes t1on. The stresses or loads in the members ot
the direction and point of application of the a truss are commonly referred to as nprimaryW
reaction Re, leaving magnitude as the only and wsecondarJn stresses. The stresses which
unknown, Thus there are 3 unknowns Re, He and are found under the following assumptions are
RO and with 3 equations of static equilibrium referred to as primary stresses.
available, the structure is statically deter.ni-
nate with respect to external reactions. The (1) The members of the truss are straight,
sense of each of the 3 unknown reactions has weightless and lie in one plane.
been assumed as indicated by the vector. (2) The members of a truss meeting at a'
point are considered as jOined together by a
To find Rn take moments about point B:-
common frictionless pin and all member axes in-
ZME = - 10000 sin 30° x 36 - 10000 cos 300 x 12 tersect at the pin center.
- RO (12/17) 2~ = 0
(3) All external loads are applied to the
whence, RO ~ - 16750 lb. Since the result truss only at the jOints and in the plane ot
comes out with a minus Sign, the reaction RO the truss. Thus all loads or stresses produced
has a sense OPPosite to that shown by the in members are either axial tension or compres-
vector in Fig. 2.17. Since the reaction RO is sion without bending or torsion.
colinear with the line DE because of the pin
endS, the load in the brace strut DE is 16750 Those trusses produced in the truss mem-
lb. compression. In the above moment equation bers due to the non-fulfillment or the above
about B, the reaction RO was resolved into assumptiOns are referred to as secondary
vertical and rorizontal components at point D, stresses. Most steel tubular trusses are welded
and thus only the vertical component which together at their ends and in other truss types,
equals (12/17) RD enters into the equation the ~embers are riveted or bolted together.
since the horizontal component has a line of This restraint at the joints may cause second-
action through point B and therefore no moment. ary stresses in some members greater than the
He does not enter in equation as it has zerO primary stresses. Likewise it is common in
moment about B. actual practical deSign to apply torces to the
truss members between their ends by supporting
To find RE take ZFv = 0 many equipment installations on these truss
members. However, regardless of the magnitude
ZFv = 10000 x cos 30° + (- 16750)(12/17) + HE at these so-called secondary loadS, it is
(24,/26.8) = 0 cammon practice to first find the priwary
Whence, HB = 3540 lb. Since Sign comes stresses under the assumption outlined above.
out plus, the sense is the same as assumed in
the figure. The strut load BF is therefore GENERAL CRITERIA FOR DETERl'lINING WHEI'JlER
3540 lb. tension, since reaction HB is collnear TRUSS STRUCTURES ARE STATICALLY DEI'ERMINATE
wi th line SF. WITlf RESPECT TO INl'ERNAL STRFBSES.

To find He take ZH : 0 The simplest truss that can be constructed

ZH = 10000 sin 300 - 3540 (12/26.8) + (-
(12/17) + He 0 =
whence, Rc = 8407 lb. Result is plus and
therefore assumed sense was correct.
To check the numerical results take
~aments about point A for equilibrium.
ZMA = 84,07 x 36 + 354,0 (24,/26.8) 12 - 3540
(12/26.8) 36 + 16750 (12/17) 12 - 16750 (12/17)
36 = 303000 + 38100 - 57100 + 142000 - 426000
:. 0 (Check)
A2. B Stresses in Coplanar Truss Structures Under
Coplanar Loading.
In aircraft construction, the t~~s type
of construction 1s quite co~on. The ~cst
common is the tUbular steel welded tr~sses :h~t
make up the fuselage frame, and less freq~~rr:ly,
is the triangle which has three members m and
16750) three joints j. A more elaborate truss consists
of additional triangular frames, so arranged
that each triangle adds one joint and two mem-
bers. Hence the number of members to insure
stability under any loading is:
m = 2j - 3 -----(2.8)
A truss haVing fewer members tillL~ required
by Eq. (2.8) is in a state at unstable eqUili-
brium and will collapse except under certain
conditions of loading. The loads in the members
of a truss With the number of members shown in
equation (2.8) can be :ound With the available
equations of statiCS, since the forces in the
me~bers acting at a paint intersect at a common
point or form a concurrent force system. For
thiS type of force system there are two static
equl11brium equations available.
Thus for j number of joints there are 2j

I .
, ~.
/'
A2.10 EQUILIBRIUM OF FORCE SYSTEMS. TRUSS STRUCTURES.

equations available. However three i~dependent 500

equa~ions a~e neceSsa~J to determine the ex~er­
2 2
nal reactions, thus the number of equations L, L, , i

necessary to solve :or all the loads in the

members is 2j - 3. Hence if the nw~Jer of truss
-I lL"

~embers is that given by equation (2.8) the

truss is statically deter.ninate relative to the
40" T 40

primary loads in the truss ~embers and the

truss is also stable. -l ""{'--~---T-'
- U
c
U
\
\ U 3
3
30
-::,I('
-,
Fig. A2. 18
1

f--- 30" -+-

1 2

It the truss has more members than indi- 30" -+- 30" -----.j
cated by equation (2.8) the trusS is considered i i i
redund~~t and statically lnGeterminate since ',1000
the member loads cannot be found in all the
members by the laws of statics. Such redundar.t
(1
L~~ ~_:)oo
, -
structures it the members are properly ?laced Fig. A2. 19 \ - ,
are stable and will support loads of any u,~
arrangement.
or tension. In equation (0) the load of 1250
ANALYTICAL IlE:rHODS fOR DE:l'SRtrINI'IG in L~L:;l was substituted as a minus value since
PRIMARY STRESSES IN TRUSS STRUCT1JRES it was foune to act opposite to ~hat sho~m in
Fig. A2.l9. Possibly a tetter p r-ccedur e would
In general there are three rather distinct be to change the sense of the ar~ow i~ ~he :ree
methods or procedures in applying the equations bodY diagram for any solved ~embers Jefore writ-
of static equilibrium to finding the primary ing further equilibriUW equati~ns. ~e ~ust
stresses in truss type structures. They are proceed to joint L:::I instead c r joint U2, as
otten referred to as the method of joints, three ~~~own nembers still exist at jOint U2
moments, and shears. Whereas only ~NO at jo~nt La. Fig. A2.20 shows
f~ee body of jOi~t La cut ou~ by section 2-2
A2. 9 Method of Joints. (see ~lg. AS.IS). The sense of the urJL~own
It the truss as a whole is i~ equilibrium
member s~ress L:;lU:;l has Jeen assw~ec as COID-
pressic~ (pUShing toward jOint) as ~t is ob-
then each member or joint in the tr~ss must
likewise be in equilibrium. The forces in the Viously act~ng this way to Jalance :he 500 lb.
members at a truss jOint intersect in a common load.
point, thus the forces on each joint form a 500
concurrent-coplanar force system. The ~ethod
~, ~H
of joints consists in cutting out or isolating
L. 250
a joint as a free bOdy and applying the laws ot u.,
equilibrium for a Concurrent force system.
Since only two independent equations are avail- tu. Fig. A2. 20
able for this type of system only two unknowns
can exist at any jOint. ThUS the procedure is For equilibrium of jOint La, ZH and ZV 0 =
to start at the jOint Where only two unknowns
exist and continue ,rogressively throughout the
ZV =-500 + LaUa = 0, whence J LaUa = 500 lb.
Since the Sign came out plus, the assumed sense
truss joint by joint. To illustrate the method in Fig. A2.20 was correct or compreSSion.
consider the cantilever truss of Fig. A2.18.
From obse~lation there are only two members ZH = 250 - LaLl : 0, wr.ence LaLl =250 IJ.
With internal stresses unknown at jOint L~.
Fig. A2.19 shows a free body of jOint L~. The Next consider joint Ua as a :ree jody c~t
stresses in the members L~ La and L~ U~ have out by section 3-3 in Fig. A2.18 and erawn as
been assumed as tension, as indicated by the Fig. A2.2l. The known member s~resses are showr.
arrows pulling away from the jOint L~. with their true sense as ~reviouslY found. The
two ~own member stresses U2L l and UaU l have
The static equations or equilibrium for Jeen assumed as tension.
the forces acting on joint L~ are ~~ and ZV =o.
!V =- 1000 _ L,U. (40/50) =a - - - - - - -fe )
whence, L ~U:;l =-
1250 1'0. Since the Sign Fig. A2. 21
came out minus the stress is opPOSite to that
aSStooed in Fig. A2.l9 or compression.
Eli =- 500 - (- 1250)(30/50) - L,L, =0 - -(b) ZV = -500 - 1250 (40/50) + UaLl (40/50) = 0
Whence, LaL~ = 250 lb. Since Sign comes whence, UaLl = 1875 is. (tens 1on as
out plus, sense is s~e as assumed in figure asaumed ,.)
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A2.11

ZH = (-1250) (30/50) - 1875 (30/50) - U.U. =0 The algebraic sign of all unknowns came out
whence, U..U a = - 1875 lb. or opposite in
positive, thus the assumed direction as shown
sense to tr2t assum~c ana therefore compression. an Fig. A2.22 was correct.

Note: The student should continue with succeed- Check results by taking &~B = 0
ing joints. In this example involving a canti- ~B = 1400 x 150 + 500 x 30 - 500 x 120 - 500 x
lever truss it was not necessary to ~lnd the 30 - 1000 x 90 - 1000 x 60 = 0 (Check)
reactions, as it was ~oSSlble to select joint
L 3 as a joint involving only VNO unknowns. In To determine the stress in member Fl , Fa and F"
trusses such as illustrated in Fig. A2.22 i t 1s we cut the section 1-1 thru the truss (Fig.
necessary to first find reactions R:l. or R a wni cn A2.22. Fig. A2.23 shows a free bOdy diagram of
then provides a joint at the reaction point in- the portion of the truss to the left of this
volving only two ~cwn forces. section.

f
, 'f~ _

Fig.A2.22

1400
a
o ,
500 1
500
~:
~; •

1400 1

A2. 10 Method of Moments. Fig. A2. 23 FIg. A2. 24

For a coplanar-non-concurrent force system The truss as a whole was in equl1ibrium

there are three equations of statics available. therefore any portion must be in eqUilibrium.
These three equations may 08 taken as moment In Fig. A2.23 the internal stresses in the mem-
equations about three different ~oints. Fig. bers F l , F~ and F~ which existed in the truss as
AZ.22 shows a typical truss. Let it be re- a whole ~ow are considered external forces in
qUired to find the loads in the members F l , F a J holding the portion of the truss to the left of
F", F. J r , and F o ' sectien 1-1 in e~Jil1brium in combination with
the other loads and reac~ions. Since the ~em­
Fig.A2.22
ioo""
3 2
'F"
1100M500#
bers a and b in Fig. A2.23 have not been cut the
loads in these ~embers remain as internal
stresses and have no i~fluence on the equilib-
10"
rium of the portion ot the truss shown. Thus
the ~ortion at the truss to lett ot section 1-1
HA----"'(----+-\-"''-+-''--.''-'-'-----''!---''>I could be considered as a solid block as shown
, 50~f i~ Fig. A2.24 without affecting the values at
Fl , F. and F". The ~ethod of ~oments as the
name unpIles involves the operation of taking
moments about a point to find the load in a
The first step in the solution is to find the particular member. Since there are three un-
reactions at pOints A and B. Due to the roller knowns a moment center must be selected such
type of support at B the only unknown element of that the moment of each of the two unknown
the reaction force at B 1s ~gnitude. At paint stresses will have zero moment about the selected
A, magnitude and direction of the reaction are moment center, thus leaving only one illL~Own
l~own giving a total of three ~~cwns With ~orce or stress to enter into the equation tor
three equations of statics available. For con- moments. For exa~ple to cetermine load F 3 i~
venience the wllic~own reaction at A has been re- Fig. A2.24 we take moments about the inter-
~laced by its unknown H ~~d V components. section of forces F l and Fa or paint O.
Takir~ momer..ts about ~oint A, Thus Z~O = 1400 x 30 - 18.97 F, =0
L~A = 500 x 30 + 100e x 60 + 1000 x 90
18.97 = ~")21:::'-' ,,,
T 500 x II ... = 42000 ( or
30 + 500 x 120 - 150 VB = a il.ence "3 ~..;. c ompr-eas t cn
acting as assumed)
Her-ce Vg = 1600 lb.
To ~ind the a~ of the force F 3 tram :he
Take ZV = 0 momen~ center a involves a s~~ll a~o~~t of cal-
z.v -, - 1000-1000-500-500
:= '/. + 1600 = a there- Clllat!on. thus in general i~ 13 Simpler to re-
fore VA. = 1400 lb. solve the ~~nown force into H and V components
at a ~oi~t In ~ts 11ne Ji action such ~hat one
Take ZH = 0 of these ~omDonents passes thru the ~oment
ZH = 500 - nA = 0, therefore HA = 500 lb. center and the arm of the other cJmponent can
~sually be dete~1~ed Jy inspecticn. Thus in
A2.12 EQUILIBRIUM OF FORCE SYSTEMS. TRUSS STRUCTURES.

Fig. A2.25 the ferce F 3 is resolved into its the other two ~Jcnowns F~ and ~s lles at i~flni­
component F 3V and F3H at point 0'. 7hen taking ty. Thus for conditions where two of the 3 c~t
members are parallel Ne have a ~ethod of solving
~or the web ~ember of the t~uss ccmr.,only re-
ferred to as the ~ethod cf shea~s, or the sum-
mation of all the forces nO~ial to the tNO
parallel ~{nown chord members ~us~ equal zero.
Fig. A2. 2S
Since the parallel chord ~embers have no com-
ponent in a direction no~al to their line of
action, they do not enter the above equatton of
equilibrium.
1000lii
!
Fig. A2. 26

50'"'

1400lii
Fig. A2. 27
/IF,
500 ...."::,,,---¥--7"- F ,
500 3
moments about point 0 as before:- 11400
Fig. A2. 28

Referring to Fig. A2.27

whence, FeH ~ 2100 lb. and therefore
ZV = 1400 - 500 - 1000 - '.I1/v'2): a
F. = 2100 (21.6/30) = 2215 lb. as pre-
Viously obtained. whence F. = - 141 lb. (tension or OPPOSite
to that assumed in the ~igure.
The load Flo. can be found by taking moments
about point Ill, the intersection at torces F. To find the stress in illember F? , we cut
and F. (See Fig. A2.23). section 3-3 in Fig. A2.22 and draw ~ free bOdy
diagram of the left portion in Fig. AZ.28.
ZMm = 1400 x 60 - 500 x 30 - 500 x 30 Since Flo. and Fs are horizontal, the member F?
-30Flo.=Q must carry ~he shear on the truss on this sectior.
3-3, hence the name ~ethod of shears.
whence, Flo. = 2800 lb. (Tension as assumed)
ZV = 1400 - 500 - 1000 • F, =0
To tinct rcrce F. by using a moment equation,
we take moments about point (r) the inter- l~ence F? = 100 lb. (compression as assumed)
section ot rorces F J, and F 3 (See Fig. A2.26).
To eliminate solving for the perpendicular Note: The student should solve this example il-
distance trom paint (r) to line at action of lustrating the methods of moments and shears
F., we resolve Fa into its n and V components using as a tree body the portion of the truss to
at point 0 on its line at action as shown in the right of the cut sections instead of the
Fig. A2.26. lett portion as used in these illustrative ex-
amples. In order to solve for the stresses in
ZMr = - 1400 x 30 + 500 x 60 + 60 F.V =0 the members of a truss most advantageously, one
usually makes use of more tha.~ one cf the above
whence, FaV = 12000/60 = 200 lb. three methods, as each has its advantages for
certain cases or members. It 1s important to
Therefore F. = 200 x ~ =282 lb. com- realize that each is a method of sections and in
pression a great many cases, such as trusses with paral-
lel chords, the stresses can ?Tactically be
A2. 11 Method of Shears found mentally without writing down equations of
equilibrium. The following statements in gen-
In Fig. A2.22 to tind the stress in member eral are true for parallel chord trusses:
F. we cut the section 2-2 giving the free bOdy
for the left portion as shown in Fig. A2.Z7. (1) The vertical camponent of the stress ~n
the panel diagonal members equals the vertical
The mebhod ot moments is not sufficient to shear (algebraic sum of external forces to one
solve tor member F. because the intersection ot s1de of the panel) on the panel, since the chord
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
members are horizontal and thus have zero verti-
cal component.
(2) The truss verticals in general resist
the vertical component of the diagonals plus
any exter~l loads applied to the end joints ot
the vertical.
(3) The load in the chord members is due
to the horizontal components ot the diagonal
members ana in general equals the summation of
these horizontal components.
To illustrate the simplicity at determining in the index triangle or - 133 lb. It is nega-
stresses in tITe members of a parallel chord
truss, consider the cantilever truss of Fig.
A2.29 with supporting reactions at points A and against the joint for equilibrium.
J.
150
550 36"-+- 40"
Fig. A2.29
~lrst, compute the length triangles in
each panel of the truss as shown by the dashed
triangles in each panel. The other triangles
in each ~anel are referred to as load ar index
triangles and their sides are directly pro-
portional to the length triangles.
The shear load in each panel is first writ-
ten on the vertical side at each index triangle.
~hus, in panel EFGD, considering forces to the
right of a vertical section cut thru the panel,
the shear is 100 lb., which is recorded on the
vertical side of the index triangle.
For the second panel from the tree end, the
shea~ is 100 + 150 ~ 250 and for the third panel
100 + 150 + 150 = 400 Ib , , and in like manner
550 ~or fourth panel.
The loads in the diagonals as well as their
horizontal components are directly proportional
to the lengths of the diagonal and horizontal
side at the length triangles. 'Thus the load in
diagonal member DF = 100 (50/30) = 167 and ~or
member CO = 250 (46.8/30) = 390. The hori-
zontal compor.ent of the load in OF = 100 (40/30)
= 133 and :or CG = 250 (36/30) =
300. These
values are shown on the index triangles tor
each truss panel as shown in Fig. A2.29. We
start our analysiS for the loads in the ~embers
8f the tr~ss by considering joint E tirst.
USing LV = 0 gives EF = 0 by obser~ffition.
A2.13
since no external vertical load exists at joint
E. Similarly, by the same reasoning for LH ~ 0,
load in DE = O. The load in the diagonal FO
equals the value on the diagonal of the panel
index triangle or 167 lb. It is tenSion by
observation since the shear in the panel to the
right is up and the vertical component of the
diagonal FD must pull down for eqUilibrium.
ConSidering Joint F. ZH = - FG - FOR = 0,
which means that the horizontal component of the
load in the diagonal OF equals the load in FG,
or is equal to the value of the horizontal side
tive because the horizontal component of DF
pUlls on Joint F and therefore Fa must pUSh
Considering Joint D:-
ZV = DFv + DO = O. But DFv = 100 (vertical side
of index triangle)
• ", DG = - 100
LH = DE + DFR - DC =0, but DE=° and DFa =
133 (from. index triangle)
DC = 133
Considering Joint G:-
ZH=-GH - GF - GCR = O. But OF =- 133, and GCR
= 300 from index triangle in the second panel.
Hence OH =-
433 lb. Proceeding in this manner,
we obtain the stress in all the members as shown
in Fig. AZ.29. All the eqUilibrium equations
can be solved mentally and with the calculations
being done on the slide rule, all member loads
can be written directly on the truss diagram.
Observation at the results at Fig. AZ.29
show tr~t the loads in the truss verticals equal
the values of the vertical sides or the index
load triangle, and the loads in the truss di-
agonals equal the values of the index triangle
diagonal side and in general the loads in the
top and bottom horizontal trJSs members equal
the summation of the values ot the horizontal
sides of the index triangles.
The reactions at A and J are found when
the above general procedure reaches joints A
and J. As a c~eck on the work the reactions
should be determined treating the truss as a
whole.
Fig. A2.3C shows the solution for the
stresses in the members of a Simply supported
Pratt Tr~ss, symmetrically loaded. Since all
panels have the same width and height, only one
length triangle is drawn as shown. Due to
symmetry, the index triangles are drawn for
panels to only one side at the truss center
line. First, the vertical shear in each panel
is written on the vertical side of each index
triangle. Due to the symmetry of the truss and
~ ,/
.,/;c
 A2.14 EQUILIBRIUM OF FORCE SYSTEMS. TRUSS STRUCTURES.

loading, we know that or-e t~l~ of the exter~al

loads at jOints U3 and L 3 is supported a~ re- 312 + 127.5 = ~SS.S. T~ere~2re, lsad i~ U.~2 =
action R:I, and 1/2 at r-eac t.t cn R a , or shear- in - 499.2. Si:nllarl~' 3.':: jo inc La, :"'aL 3 = 312 .,.
127.5 = 499.5. At ~oint U~, ete ~o~iz:~tal
Length
Fig. A2. 30 co~~onents of U:l,U~ an~ U~U3 = ~99.S ~ 52.5 = 562

~
5" Triangle
30" which must be balanced ,:,y - 562 tr; member' U aU 3 '
33" 50 50 50 50
The reac:icn R~ equals t~e val~e on the
-499.5 U -562 lu, vertical side of our i~cex triang:e in the 91.j
j .9;> 0 panel, or 375. This should oe c~ecked usi~g
;.= 1'.9 .s '1 the truss as a whole and ta~lng ~c~ents ~jout
';~~ R, •
312 312 ~L~ 499.5 p..3 jL. If a truss is lcacec unsJ~etr~cally, t~e
i!R, 100 100 100 100 100 reactions should be ceter.r.ined ~irst, atter
6 Panels @ 25" = 150" which the index triar.~:es can be dra'HTI, sta:-t-
ing with the end ~anels, since the ;anel shear
1s then readily calculated.
center panel = (100 + 50) 1/2 = 75. The verti-
cal shear in panel U:I,UgL1L a equals 75 pl~s the A2.12 Aircraft Wing Structure. Truss Type with Fabric
external loads at Ua and La or a total of 225 or Plastic Cover
and Similarly for the end panel shear = 225 +
50 + 100 = 375. With these values kTIQwn, the
TI:e metal covered carrt t Lever' wing 'N1t;'1 i ts
other two sides of the index triangles are di- better overall aerodyna~lc e~:iciency and s~f­
ficien~ torsion~l ri 6 ! : l t y has ~ractlcal1y re-
rectly proportio~41 to the sides of the length
triangles tor each panel, and the results are as placed t~e externally braced wing excep~ for lew
sho~m in Fig. AZ.30.
speed cornmerc~al or ?rivate )i:ot aircraft as
illustrated by the aircra:t 1~ tigs. A2.31 a~d
32. The wing covering is usually fabric and
The general procedure fram this paint is to
find the loads in the diagor21s, then in the therefore a drag tr~ss inside the wi~g is
verticalS, a~d finally in the horizontal chord necessarJ to resist loads in the drag t7USs
member's •
direction. Figs. A2.33 and 34 shows the gen-
eral structural layout a: such wings. The two
The loads in the diagonals are equal to the spars or beams are ~etal or wood. Instead of
va.Iues on the hypotenuse 0 f the index triangles.
USing double wires in each d~g truss bay, a
The sense, whether tension or campression, is Single diagonal strut capable ot t~<ing either
deter.nlned by inspection by cutting ~ental tension or compr-ass tve loads could be -rsec ,
sections thru the truss and noting the direction The external brace s t r-ut s are stream j rne tuocs .
ot the external shear load which ~st be bal-
anced by the vertical camponent of the diagon-
als.

The loads in the verticals are deter.nined

by the method of joints and the sequence of
jOints is so selected that the stress in the
vertical member is the only unknown in the
equation ZV = 0 for the joint in question.

Thus for jOint U3 , ZV = - 50 - U3L 3 = a

or U 3L3 = - 50.

For jOint U~, ZV = - 50 - UaL 3 v - U~L~ = 0,

but U~13V =
75, the vertical cornpo~ent of Ua L3 Fig.A2.31 Piper Tri-Pacer
rr-cm index t r tang Le. . ", U ~L~ 50 - 75 = =-
- 125. For JOint L~, ZV = - 100 + L:I,U 1 = 0,
hence L1U 1 = 100.

Since the horizontal chort me~bers receive

their loads at :~e joints due :0 horizontal
components of ~he diagonal members of the tr~ss'l
we can start a: La and add u9 these horizontal
components to obtain the chord stresses. Th~S, I
LoLl = 312 (from index triangle). L1L 2 = 312 I
from ZH = a for joint L1 • At joint U"I.' the i Fig. A2. 32 Champion Traveler
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A2.l5

Wing Tip (End) Bow in all ~enbers of the lift anc drag tr~sses will
11e / be dete~,ine~. A Si~plified air loac:ng Nill be
Fig. A2. 33 ,; ~ (0., Plywood Tip Fairing ass~~ed, as the ?ur~ose of this problem is to
;iv8 the st~dent practice in solving sta:ically
, Aileron determinate space t~uss str~ctur8S.
~
, t: ~\ Aileron Spar

Plywood Leading
Edge Fairing --l........;'1
I
I
,

/
X
\.,r:=r
.»"----.ll
~'I Compression Rib

Drag wtre
ASSL~D AIR LCADING:-

(1) A constant spanwise lift load of 45

Ib/in from hin~e to strut point and then taper-
D"gwire~ "! -/ ing to 22.5 Ib/tn at the wing tip.
:1 Anti-Drag Wire
Fitting \ :: :~' \
Aileron Rib
, (2) A [erHard uniform distributed drag
,
:1 V 11'1

\,i. Aileron Hinge load of 6 lb/in.

: / I:
, Drag Strut or
Leading '1\ " I Compression Tube
Edge- ': The above airloads represent a high angle
I
~ Forming or Plain Rib of attack condition. In this condition a for-
"
/ i :-+r ~ Trailing ward load can be placed on t~e drag truss as
Edge
I : A I Ir-- illustrated in Fig. A2.36. Projecting the air
I, I

Wing Hinge / ' :

"
"

" Butt Rib

Beam or Spar

-
Fitting ~
Airstream ~--,...-,-/i L ift
Fig. A2. 36

Drag
Drag Truss
Front
Beam
,
Center of
Pressure Rear
Beam

litt and drag forces on the drag truss direction,

~a~~le Problem 10.
plane ';ing Structure
A2.35 shows the structural d~mensional
~ig.
j~agram of an externally braced nonoplane wing.
The wing is fabric covered between wing beams,
ace thus a drag truss composed of struts and
tie rods is necessary to prov~de strength and
r~gid1ty in the drag direction. ~he axial loads assumed as shown in Fig. A2.37.
DRAG TRUSS
the fonvard projection ~ue to the lift 1s great-
er than the rearNard projection due to the air
drag, which dl~ference ~n our exa~ple problem
Fig. A2. 34 has been assumed as 6 Ib/in. In a low angle of
attack the load in the drag truss direction
~~e~nally Braced Mono- would act rearward.
SOLUTION:
The ~~ing loads on the front and rear
bea~s will Je calculated as the first step in
the solution. ?or our flight condition, the
center of pressure or t~e air~orces ~ill be
1------- 72" !
I

=:i"':
-'"
;; Wlng Chor-d e 72"
Cr
lO~'+- 36" ---1

F.Bl
e p
. •
R.B.
~
I

1 3 (3) 1.1 24.21" I

--,--- II
22.52" Fig. A2. 37
3° Dimedral
N '
[SRI T~e running load on the front bea~ will Je 45 x
, It "" 24.2/36 = 30.25 Ib/:n., a~d the re~alnder or
Airplane I 1
?. -i-(5lfi.6 ,..-- -!(6) 45 - 30.25 = 14.74 Ib/in gives the load on the
Fuselage --l ::-ear beam.
Fig. A2.35

- ' ,1..
I
A2 16 EQUILIBRIUM OF FORCE SYSTEMS TRUSS STRUCTURES

To solve ~or loads in a t~uss system by a

method of jOints, all loads ~ust be transferred
to the truss joints. The wing bear,s are sup-
~orted at one end by the f~sela~e and o~tooard
4
'I"A
1/
/1' x'\.k/!\I x/1"
\.,/
R. B. 3'---_71

'\.1/
X /11
"'-I
Drag Truss

by the two 11ft st~Jts. Thus we calc~late the

reactions on each beam at the strJt a~d ~lnge
paints due to the
beam.
~~~l~g lift load on each
:;;/.::B.,r R. B.
3

F Fl"Qot Lift R Rear

5 Truss / ; Lift
Front Seam
Truss

~ember lengthS L and :he ~ar.poner.t ra.t ; os

fellow by Simple ca:culat~on.

Table .\2.1

Taking moments about pof nt (2) lIember gym.: V D , t, I V;1. I D;1. I ';1.
114.5R .. - 114.5 x 30.26 x 114.5/2 - 15.13 x Front Be" FB
I 5.99 o 114.34 114.501.05231 0 1.9986
70.5 x 149.75 - 15.13 x 35.25 x 138 = O. Rear Beall. RB , ~.99 o 114.34 114.501.05231 0 1.9986 I

hence Front Strut

'F 15 7• 9 9 11 114.34 128.791.45011.08541.8878 i

RJ.= 3770 lb. Bear Stru.t

I '. j 57.49 o 114.34128.00i.44861 0 1.89301

ZV = 0 where V direction
V ... vertical direction,
Take 1s taken normal to
beam I D • drag direction,

ZV =- R~
- 3770 + 30.26 x 114.5 +
S ~ side direction,
1
(30.26 + 15. 3 ) 70.5 =0 L ~ Vv2+ D2 + 32

2
hence R~ =
1295 lb. We stare the solution of joints jy sta~i~g
(The student should always check results by w~th joint (1). Free bOdy sketches cf joint (I)
taking moments about pOint (1) to see it ZM .. are sketchel below. All members are considered
=0) two-f or-c e r.ember-s or havtng pf ns at each end,
thus magnt .ude is the cn.Iy unknown character-
Rear Beam istic of ~~ch member load. The jrag truss ~em­
bers coming in to jofnt (l) are replaced :,y a
single reaction called Jl.' Ar t er- Dl. is r ounc ,
(W'" 14. 7#/in. ~
its influence in causlr.g loads in drag truss
(4) 1I ! Ilftltll;,,/2
114. 5" ija 70. 5,,--1
~embers can then be found when the drag :russ as
a whole is treated. In :he joint solution, the
R. R, drag truss has been assumed parallel to drag
direction which is nct quire true frc~ Fig.
The rear beam has the same span dimensions but A2.35, but the error on member loads is negli-
the loading is 14.74 Ib/in. Hence beam re- gible.
actions R~ and R3 will be 14.74/30.25 = .4875
tilnes those for front beam. JOINT 1 (Equations of ~quilibri~~)
hence R s = .4875 x 3770 = 1838 lb.

~
770
R~ = .4875 x 1295 = 631 lb.
(1) D,
The next step in the solution is ~he B ~v
solving tor the axial loads in all the members. SF 4+D
We will use the ~ethod of joints and consider

zs == - 3770 x .0523--
the structure made up of trIee truss systems zv 3770 x .9986 .0523 FE- .4501 S" = 0- ( i
as illustrated at the top of the next col~. .;966 F& - .36'7'8 SF = 0- ( i
namely, a front li!t truss, a rear lift truss
and a drag truss. The beams are co~on to ~oth
ZD = - .0854 SF + 0, = 0 ----- - ----! , I

lift and drag trusses. SolVing equa t ions 1, 2 and 3, we obtain

Table A2.1 gives the V, 0 and S ~rojecticns F3 =- 8513 lb. (compression)
or the lift truss ~embers as dete~ined from
information given in Fig. A2.35. ~he true
SF = 9333 l~ . (tensior.)
0, = 798 lb. (aft i
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A2 17

Joint (3) (Equations of equt Lrbr-i ua) points (2) and (4). In the design of the beam
and fittings at this point, the effect of ~he
(drag truss actual conditions of eccentricity should of
t 1838

~
1 83 8 reaction on
RL.-J(3l (3) ---D, ptn (3\)
course be considered.
~ RB
R V-Splane SR V-Dplane Combined Loads on Drag Truss

:V = 1638 x .9986 - .0523 RB - .4486 SR = 0 -(4) Adding the two load systems of Figs. A2.38
:5 = -1838 x .0523 - .9986 RB -.8930 SR = 0 -(5) and A2.39, the total drag truss loading is ob-
ZD = D, + 0 = 0 - - - - - - - - - - - - - - -(6) tained as shown in Fig. A2.40. The resulting
member axial stresses are then solved for by the
Solving equations 4, 5 and 6, we obtain method of index stresses (Art. AZ.9). The
RB = - 4189 lb. (compression) values are indicated on the truss diagram. It
SR = 4579 lb. itension) is customary to ~ake one of the fittings attach-
03 = 0 ing wing to fuselage incapable of tranSferring
drag reaction to fUselage, so that the entire
Fig. A2.38 shows the reactions of the lift drag reaction from wing panel on fuselage 1s
struts on the drag truss at Joints (1) and (3) definitely ccnrtnec to one paint. In this ex-
as found above. ample paint (2) has been assumed as point where
drag is resisted. Those drag wires which would

,J><1Xt~X
be in compreSSion are assumed out of action.

~~3G---f1-36~--i
39.5 37.5 58.5
118.5 231 225 281.5 254
Fig. A2.38 798 1191 114\ I j 4189 j

Drag Truss Panel Point Loads Due to Air Drag

Load.
-13 893
~!~::o il~"'~J~I~
\~~~II·
ij~8' ~
1 ""I~~~ 162il 13
It was assumed that the air load components i(2) ·13,893 -11,933 -10,313 8513~ -413
in the drag direction were 6 lb./in. of wing
1908 '" Drag Reaction 798
acting r orwar-d,
Fig. A2.40
The distributed load of 6 lb./in. is re-
placed by concentrated loads at the panel points Fuselage Reactions
as shown in Fig. A2.39. Each panel point takes
one halt the distributed lead to the adjacent As a check on the work as well as to obtain
panel point, except for the two outboard panel reference loads on ~lselage from Wing structure,
pOints which are affected by the overhang tip the fuselage reactions will be checked against
portion. the exterr~llY applied air loads. Table AZ.2
gives the calculations in table form.
Thus the outboard panel paint concentration
Table A,2.2
of 254 lb. is Jetermined by taking moments about
(3) of the drag load outboard of (3) as follows:
.Point lIeliber Load V D s
P = 70.5 x 6 x 35.25/58.5 = 254 IJ.
2
III
Dn,
-13893
- 1908
- 726
0
0
-1908
_13870
0
Reactio..
To Si~pllfY the drag truss SOlution, the dra~ Jt2(React1on) - 1295 129'" 0 - 67
strut and drag wires in the inboard drag truss RB 1191 62 0 1190
panel have been modif~ed to ~ntersect at hinge • R.. (R.eaeUon) - 631 630 0 - 33
~/in.
,
j
s
6
Totals
..
F, 9333

4579
4205

2055

7520
"8
0

-1110 -
8290

4090

.00

Applied Air Loads:

V component = (3770 + 1295 + 1838 + 631)
.9986 = 7523 lb. (error 3 lb.
D component = -185 x 6 =
1110 lb. (er-r-or e 0)
S component = -(3770 + 1295 + 1838 + 631)
.0523 = 394 lb. (error 6 lb.
 A2.18 EQUILIBRIUM OF FORCE SYSTEMS. TRUSS STRUCTURES.

The wing bea~ ~ue to the distribu~ed l~~t Lengths k Directional Components of Cabane Struts
al~ loads acting upon them, are also s~bj~cted
to bending loads in add! t t on to the axf.a ; Loads . Member Sym. V D I s L VIL DIL 'IL
The wing be~s thus act as bea~-cJll~s. ~he Front Cabane CF 30 10 12 7 1, 4 1 . 59 1 .721 .240 .648
sUbject of beaT.-col~lh~ action is treated in Diagonal Caba,ne CD 30 30 i27150.17 .597 .597 .538
another chapter of this book. Strl.\t

Rear Cabane CR 23.5 6 27 I 40.42 .731 .1485 .668

If the ~lng 1s covered with
skin ~etal Strut
instead of fabriC, the drag truss can be omitted L ;./ vO: + D' + "
since the top ~~d batto~ skin act as webs of a
beam which has the front and rear be~ as its
flange ~embers. T~e wing is then considered as The air loads on the outer ~anel are taken
a box beam subjected to combined bending ~~d identical to those in exa~ple ;roblem 1. Like-
axial loading. wiSe the dihedral and direction ~f the lift
struts SF and SR have been made the sa~e as in
Exam~le Problem 11. 3-8ectlon Externa:ly Braced exauple ~roblem 1. Therefore the analysis :or
l,.jing. the loadS in the outer p~~el Gr~g and :i~t tr~s
trusses is identical to that in )roblem 1. T~e
Fig. A2.41 shcws a high wing exterr4lly solution will be cont1~ued assumi~g the r~~,i~g
braced wing structure. The wing outer panel has 11ft load on center panel of 4.5 10./:n. and a
been made identical to the wing panel of exa~ple forNard drag load of 6 Ib./in.
prOblem 1. This outer p~~el attached to the cen
ter panel by Sir~le pin fittings at points (2) Sol~tion 0: Cente~ ?~nel
and (4). Placing pins at these points ~ake the
structure sta~lcally determinate, whereas if the Center Rear Beam
beams were made continuous trtlo~gh all 3 panels,
692# (630-;.62)",692#
the reactions of the lift and cabane struts on
the wing beams would be statically indeterminate w'" 14. 7%/in.
since we would have a 3-span continuous bea~
resting on settling supports due to str~t de-
formation. The fitting ?in at ~oints (2) and
(4) ~~~~~§§~~~~~~r~~(4)
~20 8' 90' '8
20-1
(4) can be made eccentric with the neut~al axis R a",1650if R a",1650.f
of the beams, hence very little is gained by Fig. A2. 42
making be&~ continuous for the purpose of de-
creasing the lateral beam bending mo~ents. For Fig. A2.42 shows the late~al loacs on the center
assembly, stowage and shipping it is convenient rear beam. The loads consist of the d1str:j~teG
to build such a wing in 3 portions. If a air load and the vertical component ot t~e ~or­
~ultiple bolt fitting is used as ~oints (2) and ces exerted by outer ?anel on center panel at
(4) to obtain a continuous beam, not much is pin point (4). From Table A2.2 of exarr.ple ;rob-
gained because the design requirements of the lem 1, this resultant V reaction eq~ls 630 +
various governmental agencies speCify ~hat the 62 = 692 lb.
wing beams must also be analyzed on the as-
sumption that a multiple bolt tltting provides The vertical component of the cabane re-
only 50 percent of the full continuity. action at joint (8) equals one hal: the tetal
1', be~ load due to symmetry of 1cad1~g or 55 x
14.74 + 692 = 1650 lb.
(8)
,"
"'.=
\4) J...
Solution of force system at Joint 8
36"
1650
1650

C~ D1 ~D
/L:.. S
CR C

ZV - 1650 - .7)1 CR = a
(1)(3)
whence

CR = 1650/.731 = 2250 lb. (tens~on)

(5) (6) zs = - eRE - 2250 x .565 = 0

Fig. A2. 41
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A2.19

'whence (4) are taken fram Tabl~ A2.2 of problem 1. The

drag loac of 336 10. at (8) is ~~e to the rear
eRE 1510 lJ. (compress ion) cabane strut, as 1s likewise the bea~ ax1al load
of - 1510 at (8). The axfa I beam load of
ZD = De - 2260 X .1485 =0 - 2281 lb. at (7) 1s cue to ~eact1on of ~rc~t
caba~e truss. The ,a~el ,oint loads are d~e to
whence the given running drag lead of 6 lb./in. acti~6
forward.
De = 336 lb. drag tr~ss reaction
The reaction which holds all these drag
Center rront Beam truss loads in equilibrium is supplied by the
cabane truss at pOint (7) since the ~ront and
568# (1294 - 726) ::: 568* ciagonal cabane struts intersect to form a rigid
w'" 30.26 iii/in.
triangle. ~hus the drag reaction R equals one
(Ref. , Table A2. 2) half the total drag loads or 2634 lb.
fo- 20 71 90 Solving the truss for the loading of Fig.
;
R.,; 2535. A2.44 we obtain the member axial loads of Fig.
Fig. A2.43 A2.45.

rig. A2.43 shows the V loads on the center front 1157 568 568
beaw and the resulting V component of the cabane
reaction at joint (7).
:;:,
Solution of force syste~ at Joint 7
-17308 -1730B -1502

~
2535
(11)
CFB 7 Fig. A2.45

Co VS Plane LoadS in Cabane struts Due to Drag Reacticn at

CF Point 7

ZV = 2535 - .721 CF - .597 CD : 0 263~

zs ' - CFB - .648 CF - .538 Co , 0
ZD .240 CF + .597 CD =0
kF "'.
v-n Plane CD

solVing the three equations, we obtain

ZD = 2634 - .240 CF' + .597 CD =0
eFB = - 2281 (compression)
Zv =- .721 CF - .~97 Co =0
C:<, = 2635
Co , 1058
Solving for CF and CD, we obtain

solution for Loads 1n Drag Tr~ss ~embers

CF - - 2740 lb. (compression)
Co = 3310 (tenSion)
Fig. A2.44 shows all the loads applied to
the center p~~el drag truss. The Sand D re- adding these loads to those previously calc~­
actions fry,n the outer ~anel at joints (2) and lated for 11tt loads:

Panel point 336 CF , - 2740 + 2635 , - 105

I CD , ,
Drag Load 33 6
1
1058 + 3310 4368 lb.
for 6-#/in.-") J 270 t 60 ,
eR
f ~5~
1157 i6
1~Jl95 ~1 15.. 1157
fusela~e
2250 It.

Reactions
-~I-/1\I~ As a c~eck on the ~ork the ~~selage re-

13m.~/L\J~7
actions will be checked against t:'e applied
loadS. ~ajle A2.3 ~i7es the '1, D a~d S com-
~on2nts of t~e ~~sela~9 rec~t10~S.
t I 122IT 2.rr- (211
190a' n,
26341 Fig. A2.. 44
'R,
2634#
1190a

I !
,-
 A2.20 EQUILIBRIUM OF FORCE SYSTEMS. TRUSS STRUCTURES.

Table A2.3
36
Poiut lIeaber c=, V D S •
sl
9 Protlt 5tr\,\t
C.
_105
-" -'5 _68

,. Rear Strut CR
D1a. Strut CD
2260
4368
1650
2610
335
_2610
1510
2356 72
I, 24
e Front
S.
L1~t Strut 9333 4205 798 8290 i
E
e Rear Lift Strut 4579 2055 o 4090

Total.
'. 10444 _1502 16178
F
G
Applied Air Loads
Fig. A2.46
V component = 7523 (oute::- ~anel) + 65 x 45
= 10448 (check)
D component = - 1110 (outer pane L) 65 x -
6 -- 1500 (er::-or 2 lb.)

The total side load on a vertical plane trIU

centerline of airplane should equal the S com-
ponent of the applied loads. ~he applied side o R S ;g
":::!l
loads = - 39~ lb. (see problem 1). The air load
T / Fig.46a 1 I
on center panel is ver~lcal and thus has zero S
component.

From Table A2.3 for fusela~e reactions

i
36"
(lVTC I I --
\
\
1-
1

\1
,

T
/
I

-
\

'\

'c!
/
l.s

II;';
*'1
"'
I

I~Ac,
)'"1/ •
have ZS = 16178. From Fig. A2.45 the load in
L V / \/ (Q~
the front beam at b of airplane equals - 17308
and 568 for rear beam. The horizontal component ABC
ABC
D
D
E
E
,=
~
at the diagonal drag strut at joints 11 equals
216 x 45/57.6 = 169 lb.

Then total S components = 16178 - 17308 ~

~ 1/I~l2Is:vt;>,,1
N L J
~ 7 panels @ 12" '" 84"
G
- _-----J
F
c>;
I
r--
36"
----..,
568 + 169 = - 393 lb. which checks the side
Fig.46b Fig.46c
component ot the applied air loads.

Example Problem 12. Single SDaT Truss Plus SOLUTION:

Torsional Truss System.
In small wings or control surfaces, fabric
is often used as the surface covering. Since
the fabric cannot provide reliable torsional
reSistance, internal structure must be of suc~
design as to ~rovide torsicnal strength. A
single spar plus a special type of truss system
is often used to give a satiSfactory structure.
Fig. A2.45 illustrates such a type of structure,
namely, a trussed Single spar AEFrJ plus a tri-
angular truss system bevNeen the spar and the
trailing edge OS. Fig. A2.46 (a, ~, c) shows
the three prOjections and dimensions. The air
load on the surface covering of the structure
is assumed to be 0.5 Ib./in. a i~tensity at spar
line and then varying linearly to zero at the
trailing edge (See Fig. d).
The problem will be to dete~ine the axial
loads in all the members of the structure. It
will be assumed that all members are 2 force
illembers as 1s usually dcne in finding the
pr1ma~J loads in t~ssed str~ctures.
The total air load on the str~cture equals
the average intenSity per squa~e inch ~1mes :r.e
surface area or (0.5)(.5)(36 x 94-) = 756 lb. In
order to solve a truss system by a method of
joints the distributed lead must be replaced by
an eqUivalent load system acting a~ the Joints
of the structure. Referring to Fig. (d), :he
total air load on a strip 1 l~:h wide and 36
inches long is 36(0.5)/2 = 9 lb. and its 0'5.
or resultant location is 12 inches from line AE.
In i":g. 46e. this z-esu'l tarrt load of 9 ~b./in. is
imagined as acting on an ima~inary bea~ located
along the lIne 1-1. This rur~lr.g load a~~llec
a:ong this line is now replaced jy an equivalent
force system acting at jOlr.ts OP~}BED8CA. The
results 0: :hls joint distrlbution are shown by
the jOint loads in F~g. A2.46. !o illustrate
how these joint loads were obtained, t~e calcu-
latior.s for loads at ~oints SSDR will be given.
F~g. A2.48 shows a portien of :he struct~e
to be c ons t der-ed. For a run....n tng load 0: 9
Ib ./"!..n., along Ltne 1-1, r-eact t cns -vt Ll. be for

,.g 'JLI!
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A2.21

my :: (6 + 66 + 72 + 72 + 36 )36 - 36 Oz :: 0

whence Oz = 252 lb. acting down as assumed.

To find Oy take moments about z axis through

point (A).
mz = 0 + 36 Oy = 0, Oy ~ 0

To find Ay take moments about x axis 'thr-ough

Fig. A2. 48 point N. The moment of the air loads was pre-
viously calculated as - 31752, hence,
si~ple beams re~ting at ,oints 2, 3, 4, 5, etc. ZMx = - 31752 + 9 Ay = 0, whence, Ay =
The distance bet~een 2-3 is 8 inches. The total 3528 lb.
load on this distance is 8 x 9 = 72 lb. One
half or 36 lb. goes to paint (2) and the other To find Ny take l.Fy =a
half to paint (3). The 36 lb. at (2) is then
replaced by an equivalent force system at E and ZFy =3528 - Ny =0, hence Ny ~ 3528 lb.
8 or (36)/3 = 12 lb. to 8 and (36)(2/3) =
24 to
To find Nz take l.Fz 0 =
E. The distance between points (3) and (4) is
8 inches and the load is 8 x 9 = 72 lb. One l.Fz :: - 252 + 756 - Nz :: 0, hence Nz =
half of this or 36 goes to pOint (3) and this 504 lb.
added to the previous 36 gives 72 lb. at (3).
The load of 72 1s then replaced by an equivalent The reactions are all recorded on Fig.
force system at Sand D, or (72)/3 = 24 lb. to A2.46.
Sand (72)(2/3) ~48 to D. The final load at 8
1s therefore 24 + 12 = 36 lb. as shown in Fig. Solution of Truss Member Loads
A 2.46. Due to symmet~J of the triangle CRD,
one half of the total load on the distance CD For simplicity, the load system on the
goes to points (4) and (5) or (24 x 9)/2 = 108 structure will be considered separately as two
lb .. The distribution to D is therefore (108) load systems. One system will include only
(2/3)= 72 and (108)/3 = 36 to R. Adding 72 those loads acting along the line AE and the
to the previous load of 48 at D gives a total second load system will be remaining loads
load at D = 120 lb. as shown in Fig. A2.46. which act along line as. Since no bending
The 108 lb. at po tnt (5) also gives (108)/3 = moment can be resisted at JOint 0, the external
36 to R or a total of 72 lb. at R. The student load along spar AE will be reacted at A and N
should check tee distribution to other joints entirely or in other words, the spare alone
as shown in Fig. A2.46. resists the loads on line AE.

To check the equi,mlence of the derived F1g. A2.49 shows a d1a~ of this spar
joint load system wl~h ~he original air load with its joint external loading. The axial
system, the magnitude and ~oments of each loads oroduced by this loading are written on
system must be the same. Addi~g up the total the trUss members. (The S~Jdent should check
joint loads as shown in Fig. A2.46 gives a total these ~ember loads.)
of 756 lb. which checks the original air load.
The moment of the total air load about an x 72 144
, 144 ,
24

~i~-"23,,,3,,6"""==-i-'-""7<?"'---;-:='-:n,,",C"'""",,--"--i
~xis at left end of structure equals 756 x 42 ::
t
-11841 -1184 0 I
31752 in. lb. The ~oment of the joint load 2336
':» I "0 i~
system in Fig. A2.46 equals (66 x 12) + (72 x <\"~ :.?o~~
36) + (72 x 60) + (56 x 84) + l?4 (24 + 48) +
, IN

(120 x 72) + (24 x 84) = 31752 in. lb. or a
check. The moment of the total air load about
line AE equals 756 x 12 = 9072 in. lb. The
moment of the distributed joint loads equals
(6 + 66 + 72 + 72 + 36)36 = 9072 or a check.
Calculation of Reactions
The structure is supported by Single pin
fittings at pOi~ts A, Nand 0, with pin axes
parallel to x axis. It will be assumed that
the fitting at N takes off the spar load in
Z direction. Fig. A2.46 shows the reactions
Oy, 0z, Ay, Ny, Nz• To find Oz take moments
about y axis along spar AEFN.
BOO BOO 224 224 32
Fig. A2.49
TRI.ANGUUR TRUSS SYSTE:1
The load system along t~e trailing edge as
causes stresses in both the spar truss ~~d the
diagcr~l truss system. ~he support fitting at
paint a provides a reaction in the Z direction
but no reacting ~oment about the x axis. Since
the loads on the trailing edge 11e on a y axis
throu~~ 0, it is obvious that all these loads
flow to ?oint O. Since the bending strength of
the tra~llng edge member ~s negligible, the
A2 22 EQUILIBRIUM OF FORCE SYSTEMS TRUSS STRUCTURES

load of 36 lb. at Jol~t S in order to be trans- reaction on the s;ar but does produce ~ couple
fer~ed to pOint 0 through the dlago~al truss force on the spar <~ the v dir~ctlcn ~r.lch ~ro­
system must follow the path SDRC~PAO. In like duces compression in the ~0P chord of t~e s~ar
~nner the load of 72 at R to reach a must take truss and tension ~n the bottom c~ord.
the path RC~~PAO, etc.
Consider Joint R
Calculation of Loads in Dlaeonal Tr~ss Members:-
The load to be transferred to truss ~CJR
Member , y x L z/L y/L x/L is equal to the 72 lb. at R ~lus the 36 1J. at
S wn.tcb CCII'.es to jot nt R from tr-cs s DRG.
All Diagonal 4.5 12 36 38.2 .118 .314 .943
Truss Members Hence load ~n Re = (~2 + 36)0.5 x (1/.118)
4.5 0 36 37.5 .120 a .960
457 lb.
AO, NO
Tdhence RJ = 457, C~ = 457 and JQ =- 457
Consider Joint S Joint Q
The truss SEF carL~ot assist in
trla~Jlar
any portion of the 36 lb. lead at Load to be transferred to truss ~BL = 72 ~
transterrlr~
S because the reaction of this truss at SF 72 + 36 =
180 lb.
would put torsion on the spar and the spar has Hence load in ~ = (180 x G.5)(1/.118) =
no appreciable torsional resistance. - 762.

Considering Joint S ~nence QL =762, BP = 762, LP =- 762

X~~:-
as a free body and writing
the equilibrium equations; Joint P

l:Fx = -.943 DS -.943 GS =G G !Z

Load = 180 + 66 =246
.....hence, DS ::: - GS Load in PA = (246 x O.5)(l!.l18) =- lG40
ZFz = 305 + .118 OS - .U8 GS = 0 Whence PN = 104-0
Consider JoInt (A) as a ~ree :ody.
Subt. DS = - GS and solving for OS, gives
as = 159 lb. (tension), DS = -159 (compression) ZF x = - 1040 x .943 + .960 AO = 0, AO =
lG22 lb.
Consider Joint D In like manner, considering Joint ~J gi'l8S ~O
= - 2G22 lb.
Couple Force Reactions on Spar

As pointed out ?reviously, t~e diagonal

L. y torsion truss produces a couple reaction on the
spar in the y direction. The ~gnitude of the
torce of this couple equals the y component ot
Let Ty and Tz be reactions ot diagonal the load in the diagonal truss ~embers ~eeting
truss system on spar truss at Joint D. at a spar jOint. Let Ty equal this reaction
load on the spar.
ZFx =- 159 x .943 + .943 OR = OJ hence DR =
159 lb. At Joint c:-
ZFz = - 159 x .118 + 159 x .118 - Tz = a Ty =- (457 + 457) .314': - 28'7 1::'.

whence Tz =
0, which ~eans the diagonal Likewise at Joint J, Ty = 287
truss produces no Z reaction or shear load on
spar truss at D. At Joint s:-
ZFy =- .314 x 159 - .314 x 159 - Ty =0 Ty =- (762 + 762) .314 =- 479
, T
whence Ty =- 100 lb. Likewise at Joint c, y =479
If joint G is investigated in the same manner, At Joint A:-
the results will show that Tz a and Ty 100.= = m
'y =- (1040 x .314) =- 326
The results at joint D shows that the rear Likewise at Joint N, Ty = 326
diagor~l truss system produces no shear lead
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
These reactions of the torsion truss upon
the spar truss are shown in Fig. A2.50. The
loads in the spar truss illembers due to this
loading are written adjacent ~o each truss
~ember. Adding these illember loads to the loads
in Fig. A2.49, we obtain the final spar truss
~ember loads as shown in Fig. A2.5l.
Fig.A2.51
If we add the reactions in Figs. A2.50 and
AZ.51 J we obtain 3528 and 504 which check the
reactions obtained in Fig. A2.46.
A2. 13 Landing Gear structure
The airplane 1s both a landborne and air-
borne vent c l,e, and thus a means ot operating
the airplane on the ground must be prOVided
which means wheels and brakas. Furtharmore,
provision must be made to control the impact
forces involved in landing or in taxiing over
rough gro~~d. This reqUirement requires a
speCial energy absorption unit in the landing
gear beyond that energy absor?tion prOVided by
the tires. The landing gear thus includes a
so-called shock s~rut cammonly referred to as
~~ oleo strut J which is a member composed ot
~HO telescoping cylinders. 'dhen the strut is
compressed J oil inside the air tight cylinders
Is forced through an orifice from one cylinder
to the other and the energy due to the landing
impact Is absorbed by ~he work done in forCing
this oil through the critice. The orifice can
be so designed as ~o provide practically a
uniform resistance over the displacement or
travel of the olec strut.
An airplane can land safely with the air-
plane in various attitudes at the instant at
ground contact. Fig. A2.52 illustrates ~he
three altitudeS of the airplane that are
specified by the government aviation agencies
for deSign of landing gear. In addition to
these symmetrical unbraked loadings, special
loadingsJ such as a braked condition, landing
on one wheel condition, side load on wheel, etc.
are reqUired. In other words, a landir~ gear
can be subjected to a cor~iderable number of
different loadings ~~der the variouS lar.ding
conditions that are encountered in the normal
operation of an airplane.
A2.23
Level Landing with Inclined Reactions
Level Landing with Nose Wheel Just
Clear of Ground
Tail Down Landing
Fig. A2.52
Fig. AZ.53 shows photographs at typical
main gear units and Fig. A2.54 tor nose Wheel
gear units.
The successful deSign of landing gear tor
present day aircraft is no doubt one or the most
difficult prOblems which is encountered In the
structural layout and strength design of air-
craft. In general, the gear tor aerodynamic
efficiency must be retracted into the interior
of the wingJ nacelle or fuselage J thus a re-
liable J safe retracting and lowering mechanism
system is necessary. The wheels must be braked
and the nose wheel made steerable. The landing
gear is SUbjected to relatively large 10ads J
whose magnitudes are several times the gross
weight ot the airplane and these large loadS
must be carried into the supporting wing or
tuselage structure. Since the weIght of land-
ing gear may amount to around 6 percent ot the
weight at the airplane it Is evident that high
strength/weight ratio is a paramount design
requirement ot landing gear, as inefticlen~
structural arrangement and conservative stress
ana.Lyst a can add many unnecessary pounds of
weIght to the airplane and thus decrease the
payor useful load.
A2. 14 Example Problems of Calculating Reactions and
Loads on Members oC Landing Gear Units
In its Simplest fonn, a landing gear could
consist of a single oleo strut actlr~ as a
cantilever beam wIth its fIxed end being the
upper end which would be rigidly fastened to
the supporting structure. The lower cylinder
of the oleo strut carrieS an axle at its lower
A2.24 EQUILIBRIUM OF FORCE SYSTEMS. TRUSS STRUCTURES.

McDonnell Aircraft
(Military Airplane)

Douglas DC-7 Air Transport

Douglas DC-8
Jet Airliner

FIg. A2. S3 Main Landing Gear Illustrations (One Side)

 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A2.25

Beechcraft Twin Bonanza

Piper-Apache
Piper Tri-Pacer

Navy F4-J
North American Aviation Co.

Douglas DC· 7 Air Transport

Fig. A2. 54 Nose Wheel Gear Installations
A2.26 EQUILIBRIUM OF FORCE SYSTEMS. TRUSS STRUCTURES.

end tor attaching the wheel and t~re. ~his SOUJ'I'I:JN

cantilever beam is subjected to bending in two
directions, torsion and also axial loads. Since The supporting ~eactlons upon ~he ~ea~ at
the gear is usually made retractable, it is paints F, H, and G will oe calculated as a
difficult to design a single fitting unit at beg~nn1ng step. T~ere are six urJcnowns, na~ely
the upper end of the oleo strut that will FS, F'V, FO, Hy, HO and G (See Fig. AZ.56). Wlt~
resist this combination of forces and still 6 equations of static equilibrium available tor
per.nlt movement tor a simple retracting mechan- a space ~orce sy~tem, the reactions can be found
ism. Furthermore, it would be difficult to by statics. Referring to Fig. AZ.56:-
provide carry-through supporting wing or fuse-
lage structure tor such large concentrated To find FS take ZS =0
load systems.
Fs + a = 0, hence FS =a
Thus to decrease the magnitude of the
bending moments and also the bending fleXibility To find react10n Gv take moments about an S
ot the cantilever strut and also to simplify axis through paints F, H.
the retracting problem and the carry-through
structural problem, it is customary to add one
l:l1s =3119 x
50 - 24 Gv 0 =
or two braces to the oleo strut. In general, Whence, GV = 6500 lb. with sense as assumed.
effort is made to make the landing gear (The wheel load of 15000 lb. has been resolved
structure statically determinate by USing into V and D components as indicated in Fig.
speCially designed fittings at member ends or A2.55) .
at support paints in order to establish the With GV known, the reaction G equals (6500)
torce characteristics of direction and paint at (31.8/24) =
8610 lb. and similarly ~he compon-
application. ent GD =
(6500)(21/24) 5690 lb. =
TWo example problem solutions will be pre-
sented, one dealing with a gear with a Single
1 , 1-.-
6

J6" !
wheel and the other with a gear involving two ;0 cJi:~~ <;
wheels.
. "
Example Problem 13

Fig. A2.55 shows the projections of the

landing gear configuration on the VS and VD
l II
f
planes. Fig. A2.56 is a space dimensional
diagram. In landing gear analysis it is cammon
to use V, D and S as reference axes instead of
the symbols Z, X and Y. This gear l~it 1s
assumed as representing one side of the main
gear on a tricycle type of landing gear system.
The loading assumed corresponds to a condition
of nose wheel up or tail down. (See lower
sketch or Fig. A2 .52) • The design load on the
wheel 1s vertical and its magnitUde for this
problem is 15000 lb.

The gear unit is attached to the supporting

structure at points F, H and G. Retraction of
the gear is obtained by rotating gear rearward
and upward about axis 'thr-ough F' and H. The
fittings at F and H are deSigned to resist ~o
bending moment hence reactions at F and Hare
UIL~own in magnitude and direction. Instead of
USing the reaction and an angle as unknowns,
the resultant reaction is replaced by its Y and
D components as shown in Fig. A2.56. The re-
action at G is unknown in magnitude only Since 10063 6500
the pin fitting at each end of member GC fixes
the direction and line of action of the reaction F
1~3004
!--5690
G
at G. For convenience in calculations, the lll09
reaction G Is replaced by its camponents GV and
GD' For a stde load on the landing gear, the \'D I 433
reaction In the S direction is taken off at a
paint F by a special designed unit. Fig. A2. 56 Flg. A2. 57
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A2.1.7

the structure as a whole by taking moments about

D and V axes through point A.
lEy 21

> ED~'b
-r-E = - 10063 x 14 + 6500 x 8 + 11109 x 8
,, I"u
I ...
,,>'
=- 140882 + 52000 + 88882 O(check) =
24" 1 <> = 5690 x 8 - 433 x 8 - 3004 x 14
I I 31 = 45520 - 3464 - 42056 = 0 (cheek)
~ _ ·C--CGO=31 a e G
I "I"'- •
The next step in the solution will be the
26" I L. calculation of the forces on the oleo strut
unit. Fig. A2.58 shows a free body at the oleo-
lel-3u9* 0
strut-axle unit. The brace members EI and CG
are two force members due to the pin at each
end, and thus magnitude is the only unknown re-
action characteristic at points Band C. The
fitting at point E between the oleo strut and
the top cross member FH is designed in such a
TE=24952 manner as to resist torsional moments about the
Fv
I--12"-
I By oleo strut axis and to provide 0, V and S force
reactions but no moment reactions about 0 and S
'.JI
FS-S :E~s=3920LI~- axes. The unknowns are therefore BI, CG, Es,
~ " •
ro"rEV=13332
14"--l...f
:;:@-BIS=3920

BIV=7840
EV, ED and TE or a total of 6 and therefore
statically determinate. The torsional moment
TE 1s represented in Fig. A2.S8 by a vector
with a double arrow. The vector direction
represents the moment axis and the sense of

FS--S
to
1@)_3920
TE 39,20. HO
----~
rotation of the moment is given by the right
hand rule, namely, with the thumb of the right
F. H
hand painting in the same direction as the
arrows, the curled fingers give the sense of
i . E~.2571 I rotation.
~. +- 16" ---t Fig.A2.59
T~ fInd the resisting torsional moment T~
take moments about V axis through E. .....
To find Fv, take moments about a D axis
through point H. ZME(V) = - 3119 x 8 + TE = 0, hence TE
= 24952 in. lb.
l:l'lH(D) = 16 14672 x 8 - 22 FV = 0
Gv +
= 16 x 6500 + 14672 x 8 - 22 Fv = 0 To find CG take moments about Saxis
tnr ougn E.
:~ence FV = 10063 lb. with sense as as-
sumed. ZME(S) ~ 3119 x 50 - (24/31.8) CG x 3 -
24 (21/31.8) CG ~ 0
To find Ho, take moments about V axis Tdhence, CG = 8610 lb.
tihr-ough F.
ZMF(V) =- 6 On - 22 Hn + 3119 x 14 = 0 This checks the value previously obtained
when the reaction at G was found to be 8610.
- - 6 x 5690 - 22 HD + 3119 x 14
=0 The 0 and V components or co thus equal,
~bence, HD = 433 lb. CGD = 8610 (21/31.8) ~ 5690 lb.

~o find FD, take ZD = 0

CGV = 8610 (24/31.8) 6500 lb. =
!D =- ?O + HO + GO - 3119 =0 To find load in brace strut 8I, take moments
=- FD + 433 + 5690 - 3119 = 0 about 0 axis through po irrt z.

Whence, FD =3004 lb. ~E(D) = - 14672 x 8 + 3 (BI) 22/24.6 +

24 (BI) 11/24.6 0 =
To find take ZV = 0
HV T~ence, BI = 8775 lb.
'i.y. = - FV + Gy - Hy + 14672 = 0 and thUS, SlV = (8775)(22/24.6) = 7840 lb.
= - 10063 + 6500 - HV + 14672 = 0 BIS =
(6775)(11/24.6) = 3920 lb.
~nence, Hv =11109 lb.
To tind Es take ZS = 0
Fig. A2.57 summarizes the reactions as found.
The results wl~1 be checked for equl11brlum of 'i.S = Es - 3920 =0, hence Es =3920

-~"'-~,~.~

.~'-"
A2.28 EQUILIBRIUM OF FORCE SYSTEMS,
To find Eo take ZD =0
ZD = 5690 - 3119 - ED = 0,
To find EV take LV = 0
LV = - EV + 14672 - 7840 + 6500 : 0, hence
Ev = 13332 lb.
Fig. AZ.59 shows a tree body of the top
member F1i. The unknowns are Fy, FO, FS' Hyand
HD. The loads or reactions as found tram the
analysiS of the oleo strut unit are also re-
corded on the figure. The equations of
equilibrium for this tree bOdy are:-
1:3 =0 3920 + 3920 + FS
l:l'lF(D) = 22 HV - 3920 x 2
13332x6=0
Whence, HV =
11110 lb. This check value
obtained previously, and therefore is a check
on our work.
lJ'lF(V) = 24952 - 2571 x 5 - 22 HD
whence, Hn = 433 lb.
zv - - Fv ... 13332 ... 7840 - 11110
whence, FV = 10063
ZD - - Fn + 2571 ... 433 = 0
whence, FD =3004
Thus working through the tree bodies or
the oleo strut and the top member FH, we come
out with same reactions at F and H as obtained
when finding these reactions by equilibrium
equation for the entire landing gear.
The strength deSign of the oleo strut unit
and the top member FH could now be carried out
because with all loads and reactions on each
member known, axial, bending and torsional
stresses could now be round.
The loads on the brace struts CO and B1
are axial, namely, 8610 lb. tension ana 8775
lb. compression respectively, and thus need no
further calculation to obtain design stresses.
TORQUE LINK
The oleo strut consists of ~NO telescoping
tubes and some means ~ust be provided to trans-
mit torslor.al moment becNeen the two tubes and
still permit the lower cylinder to move upward
into the upper cylinder. The most cammon way
of providing this torque transfer 1s to use a
double-cantilever-nut cracker type ot structure.
Fig. AZ.60 illustrates ~ow such a torque length
could be applied to the oleo strut in our
prOblem.
TRUSS STRUCTURES.
hence ED = 2571
Lower
Cylinder
Fig.A2.60
Torque
(l )~;~O::::iAxle
= 0, or FS =0 LinJ< ~
7840 x 20 -
=0 The torque to be transferred in our ~rob­
lem is 24952 in. lb.
= 0
The reaction R~ between the bvo units 0:
the torque link at point (2), see Fig. A2.50,
thus equals 24952/9 =2773 lb.
The reactions R 3 at the base of the link at
lb. pOint (3) = 2773 x 8.5/2.75 = 8560 lb. ""th
these reactions %nown, the strength design of
the link units and the connections could be
made.
Example Problem 14
The landing gear as illustrated in Fig.
AZ.61 is representative of a main landing gear
which could be attachea to the under 51de of a
wing and retract forward and upward about line
AB into a space provided by the lower portion
of the power plant nacelle structure. The oleo
strut OE has a Sliding attachment at E, which
prevents any vertical load to be taken by
member AB at E. However, the fitting at E does
transfer shear and torque reactions between the
oleo strut and member AB. The brace struts
GD, FU and CD are pinned at each end and will
be assumed as 2 force members.
An airplane level landing condition with
unsymmetrical wheel loadii~ has been assumed as
shown in Fig. A2.6l.
80LUTION
The gear is attached to supporting struc-
ture at points A, Band C. The reactions at
these pOints will be calculated :irst, treating
the entire gear as a free body. Fig. A2.62
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A2.29

14- 19" ---'- 19" --: ZMA(D) =- 60000 x 9 - 40000 x 29 - 66666

~ r17"----17.. ~!B x 19 + 38 BV 0 =
, , whence, BV = 78070 lb.
"
"" To find AV' take ZV ~ 0
28"
I
""'
B~ace
4" r,

"
zv =- 78070 + 60000 + 40000 + 66666 - VA
~oDo Fig. A2. 61
=a
whence, Ay = 88596 lb.
I tV
L. 5
To find BD take moments about Y axis through
-0
po trrt (A).
36"
ZMA(V) =57142 x 19 - 15000 x 9 - 10000 x
15000
Whee11 __
I 29-38BD a=
=
whence, So 17386 lb.
10000 I

ZD = 0
Wheel
40000 To find An take
6000
ZD =-57142 + 15000 + 10000 + 17386 +
AD = a
whence An 14756 lb. =
To check the results take moments about Y
and D axes through pOint o.
ZMo(V) =5 x 10000 + 14756 x 19 - 17386 x
19 = a (check)
&~O(D) = 20000 x 10 - 88596 x 19 + 78070 x
19 =a (check)

REACTT ONS ON OLEO STRUT OE

Fig. A2.63 shows a tree body of the oleo-

64"
strut OE. The loads applied to the wheels at

ITE
r 17"'-1- i7"'
ITE
-1
E
.:l;G_ - -
Fig.A2.62
EO-
I 00
E _ES-J:-
' r
Yi
~

'\! 9
~

shows a space diagram with loads and reactions. 571~~ 28"

~OF5
" 28"
L
~t
0
005-'
The reactions at A, Band C have been replaced I 1---"
by their V and 0 components.
4~~ 41/4-1 1-4 1/4
To find reaction Cy take moments about an
S axis through paints AB.
L 36"
VL s Fig. A2. 63
0
ZXAB =- (15000
Cv = 66666
T~ence
+ 10000) 64
l~. ~ith
+ 24 Cv = 0

sense as
"0" -l
25000 M'o"
assumed in Fig. A2.62. 100,000 lb. 200,OO@ -, 100,000 lb.

c
t 50000"lb. t50,OOO"lb.
The reaction at C must rave
a line of action along the line ~~,
CD since ~ember CD 1s pinned at 28G the axle centerllnes have been transferred to
each end, thus the crag compon- 24 0 point (0). Thus the total V load at (0) equals
ent and the load in the strut 60000 + 40000 = 100000 and the total D load
2D tollow as a matter of geometry. Hence, equals 15000 + 10000 = 25000. The moment of
=
Co ~ 66666 (24/25) 57142 lb. these forces about V and D axes through (0) are
CD = 66666 (36.93/28) =
87900 lb. tension Mo(V) = (15000 - 10000) 10 50000 in.lb. and =
Mo(D) = (60000 - 40000) 10 200000 in.lb. =
To find BV take moments about a drag axis These ~oments are indicated in Fig. A2.63 by
through point (A). the vectors with double arrows. The sense of
the ~cment 1s determined by the right hand thumb
A2.30 EQUILIBRIUM OF FORCE SYSTEMS. TRUSS STRUCTURES.

and finger r~le. AD BD

~ 19 -+-19 ~
The fitting at point E is Ges~gned to I &50000 I
resist a moment about V axis or a torsional ""-A IV EB
moment on the oleo strut. It also can ~roVlde
shear reactions Es and ED but no bend1ng
I
,
32143
resistance about S or D axes.

The unknowns are the forces Es, ED' OF,

AV~~17
A
t- li~BV
_,. _ B
DG and the moment TE' G.. 4164 47023, .IF
I I I!
To tind TE take moments about axis OE. 89215 \ ' / 77451 Fig. A2. 64

~~E = - 50000 + TE = 0, whence T~ =

500CO In .u ,
ZMA(D) =- 89215 x 2 - 77451 x 36 - 38 BV
To find ES take ~oments about D axis =a
through point D. Whence, By = 78070 lb.
ZMD(D) =200000 - 28 Es = 0, whence
To r md AV take ZV =0
sa = 7143 lb.
ZV = 89215 + 77451 - 78070 - Ay =0
To tind torce DFy take moments about D Whence, AV = 88596 lb.
axis through point G.
lMo(D) =200000 - 100000 x 17 - 66666 x To tind BD take ~oments Y axis t~rougb A.
17 + 34 DFy :: 0
=77451 lb. ZMV(D) = 50000 + 32143 x 19 - 38 aD =0
whe~ce, DFV Nhence, SD = 17366 :0.
Then DFS =77451 (17/28)
17 To find AD take ZD = 0
= 47023 lb.
and DF =77451 (32.72/28) 28~
~:,\l>
ZD = 17388 - 32143 AD =O. or AD =
= 90503 lb.
14757 lb.

To rtnc IXlV take ZV =0 These tour reactions check the reactions

obtained originally ~hen gear was treated as a
Zy =
100000 - 77451 + 66666 - DGV =0, tree body, thus glvln~ a numerical check on the
or lXly = 89215 calculations.
Then, lXle =89215 (17/28) =54164 lb. the forces on each ?art of the gear
~lth
DG = 89215 (32.73/2B) = 104190 lb. ~~own, the parts could be designed for strength
and rigidity. The oleo strut would need a
To r mc ED take moments about Saxis torsion link as disc~ssed in example problem 13
through point D. and Fig. A2.60.
l:!1D(S) =- 25000 x 35 + 28 ED = 0,
ED =32143 lb.
A2.15 Problems

(I) For the structures numbered 1 to 10 deter-

The results will be checked tor static ~ne whether structure 15 statically deter-
equilibrium of str~t. Take moments about D minate with respect to external reactions
axis through point (0). and internal stresses.
l:!1o(D) = 200000 + 54164 x 38 - 47023 x 10 20
=
fIB
38 - 7143 x 64 200000 + 1949904 Pin Pin Pin
- 1892B28 - 457150 a (check) = 'A ~ ],8 J
A ~D AB
E in"
l:!1o(S) =32143 x 64 - 57142 x 36 =O(check) (1)
:;-,0
10 (2) 1- 10
(3)

REACTIONS ON TOP MEMBER AS 10

*
Fig. AZ.S4 shows a tree body at member AB ~-10 I I (6)
~
10tr: (5)
with the known applied forces as found from I A (4)
A ; t B
the previous reactions on the oleo strut. A 10 10
A
'*'
B

The unknowns are AD, BD, A~ and By. To

tind By take moments about D axIs tr~ough A.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A2.31

,
1000*
I

(2) Find the horizontal and vertical components J'

of the reactionS on the structures illus- Wing
.
-
trated in Figs. 11 to 15. +V
..J+V 52
,
L
tt4'~0 +D Front vtew +8
i 1-8,,1:00.
10"'" ~10-4
10
Slde vtew
T Pin ~!~ 10" B Flg.20
21:' (11) Pbi]: Bi"'!>c oj, l~O
1-16''-'1 - A
(I')
ho" ~O' (13) (6) Fig. 21 shows the wing structure of an ex-
ternally braced monoplane. Determine the
200f 500* axial loads in all members or the lift and
Cable k:c---r---,,,-i<'- -,..,..100/1 drag trusses for the following loads.
Pulley
(14)
#~';=f:d~I" I 'YO" Front beam lift load =
30 lb./in. (upward)
'O~' Rear beam lift load =
24 lb./in. (upward)
100
A
Wing drag load =
8 lb./in. acting art
PLANV!EW
(3) Find the axial loads in the members of the Wing Drag Truss
trJssed structures shown in Figs. 16 to 18 .

...J. jl0 r O
t10 (16)

15"~'
T 30"
10
12 4 (
24
124"1
....J..
36" -+- 36" -+- 36" ~ 45"------1

B B'
1'000 30 dihedr

~'~:£::
CF
CD
CR
...L _800 10 .i- D,D'
I l- I 1 30'....
I
1-30" -+
2000 (IB) I . I I
15"+15'..-.l
:~tL Fuselage
% A.A'
Flg.21
(4) Determine the axial loads in the :nembers
of the structure in Fig. 19. The member-a (7 )
are pinned to supports at A, Band C.
I-- 5 Bays @ 30" --.j
rl C D
T ~-DragTrUSSLS
3I
At t t t l t t t t t f t t tt t 25 lb/in.

Pin Reactions
v
t D
" '" B "'0.
C
~"'.
B
at A, B, C '-D 60" Fuse-
.L lage
~v."'~
::,.... '5
fL
t.-s
L:
I
D

Flg. 19 1100/1 E \> -- ,' E

Flg. 22
--110'\--
(5 ) Fig. 20 shows a trl-pod trame for hoisting Fig. 22 shows a braced ~onoplane wing. For the
a propeller for asseT.bly on engine. Find given air loading, tind axial loads in 11tt and
the loads in the ~rame tor a load at 1000 drag trusS members. The drag reaction on drag
lb. on hoist. truss is taken oft at point A.
A2.32 EQUILIBRIUM OF FORCE SYSTEMS. TRUSS STRUCTURES.

(8 ) (9 )
The r i tt tng at In the
6"
,
....,.-_-,A B C
20"
points A and B ~or
the Landfng gear
structure in ?ig. 23
prOVides resistance
E F
t
L22"----';
C F
landing gear-
of :'1.>;.
the br-ace
member's
2~,

BC
I to VJ D and S :-8- and SF are
24" actions and moments B f,'IW r or-ce
-13000 v member's • The
lb. about D and V axes. '0" LD .::'itt::-:1g at E:
Fig. 23 Find the reactions at
Wheeo wheel wheel

~
CD~
10000 A and a and the load cr-evices ::-e-
10000 lb.
in member CD for . .J500' sistance tc
333 A V, 0 ar:.d S
given wheel loading. wheel 2
"113'
res.c1:1.ons but
15000 10000 only momentFig. 24
resistance
about 1J axis.
Find reactions at ~ and loads in :::embers SF and
BC under given wheel loading.

Cessna Aircraft Douglas DC-8 Jet Airliner

Nose Wheel Installation (Model 182) Main Landing Gear Unit
 CHAPTER A3
PROPERTIES OF SECTIONS· CENTROIDS
MOMENTS OF INERTIA ETC.

A3.1 Introduction In engineering calculations, Then

two terms, center of gravity and moment ot iner-
tia, are constantly being used. Thus, a brief Ix =J;tJdA, I y =/XQdA, I z =/radA
review of these terms Is in order. where IX! I y and I z are moments of inertia of the
area about the axes xx, yy and zz respectivelY.
A3.2 Centroids, Center of Gravity, The cen-
troid of a line, area, volume, or mass Is that
pOint at which the whole line, area, volume, or
mass may be conceived to be concentrated and have
the same moment with respect to an axis as When a
distributed in its true or natural way. This
general relationship can be expressed by the
principle of moments, as tollows:
Lines: - XL ZLx, hence x = ZLx = J xdL
L --L- Fig. A3.1

jc-eas r- XA Sax, hence X = Zax =fXdA A3.5

Polar Moment of InerUa In Fig. A3.1, the
A moment of inertia I z =fr~dA about the Z axis is
A
vcrueeee- iv = ZVx, hence i = ZVX '=JXdV referred to as the polar moment of inertia and
T --Y- can be defined as the moment of inertia of an
area with respect to a point in its surface.
Masses:- XI1 = zxm, hence X Zmx: = JXdM Since r~ = X~ + ya (Fig. A3.1)
M M Iz <Fir» + x ») dA = Ix + I y or; the po Lar- moment
If a geometrical figure 15 symmetrical with re- or inertia Is equal to the sum of the moments of
spect to a line or plane, the centroid of the inertia with respect to any two axes in the plane
tigure lIes in the given line or Plane. This is of the area, at right angles to each other and
obvious tram the fact that the moments of the passing thru the point at intersection of the po-
parts of the figure on the opposite sides of the lar axis with the plane.
line or plane are numerically equal but of op-
posite sign. It a figure is symmetrical to ~NO A3_5 Radius of Gyration The radius at gyration
lines or planes, the centroid of the figure lies of a solid is the distance tram the inertia axis
at the intersection or the two lines or the two to that point in the SOlid at whiCh, it its en-
planes, and likewise, if the figure has 3 planes tire mass could be concentrated, its moment of
of symmetry, the centroid lies at the intersec- inertia would remain the same.
tion of the 3 planes. ThUS,fr a dl1. =faM, where ~ is the radius at
gyration J
A3.3 Moment of Inertia The ten moment at in- Since, r"ciM = I, then I =f''' I1 or f =\ IT
ertia is applied in mechanics to a number at Vl1
mathematical expressions which represents sec-
By analogy, in the case ot an area,
ond aonents at areas, volumes and masses, such
as
fY'dA, f
r'dY, fr'd1'! etc.
I =f'A or =Ift e
A3.4 M01Ilent of Inertia of an Area As applied A3,.7 P&rallel Axis Theorem In Fig. A3.2 let I y
to an area, the term ~ament of inertia has no be the moment of inertia of the area referred to
phYSiCal sig~ificance, but represents a quantity the centroidal axis y-y, and let the moment or
entering into a large number of engineering inertia about axis y~y~ be required. y~y~ Is
prOblems or calculations. However, it may be parallel to yy. Consider the elementary area dA
~onsldered as a factor Which indicates the in- with distance x + d from y~y~.
r luence of tue area i t.se Lf in determining the
total rotating moment of uniformly varying for- Than, Iy~ = (d + x ) IlidA J
ces applied over an area. =JX'dA + 2d·JXdA + d fdA
Let Fig. A3.1 be any plane area referred to
three coordinate axes, ox, oy and oz; ox and oy
being the plane of area.
Let dA represent an elementary area, with
coordinates x, y, and r as shown.
A3.1
.3 2 CENTROIDS, CENTER OF GRAVITY, MOMENTS OF INERTIA

y TABLE 1
I .c:": Section Properties of Areas
d --1- x ....,
I >dA Rectangle Area = bd
fe. g. Fig. A3.2
Y =
d
2
I I1._1. = bd 3 / 12
I _
a a = Oct'
2dA.
3"
The first term,! x represents the ac-
ment of inertia of the body about its centroldal f ,-, = .289 d
axis y-y and will be given the symbol I. The f------------:.-------------l
second term 1s zero because~xdA 1s zero since Area = bh
TJ 1s the centroldal axis of the body. The 2
last tern, d2~dA = Ad 2
or, area of body tl~es Y = h/3
the square of the distance between axes yy and

,~3
11._1. = bh 3
Y1.Y1." ~ Triangle "V"
vO
Thus in general,
I=I+Ad a
1 ---r- - - I 1,;1_a = bh 3
This expression states that the amount ot 12
inertia of an area with respect to any axis in 2 1 _ 'l 2
3 =
bh'
the plane of the area 1s equal to the moment ot
inertia ot the area with respect to a parallel
3
f---- ---i
"4
centroldal axis plus the product of the area and = .236 h f ,-,
the square ot the distance bec#sen the two axes.r-----------------~-------------------------_1
Area = bh
Parallel Axis Theorem For Masses. If·instead of '2
area the mass at the body is considered, the
parallel axis can be written: x = 2/3 h
h~
I = I + Md a , where M reters to the mass at 11.-1. = 4,8
.3

the body.

A3-.7a hss Jloaents at Inertia The product of

the mass or a particle and the square at its Trapezoid = d(b
distance tram a line or plane is referred to as
the moment of inertia at the mass at the parti-
r- bl-j Area

'j = d(2b • be)

+ b 1.)
2

cle with respect to the line or plane. Hence,

3(1) + bl.)
I = ZMra. It the summation can be express-
ed by a definite integral, the expression may be = 0.3 (b 2 + 4bb , +b 2)
wr-L tten I =J r-e dl1 36 (b + b, ~

Moments of Inertia of Airplanes. In both ilying

and landing conditions the airpl~~e may be sub-
jected to a~~lar accelerations. To determine
the magnitude ot the accelerations as well as
the distribution and magnitude or the ~ss iner-
tia resisting torces, the moment or inertia ot
the airplane about the three coordinate axes is
generally required in mEL~ing a streSs ~~lysis
of a particular airplane.
The mass moments ot inertia or the airplane
about the coordinate X, Y and Z axes through the
center or graVity at the airplane can be expres-
sed as tollows:
Ix = Z/fY2 + Zwz 2 + Zl.I x
Iy = Zwx' • Zwz ll + Zl.I y
Iz = Zwx 2 • r,wy2 + Zl.I z
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A3 3
Elliptical
TABLE 1 - Continued Ring ~~l:
I---- -----j
Area := ted + za)
~ I '~f Ix := n (a 1 3b 1 - aa3b:a)
x-l -i~X 4:
y = d/2
bd 3 - a(d _ 2t)3 ex = V A}ea
Ix :=

12 W
20
deb +a)3 _
= (dt-t') (b'-bt)
zaec -

Ix - I y
and I:a- II see equa-
given for angle.
6ab:ae

Tl1
l::
a
~al1

x Circular
Fillet
Area ::: .215 a a
X ::: .223 a

~
u-sectton Xl ..L
-jtJI- a~r.o-
Area A ::: 2dt 1 + at
-
-r-
.x ==rr I ,>,

r:n.n
lIT:±[b1
¥
y ::: d a t 1 ~ .5t a a
A
~

1
I'

'I_it- Parabolic
Fillet
x ::: 3
4:
y = .3 Y.
Xl

r-b~=r 3

Area = nr'
Circle
= nr'
2--;-~-2 1 1- 1
T y ::: 2 y.
" 'i
.... -; la_Ii = 5 n r-
l\~l 4: TABLE 2
e1-l = r- Properties of SolidS
2
Ring Area ::: n (roll - rill) Solid Cire. Cyl.
Vol. - n r-'l-L,. (r - radius)
~ M = W (Total wt.)
.. '/~
',I
1 ~ ~ -----:..-.l- 1
"

.~
ro '
Jo. l .'
11. -1 ::: n (ro" - rl")
4:

,
)5:5'
'" .-
~"
'\ .'

11 -1
s
= ~.r.a
"2
I,~ = M [r a + (L f31])'4
Area _ n r "
Serm-crrcfe 2 Hollow Circ. Cyl. ~ Vol. =- IT L{r1 a - ra~

~ y::: .424r
\~ . 11. -1 ::: M(r1 a + r aa)/2

~~~lx 1a _ liI I1(r1 a + r a ll + La/3)/4

tx
t:x
v .1098 r<
= .264 r-
, q=outside
:::

For thin hollow eire. cyl.

1 _ - Mr a
radius 1 1
Area - rt (R a _ r ")
2 Vol. ::: abL, I1 ::: W
g
semt-cu-curar Ring y ::: 4 (R '1+ Rr + ra) •appr-ox. y:::2r
3n R+r IT 11 - 1
::: M (a:a + bll)/12
Ix:::.1098(R"-r")-.283Rll!"" (R-r) 1"1-'01 ::: MLa/12
R+ r 13 -3 :::ML
ll/3
I x(approx):::.3t(r +R)::S when ~+-----------------------I
2 r Vol. = (4nr'I/3
S o lid
Lr Sp h ere
i s sm8l1 Sphere
Iabout axis =
Ellipse Area ::: nab 2M r "1/5
I
Ix = -
n a sb Thin Hallow Sphere Vol. ::: 3 rt

~~
4-
II' j (r1 :J -ra 3 )
xf'x =a r 1 ,. radius of sphere

~~
'2 d1a.
=
2Mr, '
"""""3""
13.4 CENTROIDS, CENTER OF GRAVITY, MOMENTS OF INERTIA

TABLE 2 - Co~t~nued
I -- r-e t ~-
sIn 20
- (l-cosa 0) •i
NAx 2 4
Ring with Circular
Section Y M= w = mass per unit volume

J;tf'x y
g of body.
1:0: =mnaRaa[R' + (Sa a/4)J
a
I yy =! ml1 a Raa (4R + 3a )
2
1I
TABLE 4
Centroids ot ':'rap1zaidal Areas

Lc1JT
b I J.
TABLE 3 ll-y -!-x I-
Section Propert1es of Lines
Distal1Ce x Distance y
(t 1s small in comparison to radius) Ratio a
b
0.40992 0.5008
Circular Arc Area = 2 11 rt 1.01
0.4098-4. 0.5016
1.02
1. 03 0.40976 0.50240
11._1. = 11 rat 1.04 0.4988 0.5032
1.05 0.40960 0.5040
I
pol ar = 2 11 r"t 1 06 0.40952 I 0.5048
0.5056
1.07 0.40944
1.08 0.40936 0.5064
E'x = .7(f7r 1.09 0.4928 0.5072
1.10 0.4920 0.5080
E' polar = r- loll 0.40912
0.40905
0.5088
0.5095
1. 12
1.13 0.40898 0.5102
seeu-ctrcte Arc A = n rt 1.140 0.4890 0.5ll0
1.15 0.4883 0.5ll7
y = .6366 r 0.4877 0.5123

2~2
1.16
1.17 0.40870 0.5130
= 11 rat 1.18 0.4862 0.5138
I.::L-__L.LI -2-
t:~~ o:;~~
0.5145

1._. = .2978 r >t

1. 22
1. 24
1. 28
0.4835
0.4822
0.40809
0.5151
0.5165
0.5178
O• .5191
0.5205
1 28 0.4795
Quarter-circular Arc Area = n rt 1.30 0.4782 0.5218
-2- 1.32 0.4770 0.5230

~r4-.--'-1
t l 1.3< 0.40758 0.5242
1.36 0.4746 0.5254
y = .6366 r 1.38 0.40733 0.5267
1---;2 1.<0 o 4721 0.5279
= 11 rat 1.405 0.4693 0.5301
-4- 1, 50 0.4667 0.5333
I 1 • .5.5 0.4641 0.5359
2 I 1. 80 0.4616 0.5384
I._iii = .149 r"t 1.6.5 0.40.592 0.5408
I
1. 70 0.4568 0.5432
1.75 0.4545 0.5455
1.80 0.4523 0.5477
CIRCULAR ARC 1." 0.4502 0.5498
1.90 0.4482 0.51H8
1.95 0.4462 0.5538
2 00 0.4443 0.55.57
2.10 0.4409 0.5591
2.20 0.40375 0.5625
2.30 0.4343 0.5657
2.40 0.4312 0.5688
0.4284 0.5716
;.~~ o 4259 0.5741
2.70 0.40233 0.5767
2.80
2.90
3.00
0.4209
0.4188
0.4168
- 0.5791
0.5812
0.5832
3.20 0.4128 0.5872
J 40 0.4090 0.5910
~-
3..60 0.4060
J.80 0.4030 0.5970
Area = art a in Radians 4.00
4.20
0.4000
0.3975
0.6000
0.6025
0.3950 0.6050
i =r sin a
4.40
4.60 0.3928 , 0.6072
a ' 4.'~
'.00 ~:~:; ~~~~
6:0.6152
5.50 0.3848
(a + sin 2(1) ~. ~810 0.6190
2 ~.~ o 778 0.6222
7.00 0.3750 0.6250
7.50 0.3725 0.8275
8.00 0.3702 0.6298
9.00 0.3668 0.6332
10.00 0.3636 0.63640

Y =r (1-008
c 0) ,l1xx=.W- = rlilt (j-cos a)

- r"t (0
I xx - sin 2(1)
2 - 2
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A3.5

Table"
HorizOIItal Vertical

I" Bo. Bailie

'eight IArJaSX
fI/I
lIo.ent
S .x ArJa .. 1 lIo.eut
.Jy

1 Propeller 180 0 in, o o 0

2 Engine Group 820 46 37120 o 0
3 Fuselage Group 800 182 145600 " 3200
4 .ing Group SOO 158 94800 ~18 -10800
5 Hori. Tail 60 296 17760 8 480
6 Vert. Tail 40 33.:1 13400 26 1040
7 Tail Wheel 50 328 16400 _20 _1000
8 Pront Land. Gear 300 115 34500 _30 _ 900
I 9 Pilot 200 16.:1 33000 10 2000
10 Iladio 100 240 24000 5 500

I"
where Ix, I y, and I z are generally referred to
Totals 3150

Example Problem 2, Determine the moment ot iner-

411180 .5480

as the rolling, pitching and yawing moments of
inertia of the airplane.
w = weight ot the items in the airplane
x, y and z equal the distances from the
axes thru the center of gravity ot the airplane
and the weights w. The last term in each equa-
tion is the summation ot the moments of inertia
of the various items about their own X, Y and Z
centroidal axes.
It w is expressed in pounds and the distan-has been divided into the tive parts, namely, A,
ces in inches, the moment or inertia is express-
ed in units of pound-inches squared, which can
be converted into slug feet squared by multiply-
ing by 1/32.16 x 144.
Example Problem 1. Determine the gross weight
center ot gravity of the airplane shown in Fig.
A3.3. The airplane weight has been broken down
mt c- the 10 items or weight groups, with their
individual e.g. locations denoted by the symbol moment at inertia from axis x'x' to centroidal
+.
Solution. The airplane center at gravity will be
located with respect to ~NO rectangular axes. In
this example, a vertical axis t~~ the center-
line at the propeller will be selected as a ref-
erence axis tor horizontal distances, and the
thrust line as a reference axis for vertical dis-
tances. The general expressions to be solved
are:-
x = Zwx
distance to airplane cvg , from
ref. axis -S-i5
Zw
y = Zwy distance to airplane c vg , from
ZW ref. axis X-X
Table 4 gives the necessary calculations,
whenee
x = 417180 = 133.3" aft ott propeller
3150
y 5480 -1.74" (below thrust line)
3150
tia about the horizontal centroidal axis for the
area shawn 1n Fig. A3.4
Solution. we rrrst :C1nd t he moment of inertia
about a horizontal reference axis. In this so-
lution, this arbitrary axis has been taken ~s
axis x'x' thru the baae as shown. Having this
moment ot inertia, a transfer to the centraidal
axis can be made. Table 5 gives the detailed
calculations for the moment ot inertia about
axis x'x'. For simpliCity, the cross-section
B,C,D, and E~
lex 1s moment at inertia about centroidal
x axis at the particular part being conSidered.
Distance tram axis x'x' to centroidal horizontal
Y
axis = = :Av = 17.97 = 2.9l w
ZA 6.182
BY parallel axis theorem, we transter the
axis xx,
Ixx=Ix'x' -Ajlt=79.47-6.18x2.9P.=
27.2 in"
Radius of Gyration, Pxx = xx =lj27 .2 = 2.1" VI
A 6.18
Centroidal
Fig. A3.4 Fig. A3.5
Example Problem #3. Determine the moment at in-
ertia of the stringer cross section shown in
Fig. A3.5 about the horizontal centroidal axis.

11)\ I
Solution. A horizontal reference axis x~xP is
assumed as shown. The moment of inertia is
first calculated about this axis and then
.......,F=""-,---- - - - transferred to the centroidal axis xx. See
7 • Table 6.

Fig. A3.3
, Ref. line
Z
.3 • ,
CENTROIDS , CENTER OF GRAVITY , MOMENTS OF INERTIA

Portion 3 + 3' (ref. Tcble 1 )

Table 5
", y= .375 + 4 X (.0841 + .0725 + .0625 ) = .375 +
,~ 3n .54
•e ". 0

·-" •·•"
e
•
•
- ,
• .
", • _.ent of inertia of
portlon about centroldal
lI: uia.
•,
! ,
I
I appr ox ,
.~72 = .547"
y = .375 + 2 :OJ. = .375 + .172 =547"
A 0.'" 2.7li 15.13 1/12 x .5 • 13 .. .0< 15.17 n
Al 0.50
'"
a. , 2.75 15,1:3 .. . .0< 15.17 lex= [.1098(.0841+ .0625) + .04x .54-

,• 1.00 '.25 5.2S 27.50 1/12 x 2 z .53 • .ca 27.52 .283x .0841 x .0625 x .04] = .002375
'0 2.12 2.875 6.10 17,56 1(12 x .5 x 4.25 3 • 3.19 20.75 .54
0 0.281 1.00 0.28 0.28 bb 3/36 .- .75 4(36.0.01 0.29 appr-ox , lex = .s s .04x .27"= .002365 In~
D1 0.281 1.00 0.28 0.28 .. .. 0.01 0.29
Problem #4. Determine the ::J,oment of inertia of
• 1.50 0.375 0.56 0.21 1/12%2x .7:i 3 • 0.07 0.28
the flywheel In Fig. A3.5a about axis of rotatlc n.
i5 \1116 . 182 17,97
I 179.47111.4
The material 15 aluminum alley :::ast1ng (weight =
• 1 lb • per cu. Inch. )
Table 6 AI Sec. AA
PartiOD Area , A, A,' lex Ix 'x' .. lex... Ay2 V/// / ~-------,
r/

u~
I + l' .0. 0 0 o .00281 .00281
, .06~4 .706 .04617 .0326 .00168 .03428
f;i
I l,.,.,f/ ""'~
3 + 3' .0679 _.547 _.0371 .0203 .000475 .02078
I cb •.. 5" i 1" I I
I
2" 10"

w~
s~ .1933 .009 • O~787
\ T I I I
,

I
I
y = l'.Av = .009 = .0465" -'--
l'.A :T933 r/////il 1" i
. -i--'
.
1= = lx~x'
_ Aj3
= .05787 - .1933 X IA Fig. A3.5a
:--3'l
( .0465)' = .05745 in~
Solution: ~he spokes ~y be treated as slender
Radius or gyration i.ox-x
= V· 05745 = .55"
.1933
round rods and the r-rn and hub as hollow cyl Ind-.
ers , (Refer to Table 2)
Detailed exptanat i on of Table 6:- Rim,
Portion 1 ... l' weight ," == n(5"'-4.l1)
n(R 3_r a)x3x
.1
~2' 75')
bd' .3=8.48 lb.
lex = 1 x 2
= (-04 ) = .5x8.48 (5
2"'r 2 2
12 I = .5W(R "' 4 2 ) =
173.7 lb. In'
= .00281 in- •
Hub,
Portion 2 (ref. table 1) = n(la_ .5 2 ) x 3 x .1= .707#
y = .375 + 4 Rlil ... Rr -+ r-« = .375 + 4 'I" = .5x.707 (l a .... 5:1 .44 lb. in~r=
3n R+r 3n Spokes,
(.5 3
.... 5x .54 .... 54 2 ) = .375 .. 331 = .706" Length of spoke = 3"
( .0 + .54) '""eight of spoke = 3 x .25 2 X .1 =
.588#
By approx. method see Table 1. I or one spoke = ',.rL:a/12+·>"j'r~ .588 x
y = .375 + g rl, = .375 + z x (.52) = .706" 3'1/12+ .588x2.S a =
n 4.10 lb. 1n 3
rt
I for 4 spOkes = 4x 4.10 = 16.40
Icx = .10ge x (R a + r-") t (R + r) - .283 Ra + r.llt
Total I of wheel = 16.40 + ;44, + 173.7 = 190.54
R + r
10 • In'
• . 1098(.2916+ .25)x .04x1.04- I = 190.54/32.16 x 144- = .0411 slug ft~
.283 x .2916 x .25 x .04 ~ample Problem #5.Moment or Inertia of an
1.04 Airplane.

Approx.
= .002475 - .000793 = .001682 in ~.
lex = .3trl," = .3 x .04 x.52 -. = .001688 in".
':'0 calculate the :nO:r:lent of inertia or an
airplane about the coordinate 9.xes through the
gross weight cen.ter of graVity, a breai< down or
the air?1ane weight and its distribution 15
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A3.7

necessary, which is available in the weight and axes, the prOduct of i~ertia of the weight about
balance esti=ate of the airplane. the reference axes 1s necessary. Column (14)
gives the values about the reference axes. To
Table 6a shows the complete calc~lation of
transfer the product of inertia to the c.g. axes
t~e moments of inertia of an airplane. This of the airplane, we make use of the parallel
table is reproduced. 'from :i.A.O.A. 'I'ecnnt ca L no te axis theorem. Thus
#575, ~sstimation of ~oments at inertia of air-
planes from Design Data.~ Z'NXZ C• g • -=Z'NXZ(Rer. axes) -Zwxz =48,857,589-
Explanation of Table 5325.3xl15.9x77.8=B39,253 lb. in.a
Fig. A3.5b shows the reference of planes To reduce all values to slug ft.a multiply 1
and axes which were selected. After the moments 32:l7
of inertia have been determined relative to x.J:....
these axes the values about parallel axes 144
through the center af gravity of the airplane Hence Ix -= 3061, I y = 6680, Iz -= 9096, Ixz -= 181
are found by use of the parallel axis theorem.
Having the inertia properties about the co-
Column (1) of Table 6a gives breakdown of air- ordinate c.g. axes, the moments of inertia bout
plane units or items. the principal axes are determined in a ~anner as
Column (2) gives the weight of each item. explained for areas. (See A3.13).
Columns (3), (4) and (5) give the distance or
the c.g. of the items from the references planes The angle 0 between the X and Z axes and the
or axes. principal axes is given by,
Columns (6) and (7) give the first moments of tan 2 0 = 2I xz = 2 x 181 -= .05998 hence >:1 =
the item weights about the Y' and X' reference Iz- Ix 9096-3061
axes. 10 43"
columns (8), (9) &id (10) give the moment of
inertia of the item weights about the ~eference
axes.
Columns (11), (12) and (13) give the moments ot
inertia of each ltem about its own centroida1
axis parallel to the reference axes. Such items
as the fuselage skeleton, wing panels and engine l
.
, '
have relatively large values for their centroid-
al ~oments of inertia.
/ ". " ' \ ~ .
The last values in columns (3) and (5) give
\ \
, '. .
....
the distances from the reference planes to the ~ \
. " , ,
center of gravity of the airplane.
Xc •g . -=Z'~·'X-=617.024 (col. 6) ==115.9 in.
~';~

l
.
Zn 5325.3 (col. 2)
ze ~
'0'
-= ~wz-=
_
414.848 (col. 7) -= 77.8 in.
. \
Zw 5325.3
The last values i~ columns (8), (9) and
(10) were obtal~ed by ~se of the parallel axis i
:
\\ , (

theorem, as follows:-
Zwxa about e.g. of airplane -= 97,391,595 - E325.3 (

x 115.9 2 -= 26,691,595.
, ',0--_.-- .
(

zwa 2 == 33,252,035 - 5325.3 x 77.8 2 -= 992,035 ~

\---:...-,'-- -.---'1-}' ~
The tmru :'rom the last value in columns
(11), (12) and (13) give tr..e .noment.s of inertia
or the airplane about the x, y ar.d z axes \\t-"'; /~ ~~
through the e.g. of airplane. T~e values are \'~ ! .: :
obtained as follows: /
1y =Zwxa + ZWZli + Z<lI y -= 26,691,595 + 999,035 +
3,120,384 -= 30,304,014 lb. in s
Jx = Zwy~ + Z'Nz. 3 + Zl;I x -= 10,287,522 + 992,023 +
L-'-
.
~
\

\
'\
-"'y,
"\' ~ l
:
!
~

2,899,470 -= 14,179,027 lb. In.'''

I z -= Zwy a + Z':.,rzl1l + ZIlIz = 3..0,287,522 + 26,691,595 +
_ ~:f
...
........
>="
~
,,:e s
5,157,186 = 42,136,303 lb. 1~.2 (REF. N.A.C.A. TECH. NOTE No. 575)
In order to dete~ine the principal inert:a Fig. A3.5b
 A3.8 CENTROIDS, CENTER OF GRAVITY, MOMENTS OF INERTIA
~'ole 6a

,,-
i . ,.
elillt
,.
•
..
Y
S.

• "
••
"
,. .,,'
COMPU!ATIOHS 0' MOMXiTS OF IBERTIA
. ••
"Y' ..'
10 .. 11.

."
12 •

t
lJ •
.,,
H.
m

C.ll~e:r ,eCl't1oD. 106.6 102 - " 11,098 6,20a 1,131,955 - 353,4;;1 361.339
·'
- 251,229 532,563

.....
no,......C1bly
Clln~ll:r oectloQ
• tc .
....
204.6 121

H.
- "
· ee
.
24,751
141.462
11,662
4,631
2.995,549
1.844,317
0 664,745
254,705
491,24.5
aoZ,164
-
33,680
4091,245
235,844
1,411,126
685,388

...... ....
Center .ecUon 0

rlbl, eee .
Flap ".0 ,.0 - .553
lsa
3,960 1,166 71a,8OO - 61,796 48,598 -- 48,596 209,880
Outer p&l1el ~
Outer paDfll
104.6 105
155.5 120 lsa '5
10,983
18.672
6,799
10,114
1,153,215
2,240,640
2,545,546
3,186,682
441,935
657,4ol0
164,514
274,478 0
I 184,514
374.,4078
713.895
1,213,680
Outu p&l1.1 rib' 139 ISS 54 12,462 5,747 1,735,Oae 2,185,373 367,621 158,407 17,601 175,008 798,861
Al1nona 31.4 172 lsa sa 5,401 1,947 928,938 764,150 120,702 55,390 0
55,390 334,850

Ho:rlEOll~al tall 87.1 ,., -- 96.7 31,966 8,423 11,731, ua

· 814,~3 176,378 - 176,378 3,091,083
Ve:rtlcal tall 31.4 OS. 125 11,053 3,925 3,890,586 0
490.625 110.200 10,174 .n 1,381,600

rUI.lac· oklle- 3H.0 1.,. - .1. 55,4&' .35,434 9,726,484 - 2.060,154 69.394 1.5713,594 1,5'ro,OOO 4,4075,384
to.
Eng1Il..

ruztl.~k
_ _t

( fa.1:rll'11)
40.5
".5 ...
'0 0

-
·ec
.0
.0 '.""
141,319
3,4140
a.aac
145.800
3.139,026

-
0
- 259,200
310,400
5,184
15,181,
5,184
57,861
5,164
56,648
194,400
985,520

f1l'n&ll
Step.
11.0 '0
'.0 170
.0
'0
'70
:l4O
••0
140
53.iOO
51,800 600
70,400
'.800 - .3,200
0
1,100
0
1,100 61,600
23.800

If . .l.C . .l. c~l1GS -c .0 SO

,",. -s
0
.0
'.709 '.600
',500 1",000 - .... 000 16,940 15,470 15,470 280,000

..
C.. bln &lid. 66.5 10. 7,18a l,417,5l4t 0
77S.e56 2,394 106,400 108,794 l,048,S7<l
_1lll11.bi ..111
foo-c tI'O~U
floor, ,,~
'.0
'.S 210
•• 0
1M
1,995
I,.
62'
11,858
418,950 0
SO ·9,248
4l,3S2
0

'0.
0

1.Z69 ·
1.976
10,.72
131,670

.. ·
"'tllg flllOlU 18.:1 142 20 58 .,Sa7 1.073 373.034 '.400 62,234 18,944 18,944 152,366
8o'Cto.. eow11Jl4 27,0 140 11 '5 3,180 2,02:1 529,200 3.267 151,875 0
24,300 24,300 Z8:3, .500
~ 0111e fram••
,,., "'. sa 104,853 5,160
• s 23,250
Al'l'el-C1JWJ ;;1001' 1.3
"
0

-
0 0

'l'a11·• .cl.el parI.. '.0 0

54 I._ ase 53.,900 0
za.aa4 100 100 raa ,640
e-Cc.
Sld.e doora
ae.aase door
fabrio &D4 40pe
Ta11 cone
17.0 143
1 .a ,.5
13,0
'.5
...
,., --
ie
0
••
.3
.0
91
2.431

"'"
3,302
a,8S8
1.394
11.3
1,040
••3
347,633
49,005
83S.708
1,111.688
0

0
5.se8 114,308
7,144
83,200
62,108
1,088
720
".394
120
28,288
·
15,509

--
av.acc
720
15,509
120
199,342
18.111
z64,l60
::>S2,763

.... .. ••
Co_Unr, ata- 12.0 110 0
se 1,320 1.140 145,200 0
108,300 1.200 1.200 :1...35.400
dOI18 1-3

CIla,e1. (1'_ 332.4 115 Sl. 26.726 11,6:12 3.073,490 e77,678 604.472 0
23,240 23,240 1,383.0<36

.. ....
trected)
Retr..cting "ocb-
""
- "
no as
sa
3,146 1.916
S._
see.ceo 17,875 128,JS5 0

- 0
210.7841

_.e '.-
Wb••l., 91.0 lU 54 U.83l 1,809,171 285.356 385,378 0 0 0
718,536
1'atl _.bee.l ".0 "0 1.924 3.369,600 0
142.376 ee 0
ee 692.640

l:J:1g1n. acc••-
aorl ••
Eng1A. control.
PTOp.ller
1049.0
90.' ,
,.
11.0 1.03
aaa.c
37.0 sa
J3
52

'.3
0

10
0
.0
.3

SO
M,617

1.133
a.065
83,920
'.71.1. '.520
.,.
17.760
1.142,~1
244.9941
118.699
19,201
·-
I,l.OO
·-
6, '7l3.600
ea4,143
83,538
1,420,300
253.658
9,060

-
177,600
a.
3:13,858
·
0

ea.sce
253.858
'.060
0

69,300
..,
2.769,3&0
391.030
88,109
185,168
S'ar1:1~ .ye-c_ .5 .3,0'73 3,145 116,032
- 361,325
'" 116,120
0

,.
Ll,lbr1ca'Uq aye· ae,c
•• 0
sa I._ .3.132 U3,798 l'U.834 1,374 0 1,.374 147,J.08
fuel .,.,to_
InetZ'WIIl!'l::l'.
841.0 12.
38.0 10. ·
0
.0
sa
10,496
3,8'16
'.sao
3.498
1.343,488
395.3541 ·- 534,300
331.6341
.32,008
.3,4341
20,008
SO. 600
35,668
63,233
839,680
356.592
81U'face COA- 81.5 160 0
'1. 13.040 5,187 .3.088,400 0
4010.842 U13,ll84 130,4.QO 254,:562 9415,840
trol.
F\lnl1a.bio.p. 160.0 lsa - SO .34.960 13,800 3,893,160 - 1.0M,OOO 108,160 400.000 508,1150 1,996.600
Slectr1aa.l
.qu1~t
Koht 011c.g
UJU!'L.,t.U
. IKP'n'
100.7 128

---'.0 US
3867 •• 106.6
IS
0

0
SI
10.
---
",""
,go
----81.'
8,157
7i,35O -
1,849,869
82.424
74.8 41.41,199 .2a9,l38 67,342.573 10,283,443 aa,a63.738
32,556 560,593
·
.3,881.572
ea.
0

·
2,890.«14
·
...
4.980,508 31,090,714
l,04.,O!8
70,380

.,....
Pllot 200.0 10' - go 21,000 1.8,000 .3.205,000 1,6ao,OOO 33.600 28.800 '.000 1,890,000
- •• -
0

Ob.e1"W'.r 200.0 205 41,000 17,800 8.405,000 1,584,.300 33,800 38,600 '.000 3.649,000
Fuel
OIl
Vlry P1eto1 , ..
780.0 132
75.0 Yl
195
0

·U
85.S 1041.990
.5

"
5,3a5
751.
",.
66,690 13.590.720
6,375
., 378.075
148,296
0

..
,t ••
;',701,996
541,875
21,938
141,180
3,615
-
172,380
0
157,560
3,675
--
8.603,Oao
452,625
57,038

... -, ."
0

se
.....
3.:)118 eaDd.l.,

.,.
'.0 IS' 14 13:1,424 39.864 0 0
70,6sa
.....,.
Flo.. t l1rttta '.0 190
148.7 l7S
14 72
.5
I, 7l.O
25.«11
,
324,900
12.130 4,5.31,307
1.7&4
0
46,656
1.031,008
0

5.U7 ·
0
·5,13'1 143,120
3,159,051.
ese ·
·- ,as
Cbart board.,
Or1!t 11ght '"
1.6
'0
I,.
23.680
m
J3 32,693 0 0
.37,824
Tlre-C a.1d.
Life :raIt
'.0 I.,
34.0 1.'"
14
10
0
54
101.
'55
"0
4,S.34 3,434
78.854
108,900
628,864 - '00
U • .390
12,996
3046,834 --
0 0

· see
aQ,837
37,540
461,024
U8U1JL LOAO 1457.9 140.5 0 88.7 204.828 1.26.450' 30.549,023 4 ,079 ~88 ,299 311,898 229,980 1'16,878 17,786.87:1
fOULS 5335,31 817,0.24 414,848 97,asl,595 10,.387,saa ~S.2,035 2,899,470 3,120,384 5,157,188 48, 857,58a
COflR.u::,.IO. 53a5.3 1+5.9 0 77.8
,,
r71 200,000 Or~O 000
IE.t\9i,5il5 10.Z67.5aa l~a,oj5
48, 0 1 33t1
a~ ,aS3 1
1
,
, , 10.287,5412 as. 691. see,26,691,595 ,
,,
,
,
,, I ,
99. S'~ 36,804,014 4,136,303

.
i 292 035 1~ 287 52.3
,
,,
," ..
,
I • ,
,, I 1",179, , a ,, ,,
,
,
,
, , I .3,662
,, 51.8 , 76.1
5,784
, 7,912 I
, , ,
J:c.g. a ~--"'~'i" ,,
,I
1'x
e.g. I a.g.J

, : :_~ :'~·---~'3 ..,,< .... '.~y ' ". • ~. "~':;, >.... ,..,..r"_.~"'-~~"_. t, -'): ~",. <:», ·,...:<··: . . . >:';.,~··~~··-~;-"·'-::': ...~·:·~:~i~~ . . .;;.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A3.9

The principal moments of inertia are given by will be given the symbol Ixy, hence
following equation.
Ixy = I xydA - - - - - - - -(1)
I xp = 11: cos" £I + I z sin" 0 - Jxz sin 2 £I. (see The unit, like that of moment of inertia, 1s ex-
Art. A3.lll pressed as inches or feet to the 4th power.
Since x and y may be either positive or negative,
I yp = I y
the term Ixy may be zero or either positive or
I z = Ix sin" 0 + I z cos" .0 + I xz sin 2 .0 negative.
SUbstituting ~oduct of Inertia at a Solid~ The product or
inertia or a solid is the sum or the prOducts
Ixp = 3061 x (0.9996)" + 9096 x (0.0300)' - 181 x obtained by multiplying the weight or each small
.0599 = 3056 portion in which it may be assumed to be diVided
Izp = 3061 x (0.0300)' + 9096 x (0.9996" + 181 x by the product ot its distances trom two or the
. 0599 = 9102 three coordinate planes through a given point •
Iyp =6680 Thus With respect to planes X and Y
Ixy = I xy dW
A,3.7b Problems
Jxz = I xz dW

n r Zi ""i I yz =I yz dW

'"i A3.9 Product of Inertia for Axes at S~try.

T It an area is symmetrical about two rec-

1· tangular axes, the product ot inertia about these
I
axes is zero. This tallows tram the tact that
I kR ,~ symmetrical axes are centroidal x and y axes.
8 .f0~
'1- It an area is symmetrical about only one of
1$ It ----IT, two rectangular axes, the product or inertia,
!xydA, is zero because tor each product xydA tor
Fig. A3.6 Fig. AJ.7
an element on one side at the axis or sYmmetry,
(1) Determine the moment of inertia about there is an equal product at opposite Sign tor
the horizontal centrOlda1 axis for the beam the corresponding element dA on the opPosite side
section shown in Fig. A,3~6. or the axiS, thus making the expression !ydA
(2) For the section as shown in Fig. A3.7 equal to zero.
calculate the moment or inertia about the cen-
troidal Z and X axes. A3.l0 Parallel Ax18 Theore.
The theorem states that, "the product at
inertia ot an area with respect to any pair at
T, co-planar rectangular axes is equal to the prod-

Z
1
S
T I
' 1-'"
1<.'-1"
+ uct or inertia at the area with respect to a pair
at parallel centrold.al axes plus the product at
the area and the distances at the centroid at the
Fig. A3.9 total area tram the given pair.ot axes", Or, ex-
A3.S pressed as an equation,
Ixy F Ixy + AiY- - - - - - - - - -(2)
(3) Dete~ine the moment or inertia about This equation is readily derivable by re-
the horizontal centroidal axis tor the section terring to Fig. A3.l0. yy and XX are centroidal
shown in Fig. A3.8. axes for a given area. YY and XX are par-a.lLeL
(4) In the beam cross-section at Fig. A3.9 axes passing through point O.
assume that the tour corner members are the only The product or inertia about axes YY and XX
effective material. Calculate the centroidal is I XY = I(x + x)(y ~ Y) dA
moments of inertia about the vertical an~ hori- = IxydA + 'is I dA + x I y dA + Y I xdA
zontal axes.
The last two integralS are each equal to
A3..8 Product of Inertia zero, since !ydA and IXdA rerer to centroldal
axes. Hence, I xy = IXYdA + i j IdA, which can be
In various engineering problema, particu- 'Nritten in the tarm or equation (2).
larly those involving the calculation of the
~oments ot inertia at unsymmetrical sections,
the expression I xy dA is used. This expression
is referred to as the product of inertia at the
area with respect to the rectap~~lar axes x and
y. The term, product ot inertia of an area,
A3.10 CENTROIDS, CENTER OF GRAVITY, MOMENTS OF INERTIA

By adding equations (3) and (4), ~e obtain I x1.+

1r I Y1. = Jx + I y- - - - - - - - - - - - - - - - - (.5 )
or the sum of the ~oments o~ inertia of ar. ~rea
with respect to all pairs of rectangular axes,
thru a common pOint of intersection, ~s c~nstant.
13.12 Location of Axes tor which Product of Inertia
is Zero.

In Fig. A3.ll
·_-+-x I X1. Y:l, =!X1Yl dA=!(xcOs~+:;sin,2j)(ycos.z;­
xsln0)dA
= (cos s 0 - atn- ~ / xyG.A'" cos Zl sin 0
y Fig. A3.10 (y:a..x a ) d.A

.l3.1l Moments of Inertia with Respect of Inclined

=Ixy cos 2 ¢ + ~ (Ix - I y) srn 2 "
Axes
2
Therefore, IX:l,Yl is zero when
unsymmetrical beam sections are very com-
mon in aircraft structure, because the airfoil tan 2 0 = ~ - - - - - - - - - - - -(6)
shape is generally unsymmetrical. Thus, the Iy_I X
general procedure with such sections is to first
find the moment at inertia about some set of A3.13 Princ1pal Axes.
rectangular axes and then transfer to other in-
I~ problems involving ~s~etr~cal bendl~g,
clined axes. Thus, in Fig. A3.11 the ~oment at
inertia of the area with respect to axis X~X~ is, the moment of an area is frequently used with re-
spect to a certain axis called the principal
Ix , = Iy ,, dA = I(y cos 0 - x sin 0)'dA axis. A principal axis of an area 15 an axis
cosa0! y·dA+sln a£l!x adA-2sin0 cos .0 about which the ~oment of In~rtia of the a~ea is
either greater or less than for any other axis
I xydA paSSing thru the centroid of the area.
Ix cos' 0 + I y sin' 0 - 2 Ixy sin 0 cos 0 (3) Axes for which the product of inertia 1s
zero are principal axes.
and likewise in a s1m11ar manner, the following Since the product of inertia is zero about
equation can be derived: symmetrical axes, it follows that 3~etrical
I y1. = Ix sin a 0 + I y cca- ¢ + 2; Ixy sin ¢ cos axes are principal axes.
The angle between a set of rectan~~lar
0---- --(4) centroidal axes and the principal axes 15 given
by equation (6).
y
Example Problem 4.
Determine the moment of inertia of the ang-
le as shown in Fig. A3.12 about the prinCi?al
axes passing through the centroid. Solution:
x:_-I,.- ='*"'""'O----''--Ir..-_ x Reference axes X and Yare assumed as shown in
Fig. A3.12 and the moment of inertia is first
calculated about these axes. Table 8 gives the
calculations. The angle is divided into the
~NO portions (1) and (2).
Fig. A3.11

Port
.. ,
'ox '0, ~
A~' 'x
Ax_ Iy !
A x r A, Ax A" Ax'
1 .3705
." ,1" ,0469 ,281 ,OO~8 ,211 ,03~1 fi:z 1.5U~3 e , 0019 1 Jl:1xl 53 •• 070 ,0077 ,281 !

.
12 4 '
a .500 ,1205 1.205 ,625 .06205 ,7800 ,0078 .0781 1 1 3
12 :z,,:Z2 ,167 ~2X2J:.2S3 •• 0026 ,9470 ,010
i
,6 19 .J.oli35 ,7858 .2188 ,1132 ,95.1 .291
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A3 11

-lex ~nd ley of inertia of each portion

~ ~oment section has been broken down into 16 stringers
about their own X and Y centroid- as listed in co lumn 1. For the top surface, a
ar axes. width of 30 thic~~esses of the .032 skin is as-
Location at centroidal axes:- sumed to act Nith the stringers and a Nidth of
25 thicknesses or the .04 skin (see Col. 3). On
-Y = ZAy = .6719 = .767" the lower surface, the skin half Nay to adjacent
l:T ~ stringers is assumed acting with each stringer,
x- = l:Ax = .3435 = .392"
or the entire skin is effective. Col~ 4 gives
the combined area of each stringer unit and is
l:T .875
considered as concentrated at the centroid of
~ransfer ~oment of inertia and product of iner- the stringer and effective skin. All distances,
tia from reference X and Y axes to parallel x and y, columns 5 and 8, have been scaled from
centroidal axes:- a large draWing.
Ix = Ix-J.jg = .955- .875x .767:1 = .440
-I y = I y - Ja.-, ~ .291- .875 x .392
-- Yp Y
2
= .157
I I .040 Skin
jef.
1XY ~ ZAxy - AJ:.y ~
- .150
.1132 - .B75 x .767 x .392 =

Calculate angle between centroidal X and Y axes

and principal axes through centroldal as fol-
xp
_X
F,r.
:7 -::1 _
......2 3

-- -
"i
5,
I
6' Ti:!--f~

r f XRe f.
lows: - x
.---=-- - ---L... -,-. x

~~I
0

,!
tan 2 0~2 Ix" ~2(-.150 =-.30 =1.06 x
12 13
I y _ Ix .157- .440 - .283 .r ~4 ,
15 I~
T ,
2 0 ~ 46 0 - 40' 0 = 23 0 20'
-
I I
Calculate moments or inertia about centroidal
Fig. A3.13
I. I
principal axes as follows:- Yp Y y

1xp=1 x cos" .0+1'1 Sin2121-2Ixy s m c cos a Table 9

, a a , ; , , • •
= .44 x .918 2 + .157 X .3965 2 - 2( -.150) x " U

x .918 m'
·,• ..• :::':,·..., '"" I.....
~

.. . '
.3965 504 Area
" I
-
I Y1l ~
-
Ix sin 2 .0 + -I y cos 2 0 + 2I- xy sin 0 cos el , ", • lx,
"
~f.
"<
o~
<
" ~.,

~ .44x .3965+ .157x .91a + 2( - .150 ) x

2
•, " • "<
.3965 x .918= .092 tn ;" .U _4.6041 153.85 1-18.56
a .U ·
03
.03 1.14.14
4.00
6.05
0.560 2.240 -J3.1.5
0.8471 .5.124 _29. 28j _1. 099 120.02 _24.80
a .30 .011 I .38 7.00 2.660 L8.620 _24.8.5 _9.443 234.66 _66. 10

,••r ." .,
q:-:'!- •o-t • 17 7.37 1.253 9.234 -21.18 _3.600 18.205 _26.53

T' ? t~
:.ra .~
.0< .U
.>0
1.50
~.288 9.68 I-~~' 60
1. 215 9.563 -12.60
-
_2.8::
_2.142 ~:::: :i~: ~~
I fL a .ra
.0<
· 0<
.U
.U
1.30
6.90
1.241 9.0:19 8.60
1.111 8. ceo _ 4.00
_1.462
_0.680
12.51 _10.61
2.12 _ 4.69

." ."
." ."
s .J<
~t
2f ,\1' ';""-"
,"'~i
6.50 1. 885 12.252 - 0.35 _0.100 0.04 -00.65
io .U -a.nc _0.561 1. 8.51 _33.2:1 _5.6:13 181.96 111.611
U .ur
." .U -4.90 _0.83 4.0112 _29.28 _4.918 1411.16 24.39

ra .U ." ."
ra .La _5.911 _1.6661 9.913 -24.8.5 _6.9581112.90 41. 40

, ~-~; ""ie .n
." .n
.n
.30
_1.40
_8.13
_2.2941 16.916 _LS.10
_2. S2~ 20.481 _12.42
_1l.191 108.40 42.90
_2.850 4.1.82 31.30

I-L -
•

'+-,(
, ...L
.n
.J<
ata s
.n
.n ."
a.
.30

-
_8.62
-8.81
...
_2.612 22.432 _ 6.10
-3.094 21.444 -00.35
.0<
_1. 891
_0.122
-,
11.54 16. ;10

• .,
O. il4 1. 08
- a. s

;,---I Location of centroldal axes with respect to ref.

y :> axes,_-
Fig. A3.12 y='Z.Ay=- 1.465 = - .396"
l:T 3.70
~~~ple Problem 5. X=,ZAx = - 58.238 - - 15.74"
Fig. A3.l3 shows a typical distributed l:T 3.70
flange - 2 cell - wing bean section. The upper'
and lower surface is stif:enec by Z and bulb I Ix = 187.04- 3.70x .396:1 =186.5 in ~
angle sections. _Determine the ~oment of inertia 1y = 1348.36-3.70x 15.74 .431.7 rn,' 2=

~f the sect:on aDout the principal axes.

SoLut t on : I I XY = -13.35 - (3.70 x .396 x 15.74) =36.41 In ,»
The properties of the cross-sectlon depend tan 2 0=2 I xy = 2 (-36.41) =-.29696
upon the effective ~ateria1 Nhich can develop
resis~ing axial stresses. The question of ef-
Iy-Ix 431.7-186.5
fective ~aterial 1s taken up in later c~~pter. 20=16°-32.5', .0=8°-16.25'
Table 9 shows the calculations for the moment of 1xp=Ix cos" .0+ 1y stn" ~-2 1XY sm e cos e
inert~a about the ass~ed rectan~Jlar reference
= 186.46 x .9896 2 . 4 3 1 . 7 x .1438 8 -2
axes XX ~nd YY (see Fig. A3.13). The cross- E36.41X .9896 x (-.1438)J = 181.2 10.-
 A3 12 CENTROIDS , CENTER OF GRAVITY , MOMENTS OF INERTIA

-I y p = -Ix e-
s In e 0+ 'y c os e el+2 I xy s t n 0 A3.14 Section Properties of Typical Aircraft Structural
Sections.
cos 0= 186.46 X .:~822 + :;)1.7 X .2695 2 + Table A3. :0 'thr-ough A3.15 and Char-t ......' 0 ; v.~

2 -36.41 x ,9::95 x (-,1"32) =437 in ... si'le the section pr-oper t t es of a few st:-uct.. . ral
shapes G8rrnOn to arrcra rt . Use of tr-ees tab Ie s
will 08 :TIade in later cha;Jters of th:!.s jook.

, ,
f--e-J ,~
n"TI
---.-.z~1
'I
I R
£=T ,il
-'-T
TaDI• ..1.3. 10 TB I I 'I
- r+ ~T Table ..1.3.11
x-' 'i
X PfDlI"rlies of
zee Secl'Q/\S
X ! I !
!
X
Properties ot
Channel ~ectlons
, ,::,/-R AS<T I V-R I R--) . )
---..'-
I J ..J,_ __Ai2==:jI
~a-- -T
, -T
I
I
I I
y
A

NOMINAL DIMENSIONS SECTION ELEMENTS

"
Z
I I r I R AREA
'nInches
I '" I tOyy "
Z
NOMINAL DIMENSIONS
.">REA
, :>ECTlON ELEMENT:>

"""
U
A B
IDcMa Square
Inches
Puc
Inches
"""
U
A I B
InChes
r I >-
:>q..are
InChU
Inch
'n
Incnes'
I Iyv I '"
I
"0'< I ~v,'
Inche.

". on
=t:
-+ . '00
.eas · 050
· 06J
.050
· ".
·
· 125
· 0852

· 0977
· 0035
.00<12
00"
· 00<.
.0062
· 0122
· 203
.197
""
'"
rn
· 353
1-+
I-}
· 375
· 500 '"
· 050 ".
.0127
· 0852
· 106
· 153
· 0009
· 0020
.0060
· 0075
I .111

'" '""
~96

", ". ,,,

.750 ,I><, 0," · 0 17 204 · 003, 0090 · 195
.063
''''
--+ .625 ,,,
· 125

· 125
· 1231
0 o
, 0071
· 0146
· 0069
20'
· 255
.346

'" H:
H-
.750
'" ~
· 063
· 155
,n .207
JI5
· U045
.oi rr
.010S
.0!-l7
· 190
275 ,,.
· 296

j+ .625
'"
so •
• 115
0 0070
· <lOS3
cia '"
· 259
· 268

'"m 1-+
· 875
· 315
· 094
",
' 125 .aao
· 0790
.327
· 0995
0159
0009
· J193
· 00 7
,;;
106 ,
290

H- 1.000
.063
00,
.125
· 1~5
· 131
.163
0084
.0109
0148
.0367
.253 · 336
H- .500
.050 .094 .09\6 .146 .0020 .csro .148
, 347

.. '" ". .063 · 115 0025 , 0132 144 J~9

I i1 . 625
.050

'"
·.cso
.094
,
· 125
.0977
· 123
.0090 .00l>S
.0082
;0 •
.298 ~ .625 ~
.125 .131 '"
· 196 · ,}047 015"
'" 347
· Olla
.no
· 259
pt .076
, · 158 · 201 0055 0183 1ST
, · HI

~
.750 .110 .010 .0122 ·~ 3

"
.750
,a ." ra us 0148 · 304 .n
i-#-
.875
.750
--*l-
· '0 4 .11.
.147
245
.247
0065
0079
0151
.OIS4 ,
2JT
aaa
'"
0
1. 000
,,, ".
.125 · IT .0165 .0367 .312 · +64 .or .17 • · ~53 .0094 0214 ",
Ri:te
;0
.050
r*-" .S2S
"', ".
· 125
·, 1~5
· 104
· 131
· 111
0130
· 0158
· 01 1
00"
.0122
ce
· 35J
· 347
.360
· 251
· 251
aaa
I. 000 '~
us
· 125
.328
aie · 362
.319
· 1~0
· 0~13
.03ll4
OG26
· (;276
0376
· .14
.304
.;H
;)58
aaa
304

"*"* 4*
18 .750 ,n61
1
Hl-- .87S .078'" .125
.147
.177
.0164
0214
O14B
.0175
254
.348
· ;)17
.314
.-g.
.500
". .H5 0029 .0211 HI 377

H¥-- ". ~
.139 187 0049
'" · 187
, 394

... *
.050
'" ~.
· 0172 · 0197
'"
· 365
--li'-
. '" '"
· 193 · 0058
Hl--
f--#-
.875
~
.076
.153
· 197
.caro
,0245
· ,0241
.0287
.359
.353
.385
-+.\- ~ · 155 .229 0083
· 0252
.0250 · 231 '"
· .02

--m- .ras .282

~ , .137 m 0096 0294
,
zas 392
I~
... t
1.000 · 178 · 023 .CO .H I. 000 .~. )
.125
.018
•
.216 .0276
, .0439 .358
." - · 220 · 243
es
· 0113 0335
ts
.390

~
ceo .0285
.0123
'"
~
.625
· '"
.U 00' .400
, ."" ". ~152

,"
· 294 .0335 ~03
· 125 · 139 oe .875
• ., " .244
4 · 094 · 244
." '00 · 01,6 OJ83 '159
t
397
.'w
'" , -*ae -1E- · 2. · 041
",0643 .417

tl ~ .0251 .0147 · 306 1. 250

m e9a
· J78 .187 .02S4 .0175 .396
'" 061J7 · 087

.. m
1.000
~ .170 .0285 .0242 · ~1J9 ~ '00 ~
· 131 134 02
" "
"ae H\!- .158 · 139 .0031 · 0281

.,.,
· HO · -122
;::ll:: .875
~ .125 .206 .0335 .0287 · ~03 · 373 1.12..5 .125
,,.
·". ." ;86
.0363 .0332 .397 -+.\- ~
· 141 .178 · 805l nB5 H'
f--#- 1.00< ~ ao .036 .415 .444
aa .078 .171
, · 0060 03:H lH
'"H
*
01(12
I : ~94
232 · _3
"
as sas
· 076 .226
'"
0 sea • .750
2n .237 .0117 .0444 , oa225

~
'" .050
· ".
· 125
'"
• 142
0265
, 0304
0083
· 0197
..
.441
· 464
· 237
· 374 *
:j
as
~
~
· 181
.216
.278
.283
.0135
0156
0319
0441
.2,.
27\ ~52
,"
~
,875 .176 .0374 .0241 .368 .875

s.t""*"" ~
.256 .289 .0183 .0507 :68 .. 5
., ·, , ~
.216 .0441 .0287 m ,364

.. .
1.115
0< a .445
L 125 · 125 .125 .328 ase · 0225 · '-1616 sea .33
.125
! .,.
Pi!:
· 236 · 238 0232 .0495 .214 I 4Je

=ll!:
.210 .0467 .0 30 , .72 .503
1. 12.5
." ". .0637 · 464 · 500
L 000 · 279 · 341
asa
0269
,""
.0569 .310 .52

"*"*"
.125 .359 0695 303
, '"
· .303 .0632 .0T44 · 457 · .96 m · ,66
';0
." ." ~~5.
· 508 0661 0757
~ .150 ~ .," · (1421
.04 7
,0148
.01.
1. 375
I .
,. · .53
,,
SI9 .0832 I 0930 · -123 · .53
, ,

'sae
"
! '" sa
1. 2~O LOOO ~ .202 · u~32 o · 514 --11- · 220 0071 0493
·· ~57 "
.
.125 .267
.094
~
.291
".
.0729 0511 · 501
'" Hi-
~
,875 ~
.125 .344
.278
288
.0190
.0234
.0650
'J79 · .51 481 '"
I. 'W ". ,
· 0765 .0887 · 519 · 558
,..g. 1.250 .265 355

~
.338 0886 · 1039 512 · 125 · 377 · 0334 0698 5n
· 555

..... ". ·
I. 125 .314 .383 .0389 0607 ,07
=iso
· l52

*=1t ."
0
" · 543
12 csa · 'J484 · (>994 .345 -ss

...
·,500
:jf ::$.: · 210 0662 .0367 .562
·, '" ,

..
1. 500 .560
· 25~ 07 4 · 125 · 1106 1292 · .70 · J08
,2,
,, "
1. 375 1.000 .125 · .15
~
.197 · 169 .00 3 0540 .179

"*-*4-
· .103 .0911 .0511 .549 .HI .525
.094 .232 · 115 .00B · i)621 171 · 517
~ ·" o.

.
I. 250 us ~ · ~68
.2.16 .263 0169 0705 547
· ,53
. ., s
1365 · 1308 · 549 .537 -fr
t
aae
."
. ~
· 8'15 279 ~68 .0198 · 0818 · :66 543
," .07 5 .:; 2

~
1. 375 · 125 ' .359 · 278 .0242 .1002 ' · 359 S2S
~ · 265 .0961 0439 .602 ,407 · 125
I. 500 000
· .115
,,
· .114 .111 0511 .596 .403
, ~ · 326 · J71 .0403 1010 I .352 557

=!t" I. 750
I. 2 0
,c
L .00
.094
,
". ras '"
.284
,
.134
· 11 4
1377
,
.1040

.0439
.0\1
u
.096
a
· 393
H}
ss
1.125

L sas
sae ~
~
.,
· ;56
..
· ~22
.513

.244
· 31l'1
.390
."
.0504
059\
, \710
.0076
1246 I .346
.1451
· 2035
.J76
.339

.177
SH
,
.532
,52

f:i!:
L 'SO .094 · 365 · 1925 .1039 707 .520 · :66 · ;0 I
ns
:::g: 2.000 L .00 .094 361
.12.
2194 .0511 " · 779 ate ~ \.;00 I avs ~ .125 m .269
-a 0090
.0249
0939
· 1236
:7
· 258
HB
;H

-*" L 'W ". .125

· 408 · 2621 1040 · dOl
•
· 505
'"
.".
· .55 · 278 · 0289 · 1485 , · 257 .e rz
",
1. 125
L 0 · ;94

"
3852
,. 2346

"
.029
" · 338 · 380 · 0414 · 1233 · 350 004

=H: 2,500 m ~
'" · I2S
· J23
· 385
.2890
· .1383
_0288
0331 """ ·'" ,298

=t 1.000 ~
"
00.
·;
.3723
.4707
0512
"00
.955
.941
"
.354
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A3 13

t 11"'7Y Table A3. 12 j-i---t Y ,.y

~,~ p
w , '
'~-Xit
Ia! ---L
----:Ld ~ z r
Properties of Extruded. Aluminum
Alloy Equal Leg Angles. (Ref. 1938
Alcoa Handbook)
T• I ~R
T..ble A3.14
I'rO,",rlies 01
UnequaL Angles

w
Dimensions
, R
Area
~.ID. I
Axis XX or YY
P d I
Axis ZZ
P , ~:
,x
'I'
3/.
3/'
3/32
1/16
3/32
1/'
1/'
1/'
.111
.089
· 132
0.004
0.004
0.006
0.183
0.220
0.219
0.181 .0015
0.199 .0018
0.214 .0026
0.111
O. 142
0.141
'"'"
",
1.-
I
y
, A~~
II' II' .171 0.217 0.227 .0034
1
3/'
1/16 1/16 .122
0.008
0.012 0.311 0.211 .0048
0.141
0.199 '"
",
,,0
a
•
NOMINAL DIMENSIONS

R AREA
SECTION EL.£MENTS

I
'" "'xx i
1 3/32 II' · 178 0.016 0.301 0.276 .0066 O. 19:J A a Y '------- i i ~ t"yy
1 1/' II' .23' 0.021 0.298 0.290 .0085 0.191 ", "
w
0 Incbe.s
Square
1 3/16 II' 0.029 0.293 0.314 • 0124 0.192 "
,
Inchel
"''' lJIChell
"'''
1~ 1/4 3/32 3/32
· 339
.230 0.033 0.38 0.34 .014 0.24 '"", .oea
, .,
", " · "'''
· 625 · 500 .050 · 0535 .126 · 187 .OOIl .H6
1-1/4
1_1/4
141/4
II'
3/16
1/.
3/16
3/16
3/16
.30
.43
.56
0.042
0.059
0.074
0.37
0.37
0.36
0.35
0.37
0.40
.017
.025
.032
0.24
0.24
0.24 '"'" H=
f-t .751l
·,
· 625
s
· ""
a .083
· 0818
aa
.
· 162
· 150
· 222
'"
·· 250
• · 0027
· 0029
" ., ,
· 01144
· 0069
"
·· 221
i· 274

~
...."
· 875 .US .100 0049
'" · 0073 · 270

..
1_1/2 .28 0.46 0.40 0.30
1-1/2
3/32
II'
3/16
3/16 .36
0.058
0.074 0.45 0.41
.024
.031 0.29 '" q
",
· 750

.625
~
.184
.<0
• 211
.,
I . l40
· 272
.320 .,
.00 2 .0123
, .,., .,.,.
.211 .258

.53 0.107 0.44

1-1/2 3/16 3/16 0.45 .044 0.29
'"
.0050 .017
1-1/2 II' 3/16 .59 0.135 0.44 0.46 . 057 0.29 ", ~
.108 .178
, '"
·· 322 · 00:'1 .011)6 na
• 217
.,
", atf#
1.000 · 750 .US ·
1_3/4 3/32 3/32 .32 0.096 0.55 0.47 .039 0.35 , · 125
.n ."
· 008 .01 2 .1
~ "
i .0111 · 2 2 ' · 310
1-3/4 3/16 .42 0.121 0.53 0.47 I .050 0.34
'"'"
1/' ta · 875
1_3/4
1-3/4
3/16
1/.
3/16 .62
.81
0.174 0.53 0.50 .072 0.34
.750
· 094
~
· 167
.124
· 231
· 159 ·ro,
· 292 ,DUO
.0054
,0074
.0155
.0197
· 257 · 305
.209 1 .400
3/16 0.223 0.52 0.52 0.34 ~ ~
'"",
.093 .179 .170 .416 .0278 .203 ' .393

,a
3 II'
3/16
II'
1/4
.49
.,.
.n
0.18
0.27
0.61
0,61
0.53
0.56
.08
.11
0.40
0,39
'45'
"
~
f#
1.250
.125
~
.125 .231
.139
.181
.239
.427
"I
.0092
.0124
.0346
.0217
, .291
00
~,
.381
, ".
~
1.000 .202 .250 .0\71
1/4 1/' 0.34 0.60 0.58 .14 0.39 fil-
ia ra .262 · es
· 373 .0306 .389

a 5/16 1/4 1. 16 0.41 0.60 0.61 .17 0.39 45'

.,.
· 384
aa
· 0215 · 0383
'" · 382

~ ~
.15 .0077 .0463 .194
.750
.125 .167 .534 .0096 .0584 .l91 .465 "
..... ~
.......
.228 .228 .472 . 0182 .0512 .282 • 474
,.:
fielf 1. sec '.300 -:.m-
· 156
· 156
'"
·.361 · 239
. 249
• 48~ .0226
.0265
.0647
.0712
.276 · ~68
.271 , csa
c#
~
1.250 ~
~
.251
· J27
.310
.321
.433 .0347
.0436
.0550
.0698
.312
.365 ·.462 '"
1. 00
· 156 .401l
, .331 .455 .0518 .0834 .360
'"
s
~
.125 .591 .02 6
1.250 .125 , · 3~8 .300 "" ',"
".
'" · 0465 · 1078 · 360 '

'~W
1.750 ~.m .389 .383 .506 .0775
ti&- I. 500 .156 .478 asa · 517 · 0928
.1144
1377
· ~47 ' .542
· «1 I

,
'"
y
,
--'--
,
~0il~
i

, 'r

I , n -11
A T
Table A3.1S

X...Li ~ik-R C
X I'ro,",rllell of
l"Y
I , T ..ble A3.I3
PrOpertin of
6~
1.8e<:tionll

' CJ l-' .:.

Tee~6eC<lOtl3 !
I -r- ,

I ~ ~ ~
:7 ,~!a-j ,
'----"
"T

I,, ,_A~oT

•a
SOMINAL DIMENSIONS SECTION ELEMENTS NOMINAL DIMENSIONS SECTION ELEMENTS
C
z
• , Y I R AREA
, ", "'V< a Y I R AR<A
", 'u '"
"
,~
j Ip'IY A ~
5q1JaI"e W Squate
0 Inches Inch [nches Inches 0 incheS 1JIches"' Inches
Inchell
" " Inches

-+ '"
R=j
L
· 625
· 7sa
, eo 0
· 050
.063
.0'
.063
,125
.
". , · 094
,
· l27
,
.H
.228
,'" · 0028
0055
.0046
.Ol31
· 0011
.0098
or
.0129
· 173
· 209
.209
.296 ,
· 287
ie
ia
-"
, ,H=, .750 L
L '00 '" ~
078
· 07
· 125 '"
· 250
· 289
-,
· 0206
· on9
0284
· 0193
.0235
OHI
. 118
.310
.314
'"
.307
.318
-
~1.
· bO
~
0

'"ec ~
I. 375
375 L '00
r. 00'
on
· 'J94
ias
'"
ne ,
,:H2
aas
.0182
0<, · 0158
O1B8
· 292
.290 ,
294 ats
L eas 094
m
,aa ,
· 0~2
,•
0375
; '"• •.u · 338

H=,
094 240 .327 .0345 .0 ISS · 379
L 2~0
,;0 .063 .lH
, ," .141 .0052 .0169 .192
" I. 000
L 370
l. 625 "
.094 · 125
.341
.388 0676
.0375
0626
HI

!-t: "'" '"

, ". 0244 I. ,,0
· 750 204 · 0080 · 198
aea " .H
•
HI 0725 .078
a
.431

$:1.
., ;00
I. 000
t. 000
t , 250
,078
'J94

I. 250 ; .094
. 125 '"
.228
ua
.252
.225

.3H
,
,0 60
.232 I · J186
· 0301
Q354
'
0200
.0245

.0245
,
,
286
· 371
ate
· 321
.328
· 312
tz-
I::±tJ
~l.
II

13
250
l. J "
l. 02
L B7:;
2. 125
· ,)94
.094
.oec
· 125
H
",
458
50S
.11
1285
1442
cI
.0911
H .,H
516

"530
.321
as
· 4~0
r*- 120 .ca n, .0449 .371 ra
!
, • .b
•
· 531
I.l~
"
"
~
, tea .180 II
, J2ij H' 't I. :50 .125 .5H 2268 .1023
0
.61; "
H;
H!- ! "
azs oJI :' : 1.500 ia 'OS a ,
'" 0407 · 234

"." Ht-
2. COO .025
f.li.-
Ht- I. 62~
I. l25 ,
,
· l25
.878 '"no
.212 .254
314
.0228
· ,; 39
.0263
· o~oa
na
322
rsa [7 , 2. 500 us · 781
,,'
3106
· 1543
305~ 636
· .65
626
L 125
" '" , " H¥s- 3922 1183
..,44 , ,.'"
· ~63

! .094 2,5 .3H .047; 0314 H5 .338 2.000 2.000 · l25 719 .4946
1. J75 ';0 464
, ,
L. 3,
L.1 ,
· l25

"•,
a
.n,
a
,,.0 0,
,
,)40~

sa , '"
0";
.Hl ,

'"sa ,,""* :!.500 2.000 ~ .156 .789

6571
an
.1104
L ua
HI

,
' , , 240 . 17 013l 234 · 40~
"
· 1546
· II 4 L z '"
I~L
'--"-' 2. ~OO
f--iIT- , · 156
; 23 I .67>
,
· 125
894 I
.31l .191 .01M , 0512 230 ,," , aa
3.000
\ira
· ~51 L 262
, a
l5~1 '
,. "
L aa ."
24 , I L 125
'"H"
252- · 0273 J393 .322
"" I. SOO ;H1 I,

~'
, ias , 1,8 , ,- .314 I · .167 '"
740
'0 , · iOO
<0'
~ ".
L .0412 062 · 125 I. l88 · 1184 • L
750 L 315 , ". ,os -, . 333 04 4 0393 HI , .370 2. 000 2, ~24 · l547 i t. se

" ,
L .175
,,25
156
')9.
.45 .3,8
'"
· on: , · 0627 · .03
.500
371
, a " 4.000 3. coo
""
Hm .188 , l. 52 4. 136
L a
S5d6 I L ;;
, ss
550

7'"
,
, , ,
.310
'" , · 'J394
)5l3
· J56
se 3.500
2.00 5. 521
• 368 ,
,,
L 794

"'"
12:; 4.14 .0995 2.60 t , 63d .794
L 625
625 l56
· 405

'" ,," 1200 G6 9 ss r--#- •. 000 • J 13 I 3. ~3 9. U5 a. 037 "ea .aaa

 A3.14 CENTROIDS, CENTER OF GRAVITY, MOMENTS OF INERTIA

Table A3. 16
SXC'1'IOI'f PROPERTIES OF TYPICAL .URCRAFT RX'I'RUDiD SEC1'IOI'fS

I
'Sect. Dimensions ATOll 'rop.,u•••bM x-x' src•• rU••• bou~
i xo. , A 8 I c 't R in. AlI:1s AlI:l.S
I ! I I
'1
I
'2 !
IlI:X ,
I

e lI:X e I I rY e rr e'

, 1 17/32 13/16 I 1/' ,040 ,050 1/16 .068 .00686 ' .317 .319 .00150 .148 ,112
"

,
- ...C;..-..- 2 I liZ 1 ,
, 7/32 ,050 ,050 .050 .113 .01515 ,3665 ,510 ; ,OO15~ .117 ,110

~::-[:
,
, 3 lI2 7/' 1 3/16 1/16 1/16 I 3/32 .1025 i .0100 i ,313 ,
.383 ,00162 .125 ,116

• ,
,
9/16 7/. I 3/16 .060 .060 3/32 .1031 , .0102 ,313 ,383 .00215 I .114 ,130
-, ,I
, , ,

_e_
..l. 'T.

,
-*
5

e I
9/16

3/4
7/'
1-l/8
3/16

3/16
,050

.0751.075
,078 3/32

1/16
.1061 !

.1514 I .0239
.0120

I
.329

.397
, .345

.435
!

I .00550
.00302 .165
,190
,

: .182
,133 ,

I
"
,

,
'-- A---- 7 3/4 1_5/16 9/32 3/32 3/32 9/64 I .230 I .0507 , ,471 ! .574 , .0080 .181 .171
,
, I
,•
7/8 3/8 i
,
1/16 1/16 I 1/16 i .095 , ,0101 .326
,
,4315: .00111 .108 ,
,
,
,108

7/8 3/' . 3/32 3/32 I 3/32 .1375 .0136 .314 .4375: .00148 .104 .118
-r-F I
!

! 10
. '-'l-'
1 ,I lI2 13/32 3/32 ; 3/32 i .1736 i .0240 ; .37l
,
.500 .00376 .147 I .158 I

II 1-1/4 11/16 3/32 3/32 3/32 I ,2320; .0536 I .480 .625 .00992 ,206 ,210
~ I 1'0 ,_'t\ I
,

~
12 1_13/16 11 ,
3/1' 3/16 ·3/32 ."4
,
.2920 I .677 .906 : .0555 .295 .326
_L-L., ,
, .338 ,I
-. 4
C
13 I 2_1/2 1/. I 1/8 illS .581 .525 ! .950 1.25 I, .0662 .323
l -
11 ,

,
14 : 2_1/4 1.1/8 l/S 3/16 ' 1/8 ! .666 ' .524
,
.888 1.125
,
I .0812 .362 I .375 ,

15 1_3/8 , 11/16 1/. 1/16 1/16 I 1/16 : .131 i .00425 .176 .133 .0178 ~ ,360 ,I .687

rr6
r
IS 1_5/8
I i ! .168 I .812
I~r-
3/4 5/32 1/16 1/16 ! 1/16 ,
.00580 1
,
.186 ,133 , .0335 .446

,ee
1 17 1~5/8 3/4 5/32 i 1/16 3/32 ! 5/64 I .206 .00589 I ,170 ,121 I .0458 i .474 i .812
ir+-r- §...-/C IS I 2 1- 1/ 1 6 ,
3/1' ! 3/32 ! 3/32 i 3/32 .306 : .0252 , .287 .201 i ,0191 I .509 1• 0

i~ .L.
1
1

18 ,
2 i 1_1/4
,
3/16 1/8 I 3/32 I ' 3/32 .359 .0475 .364 .300 : ,0821 i .479 ,1. 0
~--"
;-~ -; 2.
I 2 1-5/32 1/4 3/32 I 1/8 i 1/' ~ .378 ! ,0291 : .279 .217 i .1095 i .541 i1. 0 I

21 2_5/16 1-1/2 : 5/16 5/32 I 5/32 ! ' 5/32 .668 , .1075

,
, .401 .3::18 .2326 .590 11.156
,
,

22 I 1_1/2 i 1-l/8
,
1/16 1/16 I ; 1/16 , ,161 .0186 ,
.342 .255 ; ,0171 .326
I .750
j"-:-J:' 23 1-1/2 2 I l/S I .384 .1562 .641 .658 ,0346 .302 ,750

Tf"'-f~T
3/32 i 1/8 I i ,

,. ,
I 1_3/4
1

2
,
9/32 1/. ; 1/8 .760 .315
I .643 .757
,
I ,0599 .281
, ,875
I'a
! 1 I
;
I as
'C
I, 1-3/4 2_9/16 : 1/4 .5/32 I ; 5(32 .918
,
.600 I .809 .959 , .0730 .282 , .815 I
I
i.,.e.'J, ,,
26
!2 i 2_1/2 ! 5/32 1/. 5/32 .605 ! .385
; .800 .793 .159 .514 i 1. 00 i

27 11/16 1 l/S .050 .050 ; .050 3/32 .135 .0218 , .404 ,500 .0163 .348 • 6625!
,..A-! I i

28 I 13/16 1 1_ 1/ . 3/16 .060: .060 i .050 3/32 .180 I, .0315 ! .456 .562 .0295 .405 .78251
i--r~
!

, .595
I'8 - - ' -e - T 28
! 5/S 1~~/2 13/16

! 1/4
.050 1/16 1/16 ; 3/32 .166 , .0603

,, .793
.75 .0151 .298 .600

I ·'r·''.,-J
,
3. 3/4 2 .050 1/16 : 1/16 3/32 .214 .1382 1.00 .0281 .358 ,725

I..L- I ~ ,L 'I 31 i 3/4 2 i 1/4 1/16 1/16 : 1/16 3/32 I .2378 i .14.49 i .773 i 1.00 ; .0282 .341 .719

't-e'..I 132 , 3/4 ,2 11/4 1 1/ 1 6 3/32 ! 3/32 3/32 .2855 ! .1820 i .795 1.00 .oaes .356 .719 i

: 33 I '/1' : 7/8 I .050 .050 i ,1/16 I .099 ! .0120 I .349 : .437 , .0056 .238 .531

...
:
'--A ;:I
.~,
1.,,10-
" e
I 34
35

38
I
1/2

9/16
9/16
I
I1
,1 1/4
7/8 : 1/16

1/16

.ove
1/16
.080
.075
I
i
,
1/16

' 3/32

3/321.170
.115

.151
,, .0239 ,!, .397
:,

I
.0135

.0384
.341

, .475
:

,
.437

.500

.625
I .00446
I

I
,00828

.00876
.197

.234

I .227
.469

.531 I

.525 1

a ...- -~ 31 11/16 1
1 _ 1/ 4 1/18 .100 3/32 : .213 : • 05~0 I .508 .625 I .0194 .302 .655

I
L'" e
LII~J-
I

I ~
38

3.
,, .825

3/4
1.1/2

2
i
i
.075

.072
.120 I
i .012 ,
3/32 : .302

1/' : .249
I .110
.1446
,
: .604
,, .757 i
.750

1.00'
I .0391

I .0180
.360

.267
I
,787

.714
I

,,
i ' 4. , 3/4 2 , .102i· l 0 2 i 3/32 .330 .1826 .750 1.00 I .0228 .265 .700
 CHAHT A3. 1 PHOPEHTIES OF FOHMED CHANNEL SECTIONS

1.0

0.35 .1 illi J!:..I.IIIJ I ;~I LIIIIII :'. Iii J 111m II ! iii 11 1

1P:llili/lli il1mHW Imll! Ii!1~IWilf 'I«Wi' Ii 1II1Iill' ~1.ljjj1!l!!I!lIljjllHmmnn ~
jli

.. ., ,.
w JIi I'
~ .1

11:1Mo.3O ..
,:1: i I 1, ,.r-
Z
.. , ! II
1
...
~ ., II ., ~
~
.'_, i: , . , I'I
:'1' ;':. :1
"" ~
~ '+, • • i: 'i~
I .:: I T1 ::1. ,i 1
,.
Z
0.25 F.itt0'l!I: i1ii !!II:1 !l~i :II!: ill~~' ;:11 ~ -1" I-!- I~; Iiii1: ill:i : IilliIiII: 1111~1 illIlllllli! ~ '1II: i III Ll!: I!il;:~
.. , I
tl
tl
'"
~
~

030 i
~
~
~
~
~
"z
.,
0

.,
0
Ii
~
0
Iiw -.,"
r-

0,15 ~ ~
0
~ '"
<:
'"
'"
~

0
r-

0.10 1";'I'~!I!"!Im!I'!'T"!lIIWI'III'II'I~j0P;IPI'':¥'IU;;ni'W!UUwnUI'~p!!1
till r._. 1m .. , .m itT. !I. I:-p. Itt! ....... " ", ." .... ','.". "r. " I " . . . " "" . " . "
'".,
I~

'0c"t
.,
0.05 1II111l1\l!llHllllllllml!l!!Yf111 IY!!'III1i~:iIIlIJl!lIiHIM1'I!l"J!ml!!! 'III ::1 !I,'I II! II i!!!l1i 111:11 ·',IIII! I: !I .,
"
I~
~
.'-c.
'.

~.
1.0 1.0 2.0 2.5 3.0
Mom. ollnarUa.:-

I
,
I;
II - IIEIGIiT OF CHANNEL .025
U
II; HEIGIIT OF CHANNEL
3.5
•
V~
\1 3
A3.16 CENTROIDS, CENTER OF GRAVITY, MOMENTS OF INERTIA

A3.15 Problems (5) ~or the ~ean Beetien in ~i~. ~3.1S,

calculate the ~oment oi inert~a ~bo~t t~e ;r~~­
ci~al ~xes ~ssu~i~g :~e fJUY st~i~;~rs ~s t~e
only effective ~aterial.

-ll-/.,
l:~;; l.- 4-- 8 ..;... ...+
Ir..
I '
...
IS I. ,
Il"
' I
10

J I·'.T L
'i)t. "
0·8
., i
Ii /0_ I zo·'
~~t•
I-- z" -.J ~ Fig. A3.19

,,..
I
o.? Od~
Fig. A3.14 Fig. A3.15 1'- t +• 15
1"8 i
(1) For the section of Fig. A3.14 determine
the moment of inertia about each of the or-mer- (6) F~~. A3.19 shows a wing 0e~~ section
pal axes Xp and Zp. with a cut-out on the lcwe~ surface. Jeter~ine
(2) Calculate the moment of inertia of the the ~oments of inertia about the ~r1~:i~~1 axes
section in Fig. A3.l5 about the ?rincipal axes. assuming the eight stringers are the cn:y ef~2c­
t ive :naterial.

It ~ l : D~
I 5 spaces @5'" 35" I F,g. A3.20

Fig. A3. 17
(7) Fi~. A3.20 shews a 3 cell ~ulti,le
(3) Fig. A3.l6 illustrates a box type beam flan?e -beam. !he 7 flan~e T.embers on t~e upper
section with six longitudirral stringers. De- face of beams have ~~ area of .3 sq. in. each
termine the ~oment of inertia of the beam sec- and those on the bottom skin 0.2 sq. in. each.
tion about the principal axes tor the follow- The bottom skin is .03 inches in thickness.
ing assumptions:- Compute the moments of inertia about the
(a) Assume the beam is bending upward principle axes assuming that the flange :nembers
putting the top portion in compression and the and the bottom skin comprise tr.e effecti~e
lower portion in tension. Therefore, neglect material.
sheet on the top side since it has very little
resistance to compressive stresses. The sheet
on the bottom side is effective since it is in
tension. For simplicity neglect the ver~ical
webs in the calculations.
(b) Reverse the conditions in (a) thus \
placing top side in tension and lower side in \_Cutout [or door
compression.
(4) For the three stringer single cell boX
beam sectien in Fig. A3.17, calculate the mo- Fig. A3.21
ments of inertia about the principal axes. As-
sume all 'Neb or wall material ineffective.

(8) Fig. A3.21 shows the cross-section of

.. a small fuselage. T~e dashed line represe~ts
S a cut-out in the structure due to a Goor. As-
( Fig. A3.18
sume each of the 13 stringers have ~~ area of
0.1 sq. in. Consider fuselage skin ineffec::ve.
Calculate the :noment of inertia of the effective
r--- /9 ·'
section about the principal ~es.
 CHAPTER A4
GENERAL LOADS ON AIRCRAFT

A4. 1 Introduction. engers safely, efficiently and comfortably over

Betore the structural design or an airplane various distances between airports. on the
can be made, the external loads acting on the other hand the Air Force Fighter type of air-
airplane in flight, landing and take-ott con- craft has a job at shooting down enemy aircraft
ditions must be known. The complete determin- or protecting slower friendly aircratt. To do
ation ot the air loads on an aIrplane requires a this job efficiently requires a far difterent
thorough theoretical knowledge or aerodynamics, contiguration as compared to the DC-8 transport.
since modern aircraft tly in sub-sonic, trans- Furthermore the Fighter type airplane must be
sonIc and super-sontc speed ranges. further- maneuvered far mere sharply to do its required
more, there Is a wide range ot Wing configur- job as compared to the DC-8 in do Ing its re-
atIons, such as the straight tapered wing, the qurr-ed job.
swept wing and the delta Wing, and many at In general the magnitude of the air forces
these wings otten include leading and trailing on an airplane depend on the velocity at the
edge devices tor promoting better 11ft or con- airplane and the rate at which this velOCity Is
trol characteristics. The presence ot power changed in magnitude and direction (acceleration).
plant nacelle unt ts J external fuel tanka, etc. The magnitude of the flight acceleration factor
are units that etrect the airflow around the may be governed by the capacity at the human
wing and thus effect the magnitude and distri- body to Withstand these acceleration inertia
bution at the air torces on the Wing. Likewise, torces without injury which is the situation in
the fuselage or airplane body itself influences a tighter type at airplane. on the other hand
the airflow over--the wing. The theoretical cal- the maneuvering accelerations for the DC-8 are
culation ot the alrloads on the airplane 1s too not dictated by what the human body can with-
large a subject to be covered in a structures stand, but are determined by What 15 necessary
book and it is customary in college aeronautical to sately transport passengers trom one airport
curricula to provide a separate course for this to another.
subject. Designing the airplane structure tor loads
In most airplane companies the loads on greater than the airplanes sutters in the per-
the airplane are determined by a group at en- formance of its required job, obviOUSly Will add
gineers aSSigned to the Structures AnalysiS considerable weight to the airplane and decrease
Section and this group is otten reterred to as its pertormance or over-all efficiency relative
the Aircraft Load Calculation group. While the to the job it is designed to do.
work of this group is primarily based on the Te particularly insure safety in the alr-
use at aerodynamics, it is that phase at aero- transportation, along with uniformity and ef-
dynamics which is conserved with determining ficiency of design, the government aeronautical
the magnitude and distribution at the air loads agencies (civil and military) bave definite re-
on the airplane so that the airplane structure qUirements tor the various types of aircraft
can be properly deSigned to support these air relative to the magnitUde of loads to be used in
forces safely and efficiently. The engineering the structural deafgn of aircraft. In referring
department at an airplane company has a distinct in general to these specified aircraft loads two
or separate aerodynamics section, but in general terms are used as follows:-
their responsibility is the use at the SUbject
at aer-odynamrcs, to insure or guarantee the per- Limit or Applied Loads.
tormance, stability and control at the airplane. The terms limit and applied refer to the
A basic general over-all knowledge at the same loads with the civil agencies (C.A.A.)
loads on aircratt is desirable in the study at USing the term limit and the military agenc i es
aircraft structural theory, and hence thiS USing the te~ applied.
chapter attempts to give this information. In Lim1t loads are the maximum loads antici-
a later Chapter dealing with wing design, this pated on the airplane during its lifetune of
Subject will be further expanded. service.
The airplane structure shall be capable or
A4. t Limit or Applied Loads. Design Loads. supporting the limit loads withuut suffering
Because an airplane Is des Igned to carry detrimental permanent deformations. At all loads
out a definite job, there result many types or up to the limit loads the deformation of the
aircraft relative to size, conf1guration and structure shall be such as not to intertere with
performance. For example, a commercial trans- the sate operation of the airplane.
port like the Douglas DC-8 is designed to do a Ultimate or DeSign Loads.
jab of transporting a certain number of pass- These 17NO terms are used in general to mean
A4.1

>4
A4.2 GENERAL LOADS
the same thing. Ult~mate or Design Loads are
equal to the limit :oads ~ultiylled by a ~actcr
of safety (F.S.) or
Design Loads = Li~it or Applied Loads times F.S.
In general the oV9r-all factor of safety is
1.5. ~he goverr~ent ~eQuirements also specify
that these design loads be carried by the
structure without failure.
Although aircraft are not s~?pOSed to
undergo greater loads than the specified 11~lt
loads, a certaln amount of reserve strength
against complete structural tailure of a unit is
necessary in the design of practi8ally any ~­
chine or structure. This 1s due to many factors
such as:- (1) The approximations involved in
aerodynamic theory and also structural stress
analySiS theorYj (2) Variation in physical
properties of materials; (3) Variation in fab-
rication and inspection standards. POSSibly
the most important reason for the factors of
safety for airplanes is due to the fact that
practically every airplane is limited to the
maxtmum velocity it can be flown and the maxi-
mum acceleration it can be SUbjected to in
flight or landing. Since these are under the
control of the pilot it 1s poss~ble in emerg-
ency conditions that the limit loads ~y be
slightly exceeded but with a reserve factor at
safety against failure this exceeding of ths
11mit load should not prove serious from an
airplane safety standpoint, although it might
cause permanent structural deformations that
might require repair or replacements of small
units or portions ot the structure.
Loads due to airplane gusts, are arbitrary
in that the gust velocity 1s assumed. Al-
though this ~~st velocity is based on years of
experience in measuring and recording gust
forces in flight allover the world, it is quite
POSSible that during the lifetime of an air-
plane, turbulent conditions near storm areas or
over mountains or water areas might produce air
gust velocities sli~~tly greater than that
speCified in the load requirements, thus the
tact or of safety insures safety against failure
1: this situation would arise.
The broad general category of external
loads on conventional aircraft can be broken
down into such classifications as follows:-
Due to Airplane Maneuvers. (under
the control of the ~ilot).
(1) A1r Loads
{ Due to Air Gusts. (not under
control of pilot).
Landi ng on Land. (Wheel or
l
sk1 type).
(2 ) Landing.Loads Landing on ~~ter.
Arresting. {Landing on Air-
craft Carriers).
ON AIRCRAFT
Thrust .
(3) ?cNer Plant Loads { Torque.
(4) Take ott
HOi s t i ng Airplane.
, 1 1 I "
(5 ) upee a ....oa"'s Towim:: Atr-ul.ane.
- ~ " 1" I. <.,.. .... ~ "
{ Beaching of nu_l~y~e ~~~~l~r.~
Fuselage ?reSsurlzi~g.
(5) Wei~~t and Inert1a Leads.
In resolving external loads :or stress
analysis purposes, it is convenient to ~ave a
set of reference axes. ~he rs!erer-ce axes
XYZ paSSing throug~ the center o~ ~ravity of the
air?lane as illustrated in Fig. A4.0 are these
no~lly used in stress analYSiS work as well as
for aerodynamic calculatior.s. Fer ccnve .. ience
the reference axes are often referred to another
origin other than the air;lane ~.~.
z
y
x
x
y
Z
Fig. A4.0
A4.4 Weight and Inertia Forces.
The term weight is that constant force, ,ro-
portional ~o its mass, which tends to draw every
phYSical body toward the center of the earth.
.~ airplane in steady flight (uniform velocity)
is aoted upon by a system of forces in equilib-
rium, namely. the weight of the airplane, ~he air
forces on the complete airplane, and the power
plant forces. The pilot can change this bal-
anced steady flight condition ~y cr~ging the
engine power or by operating the surface controls
to change the direction of the airplane velocity.
These unbalanced forces thus cause the airplane
to accelerate or de-accelerate.
Inertia Forces For ~ot:on of ~Jre Translation
cf Rigid 80dles.
If the unbalanced forces acting on a rigid
body cause only a change in the magnitude of the
velocity of the body, but not its direction, the
motion is called tranSlation, and from basic
Physics, the accelerating ~orce F = ~, where M
is the ~ss of the bOdy or ~/g. In Fig. A4.1
the unbalanced force system F causes the rigid
body to accelerate to the rl~ht. Fig. A4.2 shews
the effect of this unbalanced ferce in ;roduclng
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A4.3

F :: unbalanced Effective Force

at = ra, and a =
n
rw
lf an acceleration normal to
external force =1:ma=Ma the flight path at A and directed toward the
/mla/ ma , center ot rotation (0). From Newton's Law the
Motion
-- -- -- ---- Motion
~
effective forces due to these accelerations
are:-
F = !".rwlf = Mv /T _
T
w
-- -- -- -- -- n
2

- -- - -
\
tw where w = angular velOCity at the paint A.
a angular acceleration at paint A.
Fig. A4. 1 Fig. A4.2 r radius at Cllrrnture at :light path
at point A.
The inertia forces are equal and OPPOSite
a force on each mass particle of mLa, m~a, etc., to these etfective forces as indicated in Fig.
thus the total effective force is zma = Ma. I f A4.3. These inertia forces can then be con-
these effective forces are reversed they are re- sidered as part at the total force system on the
terred to as inertia torces. The external airplane which is in equilibrium.
forces and tha inertia forces therefore torm a It the velocity or the airplane along the
force system in equilibrium. path 15 constant, then at = zero and thus the
From basic Physics, we have the following
inertia force F = 0, leaVing only the normal
relationships tor a motion of pure translation t
it the acceleration is constant:- inertia torce Fn•
(1) It the angular acceleration Is constant,
v - Va= at - -
lf
the following relationships hold.
s ve t -+- iat - - (2 )
v" _ va If = 2as W - WQ= at - - (6)
(3 )
Q = wot ... ~a.tlf (7 )
where, wa _ W o If = ZaQ (8)
5 = distance
moved in time t ,
where Q = angle of rotation in time t.
Vo = initial velocity w = initial angular velocity in rad/sec.
v = tinal velocity atter time t. w = angular velocity after tune t.
In Fig. A4.3 the moment To or the inertia
Inertia Forces on Rotating Rigid 80dies. forces about the center of rotation (0) equals
A common airplane maneuver is a motion I'tra(r)= MTaa. The term MT a is the mass moment
along a curved path in a plane parallel to the pf inertia of the airplane about point (0).
XZ plane of the airplane, and generally referred Since an airplane has considerable pitching
to as the pitching plane. A pull up trom steady moment of inertia about its own center ot gravity
flight or a pullout from a dive causes an air- axiS, it should be included. Thus by the
plane to follow a curved path. Fig. A4.3 shows parallel axis
an airplane following a :urved path. It at
point A the velOCity is increasing along its To = loa -+- I e.g a - - - - - - - - - - - - - (9)
path, the airplane 1s being subjected to two where I = Mfa and I e.g = moment or inertia of
Q
accelerations, n~elY, at' tangential to the
curve at point A and equal in magnitude to airplane about Y axis through e.g. at airplane.
Inertia F0rces For Pitching Rotation of Airplane
Center of Curvature about Y Axis Through e.g. airplane.
01 Flight
Path
r In flight, an air ~~st may strike the hori-
I zontal tail prodUCing a tail torc~ which has a
I moment about the airplane c.g. In some landing
an=rwlf=Vlf / ? conditions the ground or water forces do not
pass t~xough the airplane e.g., thus producing
=E2~~-MTa a moment about the airplane e.g. These moments
AI '.g. cause the airplane to rotate about the Y axis
M? w lf = Mv a / ? through the e.g.
Therefore for this etfect alone the center
Fig. A4.3 or rotation in Fig. A4.3 is net at (0) but at
A4.4 GENERAL
the e.g. at airplane, or r o. Thus F and F
n t
equal zero and thus the only inertia ~~rce tor
the pure rotation is I g C, (a couple) and
c. .
thus the moment of this inertia couple about the
e.g. = Tc. g• = Ic.g.~'
As explained before if the inertia forces
are included with all other applied forces on
the airplane, then the airplane is in static
equilibrium and the problem is handled by the
static equations for equilibrium.
A4.5 Air Forces on Wing.
The wing at an airplane carries the major
portion of the air forces. In level steady
flight the vertical upward torce of the air on
the wing, practically equals the weight of the
airplane. The term airfoil is used when re-
terring to the Shape of the cross-section at a
wing. Figs. A4.4 and A4.S illustrate the air
pressure intensity diagram due to an air-
Angle of Attack
• 12"
Angle of Attack
'" _ 60
Fig. A4.o4 Fig. A4.5
stream flowing ar-ound an airfoil shape tor 'both
a positive and negative angle of attack. The
shape and intensity at this diagram is in-
fluenced by many factors, such as the Shape of
the airfoil itself, as the thiokness to chord
ratio, the camber of the top and bottom sur-
faces etc. A normal wing is attached to a
fuselage and it may support external power
plants, wing tip tanks etc. Furthermore the
normal wing is usually tapered in planform
and thickness and may possess leading and
trailing slots and flaps to prOduce high lift
or control etfects. The airflow around the
wing is affected by such factors as listed
above and thus wind tunnel tests are usually
necessary to obtain a true picture of the air
torces on a Wing relative to their chordwise
spanwtse distribution.
Resultant Air Force. Center of Pressure.
It is convenient when dealing with the
balancing or equilibrium of the airplane as a
whole, to deal with the reSUltant of the total
LOADS ON AIRCRAFT
air fOrC(3 on the wing. For example, consider
the two air pressure intensity diagr~~ in ?lgs.
A4.6 and A4.7. ~hese distributed ~orce systems
can be replaced by their resultant (R), which
of course must be known in magnitude, direction
and location. ~~e location is specified by a
term called the center of pressure which is ~he
point where the resultant R intersects the a1r-
foil chord line. As the angle at attack is
Changed the resultant air force Changes in mag-
nitUde, direction and center of pressure
location.
R
Fig. A4.6 Fig. A4.7
Lift and Drag Components of Resultant Air Force.
Instead ot dealing with the resultant ferce
R, it is convenient for both aerodynamiC and
stress analysis considerations to replace the
resultant by its two components perpendicular
and parallel to the airstream. Fig. A4.8 il-
lustrates this resolution into lift and drag
components.
:
~
..
1 rAUlODTN..... IC
I : ,J Cl1:NTEII
o.p '0
'UGHT DlllltCTIOlI
'!'?5?:>
FLIGHT DllIECTIOIII
Fig. A4.8 Fig. A4.9
Aerodynamic Center (a.c. ). Since an air-
plane tlies at ~y different angles of attack,
it means that the center ot pressure Changes for
the many flight design conditions. It so hap-
pens, that there is one paint on the airfoil
that the moment due to the Lift and Drag torces
is constant ~or any angle ot attack. This
point 1s called the aerodynamic center (a.c.)
and its apprOXimate location 1s at the 2S percent
of chord measured from the leading edge. ~hus
the resultant R can be replaced by a 11ft and
drag force at the aerOdynamic center plus a wing
moment Ma.c. as illustrated :n Fig. A4.9.
A4.8 Forces on Airplane in Flight.
Fig. A4.l0 illustrates in general the main
forces on the airplane in an accelerated flight
condition.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A4.5

z tiplying factor by which the forces on the air-

plane in steady flight are multiplied to obtain
r/o-- e
.., a static system ot forces equivalent to the dy-
E namic force system acting during the accelera-
tion of the airplane. Fig. A4.ll illustrates
Z
LA1 total lift (Wing &l Tail)

Fig. A4.1O • 0
T_

Fig. A4.11
W
T =engine thrust. forces in steady horizontal flight. L repre-
L = total wing lift plus fuselage lift.
sents the total airplane lift (Wing plus tail).
D = total airplane drag.
=
Ma moment of Land D with reference to wing Therefore L = W. Now assume the airplane is ac-
a.c. (aerodynamic center) celerated upward along the Z axis. Fig. A4.12
shows the additional inertia torce wag/g acting
W = weight of airplane.
I L = inertia force normal to flight path. downward, or opposite to the direction ot
accaleration. The total airplane lift L tor the
In inertia force parallel to flight path. Z
I
1
m rotation inertia moment. ",L
E tail load normal to flight path.
For a hortzontal Constant velocity flight
condition, the inertia forces I
L, 1D, and 1m "...- e.g.• - 0
T
would be zero. For an accelerated flight con-
dition involving translation but not angular w
Fig. A4.12
acceleration about its own e.g. axis, the
inertia moment 1 would be zero, but I and 1 hg.•,
m L D
unaccelerated condition tn Fig. A4.ll must be
would have values.
mUltiplied by a load factor n to produce statiC
a
Equations of Equilibrium For Steady Flight. equilibrium in the ~ direction.
From Fig. A4.10 we can wrlte:-
Thus, n L - W - ~ e., ::; 0
a g a
l:F D + W sin Q-T cos ~
x 0,
° Since L =W
ZF ~.= 0, L
- '" cos Q+T sin ~ - E =° Hence n ::; 1 + a~
~
ZMy = 0, -l1a- La - Db + Tc cos ~ + te = a g

An airplane can at course be accelerated along

Equations of Equilibrium. in Accelerated Flight. the X axis as well as the ~ axis. Thus in
l:F 0,
x D sin Q - T cos ~
+ '"
I - D °
Fig. A4.l3 the magnitude ot the engine thrust T
is greater than the airplane Drag D, which
..L
ZFa = 0, L
- '" cos Q + T sin ~ - IL - E =°
0, - Ma - La - Db + Tc cos ~ + Ee + Im =0
r Forcee - Plus ts UD and toward tail. ~.g.

Signs used: lMoment - C'Iockwf se is positive. w

ng. A4.13 T is greater than D
Distances tram e.g. to force -
Plus is up and toward tail. ~waz
g
A4.7 Load Factors. causes the airplane to accelerate forward. It is
The term load factor normally given the convenient to express the inertia force in the X
Symbol (n) can~dettned as the numerical mu1- direction in terms of the load factor n and the
x
• The bar through letter Z has no significance. Same mean-
ing without bar.

A4.6 GENERAL LOADS
weight W of the airplane J hence
'~
n W ax (See Fig. Al4.l3)
x g
OJ whence T-D-n xw o A4. 9 Gust Load Factors.
T-O
Hence n
x W
Therefore the loads on the airplane c~~ be dis-
cussed in terms of load factors. The applied or
limit load factors are the ~axi~~T. load ~actors
that might occur d~ing the ser7ic3 of "the ?ar-
ticular ai~plane. These loads as jisc~ssed in
Art. A4.2 ~us~ ~e taken by t~e ~ir?lane s:r~c­
ture without appreCiable per.Tanent deformation.
The design load factors are equal to the
limit load factors multiplied by the factor of
safety. and these deSign loads must be carried
by the structuta without rupture or collapse J or
in other words J ~ornplete failure.
A4.8 Design Flight Requirements for Airplane.
The Civil and Military Aeronautics Author-
ities issue requirements whic~ specify the
design conditions for ~he various class1ficat1on
of airplanes. Generally speaking J any air~lane
flight altitude can be defined by stating the
existing values of load factors (acceleration)
and the airspeed (or mere properly the d}~ic
pressure) •
The accelerations en the airplane are pro-
duced fram ~NO causes J nanelYJ ~~euvers and air
gusts. The accelerations due to maneuvers are
subject to the control of the pilot who can
manipUlate the controls so as not to exceed a
certain acceleration. In hi&~ly maneuverable
military airplanes J an accelerometer is in-
cluded in the cockpit instruments as a ~ide to
limit the acceleration factor. ror commercial
airplanes the maneuver factors are ~de high
enough to safely take care or any maneuvers tr~t
would be reqUired in the necessary fl~ght opera-
tions of the particular type of airplane. ~hese
limiting maneuver factors are based on years of
operating experience and have given satisfactory
results fram a sarety standpoint without pen-'
alizing the airplane from a weight deSign con-
sideration.
The accelerations due to the airplane
striking an air gust are not under the control
of the pilot since it depends on the direction
and velocity of the air gust. From much ac-
cumulated data obtained by instal:ing accelero-
meters in commercial and military aircraft and
flying them in all types of weather and loca-
tions. it has been found that a gust velocity or
30 ft. per second appears sufficient.
The speed or velocity of the airplane like-
wise effects the loads on the ai~plane. The
higher the velOCity the higher the aerodyna~ic
Wing ~oment. Furthermore the gust acceler-
ations increase with airplane velocitYJ thus it
is custcmary to limit the partic~lar ai~plane to
-.Zq
ON AIRCRAFT
a defin~te ~ax~~w~ r:~;~: ,eloc~ty :ar ~om­
mer-c Ia l airplanes the ve Loc i t y is imi t.ec to 2.
reasonable ~lide speed whic~ ~s su ficient to
take care of ~easo~a8le flis~t opera~icns.
~~en a s~a~, edge ;ust strikes :~e ~ir?l2.ne
in a di~ection normal to the thrust line (X
axis)J a sudden change ta~es place in :~e wing
angle of attack with no sudden change in ~ir)la~2
sDeed. The cOrTal force coeffic~ent (C~ ) can
. -,"
be ass~ed to vary linearly wl':h :he a~gle of
attack. Thus in rig. (a)J let ?Ol~t (3) repre-
sent the nomal air?lane :orce ccef:icient C7
'-')..
necessary to ~inta~n level ~l~sh:
velOCity V and point (s) the 'ffilue Softer
a sharp edged ~ust KU has caused a s~dden change
6a in the a~gle of attack wit~out c~ange in V.
The total lr.crease in the airpl&~e load in ~he Z
direction can therefore be expressed jy the
ratio C7 at 8.
~,
~.
From ?!g. (b) for small angles, 0" KU~l
and from Fig. (a) J 6C = m 6.0., wnere cn the
ZA
Slope of the airplane no~~l force curve (C 7
~A
per r-anian) .
KU~
V
Fig. b Fig. a
Tr4e load factor increment jue to the gust
KU can then be expressed
KUVS
m (A)
575 ',.,[
U gust velOCity in tt./sec.
K*= gust correction factor depending on
wing loading (Cur7es for K are prOVided
by Civil Aeronautics Authorities).
V indicated air speed in miles per hour.
S wing area in sq. ft.
W gross weight of airpla~e.
• NACA Technical Note 2964 (June 1953), proposes that the
alleviation factor K should be replaced by a gust factor,
Kg :: 0.88 Mg/(5. 3 + Mg}. In this expression Mg is the
airplane mass ratio or mass parameter, 2 W!o.pcgs. In
which c is the mean geometric chord in feet andg the ac-
celeration due to gravity.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
It U is taken as 30 ft./sec. and ~ as the
c~ange in C wit~ respect to angle of attack
zA
in absolute ~~its per degree, equation (A) re-
duces to the following
on • ~V _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Therefore the gust load factor n when air-
plane is flying in horizontal altitude equals
n ;: 1 ... 3KmV
- I,.J/S
and when airplane 1s in a vertical altitude
3KmV
n •• - - - - - - - - - - - - - - - - -
- ',,/s
A4.10 illustration of Main Flight. Conditions.
Velocity-Load Factor Diagram.
As indicated before the main design flight
conditions for an airplane can be given by
stating the limiting values of the acceleration
and speed ~nd in addition the maximum value of
the applied gust velocity. As an illustration,
the design loading requirements for a certain
airplane could be stated as follows: ~The
proposed airplane shall be designed !or applied
positive and negative accelerations of + 5.0g
and -3.5g respectively at all speedS tram that
corresponding to CL up ~o 1.4 times the
max•
maximum level flight speed. Further.TIore, the
airplane shall Nlthstand any applied leads due
to a 30 ~t./sec. gust actl~g in any direction
up to the restricted speed of 1.4 times the
maximum level flight speed. A design factor of
safety of 1.5 shall be used on these applied
loads" .
In graphical form these design require-
ments can be represented by plotting load fac-
tor and velocity to obtain a diagram which is
generally referred to as the Velocity-accelera-
tion diagram. The results of the above speci-
fication would be similar to that of Fig. A4.14.
Thus, the lines AS and CD represent the re-
stricted pOSitive and negative ~euver load
factors which are limited to speeds inside line
BD "Nhlch is taken as 1.4 times the maxtmum
level fli~~t speed in this illustration. These
restricted maneuver lines are ~er.TIinated at
points A and C by their intersecticn With the
:nax1m.um C values of the a ir-cjcne • At speeds
L
beCNeen A and B, the pilot ~ust be careful not
to exceed the ~neuver accelerations, since in
general, it would be posSible tor him to ~n­
ipu1ate the controls to exceed these values.
At speeds beloW A and C, there need be no care
or the pilot as far as loads on the airplane
are concer~ed since a ~neuver prodUCing C~.
A4,7
would gl ve an acceleration less than the lim-
ited values given by l1nes AB and CD.
The positive ,and negative gust accelera-
tions due to a 30 ft./sec. gust normal to f:i~ht
path are shown on Fig. A4.l4. In this example
diagram, a positive !list is not critical within
(E) the restrictAd velOCity of the air?lane since
the 6Ust lineS intersect the line BD below the
line AB. For a negative gust, the zust load
tactor becomes critical at velocities between F
and D with a maximum acceleration as given by
paint E.
(e) For airplanes 'Nhich have a relatively low
required maneuver factor the IUst accelerations
may be critical for both pOSitive and negative
accelerations. ~ination of the 3Ust equation
indicates that the most lightly loaded condition
(smallest gross weight) prOduces the highest
(D)
gust lead factor, thus inVOlving only partial
pay load, fuel, etc.
On the dia~am, ~he points A and 3 corre-
spond in general to what is referred to as high
angle ot attack (H,A,A,) and low angle of attack
(L.A.A.) respectively, and points C and 0 the
inverted (H.A.A.) anrt (L.A.A.+ conditions re-
spectively.
Generally speaking, if the airplane is de-
Signed for the air lOBds produced by the veloc-
ity and acceleration conditions ~t paints A, 8,
E, F, and C, it should be sate from a structural
strength standpoint if flown Within the speCified
limits regarding velocity and acceleration.
Basically, the flight condition reqUire-
ments of the Civil Aeronautics Authority, Army,
and Navy are based on consideration ot specified
velocities and accelerations and a consideration
ot gusts, thus a stUdent understanding the basic
discussion above should have no diffiCUlty un-
derstanding the deSign reqUirements of these
three government agenCies.
For stress analySiS ~urposes, all speeds
are expressed as indicated air speeds. The
"indicated~ air speed 1S detined as the speed
which would be indicated by a ~erfect air-speed
indicator, that is, one that would indicate
true air speed at sea level under standard at-
mospheric conditions. The relation between the
actual air speed Va and the indicated air speed
Vi is given by the equation
VI .~ Va
where
Vi indicated airspeed
Va actual airspeed
Po = standard air denslty at sea level
Pa ;: density of air in which Va is attained
A4.8 GENERAL LOADS ON AIRCRAFT

the loads on the wing should be checked for

cases where the engines are attac~ed to the wing
and ~e located ~orward of the leading edge.
';''1'''1'''''' I,· I' ,,""' In cases where the land1ng gear is attac~ed
;'1',:: ,.. !' ',,;, It;l"" to wing or when the :uel and engines are carried
~ 1Cl:T in and on the wing, the loads produced on the
, ,~~. wing structure ~n a landing condition nay be
critical for some portions of the wing structure
L ;" 1,/ '''.l'''l:;;'~ I,; D."" inboard of landing gear and engine attac~~.ent
I J" L [,,, 'Qi h~RT [,,''I Y1"·I'; ", ~ points.
'.__ i!i;Wf~ :~iI;; h'::+:; ',:}J:.- l>i' \,,; i:,=: :E,
~ ',. :~:~·~tV:; .:._:~ ·':~t$.~ r;: r': :~;::.: A4. 12 Example Problems Invcivtng Accelerated Motion of
E "'T <: ...
i.l '1;;, ","" ,f'" <L L ~1:-" Rigid Airplane.
t ~:~ ;~"L~ :;.t.r' : ._::: ,:::~::::~:~T:: U~~:: i::: k .::;~: As previously explained, it is general

i!~;~~~~;i:;'"lli~~,~;["!
practice to place the airplane under accelerated
conditions of motion into a condition of static
equilibrium by adding the inertia farces to the
applied force system acting on the ai~plane. It
is usually assumed that the airplane is a r~gid
body. Several example problems will be pre-
sented to illustrate this general proceaure.

~ ~rr: ~~~;U ~~ 2; ;lli::" '~ii~l, "'?'. ,;S I , Example Problem 1

Fig. A4.l5 ill~strates an airplane landing
~~~~~~ ~~imI; ~il:I~~ ~~;t~~~~~~~:~: '~~J:~A:EfSS:! H:' on a Navy aircraft carrier and Jeing arrested by
a cable pUll T en the airplane arresting tcok.
\~~;~ i~~~~ ~~i~ m~;~~~~g~~~;g~ i;~:!~mlli~~lli~tiTIHr~ li::~i It the airplane-weight is 12,000 Ibs. ar.d the
airplane is given a constant acceleration of 3.5g
:1',::[" ""W i :~I"" :!i:lFi IeJ!iijjJijjjij'iitljijrJlic;'J' Fi~:t4:' I.' (112.7 ft/sec~), ~lnd the hook pull T, the wheel
A4.11 Special Flight Design Conditions. reaction R, and the distance (d) be~Neen the line
of action of the hook pUll and the airplane e.g.
There are $anY other rli~~t conditions If the landing velocity is 60 ~.P.H. what is the
'which may be critical ror certain portions at
stopping distance.
the wi~ or fuselage structure. Most airplanes
are equipped with flaps, to decrease the land- w" 12000 lb.
ing speed and such flaps are lowered at speeds
at least twice that or the m1n1mUm landing
speed. Since the flapped airfoil has ditferent
values for the magnitude and location of the
airro1l Characteristics, the Wing structure
Lx
must be checked for all possible flap conditions T
within the specified requirement relative to 24"
maximUm speed at which the flaps may be oper-
R Fig. A4. 15
ated. Generally speaking, the flap conditions
wtll efrect only the wing portion inboard or the
flap and it is usually only critical for the Solution: -
rear beam web or shear wall and for the top and On contact of the airplan~ with the arrest-
bottom walls or the torsion box. This is due to
the fact that the detlection flap moves the ing cable, the airplane is decelerated to the
ri&~t relative to ~ig. A4.15. The motion is pure
center ot pressure considerably aft thus pro-
ducing more shear load on the rear shear wall as translation horizontally. The inertia ~orce 1s
well as torsional moment on the conventional
cantilever box metal beam. l1a =~ a = (12~00) 3.5g = 42000 lb.
The a1rplane must likew1se be investigated
for aileron conditions. Operation of the ailer- The inertia torce acts opposite to the direction
ons produce a different air lcad on each side at acceleration, hence to the left as shown in
or the airplane wing which produces an angular Fig. A4.l5.
rolling acceleration of the airplane. Further- The unknown forces T and R can now be solved
more, the deflected ailerons change the mag- tor by using the static equations of equilibrium.
nitude and location of the airfoil character-
istics, thus calculations must be carried aut to ZFx = -42000 + T cas 10° = 0
determine whether the loads in the aileron con-
ditiOnE are more critical than those for the hence, T = 42700 lb.
normal flight conditions.
For angular acceleration resulting fram ZFe = -12000 + R - 42700 x sin 10° a
pitChing moments due to air gusts on the tail,
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A4.9

hence, hence,
R = 19420 lb. R, = 29800 lb. (up)
~o find the distance (d) take moments about c.g.
at airplane,
ZF~ = 29800 - 9000 + R1 =0
hence, R, = - 20800 lb. (acting down)
ill
c.g.
19420 x 24 - 42700 d = 0
hence, The velocity at end ot catapult track can
d = 10.9 In.
be found trom. the rof towtng equation
Landing velocity Vo ~ 60 M.P.H. = 88 tt/sec. V·_Voll~2as

V" - 0 ':: 2 x 96.6 x 35

or
Subt: - v= 82 tt/sec. = 56 M.P.H.

hence stopping distance s = 34.4 ft.
Example Problem 2
An airplane equipped w1th float is cata-
pulted into the air tram a Navy Cruiser as il-
lustrated in F1g, A4.l6. The catapulting torce
P gives the airplane a constant horizontal ac-
celeration of 3g{9S.6 tt/sec·). The gross
weight ot airplane 9000 lb. and the catapult
track is 35 ft. long. F1nd the catapu rt ing
torce P and the reactions R1 and R. tram the
catapult car. The engine thrust 1s 900 lb.
:~at is airplane velocity at end at track run?
9000 lb.
Example Problem 3
Assume that the transport aIrplane as il-
lustrated In Fig. A4.l7 has Just touched down in
landing and that a brakmg torce at 35000 lb. on
the rear wheels is beIng applied to bring the
airplane to rest. The landing horizontal vetoo-
1ty Is 85 M.P'.H. (125 rn/sec) , Neglecting air
forces on the aIrplane and assuming the propeller
forces are zero, what are the ground r-eac t i ons
R1 and R.. What is the landing run distance With
the constant braking forcs?
W ,. 100,000 lb.
Lx

Max.....--c.g.- T--
-l5",J~,"""===ThrU8tLine _ _ -<
c·g:--Ma
A

I 78"
t,
A Solution: -
R R,
~85"-j The airplane is being decelerated horizon-
Fig. A4.16
tally hence the inertia torce through the air-
plane e.g. acts toward the tront ot the airplane.
Solution: - Since the braking torce is given we can solVe
tor the deceleration factor by the equilibrium
The torces will be determined just atter equation,
the beginning of the catapUlt run, where the
car velocity 1s small, and thus the lift on the l:Fx = 35000 - Ma
x 0
airplane wing and the airplane drag can be
henes,
neglected. Max = 35000
Horizontal inertia force acting toward the air- or
plane tail equals, HIIg a
X= 35000
whence
/"", = (9000) 3.0g = 27000 lb. 35000
g (100000) 32.2 = 11.27 tt/sec'

From statics: -
To find landing run (s),
ZFx = -900 - P + 27000 ~ a, hence P = 26100 lb.
v· - Vo • = 2 a x s
To find R~ take ~oments about point A,
o - I25' = 2 (-11.27) s
2:1: = 9000 x 55 + 27000 x 78 - 900 x 83 - 85R li hence,
A =0 s = 695 ft.
A4.10 GENERAL LOADS ON AIRCRAFT

To find Re take moments about point (A) Airplane Load Factor = AirPl~e Lift
ZM
A
= 100,000 x 21 35000 x 9 + 38 Ra o
111800 - 4100 7.7
Re = 47000 lb. (2 wheels) 14000

ZF = 47000 - 100,000 + R:l. = 0 ~ple Problem 5

S Assume the airplane as used in example
R1 = 53000 lb. problem 4 is in the same attitude as used in
that example proo tem. Now the airplane is
Example Problem 4 turther maneuvered by the pilot suddenly push-
The airplane in Fig. A4.lB weighs 14,000 Ib ing the control stick forward so as to give
It is flying horizontally at a velocity ot 500 the airplane a pitching acceleration of
M.P.H. (732 ft/sec) when the pilot pulls it up- 4 rad/sec ".
ward into a curved path with a radius ot curva- (a) Find the inertia forces and the tail load T,
ture of 2500 tt. Assume the engine thrust and assuming the lift force on the wing does not
airplane drag equal, opposite and colinear with change.
each other (not shown on Fig. A4.l8).
(b) Find the torces on the jet engine which
Flnd: - weighs 1500 lb. and whose c.g. location is
(a) Acceleration ot airplane in Z di- shown ln Flg. A4.19.
rection
(b) Wing Llft (L) and Tall (T) forces Assume moment of inertia I y (pitching) of the
(c) Airplane Load tactor. airplane equals 300,000 lb. sec a • in.

93700 lb.
T

,
L
210" J ..
tIll. 800 lb.
I
210'

- D c :. . • Engine Thrust
z
t"'so'""",
L. w 14000

Fig. A4.18 Fig. A4.19

Solution: - Solution: -
•
V' 732 Fig. A4.19 shows a tree body at the air-
Acceleration aa = r 2500 = 214.5 tt/sac a plane With the lift and inertia forces as found
or 214.5/32.2 = 6.67g (uj>'IaI"d). in Problem 4.
The inertia force normal to the flight path The additional inertia torce due to the
and acting down equals angular accelerat10n a = 4 rad/sec a • equals,

Ma - (14000) 6 • 67g = 93700 lb. Ia=300000x4 1,200,000 in. lb.

z - g y
which acts clockwise or counter to the direction
Placing this torce on the airplane through the at angular acceleration.
e.g. promotes static equilibril~, hence to find The airplane is now in static equilibrium
tail load T takes moments about wing aerody- and to tind the tail load T take moments about
namic center (c.p.) airplane e.g.
ZMc.p - (14000 + 93700) 8 + 210 T = 0 ZMe.g. = 1,200,000 - 111800 x B - 218 T o
hence
T = 4100 lb. (down)
T = 1400 lb.
To find Wlng Llft (L) use
To find Mas take,
l:F
z
= - 4100 14000 - 93700 + L o
14000 + 1400 - Mas = 0
L = 111800 lb.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A4.1l

hence, 992 00) _ 'I' Find:

( 14000 g - 7.1 g . . t sec • (a) The inertia forces on the air-
plane.
The e.g. of the engine is 50 inches aft of the It) The resultant load on the pilot
airplane e.g. as shown in Fig. A4.19. The whose weight is 180 lb. and Whose
force on the engine will be its own weight at
location 1s shown in Fig. A4.20.
1500 lb., and the inertia forces due to a~ and
~.
IPilot c. g.
Inertia force due to a g equals,
\ 1-372"-1
Ma. = (15 7.1g = 10630 lb. C• .I:....--
g00)
_7' ~ T::.
Inertia force due to angular acceleration a
equals, oopooc!-l-=J::=-,,--,-,_

Mra = 3Z~~0~ 12 x 50 x 4 = 778 lb. (dawn)

'''Y
319000 lb. Fig. A4.20

Then the resultant forca on tha engine equals

Solution: -
1500 + 10630 + 778 = 12908 lb. (down)
The wing l1tt will be neglected in this
example problem.
Note if the engine had been forward of the air-
plane e.g., the inertia force of 778 lb. would The inertia forces on the airplane are
act upward instead ot downward. rcrces Max and I"'.a~ and the couple I c.g. u ,
In calculating the inertia force on a To find I"'A take,
certain airplane item due to angular acceler- x
ation,the equation F = Mra assumes that the
particular item had negligible mass moment at
lF
x = 100,000 - l1ax =a
inertia about its own centroldal Y axis. In or
the case of a large item this centroidal mass Max = 100,000 lb.
moment of inertia may be appreciable and should
hence,
be .included in the I y of airplane.
ax =
100,000
11
= eOO,OOO)
100,000 s 19 ~
Then to find the inertia torce tor such
an item the equation F = Mra. Should be ncct r iec
To find Ma~ take,
to be

F=(I e.g. ~)/r Where ZFs = 300,000 - 100,000 - Ma a = °

Ma. = 200,000 lb.
r = distance or a~ trom airplane e.g. to e.g. hence
ot item.
= 200,000 = 200,000) _
I = mass
moment of inertia at item about a iS 11 ( 100,000 g - 2g t
e.g.
airplane e.g. equals 10 + Mr il
where 1 0 is mass moment of inertia Of To find the inertia couple I a, take moments
item about its own eentroidal Yaxis. about airplane e.g., e.g.
F = inertia force in lbs. no~l to radius i .
ZM
c.g.
=- 100,000 x 120 - 300,000 x 84
Example Problem 6 +1 0.=0
Fig. A4.20 shows a large transport air- e.g.
plane whose gross weight Is 100,000 Lb, The
airplane pitChing mass ~oment of inertia I a. = 37,200,000 lb.
I y = 40,000,000 lb. sec il • in.
c,g,
~ _37,200,000=
The airplane is making a level landing hence angular acee 1erat . . on a. - 40,000,000
with nose wheel slightly otf ground. The re-
action on the rear wheels is 319,000 lb. in- 0.93 ran/sec>.
clined at such an angle to give a drag com-
ponent of leo,Ooo lb. and a vertical component
of 300,000 lb.
A4.12 GENERAL LOADS ON AIRCRAFT
Calculations of resultant load on pilot: -
x = 372"
c.g. of pilot
iL _ considerable portion of the wing would be stressed
/'
a c s g , air- the plastic range where the st i rrnass modulus rs
= 0.93
Fig. A4.21
Fig. A4.2l shows the airplane e.g. accelerations
The torces on the pilot consist of the
pilots weight of 180 lb. and the various inertia
forces as indicated in the figure.
!lax = e~O) 19 = 180·
!las = e~O) 2.0g = 380·
The inertia torce due to the angular ac-
celeration a acts normal to the radius arm
between the airplane e.g. and the Pilot. For
convenience this normal force will be replaced
by its i§ and x components.
- 180
Fx = Mza = 32.2 x 12 x 40 x 0.93
F =.Mia = 32.2180x 12 x 372 x 0.93 = 161 lb.
s
Total torce in x direction on pilot equals
180 - 17 = 163 lb.
Total torce in 8 direction =
380 + 180 - 161 = 379
Hence Resultant force R, equals
A4. 13 Effeet of Airplane Not Being a algid Body.
The example problems of Art. A4.12 as-
sume that -tne airplane is a rigid body (eur.rer-s
no structural der ormat tcn) • On the basis of
this assumption the applied loads on the air-
plane to either tlight or landing conditions
are placed in equilibrium With the inertia
forceS which occur due to the acceleration of
the airplane. It is obvious that an airplane
structure like any other structure is not a
rigid bOdy, particularly a cantilever wing which
undergoes rather large bending deflections in
both flight and landing conditions. Figure
A4.21 shows a composite photograph taken of a
test wing for the Boeing 8-47 airplane. The
maximum upward and downward deflections shown
are for deSign loadS, which in general are 1.5
times the applied loads. It would not be correct
to say·that the Wing deflections under the ap-
plied loads for these two High Angle of attack
condltior~ would be 2/3 the deflections sh0'Nn in
the photograph since under the deSign loads a
beyonu the elastic limit of the mater!al or into
plane
17 lb.
Fig. A4.21
conSiuerably less than the modulus of elasti-
lb.
city, hence the deflections under the applied
loads would be somewhat less than 2/3 those
shown 1n the photograph. This photograph thus
1ndicates very strikingly that a wing structure
is far from being a rigid body.
Static loads are loads which are gradually
applied and cause no appreciable shock or Vi-
bration of str~cture. On high speed aircraft,
air gusts, flight maneuvers and landing re-
actions are applied quite rapidly and thus can
be classed as dynamic loads. Therefore when
these d~iC loads strike a ~lexible (non-
rigid) airplane cantilever Wing, a rather large
Wing deflection 1s produced and the wing tenas
to vibrate. This vibration therefore causes
additional accelerations of the mass units of
the Wing which means additional inertia forces
on the wing. Furthermore if the time ~te of
application of the external applied forces
approaches the natural bending frequencies of
the Wing, the Vibration excited can ~roduce
large additional wing stresses.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A4.13

Up until World ~ar II practically all air-
planes were assumed as rigid bodies for struc-
tural design purposes. During the war failure
of aircraft occured under load conditions which
the conventional design procedure based on rigid
body analysis, indicated satisfactory or safe
stresses. The failures were no doubt due to
dynamic overstress because the airplane 1s not
a rigid body.
Furthermore, airplane design progress has
resulted in thin wings and relatively large
wing spans, and in many cases these wings carry
concentrated ~sses, such as, power plants,
bombs, wing tip fuel tanks etc,. Thus the
flexibl1i ty Of_ wings have increased which means
the natural bending frequencies have decreased.
This fact together with the fact that airplane
speedS have greatly increased and thus cause
air gust loads to be applied more rapidly, or
the loading is becoming more dynamic in char-
acter and thus the overall load effect on the
wing structure is appreciable and cannot be
neglected in the strength design of the wing.
General Dynamic Effect of Air Forces on
'..ling Loads.
The critical alrlaads on an airplane are
caused by maneuvering the airplane by the pilot
or in striking a transverse air gust. A trans-
port airplane does not have to be designed tor
Sharp maneuvers producing high airplane accel-
erations in its job at transporting passengers,
thus the time of applying the maneuver loads is
conSiderably more than a fighter type airplane
pUlling up sharply from high speeds.
Fig. A4.22 shows the result of a pUll-Up
~neuver on the Douglas D.C.3 airplane at 180
M.P.H. relative to load factor versus time of
application at load. As indicated the peak
load of load factor 3.25 was obtained at the
end of one second of time.
4
~ 3 / ' -,
for applying the load on the wing when striking
the air gust.
NACA Technical Note 2424 reports the flight
test results on a twin-engine Martin transport
airplane. Strain gages were placed at various
points on the wing structure, and strains were
read. for various gust conditions for Which the
normal airplane accelerations were also recorded.
Then slow pUll-Up maneuvers were run to give
similar airplane normal accelerations. The wing
had a natural frequency of 3.8 cps and the air-
plane speed 'HaS 250 M.P.H. Two of the con-
clUSions given in this report are: - (1) The
bending strains per unit normal acceleration
under air gusts were approximately 20 percent
higher than those of Slow pull-ups for all mea-
suring POSitions and flight conditions of the
tests, and (2) The dynamic component ot the wing
bending strains appeared to be due primarily to
eXcitation of the tundamental wing bending mode.
These results thus indicate that air gusts
apPly a air load more rapidly to a wing than a
maneuver load gl vtng the same airplane normal
acceleration for a commercial transport type at
airplane, and thus the dynamiC strain effect on
the Wing is more pronounced for gust conditions.
Figs. A4.23, 24 and 2S show results of dy-
namic effect at air gusts on a large wing as de-
termined by Bisplinghotr·. The results in these
tigures show that dynamic effects tend to con-
Siderably increase wing forces on same portions
of the wing and decrease it on other portions.
Fig. A4.23
-000
I Comparative Shear
Dtstribution
:; -400 , 1\ ,
•
t5
< Dynamic Analysis
-200
- - - Rigid Airplane
<, Analysts
i'..
o .2 .4 .6 .a 1.0
Fig. A4.24

.
<;
•2
1
./
<,
-, Fig. A4.22
c
e -20
-30
,
-,
Comparative Bending
Moment Distribution
0
1
I'
.5 1 1.5 2 ~ -10 I" I
Pull-up of DC-3 Airplane at 180 mph. .; i
The author estimates the natural frequency of
the 0.:.3 Wing to be around 10 to 15 cycles per
o .2 .6
<, ,
-
.6 l.0
secor.d, thus a loading time of 1 second against Fig. A4.25
a tDne of 1/10 or 1/15 for halt a wing deflec- !
tion cycle indicates that dynamic overstress -3 ,
I I
~ c~~~:a~~v;o~~~~i-
should not be appreCiable. In general, it can
be said that cynamic over-stress under maneu-
-2
•g.-1 , -. I
!:Ill
c
Abou.. Elaatic Ax1s.

vering loads en transport airplanes is not as

great as from other conditions such as air
~
.. 1
a
-
['''' ,
, I -ii General Data:-
Wing Span "" 189 ft.
GroS8 wt, "" 184000 lb.
gusts or landing •. 2 I Airl.a.ne Vel. :z260 mpb.
, 0 .2 .4 .6 .8 .10
Dynamic Effect of Air Gusts. Fraction of Semi-Span

The higher the air gust velocity and the • Report on an Investigation on stresses in Aircraft Struc.
higher the airplane velocity, the less the time tures under Dynamic Loading. M.l. T. Publication.
A4.14 GENERAL LOADS ON AIRCRAFT

It has also been four.d ~hat landing loads W: 15000 lb.

applied through the conventional land:ng gear or
by water pressure on a flying boat are applied
rapid enough to be classed as dynamic loads and
/
--
I
such loads applied to wings of large span pro- e c.g.,
duce dynamic stresses which cannot be neglected (
in the sate design of such structures.
A4.14 General Conclusions on Influence of Dynamic
Loa.di.ng on structural Design of -Afrpkane.
, .
-120'9;
Rl
65"
I ,,

The advent of the turbo-jet and the rocket
type engines has opened up a range of possible
airplane airspeeds hardly dreamed of only a few
years ago, and already trans-sonic and super-
sonic speed airplanes are a cammon development.
From. an aer-odynamtc standpoint such speeds have
dictated a thin airfoil section which has thus
promoted a high density wtng. Thus tor air-
planes With appreciable Wing spans like Mili-
tary bombers and near future jet commercial
transports, which usually carry large concen-
trated masses on the wing such as engines, fuel
tanks etc ,; the assumption that the airplane is
a rigid body is not sufficiently accurate enough
because the dynamiC stresses are appreciaDle.
The calculation of the dynamiC loading on
the Wing requires that the mass and st 1tfness
90" ---I Fig. A4.26
zontal deceleration and the stopping distance
for the airplane?
(3). The flying patrol boat in Fig. A4.27
makes a water landing with the resultant bottom
water pressure of 250,000 lb. as shown in the
figure. Assume lift and tal~ loads as shown.
The pitching moment of inertia at the airplane
is 10 million lb. sec.~ in. Determine the air-
plane pitching acceleration. what is the total
load on the crew member who weighS 200 lb. and
is located in a seat at the rear end of the
r.ull?
1000 lb.
Fig. A4. 27 25000 lb.

20"~ L .
distribution or the wing structure be known.
Since these factors are not known when the g. of Crew
structural design of a wing is started, the • ~ Member
general procedure in des tgn would be to first 30"- / __ __ 0"
base the design on the assumption that the wing '1 c. g, 'LJ'
is a rigid body plus correction factors based on r--'oo" --rr-
W: 50000
past deSign experience or available research in-
formation to approx1mately take care of the in-
fluence of the elastic wing on the airplane
aerodynam1C characteristics and the build up
dynwmic inertia forces. ~lth the wing thus in- (4). The jet-plane in Fig. A4.28 is diving at
itially deSigned by this procedure, it then can a speed at 600 X.P.H. when ~ilot starts a 8g
be checked by a complete dynamiC analysiS and pUll-out. Weight of airplane is 16,000 lb.
modif~ed as the results dictate and then re- Assume that engine thrust and total airplane
calculated for the modified elastic Wing. This drag are equal, opposite and colinear.
procedure is now practical because at the avail-
ability of high speed computors. (a) Find radius of flight path at start of
pUll-out.
A4. 15 PROBLEMS. (b) Find inertia force in Z direction.
(c) Find 11ft L and tail load T.
(1). The airplane in Fig. A4.26 is being
launched from the deck ot an aircraft car~ier by
the cable pull T which gives the airplane a for-
ward acceleration of 3.25g. The gross ~eight of
the airplane 1s 15,000 lb. T
;,
~
(a) Find the tension load T in the launching ~

cabls, and the wheel reactions R1 and

R.o
(b) If the flying speed is 75 M.P.H., what
launching distance is required and the
.-l
•
• L
t'. "
<i

launching time t?
(2). Assume the airplane of Pig. A4.26 is
landing at 75 M.P.H. on a runway and brakes are Fig. A4.28
applied to the rear wheels equal to 04 of the
vertical rear wheel reaction. ~hat is the hori-
 CHAPTER A;
BEAMS SHEAR AND MOMENTS

AS.l Introduction.
In general, a structural ~ember that sup- 30"
~orts loads perpendicular to its longltudlonal .1
axis is ~6ferred to as a beam. The structure of
aircraft provides excellent examples of beam
units, such as the wing and fuselage. Very .1----15"'
25 7S lb.
seldom do bending forces act alone on a ~jor
aircraft structural unit, but are accompanied by Fig. AS. 1

g
axial and tors1onal forces. However, the bend- . I OO . P
ing forces and the resulting beam stresses due 10 -1
to bending of the beam are usually of primary :i
~
importance in the design ot the beam structure. N

"
AS.2 statically Determinate and statically Indeterminate >1
Fig. AS. 2 a Fig. AS. 3
Beams. 75 75
A beam can be considered as SUbjected to
known applied loads and unknown supporting re-
100 ... 1
actions. If the distribution of the applied

~
I OO ' P

tclM
known loads to the supporting reactions can be :i
determined trom the conditions of static equil- C .....

ibrium alone, namely, the summation of forces
and ~oments equal zero, then the beam is con-
sidered as a statically d~terminate beam. How-
ever, if the distribution of the known applied
loads to the supporting beam reactions is in-
fluenced by the behavior of the beam material
during the loading, then the supporting reactions
cannot be found by the statical equilibrium
equations alone, and the be~ is classified as a
statically indeterminate beam. To solve such a
beam, other conditions of tact based on the
beam deta~tions ~ust be used in combination
With the static equilibrium equations.
AS. 3 Shear and Bending Moment.
A given beam is subjected ta a certain ap-
plied known loading. The beam reactions to hold
the beam in static equilibrium are then calcu-
lated by the necessary equations of static equi-
librium, namely: -
ZV = 0, or the algebraic summation af all verti-
cal forces equal zero.
ZH = 0, or the algebraic summation of all hori-
zontal forces equal zero.
ZM = 0, or the algebraic summation af all the
moments equal zero.
~ith the entire beam in static equilibrium,
it follows that every portion at ~he beam ~ust
likewise be in static equilibrium. Now consider
the beam in Fig. A5.1. The known applied load
of P ::ill 100 lb. is held in equt Itbr-tua by the two
reactions of 25 and 75 Ibs. as shown and are
calculated from simple statics. (Beam weight is
neglected in this problem). New consider the
beam as cut at section a-a and consider the
H
d >-
T
::; I. 75 I. 75
Fig. AS. .o4 Flg. AS. S
right side portion as a free body in eqUilibrium
as shown in Fig. A5.2. For static equilibrium,
~V, ~H and ZM rouEt equal zero for all forces and
moments acting on this beam portione Consider-
Ing ZV = 0 In Fig. A5.3: -
ZV = 75 - 100 = - 25 lb. - - - - - - (1)
thus, under the forces shown, the force system
is unbalanced in the V direction, and therefore
an internal resisting force Vi equal to 25 lb.
must have existed on section a-a to produce
equilibrium of forces in the V direction. Fig.
A5.3 shows the resisting shear force, Vi =
25 Ib, which must exist for equilibrium.
°
Considering ZM = in Fig. A5.3, take
moments about seme point a on section a-a,
ZM, =- 75 x 15 + 100 x 5 =- 625In.lb.(2)
or an unbalanced moment of - 625 tends to ro-
tate the portion of the beam about section a-a.
A counteracting reststing ~oment M • 625 must
exist an section a-a to provide equilibrium.
Fig. A5.4 shows the free body with the Vi and
!'Ii acting.
Now ZH must equal zero. The exterr.al
forces as well as the internal reSisting shear
Vi have no horizontal components. Therefore,
the internal farces prodUCing the resisting
moment Ml must be such as to have no horizontal

AS.I
A5.2 BEAMS -- SHEAR AND MOMENTS

unbalanced force 1 which means that the resisting 1000 lb. ,

300 lb.
moment Mi in the fo~ of a couple, as shown in .,[ 5 6, ,7 B,'
Fig. AS.5, or Mi = Cd or Td and T must equal C

-+- 5'~5'
to make ZH • O. I' 1 2 '3 6 '1'7 B'I
The tendency of the loads and reactions 1"--4' -----L.... 3' .-!
acting on a beam to shear or move one portion of RA = 1110 lb. RBi", 690 lb.:
a beam up or down relative to the adjacent por-
tion ot the beam is called the External Vertical Fig. A5.6
Shear, or commonly referred to as the beam Vert-
leal Shear and is represented by the te~ V. Calculations for Shear Diagram: -
From equation (1), the Vertical Shear at We start at ~he lett end of the beam.
any section at a beam can be defined as the al- ConSidering a section just to the right of the
gebraic sum of all the forces and reactions 500 lb. load, or section 1-1, and ~onsldering
acting to one side of the section at which the the portion to the left of the section, t~e
shear is desired. It the portion of the beam to Vertical Shear at 1-1 = ZV =_SOO (negative,
the lett of the section tends to move up rela- down on left.)
tive to the right portion, the Sign of the + 610 lb. + 610 lb.
Vertical Shear is taken as positive shear and
negative if the tendency is opposite. Or in .,~ ~b ·3001b,

~
other wordS, if the algebraic sum of the forces
is up on the lett or down on the right Sids,
then the Vertical Shear is positive, and nega- _ 500 lb. _ 500 lb. - 390 lb. - 390 lb.
tive tor down on the left and up on the right.
Fig. A5.7 (Shear Diagram)
From equation (2), the Bending Moment at
any section at a beam can be detined as the al- ~in.lb.
gebraic sum of the moments of all the forces o ~O
acting to either side of the section about the
section. It this bending moment tends to pro- ~ -900 in. lb.
duce compreSSion (shortening) ot the upper fib-
ers and tension (stretching) ot the lower' fibers' -2000 in. lb. Fig. AS. 8 (Bending Moment Diagram)
or the beam, the bending moment 1s classed as a Next, consider section 2-2, just to left of
positive bending moment, and negative tor the reaction RA'
reverse condition.
ZV - - sao, or same as at section 1-1.
AS.4 Shear. and Moment Diagrams.
Next, consider section 3-3, just to ri&~t of
In aircrart design, a large proportion ot RA'
the beams are tapered in depth and section, and
also carry a variable distributed load. ThUS, ZV =-
500 + 1110 =610 (pOSitive, up on
to design or check the various sections of such left side at section).
beams, it is necessary to have a complete piC-
Next, consider section 4-4, just to left ot
ture as to the value ot the vertical shear and 1000 load.
bending moment at all sections along the beam.
It these values are plotted as ordinates trom a ZV =-
500 + 1110 = 610 (same as at
base line, the resulting curves are referred to section 3-3).
as Shear and Moment diagrams. A tew example Section S-S, to right of 1000 load:
Shear and Moment diagrams will be plotted, to ZV =-
500 + 1110 - 1000 390 (down =-
retresh the' students kncw.l edge regarding these on lett).
diagrams.
Check this shear at section S-5 by using the
Example Problem 1. portion of the beam to the right of 5-5 as a
Draw a shear and bending moment diagram tor free body.
the beam shown in Fig. AS.6. Neglect the weight ZV =-
300 + 690 = 390, which checks
at the beam. (Sign of shear is minus, because ZV is up on
In general, the tirst step is to determine right). Section 6-6, use the portion to right
the reactions. as a free body:
To find RB, take moments about point A. ZV = - 300 + 690 = 390 (minus Shear).
ZMA = - 4 x 500 + 1000 x 5 + 300 x 13 - lORS = a
Section 7-7:
hence RB • 690 lb.
ZV =- 500 + RA - 1000 - 300 + 690 = 0 ZV =- 300 (positive shear, down on right)
hence RA = 1110 lb.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A5.3

Section 8-a: Vx = 270 - lOX, and hence, the shear de_

ZV = - 300 (positive shear). creases at aconstant rate of 10 Ib./in. frem 270
at A to 180 at C.
fig. AS.7 shows the plotted values on the shear The vertical shear at section D, just to
diagram. the right of load is,
Calculation of the Moment Diagram. VD = ZVleft = 270 - 10 x 9 - 120 = 60 up,
Start at section 1-1, and consicer the or positive.
forces to the left only:
=- 500 =0 I- 9" -1'20hP
ZM x 0
Since sections 2-2 and 3-3 are only a differen-
tial distance apart, assume a section just above
A Ij j !
c+n
I l j '3:
w=10Ib./in.
I I I j I I 1 1 i
I
B

RA and consider the forces on the left side only: RA=270 RB=210
Fig. AS.9
ZM = - 500 x 4 = - 2000 in. lb. (Negative 270 lb.
moment, because of tension in the top fibers). 'n-,.,._Vcs180 lb.
V =60 lb. Fig. A5.l0
Consider the section under the 1000 in. lb. load:
A;~ 1--6"
ZM to left = - 500 x 9 + 1110 x 5 = 1050 in.
lb. (positive moment, compressing the top ~ 15" Shear Diagram -210 lb.
fibers). Fig.A5.~.~
~
Check by considering the forces to the right:
ZM right = 300 x S - 690 x 5 = - 1050 in. lb. Moment Diagram
Next, consider a section over RB:
ZM right = 300 x 3 = 900 in. lb. (Negative The vertical shear be~Neen points D and B, when
moment, tension in top fibers). x is the distance of any section between D and
B trom A:
At Section 8-8:
ZM right = 300 x 0 =0 VDB =270 - 120 - lOX - - - - - - - - - - (1)

Fig. AS.S shows the plotted values. At point BJ x = 36:

From the above results it may be noticed
that when the bending moment is obtained fram
the forces that lie to the lett at any section,
the bending moment is positive when it is clock-
wise. If obtained fro~ the torces to the right,
it Is positive, when counter-clockwise. The
student should sketch in the apprOXimate shape
of the deflected structure and determine the
signs from whether tension or compression exists
in the upper and ~ower fibers.
Example Problem 2
Calculate and draw the shear and moment
diagrams for the beam and loading as shown in
Fig. A5.9.
First, dete~ine the reactions, RA and RS:
ZMA = 36 x 10 x 18 1- 120 x 9 -
hence RS = 210 lb.
ZV = - 120 - 36 x 10 + 210 1-
hence RA = 270 lb.
Shear Diagram: -
The vertical shear just to the right of the
reaction at A is equal to 270 up, or positive.
This is plotted as line AE in Fig. A5.10. The
vertical shear at section C just to the left of
the load and considering the forces to the left
cf the section = 270 - 9 x 10 = 180 lb. up, or
positive. The vertical shear for any section
between A and C at a distance x from A is:
hence VB = 270 - 120 - 10 x 36 = - 210 lb.,
which checks the reaction RE'
Since the Vertical Shear decreases at a
rate ot 10 Ib/in. from D to B, it will be 6"
fram D to a point where the shear is zero, since
the shear at D 1s 60 lb.
This point could also be located by equa-
ting equation (1) to zero and solving for x as
follows:
o ~ 270 - 120 - lOx, or x ~ 1 = 15" frem
1500
A.
If the shear diagram has passed through
zero under the concentrated load, then the
method of equating the shear equation to zero
- and so~ving for x could not be used, thus in
general, it is best to draw a shear diagram to
36RB = O.
find when shear is zero. Fig. AS.IO shows the
plctted shear diagram.
RA ; O.
Moment Diagram: -
At section A just to the ri~~t of reaction
RA the bending moment, considering the torces to
the left, is zero, sinCe the arm of RA is zero.
The bending moment at any section be~Neen
A and C, at a distance x from the lett reaction
RA, is,
(2 )
In equation (2), x can not be greater than
9.
A5.4 BEAMS -- SHEAR AND MOMENTS

The equation for the bending moment between

D and B (x greater than 9) 1s
wx·
Mx = RAX - P (x-s ) - 2""
= 270 x - 120 (x-g) - lOX'
---2-
= 1080 + 150 x - 5x~ - - - -
At section C, x = 9", substitute in equation (4)
Me = 1080 + 150 x 9 - 5 x 9~ = 2025 in. lb.
(positive, compression in top tibers).
At the paint or zero shear, x 15". = 270 + 180
n = 1080 + 150 x 15 - 5 x 15~ = 2205
Thus, by substituting in equation (2) and
(4) the moment diagram as plotted in Fig. A5.l1
is obtained.
AS.5 SectioD. of Muimum Bending Moment.
The general expression for the bending mo-
ment on the beam at example problem 2 is from
equa t Ion (3):
Now, the value of x that will make
ma.x1mum or minimUm is the value that will make
the first derivative ot Mx with respect to x
equal to zero, or
Therefore, the value or x that will :nake l1x a
maximUm or minimum may be found trom the equation
(3 )
Thus, the area at the shear diagram between any
two points equals ::le Change in ·~ending Jloment
between these ~NO points.
(4 ) To illustrate this relationship, 80nsider
the shear diagram in example problem 2 (Fig.
A5.l0). The change in bending moment between
the left reaction RA and the load is equal to
the area of the shear diagram between these ~NO
points, or
2 x 9 = 2025 In. lb. Since the bending
moment at the left support is zero, this change
therefore equals the true ~oment at a section
under the load P.
Adding to this the area at the snall tri-
angle between point D and the point of zero
shear, or ~o x 6 = 180, we obtain 2205 in. lb.
as the max~um ~oment. This can be c~ecked by
taking the area of the sh6a~ diagram between
the point of zero shear and point 3 =
=
2~0 x 21 2205 in. lb.
Mx a
Example Problem 3.
Fig. A5.12 illustrates a landing gear oleo
strut ADEO braced jy struts BD and CEo A land-
ing ground load of 15000 lb. 1s applied th:cugh
the wheel axle as shown. Let -1 t be required to
(5 ) find the axial load in all ~embers and the shear
and bending moment dia~ for the oleo strut.
vcr- 15.0 -j-5.77iVB

But, observation or this equation indicates

He.....-.e- HA -.rrx.o , I";
B HB T
10"
I i
that the term Ra - P - wx is the shear tor the D j
section at a distance x tram the lett reaction.
Therefore, where the shear is zero, the bending ~<f' I I
moment is maximum. Thus, the shear diagram
which shows where the shear is zero is a con-
venient medium tor locating the points of maxi-
mum bending moment.

AS.6 Relatian Between Shear and Bending Moment.

Equation (5) can also be written :- :II V,
Fig. AS. 12

~~~.at~~~~~t~~al
Resistance at Point
~
tI
I

I
II

0
--I·I
15000 lb.
16"

16"
l
since the right hand portion of equation (5) is "0"
equal to the shear.
Hence, dM = Vdx - - - (6 ) SOLUTION: -

Which means that the difference dM be~Neen
the bending ~oments at ~NO sections that are a
distance dx apart, is equal to the area Vdx
under the shear curve between the two sections.
Thus, for two sections Xl and Xu
To find Vc take moments about point 5,
2MB - - 15000 x 0.5 x 42 + 15000 x 0.866 x 5.77
+ 20.77 Vc = O.
hence, Vc ~ 21550 lb.
The axial load in member CE therefore equals
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AS.'
11550/eos 30° or 13330 lb .•'. He = 11550 x 15/26 Bracket _____
= 6660 .' ~[;,""''l--500
To find HA take moments about point D, 1 00 lb. E ,:, II
Bracket
ZMD = 11550 x 15 - 6660 x 10 - 15000 x 0.5 x 32
... 10 HA = O.
hence, HA = 13340 lb.
To find VB take ZFV =0,
ZFv • 15000 cos 30° ... 11550 - VB - O. Fig. AS. 14
hence, VB = 24550 lb.
The axial load in member ED therefore equals
24550/cos 30° = 28360 lb. (compression). The SOLlJrION: -
reaction HB therefore equals 28360 x sin 30
14180 lb.
Calculations of reactions at A and B: -
Fig. A5.13 shows the oleo strut as a free To find VB take moments about paint A,
body with the reactions at A, D and E as calcu- :MA =-500 x 7 - 500 x 6 ... 1000 x 20 ... 1000 x
lated. Fig. A5.l3 also shows the axial load, sin 45° x 10 + 1000 cos 45° x 2 - 22 VB = O.
vertical shear and bending moment diagrams. hence, VB • 999.3 lb. (up).
The bending moments due to applied loads
without regard to bending'deformation of the To r Ind VA take ZV = 0,
beam are usually referred to as the prtmary zv = 999.3 - 1000 - 1000 sin 45 Q 500 + VA = O.
bending moments. If a member carries axial hence, VA = 1207.8 lb. (up).
loads additional bending moments will be pro-
duced due to the axial loads ttmes the lateral To r inc HB take ZH = 0,
deflection of the beam, and these oe, :ing mo- ZH = - 500 ... 1000 cos 45° - HB 0, hence =
ments are usually referred to as seC)ndary bend- HB • 207.l.
ing aomerrts , (Arts. A23-30 covers t.he calcula- With the exception of the 1000 lb. load at 45°,
tion of secondary moments). all loads are applied to brackets which in
turn are fastened to the beam. Therefore the
13340
next step is to find the reaction of the loaded
brackets at the beam centerline support points.
6660

ic= 11550 !L:K:J50

IIA
The load at E and the reaction Hg at B will be
also referred to beam centerline.
Fig. A5.I5 (a,b,c,d) show the cantilever
16 -r.1 1O ' -1 brackets as free bodies. The reactions at the
base of these cantilevers will be determined.
14180
·24550 Axial Load These reactions reversed will then be the applied
~
-13000 lb. Diagram loads to the beam at points C, D, F and E.
wfl0//@~

F7/?/?/'?//hi'T,;==:z:z:J~~~:am
133~O
e-e.Compz-eaefon

,o-rl ,'.
~:84b
-7500 lb.

~~'Moment
Diagram
Bending
,. !I
2'j
50
8"

~OOO f..l C:~rib=50

Fig. a
I
8" I j)

L MlY'40ci~P"'500
Fig. b
...
1000 5"

HF"'500
VF"'I000

F1g. AS.15
FIg. c
000
L~~'O'.1.
~~.,
E-
Fig. d
,,-:.
,,<;:)
~

Fig. AS. 13 -120000 -133440 in. lb.

Example problem 4. For bracket at C, to find He take ZH = 0,

Fig. A5.l4 shows a beam loaded with both
transv~rse and longltudional loads. This beam
loading is typical of interior beams in the air-
plane fuselage which support all kinds of fixed
equipment. The reactions ~or the beam are at
points A and B. Required: - Shear and bending
moment diagrams.
or obviously HC ~ 500 lb. In like ~er use
Zv = a to tind Vc = 5CO lb. To tind Me take
moments about point C. :Me =-
500 x 2 + 500 x
8 - Me = 0, hence Me = 3000 in. lb. The stUdent
should check the reactions at the base of the
cantilever brackets at D and F (See Fig. h,c).
AS.' BEAMS -- SHEAR AND MOMENTS

The load of 1000 at 45° and applied at point E"
will be referred to point E the centerline of
beam. Fig. d shows the reaction at E due to the
load at E'. The reaction at B should also be
referred to the beam centerline.
shows the beam with the applied loads at points
C 0 E' F and 8'. Figs. AS.17, 18 and 19 show
the axial load, vertical shear and bending mo-
ment diagrams under the beam loading of Fig.
A5.l6.
500 707.1
have rraximum peak moments without the Vertical
Shear paSSing through zero. To illustra:e
this tact, consider the bean of ?~g. A5.20,
namely, a Simple supported beam with an ex-
Fig. A5.16 ternally applied couple ~o~ent of 10 1n. Ij.
magnitude at paint C the center point of the
beam. The shear and bending moment dlagr~~
are as ind~cated and a maximum bending ~ome~t
occurs at C but the shear diag~am does not pass
through zero.
,....-..,MC: 1O" -II'

t,
3000 4000 'e '
e -0 I
'R:l
10"
1207.8
-a l I Shear Dia.

±'"
Fig. AS.16
Bending
-500 lb. -707.1 Axial Load Dia.
Fig. AS.20
~"-II' Moment Dia.

Fig. AS.l7 L_ _ --'-----"-_I__

a , ...:I-=20=7.~l

Fig.A'.lS
d -
-----
707. S,-,lb",,-. _

bl ==~O. 7
[I-
Shear Dia.
A couple is two equal and opposite forces
not in the same straight lir.e. Let it be as-
sumed that the 10 in. lb. couple is made up of
torces equal to 100 lb. each and an a~ Jetween
-~ -999.3----' them at 0.1 inch as illustrated in Fig. A5.21.
4285 5788.7
3000
Bending 1~ ~O 1"
Moment Dia. e
I
-207.1
'1 + lIb.
Fig. AS. 19 11

Fig. AS.21
U99 Shear Dia.

The shear diagram Is determined in the same

manner as explained betore. The applied exter-
nal couples do not enter into the vertical shear
calculations. The hending moment diagram can be
calculated by taking the algebraic sum ot all
couples and moments ot all torces lying to the
one side or a particular section. It it is de-
sired to use the area at the shear diagram to
obtain the bending moments, it is necessary to
add the couple moments to the shear areas to ob-
tain the true bending moment. For example, the
bending moment just to the lett ot point E will
be equal in magnitude to the area at the shear
diagram between C and E plus the sum or all ao-
plied couple moments between C and E but not in-
clUding that at E.
To illustrate the calculations are: _
(- 500 x 5) + (707.8 x 10) 4578 In. lb. (tram=
area ot shear diagram).
(3000 - 4000) = - 1000 In. lb. (tram sum or
couple moments).
Thus bending moment at E • 4578 - 1000 =
3578 In. lb. lert
The bending moment at Bright will equal that at
Eleft plus the couple moment at E or 3578 + 707
= 4285 In. lb.
The student should realize that When couple
~oments are applied to a beam it is POSSible to
The shear diagram is as shown in Fig. A5.21 and
now passes through zero 11TIder each of the couple
forces. Thus i! we assume the couple ~oment has
a dx arm the shear to the right of C is one lb.
and then changes to some unknown negative 'falue
and then back to one lb. pOSitive as the dist-
ance dx is covered in going to the lett. Thus
the shear goes to zero twice in the region of
paint C.
A5.7 Moment Di2iTams as Made up 01 Parts.
In calculating the deflection of statically
deter.ninate beams (See Chapter A7) and solving
statically indeter.ninate structures (See Chapter
AS), the area under the bending ~oment curve is
required, thus it is often convenient to treat
each load and reaction as a separate acting
force and draw the moment diagram for each force.
The true bending moment at a particular paint
will then equal the algebraic summation of the
ordinates of all the various moment curves at
this particular point or adding the various
separate moment diagrams will give the true
bending moment diagram. Figs. A5.22 and A5.23
illustrate the drawing of the bending moment
diagram in parts. In these examples, we start
tram the lett end and proceed to the ri~~t end
and draw the moment curve for each force as
though the beam was a cantilever with the fixed
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A5.7

0\'0' at B and thus leaVing only 3 unknown elements

~ ~\.Q
I' \. .;z '" ,0°
~
P.",10*"
I
of the reaction at B. Fig. A5.25 shows the
r"+5"f5"~
w; 10 Ib/ in. bending moment curves for each load acting sep-
• 1t t t * *~ arately on this cantilever fr~~e. Fig. A5.26
r - - - - 15"------l r2''-t--- 10" shows the tr~e bending moment as the summation
R~",62 R lI"'4.8
R ..",100 lb. Rll""lOO Due to P of the various moment curves of Fig. A5.25.
Due to il, + 11500 As another solution ot this fixed ended
in. lb.
Due to Rl :rarne, one could assume the statically deter-
Due to -====:::=-::-=4000
p. minate modification as a frame pinned at A and
Due to Pli pinned With rollers at B as illustrated in Fig.
A5.27. This assumed stricture is statically

........---
500
Final Moment Dia.
-- ceterainet;e because there
are only 3 unknown elements,
namely the magnitude and di-
rection ot the reaction at A
Pa ::rIO

!!. =10
Fig. A5.-22 Fig. AS. 23 and the magnitude ot the re- P..",10
action at B. For conven-
ience the reaction at A is
support at the right end. The final bending
resolved into two magnitudes Pin~HA"'20 B
moment curve for the true given beam then equals
the sum of these separate diagrams as illustra- as H and V components. The
reactions VA' HA and VB can ""A"'10 B::r20
ted in the figures. Fig. AS. 27
then be found by statics and
STATIC MOMENT CURVES IN SOLVING STATICALLY the results are shown on Fig. A5.27. Fig. A5.28
shows the bending moment diagram on this frame
INDErERMlNATS STRUCTURES due to each load or reaction acting separately,
The usual procedure in solving a statically starting at A and going clockwise to B. Fig.
Indete~1nate structure 1s to first make the A5.29 shows the true bending moment diagram as
structure statically determinate by removing the the summation at the separate diagrams.
necessary redundant or unknown reactions and -=::;::::::I-50-Due to P a
then calculating the deflection of this assumed - 60 -60
o ?'--'"'
statically determinate structure as one step in .....1J.e v t->" Due to V ..J~I00
y A
the overall solution of the problem (See Chapter
AS). In the solution of such structures it is
likewise convenient to treat the bending moment
diagram as made up of parts. To illustrate,
Fig. A5.24 shows a loaded rectangular frame
fixed at paints A and B. The reactions at both

Due to P li

Due to P ,
--=::::J -50
-601 \-60 Fig. A5. 28

j~nu"~
~"r-_,,'O
'i

.I'"
180 30 in. lb.
..1 B Final Bending
-60 Fig. AS. 24 Moment Diagram
-110 Free 60 - 50 90 (Tension on inside
-6 .110 of frame is posi-
-80 FIg. AS. 25 tive moment)
A B
Fig. AS. 29
5" A5.8 Forces at a section in Terms of Forces at a
100 ...L
Previous Station.
STRAIGHT BEA11S
paints A and B are unknown in magnitude, di- Aircraft structures present many beams
rection and location, or each reaction has 3 un- which carry a varying distributed load. tlinl-
known elements or a total at 6 unknowns for the mum structural weight is at paramount importance
two reactions. With 3 static equilibrium in aircraft structural design thus it is de-
equations available, the structure is statically sirable to have the complete bending moment
indeterminate to the third degree. Fig. A5.25 diagram for the structure so that each portion
illustrates one ~er in which the structure of the structure can be proportioned effic-
can be made statically dete~nate, by freeing iently. To decrease the amount of numerical
the end A to make a bent cantilever beam fixed work required in obtaining the complete shear

'( i:t",
~.
A5.8 BEAMS -- SHEAR AND MOMENTS

~nd bending ~oment ciagr&~ it usually saves

time to ex?ress the shear and ~oment at a given
station in terms of the shear and moment at a
previous station plus the effect of any loads
lying bet~een these two stations. To illustrate,
?i~. AS.3D shows a cantilever beam carrying a /

considerable ~umber of transverse loads F of dif- ,/ • I

,

ferent ~~ltudes. Fig. A5.31 shows a free body H1 ~ !- - --I

Q M1 r d----o-;
V,

MlC
(1)
Lr~v"OM" (2)
Fig. AS. 32 Fig. AS. 33

Fig. AS.31
Then from Fig. A5.33 we can 'Nrite for the
resultant torces and moment at point (2) at
station 2: -;
at the beam portion 8evNeen stations land 2.
The Vertical Shear V1 at stat~on 1 equals the
summation of the forces to the lett of station 1
and M1 the bending moment at station 1 equals
the algebraic sum of the moments at all forces
lying to lett of station 1 about station 1.
Now conSidering station 2: - The Vertical
Shear V" = Vl + F 1 _ " , or stated in wordS, the
Shear V" equals the Shear at the previous sta- Having the resultant forces and moments for a
tion 1 plus the algebraic sum of all forces F given point on a given station, it is usually
lying between stations 1 and 2. Again consider- necessarJ in :inding beam stresses to resolve
ing Fig. A5.31, the bending moment M" at station the forces ~nto components no~~l and parallel
2 can be written, MOl = Ml + Vld + Fl_"a, or to the beam cross-section and also tr~sfer
stated in words, the bending moment tl" at sta- their location to a point on the neutral axis
tion 2 is equal to the bending moment Ml at a of the beam cross-section.
previous station 1, plus the Shear V at the For examp ;e Fig. A5.34 shews the resultant
pr-evious station. 1 t1.mes the arm c , the dist-
ance between stations 1 and 2 plus the moments
at all forces lying be~Neen stations 1 and 2
about station 2.
AS.9 Equat.1an.s for Curved Beams.
Many structural beams carry both longitud-
ional and transverse loads and also the beams
may be made of straight elements to r orn a frame
or all beam elements may be curved to form a
sP W v
Fig. AS. 34
s
Fig. AS. 3S
" "J6 ~~
Mo.:Ml wNe
s
Fig. AS. 36
curved frame or ring. For example the airplane
tuselage ring is a curved beam SUbjected to
torces of varying magnitude and direction along
its boundary due to the action or the fuselage
skin torces on the frame. Since the complete
bending moment diagram 1s usually desirable, it forces and moment at point 1 of a oe~ cross-
is desirable to minimize the amount of numerical section. They can be resolved into a normal
work in obtaining the complete shear and bending terce N and a shear for S ~lus a moment Ml as
moment values. Fig. A5.32 shows a curved beam shown in Fig. A5.35 where,
loaded with a number of different vertical loads
P and horizontal loads Q. rig. A5.33 shows the N = H cos a + V sin a
beam portion 1-2 cut out as a treo bOdy. Hl S = V cos a - H sin a
represents the resultant horizontal force at
statton I and equals the algebraic summation ot Later on when the beam section is being de-
all the Q forces to the left of station 1. Vl s igned it my be round that the neutral axis
represents the resultant vertical torce at 11es at point 0 instead of point 1. Fig.
station 1 and equals the sum of all F :orces to A5.36 shows the forces and moments referred to
lett of station 1, and Ml equals the bending point 0, With Mo being equal to Ml - Ne.
moment about point (1) on station I due to the
moments of all forces lying to the left of
point 1.
 • s

ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A5.9

AS.10 Tw:sional Moments.

The loads which cause only bending of a
beam are located so that their 11ne at action
passes through the flexural axis of the beam.
QUite otten, the loading on a beam does not act
through the flexural axis of the beam and thus
the beam undergoes both bending and twisting.
The moments which cause the twisting action are
usually referred to as torsional moments. The
airplane wing is an excellent example of a beam
structure that is subjected to combined bending
and torsion. Since the center ot pressure of
the airtoil torces changes With angle ot attack,
and since there are many flight conditions it is
impossible to &liminate torsional moments under
all conditions ot flight and landing. For the
fuselage, the Vertical tail surfaces is norm-
ally located above the fuselage and thus a load
on this tail unit causes combined bending and
twisting of the fuselage.
Fig. A5.34 illustrates a cantilever tube
being subjected to a load P acting at point A on
a fitting attached to the tube end. The flex-

velocity With the maneuvering Ilmited to certain

Fig. AS.34
ural axis coincides with the tube centerline, or
axis 1-1. Fig. AS.35 shows the load P being
moved to the paint (0) on the tube axis 1-1,
however the original force P had a moment about
(0) equal to Pr, thus the moment Pr must be
added to the load P acting at (0) it the force
system at point (0) is to be equivalent to the
original torce P at pOint A. The torce P acting
through (0) causes bending without twist and the
moment Pr causes ~Nisting only.
For the resolution of moments into various
resultan~ planes of action, the student should
refer to any textbook cn statics.
AS.11 Shears-and Moments on Wing.
Arts. A4.S and A4.6 of Chapter A4 discusses
the alrloads on the wing and the equilibrium of
the airplane as a whole in flight. As expla~n­
ed, it is customary to replace the distributed
air forces on an airtoil by two resultant
torces, namely, lift and drag forces acting
through the aerodynamic center ot the airfoil
plus a wing ~ament. The airflow around a wing
is not uniform in the spanwise direction, thus
the airfoil force coefficients CL, Co and CM
vary spanwise along the wing. Fig. A5.36 shows
a typical spanwise variation of the CL and CD
torce coefficients in terms of a uniform span-
wise variation CL and CO'
Any particular type of airplane 1s designed
maxtmum accelerations. These limiting acceler-
ations are usually specified with reference to
the X Y Z axes ot the airplane. Since the di-
rections of the lift and drag forces change with
angle of attack it is simpler and convenient in
stress analysis to resolve all forces with ref-
erence to the X Y Z axes which remain tlxed in
direction relative to the airplane.
As a time saving element in wing stress
analysis, it is customary to make unit load an-
alysis for wing shears and moments. The wing
shears and moments for any deSign condition
then follows as a matter of simple proportion
and addition. For example it is customary: -
(1) To assume a total arbitrary unit load act-
ing on the wing in the Z direction through
the aerodynamiC section of the airfoil
section and distributed spanwise according
to that of the CL or 11ft coefficient.
(2) A similar total load as in (1) but acting
in the X direction.
(3) To assume a total unit wing load acting in
the Z direction through the aerodynamic
center and distributed spanwtse a~cording
to that of the CD or drag coefficient.
(4) Same as (3) but acttng in the X direction.
(5) To assume a unit total Wing moment and
distributed spanwise according to that of
the Cm_ or moment coef~icient.
~.c

to carry out a certain Job or duty and to do The above unit load conditions are for con-
that Job requires a certain ~lmum airplane ditions ot acceleration in translation of the
A5.I0 BEAMS -- SHEAR AND MOMENTS
airplane as a rigid body. Unit load analyses
are also made for angular accelerations of the
airplane which can also occur in flight and
landing maneuvers.
The SUbject of the calculation of loads on
the airplane is far too large to cover in a
structures book. This subject is usually cover-
ed in a separate course in most aeronautical
curricula after a student has had initial
courses in aerodynamics and structures. To il-
lustrate the type at problem that Is encountered
in the calculation of the applied loads on the
airplane, simplified problems concerning the
wing and fuselage will be given.
AS.12 Ezample Problem 01 Calculating Wing Shears and
Moments for One Unit Load. Condition.
Fig. A5.37 shows the half wing plantorm of
a cantilever wing. Fig. AS.38 shows a wing
section at station O. The reference Y axis has
been taken as the 40 percent chord line which
happens to he a straight line In this particular
wing layout.
xL.
aer~mic J at the preVIous statIon (230) tImes the dist-
center a.c" 0' ...... Ref.
I i 96" I Fig.A5.38
"? Ref. Axls
The total wing area Is 17760 sq. In. For
convenience a total unit distributed load of
17760 1hs. wIll be assumed acting on the halt
Wing and acting upward in the Z direction and
through the airfoil aerodynamic center. The
spanwise distributIon at this load will be ac-
cording to the (CL) 11ft coefficient spanwise
distributIon. For simplicity in this example
it will be assumed constant.
Table AS.l shows the calculations in table
form tor determinIng the (Vz) the wing shear in
the Z direction, the bending moment Mx or mo-
ment about the X axis and My the moment about
the Y axis tor a number of stations between the
wIng tip statIon 240 and the centerline station
O.
Column 1 of the table shows the number ot
stations selected. Column 2 shows the CLIC
L Fig. A5. 39
ratio or the spanwise variation of the lift
coeff~cient C
L in terms ot a uniform dIstrIbu-
tion CL' In Ehis example we have ~aken thIs
ratIo as unity sInce we have no wind tunnel or
aerOdynamic calculatIons for this wing relative
to the spanwise distributIon of the lift force
coeffIcient. In an actual problem involvIng an
airplane a curve such as that given in FIg.
A5.36 would be available and the values to
place in Column 3 of Table AS.l would be read
from such a curve. Column (2) gives the Wing
chord length at each station. Column (4)
gIves the wing running load per inch of span
at each station Doint. Since a total unit
load of 17760 lb. was assumed acting on the
half wing and since the wIng area is 17760 sq.
in., the running load per inch at any station
equals the Wing chord length at that station.
In order to find shears and moments at the
various station points, the distributed load is
now broken down into concentrated loads which
are equal to the dIstrIbuted load on a strIp
and this concentrated strip load is taken as
acting through the center of gravity at this
distributed strip load. Columns 5, 6, and 7
show the calculations for determining the
(~Pz) strip loads. Column 8 shows the lo-
cation of the ~Pz load which is at the centroId
ot a troplzcidal dIstrIbuted load whose end
Yalues are given in Column (4). In determin-
ing these centroid locations it is convenient
to use Table A3.4 of Chapter A3.
The values of the shear Vz and the mo-
ment Mx at each station are calculated by the
method explained in Art. A5.S. _Columns 9, 10,
11 and 12 of Table AS.l gIve the calculations.
For example, the value of Mx = 9884 in Col-
umn (12) for statIon 220 Bquals 2436, the Mx
moment at the preViOUS station in Column (12)
plus 4908 in Column (10) which Is the shear
ance 10 Inches plus the moment 2540 in Column
(9) due to the strip load be~Neen statIons
230 and 220, which gives a total of 9884 the
value in Column (12).
The strip loads 6Pz act through the
aerodynamic center (a.c.) ot each airfoil strip.
Column (13) and (14) give the x arms which is
the distance trcm the a.c. to the reference Y
axis. (See Fig. A5.38). Column 15 gives the
My moment for each strip load and Column 16
the My moment at the various stations which
equals the summatIon of the strip moments as
one progresses fram station 240 to zero.
Fig. A5.39 shows the results at station
(0) as taken from Table A5.1.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A5.11
TABLE A5.1

CALCULA -nos OF WING SHEAR Vz AND WING MOMENTS My AM) My

DUE TO TOTAL UNIT DISTRIBUTED HALF WING LOAD OF 17760 LBS.,
ACTmc; UPWARD m Z DIRECTION AND APPLIED AT AERODYNAMiC
CENTERS OF WING SECTIONS (See Figs ~ 37 38 for Wing Layout)
, 10 11 12 ra 15 15

=
. .•""
C
L/_
CL
• · i ..-
~

.->

Ratio,
Assumed

~L
Unity.

a o
II

~;:
g.

Q,,"''S
~g

o 7.2
.u
>-

~g
.
..·· . :;>-:8
.-
~
~.
~S...;
. s-:
o
'40
"
-49.0 1.0
"
49.09
-48.54 >.0 <2. 2.49
242.7 o 60s 605 7.36
7.28
1787 1767
49.63 .0 48.1 2.49 7."
230 50.18 1.0 50.18 -490.8 1213 618 2438 7.53 1846 3613
51.27 10.00 512.7 4.97 7.89
220 52.36 1.0 52.36 1003.5 4908 2540 9884 7.85 '940 7553
54.00 1 .00 810.0 7.4 11. 09
205 55.64 1.0 55.6'" 1813.5 15052 6020 309S8 8.34 65'iO 14093
57 27 8 9
7 "
190 58.92 1.0 58.92 2672.5 27020 6490 64488 8.85 7390 21483
60.56 15.00 908.4 7.43 9.09
175 62.20 1.0 62.20 3580. 9 40087 6760 111313 9.33 8260 29743
.84 1.00 .4J
160 6 ., 1.0 65. '" 4538.5 53 100 17 1 9170 38913
67.13 1 .00 1007.0 7.43 10.06
14' 68.78 1.0 68.78 5545.5 68077 7480 247683 10.31 10100 49013
70.39 1.00 1055.5 7. -43 10.55
"0 72.00 1.0 72.00 6601.0 83782 7840 338705 10. 80 11100 50113
73. 15.00 1104. 7. "'3 11.04
75.23 1.0 75.23 7705.3 99015 8200 445920 11. 28 11480 71593
76.89 15.00 1153.4 7.43 11.54
I 100 78.55 1.0 78.55 115579 8570 570060 11.80 13270 84863
80.18 15.00 1202.7 7. i3 12.04
as 81. 81 1.0 81. 81 10061 132880 8940 702949 12. 28 14480 99343
83."'8 1.00 1252.2 7.43 12.52
70 85.10 1.0 85.10 11313 150915 9310 863174 12.76 15600 114943
1 .00 1300.9 .4J 13.01
ss 88.35 1.0 88.35 12614 169700 9660 1042534 13.27 16930 131873
89.98 15.00 1349.7 7.42 13.49
91. 62 1.0 91. 62 139 4 189210 10000 1241744 13.71 18200 150073
9 • 1 .00 1 9. a 14.05
o .00 1.0 . 0 ie , 9 1 00 1539 4 14.40 26350 176423
96.00 .00 1920.0 10.00 14.40
98.00 1.0 96.00 17760 318808 19200 1875630 14.40 27848 214011
Sum : 17760 Cheeks Total Limit Load. Assumed all HaJ! Wing.

l~en the time comes to deSign the structural
make-up of a cross-section to withstand these
applied shears and moments, the structural de-
Signer may wish to refer the forces to another
Y axis as for ex~ple one that passes through
the shear center of the given section. This
transfer of a force system with reference to an-
other set of axes presents no difficulty.
SHEARS AND ~OMENTS eN AIRPLANE BODY
AS. 13 Introduction.
The body of an airplane acts essentially as
a beam and in some conditions of flight or land-
ing as a beam column which may be also subjected
to twisting or torsional forces. Thus to deSign
an airplane body requires a complete picture of
the shearing, bending, twisting and axial ~orces
which may ~e encountered in flight or landi~g.
Tn the load analysis for Wings, the direct air
!orces are the major forces. For the body load
analySiS the direct air pressures are secondary
the major forces being of a concentrated nature
in the iorm of loads or reactions from units
attached to the bOdy. as the power plant, wing,
landing gear, tail, etC. In addition, since the
body usually serves as the load carrying medium
Unportant forces are produced on the body in re
sisting the inertia forces of the weight of the
interior equipment, installations, pay load etc
As in the case of the wing, a large part
of the load analysiS can be made without much
consideration as to the structural analysis of
the bOdy. The load analysis of an airplane bOdy
involves a large amount of calculation, and
thus the treatment in this chapter must be ~t a
slmpli:led nature, and is presented chiefly fer
the purpose of Showing the stUdent in general
how the problem of load analysis for an air-
plane bOdy 1s approached.

I _
..,(
A5.12 BEAMS -- SHEAR AND MOMENTS
A5.14 Design Conditions and Design Weights.
The airplane body ~ust be designed to with-
stand all loads ~rom specified ~light co~ditions
for both maneuver and gust conditions. Since
accelerations due to air gusts vary inversely
as the airplane weight, it customary to analyze
or check the body for a light load condition for
flight conditions. :n general, the design
weights are specified oy the government agen-
cies. For landing conditions, hcwever, the
normal gross Hei~~t is used since it would be
more critical than a lightly loaded condition.
The general design conditions which are
usually investigated in the design of the body
are as follows:
Flight Conditions:
H.A.A. (High angle of attack)
L.A.A. (Low angle of attack)
I.L.A.A. (Inverted low angle of attack)
LH.A.A. (Inverted high angle of attack)
The above conditions generally assume only
translational acceleration. In addition, it 1s
sometimes specified that the forces due to a
certain angular acceleration of the airplane
about the air~lane e.g. ~ust be considered.
The bOdy is usually required to withstand
speCial tail loads both symmetrical and unsym-
~etrical which may be produced by air ~~sts,
engine forces, etc. Also, the body should be
checked for forces due to unsymmetrical air
loads on the wing.
Landing Conditions:
In general, the body is investigated ~or the
following landing conditions. 'The detailed re-
quirements for each condition are given in the
government specifications for both ~illtary and
commercial airplanes.
LandpLanes : Level landing.
Level landing with side load.
Three point landing.
Three point landing with
ground loop.
Nose over or turn over
condition.
Arresting. (Usually for only
Navy Carrier based air-
planes) .
Seaplanes or Boats:
Step landing With and without
angular acceleration.
Bow landing.
Stern landing.
~NO 'Nave landing.
BeaChing conditions.
Catapulting Conditions (Navy
airplanes) .
Special Conditions or Forces:
Towing of airplane.
BOdy superCharging.
AS.i5 Body Weight and Balance Distribution.
The resisting inertia ferces Que to the
dead weight of the body ~nd its contents plays
an important part in the load analysis for the
airplane body. ~~en the initial aerodyna~ic
and general layout and arrangement of t~e air-
plane is ~de, it is necessary that a complete
weight and balance estinate of the air?lane be
made. This esti~te is usually ~ade by an en-
gineer from the weight control ~ectlon of ~he
engfneer tng cepar tmerrt who has had experience
in estimatin~ :he wei~ht and distribution of
airplane units. 7:'i8 estimate which is pre-
sented in report fo~, gives the wei;hts and
(e.g.) locations of all major air;lane units
or instal:ations as well as for many of the
minor ~~its which nake up these ~jor airplane
assemblies or installations. ~~is weight and
bal~~ce report forms the basis for the dead
weight inertia load a~alysis which forms an
important part ~n the load analysis of the air-
plane body. The use of this weight and balance
estimate will be illustrated in the ex~~ple
problem to follow later.
AS.16 Load Analysis. Unit Analysis.
Due to the many design conditions such as
those listed in Art. A5.14, the 5eneral pro-
cedure in the load analysis of an airplane body
is to oase it on a series of unit analJ~es.
The loads fo~ any part1cular deSign condition
then :cllows as a certain combination of the
unit results with the proper ~ultlplylng fac-
tors. A Simplified example problem follows
which illustrates this unit method 8f approach.
AS.17 Example Problem lllustrating the Calculation of
Shears and Moments on Fuselage I:e to Unit
Load Conditions.
Fig. A5.40 and A5.4l shows a layout of the
airplane body to be USed in this example er-e-
Lem, It happens to be the body of an actua...
airplane and the wing used in the preViOUS ex-
ample problem Nas the wing that Nent with the
airplane.
II
($'-1\"1~=r=~~~I-1 hrust Line
• _. - -- - _ £.jBo!_.!..xi.»... - - -
~c sc. G. (groSS wt. ) I 11'2"
~
~
= I I
I
:1
I '
~,.L o Fig. AS. 40
I-~==::::===- 2:T7·2;-,,---=~:.....===:j1
Front View
Fig. AS. 41
- - - - 20'·0"'----_
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AS 13

Table A5.2 gives the Weight and Balance TABLE A5.3

esti~te for the total airplane. This table 1s
usually fo~ulated by the ~ei~~t and Balance WEIGHT AND BALANCE OF AIRPLANE LESS WING GROUP
AND INSTALLATIONS IN AND ON WING
Section of the engineering department and it is
necessary to have this information before the z. equal &rill frnm thrust line. • 1a up,
Ref. Axes: III distance from z. Ref . .Ax1s 5" !orwarQ at.
airplane load analysis can be made. ( xpropeller. .. L8 a.ft.

i
TABLE A5.2

AIRPLANE WEICHT AND BALANCE

.. ~

a
~
'.m

1
3
N=.
PawerpiaDt grOlq)
Fuselage groull
w."",
y

1100
-
Boriz.
,
19
113.5
Bori:l:.
""'m.
~

30900
39700
Vert.
.um
,
o
Vert.
""'m.
~

Reference
Vert. (Z) arms Ill_red from thrUst line
(+ II up)
; ,• Tail group
'50
11' 38'
121
31550
"
1
'"
3690

"". \ BorU:. (Xj arms meuu:red froID Z ms 5"

!ot'T.l.rd ~ro-p. ~ (+ is aft)
-
e
.,
e
Surface COlltro18
'Electrical Iylltem
T1I.i1 Whee! gro~
130 " er
308
10800
7930
-19

,•
-1190
'30
, , " 118
10700 -10 - 350

'.m
No. N=. w. I
wt. it
Boriz.
.um
Bortz.
Moment
Vert.
.um
Vert.
Mom,
"~
~
10
Furn1ab.iDgs

"'"'"
Wela:ht em~ty •
'"
136
d55
191
25520
22600
1t197oo
10
~
1100

1 Power Plant 1100

(X) (iD,)

19
~

20900
(Z)

n
~

o
'50
~
0
. 11
12
Plio<
"""~,
GroSl weigllt " Z
200
200
es
151

"Z
33200
19800
19700
••
Z
'00
SO,
a Fueelaie Gro~ 35' 113.5 311700 1
"'"
3
•s
WiD( Grou-p
Tail Grou~
SDrface Controls
'50
11'
ae
" 38'
72750
31550
10800
-18
"•
-1:ui00
'04'
_ 1190
Emllty ~ z 169700 • 78.7"
""
Z.~.2..00"

", '""
-l<
e 133 , 7930 2~~~ ". ~
- '"
Electrical System With \18efUlload. i • • BG.O" • 2.32"
t Chu;ais Front 73 16450 -52 -12200
,a Tall Wheel Gro~ 35
'" 10700
'50"
-10 35'
5 I 1100
10
FuruilItw1ga
Radio '"
'" 131
11'
22600 I 10 "50
11
ra
13
Weipt emllty •
PUot
Student
Fuel Syllt.em
,,
Grosl weight ·1
,
3150
20'
'00
'"
..
151
88 ,
"8900
'0200 [
196OO! "
67600 ) _27
·21380
• ."
." _20500
weight distribution should study the inboard
prOfile drawing of the airplane which shows the
general arrangement of all the installations and
4300 37~00 ~40280
equtpment , Furthermore, he should study the
overall structural arrangement as to its possi-
.CalcuJ.ation of C, G. locations:
GroSI wt. x z 376500/4300 z 98,S" a,ft of Re! • .Mia ble influence on fuselage weight distribution.
i z. z -40280/4300 • 9.4" below line thrust
The whole process involves considerable common
sense if a good apprOXimation to the wei~~t d1s
SOLUTION: trlbutlon is to be obtained. Fortunately the
large dead weight loadS, such as the power
WEIGHT At'IITJ BALANCE OF BODY ITEJ"!S. plant, tail, etc. are definitely located, thus
~~IGHT DISTRIBUTION. small errors in the distribution of the minor
distributed weights does not change the over-
Table A5.3 gives the weight and balance all shears and moments an appreciable amount.
calculations for all items attached to fuselage In order to obtain reasonable accuracy, the
or carried in t~e fuselage, except the wing and fuselage or body is diVided into a series of
items attached to the wing as the front landing stations or sections. In Fig. A5.42, the Sec-
gear and the fuel. tions selected are deSigned as stations which
In order to obtain a close approXimation to represent the distance from the Z reference
the true shears and ~oments on the fuselage due axis. The general problem 1s to distribute the
to the dead weight inertia loadS, it is neces- concentrated loads as shown in Fig. A5.43 into
sary t~ distribute the weights of the various an equivalent system acting at the 7arious
items as given in Table A5.3. Fig. A5.42 shows fuselage station pOints.
a side view of the airplane with the center of Obviously, if a weight item trom Table A5.3,
graVity locations of the wei~~t items of Table represents a concentrated load such as a pilot,
A5.3 indicated by the (+) Signs. In the various stUdent, radio, etc., the weight can 08 dis-
deSign conditions, the direction of the weight tributed to adjacent station points inversely
inertia forces changes, thus it is convenient as the distance of the weight (c.g.) fram these
and customary to resolve the inertia forCes into adjacent stations. However, for a weight item
X and Z components. ThUS, in Fig. AS.43, the such as the fuselage struct~e (Item 2 of Table
weights as given in Table A5.3 are assumed act- AS.3) whOSe e.g. location causes it to tall be-
ing in the Z direction through their (c.g.) lo- ~Neen stations 80 and !20 of Fig. A5.43, it
cations. The loads as shown would not give a would obViously be wrong ~o distribute this
true picture as to the shears and ~oments along weight only to the two adjaCent stations since
the fuselage, thus these loads should je dis- the wel~~t of 350= 1s for ~he entire fuselage.
tributed in a ~nner which should si~ulate t~e This weight item of 350~ shou1 4 thus Je 11s-
actual weight distribution. In ~ost weight and tributed to all station points. The contrOlling
balance reports, the weight items are broken require~ent on this distribution is tr~t the
down into considerable ~ore detail than that moment of the distributed system about the ref-
shown in Table A5.3, which ~kes the wel&~t dis- erence axes must equal the moment of the orig-
tribution ~ore evident. The ~erson maKi~g the inal Height about the s~e ~~es. Fig. A5.44
 A5.14 BEAMS _. SHEAR AND MOMENTS

WEIGHT DISTRIBUTION TO FUSELAGE STATIONS

STATION
so 80 120 170 200 220 260 290 315 0 11 50 80 120 170 200 230 250 290 315
I
i 1 I I I I I
I I I
I I 110 I

!I Fig. AS.42. Location of weight items of Table AS. 3. Fig. AS. 46. Weight items from Table AS. 3 acting in X direction.
Z.

011 50 80 120 170 200 220 260 290 011 50 80 120 170 200 230 260 290 315
I
I I I 1 I I I 1
I
I I

I I I 1
I I
'" I
~
1- ~ --1-r1--ig- T

I -
,,I
.... ~
.... '" ~
~M

I
l~
<:cl N

I
-
'"
N

~; i
, ,
I I I
Fig. AS.43. Weight items of Table AS. 3 acting in Z direction. Fig. AS. 47. Vertical distribution of fuselage dead weight.

011 SO 80 120 170 200 220 260 290 011 so 80 120 170 200 230 260 290 315

I 1
I
I
I I i
I I
I I
I I' i
I I
I I
" I I I I

190
--II
80< 780 6511
j -
----i!-- j 30# 100
I
rpl 90
90 .-r-80~
3:0# 2':34#

I I
I I
Fig. AS. 44. Results of fuselage weight distribution to stations. Fig. AS.48. Fuselage weight referred to X axis plus couples.

SO 80 120 170 200 220 260 290 215 o 11 50 80 120 170 200 230 260 290 315
I I I I I I I
I
I ' I I I I I I I
I I I I I I I
I I I I' I

',9'
409*
I
311*
I
7~#
I
100 11191
11i93
8 758 882 1850 '90 10 '274 - 2
-(_288(;-307 -(-'09-C-311G"C- 10 (_21 'tf: 12e
I I
Fig. AS.45. Final weight distribution to station points. Fig. AS.49. Final weight distribution in X direction referred
to X axis plus proper couples.
 ANALySIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A5 15

shows how the dead weight of 350i: was distribu- aanner as to have the same resultant e.g. 10-
ted to the various station points considering cation as the original we Ight system which [s
the weights to be acting [n the Z direction. illustrated in Fig. A5.46. Fig. A5.47 illus-
Table A5.4 shows the results of this sta- trates how the fuselage weight distributed
tion point weIght distribution rcr the weight system as shown in F[g. A5.44 [s distributed
item.s of Table A5.3. The values in the hori- in the vertical direction at the various
zontal rows opposite each weight item shows the station points so that the moment at this sys-
distribution to the various fuselage stations. tern about the X axis [5 equal to that of the
The summation at the weights [n each ver-t tca.L original fuselage weight or 350#. For con-
column at each station point as given in the venience, these distributed fuselage weights
third horizontal row rz-cn the bottom or the can be transferred to the X axis plus a moment
table gives the final station point weight. as shown in Fig. A5.48.
These weights are shown in Fig. A5.45 for Table AS.4 shows the vertical distribution
weights acting in the Z direction. The moment at the various items at the various station
at each total_station load about the Z axiS [s points. The bottom horizontal row gives the
given in the second horizontal row tram the moment about the X axes at the loads at each
bottom of Table A5.4. The summation at the station point, which equals the indiVidual
moments in this row aust equal the total WX loads times their Z distances. The summation
moments of Table A5.3 or 219700"*. This check at the values in this horizontal row must equal
[s shown in the last vertical column of Table the total wz aoment ot Table AS.3. This check
A5.4. is shown at the bottom of the last vertical
The distributed system must also be distri- column. Fig. A5.49 shows the results as gtven
buted [n the Z or vertical direction [n such a in Table AS.4 for the weight distribution in
the X direction.

TABLE ~ ••

PANEL POINT WEIGHT DISTRlBtn'lON

(Monocoque 1';>98 "J FV.lleia&el
~ No. : di8U.llce " lro... Ref. '" Az1.s
I
I
I
,.~ i
~o.
N=, I we, I
I W ., .
u I so
, . " , , • '", , I i
I
,
•
ire
, I
. aoc
,
i
i
."" , • "" , • "" ,
I
.'" ,
.", •
i Pow. l'\Il.a.nt "fOUP 1100 8741 c I no o
-. .
I
, I
,
.
I i UOO
m" . 1

,,
roUD ,
, , - ,- , ""
urlace contro
, ."
II
, "
io
I ElllCtrical s !item

Radio
Furni •
raUl)
0

12$ i ,
• •
,
to ..
rs rc •
,

,
,
, -" , , .10 I
,
.W

".""
"", ,
•
...
, , ,
U i Pilot .OO
"",
." "'.
ra ""'.m ,
• 2.47 , " • au I te 10 1.0 11 uml ,'oo
"'''"
2555 I 893 .02
ToW ..
Horlo;o"l::I,I MomlMlt • :. ""'. 9820 r
3881
\9390 "" 24560 I
."
491001
1.67
52800
5. J
isace
6.45

"""I
1181 19.1
134200 I
.",

, ' S'l6O 21W100

Ven.1cal Marnellt • Z w%..
"I a 758 , 682 1 1 1650 I
"" ic
'" '''' - "" I 5920 I
I J: = dlKtaAce 0( statiol1 lro.ll;l
reterence U111 _mell III 5" fonra.rd 01 propeller.
Z;
z; : disbnce aeeve or beloW tllrust l.1D.8 or x wa.
I

A5.18 Unit Analysis for Fuselage Shears and Moments. (1) Unit acceleration or load factor tn Z
Since there are many flight and landing direction and acting up.
conditions. considerable tl.:ne can be saved i t a (2 ) Unit acceleration or load factor [n X
\mit analysis is made fer the fuselage shears.
direction and acting rorwarc .
axial and bending forces. The design values in
general then folloW as a sunnnation of the values (3 ) Unit tail load normal to X axis acting
in the \m[t anaLys t s t imes a ~roper multiplica- down.
t~on factor.
The loads on the fuselage [n general con- Unit analyses are also usua.Ll.y carried out
sists ot tan loads. eng:!.ne loads, Wing re- for engine thrust and engine tor-que , side load
actions, landing gear reactions If attached to on ~ll and angular acceleration, but to keep
ruse rage and tner-t i.a fore es due to the airplane the example calculations from becoming too
acceleration which ~y be due to botn trrans Ia-. lengthY only the above 3 unit conditions will
tional and an~ular acceleration of the airplane. be carried out in detail. The others will be
For s 1.:nplic1t.y, these loads can be resolved into discussed in detail [n later paragra phs •
cGIt::onents parallel to the Z and X axes.
To illustrate the unit analysis procedure. Solution for Unit Load Factor in Z Direction.
a unit analYSiS tor our example problem will be
carried out for the following unt t conditions: Fig. A5.50 shows the dead weight loads
 A:) 16 BEAMS -- SHEAR AND MOMENTS

act~ng In the Z direction as 'caken :':rorr::'!'3.ble

TABLE A5.5
A5.4 or Fig. A5 . .:i-5. T:'18 'Nin~ ~s attached to t:le
rusejage at s--cations 73 ::ond 116 as s::;'~wn en ?~,£. I FUSELAGE SHEARS AND MOMENTS FOR ONE
AS.50. The fittings at these points are assumed I LOAD FACTOR IN Z DiRECTION
as designed to cause all the :.rag or reaction in I (Dead Weight Acting Down)
the X ::ilrectlon to be taken of f er.t tr-aj.y at the
front fitting on station ';'"3. 1
i 2 3 I 4 , 6
To place the r ....selage In equt I i or-tum, t'::e Load or V : shear i ex : Dist. i ..:oM Moment
wing reaction '1'1111 be calculated: sta,
Reaction : L w ! between I = VLlx M'
No. w (lbs.) stations i
(in. lbs.) (in. lbs.)
ZFx = 0, RH + 0 = 0, nonce RH = 0
0 0 I
I 0 0
lJ"f"S ta t ; or. 0 = 219700 - 116 RR - 73 R!i' = 0-- (;, ) 31' :1
- 22 22
I
25
,

(Note: 21S700 from Table A5.3) i ,

•
ZF z . - 2555 + £iF + RR • 0 - -- - --- -- (3 )
290
- -
0
118
22
140
I
30
! - 550
I
- 550 j

Solving equations (A) and (3 ) for RF and , 260 •

- -
0
21
140
161 - 4200 II - 4750 I
30
RR:
230 • 0 161 I
- 4830 9580 I
- - 10 171
I
I -

RF == 1780 lb., RR = 775 lb. 30

I
200 • 0 171
- 5130 -14710
Table A5.5 gives the calculations for the
- - 76 247
I 30
1
fuselage shears and bending moments 3.t the var-
- - 31~
170 • 247
558 - 7410 I -22i20
ious station points. I
;0
,
sta.O 11
I I
'0 60 120 170 200 230 260 290 315 120 ·- - 40:
556
967
,,I
,, -27900 I -50020
I I

VN
I I
I I
I i
I I I I I I I I
4
I
I
I
I
~
I
I
I I i
I :
'
I
I
I
22
' 116 •
-
0
77' 1I
967
192 i
36J
- 3668 I, -53888
I
II
,I ,
* ·-
I
893 388 307 409 311 76 19 21 ~ 0 192
. - --.--.: .-t--. --t-->: - - t - . +- •- +- . .- 80
- 307 499 , i
,
- 6912 1
-60800
I 17" 20" I
.I '--- ~--- --f'-7::. •- 0 499
7
I
I

i:i' RW O .~.43"~,R",775 I
73 1780 -1281
I
- 3493 I -64~90
Ry=1780 Fig. A5.50
"" I 23

-•
STA.73 STA.1l6 388 -1281
'0 -29463 -34827
0 - 893 I
39
Solution for Unit Load Factor in X Direction. II •
-
893
0
- 693
a
-348~
Fig. AS.51 shows the panel point dead (3) Up on left and down on r-ight side of a section IS I
weight distribution for loads acting in the X pcemve shear.
direction and att, as taken from Table A5.4 or (8) Tension in upper fuselage portion is negative bending
Fig. A5.40. To place the fuselage in equlli- moment.
brlum the wi::lg reactions at points (A) and (8 ) (I) •
refers to aft side of station.
- refers to forward side of station.
will be calculated.
• M = M at previous station in col. 6 plus 1M in col. 5•
lF x = 2555 - RH = 0, hence RH = 2555 lb. (forward)
Take momentis about point (A)
II 80 120 173 200 230 260 290 31'
0 I 50 I I
5920 - 4, RR = 0, I I I I
lI1A ,. 2555 x 17 + I I I
i I I I
I

i I i i
hence RR - 1147.8 (up) I I I ;",
I
I
I
r
i
I

'\
I
I
(5920 equals the sum of the couples from Table 19 I,'
~93t ..(-388(-307 (-409
758 662 1650 490 10 274 2249
(-311(-760-10 (- 21 (.....],[3.<:-2 2
A5.4.
1. R':=t~~ ___.B
I
220
I -4' ---- 43"
lF z = 1147.8 - RF = 0, I
hence Rr = 1147.8 lb. (down) I Rp1l47.8 iRR"'1147.8 Fig. AS.51

STA.73 STA.116
Table A5.6 gives the calculations for the shears,
moments and axial loads for the loading of Flg.
,- ,
.<>..:>.5....
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A5 17

TABLE AS.S 1
TABLE A5.7
fUSELAGE SHEARS, MOMENTS" AXIAL LOADS FOR
FUSELAGE SHEARS &; MOMENTS FOR UNIT
ONE LOAD .-ACTOR IN X DIRECTION
I
, a , a I
lnerua Loada Acting Alt
s • ; a , s
HORIZONTAL TAn.. LOAD IN Z DIRECTION

...
,
, (Load Acting Down)
i w" : :w::iP":.~w" v • z w: "'Mt = r c x • I.::.M>l·- I
mX ,load. ~
'"" •
Couple i D,S!..
I
So. i 'wa.l SlIear Moment Ibetween vt:,,: i
1 2 3 4 5 6
Dlr. !ZDlr. LGad ;htloD. I
, , Load or V , shear ~X :: di st,
o: Moment
" u
,,-- ~M
c
-j ,I· c
1 315 •
! . 22.! '1 "aa 220 !
1
1
I
22g I sta. Reaction
w lbs.
: Z w
Ibs.
between
stations , V~X in. tbs.
as
t asc : I 11~ I cI ,<0 u • 22411 1
o' sac
- 2029
,I
2l\(J • i
-, " ,I ,<0
0
I
- 37~ I
30
1
I c 1_ 2029
,- 2303
315 -- 0
0
0
0
0
0
0
2li
,., '" 1 30
25
230 : i
,
o:
10 i
"
161
m c 1
1
-
c !
10 !
30
c :-
2303
• _ 2313 1 290 ·1
- I
0
0
0
0
0
0
0

L
,
12.5
200
-.! 01 rn c 1- 49~
, , c ~_23131
,'.38031, 0 0 0
,,
" ! u1
'" 272.5 0
170 : I
,I
311 : oI ...
'" i
1
o 1

,
· l6Sg L
30

su
!
I
,
01.28031
:- 445 3 1
=1
- I
-100

0
100

100
17.5
- 1750 -- 1750
a

I 'I 358
· 68~ I
c i [- 44531 260
- 0 100 1750

'I
120 _ 40IIj
o o !_ 3135 1
'"
0

4 , 30
• - 3000 - 4750
,
~ :! '" ,
,
i-' - SUS 0 100 4750
-
115
: I
I
1147.6!
'"sat , -IH7.6
_1141.8
, 1
ae
Q
5135
1 230
- 0 100
30
"" ~:~:; i,
60 •. Q 1 41320

- 3000 -- 7750
,
- rse !
:i
3071

c
1274

1214
_1147.8
I
-1147. a i
_4343~ ~- a035
Q
,

43462 i
200 -- 0
0
100
100
7750

I
73 2555'-1147.al " -1281 u 7
o 27 ! 30
, " • 0 100 ·10750
1
50 ::
3861
o1 01
-t281
- a93 !
o I :
'
I
I
c ~I
170
- 0
I 100 I -
3000 -10750

.- a9~ I
,
I I~ " i
50

: "1" •
!, u - 893 o ,I c 0 100 -15750
"I o 1
I
I
120
- 0 100 I- 5000
-15750
1 • :' 17 x (2555) : 4343:5
•
1 Col.
,..
• refers to alt aide 01 staJ:ion
1 _ refers to forward side of station
te pJ,u lor ten!lion in -fllHlage.
116 -- 0
475.6
100
-375.5
36
I! - 400
-16150
-16150
i ICoL S) M • M iU ~rl!Vlou. station In CoL. 9. ll1u .; Ml of col. 6 ~l...

-- .- 2629
:'M2 ill 001. a I 0 -37!5.6 i
80 13521 2629
0 -375.6
7
Solution for Unit Horizontal Tail Load Acting Down. a
73 -- 0
-375.6
-375.6
0
23
2629
0
The fuselage shears and :noments will be
C8mputed for a unit tail load of 100 lb. on the 50 -I
-,
0 0 0
tail acting in the Z direction, with balanc Ing , 0 0
39
0
reactions at tna wing attachment pOints. The
-i-I
0 0 0
center of ;Jressure en the hor:!.zontal tail is at 11
0 I 0 I J
station 277.5. Fig. AS.52 shows the fuselage
loading. To r ind wing reactions at (A) and (B) : (Col. 8) , M= M at previous station in Col. a plus
6M in Col. 5.
ZMA = 100 x (277.5 - 73) - 43 RR = 0,
hence RR = 475.6'" (up)
ZF z .
- 100 + '::'75.6 - RF 0, . CALCtru-TION
MOMENTS ANIl
OF APPLED FUSELAGE SHEARS,
AXIAL LOADS FOR A SPECIFIC
hence Rr = 375.6# (down)
FLIGilT CONDITION.
Table A5.7 gives the detailed calculations
for the shears and moments at the various sta- Using the results in Tables AS.5, AS.6, an d
tion :;)oints. .- 7 the applied shears and ~oments for a g:!.'l8 n
,,"0., J

0 11 50 80 120 170 200 230 260 290 315 flight condition follow as a m.tter of pr'cpor'-.
I I I I I I 127~. , i tion and addition. To illustrate, the applied
I, ! I I values ~or one flight condition 'Hill be given.
,I
1 1
1 1 ! 1
I I 1-'100* , It will be assumed that the aerodynamic
I
! I , ,
1
i calculations for this air?lane for the (H,A.A.)
! 1
hfgh angle of attack condition ;ave the follow-
+---------------- - - - - tng r-esu Lt s , which the student will aave to
; ~ RH-~~4 --2~-_ accept wt -nout knowrenge of how they were
I 43" -- -- cotameo •
R F =375. a RR=475.6 Fig. A5.52
STA.73 srA.116
 A5.I8 BEAMS -- SHEAR AND MOMENTS

Applied load factor in Z direction - 6.0

down
Applied load factor in X direction 1.333 ~:t.

Applied tail load ~ 110 lb. up.

-: ~c.G-.-
c
Thus with the load factors in the Z and X
directions and the tail load :<nown, 7able A5.S
can be !111ed in as illustrated. In a similar Fig. A5. 53
manner the values for other flight conditions
can be found, the only difference being a new -----
set of ~ultiplying factors since the applied
loads would be different.

TABLE A.5.a

APPLIED FUSELAGE SHEARS, !>IOMENTS " AX1A1. LOADS FOR

niGHT COIQrrlON I. (H.A.A.) The resistance to these X ~~d Z components
of the grcund react~on R is ?rOvided by ~he in-
ertia forces of the airplane in the X ar.d Z di-
... rections •
Tables AS.S and AS.6 show the tuse l.age
shears, moments and axial loads for inert~a
,
,
,...
~60 • j
0
c
0
c
"
c
0
c
c
co
c
c
5Q1O "
921:15
0
29. 3
29.3
186.8
16 .1:1
loads due to one load factor in the Z and X
directions respectively. Thus to obtain the
fuselage forces for this given landing condition,
"" .- ,
0 192~ '_ 29650 i ns it is only necessary to mul.t t p.Ly tne va Iues in
e _ :;~335 as
, 0 "'"•
5220 • 52030 no these two tables by the ?roper factor an~ add
200 : I 'I: 834~ I aac Ute results.
110 •
3740
- 3140
S5-10
11620 1_124640 ; ,,, Thus fuselage forces due to vertical load
- 5Q85 118ZO 1_12898~ I

'" .i "
• 6850
: 173 0 _ 86855 I
17J2.0 ·286650 1 :290
'"
'" factor ot 7 would equal 7 t1mes the values in
columns (3) of Table AS.S to obtain shear ~.d 7
',:;i~:g: ~~:~
116 • - 6850 ' 17770
_ 6850 1 17770 I
",., 2910 _313690 I 1290
times column 6 to obtain bendi~g moment.
I
, .,
80 : 47250 mo ,_314640 : 1700 Likewise the farces due to the 2.98 load
73 • I H900 0 40 I 1700
factor in X direction would equal (-2.98) ti~es
• ,.,
'_385704 1_1710
" _208937 ;-1190
0
the values in col~s (4)7 (5) and (9) of
-1190
o
Table AS.S to obtain axial loads, shears and
bending moments respectively.
The final or ~r~e for~es woult be the
Col.. 1 _ 6 :l valu•• In '01""",,, 3 at Talli. A.5.5.
Col. 3 • 1.333 '" valu'" ill .00umn 5 01 Tabl .. .405.B. algebraic sum of these results.
Col. 4 _ -1.10 '" valu" ill .00""",n 3 01 Ubi. A.5.7.
Col. 5 • eol""",,, (1) • oolWIIII (2) •• olumll (3).
Col. B • 8 '" valUH III .01"""," B af 1':lQI. A.5.5. Landing with .~~gular Acceleration
Col. 7 _ 1. 3;13 :l VlI.1ues In .0IW1111 9 of T:lQI.. -'5.6.
Col. 8 _ to 10 '" vaiUH ill .01"""," 6 at TaDl. AS.7.
Col. 9 _ col""", ... {B) • (7) • (8). In a level UL~ding condition. i t 1s some-
Col.10 _ 1.333 '" ..uuu In .01 ....... 14) of Tabl. AS.6. times specified that the horizontal component
ot the ground reaction must be a certain pro-
AS.19 Example of Fuselage Shears and Moments for portion of the vertical componeht, which causes
Landing Conditions. the line of action of the grO~~d reaction ~ in
Fig. AS.53 illustrates the airplane in a rig. AS.53 to not pass tr~ough t~g e.g. of the
level landing condition. The ground reaction is airplane, which creates an external pitching
assumed to pass the center at landing gear wheel moment on ~he airplane. This ~oment is us-
and c.g. of airplane. The fuselage Shears, mo- ually balanced by the inertia forces cue to
ments and axial loads are r-ecurr-ed when the the angular acceleration produced by the ..in-.
vertical ultimate load factor is 7. (Gross balanced moment about the e.g. The shears and
weight = 430ot). moments on the fuselage cue to this externa;
moment could be :aund as ex~la1ned in p~t.
SOLtlrION: AS.20.

The vertical or Z component of the ground
reaction R is specified as 7 load factors which
equals 7 x 4300 = 3010oi. One halt ot this is
acting on each Wheel.
The horizontal or X comDonent of R is 30100
tan 23° 425 x 30100= 12800# and acting aft.
The horizontal load factor on airplane
equals 12800/4300 = 2.98.
AS.20 Inertia Loads Due to Angular Acceleration.
In some of the flying conditions, it is
samet~es specified that the a1rplane ~ust be
SUbjected to an angular acceleration as well as
translational acceleration. ~is angular ac-
= cele~ation of the airplane produces :nertia
forces which must be calculated if the air-
plane is to be treated as a body in statiC
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A5.19

equilibrium. In some cases, a tail load due to

a gust on the tail is specified which produces a Fz ='M:y w Xc - - - - - - - - - - - - - - - (1 )
ly
moment about the airplane c.g. which produces
~~ar acceleratio~ of the airplane. In cer-
tain landing conditions, the ground forces do =My
I
w zc - - - - - - - - - - - - - - - (2)
y
not pass through the airplane e.g. thus pro-
ducing a moment about the e.g. which for stress From Table A5.9, I y = 16097600
analysis purposes is balanced by inertia forces.

Moment of Inertia of Airplane My was assumed as 100JOOO

The calcUlation of the moment of inertia at hence

an airplane about the center at gravity axes
was explained an page A3.5 at Chapter A3. A
detailed example solution was given in detail in
Table 6A or Chapter A3. The general equations
tor t ne moments ot inertia or the airplane about
the reference axes are:
Iy = Z WX ... Z wz
Ix =Z wy ...!. wz
I z =Z wy ... Z wx + Z A Iz
The last term in each at the above equa-
tions represents the moment of inertia of each
weight item about its own centroldal axes par-
allel to the reference axes.
AS.21 Solution for Inertia Loads Due to Unit 100,000 In.
Lbs. Pitching Moment.
To illustrate the general procedure ot de-
termining the balanCing inertia loads when the
airplane is SUbjected to an unbalanced moment
about the c.g., an ar~alYSls will be made for a
unit 100,000 in. lb. moment. Table AS.9 gives
the necessary calculations.
From kinetics:
Pitching angular acceleration a
I rac.zsec • :Z)
where
l'ly = unbalanced external pitching moment
" = 100000 w -r.; = .00621 w Xc
,z 16097600 ""
FX = .00621 w Zc •
Where Zc and Xc are the z and x distances of
z A Iy wei~~t w to the airplane e.g.
Columns No. 9 and 10 of Table AS.9 gives
Z A Ix the values 9f these inertia components. Fig.
A5.54 shows these inertia loads applied to the
fuselage. The reactions at Wing attachment
pomts Should be computed and then a table of
fuselage shears J moments and axial loads should
be made up. This unit table could then be used
for all conditions involving angular acceler-
ation ot the airplane.
It should be realized that the inertia
resisting loads in Table A5.9 are only approxi
mately, since the moment ot inertia neglects
the centroidal moment at inertia of the big
items, SUch as the power plant, wing, etc. The
example is only for the purpose) or illustrating
the general procedure at determining the inertia
resisting loads due to angular acceleration.
The same general procedure can be tollowed in
considering unbalanced external moments about
the Z and X axes, commonly referred to as yaw-
ing and rolling moments.
TABL.E A5.i
I 1
DJo1.AlCING INERTtA F'ORCES FOR maT 100,000 IN. loB. YOWENT
ABOUT Y AXlS TIlROUClll: AIRPLANE C.G. (PrI'CKlKG HOhlEHTJ

13j 3 1 4 $ 5 1 I • 10
!
about e.g. of airplane.
= pitching moment of inertia of air- II ~~.
~.
!~!I Arml
I"
Arml
Z :XI
Arm
z.,
i Arm
I ~ 1
". ~;:~I~;:;'I'
!x:.J. (~.~l""
I.. IFlI:
a
plane about airplane e.g. = Zwr 1.~1I 51. 3.1 I
i· ~.nn.s!
- 14.3l<u.e I
The tangential inertia force F for a mass 8:1.$ -1I.:l
wig due to an angular acceleration a equals J 110
200
IJlll
,711
~.~!170:
[4.701
5.4'1.00;
IS. il IU.SO e: n,sl 1011 \
1S50
2180000
~UOOO
130 : 10
: l 1 O ' 21
I 1. 00 1 .so i
10.40 142.50
13.0GI 250:
2.1.ta 173.$0
I 20460 i
30:nO I
204100
535000
i.1y
F = :!!..
g r-c , but a = 1y g 290 ,UBI 19. 10 I. '"
I 28.
501 :W3.50 JlGU
I
, 3110000
141.3 1_20.11

hence
'"
W'l..!lI
rroutC18IUi..12351_~'
Fual.
1:r.2. <10.0 • 315 - 0.1 3:17.50
1S01_[., 911_ I.G;

i
11.50
10;11-41.5 i.<11.50
710 1 - :n
i
I n <17.51 1.50
51100
1M
Zl20
313 i~
u"",",
133000
4il5ooo
::: l'"~:: I
• :5.5
1.1
51.3
U.3
My I
F = Ty 'N ~, wher-e r 1s the distance from I ,I,. : 11091100 1 •

; C<l1........ (2J, (3), 6 (4l T\U<8:Q fram

Tabla A:i.3.
z., , Z - 9.4 1 9H Tabla A:i.3 for e.,.
the weight w to the airplane e.g. I C<>lumQ {$)
Colwall
I C<l1W1111
(6)
(ill
x.., • X - 51.5 I
F• • . ooeu "lie
loeat1lla

~ C<l1W1l11 (10) F ·,oo51l .. Sc

It 1s convenient to t~eat the inertia force
F as resolved into two components Fx and Fz•
hence,
A5.20

III. Fig. AS.62 shows t~e ~lan f8~ sf ~ ca~:~­

11 50 aD, 120 170 230 lever wing. Ass~,e a ccnsta~t ~0~al
, , ,

, , distributed loac on :~e sur~~ce equal to

50 'cb . /sq. ft. ,";rit.e express i ens ~ cr shear
and bending moment en '.'ling and r inc 1[a2.':":'25
at 25, 100, lSO and 200 inches ~~JTI ene .
.1 I
r~ ~O"
\L~~~==-==:-=::::,~to
O"

1:12.0".
Planform Fig. AS. 62
10L
Fig. A5.54 _
1----- 200" --.....;
A5.22 Problems

(I) Draw the shear, bending moment and ax:al jX

load diagra~s for loaded structures 1~ -r- AS.63
."
Figs. 55 to BOc . I so"
90" Plan!orm
100 200 ~in. '5
rIO 1" 101 5,0 l"L -r-rTl.i
in
in
A,l.,=a, . -.l 0
a: .: Leading Edge

~ i 10'~soo 150"
I--- 20'~
o~i=:I:::J'=C=IIII::t[Il
30" (55) (56)
100 00 400 SO
1.
.L
t 1 ]lj'kS·3,0
I
Relative Span-
wise Distribution
1 1-2,..,' --I10j- }-2H I """'""1:20"1--

A AI L (56a)
~1O"f-
Fig. A5.63 shows plan ror-a of a cantilever 'Iring.
60 --j 10"
The total distributed air loa~ nc~al to sur~ace
.Cable is 10000 lb. The relative spanwise ~istriJut1cr.
is shown. Take center of ~ressure at 24 percent
400 (57) o~ chord fro~ leading edge. Divi~e wing i~to
'\.-.600 a" 10 inch width strips and calculate VZ 1 Mx and
~=;r===±i:==~' -L My, ane plot curves for same.
01" r't-12" 9"
AI
l' ~2 ,
IV. F:g. A5,64 shows an externally Jraced
monoplane wt ng , Take an aver-age 'Nir.g 11ft
500 lb. load of 90 l'J./sq.tt. norma I to wing with
center of pressure at 27 percent o~ the
chord from leading edge of wing and calcu-
late anc draw the tron0 and rear beam prl-
~ry shears and bencing ~c~ent d1a~rams.

100

M
I
Ii
60c
4 T
2.0"
,e..L
c
F.B. R.B.
(II) Draw bending ~oment diagram for structures
and loading in Fig. AS.51, abc •

•40011 Fig. A5. 64

'o'!....M'a. (0) 16"

100

4~~5°
\ 10';-
100f

,j I~O
/ 100
_ _ ::\450
iii .L (b) 1-12""",
1--16.. --l (c)
Fig. AS. 61
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A5.21

BENDING MOMENTS - BEAM - COLUMN ACTION

A5.23 Introduction deflections will gradually converge and the memb-

A beam-column is a member SUbjected to er will reach a state or equilibrium. These
transverse loads or end moments plus axial loads. secondary bending moments could be found by suc-
The transverse loading, or end moments, prOduces cessive steps by the various deflection princi-
bending moments WhiCh, in turn, produce lateral ples given in Chapter A7. However, for pr i smat i c
beams this convergency can be expressed as a
bending deflection of the member. The axial
~athematical series and thus save much time over
loads produce .~econdary bending moments due to
the axial load times this lateral deflection. the above successive step method. For ~embers of
Compressive axial loadS tend to increase the variable moment of inertia, the seCOndary moments
pri~ary transverse bending moments, where as
will usually have to be found by successive steps.
tensile axial loads tend to decrease them. If the end loads P are tension, they will
tend to decrease the primarY moments; tnua, in
Beam-column members are qUite crnmnon in
airplane structures. For example, the bearnE of general, the case or axial compression is more
externally braced wing and tail surfaces are important in practical design, since buckling
typical examples, the air loads producing trans- and instability enter into the problem.
verse beam loads and the struts introducing ax-
ial beam loads. In landing gears, one member is A5.25 Equations for a Compressive Axially Loaded
Strut with Uniformly Distributed Side Load.
usually subjected to large bending and axial
loads. In tubular fuselage trusses, ~ateral Fig. A5.66 shows a prismatic beam of length
loads due to installations supported on members L subjected to a concentric compressive load P
between truss joints produce beam-column action. and a uniformly transverse distributed load W,
In general, beam column members in airplane wi th the beam supported laterallY at each end,
structures are conrerat ivetv long and slender and with end restraining moments M1 and Mg. It
compared to those in buildings and bridges; is assumed that the general conditions for the
thus , the secondary bending moments due to the beam theory hold, namely; that'plane -sec t tons
axial loads are frequently of considerable pro- remain plane after bending; that stress is pro-
portion and need to be considered in the design portional to strain in both tension and compres-
at the members. sion.
This chapter deals briefly on the theorJ At any point a distance x tram the beam end,
of single span beam-column aember-e , A summary the moment expression is,
or equations and deSign tables is included to- 2
gether with examples of their use. The informa- 11=1'1.+ (118 L- M1 ) x -wLx
T +wX T - P y - - - - - (A5 . 1)
tion in this chapter is used frequently in other
chapters Nhere practical analysis and design of From appl1ed mechanics, we know that
beam-column members is considered. For a com- M =EI day
pleted and comprehensive treatment of beam-col- dx 2 therefore, differentiating equa-
umn theory and derivation of equations, see tion (AS.I) twice with respect to x gives
Niles and Newell-"Alrp~ane structures~.
2M
d P
dxa + EI M = w - - - - - - - - (A5.2)
A5.24 General Action of a Member Subjected to
Combined Axial and Transverse Loads. w w Fig. A5. 65
SUb-figure a of Fig. AS.6S shows a member p_ f f _pial
subjected to transverse loads wend axial com-
pressive loads P. The transverse loads W pro-
i // p,;1, .....,.., Ml I=:j L I,M2
"w//#&0z"" "-~t (bl P~I 4 \ n-p
duce a pr~ary bending distribution on the memb-
er as shown in Ffg. b. This bending will pro- P-r""--u-'i'''''"!-P (el ,,"',.L I - - l
, I ws Rua

duce a tr~~sverse deflection c~ve as illustrat- , "?;,.J"j "'''<.t'Ol'! l. -r I rnng Load

0<2<:;' . Jd l ~2la.stic Curve
ed in Fig. c. The end loads P now produce an
additional second~J bending moment due to the
rt:.. . . ·"JM~e",-:£ p_~_~_p

end load P tUnes the deflection 6 , or the bend-

P -,0.,0
I
_---_L-='l-p (el
\ • I
1
ing moment diagram of Fig. d. This first sec- o.ef~tl.,d .... t.,l.t s.."'><'d~1 H."'erot Fig. A5: 66
onda~J moment distribution produces the addi-
For s1mpl1t1catlon, let j =1 !fJ.pI; hence P 1
tional lateral deflection curve ot Fig. e and V-p EI ji"'
the end load P will again produce further bend- which, substituted in (A5.2)J gives
ing moments due to this defiection. If the ax- daM 1
ial load is not too large, these successive
a:xa +
F·!1 = w

-, ------~----------'~~~-
AS.22 BEAM -- COLUMNS

solution of this differential equation gives. 1 1M Ma -!"!:l. 'NLx 'NX a

x x y=~ -:l.+--L- X - T + T
I1=C:l. Sin] + c , cos] + wjll (A5.3)

where C1. and ell are constants of integration and x

sin x and cos x are ~~e limits at an infinite L
sin j
J J sin J
series of variable x. When x = 0, !1 = !"!:l. and (A5.7a)
J The slope at the elastic curve at any paint
when x = L, I1 = Ma , therefore: is given by the first derivative of equation
C _ Mil - wja !"h - wj.lll (A5. 7al
1. - si-n L
J
tan L
J i =
1(Mi
P - l-- I1:l. wL
L - - 2" +>'HX -
C:\.
Teas x c,
J + T 5i:1 x)
} - (_ ~\
\A;:;.Cj

(M:l. - wj.lll) cos L

J AS.26 Formulas for Other Single Span Loadings
sin L
J In investigating other transverse loadings
and Ga = M:l. _ Wjll
ror a single span carrying axial compression, it
is round that the expression tor bending moment
Let D:l. = I1:l.
- w.1:ll and Da = Ma - wj.lll. Then, in the span always takes the torm:
substituting in equation (AS.3),
M = C 1. S i n xJ + x + !'I'w)
C:a cos} ~ - - - I'~ 'I
~;:;.~

M where
t(w) is a ter::n which does not include the
sin L
load P or the end moments l11. and M:a. The
axial
J expressions tor r(w), C:\. and CII depend on the
To find the location of the maximum moment, dit- type of the transverse lead.
terent1ate equation (AS.3) and equate to zero. Table AS.I gives the ~alue of these 3
terms tor types or transverse loading on a Single
d11 _ c, X Ca sin '!:
dX - 0 = T cos J - J J span which are frequently encountered in airplane
structureS. The Table also gives equations for
whence the pcfrrt of maximum bending moment; and its mag-
nitUde.
L
x C D. - D1 cos - Table AS.II 15 a table or slnes, cosines,
tan - = --" = --,--"J IA5.5) and tangents tor L/.1 in radians which is more
j c, L
D1 sin 1 convenient to use than the usual type at trigo-
nometric tables. This table is based on values
The value or x must tall within x = 0 to x = L, given in Appendix ~ of Air Corps Information
otherwise I11 or Ma is the maximum value. Circular #493. The A difference have been added
The value of the maximum span moment can be to facilitate rapid use of the tables.
found by substituting the value trom equation For single span beams, the critical value
(A5.51 in (A5.41, which gf vee
of L/j is fi; that is, it the axial compressive
load is such that the term L/j = fi, the center
n.ax = D__ + wJ' - - - - - - - - - (A5.5) region of the beam will tend to deflect until the
cos x combined stresses equal the tailing stress of the
J material.
The moment M at any point x along the span can
also be written:
AS. 27 Moments for Combinations of the Various Load
I1=D:l. [(tan ~ . sin ~rcos ~]+Wjll - (AS.7) Systems as Given in Table AS. I, Margins of
J J J Safety. Accuracy of Calculations.
where Xm refers to the value at x where the span
moment is maximUm, or equation (AS.S). Since it The principle or superposition does not ap-
is customary to locate the point of maximum span ply to a beam-column, because the sum of the
bending moment and its value before investigat- bending momenta due to the transverse loads and
ing other S~l points, the value at tan !m is the axial loads acting separately are not the
J same as the moments when they act simultaneously.
known from equation (AS.S) and thus is available In combining several transverse load systems
to use in equation (A5.7) tor tinding moments at with their accompanying axial loads, the principle
other points along the span. ot superpOSition can be said to apply it each
If the equation tor the beam detlection is transverse loading is used with the total axial
desired, it can be found by SUbstituting the load for the systems which are being combined.
value of M trom equation (A5.3) in equation ThUS, in Ta~le A5.I, to tind the :noments tor
(AS.l), which gives: several combined loadings, add the values ot C:\.,
ell and t(w) for the several loadings and use
 ---_:.
ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AS.23

Table AS. I

Values of Terms ell Cz. and f(w) in Equation

M· C 1 sin..!. + Cz cos~+ f(w)

j j

Single Span - Axial Compression - Uniform Section

Eq. for Point of Max Eq. {or MaX. Span

Loading C, C2 f(w) Bending Moment Bending Moment
Equal End Moments. No
Ml,. Side Load M max = Ml
P"""'1r-- x _ L
.~2 o
x=L
2 ~
2j
i
Unequal End Moments. No Tan..!. = M2 - M1 cos~
M Side Load M2 j j

P+~~j====I~P
o
sin L M 1 sin..!::.
I- ~L . T j

Uniform Side Load. No

End Moments w ;,/ in. wj2 (cos L-l)
ttttttttrtTttft T x: .5 L
p- -p sin L
C:x~ .
l L • T
Uniform Side Load Plus DZ - D cos L/i
End Moments
em I J1 Tan..!.: D2 - D1 cos..!:.
j J Mmax'=~twi2
M"jf t II t t t f i I tit hM2 where wj2
-eD1 sin.,h
cos x
P+'!=~
o L
I""'P Dr:; Ml- wj2
D2 s M2- wj2
j T

x <: a, .Wjsinb
T o o
sin L
T
x o a, + Wj sin a - Wj sin.!. o
1"
tan L
)

T
Triangular Loading. No (NOTE A)
End Moments ..l. o To obtain Max1mum Moment, compute
p__ -c:rT r 1! t ~JI
w~in. moment at 3 or. " points in span. Draw a smooth
curve thru plotted results.
t--X~ L
Triangular Loading. No End

U ro-::::r
~ Momems
w./p'"iJTi I f
'Fx L
Couple Loading (Clockwise) x < a, - m cos b o o (See Note A)
T
sin L
T
X) a. - m cos..!.. m cos.!.. o
) J
tan L
T

w or W is positive when upward.

M is positive when it tends to cause compression on the upper
fibers of the beam at the section being considered.

Reference: ACIC '1'493; Niles, Airplane Design; Newell and NUes

Airplane Structures

For Table of many other loadings, see NACA T. M. 985.

 __________ 1IIl..,'~'1_

TABLE AS. II

llATUllAL SIlIES, COSIns , AIID TAIlGEllTll OJ' AIlGLEl! III IlADUIlI

t:. f::. f::.

~jj
L ill
jj Sill Sill Lj Cos Lj Cos Lj Tall Lj Tall Lj
Rad1allS j j j j j

3.00 0.14112 -0.98999 -0.14254

0.00991 0.00136 0.01019
3.01 0.13121 -0.99135 -0.13235

3.02 O. 2129
.00 92
-0.99262
O. 0127
-0.12219 · 10 6
0.00993 0.00118 0.01013
3.03 O. 1136 -0.99378 -0.11206
0.00994 0.00108 0.01011
3.04 0.10142 -0 99484 -0 10195
0.00996 0.00097 0.01010
3.05 0.09146 -0.99581 -0.09185
0.00996 0.00086 0.01008
3.06 0.08150 -0.99667 -0. 817
0.00997 0.00077 0.01006
3.07 0.07153 -0.99744 -0.0717
0.00998 0.00066 0.01004
3.08 0.06155 -0.99810 -0.08167

3.09 0.05156
0.00999
-0.99867
O. 0057
-0.05184 · 003

0.00997 0.00046 O~
3.10 0.04159 -0.99913 -0.04162
0.01000 0.00037 O.
3.11 0.03159 -0.99950 -0.0318
0.00999 0.00027 0.01001
3.12 0.02160 -0.99977 -0.02180
0.01000 0.00018 0.01000
3.13 .0 16 . 9 3
o 00007
-.
3.14 0.00160 -1.00000 -0.00160
0.01001 0.00003 0.01001
3.15 -0.00841 -0.99997 0.00841
0.0100e 0.00014 O.
3.16 -0.01841 -11.99984 00.01841
0 9 3
3.17 -0.02840 -0.99960 00.02841
,
3.18 -0.03840
0.01000
-0.99926
0.00034
0.03843 i · 2

3.19 -0.04839
0.00999
-0.99883
0.00043
0.04845 I ·
0.00999 0.00083 0.01003
3.20 -0.05838 -0.99830 0.05848
0.00998 0.00064 0.01004
3.21 -0.06836 -0.99766 0.06852
0.00997 0.00073 0.01005
3.22 -0.07833 -0.99693 0.07857
0.00996 o 00084 0.01007
3.23 -0.08829 -0.99809 0.08884
0.00996 0.00093 0.01009
3.24 -0.09825 -0.99516 0.09873
0.00995 0.00103 0.01010
3.25 -0.10820 -0.99413 0.10883 I
 c
65

AS.28 BEAM -- COLUMNS

these values in the general expression tor M as of the bea~, or 4420 x .75 -= 3315"# positive be-
given at the top of the Table. cause it procuces compression in the top tibers.
In a beam-column member, the bending mom- The moment at (2) due to the cantilever overhang
ents do not vary directly as the load 1s i~crea equals (20 + 10) 36 x 16 = 8640"#. Fl§;. A5.:58
ed. Thus, the student should realize that marg- 2
ins or safety based on direct ?roportton of mom- shows the beam ?ortion between points (1) and (2)
ents to loads are incorrect and lie on the un- as a free body.
safe side. From Art. AS.Z5, we have the follOWing pre-
It is recommended that four significant cise equations for a beam carrying a tr-ansverse
figures be used in computations, making use ot uniform distributed load with end compressive
the so-called precise equations, since the re- loads.
sults in many cases involve small differences x L
between large numbers. tan J D~ - Dl cos J - - - - - - - - -(A)

AS. 28 EDmple Problems D.. sin L

Example Problem #1 J
and
Fig. A5.S? illustrates a typical upper, •
outer panel Wing beam of a biplane. Let it be !'l
max
= ...!2>..-
cos x
+ w.:; - - - - - - - - - - -(E)
required to determine the maximum negative
bending moment between paints (1) and (2), gen- J
erally referred to as the maximum span moment. Evaluating ter:ns tor substitutIon in these equa-
To obtain the true bending moments on the beam, tions, we obtain,
the axial beam load as well as the end moments Ml 3315"#
at (1) and (2) are necessary since they influ-
ence the deflection of the beam. Mil 8640"#
Solutlon:- P = 4420# compression
To obtain the horizontal component Th or I -= 10 in~ given and assumed constant
the lift strut load, we take moments about the throughout the span.

roo- ----I-- 36''---1 =1/l.3XIoeXIO = v294I 54.23

V- 4420
w= 2Q.IP/in. I lOll/in

P-';"~I~~;b~g;~g~~~~~t'~~B~~.J...
.A. Wood Sea 2
Wj

D1.
"= 20 x 2941
M1 - wja
= 58820
= 3315 58820 55505

~...<;e lSOQ.IP=Reaction
DII = M.. - wjll 8640 58820 50160

'S:'~
,0' from Strut to
Lower Wing
L 100
J = 54.23 = 1. 844
'\..!~'\.
From Table A5.II sin J = .96290 and cos J -= .... 26981

v Fig. AS.67 Substituting In equation (A)

x L
tan J = 011 D1 cos J

...... w= 2Of/in.
D1 sin J
50180 - (-55505 x - :26981) -65156
1.2192
-5~505 x .96290 -53441
x tan-' 1.2192 = .88383
J
Fig. AS. 68
Hence, x = .88383 x 54.23 = 48 ft , which equals the
distance trom the lett end at the beam to the
hinge point at the lett end. point at maximum span moment.
l:l'lb 2000 x 50 - 540 x 116
- 1500 x 100 + 70.75 Th = 0
= ..lL:-
cos x
+ «r , cos ~ = .63419 from Table
j
A5.II
hence J
Th = 442Qj1
Hence 55505
The axial compressive load induced by the lift
strut at pOint (2) then equals - 4420#. ~ - :63419 + 58820 28,700"#

Taking ZH = 0 tor the load system or Fig. To obtain an idea as to the magnitUde at the
AS.S? gives P = - Th -= 4420#. The end moment on secondarY bending moment, that is, the ~oment due
the beam at (1) equals the end load timeB the to the axial load times the lateral beam deflec-
eccentricity or the hinge :ram the neutral axis tion, the primary bending moment at a point 48~
 -,,;; . r

ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AS. 29

tram the left end w111 be computed. substituting values of C1 and Cm and f(w) :rom
Table A5. I in the above equa.t i ons ;
M..II := 3315 + 48x20x 24- 940:<48:= -18765"#
Thus the secondary bending moment equals I1= (M .. -!11. cos L/j) sin x/j +M cos x/j
sin L/j 1.
- 28700 + 18765 = - 9935"# which 1s a large per-
centage of the primary moment. The transverse
deflection of the beam at the po1nt of max. span 20, OQQ!...lF==={~2='=!=2="=0=6=3=8=,,="=1=T=U=be=J~02~~000t
moment then equals - 9935 := 2.25 inches upward. . L: 41. 762 ,(
- 4420
Bending Moment at any Point Along Span Fig. A5.70
(j) ®
Let the moment at a pcfrrt 10" from paint
(2) be required. In this case, x = 100 - 10 = 90 But, M:a. = 0 in our problem, henes,
M:= Dl.~tan ~ . sin })+ cos
L .
11
+ wja (Ref. Eq. I1 = Masin x/j
A5.?) sin L!j
x 90 x
- - = 1. 6596, sin j- .99605 =" {Ei:~ /"'2"'9"'X'l"07''''X-."746' =' ~ = 25 826
J 54.2:3 VP - V 20000 V66'l •

cos x = - .08867 =~=1

J Llj 25 .826 • 617 - - - sin L/j = .99892
tan .s =1.2192 =value tor x at the point of
x = Ll2 = 20.881
j maximum bending moment
Hence, x!j -- 25.80
20.881 -- .8085 ern x/j = .72327

58820 = - 3664'"
.
M =-55505 [(1.2192 x .99605) +-.08867] ...
suost i tutlng in the above equations tor 11 /'
E:x:ample Problem #2 M = 36000 x .72327 = 26066""
Fig. A5.59 shows a simplified landing gear .99892
structure carrying a vertical load of 12000# on This compares with a prtmarY moment of
the axle. M.ember ABC is continuous thru B and 36000/2 ~ 18000"#. The deflection at the mid-
ptnned at C. Let it be recurred to determine the point or BC := 26066 - 18000 = .403 In.
bending moment at the midpoint or member BC and 20000
its lateral deflection due to the 12000# verti- The maximum moment is gl ven by the equation:
cal design load.
~ =s~n L and it occurs at x =n2j (See
~ 1 26" ~ Ie J Tabla A5.I)
~D ~
"0:"------;;:86:::2"."'" "--------1- A5. 29 Stresses Above Proportional Limit Stress of
Material.
12
The equations as presented in this chapter
4 assume that E is constant or in other words the
stresses are Within the elastic range. In air-
crart structural destgn the applied or limit
21/2-083. loads must be taken without suffering permanent
Round
Steel 10 deformation, hence E 1s constant under such
loads. However the aircraft structure must
Tub" I take the deSign loads which equal the limit
loads times a factor at safety (usually 1.5)
Fig. A5.69 without failure. In many cases structural
failure will occur under stresses in the plastiC
Line of action of DB goes through E
AxlJL range where the material stiffness Is less and
BiE,,( A not constant.
"""'"3 2000*
A good approximation for ~~ effective
Solution:- modulus E' Is obtained as :ollows:-
Solving for reactions at C by staticS, we
obtaIn :he axial load in Be = -20000. The (1) Compute Fc = PIA for the given number.
bending momen~ at 8 due to 3" eccen~ricity of
the 'wheel load = 3 x 12000 := 36000"#. (2) With this value of Fc enter the basic
Fig. A5.70 shows a r~ee body of portion BC column curve diagram for the given material (for
of member ABC. From Table A5.1 end fixity C = 1) and find value of L'/p cor-
M = C~ sin xj + ,v~ cos x
} + '()
~ w responding to the stress Fc'

~~--~-------~~
 AS. 30 BEAM -~ COLUMNS

(3) USing these values of L' /p and FC J (4) Determine the bending ~oment at the
compute centerline of the beam-c01~~s shown in Fig.
Assume ~I =
64 J OOO, 000 lb. in. sq.
"' -""Tti"
to
_Fe (L')'
P
A5.74.

(4) Then i > (E~ry/'

Basic column curves for various ~terials 1!
are given in another chapter of this book. I
If
AS.30 Problems •
800<1
Fig. A5.75
500< ·(5) For the beam-column in Fig. A5.75
H155"'-!-- 35" ---I W= 30<1/in. calculate the bending moment of the centerline
6000* L--l.-- I sooo " t t t t t .. ·BOOot of the member. Assume E = 1,300,300 psi. and
1 - - .JooOit---4 . - ;-- I = 10 in~
I 50" I 1-30''t- 80''---,
300<1
Flg. AS. 71 Fig. AS.72
(1) Fig. A5.71 shows a 1-1/2 - .065 steel
tube subjected to both end and lateral loads.
Determine the max~um bending moment on the j ;,000#
tube. Campare the result with the bending
moment due to the side load only. E = 29 X 10 8
,

~
50" ----1 18000"-IF

ps i . I of tube = .075 in. 40 Compute lateral 500.

deflection at point of maximum bending moment. Fig. AS. 76
(2) The beam column member in Fig. A5.72 is (6) For the beam-column loading in Fig.
made of 24ST aluminum. alloy. Calculate and A5.76, calculate bending moment at center
plot a curve of the bending moments on the point ot beam. Take S • 1,200,000 psi and
member. Also plot bending moment due to I = 10 In~
lateral loads only. E = 10.3 x 10" psi. I::::
5.0 in .... AS.31 Beam-Columns in Continuous Structures.

(3) Determine the maximum. bending moment The secondary moments in a partic~lar
for the wood wing beam and loading of Fig. member due to beam-column~action also effect or
influence the deflections in adjacent members
A5.73. I of beam section = 17 in.· E 1.3 x = of a continuous structure. This rat~er involved
10".
problem can be handled quite simply and rapidly
by the moment distribution method as explained
and illustrated in Arts. All.l2 to 15 of
I cnap'ter All.
t f ! t
p-r* V

10001 20007

s.
"000..
.
t-<."+-:""--.
#rlll
:-
.
I I I tIl I
""-j
tyeoo IO<I/in.
tt I tl l!
0'"

1~ I 100001
200" - - - - . - . "

Flg. AS. 74
 CHAPTER A6
TORSION.. STRESSES AND DEFLECTIONS

A6.1 Introduction. sumed that any radial line undergoes angular

Problems involving torsion are common in displacement only, or OB remains stral&~t when
aircraft structures. The metal covered airplane moving to OB' •
wing and fuselage are basically thin-walled The unit shearing strain in a distance L
tUbular structures and are subjected to large equals,
torsional moments in certain flight and landing
conditions. The various mechanical control sys-
tems in an airplane often contain units or
various cross~sectlonal shapes which are sub-
1ected to torsional forces under operating con- Let G equal modUlus at rigidity at the
ditions, hence a knowledge of torsional stresses ~terial and let ~ equal the unit Shearing
and distortions of members Is necessary in air- stress at the extreme fiber on the cross-sec-
craft structural design. tion.
A6.2. Torsion of Members with Circular Cross Sections. Hence, ~ = E G = r 9 G _ _ _ _ _ _ _ _ _ _ (1)
L
The following conditions are assumed in the
derivation or the equations for torsional
stresses and distortions: - In Fig. A6.2 let ~p equal the unit Shear-
ing stress on a circular strip dA at a distance
(1) The ~ember is a circular, solid or hollow p from O. 'fhen
round cvi incer .
p __ prQG = p 9 G
(2) Sections remain circular after application "0 =-'t" -L-
of torque. , r r L

(3) Diameters remain straight atter twisting of The ~oment of the shearing stress on the
section. circular strip dA about 0 the axis of the bar
(4') Material rs homogeneous, isotropic and 1s equal to,
o'ig}dA
elastic. dM .. -r pdA
c = '--L- and thus the total in-
(5) The applied loads lie in a plane or planes
perpendicular to the axis of the shaft or ternal torSional resisting moment is,
cylinder. p
ao"QdA
\nt.
,
\
L

B'
=
r--L-

\
For equilibrium, the internal resisting
A moment equals the external t or-ai onaj moment T,
- i
and since GQ/L is a constant, we can write,
,,
J
Fig. A6.1
,
p
-- .T = Mint. =
GQ
L Jr
0
2
p ctA = T
GQJ (2)

Where J = polar moment of inertia of the sr~ft

cross section and equals twice the moment of in-
ertia about a diameter.
Fig. A6.2
From equation (l) GQ = ~
L r
-r J
Fig. A6.l shows a straight cylindrical bar Hence, T = r (3 )
subjected to two equal but OPPOSite torsional
couples. The bar twists and each section is or (4 )
SUbjected to a Shearing stress. Assuming the
left end as stationary relative to the rest of also from equatfon (2), solv:ng for the ~Nlst 9,
the bar a line AB on the surface will move to
AB' under these shearing stresses and this ro- Q = GJ
TL (5 )
tation at any section will be proportional to
the distance from the fixed support. It 15 as- (9 1s measured in radians).
A6.1

C-,.'
vb
AB.2 TORSION

AB.3 Transmission of Power by a Cylindrical Shalt. ?;,obl~m 2.

The done by a tWlst~n~ 8cu~le T 1~
~ork ?i s : AE.4 111ust~ates a~ a:lercn control
~ovl~g tn~ough an a~~~lar jis~l~c?~~~t :s ~qual s~r~~C3, consisting of a circ~lar tor~~e tube
to th3 ~roduct of the ~~gr.it~de :f the cJu~le (1-1/4 - .C48 in Size) suppo:,ted on :~ee
and the angu lnr c i s cracement in rao Lans . I~ tne hi~ge ~rack3~s ~~d ~lt~ tte c:ntr:l ~od ~:tt:~g
angular dt s pLaceasnt t s one r-ev -jut t on , ::18 ',II'::'!:. attac~~c ~c ~he tor~ue t~be above the center
done equals 2 n T. I: T is ex:ressed in lr.ch- SUDJOrt bracket. F:nd the 8~xi~lli~ torsional
pounds and N Is the~~ul~r ,elccity in revol~­ shear1ns stress in :he tu~e i t the air lo~d on
tions per ~inute, then the horsepower tr3ns- the ailer:n is as indicated in Fig. A6.4, and
~itted by a rot~ting shaft ~y be written, also compute the an~le sf twist of tute be-
tween horn section anc end of aileron.
. , r----------------" Pressure
2 n N T
H, P. = 396000 ( 0,
istributiOn4
Average load on surfac e e 40#/0' " "'-' i
rr
.
1\ :
! I\ :::
where 396000 represents inch ~o~~ds of work of
one horsepower fer ~ne min~te. Equation (6; _ .:'li· '0495ETUbe(.A~~~~~ _~_ Tube t ,I I, I

may be wr-ttten: . c ~ - - - - -=-=r~e.:r

T = H.P. x 396000
2nN
63025 H.P.
N
(7) 5" L
I
24" i 24" -15" L
Fig. AB.4
EXAMPLE PRCBLillS.
Problem l. SOLUTION:
Fig. A6.3 shows a conventional central
stick-torque tube operating unit. For a side airload on the sur~ace ~er.ds to rotate
~he
the ailer8n ~round the tC~Gue t~be. ju~ nove-
load of 150 Ibs. on stick grip, determine the
shearing stress on aileron torque tube and the ment is prevented or crea~ed 2y a control rod
angle of ~Nist between points A and B. attached to the torque t~be over the center
supporting oracket.
SOLUTION: The total load 0n a 5t~ip of aileron cne
lucre wide = 40(15 x 1/:'44) ~ 4.16 lb.
Torsional moment on tube AB due to slde Let w equal intensity of loading per inch
stick force of 150# ~ 150 x 26 = 3900 in. lb. of aileron span at tha.,leading edge point cr the
The resistance to this torque is provided by the aileron surface. (see ~ressure diagram in
Fig. A6.41.
150*
Then 3. + (0.5 w)12 = 4.16
J24 ST At Alloy
,...control Stick l G ::: 3,800,000 psi
hence w = 0.463 lb.
The total load forwarG of the center-
p~

ilt----2.'· - - - 1 line at torque tube = 0.463

x 3 = 1.389 lb. and
P a the load on aileron portion att of hinge line
.. EI~v. Control Wire I ",.IAil 0 Ho ~ 0.463 x 0.5 x 12 = 2.778 lb .
....--- 356J1J1 II er n rn
Ai}eron Operating Tube ~ l
B r .J:
? ~
Bearing
Fig. AB.3
aileron operating system attached to aileron
horn and the horn pull equals 3900/11 ~ 356 lb.
The polar ~oment of inertia of a It - 0.058 The torsional ~oment per r~nlng inch or
round tube equals 0.1368 in 4 • torque tube: = - 1.389 x 1.5 + 2.7~8 x 4 = 9.0
in. lb. Hence, the maximum torque, which occurs
Maximum Shearing stress = ~ - Tr/J = at the center of the aileron. equals 9.0 x 29 =
(3900 x 0.75)/0.1368 = 21400 pSi. 261 in. lb.
The angular twist of the tube between
J = 2510.06678
points A and a eq~als
_ Tr x 0.625 = 2450 ps t .
1:(:nax.) -

~ - TL _ 3900 x 28 = 0.21 radians (J 0.06678 in 4 . )

- GJ - 3,800,000 x 0.1368
or 12 degrees. Since the tube section Is constant and ~he torque

ANALYSIS AND DESIGN OF
varies cirectly as the cistance :rom the end of
the ailerons, the angle of twist g can be CQm-
?uted by using the average torque as acting on
entire length of the tube to one side of horn or
a distant J.... = 29", :tence
261 x 29
2 x 3800000 x 0.06678
AB.4 Torsion of Members with Non-Circular erose-
sections.
The for-~las derived in Art. A6.2 cannot be
~sed for non-c1rcular Shapes since the assump-
tions made do not hold. In a circular shaft
subjected to pure torsion, the shearing stress
distribution is as indicated in Fig. A6.5,
namely, The maxi~um Shearing stress is located
at the most remote fiber from the centerline
axis of the bar and is perpendicular to the
radians to the stressed paint. At a given dist-
ance from the axis of rotation the shear stress
Fig. AB.5
-E-P=+ Fig. AB. B
Ie :;;;::-_-~~I
"': max.
is constant in both directions as illustrated in
Fig. A6.5, Which means that ends of segments of
the bar as it twists remain parallel to each
other or in other words the bar Sections do not
warp out of their Plane when the bar twists.
If the conditions of Fig. A6.S are applied
to the rectangular bar of Fig. A6.S, the most
stressed fibers will be at the corners and the
sGress will be directed as shown. The stress
would then have a component no~l to the sur-
face as well as along ~he surface and this is
~ot true. The theorj of elasticity shows that
the maximum shear stress occurs at the center-
line of the long sides as illustrated in Fig.
A6.S and that the stress at the :orners is zero.
Thus When a rectangular bar twists, the shear
stresses are not constant at the same distances
from the axis of rotation and thus the ends of
segnents cut trIough the bar would not remain
~arallel to each other when the bar ~Nists or in
other words, warping of the section out of ~ts
plane takes place. Fig. A6.7 il1ust~ates this
action in a twisted rectanb~lar bar. The ends
of the bar are warped or suffer distortion
normal to the original unstressed plane of the
bar ends.
Further discussion and a summary of equa-
tions for determining the shear stresses and
'n
FLIGHT VEHICLE STRUCTURES AB.3
twists of non-circular cross-sections is given
in Art. A6.5.
A6.5 Elastic Membrane Analogy.
The \Shape of a~watD~a cross-section of a
non-circular cross-section in torsion is
needed in the analysiS by the theory of elas-
ticity, and as a rasult only a few Shapes stich
as rectangles, ;eJ:,lJ2.-seS, triangles, etc., have
been solved by the theoretical approach. How-
ever, a close approximation can be made ex-
perimentally for a Iznos't any shape of cross-
section by the use of the membrane analogy.
It was pointed out by Prandtl that the
equation of torsion of a bar and the equation
for the deflection of a ~embrane subjected to
uniform pressure have the same form. Thus if
an elastic membrane is stretched over an open-
ing which has the same shape as the cross-
section of the bar being considered and then if
the membrane 1s deflected by subjecting it to a
slight difference of pressure on the two Sides,
the reSulting deflected shape of the membrane
provides certain quantities Which can be mea-
Sured experi~entallY and then used in the
theoretical equations. However, pOSSibly the
main advantage ot the membrane theory is, tr~t
it provides a method of visualizing to a
considerable degree of accuracy haw the stress
conditions vary over a complicated cress-section
Fig. AS.7
of a bar in torsion.
The membrane ar~alogy provides the follaN-
ing relationShips between the deflected ~em­
brane and the twisted bar.
(1) Lines of equal deflection on the membrane
(contour lines) correspond to Shearing
stress lines of the twisted bar.
(2) The tangent to a contour line at any point
on the membrane Surface gives the direction
of the resultant Shear stress at the corre-
sponding ~oint on the cross-section of the
bar being twisted.
.,,,.>~
(3) The max~ slope of the deflected membrane
at any paint, with respect to the edge
support plane 1s equal in ~gnitude to the
shear stress at the corresponding point on
the cross-section of the twisted bar.
(4) The applied torsion on the ~Nisted bar is
proportional to ~N1ce the volume inclUded
be~Neen the deflected membrane and a plane
through the supporting edges.
To illustrate, consider a bar with a
rectangular cross-section as indicated in Fig.
A6.B. Over an opening of the same sr~pe we
stretch a thin membrane and deflect it normal
to the cross-section by a small uniform pres-
s~re. ~qual deflection contour lines for this
deflected ~embrane will take the sr~pe as il-
lustrated in Fig. A6.9. These contour lines
Which correspond to direction of shearing
stress in the twisted bar are nearly cirCUlar
near the center region of the oar, but tend
Sf
A6 4 TORSION

to take the shape of the bar boundary as the From Table A6.l it is ~otised that :or
boundary 1s approaChed. ~lg. A6.Sa shows a sec- large values 0:' bit, tne values 0-:: t he con-
tion through the contour lines or the ~erlected stants 1s 1/3, and t~us for such narrow rec-
membrane along the lines 1-1. 2-2 and 3-3 of Fl~. tangles, equations (6) and (7) reduce to,
AS.9. It is obvious that the slopes of the de-
tlected surface along line 1-1 will be greater _ _ 3 T
bt:l (5 )
than along lines 2-2 or 3-3. From this we can "MAX -
conclude that the shear stress at any point on
line 1-1 will be greater than the shear stress _ 3 T
g - bt 3 G (9 )
tor corresponding pOints on lines 2-2 and 3-3.
The maxlmu~ slope and therefore the maximum
, , •
,
,
,:>
,
Although equations (8) and (9) have been
derived for a narrow rectangular shape, they
can be applied to an approx~mate analysis of
I ,

~
~. __I ._.
\ \
shapes made up of thin rectangular members
such as illustrated in Fig. A6.10. The ~cre
generous the fillet or corner radius, the
I smaller the stress concentration at these jun-
ctions and therefore the more accuracy of these
,,! ,,

/
approximate formulas. Thus for a section made
3
I .J up of a continuous plate such as illustrated in
•
Fig. A6. a Fig. A6.9

.LSlope stope Slope

l~l 2~ 3~
Fig. a
Fig. A6.9a b can be taken as centerline length for above type
of sections

stress will occur at the ends of line 1-1. The .r-: b " --! ..J::t "
slope of the detlected membrane will be zero at
the center ot the membrane and at the four ~ET
b.
corners, and thus the shear stress ~t these
points will be zero.
...
fo-bl.~~tl.
A6.6 Torsion at. Open Sections Composed of Thin Plates. Fig. A6.10
Members having cross-sections made up of
narrow or thin rectangular elements are some- Fig. (a ) of Fig. A6,.lO, the width b can be taken
times used in aircratt structures to carry tor- as the total lteiigthof the cress-section. For
sional loads such as the angle, channel, and Tee sections such as the tee and H section in Fig.
Shapes. A6.l0, the polar moment of inertia J can be
_;"'~'Ir-For a bar of rectangular cross-section or
taken as Z bt~/ •. Thus for the tee section of
width b and thickness t a mathematical elastiCity Fig. A6.10:
analysis gives the following equations tor maxi-
mum shearing stress and the angle of twist per 9
m
=-'- = G 3Z Tbt 3
un1 t length. OJ

= T a (6 ) = 3 T
a b t G (blt~ + bat~)

"'0' -- ~ b T For the maximum Shearing stress on leg bl

t'" G 1n radl ana - - (7 )
_Tt l _ 3 T t l 3Tt l
Values or a and ~ are given in Table A6.1.
'tb l --J--
.fb'tT=blt l S T b:lt a " - - - - (10)
and tor the ?late bar
_Tt a _ 3Tt a (
'toa - - J - - bltl S + b a t a 3 - - - - - - - - 11)
TABLE AS.l
CONSTANTS (lAND ,
It t 1 = t , = t, then
bit 1.00 1.5Q 1.'1'S .2.00 2.50 3.00 4 S a 10 I 0 0 9 = 3(b
3 T - - - - - - - - - - - - (12)
c, O.:W8 0.231 O.~O.2~~~O.~o.m~~~WO.3130.m G t l + ba)
0.141 0.19 0.214 O. O. 49 O. 3 O. 1 O. .299 0.307 0.313, O. 333
1:' = 3 T - - - - - - - - - - - (13)
t 1 ( b 1. + 011)
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A6.5

EXAMPL" PROBLEM SHOWING TOPBIONAL STIFFNESS OF
CLOSED THIN ',.JALLED TUBE COMPARED
TO OPEN OR SLOTTED TuBE.
Fig. A6.lla shows a 1 inch diameter tube
with .035 wall thickness, and Fig. A6.11b shows
the same tube but with a cut in the wall making
it an open section.
For the round tube J. = 0.02474 in~.
For open tube 3 2 = ~ x 3.14 x .035~
Fig. AG.lla
Let Q1 equal twist of closed tube and Q 2
equal twist of open tube. The VNist, will then
be inversely proportional to J since
Therefore the closed tube is Jl/J a = 0.02474/
0.000045 = 550 times as stiff as the open tube.
This result shows why open sections are not ef-
ficient torsional members relative to
torsional deflection.
TABLE Ae.3
AG.7 Torsion of Solid Non-Circular Shapes and Thick-
Walled Tubular Shapes.
Table A6.3 summarizes the for.nulas for
torsional deflection and stress for a few
Shapes. These for~ulas are ,based on the as-
sumption that the cross-sections are free to
warp (no end restraints). Material is homo-
geneous and stresses are within the elastic
range.
A6.8 Torsion of Thin-Walled Closed Sections.
The structure of aircraft wings, fuselages
= 0.000045 and control surfaces are essentially thin-walled
tubes of one or more cells. Flight and landing
loads often produce torsional forces on these
major structural units, thus the determination
at the torsional stress and deformation of such
structures plays an important part in aircraft
structural analysis and design.
Fig. A6.12 shows a portion of a thln-
walled cylindrical tube which is under a pure
Fig. A6.l1b torsional moment. There are no end restraints
on the tube or in other words the tube ends and
tube cross-sections are free to warp out of
their plane.
Q = G~J'
~qds

T
FORMULAS FOR TORSIONAL DEFl.ECTION A.ND ,rrRESS

e :: KG :: tw1.1lt l.n I"ll.dJans per meh of length.

T :: Tors1onal Moltlent (In. lb.).
~ q constant

G :: Modulus oJ. R~id1ty.

K (in4) Froltl Ta.b1e. Fig. A6.12 Fig. All. 13

FORMUloA FOR
SECTION K SHEAR STRESS

SO
SOLID ELLlPITICAL

'T
'tMAX :: ft'iijr (1Lt elida
of minOr ana).
(b) •
Let ~ be the shear force intenSity at point
(a) on the cross-section and qb that at point
Now consider the segment a a ~ b of the tube
wall as shown in Fig. A6.l2 as a free body. The
applied shear force intensity along the segment
edges parallel to the y axis will be given the
SOLID SQUARE.
values ~y and qby as shown in Fig. A6.l2. For
T

D1
SOLID RECTANGLE.
K tO.141a 4 't MAX :: O.208 aJ (at
m1dp::lint of each 'ide).
a plate in pure shear the shearing stress at a
point in one plane equalS the stress in a plane
at right angles to the first plane, hence
qa = qa and qb = qb .
Y Y
ao3[~ 3.3e~(1 = T(3a Since the tube sections are free to warp
CJI K t _ • 'tM.\X - 1. 8bl
8a.2t)%
mtdpo1Dt of 10ll(side. there can be no longitudional stresses on the
2a--., tUbe wall. Considering the eqUilibrium of the
~
~] segnent -in the Y direction,
SOLID TRIANGLE. ~Fy = 0 = Qa~qb;"
- =
0, hence Qa y qby and =
6~ therefore Ga = qb or in other words the shear
torce intenSity around the tube wall is con-
stant. The shear stress at any point T = qjt.
It the wall thickness t Changes the shear stress
For an extensive U,t of formula. for many ~, both .soUd and bollo..,
refer to book, "Formulas For StreBB and stl"ll.in" by Roark, 19!14 Ed.itiOQ.
A6.6 TORSION

changes but the shea~ force q dOes not change, T~e elas~lc strain energy dU stored in :his
or element therefore equals,

~ata = ~btb = constant. dU = 24CS -:5

The prOduct ~ is generally refe-rred to as

the shear flow and is given the symbol q. The nowever the shear strain 6 can be written,
name shear flow possibly came from the fact that
the equation ~ = const~~t, resembles the equa-
tion of continuity of fluid flow qS = const~~t
6 =~G :::;...9:....
Gt' but 4 :=
T"
2A
where q is the flow velocity and S the tube
cross-sectional area.
hence em : :; --
8A.L'IGt cs

ds
We will now take moments of the shear flow
q on the tube cross-section about some point (0).
In Fig. A6.l3 the force dF on the wall element
=qds , Its arm from the assumed ooment cen- or U =
.- cs , 'ehe mt egraI y
f 8Jl:""Gt
ter (0) is h. Thus the moment of dF about (0)
is q cs h. However, ds times h 1s twice the area is the line integral ar8und the periphery of
of the shaded triangle in Fig. A6.13. the tube. From Chapter A7 from Castigliano's
Hence the torsional moment dT of the force theorem.
on the element ds equals, au _ ~ cs
dT • qnds • 2qdA
f"I
'7
_
- aT - r 4A,.T OIGt ~ - - - - - - - - - - - (17)

since all values except t are constant. equa-

and thus for the total torque for the entire tion (17) can be wTitten,
shear flow around the tube wall equals,
Q ~ 4A,.IlG
T
JL~ 15t - - - - - - - - - (~8)

T = fA 2qdA and since q is constant and since T =2 qA, then slso,

Q=~Gfd:- - (-19)
T = qU (14) Where Q is angle of twist in rcdlans per lli~it
or length ot one ir.ch cf tube. For a tube length
at L
T \
q = 2A (15)
Q=i<~f~s-------- ----(20)

where A is the enclosed area of the mean periph-
ery of the tube walL
The shear stress ~ at any point on the tube
'NaIl is equal to q, the shear force per inch ot
wall divided by the area ot this one inch length
orlxtor
(16)
TUBE TWIST
Consider a small element cut from the tube
wall and treated as a free body in Fig. AS.14,
With ds in the plane at t~e tube cross-section
and a unit length parallel to the tube axis.
Under the Shearing strains the plate element
AG.9 Expressi.on for Torsional Moment in Terms of
Internal Shear Flow Systems for Multiple Cell
Closed Sections.
Fig. A6".l6 shows the internal shear f Low
pattern for a 2-cell thln-walled tube. when
the t~be is subjected to an external torque.
q~, qa and 43 represent the shear load per inch
on the three different portions of the cell
walls.
For equilibrium of shear ~orces at the
junction point of the interior web With the out-
Side wall, we know that
------(21)
m

T ,
<Is '0----1
to- 1 -0\

1. t
I
>0- 1 --<

qds~D: t ,
t
\
----I Fig. AS. 16
Fig.A6.14 Fig. AS. 15 n
detorma as illustrated in Fig. AS.l5, that is, Choose any ~oment axis such as paint (0).
the tace a-a moves with respect to face 2-2 a Referring back to Fig. A6.l3, ~e found that the
distance O. The torce on edge a-a equals q ds moment of a constant shear force q acting along
and it moves thrOU~~ a distance O. a wall length ds about a pOint (a} was e4~1 in
 _ _ _ _ _ _ _ _.........··iII'itMTlllil'i

ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A6 7

magnitude to twice the area ot the geometrical
Shape formed by radii from the moment center to
the ends of the wall element d5 times the shear
flow q.
Let To • moment ot shear flow about point
(0). Then from Fig. A6.16,
inch in the web 1-2 and (qe - q~) ~or web 2-3.
For eqUilibrium, the torsional moment af
the internal shear system must equal the ex-
ternal torque on the tube at this particular
section. Thus. from the conclusions of article
A6.9, we can write:

(24)

For elastic continuity, the twist of each cell

must be equal. or 01. = c, = Q~.
But from equation (21), qa = q~ - qao From equation (19), the angular ~Nist of a cell
Substituting the value of qa in (22) Is
Q=-q_!~ or
2AG j t

2GQ=.s.i~ (25)
A j t
Hence, To • 2q,A, + 2q,A, - --- - - - - - - (23)
ThUS, for each cell of a ~ultiple cell struc-
tf
where A1 = area ot cell (1) and A3 = area of
cell (2). Therefore, the moment of the internal ture an expression ~s can be written and
shear system of a ~ultlple cell tube carrying equated to the constant value 2OQ. Let a l Q•
pur~ torsional shear stresses 1s equal to the
sum of twice the ~nClosed area of each cell
times the shear load per inch Which exists in
represent a line integral f ~s for cell wall
1-0, and ala, a~o, a2~ and a~Q the line in-
t~e outside wall or tr~t cell. (Note: The web
mn 18 referred to as an inside wall of either
cell) •
tegrals f
~s for the other outside wall and
interior web portions of the 3-cell tube. Let
AS.10 Distribution of Torsional Shear Stresses 1n a clockwise direction of wall shear stresses in
Multiple-Cell Thin-Walled Closed Section. any cell be pOSitive in sign. Now. substituting
Angle of Twist. in Equation (25), we have:

Cell@ cell (1) A~ [q~ au + (qJ. - qa) au] =209 (26)

j t a

=2GQ - - - - - - - - - - - - - - - - - - (27)

Fig. AG.17

Equations (24. 26, 27 and 28) are sufficient to

Fig. A6.17 shows in general the internal
shear flow pattern on a 3-cell tube produced by
a pure torque load on the tube. ~he cells are
~umbered (1), (2) and (3), and the area outside
the tube 1s designated as cell (0). Thus. to
designate the outside wall ot cell (1), we re-
ter to it as lying between cells 1-0; for the
outside wall of cell 2, as 2-C; and for the web
between cells (1) and (2) as 1-2, etc.
q~ = shear load per inch = ~~tl in the out-
side wall of cell (I), where ~l equals the unit
stress and t • • wall thickness. Likewise,
qa • ~ata and q3 • ~3t3 = shear load per inch in
outside walls ot cells (2) and (3) respectively.
For equilibrium of shear forces ~t the junction
points of interior webs with the outside walls.
we have (q. - q~) equal to t~e shear load per
determine the true values of ql, qa, q~ and Q.
ThUS, to 1ete~ine ~rre torsional stress
distribution in a multiple cell structure, we
write equation (25) fer each cell ant these
equations together with the general torque
equat:on, s1milar to eq~tion (24), proViQes
sufficient conditions for the solution of the
shear stresses and the angle of twist.
A6.11 stress Distribution and Angle of Twist for 2-Cell
Thin- Wall Closed Section.
For a two cell tube, the equations can be
Simplified to give the values of ql' qa and 0
directly. For tubes with ~ore than two cells,
the equations become too complicated, and thus
the equations should be solved S1~ultaneously.
Equations for ~No-cell tube (Fig. A6.18). -
AS.S TORSION

Cell 2
----(29)
_ 1 [
2GQ-367.4 -

.865 q1. - 5.34 qa - - - - - - (34)

Fig. AS.18
Equating (33) and (34)
- 14.195 q1. + 8.505 qa = 0 - - - - - - - - (35)

- - - - (30) The summation o~ the external and internal re-

Sisting torque must equal zero.

1 :: 25. 25"
.040"

External
g =GTJ - - - - - - - - - - - - - - - - - - (32) Torque
83450"*

where A = A1. Aa • Cell @

+
1::25.30"
II
.050"

A6.12 &ample Problems of. Torsional Stresses in ='

MulUple-Cell-Thin-Walled Tubes. , .032"
Fig. A6.19
Erample 1 - Torsional Stresses in Un-symmetrical Two-
Cell-Tube.

Fig. A6.19 shows a typical 2-cell tubular
section as formed by a conventional airfoil
shape, and having one interior web. An external
applied torque T of 83450 in. lb. is assumed
acting as shown. The internal shear res ist tng
pattern is reqUired.
Calculation at Cell Constants
Cell areas: - A, = 105.8 sq. In.
A. = 387.4 sq. In.
A = 493.2 sq. In.
Line integrals a = ~s : f -
=26.9
•025
= 1075; au
83450 - 2 X 105.8 q, - 2 x 387.4 q. =0 - (36)
Solving equations (35) and (36), q1. = 55.S#/in.
and q. = 9Z.St/in. Since results come out
?OSitiv6 1 the assumed direction of coun~er­
clockwise was correct for q~ and q~ or true
signs are q1. ~ - 55.6 and qa - - ~2.5.
q1.~ =- 55.6 + 92.5· = 36.9#/in. (as Viewed
from Cell 1).
Fig. A6.20 shows the resulting shear ,at-
tern. The angular twist of the complete cell
can be found by SUbstituting values of q1. and
~~ in either equations (33) or (34), since twist
ot each cell ~ust be the same and equal to
=~
.04 = 335 twist of tube as a whole .

aao = ---:o:;r
25.25 + ~ + 25.3 = -''''35
•05 •032 (
I
Solution by equating angular ~Nist of each cell.
General eQuat~on 2GQ ~ ~s. Clockwise tf
tlow at q is pOSitive.
Cell 1 Subt. in general equation

2Gg = 1
105.8

q1.
r- x 1075

qa
+

SOLUTION BY SUBSTITUTING IN EQUATIONS (29) & (30)

- 13.33 + 3.165 - - - - - - - - - - (33)
 ··~&iwr'

ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A6.9

•• 002456 T

1735 x 105.8 + 335 X 493.2 ] T

[ 1735 X 105.a u + 335 x 493.2~ + 1075 X 387.4 3 2

•• 000665 T •• 000665 x 83450 • 55.St/ln.

=4 r.La66.7
916.7 x 100a
x 333.3
+
+
333.3 x: zona + 866.7 x 100
333.3 x 916.7 + 916.7 x 866.7
a J
= 89.76

1075 X 387.4 + 335 x 493.2 ] T Q=JT

G
= T
G X 89.76
= .01116 ~ (rad.) per unit
[ 1735 X 105.8 3 + 335 X 493.2~ + 1075 X 387.4 3 "2 length of cell.

= .001107 T = .001107 X 83450 = 9Z.5t/ln. A6.13 Example 3 - three-Ceil-Tube.

Fig. A6.22 shows a thin-walled tubular
Example Problem 2. section composed of three cells. The internal
shear flow pattern will be determined in re-
Determine the torsional shear stresses in Sisting the external torque at lOo,OOO~t as
the symmetrical 2 cell section of Fig. A6.21 shown.
when subjected to a torque T. Neglect any re-
sistance of stringers in resisting torsional
moment.
~ Flg.A6.21 T
SOLUTION:
( .03" \ ,.03" q. Iq05To' o:l<XO'i
Calculation of .031 CeU@ ~
terms
Area of cer.fa: - .03 . 03

A l = 100 Ali 100 Fig. ,A6. 22

SOLUTION:
A • Al + A3 200
Line integrals a = Calculation at cell constants
fdtS:- Cell areas: -

Al. = 39.3 As ~ 100

a - 10 + ~ = 866.7
JL~t
1.0 - .05 .03
Line integrals a = •

au = .~~ = 333.3 10
a" = 2nxx .025 = 629 au = 10
.05
= 200
- 20
altO -:03 .04 = 916.7
+..l2... _ 20
=.l2...=
a a e - .03 = 667 a ••
.03
333

solution of Equations from Article A6.1I: -

a;so = ZO +.lQ.. = 917
.03 .04

Equat1n~ the external torque to the internal re-

sisttng torque: -
1 r. 916.7 x 100 + 333.3 x 200
q.l.=ZL916.7 X 100 2 + ~.3 x 200 2 + 866 . 7
subat i tut Ing r
= .00254D T
.
78.6 q.l. + 200 q~ + 200 q~ - 100,000 = 0 _ (37)

Writing the expresston for the angular twist

of each cell: -

-.1[ 866.7 x 100 + 333.3 x 200

;::; 916.7 x 100 2 + 333.3 X 200 5 + 866.7
Cell (1)
2G1d = All. [q;>al.O • (q, - q,) a,.]
A6.10 TORSION

SUbstituting: A6.14 Torsional Shear Flow in Multiple Cell Beams by

Method of • Successive Corrections.
trend in airplane wing struct~al de-
~he
1.3 sign particularly in high speed airplanes is
209 = 39 [629 q, + 200 q, - 200 q,J - - - (38)
toward ~ultiple cell arrangement as illustrated
in :1g. A6.24, n~.ely a Wing cross-section
Cell (2) made up of a relatively large number of cells.

c:=:ITO 0,---,---,---1,---
Fig. AB.24
Subat I t.ut tng r
With one unknaNn shear :low q tor each cell,
the solution by the previous eq~tions becomes
roe = l~O [200 qlil - 200 q , + 667 c, + 3:33 q , - Quite labor! ous ,
The method of successive approximations
provides a simple, rapid method :or fl~dlng the
- - - - - - - - - - - - - - - - - (39)
shear flow in multiple cells under pure torsion.

EXPLANATION OF SUCG~SSI'lE COFL~ECTION ~HCD.

Cell (3) Consider a ~HO cell tube as shown in Fig.

a. To begin with assume each cell as act1ng
independently, and
subject cell (1) to
such a shear flow QJ.
as to make GQJ. = 1. Cell Iceu~
SUbstituting: From equation (1) (2)
.04 .05 "'J )0"
(19) we can 'NTite, I ~s-
2i}Q = l~ [333 q;) - 333 q.a + 917 q~J - - - (40) Gg=..s..~dS
2A J t '-. . __-...l.'~_"/'U
A ..=89.3

Solving equations (37) to (40), we obtain, Now assume GQ = 1,

!- 5" -1
then _ Fig. a
q, = 143.4#/in.
q, = 234.1Vln.
q, = 208.e*/ln.
q, q, = 90.7*/ln.
Since practically all cellular aircraft
q. -
q, = 25:3*/ln. beams have waLl, and web panels of constant
thickness for each part~cular unit, the term
p d: for s tmpt tc t tv will be written Z ~, where
L equals the length of a wallar web panel and
t its thickness. Thus we can write,
Fig. AB. 23
:2 A
q = -----;- (41)
L
:E cel l t

234. H/in.
Fig. A6.23 shows the resulting internal
shear flow pattern. ~he angle of twist, if de-
sired, can be found by substituting values of
shear flows in any of the equations (38) to (40).
Therefore assuming GQ~ = 1 fer cell (1) of
Fig. a, we can write ~rom equation (41): _
• Based on Paper, "Numerical Transformation Procedures
for Shear Flow Calculations" by S. U. Benscoter.
Journal, Institute of Aeronautical Sciences, Aug. 1946.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
A6.1l

=-
2 X 89.3
25.7
--
.04
10,
.05 .-- 0.212 lb./in.
(1-2) is SUbjected to a shear flow qJ1l
lin. (counterclockwise with respect to cell (1)
and therefore negative), in addition to the
shear flow q1. = .212 of cell (1). The negative
.109#

Fig. b. shows the results. shear flow qa =-

.109 on web 1-2 decreases the
~Nist of cell (1) as calculated above with the

r-:---
resulting value tor G91. = 0.875 instead of 1.0
. I
,.-
212
I~
~
q '" . 109
as started with.
Thus in order to make 091. = 1 again, we
j (11 I I (2) will have to add a constant shear flow q'l to
cell (1) which will cancel the negative ~Nist
~ j due to q~ acting on web (1-2). Since we are
I considering only cell (1) we can compare cell
Fig. b Fig. c wall strains instead of cell twist since in
equation (41), the term 2A is constant.
In a similar manner assume cell (2) sub- Thus adding a constant shear flow qi to
jected to a shear flow qa to make GQ 3 = 1. Then cancel influences at qJ1l on web l-?, we can ve-tta:

l.Q.
.05
2 x 39.3
+ 15.7
.03
=
78.6
723 :. 0.lOS#/1n. hence
qi (z n cell (l) - q, (~) 'Neb 1-2 = 0

Fig. c shows the results.

Now assume the two cells are joined to-
(If) 'Neb 1-2 ] - - - - - - - - (42)
gether with the ~nterior web (1-2) as a common
part at both cells. -
[ (z i') cell (1)
See Fig. d. The in- q w • 109
ter i or; web 15 now SUbstituting values in equation (42)
Subjected to a re-
sultant shear flow
at q1. - qJ1l = (.212 -
.109) =.103t/ln. .- [ .~+....!2..J_200
q1. - q~ ~ - 842 q3 =
Obviously this change .04 .05
of shear flow on the Fig. d
interior web will cause
the cell twist to be different tor each cell in- Thus to make GQ1. equal to 1 we must correct
stead of the same when the cells were considered the shear flow in cell (1) by adding a constant
acting separately. To verify this conclusion shear flow equal to .237 tl~es the shear flow qJ1l
the twist measured by the term OQ will be com- in cell (2) which equals .237 x .109 = .0258~/in.
,uted for each cell. Since this shear :low is in te~ of the shear
flow qJ1l of the adjacent cell it will be referred
1 L _ to as a correction carrv over shear flow, and
Cen (1), GQ. ' -
2A, Z q-t - will consist of a carryover correction tactor
times qJ1l.
[212 x 25
-- 2 x 189.3·
7
. + (2'
.C4 ..... 2 - .109) .~;] =
Thus the carryover factor tram cell (2) to
cell (1) may be written as

= 0.375 Jf) 'Neb (1-2)

which equals
C.O.F. (2 to 11
Cen (2), -(z ~) cell (1)

.237 as found above in substitution in equation

• 2 x ~S.3 [.109 x -
15.7
- - (.212 - .109)
• 03 .~;J (42) •
Now consider cell (2) in Fig. d. In
• 0.4375
bringlr.g up ane attaching cell (1), the common
web (2-1) is subjected to a shear flow of Q1. =
- O.212*/ln. (co~~terc1~c~vise as viewed from
Since 91. must eq~l QJ1I if ~he cress-section cell 2 and ~~eretore negative). This additior~l
is not to c~stort from its original shape, it is shear ~low c~~nges GQ 3 twist of cell (2) to a
eVident ~hat the above shear :lows ~e not the relative "al~e of 0.4375 instead Jf 1.0 (see
true ones w~en the two cells act together as a preVious GGJ1I calculations). Therefore to ~ke
unit. OQ 3 equal to 1.0 again, a cor~ectlve constant
Now consider cell (1) in Fig. d. In br-tng-. shear flow q~ ~ust be added to ~ell (2) to
ing up and attaching cell (2) the common web

-----_. - --- -----------------------,-..,..,.,---,=

AS. 12 TOR S ION

cancel the twist effect of q~ = 0.212 on web Considering Cell (2)

(2-1). Therefore we can write,
q: = q~ (C.O.F.) 1 to 2 = .0256 x .227 =
q. (z ~) cell (2) - q, (~) web (2-1) = 0
.00717#/in.

Fig. f shows the result1ng second set ct

hence corrective constant shear flows for each cell.
Since our corrective
q, (~) web (2-1) shear flows are
q. = (43) rapidly getting
(z~) cell (2) sffialler, the con-
tinuation of the
{"', •. 0139
q'k '"' .00717 ;:;.-- - -

II \
I.

SUbstituting in equation (43): -

process dependS on \~ (il
the degree of ac-
(~' jI
,_ (200)
q. - q~ \723 =
.277 q, =
.277 x .212 .058711!1n.
curacy we wish for
=
the tinal results. Fig. f
---
Suppose we a dd one
Thus the carryover factor trom cell (1) to more set of cor-
cell (2) in terms at q~ to make GQ~ = 1 again can rective constant shear flows q~ and q:, Us tng
be written the carryover correction factors previOusly
found we obtain,

C.O.F. (1 to 2)
= (-¥-) web (2-1) = 200 = .277
qT .237 x .00717 = .0017~/in.

(z ~) cell (2)
723
q~ = .277 x .0139 = .00385#/in.

Fig. e shows the constant shear flow q~ and Fig. g shows the result s .
qt that were added to make GQ = 1 for each cell.
However these cor- The final or re- :;....-_ _ q'~ '" . 00385
rective shear flows q'~ e , 0587 sulting cell shear 0i,~=.OOi71j
were added assuming flows then equal the
the cells were again
independent ot each
original shear flows
plus all corrective
\ (I) t!(21)
other or did not cell shear flOWS, or ~
--- »
have the cammon web Fig. g
Fig. e
(1-2). ThUS in
bringing the cells
together again the Interior web Is SUbjected to
be resultant shear flow of q~ - q~. In other q = q. + q'. + q". + qm.
wordS it we were to add the shear flows of Fig. e • (r inat )
to those of Fig. d, we would not have GQ 1 and GQ~ Fig. h shows
equal to 1. The resulting values would be closer :;;..- 1 r . 1787
the final results.
to 1.0 than were r OUIld tor the shear flow system
or Fig. d.
To check the
final twist at each
I'q,= .2534 fll I
Considering Fig. e, we will now add a second
set ot correctIve shear flows q~ and to cells q: cell the value GQ
will be computed for
I~ (1) ll(~
(1) and (2) respectively to ~ke GQ, and GQ~ = 1 each cell using the ~--
tor cells acting independently. q values in Fig. h. Fig. h
Considering cell (1), and proceeding with
same reasoning as betore~ cen (1)

1 [~534 25.7
'-' ..... .- ~ x -04
x "9::; • + .0747 x .~~J '" .997
Hence

1-2 GQ. = 2 x 139.3 [ .1787 x 15.7 10]

.03 - .0747 x .C5 = .997
q~ =-,..,.,c'-""--- = .0587 x .237 = .013911!in.
AS.15 Use o! Operations Table to Organize Solution by
Successive Corrections.
or Operations Table 1 arranges the calculations
so that the steps jealing with the corrective
q~ = (C.OS.) 2 to 1 times q~ = .237 x .0587 shear flows can be carried out rapidly and with
= .0139 a minimum of thought.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A6.13
OPIRA T IOKS TABU 1

Cdl 1 Cell 2
(~) web 1-2
10
.05
C.O. (1-2) = = = .40
1
2
C.O. FactoX'
AssUIIled
.277
a a
237 z (t) cell 2 ..lQ. + ...2Q.
.05 0.1
~ ~~
J CO 02~8

•
m
C.O. 3
. 00
• web 2-1 10
T • 2A <15. zo·" 14 1".
C.O. (2-1) = = .~5 = .25
(~l
0," T .3""
Z cell 1 :05
.Explanation of Table 1
Line 2 gives the shear flow q in each cell,
Line 1 gives the carryover factors for when it Is assumed each cell is acting separ-
each cell, computed as explained before. Line 2 ately and is subjected to a unit rate at twist
gives the necessary constant shear flow q in or GQ - 1.
each cell to give unit rate of twist to each The calculations for the q values are as
cell acting independently. (GQ = I). Line 3 follows: -
gives the first set of constant corrective shear
flows to add to each cell. The corrective q re- For cell (1 )
ferred to as the carryover q or C.O.q in the
table consists of the q in the adjacent cell q. • 2A. . 2 x 100
• •25
times the C.O. factor of that cell. Z (~) cell 1
40
.05
Thus .237 x .109 = .0258 is carryover from
cell (2) to cell (1) and .277 x .212 = .0587 is
For cell (2 )
carryover from cell (1) to cell (2).
Line 4 gives the second set of corrective
carryover shear flows, namely .277 x .0258 =
.00717 to cell (2) and .237 x .0587 - .0139 to
q. •
Z (~)
2A.
cell 2
- 2 x 100 • • 4
10 + ~
.05 0.1
cell (1). Line 5 repeats the corrective carry
over process once more. Line 6 gives the final A. and A. equal cell area of cells (1) and (2)
q values which equal the original q plus all respectively. L - length of wall element and t
carryover q values. its corresponding thickness.
Example Problem 1 (2 cells) In order not to start out with dec1lllal values
Determine the ~nternal shear flaN system of q, the values above will arbitrarily be
for the ~NO cell tube in Fig. A6.25 when sub- multiplied by 100 to make q. = 25#/in. and
jected to a torque of 20,000 in. lbs. q2 :; 4Oi/in. Since we want only relative
I-- 10" -----.J.-- 10" -----I values of terms this is permiSSible. These
values are shown in line 2 of Table (2). The

r
10" .05
.05

Cell
1 .05
0.10

Cell
2 0.10
corrective carryover process proper is started
in 11ne 3 of table (2). In cell (1) the amount
carried over of q. = 25 to cell (2) equals the
C.O. factor times 25 or 0.4 x 25 = 10 Which is
written along the vertical line under cell (2).
1 .05

Fig. A6.25
0.10
Likewise in cell (2) the amount of q. - 40 that
Is carried over to cell (1) equals the C.O.
factor t !mes q. or .25 x 40 = 10, ·.mich is
written along the vertical line under cell (1).
OPERlTIOMS TA8U 2
The second set of corrective carryover
constant shear flows are given in line 4 ot
Cell 1 Cell 2 Table (2), thus, .4 of the q. :: 10 :; 4.0 is
1
,
CO. l"ac'toJ:"
l!IU11ed
0.'
0
O.2~
.0 0
carried over to cell (2) and .25 x value of
q. = 10 = 2.5 is carried over to cell (1). Line
J C.O 10 10
• C.O•
•0.
2. S
• 5 repeats the process namely 0.4 x 2.5 = I to
a cell (2) and .25 x 4 = 1.0 to cell (1).
7 . . 0 10
, This process of carrying over 'lalues of q
>n.
2A
•

7770 11100 is continued until the values are small or

0 T 0," 18870
11 for ~20 000 41 a 9
negligible. Line 8 gives the final q values in
each cell as the summation of the assumed q
Explanation of solution as given in ~able (2): value plus all carryover values of q. Thus in
cell (1) q • 38.85#/in. and for cell (2)
Line 1 gives the values of the carryover q = 55.5. Line 9 gives the ~orque that these
factors. The values are calculated as follows: values of cell shear flow can produce.
C.O. factor cell (1) to cell (2): T = 2Aq
AG.14 TORSION

For cell (1) T = 2 x 100 x 38.85 = 7770 in. lb. OPERATIONS TABLE 3
For cell (2) T = 2 x 100 x 55.5 = 11100 in. lb. Cell 1 Cell 2 Call 3 I
,
Line 10 gives the sum at th9 above two val~as 1 co Factor 23 .21 .'88 .288
a Assumed 25.30 18.45
which equals 18870 in. lb. 3 CO 3 88 2 3. o.
The original requirement of ~he problem was •5 0
C 0 °
1 "
505
° 394 1 51
0.430 0.13
2.53
0.685
the shear flow system ~or a torque of 20,000"#. , 6 C.O. 0.244 0.116 O. .33
co.
, ,.
Therefore the required q values follow by direct 7
, 0.072 0.056
.0
proportion, whence 9200 6020 2215
10
11
"
T Total ~ 19000 + 6020 + 2215 ~ 1 43S'"1lI
q _ to;,: T=100,000 183 '"0 112.5
20000
=I88'iO x 38.85 = 41.2#/ln.
20000 Explanation of sol~tion as given in Table 3: -
q~ = I8870 x 55.5 = 58.9i/ln.
Cell (1) Cell (2 ) Cell (3 )
These values are shown in line 11 of Table
2. Check on twist of cells under fir~l q values.
A~ .: 144 = 96
A. A. = 56. 5
The relative total strain around each cell
1
"i = 1140
L
"t= 1040
L
Zt::; 1055
boundary is given by the term. A z q tL for the
Shear flow q for GQ • 1 for each cell acting in
cell.
Thus tor cell (1) dependently: -
Cell (1) q = 2A, = 288 =
"~ 1140 .253
t

Cell (2) q .: 2A.~ = 192 = .1845

= 212 Z L 1040
t
For call (2),
Cell (3) q = z2Ai' = 113 =
1:.. 1055
L z q~ = 1;0 [(58.9 - 41.2) .~~ ,. 58.9 x ~03.01)J t
.107

To avoid small numbers t~ese values of q

=212 are ~ultiplied ~y 100 and er.tered on line 2 of
the table. Calculation of carry factors as
Thus both cells have the same twist. given in line 1 of Table 3.
In the above calculations q~ and q~ act Cell (1) to (2)
clockwise in each cell, hence the shear flow on
the !nter1or common Neb 15 the difference of ~~e
two q values. _(f) web 1-2 - 240 - .23
C.O. (1-2) - 1040 -
Example Problem 2. Three cells
The three cell structure in Fig. A6.26 is
"( ~) cell 2

Subjected to an external torque of - 100,000 in. Cell (2) to (1 )

lb. Determine the internal reSisting shear flow
pattern.

, C.O. =
m web 2-1
;:rw
240
= .21
I - 12"
I
I
8" ---I,
(2-1)
!(~)cell 1
T .04 .032

i I Cell (2 ) to (3 )
.041 \
12" .04 Cell 1 .0' Cell 2 'Cell 3 I (~) web 2-3

!y
_ 300
C. O.
(2-3) = - 1055 = .234
z(~)ce113
I

i
...i... i .04 0032
I

Fig. A6.26
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A6.15

cell (3) to (2) as given in TablA 4. Line 10 shows the cor-

rection of q values to develop a reSisting
:5
_(i-) web 2-3 torque of 100,000n#. The ~ultlplylng factor
100,000/80630.
c .0. (3-2)
- Z (~) cell 2 A6.16 Torsion of Thin-Walled Cylinder HaVing Closed
Type Stiffeners.
The balance at the solution or procedure in The airplane thin-walled structure usually
Table 3 is the same as explained for Problem 1 corrta tns ron-t tuc inat stiffeners spaced around
in Table (2). It should be noticed that cell the outer walls as illustrated in Fi~:. A6.28
(2) being between two cells receives carryover and A6.29.
q ~lues from both adjacent cells and these ~NO
values are added together before being distribu-
ted or carried over again to adjacent cells.
?or example consider line 3 in Table (3) and
cell (2). The q value 5.82 representing .23 x
25.3 is brought over fr~ cell (l) and the q
value 3.CB • • 288 x 10.70 from cell (3). These
two values are added together or 5.82 + 3.08 -
2
Open Type StHfener
2
Closed Type Stiffener

8.90. Pne car-r-y over q to cell (1) is then .21
x 8.90 = 1.87 and to cell (3) is .224 x 8.90 =
2.53.
In line 8, the final q in cell (2) equals
the original q of 18.45 plus all carryover q
values from each adjacent cell.
Line 10 in Table (3) shows the total torque
developed by t~e resultant internal shear flow
is 17435"~. Since the problem was to find the
shear flow system for a torque of - 100,000 in.
lb., the values of q in line 8 must ~e multipli-
ed by the factor 100,000/17435.
Line ~l shows the final q values.
Example Problem 3. Four cells.
Determine the ir.ternal shear flow system
for the four cell structure in Fig. A6.27 when
SUbjected to a torsional moment of - 100,000 tn.
lb.
(1600)
16"
(1600)
(500) (625)
cen i Cell 2 Cell 3
Fig. A6.27
For the open type stiffener as illustrated in
Fig. A6.28, the torsional rigidity of the in-
dividual stiffeners as compared t 0 the torsional
rigidity of the thin-wal!ed cell 1s so small
to be negligible. However a closed type
stiffener is essentially a small sized tube and
its stiffness 15 much greater than an open
section of similar size. Thus a cell with
closed type stiffeners attached to its outer
walls could be handled as a multiple cell struc-
ture, With each st_tfener acting as a cell with
a common wall With the outside surrounding cell.
Since in general the st11'fness provided by the
stiffeners 1s comparatively small compared to
the over-all cell, the approximate Simplified
procedure as given in NAeA T.N. 542 by Kuhn can
be used to usually give sufficient accuracy. In
this appraximate method, the thin-walled tube
and closed stiffeners are converted or trans-
40 formed into a Single thin-walled tube by modi-
fying the closed stiffeners by either one of the
follOWing procedures: -
.025
(1) Replace each closed stiffener by a doubler
14001 plate having an effective thickness
te=tSTs!d, and calculate ¢ dslt with these
doubler plates in place. The enclosed area
Cell 4 of the torsion tube still re~lns (A) or
the same. See Fig. A6.30.

OPDATIO!(S TABLE .oj, d

Cell 1 Cel]. 2 cell 3 Cell .oj,

1 CO, 17' 20. .153 181 .078 .16",

2 22.4 38.8 28.4 20.5 Original
3
•s
0
0
1 e
1 69
3 88 , 1<
2 00 188
s •
3 as
1. 38 0.36
2 22
O. 3
~iiiener
0 1 10 0.47 022 0.57 0 2 J
o '0 o O. o 02 o rs Transformation

,-
, 12
, Transiormation Procedure (2)
a
9
10
", .
7620

47.1
31560

65.3
, 22550

49.9
8900

29.41
by Procedure (1)
I -,- d ----oot ,.tsK
by

I •

In Fig. A6.27 the values in the rectangles lte:::tsK+tst ~

represent the cell areas. The values in ( )
Fig. AG. 30
represent ~he Lit values for the particular wall Fig. AG.31 t:::tst ... t s K ~
or web. After studying example problems I and 2
one should r2ve no trouble checking the values
A6.16 TORSION

(2) Replace the skin over each stiffener by a

"liner" 1n the stiffener having a thickness
Lateral shear, H , - -
t e = tSK dis' (See F'ig. A6.31.) The en- mainly ; :--
closed area (A) of the cell now equals the
original area less that area cross-latched
in Fig. A6.31. Partiy lateral shear
and partl y torsional shear/:-"""",V
Procedure (1) slightly overestimates and fiW
procedure (2) slightly underestimates the stiff-
ness effect of the stiffeners.
The corner members of a stiffened cell are
usually open or solid sections and thus their
torsional resistance can be simply added to the
torsional sti:fness of the thin-walled over-all
cell.
A6.1'l Effect of End RestraJ.nt on Members Carrying
Torsion. Fig. A6.32
The equations derived in the previous part
at this Chapter assumed that cross-sections
throughout the length of the torsion crembers
were free to warp out ot their plane and thus
there could be no stresses normal to the cross-
sections. In actual practical structures re- AS.IS Example Problem illustrating Effect ot End ..
straint against thiS free warping at sections Restraint on a Member in Torsion.
is however Often present. For example, the Fig. A6.33 shows ~ I-bea~ s~bjected to a
airplane cantilever wing from its attachment to torSional moment T at its free e~d. The prcb-
a rather rigid fuselage structure is restrained lem will be to deterT-ine ~rEt proportien of the
against '~rping at the wing-fuselage attachment torque T is taken by ~he flanges in bending and
pOint. Another example of restraint is a heavy what proportion by pure shear, at two different
wing balkhead such as those carr-ying a landing sections, namely 10 inches ~nd 40 incheS trom
gear or power plant reaction. The flanges of the fixed end of the I-beam.
these heavy balkheads Often possess considerable .,.u"",""-,+- Fixed
lateral bending stitfness, hence they tend to End
prevent warping ot the wing cross-section.
Since only torsional forces are being considered r--L 75''-1
here as being applied to the member, the stress- F
es prOduced normal to the cross-section of the
member namely, tension and compression must add '1 3 T'
up zero tor eqUilibrtum. Thus the applied tor-
que is carried by pure torsion action at the
member and part by the Iongt tuo i ona.i stresses
Fhi .4";#-
normal to the member cross-sections. The per- ~~
centage ot the total torque carried by each Fig. A6. 33 FIg. A6. 34 FIg. A6. 35
action dependS on the dimensions and shape at
the cross-section and the length ot the member. Fig. A6.34 s~ows the torque diViding into VNO
parts, namely the couple force F-F ~armed by
Fig. A6.32 illustrates the distortion of an
open section, nanely, a channel section subject- bending of the flanges of ~he I-bea~ and the
pure shearln~ stress system on :he cross-
ed to a pure torsional force T at its :ree end
and fixed at the other or supporting end. Near
section. :<'ig. A6.35 shows the twisting 0:' tne
the fixed end the applied torque is practically Section through a distance 5.
all resisted by the lateral bending at the top The solution will consist in computing the
and bottom legs of the channel acting as short angle of twist 9 under the two stress con-
cantilever beams, thus torming the couple With n ditions and equat1nz them.
forces as illustrated in the Figure. Near the Let Ta be the proportion of the total
!ree end of the ~emberJ these top and bottom torque T carried by the fl2nges in bending
legs are now very long cantilever beams and thus forming the couple F-F in Fig. A6.34. From
their bending rigidity 1s small and thus the Fig. A6.35, the angle of tNist can be written
pure torsional rigidity of the Section in this
region is greater than the bending rigidity at
the channel legs. 2 Tg L"
3 na-IT

~ .. - ,.
~- . ~. , ... ~ ~. .
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A6.17

Note: The deflection of a cantilever ~e~ with 7he shaft rotates at constant Soeec. The dif-
a load F at its end equals FL~/3EI, and I the ference in be!t pull on two sides of a p~lley
moment of inertia of a rectangle about Its c~n­ are shown. an tne figure. Calculate the ::2...'(1~~
ter axis tb~/12. torsional shearing stress in the 1-3/4 inch
shaft bet~een pulleys (1) and (2) and between
hence (2) and (3).
(2) rl. 1/2 HP. motor operating at lOCO RR~
Now let Tt be the ?ortlon ot the tctal torque rotateS a 3/4-.035 al~~lnum alloy torque tuee
carried by the ~e~ber in pure torsion. T~e ap- 30 inches long which drives t~e gear =ectan~~
proximate sol~tion for open sections co~posed of tor ~p&ra~ln; a Wing flap. Dete~ine the ~~<:­
rectang~lar ele~ents as given in Art. A6.6, ~~ torslcnal st~ess in the torque sha:t ~cer
equation (12) will be used. fUll ~~wer and RPM. Find ~he ang~lar de:lec~ian
of shaft in the 30 lr.ch ~ength. ?olar =c~ar.~ ot
= 3 Tr L inertia of tube = .01 in. ~odulus ot r:gidity
9t Gt'(b + b w) G = 3300,000 psi.
Tni.n.lb.
Equating GS to tit we can ''iT' i te

3 E h
2
b3 I I

8L 2Ge(2b
+ b.n ) / I
/
,/ I
Substituting values when L : 10 inches. 'f--- I- "
T
I I
3 x 10.5 X 10 4 X 3.4 x 1.75 I /
/
io- X 3.7 X 10 4 X 0.1 11(2 X 1.75 + 3.3}

= 9.65
Fig. AS, 36
'rner-ercre , percent of total t cr-s tona L T -15" :-
taken :y :e:'.d.ir.g of flanges e~~als,
'!'c ':'8
.... '1' (100) =,,~:-+"='...-(lCC): 90.5 percent.
Ide t.ertafne t:r.etne crs ona ;
(3) I ." . .
~,

t
__ ..1."" r::
__ " "," e ..,..'!!: • •"\._
"01'1""'''' """' '{:: • .;.0
snea r :lo'iJ :::. :-es:~'::::g
t
~ -ee
.J_ • C

.T
3 . -t -& I tne ext er-na L t orque 0: scccc :n.lb. -,.;,:2: ::.:.:.
wall t'u cxnees ar-e gi '''e:: or: tno r:;,;u.:-e. As "'\.::",e
the tube 1s 100 In. long and :in~ :~e :c::::~3.1
It we ccnsid.er the section 4Q inches :~c~ de f t i cn , za cer 15 <:ll:"''"ll.L'1.u::l.
L e c (J-=
i a t a.I L o y ,

the fixed end, then L = 40 inches. Thus if 40· 32CO , COO ps 1. )

is orac ec in the above subs t I t at Lcn inst eac of (4) In fig. A6.36 :"2r.:lC':,9 :~e Irrt er-i cr- .035
10 11 • the r~s~l:s for TS/T t WQuld be 0.602 and web and cv~pute torsional shear flow and ~e­
the percent or the torGue carried by ~he rl~~ges rt.ect ions .
in bending wOuld be 36 as com~ared to 30.5 per-
cent at L = 10 inc~es fro~ support. Thus in T '" 100,000 in. lb.
general the e::ect of the end restraint decreaSes
rapidly with increasing value of L. f\
The ef!9c~ af end restraint on thin walled
tUbes with lc~glt~dlcnal stiffeners 1s a more
Invc Lved pr ob Lera and cannot be nancj.ec in sucn
a si~p1::ied =a~ner. This ?roble~ 15 ccnsi~erec I
in other Chapters.
A6.18 Problems.

Fi.g. AS. 37

(5) In the 3-cell structure of fig. A6.37

320*
Fig. A6.35 detieratne the internal res t st tng shear ilow due
to exter~.al torque of 100,000 in. lb. ?or a
(1} In Fig. A6.36 ?~lley (1) is ~~e 1rivl~g length of 100 inches calculate twist of cellular
PUlley and (2) and (3) are the driven pulleys, structure i: G is assumAd 3,800,000 psi.
A6.18 TORSION

(6) Remove the .05 interior web of Fig. ~14' 14"---i

A6.37 and calculate shear flow and twist. .03
I I
.0' 05 .0' .0>" 10"
(01
(7) Remove both interior webs of F:g. 1.
.03 .03
i I
A6.37 and calculate shear flow and twist. -L
fo-6" --t- 8" 10" 10" ......l
(8) Each of ~he cellular structures in ,
Fig. A6.38 is subjected to a torsional noment T .03 .03
!
.03 .03

of 120,000 In.lbs. USing the ~ethod of 10" .04 04 .035 ~035 .03 .0' .03 I (b]
successive approximation calculate reststing
shear flow pattern.
1 .03 .03 I .03 .03

I- 12" '-1-10"-+-8" ........... 8" +8" -+-8"""';

1-==::::::::===
All interior wece e . 05 thickness
I i i I'"
Top skin .084" thickness
Bottom skin .064" thickness
Fig. A6.38

'. ,
,.

The big helicopters of the future will be used in ~ny important industrial and military opera-
t t ons , The helicopter presents many challenging problems for the structures engineer.
(Sketches from United Aircraft Corp. Publication "BEE~HIVE", Jan. 1958. Sikorsky Helicopters)
 •

CHAPTER Ai
DEFLECTIONS OF STRUCTURES
ALFRED F. SCHMITT'

A7. 1 Introduction.
Calculations of structural deflections are p
i~portant for tNO reasons: - ,
(1) A knowledge of the load-defo~ation Work =- L: Pi
I~l
6 Oi
characteristics of the air?lane 1s of prllnary in the limit as
ireportance in st~dles of the influence of
structural flexibility upon airplane perform- 60-0
ance.
Work =Lt'J P dO
(2) Calculations of deflections are neces-
sary in solving for tt.e In~ernal load distri-
butions of complex redundant structures.
Fig. A7.1
The elastic deflection of a structure
under load i s the cumuLat;'... ''j result of the curve ls a straight line whose equation 1s
strain deformation of the individual ~lements P = k6 and the strain ener~J is readily com-
composing the structure. AS SUCh, one method puted to be
of solution for the t~tal deflection ~ight in- p'
,o~ve a vectorial addition of these individual U equally, U = 2k'
:ontriJutions. ~he involved geometry of most
?ract~cal str~ctures ~kes such an approach
A7.3 Strain Energy EXl)ressions for Various Loadings.
prohiJitively ditf~c~lt.
For complex structures the more popular STRAIN SNERGY OF !ENS I ON
tec~~lques ~re analytical rather than vectorial.
7hey::ieal dt r sc t Ly 'Nith quarrt t t Les which are not A tens 11e load S acting at ~he end of a
themselves ceflections but from which deflec- aniform bar of length L, cross sectional area
tions ~y be obtained by SUitable operations. A and elastic ~odulus E causes a deformation
~he nethods employed herein tor deflection cal- a = SL/AE• Hence S AE a and
L
culations are analytical in nature.
A7.2 Work and Strain Energy. U • ~ Sd5 • AE 5' (1)
'Q L 2
~ork as defined in ~echanlcs 1s the prod-
uct of rcr-ce t ines cr stence . It the rorce Alternately.
v~aries over the distance then the work 1s com-
puted by the integral calculus. Thus the work S'L
U' (2 )
cone by a varying force P in deforming a body 2AE - - -
an aaount 6 is
',.Jork =.rPd.a and is represented
by the area ander the load deformation (P-6)
Squations (1) and (2) are equivalent ex-
pressions ~Jr the strain energy in a Qni~o~ bar,
curve as shown in Pig. A7.1. the farner expressl~g U in terms of the de-
If the tefor.ned body 1s per~ectly elastic fo~~tior. and the latter expressing U in~e~
the energy stored i~ the body nay oe co~pletely of load. 7he secand form of expression is ~ore
r-ec over-ed , the bocy un Loac tng a Lcng the same 80nvenlent for ~eneral usage and succeeding
?-6 c'Jrve :ollowee for increasing load. T~is strain energy :ornulas will be ~ut in this form.
energy is ca l Ied the elastic ~ energy q£ ?cr a tensile bar haVing non-uni:orm prop-
GeforT.aticn (hersa~ter ~he strain energy, for erties (varying AE). or for which the axial load
brevi~y) and is denoted by ~~e s~bol U. Thus S varies, the strain energy is computed by the
U = t Pdc. Should the cody '_be .Li near Iy
calculus. Thus the energy in a dir:erentlal
element of length dx is atven by eq , (2) as
elastic (as are ~ost bodles jealt w1~h in
structural analYSiS) t~en the load-defo~tion s-ex
dU :: ;;'2AE Where S and AE

(1) Design gpec iarist, Convair-Astronautics are average ';alues over the ~ength of element.
A7.1
A7.2 DEFLeCTIONS OF STHUCTUReS

The total energy in the bar is obtai~ed oy

summ:ng over tr.e length of the ba,.
U
2
~ '" JL (
n"'AE
1- , 2L
S.Ln nx)' ex
. =
~.adx c
U =j dU =j "2AE (3)
01'5" 1....
'0"
'-.:; I ) (22'7\(4-0) '" 0.0920 in Ics ,
9.87(2)(10 x 2.0 5 ) ' !
Example ProbleT. 1
A linea~ly tapered al~nin~ll Jar is under STEArN SN~GY C? FLZ:('l)ES
an axial load of 15Ca Ibs , '1S snovr. .n Fig ..:;'7.2.
Find U. A ~~ len;th L ~~er the actien
~ifo~ bea~
of a pure couple ~~jer~oes an angular ro:atlcn
A, proportional :0 the couple. Thus, from elemen-
tary strength of Taterials
TU>AIXI =A,II _A)
(E-a~
40" x 80
L A.,= 2 in"

t 1500. Fig. A7.2 ~:=..1 f

Solution: where the constant S1, the product of elastic
modulus by sect~on ~oment of inertia. is called
Fram statics. the internal load S -15CO# the ~bendlng stif:ness R •
at every section. Since the rotation 0 builds up linearly
with M the strain energy stored is
S'dx
2AE =
( h::tt:/40
U = 2:1 MO
intl'odl.leed or
= 21-1500)'],' d~ = 4 5 In 2 I h 1b
tal' conven_
M'L
10' ,
2-1:' ncr s• ience)
U =
1
2 Ef 14,)

Note that although P was a negative (com-

preSSive) load the strain energy remained Fo~ a jeam haVing ncn-~~i~orn prcpert~es
positive. (varying El) and/or for which M varies along
the beam, the strain energy 1s com;uted by the
Example Problem 2 calculus. From sq. (4) the str~in energy in a
Find U in a uniform bar under a ~inlng beam element ot 1ifferentlal length is
load
_ nx(
q - q , cos 2I Fig • .40.7.3).

Hence summing over the complete beam to get the

~WT
total strain energy one has
1. • 40·
2
.to • 2 in U = j dU = ~ jl1'dx

q(~ m1
40 (5 )
J: _ lO::do 6 pd 2 EI
" 'to. 2' lb/in
Exa~ple Problem 3

I. Flg~ A7.3
For~
the beam of Fig. A7.4 derive the strain
energy expression as a function of Mo.

r;.M c. GEl = constant

Solution:
1 !
The equation for S (x) was found by statics. I-- L----.J
Fig. A7.4

Solution:

= q , 2L
n sin 2L mI' The bending ~oment diagram (Fig. A7-4a) was
found and an analytiC expression 'Nrltten ~or v

Substitution into eq. (3) gave

•
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A7.3

L
~=GJT -------------- (6)

where the constant GJ, the product of the elastic

modulus in shear by the cross section polar mom-
x ent of inertia, 15 called the "torsional stiff-
ness" •
Fig. A7.4a
Since the CNist ~ builds up linearly with
T the strain ener~J stored 1s

Then
or
(7 )

For a shaft of non-uniform properties and

varyin~ loading one has
Example Problem 4
Determine the strain energy of flexure of
the beam of Fig. A7.5. Neglect shear strain
energy.
U = HT~7' -- (8)

In passing it is worth remarking that one

often encounters the group symbol "GJ" in use
GEl =Constant for the torsional stiffness of a non-circular
shaft or beam such as an aircraft-wing. In
z: such a case the torsional stiftness has not been
I computed literally as G x J, but rather as-de-
i-~ - .....- L
2"
--01 tined by eq, (5), viz. the ratio ot torque to
Fig. A7.5
rate of twist.
Solution: Example Problem 5
For a certain flight condition the torque
The bending moment diagram was drawn first on an airplane wing due to aerodynamic loading
(Fig. A7.5a) and analytic expressions were is given as shown graphically in Fig. A7.6. The
written :cr 11. variation of torsional sti:tness GJ 1s given in
like manner. Find the strain energy stored.

1. 01"":~- __

Fig. A7.5a

Fig. A7.5
Inspection of the diagram revealed that o !_ _ ~ __ ~_~ __ ~_~_

the ener 5J of flexure in the right half of the y=L

Y=O
beam ~ust be identical with that of the left Root Tip
half. Hence
Solution:
U =2 •
A numerical integration of eq. (8) 'NaS
performed using Slmpson's rule. The work is
shOwn in tabular ferm in table A7.1.
Values of T and GJ ror selected wing
stations were taken tram the graphs provided
STRAIN =G1 OF TORSION and were entered in the table. For convenience
in ha~d11ng the numerical work all the variables
A uniform circular sr~ft car~ying a torque were treated in ~on-di~ensional :OrTI, eq. (8)
T experiences a total CNist in a length L pro- being changed as follows
portional to the torque. Thus from ele~entary
strength ef ~terials
U -- ~
I
.:::
i~'
0
-
GJ
-
dy - -1 LTRaJ'lila
--
2 GJR 0 GJ
c-r

~:-~-:.i.?1:
.
;,::~ ~.~<~--~
---~'. ':";\':,,,.
 ,

A7 • DEFLECTIONS OF STRUCTURES

where strain ~~er~J ~~ ~ beam under the act~o~ of ~

T - TlTn , 7=y/L transverse shear loading V(l~s)·. lor this
purpose V = ~ x dy x t and t x dy is called A,
GJ = GJ/GJR the ~eaw. cross sectional area. nance

The subscript R denotes Wrcot~ val~e (y=o).

The coefficients for 31~pson's ~Jle appearing 1~ dU = V'dx
ZAG J and
colunn (6) were taken from tte express~on for
odd n t~erefore t~e elastic strain energy of shear fer
I =;'r (r 0 + 4f J,. + 2f 2 + 4f 3 + - - - -
the enttre bea~ is ~lven by

V'dx
_ - - _ + 15 f
4- n-z
+ 3f
n-L
+ ~f )
4- n
U=
J-2AG - - - - - - - - - - - - - - - - (10,

Sxa:nple Problem 6
Determine the strain energy of shear in
TABLE A7.1 the Jeam of Example Problem ~.
(T)'
STA.
7
T (T)~ GJ
ill
Coe£!. m~ Coeff. Sc Iut t on :

a 1.0 1.0 1.0 1.0 . 0667 .0667 The shear load diagram was drawn first .
.98 .98 .78 1. 26 .267 .336 L L
•• .9' .88 .54 1. 63 .133 · 216 ~ 2 1 2 =-jP o
.8
.8
.07
• 2• ."
. 084
. 37
.24
1. 22
.35
. 250
' . 200
· 305
· 070 Po Ee" . . . . . . AG • CONSTANT
1.0 a a .14 a . 0833 0.0 V (xli I! iii I i_!J~L! i ! ! Ii i ii iii j

z= · 994

Then
Therefore the strain energy was ,L/2
2 x J 0 Poa dx
u = LTR'
_.- x .994
GJR ::: Po aL
2AG
STRA IN ENERGY OF SHEAR
Eq , (9) may be used to comput.e the shear
A rectan~~lar element "dx" by "dy" of
thickness "t" into the page under the action of strain ener 5J ~n a thin sheet. The elemen~
uniform shear stress ~ (psi) is shown in Fig. dx x dy Is visualized as but one of many in the
A7.6a. sheet and the total ener~J is oJtained by
summing. Thus
From eIementar-y strength
ot materials the angle of
shear strain J is pro- U =~ SS7't gx dy - - - - - - - - - - - - (11)
portional to the &hear
stress -r as Here a double integration 1s required, t~e
summation being carried out in both directions
~ = ~ where G is the over the sheet. In dealing with ~hin sheet the
material elastic modulus use or the shear flow q Tt 1s often convenient
in shear. For the dis- Fig. A7.6a so that eq , (11) rewrites
placements as shown in
the sketch only the down U = 2~ Sf q'dxGt dy (l1a)
load on the lett hand side does any work. (In
general all four sides ~ove, but if the mot~on A very important special case occurs when a
is referred to axes lying along ~NO adjacent homogenous sheet or constant thickness 1s
sides of the element, as was done here, the de- analyzed assuming q 1s constant everywhere. In
rivation is Simplified). This load is equal to this case one has
't' x dy x t and moves an amount a' x ex, Then _ q II _ qas
U - 2Gt JJ dxdy - 2Gt - - - - - - - - - - - (12)
dU =~ 7Et dx dy
where S = J5 dxdy is the surface area of the
1 7' sheet.
=2 G tdxdy - (9 )

• The sketch is visualized as a side view of an element of

Eq. (9) may be used to compute the shear length dx taken from a beam of total height dy.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
Example Problem 7
Find t~e shear strain energy stored in a
cant~lever jox bea~ of uniform rectangular
cross section under the action of torque T.
(Fig. A7.8).
j-~7
T" f.-b--l
,
Fig. A7.8
Solution:
The shear flow was assumed given by
Sredt's formula (Ref. Chapter A-5)
q=.1.= T In words,
2A Zbe
constant around the ?erlmeter of any section.
Then
= q"S
U 1: 2Gt
T'L(b+c)
4Gb lie it
THE TOTAL ELASTIC STRAIN
~GY OF A STRUCTURE
The strain energy by its ~efinition is
always a positive quantity. I~ also is a
scalar quantity (one having TSgllltude but not
direction) and hence the total energy of a
composite structure, having a variety of ele-
ments under various loadings, is readily found
as a simple sum.
+ J'{"d:<
ZAG
+Ii q",j.xc.y
2Gt
- - - - - - - - - - - - (13)
The integral symbols and comacn use of "XiI
an an index of integration should not be taken
too 11terally. It is probably best to read
these terms as "the sum of so and so over the
str~c~ure" ra)her than "the integral of It, for
quite often the te~ are formed as simple
sums without resort to the 8clculus. ~he 8al-
culus is only ~sed as an aid in Some ~pp1ica­
tions.
It is seldom that all ~he terms o~ eq.
(13) need Je employed :n a calculation. ~any
of the loacings, i~ actually ~resent, ~y be of
a localized or o~ a secondary nat~e and their
energy contribution ~ay be neglected.
A7.5
A7.4 The Theorems of Virtual WorK and Minimum
Potential Energy.
An important relationsr.ip between load and
deformation stems di~ectly from the definitions
of work and strain energy. Consider Fig. A7.9(a).
p - ---------
6 U= PM
~u
Limit flo .. 0,
dU = Pda
"------...JL.j- a
Fig. A7.9a
Thus
dU = P (14)
do
ItThe ~ 9S chang;e ~ ~ energy ~
respect to deflection is equal t9 the associated
load" .
---- Sq. (l4) and the above quotation are state_
ments of the Theorem of Virtual t.or-k. The
reader may find this theorem sta~quite dif-
ferently in the literature on rigid body me-
chanics but should be able to satisfy himSelf
that the expressions are nevertheless compatible.
A useful restat~ment of the above theorem
is obtained by rewriting eq. (14) as
dU - Pda =0
It is next argued that if the Change in dis-
placement dO 1s sufficiently small the load P
re~Alns senSibly constant and hence
dU - d(PO) =0
diU-po) =0 - - - - - - - (15)
The quantity U - PO is called the total
potent:'..al or' the system and eq, (15), r-esemb Ltng
as it 10es the ~th~~atical condition for the
~inimum value of a function, is said to be a
statement of the Theorem of Minimum Potential.
From the foregoing i~is clear that the
Theorem of Mini~um Potential 1s a ~estatemen~ Of
the 1':leorem or 'iirtual .vor-x .
In structural analysiS the ~ost Dnportant
~ses of these theorems are T~de in problems con-
cerning buckling instability ~d other non-
linear1ties. No a?plications will be ~de at
this poInt .
A7.5 The Theorem of Complementary Energy and
casngnaoo'e Theorem.
Again in t~e case of an elastic bOdy, ex-
aoinatior. of the arsa above the load-deformation
cu:ve shows that increments in this area (called
(',- ,.-

A7.8 DEFLECTIONS OF STRUCTURES

the complementary energy,U·), are related to the Force {Ibs .") Translation (inches)
load and deformation by (see Fig. A7.G(b). Moment (in. Ibs , ) Rotation (radians;
Torque (in. Ibs.) Rotation (r~dians)
Pressure (los/in"') vojuae (Ln ")
L' r Ou- Shear Flow (lbs/in) Area (in"')
TI;/'
III I/U:,
u- dU· ::6dP Any generalizations of t~e ~eanings of
. "P ,
,
' ~force~ and ~deflection~ are possible only so
, long as the units ~re such that their product
,, yields the units of strain errergy (in. IbS).
Once again !or emphasis it is repeated that,
while the complementary ~ature ot eqs. (14) and
(17) are clearly eVident, the use of eq. (17)
dU' (Castigliano's T~eorem) is restricted to linearlY
iiP =6 - - - - - - - - - - - - (16) elastic structures. A brief exa~ple will serve
tor illustration at the pOSSible pitfalls.
This is the Theorem or Complementary STIergy. The strain energy stored in an initiallY
Now tor the linearly elastic body a very straight uniform column under an axial load P
tmportant theorem follows since (Fig. A7.9c) when deflected into a hal! sine wave is
dU = dU· so that

P~U'
, ,
~ a. u
. rt x
Y :: Y oSlDT; M::PY
y Consider Flexural Energy Only
;.., 6 U .fM""" • P"Yd L
2EI 4EI

Fig. A7.9c •• 0 • -If(§LY'

2 di.)
dx ; Y3 n:
4L
- P~6L~
., U -~
dU = 0 - - - - - - - - - - - - - (17) Fig. A7.10
dP
where 0 is the end shortenir.g due to bowing.
In worne , Because the deflections grow rapidly as P ap-
~The rate of change ot strain ener~J with proaches the critical (buckling) load the
respect to lOad is equal to the ass0"CTat'8d-- problem is non-linear. The details of the -
detlection".-- - - - calculation of U are given with Fig. A7.l0.
Eq. (17) and the above quotation are state- Now according to the Theorem of ViTtual
ments of Castigl1ano's Theorem. ~ (eq. 14) -
For a bOdY under the simUltaneous action of dU _ P
several loads the theorem is writte~ so as to ,j3-
apply individually to each load and its associ- but
ated detlection, thus·:

=0 Theretore
1 (17a)
P"L"
nilEI = P
The partial derl~tive sign in eq. (17a)
or
_ n 2 EI
indicates that the increment in strain energy is
due to a small change in the particular load P ,
P - r::r
all other loads held constant. i (Euler formula for uniform co nem ) .
Note that by "load~ and "detlection" may correct r-esuj.t ,
be meant: Application at Castigliano's Theorem.
eq. (17), leads to the erroneous result:
Load ASsOciated
Detlection dU = 0
dP

dU 2P6L~
dP = nAEI
• The proof of the theorem for the case of multiple ioads is
generally formulated more rigorously. appeal to a simple
diagram such as Fig. A7. 9c being less effective. se e, •• See Art. A17.6, Chapter A-I7 for detailed
for example, "Theory of ElaSticitY" by S. Timoshenko. derivation of this equation.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A7.7

(incorrect) Then

Moral: Do not use Castigliano's theorem for

non-linear problems. 6p = dll
dp = 43.26
•
x 10- P tt.
Fortunately the above restriction upon ~he
use of Castigliano's theorem is not a very Note that in this problem the only use made of
severe one, the majority of avery day structural the calculus was in the differentiation. A
problems being linear. Castigliano's theorem Simple sum was used to form U, as per eq , (2).
Is quite useful in performing deflection calcu-
lations and a variety of applications will be £xample Problem 9
made in the following section~. Derive Bredt's to~ula for the rate at
twist of a hollow, closed, thin-walled tUbe:
A7.6 Calculations of Structural Deflections by Use of
Ca8tigliano's Theorem. ds
Ref. chap. A6.
As the examples of Art. A7.3 have illus- t
trated,the strain energy of a structure can be
expressed as a function of the external loadings The strain energy is stored in shear ac-
prOVided the internal load or stress distribu-
tion is calculable. Having the strain energy so
cording to 1
U=zJ
'd>:d rJ Y .
expressed the deflections at points ot load ap-
plication can be determined with the aid of eq. This is the only energy stored, secondary et-
(17), Castigliano's Theorem. fects neglected. OVer a given cross section q
In the examples to follow the deflections 1s a constant and is given by Bredt's equation
ot a variety of statically determinate struc- q = ~ , where A is the enclosed area ot the tube
tures are computed. Methods of handling redun-
dant structures are considered in subsequent (Ret. Chap. A.5).
articles.
Solution:
Example Problem 8
F1nd the vertical deflection at the point Since the twist per unit length was desired
01' load application of the crane of F1g. A7.ll. the strain energy per unit length only was
Cross sectional areas are given on each written. Thus, assuming y in the axial direction,
member. The stranded cables have effective no integration was made with respect to y. The
~odul1i of 13.5 x 10" pSi. E = 29,000,000 for integration in the remaining direction was to be
carried out around the perimeter ot the tube and
other members. ~'1I5'\.l\'~C. 'I so the index was changed tram "x" to the ~ore ap
A' •
propr-tat;e "e ", Hence (compare with Example Pr'ob , 7)

'r1-i-1.-
~"~'~:
r
~~
.-$:' p 60'
I_
u = J:..
2GJ'
t q ads
t
=~ ~ cis
8AG~ t

E 0 Th~ symbol ~ means the integration is carried out

~30'+- 60' ~ Fig. A7,1l around the tube perimeter. Therefore

Solution:
USE OF FICTITIOUS LOADS
The strain energy considered here was that
due to axial lo~ding in each of the four mem- In the follOWing example the desired detlec
bers. The load distribution was obtained fram tion is at the free end of the bar where no load
statics and the ener 5 J calculation was made in is applied. A fictitious load will be added for
tabular form a~ follows: purposes of the calculation. The rate of change
of strain ener~J with respect to this fictitious
load will be found atter which the load will be
TABLE A7.2 set equal to zero. This technique gives the de-
si~ed result in as much as the deflection 15
MEMBER S LaS
AE X iO"'5 s.aL x 10" equal to the rate £f change at stratn energy
LBS AE with respect to the load and such a rate exists
I OA -1. 50 P 40.0 136 0.66 P Ii even though the load itself be zero.
AS 2.50 P 50.0 11.8 26.48 P
AC 1.58 P 63.0 11.8 13.33 P Example Problem 10
DC -2. 12 P 4,5 13 2.79 P Compute the axial motion at the free end at
I;:; - 43.26 P Ii
the tapered bar of Fig. A7.12.
A7.8 DEFLECTIONS OF STRUCTURES

q - -flA'"
constant J
~
Ao u- 2L
x) q
A.o '" 40.,/in.
= 4 in. 4
L = 40 in.
erally are easier to

:Q:a'nple ?roblem 11
eval~ate.

~ind the deflection at ,oint 8 of the beam

j E '" 10 x 10e psi of rig. A7.:3.
, Flo
p
EI constant
x Fig. A7.12
A, e

Solution:
--- -:-L
L
3
: L
3 3"

Fig. A7.13
After addition of the ~istitious end load
R the axial load frem statics was found to Je
Solution:
S(x) = p. -I- q (L - x)
A fictitious load 21. ·NaS added at point B
Hence, since loadings other than tensile are of and the bending moment dia;ram was 1rawn in VNO
a secondary nature parts.

U = !
2
t0
S'dx
AE Mp
0<X<:L/3

I~
1 [Po + q IL - x)]'dx 0<Y<L/3
2 AoE (1 - 2
XL)
0< ;;<L/ 3

Po'
2A,E
I L

o
dx p.q JL 1L - x ) dx
(1 _ -"-)+ A,E
2L
- 1(_ X-)
0-2L
Fig. A7.13a

+ ~ IL 'L- x)'dx ~en neglecting the energy at shear as being

small-
2A,E ( 1 _ -,,-)
, 2L
L/7
Betore evaluating the integrals it was ob-
2~I
served that the steps to follow, in which U was U =
I c
V[(2P +

r/'
to be differentiated with respect to P, and the
subsequent setting of R = 0 would drop out both ,
the first and last terms. Hence only the second 1 ~ P + P, (L +
term was evaluated. + 2EI , [gIZL - 3Y) 3YIJ dy
9

L~

r
II
u = X + 2L Rq (1 _ 2) -I- X ,
A,E
+ .z.
2E!
3 [IP 2P1.) (L- 30)J do.
-I-
9
Then c

Differentiation under the integral sign With

•
6 = dU = 2L'q 11 _ L~ 2) respect to ?~ gave
dP, AoE L/ 3
DIFFERENTIATION UNDER THE INTEGRAL SIGN 0B=~~~ ir J e
[(ZP+
Animportant labor savings may be had in
the calculation of deflections ·"by Castigllano's
theorem. P'IL " 3 y)J(L 3y) dy
3yl + +
In the strain energy integrals arising in 9 ~
this class of problems, the load Pi' with respec
to which the deflection is to be round, acts as
an independent parameter in the integral. Pro-
vided certain requirements for continuity of the
functions are met - and they invaria~ly are in
- For beams of usual high length - to - depth ratios the shear
these problems - the differentiation with re- strain energy is small compared to the energy 01 flexure.
spect to P may be carried out before the inte- Neglecting the shear energy is eqctvarent to neglecting the
gration is made. The resulting integrals gen- shear deflection contribution. (see p, A7.14)
 . . ._ ....
··7 _

ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES

A7.9

T~e fictitious load P L , havin~served its pur- The total loadings

;058, was set equal to zero Jefore complet~ng 'Nere
the wark.
Plan View

T=T1.+T:a= TaCOS Q ... PR

(1 - cos G)

1 rL
/3 T:!1.us P
+ E:I :1), (2L - 3y) (L + 3y) dy Fig. A7.14b

1 _ PR)lI s in" Q Rd g
U = 2EI

+ 2;J )n/·02[T oCOS Q + PR(l - cos Q)J.:lI Rd g

(Note use of ~RdQ~ for the length of a dif-

Example Problem 12 ferential beam element instead of "dx"). Dif-
Fig. A7.14 shows a ferentiating under the integral sign
cantilever round rod of Fig. A7.14
diameter 0 formed in a
quarter circle and acted au (T, - PR)R" IT/ 2
o sin.:ll Q dQ
upon by a torque To. TI = - EI \
Find the vertical move-
ment of the free end.
R"
Solution: + -
GJ

Fig. A7.14a shows Putting P J the fictitious load equal to zero and
the vector resolution of Plan View integrating gave
ene applied torque To on
beam elements. TL (Q) =
ToGas Q and the moment au = ToR
li
4 - IT
( ~-EI
IT )
OVERT = IP 4
nl,{Q) = To sin 'g. AP-
plication of a fictitious E
vertical load'? (dOwn) at Since J = 21 and G ~ 12.6 the deflection was
the point of desired 1e- negative (up).
flection gave the loadings Fig. A7.14a
shown in Fig. A7.14b.

A7.7 Calculation of Structural Deflections by the Method of Dummy-Unit Loads (Method of Virtual Loads).
The strict application of the calculus to Castlg11ano Js theorem as in Art. A7.6 J leads to a
number of clli~bersame techniques ill-suitee to the solution of large complex structures. A more
flexiJle approach, readily adapted to improved ~book keeping~ techniques is the ~ 2f ~­
Unit Loads developed lr.dependently JY J. C. ~arNell (lB64) and O. Z. Mohr (1674).
--- That the eQuations for the Method of Dummy-Unit Loads may be derived in a number of 'Nays is
attested to by the great variety of names applied to ~~is nethod in the literature ~. Presented
below are t~o deri'lations of the equations stemmir.g ~rom different viewpOints. One derivation ob-
tains the equations by a reinterpretation of the S~bols of Castigliano's theorem - essentially an
a?peal to the concepts of s~~ain er.er~J. The other derivation uses the ~rinciples of rigid ~ody
mechanics. Based as they are upon a commoe set of consistent assumptions, all the ~ethods must, of
course, yield the sa~e result.
Derivation of Squations for the Method of Dummy-Unit Loads
(Virtual Loads)

I - From Cast1glianc's TO_eoram II - From the ?r1nciple of Virtual work

Beginning with the gener-a.I expression for ~nen a system J:' ~orces whose ~esultant is
strain energy, aq. (13) zero (a system in eQ~1Iibr1um) is displaced a

~ Variously called the Maxwell-Mohr Method, Method of Virtual Velocities, Method of Virtual Work, Method of Auxilliary
Loads, Dummy Loads, Method of Work, etc.
A7.10 DEFLECTIONS OF STRUCTURES
I Cont'd. IT Cont'd.
U = ~
2
J S "dX + ~
AE 2
J!'l"u
:::1 + ~ J T~f + - - - etc. small ~ount withcut dis~urbing the equl:1br1~
balar.ce, the work done is zero - obv1ously,
differentiate under the integral sign with since zero res41tant torse ~ovlng through a
respect to Pi to obtain distance does zero work.

6
1
= IS~dX
~
i
+ J M~dX
1
,1
+ - - - - - - - etc.
Consider now the set of equilibrium force~
applied to the truss of Fig. A7.15(b). The set
may be divided into ~~o parts: the "external
system", conSisting of the ~~it load applied at
Consider the symbolS the point whose deflection is ~eslred ane the
~S ~M ~T two reactions fiXing the line of referer.ce. ~d
<lP '"'JP'1'P'"" - - -- - etc. the "internal system" consisting of the axial
1 i i loads acting on the truss members to produce
Each of these 1s the "rate of change of so equilibrium. T~ese latter are denoted by the
-and-so with respect to Pi" or "how much so- synbol ~u". T~is set o~ forceS 15 considered
and-so changes when Pi changes a unit a~ount" OR small enough so as not to.affect t~e actual be-
EQUALLY, "the so-and-so loadln~ for a unit load havior of the structure during subsequent ap-
PI' plication of a real set of ~Jor loads. This
Thus, to compute these partial derivative unit load set is present solely fer ~thematical
terms one need only compute the internal load- reasons and is called a "virtual loading" or
ings due to a unit load (the virtual load) ap- "dummy loading".
plied at the point of desired deflection. For Assume now that t~e struct~e undergoes a
example, the term elM/elPt, could be computed in deformation due to application of a set of real
either ot the ~HO ways shown in Fig. A7.15a. loadS, the virtual leads ftgoing along for t~
ride n • Each ~ember of the structure suffers a
RATE METHOD UNIT METHOD deformation denoted by ~ ~ • The Virtual load-
ing system, bei~g in equl11b~1~ (zero reSUltant;
by hypo'thes t s ,10es work (I'virtual work") equal
to zero. Or. conSidering the s~bdivision of the
virtual loading system, t~e work done by the
external Virtual load must equal that absorbed
Fig. A7.15a by the internal vrr-tuaj loads. The work done by
the external virtual forces is equal to one
pound times the deflection at joint C, the re-
actions Rl. and Ra not moving. !~at is
I1 = Pl.X m = dummy loading
Exter~al Virt~l Work = l' x 6 c
(=~~J
oM
»l. = x x
The
=
internal Virtual Nark is the sum over the
structure of the products of th~ ~ember vir~ual
Likewise, d s~ Pc, where Pc is a load (real loads u by the member distortions~. That is,
or fictitious) applied at jOint c ot Fig. A7.15b, Internal Virtual ~ork = ~ u.~
is given by the loadings for the unit load ap-
plied as shown. Then equatin~ these works,
In practice the use at the WIlt load is
most convenient. Using the notation 1 x 0c = Z u·d .
_ oM If the deformations 6 are the result of
,m =~, elastic strains due to real member loads S then
1 Do = 8L/AE for each member and one has

v a
Pi'
q = aq
- ~
:V 6 = Z u 8L
c AE

for the unit loadings, the deflection equation The ar~~ent given above may be extended
becomes QUickly to include the internal virtual work of
6 =
1
Sud:< +
AE
IlmdX +
EI
~
GJ
f J J
virtual bending moments (TIl), t or s i on loads (t),
shear loads (v), and shear flows (q) doing work
during defo~tions due to real moments (M),
+ J If
~ + q qGfdY - - - - - - - - - (IS) torques (T), shear loads {Vr;-ind shear flows
(q). The general expression becomes
-1.0 -1.0 6 f Sud=< ~ J Ttc.x
1
=
AE
+ ;
EI
...
GJ

+J AG +J;q q
VvdX dXdy
Gt - - - - - (18

Fig'. A7.1Sb ~ Note that the deformations are not restricted to those due
to ~lastic strains only. They may be the result of elastic
"un loads due to a unit (virtual) load. or tnetesuc strains. temperature strains or misalign_
ment corrections.
 . '.?:'"B.'P!

ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES

A7 11

In applying eq. (18) the labor of a =eflec-

t~on calculation divides conveniently into
several steps:
1. Calculation of the real (actual) load
distribution (S,M,T, etc.)
11. Calculation of the unit (virtual) load
distribution (u,m,t, etc.) due to a unit 111
(virtual) load applied at the potnt of desired "U1," loads "U.' loads
deflection and reacted at the reference polnt(s).
Fig. A7.16a Fig. A7.1Bb

111. Calculation ot flex!bl11tles,

1
'. ~
1
-A~ , "'1 J etc. ...... Lbo
AIls 10·· toslO·
TABLJ: AT.3

• ., II.
.....,. """""
~s 10· ~JCIO.
iv. Summation and/or Integration.
Here again note the general nature of the
AS .. Ibo

4.1•
AI ""Ib
8.2'1'0 10,500 1.' a a.. " a

terms "load" and "ceriece i on". (See p. A7.6)

The following examples apply the method of
BC

..
CD ...." 3.074

3.0T4
g.T5'

g. '1'1
2,~0

2,250
a•
• • •
... "
a a
a
dummy-unit loads to the determination of ab-
solute and relative deflections, both rotation
and translatioo.
,., .. 5.385

5.385
5.5111

5.1111
_6,~

-5,2llO .0. T5 1
1 22.01

12.01
_21.~

_21.31

aB 3." 1.821
• c 1 a •
Example Problem 13
The pin-jointed truss of Fig. A7.l6 is
•• 50 lO.a 4.1128 _8, 'I'1lO 1-1.25
• 53."
•
acted upon by the external loading system shown.
BQ 50 3.4' 14.3. _.000 1." • a 111.80
•a
The ~ember loads are given on the fi~~re. Mem-
ber properties are given ~n Table A7.3. Find
00
SF ....
50 5.3G

3.0'1'4
, 11.320

13.012
.3, 'liO
3,000
•
•
•
• a a
the vertical movement of joint G and the hori-
zontal ~ovement of jOint H.
co
DB .. 3.074

3.0'1'4
13.012

13.012
_l,OOO

3.000
•
•
• •
• a
•
a

.
A 10
1000"
sods 2250
1000*
Ie 2250
1000'*
10
fj~gl~(?'bgl~~~ l
Answer-s : oG ::: 0.286"
VER
~"
0
-;o:"g40"
'-5250 .5250 (l N 1 O~OR ::: -.0587" (the negative Sign means

\~30" t30"'+ 30"1 the joint moves to the LEFT since the unit load
was drawn to the RIGHT in Fig. A7.16b).
2000t 2000t
Example Problem 14
Fig. A7.16 For the truss of Fig. A7.16 rlnd the fol-
Solution: lowing relative displacements of joints:
c) the movement of joint c in the direction
only the energy of axial loadings in the of a diagonal line joining c and F.
members was considered. Unit (virtual) loads
were applied successively at jOints G and H as d) the movement of joint G relative to a line
shown in Figs. A7.15a and A7.16b. All Sand u joining points F and H.
loadings were entered in Table A7.3 and the
~elatlve deflections are determined by
applying unit (Virtual) loads at the points
calculat~on ~or 0 = J Sudx
AE
was carried out JY where the deflections are desired and by support
ing such ~nlt load systems at the reference
forming the sum ot SuL/AE terms fer the ~embers poir.ts of the motion. ThUS, ror solution to
,art (c) a unit load system was applied as
UL
, e.
of the truss, .. 6 = ~"
'" SAE
_ shown in Fig. A7.16c and for the solution of
A7 12 DEfLECTIONS OF STRUCTURES

part (d) the system of l~it loads of r1g. A7.l6d
.cas used. Table A7.4 completes tne solution,
the ~eal loads and member :leXibilities
(A~) being the s~e as for ex~ple problem 13.
o -.6 roO
Rotations, both absolute anQ "elative a~~
dete~ined by a~ply1ng unit (virt~al) couples
to the member or port~cr. of struc~ure ~hose
rotation is desired. The unit ccuple is ~esis~­
ed by reactior.s placed en t~e line of reference
for the rotation. Thus Figs. A7.16e ~nd A7.l6f
-.375 -.375 show the unit (virtual) loadings fer par t.s (e)

o I~,,;>°t"
and (f) respectively. T~ble A7.5 completes the

iJ; '_. ~Jl f

il&
calculation, the real loads and ~ember ~lexib~l-
>;) ':c: , '..,i' " a O..n -r :0 It:I
lties (L/~) be Ing the same as fer examp.Le
Ac
6' 0 0 0 ,. FOG 0" H
problem 13.
I I .5
R=U' .5 1jjo

/\/0
1/50*, 1/50# R=1/40'"

"~~?@?1
"us" loads "u ..." loads

Fig. A7.16c Fig. A7.16 d

R=l!40'll' -.025 -. 02~ G'l/50# /G R=1/40#

1/50*
TABLE A'l.4 "u e" loads "u.!l;" loads

...h. x las Fig. A7.16e Fig. A7.161

MEM. AE In/lb ~x1a" Su ...L x 10"
{See Table
S,b. '. '. 1". A'r1.llcb.
A7.3}

AS 8.210 10,500 0 - 0 0 0

Be 9.159 2, aso -. , -.315 -13.11 -8.23

CD 9.7:19 1,%50 0 ·.31:1 0 -8.23

TABLE A7.5 -
E" ~.591 ·5,250 0 0 0 I 0
..k. re-
FG 5.591 -5,2:la -.' 0 17.61 0 AE
:I:
WIb ~xlO.s ~X10S
GB S,621 0 n c c , blEM.
(See Table
A'l.3)
S,b a&lnch-J. Il....lncb.-l.
AE

BE
eo
4.926

14.368
-8,750

'.000 ! 1.0
0 o
.e"
o
71.84 ....
0 AE
Be
6.a70

9. 7S~
10,500

a. 450
. 0.5

. D25
0
_. D.5
1."
.55
0

.0.55
DO

BF
.9.320

13. 012
-3,~0

2.000 -
-
....
0

·.50
.'20 0

·20.82
-21.8

·13. a
CD
EF
9.

~.591
7~9 2,250

-5,2:10
.020

·.025
0

0
.55

•. 73
0

0
CG 13. on ·1.000 0 .10.41 0 FG 5.591 ·5,2:10 ·.025 i 0 •. 73 0
DB 13.012 2.000 0 _.50 u ·13. a GB e.eu 0 0 0 0 0
z = 65.87 Z =·19.38 BE 4.926 -8, 750 0 0 0 0

BG 14. see '.000 0 .OU 0 1. 08

DO 9.320 ·3.750 ".015 0 .52 0

BF 13.012 2,000 0 0 0 0
,
CG 13. 012 -1,000 0
I
0 , 0 0
Therefore the movement of joint c towards
, DB
joint F was 0 = .06587 inches and the motion of 13.012 2,000 0 0 0 0

joint G relative to a line between F and H was Z = 4.73 .z =0.53

6 = -.0194 inches, the negatlve sign indicating
an upward movement. .
Example Problem 15
For the truss or Fig. A7.l6 determine
a) the absolute rotation of member DG ~herefare the absolute rotation of member
t) the rotation of member 80 relative to DO was Q
DG = .00473 radla~s and the rotation of

member CO. 80 relat~ve to CG was Q

3G
_ CG = .00053 radians
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A7.13

a.ample fToblem IS points C, DJ ~, the attachment ~oints of an

tine the vertical eeflect:on of point C for aileron or a flap. The wing beam deflection
the cantilever jea~ of ?lg. A7.17 carrying a bends the aileron or flap structure by applying
concentrated load P at its end. Al~o find Slope a lcad at 0 thru aileron s~pporting bracket. To
of elastic ~urve at 8. know this force the deflection of the wing beam
at D relative to li~e CE ~ust be known.
Solution:
Solution:
Mmdx
f -W- Origin at 8: -
with origin at 9
Ii = 30 x ~ = 15 x"
M = - Px (Fig. A7.17)

For virtual load1~g (?ig. A7.17a)

I!l. = 0, far x cb
-1 (x - b), for x)b

,±r '°'1
a Fig. A7.18

Hence !1mdx =- Px(-x + b Idx = (?x'" - Pbx Idx

I'"
P rL a = 1.[ x' bX'J L -----+I c,....-- 0 s I
EI ) b (x - bxidx EI 3 2 b
Fig. A7.l8a
a c· .' Rd··'
<
P [, 3L'b b'
3EI L - -2- + "2
J ~ OJ x 20
u = -.5 (x-20) =-.5x+10,whenx=20to50
m -.5(x-20)+1(x-50)=.5x-40 J when
If b = zero, then 6
3
= PL
3/3EI
x = 50 to 80
l1mdx
) A

,
C
, Fig. A7.17 J E1

l== L
c
4p =F:r
1 k 50
15 x~ (-.5X + 10)dX + ;;'1
1 h 80
15 x"
1' It
a Fig. A7.17a .... .1. 20 " 50
(.5x-40jdx
c r- B
)' .,l un1 t.
Fig. A7.17b
150x,l50 + l. [7.5 x' _ 600 x' l80
ooupl_
3 J20 S1 4 3 jso

ec,
=JMmdx
2;1 = ISO; [_ 11.72 + 6.25 + .300 - .400 +

For vi~tua1 loading See Fig. A7.11b 76.8 - 102.4 - 11. 72 + 25.0]

a = 0, x cb , - I, x>b
17.9 X lOll
EI
Hence !1mdX = (-?xi (-Ii ex ?x ex
Therefore deflection of point 0 relative to line
joining ~E is down because result comes out nega-
J Mmdx_PfL
SI - ~ b xcx 2EI
o
tive and therefore opposite to directicn of Vir-
tual load.
0' , Example Problem 18
If b = 0, ~ = 2EI x~

j:~~lr,1
1'''' lb/rt
'; II i i i
a.~ple Problem 17
Tor tne um r orm.Ly loaded cantilever beam
of Fig. A7.18, f~nd the deflection of point 0
relat~ve to the line joining points C and ~ on
the elast~c curve of the beam. This is repre-
sentative of a yr~ctlcal ?rOblem in aBronauttcs,
in that A3 :n.1.~ht represent a rear wing beam and
tn ~L-

(a)
•
",to
'2
L
2

(b)
Fig. A7.19
x-'1
m diag'l
wx e

• u
'oiL'
'2
( c)
R=l-1i' ~ '\
R=l/L unit
couple
1

Fig. A7.20
R=l/L
A7.14 DEFLECTIONS Ol" STRUCTURES

Find the horizontal deflection ot point C V

dy = J ex and ~ = A-' J where A cross sectional
for the frame and loading of Fig. A7.l9. Also ~s
angular deflection af C with respect to line CD. area of beam at section and ~s ~cdulus of
rigidity: and assuming that the shearing stress
Solution: V
A is uniform over the cr~ss-sectlon.
Fig. b shows the static moment curve for
the given loading and Fig. c the moment diagram Therefore 1 x 0 = Vvdx
----A-.
~s
• ~hen the total detlec-
for the virtual loading of a unit horizontal
load applied at C and resisted at D. tion for the shear slipS of all elements of the
beam equals
-_JMmdx
- - ,.
EI
M = wLx
2
wx
2
a
:n = h
e total = L VvdX
, AEs
f _______ (a)

hence
where V 15 the shear at any section due ~o g~ven
L(WLx WX')
2 - T h d x =r,(WLx'_'/IX') 1!. JL
loads. v = shear at any sect10n due to unit
•f EI
~ 4 6. EI
c
hypothetical load at 'tae point where tne aer iec-,
tion is wanted and acting in the desired direction
of the deflection. The reactions to the hypo-
1 whV' thetical unit load fix the line of reference for
• 12 EI the deflection.
A is the cross-sectional area and Es the
To tind angular deflection at C apply a unit
imaginary couple at C with reactions at C and D. modulus of rigidity. Equation (a) is slightly
F.ig. A7.20 shows the virtual m diagram. in error as the Shear:ng stress is not ~~i~orn
over the cross-section. e.g. being ~raballc ~cr
a rectan~~lar section. However. ~he average
So =JL~EI
1.
EI
= fL (WLx _
2
WX') ~ dx
2L J;I,
= ijL
shearing stress gives close results.
For a unito~ load of w per unit le~gth, the
• , L center deflection on a s~ply supported beam 1s: -
(W~' - w:: ) dx = ,;\

1
= 24 Ef
[~' - w:~ J
wL;'
c
°center = 2f~~=2
L AE s
,-
~
wL'
r
e
(wL
2
wx)

AE s
1
~
ax

Linear Deflection of Beams Due to Shear by Virtual Work.

SAEs
Generally speakingJ shear deflections in For bending deflection for a simply supported
beams are small compared to those due to bending beam uniformly loaded the center deflection 1s
except tor comparatively short beams and there-
tore are usually neglected in deflection calcu- •
lations. A close approximation is sometimes
made by USing a modUlus of elasticity slightly Henae
less than that for bending and using the bending wL'
deflection equations. 8AE s
The expression for shear deflection of a
beam is derived tram the same reasoning as in 5wL"
= 24 (Er J using is = .4:E; I
384EI. r = radiUS at gyration.
previous derivations. The virtual work equa-
tion tor the hypothetical unit load system for a
shear detrusion dy (Fig. A7.2l) conSidering only For beama and cr~els r is apprOXimately ~ d
dx elastic is 1 x O=vdy where v is shear on
section due to unit hypothetical load at paint and tor rectangular sections r = v12
c (1 = depth)
OJ and dy Is the shear detruslon at the element
dx due to any given load system or any other
cause. In aircratt structures a ratiO of d 1s
L

"'~dLdy
seldom. greater than 12
~--
6

-f-5-' ----ft
T,.
Thus the Shearing deflection in percent of
the bending deflection equals 4.1~ for a ~ ratiO
Fig. A7.21

• Sometimes "G", instead. Seep. A7.4

ot iz for I-beam sections and 1.4 percent for
rectangular sections.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A7.15

Example Problem 19 t ; torsior~l ~oment at any section due to

a unit virtual couple acting at section
501' 10011"
where angle of ~N1st is desired and

t '+ lO'4
'O
A
IF==='
Unit.
Load.
acting in the plane of the desired de-
flection. (inch Ibs/inch lb)
Example Problem
Fig. A7.22
Example Problem 20
Find the vertical deflection of free end A Fig. A7.Z3 shows a cantilever landing gear
due to shear deformation for beam of Fig. A7.22 strut-axle unit ABC lying in XY plane. A load
assuming shearing stress uniform over cross- of 1000* is applied to axle at point A normal to
section, and AE constant. XY plane. Find the deflection of point A no~l
s to XY plane. Assume strut and axle are tubular
"' _ I Vvdx and of constant section.
vA., - -
AE-s Solution:
V = 100' for x = 0 to 10
V 150 for x 10 to 20 The loading shown causes both bending and
twisting of the strut axle unit. First find
v = 1 for x = a to 20 bending and torsional moments on axle and strut
hence due to 1000* load.

°A 1
AEs to ...L
0 100·1 dx + AE
s
fO
10
150·1 dx
I
Y

10 20
1
= IE:s [lOOX] + ~.[150X] 2500 I / 36" 3°1
0

1/ \\·~~3'i s:
AE s
0 10 ,.
Method_of Virtual Work Applied to Torsion of Cylindrical Bars. B I
B A
The angle of twist of a circular shaft due Fig. A7.23 Fig. A7.24
to a torsional moment may be found by similar
reasoning as used in previous articles for find- Ttdx
ing deflection due to bending or shear forces.
E J
The resulting expressions are: - s

o = JTEts~ - - - - - - - - - - - - - (A)
Member AS M = 1000 x, (for x
T = 0
o to 3)

= 3000 sin 20° + 1000 x,

Member BC
(tor x = a to 36)
(B) Tse ; 3000 cos 20 0 constant
between 8 and C.
In equation (A), for translation deflections,
T twisting moment at any section due to Now apply a unit 1* rorc~ at A normal to xy
applied vNisting forces. plane as shown in F!g. A7.24 and find bending
t = torsional ~oment at any Section due to and torsional moments due to this 1* force.
a virtual unit 1 lb. force applied at
the point where deflection is wanted
and applied in the direction or the Member A8
desired displacement. (in lbs/Ib)
Es Shearing modulus of elasticity for the m = 1- x = x, (r or x =o to 3)
material. (also nGn)
J polar moment of inertia of the circular t =0
crcss-eect ron.
In equation (B), for rotational deflections, Member Be
Q = angle in twist at any section due to
the applied vNisting moments in planes m =3 sin 20' + 1- x (for x =0 to 36)
per?endicular to the shaft axis.
Angle in radi~~s. t • 3 cos 20' constant bevNeen B and C.
A7.16 DEFLECTIONS OF STRUCTURES

Subt. The shear :lows shown on ~he (nearly) horizontal

edges of the web panels are average values. ?i6'

6 = J I'!mdx -+-
81
JTtdx
E J
A7.27 ~s an explOded vi~N of the beam shaNing
the unit (Virtual) loads.
s

= tr J: 1
1000 X . X dx • Er f: 6
K1000 x • 1026)

(x + 1.026)J dx -+-
1
E-J
f36 (2820) (2.82) dx
s 0

Virtual loading. Fig. A7. 27

Since both axial loads and shear flo'~

were considered, ~he form of deflection
eq~tion used was
1 (16,92500) •
EI
q q dxdy
Gt
_1_ (286200)
EsJ
I~tegrations in the ~langes were made
assuming linear load variations. Such ~~ in-
Note: A practical landing gear strut would in- tegration carried out over a uniform :lange of
volve a tapered or reinforced section involving length L whose real load varies from 8 1 to 5 j
a variable I and J and the integration would
have to be done graphically or numerically. and WhOS8 virtual lead varies :rcm ~i ~o Uj
yields
Example Problem 21
For the trnn-weo aluminum beam of Fig. L Sudx =
A7.25 determine the deflection at point Gunder
the loading shown. Stringer section areas are J Q AE
given on the figure.

A ... A=.15 B A=.15 C ,..A-.15 D The integrations in the trapezoidal sheet

rL I
20"
t=.0
32
1l t=.032 1
A=.08
t=.032
A=.08
A-.05"
TIS"
--..L
panels were made using the shear flows en the
(nearly) horizontal sides as average values,
assumed constant over the panel. with this
E
~ 20" ----..+- -20"
rF I G
-+- 20"
H+ P.j80"*
-4 Simplification
Gt
= qAV 55 qAV q,w :ixdy
Fig. A7.2S
where S 15 the panel area.
Solution: The calculation was completad in Table
A7.6.
It was assumed that the webs did not buckle
and carried shear only. Fig. A7.26 is an ex- c==.:===;::::~~~====
•
ploded view of the beam Showing the internal
real loads carried as determined by statics:
3. OP_ 2. 184P 1. 201P : :.:.,. p, , ".'"

~ID·
-··*~i ;~~
·'.00 P ,_L,H P, ·1.00,_:.011 .un I .,.,., ,
.'.'M "'-'-:10""'.01' 'u. ".0 P

.04
3.0P...".. 2:"I84P
.O!~~~~. W I 1.201p P ...:,:" ,-',W")
... ~.r'._.·
",,"'''1 i •

.,.\< ""I
i.,'. • ",.i,~......5. •".,
I : • 171.0 P."

Fig. A7.26 a.c_r...... <>U'. I

C·D-<l-HI._P' : .1
• The equations of statics for tapered beam webs are
derived in Art. AIS.18, Ch. A-IS.
 ANALYSIS AND DESIGN OF FLIGHT VEH LE TR C RE A7.17

6 = (271.0 + 13.4) P x 10- 11

G 2 -j ~hr;l~Tdx
.147 (1800) 10-~ in = .265 in.

A7.8 Deflections Due to Thermal strains.

II5'~ 0

As noted in the "virtual work derivation" ;-t I-dx

of the dummy-unit load deflection equations, the
real internal strai~s of the structure may be
due to any cause including ther.nal effects.
Hence, ~rovided the temperature distribution and
thermal properties of a structure are known, the
dummy-unit load method ?rovides a ready means
for computing thermal deflections.
Example Problem 22
Find the axial movement at the free end of
a uniform bar due to heat application to the
fixed end, re~ulting in the steady state tem-
perature distribution shown in Fig. A7.28. As-
sume material properti9s are not functions of
temperature.
T temperature above
ambient tempera-
tlire
an empirical con-
stant depending
upon the~l prop-
erties and rate of
Fig. A7.28 heat addition.
L
a b
Fig. A7.29
Solution:
T~e axial deformation of a differential
element of the u?per flange (sUbSer}pt u) was
ass~~8d given by Au
= a T ax where a was the
material ~hermal coefficient of expansion.
The lower flange, having received no heating
underNent no expanSion.
Inasmuch as a thermal expansion is uniform in all
directions no shear strain can occur on a material element.
Hence no shear strata occurs in the web. The apparent
anomaly here - that web elements appear to undergo shear
deformations J = a~ (Fig. A7. 29b) - is explained as
follows: The temperature varies linearly over the beam
depth. The various horizontal beam "fibers" thus undergo
axial deformations which vary linearly also in the manner
of Fig. A7. 29b giving the apparent shear deformation. No
virtual work is done during this web deformation since no
axial virtual stresses are carried in the web.
~ith ~he addition of a unit (virtual)
load to the free end, the virtual loadings ob-
tained in the flanges were:

L - x
Solution: u
U h
(~ - "L)
The thermal coefficient of expansion of the
rod material was u. Hence a rod element of T~en ~he ceflection equation was
length ax experienced a thermal defo~tlon
a = a . T • ax. Application of a unit load at
the bar end gave u = 1. T~eretore ~J~
o n
• a T

T . dx

23
2.X8.!l:p19 Proj19m
~heidealized ~No-flange cantilever beam
of ?ig. A7.2Sa undergoes rapid ~eatlng of the
upper flange to a ~emperature T, uni:orm span-
Nise, above that of the lower flange. Deter-
xlne the resulting displaeemer.t of the free end.
SXa.m.ple Pr::;ble:n 24
~he ~irst step in computing the the~l
stresses i~ a closed ring (3 times indeterminate)
involves cutting the ring to ~ke i~ statically
determinate and r1~d1ng the relative movement of
the ~NO cut faces.
Fig. A7.3Oa s~cws a unifo~ circular ring
whose insi~e surface is heat~d to a temperature
T above the outside surface. The temperature is
constant ar8und the c~rcum!erer.ce and is assumed
A7.l8 DEFLECTIONS OF STRUCTURES

to vary linearly over the depth of the cross

section. Find the relative movement of the cut
o
surfaces shown in Fig. A7.30b.

• b
Flg. A7.30
Solution:
An element of the beam of length Rd¢ is
shawn in Fig. A7.30ce Due to the linear tem-
perature variation an angular change dQ =
RaT
~ d$ occurred in the element. The Change in
length of the midline (centroid) ot the section
was 4 = R~T d~. Unit redundant loads were ap-
plied at· the cut surface as shown in Fig. A7.30b
giving the following unit loadings around the
ring.
From. unit redundant couple (X)
Remarks:
In the three elementary examples given
above no stresse~ were developed inasmuch as the
idealizations yielded statically dete~inate
structures which, with no loads applied, can
have no stresses. Indeterminate str~ctures are
treated in Chapter A.a.
A7.9 Matrix Methods in Deflection Calculations.
Introduction. There is much to recommend the 'lse of
natr rx methads'~ for the hand Ling of 'the quarrt Lty
of data ariSing in the solutions ot stress and
deflection calculations ot complex structures:
~he data is presented in a form suitable for use
in the routine calculatory procedures of hi~~
speed digital computersj a fleXibility of opera-
tion is present which permits the solution of ad-
ditional related proble~ by a simple expansion
of the program; The notation itsel: suggests new
and improved methods both of theoretical ap-
proach and work division.
The ~ethods and notations employed here ar.d
later are essentially those presented by Wehle
and Lansing@ in adapting the Method of D1..1L1!IlY-
Unit Loads to matriX notation. Other appropriate
references are listed in the bibliography.

~ =1 (m positive if it tends to open BASIS OF METHOD

the ring). Assume the structure to be analyzed has been
U (= axial load) = a (positive of tensile)
idealized into a truss-like assembly of ~ods, ,
x bars, tubes and panels (sheets) upon which are
From unit redundant axial (horizontal) load (Y) acting the external loads applied as concentrated
loads Pm or Pn, each with a different numerical
my -R (1 - cos $)
U
y
= cos $
From unit redundant shear (vertical) load (~)

m
a
-R sin $
ua - sin $
The deflection equation by the dummy-unit load
method is (.) (b)

e = Ju • 4 + f ru . dQ Flg. A7. 31. Idealization into an

assembly of bars and panels.
Then

Q,)For the reader not famUia.r with the elementary arithmetic

rules of matrix operations employed here, a. short appen-
dix has been included.
iL. B. Wehle Jr. and Warner LanSing, A Method for Re-
ducing the AnalySis of Complex Redunci:Uit""""StrUCtures to a
2n R-aT (negative indicating movement to Rouune-Procedure, Journ. of Aero. Sciences, 19, october
= - h the right) mr.- -

4Z
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A7.l9

subscript. T~us the system of Fig. A7.3la 1s THE STRAIN ENERGY

idealized into that of Fig. A7.3l~.
It is next desired to write the strain
With the above idealization an improved energy as a function of the q's. Continuing the
scheme may be employed :0 systematize the com-
putation of deflection calculations. The fol- illustrative example, write
lowing steps summarize the procedure wh t ch is
discussed in detail in succeedi~g sections.
I. A set of internal generalized forces,
denoted by qt or qj (i, j are different numeri-
cal subscripts), is used to describe the inter-
nal stress distribution. The q's may represent 1 q.j,'X'dx
axial loadS, ~oments, shears, etc. In con- 2 1. ~
junction 'Nith a set of member flexibility coef- '-
ficients, Uij, the q's~employed to express
the strain energy U. Ulj ~ives,the displacement
of point! per unit :orce at point ~ ••
II. EqUilibrium conditions are used to Observe that each of the integral terms in
relate the interr~l generalized forces qi, qj to the above expression is a property of the struc-
the external applied ~ , Pm or Pn • With this tural element (variation of El) and of the nature
relationship the strain energy expreSSion ob- of the associated generalized force (exponent on
tained in I, above is then transformed to give variable). Introducing the notation
U as a function of the P's.
III. Gastigliano's Theorem 1s used to = IL'Y'd Y
compute deflections. - Ela
c
CHOICE OF GENERALIZED FORCES
Consider for example the problem of writing =
dy
jL.E1.
the strain energy (Of flexure) of the stepped c
c~~tl1ever beam of Fig. A7.32a, assuming ex-
ternal loads are to be applied as transverse the strain ener~J becomes
paint loads at A and E. The set of internal
generalized forces ot Fig. A7.32b will com-
pletely determine the bending ~oment distribU-
tion in the be~ elements and hence the strain
energy. s~t (b) then is a sat1stactory choice
of generalized torces. - - (19)
It should be pointed out that sat (b) is
not a unique set. Other satiSfactory choiceS Equation (19) is an expression for U Which
(not an exhaustive display) are shown in Figs. could have been written lmm.ediatelY from physi-
A7.32c, d and e. The final selection may be cal considerations. Each coefficient aij is the
"made for convenience or personal taste.
Note that only as ~ny generalized forces displacement at paint i per unit change in torce
are used per element as are required to deter- at point j. This identity is eaSily seen by
~ine the significant loadings in that element.
applying Castlgliano's Theorem to eq. (19).
With this interpretation the first ter.n in eq.
(19), representing the strain energy in the
£ EI. A EI, B
{I
q,
I)
q.
(
q,
outer beam portion, is written by analogy to
2: S'L)
eq , (2) of Art. A7.3 ( U = 1 AE •The remain-
LaY::-L\~{a) (e)
ing three terms, representing the energy stored
~q1. q3 ~n the lIh~er Jeam segment by qa and q~J are like-

~ ql)", JI
~qg
\ h
q1.
wise readilY written, With proper account taken
q, (b) (d) tor the cross influence at one :orce upon an-
other (the na a3 qa q~~ term).
M = q1.X O<x<L1. q. q'f No'te that
qtj I)
M;Q3+qaY O<y<L. (e)
= = "jl - - - - - - - - (20)
Fig. A7.32. Some possible choices
of generalized forces.
Hence
~ •• (Maxwell's Reciprocal Theorem)
-("Relative displacements in the individual member") ,.
A7.20 DEFLECTIONS OF STRUCTURES

The general form of strain energy express-

ion is (ex,anding sq. (19) by ir.duction)

2U = qlql all + qlq~ al~ + - - - - - + ql qN a(N

,
,
I

.. qaql <l<ill • q~qa 0.a a ... - - - - - - - - - -I

I

,,
+Q"q1. (l"l + q"q" (l"3 ... - - - _ - _ - - - _I
,
,I
+qfoQ1. au .
,,
- - - - - - - - - - - - - - -1

,, Fig. A7.32f
,,
,

I
I I
+ qNql aN' + - - - - - - -- - - + qN qN (lNN
In matrix notation

In matrix notation this equation is written q, 1 o.

(see appendix) P,
q. • 1 1
P.
Ql2---a1.~ r
q. ~, 0
,,
C;3l Cl3 '1------:
,, q. (21)
: Symbol:cally these relationships are 'NT1~tan
,, ,,
,
, ' ,,
• - - - - - (22)
~~------- ,\';' qN

or, more concisely, The matrix [Jim] 1s called the ~1t load
dlstr1butlon~ inasmuch as any one col'~ at

[GimJ ' say the mt h column, gives the values of

In the matrix [Cl i j ] many (if not most) of
the generalized forces (the Q's) for a unit
the elements are zero. In the specifiC example, value of load P~, all other external loads
eq. (21) would be 'Nritten zero.

o 0 q, THE STRAIN ENERGY IN TSR11S OF APPLIED LOADS .

If eq. (22) and its transpose are used GO sub-
o q. st rtuts for the q.. "s in eq. (21) one gets
•
o q.
2U (23)

The problem of computing and tabulating

various (lij'S is considered in detail later. In the here i and j are used intar-
notat~on
Chan~eably as are ~ and n. Also [~lJ is the
RELATING THE INTERNAL GENERALIZED transpose of [G1m] I i.e. lntercr~nge of sub-
FORClli TO THE EXTERNAL APPLIED LOADS scripts denotes transposition (see appendiX).
For the statically determinate structures It the matrix triple product in eq. (23) is
considered in this chapter the internal forces fo~ed and defined as
are related to the external loads by use of the
equations at static equilibrium. A set of
- - - - (2?)
linear equations results. ThUS, in the speCific
example conSidered, it Pl and Pa are the ex-
ternal loads applied as in Fig. A7.32f, then by
statics (reter to Fig. A7.32b) then
 _ ~ . -' ...._\:.
._.._m;¢ '

A7.21

L J r,L':rJ1J)1:1 i,I
plication itself yields a summation to complete
2U p
m
P - - - - - - - - - (25) the calculation.
?rcm this last discussion it is clear that
eq , (26a) also may be derived by formulating the
Sq. (25) expresses the strain energy as a Dummy-Unit Load equations (Art. A7.7) in matrix
function of the external ~pplied loads. notation. -
In the ~pecific example being used To illustrate the application of the ~trix
methods presented thus far, a brief and element-
ary example 1s worked with those tools already
~ rLa1 1L'J
10
developed.
Example Problem 25
Determine the influence coefficient matrix
rorxransverae forces to be applied to the uni-
DEFLECTtONS BY CASTIGLIANO'S THEORE!1 form cantilever ot Fig. A7.33 at the three points
Application of Castigllano's theorem to indicated.
eq , (25),
,@ ,lID ,®
(25)
t b I.
3"""" 3 -e- 3
b, b J
c:::::::=::=J.~ q1.
=:J~)q"
yields q.
="q
q. '
I~ ) (26 )
I ~p n .f Fig. A7.33

The steps in passing to eq. (26) may be
demonstrated readily by writing out sq. (25)
for, say, a set of three loads (a, n = 1,2,3),
differentiating successively with respect to
Pl.. ,'Pa and P" and then re-collecting in -catrr tx
rorn,
The matrix ~mnJ gives the deflection at
the ext ernal points "a" ror urn t values of the
loads? and 1s therefore, by definition, the
n
Solution:
T~e choice and numbering at genera11zed
forces are shawn on the figure. These forces
were placed so that pr~vlously derived express-
ions for the a's could be used. The following
member flexibility coefficients were computed.
Note that the only non-zero coefficients ot
mixed subscripts (i not equal to j) are those
for loads common to an element.

l
of influence coefficients.
~trix
---- L/~.dx L'
COt1PARISON ·.ITH DUMt'lY-UNIT LOADS SQUATIONS
al,l = Q ~ = BIEr = aa::a = a.....
It is instr~ctive to Nrtte eq. (26) out as This expression was adopted from that de-
veloped for a transverse shear force on a canti-
lever beam segment in the preceding illustrative
example.

and compar-e the expression with a typical term

from the dumny-unlt load ~et~od equations, say
SL
6=zu E: ' :n the ~~x equation (25a) ~he
A This expreSSion is for a couple on the end
[t~iJ Is the unit (virt~al) matrix correspond- of a cantilever segment (of length L/3).
Ing to the symbol
, TIU" in the simple sum. The L/ 3
[2. i .:1 J are the member' r Lexfb i Lt t Les correspond-
Ua " - au -
__ ydY_L a
EI - 18EI
J
0

ing to
Ai:
L...,. T~e ma tr tx pr'oduct l'J:l] :Pn (
This expression is for the cross influence
;ives the :nember load distrlb~t~ons due to the af a couple ~~d a shear load an a cantilever
real applied loads, hence these are the "B" segment. Collecting in matrix form,
leads. Finally the operation of matrix :nulti-
A7.22 DEFLECTIONS OF STRUCTURES

as per eq • (26).
1
27 0 0 0 0 T'he matrix G!Il11] is seen to be svmetc-rc

0
1 1
0 0 about the main diagonal as it ~ust be: from
27 6L reciprocal theorsm A
~~Nell's .c.'
nm (see
[a !] l 3EI
L'
0
6L
1 1
V 0 0
eq. (20)). mn

1 1 A7.10 Member Flexibility Coefficients:

0 0 0 27 6L Compilation of a Library.
1 1 Several ~ember ~lexibi:ity coefficients are
0 0 0 -r-r
6L w derived below for various members ~d leadings.
A more comprehensive listing 1s available in :he
The unit load matrix [G imJ was computed by paper by Wehle and Lansing referenced ea:'lier.

successively applying unit leads at ~oints 1, 2 BARS

and 3 and computing the values of the q's by The energy in a uniform bar under var-ytng
statics. axial force (Fig. A7.34l is
1 0 0
1 1 0
[G1m J = L/3 0 0
1 1 1
2L/3 L/3 Then referring to eq . (20), Fig. A7.34
0

Note that the first column of ~ imJ gives the

values at the q's obtained ror a unit load at
point "1" with no other loads applied. The
second column gives the q's tor a unit load at = ...!c..
3AE
point "2" only, and so forth.
Finally,
and,

= ,q~ -u~q = Ai:

L J~
1..
'Z1
0 0 o 0 1 0 0
a1 j L

~
~
1 L/3 1 2L/j 0 1..
27 1..
6L
o 0 1 1 a 1 j 0

~~=3iIOl 0 1 L/ 06iL~ o 0 L/3 0 o

o 0 0 1 0 1 1
o 0 0 276" 1 1 1
An equally likely choice at generalized
o 0 0 .l. ,];1 2L L/3 0 forces for the above case is shown ~n Fig.
6L ~ 3 A7.34a. The strain energy is (x measured from
.free end)
Multiplying (see appendix),

Fig. A7.3-4a

Should the deflections be desired at the

t~xee points one torms

-- d•q/"
'U _ 1 JLx , dx
- AE =
L'
3AE
" 0

~'U = AE
= oqi)qj IJLxdx 0 = 2AE
L'

•
.Y!
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A7.23

In ~he case of tapered ~embers the

coefflc1ents are determined by evaluating in-
tegrals of the form U = 2~I gql •
and from which
o. - 1 JL!(X)dx
Ij -]" A"1Xi Fig. A7.34e
o a11 =...h...
• EI
NOTE: The coen-
Such a quadrature can always be made in these icients for bars and
problema. For the linearly tapered bar the beams are directly
results may be obtained as functions of the end analogous (compare
area ratios. ThUS, Wehle and Lansing give several cases) so
ajj
L' that for tapered
= 3EI beams one uses the
L results {or tapered
a11 =~. ¢11
1" bars with EI in place
SHEAR PANELS of AB.
L
alj = 6A E • ¢!j For the rectangular shear panel with a
1 uniform shear flow Q on all edges (Fig. A7.34t)
i
Fig. A7.34b L
0jj 3AE . ~jj
1
S = surface area
~ 3
, : Fig. A7.34f
I The trapezo idal shear
panel (Fig. A7.34g) is treated approximately by
',j\.. using the average shear flow on the non-parallel
sides as though it were constant throughout the
sheet. Thus

S = surface area
a11 =..§...
Fig. A7. 34c::::::!:d Gt
~e~ v~tes lin~rlY :J: ~
Since by statics qj =
=
- h, Fig. A7.34g
Ql -.-, one could use Q as
a. j
an alternate choice of generalized torce and
BEAm
The energy in the uniform beam 0: Fig. ajj -
_(h.)' S.
h. lit
A7.34d is given by
TORSION BAR
qj r1---====<'
1 1 q,
A uniform shaft under torque qi has strain
L energy

L
Then Fig. A7.34d Then 0.11 = GJ •

A7.11 Application of Matrix Methods to Various

a11 =~'U=_L_(
,ql' 3El = Ojj
)
Structures.
Example ?rob 1em 26
O'u =_L_ (= Ojl)
The tubular steel truss ot Fig. A7.35 is to
aQl,Qj 6EI be analyzed for vertical deflections at points ~
and F under several load conditions in which
vertical leads are to ~e applied to all joints
An alterr~te choice of generalized forces excepting A and D. The crass sectional areas at
tor the beam of ?ig. A7.34d is shown in Fig. tube member~ are given en the figure.
A7.34€l. Set up the matrix form at expression tor the

7-£
A7 24 DEFLECTIONS OF STRUCTURES

Con.tant Axial
Load; 'l.i-'l.j 'K 1 2 3 4
-=~
1 0 1.0 0 1.0
!ben
"'1.- e(i1'" ~jj'" 0 0
2 0 0
OC1j+<:o.j1
.. L!AE 3 0 -1.0 C 0

~i~
Fig. A7.35 Fig. A7.35a 4 0 1.25 0 1.25
=
deflections at points E and r. 5 0 -1.03 0 -1.03
Solution: 6 -1.0 - .75 0 - .75
The member flexibility coefficient for a 7 1.13 .848 1.13 .848
unitorm bar under constant axial load is
L/AE. Fig. A7.36a gives the ~umberlng scheme 8 -.825 -.165 -.825 -1.55
applied to the members and the q's (these being
one and the same, since q is constant in a given 9 .20 0.40 .20 0.40
member). Fig. A7.36b shows the numbering scheme
adopted for the external loading points.
Then the matrix triple product
plo Pa

~Jn]
tsl3J3 DJsL
1 2
t=] = Emi] [OiJ]

was formed giVing, pe~ eq. (26) •

8 Ip,
rs, -440 389 257
-
389
Fig. A1.36a Fig. A7.36b
o. 389 927 252 789
p'l
P.
1
=-
Member flexibility coefficients were col- E
Il, 257 252 257 252
lected in matrix form as P, J
6, 389 7Sti 252 789 P,
e..- - 1nche s
~ 1 2 3
• 5 6 7 8 i 9

1 92.2 a a a a a a a a The results here give the deflections of

2 a 92.2 0 a a a 0 a a all four points. Since only the deflections of
points 3 and 4 were desired the ~lrst two rows
3 a a 138 a a a a a a
["1J] ·i • 5
a
a
a
a
a
a
1<6
a
a
55.7
a
a
a
a
a
a
a
a
of tmnJ may be dropped aut. The same result

6 a a a a a 185 a a a
could have been achieved by leaving out the
7 a 0 0 a 0 a 155 a a
8
9
a
a
a
a
0
a
0
a
0
a
a
a
a
a
55.7
a
a
229
first two raws of @",i] 0he transpose of @bl])
The matrix fo~ of equation above ~s useful
in organiZing the computation of deflections for
In the case of a Pin jointed truss, where a number of different loading conditions. Thus
only a single generalized force is required to should there be several different sets of ex-
describe the strain energy per member, the ternal loads P corresponding to various load-
matrix or member fleXibility coeffic18nts 15 a n,
diagonal matrix as above. ing conditions, each set is placed in column
fo~ giving the loads as the rectan~~lar matriX
Unit lead distributions were obtained by
placing unit loads succeSSively at external
loading points 1, 2, 3 and 4 (Fig. A7.36b). ~nk] , k different numerical :ubscr'1pts for the
The results were collected in matrix form as
load conditions. The matrix product
 ,
tip'

A7.25

LAB~J
L'
~llJlJ
iL-':c-J<.......
."- = -"
. 'nn
P
nk
(26b) au a"
AB;SEI
a••

L;~
r1SW 21'1e5 t.ne 'deflections at each point (:r,; for
t~e ~arlous load conditions (k). a" a .. a" = L8~
3EI GJ
~a.T:J.ple ?roblem 27

= a,. = La~
Deflections at points 1, 2, 3 and 4 of the
truss of ?lg. A7.35 are Geslred for the follow- a,. BO a •• a .. = L8~
ing loading conditions: 2EI EI

Ccndf t t on No. P, P. P, P, ( see Fig. Collected in matrix form. (noting that

1 2500 2000 eco 450 A7.36bl GJ =1.3
EI )
2 -1200 -800 -2100 01750
3 1800 1470 -1200 -1100
1 2 3 4 5 , 7 8
Solution: 1 '.0 0 0 0 0 0 0 0
2 0 3,' 0 0 0 0 0 0
T~e matrix product formed per eq. (26b) was 3 0 0 '.0 0 0 0 0 0
set up as 4 0 0 0 15,500 0 0 0
5 0 0 0 see ""
36,0 0 0 0
,
440 389 257 389
389 927 252 789
.
2500 -1200
2000 - 800
1800
1470
7
0
0
0
0
0
0
0
0
0
0
36.0
0
0
40.8
sae
0

~'"'J
1 8 0 0 0 0 0 sae 0 15.500
;; 257 252 257 252 800 -2100 -1200
389 789 252 789 450 -1750 -1100

The unit load distributions ~1~ were obtained

~ample Pr"oblem 28
Tor the landing gear unit of SXanple Prob- by applying unit external applied loads, IT~bered
lem 20, Fig. A7.23 find the matrix of influence and directed as in Fig. A7.37b.
soef~lclents relating deflections due to 11ft
- -
~
and drag loads acting at point A and torque
1 2 3
about t~e axle A-B.
1 1.0 0 0
Sol:ltion:
2 0 1.0 0
st~ucture was divided into elements
7he
3 0 0 LO
~nd ~~e set of internal generalized forces ap-
~lied as shown in ~~g. A?37a. (Torques and 4 ,342 0 0
noments are shown vectorlally by R.H. rules).
5 3 0 0
Axial stresses were neglected in C-B.
6 0 -.937 1.026
7 0 .342 2.311
3 0 0 l.G
- L.. -
At this point the e~g1neer ~ay consider t~e
problem as solved, for the re~2inlng computation
is a ~outine cperation:

Fig. A7.37a Fig. A7.37b

Exa."'11ple Problem 29
~he following ~e~ber flexibility coef::cients T~e bea~ at example problem 21 is to be re-
'Nere oaterainec solved by the T~trlx wethoas ?resented ~ereln.

•
A7.26 DEFLECTIONS" OF STRUCTURES

Influence coefficients for points ?, G and H The matrix triple pr-cduct

are to be found.
Solution:
Fig. A7.38 shows the choice and numbering
of generalized forces.
completes the calculation.
/'" (caef!'s
a 33 '" a~3 -;. a ~3/
Example Problem 30
call'td
q. \ ' fr.two
-",==="'_q.. . Deflections of statically indeterminate
===='~9 ~'?....;, members)
structures often may be computed successfully
by the methods of this chapter proVided that
some auxiliary means is employed to obtain an
approximation to :he true ~nternal force dis-
tribution. The exac~ ~nternal ~orce distribu-
tion is not necessarily required in making de-
flection calculations inasmuch as such a calcu-
lation amounts to an integration over the
Fig. A7.38 structure - an operation which tends to average
out any errors. Thus one may use the er~ineeri~g
theory of bending (E.T.B.), experimental data,
previous experience, etc. to obtain reasonable
No torces were shown applied to the lower tlange esti~tes of the internal force distribution for
elements as these were known to ue equal to unit loadings.
those ot the upper tlange due to symmetry. In the folloWing probl~ ~he ~trix of in-
Entries were made for ll in matrix form as be- fluence coefficients is dete~ined tor a sin~le
iJ cell, t~Iee-bay box beam (3 times indetermi~2te)
low. Entries tor !l33 and !l88 were quadrUpled as by using the ~.T.B.
these occur in ~ identical members each on !QQ Fig. A7.39a shows an idealized doubly
and bottom. Entries tor C1e1l' au and c •• were symmetric single cell cantilever box beam having
doubled. (See Art. A7.l0 tor coefficient three bays. Determine the ~trix ot influenCe
formulae. ) coefficients tor the six point net indicated.
a a • , I , ,
1 10
a
a
aoes
17.8 .... 10"
t,~ . •, e.se L
10-

.... =0'

1
17 8
•
'-'.
....
..... 2500
8.89

®
Unit load Yalues were obtained as in Fig. Stringer Areas
A7.27, considered to be external loading num- Constant", . 33 in a 20" 1:-5"
ber "two". S1ml1ar diagrams were drawn for unit '-J.-
, 15.. -J
'
T
loads at points "one" (H) and "three" (Ft(llt'. rig.
A. 7-2"
1 2 3 Fig. A7.39a

1 1.0 0 a Solution:
2 .0667 a a
Fig. A7.39b is an exploded view of the beam
3 1.201 a a showing the placement and numbering of the in-

~~
4 a 1.0 0 ternal generalized forces. Note that only the
upper side of the beam was numbered, the lower
5 .0540 .0600 a side being identical by symmetry.
6 2.184 a Member fleXibility coefficients were com-
11.0:
7 a 1.0 puted by the formulas ot Art. A7.l0 and entered
in ~trix form as below. Note tr2t all entries
8 .0447 .0496 .0545 tor Which there Nere corresponding leads on the
9 3.00 2.00 1.00 lower surface of the beam were dOUbled. ay
this ~eans the total strain ener~J of the beam
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
A7 27

flow (q = ~) due to the torque developed in

transferring the load to one side.
The matriX triple product

completes the solution.

A7.i2 Denecncne and Angular Change ot tbe Elastic Curve

of Simple Beams by the "Method of Elastic Weights"
(Mobr's Method).
Fig. A7.39b
In the calculation of structural deflec-
was accounted ror. Note also that entries tor tions there occur many steps involving simple
a~~, a 8 8 , a•• and a 1 0 • 1 0 were re-doubled as integral properties of elementary functions.
each at these q's act on two (identical) mem- The Method of Elastic Weights (and the Area
bers. Moment Method to folloW in Art. A7.l4) owes its
popularity in large measure to the fact that it
enables the analyst to write down many of these
integral properties almost by inspectIon, relY-
ing as it does upon the analyst's familiarity
with the properties of Simple geometric figures.
For finding the deflection of a paint on a
simply supported beam relative to a lIne joining
the supports, the Method of Elastic Weights
states:
The deflectIon at point A on the elastic
curve £f ~ simple beam 1s equal-to the bending
moment at A due to the ii
diagram acting as ~
distributed beam load.
Note; VOID SPACES INDICATE ZEROS. -- --
Spelled out in steps;
Unit load distributions were obtained tor i-The ~ diagram is drawn just as it
successive applications at unit loads to points
one through six (Fig. A7.39a). The internal occurs due to the applied beam load
torces Dredicted by the E.T.E, for a load ii - This dIa~ is VisualIzed as being
thraugh- the shear center (center of beam, due to the loading on a second beam (the conjugate
symmetry) were superposed on the uniform shear beam) supported at the paints of reference for
the deflection desired
~ 1 Z 3 4 5 5 iii - The bending moment iro this conjugate
1 0.3 0.1
beam is found at the station where the deflec-
tion of the original beam was desired. This
2 0.1 -0.1 bending moment is equal to the desired
~L~ =
3 0.1 0.3 deflection.
4 Z 2
5 Z 2 To prove the theorem, consider the dummy-
6 .0875 .0l25 0.15 0.05 unit load (Virtual work) equation
7 .0375 -.0375 0.05 _0.05
a .0125 .0875 0.05 0.15
9 2.67 2.67 1.33 1.33
10 2.67 2.67 1.33 1.33 This expression equates the exte~al virtual
11 .osea .0023 .0713 .0179 0.100 .oasa work done by a unit load, applied at a point de-
12
13
.020
• 0023
1-.020
I .0423
.0267
.0179
_.0267
.:J713
.cess
.csss
-.=
.100
:lectlng an amount 6, to the internal Virtual
work on a beam element experiencing an angular
14 a Z 1 1 change d g = ~~. The sum (integral) of such
3 'I 3
15 3 3 2 Z 1 1 expressions throughout a beam gives the total
A7. :.::8 DEFLECTIONS OF STHUCTURES

deflection at tne ccmt (c r . eq , 103). ..~e now In rig. r co~sider Mex ~3 a load on a si~~ly
show tr~t the ceflecticr- expression, uSin: :he 81
above equat~on, is the SaT.9 as the bending supported bea~ a~c dete~ine the benG:n; 200ent
moment expreSSion fer ~ simple beam loaded by at pOints band c G~e ~o ~dx acting at pOl~t c.
--"- ~ Md.x "IT
an elastic weight 21 Mdx L XLdx
In Fig. A7-40, the la~cing at (a) pr ocuces
4EI :2 BEl
the real moments of (0). Consider the ce r jeo- L
tlons ot points 3 and C due to the ar.gular
ch~nge Mdx
~ in a 0 ea~ ~~emen t
",1 a tA" (Fig. A'7-4C':).
"
values of ~~e te~ bending ~o~e~ts at
~hese
points 0 and care ldentica: to :~e deflections
at band c by the virtual ~ork equa:ions. ~~e
noment diagram ~ for a unit ~oad at b a~d c
(Figs. d and e) is numerically pr sc t se Ly tne sane
B
as the influence line for illo~e~t ~t ~oi~ts b
41--, -....--
L I L
4
l.h-Lhj
-1- 4 ' 4
and c.
Therefore deflections of a si~ple bea~ can
I') be determined by cons Ider mg the :1 cur-ve as an
81
~WEI diagram ~ imaginary beam loading. ~he bending ~oment at
any pOint due to this !1 load.ing equals the 1e-
.., !"odx
(h) EI
flection of the beam ili~der the gi7en loads.
Mdx
EIpj Likewise it is easily pr)ved :~zt :he ~n~~­
lar change at any section of a sl~ply supported
(c) beam is equal to the shear at that section due to
~ !'! dt azr-am acting as a beam Lcac .
I Sol
-= L~4 ~ A1.13 Example Problems
(d) mb, moment diagram for unit
load acting downward at point B Examnle Problem 31. ~ind :he ~2rt~cal
3 deflection and slope of paints?. and b :or be~
r16 L and loading sho'Nn in Fig. A7.41. T~e lower F~g.
shows the :noment c Iagr-am for load? acting at
f ----
center at a sl~ple beam.
(el me" unit load at point C

r-~1 IP
~-;P
Mdx
lEI
EI Constant
If) a b
3 Mdx
4EI
Fig. A7.40
lMdx
4EI
P f L
2
2
For a unit load at point b, Fig. d shows the m
PL
T Fig. A7.11
diagram. The value of m at the midpOint of dx
(point a) = LIS. Hence -6"-
L L
2
°b = ~~ . ~ ~~ PL'
16 PL'
16
For deflection ot paint c, ,1raw m diagram ror a Deflection at point a 8Guals bending ~cm­
unit load at C (see Fig. e). Value or ~ on ele- ent due to M diagram as a load di vtoec by EI.
aerrt dx = L (See lower Fig. at Fig. A7.~1)
16
PL2 L PL2 L) 1 11 PL3
Hence 6a ( 16 4. - 64 12 EI = 768 EI
00
Db = (~2 ~ i~2 ~ )Elr = 1a ~~3
.. For srmoncrty the points A, Band C were placed at the one-
quarter span points. The reader may satisfy himself with The angUlar c~nge ot any point equals the
the general character of the proof by substituting xA, xB and shear due to ~JEI diagram as a load.
XC for the point locations and then following through the
argument once again. = (PL2 _ PL2)2:.. = 3 PL2
Ua 16 64 E1 64 EI
 -, 7

A7.29

o (Slope is horizont- I<i Airplane

al cr- no change
from origi~al di-
1 50 " - ~:)O"130"

rection of beam
axis. ) rHear Beam
I Anercn E FIA
A B C
?rcbl,:::n ·32. De t cr-ntne ':;;'a cer Iec t Lcn
~x":.:nJ'"'!
,
of a sixple be~~ loaded unifo~y as shown in ..-- Ercnt Bean-
Fig. A7.42. The bending moment exor ess i on for /
a unlforn loac M = wlx - ,NX2 or ~arabolic as ,
2" 2" Plan View I: 2 Wmg
shown in F'1;. A7.42a. The deflection at mfd-
?oint equals the Jending J10ment due to M dia-
gram as a load.
r 0"1
.,.....Air Load
IiffffHHHflH fJ HJ f OHm!! H Hbj"
25li'iin.
3

i- Load on Rear Beam )

15~/in.
wz in.

r' lll! 'LPl ! ! If Y+ Fig. A7.43

Solution:- Due to the beam variable moment ot in-

wL wL
2"" T ertia the beam length between A and C will be
divided into 10 equal strips of 10 inches each.
Fig. A7.42 The bending moment M at the midpoint at each will
be calculated. The elastic weight for each strip
will equal Mds, where ds = lOW and I the moment
.-?'7A-.<(..M Diagram T
of inertia at midpoint of the strip. These elas-
tic loads are then considered at loads on an 1~­
aginary beam of length AC and simply supported at
A and C. ~he bending moment on this Dnagi~ary
....!....wL~""Area 1 J3am at ~o:nt B ~ill equ~l the deflection of 8
24 24 wL"
with r3spect to line jOining AC.
Fig. A7.42a The bending moment at C = 15 x 30 x 15 ?
10 x 15 x 10 = 8250"#
The shear load at C = (15 + 25) x 30 =600#
6center =(2~ NL3 . ~ - d4 wL3 . ~ ~) SlI -2-
Bending moment expression be~#een points C
c wL4 and A equals, :M: = 8250 + 60Qx + 12.5x2 , where
384 sr x = 0 to 100.
Table A7.6 gives the detailed calculations
ccent.er =(2~ 14
'NL3 - 2 WL 3) Ell = 0 tor the strip elastic loads. The I values as-
sumed are typical values tor a aluminum alloy
1 wL3 beam carrying the given load. The modulus ot
Slope at supports = the reaction
24 EI elasticity E = 10 x 106 1s constant and thus can
SXP~;le P~QbleJ1 33. be omitted until the final calculations. The
Fig. A7.43 shews the plan ~iew of one-half figure below the table shows the elastic loads
of a cantilever wing. The aileron is supported on the imaginary beam.
on brackets at ?oints D, E and F with self-align- Table A7.6
ing bearings. The brackets are attached to the
wing rear beam at ?oints A, B,and C. ~hen the
Wing bends under the air load the aileron must
likewise bend since it is connected to wing at
Strip
'0.
d.
'0.
e ,
lIo_nt
at mid_
point
I
at JlIid-
point ...
Elastic
load

rt:
I IO US63 5.S 21000
t~~ee ?oints. In the desi~ of the aileron beam 2 IO 20063 6.5 30900
and similarly for cases of wihg flaps this de- J IO 31063 7.S 41450
flection produces critical bending ~oments. As- •
·5
IO
IO
44580
60580
'.5
9.5
52300
63850
suming that the running load distriblted to the 6 IO 79050 12.0 6.5940
rear beam as the wing bends as a uni~ is as shown ,s
7 10
10
1000.50
123$50
16.0
20.0
62530
61770
i~ the Fig., tind the deflection of pOint 3 with 10 149550 24.0 62320
10 10 118150 28.0 63600
respect to straight line joining ~oints A and C,
0 0 0 0 0 0 0 0 0
which will be the deflection of £ with respect to 0 OJ e- eo ~ en 0 en 0
line jOining D and F it bracket deflection is co eo c- 0> co eo ~ 0
ro
.o
OJ ~
so
"':'J
-o 0'_0
eo OJ
.c l.D
~ ~
OJ
neglected. 7he moment or inertia of the rear
beam between A and C 'varies as indicated in
'" '"
:he Table A7.6
A7.30
Bending ~oment at paint B due to above elastic
loading = 7,100,000 .'. de~lection at B ~elatlve
to line AC = 7,100,000 = .71 inch
E::IO,OOO,COO
Example Problem 34
Fig. A7.43a shows a section of a cantilev-
er wing sea plane. The wing beams are attached
to the hull at paints A and B. Due to wing
loads the wing will deflect vertically relative
to attachment points AB. Thus installations
DEFLECTIONS OF STRUCTURES
A7.14 Deflections of Beams by Moment Area Method·
For certain types of beam problems the oeth-
ad of moment areas has advantages and this ~eth­
ad is frequently used in routine analysis.
Angular Change PrinCiple. Fig. A7.44 shows a
cantilever beam. Let it be required to deter-
~ine the angular change of the elastic line be-
~eent~pointsA~8~to~givenlMdl~.
From the equation of virtual work, we have

.,
such as piping, controls, etc., ~ust be so lo-
cated as not to interfere with the wing deflec-
tions between A and B. For illustrative pur-
poses a simplified loading r~ been assumed
as shown in the figure. EI has been assumed as
constant whereas the practical case would in-
volve variable I. For the given loading deter-
~ine the deflection of point C with respect to
the support pOints A and B. Also determine the
vertical deflection of the tip ppints D and E.
0
0
::.
0
0
0

::,
0 0
'\ ,

\
N 0
~r400"
~
300'~1w
, 40" I Fig. A7.43a
D A 0 B EI Constant E

a
Moment Diagram
o!
o
in.'
1000

""
0,
0,
, Fig. A7.43b
\ c

.,
0·1
,
Fig. A7.43c

!
Fig. A7.43d

Solution:- Fig. A7.43b shows the bending moment

diagram for the given wing loading. To tind the
deflection of C normal to line joining AS we
treat the moment diagram as a load on a 1lllaglnary
beam of length AB and simply supported at A and B
(See Fig. A7.43c.) The deflection ot C is equal
numerically to the bending moment on this tlcti-
c i ous beam,
Hence EIO c = 25920 x 40 - 25920 x 20
or 6c 518000
EI
To tind the tip detlection. we place the elastic
Fig. A7.44
loads (area or moment; df agram ) on an imaginary
beam Simply supported at the tip D and E (See
Fig. A7.43d). The bendlngmoment on this imagi-
nary beam. at points A or B will equal numerically Clb = (A ~:n Where en Is the moment at any
the deflection of these points With respect to )B EI ' sect i cn, distant x trom 3 due
the tip points D and E and since points A and B to unit hypothetical couole
actually do not move this deflection will be the applied at B. But ~ = ~lty
movement of the tip paints with respect to the at all points between Band A.
beam support points. Therefore as =(A ~
Bending moment at A = 193420 x 700 - 40000 x )B EI
433 - 127500 x 124 = 102200000. .'. Ot1p (·Moment area first developed by Prof. C. E. Greene and
102,200,000 published in 1874. )
EI
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A7.31

Referring to Fig. A7.44, this ex~ressian repre-

sents the area ot the M diagram between points Henceos = (_ PL2 § L) ~
03
= _ PL3
EI Z EI 3EI
8 and A. Thus the first principle:- "The Example Problem 36
:hange 1rr slope of the elastic line of a beam Fig. A7.46 illustrates the same simplified
between any ~wo ~o1nts A and 8 is nw~erically wing and loadiP~ as used in example prOblem 34.
equal to the area of the :1 diagram betw"een Find the deflection of paint C nor~l to line
EI joining the support points A and B. Also find
these two pol~ts.n the deflection of the tip paints D and E relative
Deflect~on Principle to support pOints A and S.
In Fig. A7.44 dete~ine the deflection of
paint B normal to tangent of elastic curve at A. g
In Fig. A7.44 this deflection would be vertical
since tangent to elastic line at A is horizontal. 1 T
From virtual work expression Os = I~ ~~ ~, o ~B EI Constant E
Fig. A7. 46
where m 1s the moment at any section A dist~~ce = Moment
x from 8 due to a unit hypothetical vertical
load acting at B. Hence m = l.x = x :or any
oo
~!,ir!'J;
h ~
= ,
o Diagram
o in.~
TijijQ
paint between Band A.
I~67"-'" ~ 21 ifi:!it:I
0 11'0 0Oi

° fA
~ a;:
~i ~ ,~o ~4
N
Hence Max
r- 576 " -
8 = B This
-EI x r-- 720" ...
eXDression ~epresants the 1st noment of the
M uiagracaoout a vertical t.hru B. 'fhus the Solution:-
~I Due to symmetry ot loading, the tangent to
deflection principle ot the moment area method the deflected elastic l1na at the center line of
can be stated as follows:- "The deflection of a airplane is horizontal. Therefore, we will find
point A on the elastic line at a beam in bending the deflection of points A or B away tram the
normal to the tangent of ~he elastic line at a horizontal tangent ot the deflected beam at pOint
point 3 is equal numerically to the statical mO- C ~hich is equivalent to vertical deflection of
ment or the !1 area between points "An and "B" C with respect to line AB.
. EI Thus to find vertical deflection or A with
about point A". respect to horizontal tangent at C take moments
of the M diagram as a load between points A and
Illustrative Problems
EI
ExaT.ple Problem 35. Determine the Slope C about point A.
and vertical deflection at the rree end B of the
cantilever beam shown in Fig. A7.45. £1 1s con- Whence
stant. (area) (arm)
1~~0 (tangent at C) =E~ (650 ~ 646) 40 x ZO =

1--- L ------i
1
EI (518400) = deflection of C normal to AB.
Fig. A7.45 To tind the vertical deflection of the tip point
~~am D with respect to line AB. tirst rind deflection
of D with ~espect to horizontal tangent at C and
~ PL2 __ .!L--J subtract deflection of a ~ith respect to tangent
2' 3 at C.
Solutlon:- The ~oment diagram ror given load is
triangular as shown in rig. A7.45. Since the ~ (respect to tangent at C) = E; (40000 x 267 +
beam is fixed at A, the elastic line at A is
horizontal or slope is zero. Therefore true 1Z7500 x
576 • 25920 7Z0) = x
(10Z700, 000) ir
slope at 8 equals ar.gular c~~ge becNeen A and 8
(See Fig. A7.46 for areas and arms or M/EI dia-
wm cn equals area of 'nomerrt diagram between A gram). subtracting the deflection of A with ~e­
and B diVided by E1.
spect to C as found above we obtain
Hence 1 PL Z
"8 = (-PLo LIZ) £I - 2EI l~gO (respect to Ltne AS) = Ell (102,700,000
The vertical deflection at B is equal to the 1st
518400) = SI (102,180,000)
:no~ent of the :noment diag~am about ~oint S di-
Vided by S1, s~nce tangent to elastic curve at A
1s norlzontal aue to ~ixed support.
A7.32 DEFLECTIONS OF STRUCTURES

A7.15 Beam Fixed End Moments by ~e~hod of Area :1A =- Pab 2 and MB =- ?ta 2 / L " l ' o ' h e r e b:: (L-a)
lIOllIents L'"""
From the rNQ prinCiples of area ~o~ent5 ~s To find the fixed end ~owents ~or a te~~ ""itt
given in Art. A7.14, it is evident that the de- variable moment of Iner-t-Ia use the t""iI crucrana
flection and 510;e of the elastic curve depend in place of :he ~oment clagraos.
on the amount of bending ~oment area and its lo-
2xa~ple Problem 37
cation or its center of graVity. Fig. A7.48 shows a :1x-ended je~~ carry~ng
Fig. A7.47 shows a beam r1xed at the ends
two concentrated loads. finS the f~xed-end
and car~Jlng a single load P as shown. T~e oend
~ng moment shown in (c) can be considered as
moments MA and roB'
made up or two parts, namely that for a load P 200*
acting on a simply supported beam wmcn gives j 100. I t:',
the triangular diagram Hlth value Pa (L-a)/L
tor the moment at the load ~otnt, and secondly
a trapezoidal moment diagra~ of negative sign
4t=
M9"......:.-....12"~9" +.-
__ I ~tB ; (a)

with values of MA and ME and of such magnitude

as to make the slope of the beam elastic curle
zero or horizontal at the support paints A and Ih)
B, since the beam is considered fixed at A and B.
The end moments MA and ~B are statically
indeterminate, however, with the use of the ~NO Fig. A7.48
moment area principles they are easily determin-
ed. In Fig. b the slope of elastic curve at. A 15MA
and a is zero or horizontal, thus the change in -l1-lO'~
slope becNeen A and B is zero. By the 1st ~L i
(e)
,

Solutlon;- Fl~. b shows the static ~o~ent dla~ao

A ass~i~g the beam simplY supported a~ A and 3.
~-----~B (b]
~lastic Curve
For SImpliCity in finding areas and tak~r.g moments
of the moment areas the :noment Qia~~ ~~s been
Fig.A7.47
divided into the 4 simple shapes as ShO','iC. "he
--, ~a(t-a) lEI'' ' (c}
centroid of each portion is shown together ~ith
the area which is shovm as a concent~ated load at
~~- the centroids.
Fig. C shows the moment diagrams due to .~~­
T known ~oments MA and Mg. ~he area of these tri-
anglas is shown as a concentrated 10aG at t~e
principle of area moments, this ~ans that the
algebraic sum of the moment areas be~Neen A and centroids.
8 equal zero. Hence in Fig. c Since the change in slope of the elastic
curve bevNeen A and B 1s zero, the area ot these
(-l1A -!'lbl L (Fa (L - a) moment diagrams must equal zer~, hence
-2- + L .2L 0 - - - - -
_ -(AI
5265 + 14040 + 2160 + 6885 + lSMA + 15MB:: 0
In Fi~. b the detlection or B away tram a tang- or
ent to elastic curve at A is zero, and also de- 15MA + 15MB + 28350 = 0 - - - - - - - - - - (1 )
tlectlon ot A away fram tangent to elastIc curve ~he deflection of point A away :Tom tangent to
at B is zero. elastic curve at B is zero, therefore the ~irst
ThUS by moment area prinCiple, the ~oment moment of the moment diagrams about point A
of the moment diagrams of Fig. C about points A, equals zero. Hence,
or B is zero. 5265 x 6 + 15 x 14040 + 17 x 2160 + 24 x 6855 + l5CMA +
30G Me = 0 or 150 MA + 300 ME + 444600 ,. 0- - - -(2)
Taking moments about point A:-
solving equations (1) and (2), »e obtain
MA = - 816 in. lb.
ME = - 1074 in. lbs.
With the end moments known, the ~etlec~ion or
~ L: 0 - - - - - - - - - - - - - - - (E) slope of any ~oint on :he elast~c curve between
2 3 A and B can be found by use of the 2 princ~ples
of area :n.oments.
solVIng equations A and B for MA and ME
 A7 33

A7.l6 Truss Deflection by Method of

Weights
Xla~tic

Ir the deflect~on of several or sll the

c~~-- ,yf1J\
1M
~;.-'- . ;Jl90
,- -
,'l.0\c
I
., I
d

I
(I)
jOints of a trussed structure are required, the
~ ,/1'1\. I
g
method of elastic ~elghts ~y save considerable : d <ltokih
::~e over the method of virtual work used in r-rr : 6 Panels @ p
previous articles of this chapter. !he illethod Bar be (2)
in general consists of finding the illagnitude and
location of the elastic ve i gnt for each member u6L _ 4P6~L.L Shortening
~lection Diagram

of a truss due to a strain from a given truss "F" - ---sr-- .: . Lbc of Member be
loading or condition and applying these elastic o Elastic Load on Imaginary Beam
weights as concentrated loads on an imaginary I.!..
3r
uLb
c
t .::.Lbc
3r (3)
beam. The bending aoment on this imaginary beam
due to this elastic loading equals numerically ~~entCurve (4)
the deflection of the given truss structure. for Elastic Loads
Consider the tr~ss of dia;ram (1) of Fig.
A7.49. Diagram (2) shows the deflection curve Bar ck 6L(d+6p)
for the truss for a ~L shortening at member be,
all ether members considered rigid. This de- /?\~ (5)
flection diagram can be determined by the vir-
tual work ex pression 6 = u~L. Thus for deflec-
~ection Curve
for !J. L Lengthening
tion of jOint O. apply a uhit vertical load act- 6L • djip t
SL, 1M. d+2p of Member ck
ing down at joint O. The stress m in bar hc due f. rl, L"'[':'" "'--Ela~tiC Loads on
to this unit load = 2 2P = 4P. Therefore Rigid Arm t a K .::. Ld Imaginary (6)
:3 r- 3r t:. L! d+6W ~ Beam
6pr~ l:.L(d+6p)
°0
=
3r
4P The deflection at other
lower chord jOints could ~iiiz:
be round in a si~ilar
~anner by placing a unit load at these joints.
Diagram (2) shows the resu~ting deflection curve. Fig. A7.49
7hls diagram is ~lalnly the influence line for
stress in bar be multiplied by ~Lbc' Table A 7.7 Equations for Elastic Weights
Diagram (3) shows an imaglnarJ heam loaded Elastic- Weight for Chord Members
~ith an elastic load 6Lbc acting along a verti- (See Member ab)
r Upper Chord Lower Chord
cal line thru joint O. the moment center for ob- w=~L
taining the stress in bar be. The beam reac- h
tions ~cr this elastic loading are also given.
Diagram (4) shows the bea:n bending moment dia-
gram due to the elastic load at point O. I~ is &1:0""i
tab t
noticed that this moment diagram is identical to
the deflection diagram for the truss as shown in elL
w=-
dt agr-am (2). I r
a b

~\t
The elastic we Lght of a member' is therefore
equal to the =ember defo~ation diVided by the
a~ r to its ~oment center. If this elastic
load is applied to an i~gi~ary beam correspond-
/y\/\r
f 0 j,L tab
w=-
ing to the truss ~ower chord, the bendi~g ::noment r
on this Imagfnar'y beam. ,/.111 equal to the true
tr~ss deflection. b

i/NJ\I~
Diagram 5. 6 and 7 of Fig. A7.49 gives a
si~ilar study and the resul~s for a ~ lengthen-
lng of ::nember CK. The stress moment center :or
.~~'
I aL
this ciagonal ~ember lies at point O~ which 11es ~­
,
outs1~e the truss. The elastic Neight ~ at w= Member Axial Deformation J.L
Perpendicular Arm to Moment Center r
rl
point 0 1 can be replaced by an equivalent system The moment center 0 of a chord member is the
at points 0 and ~ on the imaginary beam as shown intersection of the other t ..... o member-s cut by the
in Diagram (6). These elastic loads prOduce a section used in determining the load in that memb-
bending ~oment diagr~ (Diagram 7) identical to er by the method of moments
the deflection dia~am of diagram (5). The Signor the elastic weight w for a. chord
Table A7.7 gives a s~~J of the equations member is pj.ua if it tends to produce downwardde-
for the elastic .vefght.s of truss chord. and 'Neb r Iec t t on of its point of application. Thus for a
~embers together Nith their location and sign.
simple truss compression in top chord or tension
in botton chord pr-oduces downward or positive
elastic weight
 • tt rmn=rry

A 7 .34 DEFLECTIONS OF STRUCTURES

1000 1000 1000 1000

Table A7. 7 WEB DIAGONAL MEMBERS ~(.5630 1.::1-7500 if"' ~5630;D
(continued) (See Member ab}
/ '\
'~~t.:~0·$"¢
/3120~ ~
6870 \ 6870 \/ 3120 ~ --.1...
~ W"~:i;P.

/<'_:'~
- ~/"</\~
I N
a
~( bI CI d~ ...<iaI>...e
wO 6L 000 '2000 2000 '2000 i5000
r ...... __ - ~ ~
1,_-
k:-::.._____ ,i'''l,e, b 4@30" = 120"
,
Id
I a

p. bw=~=2.L
b
Q
FIg. A I
this step are given in the figure, the stresses
Fig. A7.50

a bJ,Sln 9 r:, being w rl tten ad~acent to each member. The next

a AL "L step or steps is to compute the 3ember elastic
Q=-a w =~=;'1
weights, their location and their sense or dlr-
r. p ection. Tables A7.8 and A7.9 gives these ca.Lcu-

~
lations. Table ..1.7:8
I\&i\l\
t \Q i fQ r
Chord Member E1as~ic Weigbts

6L Kellber Length Area L~' ~ • ~L IA.rJI i Elastic Point of

P=Q=T P • Q • 6L
r
L A
• ~:29z106 I r l , ~·~L
,
allplica-
tiol1
, 1 I joint

For a truas diagonal member the elastic AB ,I ao .2172 _5630 _.0268 24,I·Q011171 , I
weights P & Q have opposite signs and are as-
~i,
so
ao
.2172 _7500
.2172 _5630
_.0357
-.0256 I 24 .0014871
24 .001117 ,c I

sumed to be directed toward each other or ee , ,.

ac .U98 3120
1·~~98 ".
.0259
.0592
,
24 .001117 A
•
I ""
.OO~465

away according as the member Is in compres-

sian or tension. In t1g. a, P is greater
,. I
cd ac
ac
.1198 6870
. 1198 3120
.. 0592
.0269
.002465
24 i .001117
C
c
,
than Q and P is located at the end at the di-
agonal nearest the moment center O. Downward
elastic weights are plus.
Table A7.9 .eb Mesber Xl.ati~ lfe~ght5.
TABLE J,.7. 7 (CONTINUED)

<
Truss Verticals
(See Member ab)
· · . ,I
•" •
• •• w.d
•
'L •
~ I
.-
il
~I
~!
.,
I
I Q -
II
~
,
:1 " II "
. •• e,

~
'1 '1
•• •"
e,
} I •
,.
A'
28.25 .2421_5880 ._.0238112.75
.146 47101.0314 ....
_.00185
_.00246
a :12.75 . 0018~
:1 ""
.00246
A
,
P,
d
I
~L Bo
"
"
.1461_~·3!>51_.01571
.093 11771 .0123
..
"
_.00123
_.00095.5
•, "
.00123
.000955
•,
PoQ'"
,c " .09311177 .0123 " _.00096.5 C " .00096.5 c
c, " .146 _23.55 _.01.57 " _.00123 " .00123 C

<, " a - -on J

Cd
De
"
"
.146. 4710 .0314
.242 _.5380 _.0236
"
"
_.00246
_.0018.5 •, " .00246
" .00185
d
n

Ai~l~ Fig. A7.51 shows the elastic weights obtained

tram Tables 8 and 9 applied to an tmagtna ry beam
I br d '""? r whose span equals that at the given truss.
P Q
p.'" OL Q=8L These elastic weights are the algebraic sum ot
ho d d the elastic weights acting at each truss jorrrt •
The elastic weight P acts at toot at verti- ~ c-: c- c- r-.
co
cal and downward it vertical Is in tension.
Q. acta opposite to P at tar end at chord memb- "'co0
~

0
"13
0
0
;i
g g
0
r--
~

:3
0
c-
g
;i
OJ
0
0
130 co
~

0
er cut by index section 1-1 used in tlnding 0 0 0 0 0 Fig. A7.5l
stress in ab by method at sections.
1 1 1 1 J' I I'
a
A b c C d D- e
B
.17.17 Solution of Z%aaple PrOble... ~ 8 spaces at 15' _
The method at elastic weights as applied to Ra = .005412 Re = .005412
truss detlection can be best explained by the
solution at several Simple typical trusses. The detlection at any joint equals the bending
Example Problem 38 moment on the imaginary beam of Fig. A7.51.
Fig. A7.50 shows a simply supported t~JSs Den. at J,. = (.005412 + .00185)15 = .007252 x 15 =
symmetrically loaded. Since the axial detorma- .109"
tiona in all the members must be tound, the Den. at b = .109 + (.007262 - .000507)15· .109 +
tirst step is to tind the loads in all the memb-
ars due to the given loading. The results ot Den. at B .
.006755 x 15 = .209"
.209 + (.006755 - .002347)15 = .209 +
.004408 x 15 . .275
,
 ANAT·YSIS AND DESIr.N OF Fl.Tr.HT VEHICLE STRUCTURES A7.35

nen. at c = .275+ (.004408- .0027)15= .301" beam loaded with the elastic weights from Table
The slope of the elastic curve at the truss join A7.1l. Table A7.12 gives the calculation for the
points equals the vertical shear at these points jOint deflections.
for the beam of Fig. A7.5l.
Example Problem 39
Find the vertical deflection of the joints
of the Pratt truss as shown in Fig. A7.52. The
member detormationsAL for each member due to the
given loading are written adjacent to each memb-
er. Table A7.l0 gives the calculation of member
elastic weights. Fig. A7.53 shows the imaginary
beam loaded with the elastic weights from Table Fig. A7.54
A7.l0. The detlections are equal numerically to
the bending moments on ..t his beam. Tabb A7.11

0b = .01855 x 25 = .465"
0B = .465 + .053 (~in Bar Bb) = .518" Ilellber 6 L , ... -.ll
, Joint
0c = .01855 x 50 - .00387 x 25 - .833"
....
••
0c = .833 + .031 (~ of ecl = .864 .O82~ 15.0 .00550
BC -.056 17.17 .00320
On = .01855 x 75 - .00387 x 50 - .00623 x

..
U .096 20.0 .00.480 C
I
25 = 1.03"
b ·.083 c •. 091 d -.091 e .083 !
CD
.n
18
_.069
••06
-.0738
.113
21.1-6

~i:i~
20.0
....."
.00322

.00348
.00565
D
J

1
a
•r
~~I~o ~I~~I
BE _.015 17.17 .00437
. '" ~ "0> '=, ~o Q ~e:::,' ~ ,~.,,~lI' g BG .108 15.00 .00120

A
'T
.061
r.-l BCD
.061 .066

6@ 25" • 1S0"
.066'
E
.061
F
.061
G
lfeiRDer 6L
EIa.tie Weight ot Web

'1 P -A!- JOin

»eaber~

" Q -ll Joint

'1 "
AD -.0432 9.60 _.00450 A 9.60 .00450 B
Fig. A7.52 BE
tc
CJ
.072 9.60
_.063:1 9.27
.085 10.60
_.00748
_.00685
_.00802
••
C
11.20 .00643
10.60 .00600
11.26 .007.55
•
C
J

.. .... ••
Table A7.10
JD .oraa I ~0.33 .00134 J ~l?~3 _. ~0126
ID .0496 10.33 .00480 1 10.93 -.00454
El 10.80 _.000115 I 11.26 .00891 1
Elzstic Weight Chord Ke~erB

r•
_.0254 9.27
.?~!,5 '.60
_.00274
_. ?~?28 •r 10.60 ,00240
1~.20 .00878
I
•r
lteaber H r
,
... =il Joint G7 _.0585 9.60 _.00610 G 9.60 .00610

AD • 061 3• .00203 b
ac • 061 3• .00203 b
CD
be
• 066
-.063
3•
30
I .0022
.00277
c
C
CO _.091 3. I • 00304
•
.lzstic Weigbt ot Web Me.bers
,
I p Q
I

lfellber
I HI, '1 p .. 6L
"l
Joint r2 Q .~

"
JOint
Fig. A7.55

i Ab
.b .121119.2
.053 dJ
-.00648 19.2 .00648 b Table A7.12

.
00
I be I .128 19.2 -.00667 b 19.2 .00667 C
eC
I-'O~~ i 25
_.00124 I
i
C 25 '
:gg~f~ • hllel PaDel Sher.r Z Shear
.IIloael1t
Shear " de-
12.5 z
Point
CD
ad
, .08g 119.2 _.00417 0 19.2
° ! ileet10D

o , o A
= =
=
~
~ M
N
=
0
~
=
M
N
=
~
= =
~

=
AD
,
• 02158
• .02 58
•• .508
••
.
M M
0 0 0 0 0 0 6
q q 0 q .056 2 .710
ql q CJ .01350
• .8. J
.
~
I I

.
I I I .07181 .900 D

~r
qt1;G
B C D E F IB -. .07429 .930 I
N
~

q Fig. A7.53 -
N

q YG
-.
-.
_.02 04
~
.638
,
• error
Example Problem 40
Find the vertical jOint deflections ror the Example Problem 41
unsymmetrically loaded truss of Fig. A7.54. The
Fig. A7.56 shows a simply supported truss
~ deformations ror all ~embers are given on the
Figure. Table A7.1l gives the calculation of with cantilever overhang on each ena.' This sim-
the elastic weights, their signs a nd points of plified truss is representative of a cantilever
wtng beam the ruseaage attachment points being
application. Fig. A7.55 shows the imaginary
 A7.36 DEFLECTIONS OF STRUCTURES

at e and .'.
The ilL deformation In each truss
memoer- due to the given external loading Is giv-
lever over-hang portion of the truss r eIat t ve to
support points e and e l •
en on the figure. The complete truss elastic Since the cantilever :Jor1:iOn is not fixed 2. t
loading 'Nill be determined. vt th the elastic e since the r-es tr-a i rrt Is ceter.nined by the truss
loading known the truss deflections from various between e' and e, this fact must be taken into
reference lines are readily determined. account In loading the cantilever portion. The
reactions on the beam of Fig. A7.S8 represent
< the Slope at e due to the elastic Ioad Ing be twee
A' .M""·.""", ·_~~...n'.M...., .o.a E ·.GSA ;)·DS6C·D6Q8·1170A

n"i
~ " tf},(~
. '01n74.~~ .. ~".
:;.. 'if' '" \~"":"''' ",- <at. ~i;~,'
".
,;~" 0.c:.~
~,("p
\072 '(', . .072 "(,"::,U74\ .U70\"~74\ .:l..-
"",; e and e' . This elastic reaction in acting in the
reverse direction Is therefore applied as a ::"oad
r.---:- ~, d' .e' [ ,,4 c ~~
Q

to the imaginary beam between e and a as shown 1

.~ • ~ . .. . •s . .s •g •~ ..
. . ::
4 Q 20" , - + - 2 i! 30" - - - 4 @ 20" Fig. A7.56 Fig. A7.59.
•; .-
g, g
, , ~, , ," ", ,, ,I , I 3,, ~,,
0_
g 0
~
g g
, ,
N N
• , , , ,
. .
0 0
, ~.Oj 0
,
~ ~, ~

, ,
"
N
0 01 N
0
a' A' b' 8' c' C- D' e' A a ~.
"
Elutic W"igj'lt Loadinl
0
Fig. A7.57
e'
0'
, ~.
,
e
I

Table A7 .13 I E' I

,
E ,
Ilastic Welgbta of Chord »embers R:.OO:l12 R", .00312
IIember dL , • =& Apply at Fig. A7.58
I, r I Joint
In finding deflections this overhang elas-
••
AD
.oao
_.070
ac
20
.0040 A
tically loaded portion 15 considered as f1xed at
.0035
•
be
Be
.074
-.060 to
.0037
.0030 I, B0 20 a and tree at e. The bending moment at any po in t
od .070 20 .0035 C on this beam equals the magnitude of the 'ler~ica1
-.~~6 20 .0028 , detlection at tnat point.
~
d

..
.074 20 .0037 0
IlK _.056 20 .0028 0 Thus to tind the cer l ect i on of the truss er
of .012
-.0""8
20
20
.0036
.0024
E (Joint
I ,
a) we tind the bending moment at pOint a
of the imaginary beam or ?lg. A7.59.

..... Ilastic Weights of Web ~ • .oera

, dL '1 P ~ I apply '2 Q .<\1, apply wise pos 3
Hence
deflection at a ; Z!1a (calling counter:loc'K-
i t i ve ) , ; ( .01922 - .00312 ) 80 + .00248 X

....
rl at
, joint
'2
"
Jo1ut 70 + .00333 X 60 .. . 00239 X 50 + .00468 X 40 +

,. -.031 8.95 -.003406 I

• 8.95 .00346 .00234 X 30 + .00543 X 20 - .0014-9 X 10 ; 2.13"
••
Be
.080
_.063
.075
8.95
8.95 _.00702 I
8.95 ~.00838
A

B
8.95
8.95
8.95
-,00":
,
• 00895
.00702
.00838
B
e
upward.
•
the truss
Due
we
to symmetry of 'truss and leading o•
know the slope of the elastic curve
'C _.060 8.95 _.00670
, o;~ 8.95 -.00781 ! C
c 8.95
8.95
,00670
• O~L81
C
,
at the center line ot the truss Is horizontal 0 r
zero. Thus to find the deflection of any potnt
dO -.0605
o.! f .02.
8.905
-. 00 72 8 1 d 8.95 .00728 0
I With reference to Joint f we can make use ot the
"" .076
.09' 12.0
8.95 ~. 008.50
.00792 I 0
D 8.95
12.0
.00850
-.00792
0
II deflection •,
12.0 .0020 12.0
• principle of the moment area method.
~.0020

Thus In Fig. A7.60 the vertical ce ri ect i on of a.

paint for example joint a, relative to Joint t
Table A7.l3 gives the calculations tor the nag- equals the moment of all elastic loads bet-....een
nltude of the member elastic weights. The signs a- and t: about a.
and also the jOint locations for locating the
= =
- "',
N ~ ~
elastic loads are also given. Combining alge- ~ ~ ~ ~

~I
~ ~ ~ ~ ~ ~ ~
N ~
N N ~ ~ ~ ~
braically the elastic weights tor each jOint t:ro~
~I
0 0 0 0 0 0 0
Table A7.13 the beam elastic loading as shown In
Fig. A7.57 Is obtained. Freete 0
~I
d
0,
"

C
,
0

c
0
'I " 9 , o.
B b A
I, Fixed
ar
.
Let it first be required to determtne the R", .00312
vertical deflection at Joint r relative to the (From Span ee') Fig. A7.59
truSS support pOints at e and e ' .
To detern,ine the deflections of the truss l:l1a = 2.077" (student should make calcula-
between the supports e a d e' It Is only neces- t i ons ) • Previously the deflection of t with r-e
sary to consider the elastic weight loading be- spact to e was found to be -.0586" . Thus de-
tween these paints. Fig. A7.58 shows the portion flection of ~ with respect to paint e ; 2.077 +
of the 1Icaginary beam of Fig. A7.57 be'tNeen these .0586 ; 2.135" which checks value found above.
paints. The deflection at f relative to line ee' Let It be required to f'1nd the deflection of
.
Is equal to the bending ~oment at , for the por- joint c relative to a j rne connec t Ing joints b
tion of the imaginary beam be~Neen joints e and and d. N
e' and simply supported at these points. "N 0=N 0 0N =0 N0 0 0 0 ~ ~ ~ ~ ~ ~

~I~,
~ ~ ~ ~ ~ ~ ~ ~ ~
N

~I
~ ~
Hence deflection at t ; - .00312 x 30 + 0
0
~ ~

.00232 X 15 ; .0586" (upward since >n.inus Is up) • 'I ,

0
" OJ, 0 0 " , ~ 0 , ~ 0 " ,
i I !
Find the vertical deflection of the canti- l E e 0 d C 0 B b A a

!<t Fig. A7.50

 ________...;, ••••.•'-'.··-:~.-:~li--~~.~;.,r:.-~~F.,~:i!!::_

ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A7.37

For this problem we need only t~ consider (7) For the truss in Fig. A7.66 calculate
the elastic loads betN8en pOints band d as the deflection of jOint C along the direction CEo
loads on a simple beam supported at band d E = 30.000,000 psi.
(See Fig. A7.61) ~he deflection at C with re-
spect to a line bd of the deflected truss
equals the bending moment at point c for the C --.-20,000*
loaded beam of Fig. A7.61. S'
hence
0c = .004743 x 20 - .00239 x 10 =,07 inches B 0 ~
8' 2 S'
A
~
M
=
= ~
M B
L
N ~ N
20' '7'
d ~l ~l ~1 b
A E

~
Fig. A7.67 Fig. A7.68

~=.004743 Rb=.004667 (8) Far the truss in Fig. A7 .67, tind the
vertical and horizontal displacement of jOints C
Fig. A7.61 and D. Take area of all members carrying tension
A1. 18 Prob 1ems as 2 sq. in. each and those carrying compression
as 5 sq. in. each. E = 30,000.000 pSi.
(9) For the truss in Fig. A7.68, determine
A~

~
O"
B the horizontal displacement of points C and B.
E = 28.000,000 psi.
3~'

J
45' \. G 5000 5000 500*
t- - 1000 1000

~
~ 0 A IB Ie 0 ~-r--r.,--r-.,---,
I I i-
to'
H +- t T 10"
C
Fig. A7.62 1>\)(): 3~ A 20" ::i ::'F 2 ,000< 30" T
A B J.. G414~E I e B
40" . 5060 5000 ~ D 4 @20 = 80"
Fig. A7.63 Fig. A7.64
Fig. A7.59 Fig. A7.70

(1) Find vertical and horizontal deflec- (10) For the truss in Fig. A7.69, rind the
tion of jOint B for the structure in Fig. A7.62. vertical deflection of jOint D. Depth at tr~ss =
Area ot AS = 0.2 sq. in. and Be = 0.3. E = 180". l..tIdth ot each panel is 180". The area of
10,000,000 psi. each tr~ss member is indIcated by the number on
(2) For the truss in Fig. A7.63, calculate each bar in the figure. E = 30.000,000 pSi. Al-
the vertical deflection of jOint C. Use AE so calculate the angular rotation at bar DE.
tor each member equal to 2 x 10 7 • (11) For the truss in Fig. A7.70, calculate
(3) For the truss of Fig. A7.64 determine the vertical a.~d horizontal displacement of
the horizontal deflection of jOint E. Area of joints A and B. Assume the cross-sectional area
each truss member = 1 sq. in., E = 10,000,000 tor members in tension as 1 sq. in. each and
psi. those in compression as 2 sq. In. E = 10,300,000
(4) Determine the vertical deflection of pat ,
joint E of the truss in Fig. A7.64 (12) For the truss in Fig. A7.70 calculate
(5) Determine the deflection at joint D the angular rotation at member AS under the given
normal to a line jOining joint CE or the truss truss loading.
in Fig. A7.64. 1000* 5000
I i o
B iii
b
12'
r
" 10
' T
A
10' is'1
c Fig. A7.71 Fig. A7.72
h
g I
(13) For the beam in Fig. A7.71 deter.nlne the
d 1.-
-
e
B C
deflection at points A and B using method of elas-
~a' ---.-a' ~9'
, , tic 'N8ights. Also determine the slopes ot the
I SO,OOQt- elastic curve at these points. Take E = 1,000,000
Fig. A7.65 pSi and I = 1296 In.~
(14) For the beam in Fig. A7.72 find the de-
(6) Calculate the vertical displacement of
flection at points A and E. Also the slope at the
joint C tor the truss in Fig. A7.65 due to the elastic curve at point C. Assume EI equals to
load at joint B. tlembers a, b, c and h have
areas of 20 sq. in. each. Members d , e , t , g 5.000.000 Ib in. sq.
and i have areas ot 2 sq. in. each. E =
30, 000, 000 psi.
 A7.38 DEFLECTIONS OF STRUCTURES

500Jl
10"---.1
100,*
A
,--- t=25"
"
-j.<a'
;r.~1
t 2"1

'I
~A
, 5"
5"
50"
lOOOf
Fig. A7.81 Fig. A7.B2
(15) Fig. A7.73 illustrates the airloads on (23) In Fig. A7.81 find the vertical ~ove­
a tlap beams ABCDE. The tlap beams is supported ment and the angular rotation of point A. Take
at B and D and a horn load or 500# is applied at EI = 12,000,000.
C. The beam is made tram a 1~-.049 aluminum (24) Determine the vertical deflection or
alloy round tube. 1= .01659 in~; E = 10,300,0 point A for the struc~~e in Fig. A7.82. EI =
psi. Compute the deflection at pOints C and E 14,000,000.
and the slope of the elastic curve at point B. 100#

F ...
B 3"
lOOt 100
"'3" 1"3"
100
3 ., F91t -;- 100
9" + 9 "
100*
j
100
~~1=t=-.~a=58~st=e~el=tU=b~e~~" Fig. A7.83
t 0 1 B A ri:1 20" Y
A
Fig. A7.74 Fig. A7.75 100ill'
(16) For the beam or Fig. A7.74 determine V
c
the dert.ectaons at paints C and D in 'terns or EI
which is constant. Also determine slopes ot the
Fig. A7.B4 L. o r a•
elastic curve at these same paints.
(17) For the cantilever beam ot Fig. A7.75
determine the detlections and slopes or the
elastic curve at points A and B. Take EI as
constant. ExPress results in terms or EI. Front View Side View

(25) The cantilever beam of Fig. A7.53 is

p." T·" "1~
1001 400f
loaded normal to the plane of the paper by the
two loads of lOCI each as shown. Find the de-
BEMA flection at point A no~l to the plane of the
Fig. /07.77 paper by the method of virtual work. The rec-
tangular moment or inertia tor the t~be 1s
(18) For the loaded beam In Fig. A7.76 de- 0.0277 in·. E = 29,000,000.
terame the value at the t Ixed end moments MA (26) The cantilever landing gear strut in
and MB. EI ts constant. Also tind the detlec- Fig. A7.84 is SUbjected to the load or 500# in
tion at points C and D in terms at EI. the drag direction at point A and also a torsion-
(19) In Fig. A7.77 determine the magnitude al moment of 2000 in. lb. at A as shown. De-
at the tixed end moment I1A and the simple sup- termine the displacement at point A in the drag
port RB. direction. The tube size tor portion CB is 2~­
.083 and tor portion EA, 2~-.065 round ~~be.
p
p Material is steel with E = 29.000,000 psi.
P

IF 11\\ ~
Flg.A7.78 Fig.A7.79
A
tJA RI
I

Fig.A7.80
'11 2
[ aij] {qj} compute the strain energy in the'
truss at Fig. A7.63 (Problem 2). The ~e~ber
flexibility coefficient fer a member under
uniform axial load is L/AE (see Fig. A7.35a).
An". U = 22.4 lb. In.
(27) Using the matrix equation 2U = Lqi I

(20) In Fig. A7.78 EI is constant throughout.
Calculate the vertical deflection and the angu-
lar rotation ot point A.
(21) For the curved beam In Fig. A7.79 rind
the vertical deflection and the angular rota-
tion at pOint A. Take EI as constant.
(22) For the loaded curved beam at Fig.
A7.80, determine the vertical detlection and the problem 25 tor a staowed cantilever beam whose
angular rotation at the point A. Take EI as
conatant ,
(28) Using matrix equation (23) compu~e
the strain energy in the beam at Fig. A7.71.
Note: the choice of generalized forces should
be made so as to permit computation at the
member fleXibility coef~iclents by ~he equations
of p . A7.l9. Ans. U =- 3533 lb. in.
(29) Re-solve the problem of example
I doubles at point "2 and doubles again at "3".
(HeaViest sect10n at built-in end.)

_ " . .; . /' : _:;. i' .:.", :,.>.:-:~.::~ -:-:-.. :.~,~ :'::::;:<; ~ ..": <:' ~ -,. -- -

-<";:. ;','" _~ -::!,~. :',." ~2::'':~'' .....<:\-~. ~-: ';0,\, . ._~r '3 :./"-:- "'-~. ~ ,:t ._~ <-.' l: /}",-.f,~~7'f:": N" - ~
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A7.39

(30) For the truss of Fig. A7 85 determine

T
the influence coefficient ~trix relating
vertical d.eflections due to loads P1.' P 'U P~
and P~ applied as shown. Member areas are
shown on ~he f~gure.
B t:.051"
A 1.0 m 1.0 G : 3.85 x 106 psi
;

1.0 10"

2.0 1.5 n-l

r, L Fig. A7.85
10" -Jrn Fig. A7.86

Answer.
(34) Find the influence coefficients re-
lating deflections at points land 2 of the

[" ~.~
44.67 44.67 simply supported be~ of Fig. A7 87. Use
44.67 106.0 99.34 38.0 matrix methods.
[3,nJ. E 1
44.67 99.34 99.90 38.0
33.00 38.00 ~8.00 33.0

(31) For the truss ot problem (30) ceter-

~ine which of the following two loading con-
ditions produces the greatest deflection of
point 4, (All loads in pounds},
(j) @+ ( AE:3XI~~lbS
AE:8 x 10 6 lbs
Condition )10. p. p ,

1 1000 500 800 400 Fig. A7.87

2 300 700 400 600

Ans.

(32) Determine the matrix of influence 7.10~

coefficients relating drag load (positive aft),
braking torque (positive nose up) and moment
in the V-S ~lane (positive right wing down) as
12.~
applied to the free end of the gear strut
assembly ot Problem 26.
Note to student: It will be highly instructive to re-work
Answer. problems 33 and 34 using the alternate choice of generalized
forces in the stringers from those used in your first solution.
See p. A7.22 for alternate generalized forces on a stringer.
f2S00 -50.7
I -50.7 6.97 References for Cr~pters A7 J AB.
Lo o
T = BOOKS ON STRUCTURAL TOlEORY

"Advanced Mechanics of ~~terialsn, F. Seely and

the 1eflection of the load
(33) F~nd J. O. Smith J 2nd Ed., John ~ileYJ New York.
applied to the cantilever panel of F~g. A7 86.
(Assume the web does not buckle). Use matrix "Advanced Strength of Materials", J. P. Den
notation . .~s. 0 = 7.94 x lO-~ inches. Hartog J McGraW-Hill, N. Y.
"Theory of tna st ic rtv", S. T1m.oShenk0 J rrccraw;
arn , N. Y.
A7.40 DEFLECTIONS OF STRUCTURES

TEXT BOOKS wITH MATRIX APPLICATIO~~ ~ehle, L. and Lansing, ~'J A Method for Reducing
the Analysis of Ca~p1ex RedUr..~tructUIes tc
"Elementary ra cr-ices", R. A. Grazer, v : J. ~ Routine Procedure, Jo~n. of Aerc. S:~.
Duncan and A. R. Collar, Cambridge University Vol. i9, 1952.
Press. Langefors, B., ~~alySis of ~las~lc Structur~5
"Introduction to ~he Study of Aircraft Vibration Matrix Transformation with Spec ~al ?,eg-a.r1 ~
and ~lutter", R. Scanlan and R. Rosenbaum, Sere1 Monocoque Structur~ .jour-n , of Aero. Sci.
Mac Millan, New York. 19, 1952.
Langefors, 3., ~trlx r.ethods ~or Recundant
Structures, Journ. of Aero. Sc~ Vol. 2C, :;53.
TECHNICAL PAPERS Falke~~einer, H. Systematic )~alysls o~ Redundant
21astic Struct-:.rres 'Jy :-:ea:;s of ~aL:rix Ca:c'.!lus,
Benscoter, S. U., "The Partitioning of ~atrices .Jour-n , or .Aer-o, 3ci:-: 20, ::''.753.---
!E Structural AnalySTS", .rourn. of Appl , hechs ,
Vol. 15, 1948. Argyris, J. and Kelsey, S., 2nergy Theorems in
Structural Ahalysis, Aircraft ~~gineeri~g,
cct • lS54, e t , seg ,

. .
All.
;..~~
... ~
....-.
• a';~'~ .

"-\'

Many problems involving calculation of deflections are encountered in the structural

destgn of a large modern airplane such as the Douglas DC-a.

 CHAPTER As
STATICALLY INDETERJ.\1INATE STRUCTURES
ALFRED F. SCHMITT
AB. 00 Introduction.
A statically indeterminate (redundant)
problem Is one in which the equations of static
equilibrium are not sufficient to deter.nlne the
internal stress distribution. Additional re-
lationships between displacements must be
written to permit a solution.
The "Theory of Slastlclty" shows that all
structures are statically indeterminate when--
analyzed in amut e detail. The engineer how-
ever, Is otten able to ~ke a number of as-
sumptions and coarse approximations which render
the problem determinate. In addition, auxiliary
aids are available such as the Engineering
Theory of Bending (MJ) and the constant-shear-
flow rules of thumb (q = T/2A) (see Chaps. A-5,
A-6 and A-!3 through A-IS). While these latter
are certainly not laws of "statics", the en-
gineer employs them often enough so that prob-
lems in which they are used to obtain stress
distributions are often thought of as being
"cetermnace" .
It is frequently the case in aircraft
structural arzlysis that, in view of the re-
quirements for efficient design, one cannot ob-
tain a determinate problem without sacrificing
necessary accuracy. The Theory of Elasticity
assures the existence of a sufficient number of
auxiliary conditions to permit a solution in
such cases.
This chapter employs extensions of the
methods of Chapter A-7 to effect the solution
of typical redundant problems. SpeCial methods
of handling particular structural configurations
are shown in later chapters.
AS.O The Principle of SU'Pel"?osition.
The general principle of superposition
states that the resultant effect of a group of
loadings or causes acting Simultaneously is
equal to the algebraic sum of the effects acting
separately. The principle is restricted to the
condition that the resultant effect of the
several loadings or causes varies as a linear
function. Thus, the principle does not apply
Nhen the member material is stressed above the
proportional limit or when the member stresses
are dependent upon member deflections or de-
formations, as, for example, the beam-column, a
~ember carrying bending and axial loads at the
same time.
AB.! The Statically Indeterminate Problem.
Several characteristics (and interpreta-
tions thereof) of the statically indeterminate
prOblem may be pointed out. rhese character is-
AS.!
tics are individually useful in forming the bases
for methods of solution.
C[J There are more members in the structure
than are required to support the applied roacs ,
If n members may be removed (cut) while leaving
a stable structure the original structure is said
to be "n-times redundant".
-eOROLLARY-
In an n-times redundant structure the mag-
nitude of the forces in n members may be assigned
arbitrarily while establiShing stresses in equi-
librium with the applied loads. ThUS, in Fig.
AS.I (a singly redundant structure), the internal
force distribution of (a) is in equilibrium with
the external loads for any and all values of X,
the force in member BD.
~. ~12l
c-: c-: Nt-to:
_A D -:'Jlc----'!J!' A D
20004;' 2000# '"' .707X •
\.00"* 1200"* hoo,,* 200,,*
<a) (b) (e)
Fig. AS.!
SIngly redundant stress distribution, (a) consisting 01 a stress
in static equilibrium with the awlied loads, (b), with one
zero-resultant stress distribution, (c), superposed.
Only the system (b) is actually required to
eqUilibrate the external loads (corresponding to
X = 0). Note that the system (c) has zero ex-
ternal resultant.
[[j Of all the pOSSible stress (foroe) dis-
tributions satiSfying static equilibrium the
one correct solution is that one which results
rn-kinematically pOSSible strains (displace-
ments), i.e. retains continuity of the struc-
ture.
ThUS, for example, there are an infinite
number of bending moment distributions satis-
fying static equilibrium in Fig. AS.l (d) since
~ can assume any value. Of these, only one
will result in the zero deflection of the right
hand beam tip necessary to maintain structural
continuity witn the support at that point.
 AB.2 STATICALLY INDETERMINATE STRUCTURES

the Theorem of Least Work. In words; Rthe rate

--
of Change of -strain energy with r9S?ect to a
ilxed redundant reaction 1s zero R.

M. -I===:r-<:'.
M

PL/.
AS. 2. 1 Determination of Redundant Reactions by
Least Work.

Example PrOblem A
Fig. A8.1d
By way of illustration, the proble~ posed
by Fig. A8.2 'NaS carried to completion. The
Singly redundant beam with root bending bending moment was given by (x, YJ ~ measured
tram the left ends of the tr~ee beam dlvls~Qns)
moment Me undetermined by statics.
M

-eOROLLARY- (500 + RxJ L

M + (Rx-500 J Y
It n member loads have been assigned ar- 2
bitrarily ~hile establishing equilibrium with
the external loads, relative ~ovements of the M • Rx (L - g) O<;;;<L
elements will result, violating continuity at n
points. n zero-resultant stress (force) dis- Then
tributions may then be superposed to reduce the
relative motions to Zero. The resulting stress
distribution Is the correct one.
AB.2 The Theorem of Least Work.
U' 2
1 f M' dx
EI '2EI
1 JLit(500+Rx ) "y ' 1x
0

A theorem extremely uset~l in the solution

of redundant problems may be obtained trom
Castlgliano's Theorem. Consider first the
problem ot redundant reactions such as in a
2EI L
+ -1-ffYT500+RxiL + (Rx - 500 Iy]'d Y
2
o
beam over three supports (Fig. A8.2). One ot
the reactions cannot be obtained by statics.
+ 2~I EX'(L - g) 'ds

&
....El = constant Differentiating under the lntegra~ 51'S!:, (see
p , A'7.8)
--t---
500-2Rx
L
au
Fig. AS.2 oR x •
A singly redundant beam with one reaction given an ar-
bitrary value (Rx)'

It the unknown reaction (say tr~t on the :ar

right) is given a symbol, Rx , then the re~zining
reactions and the bending moments ~y be de-
termined from statics. The strain energy U ~y
then be written as a function of Rx , i.e., L
U = r (Rx). Next form + _~ , fR'X (L - 0,. , \ 'I
-"
'"'

'1
o

This is the deflection at rtx due to Rx o But

this ~ust also be zero.-Slnce the support is
rigid. Hence

~. a - - - - - - - - - - - - - -
aR (1) .. (500 +
x 2
Eq. (I) is true for all redundant reactions
occurlng at fixed su;po~ts. 3ecause it corre_
sponds to the T~the~atlcal condition for t~e
minimum of a function, sq. (l) is said to state

BE.""...._ _IIliIIiiii......... ..
2 a
_
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AB.3

Completion of the computation gave

Rx =- 1500
~ lbs. =- 93.8 lbs., au
ai1= 0
J ~
= [Ll1L+ • + 2PL - l1L X] -Xd y
R 3L 3L
t~e negative Sign indicating that Rx was down. c

Example Problem E
Determine the redundant fixed end moments
for the beam of Fig. A8.2(a).

o =l1L J~ -3~) dx

Fig. AB.2a
~+2PL-'\JL x
A doubly redundant beam with two reactions given + 3L oX(1 - 3L)dx
two arbitrary values.

Solution:
+!lR
2Iy -dY+
'\ + PL - ~ J2~.
-dy
The redundant end moments were designated
as ML and MR for the left and right beam ends
J , 3L 3L
, 3L

~ + 2PL -
respectively and were taken positive as shawn.
The moment equations for the two beam portions
(x trom left end, y from right) were O'l1L
\-dx+ '\ J\.
-r-r dx
3L 3L 3L
~+2PL-,\ Je c
11='\+ x O<x<L
3L

O<Y<2L

Then

Evaluating the integrals and solVing sDnultane-

ausly gave

4 PL
9
+..l.
2EI 2
~ • - - PL
9

Differentiati~g ander the integral Sign (see
remarks on p. A7-8)
AB. 2. 2 Redwldant Stresses by Least Work.
The Theorem of Least Work may be applied to
the problem of determining redundant member
forces within a statically indeterminate structure.
ThUS, in-an-fi-tDn8S redundant structure if the
redundant member forces are assigned symbols X,
Y, ~, - - - etc., the values which these forces
~ust assume for continuity of the structure are
such tv~t the displacements associated with these
forces (the discontinuities) must be zero.
Hence, by an argument parallel to that used for
redundant reactions, one writes,
A8.4 STATICALLY INDETE.RMINATE STRUCTURES

au o
50
.~.
ax
au
ay = 0 __ - - - - - - (2)
+
':;=".l.BC j c
(lOOe
\:

+--
1
2EI~e
etc.
In words, "the rate of change of strain eneT~Y 60
~lthrespect to the redundant forces is zero". 1
+-- o (50,000):ld6'
2EICD
Eqs , (2), like eq, (1) I are statements of the
Theorem of Least ~ork. They provide n equations 1
for the n:tlmes redUndant structure. The simul- Obviously there was no need to ccnsider the
taneous solution of these equations yields the energy in CD as its loading did not depend upon
desired solution of the problem. X and hence could not enter the problem. Dif-
ferentiatln~ under the integral sign
Example Problem C
The can~11ever beam and cable system of au 58.3 X
Fig. A8.3(a) 1s singly redundant. Find the
member loadings by use of the Least ~ork ax ~.U3
Theorem. 50
so
v
(1000 - .30_ x\
'_ _...";,,0,,8.:..,,,,_'-)
1x a ex
58.3 EI BC 0

+ -0 )"
_0- o
(b) ( 50.3
Fig. AS.3
A singly redundant structure with one member force given an
arbitrary value (X).
58.3 X ... 11,032 X • 6620 X
Solution: EA trac S; AC
AB <

The tensile load in the cable was treated

as the redundant load and was glvzn the symbol = 21.44 X 101!!
X (Fig. A8.3(b)). The strain energies con- ELse
Sidered were those of flexure in portions AC,
CD and BC and that of tension in the cab~e AB. Putting
~ergies due to axial forces in the beam ~ort­
ions were considered negligible. AAB 0.025 in a
The bending moment in Be (origin at B) Nas

!'lac = (1000 - 5~~3 XJ x

10.0 in"
InAC, (origlnatA):

50 gave
MAC = 58.3 X • Y
X 613 lts.
In eD: #
Then
l""cD = 50.000
~e [1000 ~ 30
vv.,,)
("3)J
0.1. x = 685x
The strain ener~J was therefore

M..',.c
50
52:3 613 .y 520 y
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AS.5

Exa~ole P~0b13~ D Dif:erentiatins under the integral sign

A semlci~cular pin-9nded, uniforn ring is
support ad and leaded as shown i~ Fig. A8.3(c). 0

As a first ~~prcxi~tlon the ~orlzontal flcor au

ax =
R'
EI x J60
0 sin
,
Q d Q
tie is te Je assumed rigl~ aXially. Find the
~ending moment distribution in the ring.

p p -
R'P
~
j 60'
0 (1 - cos g) sin g d g

0
R' X f90
+ ""-r sin II g d g
"" 6Qo

0
101 Fig. AS.3 .... (d) R'P )90
- 2Ei' sin g d g
60'

60 0
Solution: + ~ locOS g sin g d g

The axial lead in the floor was taken to

be the redundant (since the floor was assumed 90 '
rigid, this could have been thou&~t of as a re- II Q d Q + ~ sin II Q d Q
dundant floor reaction from fixed supports). ] 60'
The loading is shown in Fig. A8.3(d).
The bending moment distribution 'NaS
Evaluating,
11 X R Sin g - PR (1 - cos g) 0(g<60'

X R Sin g PRIZ

The axial loadings were +gP

3 (RAE-TI
R')
s P ccs Q + X sin Q
Therefore
S X sin Q

The strain energy (for only half the

X
3P (R' R)
rr-TI:
structure) was « 2n
(R
-' + -
R)

t
EI AE
60'
U = zir E sin Q- PR R (1 - cos Q)J' Rdg
Example Problem E
The portal frame 1. 50
of Fig. A8.3(e) is
three times redundant.

+ zir
90'

160(X R sin g - PRIZ J Rdg

Set up the simultaneous
equations in the re-
dundant forces. The
relative ~ending stiff-
-
1 000lf
B

2.25 2.25
c
1
50"

nesses of the segnents A

D...1

+ X sin Q r RdG
are given on the figure.
Solution:
Fig. AS.3e

The redundant forces

90 ' selected were the bendlng'moment,-the transverse
+ ~E [x sin QJlI RdG shear torce and the axial force, all at point A.
The four figures A8.3(f) tr~ough A8.3(1) show
the bending ~ament diagrams or the structure due
J60 0
to applied loads and due to redundant forces
• Zero strain energy in the rigid floor (AE -+ ....).
AS.6 STATICALLY INDETERMWATE. STRUCTURES

acting individually (it being easier to com-

pute the loadings in this fashion). The com- .1[-50' ;0 Sln g .. ~ .. 50'1(1. sIn !il) -5crl'll - '~~5!ilIJ(,_ . .s....•
~", """'"
"we...
• -'..:>0
plete loading ,v,as obtalr.ed Jy superposition.
..

I".
(l000ll'·!"l. '{a' .. SO '{ ~ ~OO ":':(~C - s': 1S'
LZ5

Fig. AB.3f 50,000 Fig. AS.3g After evaluation of the ~nte~31s the equations
obtained were

9.682V 7.459T 1.112 x 10'

{'1492!l +
9.682!l + 763.1V 484. a!' 288.3 x 10 3
-7.45911 - 484.0V + 614.9T = 111.15 x 10 3
AB.3 Redundant Problems by The Methods of Dummy-
Unit Loads.
50V
Q.6. ',.Jhile the Theorem of Least ',.jork may be nace
T the basis of red~dant problem analYSiS, its

-
s s'
v
l 1 ~
• lOOT
direct application by the calculus, as in Art.
A8-2.1 and A8-2.2, is often impractical. ~or t~e
Fig. A8.3h lT Fl •. AB.31
majority of problems the work is facilitated if
carried out by the techniques of the Method of
Dummy-Unit Loads.
The follOWing derivation is for a doubly
redundant tMISS struct~e. The extension te a
more general n-times redundant structure, in
The composite bending moments as funct10ns ot which ether loadings in addition to axial (flex-
5, Q and 5' were
ure, t crst c.i and Shear) are present, is indicated
later.
I1 Consi~er the doubly redundant truss of Fig.
AB=Il+Vs
A8.4(a). It may be made statically dete~inate
MBc 50,000 sin 9 + 11 + 50 V + 50V sin 9 by "cutting" two members such as the diagonals
indicated. Application of the external loads to
- SOT (1 - cos g) this determinate ("cut") structure gives a load
distribution, ~S", computed by satisfying static
MCD = 1000 s' + Il - Vs' + 50 V-100 T equilibrium. At this time discontinuities appear
Then since at the cuts "xn and "yn due to the strains
developed.
U = :2
l J Il
~an
' dS d au
aR=;j1i
au au
aT 0, IP, ,P, ..
one haS,

au
aM = ° f
) 0
so
III
2.50
+ vsl ds M x

S loads
Fig. AS.4a
y
y

ux loads
Fig. AS.4b
Uy loads
Fig. AB.4c

.j ["-50.000 sin;"!"l" 5OV(1 .. sIn ;1

To compute these and subsequent displacements
. ,so
1.50
the Method of Dummy-Unit Loads may be used (Art •
A7-7). For this purpose Virtual loads are placed
• J .(10009' .. 11 - 'Is' .. SO '( - 100 ":') d '
Q.= S
alternately at the x and y cuts as in Figs. A8.4(b)
and (c). From the dummy-unit load equations
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AB.7

z S~t} (3) For continuity these net relat~ve displacements

must be zero. Equating the above expreSSions

l
each to zero, and rearranging, gives the simul-
Z SuyL taneous equations
AE
~ - z "",C
llL uxuyL
X Z UX + Y Z
u and u are the unit-redundant stress dis- AE AE AE
x y
tributions as indicated in Figs. A8-4b, c. The u 'L
--- - - (4)
X Z UyUx + Y Z -y-= _ Z SuyL
L
subscript wow indicates these relative dlsplace- ICE AE AE
~ents occur in the determinate (Wcut W) struc-
ture with the Wor i gl na l Wstress distribution. Eqs. (4) are two simultaneous equations in the
It is now desired to close up the dis- two unknowns X and Y. Upon solution for X and Y
continuities by application ot redundant loads the true stress distribution may be computed as
X and Y to the x and y cuts, respectively, as
In Fig. A8-4(d). Load X (5)
causes a stress distri-
bution XUx and, likewise,
~ For a structure which is only singly re-

M
Y causes a distribution dundant, eqs. (4) and (5) are applied by setting
YUy. The relative dis- Y = a giVing
placement at cut x due to
redundant load X is given xZ ux aL = _ Z SuxL
by (6 is read wdisplace- Fig. AB.4d AE AE
xx
ment at x due to X W) . or, Simply,

SUxL
Z Ai:
X - - - - - - - - - - - - - - - (4a)
and at cut.y by (read 0yx as wdisplacement at u 'L
Z ...L-
y due to X W ) . AE

= Z XUx • uyL and,

AE
- - - - - - - - - - - - - (5a)
Similarly the load Y causes displacements at
the cuts y and x given respectively by
AB.4 Example Problems· Trusses With Single Redundancy.
e ~ Z YU y ' uyL =Y Z uyaL
Example Problem #1
YY AE AE

and - 30 - po.o* 0 1000* -.707

Z YU1 • llX L = Y Z uyuxL T~a O'~-g~g

AE AE
-h~l~~ "~,,
Now the net relative displacement at each
cut under the slmultar.eous action of the three hoo"" \000'
Fig. AB.5 Fig. AB.6 Fig. AB.7
stress systems 5, X-u and Y.u y is
x
Fig. A8.S shows a single bay pin connected
6 + 6 + 6 = Z SUxL + X Z uxuxL truss. The truss is statically deter.nir~te With
xo xx xy AE AE respect to exter~l reactions, but staticaily
indeterminate with respect to internal member
+ Y Z u y uxL loads, since at any joint there are 3 unknowns
AE with only ~NO equations or statics available for
a concurrent force system. The truss is there-
and fore redundant to the first degree. The general
Z SuyL + X Z uxuyt- procedure for solution is to make the truss
AE AE statically deter~inate by cutting one or the
~embers;on Fig. A8.6, member bc has been selec-
ted as the redundant ~ember, and it 1s cut as
+ Y Z uyuyL shown. The member stresses S for the truss ot
AE
A8 8 STATICALLY INDETERMINATE STRUCTURES .
Fig. AB.6 are then deter.n1ned, the results being A B C A B C
recorded on the r:le::nbers and also entered in
Table AS.L In Fig. AS. 7, a unit If tensile
dummy load has been applied at the cut section
of the r-edundant member be, and the loads in all
~/ ~/ V'...; Iiii'

the m.embers due to this unit load are calculated 0 0

1,000
The results are recorded on the figure and also
in Table AS.l under the head of u stresses. The Fig. AS.9 Fig. AS. 10
solution for the redundant load X in the redund- S loads u loads
ant aemoer- be is given at the bottom of Table
AS.!. The true load in any member equals the S Example Pr'cb Lem s I-A; Deflecti0n Ca:c~l~tiJn 1:1
stress plus X times its u stress. a Redundant Truss
bble U.1 Calculations of the deflections under load of a redundant
structure are made by application of the methods of Chapter
»""r L
• s u
';t u 2L
T
True Stress
= s + Xu
A-7. Since, however, there are certain piUalls as regards
symbols and also some important special techniques, the
••••
d.
••
u.
30
30
30
30
42.4
1
1
1
1
2
_1000
0

0
0
0
_.107
•• 707
_.707
_.707
1.0
0
21210
0
0
0
".
15
15
21.2
.60'
_559
3"
395
3••
following examples. are given at this time. The extension of
the method to more complex structures is immediate and no
further work on de!lections of redundant structures is given
in this chapter (excepting in the case of the matrix methods of
.d 42.4 1.' 14140 1.0 40000 28.3 '55 Arts. AS. 10 et. eeq.}.
• 2
X • true 1~c1 111 redundant _.tier be Find the horizontal ~ovement of :Joint ftd"
of Example Problem 1# 1 under the action ef tne
X •
." hi
A.
~
~ -aUlO
109.S
• _5~9.
load applied there.
t"r Solution: The equation used to find the
deflection is Sq. (18) of Chapter A-7. ',iritten
Example Pr'oo.Iem t 2 for application tv t:-uss deflections it is
(see Example Problem 13, p. A7.l1)
Fig. AS.S shows a Singly redundant
3-member frame. Ftnd the member loadings.
Member areas are shown on' the r rgure , a : z~
AE
--- - - - - - - - - - - - -(A)
~~.
Now for a deflecticn calculation the symbols
T~OV·
100" -' .2
'8' and "u" ;nust be carefully reinterpreted
from their ~eanings in the r-edundant stress
J.2 .4 catcutat i cn.' For a deflection calculat1Cml the
symbols of eq. (A) , above, :nean: "s-Loacs " are
0 the true loads of the redundant structure due
ho"", to application of the real external leading;
Fig. A8.8 "u-Loads " are tae loads due to a dummy-untt
(virtual) load applied at the external ;Joint
where the deflection is desired and in the di-
Solution: Member ex: was selected as the recti on of the desired deflection.
redundant and. was cut in figuring the S-loads,
as in Fig. AB.9. Fig. AS.IO shows the u-load Thus , the '8 loads" for use in ;;q. (A) are
calculation. The table completes the calcu- the true stresses (the sclution] of Example
iaercn, Problem It L

SuL u 2L True Load The "u-loads" represent add1 t i onaj infar:na

Mem. L A S u ""A ""A = s ... X u tion which would, in general, appear to necessl
tate another redundant stress calculation. As
AO 141.4 0.2 0 1.224 0 1059 335.5 will be seen, such is ror-cunate Iv not tr.e case.
BO 100.0 0.2 1000 -1. 366 -6. 83x10 5 933 625.6
In the present pr'obIem the dummy-urn t load is
applied ldentically as is the lOOOJll real load
CO 200.0 0.4 0 1. 000 0 500 274.1 and hence the u-loads ar-e simply equal to the
"S-loads" properly scaled .jmvn. The following
Z -6. 83xl0 5 2492 table completes the calculation.
Z 8uL
A
X -- uliL =274.1 1':)•
Z-
A
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A8 •

Mem. L A S' u" SuLI A SuL

Mem. L A S' u A
ab 30 1 3.5 .395 4,680
AO 141. 4 .2 335.5 3. 355x10 5
bd 30 1 -805 -.605 10,980
BO 100 .2 625.6 -1 -3. 128xI0 5
de 30 1 3'5 .395 4,680
co 200 .4 274.1 o o
ca 30 1 3.5 .395 4,680
O.227xl0 5
cb 42.4 2 -55' -.559 6,620

ad 42.4 1.5 855 .855 20,660

:. -5 = 22,700
Z 52.300 E

• true loads from Exa'Ilple Problem #2 •

•• 6 =52,300
E By way of Qemonatration another set of
u-loads, called u l , were found for this same
• identical ','lith the "true str-ess" of Table problem, this time by cutting aember OA. The
AS.l • corresponding calculations follow:

•• si:lply l/lOOOth of the "S-loads" sfnce the

Su'L
dummy-unit load is applied exactly as is S u' rt:
the 10001 real load.
335.5 0 0
Example Problem *2-A.
625.6 .578 1. 808x10 5
Find the horizontal deflection of point 0 274.1 -1.155. -1. 583xlO 5
of Example Problem 12 under application of the
vertical 1000* load shown in Fig. AS.S. Z O.22Sx10 5

Solution; To compute the deflection use

The results are identical (allowing for
round-of! errors).

Again the symbols "S" and "u ll are to be re-
inter?reted for a deflection calculation as
explained above in Example Problem II-A. The
"Sc-Lcada" are now the "true Loads" coraout ed in
Example Problem. #2, above. The lI u_l oadS" are
loads due to placir~ a dummy-unit load acting
horizontally on the structrre at pOint O.
Since this load acts on a redundant structure
it would appear that another redundant stress
calculation is required. However, this Is not
necessary.
Theorem; For the u-10ads in a de~lection
calculation any-s6t of stresses~loads) i~
static equllT'5ri'i:iinNith the dum.lily-unit load
may be used , even :~rom the Simplest of "cut;"
S"trueture5:'" - - --- - - - --
This theorem says that to get the wu_loaasl
for this deflection calculation we may " cut "
any one of the ~hree ~embers and get a satis-
factory set of u-loads by Simple statics I
Berore proving the theorem we complete the
calculation in tabular forn as shown. The
lI u _I oa ds " were obtained by cu~ting ~ember OC
and applying a unit load hor~zontallY at O.
PrOOf of Theorem
==
To prove the theorem above we return to the
virtual work prinCiple and the argument from
which the dummy-unit loads deflection equation,
Eq. (18) of Chapter A-7, was derived (refer to
p. A7.l0). It will be remembered that the de-
flection was shown to be equal to the work done
by the internal virtual loads (u-loads) moving
through the distortions (~) due to the real
loads, r . e., 6 Z u e, =
The internal Virtual
loads are those loads due to a unit load acting
at the paint of desired deflection.
Now for the statically indeterminate
structure these internal 7ir~ual loads (u-loads
are, in general, indeterminate since the dummy-
unit load is applied at an external point ot the
structure. However, we recall that,
i-any stress df s'tr-fbut f o.t in static
equilibrium with the lI a pp l i ed load" (for the
~a::nent now we are thinking of the dummy-unit
load as the "applied load") differs from the
correct (true) distribution only by a stress
distribution having zero external resultant
(~ • .'18.1).
ii - a zero-resultant stress distribution
moving through a set of displacements does zero
work 4
AS.lO STATICALLY INDETERMINATE STRUCTURES

Mathematically expressed these points 2re;

i - ~RUE = U STATIC + ~=o

where UsTATIC 1s a u-load d1st~ibut1cn cbtained

from statics in a simple "cut" structure under
the action of the exterr~lly applied d~~y-unit
load and )~=o is the zero-resultant u-load
system which must be superpozed to give the
true u-load distribution
11- Z <I x "R=o = a
It follows, therefore, that any set of
u-laads in static equilibrium with the
externally applied dummy-unit load will do the
same amount of virtual work when the structure
undergoes its distortion as would a "true" set
of u-loads computed by an indeterminate stress
calculation. That is,
Fig. AS.12 Fig. AS. 13
Ux loads 1ly loads
We note here the rule by which the degree of redundancy
of a planar pmjninted truss can be deter mined. For a truss of
m members with p jointa, the truss is n times redundant where
n "" m - (2p - 3). For a spatial truss (3 dimensionai truss)
n "" m - (3p - 6).
In the present problem n = 11 - (12 - 3) = 2.
~~e calculation is car~ied out in tabular
form 1n Table A8.2. The member d~~ensions ~re
given in the table.

, ,
~
'";" I -..-- ~A
Member t, A 5
S\I,,1.. ..jL \I""I..! 1'rue Load
" "y
A A S"X\I".Yu.,

Z <I x "sTATIC + Z <I x "R=O •

AB 30.5 750 .. ..
-
-
0-27,000 0 21.6 0
I
I 0 1680

I
..,
Be 30. 2~ 0 0 0 0 0 i 43.2 0 .50

Z <I x "sTATIC
CD 40.5 -'000 0 - 0 64,000 0 I 51.2 0
I ·255

eo 50 .25 U50 0 0 1250,000 o 1200 0 are I

50 .s , ,,
Q.E.D. C< 0 0 0 I 0 0
1
100 0 -932

........ ,W: , ..
BE 40.1 -300( -.8 -.6,137,000'137,000 36.5 3(;.6 36.6 -1015 I
AB. 5 Trusses With Double Redundancy ED 30.5 -750 0 - 0 27,000 0 21. 6 1 0 -190

B' 0 0 0 0 62. ~ 0 0 -1550 I

....
Trusses with double redundancy are ,
-..
>0,
handled directly by Eqs. (4). By way of AE
'. 0375.000 0 I 0 0 "'0 !

illustration, the structure ot Fig. A8.4, from -,00<

-..
30.7 0 77, ZOO 0 15.41 0 0 -2070 I

which Eqs. (4) were derived, will be solved

=
tor a loading P:l. 2000,* and P a = 1000#.
40.25 0

Z
0 0

~62,
0 105.01

:tOO,478, 000 341. 11452_ 6 36.6

0 0 1240 !

Choices ot redundants were ~de identical with

those at Fig. AS.4a. Figs. AS.ll, AS.l2.and
A8.l3 show the S, u and u load calculations SUbstituting from the table into Eqs. (4)
r-eepec t i ve Ly, x y gives (cammon fac~or of E diVided out)

I
X Z ~ + Y Z uxUyL - - Z!
SUxL
A A r

X Z Ux~'{L + y Z UrL = - z S\~t

341.1 X + 36.6 Y 562,000

{ 36.6 X + 452.6 Y 478,000

So Ivtng ,
X;:: 1550*
Y--932H

Fig. AB.ll Finally (see Table A8.Z)

S loads True Load =S + XUx + YUy.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STR CTURES A8.11

A8.5 Trusses With Double Redundancy. cont'd. Solving the two equations for X and Y, one ob-
tains X = 521* and '[ = 416*. The true load in
Example Problem 3 any member = S + XU + Yu which gave the values
x y
Fig. A8.l4 shows a structure composed at in the last :olumn of the table.
tour co-planar members supporting a 2000* load.
With only two equations of statics available for AS.6 Trusses With Multiple Redundancy.
the concurrent force system the structure, rel- By induction; eqs , (4) may be extended for
ative to loads in the members, is redundant to application to trusses which are three or more
the second degree. times redundant. Thus for a triple redundancy,
Solution:
U 'L UxuyL uxUsL SUxL
X Z _x_ _ + Y z~+ bZ~ - Z
Fig. AS.lS shows the assumed statically AE \ AE
determinate structure; the two members CE and
DE were taken as the redundants and were cut at uxuyL U 'L
poin~s x and y as shown. The member stresses X Z + Y Z .:L... + b Z Uyu.,L - Z
SUyL
-(6 )
AE AE ~ AE
tor this structure and loading are recorded on
the members. Figs. AS.16 and AS.17 give the
U and U member stresses due to unit (1*) uxuoeL U UsL uoe aL
x y
tensile loads applied at the cut faces x and y.
X Z
AE +YZ ~ +gZ
AE - - Z SUsL
AE
Table A8.3 gives the complete calculations for
eqs , (4) and (5). The load in the redundant and atter solving for X, Y, b,
member CE was designated X and that in DE as Y.
True Stresses = S + Xu + Yu + bU ---- -(7)
r 4' ..,..' ,2' I"- x Y •
~[ ~/y
AS.7 Redundant Structures With Members Subjected to
Loadlngs in Addition to Axial Forces.
::;qS. (6) are extended readily to cover
E problems in which flexural, torsional. and shear
1200041 +200041 loadings occur. ThUS, for a three times redundant
Fig. A8.14 Fig. AS.I5 structure
Xa + Ya +;;a

~
N j/
~ ~,~:./
xx xy XS
""."..-::3 1*
-~V "T' - - - - - - - - -(8)
ux loads Uy loads
xa + Ya +:!a = - °so
Fig. AS. 16 Fig. AS. 17 sx .y
••
Ux aL uxuyL SUxL where
X Z -A- + Y Z - A - = - Z -A-

Substituting values from table a

xx
= Z ux'L +
AE
f :n~dX
E1
+ J tx'dX
OJ

2446 X + 2350 Y = 2,253,000.

X ~ uxuyL + UyaL _ SUyL

+ fP; :d
Y

"A Y Z -A- - - Z -A-

a
"0'
=a
yx
=ZuxUyL
--+
AE
f mxmydX
"SI
Substituting:
2350 X + 3039 '[ = 2,488,000.
TABLE A8.3
+ Jtx~ + If qXq~~dY
i_
~I
~I
IsuxL x 10• • ¥XI0·· ¥ ~ uxXyL True Load .. U 'L :ny'dX
J
Member S.Xu2:.Yuy +
L S 'x 'y --. Z _Y_- + - - - - - etc.
AE E1
AE .286.51 e .806 1. 1541 o I c 280.5 576.5 402 .00
BE .1 72 2000,.1. 564 -1. 7291 -2.2S:! ·2.488 i7S:! 2150 1948 ..,
CE .280.5 o 1. 00 c I c o 402.5 Q 0 I 521

i D' .393.7 0 0 1. 00 0 0 0 312.5 0 ue

iTOtal 1: -2.25:! -2.488 2446 3039 2350
and

".~'''~-.c~:~~;;:?":!.,_
,..',..:;)~~~
A8.12 STATICALLY lNDETERML'lATE STRUCTURES

Solution:

5xO
~ SUXL
--+ J ~xOX +
JTtxox
-"-.- The shear flow in the 2heet panels ~as
AE £1 0" chosen as redundant. Secau2e 0: s~~~et~y :~e
problem was only singly red~da~t. ?1g. AS.IS
+ jj qqxdxdy
Gt
shows the Ux and qx loadings jue to the redand-
ant shear flow X = 1. The real loadir.g in the
dete~inate structure consisted of a ccnstant
load P in the central stringer alone. The

Dye
~ --y-
Su L + J~ OX
-y--+
T\ox equation solved was (ref. eqs. (3)).
J OJ
AE E1

ax dy
x (J Ux ~"ax
...... t..
+ If 0-x' ncct '\ =
Gt
Y
,)
q<l.r
+fJ Gt
_(J SU~E ax + jJ q ~ OXd Y)
Gt

etc. where

and where S = P = constant, in central stringer

REAL S - o in s ice s'tr-fnger-s
5, M, T, q are the real loads in the de- LOADS {
terminate structure; q - 0

Ux ' rnx' t x ' ~ are the unit (virt~l) load- Ux - L - x in s1de stringers
ings due to a unit load at cut x; VIRTUAL
LOADS _ 2(x-L) in cent ra 1 s 'tr-:
r ...nger-s
uy' my, t y ' ~ are for a unit load at cut
{ ~ - 1.0
s, etc. When evaluated, (note that the double inte-
grals Simply reduce to a constant times/the
The redundant force(s) need not be an axial panel area)
force but may be a moment, torque etc.
Atter solution for the redundants, True X C: ~; + 2itL) = - (-P1E')
Axial Forces =

S·+ X u
x
+ y U
Y
+;; u. x• .E.
2L
( 1 1 ME )
+ GtLa
True Bending Moments - - - -- - - - - (9)
Therefore the true stresses were

1 +

= LX = 2~ --~,"~ln side stri~gers

etc. 1
1 + AEb
Example Problem 4 L'Gt
The symmetric sheet stringer panel of Fig.
AS.IS is to be analyzed for distribution ot load
P between stringers. As a tirst approximation,
assume constant shear flow in the sheet panels.
All stringers have the same area.
//// //
L
t
A·
const•
1
L
x. 0

...
'!-1o-:-b-t:.lLb-~..l;l
1 l1l J tj "Clamped" I- L .j.. L -'h
Fig. A8.19

x
p
Fig.A8.I8
qx loadings
Fig. A8.19
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AS. 13

Solution: x (L·/'Z ...

'(AE
L')
621
(L
+ Y \v~ +
(l + \/'2iV\
3EJ ")
!ha bendin~ momants at ~cint C in ~ember
':30 and at point 3 in member- .':'.3 were selected
as radundants yielding (when cut) the pin- (2L/[ L')
jOinted deter.ninate struct~e of Fig. AS.20.
=- PL \" ;'.2 -3£1
~te virtual loadings were as shown in Figs.
A8.21 anc. A8.22. For a specific case it 'NaS ass~~ed that

A
~ :: 100
L
g.
L
giving

'+
~ .3716 X ... •1526 Y •• 09011 PL
'<
X=1,·tcu~-1!L B 1 Y=1" *
{ .1526 X + .8121 y. .3616 PL

~L
Ux> m x loadings
P
Uy, my loadings
X .
{ Y· .456
.0645 PL
PL

Fig. AS. 21 Fig. AS. 22 Then as usual

~

Note that in Fig. A8.22 the unit redundant True Stresses :: S + X U ... Y U
loading 'NaS applied as a self-equilibrating set x Y
af unit couples. The real and virtual load- rrue Moments = !1 + Xrn. + Ym y
ings were as follows: (member portion 3D J x
haVing no virtual loadings J was omitted. It A8.8 Initial Stresses.
could not enter the calculation.) In a red'.lndant structure In~tlal stresses
I are developed if J upon assemblYJ certain mem-
Member S M Ux "'x uy !lly bers must be forced into place because of lack
of fit. In some situations intentional misfits
AB 2P$ 0 Y2/L 0 -IlL/£" 6/Lv'1l" are employed to obtain more favorable stress
y-L distributlons under load (~prestressing").
BO -2P ?y - IlL L 0 - YI L "If, in Fig. A8.4(a), the redundant member
with the "x cut~ was initially oversize (too
long) an amount 0Xi (an overSize J corresponding
The equations which were solved were (Ref. to a distortion in the positive X direction, is
eq , (8)).
a pOSitive 0Xi)J the modified condition tor
continuity at the x cut would be (compare with
the equations just preceding eqs. (4)).

Similarly if the Y redundant member were too

long

°10 + 6y1 ~ 6yx + 6YY = 0

Then using the previous notations J the appropri-

ate equations for the redundant forces ~e

y ~ Ux,\L ~ SUxL _
After evaluation of the integrals and
~lylng t~~ough by L~ these become
~ulti­
X f '_xL
_ _ ...
-- = - °Xi 1
rx:.U .
AE AE AE
- -(10)
L r'L fUA~L
= _ PL ((2 ... 4v:1"1L _ ~\
X y :E =- - °Y1
t
,-, Ai: 621 ) The ~8 loads" of eq. (10) are present because
of applied ext3rnal loads. These mayor may not
be zero depending upon the problem.
AS. 14 STATICALLY INDETERMINATE STRUCTURES

Example Problem 6 If, as is sometLmes the case, the number of

If in example problem 3 ~e~ber CE was 0.01
inches too short before assembly, determine the
stress distribution after assembly and load
application.
Solution:
Data obtained from the ~revious problem was
substituted into eqs. (10) along with
(negative because ~too short~)
misali~ents exceeds the number of redundanCies,
or it the misalignm~nt does not coincide with
the redundant cut chosen but occurs elsewhere,
one ~y use the v1rtual work pri~ciple to com-
pute the effect of these misalignments on the
redundant cuts proper. ThUS, referring to the
~virtual work~ derivation of the Dummy-Unit
load equations, {Chap. A7) one has

where 0 ~ is the initial misalignment in the

x"
determinate structure at the X redundancy-cut
to give due to initial imperfections (equivalent to
initial strains) 6 throughout the structure.
1
2446 X + 2350 Y = 2.253 X 10 8 + .0lE u
x
as before, 1& the unit loading due to a
{ 2350 X + 3039 Y = 2.488 X 10 8 virtual load at cut x.Eq. (11) and similar ex-
pressions for the Y, ~, etc. cuts may be in-
With E 29 X 10 8 the redundant forces were serted in eqs , (10).
X = 985 Ibs. Example Problem 8
Y = 57 Lbs , Referring back to example problems num-
bers 3 and 6, assume that member BE is .025"
Then, as usual, too long. Determine the initial stresses if
the other member-s are of proper length and no
True stresses = external load 1s applied.
Solution:
~ple Problem 7
Assume that in the structure of example To employ the soone equations as those of
problem 5 an angular misalignment occurred be- example prr~lem 3, the initial imperfections
tween members AB and CEO at Joint B such that occurring at the same x and y cuts used there
the end at member AB had to be rotated 2.7°. _ were computed, in this case due to the initial
clockwise to fit upon assembly. Determine the elongation at BE. Thus
moments developed Without external loads
applied. 0Xi = Z UxJl i = (-1.564)( .025') = -.0391'
Solution: 0Yi = Z UyJl j = (-1.729)(.025) = -.0432'

The initial imperfection was 0Yl

Then, use of those previQusly computed coeffi-
- 2.7/57• 3 = - .0471 radians cients in eqs. (10) gave,
The Sign was determined by noting that the
original misalignment was in the negative di- 2446 X + 2350 Y = .0391 E
rection of the redundant couple Y. 2350 X + 3039 Y = .0432 E
The equations used from the previous
problem were (noting the equations there had With E = 29 x 10' pSi
been multiplied by L8 x EI/L 3 ~ EI!L).
X = 263 Ibs.
.3716 X + .1526 Y ::: 0 Y = 209 Ibs.
{ .1526 X + .8121 Y = .0471 E
C Finally,
SolVing, True Initial Stresses = S + XU + Yu
X Y
X = - .0258 EI/L. AB. 9 Thermal Stresses.
Y = .0630 EI/L Stresses induced in redundant structures by
thermal strains may be computed by application at
True tnitia1 stresses and moments were deter- methods presented above. The problem may be ap-
mined as usual. proached from the point ot view ot computing the
 .. 3£!!!l1"fiW
_ _ _ _ _ _ _ _ _ _IIlII.IIIl!II

ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A8.1S

relative ~otlons at the cuts of the deter.nlnate Example Problem 9

structure caused by the thermal strains and then The end upright of the truss of ~ig. A8.23
restoring continuity by applying redundant ~em­ is heated to the temperature distribution shown.
ber forces to the cuts. Determine the stresses and reactions developed.
Specifically, consider a doubly redundant
truss such as that of Fig. A8.4(a). After

~
making cuts ~x" and "y" to render the structure
determinate, the application ot the temperature B;:rr A ET(~;~
distribution IS visualized. Relative displace- 30" x
ments occur at the cuts,dencted by exT and 0yTo
--l ,
These displacements may be computed by the
Dummy-Unit Load method as shown in Art. A7.8 of
Chap. A? After this calculation Is accomp-
40" C (permits
sliding
Y r-- --l r
Fig. AS. 23 Fig. A8.24
lished, the problem proceeds as for initial vertically)
strains, Art. AS.S. Thus, the continuity con-
dition at the cut gives (compare with the equa- Solution:
tions 1mmediately preceding eqs , (4) and as-
sume tor simpliCity that the external loads are The structure was made determinate by cuts
absent, making axe = aye = 0) x and y as in Fig. A8.24. The unit loadings are
shown in Fig. A8.25.
0xT + On + 0xy = 0 }

0yT + 0
yx
+ 0

In a truss, the thermal strains produce

yy
= 0
.6
"
,tSJ
1#.6 ~o
~'~
relative displacements at the cuts given by the vx loading I,.
~vlrtual work~ derivation of the Dummy-Unit Uy loading
Fig. A8.25
Load equations (ret. Chap. A7, Arts. A7.7 and
A7.8) as
The thermal coefficient a was assumed
exT = I U a T dx } constant. The calculation was set up in tabU-
- - - - - - - - - - - (12) lar form.
0yT = I U: a Tdx

where a is the material thermal coefficient of

expansion, T is the temperature above the
ambient temperature and U and U are the unit
x y ·'.11
load distributions due to virtual loads at the , ,.t
x and y cuts, respectively. The sums in eqs.
(12) are written as integrals rather than finite
sums to allow for possible variation in a and T,
lIll sol • 11.010 I'.UI
,._.Ill" 1.10.10-"- . • • 10-'1 .,,,r
".
i UI

along the members as well as from ~ember to

member. Then the final equations for the rial
stresses in a doubly redundant truss become Substituting into eqs. (13)

-JU
=9 a
r 'L 5 . 008 X - .90 Y T X lOll
x ~ uXuyL
X + y --- x
a Tdx {_ .90 X + 1.50 Y 0
AZ AE

rUL
- - (13) So Lvtng ,

X .!:....!....
AE
+ Y ~U'L
.!-
Ai;;
=
-f uy
a Tdx x = 2.01 a T x 10 II

Y = 1.21 a T x 10 1I
Equations (13) may, of course, be extended
tor application to structures other than True stresses are given in Table AB.4.
~usses. The expressions appropriate to other
loadings have been developed previously (eqs. Exa~ple Problem 10
(8) et seq. in this chapter and other equations upper surface of the built-in beam of
T~e
in Art. A7.3). Fig. A8.26 1s heated to a uniform ta~perature T.

-:
AS.16 STATICALLY INDETERMINATE STRUCTURES

Through the depth of the beam the temperature Then the equations corresponding to eqs. (13)
varies linearly to normal (T = 0) at the lower were written (see also eqs. (8)).
surface. Determine the end moments developed,
neglecting axial constraint and influence of
axial forces.
X(t:~dS J\:dS) y(J u<t Jmx;{s)
+ + +

r- L T I

~~ EI constant Virtual Loading

Fig. AS.26a Fig. AS. 2Gb

Solution:

The problem was only singly redundant be-

cause of symmetry and was made det~inate by
cutting the end bending restraints. Applica-
tion of unit couples (Fig. A8.26b) gave m = 1 =
canst. Then (see Example Problem 24, Art. A7.8)
the thermal deflection at the ~cut~ was

Evaluated, the equations were

The redundant moment; equation was (by analogy to ....!-x_...B....

EI .81 Y
+ O.~'"' = + h«:
eq, 13)
!!lx'dx -R X 1 (1 aRT
X --=-0xT EI ,. 2 AE" R')
EI Y + 0 • "0 =-
J EI

• + O·y + ( -=- =0
"Therefore O·X 1 +-'-
R ' ) ;;
)~ E1
L
X "IT =- TaL/h
Note that from the last of these equations
g = 0,
as it must because of the symmetry of ~~e
X = - TaEI ring. SolVing the first two equations
-h-

The redundant moment compresses the upper fibers X = aT:I

as was to be expected.
y = 0
Example Problem 11
Complete the problem begun in Example A non-zero value of Y would ~roduce a vary-
Problem 24 Art. A7.8, viz, that of computing ing bending ~oment which cannot Je because cf
the thermal stresses in a closed ring whose symmetry. Hence this result too, is rational.
inner surface is uniformly heated to a tempera-
ture T above the outSide. AB. iO Redundant Problem Stress Calculations by
Matrix Methods.
Solution: In the following section the Indete~inate
structural problem 1s formulated in ~~rix no-
The ring was made determinate by cutting at tation. The reader is assumed to be f~iliar
the top as in Fig. A7.30(b). The unit loadings with the matriX applications of Art. A7.9 and
~~d thermal deflections were dete~lned in the the elements of matrix notation and arithmetic
referenced example. The results of deflection (see Appendix).
calculations made preViously were The stress distribution of the structure is
specified by a set of internal generalized ~orces,
Orr = 2n RaT Ql' qj'- (ref. Art. A7.9). Unlike the cas~
h
the determinate structure, these qi' qj cannot be
0:rr = -2n R .aaT
-h- .. In the case of indeterminate structures, wherein some of the
support reactions may also be redundant, these reactions
also are denoted by qs. (see Example Problem 13a).
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
A8.17

related ianeaa at etv to the extarna I loads by the SYMBOLS - continued

equations of stat~cs. Thus a certain subset of
the qt' qj are r3dundants and are denoted by - the temperature (above nOrTAl) at
points i , j.
~, Q.s' (r , s different numerical suoscr i ots ) . - the ~ember thermal distortions as-
~nen, finally, the redundant forces, ~, sociated with qi J qj'
_ qs ,
are computed (by satisfying continuity) the true
values of all t~e qi' qj may be found by
In the notation of Arts. A8.3 et seq, the
statics. final true values of the stresses were ex-
pressed as {eq , (5) Art. A8.3)

8maOLB - (5)

G. , qJ - internal generalized forces acting

t In the notation to be employed hers, this
on structural elements and re-
act~ons at support points.
equation is restated as
'1,.' qs - redundant generalized ~orces and
redundant reactions (14)
P:nJ Pn - applied external loads
sr - the value of q , (q ) in the de- Here:
"Im ( =gJn)
terminate stru~turejunder appli-
cation of a unit external load
P,,=l(Pn=l) [gim] (=[gjn])' is the matrix of unit-load
gtr (=gjs) - the value of qt (qj) in ~he de- stress distributions in the determinate ("cut!t)
terminate structure due to appli- structure found by the application of unit
cation of a unit redundant force (Virtual) ~oads at the external loading points.
'I,. =1 (qs = 1) The product [glm] {Pm} then 6i'TeS the real loads
G (=0 ) - the true value of ql (qj) in the
1" In 1n the determinate structure, corresponding to
redundant structure due to appli- the "S!t loads of eq, (5).
cation at load P = 1 (p = 1)
" n
G~n (=G ) - the true value of ~ (qs) for a
sn [g1rJ~[gjsJ) is the za'tr-tx of unt t-d.oad
unit value of applied load
stress distributions found by application ot
P = 1 (p = 1).
m n unit (virtual) forces at the red1.Uldant cuts in
a - member fleXibility coefficient: the determinate (!tcut Tl ) structure. Hence this
ij deflection at point i for a unit 1s the matrix of ux ' u y ' etc. loads.
force, q1 = 1 (see Arts. A7.9, 10)'
The {~} of course, correspond to X, Y,
amn - influence" coefficient for the de- etc.
terminate structure: displacement Note that the [gim] and [gtr] matrices are
at exterr~l loading point ~ for a
unit applied load, P = 1. load distributions computed and arranged in much
n
(a = a ) the sarr,e fashion as was [Gim] of Art. A7.9. 'The
mn nm
a~n (=u infl~ence coefficient for t~e de-
sm) - small letter It g " 1s used to indicate load dis-
terminate structure: displacement tribution in the !tout" structure.
at redundant cut r (s) fer a unit
By way of illustration, the r ina l result
ac-n
-- i ec load, Pn = 1 (P = I), ~ for Example ?roblem 3, Ayt. AS.5 is expressed
(QrI:t = a nr) below. FIRST, in the r cru of eq , (5):
a~s - l~fluence coefficient fer the de- True Stresses = S + X
ter-aune t e (vcut") structure: dis- (u,) + Y (U )
y
placement at red~dant cut r for a
~~it ~edundant force q = 1. AE o + 521 (.806 ) + 416 (l.154)
(a
r-s
- a
sr-
) s BE = 2000 + 821 (-l.564 ) + 416 (-l.729)
A:nn influence coefficient for complete 0_
co o+ 521 (l.00) + 416 (0 )
Yedundant ctr~ct~e: ~eflect~on
at externa: lO~dl~g ~oint n for a DE 0 + 521 (0 ) + 416 (l.00)
unt t applied load, ':l = L
n • Note that within each of the sets of subscript symbols (I, j),
, (.~
:::''1
= Po. :',;::J. ) (r, s}, (m, n) the symbols may be used interchangeably.
AS. IS STATICALLY INDETERMINATE STRUCTURES

SECOND, in the form of eq , (14): of external-point ir.fluence 80efficients in the

~cut~ struct~e: deflection at point n per ~~it
load at point n.

- the :r..a -n-tx

of in~luence coeff:c~ents ~el~ting rela~iye dis-
placements at tne "cut s" :0 external loads:
Note that in this case [gim] consisted of only dlsplaceme~t at cut r per unit lead at pOint ~.

one column, inasmuch as there was only a Single (18)

external load.
In Art. A7.9 the strain energy was written - the :::a.trix
of influence coefficients relating relative dis-
(15 ) placements at the "cu~s" to ~9dundant :oads at
the Wcuts": displace~ent at cut r per unit re-
dundant force at cut s.
where [OiJJ is the matrix of member flexibility With the above notation one may Write
coefficients (Art. A7.10). If now eq. (14) and
its transpose are used to substitute into (15)
the expreSSicn becomes (note the use of (i,J),
(r,s), and (m,n) interchangeably)
(19)
2U = (LPmJ [~iJ L~J [g,.iJ)X
+

Now according to the T~eorem of Least ~ork

x ~iLJ X([!j~{qs}+ [!j~Wn}} aU/a qr = 0 for continUity. Then, differentiating
eq, (19):
Multiplying out

This last result may be verified by writing

eq. (19) out in expanded form, differentiating
and then recombining in matrix form. Rearranged,
sq. (20) gives

(21)

The reader may satisry himself that the Eq. (21) is a set of simultaneous equations
~crass praduct~ term in the ~ldd1e of the above for the redundant internal forces ~,qs' It
result is correct by Qbserving that, because of :nay be compared With sq. (6) of Art. ,1.8.6, to
the symmetry at [u1J
], which it corresponds. Eq. (21) ~y be solved
directly from the form there displayed or its
solution may be obtained by computing ~rs:JJ
the inverse of the matriX of coefficients, giVing

{qs}= - ~rs~ ~rJ{Pn} (22)

The various ~atrix triple products occurring The matrix product - fa -~ ru -,
L.:rs....J L..:rrrJ
gives the
above are assigned the following symbOls, each va Iues of the redundant forces ~cr unit values
having the interpretation given (compare With of the- external loads. This may be given the
eq. (24) of Art. A7.9)
sy:nbol ~s~ so that
- - - - - -(16)
- the matrix
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AB.19

If :lOW eq. (22) is substituted into sq.

(14) one gets (With exchanse of r ~or s ~d
a
L /3 Liz a a
a rcr n)
L/2 1 a 0

{qi} = ~i~ {p,,}- ~iJ ~sr:J ~s;::Wm} CiJ E:I

L
a
0
a
a
V'/3
Liz
Liz
1

(~~ - ~iJ ~s[] ~sJ ) {Pm} The moments q~ and q. were taken as the redund-
ants. With these set equal to zero, the internal
force distributions due to application of unit
= (~iJ+ ~iJ l:rmJ ){Pm
} values of P l and P a were determined, giving

The matrix set of~ in parentheses above, ilL a

gives the internal force distribution per unit
value of the external loads. It is given the a a
symbol
~~ =
a ilL
(23) a a
With the applied loads set equal to zero, unit
values of the redundants were applied yielding
(r =) (2 ) (4)
so that
- ilL a
(24)
1 a
Eqs. (23), (24) constitute the major re-
~!J =
ilL -11L
sult, inasmuch as they present the ~eans for a 1
computing the internal force distrib~tion in a
redundant structure.
Note that redundant load qa was applied as a
self-equilibrating internal couple, acting on
Example Problem 13
The doubly redundant beam at Fig. A8.27 (a) both beam halfs.
----- The following matrix prOducts were formed:
i5 to be analyzed for the bending ~oment dis-

r I" 'J
tribution. The beam is loaded by couples over
the supports as shown.
[SJ . [!r!J ~ti] ~J !J

EI constant ~P2 -Pi ., L'" · ., Liz

!..Iz 0

..!
:.1 0 0 -liL 1 1 0
~
f--- L L
1
----1
('I 0

0
0

0
L-/3
L/,
L:, ~L -~L
~~
q2( II
q, } (hi
L
: 6E:.1
q,( Ii
q,
Fig. AB.27
rrnJ • [!r!J rti] ~J;]

-r. J["
I' .:J
So Lut Lbn: L/, 0

£! 0 0i _lit.
." 1 Liz 1 0
The choice of internal ~eneralized forces
is shown in Fig. A2.Z7 (b) .. T~e appropriate . 0 0 1.-/3 Liz 0
~ember fleXibility coefficients were arranged in
0 0 L/2 1 0
~trix :or~ as (ref. Art. A7.l0 :or coefficient

~ ~
expressions) . L 1 2
..
: mOl ... -
~
AB.20 STATICALLY INDETERML.,*ATE STRUCTURES

The inverse of Cars] was fOillld (ref. though the ma'tr-tx of member- aexi'Ji2.1 ty cce-
appendix) fflclents was expanded to a 6 x 6, the
coefficients for q, and q, were zero. ThUS,

~rs ~
II _J. _ 5EI
- 7[ C -J
2
-1 4 VI/3

L/2
L/2

1
a
0
0

a
0

c
0

0
_ L
~lLJ
Next, the unit redundant load distribution
was found (eq. 22). - ;;:1 a a La/3 Liz 0 0

0 a L/2 , 0 0
1
"
- '7 0 0 0 r! 0 0
0 0 0 0 0 0

= r-. 286 -.428]

-.28~
With the redundants q~ and qa set equal to
L·I43 zero, successive applications of llii!t external
couples PJ. and Pa gave the stress distribution
Finally, the true unit stress distribution
was computed. 0 a
1 0
a 0
[!1;] 1 1

L· J
1 0 _1/ 0 286 -.42
L 0 0
0 0 1 0 .143 -.286 0 0
= 1 •
L 0 1 I/L - I/L
and, with PJ. and P a zero, successive applica-
0 0 0 0 tions of unit redundant forces q$ ~id q~ gave

-1 0

0'\ 1
1

1.286
2

.428 -I
L
-I
0

1
=L 2 - .286L -.428L ~lJ = 2L L
3 - .429 .858 1 0
4- .143L -.286L La I

(The tabular form at presentation of the The~, multiplying out per eqs. (17) and (18)
matrix Gim, above, is used here only to indicate

~ J
clearly the functioning of the subscript nota- L'
tional scheme. In general, it should be unnec- ["r;;] 2E1
essary to callout the subscripts in this
fashion excepting for the larger matrices, tor
the handling of WhiCh, the tabul~r form may
prove helpful.)
Example Frob lem 13a
~r~
L'
6Ei r 6
~5
51
sl
The redundant beam problem of Fig. A8.27 is The inverse was found:
to be re-solved using the redundant reactions as
the unknowns.
6EI -5]
Solution: 7L' l~
The support reactions under the loads PJ.
and P a were given ·the symbols qe and qe, respec- Finally,
tlvely, Positive up. These forces did not enter
into the strain energy expression so that, a i- rG l r:. -, ~ -, ~ ;:'l r;; -,
~
L:1m ~1~ - ~:CJ ~sr~ ~s~
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A8.21

1.286 •.l25
- .':::;CJ.., -.423L
1 - .'+<:.c .857
=L
.143L -.256L
-1.256 -.428 Loading for column {gi2} of the matrix [gim]
1.714 -.428 Fil{. AB.29

(COm'Cl2.r':! 'fIith s ojut i on to Example Problem 13).

~ample Problem 14 .5 1# .5 ~,,~

The oontinuo~s tr~ss of Fig. A8.28 is
t.vr cc redundant. :t is desired. to analyze it
fa~ stress distributions under a variety of
loading conditions consisting of concentrated
/\t:it:J%f2t\l\/V\
I t I
vertical loads applied at the four external Fig. AB.30a {gi3} loading
points ~ndic2ted.

'R2
8 panels at 20" :: 160"

~
Fig. A8.28
3 5

Solution: 20 1 0 0
20 2 .0 0
The internal generalized forces (qi,G ) 20 3 1.0 0
j 20 4 .5 .5
employed the axial loads in the various
~ere 20 0 0 1.0
~eabers. These ~ere numbered from one to 20 6 0 .5
thirty-one as ShC~~l on the fl~1re. The member 20 7 0 0
fle:Krbility 80efficients in this case 'Nere of 40 a 0 0
the form c , = L/AE (Rei'. Fig. A7.35a). The 40 9 -.25 0
• 40 10 -.75 0
coeffic1ents arf written as a column matrix be- 40 11 o~
-. , 0 -.25
low. (They 'NE::ce employed as the diagonal e le- 40 12 -.25 -.75
ments of a square ~~rix in the matrix multi- 40 13 0 -.75
plications, but ars written here as a column to
~1~
1 40 14 0 -.25
conSer.e space.) a1 =~
40 10 0 0
Membe~ loads qa and qo were selected as 22.4- 16 0 0
redundants. ~ith qa and Qe set ':!qual to zero 22.4 1" _.
0 0
(''But''), un t t loads were applied succ ess ive Iv 22.4- 18 I .56 0
at external loading paints one through four, 22.4 19 -.56 0
t~e four stress distributions-thus fo~nd being 22.4 20 • 561 0
arranged in four columns giving the matrix 22.4. 21 -.56 0
~i~ (below). By way of illustration. the 22.4- 22 -.56 .56
22.4 23 .56 -.56
loading fi~ure used to obtain the second column 22.4 24- -.56 .56
22.4 25 .56 -.,:)6
of ~i;] 1s shown in Fig. A8.29. 22.4-
m"
25
27
v
0
- -.56
.00

Next, unit forces were applied successively 22.4 29 0 -.00

at the redundant cuts "three" and "rtve'' as shown 22.~ 29 0 .56

in Figs. A8.30a and A8.30b. ~hese loads were 22.4- 30 0 0
arranged ~n two columns to give the ~trlx ~i£J 22.4 31 0 0

•
AS. 22 STATICALLY INDETERMll'{ATE STRUCTURES

I~
05 9 .111 .0065 .0035']
1 2 3 4
[ .0035 .0065 .111 .059J
1 1.0
2 .50 -.50
3 The calculation was completed as per eq.
4
(23) to give [qim] , the values of the member
0
-.50 .50 forces for unit applied external loads.
7 1.0
8 -- .50
9
10
11
12
.70
- .25
.25
.25 Ix 1 1.0
1 2

0
3

0 I 0
4
,
I
2 .529 -.445 .003 .001
13 .25 - .25
3 .059 .111
14 .25 - .75
4 .031 .059
.006
.059
.003
15
16 l.12
- .50
5 .003 .006 .111
.031
.059
17 '6 .001 .003 -.445 •.529
-1.12
18
19
- .56
.06
-.56
.56
7
8
0
- .50
0
0
0
0
1.0
0
20 - .56 .56 9 - .764 .222 -.001 0

-- .017
2 .56 -.56 I
10 - .294- .167 -.003 .002
22 11 - .045 -.085 -.032
23 12 - .017 -.032 -.oas - .04.5
24 I 13 - .002 -.003 .167 - "'G'
.':::',,7

25 I
14 0 -.001 .222 - . , ::" ...."
.....

26 -.56 .56 5 0 0 0 I - 0.00

27
~a I
.56
.56
- .56
.56
16
17
l.12
-1.12
0
0
0
0 0
29 -.56 - .56 18 - .527 -.498 .003 .001
30 -1.12 19 .527 .498 -.003 - .001
31 l.12 20 - .527 I -.622 .003 .001
21 .527 .622 -.003 - .001
NOTE: VOIDS INDICATE ZEROES 22 - .031 -.056 .058 .031
23
24
- .031
.031
.058
-.058
-.058
.058
- .031
Multiplying out gave .031
25 - .031 .058 -.058 - .031
-a. ill
26 - .001 -.003 -.622 .527
[136 27 .001 .00:3 .522 - .52""
L:a.o 136J 28 - .001 -.003 -.498 .527
29 .001 .003 .498 - .527
30 0 0 0 -1.12
31 0 0 0 1.12

Example Problem 14a

L~'O
1 -15.0
E 0
The inverse ot ~r~ was found (ret.
appendix)
-~.J
0 Fig. AS.31 shows the two bays ot a steel
-15.0 tubular tail fuselage truss Which is loaded by
tail air loads to be resolved into t~ree :on-
centrated loads applied as shown. The fuselage
bulkhead at the attach-paints station (A-E-F-K)
is heavy enough so that it may be assumed to be
rigid in its own ~lane. Hence, the truss may
[136
~.o
a·1ll be analyzed as if cantilevered from A-E-F-K as
shown. All members are steel tUbes, their
13~ lengths and areas being tabulated below.

Next, the values of the redundant forces, Solution:

tor unit values of the applied loads, were
found per eq, (22). The generalized forces were taken to be
the member axial loads, these being numbered as
in the table below. Member fleXibility coef:ic-
ients, L/A, (E set equal to unity for canven-
ience) were also tabUlated.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A8 23

than apply successive unit loads and forceS at

points m = 1, 2, 3 and r = 22, 23, 24 and then
carry each loading through the structure, an-
other procedure, often better adapted to large
complex structures, was employed.
In the method used, the 9quations of static
equilibrium were written for eaCh of the seven
Joints.* Summation of forces in tr~ee directions
on each jOint gave 21 equations in 24 q's (the
unknowns) and the tr~ee applied loads P~, P3 and
P~. Some typical equations obtained were:

1rom Z F on Joint C.

.213'. q . . . . 213'. q . . . . 18334 q • • . ,:334 q. ' -Po

.f.
A1~"~ PI q••
A 0 I P2 {

~
..... B
"' " /
C
::I....... '" -1.0236 P,

~ K
,

__
-,
-, ~~,
I"~
"I ......" P
3
From Z F on Joint B.
II- J--:::--
--
- -'.---
G
H
F --- .081759 Cl.1. - •18334 Cl.. - q. -.07617 Cl.1.. '" .5227 q•
Fig. A8.31
q•• - .91593 Cl.u , 0
{
-.1410 q1.'" '" _.8523 Cl. ..

~ER N"lJ11BER L2NGTH AREA L/ A And so forth, for the other Joints.
Note that in each case the equations were
cli 1 35.47 .565 52.8 arranged with the applied loads (P ) and the
RC 2 25.43 .49~ 51.0 n
BD 3 9.20 .165 .-
00.""~
redundant q's (q33, q33, q.~) grouped on the
CD 4 25.43 .499 51.0 right hand side at the equal Sign. This ar-
DE 5 35.47 .565 62.8 rangement was observed for all 21 equations,
?G 6 35.47 .565 62.3 atter which the equations were placed in matrix
GH 7 25.34 .555 44.85 form as
HI 5 5.00 .165 30.3
IJ 9 25.34 .565 44.85
GJ 10 9.20 .165 50.8
J"
AG
'1
12
35.47
40.4
.565
.630
62.8
64.1
BG 13 15.0 .165 90.9
i,j = 1,2, •.••••. 24
3H 14- 27.57 .500 55.14-
He 5 11. '70 .395 29.62 n = 1,2,3
Ie i5 11.70 .395 29;52
s = 22,23,24
Dr 17 27.57 .500 55. "4-
JJ ,Q 15.0 .160 90.9
SJ ..,
;
40.4- .630 64.1.
(Note that there were 24 equations ~ere,
the adcitiona1 truee equations being the ident-
20 37.36 .565
.' J
FJ
"~.
-."
22
37.36
26.23
.565
.500
66.1
66.1
52."~6
::.ties

sa 23 17.60 'Q.
.... ..,0 1.06.6
".
.U 2.:1 ,-.60 ; .-
• ....0 0 lOE.S

The structure was three ti~es redundant.

I~ a space rra~ework of p jOints, 3p-6 !nde- On the right hand side of the above matrix
pence~t equations ot statics nay 88 wT~tten equation the ~trices are shown ~partitioned~.
(p. A8.10). Here, ~oweverJ stress cetails in T~e ~irst three columnS at [OJ are the coeffic-
the ~l~ne AEKF are t8 be sacrl!iCed, six equa- Lerrt s of P and the last 'thr-ee are the
tions are lost ~hereby si~ce onlY net ~orces in n
two ~:rec~ions in :his ~lane can Je sUT~ed.
3 x 11- 6 - 6 =21 equations, 24 member' unknown.
~e~bers 22, 23 and 24 ~ere cut. • For structures other than trusses the equilibrium equations
The next steQ Nas to comnute the unit are written for the various structural elements, equilibrium
stre~s ~istri~uti~ns [glm] and [~irJ' Rather of joints alone being inappropriate.
 •
A8.24 STATICALLY INDETERMINATE STRUCTURES

coefficients of the q_. The inverse was rocnc (see appenc Lx)
o
The ::latrix equation NaB solved ~or t~e q.
by finding the " _. [5555
- .2580 l
'2~7~
Thus,
lnver~e of [C
1 j]
(see appendix).
trs-J 10 -.2550
.2775
3.C~12

2.3015
-~ '2:" -
<::'..... ,-,·1.:)

3.023

T~en (referring to eqs. (22) and (22a))

~
where .18 2259
IG "I =_Ia
LSllJ - LTS -'J LTllJ
ICi "I = 10 -~
-be."=,
","
-238 '"'n .".
-.<:v1.v

- 31A 121.~

?lnally. the complete unit stress distrlbu~~on

ThUS, the ~~lt stress distributions in the
1II8.S obta Ined as
determinate structure were found by a procedure
having as its ~ln advantage the reduction of
the work to a routine mathematical operation.
In the conduct of this work appropriate stand-
ardized techniques may be employed.·
The result in this case was
X 1 2 I, 3

1;\ 1 a
2~727 5.;'7~
3
I Z2

,
Zl
I ~
o
1
2 -
.8412
.00017
1~...,.=::

2.203
.~IV 1.296
.~38
1 .5415 -2.:!C59 0
-
, · " ..
.. · ..
.:).5<> 0 3 .01,523 .31.53 .1190 -
zr-
.' 0

·.
.,. ~:: . o
,- .0
4 .00017 -2.206 .E;39 -
_2.72:7 ,:;,,60
.. .. - -- o

:~
0
4.0.59 _7.~ e •..;..,5 5 .6529 -3.008 i -1. 557
, s s ,
-c . ."'~
.."' 6 -1.4-04 1.5247 i .332
.,
.. · .. ,
,
, 7 - .8443 .7352 I .761 -
. ·
,.
• .06<:
· .. as ·-3.n4
.1 ..l" - ...17
.0 0 8 .1094 - .095<::'. I .096::.
· - .0= -
2.," .. ,
9 .84.50 -1.260 -1. -t-50
· ..
, . -i · •.:.009

" - .aiee
.4415

.5 ,5 ,
.54.
."'"
0 ..
· .01<+5
00
.eess
-r.v-a
, ,
-1. ..35
c
0
·
r..4J.S
.3523
0
,
10
~1 -
.0973
.0494
.0205
-1.607 I -l.S'~2
.3338 -
,.·
_2.340
.zao
-.... 5
-
...
.&C
.22
0
0 12 .6638 - .3896
.3143 .733
. . .W
·1.W9.
·
-
,-
'."
.:...,27 · .o se 13 - .2350 - .2641

~
.27E"
,
z . , -L'" .0

,·
.2445
."S06
1..;,01
-'-.302
, _,.~l

1.102
14
15 -
.9185
.5116
- .7998
. , -
445-
.6223
.461
oe ·
0
o
•• 0
0
.~O'"

1.0
.
0 16
17
- .5120
.9193
- .4457 I
.B002 I -
.461
.829
18 - .1061 - .3638 .3205
The member flexibility coefficients L/ A 19
20
.1434
.2023
.2411
1.6471
- .6956
.8512
were arranged as the diagonal elements of the
matrix [u 1j ] Then, multiplying out according
21
22
- .2660
.0002
- .3'796
.543 i
.556e
2.269
,-
23 - .1364- .238 - .2010
to eqa , (17) and (18), 24 - .03 4 .176 .122.3

[177,9
-1137.0
-6430'~ ~ar.mle Problem 15
The doubly symmetric four flange idealized
~~
195.0 - 151. 7 -2263.0
box beam cf Fig. A8.32 is to oe ~n~lyzed for
-154.4 161.6 2273.3 stresses due to load applicat:on ~t the six
paints i~QicateQ. Fl~~ge areas· taper linearly
from root to ~1p while sheet thicknesses are
-114~
[ 2970 1150 constant in ~ac~ panel. The bea~ :s ~o~ted
rl~1dly at t~e ~oot, provid1ns ~u:l restra~nt
~r;] 1150 1221 -1035
-1148 -1035 1224-
• These are the "effective areas", being the flange area plus
• see references in appendix. adjacent effective cover sheet area plus one- sixth of the
web area. (The factor of one-eixm provides the same
moment oC inertia as the distributed web area).
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A8.25

against war?ing of the root CT2SS section ~ue The ~e~ber :lexlbillty coefficients were
to any torsion loadings. collected :n ~trix :crm as shown below. Note
t~at entries for u," U e e • Q v v and u~o,~o were
collected irom ~NO stringers each (as well as
being doubled as discussed above). Coefficients
for these tapered stringers were computed from
the formulas of Ar~. A7.l0.
1)< , ! a IaI , ,
, s I • . i a I 10 U !Z I IJ I
" ,! ie ,
, iaa
,
,
' I
, ,
,

,
" ace
.. ..
··
,-.o.~

,
. · ..
,
c.
,
.
.

i
.
15.2.il ::7 .5
" , .~ , .,

NOTE: VOIDS INDICATE ZEROES, E =1

Cover sheet shear flows q~, q7 and ql~ were
selected as redundants. ~ith these set equal to
Fig. A8.32
zero, and with unit loads placed successively at
loading points ~one~ trxough ~sixw, the [gim]
matrix was obtai~ed. ~~en the cover sheets are
Solution: ~cut~ the ~NO webs act independently as plane-
web beam5. The details of the stress calculation
The generalized forces employed are shown for such beams are similar to those of Ixamp.Le
on the exp Ioded View. Fig. A3.33. No forces Problem 21, Art. A7.7 and are not shown here.
were shown :or the laHer surface, its ~embers
and forces being equal to those of the upper
I~ 1 2 3 4 S 5
surface because of symmetry. In the [u 1j ] 1 ,400

matrix this fact was accounted fer by doubling -

3 •.\00
the ~ember fleXibility coefficients for the 4 4.
4,
members of the upper surface. 0
.200

8 .00 ,200
'\.'1:" ,Qle 9 5.341 2.571
10 5.341 2.571

~~ \~
.08_87
12
13 ,0444S .08887 ,1333
14 6.231 4.006 2.003
IS 5, 1 4.006 2.003

NOTE: VOIDS INDICATE ZEROES,

Calculations for [glr] were ~de by suc-
ceSSively assigning unit values to the redunc-

,'.~~,
ants q s, q.., and q u. The calculations are
illustrated by the exploded view of the end bay
in Fig. A8.34 showing the calculation in that
part of the structure for q~ = 1. Note that
q:z = 1 was applied as a self-equilibrating pair
of shear flows acting one on each side of the
q, "cut;". The r-ibs were considered rigid in their
own planes.
Fig. A8.33
?ut q~ =1
I From equilibrium of end r t o : (1: M, Z F)
A8.26 ,
STATICALLY INDETERMINATE STRUCTURES

q ",0 T:-.8 f'ollcw1r:g matrix products were ::om.ed:

I~ .;'" ~q
.~E;--'-<-S\l25
Per eo.. (18) :
•
q
.1769 .0520j
[774
\q _ lOll
~r~
,40",-25.04 .75x~=.5625 25.04 .1789 .2578 .07322

..
IS"
- E
.0520 .07322 .1254-

~\~
2.5 j 15
Per eq. (17) :

_.ro:l56 ._, •• 009318 .009318 _.001605 .001605

lX20",·75
-"
~;]
urn : 10'
T -.= .ceeec -.OU56 .OllS6 _.C02Z59 .~2259

-IX~~
_.008216 ,008216 -.005150 .005150 _.002278 .002278

5.0 q." '\

f q,-I
The inverse of Cars] was tOlUld (ref.
appendix) •
fq,=_1
q,..= 1
•
-2.562
q,. -11Cj!.__. __._...__..':;S q3 " 1 [3,862 - .113J

Fig. AS. 34
e-: = E
10 IS
-2.562
- .1137
6.137
-2.321
-2.521
9.499

Then
q, = -1.0
q, = 1.0 ~sJ =- ~rs -j ~r;;]
From eq'.l111brlum of 200 rib:
[0699 -.0699 .00598 -.00598 .00016 -.00016
q. = -t .5625 - .25) = -.3125
= .03593 -.03593 .03409 -.03409 .00403 -.0040
q. = .3125
.01568 -.01568 .01872 -.01872 .01578 -.0157
q. = 0 , by hypothesiS
Finally, the true stresses were (per eq.
So on, into the next bay. In the idealization
used here l the ribs have zero stitfness normal
to their own plane so that the axial flange-
loads are transmitted directly to the flange
~i~ = ~i~ - ~1J ~~
ends ot the adjacent bay (see q~, q" q .. q"
of Fig. A8.33). I~ 1 2 3 4 5 5

1 .330 .0699 .0060 - .0050 -.0002 .0002

-- .0060
I~ 2 7 • 699 - .0699 .0060 .0060 .0002 -.JCO 2
12 3 .0699 .330 .0060 .:JOC2 -.00'22
4 2. 6 1.75 - .150 .150 -.0040 .00'+ o
1
2
- 1.00
1.00
0 .75 2.26 .150 - .150 I .00401-.004.o
6 .08 9 .0141 .16B .0322 -.0040 . 0:40
3 1.00 .0 9 - .0359 .0341 - •C341 .0040 -.' 040 .
4 -25.04 ~
• 141 .0359 .0322 .168 .0040 -.004-o
5
6
25.04
.3125 - 1.00 0 ·
2.51
.0
2.83
1.w45
1.126
1.16 - ....22
1.5<t.5
. 22
.122 -.122
7 1.00 1
· 4 - .0010 .0776 .0112 -.0150 .015 o
~iJ = 8
9
--20.87
.3125 l.00
-29.38
12
3
.0157 - .0157
- .0010 .0454
.0187
.0112
- .0187 .0158 -.015 8
.0776 .0150 -.015 o
4 3.47 2.76 2.40 1.61 .610 "-.610
10 20.87 29.38 15 2.76 3.47 1.61 2.40 -.610 .610
11 .1387 .1955 - LOO ~

12 1.00 The reader will observe that the result

13
14
--18.79
.1387 --25.44
.1955 LOO displays the ~bending stresses due to torsion~
-31. 71 that Is, the bUildUp of axial flange stresses
15 18.79 26.44 31.71 near the root of a beam under torsion when the
root is restrained against warp1r.g. The solu-
NOTE: VOIDS INDICATE ZEROES. tion for application of a torque is readily
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AS 27

found by superposing stresses for P~ = 1 and

-1~J
P a = -1. Thus, one :inds that under this con-
44 -15.0 o
dition there 1s a root flange load of =; -15.0 38.0 o
·00 38.0
= 3.47 = 0.71 [
ql. - 2.76 Ibs. o 0 -15.0 144J
for a torque of 15 inch Ibs.
472 .888 .0525 .028
AS.11 Redundant Problem Deflection Calculations by •388 1. 665 .0975 . 0525
Matrix Methods.
Datlections (and in particular, the matrix
of influence coefficients) are readily computed
from the results of Art. A8.10.
Assume that the redundant forces { have qs}
[
~0525

.028
.0975 1.665
.0525 .888
.888
.472
J
Finally, the sum of the above eNO matrices gave
been det.er-atned from eq . (22). The total de-

~'#
~lection of external loading points is then -15.9 - .052
eaSily computed as the sum of, 01~, the de-
flection due to external loads acting on the
"cut;" structure, [amnJ {P n } (ref. eq , 16) and,
TNO, the deflections due to the redundant
~,,;;] ,. ~
- 15.9
-
-
.052
.028
-
-
36.3
.098
.052
-
-15.9
.098
36.3 -"J
-
-15.9
144
.052

forces acting on this same cut structure, Example Problem 17

[a.:nsJ {qs} (ref. eq , 17). Thus, Determine the matrix of influence coefficients
for the box beam of example problem 15.

Solution:

An alternate procedure to that shown by

eq. (25) was followed. The influence coeffic-
SUbstituting from eq. (22) ient matrix was formed as in Chapter A-7, Art.
A7.9, by the product.
{ 6,. } = ~:nnJ {n}- ~msJ ~rsJ ~~ {q
P

= ( ~mnJ - ~:nsJ ~r(] ~r;] ) {pn} This product was formed readily, inasmuch as
[G ] was evai tame tram example problem 15.
im
The matrix expression set off in paren-
theses above, giving as it does the deflections The result was
for unit values of the applied loadS, is the
matrix of influence coefficients. Let
----
4363
3500
3500
4363
1522
l243
1243
1522
29.7
-29.7 29.7
-29.71
1522 1243 897.8 594.2 33.7 -33.7
1243 1522 594.2 897.8 -33.7 33.7
- • 33.7 - 33.7 .8 -36.8
9.7 - 33.7 33.7 -36.8 36.8 1
so that
AS.12 Precision and Accuracy in Redundant Stress
- - - - - - - - - - - - (26) Calculations.
Matters of prectsion are dependent upon the
number of Significant figures obtained and re-
~ample Problem 16 tained in dealing with the geometry of the
Deter.nlne the matrix of influence coeffic- structure and in the care with which arithmetic
ients for the redundant truss of example oDerations are perfo~ed. In the discussion to
problem 14. follow it 1s assumed that all due caution 1s
exercised with regard to the preCision of the
Solution: work.
~~t~ers of accuracy'r4ve to 10 with the num-
From the previous work the following ~trix ber of Significant figures :lnally obtained in
~roducts wsr9 fo~ed the answer as influenced by the ~er of formu
A8 28 STATICALLY INDETERMINATE STRUCTURES

lation of the ?roblem. ~~e accuracy ~f the ~e­ inverse ~ar~ot be ro~c (r sr , appendix).
sult may be affected by a n~~ber 0: factors, two condition is terrib12. :::xampl-a:
of the most important of wh i ch are discussed
here.
Two factors infl~encin~ accuracy often are
considered together under the heading ~choice of
redundants n • They are:

.~- the number of Significant ~lgures

which may be retained in the solution for the
inverse of the redundant force :oefficient TIa-
tr-ix , Cars] (eq.21).
•• - the magnitude of the r~dunGant ~orces
relative to the size of the dete~,lnate forces.
These ~NO factors are concerned respective-
ly With the left hand side and the ri~ht hand
side of eq , 21, 'Viz. '--
In the matir-tx
~ents are measures o~ the str~ctural cross-
cou~l1n; o~ one red~~dant force with a~o:~er.
~he st~ength of t~is cross-coupling is de~end­
ent upon the choice of red~~dants made ~~
"cu't t tng" 'the s tr-uc tur'e to make it s ta t tca Lj.y
determinate.
JOR EXAMPLE, the do~bly reduncant bea~ of
Fig. A8.35(a) ~y be ~de statically 1eter~in~te
by "cutting n ~y two constraints.

they are discussed :n detail below. 5':::::::::::-:

~
==~
(c)

• - ACCURACY OF INVERSION OF ri
Lr~
:l ; (bl
~
(dl
THE CONDITION OF THE nATRIX.
Fig. AB.35
The characteristic of the mat-~x
.. ~ [0 rs ]
which determines the accuracy with which its in- Fig. AB.35(b) shows the c~oice o~ general-
verse can be computed is its condition. The ized torces. Only two (Ql and qa) a~Q required
condition of the matrix is an indication of the to describe the strain energy, but the central
magnitude of elements otf the main diagonal support react1or~ were also given s~bols as it
(upper left to lower r1&~t) relative to those was desired to consider them in the dis~ussion.
on. The smaller are the relative sizes of Then
elements off the ~in diagonal, the better is
the condition of the matrix. A weJl conditioned c-
2/3 1/6 0 0
-
matrix is more accurately inverted than a poorly
conditioned one. 'rNo extreme cases are now
given tor illustration: 1/6 2/3 0 0
a) the diagonal matrix. Its ott-diagcnal C:I1J =..1.EI 0 0 0 0
elements are zero, so that it is ideally con- 0
ditioned. Thus. the inverse of
,-0 0 2..
FIRST, suppose the beam was Iade deter.ninate by
selecting the support reactions q~ and q~ as
redun1ants. The ~cut structure~ in this case
may be visualized as the Jeam of Fig. A8.35(c)
whose central supports have been removed.
Application of unit redundant ~orces q3 = I
Is eaSily and accurately obtained as and q .. = 1 gave

~ 3 4

2 1
1 - 3 L - :3 L

b) a matriX all of whose elements are equal

~1~ = 0
<. - 1
;j L - 2
:3 L

in each row. All the elements cff the main 3 1 0

diagonal are equal to those on. --rhe dete~in­
ant of such a ~trix is zero and nence its 4 0 1
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A8.29

~ulti~lYi~g out, r'edundant s may be chosen simply. In most struc-

tures, however, the additional labor involved in
seeking an ortho~onal set of redundants is not
L'~.
laEILz 71
~
8
warranted.
(2) In choosin~ a set of redundants which
will yield a well-conditioned redundant matrix,
r,s=3,4 it is best to make such "cuts" as will leave a
statically determinate structure retaining as
The condition of this matrix is poor. many of the characteristics of the original
pnys i caj jv , a unit load at po trrt "3" causes structure as possible. ThUS, one may consider
a Imos't as much cer rec t ron at point "4" as at "3 11 that the structure of Fig. A8.35(d) retains more
itself. The crosS-coupling is large. of the features of the original continuous-beam
structure than 10es that of Fig. A8.35(c).
SECO~TI, suppose the moments ql and q~ had
been chosen as ::-edundants. 'rne "cut structurel! (3) The degree to Which one redundant in-
in this case being visualized as in Fig. fluences another (extent ot cross-coupling) can
A8.35(d). Application of unit redundant forces be visualized by observing how much their indi-
q. = 1 and qa = 1 gave vidual unit-load diagrams "overlapH.
To elaborate on this last point, refer once
again to the above illustrative example. The
I~ 1 2 cross-coupling of q3 with q~ may be expected to
be large if their unit-load diagrams are drawn as
1 1 a below in Fig. A8.36(a). This deduction follows
2 a 1
eaSily if it is recalled that the dummy-unit load
equation for such a cross-coupling term is at
2 1
3 -I: I: the form Jm~.:n~ ex- ,obviouslY
EI 1"
Lar-ge .. or m~,
1 2
4
L -I: m.. Strictly, the comparison is with the terms

tIultiply1rlg out,
:0:
J -W-'
dx Im:EIdx .wh'... ch rorm
' thie on-a..
.,. agcna.J..
,

elements of the ~trix [urs]

r,s=l,2
m,la
The condition of this redundant ~~trix is
obviously better than that obtained with the m
z[ ~
first choice of redundants. There is less
cross-coupling between the redundant forces. (b)

Fig. AB.36
Thus the analyst, by choice of redundants, determines
the condition of the matrix. The choice may be critical in the
case of a highly redundant structure, for it may prove im-
possible to invert a large, ill-conditioned matrix with the St~dy of t~e ~it-load j1agr~~ fo~ ~he
l1.mited number of significant figures available from the initiaJ. redundant choice ql' q~ (Fig. A5.35b) reveals
data. The following statements and rules-of-thumb may be that the cross-coupling sh~uld 88 small her~
useful in the treatment of highly redundant problems.
since an 1rrteetr-a lJ ':J".......
v • .9 r' orm J:nlffi'EIl ex can
(1) It is always possible to rind a set of ~ave a contribution from the center s~an O~ly.

r-
redundants for which tIle cross-co'J.pl1m; is~t a Hence, the J ::I1.:n'l, ct.:.:: is obvt cus Ly c ons tder-ab Ly
~inimum-zero, ~n tact. Theoretically t~en, by 0'
~roper choice of redundants, tte ~a~r!x [Q ]
rs sma11 er ·h
cnan 31ex or J r ::I~
~I ex Th
.. us, a
~y ~e reduced to a diagonal ~trix (idea~ly Visual ~nspect1on of Fig. A.36 reveals that q1.,
conditioned) . qa is a better choice of redundants ~han is
(la) The choice of red~~cants which gives q, c...
zero cr oss-c oupt inz ("orthog:Jnal func-O:ions") 1s
no'\: reaai i .' / found in :::ene p :;' . I!1 some special
structures, such as r tngs anc rr-ames, orthogonal .. see eq. (8) Art, A8.7.
AB.30 S!ATICALLY INDETERMINATE STRUCTURES

(4) Combinations of ~ed'~dants ~ay be ~m­

played to yield new '~it-reduncant stress dis-
such a way that the ~rOduct [gir]
ro
] does [a
tributions which do not "overlap" as extensively exactly 'Nt-at is jesi~ed, 7iZ., "t~ke Qn~ ti~es
as do those of the individual red~dants origin- the q3 coll~~ ~l~us one halt the q~ ccl~~ to
ally chosen. give the first column of the new distribution
For ex~ple, suppose that in the previous
illustrative problem, the choice q~, Q~ for re- [giPJ ". Then "t-ake cmua one half of the q3
dundants had been made originally, leading to
the unit-load dia~acs of Fig. A8.37(a). In- column plus one times ~he q~ column to give the
spection of the diagra~s leads one to anticipate second column of the new distribution [giO]
a stror~ degree of cross-coupling and hence a
new set of redundants is sought. Rather than Multiplying out in the above example
return to the s'tz-uc tur-e to choose new "cuts",
combinations of the m3 and m~ diagrams are
looked tor ~hich will have less "overlap" and
hence less cross-coupling.
It is observed by inspection that two new
stress dlstributions"'which have the desired
~
- 2 L

0
0

1
- 2 L
l
property may be formed trom the m~, m~diagramB
by proper combination. Thus, if one halt the
1
1
C: eJ
1
m" diagram is subtracted rr-cm the --m:
d'Iagrarnto - 2
fOrm one newstress distrIbUtion and one halt 1
the ~ diagram ~ subtracted from the m~ dia- - 2 1
gram to form the other new stress distribution,
the results are as shown in Fig. A.37b. There
ta. oovtous jy less "overlap" of the diagrams for Now form the ~trix of redundant coefficients
these new combir.ations. for the ~ew unknowns (subscripts p, 0; [giP] =
[gj,,] )

L'
24"::' .

Fig. AB.37
The condition of this ~trix is ~eatly
New unit-redundant-force stress distributions (b) obtained by improved over that obtained :or q~, q~ alone
(previously computed), ViZ••
combining previous d1st.r1butions (a).

In this way two new unknowns are introducen by

linear combination. In matrix notation. the old ~r~ = L~/18EI (r , s 3, 4)

stress distributions [gir] are transformed to a

new set [giP] by forming

(That the [apo] matrix obtained here happens

to be similar to that obtained for Ql' qa in a

previous example, is coincidental.)
Once a transformation has been perfo~ed,
leading to a new ~it redundant matrix [giP] ,

~ -:/J
, the problem ~y be completed in the "0, 0
2 L
:3 !3 L • system". The appropriate equations ~re abtai~­
ad from eqs . (14-), (21), (23) and (25) Simply
, ! L 2
L
1/2 by replac1ng all "r, 5," by "~, 0". Thus
3 - :3
1 0
(29 )
0 1

where the redundants q (= q ) ~re the solutions

The transformation ~triX [Srp] is written in of P o
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A8.31

(30) auxiliary rules, test information, or even ~n­

and where
The final unit load distribution is
and the matrix of influence coefficients is
given by
•• - THE t".AGNITUDE OF THE REDUNDANT
FORCES; SIZE OF ELZME1~S IN [Urn]
Inspection of eqs. (14) or (29) reveals
that the redundant forces act in the nature of
corrections to the determinate stress dis~ri-
butions origi~~lly assumed (the [gim]).
apparent that if these corrections are large,
then any inaccuracies in the redundants (arising
tuitive guess-work which leads to a distribution
close to the final true distribution. There 1S
no need to set the redundant (c~t) member forces
equal to zero in establishin~ the determinate
distribution. Instead, reasonable approximate
values may be employed fOr them. (The correc-
tions to these values become the unknown-----
rediiiidants : )
Mathematically, the magnitUdes of the re-
dundant forces are directly dependent upon the
magnitUdes of the elements in the matrices [arnJ
or [apn] on the right-hand side of eqs. (21) or
(31)
(30). (The right-hand side of such a set at
Simultaneous equations is called the non-homo-
geneous part.) Thus, the relative mertts-or-
several possible determinate stress distributions
may be judged by forming the matrix product earn]
(32)
With each and comparing results.
FOR 2XAMPLE, if the doubly redundant struc-
ture or Example ?roblem 3, Art. A8.5 were formu-
lated in matrix forn the [gir] and [a t j ]
matrices would be (see Fig. A8.38 for numbering
scheme)
Fig. A8.38
It is Generalized force numbering
ecneme in illustrative proc-
lem. q3 and q4 selected. 3.3

frcm the difficulties inherent in accurately redundants.

inverting [ur s] or [apoJ) will have an im-

portant effect on the accuracy at the final 432 a a 0
stress distribution. ThUS, as a matter of
general practice, It is desirable to keep the 1 a 720 a a
aagnttudes of the redundant forces as small as E 0 a 402 a
possible.
It follows immediately that one should use a a a 312
for the deter.ninate stress distribution one re-
quiring a minimum of correction, i.e., Select
as the determinate stress distribution one-which .806 1.154
apprOXimates the true stress distribution as -1.564 -1. 729
closely as posslDie.
The rule giien in COnnection With the ~1!J = 1.0 a
"choice of redundants " (rule 2, above) is an aid 0 l.0
in ~king a gOOd selection for the determinate
stress distribution. If, as suggested, a deter-
minate structure is obtained by making "cuts" Several Possible deternlnate stress distri-
which leave a system having properties similar butions [gtmj
will now be tried. FIRST, the
to the original, the stress distribution obtain-
ed therein by statics should be a fair approxi- stress distribution obtained by statics alone in
~tlon to the final true stress distribution.
the "cut" structure (q, = q~ 0)
However, it is even ~ore im~ortant to
realize that the determinate stress distribu-
tion only need be in static equilibrium with
the external applied loads and that it may be
dete~lned with the aid of any appropriate
AB.32 STATICALLY INDETERMINATE STRUCTURES

(This is certainly a poor approximation to the ExaQple P~oblem 18

final stress jistribution). Multiplying out It ~s ~9sired to increase t~e ~cc~racy of
the calculation in the case ~f the box beam,
Example Problem 15. It will Je ass~ned for this
~urpose that the initial data of that proble~
were suffiCiently preCise to warrant an insrease
of accuracy.

SECOND, a stress distribution in which loads of Solution:

0440 Ibs were guessed at for q1 and Q3' (In a
t'NO times redundant structure any two forces The first step taken was ~~ examination of
may be assigned values arbitrarily while satis- the unit-red~~dant stress distributions :~ta~ned
fying static equilibrium.) q~, Q~ were found with ~he choice pr~vlously zade of q2, ~~ and ~12
by statics. Thus, as r-sdundant.s . The VITee un i t -r-eduncant stress
distributions were represented g~aphically as in
Fig. A8.39.
0.40
0.258 }
0.40
{ 0.0672

-18.79 1B.79 -26.44 26.44 -31. 71 31.71

Multiplying out one finds

= { 9.49}
-101 •

Note that a reasonable ~ at a stress

distribution resulted in non-homogenous terms
only one-tenth as great as those obtained by
use at the ~cut~ distribution. The magnitudes
at redundants are correspondingly reduced. Fig. AS. 39
THIRD, as a ~tter ot interest, the true Unit-redundant-force stress distributions -
stress distribution, obtained in Example PrOD-
lem 3, was used. The result: axial flange forces.

Inspection of' the figures showed that the

following combinations of 11striJutions should
give new distributions likely to have consider-
ably less "overlapll.
The non-homogeneous terms are practically zero,
as they should be·. The redundant forces would
be zero also.
(1)-- lX{gl.}-~g:~~{gl,}+OX{gl'''}
(2)--- 0X{gl.}+ 1x{g!,}- ;~:~{gl'''}
• The demonstration here suggests a useful check upon the
(3)--- 0X{gl.}+ 0x{gl'}+ 1x{gin.}
final result of a redundant stress calculation. After ob-
taining the final true stresses [Gim] (.. [Gjn]), one forms In ~trix form the transfo~tion was

o
and compares the result element-by-dement with the 1
matrix previously computed,
-.8330

Using [glr] as preViously com;uted, the ~ultl­

The "true-matrix" elements ought to be zero, or
nearly so, if [Gim] is error-free. plication gave
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
AB.33

3774 .1789
'0520~
I~ 1 2 3
10' .1789 .2678 .07322
1 - 1.00 E: [ .07322
.0520 .1254
2 1.00
3 1.00
4 -25.04- It remained to select a determinate stress
5 25.04 distribution Which would reduce the ~gnitude of
6 1.023 - 1.00 the redundants. For this purpose, the engin-
7 - .7103 1.00 eering theory of bending was employed to comoute
stress distributions satisfying equilibrium for
9 - 1. 023 1.00
-29.38 each application of a unit external load. The
"
10 29.38 result was (refer to Example Problem 30, Art.
A7.11).
11 1.0285 - 1.00
12 . - . 833D 1.00
13
14 - .0lD
- 1.0265
- .026
1.00
-31.71 ~ 1 2 3 4 5 5

1 15 .0lD .026 31.71 1 .30 .10

2 .10
.10
- .10
.30
NOTE: VOIDS DENOTE ZEROES.
4 .0 2.0
5 2.0 2.0
Fig. A8.39a shows plots of the new distri- 6 .0875 .0125 0.15 .05
butions, these having greatly reduced "overlap" 7 .0375
a
- .0875
.0375 .05
.05
- .15
.05
(compare with ~ig. A8.39). .0125
9 2.67 2.67 1.33 1.33
1 2.67 2.67 1.33 1.33
11 .0423 .0713
.0023 .0179 .100 .0333
12
13
.020
.0023
- .020
.0267
.0423 .0179
- .0267
.0713
.0333 -.0333
.0333 .lOO
-31.71 31.71 14 3.00 3.00 2.00 2.00 1.0 1.0
15 3.00 3.00 2.00 2.00 1.0 1.0

NOTE: VOIDS DENOTE ZEROES.

Next the matriX [a was obtained by

pn]
multiplying out per eq. (30b)

Fig. AB.39a
95 -3495 -1841 1841 0 0 ~
[gip]
G
New unit redundant stress distributions,
~p~ =~ 1167 -1167 1554 -1554 -1450 1450
1082 -1082 1445 -1445 1802 -1802

This result compared very favorably With earn]

previously obtained, the elements being fram one-
half to one-tenth as large.
The new redundant coefficient matrix [a ] The solution ~ay be carried to completion in
po
was

10'
E
·t
obtained by :n.ultiplying out per eq , (Soa ) ,

2584
-.011'32
.0000175
-.01132
.2566
-.03124
.000017J
-.03124
.1254
the "p,o system" using the matrices [giP} Capo}
[giro] and [Opn] in eqs • (31) and (32).

A8.13 Thermal Stress Calculations by Matrix Methods.

The thermal stress problem is conveniently

!or.nulated in matrix notation by an extension ot
This mat~ix is very well conditioned, being a the teChniques presented above.
considerable improvement over that at Cars]
?irst, consider eq. (21), wTitten ~n the
round originally in Example ~oblem 15, ViZ.,
form

,-.Ae·-f
-<rtf
.:{-.J::-i!:
'~-:;:::;:-::;,?
" ~':.)"4
A8.34 STATICALLY INDETERMINATE STRUCTURES

may consist of the su~ of several contributio~s

should qi act on more than one member.
Because a Is desired at each of the re-
In light ot the physical interpretations given rT --
dundant cuts sq. (34) is expanded by wri t~:::g
to [UrsJ and [Urn] , the equation is seen to be
J

a ~trix-form statement of the condition for

continuity at the redundant cuts, ViZ., the dis-
{ °rT } = ~rJ ~iT } { (35 )

placement at each cut caused bY the redundant

torces plus the displacement at each cut caused [~i] is, at course, the transpose of [girJ,
by the external loadS, must be equal to zera.
To modify this equation for thermal stresses, the unit redundant force stress distribution in
the appropriate expressions for thermal dis- the dete~inate structure. Substitution of eq.
placements at the cuts must be added. Fallowing (35) into eq. (33) gives
the argument used in Art. A8.9 one writes

{Orr } + ~r~ { qs } + ~rnJ { Pn} =a

where 0rT is the displacement at the r t h cut
due to thermal straining in the determinate
structure. The explicit form--ror this te~ will
be derived below. Re~Titten,
Solution of eq. (36) gives the values of the
redundants qs' atter which the problem ~y be
completed in the usual fashion, viz,

(33) It is obvious that the use of combinations

of redundants (the ~p,aft system of Art. A8.l2)
is possible. One makes a direct substitution or
Eq. (33) is a modHied rom of eq. (21),
giving as its solution the redundant forces in [giP] for [gir} [J.
pn] for [Urn} etc. into eq .
an indeterminate structure under the application (36 ).
ot both external loads and a temperature distri-
bution.
To der~ve an eXPliCit expression tor 6 , MEMBER THE:R1lAL DISTORTIONS
rT
the virtual work concept may be employed to ad- It remains to establish the forms for 6 •
iT
vantage. Thus, following the argument of the Thermal strains on an inrim teet inar ele-
ftVirtual work ft derivation tor deflections (Arts. ment of homogeneous ~terial can cause uniform
A7.7, A7.8), the thermal deflection at the rth normal extensions only, so that no shear strains
redundant cut must be equal to the total in- develop. Hence only normal (as opposed to
ternal virtual work done by the rt..h.-redundant Shear) virtual stresses-need be considered in
torce virtual stresses (due to a unit rt..h- r e _ computing the internal virtual work. Note that
dundant force) moving through distortions normal stresses associated with flexure ~ust be
caused by thermal strains. included. It rollows, that only virtual 'Nark in
Since the internal stress distribution is axially loaded bars and in beamB in flexure need
expressed in terms of the internal generalized be considered. Hence 6 is zero for all qi
iT
torces Q1' qj' it 15 convenient to employ these Which are shear flows on panelS or torques on
q's in writing the virtual work at straining. shafts.
It one lets ~iT be the displacement ot internal
generalized torce ql due to thermal straining, BARB

then the virtual work done by a single general- The general expreSSion for the virtual 'Nork
done by Virtual axial loads u in a bar under
ized torce 1s qi ~iT • The quantity ~iT will be varying temperature T is
called the member thermal distortion. The total
virtual work throughout the structure is ob-
tained by summing, giving the deflection at the
w = u aT dx J
r~ cut as the matrix product Where a is the ~terial thermal coefficient of
expansion.
(34)

Where ~i Is the value at the qi due to a unit *n will be convenient later to designate by qST the solution
to eq, (36) when the mechanical loads P n are zero, the
(vtr-tua r ) load at cut r , Note that the term 6 stresses in such a case being purely "thermal".
iT
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
AB.35

Eeveral speCific cases are now treated. 8EA11S

A bar under linearly varying load with ~or a bea~ the general form or expression
linearly varying temperature is shown in Fig.
for the virtual work done during thermal strain-
A8.40.
ing by a top-to-bottom-surface temperature dif-
In this case
ference 6T, varying linearly over the beam depth,

qj~llql
is (see Art. A7.8, Ex. Prob. 24).
u = qJ +
ql - qJ x
L IYrdx
Ti llll1lUJ.llill T i -h-
Tl - Tj
T - Tj • L
x Fig. AB.40
where

Then, assuming a constant m = virtual moment (positive for compress-

ion on top fiber)
IYr = TBOTTOM - TTOP
h = beam depth.
2f 1 + T
Applied to the case at Fig. A8.42 one gets
Ti+ zr j
a L J ql +aL 6 qj
6

This expression TIay be put in the form

aL 1Yr 1) qi (1Yr1 + 20T,)

+ ct,
- J q.
wher-e 11 h 6 '

6.
jT
e a
qj~T"::1 x

h = constant
Fig. AB.42
Note that variation in the cross sectional
area of the bar does not affect the distortions or
'IT' 'jT' w=
The alternate choice of generalized forces
for the bar under varying axial load is shown where
in ,ig. A8.4l. By a
derivation similar to
that above one finds
e ' ; (31Yr 1 6+ IYr ~
aL (1Yr1 +6 20T~
AjT ;;; h
Fig. AB.41
wner-e Special forms of the thermal distortion
expressions for beams of varying depth may be
derived readily as reqUired.

The simpler cases of uniform load (qi
Example Problem 19
The upper surface of the beam of Fig. A8.43
is subjected to a temperature 6T above that of
the lower surface, varYing linearly ~s shown (i.e.,
=U

r~=-x
constant) and unifo~ T (T = Tj ) follow im-
i
~ediately by speCialization of the above forms.
For example, for a bar under constant load d (
qi = qj' and constant temperature, T i = Tj = T, ~.~on~
L -1 L

one has 6. 1,:, = a LT.

Fig. A8.43

I f',...
./'
"
~
I
 AS. 36 STATICALLY INDETERMINATE STRUCTURES

the temperatures are equal at the left end and

differ by aTo at the right end). Determine the 2EI
center reactions assuming a 1s constant.

Solution:
solving

{qs,,}=
In the illustrative example of Art. AS.12
this structure was analyzed by emploYing as EI a OT,{.267}
generalized internal forces the bending moments h .933
q, and q .. over the central supports (see Fig.
A8.35). The two central reactions were denoted The final stress distribution was
by q~ and q~. The matrix of redundant coeffic-
ients, considering q~ and qa as redundants (the
better choice, it will be recalled) was (ret.
Art. A8.12)

=0 ... EI a 6T, 1 0
L
h
6EI 0 1 .257

The unit redundant load distribution for q~ 1

- 2C 1
C .933
and q .. = 1 'NaS 1 2
C - c

I~ 1 2

1 1 0 .267 }
EI a. OT 0 933
2 0 1 h :399/L
{
-1.60/L
2 1
3 - C C Thus the reactions were
1 2
4
C - C EI a. 6T o
Lh
Member thermal distortions were computed. (Note
EI a 6To (negative indi-
that OT was negative according to the convention c, = -1.60
Lh cates DO'>,/'N)
adopted earlier).
Example Problem 20
aL
2 X 3! AT ,
v The symmetric sheet-stringer panel of Fig.
~lT 11(-) A8.44(a) is to be ar~lyzed for thermal stresses
6
developed by heating the two outSide stringers
to a uniform temperature T above the center
stringer. Assume G • O.385E.
Solution:
""h.
; - (-) (2 x ~ OT6 + ~ 0
4
T0"" (""
+ -
h
(-l The panel was divided for convenience into
three bays. The numbering and placing of gener-
alized forces is shown in Fig. A8.44(b).
'" - 3:2 l""i 0 T~ Transverse members (ribs, not shown on Figure)
were considered rigid in their own planes - a
satisfactory assumption for symmetric panels.
Then from eq. (36) Because of symmetry only one half the panel was
handled. All member fleXibility coefficients and

~ ~ {q'J ~ ~ ~!
aL6T, thermal distortions were doubled where appropri-
3h ate.
..1.
6EI 14 q, =(-) 0 - 1
01 2 2aL -T The matrix of fleXibility coefficients was
3h 0 c set up as (VOIDS DENOTE ZEROES)
C- L-
o
o
 P'

ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AB.37

177 .6 44.4 I T:J.en

44.411"7"':'.6 -.'0
44.~ 3S.
, .,.2 <8S.3 22.2
- .5 0 (;

2. 44.4 0 - .0 0
31 200
31,200 0 0
31 2 0
- .5
1.0 0 0

was
The matrix of ~ember thermal distortions
~1~ =
0 1.0 0
0 0 1.0
40 - .025 0 0
40
.025 - .025 0
20
o 0 .025 - .025
o
o MultiplYing out per eq. 18,
o
o
o
1~'8J
13.8
[2,2
Here, for example AlT = 2 x 2 x 10 = 40 (doubled ~r~ = 1
E
13.8 172.2
once because ql acts on two stringer ends and 0 13.8 86.1
doubled again to account for the other halt or
the panel).
and

~ "1I
-.025
", ""'"

1
-.005 o
.025 "

~ -."'-' ao

",,02
PCyp 1=== 60"
T
F=
I- =
=
(bl 'i.
Then the redundant equations, per eq. (36),
FIg. AB.44 were
The structure was three times redundant.
Stringer loads q~J qo and qo were selected as 13.8
redun<1ants. Fig. A8.45 shows the unit redundant
load Sketches for q. = 1, qo = land qo = 1. 172.2
13.8

Solving,

n3lf '\-
{ } = a ~ S
{
. 1082 }
.?9944
."C02

~ n~ Finally, the complete therTal stress distribu-

tion 'NaB
- 54.1 -
- 49.7 -

~ 0~ - 50.1 -

:::: x 10 -. 108.2
99.44
is 16}1 0&<I.1n~ fi. 100.2
- 2.70
Fig. AB.45 0.219
- 0.019
A8.38 STATICALLY lNDETERMrnATE STRUCTURES

The result compares favorablY ~lth "exact"

solutions made under the same assumptions (ref. l
NACA TN 2240) as far as the stringer loads are
concerned. The shear flow result is not very L/Z
satisfactory, primarily due to the use of too
few "bays ~ in this analysis. 1
Example Problem 21 L/ Z Liz
The uniform fOUT-flange box beam of Fig.
A8.46a is to be analyzed for the thermal
stresses developed upon heating one flange to a ~I£J ~

1
temperature T, unt rcr-a spanwt se , above the other
three flanges.
,. Liz ILiZ I,
L/2

"', ' 1
~~ ~ I
1\ ,\~,. "\~'''' Liz Liz L/2 L/2
). ~':\~,
~'~ •
~,~~- q\~\~
\
Mult 1plying out per sq. (16) ,
~) ~\
:\~, ~
~' .500.. B:33
, \ i:"iC7 . .625 .375 .1:5

",~,,- '\~~' .500 -83

~:,.., \1
[iki''' ... .375 .l25

\~'~
.'25
l.·k~(b'::)
~'!J . '", t
.s-e .37' ~ ...::.::.3 .:25
--L
~.--I
(aJ
Fig. AS. 46
~ ~ "\~ .125 . 125 .125 ~ . .JS33

For a spec !fie case let :-::'iLa, ~

L Tr.e
Solution: ( 'j i~verse was computed to je
The beam was df vt ded into four equal bays
giving a tour times redundant problem. Four 1.070 - .5291 .2337 .06587 - -
self-equilibrating (zero-resultant) independent
. Gt
stress distributions were taken as the unknowns, Grs::J L (0 • oj - .2337 - .3613 1.497 - .1934
- .5291 1.368 - .3613 .1018 -
these being shown In Fig. A8.46(b). Such zero-
resultant strP.ss distributions are the only oneS - .06587 -
.1018 - .1934 1.792
possible in a structure having no applied loads·.
The matrix of member flexibility coeftic-
tents was tormed by collecting coefficients from Member thermal distortions (!Ii'!') were
the several members. Un1t redundant stress d1s- computed tor loads Q2. q .. Qa and qa and were
tributtons were prepared, taking QJ., q.. Qa and collected from the one heated flange only.
Q? as the redundants and setting these equal to
unity successively.
0 (Note that if two ad-
.scc 2 jacent flanges are
.6emc" .1667lc" 0 heated equally, one aus't
~lU' , (:,o ..e c!
~IT} ~ a LT 2 set the corresponding ~i
.500
.1667lc" .667lc" .1667ic" { 8 0
2 equal to zero; this be-
.1667lc"
.eoc
.667',(" ,
.500
.1667lc" 0
1
cause the virtuai loads
being of opposite Sign
adjaCent flanges, the
.
.""".
.1;;67k~ (
Virtual Ncrk ~ust cane ,"1 ) .
where k 2 = Gt / AE (b • c) MUltiply1n~ out:

• Since externaf Ioada are not to be applied it follows from

f}
statics that the gener-al.iaed forces for adjacent webs and
cover sheets are equal, as are the loads in front and rear a T Sa 5
spar caps at any given station. Hence an economy of
numbering in the generalized force scheme is possible.
Much labor is saved in the handling of data when the same
[]rLl{"IT}=~ i
symbol can be employed on several members whose loads
are known to be equa1. Then the solu-::ion to 3q. (36) w~s ~Titten as
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AS. 39

"cut" ~king it determinate, after 'Nhich the

temperature distribution is applied producing
thermal deflections
which, when multiplied out gave

4 ' 079 }
Gt a L T 1.938
{qST} = - 16 Ib + c)
{
.8562 at the external points (compare with eqs. (35)
.2413 . and (37) ). Simultaneously the redundant cuts
will experience relative displacements.
Finally the .conptece set at thermal stresses SECOND, the redundant cuts are restored to
were (still for the case L3k 3 = 1) zero displacement by the application at redund-
ant forces (thiS problem was solved in Art.
A8.l3) .. The qs are given by eq. (36); they pro-
4.079
2.039 L duce additional deflections at points m
1.938
3.008 L
.8562
3.437 L
.2413
The total deflection of point m is then
3.558 L

The result compares favorably with an

exact solution (NACA TN 2240) insofar as flange
stresses are concerned, an error of less than But
1 percent being present at the root. Shear
flow values in the sheet are less satisfactory
due to the (relatively) crude assumption of
constant shear in~each beam quarter.
Therefore
A8.14 Thermal Deflections by- Matrix Methods.

A. STATICALLY DETElIMINATE STRUCTURES

The problem of the thermal deflections ot
a statically determinate structure was con-
sidered earlier in Art. A7.8 in non-matrix
fon. It should be apparent from the deriva-
tion, that the matrix method presented for the
calculation of redundant-cut thermal deflec-
tions may be applied equally well to the
problem of computing the thermal deflections of
external paints of a determinate structure.
Thus
(37)
where OwT is the the~l deflection at external
point ffi, and [Gmi ] is the transpose at [Gl~J'
the unit-applied load stress distribution
(compare with eq. 35).
8• REDUNDk'lT STRUCTURES
In the case of the redundant structure,
additional strains are present due to the
the~l stresses set up; the effect of these
strains upon the deflection of external paints
must be included in the calculation.
The appropriate equation is most eaSily
derived by VisualiZing the action in two
stages. FIRST, the redundant structure is
or,
The matrix quantity in parentheses is the
total strain (thermal Dlus "mechanical").
------Fa:r-detinite reasons the equation for
thermal deflections has been lett in the ·form at
eq. (38) rather than the more polished forms
which might be obtained by substitution from eq.
(36), FIRST, the qjT' the thermal stresses, will
probably have been solved tor preViously ~~d will
be readily available in explicit torm. SECOND,
and far more UnDortant from a labor saving stand-
point, the unit load distrioution [giml (Whose
transpose is used in sq. 38) ~y be any conven-
ient stress distribution satiSfYing atatrcs-rn
~s~t of "cut" structures. One need not
even use the Bam0l[ilm] distribution (and same
choice of "cuts") as employed in the redundant
thermal stress calculation; a more convenient
choice of cuts may be employed: In prinCiple,
any stress distribution statically equivalent to
the LU1it aprlied load(s) may be '..l.sed for [gimJ
!!l2.S.. (38). (See p. A8.9). --
A8.40 STATICALLY INDETERMINATE STRUCTURES

Example Problem 22
Co~pute the rotation occurring at the right
_ a L 0 To
- h 1.2 0 !.J
hand end at the beam ot Fig. A8.43.

Solution:
aL:ST o
- .289
To compute the rotation a unit couple was h
applied (positive counterclockwise) at the right
hand end. An additional generalized force, It is apparent that the values of the re-
called qa, also was added at that point. Then, dundant ~oments ql, qa could have been chosen
arJ1trar1ly in {g1~}' above, with8ut af~ecting
,
"5T--(-)
_ ct, (3g 0To + 20 T0)_- -4 a L 15 To
the result. Here, clearly, {gi;n} ror any "cut "
h 6 9h
structure visualized w1l1 lead to the same re-
and, uSing some results tram Example Prob. 19, sult, qa being equal to ~ity in all cases.

L 15 To
9n
2, 5
Note that q~ and q., the intermediate support
reactions, were omitted from consideration.
They do not enter into any expression for the
internal virtual work or the structure; or,
Example Problem 23
compute the axial ~ovement of the tree end
of the central stringer of the panel of Fig.
A8.44.
Solution:
An additional generalized ~orce, qlOJ was
added aXially to the free end of the central
stringer (ret. Fig. A8.44b). T~en

..., ......,, ...,

equally, they are not used to describe the
strain energy of the structure. Hence they are l"n.13 !, I , I I
not included in writing the total str~in. 177 .5 i I
(Their 6
iT
are zero.) -- 88.8 , I I
The member flexibility ccefficient matrix I 88.8 22.2 !2.2.2 I

was I ,22.2 88.8 22.2 I

22.2 44.41
! ! 31,2001 I

, !31.20:> !
i
i=1,2,5 i 2.2.21
,I 31,200
44.':'

40
From Example Problem 19, the true thermal 40
stress distribution was 20
o

{ q! } =
EI a 0
h
To {.267}
.933
o
a
o ! 1, 2, 5 o
o
(qa was zero by inspection) o
o
To determine [gi~J' a unit couple qa =1
Using the qjT as obtained in Example ?roblem 20
was applied. Taking as the Wcut Wstructure one
Where q. = qa = 0 one has Simply (With qlo = 0), the follOWing product was
formed:
-11.81
-13.45
- 6.655
11.81
Then SUbstit~tlng into eq. (38), 13.45
x aT

{Omr }• ts 0!J(~.
9h
{'} ....La 1n, "0 0T,{.257})
6
40
6EI 14 -
012
h .933
0
6.656
-84.24
6.833
•.5 928
2.402
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A8.41

I'hen , subs t t tut t ng into eq , (38) J a unit load qlo =1 gives (writing the transpose
of {glm})
5", = aT L&nJ 28.19
26.55
13.34 L&nl~ = L· 50 .50 .50 0 0 0 .025 0 0 1~
11.31
13.46
6.656
Then ~ultiplying out for the thermal deflection:
-84.24 O:.r = 34.3 a T (same answer).
6.833 It is apparent ~rom the above result that
.5928 the Simplest deter~inate ("cut") structure should
2.402 be used to compute [glmJ; it is completely
adequate.
It remained to find the determinate stress dis-
tribution for Qlo = 1. CLC6URE
As the determinate distribution the {gim} The general the~al stress problem is
stresses due to a unit load Q~o = 1 were com- complicated by the fact that the material prop-
puted in the "cut" structure obtained when erties E, G-and a vary with temperature. The
q, = qll = q:s = O. (Note that this is a differ- problem created ther-eby is primarily one of book-
ent choice of redundant :uts from that employed keeping - co~puting the member fleXibility coeffi-
in computing the thermal stresses in Exa~ple cients (U and the member thermal distortions
i j)
Problem 20.) rnus (d for a structure whose properties vary from
iT)
point to pOint with ths temperature. The vari-
o ations of E, G and a with T Will, of course,
o have to be known from test data.
o Two additional complications, not considered
1 here, are the lowering of the yield pOint with
1 heating (and the attendant increased likelihood
1 at developing inelastic strains) and the phenom-
o enon of "creep" (the time-dependent development
o of inelastic strains under steady loading).
o Should it prove necess~ry to analyze for
1 the~l stresses under ~ore than one temperature
distribution, the me~ber thermal distortion
and finally,
matrix { d may be gener'a Ltzed eaa i Ly into a
1T}
o",T = aT LO 0 0 1 1 1000 1J 28.19 rectangular form such as
26.55
13.34
11.81
13.46
6.656
-84.24 where d is the member thermal distortion as-
iR
6.833 sociated with rorce qi from thermal loading con-
.5928
2.402 dition R. The matrix [C f j ] of ~ ther.na.l
34.3 a T (AJlSWER) coefficients consists of the constant coef~iclents
in the d expressions preViously presented. wr-ile
As a matter of interest, a different unit iT
stress distribution was employed With a differ- T would be the temperature associated with qj
jR
ent choice of "cuts". If the forces q., q , and for condition R. (compare with eq. (26b),
qll are set equal to zero ("cut"), app Ltca't t on of Art. A7.111.
 A8.42 STATICALLY lNDETER1vfiNATE STRUCTURES

A8.lS Problems. (5) For the "King ~ost" truss in Fig. e,

Note: Problems (1) through (9) below may calCUlate the load in member SO. Members AS,
ba worked by either the Least Work or DQ~Y Unit BC and BD have area of 2 sq. in. each . ':':-le con-
Load Methods. The student will be well advised tinuous ~ember ADC has an area of 9.25 sq. in.
to try some problems both ways tor comparison. and moment of inertia of 21~ in~. E 1s same for
all members.
5000* A ..........
r"'"""'----7i A
~
E

100"
20"
"I
[-.10'j
4000*
!
A
jA
f- 100" B C,
Fig. a Fig. b
I- 60
+ 60 -j
(1) Determine the load in all the members Fig. g
at the loaded truss shown in Fig. (a). Values
in ( ) on members represent the cross-sectional
area in sq. in. tor that member. All members 500.
at same material. (6) In Fig. f, AS is a steel wire both 0.50
(2) For ~he structure in Fig. (b), deter- sq. in. area. The steel angle frame C3D has ~
mlne the load in each member tor a 700# load at 4 sq. in. cross section. Determine the load 1n
joint B. Areas at members are given by the member AB. E = 30,000,000 psi.
values in ( ) on each member. All members made (7) In Fig. g find the loads in the two tie
at same material. rods BD and CEo lac = 72 in." ; Abd = 0.05 sq.
in. Ace = 0.15 sq. in. E 1s same for all
60"
members.

1) 30" c
B~.r---aD9ct6~"
Fig. c
IE IF:o-
I
1000* 1000. II 180"
(3) For the loaded truss in Fig. c, deter-
mine the axial load in all members, Values in
parenthesis adjacent to members represent rela-
tive areas. E is constant tor all members.
so.ooos A
'F-100"-I
1-1
Fig. h Fig. i
C

Fig. d (8) For the structure in Fig. h, determine

(4) For structure in Fig. d, calculate the
axial loads in all the members. Values in paren-
thesis adjacent to each member represent relative
areas. ~ 15 constant or same for all members.
5000*
the reactions at points A, B. M:embers CE and ED
are steel tie rods with areas of 1 sq. in. each.
Member AB is a wood beam with 12" x 12" cross
section. Esteel = 30,000,000 ps1. Ew-ood =
1,300,000 psi.
(9) For the structure in Fig. 1, determine
the axial loads, bending moments and shears in
the various members. The structure is continu-
ous at joint D. Members AB', 8C are wires. The
member areas are AB = 1.2, Be = 0.6; CD = 6.0;
SDE = 10.0. The moment of inertia for members
CD = 60.0 in."; for SDE = 140 In.''
(10) Re-so1ve Example Problem 10, p. AS.15
using as rsdundants the two restraints at one end

~
(couple and transverse force). Solved in this
I D
C
way the problem is doubly redundant as no ad-

l~iI7T~
vantage 15 made of the symmetry of the st.r-uc'tur-e,
(11) Add ~NO additional members) diagonalS
~_ t FE and ZC (each with areas 1.0 ina) to the truss
Fig.e
I-- 120" + 120" --j
of Fig. A7.85, Chapter A-7. Find the matriX of

- '. ~"._:_.-,,-: . . . . ' . "'.'~'~'"":" ,. T' ~_ • •" _ • • _.. ,._

 ANALYSIS AND DESIGN OF
influence coefficients.
.ane ,
21.8 27.2 27.5
1 27.2 68.5 85.4
[Amn] = E 27.5 65.4 68.4
18.3 26.4- 26.2
(12) Re-solve the doubly redundant beam
of Example Problem B, page A8.3 by matrix
~ethods. The redundant reactions should be
given ~q~ symbols. (See Example Problem 13a,
page A8.20).
(13) Re-solve Example Problem 5, page
A8.12 by matrix methods. For Simplicity, make
your choice of generalized forces including
those designated as X and Y in the example so
that Figs. A8.21 and A8.22 can be used to give
the gir loadings.
(14) By matrix methOds re-solve Example
Problem 4, p. A8.12 using 3 equal bay diVisions
along the panel (3 times redundant). Use the
same structural dimensions as in Example Prob-
lem 20, p. A8.36. Campare the results with
those obtained from the formulas developed in
Example Problem 4.
(lS) Show that the matrix equation
sq. (21) 1s modified to cover the initial stress
Douglas Dc~a airplane. Ph.otograph. showing simulated aerodynamic load being applied to
main entrance door of fuselage test section.
FLIGHT VEHICLE STRUCTURES A8.43
problems of Art. AS.S by 'Nritlng
18.3
where A1 1s the initial imperfection associated
26.4 with force qi' Refer to the argument leading
26.2 to eq. (11) or Art. A8.8.
20.6 (16) USing the equation of problem (15),
above, re-solve Example Problem 6, p. AS.l4.
(17) Using the equation or problem (15)
above, re-solve Example Problem 7, p. AS.14.
(18) USing the matrix methods or Art.
A8.13, re-solve Example Problem 9, p. AS.15.
(19) For the dOUbly symmetriC four flange
box beam at Example Problem 15, p, A8.24, de-
termine the redundant stresses qa, q7 and q~. if
one flange is heated to a temperature T, unitorm
spanwlse, above the remainder of the structure.
nne,
EaT 20Bl}
: : } = - 10' 1599
{ {
741
qu
REFERENCES •
See references at the end of Chapter A-7.
, Cf.
( 'j
A8.44

PROOF TEST OF DOUGLAS AIRCRAFT MODEL DC-6

DOUGLAS DC-8 AIRPLANE. An outboard engine pylon mounted on a section of wing for static and nutter
tests. The steel box represents the weight and moment of inertia of the engine.
 CHAPTER A9
BENDING MOMENTS IN FRAMES
BY
ELASTIC CENTER METHOD
A9.1 Introduction
In observing the inside of an airplane
fuselage or seaplane hull one sees a large num-
ber of structural rings or closed frames. Some
appear quite light and are essentially used to
~lntaln the shape of the bOdy metal shell and
t~ ~rovlde stabilizing supports for the longi-
tudinal shell stringers. At points where large
load concentrations are transferred between body
and tail, wing ~ower plant, landln~ gear, etc.,
relatively heavy fra~9s will be observed. In
~ull construction, the bottom structural fram-
ing transfers the water pressure in landln~ to
the bottom portion of the hull fraToas which in
turn transfers the load to the hull shell.
In general the :~ames are of such Shape
an~ the load distribution of such character
that these fra~es or rings l~~dergo bending
forces in tranSferring the applied loads to the
ot~er resisting portions of the airplane bOdy.
~hese jending fcrces produce frame stresses in
gener~l which are of major importance in the
strangth ,roportioning ot the frame, and thus
a reasonable close apprOXimation of such bend-
ing forces is nec~ssary.
Such fra~es are statically indeterminate
relative to internal resisting stress and thus
consideration ~ust be given to section and
physical properties to obtain a solution of the
distribution of the internal resisting forces.
Al~D RINGS
Assumptions
In the derivations Which follow the distor-
tions due to axial and Shear forces are neglect-
ed. In general these distortions are s~ll
compared to fra~e bending distortions anG thus
the error 1s small.
In computing distortions plane sections are
assumed to remain plane after bending. This is
not strictly true ~ecause the curvature of the
trame Chan~es this linear distribution of oend-
tng stresses on a frame cross-section. (correc-,
tions for curvature influence are given in
Chapter A13.
Furthermore it is assuned that stress is
proportional to strain. Since the airplane
stress analyst TIust calculate the ultimate
strength of a frame, this assumotion obviously
does not hold with heavy frames ~here the rup-
turing stresses for the frame are above the pro-
portional limit of the frame material.
This Chapter will deal only with the the-
oretical analysis tor bending moments in frames
ar.d ~ings by the elastic center ~ethcd. FYac-
tical questions of bOdy fra~e design are covered
in a later c~apter.
The followin~ photographs of a ~ortion ot
the structural framin~ of the hull of a sea-
plane illustrate both light and heavy fra~es.

Jeneral ~ethods of Analysis:

There are ~ny ~ethods of applyin~ the

~rinciples of continuity to obtain the solution
for the redundant forces in closed rings or
frames a~d bents. T~e author prefers the one
Nh~c~ is g8ner~lly referred to ~s the ~Z18stiC
:::2::t~r~ me thod anc :,"':.s '. . sec. it for -nany year-s ~n
~Q~ti~8 ~ir~l~"e :~si~n. The 7.et~od was Qri~i­
~atec by ~uller-Breslau~. T~e main cifference
in this ~et~od as som~ared to Tiost other methods
of s~lution is that the redundant forces are as-
su~~i ac~ing at a s~8cial :oint called th~ elas-

ec...a t Lcns for t ne r edunoant.s Nhic:l a re tnc v-

~e~d2r.t of eac~ other.

'~MuUer.Breslau, H., Die Neueren Hethden der

FestigkeHslehre und der Statik der Baukon-
struktionen. Leipzig, 1886.

A9.1
A9.2 BENDING MOMENTS IN FRAMES AND RINGS

Consider a small element ds of the curved

beam as shown in Fig. A9.2. Let Ms equal the
bending moment on this small element due to tee
given external load system. The total jending
moment on the element ds thus equals,
- -(1)
(Moments which cause tension on the inside
fibers of the frame are regarded as ?osi-
ti ve moments. )
The following deflection equations for
point (A) must equal zero:-
Q = a (angular rotation of (A) = zero)
Ilx -= (, (movement of (A) in x df r-ect t on -= 0)
A9.2 Derivation of Equations. UnsymmetriCal Frame. Ily = a: (movement of (A) in y direction = 0)

Fig. A9.l shows an unsymmetrical curved From Chapter A7, which dealt with deflection
beam fixed at ends (A) and (8) and carrying theory, we have the following equations for the
some external loading p~, P a , etc. This movement of point (A):-
structure is statically indeterminate to the
third degree because the reactions at (A) and -(2)
(B) have three unknown elements, namely, sagnt-.
tUde, direction and line of action, making a
total of six unknowns with only three equations Ilx=Z~~S- -(3)
of static equilibrium available.
Oy := Z i1mds -(4)
£1
p.
P, In equation (2) the term m is the bending
P,
t I
moment on a element ds due to a unit moment
applied at point (A) (See Fig. A9.3). The bend-
\ ing moment is thus
equal one or unity
Fig. A9.1 for all ds elements A~
B
of frame. Cunit moment
A Fig. Ag.3
ds
p P Then substi-
..J I.- B tuting in equation (2) and using value of M from
equation (1) we obtain -

g := .J"!sds + MAZ l;d _ XAZl.yds + YAZl.XdS = 0

~ ... 1S EI EI
B

FIg. A9.2 whence,

In Fig. A9.2 the reaction at (A) has been

replaced by its 3 components, namely, the forces I
X and Y and the moment !1 and the structure In equation (3) the term m represents the
A A A
bending moment en a element ds due to a unit
is now trea~ed as a cantilever beam fixed at load applied at point (A) and acting in the x
end (B) and carrying the redundant loads at (A) direction, as illustrated in Fig. A9.4.
and the known external loading p~, P3' etc.
Because jOint (A) is actually fixed it does not The applied unit y
sutfer translation or rotation when structure ,
load has a positive

l~~B
is loaded, thus the movement of end (A) under sign as it has been
the loading system of Fig. A9.2 must be zero. assumed acting toward
Therefore, three equations of tact can be
written stating trAt the horizontal, vertical
the right. The ,
distance y to the ds Y
and angular deflection of ,oint (A) must equal Fig. A9.4
element is a plus
zero. distance as it is
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A9.3

measured upward from axis x-x through (A). place of XA, YA and MA respectively.
However the bending ~oment on the ds element
shown is negative (tension in top fibers), thus The axes x and y through the point (0) are
the value of m ; - (1) y = -yo The minus Sign centroidal axes for the values ds/EI of the
is necessary to give the correct bendir.g moment structure. This fact means that the summations-
Sign.
ydS
SUbstituting in Equation (3) and using M Z Er= 0 and
from Equation (1):-
The expressions Z x·ds/EI, Z y·ds/EI and
Z xydslEI also appear in equations 6 and 7.
These terms will be referred to as elastic
moments of inertia and product ot inertia ot
In equation (4) the term m represents the the trame about y and x axes through the elastic
bending moment on a element ds due to a unit center of the ~e, and for Simplicity will be
load at point (A) acting in Y direction as given the following symbols.
=
illustrated in Fig. A9.5. Hence, m l(x) x =
SUbstituting in x·ds yiJds xyds =
equation (4) and Y ds Z """'EI = I y , Z "E'r = Ix <;'

, "" EI Ixy
USing I'! from
equation (1), we x---~ Equations 5, 6 and 7 will now be rewritten
obtain, lj;- USing the redundant forces at point (0).
Fig. A9.5

hence,
- - -(7)

Equations 5, 6, 7 can now be used to solve - - - - - - - - - - - (6)

for the redundant forces MA' XA and YA. With
these values known the true bending moment at
any point on structure follows from equation
(1) •

REFERRING REDUNDANTS TO ELASTIC CENTER

For the purpose of simplifying equations
5, 6, 7, let it be assumed that end A is
attached to a inelastic arm terminating at a
point (0) as illustrated in Fig. A9.6. The
~oint (0) coincides with the centroid of the
ds/EI values for the structure. Reference
axes x and y wIll now be taken with point (0)
as the origin. The redundant reactions will
now be placed at paint (0) the end of the
Y
I element ds will be given the S~bol 0s ' that is,
The term ~Z ydslEI is zero since Z yds/EI
is zero, thus Mo drops out when substituting in
Equation (6).
- - - -(10)
The term Msds/EI represents the angle
change between the end faces ot the ds element
when acted upon by a constant static moment Ms.
This angle change which actually is equal in
value to the area of the Ms/EI diagram on the
05 : MgdsIEI. ~ith this symbol SUbstitution,
equations 8, 9, 10 can now be rewritten as
follows:-

------(ll)
elastic Fig. A9.6

inelastic bracket, as shown in Fig. A9.6. - - -(12)

Since point A sutfers no movement in the actual
structure, then we can say that point (0) must Z 0s x - XoIxy + YoIy =a - - - -(13)
undergo no movement since (0) is connected to
pOint (A) by a rigid arm. Solving equations (12) and (13) tor the
redundant forces Xo and Yo we obtain,
Thus equations 5, 6, and 7 can be re-
written USing the reduncarrts Xo' Yo' and Ma in
A9 4 BENDING MOMENTS IN FRAMES AND RINGS

•• y
"'~S -
MX(~\
---r:-:-)
"'~S
Table A9. 1
= \ y - - - - - - -(14) y= x=
IX (I -~) Por- w=ds dist. dist.
TXIY tion 1 ).05. from
x -x y-y

- [Z0s X - 60 sY (!t;-)J AB 3~:10 15

wy2= 2250 wx 2=1440
·12 -120 +150 i = 30 3 ::750 i - 30 - 0
x y-nz3-
Ye = I (I _ I 'x? ) - - - - - -(1,0) 12x3
y TXTY wy2::10800 wx 2 : 0
Be ~12
2
30 0 0 360 . - 24 3 ·~6
i x ::.-li 3= 0 'v·-_·,,
A9. 3 Equations for Structure with Symmetry About One Axis
through Elastic Center. 12x2 . 12x2 I

wy2'::2250 wx 2:1440
CD 30=iO 15 12 120 150
If the structure 1s such that either the x 3 Ix: 750 iy=~3=0
12x3
or y axis through the elastic center 48 a axis
of symmetry than the Droduct of inertia Sum 32 0 660 16800 3456
Zxyds/EI = I xy = zero~ Thus making the term
I xy = 0 in equations 11, 14 and 15 we obtain,

=-z ¢s - - - - - - - - - - - - -(16)
tds/EI
_ Z¢sy The terms i x an1 1 are the elastic ~OT.e~t
-r;- - - - - - - - - - - - - - -(17) of inertia of each portIon of the fra~e about
its centroidal x and y axes. Si~ce I is con-
_-Z¢sx stant over each pert ian the centroidal T.oment
Yo - -r;;-
y
- - - - - - - - - - - - - - (18) of inertia of each portion ~s identical to that
of a rectangle about its centroidal axis.
A9.o1 E:m.mple Problem Solutions. Structures with at least
One Axis of Symmetry. To exp'ln In for member AB:- ~ '_ +=t= b

,Exa~ple ~oblem 1 Re re rrv.ng to Fig. a,

ll-----O
liT
B!7:JC I
10 lb.
Fig. A9.7 ~'~18"_ ix =121 bh3 = 121 x 3I x 30~ = ·7:':
_,,0,0 <-,I -IX
I ,ds' 30' h
shows a rectang-
ular frame with
!lxed supports •Rf
e erring to Fig. b , A uI '
!
--1-
I

Fig. a
at ;:JOints A and
D, and carrying ly = ~bh3
12
=.l:.- x ~o
12'" x3!'
1 :;:: .09 (negligible)
a single load as 30"
Shown. The Y
The distance from the

8T
problem Is to de- I =3
termine the bend- two reference axes to the
ing moment dia- elastic cer.ter can now be
calculated:-
gram under this
loading.
! I ds=30=b
X •- 'fT11"TT]
A IY- D
- - - - - rrrtrrrr-
=~ : 660 = 20.625 In. ~

-
y I I
T'ne r irst;
step in the
Fig. A9.7 Zw 32
U...l.
-oJ yl ~.!=.!= h
solution is to 0 - I , I 3
x = J::N • 32 = 0
Fig. b
find the location of the elast1c center of the
frame and the elastic moments of inertia Ix
and I y . Having the moment of inertia about axis x-x- we
can now find its value about the centroidal
Due to symmetry of the structure about the axis xx ot the frame, by use of the parallel
Y axis the centroidal Y axis is located midway axis theorem.
between the sides of the frame, and thus the
elastic center (0) lies on this axis. Ix ~ Ix - ~w(y~) = 16800 - 32 x 20.625~= 3188

Table A9.l shows some of the necessary I y : I y - "w(x') = 3456 - 32(0) : 3456
calculations to determine the location ot the
elastic center and the elastic moments of The problem ~ow consists in solving eq~­
inertia. The reference axes used are x·-x~ tions (16), (1:) and (18) for the redundants at
and s-s, the elastic center, namely
 - ..
--~

-~-
. ..

ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A9.5

_ ~ _ Area of static Mil diagram along the neutral axis of the frame members.
Mo - ZdS!I - Total elastic weight ot structure
In Fig. A9.10 the area at the Ms/I diagram
Moment of' static Mil diagram about equals 0 s = 22.5 x 24/2 = 270. The centroid by
X - ZTsY - x ax1s siillPle calculations of this triangle would fall

'
o - x - Elastic moment of inertia about x 10 inches from B. Fig. A9.ll now shows the
ax1s irame With 1ts MslI or lts 03 load. Os is
0 s = 270 -l' .. positive since I1s is
Moment of static Mil diagram about B-...;;:,·:-2 C pos rt ive . The next
y axis
Yo • -Z¢sx = ==""'-==::'-::T;~="'--;=::'-:;­ 1·375" step is to solve the

fl
Iy Elastic moment ot inertia about y x---- ..:- x equations for the
axis 0: r-edundant at the
: 20.625" elastic center. The
Thus to solve these three equations we ; I signs of the distances
~ust assume a static frame condition consistent A 'Y D i x and y from the axes
with the given irame and loading. In general I x and yare conven-
there are a nlli~ber of static conditions that can Fig. A9.1l t i ona L,
be chosen. For exanple in this problem we
~ight select one ot the statically determinate Thus,
conditions 11l~strated in Fig. A9.8 cases 1 to
,. - (270)
= 32 (from Table A9.1 = -8.437 in. lb.
, 10

,~I Xo
= Z¢sY = 270(9.375)
Ix 3188 =0.7939 lb.

I I
-:-:T1
--ZTsx _ -(270)(-2)
Yo - y - 3456 = 0.1562 lb.
Case 1

Fig. A9. S Fig. A9.12 shows these values of the re-

dill1dants acti~g at the elastic center.
To illustrate the use at different static
conditions, three solutions will be presented
with each using a different static condition. B
~17. 75
;:8
T
9.375

t -F'9
Solution No.1 8.437

In this solution we will use Case 3 as the

static fr&~e concition. The bending moment on
the fra~e for this static trame condition Is 20.625" ):1562
given in Fig. A9.9. The equations
~~.81 J
~ s diagram 22.5 Ms/l diagram
L06'AJ 6.06 ~81
e

A
B[rig.lo. ,e
DolA9.9~
Bn Ms is posrnve ,
on inside of frame.
tension
Fig. A9.12

The bending moments due to these redundant

forces will now Je calculated.
FIg. A9.13

,
7.5#
,
,""'- A
'2.j~ Fig. A9.10
MA =- 8.~7 - .1562 x 12 + .7939 x 20.625 :;:
6.06 in.lb.

Me =- 8.437 - .7939 x 9.375 - .1562 x 12 =

for the redundants require 0s the area of the -17.75 tn Lb. s

~s/I ~lagra~. Fig. ~9.10 shows the ~s/I curve

which is obtained by divlcing the values in Me =- 9.437 - .7939 x 9.375 + .1562 x 12 :;:
~lg. A9.9 by the te~ 2 which is the moment of -14.00 In. lb.
inertia of ~ember Be as given in the problem.
Since the equations for Xo and Yo require the NO =- 8.437 + .7939 x 20.625 + .1562 x 12 =
moment of the Ms!I diagram as a load about 9.81 In. lb.
axes trIough the elastic cencer of frame, the
area of the Ms/I diagr~n will be concentrated These resulting values are plotted on Fig.
at the centroid of the diagra~ and along t~e A9.12 to give the bending ~oment diagram due to
centerline of the frame, or more accu~telY :he redundant forces at the elastic center.
A9.6 BENDING MOMENTS IN FRAMES AND RINGS

Adding this bending moment diagram to the static y = -Z0s x = -[-45(-10)-4D5X6-300(-12)-90OX12]

bending diagram of Fig. A9.9 we obtain the final o ly 3436
bending moment diagram of Fig. A9.13.
= -(-9180) = 2.656 lb.
The final bending moments could also be 3456
obtained by substituting directly in equation
(1) ua tng subscript (0) instead of (A). ThUS,
The final ~cments at any point can now be
M = Ms + Mo - Xoy + YOX - - - - - - -(19) found by equation (19).

For example, determine bending moment at Consider paint 3:-

pOint a. lis = -30 from Fig. A9.14
For point B, x = -12 and y = 9.375, Ms =0 x ~ -12, y ~ 9.375
substituting in (19)
suot , In (19)
liB =0 + (-8.437) - .7939 x 9.375 +
liB =-30+51.56-.7939X9.375+2.656(-12)
.1562 (-12) = -17.75 as previously
round = -17.75 in. lb. which checks fi~st
solution.
AT ~OINT D. x = 12, y = -20.625, Ms =O. Consider point D:-
liD = 0 + (-8.437) - .7939 (-20.625) +
.1562 x 12 =
9.81 in.lb. Ms = -90, x = 12, Y = -20.625

Solution No.2 Subt. In (19)

In this solution we will use Case 4 (See MD =-90+51.56 - 7939(-20.625) + 2.656 x 12

Fig. A9.8) as the assumed static condition, = 9.80 in.lb.
that is two cantilever beams with halr the
exter-ual load or 5 lb. acting on each canti- Solution }Io. 3
lever. Fig. A9.14 shows the static bending
mcment diagram and Fig. A9.15 the Ms/I diagram. In this solution we will use Case 5
• "5=-45 "~-405
(Fig. A9.8) as the assumed static condition,
namely a trame with 3 hinges at points A, D and
-45 ~, as illustrated in Fig. AS.16.

Ill< Betore the

I -1"" ~18"--I bending moment cla-
-"62515. 6~5­ E gram can be calcu-
T':: I -'-
r- 12" lated the ~eacticns
~:- 30P
Fig.A9.14
.08=- 00 at A and Dare
1
t-x necessary.
I 30"
I
Fig. A9. 16 To tind VD
-30
!1 take moments abou~
·90 -10 -30
Fig. A9.15 point A.

Fig. A9.l5 also shows the results or

calculating the ~s value for each portion of
,o
VD=2.5
HD"'l ZIlA = lOx6
=0
- 2~VD

the Ms/I diagram and its centroid location.

Substituting in the equations for the re- hence, Vo = 2.5
dundants we obtain,
To tind ~A t~ke ZFy = 0 =-10 + 2.5 + VA
=-Z0, = - (-45-405-300-900) _ _ , =0 hence VA - 7.0.
I'1o Zds I 3.2 - 01....,6 in .u ,
To find HD take moments about hinge at E at
all rorces on frame to right side of E and
X - Z0sY - (-45-405)9.375+(-900-300)(-5.625) equate to zero.
o - ~ - 3188
ZME = -2.5 x 12 + 30HD = 0, hence nD = 1.
=0.7939 lb.
Then USing ZFx =
0 for entire frame, we obtain
ZFx = -1 + HA = 0, hence HA = 1.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A9.7

The ~rame static bending noment Q~a~ram axes through the fr~e elastic cer.ter are then
~an now te calc~lated and dra'NTI as shawn i~ calculated as -12 ~d -0.625 inches ~espect­
Pig. A9.17. ively.
-30 ; - E0s _ - (-390) ;

I
(5) -30 Eds/I - 32 12.19 in. lb.

T
9.375 \ .
1
x = 12 ...J
(6) Xo ; Z0sY
Ix
= -548
3188
= -0 • 203 Ib
c ,
I" C
I ,,0s'-150- "'l50 = -Z~;X ; - t.;~0) = 0.1562
t\I I
1[1:1

I
.- -
,y=0.625'
-
~:i;c:n~er
i
I
Yo

The final ~oments at any point can now be

found by use of equation (19), namely
lb.

20.625
I 20 - - -(19)

Consider point B:-

y
x = -12, Y = 9.375, Ms = -30

1 Pin
Fig. AS. 17
D
Pin
SUbstituting -

Mg =-30+12.19-(-0.203)9.375+0.1562(-12)
= -17.76 in.lb.(checks previouS solu-
tions)

The Qoment diag~~ is l~beled in 6 parts Consider pOint 0:-

1 to 6 as indicated by the values in the small
circles on each portion. Most of the calcu- x = 12, Y = -20.625, Ms =0
lations frem this ~oint onward can te done
conveniently in table form as illustrated in Sub't , in (l9)
Table A9.2.
MO = O+l2.l9-(-0.203) (-20.625)+0.1562x12
Table A9. 2 = 9.82 tn Lb , (checks previous solu-
v

1 2 3 4 5 6 7 8 tions)

Mom. A' x = dist. Y = dist. The final complete bending moment diagram
I Os = from Y from X
Dia- Area would of course 08 the s~me as dravm in Fig.
gram of
Por- Mom.
for
Beam A
Section I
axis to axis to
C.g. of C.g. of
° 5, A,y A9.13 for the results of Solution No.1.
tion Dia. A, Os
Example ?~oblem 2.
1 ~450 3 -150 _12.00 _0.625 1800 93
Fig. Ag.1a shows arectan~~lar closed

I 2 . 60 2 . 30 _10.67 9.375 320 -281 fra~e supported at pOints A and B and carrying
the external loads as shown. The reaction at B
3 15 2 7.5 - 6.67 9.375 - 50 70
due to rollers is vertical. The ~rame at po~nt
4 45 2 22.5 . 4.00 9.375 . 90 211 A is continuous through the joint but the
react~on is applied through a ?in at the center
5 -180 2 - 90 8.00 9.375 . 720 -844 of 'the j oint . The pr obLem is to d et.ermtne the
6 _450 3 ·150 i2.00 -0.625 -1800
bending moment; c tagr-an.
93

I
'
Sum ~390 - 540 -658 w = 10*.1 in. .2401 (2) ~240

r
I
I .I ! ,L,L I ii'~""""-----7
IB • j
L :: 30 c: \1 B C 3
In order to take mo~ents of the ~5 values
I' 1
y4
L'24I~2f)·\ y
ii,
I
in cojumn (4) of the te.oIe , the centroid of
~~;4
I n
,P1"2M I 1=3

-~E. z 1
each ~ortion at the diagran must be ce~e~ined.
' T - - - ~I .-C:- x~
18
For example, the centroid of the ~wo triangular
bending morr.ent corticns marked 1 and 6 1s
12' y=9f6;'L=30 9.67 cut
x
,
.667 x 30 from the lower end or 20 inches as
I 1,A t
r I:2 D .1. 1 (5' 240
sho'Nn in ?lg. A9.l7. Thus the 1istanceS x "'" '""*' A\oo--15~ 240D
enc y from this 05 location to the y and x I Fig. A9.18 Fig. A9.18
Ms diagram for load Pi

I-I c
"~~""':'O:i'";;"

-: '~~:.-
A9 8 BENDING . :>10YlENTS, .
IN FRAMES ,
AND RINGS

T:,2 next 5,:e;:; :s ". C~.'J:Jse c' 5':;.<:1: .::':-:o..:T.e

----<~4500
00

c cnc t t ion en- ce t erntne stc '::1'.:: \"5 j~:-;;i '.:.; .' 'If

IE ~ ,a c moment ::2.::I""::.;7:. -. '::-,8

CC i s 5 ':=:L::. Lm ':7'.'::' frq-;:s _0
I
I i
assur-o-r cut on membe r .'B ~ "':'3~ -,:-::·;e -c i:::"': .. :3
l:il...ustr?..ted 1" ::.';s. •"::' ....10'-' J :'9 ar;c GiJ • ?':r

·~1
1 r ,
[E7. (7)
S1:::c:li:ity ::12 ,"._" me-tent c ur-ve :',2.5
' : ; - ' 0 " · ...
~ ~

cr-own in "'c:-...r-ee ~ar-:;5, ..' -'~ 83.C:".

\...
-;.~

. ':::=:::5:::::'-2,-
h ... ~ . .
"",.,..~
,,-~
~o"',.,

out I
--:- cut 9 l!1g only one of tns ... ext.er-nn l - ' t_",,,, i. -:;"'3.::'.5
.~,,-
,.,~o""
-~ 0
0"
, ,_.,

A
~-4500
D1
I,
iA
.D 3b ...
360
on the s t.r-uc tur e . ?L;s. A.~.::'2, ..
these resulti::~ bend tng ::J,c;;:snt
• C
, 20 S:~:'N
C::'::;V,:;S. ::- 9
po!'tions :J1' tnes a benc t ng ::lO::;:.en: ,j 1 '-':.:;-:'·:o..'~.s
Fig. A9, 19 Fig. A9.20 numuer-ec '- to 10 2.:-8 5 ['1 own tnc par-errtnes ~2 ,~

M, diagram fa, w loading M, diagram for P2 load. on each ;Jo:'':icn.

....
:''..18 nex"': sts'J '" ;:,'ae s-:lutioD ::':'.3:'3t3 '::l~
SoL;t~:n: r Inc tng the area of the M~/I 2i~;r~~s '"-7'.': 'tr;e
first ::lO!1!en'C Gf t ae se j .: 3. ~ :2..i7_S 2.::'0'..1.': tne x 3.:'.:::
T~,e first step is to find the j ocat i cn of y axes t nr-ougn t::e 8l3.S~~:: c ent cr-. T;;'858 s i:::~l
the elastic center of the rr-ame . Due to calC'-llations can jest '.::e cone in tab l e "er-n 3.3
sy.mnetry o~ rr-ane about the y ex.;s , 'the elastic i Ll.uat.ra ted in ~2.ble .'.9.3.
center 1.'111 be on a y axis 'thr-ough tt).~ middle
Table A9.3
o~ the rraae . The ver-t.t ce.I d~s"tance y
measured from a ax i s through AD equals, Portion Area Os ,
, dist. y , dist.
of M, of Ms to Y to X
Axis °sx I. OsY
3°x24~ Diagram Portion A Axis
-y = Z(dIS) Y = 4 ::>
3 x...1'",::;X<::, 372 ( ) , A I T ,
os = 33.5 = 9.67"
Z
T
30+( 24)2.30
4 3, 2 1 - 1440 3 ~ 480 -15 10. 33 7200 - 4958

rhe next step IS to deter::l.ine the elastic

2 - 72004- 1800 a 14.33 01· 25794

moment of inert ta of the rr-azae a::: out x and y

axes throug~ the elastic center of the f:-ame.
3 - 14403- 4801 15 10.33 - 7200i- 4958

4 14403 480 15 - 5.67 7206 i- 2726

~oment of inertia about x axis = Ix :- 5 36002 1800 5 , . 9.67 9000 1- 17406

Members An and DC, 6 - 45000 4-112501 , 7.5

i 14.33
- - 843751-161212
, 7 -108000 3 -36000 15 2.33 -540000 - 83880 i
Ix =( ~ x '3 X
k
14.33') 2 = 653.9
. I
1
8 67500 2-33750 5 - 9.67 -168750 326352
(1' x
'3 x 9.67') 2 = 2()0.9
9 32403 1080 1 15
I - 3.67 16200 1-
,
3064 !
Member BC 10 54002 2700 5 - 9.67 13500 - 26109

Ix = (30/4) (14.33') = 1540.0 Sum -77700 -747225 - 4645

(30/2)(9.67") = 1402.7
Solving for the red1.U1dants 'It tte i:la;:-t1.c
Ix = 3797.5 center,
Moment of inertia about y axis =Iy ,-
l'!:J
- .621<:: = -(-77700) = 20:;.2
=1:dS/T in.la.
38.5
Me:nbers AD and BC.
1 Xo = Z0IxsY = -4645
3797 =
-:;'.22 lb.
Iy = 12 x '41 x 30' = 563
1 1
= -~0'Jx
-(-74-7225)
'-2 x 2' x 30 3 = 1125 Yo Iy = ~289 = 14.1.28 :"b.

Members AB, CD, Fig. A9.21 shows these redundant r,JrC2S

acting at
. .. ';'9.22 .
t:-. 8 e las c ;c cen:er. .;;":J" ~.

24 ::>hONS tne bend.ing

j.
"'-;-- am cue ",,0 -0
moment c Iagr cnese ~

= 3600
r-edundant for-cas. 1';:',8 calculations 'Nith r-er ar-,
= 5289 ence to fig. A9.21 are -
'y
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A9.9

Total elastic weight of ring =

z ~ = It (36) ~ 113.1
I 1
i
The elastic moment of inertia about x and y
IT 9 67
axes through center paint of ring are the same
1 . for each axis and equal

4124
,'-------'r0 1 Ix = I y = nr~ = n x 18~ = 18300
Fig. A9.21
The next step in the solution 1s to assume
a static ring condition and determine the static
MA = 2018 14-1. '28 x 15 1.22 x 9.67 = -113 (Ms) diagram. In general it is good practice
to try and assume a static condition such that
M8 = 201S 141.28 X 15 + 1.22 X 14.33 = -84 the Mg diagram 1s symmetrical about one or 1t
~ = 2018 + 141.28 X 15 + 1.22 X 14.33 = 4153 posSible about both x and y axes through the
,- = 4124 elastic center, thus making one or both of the
MD = 2018 + 141. 28 x +0 1.22 X 9.67
r-edundarrt a Xo and Yo zero and thus r-ecuc Ing
considerably the amount of numerical calcula-
Combining the bending ~oment diagrams of tion for the solution of the prOblem.
Figs. A9, 18, 19, 20 with Fig. A9.22 would
give the true or final bending moment diagram. In order to obtain symmetry of the Ms
diagram and also the Mall diagram since I is
Example Problem 3. Circular Ring. constant, the static condition as shown in
Fig. A9.24 is assumed, namely, a pin at (e) and
Fig. A9.23 shows a circular ring of rOllers at (f). The static bending moment at
constant cross-section sUbjected to a sym- points (a ) , (b), (c) and (d) are the same
metrical loading as shown. The problem is to magnt tude ann. equal,
determine the bending moment diagram.
50 50. Ms =50(18 - 18 cos 45 0) =265 in.lb.

The sign is pOSitive because the bending moment

produces tension on the i~lde of the ring.

The next step is to determine the ~s, ~sx

and :2l s Y values.
¢s 1s the area of the Ms/I dtagr-am, how-
ever since I is unity it is the area of the ~
diagram. The static Ms diagram of Fig. A9.24
is divided into similar portions labeled (1)
Fig. A9.24 and (2). Hence
Fig. A9.23
0s(~) = area of portion (1) = Pr~(c -sin ~),
M=88"* where P = 50 lb. and a = 45°.

I~
i0~rack:~r~ ~\\', ~
SUbstituting and multfplying by
are four portions labeled (1).,
4 since there

e ---+-Xo=CX: M=-177"# -1~­ 05(.1 ~ 4[50 x IS" {0.785 - 0.707)J ~ 5052

0' J I- j
.
Fig. A9.25
"~MO"_177".
,
'
-;
.
~~
.,~

M=88"#
-,....-,_.-,c-,-;"o:
"
\~
The area of portion labeled (2) equals,
Prae(l - cos a).
Moments due to redundant
forces. Fig. A9.26 Since there are two areas (2) we obtain,

Solution. Due to s~etry or the ring

05 {" )= 2 [ 50 X IS" X; (1-0.707IJ = 15000
structure the elastic center falls at the Hence, total 0s = 15000 + 5052 = 20052
center of the ring. Since the ring ~as been
ass~ed with constant cross-section, a relative
Since the centroid of the Xs diagram due
value of one will be used for 1. to symmetry about ~otn x and y axes coincides

( 1·1
A9.10 BENDING MOMENTS IN FRAMES AND RINGS

with the center or el~stic center of the fra~e, Solution:

the te~s Z¢sx and Z¢sy will be zero.
The fi~2t step 15 to 1ete~lfle the elastic
SUbstituting to determine the value of the center of the f~ame and the elastic ~o~ents of
redundants at the el~stlc center we obtain, inertia. ~able AG.4 shows the calc~latlons. A
reference axis x'-x' has jean selected a~ the
Me
= :Zd0S"
~ ~
= - (20052)
113.1 = - 177 in. lb.
~icpeint of the side AB. Since a static :r&T.e
condition has been selected to ~ke the Ms
diagram sJ:mnetrical ~bout y axes t;Iough the
= Z0,y = 20052(0) = 0 elastic center (see Fig. A9.2S), it is ~ot
Ix 18300 necessary to determine I y si~ce the redundant
Yo will Je zero due to this sJ~T.e~ry.
Yo =-l:0,x
Iy
=-(20052)(0)
18300
=0 Table A9. 4

Length cis
Fig. A9.2S shows the values at the acti~g Member cis I w' "f y 'NY IX'X' : tx .. wy2
elastic center and the bending moment diagram
prOduced by these forces. Adding the bending BDB' 94.25 1.5 62.8 t!: .. 30 3080 54.00.. 151200: 156600
moment diagram at Fig. A9.25 which is a constant n
value over entire trame of -177 to the static : 49. i
moment diagram at Fig. A9.24 gives the tinal
bending moment diagram as shown in Fig. A9.26. AB 60.0 1.0 60.0 0 0 18000 ... 0 , 18000

Example Problem 4. Hull Frame A'B' 60.0 I. 60.0 0 0 18000 ... 0 , 18000

AC 38.4 2. 19.2 ·42 806 922 .. 33850 , 34772

Fig. A9.27 shows a closed frame subjected
to the loads as shown. The problem is to CA' 38.4 2. 19.2 -42 806 922 .. 33850 , 3477~
deter.nlne the bending moment diag~.
Totals 221. 2 1488 262140

odl){)"
lJ
In the last col~~ of Table A9.4 the te~
l x is the moment of inertia of a particular
member about its own centTQidal x axis. T~~s
for member 80B;
D

static moment z.~, i a

x ;: .3!a = .3 x 30 /1.5 = 54C0
curve

For members AC and CA ,

T A~ L'38.4 _ 1

~ ~:_jY
tx - 12 b L h"
h'"''
Fig. A9.28
-L ~~~~. = 12x2".....
1 1""3' '''04'-920
'-'. "±A~ - -

Let y = distance from X'X' ret. axis to

centToidal elastic axis X-X.

"wy
y=-zw= 1468
221.2 ;z 6.64 in.

By parallel axis theorem,

Ix = Ix! - 6.64"(Zw)
262140 - 6.54"(221.2; = 252400

The next step in the solution is to ccm-

puts the static moment elastic weights 03 and
Fig. A9.29 their centroid locaticns. In Fig. A9.28, the
Fig. A9.30 static frame condition assumed 15 a pin at point
A and rollers at point A' , which gives the
 ANALYSI~ AND DESIGN OF FLIGHT VEHICLE STRUCTURES

general shape of static ~oment curve as shown 38 . 4.

in the f Igur-e ,

Cons i der- member- BDB P t

x=
1/2
10o (4690 x 100 x")xdx

78400

T~e bending moment curve will be con~ 38.4-

sl~ered
t.er-n ~3
in two parts, namely (1) and (2). The
represents the area of the Ms/I
= 1/2 [469~ x' . 10~ xj , 21.77'
c tagr-an. o
784.00
Thus for pottion (1) and (1' )
Vertical distance from line AA' to centroid =
~3('l + ~s(") • 2b"(d. sin 0] 21.77 x 24/38.4 13.~". =
The static moment
weight for Ale' 1s same as for AC, thus
= 2@00~.~ 30"(0.524 - 0.5~ "172800
~AC + ~A'C' ~ 2 x 784.000 = 1568000
1he vertical dlsta~ce from the line BS! to
centroii of !"'Is curve tor portion (1) and (1') rig. A9.29 shoWs the trame With the moment
Is, weights ¢a located at the centroidS, together
WIth the redundant forces Me and Xo at the elas-

y ,
r{l ... cos a - - - - )
sin au
2. ,
30(1 • 0.867 - ¥) tic center. It makeS no dltterence where the
frame t s cut to form our residual cantilever,
a - sin u 0.524 - 0.5 it one or the cnt races 1s attached to elastic
center and the other Is considered fixed. With
= 10 in. the elastic properties and moment wel~~ts known
the redundants can be solved for:
For portion (2) of the Ms diagram the area of
the ~/I diagram Which equals ~s 1s _(,.=.10",Cf7=0:::00~+ .=1,;;72;:;8,.,0;;0...:+...:1:::5",6",80",0'-".) ,
- 221.2

,05(.:1) =PTrlitj (1 - cos a)= 6000Xl~~"X2.1o._a.967) 124.30'#

= 1,007,000
X
o = ZX'sy
I xx
::

Consider member AC, =,,10",Cf7=0"0"0-,x,-,48=.-=1,,6...:+~1,,7.;;2:;;8*00,,,,,xC'33=. 3"6::...:.+-=1"5,,6,,60,,,0,-x,,,--,,5"'.2=4

252400
~rom free body clagrarn of bottom portfon
of frame (Fig. A9.31~ the equat~on for bending = -98.5#
moment 6000 ;4690 6000 Yo 1s zero because of the symmetrical frame
. . . . . . . . 11 a.~dloading. The f1nal or true bending moment
A -- A'
I at any point equals
I~ C
>0.,"'7 Thus for point 8

on ~ember AC equals: MB = 0 + (-12430) - (-98.5 x 23.36) =

-10130'~
Mx :: 4690 x -100 x Q
For point C
Area of ~/I c~rle between A a~d C when I = 2
aq\..ials, Ms at paint C = 4590 x 38.4 - 20OX{38.4)a/2
= 32700
1/2
% o
38 . 4
Hence Me = 32700 - 12430 - (-98.5 x -60.64)
14.300'~
=
0]38.4.
= 1/2 [4690 x" _ 100 X = 784000
fig. A9.30 shows the general shape of the true
L 2 3 Jo ~ra~e bending ~cment diagram.

Example Problem 5
Distance to centroid of ~/I curve along line
;"C iroJ1 A. ?ig. A9.31 shows the general details of one
A9.12 BENDING MOMENTS IN FRA:YIES AND RINGS

The moment of inertia of the frame

half of a SYTI'-T.8t:'iC3.1 hull rreme that was used cross sections are given in column
in an actual seapl~ne. The ~ai~ external load 2 of Table /1.9,6 which have been
on such frames is ~he water ~ress~e cn ~he ~ul: determined from a consideration
bottom pl~ting. The hull bottom s~rtngers of the actual dimensions of the
transfer the bottom pressure as cancentrations frame members.
on the ~rame bottom as sho,vn. The resistance to
this bottom upward load on the fr~~e is ;rovlded
by the hull metal covering Nhich exerts tangen-
tial loads on the frame contour. The q~estlon
as to the distribution of these reSisting
forces is discussed in later chapters. In this
problem the reSisting shear flow in the hull
sheet has been assumed constant between the
chine paint and the upper hea~J langeron. ~cr
analysis purposes the frame has been divided
into 20 strips. The centroid of these strips
located on the neutral axis of the fra~e sec-
tions are numbered I to 20 in Fig. A9.2l. The
tangential skin reSisting forces are shown as
concentrations on frame strips #6 to #16. On
the figure these tangential loads have Jeen
replaced by their horizontal and vertical com-
ponents. The sum of the vertical components
should equal the vertical component of the
bottom water ~ressure.
Table A9.5 showS the complete calculations
~or determining the bending moment on the f~ne.
Col~s 1 to 7 give the calculations for the
elastic properties of the frame, namely the
elastic weight of the frame; the elastic center
location, and the elastic centroidal moment of
inertia about the horizontal ~entroidal elastic Bottom member of frame
axis. A reference horizontal ~is X'X' ~as
been selected as shown. All distance recorded
in the table have been obtained by scaling fram Floor Line
a large drawing ot the fr~~e.

The static condition assumed for computing

the Ms moments is a double symmetrical canti-
lever beam as illustrated in Fig. A9.32. The
trame is cut at the top to form the free end
of the cantilever bea~, and the fixed end
has been taken at the centerline bottom frame
section. The static bending moment diagram
will be symmetrical about the y axis through
the elastic center of the frame and thus the
redundant Yo at the elastic center will be
zero since the term Z¢sx will be zero. The
calculations tor determining the static
moments Ms in Column 8 of Table A9.5 are not
shown. The student should refer to Art. A5.9
I' Fig. A9.31
I~

of Chapter AS to refresh his thinking relative

to bending moment calculatlor$ on curled beams.
eutra
Axis
I~
~~
l~

Columns 9 and 10 give the calculations of

the 0s values (area at Ms/I diagram) and the
first moment (0 sx values). The summations of Bending Moment
columns 3, 9, 10 permits the solution for the Diagram
red~~dants Me and X as shown below the table.
o
The final bending ~oment M at any pOint on the Fig. A9.33
frame is by Simple statics equalS, Fig. A9.32
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A9.13

Table A9. 5

Calculation of frame Elastic Properties Calculation of Moment Weights and Solution of Redundants

CoL 1 2 3 4 5 6 7 8 9 10 11 12 13

0 Elastic Arm w y' w y'2 Arm Static Moment es y ~10 -Xoy -Total

- " =.fiE: :
Weight to '0 Moment Weight Moment

- I=~
0
Ref. Axis Os :: M
~,j = ds
~
w Axis XX MS
~
" ~.= 1 X'X' ~ Y 1 0
ds I :: y'

1 7.6 .065 117.00 26. 1 3050 79700 17. 2 0 0 0 -9020 16650 7630
0 0 0 6000
2 7.6
7.6
.065 117.00 24.4 2850
2497
69500 15. 5
0 0 0
"
.. 15020
12100 3080
3
7.6
.065 117.00 21. 4 53300
35020
12. 5
8.4 0 0 0 . 8130 --6020
890
4 .065 117.00 17.3 2025
3. 1 0 0 0 ..
6
5 7.6
6.4
.065
.25
117.00
25.60
12.0
5.9
1403
151
16820
892 - 3.0 1220 31. 2 - 637 93.7 .. -
3000
2910 ·10710
- 8. 9 - .. - 8610
7 6.4 . 25 25.60 0
- 6. 3
0 I 0 2880 71. 6
- 6160 .. -14840
8 6.4 .25 25.60 -161 1017 -15. 2 6230 159.3
286.5
2420
- 13220 .. -14720
-20830
-17510
-18590
9
10
6.4
6.4
. 25 25.60 -12.6 -323 4060 -21. 5 11260
18470 473.0 - 21250 . -27100 -17650
6.4
.25
.30
25.60 -19. 1 -488
-543
9320
13850
-28.0
-34.4 29020 618.0 - 33100 .. -33300 -13300
11 21. 30 -25. 5
..
.- 52400 -
12 6.4 .34 -32.0 -602 19250 -40.9 43180 814.0 -39700 5540
6,4
18. 82
27600 58980 1110.0 .. -45600 4360
13 .34 18. 82 -38.3 -720
_840
-47.2
--111300 ..
14
15
6.4
6.4
.34
.3'
18. 82
18.82
-44.6
-50.8 -959
57500
48750
-53.5
-59.9
77330
99030
1455.0
1865.0
77800
.. -51700
-57900
16610
32110
111700 -61100 41600
.....
~.hine -63.0
16 7.0 12.0 O. 58 -55.4 - 32 1780 -64. 3 149700 86.7 - 5570 -62200 78500
f-: 7, , 7.0
7.0
12.0 O. 58
O. 39
-58.4 - 3'24
- 131
1980
1470
-67. 3 221400 128. 3 -
-
8650 -65100
.68100
147300
194100
18 18.0 -61. 4
-
-70. 3 271200 105. 8
-
7440
.
19
20
7,0
7.0
34. a
50.0
o. 21
O. 14
-63. 2
-64.8 - 9
840
588
-72. 1 I 3086()0
-73.7 327900
64.9
45.9 -
4680
3385 .. -70000
-71400
229600
247500
Keel -75.0 332900 I -9020 - 72600 251280
Totals 811.48 7228 423237 7315.0 -348106

Y = 7228/811. 48: 8.9" M = -~Os :: • (7315.7 x 1000) = -9020"#

o '-'w 311.;:1
IXX:: 423237 - 811.48 x 8. 9 2 :: 358940
x, :: Z?s y::. [ 348106 x 1000J:: -968#
XX 35894

"Total. Moment M at any statton e Ms ... Mo - XoY

2401b
Columns 11 and 12 record the values of l.-. 6' 6" ------ooi
Mo and -Xoy for each station poIrrt . For
examp.Ie , trie va Iue of -Xoy for station (1) c , -r-r-
equals - (-368 x 17.2) : 16650 and for stat~on
(20) = -[-968(-73.7)J = -71400,
T 6"
,
1

Y::5.242
I

Fig. A9.44 shows the shape of the final I !_l_!..._x I 10"

'1
~cment carve as the result of the values in 120~ :E.C.
c ojunn 13. ~;1
I
I x- --.. h
A9.5 Unsymmetrical Structures. Example prootem Solutions. 9" I 5. 032
.L

Fig. A9.34
Fig. A9.34 shows ~n l~~s~~etrical frame
carrying ~he leads as shewn. Dete~ine tte
jenQ~ng no~ents at p01~ts A, B, C and D.
Soluti.::m: -

The elas:ic weight of ~ra~e = Zds/1 =

(15/1) + (12/2) + 10/1 = 31

.' I '1' ...,

A9.14 BENDING MOMENTS IN FRAMES AND RINGS

The distance x f r'cm tne j me AB to the Fig. A9.36 shows the centraid loc~tlQns uf
elastic center is, the 05 values along the center line of the
frame. The ~cment of these 0s values about tte
x 6 10 x 12
x = (15)0 + 6
31
+
= 5.032 in.
x and y axes will now be calc~lated.

Z0sx = (-2160)(-3.032) + (-21600)(-5.032) +

The dist~nce y from line Be to elas~lc (-4260)(-5.032) = 139700
center equals,
Z0sY = (-2160)(5.242) + (-21600)(-2.258) +
y = 15 x 7.5 + ~lX 0 + 10 x 5 = 5.242 in. (-4860)(-6.758) = 70290
The 'ffilues of :he recunGants at t~e elastic
The moments of inertia Ix and I y
elas~ic center can now be cal~~lated USing equations
and the product of inertia I xy are ~eGuired. (ll), (14), (15) J nameIy

Ix = (~)' + (15)(2.258)' + (1:)(5.2421" + (i~)' Mo = ;;~~7I = - (-;~620) = 923 In. lb.

+ (10) (0.242)' =606.51

1'0
J
= (15)(5.032)' + (12)' +(E)(o 968)' +
2x12 2 •

(10) (6.968)" = 942.96 70290 _ 139700 (217.74)

IxY = (15)(-5.032)(-2.258) + (~)(0.968)(5.242)
= 942.~6 = 68.46
217.74' \
606 51 (1 - (606.51)(942.96)'
.
+ (10){6.968)(0.242) = 217.74
The next step 1s to assume some static
= - [Z0sx - Z0sY (~)J
Yo I 4
trame condition and draw the static bending I y (1 _..::s:L)
moment diagram. Fig. A9.35 shews that the rxry
frame has been assumed cut near pOint C which
gives ~NO cantilever beams. The oending _ 1139700 _ 70290 ,217.74)J
moment diagram in three parts for this static = L.: '606.51
217.74""
;
-132.36 lb.
condition Is also shown on Fig. A9.35. 942.96 (1 - (606.01)(942.96)

B ISl=~2160 C The bending moment at any ~olnt from eq.

(",~I44l:ii£0s:S:i(I,.)_ --1,cut
T'i . C T (19) equals,

1_f:Jx
5.242

t U· l
6"
11 • Ms + ~o - XoY + Yox, for exampl e ,

t'" 2)
D 9.758
'1-
258
OS2'::-
. C.
600 6 750
Consider point A.

x = -5.032, Y = -9.758, Ms = -2520

1 "I ~1080
M s diagram 10S3"'-~60
A .o~
MA = -2520 + 923 - 68.46 (-9.758) +
(-132.36)(-5.032) = -263 in. lb.
Fig. A9.35 FIg. A9.36 Point 8. x = -5.032, y = 5.242, ~s = -1440
The 05 values which equal the area at
the Ms diagram divided by the I values of the Ma = -1440 + 923 - 58.40 x 5.242 + (-132.36)
part1culqr portion will be calCUlated. (-5.032) =- 209 in. lb.

0s, = 3(-1440)/2 = -2160 Point C. x = 5.968, Y = 5.242, Ms =0

~s. = -1440 x 15 = -21600 Me =a + 923 - 58.46 x 5.242 + (-132.36)
(6.968) = -357 In. lb.
0s, • 4.5(-1080) = -4860
ThuS, the com~lete be~ding ~cment diagr~
= -2160-- - could oe determined by computing several more
:0s 21600 4860 = -28620
'/alues such as ~oint D ar.d the ~xternal load
points.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A9.15

Exa~ple Problem 2. Fig. A9.37 shows a~ ~n­ Total Ix =6451.07

symmetrical closed ::."rame. The bending ~oment
diagram will be deter.nlned under the given Calculation ot I y -
rr-ame loading.
Member AB, I y = 13.~ x IS' +.lx
12
30(0.667)' :

4500.7
P2 = lOOt _ 1
I Member BO, I y - 'j] x 31 367 x 30' = 791.7
Member CD, I y -T
- 20 X IS' + {z x 20(11" : 4501.7

Member AD, I y -_ 12
1
x:r1 x 30'
= 1125.0
Totally = 10919
Calculation of Ixy -

11ember AB, I xy = 13.~ (-151(3.343) = -1002.9

,~---30"'---
ly' Fig. A9,37
31.67 1 31.67
Hemoer BO, Jxy = -3-(13.343)(0) -12 x - 3-
Solution: -
(30)(10) = -253.92
The elastic weight of frame equals
~
'<' =1.5
30 -+- 31. 67 + 3.Q. -+- 30 = 6!::; ,1:8
Member CD. Ixy = 21° (15) (-1.657) = 497.1
"I 3 1 2 ~.,
Member AD, I~J =0
The location of the elastic axes will be
determined with ~eference to assumed axes Tot~l I~J = -17G3.9
x'x' ~~d y'yl as shown on rig. A9.37.
The next step in the solution 1s to assume
a static fra~e condition and draw the Ms dia-
= ~(0)+(¥)(151+(¥)(301+(4f1(151 = gr~. Fig, AS.38 showS the assumed static
x 65.58 15' condition, namely pi~~ed at paint A and supported
as rollers at ?oint D. The bending ~oment
dia 6 raID is dra~m in parts as sho.vn.
(fj (151+(¥) (251+FfI (101+(4f1 (0) 750"
y = 65.58 = ~uetoP2

11.657'" 600 I' 121 <Due to Pl

600
These distances x
and y locate the x and
"",1111
y e1ast!c axes as shown in Fig. A9.37. \ I

Due to Pl~l,
"

The elastic ~oments of inertia and ~he c

elastic product of inertia will now be cal-
culated. Fig. A9.38

Calculation at Ix -

dem
Y bar·AB , -x
T-lx-l..-(,e3'''+'1'''-7'')=1'7.
3 1.5 ~ . "'tV ..l. .<)~. . . .-;;:~1
0 . .:::. . . . .
=D

Member CD, Ix=~ x i (8.343"+ 11.657") =721.58 ~

{41 ~
~---=m"""'--"- Due to P 3 and P4
_50 11 -;)UU

Member Be. Ix = (31367}(~3.343<i) + l~ (31;367)

=1967.48
(10" I
The ~ext step is to compute the '~lue of
Member AD, Ix = (3 (l1.657::1) 1
+12x 30 x
0s for each ?ortion of the ~oment di~gran. 0 5
20) is the area of the !'1s / I diagram. For reference
(0.5)" =2038.·30
the ~ortions of the Ms diagram have been labeled
1 to 4.

.t»
A9.16 BENDING MOMENTS IN FRAMES AND RINGS

~co 30 The ~~j ~oir.t e~~als the

b~r.ding ~om~~: ~t
?ortlon (1) • 03 = -.~-
G
x -.-.- = ::000
1.;J o~i~:na1 sta:::.c Ms ~lus the ~omen: due to the
r-eduncant forces as S:lO'.\TI in !"i.;;;. ).'; ... 0.
600 31.67
3-
03 =-Zx-- =
Portion (2) • ~'t:~
o,~

B
" ,
Portion \0). e;s ~
31.67
-2- x ---t-
T-O
:: 39:~8
I~
-165

18.343! ~l~C-. ~
Portion (4). - el s

z0's = 13125
= -5CO(120) _500(1 ° )
2
= -,3000 !
L.-i
:

~ :4.674
~
. 8.343
"31. 1)71-;
.

11.657 \
- 5000 :: 8125
-Ai
;""-15
:
11--"":""'-15''------1
iD.L- 167,§,
167'4;j[~307
fi'"1307

The individual 0'8 values are now con- I . • 286 240

centrated at the e.g. of each of the Ms dia- Fig, A9.40 Fig. A9.41
grams. Fig. A9.39 shows the location of the Consider ~oint A: - Ms :: 0
~s values with respect to the x and y axes
thro~~ the elastic center.
MA = a - 124.8 - 4.674 x :5 + 31.07 x 11.657
= 167 in. ::''::1.
0s:3167
I
I I
I
Point B. Ms :: 600
'....5_ I
Ds~6000~ ,
Ma = 600 - 124.8 - 4.674 x 15 - 31.07 x 16.3~
; ~ 15 13.343
l8. 34 31 I = -165 in.lb.
I ' •
,
I' I
!
11. 657
~

Point c. !'Is =0
. .
i
~ 0 s .:5000
---
Me =0 - 124.8 - 31.07 x 8.343 + 4.674 x 15
= -314 1::1.1'J.
r-- 15'-L.........j
Fig. A9.39 Point D. Ms:: 0
Z0sx = 6000(-15) + 3167(-5) =-105835 MD = 0 - 124.3 4.574 x 15 + 31.07 x 11.657

l0sY = 6000(8.343) + 3167(15) + =307 in. lb.

3958(13.343) -5000(-11.657) = 208659 Fig. A9.41 shows the true bending moment
diagram.
The values of the redundants at the elastic
center can now be calculated. AS.6 Analysis of Frame with Pinned Supports.

MO =~
Zds/I
~ -(8125)
65 •.58
= -124 • 8 1
n,
1b
•
Fig. A9.42 shows a rectangular frame and
loading. This frame is identical to example
problem 1 of Art. A9.3, except it is pl~ed at
I points A and D instead of fixed.
= Z0s y - Z0s x ~)
x., Ix (1 _ ;xy') The first step will be to dete~ine the
x"y elastic weight of the fr~ne, the elastic center
location and the elastic moments of inertia
=208659 - (-105835)(-~) = about axes through the elastic center.
6451 [1 -
10919
(-1753.9')
x 6451
J 31.07 lb.
The term dslEr of a 'Jeam element of length
ds represents the angle change be~Neen its ~#o
end :aces when the element 1s acted upon by a
=- [l0sX - l0sY (-¥;~ unit moment. In this chapter this tern has
Yo I J been called the elastic weight of the element.
I y (1 - IXly ) PhYSically, the elastic weight is the ability
of the element to cause rotation when acted
upon by a ~it moment. When a unit ~oment is
= -[-105835 208659C1753.s11
6451 ~_
10919[1
-
(-1753.9')
10919 x 6451
4.674 lb.
J- applied to a rigid support, the support suffers
no rotation since the support is ~igid, there-
fore a rigid support has zero elastic weight
and therefore does not figure in the fr~e
elastic properties.
 A9.17

If a su~port is pi!h~ed or hinged it ~~s

no resistance to rotation and thus a unit
_14~!30'54~14.46
moment acting on a hinge would have infinite 4 46
angle change or rotation and therefore a hinge -14'n
or pin possesses infinite elastic weight.
10 lb.
,-:»:
oo!!J "
-:
~s=270
~' I

"F' 1I"
L
230
i !
,I
Fn
-ld(~L
Fig. A9.45
Fig. A9.45 shows the bending moment dia-
gram due to the redundant XO ' Adding this
diagram to the original static diagram gives
Fig. A9.46

A B '" yJE.C.
Y
the final bending moment curve in Fig. A9.40.
Fig. A9.42 Fig. A9.44 Fig. A9. 43 Ag.7 Analysis of Frame with One Pinned and One Fixed
Support.
Due to symmetry of structure about the
centerline y axis the elastic center will lie Fig. A9.47 shows the same frame and loading
on this axis. Since the two hinges at A ~d as in the previous example but point D is fixed
B have infinite elastic weight, the centroid instead of hinged.

n4i
or elastic center of the frame will obviously
lie midway becNeen A and B. Fig. A9.43 shows The support D
the elastic center 0 connected to the point A has zero elastic
by a rigid bracket. weight and the pin
at A has infinite
tor frame is infinite because or
~ds/EI
elastic weight,
the hinges at A and B. I Fig. A'. therefore the elastic
center of the frame
The elastic moment of inertia about a y
axis through elastic center is infinite since
x-Ai - - x D lies at point A.
Tne total elastic
y wei~~t of frame 1s
the hinge supports have infinite elastic weight.
infinite because of
Ix is calculated as follows: - pin at A.

=..lx.l.x The elastic moments of inertia will be

For Portion AB
3 3
30' = 3000
calculated about x and y axes through A.
For Portion CD :J:..xl x 30' = 3000 Ix = 16800 (Same as previous eXaID?le)
3 3

For Portion Be = 24 x ~ x 30A • 10800 I y = (3 x24' )+(; x ~ x 24') = 8064

30

Ix =- 16800 Ixy = (;0 x 15 x 24-) + (224 x 30 x 12) = 7920

Fig. A9.44 shows the static frame con- The static frame condition will be assumed
dition assumed to obtain the Ms values. the same as in the previous example problem,
hence 0s = 270 and acts lOW from B (Fig. A9.48).

r,I
The value of 0s for member Be equals the
area ot the Ms curve divided by I tor 8e, / ' " 5270
hence ¢s = 4S x 24 x 1 = 270. The centroid s - l1S ' n-11.92-~ A-11.92
2' 2' r~lO...i "' W1l
of this ¢s value is 10 inches from pOint B. rw ~

t
'I I 28.45 'a
'~e redundant forces at the elastic center can 3~' r ~
now be solved for ' !MO
tl;!- Xo """ 5038#
~
'i 19
~1 •
~6.1'
, ~.
-270
ir.f:ni ty 'Yo i. 2582* Fig. A9. 50

Fig. A9. 48 Fig. A9.49

y
a
= -Z¢5
1y
X = -(270)10 = 0
infinity
Since the frame is ~~symmetrical, the x
and y axes through the elastic center at A are
not prinCipal axes, hence
_ ~0sY _ 270 x 30 =0.482 lb.
Xo - iX - 16500
 CHAPTER AlO
STATICALLY INDETERMINATE STRUCTURES
SPECIAL METHOD . THE COLUMN ANALOGY METHOD
AID.l General. The Column Ana Logy- method is a
method tr~t 1S widely uSed by engineers in ce-
termlnlng the bending moments in a bent or
~lng type structure. The method considers
only distortions due to bending of the
structure.
The numerical work in using the column
analogy method 1s practically identical to
that carried out in applying the elastic center
method of Chapter A9.
AID.2 General Explanation of Column Analogy Method.
Fig. AlO.l shows a short column loaded in
compression by a load P located at distances
(a) and (b) from the principal axes x and y ot
the column cross-section.
Fig. AlO.!
To tind the bearing stress between the
supporting base and the lower end of the
column, it 1s convenient to transfer the load
p to the column centroid plus moments about
the principal axes. ~hen it we let cr equal the
bearing stress intensity at same point a
distance x and y rr-ce the yy and xx axes, we
can wr t te
wnere A is t~e area of the calumn crass-
section and pa and Pb the moment of the load
P about the xx and yy axes respectively. If
we let Pa = Mx and Pb = My, the above equation
can be written,
- - -
-Method of Analysis due to Prof. Hardy Cross. See "The
Column Analogy" Univ.lll. Eng. Expt. SU. Bull. 215.
Now assume we have a frame whose centerl~ne
length and Shape is identical to that of the
column section in Fig. AIO.I. The width of each
portion of this frame will be ?rOportional to
lIE! of the member. Fig. AlO.2 shows this
assumed ~e. Furthermore, assume tr~t end B
of the frame 1s fixed and that a rigid bracket
is attached to the end A and terminating at
point (0) the elastic center of the frame. The
frame is SUbjected to an external loading,
W"l.1 w{I, etc.
Y
Ell I
w,
x-
.i,
EX
:f),Mo
'-Xo
jyo --r
-ly
A
f-x :11-
Fixed
Fig. AlO.2
This cantilever structure wIll sutfer bend-
ing distortion under the exterr~l load system
W)., Wv e'tc , , and point (O) will be displaced.
Point (0) can be brought back to its original
undeflected position by applying a couple and
two forces at (0), namely, Mo, Xo and Yo as
shown in Fig. AlO.2. Since point (0) is attached
to frame end A by the rigid bracket these three
forces at the elastic center (0) will cause
point A to remain stationary or in other words
to be fixed. Therefore, for the frame in Fig.
AlO.2 fixed at A and 8, the moment and two forces
acting at the elastic center cause the static-
ally indeter.ninate moments M1 when reSisting a
given external loading causing static ~oments
MS. The fir~l true bending moment M at a paint
on the frame than equalS
M=!'1s-+!"!i'
From ?ig. AIO.2 we can 'Nrite for a point
on the frame such as 8 t~at the lndeterminate
bending moment Mi equals,
M. i =:1
0 + Xoy -+ Yox - - - - - - - - - - (2)
(1) In Crapter A9, Art. A9.3, the equations
~or Mo, Xo and Yo were derived. They are,
AlD. I
AlO.2 THE COLUMN ANALOGY METHOD

(4-) M at any
~ - Z0sx
Mo = Zds/I ' YO = Iy

The ter.n z0s represe~ts the area of t~e

static Ms/I curve. (E r2s been assumed con-
stant and therefore emitted). Let :~e t&~
Z0s be called the elastic lead and s~ve it a
new synbel P. The term Z15/1 equals the ?ig. AIO.3 shews a rectar.g~la~ ~r~~e with
elastic weight of the fr~~e and equals the s~ ~~xed supports ~t ~oints A anj D. The bend~ng
o~ the length of each ~ember tl~es its width ~o~ents at ~oin:3 A, 5, C and D will Je de-
which equals 1/1. Let this total ~~~~e elastic te~ined JY the cc~~~n a~alogy ~~t~ed. T~is
weisht be given a new 5J~bol A. ~r~~e and load~ns is identical to ex~~?le
problem 1 of Art .."-~.? or c-a-ter ;:.g'Nnere the
In the exoressions ~or Yo a~c Xo the sclution was ~aQe JY the elastic cen~er =e~hoG.
terms Z0s x and Z0sY represent t~e TIcnent of
the static Mil curve acting as a load abcut 10 lb. 45"#
the y and x axes respectively passing t:;rcugh I
~
-16" 18"--1
the frame elastic center. Therefore let Z0s x BI;-.L_ _ ---:c .. ~
be given a new symbol My and Z0sY a new I : 2
symbol Mx • With these new symbJls, equation L:. 24 I I

(2) can now be re'NTltten as follows: -

L'30,,1
M1 - p +
-A
E;£.
-x
+ M{X __
Y

Comparing equations (1) and (3) we see

- - (3) 1-3

~
10
j7. 5
they are similar. In other wordS, the inde- 2.5
Fig. AlD.3
ter.ninate bending ~oments Mi in a trame are Fig. AlO.4
analogous to the column bearing pressures 0,
hence the name nColumn ~~alogyn for the ~ethod SOLUTICN NO. i
using equaticn (1). ~ith this ~eneral ex-
planation, the method can now be clearly ex- ~e first c8r.sider the ~r&~e centerline
plained by g~ving several example prablem shape as shewn in Fig. AIO.3 as the crcss-
solutions. section of a shart column (see rig. AlO.5). The
width of each portion of the column section is
AlO.3 Fra.mes with One Axis of Symmetry. equal to llEI of the ~ember cross-section.
Since E is constar.t, it ~i1l be ~de w~i~y a~d
From the previOus discussion we can write, the widths will then eq~al 1/1.

(1) The cross-section ot the ar.alogous column

consists ot an area, the shape 0f which
1s the same as that of the given frame
and the thickness of any part equals
liE! of that part.
(2 ) The loading applied to the top of the
analogous col~ is equal to the Ms/EI
diagram, where Xs 1s the statical moment
in any basic dete~inate str~cture de-
rived from the given frame. If Ms
causes bendi~g compression on the inside
face of the frame it is a positive bending
moment and the analogous load P on the I Fig. AlD. 5 Fig. AlO.6
column acts downward. i
The first step in ~he ca:culations is to
(3) The indeter.nlnate bendln~ ~oment M1 at a compute the area (A) of the column cross-section
given frame ~oint equals the base - I in Fig. AlO.S and the moreents of inertia of the
pressure at this same poi~t on the column cross-section about x and y centroidal
analogous col~14~. Thus the indete~inate axes.
moment at any paint on the frame equals
(fromeq.3),

The calculation of the location o~ ~he

centroldal axes and the moments at :nertia Ix
 AlO.3

and I, woulc be identical to the ccmnlete = =

Mx 270 x 9.375 2530 (positive because
calcufations given in Art. A9.4 J and-Table A9.l 8ase pressure is compressive on column portion
where this sa~e problem is solved by the elastic above x axis).
ce~ter ~ethod. These cal~ulations Nill not be
re?eated here. The results Nere J Ix = 3188 My =- (270x2) =-540 (negative because
and 1y =
3456. baSe oressure is tension on column ~ortion to
right" of yaxis).
Since the fra~e is statically indetermin-
ate the next stea 15 to assume a static frame Substitution in equation (3).
condition consiste~t with the given fra~e and
loading. Fig. AIO.4 shows the condition Frame Point A. x =- 12"J y =- 20.625"
assumed for this solution J namelYJ pinned at
point A and a pin with rollers at point D. I1i =1:A + MxY + MlX
The static Ms diagram is therefore as shown in Ix y
Fig. AIO.4. ~e now load the column cross-
section with the Ms/1 diagram as a load as - 270 2530 (-20.625) + (-~5~-12)
shown in ?ig. AlO.S. The static mocent Sign l'l1-32+ 3188
is positive because the sGatic condition causes
= 8.44-16.38+1.88 =-6.0. tn Lb ,
tension on the inside face of the frame. In i

the column analogy method a positive Ms/I

loading is a downward or compressive load on The true bending moment fram equation (5) J

top of cOlumn J and therefore a negative Ms/I

value would be an upward or tension load on
the column.
M£ =OJ see Fig. AIO.5
Equation (3) requires the values of the
whence, l'lA = 0 - (-6.06) = 6.06 in.lb.
~x and My the moment of the ~s/I diagram as a
lead about the x and y axes. Equation (3)
also requires the total column load P which Frame Point B. x =- l2"J Y = 9.325"
equals- the ar-ea of the Ms/I diagram.
l'l = 270 + 2530 x 9.375 (-540) (-12)
i 32 3188 + 3456
For this problem the value of P from
Fig. AIO.S equals J = 8.44+7.44+1.88 =17.77
P =22.5 x 24/2 =270 l'lB = l'!s-l'li = 0-(17.77) =-17.77 in.lb.

The centroid of this trianG~lar loading Frame ?olnt C. x = 12 J y = 9.325

is 2 inches to left of yaxis. Fig. AlO.6
shows the column section with this resultant 270 2530 x 9.325 (-540)12
load P. l'li = -32- + 3188 3456
= 8.44+7.44-1.88 = 14.0
~enow USe equation (3) to find the in-
determinate moments Ml which are equal to the
base pressures on the column. Bquation (3) Me =l'ls-l'li = 0-(14.0) =-14.0 in.lb.
involves bending moments Mx and My and
disGances x and YJ allot which must have Frame Point D. x = 12 J Y =-20.625
Signs. The Signs will be determined as follows:
270 2530(-20.625) + (-540)12
Mi = 32 + 3188 3456
When moment ~x produces compression on
base on that ~ortion above x axis J then Mx is = 8.44-16.38-1.88 =-9.82
pos Lt.tve ,
l'lD = l'ls -l'li = 0 - (-9.82) = 9.82 in.lb.
~nen ~oment My produces c~mpress1on on
~ase on that portion to right ot y axis J then Fig. A10.7 shows the 28.17
My Is pos i t t ve . final bending ~ament d1a-
gram, which of course
A distance y ~easured upward from x ~1s checks the solution by the
is posit~ve; ~easured downward 1s negative. elastic center ~ethod in
Art. A9.4. The student
A distance x measured to right fram y should note that the
axis is positive J to left negative. numerical Nork in the
column analogy ~ethod is
From Fig. AIO.6: - practically the same as
in the elasGic center
p = 270 method. Fig. AIO.7
AID. 4 THE COLUMN ANALOGY METHOD

Solution No.2 MA = ~'s -11 1 :: - 30 - (-36.06) = 6.05 in . It: .

In this solution a different static ~raoe which checks solution 1.

condition will be assumec as shown in Fig.
AIO.8, namely the frame is cut under the load Frame Point 8. x =- 12, s » 9.375
and one-half the 10 lb. load will be assumed
as gOing to each cantilever part. Fig. AlO.9 M - -1650 + 2531x9.325 + (-9180)(-'2)
! - ~ 3138 3456
shows the static ~oment diagram and Fig.
AIO.lO the static no/I diagram with centroid =-12.23
locations or each ~ort1on or diagram which
Me =M:s-Mi =-30-(-12.33) =-17.77in.lb.

-30 -90 Frame Point c. x :: 12, Y = 9.375

_-1650 + 2531x9.375 + (-9180)12
M1 -~ u.:: 3188 3'--
400

=-76.0
-30 .90 Me = Ms-Ml =-90- (-76.0) =-14.0 In s Ib ,
Fig. AlD.9
4' Fraille Point D. x = 12, Y = -20.525
•
::~ + 2531(-20.525) (-9160 )12
Ml 32 3186 3456
(4) ;' -99.82
(1)

-10 I -30
MD =Ms -M1 =- 90 - (-99.82) = 9.82 1::1. lb.

h2''1 Thus solution 2 checks sol~tlon 1. The

student should solve this problem uSing other
Fig. AIO. 10 Fig. AIO. 11
static conditions.
are numbered (1), (2), (3) and (4). The area AIO.4 Unsymmetrical Frames or Rings.
of each ot these portions will represent a
load P1 , P a , etc. on the column in F1g. AlO.ll. In applying 0he column analogy ~ethod to
Since the static moment 1s negative on each unsymmetrical frames and r1ngs, the moment of
portion the load on the column section will be the H/EI dia~ram ~ust be taken ~bout principal
upward. axes and the ~aments of inertia ~ith resnect to
principal axes. .
Pl. =-10(30) =-300, P3 =-45x9 =-405
However, as explained for the elastic cen-
p. =-15(6)/2 =-45, p. =-30x30 =-900 ter method in Chapter A9, the moments and
section properties with regard to centroidal
p = Zp = - 300 - 45 - 405 - 900 = - 1650 axes can be used if Mx, My, Ix and I y are modi-
fied to take care of the dis-symmetry of the
From Fig. AIO.ll structure. In Art. A9.2 it was shown that the
redundant forces at the elastic center to ~~­
s~etrical frame sections was, (see equations
Mx =- (45+405) 9.375+ (300+900) 5.635
11, 14, 15 of Art. A9.2).
= 2531
My =300 x 12 + 45 x 10 - 405 x 6 - 900 - 12 M
o =~-p(
~ds/EI - T same as ,. . or symmetrical
t:-ame)
= -9180

111 =!. + MxY + Mt{

A IX y

POINT A. x > - 12, y = - 20.625

Ml :: -1650 + 2531(-20.525) (-9180) (-12)

32 3188 + 3456 - - --- - - (~)

= -36.06

s
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AlD. 5

As prev~ously done in Art. AlO.2 let, SOLUTION:

Fig. AlO.13 shows the cross-section of the
analogous column. The length of each ::nember of
Fu:thermo!e, let, the column section being the same as in Fig.
AlO.l2, and the width of each portion being 1/1.
The first step in the solution is to find the
(4:!-1
y
and !'Ix = M,; (2f!--)
x column section properties.
I • The total area (A) of column section equals
and k = (1 -~I
Ixl y (15/1) + (10/1) + (12/2) =31.
Then substituting in equations (a) and (b) The other properties required are: - Ix,
we obtain, I y and I xy•

The calculations to determine the location

From equation (2) we have,
Substituting values of Xo and Yo into this
equation, we obtain as the equation for 11i the
indeterminate moment in the column analogy
::nethod, the following -
Ml =~+ (l1x - 11 y)Y + (Mv -
A :'::Ix
The true moment is the same as for the
symmetrical section, namely,
Thus the solution of an unsymmetrical
frame by the column analogy method tollows the
same procedure as tor a ~etrical section
except that equation (6) 1s used instead of
equation (3).
AlO.5 Example Problem - Unsymmetrical Section.
at the centroidal x and y axes and the above
properties about these axes would be identical
to that given in example problem 1 of Art~ A9.5
at Chapter A9, where this same problem was
solved by elastic center ~ethod. The results
were Ix =
606.51, I y = 942.96 and I xy 217.74. =
The location ot the axes were as shown in Fig.
AIO.13.
The next step in the solution is to choose
a static frame condition and determine the
i1X )x - - - (6 I
static (~) diagram. Fig. AlO~l4 shows the
i<I y assumed static condition, namely, two cantilevers
because ot the frame cut as shawn~ The figure
also shows the static bending moment diagram
made up of three portions labeled (I), (2) and
(3). The area at each portion of the moment
diagram divided by the (I) for that frame por-
tion will give the loads? on the c o Iumn,
(I)
240
-1440"
\.Bh, \~cfn:",,","---1\
cut

Fig. AlO.l2 shows a loaded unsymmetrical

frame fixed at pOints A and D. Required, the
true bending moments at points A, B, C and D.
This problem is identical to example problem
1 of Art. A9.5 where a solution was given by
the elastic center ~ethod. D

2401b

-1~40
L 6" -.;... 6" .., Fig. AlO.14
at , C
I
JI
, I
120'--r:
1=2
I' I
L'T I
10"
1-1
·1080
_5"'00"'3"'2 , Y

9" I
1= 1 1

!
P 1. =- 1440 x 3/2 = - 2160
1 I -'- P II =- 1440 x 15/1 = -21600
-l IA
'::T:7
p, =- 1080 x 4.5/1 =- 4860
Fig. AlD. 12
Fig. AlD. 13 zp =- 28620
AIO.6 THE COLUMN ANALOGY METHOD

These loads act on the centerl~ne of the Fr~e Po1rot C. x = 6.968, Y = 5.242
frame members and t:rough the centro1~ of the
geometrical moment d1agram Shapes. These M1 =-923+58.37x5.242+132.36x6.S68
centroid locations are indicated ~y the heavy
dots 1n Fig. AI0.l4 and their locations are = 357.7
given With respect to the centroidal x and y
axes. The loads Pl, Pa and 23 are now placed ~ =Ms - Mi = 0 - (357.7) =- 357.7 tn .lb.
en the column in AIO.13, acting upward because
they are negative. FY"ame Point D. x = 6.968, Y =-4.758

we now find the ~crments Mx and My wh1ch Mi =-923+68.37 (-4.758) +132.36x6.968

equal the moments of the loads P about the = -326
cer-troldal axes.

i1:x =- 2160 x 5.242 + 21600 x 2.258 + 4860 ?'In = lis -I'1i = 0 - (-326) = 326 1::.10.
x 5.758 = 70290 The above results check the solution of
this same problem by the elas~1c centeY" method
= 2160 x 3.032'" 21600 x 5.032 + 4860 in Art. AS.S of Chapter A9. The st~Gent should
x 5.032 = 139700 solve this problem by chOOSing other stat~c
frame conditions.
To solve equation (6) we must have the
terms ~' MY and k ,

(*)
AlO.6 Problems

~ = l1x = 70290 x 217.74/606.51 (1) Deter.n1ne the bending moment dlag~am ~or
the loaded structures of Figs. AIO.IS to
= 25234 AIO.20.
400 400
11'Y =11Y (~)
I -
y = 139700 x 217.74/942.96 ~ lSi 15 +151 w=lO Ib/in.

= 32258
L_45"
I 3 = f H~'; I
IL=30"
ix(
L=30"
11 L = 24 L = 24
= (1 = (1
-
_ 217.74' ) I =2 =2
k -
x y
) 606.51 x 942.96 II =6 I I =6

= .9171 Fig. AlO.l5 ""'" Fig. Ala. 16

SUbstituting values in equation (6) we obtain,
200 300 lb.
-28620 + (70290 - 32258 )y + ~12" j
31 0.9171 x 606.51 L 224 , '5"~
I ; 6 i +-:0 \Cl 1<> lCrJ
(139700 - 25234 ) x
0.9171 x 942.96
100 L=20
1;4
L=20i
1=4:
Ir O I

29-
I
:L== 20
I

1L== 20
10 L; 24 i 10 I II 5 I 5
whence, 1=4 i-.i. ~ .....
1-8 8+i""; i---- 24" -------!
!'1 i =- 923 + 68.37y + 132.36x -----(7) 100 100
Fig. Ala. 17 Fig. Ala. 18
For Frame Point A. x =-5.032, Y =-9.758

Mi = - 923 + 68.37(-9.758) ... 132.36 (-5.032) 100 100 400

f.., 1=~
= - 923 - 667.15 - 666.0 = - 2256 1/,-;--"

I1A = 11. -111 =-1440-1080- (-2256) =

,

4~",~~~
I
. 1-~-2 /,10" \ I
1 1
-264 i'l.lb. 2
.
,~4'~12'~''''

For Frame Point 8. x =- S.032, y = 5.242 -..~~I=

I
1.5 200 200

Mi =- 923 + 68.37 x 5.242 + 132.36 (-5.C32) 200 lb. Fig. AID. 20

= - 923 + 358.39 - 666.0 = -1230.6 Fig. Ala. 19

ME = 11 8 -~ =-1440- (-1230.6) •
(2) Solve ?rob:ems (2) and (3) at ~he end of
-209.4 l.n.l'tl. C~~pter A9, Art. A9.9.
 CHAPTER All
CONTINUOUS STRUCTURES· MOMENT
DISTRIBUTION METHOD
All.l Introduction. The moment distribution
method was originated by Professor Hardy Cross.*
The method 1s Simple. rapid and particularly
adapted to the solution of continuous structures
ot a high degree of redundancy, where i t avoids
the usual tedious algebraic manipulations of
numerous equations. Furthermore, it possesses
the merit of giving one a better conception at
the true physical action ot the structure in
carrying its loads, a tact which Is usually
quite obscure in some methods of solution.
The method ot procedure in the Cross method
is in general the reverse of that useo -tn the
usual methods where the continuous structure Is
first made statically determinate by removing
the continuous feature and the value of the re-
dundant then solved for which will provide the
original continuity. In the Cross method each
member of the structure is assumed in a definite
restrained state. continuity of the structure
is thus maintained but the statics pt the struc-
ture are unbalanced. The structure is then
gradually released from its arbitrary assumed
restrained state according to definite laws of
continuity and statics ~til every part of the
struc~~e rests in its true state of equilibrium.
The general principles of the Cross method
can best be explained by reference to a speCific
structure.
Fig. All.l shows a continuous 2 span beam.
Let it be reqUited to dete~ne the bending mom-
ent diagram. T~e first arbitrarily assume that
each span is completely restrained against rota-
tion at its ends. In the example selected ends
A and C are already fixed so no restraint must
be added to these points. Joint B is not fixed
so this jOint is 1magained as locked so it can-
not rotate. The bending moments which exist at
the ends of eacri member under the assumed condi-
tion are then determined. Fig. All.2 shows the
moment curves for this condition. (For calcula-
tion and formulas for fixed end moments see fol-
lowing articles). Fig. All.3 shows the general
shape of the elastic curve under this assumed
condition. It is noticed that continuity of the
structure at 8 is maintained, hOwever from the
~oment curves or rig. A1I.2 it 1s found that the
internal bending moments in the beams over sup-
port 8 are not statically balanced, or speci:i-
cally there is an unbalance of 270. The next
step is to statically balance this joint, so
it is unlocked tram its imaginary locked state
and ObViously joint B will rotate (See Fig.
All.4) until equilibrium is established, that Is,
until resisting moments equal to 270 have been
( Paper - A.S.C.E. Proceedings, May 1930)
All.l
set up in the two beams at B. The question is
how ~uch of this moment is developed by each
beam. The physical condition which establishes
the ratio of this distribution to the two beacs
at B 15 the fact that the B end of both beams
must rotate through the same angle and thus the
unbalanced moment at 270 will be distributed be-
tween the two beams in proportion to their abili-
ty of resisting the rotation at their Bend thru
a common angle. This physical characteristic ot
a beam is referred to as its stiffness. Thus let
it be considered that the stiffness factors of
the beam BA and Be are such that 162 is distribut-
ed to BC and - 108 to BA as shown in Fig. AI1.4.
(The question of stiffness factors is discussed
in a following article).
Referring to Fig. AII.4 again it 15 evident
that when the elastic curve rotates over joint B
that it tends to rotate the far ends at the
beams at A and C, but since these jOints are
fixed, this rotation at A and C is prevented or
moments at A and C are produced. These moments
produced at A and C due to rotation at B are re-
ferred to as carry-over moments. As shown by
the obvious curvature at the elastic curves
(Fig. All.4), the carry-over moment is of oppo-
site sign to the distributed moment at the ro-
tating end. The ratio of the carry-over moment
to the distributed moment, referred to as the
carrJ-over factor, depends on the physical prop-
erties of the beam and the degree of restraint
of its far end. 1Carry-OVer factors are discus-
sed in a following article. For a beam at con-
stant section and fixed at the far end, the car-
ry-over factor is -1/2). In figure AI1.4 a fac-
tor at -1/2 has been assumed which gives carry-
over moments of 54 and -81 to A and C respective*
ly. To bbtain the final end moments we add the
original fixed end moments, (.the distributed bal-
ancing moments and the ca~J-over~aments as
shown in Fig. AII.4. With the indeterminate
moments thus determined, the question of shear,
reactions and span moments follow as a matter
of statics.
All.~ Def1n1t10na and Der1Tat10ns 01 Term8
1. Flxed~end ~oments:
By ~:ixed end moment 1s meant the moment
ft
which would ex1st at the ends of a member if
these ends were fixed against rota~ion.
2. stiffness Factor:
The stiffness :actor of a ~ember is a value
pnopor-t tonaj, to the tnagnf tude of a couple that
must be applied at one end of a member to cause
unit rotation of that end, both ends of the
member being assumed to have no movement at
'."".':_'~':.""'~~-:;;-
'{~1~:
.111.2 THE MOMENT DISTRIBUTION METHOD

0.5 when rar end is fixed. ~he letter ~

will be ~sed to designate carry-over fcc~or.

r-
200
15''-01
VI: 10.,/in. L 4. Distribution Factor:
Fixed~ . ! I" . 1= Fixed It a ~oment is applied at a joi~t Nr.ere :NO
, .I. i" L·1.3~ ----1 or more members are rigidly connected the distrl-
A B Fig. All. 1 C
but10n factor for each member 1s the prcport~onal
part of the applied moment that ~s resisted by
that member. The distribution factor for any
I ~ember which will be given the symbol D equals
KlZK, where K equals stif~ness ~actor of a par-
ticular member and ZK equals the sum of K values
for all the Joint members. The sum of the D
,-480 values for any joint must equal unity.
5. Sign Convention:
Due to the fact that in ~ny problems where
members come into a joint n-om all dn-ect icns as
is commonly found in alr~lane structure, the
Elastic Curve for Fixed End Spans customary sign convention for moments ~y produce
contusion in applying the ~oment distriJutlon
method. The follaNing sign convention 1s used in
Fig. All.·3 this book: a clockwise moment acti~g on the end
of a member is positive, a counterclockwise one
is negative. It tollows that a moment tending to
Elastic Curve due to Rotation of Joint B rotate a jo Int clockwise is negative. It should
be understood that when indete~lnate continuity
Balancing Joint B moments are determined by the ~oment distribution
i
-1081162 ! method using the above adopted sign convention,
,./ --=:::::::::... ! Carry Over Moments that they should be transferred i~to the conven-
54 -811 from B to C & A tial signs before proceeding with the design of
1 ~ ::..::::::..:c.:::...:....:c.::.._ _

~ .5881.588 -8311 Final Moments = Values

the member proper.
I in Figs. All. 4 plus All. 2 The follOWing sketches illustrate the a~pt­
I I ed sign convention.
Fig. All. 4 Illustrations at Sign Conventions for
End Xoments.
I Example 1 Fixed end beam with
lateral loads.

I
I
i

831
Conventional + + tension in bottom fib-

.26~~~====~~~~~~~~~~~~~~~~/~1,
moment Signs ers 1s POSitive bend-
Final Moment ing moment.
Ar- C Curve

Adopted sign moments which tend to

convention rotate end ot member
clockwise are pOSitive.
translation. The value ot the stittness :actor - ~'--i-
-~
-
will depend in part on the restraint or degree +
ot tixity of the opposite end ot the member
from which the couple is applied. The letter K
1s used as a symbol tor stiffness factor. ExaI:1ll1e 2 Translation of supports
3. carry-over Factor: of fixed ended beao.
It a beam is simply supported at one end
and restrained to same degree at the other, and
a moment is applied at th5 simply supported end,
a ~oment 1s developed at the restrained end.
The carry-over tactor is the ratio of the mom- Conventional +
ent at the restrained end to that at the I:loment Signs
simply supported end. For a prismatic beam
without axial load the carry-over factor 15 -=
Adopted s ign::E:::::::::::--.
convention ~
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES All.3

E:xam:ple 3 Pab

M t~ L~
• 1.L... _ S'3M a
Fig. All. 9
External applied ~oment
at joint in structure Fig. AI1.10

adopted sign Fig. All.ll

Joint Moments which tend to
rotate joint counter-
clockwise are positive.
+ moment Fig. AI1.12

AIl..3 Calculation of Fixed End 1I0111euts

Since the fixed end moments are statically Since the change in slope of the elastic
indete~inateladditional facts must be obtained curve bet'.. . een ends (1) and (2) is zero, theorem
from the laws or continuity i~ order to solve (2) as applying to fixed end beams can be re-
for them. In this book the theorem of area stated as follows.
moments will be used to illustrate the calcula- The sum of the areas of the moment diagram
tion of the fixed end moments as well as the must equal zero. And fram theorem (1) the stat-
other te~ which are used in the moment dis- ic moment of the areas of the moment diagram
tribution method. -(Ret. Chapter A7) about any point must equal zero or in equation
The following well known principles or form:
theorem of area moments will be used:- ZM = 0 - - - - - - - - - - ----(1)
(1) The deflection at any point WA" on the ==0 --------- ----(2)
elastic curve of a beam away from a tangent to For a beam with variable moment of inertia the
the elastic curve at another point WB" is equal conditions for fixity are:-
to the moment a r the area of the M: diagram be-
jl1dx/El = 0
El j!".xdx/EI = 0
tween the points A and 8 about point WAft.
(2) The change in slape between two points Figs. All.ll and AIl.12 show the static and con-
~A~ a~d ~Bft on the elastic curve at a beam 1s tinuity moment areas, the total area of each por-
equal to the area of the M diagram between the tion and its e.g. location.
El Substituting in Equation (1)
two paints ~A" and "9". M = Pab M1L MaL 0
7+7+2= ------(3)
The "area moment" theorems will be illus-
trated by the applications to the solution at a
simple problem. Fig. AlI.5 shows a simply sup- and rr on equation (2)
ported beam of constant ~oment of inertia and Mx about lett end = Pab L + a ~"hL L
~odulus of elasticity carrJlng a single concen- 2'-3-+2·3+
trated load. Figs. All.6 and All.? show the
static moment curve acd the shape of the beam - -(4)
elastic curve. Now assume that the ends are fix
ad as shown in Fig. All.S and let the value at The values of M1 and t1a tor any value at a or b
the fixed end moments be required. Fig. All.9 can now be found by solving equations (3) and (4)
shows the shape of the final moment curves wade Table AlI.l gives a summary at bema fixed
~p of the static moment curve and the '~~~O~trl
end moments for most at the loadings encountered
tr-api zo i ca.; moment curve for:ned by the ur...iknown in routine deSign and analYsis.
end moments. Fig. AIl.IO shows the shape of the
elastic curve, the slope at the two supports be- w TABLE All. I
ing made zero by fixity at these points.
P
'--
I
a -t--- b -----:1
1 EI is constant ,
Fig. All. 5 r;~( ,
1 L _ _--I ;.-' 12L" 4aL-3a)

@~ Fig. All.6 w

WL~_1< + !:Jr::WL
-8- 8
~ - - - - - - -..- FIg. Al1. 7 . 2 ' 2 .

Fa iP
I
L
b=1 Fig. AI1.S
,
;0 (5u ... 2v)
::::;.-:;::::rITn
v ::::' , 1 I ; I L L'
L ----{
6O(5u ... 3v)
 All.4 THE MOMENT DISTRIBUTION METHOD

Table All.l - continued tangent to the elastic curve at poi~t A relative

to a line AS is equal to the shear at A on a
_W si:nply suppor-ted beam AB cue to t ae vr curve be-
w2~~ 4Jl:Ii:Ii::r::::----=~f w3~ <a
t~een A and a acting as a load.
::;1

':'hus
(ML x 1(3) _ ML
6EI (Positive Shear
2EI
~5~" {lOL <I_ 10aL+ 3a 'i ~~
60L {5L~ 3a)
<I
Neg. Slope;
~ x 2(3 ~
"-
~Jj
= 2'E~I ,~
3EI

wa• a a.
30L<I(lOL -15aL+6a)
p/ wa'
20L a ( 5L - 4a ) Let 98 unity
Then 1 ML or Mg = :lEI stiff~ess Factor of'
3EI r::-
Seam SA of Fig. AII.13. A ~c~ent applied at 3
produces no ~oment at A since end A is freely
supported. Thus the carry-over factor for a
beam freely supported at its ~ar end is obvious-
ly zero. Consider the beam of Fig. All.14. uue
to complete fixity at end A, the slope of the
-tj~X(L-XI' ~
~o
t{WX'(L-X)dX
dx
elastic curve at A is zero.
1/3 + ~ 2(3 =0
o
(Ref.N.A.C.A. °A !1ll!: 2EI
2EI
General case - Any loading T. N. #534)
or
MA = -~
2
All.4 Stiffness 7actor; Carryover ?actar: -
Derivation af: Thus the carryover factor for a beam fixed at
its tar end Is 1/2. Using the conventional ~om­
(For definitions of these terms see page ent signs, the carry over moment is of the oppo-
A11.2) site sign as shown by the above equation. How-
Consider the beam of Figure All.13. By ever, for our adopted sign convention inspection
~Ohr's theorem (see Art. A7.l2), the slope ot a of the shape of the elastic curve as shown in
Fig. All.14 tells us that the sign of the carrJ-
EI is constant over mament is of the same sign as the rotating
moment at the ~ear end. That ls, the moment act-
i~g on each end of the ~ember 1s ir. the same di-
L
rection, and therefore of the same sign.

~BA Elastic Curve

""T
:;1"
C'"""=---+---J....l
1.L ML
3 2£1

Fig. All. 13

Fig. All. 14

IlJl = !1ll!: 2/3+~ .1/3=~+~, but :1A=- i1E

2EI 2EI 3EI 6EI 2
ThengB=~-~=~
3EI l2EI 4EI
Let 9B = urn ty, t.lten~:a = 4EI sui r rness ract or-
L
of beam SA ot Fig. All.14.
A comparison of the st1f:~ess factor of this
beam to that of Fig. All.13 shows that the stif:-
 All S

ness factor of a beam freely supported at its MA ; -

2KF ;
- 2F which is the general ex-
3+F
far end is 3/4 as great as one r ixec at its far MB K(3 + F)
end. j'ur-tnermor-e in one case the carry-over pression for carry-over rae tor for a degree of
factor is zero and in the other case it 15 1/2. fixation F.
It is t.ner-er or-e obvious that the values or these
1;NO ter:ns depends in part upon t::-te restraint All.5a Exa.ple ProblellLS
or degree of r ixat i cn of the far and of the To obtain a definite conception at the true
beam. mechanics of the "cross" ;nethoo, the reader is
advised to follow thru the detailed solution or
All.S General Express1on~ tor Stiffness and Carry- the follOwing simple problems. In these pr-ob-
over Factor in Terms ot Fixation Factor (p) lems, the moment of inertia in any span has been
at Far End of a Beam
taken as constant and all joints have been as-
In the beam at Fig. AIL14 (F) fixation surned to undergo no translation. Problems in-
factor at A was unt ty since beam had been taken volving variable I and joint translation will be
as conpletely fixed at A. It was found that:- considered later.
ME ;
4E19B and MA ;
- 2 EI 9B ,-W. H/in.
Example
E2
.~./!Il.

Te.ke 9B
L
;
L
unity and let EI = K for simplicity.
L
Problem #1 ~A
~1.5 ~
11 11
gil
]} tt tt tl
! all
W!!'J 51.51

':'hen MB; 4K jStUfIl." hclDl' ~ s K ~, iu uc

MA; -2K
IOllltriblu1OCl Facl<ll' KI.l:K ~, .51.5 nc
L~kewise the results for the beam ot F~g. A1L13
give Carry Oo'er Facwr ., .51·5 .00

Me = 3K 1'lIed: End Mo",ent. '-88 7118=.OU3..Lot.:.7118 7ell=.0833Wr..z,,_711a1al13

I1A; 0 lal BUanclng
1157<E
o
0 ~-llS!O

Figs. All.lS and AII.IS show these results. Carry 0.-.,.. 1/1 57.5 _51.5

Fig. All. 17 shows the general case, the r ixa- 2IId Bll!aJlCUlIj: c c «c
tion factor at A being F. The difference be-
." '"
tween F'fga , AII.IS and AII.I? is that the slope
at end A has changed but eB the slope at end B
FI.llaJ. Moll:lefttll

ConvellUO.... 1 MOlQIIDt SIilU' ..

.S8~ 8113 -710.51710. 5

• I•
·o.rerlla:lg MOIll.nt : .5';51.5)[15.75 • . 15x$1.5':17.11: alI3
• •
remains the same.
tlA=O Example Problem 1 shows a two-span cont1nu-
F;i~B ous be~ with over-hanging ends. '..le first as-
A· "". sume all the joints B. C and D to be locked
MA=~2K MB=4K
against rotation or the beams Be and CD are
Fig. All. 15 arbitrarily assumed in a fixed-end condt t t on ,
The !irst line in the solution gives the
stiffness factor of each beam. Fr-om Art. AIl.S
the general expression for stiffness factor is
F=O
~
A
MAsO
=.5 8B'"1,
MB-3K
- EI(3 + F)/L. where F ; fixation factor of far end,
wni en equals , since all joints have been con-
"
steer-ed rtxec. ThUS, stiffness factor K ; EI
Fig. All. is (3+ l)/L; 4. EI/L. Since EI/L is the same for
each span Be and CD, the Stiffness factor has
The change in :noment at end A when changing been wri tten as one since it 15 only the rela-
beam AIl.lS to that of All. 17 = 2K- 2KF ; tiva values that are necessary. The cantilever
2K(1-F). Since 9B is xept the same value, one- SA has zero s t t r rneas ,
narr of the moment change at end A appears at The second row gives the member distri but t on
end 8 but of cppoat te sign, or factor D at each jOint or 0 ; KlZK. ?or eXalIlple
Mg=4K-l/2 2K(l-F) ;3K-KF=K (3 + F) ;
ZI (3 +F) 0) = o and to BC
distribution factor
.
at jc ~::.t 3. the distribution factor to 3A = 0/(1
lI(c • 0) ; ;- . At joint r !:-o. e
to CB ; 1/(1 + 1) ; .5 and
..
L
eA' F(. 5) Lfkewt se to v-c• •
The thll'd row giyes the carry-over factors.
F"F~ .
From Art. All.5, the carry-over factor C ; 2;;'/(3 +
MA'= -2KF MB-K(3+ F) F'I = 2 x 1/(3 + 1) .5 assummg the far end rrxe d
or F = 1 for each member.
Fig. All. 17 The next step as shown by the 4th horizont-
Thus the ge!leral expression tor the stiffness '11 row is the calculation of the r ixed end non-
factor or a beam of constant section equals EI ants. The signs or tl:e :n.oments are according to
C- our adopted sign convention, t~t is an inter-
(3 + F') • ':'he carry-over factor from 9 to A ; nal r-es i st.mg moment wni cn tends to rotate the
All 6 THE MOMENT DISTRIBUTION METHOD

end at a member clock'.... i ee 15 positive.
We now begin the solution proper by first
unlocking joint 8 from its as suzned fixed state.
We rind a moment of -883 on one side and 765 on
the other side of JOint or a sta t1 c unba.Iance
of -115. Joint 9 therefore will rotate until a
resisting moment of 115 is set up iY1 the member-s
EA and Be.
The resistance of these members to a rota-
t10n ot Joint B 15 proportional to their stiff-
ness. The distribution factor based on the
stiffness factors 15 o for BA and 1 for SC.
Thus 1 x 115 = 115 15 distributed to SC at Band
Ox 115 = 0 to EA. Joint 8 15 now imagined as
again locked against rotation and we proceed to
Joint C, which 15 now released from its assumed
locked state. Since the jOint 15 already stati-
cally balanced, no rotation takes place and the
distributing balancing moment to each span 15
zero. Next proceed to joint D, and release it.
The unbalanced mament IS, 115 so the JOint 15
4. Calculate the carry-over factor C for each
end of a:t.l :nembers C = 2F/(3+;-). Thus ror
beams f~xed at far end F = 1 and thus C = 112.
For pinned at far end F = 0, hence C = O.
5. Calculate the fixed end :noments (t) for giv-
en transverse member loadi:Igs or support oer rec-
tions, using equations in summary ~able All.l.
~~d ~oments which 'tend to rotate end of :Jearn.
cIockwf se are ~ositive moments. (See Art. Al1.2)
6. Considering one joint at a time, unlock it
from its assumed fixed state, all other joints
remaining locked. It an unbal~~ced moment ex-
ists balance It statically by distr~butlr:.g a
counter acting moment ot oppOSite sign among 'che
connecting members according to their D or dis-
tribution factors.
7. These distributed balancing moments produce,
carryover moments at far-end of member-s equal to
the distributed moment t taes the carry-over rae-
tor C and of the same Sign as the dlst:oibuted
moment. Record these carry-over moments far ,~
_0

balanced by distributing -115 between DE and DC
as explained above for joint 8.
As painted out in Art. All.5, when we ro-
tate one end of a beam It tends to rotate the
far fixed end of the beam by exerting a moment
equal to some proportion of the moment causing
rotation al: the near end. For beams of constant
section and rtxed at their far endS, the carry
over factor 15 112 as explained before. Thus
the distributing balancing moments 1n line 4
produce the carry-over moments as shown In line
5 ot the table. This completes one cycle at the
moment distribut10n method, which is repeated
unti~ there Is nothing to balance or carry-over,
ends for all distributed moments.
8. Repeat the pr-oceedur'e of unlocking each
joint, distributing, and carrying over moments
until the desired pr-ecf s t on Is obtained, stopping
the solution e rt.er- a distr~bution. The final
moment at the end of any member equals the alge-
braic sum at the original r fxed end Jl.Oments and
all distributed and carryover moments.
~ample Problem #2
*
.SJ/lIl.. f lJ/lll..

}i.S+---~____t__72
" I' -L
EI l.o Can.w>t

pfli j t I , I ! r j it! r1111 t t! ; t 1, Ij • , ''-''''

I'
, il:~

F! q
ge------J..s l . ,~

.
. 12 :
or in other worca until all artiticial restraints I ,

have been removed and the structure rests in Its ~

.!!.~ K O.OIIl4',*' . 011l41. 013•• ...Lz. 013i1 .013il'~',OI39!.011l4 .~ • Q1Q~1 II
r:ac'lar t,
true state of equt Li br-tun, I "
I

To continue with the second cycle, go back D~~:'IOI>~ ,I. .~:l8!. S72 .s.s .sn!.Hs 110 I
to joint B and release it again tram its assumed
locked state. There is no unbalance since the
Carrr·Oo.r C
factor e .s sl.s sl.s
I
... 'I'
,

carry over moment tram point C was zero, thus fl.om E%ld _a, l&S • ..l...L1 •• ,lllliu:lOl."Lz-'_dzi43:z--.l. .. L1'_43 z
lZ ,11 _ 12
l,e8 "-!.."Ll
11
• _ 7M1883 i

-~"
I

there is nothing to distribute or carry-over.

Proceed to point C, releasing the joint, 'He flnd
IOI_C.....

Carrr-Qru
n Il5
72
X 1«lm
57. slo
::::><: 01
0
961_9&
X~
~><~
c _S7.S _121:
1t balanced Wlder the carry-over moments at 57.5
and -57.5. Thus the distributing balancing mom- :lod8ll\alw:lIl.t: u _72 X2~_5!'32_9 ~X2.9!24.6~ 7210
car,.,-Oo... ·(2.3 _3slo X;6.5116.5 0136 12.31
ents are zero. Joint D Is l1kewise in balance
since the carry-over moment tram C 15 zero. All 3"" 8IIl:uI<:..... 0,IZ.3~lS.ti20.6~~X°:.1.1S.?::<11.JIO
joints can now be released Without any rotation car,.,.OvJr 7.1 6.110 10.l!.10.3 *6,1 _:71

since all joints are in equ1ll brium. To obtain ~tlI 8IIlaAc ..... 01_1.7~_1.6 ::':'X~X.SZ.6 ~r,:io
the tinal moments we add the original raxec end e...., ..,·oou I-l.] .~ o _1.611.8 013.9 I.J!
moments plus all distributed balancing and car- Sib SOWle .... olt.3 I. 7 1. ~
," _1.1 _I. 1 _l.J10

ry over momen ts . ll.oal..,..._ .. .~318U -erc 41610,4 _3+413+4 _610. tlalO. 4 _8331883

Ie_Ilona! SilPUl • 1. .1. .1. .1• '1' i

All~S Ge 1:1.8 ra 1 SWIIII&ry of Proceedure
1. All computations should be written on, or
adjacent to the d1agram. of the structure.
2. Determine the stittness factor K tor each
end at each member. K = (3 + F) EI/L, where F
15 the degree of fiXity at far end ot member.
It all members are assumed rrxec at far end than
K 15 proportional to IlL assuming E as constant.
3. Determine the distribution factor D tor each
member at each joint of structure. D = K/l](.
Example Problem 2 15 similar to prOblem 1 but two
spans have been added. ',oj'e first assume all joints
locked against rotation. The stiffness factor of
each span 15 :?roportional to EIIL or lIL since EI
constant. The car:-y-over factor 15 1/2 as in
is
previous example. Fixed end moments are calculat-
ed as shown. Unlock jOint E, the unbalanced mo:n-
ent 15 -U5. Balance the JOint by distributing
1 x 115 to BC and zero to EA. Proceed to joint C,
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES All 7

and unlock, all ot~er joints remaining fixed
against rotation. The unbalanced moment 1s
(-768 + 432) = - 336. Balance by distributing
.428 x 336 = 144 to CB and .572 x 336 = 192 to
CD. Proceed to joint D, and release. The un-
balanced moment is zero which means that joint
D is in equilibrium. thus no distribution is
necessary. Proceed to joint E and F and balance
in a similar ~er. The distributed moments
will be the same as the values for joints Band
C due to symmetry of structure and loading,
however, the signs will be oppOSite under our
adopted sign conventions. The next step is the
carry-over moments which are equal to 1/2 the
distributed balancing moments. This operation
is shown clearly in the table. Values of all
moments are given only to first decimal place.
The first cycle has now been completed. Cycle
~NO is started by again releasing JOint B. We
find the joint has been unbalanced by the carry-
over moment of 72. Balance the joint by distri-
buting - 72 x 1 = - 72 to·BC and zero to EA.
Proceed to joint C. The unbalanced moment is
57.5. Balance by distributing - 57.5 x .428 =
- 24.6 to CB and the remainder - 32.9 to CD.
Proceed to jOint D. There is no unbalance at
this joint since the carry-over moments are in
balance, thus no distribution is necessary.
Proceed to joints E and F in a similar manner.
The carry over moments equal to 1/2 at the
2nd set of balancing distributed moments are now
carried over as shown in the table. The second
cycle has now been completed. This operation
has been repeated five times in the solution
shown, or until the values at the balancing and
carryover moments are quite small or negligible
The final ~oments equal the algebraic summation
or the original fixed end moments plus all dis-
tributed and carry over ~oments. One require-
ment of the final end moments at any joint is
that the algebraic sum must equa L zero. The
other ~equirement consistant with the cammon
slope to all members at any joint is given by
equation (5) of Art. All.S. The results at
jOint C will be checked using this equation.
.5 t. l",jc = ~
<ll'Igd -
t.l1cb - .5 Hlbc Kcb
SUbt. values
(610.4-432) -.5 (344- (-432)J _178.6-44
[-610.4 -(-768)J - .5 (883 -768) 107.6-57.3
1.343
necessarJ. These modifications usually involve
rather long expreSSions :or expressing the stiff
ness and carryover factors of a me~ber in terms
at the fixation given by adjacent members. It 1s
felt that it is best to keep the method in its
simplest fOrn which means that very little is to
be remembered and then the method can be used in
frequently Without refreshing ones mind as to
many required formulas or equa't Iona ,
There are however several quite Simple ~odi
f1cations which are easily understood and re-
membered and which reduce the amount ot arithmet
ic required considerably.
For example in Problem 12, joints Band F
are 1n reality treely supported, thus it is need
less arithmetic to continue locking and unlocking
a jOint which is definitely tree to rotate.
Likewise due to symmetry of structure and load-
ing it is only necessary to solve one half of the
structure. Due to symmetry jOint D does not ro-
tate and thus can be considered fixed, which
eliminates the repeated locking and unlocking of
this Joint.
A second solution of Problem #2 is given in
gxaap te Problem #3. As before we assume each
jOint locked and calculate the fixed end moments
Now release or unlock jOint B and balance as ex
platned in previous example #2. Before proceed
ing to jOint C, carry over to C tram B the car-
ry over moment equal to 115 x 1/2 = 57.5. Joint
B is now lett free to rotate or in its natural
COndition. Proceed to joint C and unlock. The
unbalanced moment = (-768 + 432 + 57.5) =- 278.5 or
278.5 is necessary for equilibrium. This moment
is distributed between eNO beams, CD which is
fixed at its tar end D and CB ,Yhich is treely
supported at the far end B. The stiffness fac-
tor is equal to (3 + F) EI/L (See Art. All.5).
Hence for CD stiffness tactor = (3 + 1)
EI/L = 4 EI/L. For CB stiffness factor = (3 + 0)
EI/L = 3 EI/L or in other words the stiffness ot
a beam freely supported at its far end is 314
as great as when fixed at its far end. Thus the
stiffness factor at CB at C is .75 x .0104 =
.0078. The carry-over factor C to B is zero
since B is left free to rotate. (See Art. All.S)
Example Problem #3. Simplified Solution Of Prob-
lem /12
=
..., r,:-H.
.~. l'IUI. FlU<!,
s
e
. c
12~

Ratio of stiffness factors = ~ = ~ .0104 = 134

.• o . QHl4=ft .1$,.. 01Q.p. 001"8 . ou~ ~ lin 2
"
Kcb O\:N IC<:

Thus the distribution is according to the K Oi",,.,butIOll F1cwr

o , !t/!x: uu cu
ratios of the adjacent members. Cu..,.""u lK,or (C) c s
. le .IM
c .s ,
Simplifying Modifications - Examule Problem #3 flU<!.t>d mo",onUl .;183 163 .1e8 432 -4U
1lI~'iO",'f1 'J ll~_
_
The solution as given in Problem #2 repre- Ca,..,.ov.r l,2 '0 C ~T.S

B:l!anc. 10"" C
sents the "Cross" method in its flli~damental and Ca,..,.""" to 9. 0
most elementary detailed form. Many modifica- B.t.J..:u>c. ,<lUll 0
tions of the general method have been presented,
..
Car..,_onr to C
.a831883
in the most part for the purpose of eliminating Final mOm.nUl
Con.enuon:U &lll'l_
part of the arithmetic or the n~ber of cycles

I
c, ..
~
All.8 THE MOMENT DISTRIBUTION METHOD

The stiffness factor 0: ~he fixed support 1s in- SoL.;t ion #1

finity, that is a rigid support r~s ~n:inita re- zxaco;e
sistance to rotation. Pr cb Lem ""4
Joint C 1s now balanced by distributing
278.5 x .36 = 100.1 to CB and 278.5 x .64 = 178.4
to CD. Now carryover to joint 8, a x 100.1 = Stiffness Factor k 0.0104' 1/96' .0IO'lI.0139_l/72'.0139!.OI39
a and 0.5 x 176.4 = 89.2 to O. Proceed to jOint Oistrtllutioa F:Lctor O-rt
o and release it from its assumec locked state. Carry-Over Factor 0.> 5 .5
The unbalanced moment is -432 + 89.2 = - 342.8, n"ttd End Moments .a831768 _7681<132 "
_<1321
we balance by distributing 342.8 be~~een DC 1st a..~CUlg 01115 1<141192 2l6i21~
Carry_Over 112 s7.si108
which has a stif!ness of 0.139 and the support
E which has an infinite stiffness or zero goes
:!nll BIil~clng 01·72 _70.71_94.8 '"
-481·48
Carry_Over '-35.3 ·36. QI.2<I. 0 _<17, ~.
to DC and 342.8 goes to rigid support. ~he car- 3rd. Ba~ClJ'tg Oi35.3 25.713<1.3 23.7,23.7
Cury·Over 17.,111.8 11.2
ry over moment to C from D is zero since 0.5 x <lUI BalanCing
11209

o = O. The t~na1 moments thus equals the sum-

OI·U.9 -12.61_16.9 .8.61·8.6
C.. rry·aver ,.6.3 _6. <11_<1. J _3. ~i
~tions as shown which of course are equal to thE 5th Balancinlj: 016.3 ~. 616. 1 4. 214. 2
FlAal l.lomenUi _8831883
results shown in Problem #2. Cmlventional Signs .i.
_64<4.~I~H. 3 -187.31187.3

Example Problem #4-

.5 X 216 = 108 is carried over to C. Joint D is
Problem 4 is similar to Problem #2 and #3, left ~~locked or in its true state of ~estraint.
except that the support at 0 is assumed as hav- Joint C is now unlocked. The unbalanced ~o~ent
ing 50 percent fixity. Thus 50 percent of any
moment at this point produces rotation of the solution #2
SxampLe
member DC at D. Pr-ob Iem 11'4
In continuous wing beams, which fasten to-
gether by fittings on a support, it is commonly
required that the beam be considered as being
tully continuous and also tna't the degree of
SUffn ....s hotor K
continuity be taken as 50 percent. Solution 1
of PrOblem 4 1s a detailed solution. The only Olst,..bIltlon Fa",o •• i1c
Carry_over Fac,o. liZ
change that has been made is in the stiffness Fixed End Mom..."" -a83j1Sa
factor of the support E, Nhich has been taken 9alanC. Jain' II ill t 15
Carry.o..... to C
as equal to the beam DC, thus ~~ unbalanced f
BalanCe JOIRt 0 I
moment at this poi~t is equally divided between Cury.o..... to C
the beam and the support. Bal:lnce JOin, C
C.. rry_ove.:o 0
Solution 2 is a ~odi!ied solution which Elal:lnc. 0
elimlnates considerable arithmetic. Thus it is FIn>,j Mom",,-t.o

unnecessary to lock and unlock joint Band D

since ~e know definitely that one is freely sup-
~orted and the other 50 percent fixed. ~here­
fore, once we have r'e Leaaee these joints from
their assumed fixed end state, we leave them in
their natural state. The stiffness and carry-
over factors for beams CB and CD ~ust then be
cetermined for these beams with their ~odified
end conditions.
By reference to the fundamental equations
for stiffness and carryover factors in Art.
All.5, it 1s readily seen that the stiffness
factor for C8 1s 3/4 as ~uch as when fixed at
its far end B and the carryover moment is zero.
For beam CD the stiffness ts 7/8 as much as wnen
fixed at end D, and the carryover factor is
2F/3 + F = (2 x .5)/3 + .5 = .286. ~ith these
modifications the solution Is carried ~hru with
a relatively small number of steps. ~hus in
solution #2, joint 8 is unlocked. The unbal-
anced moment of -115 is balanced statically by
distributing 115 to BC. The carrJ-over moment
of .5 x 115 = 57.5 is carried over to C as shown.
Joint B is now left unlocked or free to ~otate.
Joint 0 is unlocked next. The unbalanced moment
is -432. It is balanced by distributing 216 to
DC and 216 to E since support at E is considered
to ~iYe 50% fiXity. ~he carry-over moment of
equals (-768 + 432 + 57.5 + lOB) = - 170.5, or
170.5 is necessary for equilibrium. Joint C 1s
balanced by distributing .392 x 170.5 = 66.8 ~o
CB and the remainder of 103.7 to CD. The carry-
over moment to B is zero and to D it equals
103.7 x .286 = 29.6. The final moments in sol~­
tion i W are slightly different than solution #1.
If another cycle had been added in solution #1
the descrepancy would be conSiderably smaller.'
All.7 Continuous Beaas with Yielding or Deflected
Supports
In Wing, elevator and rudder beams the sup-
port points usually deflect cue ~o the defo~2­
tion of the supporting str~ts or Nires in the
case of a wing, or to the deflection of the
stabilizer or fin in the case of elevator and
rUdder beams. If these beans are c~ntinuous this
deflection of their support points causes ~cdi­
tional bending ~o~ents in the jeams. The moment
distribution y.ethod can of cou:se be ~sed to find
the addi~ional ~oments due to this deflection.
T~us Exan~le Problem #5, shows a solution illus-
trating a ~roblem which involves the de:lecting
of the supports of a continuous beam. Due to
sJ~etry of structure and loading, the Slope at
D 1s zero or t~e oeam way be cons1cered fixed at
 •

ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES All 9

Joint D. Since the mOment of inertia is con- Since the fixed end ~Oments are due to bot~
stant and the spans are cons tant , the relative lateral loads and support def:ect~on, the values
stiffness factor of the beam is 1. In the so- as listed in the solution table Nill be explain-
lution shown since beam ~s freely supported at ed in detail.
3 t.nt s joint is lett tree to rotate after re-
leasing and thus the stiffness factor of beam Fixed End Moment For Lateral Beam Loading
CB Is 3/4 x 1 = 3/4, when compared to one haVing The distributed airload is trapezoidal in
full fixity at B. shape. The fixed-end ~oments for a trapezolcal
Since the first step in the solution proper loading from Table All.4 are:
1s to assume the joints fixed against rotation,
it Is evident that deflecting one support rela- Ml_2=L.a (5u+2V) (See Fig. (a))
tive to an adjacent support will produce mom- 60
ents at the ends which are assumed fixed against M2_1==L.a (5u+3v) (See Fig. (a ) I
rotation. 65
All.a Fixed End Moments ..L_ -- -- -=
Due to Support
Deflections

Fig. All.IS shows a T

v
-
fixed end beam. The sup- u Fig. (a]
ML port B -Ls deflected a
1. -L
M ~

ll.6 ~~
Ef ,~
i~ ~ -~~
4EI distance d relative to
the support paint C. It
the me:nber ls or constant
f

CD
L
"
@

6
1 LJ ~~O~~~~:~~~~~ -~~~lP~~~i
at the beam mid~oint and
For Span Be:
40'
Fig. All. 18 ME - Me. !1bc (5x3+1) = 426 ~n. lb.
60
By moment area prln- 40'
cipal deflection !'leb (5x3+1.5) 440 in. lb.
a !1L L -J1L 5 L
60
4El (3 4El (3 For Span CD:
or d == ML a 40'
5El I'lcct 60 (5 x 3.5 + 1 I = 494 In. lb.

hence 40'
I'l 6EIa 'tha.magnt tude for the fixed end Mct.c = 6'Q" (5x3.5+1.5) = 507 in. lb.
-V
mO:nent due to a transverse support settlement ?lxed End Moments Due to Sunnort Movement
of d.
From Art. All.S l1 = ~Eld/L.a
Examole Problem #5. Continuous beam with de-
flected supports. For span Be:
Deflection of B relative to C = 5/16 inch. Hence,
General
data; 3 f/in. 4 '''/In. 3'/- It-·049
A1.

! ; ; t
U1;J (Alloy tube !'!be = Me'::l =6 x 10,000,000 x .03339 x .3125/40.a == 390
o ": • , I j , tt.:,1:.03339
in. lb.
I!
-jlOj-- 40 ---r--
I
40 - - 40 40 ~ --110~- F'Qr Span CD:
ABC o Defle<:,tion of Supp?rt!l
t i 3/ 16 "
DI --- ?'lcd =t""'dc :: 6 x 10, 000, oeo x .03339 X .1875/40
a
;:: 234
- -f'11e<I due 10
1::1. 10.
Solution:
A iB
&) i
C
I For si;~s of tte ~oments due to tr.ese deflections
see Art All.2. Havir~ dete~ined the fixed end
Stiffness factor K 3/4x1~. 75 1 ~ooer.ts t~e ~eneral distributing and carrYing
Distl'1.bution Factor ~ (t,i ZK 011 . i21l .571 n I over process follows as indicated in the solution
C.u-ry.over Factor 1/' o 1/2 112 table. Thus at joi~t B, the unbalanced moment ::
Ftxed
End
Due to
Lateral Load
50 426 507 (50-426 -+- 390) :: 14. Balance by df s tr-t out t ng - 14
Momentll Due to support
deflection
o 390 390 234 ,,. xl:: - 14 to Be and Zero to EA. carryover.5 x
-14 :: - 7 to C. Considering joint C, the unbal-
Balance joint B
anced moment == (440 + 390 -+- 234 - 494 - 7) :: 563. Bal-
Carry_over 10 C
Balance joint C
ance bYdistri but tng - 563 x .571:: - 322 to GD and
Carry·over ~o B. D ·161 -563 x .429 :: - 241 to CB. carryover.5 x -322 ::
Balance joint 0 o ·580 - 161 to D. A~ joint D the unbalanced ~oment =
Final Moments 50 _50 580 ·580 (507 + 234 - 161) = 580. This is balanced by dis-
Conventional Signs tributing zero to DC and -580 to the fixed support.
All.I0 THE MOMENT DISTRIBUTION METHOD

111.8 Check on F1nal Moments &1~b - .5AMha = ~ _

- -(5)
To satisfy statics, the algebraic sum at M1ac - .5AI1sa Ka c
the moments in all the oe~bers at a joint must :: equation (5) is satiSfied the accuracy or the
equal zero. This requirement alone will not moment distributions is established.
prove that the final moments are correct, as er-
rors could have been made in the various distri-
All.9 End Moments tor Continuous Frame.orks 'bose
buting and carryover moments. continuity tells )(e=bers are not in a Straight L1ne. Joint
us that the final slope of the elastic curve at Rotat1on Only
anY joint is common to all ~ambers meeting at
that joint, thus for a complete check it is Continuous truss frame works are quite com-
necessary to prove that the final moments are mon in aircraft construction. Welded steel tub-
consistent with equal slopes for the various ular fuselages are composed of ~embers ~hlch
members. The rotation of the joints from their maintain continuity thru the jOints due to the
original assumed fixed or locked condition is welding. Landi~ gears frequently :~ve ~NO
due entirely to the distributed and carryover members which are made continuous at their points
moments. The actual rotation will therefore of connection. The members of such structures
equal the rotation of the end of a simply sup- usually carry high axial stresses which cause
ported beam when subjected to end moments equal joint translation which in tur~ prOduces bending
to those produced by the algebraic sum of the of the members since the joints are rigid. ~hese
distributed and carryover moments. These end moments produce lateral deflection of the memb-
moments equal the final moments minus the origi- ers which introduces additional secondary moments
nal fixed end moments and will be referred to as due to member axial loads times the lateral memb-
AM moments. er deflections. These influences are treated in
From Chapter A7, the slope at any point on a later articles. In this article to further fa-
simply supported beam equals the shear d t h miliarize the student with the moment distribu-
M/EI diagrams as a loading. ue a ~ e tion procedure, the effect of JOint translation
and secondary moments will be neglected.
The slope QA from Fig.a!
equals the beam shear at Ex8!llple Problem #6
i--L/3=i 0 MabL A or equals the reaction Fig. All.l9 illustrates a simplified land1ng
---.--- 2EI
, at A. gear chassis problem. Let it be reqUired to de-
c.Mab ,. termine the bending moments in the two ~embers
Ell 'C.g. due to the vertical load on the axle. The prob-
-AI B
lem has been solved using ~hree different de-
1-1----- L grees ot restrain~ at ends A and B. Joint 0 is
a welded joint and full continuity is assumed
thru this joint. The solutions as given in Fig.
All.20 give only the moments due to jOint rota-
~c. g.
tion under prumary bending moments. The effect
of axial deformation and secondary moments due
i to member deflections is omitted in these solu-
-_!-I-L/3 tions. These factors are treated in later arti-
•
.o.MbaL cles .
2Er" In a practical problem the degree of re-
2/3 _ <IIlbaL ~ straint at points A and B would be dete~ined by
2E1 • 3 the type at tittlng used and also on the r1gidity
(&lab + .5 <IIlba) of the adjacent fuselage or Wing structure. As
illustrated in later ex~p~es, the moment distr1-
3EKab bution method permits the consideration at the
where rigidity of the adjacent structure without adding
I any diffiCUlty, While such methods as least work
Kab = L or AB

Since the angle QA must be the same for 1 1S" 32"

members meetir~ at A, the general relation be-
t;....e en the moment increments at any t'NO members
such as AS and AC must be,
&lab + .5 <IIlha
M""-ac + .5 &1 c a
=~
Xac
T.
N

To make this equation consistent with the assum-

ed Sign convention, that is, the carryover mom-
ent has the same Sign as the balancing distri-
buted moment, the above equation must be modi-
fied as follows:- Fig. Ail. 19
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
K :II Stiffness Factor
Values in ( ) are carryover factors
Fig. All. 20
are not practicable because of the large number
of equations that must be solved to obtain val-
ues tor the many unknowns.
Solution torCondit1on I, Fixed at Ends A and B
Referring to Fig. All.20, all joints are
assumed locked against rotation or fixed. The
vertical axle load of 6000# produces a counter-
clockwise moment of 3 x 6000 = 18000 in. lb.
about jOint O. The Sign is positive (See Art.
All.2). Release or unlock jOint 0, the unbal-
anced moment is 18000 or - 18000 is required
for static equilibrium of joint O. Joint 0 is
balanced by distributing - 18000(.464/.464 +
.369) =-10030 to member OB and the remainder or
or -18000(.369/.464 + .369) = -7970 to OA.
~hese distributed balancing moments at 0 produce
carry over moments at A and B.
Thus carryover to 8 •. 5 x - 10030 = - 5015
and carryover to A, .5 x - 7970 = - 3985
Proceeding to joint A which is a fixed joint,
the unbalanced moment of -3985 is balanced en-
tirely by the rigid support, or no rotation
takes place when joint is released from its im-
aginary fixed state. Similar action takes place
at joint 8. The final end moments are as
shown in the Fi~Jre.
Solution ror Condition II. End A and B Pinned
For this condition the ends A and Bare
freely supported. Instead of locking and un-
locking these joints which are definitely known
to be free
All.ll
to be freely supported, they will be left in
their true state. Thus the carryover moments
from end 0 will be zero. Since A and B are both
pinned, the relation between the relative stiff-
ness factors of members OA and 08 remain the
same as in condition I, thus the same K or
stiffness factors that were used in condition I
can be used in distributing moments at joint O.
Joint 0 is balanced in same manner as condition
I but with zero carryover moments to A and B.
Solution for Condition III. A is 50% Fixed and
B is Pinned.
Since each member has a different degree of
fixity at its upper end, the stiffness and carry
over factors will be considered in detail. In
condition I since both members were fixed at
their upper ends the relative stiffness factor
of each member was proportional to IlL tor the
member and this ratio was used. The general ex-
pression for stltfness factor is K = EI (3 + F)/L
carry over factor 2F. For member OA, F
3+F
equals .5 Since A is 50% fixed, and for member
OB, F is zero since 8 is treely supported. Hence
for member OA
K=EI(3+ .5) =~=3.5X .1105 E= 0129 E
L L 30 •
C.O. Factor from 0 to A = 2 x .5 = .286
3 + .5
For member OB
K = gL (3+0) =3x .185xE = .0139E
40
C.O. Factor From 0 to B = 2 x a =a
3+0
Considering joint a in Fig. All.20 the external
moment ot 18000 in.lb. is balanced by distribut-
ing - 18000 between the two members in proportion
to their stiffness factors. Hence - 18000 (.01291
.0129 + .0139) = -8650 in Ibs. is resisted by OA
and the remainder ot -18000 (.0139/.0129 + .0139)=
-9350 to OB. The carry over moment rrca 0 to A =
.286 x -8650 = -2475 and zero tram 0 to B. (See
Fig. All.20)
Example Problem #7
Fig. All.2l shows a structure composed ot 3
members. Member AO is subjected to a transverse
load of 120#. Joint A is fixed, B is freely sup-
ported C is 25 percent fixed and joint 0 is con-
sidered to ~intain continuity be~Neen all memb-
ers at O. The end moments on the three members
due to the transverse loading on ~ember AO will
be detennined.
Solution #1. Fig. Al1.21 gives a solution USing
the "Cross" method in its fundamental unmodified
state. The solution is started by assuming all
three members as fix-ended. The relative stiff-
ness factor K of each member is therefore pro-
portional to IlL of each member. These K values
are listed in ~ig. All.21. The distribution
All. 12 THE MOMENT
factor D for each ~ember at each ~oint which
equal K/~ is recorded in c=lon each member'
around each jOint. Thus any balancing ~oment
is distributed bet~een the joint members as per
these d1stribution factors. The carryover
factors for all members is 1/2. The fixed end
moments due to external loading are computed
tor the three members. For member AO, the tix-
ed end moments equal PL/8 = 120 x 20/8 = 300 in. OC=25{3~.25)=~ 3 •. 25 154
lb. The other two members having no transverse
loading, the fixed end moments are zero.
In this solution the or1er of joint con-
sideration r~s been AOEe and repeat. starting
With joint A the joint is released but since
the member AO is actually held by a fixed sup-
port, no rotation takes place and the balancing
moment of 300 is provided entirely by the sup-
port and zero by the member AO. The carryover
moment C, to 0 is zero. Releasing joint 0, the
unbalanced moment of 300 is balanced by distri-
buting - 300 between the three members accord-
ing to their D values, thus - 300 x .416 =
- 125 to OA; - 300 x .168 = - 50 to OC and
- 125 to CB. To prevent contusion it is recom-
mended that a line be drawn under all distri-
buted balancing moments, thus any values above
these lines need not be given tur~her consid-
eration and only values below the lines need
be considered in later balanCing ot the jOi~tS.
Immediately a:ter dlstribu~ing the moments at
joint 0 the proper carry over moments should be
taken over to the far end ot each member, thus
- 62.5 to A, - 62.5 to Band - 25 to C. Joint
8 is next considered. The unbalanced ~oment is
-62.5 and it is balanced by distributing 62.5
to 80 slnce the pin support has zero sti:tness,
or no resistance to rotation. A line is erawn
under the 62.5 and the carryover moment ot
31.25 is placed at O. Joint C is considered
next. The unbal~lced ~oment o~ -25 is balanced
SOL111'ION'IIl
SWmellB Factor K
[ a
OA=I:- 10 : .10
5;,lO
08= lis
DC = 225 =~
ZK·JointO=,24
C.O,F",ctor_1/2
far All Memo.
Fig. All. 21
DISTRIBUTION METHOD
Member - K. Value c. O. Factor
~(hF)
2F
3- F
2x1 =1
OA:-l..(3+1J "'.4,
20 ~ 2
2xO
OB.i·~{3+0) :.3, 3 +0 • 0
1 2x.25:.
ZK=.83
SOL tJrION .2
Fig. All. 22
by distri but.fng .75 X 25 =18. '75 to CO and the r-e-.
mafnder- of 6.25 to the support, s tnce the fix!.. ty -r;
the support at C has been assumed as 25 per-cent _ A
line is cra~m ~lde~ t~e 18.70 and the ca~ry over
moment of 9.37 is taken over to 0.- One cycle has
now been completed. Retur~ing to ~olnt A, ~e
find -62.5 below the line. This is balanced by
distributing zero to OA and 62.5 to the t!..xed
support. A line 1s drawn ~~der :he zero distri-
buted ~oment to AO and the carryover woment of
zero is placed at O. Considering joint 0 f~r
the second time the unbalanced mo~ent is 9.37 +
31.25 + 0 = 40.62 or the S~~ ot all vaiues below
the col~ horizontal lines. The joi~t is bal-
anced by 1istributing - 17 to OA ~~d 08 ~~d
-6.62 to OC. Lines are dr~Nn under t~ese bal-
ancing momen~s as shown in Fig. All.21 and t~e
carryover mcments are taken over to the far
eTIQs Jefore proceeding to joint 3.
This general process is repeated until ~oint
A has been balanced 5 times and the ot~er jOints
4 ti~es each, as indicated ~n the figure the dis-
tributl~g values have become quite ~ll and it
is evident tha~ a ~igh degrs8 of accuracy has
been obtained. The ~inal end ~oments at each
joint equal the algebraic sum of the values in .
each COIWL~. A double line ~s placed above the
r inaa noaerrts as a df s t tnguf sht ng symbol . In
the fi~~e the letters band c ~efer to balancing
and carr; over ~oments, the sUbscri~ts ~etarring
to the member of the balancing or carry over
o=eration. Any order of joint consideration can
be ~sed in ~eaching the same result.
Sol~tion #2 of ~oblem 7
Fig. All.22 gives a second solution. ~ith
the end conditions ~own at A, Band C, the ~odi­
~ied stitfness factors of the members can be
found together with the ~odit1ed carryover fac-
tors, thus making it ~ecessary to balance ~oint
o only once and car~y over this final far end
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES All.13

~oments of each member. The fib~e gives the assumed to give 50% fiXity to these joints. In
calculation of the modified or actual stiff~ess Table A11.2 a modified stiffness factor is cal-
and carryover factors. ~ith these !u~c,vn the culated tor ~embers GI, Fl, and FH using a 50
solution is started as before by computing the percent fiXity at their tar ends. The last
fixed end ooments due to transverse loading on column of Table All.2 gives the s~tion of the
~e~ber AO. Joints Band C are released and ~ember stiffneSS factors for members intersect-
si~ce no fixed end moments exist, no balancing ing at each jOint.
1s required and the joints are left in their 120*
true state 0: restraint instead of locking and 3200"-' J. l-
,- .12'", G 1
unloCki~ as in solution #1. Releasing joint 0

~:/7\ ~~~:'~d
tram its imaginary fixed state the ~~balanced
moment is 300 which is balanced by distributing /\ /
- 300 betNeen the 3 connecting members accord-
ing to the new distribution factors at joint o.
Thus - 300 x .482 = - 145 to OA; - 300 x .361 =
- 105.2 to OB and - 300 x .157 = - 47 to ce.
~~e carry over ~oment to A = - 145 x .5 =- 72.5
l~7;''t ~~-~
Engine
--t1d'r-
MOWlt 2400\
,~~
I~
A
12400".
F
r
H

'\ /
t.o B = - 108.2 x 0 = 0 and -47 x .154 = - 7.2 to

Exa~ple Problem #8
"'\ -, f -,
/
Landing
Gear
Fig. All. 23

figure All.23 shows the forward portion of

a fuselage side truss. Due to eccentricity of TABLE AiL 2
engine 2o~,t and la~ding gear members, external
~o~e~ts ~e produced on joints A, Band D as Leuct h 1. x 1000
L
shown. Fur~he~ore lateral loads due to eGuip- it1ffA••• F-''-,-...,-,----j
~er.~ installation are shown acting on ~embers BE
~nd CD. Assuming the fuselage welded joints
pr ocuce rigi::i continuity of member-s thru ttie
joint, t~e problem 1s to find the end moments in
. ·
cr
AC
1-1
1
,0< ,
- ,0<'
-
8 _

34.'
, 1.0240
.0240

- .049 19.25 .0588

_1 2 .0588
.049 41. 2~" .,,240
J:. J'aetar

0.582
61
.S18*
3.06
A
IS
C
D
S 80
4 81
6 95
6.14
all the ~a~bers due to the_eccentric joint ~om­ -- 2 -- .049 30.0 .0588 1 6 ,sa
,.. _3 - .,
en~s ana two lateral loads. The effect of joint
..
D' 1. 96 7 4.42
.0449 6 G 3.21
tr?-nslation and secondary ooments due to deflec- • - •• -- •• 34. • •
..
:I 2•

o s•
3 .033 D.
t10TIS and axial loads is to be neglected in this
ex=..m.ple. • 1_1_1 2 - .049139.5 0178
Be
- - • ". (lSBB (II: 1. eoutaJIt
ell 8 - o (I:l~141 0 •
,• fol' Lll _ -
So1.:J.tion: 1_ - >en)
~able All.2 gives the calculation or the
( .t1ffn••• fLctor _ 7/8 ~ beca~ of ~OS f1xity)
stiffness factor for each truss member. The L
~uselage truss aft of joints I and H tave been

Fig. All. 24

..
_1.98 (
.1.82 b

-,
(n5
-141
b .178 _187
. ,a ee .", "'-, --. ·,
"
• 18 b b -:....11. .286

-"
·.,84
1)---i
0-'.
c-t
-'-'
b
'-."
...:l.o
'-.-.i..
c ·n ·"',....
«n .195 0-'-
.l15
•

"'."' ......,..,....
:::~ sr- -s
" .

..
__ .834
:...JJMb
na --o...ill
~b aa .,29 b
_ _ b· 140
n•
-
70 • 140 -10 ->;.286
',0 b ~
" ,
. ~b
-,
,
~ . --1)·15
':""""55'
m" b-rot ; ~
~"
.\11.14 THE MOMENT DISTRIBUTION METHOD
Fig. All.24 gives :he solution of the
problem. The procedure in this solution was as
follows:
The stiffness factor K for each member as
computed in Table All.2 15 recorded in the cir-
cles adjacent to each truss member. The carry
over factors for all ~embers 15 1/2 except fcr
modified members GI, FI and FH for which the
carryover factor to the 50% fixed ends is .286.
The distribution factor for each member at each
jOint is recorded at the end ot each member, and
equals K/LX.
The next step in the solution is to cocpute
the flXed end moments due to the transverse
loads on members.
For member BE
MBE=Paba/L li =120x29.25x12.11/41.2S ll =-2J8"#
l1E:B =120 x 29 .25' x 12/41.25' =725"#
l1cD = 100 x 20' x 10/30' = - 445
l10c = 100 x 10' x 20/30' = 222
These moments are placed at the ends of the joints.
members on Fig. All.23 together with the eccen-
tric joint moments. The process of ur~ocklng
the joints, distributing and carrying over mom-
ents can now be started. In the solution as
given the order of joint consideration is
ABCDEFG and repeat, and each joint has been bal- first step 1s to compute the fixed end moments
anced three t1.mes.
Consider joint A:-
Unbalanced moment = 2400. Balance by dis-
tributing - 2400 as tollows:-
To AC = - 2400 x .527 =-1258. Carry over
to C =-634
To AB 2400 x .473 = -1132. carry over
to B =-566
Proceed to Joint B:-
Unbalanced moment = (-566 + 3200 - 298) =
2336. Joint 1s balanced by distributing - 2336
to connecting members as tollows:- • in balanCing joint A 1t will throw over to 8 the
To BA =- 2336 x .569 =- 1330. carry over to
A = - 665
To Be = - 2336 x .310 = - 724. Carry over to
C - 362
To BE =- 2336 x .121 = - 282. Carry over to
E = - 141
The convenient device of drawing a line
under all balancing moments is used to prevent
contus1on in later balances of the joint.
Proceed to Joint c:-
Unbalanced moment = (-634 - 362 - 44S) =-1441
The joint 1s balanced by distributing 1441 as
tollows:-
To CA:= 1441 x .44 = 635. Carry over to A = 318
CB = 1441 x .214 = 309. carry over to B = 155
CE = 1441 x .064 = 92. Carryover to E = 46
CD =1441 x .282 =406. carryover to D = 203
This process is continued tor the ~emainder or
the trJSS jo rrrts, ar tar all joints has been
balanced once, on ret~1r~ to joint A we rind
below trie lines an unbalanced moment; of (318-
665) =-347. The jotnt is balanced a second time
8Y distributing 183 to AC and 15~ to AS with
carry over mo~ents of r~l: these values to ends
C and B respectively.
The student should now be able to c~eck the
rest of the solution as given on Fig. All.24.
The solution could be ~de with any order of
joint consideration. If any particular jotnt ap-
pears to be nearly balanced, it 1s best to skip
it tor the time being and consider those joints
which are considerably unbal~~ced.
7he ~inal moments at the end 0: each me~ber
are given below the double :i~es.
Example Problem #9
Fig. All.25 represents a cross section of a
welded tubUlar steel tuselage. The top and bot-
tom members which are web members in the top ~~d
bottom fUselage trusses are subjected to the
equipment installation transverse loads as shown.
Let it be required to deter.nine the end bending
moments in the rectangular trams due to these
transverse loads assuming full cont:nuity thru
Solution:
Fig. All.26 shows the solution. The dlstr1-
bution factors based on the member sti!fness fac-
tors are shown inD at ends of each member. The
due to transverse loads, on members AS and CD us-
ing equations from 7able All.l. The ~gnltudes
are 1890"# for AS and 2025~# for CD.
Joint B is now released f~om its assumed
fixed state. The unbalanced moment of 1890 is
balanced by distributing - 1890 x .247 = - 467 ~o
SA and the remainder of -1423 to BD. T~e carry
over moment to A = - 46S-x .5 = - 233. Due to
symmetry ot struc~~re and loading only one t~lt
of trame need be considered and hence these car-
ry over moments to A are not recorded. However,
same aagrrrtuee of carry over moments as thrown
over to A from B but of opposite sign since the
original fixed end moment at B is minus. Thus
233 comes to 8 tram first balance of A as shown
in the figure. The distributing moment to a of
-1423 produces a carry Qver moment of -1423 x .5=
-712 at D. ~27"
300# 300#
I
10··--t- 7" -r-10"-j
1"-.035 Tube
T
K=· 0~~37::. 000458
~
0:1 0 0:1
36" ~ 0 ~ :::
.
. ,. 0
'~I~
<;gOM :<;g
0
11
I~ ~ 600# III~_~.J
I. :
-Llr13.S'~13.5"-W
1~118 - .035 Tube
K=L:·01782= 00066 Fig. All. 25
L 27 .
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES All. 15

A~.247 B tors will not be considered here but will be

treated in a later article.
~1 f 1890 (A) Calculation of Fixed End Moments for a
~:ill
X 1/2 233
Given Beam Load
~~
X 1/2 -27
Fig.All.27 shows a fixed end beam with a
variable moment of inertia and carrying a Single
~-'L
-15 concentrated load of 100#. The beam ~oment dia-
b.~ gram for this load is considered in three parts.
1710 0 F1g.All.28 shows the static moment curve as-
suming the beam simply supported at A and 8 and
C -2025f 660 .322 carrying the load of 100#. Flg.A1l.29 shows
X 1/2 f 2025 the other two parts, namely the triangular mom-
~-421 ent diagram due to the unknown moments MA and
X 1/2 210 MS which produce fixity at the two ends. In
~-93 this figure MA and MS have arbitrarily been taken
X 1/2 47 Fig. All. 26
~ b3 ~30 as 100, instead of unity.
-......:c.;: 15 Fig.A11.30 shows the MIl diagram for the 3
b. -10 moment diagrams of Flg.All.29. The beam is di-
1743 vided into ten equal strips, and the MII curves
are obtained by dividing the moment values at the
end of each strip or portion by the corresponding
Joint D is considered next: The unbalanced mom
I value from Fig. A11.27.
ent of (2025 - 712) = 1313 1s balanced by d1s- From the cond1tions of fiXity at the beam
tribut1~g - 1313 to the connecting members.
ends J we know that the slope of the beam elastic
~o member DB =-
1313 t .678 = - 892 and curve 1s zero at each end. Likewise the deflec-
the remainder of -421 to DC. The carry-over tion of one end of the beam away from a tangent
moment to B is - 446. The carry-over trom C to at the other end is zero. Stating these facts
D i& Dne-half the balancing moment S = - 421 in terms of the moment area principles, we ob-
but of oPPosite Sign or 2l0 J due to symmetry as tain
ex?lained before tor member AB.

~o 11~
Returning to joint B, the unbalanced as
recorded below the single lines is (233 - 446) L _ a - -(5) (Area of ll/I diagrams
= - 213. ~o balance 160 is distributed to SD
equal zero)
and 53 to SA. Carry-over 80 to D and bring
over from A to S .5 x-53 =- 27. Continue this

i
L~ -O
~rocess until joints A and D have been balanced
4 times or 4 cycles have been completed. The - (6) (Moment of the WI dia-
tinal moments are shown below the double lines. I - gram as a load about
?ig.All.25a shows the resulting moment diagr~ o either end equals zero)
on frame.
(Note: Since E is usually considered constant
it has been omitted trom denominator or the above
171

r :rwo®~1710
Final
§
z
equations. )
TABLE All. 3

_...
~ J-
Moment
~ Diagram
_.......- Static all C1U"'N Trial -
t
<:Ur"N Trial ~C1U"'N
r

A'r::' ,. "',. ..... ,. ..... , 1-;' I

S\
~17~a~1743
a::i h. .~ h. A~
Ord. lett left left •
,..
. 1 etl'1.z e.d. •
, I

,... ""
("
4050 i I
,
Fig. All. 26a
a
a
'"
,...
".
'" ,...
2.67
6.:n
U.
:l1ll0 U3 I 6.07
". r.. ...
"•
•• r 2. 67 1
,::~ ::.9 ,~~~~ U.

..
10 11l
All. 10 Continuous Structures with Members of
'"
,
•
..,.,...."
1120 14.10 111600 1130 13.i" 1810. 70.0 1O
Variable Moment of Inertia
I • 18.07 26000 1110
22 06 ,38900
17.96 li761 90.0 18.06 U30 I

...".'" , i " .,
1760 21 92 1970 110 0 :z:.:I cs
In Arts.Al1.3, 4 and 5 consideration 'HaS
• I 615 211.88 i41600 67.3 2ll.1l8 117<lO!~~~.1l 26.04
I
2lJ. 77 aeoec 40.7 29.76 1 121.5!117.0 211.94 I
given to the deri~~tion of expressions for
fixed end ~oments, stiffness and carry-over
ractors for beams of ~iform crass-section.
I •
"r
,,. .... 179Ctl1l
:i·g
33.llJ i4180 17.11 33.8.

M.5.7
, ~~I 33.1• "00

111122:2 1
743 8
•
" 1812:1:

Many cases occur in routine deSign where mem-

bers have a var~able cross section. This
,,
i -eza ..
.. 21.1.5"
....
!1!!!!! i ,,11922 ..14.1"
845.7
i . ~ 2....."
143.8
,,
article will illustrate ~he calculation of the
fixed end ~oments, stiffness ~~d carry-over . Actual area u; 4 u .... t o rnults a.II._.
For ca1c.l.at1_ ot
U;lt .... "d. _ l i t e sZc.o. facto", 0Il.1,. ...lati_ Talus are
I

)
factor for a beam with var~able ~omen~ of in- _·''IT·
ertia. ~he effect ofax:al load on these ~ac-
 All. IS THE MOMENT DISTRIBUTION METHOD

P identical to F1g. All.30e, ?or

A B
} 10"* the conditions of su~por~ ~ss~~eG,
Fixed & f Fixed the deflection 0: A away rr ce 3-
24" 16"
2,,0 in.
tangent to t~e elastic c~rve ~: 3
Fig. All. 27 1s zero. Th~s by the de~1=c:1Q~
1.0 1.2/' principle of area ~o~ents, t~e
.75 .75 ~oment ot the Mil dlagr~s of
55 .5 .5 .5 .5 . 55 I Curve
Figs. All.30 (a and b) about
( 1) ( 2) I (3) I (4) ( 5) (6) , (7) 18) 19) (101 end A eq~als zero. ThUS,
74,3.8
845.7X14.1+l'5O ~Bx24.4=0
aoo 960"*
320 480 or
Fig. All. 28 169 I • M curve 240 :1B= -66"#
I

Since MA was ass~~ed 100, then

Fig. All. 29 carryover :actor tram A to 3 =
Trial MA -66 =- .66.
: ·100 100
To find carryover factor from 3
to A, take moments about B ~~d

00
r curve equate to zer8.
Hence
845.7 M ....~,....
(a) 743.8x 15.5.;.l'CO ~.Ax .:::V.~ = 0

whence
MA = - 53.2"#
(b)
~heretore carry over :~ctor 3 to
A =- 53.2 = - .532
Fig. All. 30 100

13 (~ota: ?or t~e no~ent sign ccn-

(e) ,13.4 36 i. I ver.t~on used in this book c~~~y
~ j . 60 180 over factor Nould be ?lus.)
2.67 6.2 10:1'6 14j 18.06 (C) Calculation of Beam St~~f­
37.86 :;.ess Factors
1------ 24.4 'Jhen a ~eam :s freely sup-
ported at one end A and rixed at
743.8
the rar- end 8, the st i rrness rae-
Table A.l1.3 shows the calculations tor computing I the ~oment necessary I tor at the A end :s measured by
at 'A to ~roduce ~it rota-
the areas and the centroids ot the three Mil tion of the elastic c~ve at A.
diagrams of Fig. A11.3D. Substituting values In Art. Al1.4 it was nroved for be~s of
tram Table in equations (5) and (6). uniform section tr2t EgA; MAL/4I or MA = ~EIgAI
845.7 M 743.8 8504 0 L. In a continuous structure at any joint all
100 A + 100 MS + ::: -- -(7) members have the sa=e g, thus 4E9 is constant
and the st~f1"ness K of any ~rismatic beam is
11922 18122 proportional to IlL. For beams of variable
100 MA';' 100 i1S + 179809 ::: 0- - -(8)
section the stiffness iactor K may Je 'Nritten:-
The value 100 in the denominators is due to the K=cIo/L - - - - - - - - - - - - - - - -(9)
fact that trial values of 100 were assumed for
MA and Ms. Solving equations (7) and (8) we ~here Ie is the ~o~ent of inertia at a particu-
obtain; lar referer.ce section and c 1s a constant to ~e
found for each non-uniform aember. T~us for
MA = - 315~# and ME = - 785"#. non-uniform ~embers
(B) Calculation of carry Over ~actors EgA =MAL/4cIo - - - - - - - - - - - - (10)
To dete~ine the carryover factor trom A to By the ~oment ~rea slope ~rl~c1ple, t~e
B consider end A as freely supported and B as I slope at A when 3 is rIxec equals the ar-ea of
fixed. A moment MA = 100 Is applied at A ~hich I the MII df agr-an oet.veen A and B.
prod~ces the ~JI curve of Fig. All.30b. Due to Tak~ng XA = lOa, ~s Nas ~o~~d ~J equal
the derlec~~on or the be~u ~~der this loading a -06"#. Thus EgA = 4(2<;'5.7- .6t3x'7~3.3) =l~::'d.
restra~ning ~oment at a to cause tar~ent of ';,'he va Iue 4- is cue :0 t.ne etr ic i ..1d.t:. since
alast1c c~rve to remain ~or:zontal w~ll exist. true area is Nan:ed ~nd on:y avera;e ordinate
';,':a5 unknown fixed moment at 3 has ar-ot trar i iv was used in Tab~e All.3.
been taken as 100, uhus pr cduc i ng a :-;;'I cur-ve Equa t rng ::,15 r-esu.;'; ':0 gcua t t cn (:0)
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES All. 17

1'+' 19 = loa
ac xx40l ' whence c = .7 0 7 bent car:ying a distributed side load on one
leg as shown. The K or IlL values for each
10 was taken as the value at the A end or 1. member are g~ven on the figure.
(See Fig. All.27). Under the given loading it is obvious that
Therefore KAB = .707 IA/L the frame will sHay to the right or in other
Similarly ~or end 8:- words, jOints 8 and C will undergo considerable
100 x 40 horizontal movement. The moment distribution
;:gB = 4(743.8-.532x845.7) = '!exl method assumes that only joint rotation takes
whence place. To make this assumption true for thiS
structure we will add an imaginary support at
c = .85 and KBA = .85 IA/L joint C which will prevent sidesway of the frame
as illustrated in Fig.All.34. The end moments
All. 11 Frames with Unknown Joint Deflections Due in the frame will then be found by the moment
to Sidesway
distribution process. Fig. All.34 shows the
In the example problems so far treated the results of this process. To explain, the Solu-
~oi~ts of the particalar structure were assumed tion begins with computing the fixed end ~oment
to r8tate without tra~slation or with a definite on member AB !,.[L"'/12 =
(300x25 "')/12 =
15.63 =
amount of translatory ~ovement. Translation of thousands at foot Ibs. This value with the
the joints may however, be produced by shorten- proper sign is written at the head of a column
ing and lengthening of the members due to axial of figures on member AS as shown in Flg.All.34.
loads and by lateral sway due to lack of diag- Now considering Joint B, the unbalanced
onal shear ~e~bers. The problem relative to joint moment of 15.63 is distributed as follows:-
the effect of jOint translation due to axial To Member SA =- 15.63 (40/190) =- 3.27. Carry
stresses is treated in a later article. In over to Joint A = - 3.27/2 =- 1.63
this article only the effect of sldesway of
rectangular frames on the frame bending moments To Member 8C = - 15.63 (150/190) =- 12.36. Car-
ry over to Joint C = - 12.36/2 = - 6.18
';;111 be considered.
:1g. All.31 illustrates cases of tr~es
where only rotat1on of joints takes place (neg-
lecting axial deformation) whereas Flg.All.32
ll1~strates conditions in which sidesway takes
4 6' I 10'
1
6
'r-
place and the joints suffer translation as well 1= 10, 500 cl

T
B
as rotation. I
K=L'= 150
0 °
~

~ 0
~ g " 25'

-~~!J
~

°° "
~ -1-'
'"
A D
rnr
Fig. All. 33

1 Symmetry of Structure
& Loading
Frame prevented
from sidesway by
external restraint
Fig. A11.31

~Ll'>~II~I'~I~I~
- . ,-
(O.OO)f (0.00)
Symmetry
of Loading
<:"lOOIO'I_ -12.36 b 1 . -
+ 2.44 _ _ bl.
-6.18
4.88
... , ,,' II'"
but Structure '" " '"
.... .c.c..::::t.o .1.93b", -0.96
Symmetry of Str-uctur-e, Unsymmetrical + 0.38 . . - b 2+O.76
Loading Unsymetrical
Fig. A11 ....3.2
I;[ i ~
• O. 30 b~__

.O.04b ... -
-11.75
-0.1:>
_b,.,+0.12
_0.02
b..... 0.02
There are several ~ethods of dete~~ning -1. 53
the bending mcments Gue to sidesway. Only one
~ethod will be presented here and it can Jest
be explained by the solution of example prob- Fig. All. 34
lems.
~ample P:oblem 1.

Fig . .4.12.33 shews " single bay r-ec tangu.lar- I

All. 18 THE MOMENT DISTRIBUTION METHOD

horizontal shear reaction at the upper end of

the vertical legs of the frame equal to 3606 lb.
Bending Moments Due ta Sidesway.
ASsume the structure sways sideways as
illustrated in Fig. A1I.36, but With no
rotation of the ~pper jOints. It 'NaS proven
in Art. AII.8 that the end moments produced by
the lateral movement of one end at a beam whose
ends are fixed against rotation are equal to
6Eld/La where d is the lateral ~ovement. In
Fig. All.36 the load? causes both vertical
members to suffer the
Fig. All. 35
same horizontal dis- -idl- -jd l
placement at their .~---~-'-
: IB :C 00
upper endS, hence their
Now consider JoInt C:-
The unbalanced moment Is -6.18. To bal-
end moments due to this
displacement are pro-
portional to EI/L a. ~i
:/
Fig. All. 36
:
ance we distribute to CB = 6.18 x 150/190 :;;
4.88. Carryover to B 2.44; to CD = 6.18 x = Since EI and L are
40/190 =
1.30. Carryover to D = 0.65 the same for each vertical
Sidesway Without
The balancing and carryover procedure is now member in our s tr-acture , Rotation of Joints
repeated for jOints Band C, until the unbal- the fixed end moments will
~ced moments become at negligible magnitude.
be the same magnitude for
l' ig.AILM shows that 4 cycles have been car-
each member. Therefore, tor convenience we will
ried through. By keeping the ratio 150/190 set assume fixed end moments of 10,000 ft.1bs. are
on the slide rule the unbalanced moments at produced by the atdeswav, T..le now use the moment
Joints B and Care distributed'and chased back distribution process in permitting the upper
and torth as rapidly as one can WT'ite them jOints to rotate as illustrated In Fig. All.37.
down. Since joints A and D are assumed fixed The procedure is similar to that in Fig. All.34.
they absorb moments but do not give out any. J For example the solution is started by consider-
ing joint B. The unbalanced moment rs -10.
Fig. AlI.35 shows a free body diagram at This is balanced by distributing 10 x 150/190 =
each portion at the rrame , The end moments 7.69 to Be and the remainder at 2.11 to BA.
are taken tram the results in Fig. AI1.34. The carry over moments are 7.89/2 = 3.95 to C
Consider member AS as a tree body. To find and 2.11/2 = 1.06 to A. Due to symmetry of
the horizontal reaction HB we take moments loading and structure, the distributing and
about point A. carryover moments at joint C will be same as
at jOint B, hence it is needless work to show
Z!lA = 300X25 x 12.5 + 11750 -17570 - 25HB = 0
calculations at these joints. The carryover
moments tram C to B will be identical to those
hence He = 87930/25 =3517 lb.
tram B to C. FIg. A11.37 shows the 5 cycles
have been carried out to 0Etain the final end
Now consider member CD as a tree body. moments as shown below the double lines.
.. 7.89 b
To rmc He take moments about D. .3.95 - 1 / 2 Due to symmetry
- 3.12 b of loading and
lJ1D = 1530 + 770 - 25Hc = 0, hence He =89 1b. ·1.56 - 1 / 2 structure it is not
necessary to con-
+1.23b
We now place these horizontal forces on the _" 0.62 - sider right hand
top member Be as a tree body as shown in the -0.49 b portion as calcu-
- 0.25 - lations will be
upper portion or Fig. A11.35. The unknown
~b identical as shown
imaginary reaction He at potrrt C that; was t 8.47 for lett side.
added in Fig. Al1.34 to prevent sidesway is
also shown. To find He take ZFh = a B c

1:Fh =3517 + 89 - Be = 0, hence Be = 3606 Lb

Since the external reaction of 3606 does
Il~ I
Fig. All. 37
not exist, we must eliminate it and rind the
bending moments due to the sidesway or the o
frame. In other words, the frame will sway o
Sideways until bending ot the frame develops a O~N~=t
C:~~"":~OON
~",ooooci
I .. , .. , I
A
-
o
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES All. 19

Fig. All.38 shows a free body diagram of 100

the frame vertical members with the end moments I I ,
as found in ?ig. AIl.37. The shear reactions :--8" ......,.....8" ....,
HB and He are each equal to (8470 + 9230)/25 =
708 lb. These shear reactions are then placed
on the free body of member BC in Fig. All.38.
T B 1:=16
12"
For equilibrium the summation or the horizontal 1=10 15"
forces must equal zero. Thus P = 708 + 708 = 100 lb.
1416 lb. and acting to the right as shown in 1=24 D _J
Fig. AII.38. Since the reaction Rc = -3606 in Pin
Fig. AII.35 must be liquidated and since the 12"
liquidating force P produced by the moments in
Fig. AIl.37 is 1416 lb. it is obvious the
values in Fig. AIl.37 must be multiplied by a
1- A
Pin Fig. All. 41

factor equal to 3606/1416 = 2.545. Therefore, SOLUTION:

the final bending moment values equal those or
Fig. All.34 plus 2.545 times ~hose in Fig. Relative Stiffness Factors; -
All.37. Fig. All.39 shows the results and
Fig. All.40 the final moment diagram. KBA = 3I = 3 X 24 = 3
L 24 '
KBe = R = 4x 16
L 16
=4
11"
-c
-n =3IL = 3 x 10 = 2
15
708 +' ,..----~_=,.;tc- P' 1416 lb.
~B 708 The distribution factors for each member at
8470 Joints B and C, which equal K/ZK are recorded in
"FHB::lI708 _ 8470 the small 0 on Fig. AlL41a.
T
25" .
HC 1ll 708

Fixed End Moments: -

For Member AS = PL/8 = 100 x 24/8 = 300 Ins Lb,
Fig. All. 38
For Member Be = PL/8 = 100 x 16/8 =200
1 -6 9230
D

9230
As explained in example problem 1, the
moment distribution 's carried out in two steps,
one for jOint rotation only and the ~ther for
effect of sidesway or horizontal translation of
- 11.75 - 1. 53 joints B and C. Fig. Al1.41a shows the moment
,
~
9.81
~
+ 20.03
distribution for no sidesway by placing an
imaginary reaction Rc at joint C. The process
is started at joint A and the order of jOint
IB C
balancing 1s ABCBC. As soon as a joint 1s
Fig. All. 39 I
~~,
~~ ~ :lJ~
_-10
balanced the carry over moments are 1mmedlately
carried over before proceeding to the next
~
~,
+1, jOint. When a joint 15 balanced a horizontal
line 15 drawn.
~r
~ ~~
.. ~~~
F.E.M. ·200 200 F.E.M.

"
~ ~
d..;j . Bl. ·143 - - - 71.5
- 4 2 . 8 - - 85.7 a ,
A '7'~
mil\'( ~ " II ...
" 1I II
D B. 24.4 - -u:2
- 4.1-- -8.1 B~
9.81
~I e
~

_"='~"b7T
M M • B,-W--U
...:..., -363.2 ~ B lI

J <I~~
.... ~~ ::i

1 .
47.3
~. N g:
.-
l>.o~E-
e .~
~
.IN
r;::1~
el,-.
,
<D

.
t: "
"'''''''
.. " ..
"'''''''
4 1.07 22.73
0
• _- •0
~<r;:: ~I< ;:
0 OO~I~I='I~
o.a 0 . " .
<'") _ _ co ...., <'") C01_1"'IIM
Fig. All. 40
,_ 'cD
M '"1"'1 0 1.;
...
I
I
I
"
,
...
I

Example Problem. 2. Bent with Unequal Length

D
j jI
Legs and Pinned Ends. 00°11 0

Fig. All.41 shows a loaded unsymmetrical

trame. The final bending noments at Band C A Fig. All. 41a
will be determined.

,':);EifF"'c
,-'-..' -•.-':).:::';>it..
All. 20 THE MOMENT DISTRIBUTION METHOD

With the end ~oments ~own the reaction B),119 - 5 9 . 5

Rr can be calculated by a consideration of the 54.4-108.8 B),
free body of each member as shown in Fig. s , -31. 0--15. 5
Al1.41b. "5":2- 10.3 B 2
B" ~ _ _ 1.5

61..,. _
3.15 C
-----Rc' 61.85 ::i
144.6 ---l......Q. 53
162.6 -" .
'r
B
"'l .... I'l

"T"
363. 2 ~ H 65 HC=3.15 c-; ~ ;:Q C:U::l

~~I.''''I= cl
B" oiIMI~"'; IB
<:0 i
t-
;~<:Di~ F~ I I
10JI

35 A
!
D1_ 3 . 15 'i
0
~ ;1 0°110 QA
r l' 01
•
;ic5'
oJ
Fig. All. -lIb
. ~
Fig. All. 41c ~
;l;'"
oJ
To find Ha take moments about A, .; 6.0L _ p=6. 02" 10.85
B 10.85 =16.87 lb.
100Xl2+362.3-24Ha = 0, Ha "" 65 lb.
144.6 '0.85_")'62.6
..r:-:~ 02 C I,
To find He take moments about D, ~.

,
,

-47.3 + 15He = 0, He = 3.15 I

Placing these shear reactions on member BC
and writing equilibrium of horizontal forces
we find that Rc = 65 - 3.15 = 61.85 lb. ~e
-lA ol
• Fig. All. 41d

now must liquidate this reaction RC and per,nit
:rame to sldesway.
We will assume that the frame 15 :ixed
ended and that an unknown horizontal torce ?
at C will deflect the trame Sideways. The
fixed end moments for equal horizontal de-
flection of a and C will be proportional to
EI/L a ot the vertical member.
F ormemer I . 24 _
bAB 'P-24:
a -
To liquidate the reaction Re = 61.85 it
will therefore require 61.B5/16.37 : 3.67 tioes
the moments in Fig. All.4lc. Therefore the
!lnal bending moments equal those in ?~g.
All.41a plus 3.67 times those in Fig. All.4lc.
Fig. All.4le shows the results. T~e bendl~g
moment and shear diagram and the ~xial loads
now follow as a s iapje aat t ar of jsta t i cs .
-363.2 47.3
~ ..ll2..Jl.
0417 """'lS6:8 643. J
'

B Ci: ~1:;
F or memb er DC , r::aI -- 1510 " -- • 0445
ti gil; li~ ~I~
Mlt:Il_
",
For convenience we will assume 417 in.lb. as
the fixed end ~oments on AB and 445 on DC. I DOl O
The assumed fixed ends will now be eliminated
by the moment distribution process as shown I
in Fig. Al1.41c. The order of joint balance
was ABDCBCBC.
IA
\1 00110 Fig. All. He

Fig. All.4ld shows the shear reaction on
the vertical members at Band C. These forces
reversed on the top ~ember show an unbalanced
force of 6.02 + 10.85 16.87 lb. Therefore=
a force P = 16.87 was necessary to ~roduce the
bending moments that resulted on the frame
due to sidesway.
SXample P:oblem 3. 3enc~~g ~Oilients ~n ~~uss
Involving One Panel ~ithou~ 01a~Gnal Stear
Member.
Frequently, in aircraft struct~r~s ~ :rU2S
is used i~ which a diagonal member ~us~ os l~f~
out at one or ~are truss bays. T~e sy.:er~al
shear load on such bays ~ust je carried by :he
truss chord members in bending. Fig. All.42
shews a 3 jay truss witt no 11ag~na: ~e~ber in
the center bay. The jending ~cments on tte
truss members will be dete~1ned rJ~ the truss
 ______________________•••••••••••..,RIlIR.'.·t!liiilll

ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES All. 21

loading as shown In the figure. Then the distribution factors are: -

.,
10 10 10 To Member cb =4/7 - .57
a 'c ld
~d
Gi ® .:zJ -r I
To Member cr = 1/7 = .14

CD CD CD CD @ 60" To Member cd = 2/7 = .29

3.5 @ @
e -'- In this problem there are no loads applied
g
to the members between joints. The external
shear load on the truss to the right of the
1_ 60" --+- 60" --+- 60" -1 center truss panel which equals 10 + 10 = 20
must be carried through the center panel by the
truss chord ~embers in bending. This bending
Fig. All. 42 causes the center truss to sNay or deflect
until a resisting shear :orce equal to 20 is
SOLUTICN: developed.
The relative moment of inertia values We will assume the truss center panel is
for each member are given in the circles on fixed at joints b, c, f, and g. The right end
the truss in Fig. All.42. of this assumed fixed ended truss will be given
an upward deflection. This deflection will
The distribution factor to each member cause fixed end moments in members bc and gf
at each JOint IS then computed and equalS which are proportional to I/L 8 for each ~ember.
KIZK. The values are recorded in t ..8 0 Since L is the same for each member, the !ixed
on each member in Fig. All.43. Th; stiffness end moments will be proportional to I of the
factor K Is proportional to IlL frr the member. member.
For example for joint c; -
+ -
1 =~-
7
60 SO
~ ~
100 F.E.M. F.E.M. 100
-28.5 ~-57 -295.9~~~-14.5
11.8
-13.6 -27.2 -27.2 -13.6
9.8 4.9 4.0 8.0 4.1 2.0
--=T:"4 -2.8 -2.8 1.4 -0.9 .r.s
LO o:s ----0:5 0.9 Q.:.i 0.2
~4. 4: -0.3 ~ 36.9 -19. 4 .:.2.1.
-24.9 45. 7 -2.6
b

~~
-7.1
de
3,0
-3.4 0.5
g!1 -0.8 -0. 5
-3.6 -0.4 """0:0
-6.7 -0 1 ~
ha -0.4 -11. 8 3.0
~ -0.9 <s!
1.2 o 1.5
0.3 -0.1 Q..2.
'.3 -TI.":'T i -0.3
Q.Jl
I
0.0
" 16 I '~. !' 13 I ~
I 2.0

~
-12.2 -24.4
10.3 5. 1
-2.6 -U
1.1 0.6
~ ---l!...i
-22.3 Fig. All. 43
 All. 22 THE MOMENT DISTRIBUTION METHOD
Therefore the fixed end mcments on ~ember
gt will be 75 percent of those on member bc.
Assume 100 as the fixed end moments on
be. Then a 75 x 10 : 75 1s the accompanying
fixed end moment on gr.
w~ now remove the imaginary fixed supports
on the center truss and let them rotate to
equilibrium by the moment distribution process
as shown in detail in Fig. All.43.
The first step is to record the assumed
fixed end moments of 100 at each end ot ~ember
be and 75 at each end at member gt with due
regard for Sign. The moment distribution pro-
cess will be started at jOint C. The unbal-
anced moment is 100 or -100 is necessary :or
static balance. Using the distribution fac-
tors on joint c, we find -57 goes to Cb, -29
to cd and -14 to cf. Short lines are drawn
below each at these numbers. Fifty percent of
these balancing moments are carried over to
the tar end of each of these members. This
process is repeated at each jOint ~ntil the
remaining balancing moments are negli~ible.
In Fig. All.43 the order of jOint balance was
ctdebghaCfdebghacfdebg. If the student will
tollow this order he should be able to check
the figures in Fig. All.43.
~ig. All.44 shows
free bodies of the truss
members bc and gf with
• end moments from Fig.
60"
AIl.43. The shear re-
action at ends c and r -L
I b
by statics equals
He + Ht :II 36.9 + 45.1
45:7
60
Fig. A1L44
+ 28.7+34:=
60
1.377 + 1.045 = 2.422 lb.
The external truss shear at line cf 20 lb.
Therefore it will take 20/2.422 = 8.25 times
the final bending moments as found in Fig.
All.43 to develop a bending shear reaction
of 20 lb. Thus the tinal moments are 8.25
times those in Fig. All.43.
The solution as given neglects the effect
at axial loads upon the value of the stiffness
and carryover factors. Art. All.l2 explains
how to include these effects.
All.12 Effect of Axial Load on Moment Distribution a tension axial load diagram C becomes a ~ega­
In the preViOUS articles the effect at
axial loads upon the member end moments of a
continuous structure was neglected. The axial
loads produce bending in the members of a con-
tinuous structure in two ways. (1) The short- I loads. It i s evident that the moment diagram.
ening or lengthening of the various members due I due to P will also effect the ~gnitude of any
to axial loads ~rcduces :r~~sl~tiQn of the
joints of the structure. Since the ar~les be-
tween the members at a joint re~aln the same
due to continuity, this translation of joinGs
bends the members between joints. (2) ~~e
bending ~omen:s In ~he ~e~o~rs d~e to exte~nal
joint or lateral loads or those ~ue to ~oin~
rotation produce lateral deflection of the
members Jetween joints. The member axial loads
times these lateral deflections produce
seconda~J ~oments. These secondary ~ament3 can
be r~ndled by the general method of mcment
distribution however the stiffness and :arry
over factors and the fixed end ~o~ents are ~ct
constant but become nL~ctions of the axial
loads.
Fig. All.45 shows a prismatic beam Sl~ply
supported at A and fixed at B, Nith a moment
MA applied at A and carrying an axial compres-
sive load? Sub. Figure a, band c show the
3 parts which make up the ~cment jiagr~ on the
beam. Without the axial load P the pertien (c)
would be omitted.
28.7
Hf~f
.3
34
MOMENT DIAGRAMS
Fig. A'll. 45
By definition the stiffness factor of beam
=
AB is measured by tne necessary moment MA to
pr-oduce unit rotation (9=1) tor elastic Lme at
point A. By the moment area pr~nciples, the
slope at A equals the shear at A d~e :0 the
~oment diagrams acting as a load en a slnply
supported beam at A and B. ror a given unit
slope "at A tt is obvious that the r-eq. i r-ec MA
is less when the load P is acting as it pro-
duces a 20ffient load i~ (Fig. c) of the same sign
as the-~oment load due to MA' Thus the stitt-
ness at AB is less when carrying an axial com-
preSSive load compared to that wltho~t it. For
tive moment diagram and thus MA must become
larger to give ~it slope at A. Or in other
words a ~ember carrying a tensile axial load
has a greater stiff~ess tr~n one without axial
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES All. 23

rixed end ~oments and also the carry over fac-

tor.
, ,
•• 1
All.l3 Fixed End Yoments. Stiffness and Carry
over Factors for Beam Columns of Constant
,
, i
,

,
i i i
,
,

,
Section ! ! , i ,I :
, ,
r,
• , I i,
,
! i ! !
!
In deriving expressions for fixed end mom-
ents, stl::ness and carry over iactors the beam i
,
! I :r-, e_ - , !
coltwh~ to~ulas ot chapter AIO, must be used
, I I I I 1
i I i !
, , ,
1

in finding slopes, deflection and moments.

Mr. B. W. J~es in an excellent thesis at ,., i I I I A I
Stanford University and later published by N.A. : ! 1
I Ii ! I , , :
! , , I i
C.A. as Technical Note #534 has derived expres- I/ , , 1

I , V
0 ,
sions tor these factors and has provided graphs i ,
, ,pb- I
,
A<" I
I
of these factors for use in routine design. , ,
~~ I
Figures All.~6 to All.56 inclusive are taken •, Ii
,
i /, , I'"
:rom this thesis. The use of these Figures can / I I
, , , , i
best be illustrated by the solution of several
pro b.Lems •
.. .:.---- ,
/1
I
I I
!
,

i
i
,
,

,
,

,
i
I ,
!
j

All.l4 Illustrative Problems ..I i r- ---l I , , >t__ I j ,

i ,
i , ,
,
:......+-..' 1

, ,i
Example Problem #12
., I ! I i~ 1
,
! I
F~g. AII.57 shows the same continuous beam ,, ,

as us ed in Example Problem #4 (and solution #2). I , I I I i j 1

,
,
! I i
,

For this example it has been assumed tr~t such I ! ,! I i I I I

axial compression loads exist in spans Be and CD
0 l • • • ~l •
as ~o make the term L/j =2.5 and 2.0 for these carry-over factor. C
spans respectively, where j = v~I/P. Due to the Fig. All. 46
axial loads, new values at stittness and carry
over moments as well as fixed end moments ~ust
be determined as follows.
Span BC ,
, ,
'---'---'-1
= t = 9~ = .0104 , I I i
1
, ' , I
K I ,
,
i I
,
i i Vi
r,
• , ,

From Fig. All.47 when L/j = 2.5, correction fac- I I I ,

I /I ~/ ,

,,
,

tor for stiffness factor = .775 when tar end is I I I r_I..o.'. tu' _ 'tl_., '/ I
fixed and .36 when far end is pinned. Hence,
r. •
,
, ,, ,/ I
,
I , , , ,

y ,, , !

KBC = .0104 X .775 = .00B06

, ,
,
I
·, \ 1 i
r.
, ,
KcB = .0104 x .36 = .00374 i I I I ,
)/ I 1 I
! ~ Vt_lo:... III"':" i
Frem Fig. All.46, when L/J = 2.5 the carry over
fU _
i, ,I
!
, , V, VI , I
taetors are Cac = .73 and CCE = a it B is con- i I !
, , I
sidered freely su~ported.
r, 0
,
...--1 Vi I , ! I 1

, ,
,
NY j I
,

From Fig. All.48 when L/j = 2.5 fixed end mom-

,,
1

",-<:: 'A''''
, , ,
,
i i ,

ents ~or unito~ load = wL~/lO.67 = 1 x 96al i I

10.67 = 865 in. lb. \' 1
,
,

Span CD .. ,
"C ! \
I 'i \i·-e_.... l_. fU . . . IU"

K = IlL = 1/72 = .0139, L/J = 2.0 : \ ,

?rom Fig. All.47 correction factor = .86, hence

KeD = KDC = .86 x .0139 = .01198. carryover
factor from Fig. All.46 = .62. Fixed end moment~
from F'ig. All.48 = WL"'/l1.2 = 1 x 72"'/11.2 = 462 I
The dlst~ibutlon factor at joint C equals II

(.00374/.00374 + .01198) = .238 to CB and the

~eoainder .762 to CD.
tlJ'-
The balancing and carT:! over process is I Stiffness factor coefficient. K = C I
[
s imi Laz- to tnat in zxamrne Problem #4. Fig. All. 47

I ",
/
I~',
'r
All. 24 THE MOMENT DISTRIBUTION METHOD

, ,
,·fJ
, ,
i ,
,

",
_,1"II./I.a. ,
!",-I 'Io~
,

w
! iA i ,

!
,
! ,, 1 , V , ,
,

R
,
, I,
I ... • - .. ·,t I ,
,
,
,

u
i ,;I ,
I ,
!
i
i ! ! i r : rtr
" I, y-t··- r " ,
I
,
1Ira
:
, ,

, '\!
I 1 Y ,
, , ! 1 ! I ! '\
" ,
I A I
I
,
! 1
! ! : 1\ ,
,
1

r1
1

,
, V , 1
! 1 I I, ,
r ,
\i ,
,
u
, , , !

'" ! , , ! 1
~I! \
, , 1

u
I ! ,
,
,
, , . ,

i ,,' f . . . ..1... , ! , I

! '\J I I
! ! ,I 1 ! ;-," I
-'I-'-1--'HH-i---'+--+--+_--+--'
"
I
,

I : i\
i : ,
,
! \
" 1 I ,
"
\,
,I '\ I , I I 1 I
• ,
!\
, I
! I ! '\

I I
\i
.u ,
,
! i I

,
,
\ 1
\ I
,
,
, '"
Fixed-end moment coefficient.
", Fixed-end ~oment cceffic1ent.
uni rora Load. UnifOrmly varying load.
Axial .compression.
Fig. All. 48 Fig. All. 49

.. ~ .... _~ i
,
"
'. (" I ~Ii I ~/2
\") .., ,
I
~
~
, . ,

! ' i
, :
1
: ,
, : " • -"to • ~
T_ _ /
m
, ,I
I
"
I , ,
! i ! ,
,
! I ;~ I , , :
'Y
! " i

" ,I 1 ! , U , ,
,/
. ,

I i I I I, /1 ,
1
', , ,

..
,
I ,I i 1 V ,
,
, I ! ! ,
J ,

V, , I ! /1 I
I I 1 , 1
,

•
i I
, ! 0.
,
I : 1
,
, >,, ! I ! , ,
,

'.• I ! 1/1 ! i I : Y, ,
! ,
,

,
1

~ll
, , ,

'. ,I / i
•
,
: !
!
~

I v, 1 , ,
1
! i :~I I
,
I
,

1
I 1
,

j--{' ! ~.~--
,

»
I I ! i i ,
r
, I 1 : i
I ,
~;
I I ~ !
,
I ,
I ,
,, \, ,
"

, Y
,
,
! !
·
,
I ' , ! I 1 i , 1 ,
i
• ,
, ....--1
, ,
I
,
I ,
I 1 ! , ,
\ I
I ..:--
, , , ,
i i , ,.llJ I : '{
·
, , : !
• i I
..J...--+- 1
! I
I

i i !\
,
1
,
,
.
,
I
,! I , I I ! \
,

·
, ! , , ,
! , ,
w , 1 1
• • ,! • , ,
'" coefficient.
c
• a
'I' coefficient. •
Fixed-end moment ?lxed-end ~oment
Uniformly vary'ing load. Concentrated load at ~id-span.
Axial tension.
Fig, All. 50 Fig. All. Sl
 . --.-.-.------- IIIIIII!!l!!!!~
ze=

ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES All. 25

2.0

f-I.l.l~ a 1"0 ~~ I -,
r.e I- \~ 1-+-l-+7'f'=--f-+--l--t---+-t---+-t---j-+--1
!.fA, .. WI;
'b 2
D..L) Col
"
v I \
I-
4.0

I
J.5

i
.,
2.0~

I 1/ j .. 1 ~ a compreu1on--..

LI.i '" 1. 0 tendon ---'.

i - 2.0

. JiO
4.0 !

!
.1
6.0 .......
I
.J
t- I
.. .5
alt or bit
'.0 I
., •• .9

Fig. All. 52 Flxed-end moment coefficIent. concentrated load.

I I
I I

I I I l.~(~~ I

I I I 1:.-- i
! ' 'I 'N.
I !

! ;

1/'f'_ _
V 1/1// i
., ., . . . -. . . '''01_' !If/"
~
(/'1
1/1
.
'. ": ":
, ": 0
iii Y YI!
. I' ! I iI Ii! i I
i I I' I : I
II !II
,
•
I
, ,
,
I Ii ,
Iii Ii :
,
;1 ,

~~Irl-'--;",,,1_,,';'-'f;p,.-,,,,--:.C;,,Tr.. ~
I
I
,,
.1 ,, I
,
I : I ii I, ,i !
I' /' F l.2L._
r -r-
= -,." " , : : , , "

.1,
,
:I I
- - -"' II
. I ..... 1

,~
I , I •,
,.,:
,
Ii i II II II '! I
I : i

i~·L-.~
, i
, I
-e
'.• ,
, i
I, I I
,
i \ \ i , II ij I iI
~ ~. ,I ,, ! ,, I i Ii i
! i I

\l-~i, .,
_.-
,
. ,
\ I
; ... ."
L~ ... ! ;1
:
,
'; , ,i I ',.1 'II., , I I I ill,,
, Fig. All. 53
' '."" \ ,
\ , , ,
\ v \i '\ 'I~ I I / II
, : ,
•

,
.• " ••
"'1 , II
,
I
 All 26 THE MOMENT DISTRIBUTION METHOD

IXrnr&d~IFB
1
1.a
ri-f ~
,
i

' I:
I
,
,
,
!
I

I
I i ! III, I <;.:......eO" B' lO~" A' 84" A lOOt~60" C
j ~~. I w.3~1 w:32 ~16
L
I,
I
I I I i
i::l::O 1 ttl tIl t 1 CIT' 1 r~~
2
J
, I
.. ,
I I
! !
,i
I
,
~ 7~ 1..@j.@;2JJll.. .:L..iQ...15.J.Q:6Jg:2 ~ T FB

i I
I
I
I
1 1
,
i
! ! 1
,
C'
Fig. AI1.58
B' '~....
.. ,
............ B ,.. '
A' '....cabana /A
',strut~ "
,,,'B
v....
C

IT' ,:.,.,
a, .. " c:,....
I I ! I I ,.q_lIl_ Spruce . . ,__ uael. /:.~~\.
! i
,
II I
i

I
I
I
T
, I ,
I
r,
, i
i 6" 6"
Boam Loads
", 1 I
1 [;U75BJ -~ ~
1

• r.
i ,% I
I I 1
Y II 1=17.21 1=22 A' A B
./'! I' I I I Span·AB, A'B' Span4AA'
,. ./ ~ i : ] i I
Fig. All. 59
r-c, I i I I I
Fig. All.59 shows the total beam axial loads in
NI I I the various portions of the beam by adding the
, I
values shown in Fig. All.S8. The outer span AB
r-, I due to the axial loads 1nduced by drag wires ot
I ! ~
drag truss 1s subjected to a 'mrJing load. As is
customary in design of such wing beams the ax1a1
1

I 'I I' 1,,1 I load in span AB will be taken ~s equal to the

! I I I: I I::> average load or (-9571-8565-7559)/3=-8565#.
o 1 '~/J~ ~ •• The be~ bending ~omer.ts at the support
Culum..'"l dtstrlbutlon factor. 2~ - a. pOints will be determined by the ~oment distribu-
tion method.
Fig. All. 56
Calculation ot Factors;-
"&1*11n,\ L/j.2.5 lis Cllnata..llt, L/J=2.0
Span AB
I = 17.21, L = 100, IlL = .172
B c D E E ' 1300,000, p, 8565
!'1e·All.57
~51.5"-!---"-----+ __ 72"_ j ,~,1jl300,000 x 17.21 , 51.1
StiU. Fact. II: e .00808 "00374 .01198 .01198 _ P 8565
D b<ttiOll:le aa 76 o
c.O. F1I.eter o .73 c .52 52 L/j , 100/51.1 ' 1.96
F1ud End. Mom.
e Joint B
c.o. rc c
.883
c
lI6S
18
.865
." -4<' From Fig. All.47 when L/j = 1.96, correction
13.2
Ba.1. Jo1Dt C
Car -ceer-
146 0 • a factor = .86 for fixed far.end and .52 tor pinn
Ba1aJ:lee D ed tar end. Theretore
CAl' -cver-
FInU Mom. _88 _7 KAB ' .52 X .172 ' .0895
KBA ' .86 X .172 ' .148
Comparing these results with those of Problem #3 From Fig. A11.46, when L/j = 1.96, the carryover
we find moment at C is increased 15.7 percent fa.ctors are:-
and that at D is decreased 8.5 percent. For
larger values ot Llj the difference would be C.O' BA = .62, and zero from A to B since 5
greater. is considered in its true state or freely sup-
ported. From Fig. All.48 when L/j = 1.96 equa-
Example Problem #13 tion tor fixed end moment for uniform load =
Fig. All.58 shows the upper wing ot a b1- WL 2 / 11 . 25 = 32 x 100:1/11.25 = 28420"#.
plane. The wing beams are continuous over 3 The overhand moment Mac = 60 x 26.6 x
spans. The distributed air loads on the tront (32+16)/2' 38300"#.
beam are shown in the figure, also the axial Center Span A.A' :-
loads on tront beam induced by the lift and drag I ' 22, IlL ' 22/84 ' .262, P , - 11758
truss. The bay sections or ~he spruce are shown ~,300,OOO x 22 = 49.4
J" ='V"
in Fig. All.58. The moment of inertia in each 11758
span will be assumed constant, neglec~lng the
influence or tapered filler blocks at str~t L/j , 84/49.4 ' 1.70
points. KAA I =KA'A .262 x .895 .234
C.O. Factor .585

&
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES All 27

Fixed end moment = wL a/11.35 = 31 x §4a/l1.35 = For member Be

19300"# p = - 27150, IlL = .0083, I = .332
The moment distribution process is given j = e9,00~7~~~X .330//2 =18.82, L/j =184.°82
in Fig. All.60. It the axial loads were neglec-
ted, the bending ~oment, at support A and A' = 2.12
would be 19480, thus the axial influence in- From Fig. All.47, stiffness factor = .32 x .0083
creases the ~oment at A approximately 7.5 per- = .00681
cent. From Fig. Al1.46 C.O•. Factor = .66
...~ Ie .... n .~31 ._16
For Member AC
FlJ. an.eo ,
p = 10400#, IlL = .0029, I = .137
C' B'
IA' A

j ;;V29,OOO,000x .137 L 47.20 =

c .148 .234 .caas
10400 = 19.58, 1 =19.58
uu 2.42
a a, .52 0
·1 2 420 _2B420 383 St1tt'ness factor = .0029 x l.18 = .00342
·~880 0
,.e125 C .0. factor = .39
·:lle1 .828
Fig. All.53 shows the moment distribution solu-
1265
_918 -341

,'0 tion which includes the effect of axial loads on

saa
.3901·148
o no joint rotation. Comparing the results in Figs.
_l85
·38300 38300 ·20909 2090S1 .38300 38300 Al1.62 and 63, the moment MeA of 24050 is 29 per
cent larger than that in Fig. 52, and the moment
Exa~ple P:cble~ #14
at B is 18 percent larger. The effect on the ax
ial loads or these new final moments will be
?lg. All.51 shews a triangular t~lSS com- quite small, and thus further revision is un-
posed of two ~embers fixed at A and B and rigid- necessary.
ly joined at C to the axle bar. Let it be re-
quired to determine the end moments on the two
~embers considering the eftect or axial loads on
joint rotation and translation.

Et'fect on End Moments Due to Translation of

Joint C Due to Axial Loads.
The ~ovement of joint C normal to each
member will be calculated by virtual work.
solution:- (Reference Chapter A7). Fig. AII.54 shows the
virtual loading of 1# normal to each ~ember at C
The ~gnitude of the axial loads in the The Table shows the calculation of the normal de
members is influenced by the unknown restraining tlections at C.
~oments at A and B. To obtain a close approxi-
mation or the axial loads, the end moments in Table
the ~NO members will be deter.nined Without con_
sideration of axial loads. ~hus the external
joint ~ament of 4 x 18000 = 72000 In. lb. at C
1 L/A.! a ", I !,z.L1-.
-n- I
I I 3~bL. I
All: i
, I ""
,C 1.89 ! ,0962 I 1.5
is distributed be~Neen the two ~embers as shown
in Fig. All.52. With the ~ember end ~oments I
.00000"9
sc I, .000002"" , ''''00 I _1.60 i' .202
1060 -1.89 .1250 I
.0816
_~'71~
.207
known the axial loads and shear reactions at A
and B can be found by statics. ~he resulting Thus the deflection of joint C normal to
axial loads are I BC equals .202" in the d ir-ect i on assumed tor
PAC = 10400# and PSC = - 27150# and the the unit load, and the deflection of C normal to
shear reactions, SA = 590# and SB = 2000#. AC = .20'7"
~ith the axial loads known :he ~odifteG
beam factors can be deternined.

I"", "
f "J ~::..'
All. 28 THE MOMENT DISTRIBUTION METHOD

All. 15 Secondary Bending .Moments in Trusses with

Rigid Joints.
B A B A
Often in alrpla.~e structural design tr~sses
with rigid joints are used. Rigid joi~ts ~re
produced because of welding the ~embers together
at the truss joints or by the use of gusset
plates bolted or riveted to each member at a
truss jOint. when a truss bends l1nder loading
Fig. All. 64 the truss joints undergo 11ffer9nt amounts of
movement or deflection. SinCE, the truss member-s
c are held rigidly at the jctnt, any joint dis-
placement will tend to bend :~e truss ~ember8.
The fixed end moments due to support de- The bending moments prOduced in the truss mem-
flection equals M = 6Eld/L a as derived in Art. bers due to the truss joint (e!lectlons are
All.? However, the secondarY moments due to generally referred to as secondary moments and
axial-loads times lateral deflection modity thIs the stresses produced by these moments as
equation. ~James~ has shown that the modified secondarJ stresses.
equation is,
Since fatigue strength 15 becoming ~ore
d
i t K'L! and R' L
!'lt1xed end = L" 6Eld
(2~-a)' ' than important In aircraft structural design, the
question or secondary stresses beca.nes of more
Unportance than in the part. The moment distri-
!'l ' 6Kffi
2{l-a , bution method prOVides a Simple and rapid methoc
for determining these secondary moments in truss
Fig. All.56 shows a plot ot 2~-a against L/j. members due to truss deflection. T~e general
For Member scr- procedure would be as follows: -
R' .202/40' .00505 K' IlL' .0083,
(1) Find the horiZontal and vertical displace-
2~--<t ' 1. 08ror LI J ' 2.12. ments of each truss joint due to the criti-
Hence, tixed end moments due to translation or ~al des~gn condition. (See Chapter A7 for
JOint C equals- methods of !inding truss deflections).
!'lBC , !'loB ' 6x .0083x .~~X29,OOO,OOO,-6750"#
(2) Fram these displacements, the transverse
deflection of one end of a member With
For Member AC
respect to the other can be found fer each
R' .207/47,2' ,00438 K= ,0029, 2~--<t' .92 truss member.
when
LlJ = 2.42 and member is in tension. (3) Compute the fixed end moments on each member
hence due to these transverse displacements .
!'lAC '1'lo.l,6X .00438 x .0029x29,000,OOO
•92
(4) Calculate stiftness and carry-over !actors
-2400"# for each truss member.
The signs are minus because inspection shows
that the moments tend to rotate ends at members (5) Calculate distribution factors ~or each
counter clockwise. (Ret. Art. All.2) truss joint.
Fig. All.65 shows the moment distribution
tor these moments. The magnitude ot the moment carry out the ~oment distribution process
at B is 6.7 percent ot that in Fig. AlI.53 and to find the secondary moments at the enes
10.4 percent at Joint C, however, it is reliev- at each member.
ing in this example.
(7) Calculate the S~resses due to these second-
ary moments and cambine with the pr1~ry
axial stresses in the tr~s members ~ue to
truss action with pinned truss joi~ts.
All. 16 Structures with Curved Members.

T:18 moment d.i s tr-i but i on me thod can 88 ap-

plied to continuous str~ctures which :'..~cl~de
cur-ved as -,','ell as scr-a rmt nenoer-s . :':::e
equations :or fl~jlng the st~:~ness iac~8rs,
carryover factors and fixe~ ~nd ~e~2nts ior
Fig. All. 65 stTa~~ht ~2~bers canr.ot je ~sed ior 2UTVed ne~­
jers. ~he ~lastic center ~ethod as ~resented
 we em

ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES All. 29

in Chapter A9 provides a rapid and Simple diagram. Thus to find the distance fram E x
method for deternining these values for curved to this centroid we take moments of the MlEI
members. The use or the elastic center method diagram about B and diVide by ~ the total area
in deter.nining the value of stiffness and at the MlEI utagran, Thus
carryover factors will now be explained.
X : -2( .667L) + l( .333L) :..:...1..= L
All. 17 Structures with Curved Members. - 2 + 1 - 1

Before considering a curved member a Thus the centroid of the total ~oment weight ~
which equals one lies at paint A or a distance
strai&~t member or constant EI will be con-
Sidered. Fig. All.66 shows a beam treely L to left of B.

In USing the elastic center method to de-

Fig. All. 66 # _ ,I- .d,= 1 Had.
termine the stiftness and carry-aver factor tor
a straight beam, we aSSume that the bending
moment curve due to a moment applied at A is of
n
MA 5
..
~ -~ .-. -,~
•
deflected beam
M such magnitude as to turn the end A through an
~ A EI COnstant ' ,~ B angle of 1 radian. As shown above, the moment
L I weight ~ for this loading is unity or 1 and its
centroid location is at A. Then by the elastic
center method we tind the moment required to
turn end A back to zero rotation. The value of
this mOment at A will then equal the stiffness
M diagram
factor of the beam AB. In order to simplify
the equations for the redundant forces the
Fig. All. 68 M/EI Diagram elastic center method refers them to the elastic
center. From Chapter A9 the equations for the
4
-1:" redundant forces at the elastic center for a
structure symmetrical about one axis are; -

11 ::; -1: 0s - - - - - - - - - - - (1)

0 Z ds/EI

supported at end A and fixed at end B. A - - - - - - - - - - - - - (2)

~oment MA 1s applied at end A of such ~agni­
tude as to turn end A through a unit ar~le ot
one radian as illustrated tn F~g. AlI.66. By - - - - - - - - - - - - - (3)
Qefinition, the necessary moment MA to cause
thIs unit rotation at A is referred to as the
stiffness of the beam AB. In Art. AIl.4 it Fig. AIl.59 shows beam of Fig. All.56 re-
was shown that thi s required moment was equal ~laced by a beam with the reaction at end A
in magnitude to -4EI/L. It was also proved replaced by a rigid bracket terminating at
that this ~oment at A produced a noment at point (0) the elastic center of the beam, which
the fixed end a of 2EI/L or a moment of one due to symmetry of the beam lies at the mid-
hal~ the magnitUde and of opposite sign to point of the beam. The elastic mcment loading
t~at at A. Fig. All.6? shows the bending is ~s ::; 1 and its location is at A as shown in
~cment d~agrs~ Nhich causes ene radian rota- Fig. All-59.
:~on of end A. ?ig. All.58 shows the M/EI
c tagr'am, wn..Len equals 'the moment diagram ~ • 1 ! L ------l
1-0---- - 1
jivi~ed by SI ~hich has been assw~ed constant. <,' 2 'Mo Or
The total moment we Ight ¢ as exp Ia Ined A~ B
in Chapter A9 equals the ~rea of the MIS I
diagram. Thus 0 for the MIS I dlagr~~ in
! Y L---: o 1

f~g. All.58 equa:s,

Fig. All. 69

Solving fer redundants at (0) by equations (1),

(2) and (3).
In other words the total elastic moment
~elght ¢ equalS one or unity. 'L = - 09 -1 - _ SIlL -
"'"0 Z dS/;r a LiEr -
The location of this total moment weight
o will coincide with the centroid of the M/SI

,i ",, ,
v'./
All. 30 THE MOMENT DISTRIBUTION METHOD

Is removed and end A is cor~ected by a rilid

bracket te~inating at the elastiC center of
the structare. ~e will new find the required
forces at (0) to cancel the unit rotation ~t A
Xo a Z ~S Y = 0, because y the vertical which 'NaS assumed in Fig. All.71.
Ix
distance of ~s load to x axis through elastic
center (0) is zero. ,,Y
Fig. All.70 shows these forces acting at
the elastic center. ,0'= 1
XD1~\Xt -: Yo \ yo -d. O.6366r
5 ".... . 'J 'B
..J...!: -L-
I
EI/L A 'y
A ~rl6E~I/-L-2---{' " Fig. All. 73

Fig. AII.70
The bending moment at end A equals
By definition the stiffness factor is the
moment at A which is yeQUlred to turn end A
through an angle of I raddan , Thus 42I/L
Is the stiffness factor and thIs result
checks the value as previously derived In
Art. Al1.4.
The bending moment at B in Fig. All.70
equals,
The total area of the M/sr curve ~or the
curve in Fig. All.72 it calc~lated ~rould equal
one or unity as explained in detail for a
straight ~ember. The centroid of this M/EI
diagram would if calculated fall at 901nt A.
Thus in Fig. All.73 we 2.?ply a unit ~s load at
A and find the redundant force at O. Due to
symmetry of structure about a vertical or y
axis the elastic center lies on ~h~5 3ymnet~ical
axis. The vertical distance ITom :~se ~:ne AB
to elastic center equals Y= .6366~. (See
page A3.4 ot Chapter A3).
The elastic moments ot inert~a Ix and 1y
can be calculated or taken ~r0m reference
sources such as the table on ~age A3.~.

_ M 5EI (.b.) _ 2EI :"thence,

- - L + La 2 - L
Ix = .2978r"t J but t = SII
Hence the carry-over factor from A to B
Is .5 and the carry-over moment is ot opposite .2978r"
Hence Ix· SI
Sign to that of the moment at A.

All.18 Stiffness and Carry-Over Factors For Curved

Members.

Fig. All.71 shows a curved member, namely, Solving the equations for the redundanta
a halt circular arc of constant EI cross- at (O)J remembering ~s = 1 and located at pOint
section. The end B Is fixed and the end A is A.
freely supported. A moment MA Is applied at
A of such magnitude as to cause a rotation at = .=.....ill.::: - E: I
A at 1 radian as illUBtrated in the Fig. Me - - nr nr
Fig. All.72 shows the general shape of the ZI
bending moment curve which IS statically in-
determinate. In FIg. All.73 the support at A V
_
- Z¢sx _ - (l)(-r) _ 2EI
·0 -
Iy - nr.3 - nr"
ZSI
= Z¢sy = (l)(-.6366r)
Xo Ix O.2g7e~ 3 = -
SI
Fig. All.74 shows these forces acting at
the elastic center. T~e Jendi~g ~oment at A
equals the stiffness factor ~or the curved Jearn
Fig. All. 71 Fig. All. 72 :1xed at rar end 3 and :reely supported at near
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES All. 31

end A. The ratio of- t~e :ending ~o~er.t at The method of hew to handle this re~aining
point B to that redundant force can best be explained by pre-
at point A 'Nil:' senting some example problem solutions.
cive the carry-
Jver fac:or. All. 19 Example Problems. Continuous Structures
Involving Curved Members.
Bend.tng
:!loment at A, Ex~~ple Problem 1
Fig. All. 74 B
Fig. All.76 shows I =1
MA = nr - 2.14 (0.6366r) -
-SI;'I
;'3 (281)r
nr " a frame consisting of
both straight and
= -0.318EI 1.:;81 _ 0.536 SrI: 2.314- :::1 curved members. Al-
r r though simplified L::r60"
relative to Shape, I •2
Send~ng moment at 3, this frame is somewhat
representative of a
:13 ::: _-,,0-,-.3,,',,""8",E",I, 1.36EI + O.636EI fuselage frame with twa L::80" L=80"
1" 042 E1
r r • r cross members, one be- 1=1 I =1
tween A and C to support
T~er9fore the stiffness fac:or for a half- installations above
circ~lar arch of constant EI is 2.314 EliI'. cabin ceiling and the
other between F and 0 w: SOl/in.
The carry-over factor equals the ratio of to support the cabin I I
Mg to ~A or (-1.042 El/r)/(-2.314- EliI') = floor loads. The frame
L=60"
C.~2. It sr.ould be noticed that the carry- supporting forces are 1" 6
over mo~ent has the same sign as the applied assumed provided by
~cment at A as compared to the opposite Sign the fuselage skin as I :: 2
~or straight members. In other wor1s, there shown by the a~ows
are t~o points of inflection in the elastic on the side members. E
curve for the curved arch as compared to one Eccentricity of these
for the straight member. skin supporting forces Fig. All. 76
relative to neutral
axis of frame member is neglected in this stmpl!-
tied example problem, since the main ~urpose of
The fixed end moments on a curved member this example problem is to illustrate the appli-
for any external loading can be dete~ined cation of the moment distribution ~ethod to
q~ite rapidly jy the elastic center ~ethod as solVing continuous structures involving curved
11:~strateG ir. Cha~ter A9 and thus the ex- members.
?lanaticn will not be repeated here.
SOLUTION
The student should realize or understand
that ~hen the end moments on a straight member Due to symmetry of structure and loading,
in a ~cntinuous structure are found tr~m the no translation of the frame jOints takes place
~c~ent cistrijution ~rocess. the remaining due to frame sidesway.
end f~r:es are statically dete~inate, whereas
for a c~~~ed ~emjer· in a continuous structure, The ~rame cross members AC and FD prevent
Lr.cwing the end ~c~ents does not ~ke the horizontal movement of joints A, C, F, and 0
curvet ~e~ber statically ~eter.n~r.ate, since due to bending of the two arches. .~y horizontal
',':8 have six unknowns at the two supports as movement of these joints due to axial deforma-
il:~strated in Fig. All.75 and only 3 equations tion is usually of minor importance relative to
~~ static equili- caUSing bending of frame members. Therefore it
brl·~. ~ven when can be assumed that the frame joints suffer
tne end mcment s Fig. All. 7S rotation only and therefore the ~oment distri-
are de t erninec bution method 1s directly appllC~ble.
t: om the ncment
Calculation of stiffness (K) values for
pr-oc ess there each member of ~~a~e: -
still r-enatna
one unknown, Upper curved memJer ABC: -
n~ely the hori-
zcntaa reaction K,il,Be::: K CEA::: 2.314 E_I. mh'..e va 1ue was
,_~,
at one of the
beam encs , der:ved in the ~revlous Art. AII.IS.
All. 32 THE MOMENT DISTRIBUTION METHOD

SUbstituting: - dis~ributa1 nome~t.

2 "2~' l
K, ..."d.... X - = .0770 (E is con-
,;: K, ;:
ABC 30
:BA
s~ant and therefore omitted since only ~elatlve 'The curved :r.enbe:-s :::r' the .:'ra:J.e have ::c
values are needed for the K vaLues v ) applied external lcadin~s ~ence the .:'~xed end
caments on t~e 2urved :r.emjers is zero.
~he relative moment o! i~ertia of the
cross-section ot eac~ fr~e ~ember is given For ~embe~ AC, :lxeC end ~cment equals
on Fig. All. 76. WL 2/12 = 10 x 60 2/12 = 3000 in. lb. and tor
member F'D = 50 x 60 11/12 ;: 1.5000 in .lb •
K,.., =2.314x2=.1540
=X,.,
,'~D 30wEii' Since the supporting skl~ .:'arces on the
side members have jean ass~~ed ac~1::g along
For all straight members, the sti~fness factor centerline of .:'~2.Ee membe:-s, the ~1xed end
equals 4EI/L. Hence, moments on ~e~bers AF and CD are zero.
K
AC = KCA = 4x 2/60 0.1333
Moment DistriJution Process: -

~D = KaF = =4 x 6/60 0.4000

Fig. All.7? shoWS the calc~lations in
carrying out the successive cycles of the
KAF = K..-,A • 4 x 1/80 = 0.0500
moment distribution process. Due to symmetry
KCD = Kac = a x 1/80 =0.0500
uISTRIBUTION ~ACTORS AT 2ACH JOINT: - car ry over from C ' -.452 X CI-880\
B ,
JOINT A.

~ "'I~ . ".I~ " "I~ i~

= 0.1333 + .0770 + .0500 = 0.2603
( ' O' F. . O' 452
ZK

Let D equal symbol for distribution factor.

01'"
<:;0..".
t-''''
-=I0t-,M"l"IN-I,
'I'
f " " "I
'I' 'I
"0>
,'0
;':<:0<0
-:'
\
'"' 'I

, '- A .5121 I C
D = .0770/0.2503 = 0.296
illC

D~C = .1333/0.2603 = 0.512

AC
-3000
1536
3000
-1536
CA
I
I
"d
Ut-I
«':1
-768"--- - - "

D = .05CO/0.2603 = 0.192 -122

AF ----sr--
-82
JOINT F.
----:u--
-25
ZK = .0500+0.400+0.1540 = 0.6040 --r2--
D
FA = .0500/.6040 = .082 6 -d
-2353

D = .4000/.6040 = .663 Tr-CillJ------l

FD

D'ED = .1540/.6040 = .255

F

1
.
2 ( 09750/Y
DF
-15COO 15000 !
D

Carry-Over FactorS: - --1875

2122
For the straight members the carr-J-over -106lJll'E
factor is 0.5 and the moment sign is the same 469
-234/ Fig. AIL 77
as the distributed mament when the Sign con- 105
vention adopted in this chapter is ~sed. -52/
23
For a half circular arch at constant I, -8753
the carrJ-over factor was derived in the pre-
vious article and '~s found to be 0.452. T~e
Sign or this car-r-y-over moment; was the s ame of ~r~~e ~~d l~aGi~g, the ~r~ce5s ~eec only Je
Sign as the distributed ~cment at the ~ther ca"riec c~t .:'cr one-half 0: fra~e, t~us in ?i 6-
end of the beam. However, ~sing the sl~n ~re Al:.~~, ~n:y j01~ts A ~~d ? are cor.s1de"ed
convention as adopted ~or the ~oment ~lstr:­ since the mzme r t ca I r asu.l t.s a: D anc C 'No~11 be
bution in this book, the ca~ry-over faC~Gr the sane as at )'. and F' respec c ivatv , In the
would be ~inus or of or~osi~e Sl~, :0 the :l~~re t~e c12trlj~tl:n .:'actors are shewn in
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES All. 33

the 0 at each joint. The process is started inertia about x and y axes through the elastic
by placing the fixed end moments wi~h due re- center.
gard to s~gn at the ends of members AC and FD,
n~nely, -3000 at AC, 3000 at CA, -15000 at FD Bly
and 15000 at OF. ~e now unlock Joint A and I.. 1
- 1 - ---l
i
find an unbalanced moment of -3000 which means d:; 4.6r:-T
a plus 3COO is needed for static balance. - ---.-~ I
Joint A 1s therefore ba:anced by distributing , I 19. 1"
.512 x 3000 = 1536 to AC, .296 x 3000 =
880 to
k
ABC, and .192 x 3000 = 576 to AF. Short hori- Fig. All. 78
zontal lines are then drawn under each of these
distributed values to indicate that these are
balancing moments. Carry-over moments are
immediately taken care of by carrying over to 19.1 x n (30/1) :
joint F', .5 x 576 = 288. From A to C the y = 14.4-9"
'IT x 30 60
carry-over moment would be .5 x 1536 =768 and 1
+-
2
therefore the carry-over :rom c to A would be
.5(-1536) = -768 which is recorded at A as (NOTE:- 19.1 equals distance from line AC to
shown. For the arch member ABC, the carry-over centroid of arch member ABC.)
moment from A to C would be -0.452 x 880 =
-401 (not shown) and therefore fram C to A = Calculation of moment Of inertia I : -
-0.452 x (-880) = 401 as shown at joint A in x
the figure for arch member CBA. Joint C in r.:ember ABC
the figure has been balanced once for the
purpose of helping the student understand the Ix = -1- + -rtrd
.3r~
1-
a

sign of the carry-over moments which flow to

the left side trom the right side of the trame. = .3 X13D;! n x 30
+ --1- x 4.61 a = 10100
After balancing joint A and taking care
of the carry-over moments, we Lrnaglne A as rlember AC = .§.Q.
2
X l4.49 a = 6310
fixed again and proceed to joint F where we
find an unbalanced ~oment of -15000 + 288 = Ix total = 16410
-14712, thus plus 14712 is necessary tor
balancing. The balancing distribution is .255 Total elastic weight of structure equals
x 14712 = 3750 to FED, .663 x 14712 = 9750 to Z dsll = (n x 30/1) + 60/2 = 124.24.
FD and .082 x 14712 = 1212 to FA. The carry-
over ~oments are .5 x 1212 = 606 to A, .452 x The next step 1n the solution is to draw
3750 = 1695 from D to F by way of the arch
the bending mOme~t curves on this frame portion
~ember and -.5 x 9750 = -4875 from D on ~ember
due to the given load on member AC and the end
FD. We now go back to joint A which rzs been moments as found by the moment distribution
unbalanced by the car-ry-over- moments and repeat process in Fig. All.77. It 1s composed of
the balancing and carrying-over cycle. In the three parts labeled (1) to (3) in Fig. All.79.
complete solution as given in Figure A11.77 Portions (1) and (2) are due to the end mOments
each joi~t A and F was balanced five times. and ?ortion (3) due to the distributed lateral
The :lr~l bending ~oments at the ends of the load on member AC.
members at each joint are shown beloW the
double short lines.

The arch member ABC has 3 unknown forces

at each end A and C or a total of 6 u~~owns.
~~th 3 equations ot static equilibrium avail-
able ?lus the ~own values of the end moments
at A and C as round from the accent df s'tr-t bu-.
tion process. :he arch nember is still static- 1094 ~.UCilllli~""':liCcc..:d'-l
ally indeterninate to one degree. ~hus the 2353 Fig. All. 79
horizontal reaction at A or C as provided by
the ~xlal load in oember AC ~ust be found be-
fore the bending moments on arch ABC can be The next step is to f~nd the 03 (area of
calculated. Mlr curve) for each ~ortlon and its centroid
location.
The :irst step in this problem is to find
the elastic center of the frame portion com-
;JQSed of members ABC and AC, as snown 1:1. Fig.
Al~.78, and then :in~ the elastic moments of
All. 34 THE MOMENT DISTRIBUTION METHOD

~Sl = 1094 x n x 30/1 = 103000 46 lb. The member AC also suf:e~s an ~x~al l~~~
due to the shear ~eactions at the tJ~ o~ =e~:~"2
0., = 60 x 2353/2 = 70590 FA and DC. ?ig. All.Bl shsws ~re8 te118S :: ~he
s1de member ?A and DC wi:h the en~ =~,e~t~ as
~S~ = (-.667x60x4500)/2· -90000 found in ~ig.All.77.
The shear reaction
l:0 s = 63590 at A and C can be
found by taking
Fig. All.30 shows the ~s values concen- ~oments abo~t lower
trated at their centroid loca~ions and referred ends. Thus for FD, 80"
to the x and y axes :hrough the elastic center. RA = (1259 + 1791)/30
• 38.1 lb. Likewise iF
~-.:...
~
,

,-'
jo
He = 38.1 lb. These 1791 1791
y ~eactions react on
B Fig. All. 81
cross member AC in t~e
l • 0$ = 103000 oppOSite directions
4.61"......"
! .,.II 0
'x--~--~,---x
0 I thus Giving a compression load of 35.1 lb. i~
Gl, .70590. 14.49 member AC, which ~ust be added to the tension
A' It c-L- load of 46 l~. :rom the arch reaction to cb~ain
~:: ~90000 Fig. All. 80 the fir~l load in the cross-member.

Bending Moment in Lower Arc~ Member ?~~

The frame has been imagined cut at A and the
arch end at A has been connected by a ~igid
bracket to the elastic center. The redur,dants
at the elastic center which will cancel any
relative movement of the cut faces at A can
now be computed.
The horizontal reaction on arch ?SC must
be found before :he STue bendin~ moments on ~he
arch can be found. The proc?cure is the sase
as ~or the upper arch. 2ig5. AIl.SI, 22, 63
show the results for the lower fra~e ~ortlon.

= - ~ 05 -03590
Z d8/1 • 124.24
=- 673 Lns Ib , 8753
'I· ·····\.\/592
Fy D F~ID/
Xo = Z/8 Y = [102000 (4.61) + 70590 t "~-
3
T \ _ •IE.c.
...i....X_\_ __ 3.34" 19.1
x..:.:.L:.. ';1 1 . / ,...., :::J
x (-14.49) _ 90000 (-14.49)J / -"';: I -::/ f I 22500 .
16410 =46 lb. E
Iy Fig.AIl.SI E
3 -
Fig. AIl.82
Yo = -z1¢sx ~ 0 because x ~ zero.
y = 19.1 x n(30/2) z~ = 57.1
Y n x 30 + 60 I
The bending moment at any point on ABC 26
or AC equals that due to Mo and Xc plus the
moments in Fig. All.79. Ix = .3X + 15nx3.34'1 = 4578
For example,
Z30.3
(60/6)(15.75)' = 2478
Ix = 7056
At point A on member ABC,
2rom P:g. All.82,
MA = 1094 - 670 + 46 x 14.49 = 1091 (should
-8753 x 60/6 = - 37530
be 1094 since moment as found in Fig. All.77
is correct one. Small error due to slide r~le .667 x 22500 x 60/6 = 1500GO
accuracy.
05 3 = - 5962n x (30/2) = -326000
At paint B on ~ember ABC: -
z 0s = -265500
MB = 1094-670-45xl5.51 :::II -289 in. lb.
So Iv tng for Redundant; "or-ces i"LJ anc Xo (See
The horizontal reaction at A will not :;'::'g. ).11.33)
~roduce any bending on nember AC, thus t~e
values in Fig. All.79 are the true moments. M - - Z 2iS = - (-265500) = ..',.650
'0 - Z dS/I 57.1
The axial load in ~ember AC by statics
from F:g. AIl.BO equals Xo or a ~er.sio~ load of

22£ su E
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES All. 35

x .Z0s Y: - 8753OX15 . 75+15CCOOx15 . 76- 328000 (- 3 . 34 ) Example Problem 2. External Loads on Curved
. o -x 7056 :-lembers.
X o = 294. lb.
Fig. AIl.85 shows a frame which has ex-
y ternal loads applied to the eurved member as
F I D well as ~o the straight members. The fra~e
c~ ),Mo'll4650./
1 - ' -
l~. 76
supporting torees r~ve been assumed as acting
uniformly along the side members AD and CEo
x _ - - ;Xo=294J-X--+- The problem will be to determine the bending
i 14.24 moments at frame points ABCDE.
E
Fig. All. 83
1000 B 1000
Bending moment at F on member TED, \ .,.. \.15 I

MF = -6962+4650-29'1,x15.76 =-6962 in.lb.

-- ,.)\
ME = -6962 + 4650 + 294 x 15.76 = 2340 tnv Lb , 30° '(. /1 s 2
The axial load in member FD = 294 Lb ,
A I
• • , C
w s 101b/in.
compression plus 38.1 1bs. tension due to sfiear
reaction from side members at pOints F and D I =1 1=1
or a resultant load of 255.9 lb. compression. Fig. All. 85 L-60 L=60
Fig. All.84 shows the final bending moment
~iagram on each member of the frame.
1000 lb •
00 L
• v ,- 30'
B
I : 3, L a 60
259 Member ABC
SOLUTION:-
\ Calculation of stiffness factors K.
1094 A~
2147 Member ABC. KABC = 2.314 I/r = 2.31'1,xl.5/30 =
.1158
Member AC KAC = '1,IIL =4x 2/60 = .1333
KAD ='1,I/L = 4 x 1/60 = .0667
2353
1259qA
2353 KDE = '1,I1L =4 x 3/60 = .2000
CS'259
Calculation of Distribution Factors D.
f;7
JaINI' A. ZK = .1156 + .1333 + .0667 = .3158
\ MemberAF Member CD
DABC :: .1158/.3158 = .366
= =
~
DAC .1333/.3158 .422
DAD • • 0667/.3158 = .212
F 1791 13747
JOINT D. ZK ~ .0667 + .2000 = .2667
Member FD DDA = .0667/.2667 = .25
DDE = 2000/.2667 = .75
The carry-over factor for ABC is -0.452 as pre-
8753 viously derived for a half circle arc, and 0.5
5952
tor the straight members.
D 962
Calculation of ~ixed ~d Moments
1 1 _ I I __
2340 For member AC, riA = ME 3 12 wL:a = 12 x 10 x 60
Member FED
3000
E

Figs. All. 84. Final Bending Moment Curves on Each

?or DE, !'IE = !'IE = ?L/8 = 1000 x 60/8 =7500
Member of Frame.
 All. 36 THE MOMENT DISTRIBUTION METHOD

Curved member ABC. total elastic weight of ~emjer and its elastic
mcments o! iner~~a abo~t ~es t~rJugh :~e
The ~lxed end moments on this curvec e Ias t Lc cent.er •
~ember due to the ex:ernal loads will be de-
termined by the elastic canter method. ~he TOTAL ~~STIC ~~IGHT
nr
': Z ds/I =-- = n x ~O
l.v
assumed static frame condition will be an arch
pinned at A and su?ported on rollers at 8. = 62.83
(See Fig. All.B6)
Distance y from line AC to elastic center of
Fig. AIl.S7 arch ABC equals 0.6366r = .6366 x 30 ': 19.1 i~.
shews the general 1000 =P B 1000 : P
Ix =0.2978r 3 / I =
2978 x 30 3/1.5 = 5350

!~\I
Shape of the static
moment.cu~re. :or
the fr~~e portion I' /
between the re- I 00:" 30~_ .- .- C _---,.<~ Os~ • 167300
AA, -r-~--
actions and the load
paints, the bending Fig. All. B6 ~,0Mo
-/-.---'~ - "0-
moment equation is
M=P(r-r cos a).
(2) I i
0'""'l -Os'1 =288001 19 • 1
For the beam (1' I \ I
portion between the (1) -- - - 'lIT....-.%-.
C
two 1000 lb. loadS, Fig. All. 87
the bending moment Static Moment Fig. All. 88
1s constant and Curve.
equals, Calculation of redunda.~t forces at elastic cer--
M=P(r - r cos 30 0 ) tar.
Calculation of 0s Values. Z 09 = -(28800 +' 67300) =
110 - - Z osvt 62.83 in.lb.
The ¢g values equal the area of t~e X/I
C~les. The moment curve in Fig. AII.S? has _ Z 0sY _ 167300x5.7+28800(-9.l) =
been broken down into three Darts labeled (1) Xo -~- ;:)350
(1') and (2). - 129.3 lb.

0S 1 + .elsa = 2ti a(a - sin a~ Yo = =0 (because x = 0)

= 2~0~~5X 30 '(.524 _ .5 ~ = 28800 The fixed end ~oments at ends A and C will
equal the moment due to the redundant forCes ~~
The vertical distance y trom line AC to and Yo since the static ~cment assumed was zero
centroid of ¢S1 and .elSa values is, at A and C.

y = r(l- cos a -~) = 30(1- .867 - 0.5/2) = MA = Me = Mo + Xoy = - 3122 + 129.3 (- :9.:;
c c s in c .524-.50
= - 652 in. lb.
ro in.
Xament ~istribut1on Process
no:
):/sa
=- Prae
I
(1 _
cos a ) . e = 120 0, a• 30 0•
Having determined the fixed end ~cmen"'Cs,
distribution and carry-over tacters, ~~e =oment
¢., = 1000~30'X2.1 (1- .867) = 167300 distribution can now be carried o~t. Fig.
.0 AIl.89 shows the solution. The first cycle wi::
be explained. Starting at joint (A) the un-
Vertical distance y from line AC to centroid of balanced moment is -3000 - 652 = - 3652. The
.elSa equals, joint is balanced by dis~ributlng .~2 x 3652
= 1543 to ACj .364 x 1543 ': 1333 to AbC and
2r sin 8/2 = 2x30xO.S67 .212 x 3652 = 776 to AD. The carry-over moment
r » e 2.1 = 24.8 in. from C to A = .5(-1543) : -772; from C on
member eBA to A = - .452(-l332) ': 601; and trom
Fig. All.SS shows the 0s va1ues and their lo- A to D = .5 x 776 = 388. Now ~roceedl~g to
cation with respect to x and y ~es ~hrough joint 0 the unbalanced TIoment is -7500 + 388 =
the arch elastic center. -7112. ~~e jOint is balanced by distributIng
.75 x 7112 = 5340 to DE and the re~ainder 25
The elastic 8entar ~ethod reqUires the per :ent = 1772 to DA.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES All. 37

B
__ over from C
_.-lS2. (_6S2.\ = car·,
Static Mom.
(/r : ~
u '" ...,'" ""ao ....'''' «>''''' -,<:>,.,.....
• 366
(1)
Curve

"I'
~ :;ii/ \~
( 1)
1::1"'''''$'''''-''''''''''''''<'«'''-
<"i'~: ";'1'7'71' 'I' ",
A .422.1-",;;=------:"::0'"01 C
,.212. -3000 i.t.M. I 2513
I I Static Mom •
• 302./
151 Curve
-l:~/
.~~/
Fig. A1l.90 -r/a wLi!. = 4500

.~~/ I 0s = 2513 x 60/2 = 75390

•
~
-0&.
-2513 I
I
);0 s = 218290
o l..(illl--------' "
Ja 7500
Fig. AU.89 .7500 Calculation of elastic center location and
~340/
_2670

-~~g/ Distance y from line AC to elastic center =

602/ "
.401
-.ill/
-156
19.1 x 1\ x 30/1.5 = ~ = 12.97"-
nx30+60 92.83
---!ll/
-60 1.5 Z
=" TOTAL EL~STIC WEIGHT =92.83.
Ix = .Z978x30:'l +..1!..E2.9. (6.13)2 +.§Qx 12.97 11
1.5 1.5 2
T~e ca~y-over moments are: -
= 12713
Fr-om S to D = .5 x (-5340) 2 -2670
Fig. All.91 shows the elastic center loca-
To A ~rom D = .5 (1772) = 886 tion and the ¢s values together with their
centroid locations.
The :irst cycle is now completed. Five more
cycles are carried out in Fig. AII.89 in order
to obtain reasonable accuracy of results. The
fl~al end moments are listed below the double
short lines. IB 0., • 167300
On arch member ABC the end moments of 584
---:7f::::::::::.f/ =Il"':.,:=:., 0
are correct. However before the ene moments 11.63-,- •
at any other potnt on the arch can be found I 6.13 -. -
the horizontal reaction on the arch at A must -f---~co-'~'--'~' x, - - T - " -I- x
be determined. This reaction will be de- 'T-'------::;,;t' ~ - Ps, '2BBOO \
te~ir.ed by the elastic center method.
Y =, 12.97
I

A '<, C
?ig. All.90 shows the bencing moment tis = -90000 '" 75390 = - 14610
c~ves for ~2mbers ABC and AC as ~de up of Fig. All. 91
5 parts labeled: to 5.
Solving for redundant forces at elastic center.
Calculation of 0s val~es Nhic~ equal area
of each ~oment curve ji,1ded by r of member. 0s = -218290
Mo = z l: ds/I 92.83
=-2356
2~Sl = 25800
X =..usL= 167300 x 11.83+ 36800 x 6.13 +
0·'. = 167300 o Ix 12713
28800 (-2.971 - 14C:O (-12.97)= 181
0s , = 58~ x rt x 30/1.5 = 36800 12713 lb.

0s , =- ( .567 x 4.SCO x 60/2) = -90000

AIL 38 THE MOMENT DISTRIBUTION METHOD

The bending ~oment at ars pOin~ en T.ember we ao tb/tn.

AC or ABC equals the values in Fig. A~l.~O
~lus those due to Xo and Xo in ?ig. All.Sl.
T 1 ' ·L~h'·/ ~~O I

,,
20" 400~ I =6 1~400
]:::4 1=41 10
Thus at ~oint A on member ABC, -~
MABC = 584-2356+181x12.97 = 554 in. lb. Fig. 7 Fig. 8
which is the same as found by the moment distri-
bution process.
100 400 600 400
~_100
At point B on arch: - :-12 ~ r-I O Pin ': 4"
ME tram Fig. AIL90 = 584 + 4000 = 4584 1- 5" ..1 1·4 I I '" 6 1= 3 AI=l :"'-8"-
Mg from Fig. All.9l -2356 -181 x (30 - 12.37) I--- 16" ----'
I

[=2 ;I=3 11=2

= -5476 1 lPin J.. Fig. 10
Hence MS = 4584-5476 ,. -392 tnv Lb, r- 24"-f- 20"----i 5" Ii""""
Fig. 9
STRUCTG~ 'ITH JOINT DISPLAC~S

For unsymmetrical frames or for frames

loaded unsymmetrically, the assumption that (3 )
only jOint rotation occurs may give resulting
moments considerably in error. Joint dis- ~12"~
placement can be handled in a manner as 12 lb/in.
previously explained and illustrated for frames rt' Itt

composed of straight ~embers. A

f-20',-+---40" ---!--32" --!--- 40"_20'~

The distribution of the skin supporting , . ,

forces on the frame boundary are usually taken Fig. 11

as following the shear flow distribution for
the shell in bending as explained in a later
Chapter. In the loaded bea~ of Fig. 11, the supports
a and E of the elevator beam deflect 0.1 inch
All. 20 Problems.
~ore than su,ports C and D. Com?ute the re-
(1) Deter.nlne the bending moment diagram for sultant bendinz mOwents at the suppo~ts and fine
the loaded structures sho,vn in Figs. 1 reactions. EI = 320,000 Ib.i~.sq.
to 6.

t:.t= '
(4) Fig. 12 illustrates a continuous 3 span
1100 1100 w= 101b/in.
Wing bea~, carrying a uniform air load of 20
1f-8"-+-16"-----li "I
~
I • lb./ln. Deter.nlne the beam bending moments at
24"----1 a strut points A and 8. Take
Fig. 1 Fig. 2
IAB=171n., IAA I = 20 in., and E = 1.3x:l.O e ln~
'fo w=201b/. 200y" . . .
I::2 1=1 fll'=P ----{~
h+i·to16
In. Neglect e~fect of support deflection due to
strut axial defor.rnat:on.
I-- 16'.--1---- 12"-~ Fig. 4 ~50"~ 110,-"---100'+ 110'~50'1
Fig. 3 ;

100 f , I , • •• • .<W=.20.rb~in,'
....
,
12~ , 1~ 12 ......
t

.A '¥!
"",
hI 1_2
f- 14" 36"--~ 16" -j
Fig.S Fig. 12

100 20C w = 121b/in.

,
,

I : I *i ;
(5) Figs. 13 to 15 are leaded structures :hat
suffer jOint translation. Jeter.nlne be~ding
Ii
j

A I· I ~ I· 1.5 I" I - 2 1·2 ! moment dfagz-am ,

~6" +-12"-+12"-.L'- 16"-+ 12"-j 100
500
l~
-<5f~O~~
....." ,
200
' ,300
Fig. 6 =10 1_5
I
24" iI = 2
'N 12
111 ...... ~.:..
1-:; -_
..-c 'I
200 __I
I=4 N
10"
II

(2) Find the bending moments at all ~oints ,; 1-6" ~~ :1= 4 -~

-1..""'''''''T _
and sup~or~ ~oints Jf the ~oaced s~ruct~res 1:.:
in Figs. 7 to 10.
f- 20"-; Fig. 14 -20" -r-- 24"--": I
Fig. 13
Fig. 15
 •

ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES

CHAPTER Al2

SPECIAL METHODS . SLOPE DEFLECTION METHOD

A12.1 General. The slope-deflection method is Fig. A12. 1a

another widely used special ~ethod for analyz- I L
ing all types of statically lndeter.nlnate
beams and rigid frames. In this method as in ,.ill f-I------,=.,...--:;c=-:-----I1(2)
I EI Constant "I
the moment distribution method all jOints are d1 MJ,
considered rigid from the standpoint that all 1..- C -Y- __ -L d.
angles between ~embers meeting at a joint are
assumed not to change in magnitude as loads Fig. b
- ~ - '-'_Q,
-.<
- - - 'il'
->V-
are ~pplled to the structure. In the slope
deflection TIlethod the rotations of the joints
i) M.
are treated as the unknowns. For a nember
bounded by two end joints J the end moments can
be expressed in terms of the end rotations.
Furthernore for static equilibrium the sum of
the end moments on the members meeting at a
joint ~ust be equal to zero. These equations
of static equilibrium provide the necessary
conditions to bzndle the ~own joint rota-
tions and when these unknown jOint rotations
are found, the end moments can be computed
frum the slope-deflection equations. The ad-
v~~tages of the slope deflections method will
be discussed at the end of the chapter after
the method has been explained and applied to
problem solution.
Figs. Al2. 1
A12.2 Derivation of Slope Deflection Equation.

The problem 15 to dete~lne the relation- distance !J. which causes end moments M~ and M":., .
ships between the displacements Of the end In deriving the slope deflection equations each
supports or a beam and the resulting end of the three beam deflections will be considered
moments on the beam. separately and the reSUlts are then added to
give the final equations. Fig. A12.2 shows Fig.
Fig. A12.1a shows a beam restrained at c repeated.
ends 1 and 2. It is assw~ed unloaded and at
constant cross-section or moment of inertia.
This bea~ is now displaced as shown in Fig. b, Fig. A12. 2 L
namely, that the ends are rotated through the
angles eJ, and e a plus a vertical displacement
dJ, and d lil of its ends rr-cm its original pos i ,
tion, which ?rOduCes a relative deflection 6
of its two ends. The angle 0 representir~ the
swing of the member equals 6/L.

The problem is to derive equations for M'a b

MJ, and Ma in te~s of the end slopes 6J, and
6 3 , the length of the jearn and EI. Figs. c,
d and e lllustrate hew the total beam deflect10
in ?ig. b is broken dcvm into three separate
cer iect ions , 1:1 F':g. c , the end (2) is con-
sidered fixed and a moment Mi is applied to
rotate end (1) through an angle e 1 • In Fig.a,
end (1) is considered fixed and a moment M: is
applied at (2) :0 rotate end (2) through an
angle 6 3 • In Fig. e, both ends are considered
fixed and end (2) 1s displaced downward a
2
~M,··Ml
~ '--2-
The applied moment Mi, is pOSitive (tension in
bottom fibers). Since end (2) is assumed fixed
an WL~own ~oment M~ is produced at (2). Fig.
Al2.3 shows the bending moment diagram ~de up
of two triangles. Mia being unknown '...il1 be

A12.1
A12.2 SP ECIAL M E l' HOD S - DEFLECTION METHOD
S L 0 P'-"E--="'-'-"c:..::...:.:.:::...:.:.....:::.-"-''-''-''-''--- _

assumed also ~osltlve, as the algebraic solu- I Fig. A12~6 shows ~he ~oment diagr~ ~0r ~~
tion will deternlne the true sign of r.~. applied bending noment ~~ which rotates en~ (2)
thro~gh an angle 6 2 Nhen other ~nd (1) is ~!xed
~NO ~oment area theore~s will be used 1~ (See ?1g. A12.5). In a similar ~nner as
deriving the slope-deflection equations, desc~ibed befo~e,
namely (I) that the change in slope of the
beam elastic curve between two ~oints on the Sa = (-M'lL/2) + (-~~L/2) = :-~(~l1~"~.;+~,,;.~~~)"_L _ (5)
beam is equal in magnitude to the area of the E1 ~I 2E1
M/EI diagram between the two points, and (II)
the deflection of a paint (A) on the elastic ~ak1ng moments of M/EI diagram about (2),
curve away from a tangent to the elastic curve
at (B) is equal in magnitude to the first (l1~L)!: + (l1~L) 2L = 0
~oment about A of the MjEI diagram between (A) \2E1 3 2E1::3
and (B) acti~g as a load.
or, - - - - - - - - - - (6)
In Fig. A12.2 since Sa 0, then 6~ will =
equal the area of the M curves divided by Sl. whence,

) = (11'~+i1~)L
e =
1.
(11\L/2) +
EI
(11~L/2
E1 2E1
(1)
e. = M~L
2£1 an
d Mil - 2ElS"
~ - L
(7\.

The deflection at end (1) away from a Then from equation (6)
tangent at (2) equals zero, thus we take
moments or the moment diagram about (1) and
equate to zero. .
M" --
4E16 a
L
- - (S)

(lille)
2E1 3
h (!ill,,) 2L
+
2E1 3
Fig. A12.8 shows the third ~art of the
beam deflection, namely, support (2) is jef:ec~sd
downward a dista~ce 6 assuming both ends fixed.
whence -tl'llI =!1'J2 -- (2)

Fig. A12.4 shows the tr~e shape of moment

diagram.

SUbstituting ~esults in equation (2) in

equation (1) we obtain,

or I1' 4E1el.
,- L - - - - - - - (3)
,
Fig. A12.9
M'''.,h
, 2 M'a·~t
Then !rom equation (2)

11'. =
-
2EI6j
L - - - - - - - - - - - - - (4)
M'{' c:==:====-==
.............-- . , '"
M

Fig. A12. 10
The change in slope of the elastic curve
between the enes equal zero, thus Jy the moment
L
Fig.AI2.5 area theorem the area of M/EI diagram equals
zero.
"r (, ~_"'-=- __
+ M~ )L
I---.i; L ----r-- Hence,
2El
o
Whence M:~ =-:1':- - - - - - - - (9)
Fig. AI2. 6
Frem the deflection theorem,

M'~ (10)
Mi,': ""T e:t====="
Subt. (9) in (10)
Fig. A12. 7

= - 6EL\. (11 )
~-
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A12.3

+ 6£1:.6- (lea)
= --'i-'T'-
, 41 (12)
(lSb)
The ccmo tned effect of deflection Or , o,
an~ ~ can now be obtained by adding the separ- A12. 3 Hinged End. Slope Deflection Equation.
3.te results.
Consider that end (2) of beam 1-2 is free-
ly supported or in other wards hinged. This
means that Ma_l. is zero. Thus equation (18)
can be equated to zero.
Let K ~ ~C and 0 ~ ~/L

(13)
whence
ana ;1a_ .. = -2K (28,,+9 .. -30) (l4)
2K~ ~ -K81. + 3K0' - .SMFa
These end ~oments due to the distortion
of the supports must be added to the moments Substituting this value in equation (17)
due to ~ny applied loads on the beam when
considered ~s :lxed ended. Let these fixed
ecd moments be l~beled MF l and MF". Then A12.4 Example Problems.
equations (13) and (14) can be written includ-
ing these mcments. Problem 1
(15) Fig. Al2.11 shows a ~NO span continuous
(16) beam fixed at ends (1) and (3) and carrying
lateral loads as shown. The bending ~oments at
points lJ 2 and 3 will be dete~ined. Relative
MCCI?I~D SIGN CO~~ION
values of I are shown for each member.
Squatlons (15) and (16) have been 100 lb.
jeveloped uSing the conventional signs for ~ 12" ~ w = 24 lb/In,
~ 1 -.:21Ji!i
bendi~g ~cments. In general there are ad-
V2~ntages of uSing a statical Sign convention
2.3 '.'las ueed in the moment d t s t r tbut ton method
1 § L ; 24"
I~24
i L • 24"
1=48
in :hapter All. Therefore, the following
sign :onvention will be used in this chapter: - Fig. A12.11

(1) The rotation of a jOint at member 1s posi-

SOLUTION:
tive 1: it tu~s in a cloc~Nise direction. Calculation of fixed end mcments due to
app l I ed loads -
(2) An end mcment is considered Positive if it
tends to rotate the end of the member MF .._ a ::: -PL/8 ~ 100 x 24/8 ~ -300 in. lb.
~loc~N1se or t~e jOint counterclockwise.

This ~dopted s~gn convention is the sa~e Mp,_, = PL/8 = 300

as adopted :or the ~o~ent distribution method The S1gns as shown are deter.nlned from the
(See Art. Al:.2).
tact that the end moment at (1) tends to rotate
~Gen equations (15) and (IS) are revised the end of the T.ember counterclockNise which is
a ?ositive noment according to our adopted sign
~or :h~s new Sign ccnvention they become: - conventicn. gy similar reasoning the fixed in
moment a: (2) is positive because the moment
Ml._ a = 2K (26",+6 2 - 3 0 ) +MS- .. (17)
tends to rotate end of ~ember clockNlse. for
(18) span 2-3: -
11 41_ .. ~ 2K (29 2 + 8 l - 391) +?'!S-"

~q~a~ions and (18) are referred to as the

(~7)
MF a-3 =wV /12 = 24 x 24 /12 = -1152 a
Inc Lb ,
sl~pe-ceflectlcn equat~0ns ~here K = SIlL and
o = oiL.
K vaIues K = SIlL. s ince ~ is constant it
?or no Yielding or ~ransverse movement of will ~e considered '~ity.
sU~yorts, ~ ~ 0 a~d equa~1o~s (17) and (lB)
FOR BEAM 1-2, K = IlL ~ 24/24 ~ 1
FOR SEA!'! 2-3, K = 48/24 ~ 2
A12.4 SPECIAL METHODS - SLOPE DEFLECTION METHOD

SUbstituting in slope-def1ec~ion equations lea Ex~mp1e P~cble~ 2

and 18b. (~= a since supports do not trans-
100* 100 100
late. )

Beam 1-2. 81 = 0 because of fixity at (1).

--12·'--00.;
§ t
L8"+8"~a"~ IIIlli\l\ §
(1) ~ L ' 24
I = 24
Mlo- = 2K (28 + e.l +i'IF
. 1 1 K =1

I1 1 - ,.= 2xl (0+8.)-300 Fig. A12.12

Ml._,. = Z8,. - 300 - - - - - - - - - - - (a) Fig. A12.12 shows a loaded 3 span beam
fixed at points (1) and (4). ~hls ~~xity at
M.- = 2K (28 + 8:1.) +i'IF '1
1 a (1) and (4) causes 81 and 8~ to be zero.
M'1 _ .. = 2xl (26. + 0) + 300 Fixed end moments:

l'l~, = 48'1 + 300 ---- - ---- -- It) MFl,,_g =- PL/8 = 100 x 24/3 = - 300,
Beam 2-3. 6. = 0, because ot fixity at (3) . !iF -'1-1 = 300
11 11-= 2K (28 a + 6,,) +i'IF •
" MF-'1_::I = (100x8xlS Il / 24 ) + (lOOx16xa 2 / 2 4- 2 )

=-534
M._ a = 2x2 (26. + 0) -1152

M._ a = 86 - - - - - - - - - - - (e)
a-1152
Ml;"
'''-a
= 534
l'l_. = 2.'{ (28" + 8 +MF" a )

l'lF._. = 400
11 3 = 2x2 (0+9,.) + 1152
_ .

M,,_a: = 49 a + 1152 - - - - - - - - - - - (d)

SLOPE DEFLECTID)J ::QUATIONS (0 = 0)

At joint (2) l:l'l = 0 for static equt r rbr-tum, !11._11 = 2 X l( 0 + 8:a) - 300 = 28 2 - 300 (a)

Hence Mg-l,,:: 2xl(28 g+0)+3CO = 48 11+300 {b )

MII _ " = 2 x 2 (Z8,. + 6" ) - 534 = 86 2 + -i-8::1

- 534 - - - - - - - - - - - - - - (e)
SUbstituting trom (b) and (e)
4Ba: + 300 + S9 a -1152 = 0
+ 534 - - - - - - - - I co )
whence 8 11 71
::
!13_. =2 x 1(28" + 0) -400 :: 4e 3 - 400- (e)
With 8 11 known the final end ~oments can be
round by substituting in equations (a), (b), M~_" = 2x1(0+8,,) +400 = 28" +400
(0) and (d) as follows: _
Joint-Moment eQul1ibril~ equations: -
!11_. =2x71-300 =-158 in.lb.
JOINT (2)
M.._.. :: 4 x 71 + 300 = 584
whence
!1._3 :: a x 71 - 1152 = - 584
48 g + 3 0 0 + 8 8 a+48 3 - 5 3 4 = 0 or
M"_II = 4 x 71 + 1152 = 1436 - - - (6)
Changing the resulting ~ament signs to the
conventional sign practice, gives JOINT (3)

80:]1'"48 a+534+48.,-400 =0 or

I'1,,_g = -1436 128:] + 4.6:a + 134 = \) - - - -- - - - (n )

 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A12.5

Solving equations (g) and (h) for Sa and 8~ or free to rotate we use equation (19) in writ-
gives, ing equations for Ma _ 1. and M3 _ 4 •

8 01 = 26.15, 8~ =- 19.9 Ma_1. = 3K8 l1 +MF a_1. -0.5 MF1.-a· (0 ls zero)

Substituting in equations (a) to (f) inclusive

to !ind final end ~oments.
M:~-1. = 3 x 1 x e a + 250 - 0.5 (-250) =
za, + 375 - - - - - -- - (a)
i'!1._a:: Z8 a -300 = 2x25.15-300 = M"_4 = 3K63 + MF~-4 - 0.5 MF4-~
-247.7 in.1b.
!1~_4 = s x i x s, -500-0.5 (500) :: 39 3
Ma _ 1. =46a + 300 =4x 26.15 + 300 =404.6 - 750 - - - - - - - - - (b)
M2_~ = S8 a +48~ -534 = 8x26.15+4 (-19.9)
Using equations lBa and leb and substituting,
- 534 =- 404.6
M~_a = 88~ +48 01 +534 = 8 (-19.9) +4x26.l5
Ma-~ = 2K{28 a + 8 ... M'l,;" , , 3 )

+ 534 = 479.4 Ma- 3 = 2xO.5(Z8 + S +0 = 26 + 6 - - l1 3 ) 2 3 (c)

M3 _ = 48 -400 = 4- (-19.9) -400 =

M~_a
=2K(Z8 +-. ea ) +MIi". 3 ~-a

- 479.4 !1~_a = 2 x 0.5 (2e~ + e a ) + 0 = 28 + 6 - 3 11- (d)

!1"'_ 3 :: 28 3 + 400 =2 (-19.9) + 400 = 360.2 For statical equilibrium af joints: -

(Note: - Student should convert to conventional JOIN'~ (2) Ma _ 1. + 11'a_3 =0

~oment signs and draw complete bending ~oment
diagram) . Sutrt , 36 a + 325 ... 29,. + 8 3 =0
~ample Problem 3. Bea~ with Si~Dle Supports. or sea + 8~ + 375 =0 - - - - - - - - (e)

Fig. A12.13 shows a leaded 3 span contin- Jourr (3)

uous beam with hinged su?ports at points (1)
and (4), which ~eans that ~1.-a and !14 - 3 O. = 28 3 + ea ... 38 3 - 750 =0
The ~oments at supports (2) ane (3) wi:l be
dete:r:nlned. or 58 3 ... ell - 750 =a - - - - - - - (f')

Solving (e) and (r) for ell and 8 3 gives.

(21 131 ea - - 109.4, 6a =172

L = 20 L = 20
I ,. 10 I = 20 T~e end ~oments at (2) and (3) can now be found
K = 0.5 K=l from equation (a) to (d) inclusive.

MlI _ 1. = 36" + 375 =3 (-109.4). + 375 = 46.8

Fig. A12.13 in.Ib.

SOLUTION: M:a_ll = Z8 a + 63 =2 (-109.4):;' 172 = -45.S

Fixed end ~oments: - !'!3-OI = 26 3 ", Sa = 2 X 172 + (-l09.~) = 234
~F1._a =- PL/8 = 100 x 20/8 = - 250, M~_4 =383 - 750 = 3 X 172 - 750 - - 234

!1Fa_ = 250
1. A12.5 Loaded Continuous Beam with
Yielding Supports.
i1...
- ,~ =!'!F ~_a = 0 (No load on span 2-3)
The
~ovable contr81 surfac~s of an air~lar.e,
the elevator, rUdder ~G aileron aTe
11~
. 3-4 = FL/8 = 200 x 20/8 = - 500, na~ely
attached at several ro ints to the s tact i icer ,
!14_~ = 500 fin and wine r-esr-ec t tve Ly. T:"_ese suppor t t ng
surfaces are usua:ly ca~tl1ever strUctU:-8S and
Slope-def1ection equa~ions: - ~hus t~e suppcrting ~olnts ~o" ~~= 7.cvab~e ccn-
~rJl sur~aces suffer a ~lsp1ace~ent :hus ~"o­
Since suppor:s at (1) ane {~; ~re ~i~~ed
~o~:~~ a cJ~tin'..:.ous ~e~~ cr. Jlelc~~s cr de~:ec:ed

-"."',
A12.6 SPECIAL METHODS - SLOPE DEFLECTION METHOD

supports. The slope deflection equations can

take care of this support displacement as MF"'-3
40" (-0 x 3 .0-
=W +
; -)
_.0 = -07
0U

illustrated :n the following exa~ple ~roblem.

The be~ has a CQnstant sec:~an, hence,
E:<:ample Problem. K for all s~ans equals,

3 Ib/l.n ~ 4 Ibl in K = EI/L = (10,000,000 x 0.03339) 40 = 8347

14!1IilllilTIVIJll!,j!!
2 t3 , f \
'r Calculation of ~ val~es.

~10'~40"- ~10't-

*--1.
" I
40" -
I

I
40"--+-- 40"
'

Fig. A12. 14
\'

~/16'"
deflected sUPPJrts.
Fig. A12.15
Fig. A12.14 is representative of an ele-
vator beam attached to detlec~ing stabilizer
str~cture at the five reaction points as
indicated in the figure. The elevator beam
15 a round aluminum tube 1-1/4 - .049 in size.
The moment or inertia (I) of this tube equals
0.03339 and the modulus of elasticity (S) or
the material equals 10 million psi. The air
load on the elevator beam is variable as
indicated in Fig. Al2.l4.
Fig. A12.15 shows the Shape of the de-
flecting supporting structure, which means
that supports (2) and (3) move downward 'thr-ough
the distances indicated on the figure. The
problem will be to determine the bending
moment at the supports under the combined
action or transverse loading and settling of
supports.
tube
wi~h
Span 2-3. The settlement of support (2)
respect to SUDDor"': (3) :: (0.5 _ .':'875) =
.3125 inches = ~
---------. ¢ =- 6/L =- ,3125/40 =- 0,007812 cad,
(Since (2) noves co~terclockN1se with ~espect
to (3) the Sign of 0" is nega t tve ,
Span 3-4. ~ = .1875 inches
¢ =-/l/L =- .1875/40 =- .004688 cad,
Since the be~, external loadly.g and supper:
settlement 1s sJillmetrical about support (4), the
slope of the beam elastic cur-ve at (.;) is nor t ,
zontal or zero and theref0re 9.. = O. T~us cnly
one-half of the structure need be considered in
the solution. Due to the fact :hat (2) is a
Simple support with a cantilever overhang, the
moment M~_l is statically dete~l~ate and equa~s
5 x 3 x 3.33 :: 50 m.rc, Then for stat ;c equili-
brium of joint (2), I1 11- 3 must equal - 50.
Substitution in slope deflection equations
(17) and (18): -

80LlITION: -
M.a_3 =2K(28 11 + 8 3 - 30") + I1F 11-3
- 50 = 2X8347 [Z9.a +93 -3(,007812)J -426
Calculation of fixed end moments: -
or = 333886 11 +1669483 +15 =0 - - - - - - - (1)
For a trapeqoidal loading as shown in
Fig. a, the fixed end ~omen~s are, ~_a = 2K(29 3 + 911 - 30') + :11'\_11 = 2 x 3347
[293 +9'i1 -3(-.007812)J +440
_ L"
M:i._ - 60 (5u + zv)
Whence ~_II = 333888 3 + 166948::1 ... 831 - - - - (2)
_ L" (ao - '"\'
Ma-:i. - 60 (5u + 3v) ,'"'... - Uj

SUbstituting in these
equations tor the loading Fig. a Whence !13_.... = 333888" - 259 -----(3)
values as shown in Fig.
A12.4, M..._" = 2K(28... - 8" - 30')
40"
M'F.a_3 :: W (5x3+ 1) =-426 in.1J. ~_3 =2 x 8347 [0 + 9" - 3(-.004688)J + 507
!1 ..._" = 166940 + 742
MF3 _ 11
=:gil (5x3+1.5) = 440
whence 3 - - _ (4)

For static eqUilibrium of joint (3),

M3 _ 1I+I1 3 _ ... = 0, hence

3338893 + lE6g4e~ + 331 ... 33386~ - 259 =0

 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A12.7

whence , 567766;,+166948 11+572 = 0 - - - - - (5) MFl._.lII = (100xSx12-)! ZOa + (100x12xS·)!

ZOIl =-480
Solving equations (1) ~nd {5J for 8 g and
9;, we obtain 6,.. = 0.00439 and 6" =- 0.00967. I1F._, = 480
SUbstituting i~ equations (2), (3) and (4)
gives the moments at these points. I1F,_, =100 X 24/8 = 300.

11,_. =33388 (-.00967) + 16694 x .00439

= 10 x 20'/12 =334. 334

+ 831 = 582 in. lb. Due to symmetry of loading we know that

=- 6 g and 8"
6 10 =-
8., which fact will shorten
11,_. = 33883 (-.00967) 259 = -582 the solution.

M4 - ;, = 16694 (-.00967) + 742 = 580 Slope deflection equations: -

converting these signs to the convent1onal I11o_.lII =2K (28:1, + e.) + MF:1,-.
signs would give Mil =-50, M" =--582 and
11 4 :::: -580. M1o_.lII =2 xI (26 1 - 91 ) - 480

A12.6 Statically Indeterminate Frames.

Joint Rotation Only.
I110_. = Z6 10 - 480 --- - - ---- - - (a)

The slope deflection equations can be 11.-, = 2K (2610+8~) + MFl. ,

used in solving all types of framed structures.
In fr~~es as compared to straight continuous M1o- 3 = 2 x 1.67 (26 1 + 8 a ) • 300
be&~s, the axial leads in the members are
usually more important, however the influence M1o_ " = 6.678 1 + 3.338" + 300 -- --- - (b)
en Joint dls~lacement due to axial deformation
of ~embers is relatively small and is there- M~-1o = 2K (Z8~+e1o) + I1F"-10
fore usually neglected in most simple framed
structures. To illustrate the slope-detlect1on 11~, = 2 x 1.67 (28,,+81.) - 300
method as applied to fr~~es, the structure
shown in Fig. A12.16 will be solved.
M,,_l. = 6.678" + 3.336 300 1 - - - ---- (c)

SXample Problem 1. I1"_4 = 2K (Z6,,+8.) + I1F,,_.,

100 100 M"_4 = 2xl (28,-8,) + 334

L,.. L4.. -L,.

I I,
.J M':l-4 = Z8,,+3.'34 - - - - - - - - - (d)
1 2
L = 20 Static Joint Equations: -
I = 20
K" JOINT (1).

L: 24 L: 24 which gives,
1 0 0 - I = 40 I =40 lot-- 100
K = 1. 67 K = 1. 67 281. - 480 + 6.676 1 + 3.338" + 300 =0
whence 8.678:\, + 3.338" -180 = 0 - - (e)

JOINT (3 i.
which gf.vee ,
Fig. A12.16
6.676;, + 3.3361. - 300 + 26" + 334 = a
?lg. A:Z.16 shows a closed rectangular
fr~e subjected to loadings on its four s:des
whence 8.678" +3.336:\, +14 = 0 - - - - - - - (f)
wh:c~ hold the frame in equil~brium. The
Solving equations (e) and (f) for 8" and
bending ~oments at the frame joints will be 61. gives,
deuerrmned ,

SOLUTION: -
81. = Z6.3, 9;, =-13.94
A12.8 SPECIAL METHODS - SLOPE DEFLECTION METHOD

Subt. in equation (a) Due to the flXity a~ j~i~ts (1), (3) ~nd
(5), 81.-a = 9:5-" = 9&_., = O. "het-e r or-e , the
Ml.-e =2x25.3-?80 =-428 in.D. unknowns are e a -~, e.. _~, g e _ e :::. nd 6.

Subt. in equation (d) There are t;~ee statical ~?int equatior.s

of equilibrium available, nameLy,
M:5-" = 2(-13.94) +334 = 306 tnv Ib .

Fig. Al2.17 ShryNS the resulting bending

moment diagram.
Another equation is necessary beca~se there are
372 tour unknowns. This additional equation is
obtained by applying the eq~atic~s of statics
to the free bodies of the Jartical ~embers.
Fig. Al2.l9 shews the free bccies.

233 M e-!l
Fig.A12.17 f~
T 4 T i 6
I
Li,J I
i 15
194
.I'_a,
'-'
....1,. ,J.,--H~

A12.7 Frames with Joint Displacements.
In the previous example problems. the
conditions were such that only joint ~otations
took place, or it any transverse joint dis-
~lacement took place as in the ex~ple problem
of Art. Al2.5, these displacements were known,
or in other words the value of ¢ in the slope-
deflection was known.
Me_t'l
M~_...
Fig. A12.19
Treating each ~ember sepa~tely. we ta~e
moments about the upper end ~.d equate to zero
and then solve for the horizontal reaction at
the bottom end. The results are,

In ~~ practical frames however, the

joints suffer unknown displacements as for
example in the frame ot Fig. A12.l8. The term
~ =~/L in the slope-deflection equation was
derived on the basis of transverse dlsplace-
ment of the member ends when both ends were
fixed. Thus in Fig. A12.18 under the action
of the external loadS, will sway to the right With these horizontal reactions known the
as magnified by the dashed lines. Neglecting static equation ZH = 0 can now be applied to
any joint displacement due to axial deforma- the frame as a whale, which gives.
tion. the upper end of each member will move
through the same displacement~. Then we can (20)
write,
Equation (20) is generallY referred ~o as
the bent or shear equation and is an equation
that supplies the necessary extra conc I t j on to
take care of the additional unknown ~.

A12.8 Example Problems of Frames with Unknown

Joint Displacement.

Exa~ple Problem 1

Fig. A12.20 shows an unsymmetrlcal ~Ta~e

with ~~symmetrlcal loading. The bending
~oment c~ve will be determined.

SOLUTEN: -
Fig. A12.18 The relat:ve ~8rnent o~ :~ertia of each
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A12.9

~embeT ~s given on MliI _ ..

= 2 x 1.667(28 .. + 8 .. -0) -144
96.
-;:[';,e r rgur-e . The
•1-- -1
6
r-eIa t tve K =
IlL M.._.. = 6.6678 .. +3.3338.. -144- ---- (c)
values are also
i~dicated aGjacent I1.._ a = 2 x 1.667(26, + 6 a - 0) + 144
to eacn member".
I1.._ a = 6.6678.. + 3.3338 1i1 + 144 - ---- (d)
Relative values
of .0 due to sldesway 11,_ = 2xl(O+6,-3xO.l,0)+0
of frame: -

The angles .0
M,,_ .. = Z8, - 0.60 - - - - - - - --- (e)
Fig. A12.20
are proportional to 11_, = Zxl(Z8, +0-3xO.10) +0
the ~/L for each
member where ~ is 11,~ = 48, - 0.60 - ---------- (r)
same for member-s
(1-2) and (3-4) and Equilibrium equations: -
zero for member (2-4).
(See Fig. A12.10). JOIN]' (Z).

Hence, Substituting: -

2.6678.. - 0.26670 + 103.68 + 6.6676 11 + 3.3336...

Fig. A12.10
- 144 = 0

" .. = 10
)(.1,,_ 1 = 0.100 whence, 9.33361i1-0.2667~+3.3338,-40.32
=0 - - - - - - - - - - - - - - (g)
o JOIN]' (4). I1.._ a + I1,_;s =0
.0:a_-4.= 12 = 0
FL"'CZD 2ND :-!OMEN'I'S: - 6.6676, + 3.3338 11 + 144 + 48", - 0.60 =0
tlember 1-2. whence, 10.6676, + 3.3338.. - 0.60 + 144 =0 - (h)

I1F
.-. =-(48xo"x9)/lS" =-69.12 tnv Lb • Writing the bent equation (See Eq. 20).

- - (1)

'M:ember 2-4. In order to get H~ and Hs in terms of end

moments see Fig. A12.21.
I1F._ =- (96 x 12)/8 =-144

MF..-a = 144

~ember3-4 has no lateral loads and thus ~ixed

end moments are zero.

The slope-def1ection equations are: -

(17) Fig. A12.21

(18) For free body of member 1-2 take moments

about (2) and equate to zero.
Writir~ above equat10ns for each ~ember
and noting that e~ and 6 3 are zero jecause MII_~ + I1~_a - 6 x 48 + 15H:\. =0
frame is fixed at support points, ~lves,
h nee H - 288 - (I11i1 _ :\. +M~ a)
ne ,.~- Lc
111.-a = 2xO.667(0+e"l-3x .0667Z1) -69.12
Subt. 'ralues of MlI _ and !11._.. in the equation,
M:\._a = 1.3336.. -0.2667Z1-69.12- - - - (a) 1

:1"-1 =2xO.667(Ze"l +O-3x.0667.0) +103.68 H~ =- 0.26676 2 + .03555.0 + 16.88 - - - - (j)

?or free body of member 3-4 take moments

Ma _ :\. = 2.6678.. - 0.2667.0 + 103.68 - - - (b) about (4) and equate to zero.
Al2.10 SPECIAL METHODS - SLOPE DEFLECTION METHOD

I1e_ +!"',",-_., + lOR., =0 Ing the symme'tr t ca.L cf s tr-Lbut.ed load as shcwn .

w: 12 lb/in.
..
Substituting values of
equations, gives,
M~_~ and M~_~ in above I 3
10"

t
H, =- o.se. + 0.120 - - - - - - - - - (k)
3 4
SUbstituting values of H 1. and H~ in equation
(1), gives,

-O.2667ell-.6e~+.15550-3L12=0- - (1) 20" : 10 I : 10 I

Fig. A12.23
' 20 L = 20":
Solving equations (g), (h), (1) for the
unknowns e., e~ and 0 gives, I
---L 15
0= 196.9, ell ;;: 12.17, e~ =- 6.22 1'iiTTT1
40"
The final end ~aments can now be :ound by
SUbstituting these values in equations (a) The bending moment diagram under the given load-
to (t) inclusive. ing will be dete~lned.
M1._. =1.333x12.17- .2667x196.9 SOLUTION: -
- 69.12 = -105.4 In. lb.
Relative Stiffness (K values)
!1._1 =2.667x12.17- .2667x196.9
+ 103.68 =-83.6 Members 1-2 and 4-5, K = :0/20 = 0.5
11_~ = 6.667x12.l7+3.333 (-6.22) Members 2-3 and 4-3, K = 20/22.4 = 0.692
- 144 = 83.S
Fixed end ~cments:
M .._. = 6.667 (-5.22) +3.333x 12.17 + 144
• 142.9 I1F~. = wLV12 = (12 x 20' )/12 =-?DO m Lb• s

M..~ = ,; (-6.22) - 0.6 x 196.9 = -142.9 MFa_ = 400, MF .._., = 400, MF;,_.. = - 4-00

M3_ =2 (-6.22) -0.6x19S.9 =-130.4.5 Relative values of ~: -

Fig. A12.22 shows the bendIng moment Due to the sloping ~embers, the relative
diagram. The stUdent should also draw the transverse deflections of each member are 40t
shear diagram, fInd axial load in members as obvious as when the vertical members are con-
since all these loads are needed when the nected to horizontal members as in the previous
example. In this example, joints (1) and (5)
174.7 are fixed. Because of symmetry of frame and
loading, joints (2) and (4) will ~ove ~ut~ard
83, 6",,==0..;';;; the same distance ~ as indicated in ~~g. Al2.24.
83.6 b
Furthermore, due to s~etry, jOint (3) will
undergo vertical movement only.
80.46
In Fig. Al2.24, draw a line from ~oint (2')
130.4
parallel to 2-3 and equal tn length to 2-3 ~o
locate the point (3"). Erect a perpendicular
105.4 to 2 ' -3" at 3" and where it intersects a ver-t i ca I
Fig. A12.22 through (3) locates the point (3 ') the final
location of joint (3). T~e length at 2'-3'
strength of the fra~e members are computed equals 2-3. In the triangle 3-3' _3 1' , the s ice
anQ compared against the stresses caused by 3
1_3"
=~. The relative values of ~ which
the frame loading. are measured by 6/L for each member can now ~e
calculated.
Example Problem 2.
1
.I1h-ll = - 20 =- .05.0
Fig. Al2.23 shows a gable ~rame, carry-
 ANALYSIS AN ESI N 0 v A12.11

I1._", = 2xO.S(-29. +O-3x .050) +0

l ' /t/i E
3"1 3
\C/.
0"-0- =-
= 2K(2S",
28. - 0.150 - - - - - -

+ 8. - 30) +:1r ,-.

- (g)

2'
3'--_ = 2xO.5(O-8. -3x .050') +0
i M",_ =- e:a - 0.150 - - - - (h)
I
I JOINT (2). Equl1i~rl~~ equation.

I1 a_1 + M a~ = 0, subs t i 'tu't tng -

5
28. + O.15¢ + 3.5688. - O.5352¢ - 4CO = 0

FIg. A12.24 whence, 5.56S8 a -0.32520-400 =0 - - - - (i)

~~_ =- 0.1¢ The jOint equilibrium equations at (3) and

(4) will not provide independent equations be-
Substl:~tlQn i~ SlaDe Deflection ~auations: - caUse in the ~revicus substit~tions in the
Slope-deflection equations e~ was made equal to
We ~cow that 8 L 8", = =
0, due to fiXity zero and 8. was equal to - e:a •
at jOi~ts (1) and (5). Also 9~ = 0 due to
sjmmetry or only vertical movement of joint Shear 2quaticns: -
(3). Further8ere due to symmetry e. = - Sa
Due to symmetry ~~e ~orizontal reactions
at points (1; ~~d (5) are equal and opposite
and therefore in static balance. Si~ce we ~ave
M.._;II = 2xO.5(O+S;II+ .05x3¢) +0 ~NO unknowns 6:a and 0', we need another eq~tion
to use with equatien (i). This equation can be
Ml.-a = Sa + 0.150 - - - - - - - - - - (a) obtained by stati~g that the horizontal reaction
Ha- l on ~ember 2-1 at end (2) must be equal and
opPosite to Ha_~, the horizontal reaction at
end (2) of member 2-3.
Ma _ l =2xO.5(Z8 + .05x30) +0
a
Fi~. Al2.25 and 26 show free body sketches
I1._ l =28.+0.150- - - - - - - - - - (b) of members 1-2 and 2-3.
12*/in.
1~ ~_a
1 l I I I I 1 11 \
11._" = 2xO.3S2(29. +0- .3xO.10) -400 ~----r-.
t
H._.
M.-J = 3.5688. - 0.53520 - 400 - - - - (c) T20"
H _
______

,2
_
3

'
10"
.--L-

240 20,,-----f
=2K(28
MJ _

M~_a
a

= 2x
J + S. -30) +MF

.892(0+8. -3xO.10+400
,-. 1 1
,,-,- H 1
Fig. A12.26
M \-.11

M 3-. = 1. '7249 a - 0.53520 + 400 - - (d)

Fig. A12.25

In Fig . .1..12.25 take ~cnents aJout (:) ane

equat;e to cer-c ,
;;:J_. = z x .392(0-8;11 +3xO.10) -400
M l-a + r. a- .. - 20H a_1 =0
M
•• J - .. = -1.7848. + 0.53520 - 400 - - - (e)
.
wnence, H .\I-l = M. a
+M
20 .<li!

:1._ J =2x .892(-28. +O+3xO.:0) +400 ane in Fig. A12.26, taking ~oments about (3),

I1 .._ J
=-3.5888" +0.53520+400 - - - (f) !'! a-" + 11 3-a + 240 x 20 - 20 x 12 x 10 -lOH .\1-3 =0
"+M,, 2+24:00
wr.ence, :f a-~ = 10
A12.12 SPECIAL METHODS - SLOPE DEFLECTION METHOD

Shear equa t 1or- , H .\1_1. = H 2_" ?:~. Al2.27 ShC~2 t~2 ~~2~1~:r:~ :e~~:~;
cement c iacr-em, ~::'rs: ::'.::"'2.','.7', tn oar-as i.'cr 2S.,:::j
whence, member of t~e fr~~e ant ther: addec :0 fCrT, :~e
cO~Dosi:e d:.a~~~. ~lJt:ec. :ir~c:ly :r: :~e
f::"'2.~e.stears an::'. ~~~2.l l::acs ~cllcw as
:~e 2.
~atter of statics. ~lth t~e 2cment ~i2.~r~
Y~ovm the fr~s det1sc~sd ~~~pe ~5 ~ead~~y
Substituting values for the er:d Iome~ts: - c3.1c:11a:ed.

Sa + 0.150 + Z8!i + 0.150 :::; 3.56882 - 0.53520 A12.9 Comments on Stope-Deflection Method.
20 10
The ex~~pls problems solut~ons i4c::a~e
- 400 + 1.7848 2 - 0.53520 + 4,00 + 2~OO that the met.hoc c: s Iope-def Lect t cn nae t.vo ad-
10 vantnges , nameLy, (1) i t !'SCc:.C8S t::s number- 0:'
equatior.s to )8 solved s:~~1:2~e:uslY a~G (2)
Whence, it ,resents equations trzt are ea5:11 a~c
rap:cly ~or.nulatec..
0.38528a - 0.1220 + 240 =0 - - - - (j)
Thus for structures ~lth a hi~~ ~e~ree o~
Solving equations i and j for S.\I and [2,?;i vee red~ncancYJ the slo~e-deflectio~ !1ethod should
be considered as 80s51j1y the jest !1e~hod o~
[2 =2805, ea = 265.8 sclution. The solvin~ of the equations in :~is
method are readily programmed ~or solution jy
Substit~ting these values in equations (a) to high speed computing ~achlnery.
(h), the end moments are obtained as follows: -
A12. 10 Problems.
M1.-a = 6.. +0.150 = 265.8+0.l5x:2505
= 666.5 Ins Ib , (1) Determ,ne the bend~"g Go~ents at suoport
points A, a, C, 0, for t~e con~lnuous te~~
MlII-1. = Z9 a + 0.150 = 531.6 + 420.7 = 952 shown in Fi~. A12.28.

M lII-=3.5689.. -0.5352.0-400
3 100 100 150 lb.
= 3.568 x26S.S - 0.5352 x2S05 -400
=- 952
~7'!..l.6"...L7" +-12"
+ + I l I
w,., 101b/in,
•

!'b-a =1.'7848" - 0.53520 + 400

j~~~,,20 jB 12~..a_6_.,.;C_'_; r;;,',6d r ; ~D
=1.784 x 265.3 - 0.5352 x 2605 + 400 Fig. A12, 28
= 628

M::I_... = -1.7846.11 + 0.53520 -400:::; - 628 (2) Same as problem (1) but consi::'.er sup~crt A
as ~reelY supported instead of fixed.
M._. = 952
(3 ) Determine the bendinz ~oment d~agram ~or
M ..._ 6 = - 952 the variOUS loaded structures :n r:g.Al2.2S.
w,., 12 lb/in.
M:~_ ... =-686.5 200 lb. I : .: .. I ;

C~D
~8"'" I I
A
L:Il 16
I =32

~~ 12"
300 lb.
100~=30
I
I

.60
= 36
L=30
1·60
, '
100 lb.

oJ
1
Pin,e,t-
15 i
, I
...1.,TTm
A
M IS
I
m'\'l..L.-
B
1000 1000 lb.
,
w,., 20 Ib/in.
952
I ~5"~5"
C 1 _ 2 D

"
L,., 16

686.S • 1
,., 20 20" 1=3
, A
-L ~
Fig. A12.27
A Fig. A12.29 ,--20"---......J
 CHAPTER A13
BENDING STRESSES

Al3.0 Introduction. applied moment the beam takes the exaggerated

The bar AB in ~lg. a Is subjected to an deflected shape as shown.
axial compressive load P. If the compresSlve The applied bending moment vector acts
stresses are such that no buckling of the bar parallel to the Z axiS, or in other words the
takes place, then bar sections such as 1-1 and applied bending moment acts in a plane perpend-
2-2 move parallel to each other as the bar icular to the Z axis. Consider a beam segment
shortens under the compreSSive stress. at length L. Fig. d shows the distortion of
I, ,2 this segment when plane sections remain plane
atter bending of the beam.
It will be assumed that the beam section is
I' , 2
homogeneous, that is, made of the same material,
Fig. a and that the beam stresses are below the pro-
Pj IP portional limit stress of the material or in
,
""', ' ' "
,
,
nJL
L]
other words that Hook's Law holds.
From the geometry of similar triangles,
Sbear~
e _ Be or
---tx:+ y c
e = :i.. ec
c
(A)
/' Fig. b Fig. c
.,/ ....:.\1oment Dla. ..............
We have from Youngs Modulus E, that
In Fig. b the same bar Is used as a simply
supported beam with two applied loads P as E = unit stress =
shown. The shear and bending moment diagrams unit strain
~or the given beam loading are also shown. The
portion of the beam between sections 1-1 and 2-2 where cry is the bending stress under a deforma-
under the given loading are subjected to pure tion e and since it is compression it will be
bending since the shear Is zero in this region. given a minus sign.
Experimental evidence for a beam segment SolVing tor cry.
~ taken in this beam region under pure bending
shows that plane sections remain plane atter cry:_E ec y
bending but that the plane sections rotate with cL
respect to each other as illustrated in Fig. c,
where the dashed line represents the unstressed The most remote fiber is at a distance y = c.
beam segment and the hea'sy section the shape Hence
atter pure bending takes place. Thus the top
fibers are shortened and SUbjected to compress- = _ E ee
ive stresses and the lower fibers are elongated L
and subjected to tensile stresses. Therefore at whence
some plane n-n on the cross-sectlon, the fibers
suffer no deformation and th~ have zero stress.
cry • - crc y - - - - - - - - - - - - - - - (B)
This location of zero stress under pure bending c
is referred to as the neutral axis.
For equilibrium of the bending stress per?end-
A13.1 Location of Neutral Axis. icular to the beam cross-section or in the X
Fig. d s~ows a cantilever beam SUbjected to direction, we can write Z Fx = 0, or
a pure moment at its free end, and under this

'JM Z Z F'x '= - crc

c
f yda = o.

~
y
however in this expression. the term Oc 1s not
z-I-t.:z zero, hence the term fc
yda must equal zero and
x LJ
Beam this can only be true i ! the neutral axis coin-
Section cides with the centroidal axis of the beam
Fig. d cress-section.
A13.1
 A13.2 BEAM BENDING STRESSES
The neutral axis does not pass through the
beam. secticn centroid when the beam is nonhomo-
geneous that is, the modulus of elasticity is not
constant over the beam. section ~~d also when
Hook's Law does not apply or where the stress-
strain relationship 1s non-linear. These beam
conditions are described later in this Cr~pter.
A13.2 Equations for Bending stress, Homogeneous Beams,
Stresses Below Proportiooal Limit stress.
In the following derivations, it will be
assumed that the plane of the external loads
contain the flexural axis of the beam and hence,
the beam is not subjected to torsional forces
whiCh, it present, would produce bending stress-
es it tree warping of the beam sect~ons was re-
strained, as occurs at points of support. The
questions of flexural axes and torsional etfects
are taken up in later chapters.
m The moment of the stresses on the beam section
P moments about axis x-x, we obtain for equil-
Plane
of loads c
x-----'b-tt-.::::.:y---'- X
m
Fig. A13.1 y Fig. A13. 2
m 11 sin
Fig. A13.l represents a cross-section of a
straight cantilever beam with a constant cross-
section, subjected to external loads which lie
in a plane making an angle Q with axis 1_1
through the centroid O. To simplify the figure,
the flexural axis has been assumed to coincide
with the centroidal axiS, which in general is
not true.
Let NN represent the neutral axis under the
given loading, and let 0 be the angle between
the neutral axis and the axis X-X. The problem
is to tind the direction of the neutral axis and
the bending stress 0 at any point on the section.
In the fundamental beam theory, it is as-
sumed that the unit stress varies directly as
the distance tram the neutral axiS, within the
proportional limit of the ~terlal. Thus, Fig.
Al3.2 illustrates how the stress varies along a
line such as ~ perpendicular to the neutral
axis }I-N.
Let c represent unit bending stress at any
point a distance Yn from the neutral axis. Then
the stress 0 on da is
kYn
0= - - - - -
where k Is a constant. Since the position of
the neutral axis Is unknown, Yn will be express-
ed tor convenience in terms of rectangular co-
ordinates with respect to the axes X-X and Y-I.
Thus, Y~ • (y - x ta.~ 0) cos 0
• yeas 0 - x sin 0 - - - - - - (2 )
Then Sq , (1) bec ames
o :::; k (y cos 0 - X s tn 0) -
This equation contains three unknowns, v,
k, and 0. For solution, two additional equa-
tions are furnished by conditions of equili-
brium namely, that the sum ot the ~oments of the
external forces that lie on one side of the
section ABCD about each of the rectangular axes
X-X and y-y ~ust be equal and opposite, respect-
ively, tc the sum of the moments of the internal
stresses on the section about the same axes.
Let 11 represent the bending ~oment in the
plane of the loads; then the moment about axis
X-X and Y-Y is Mx : :; M cos g and My • M sin Q.
about axis X-X is / a da y. Hence, taking
ibrium,
J1 cos Q = / c da y
- / k (cos 0 y"cta - sin 0 xyda)
= k cos 0 / y'da - k sin 0 / xyda (4)
In similar manner, taking moments about
the Y-1 axis
Q = / 0 de x
whence
M sin Q = - k sin ¢/ x~da + k cos ¢ / xyda(4a)
Al3.3 Method 1. Stresses for Moments About the
Principal Axes.
In equation (4), the term. / y~da is the
moment of inertia of the cross-sectional area
about axis X-X, which we will denote by Ix, and
the term / xyda represents the product at in-
ertia about axes X-X and Y-Y. ~e know, however,
that the product or inertia with respect to the
principal axes 1s zero. Therefore, if we se-
lect XX and IT in such a way as to make them
coincide with the principal axes, we can write
equation (4):
M cos Q = k cos 0 Ix - - - - (5)
p
In like manner, from equation (4a)
I1 sin g k sm 0 Iy (6 )
P
To find the unit stress 0 at any point on the
(1) cross-section, we solve equation (5) for cos 0
and equation (6) for sin 0, and then substitut-
ing these values in (3), we obtain the follow-
ing expressing, giVing 0 the subscript b to
represent bending stress: _
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A13.3

moment of inertia about the neutral axt s, hence

_ [11 cos (g-~)JYn=
°b - (9 )
Let the resolved bending moment M cos g and In
M sin Q about the principal axes be given the
symbols Mx p and M,f'l..' Then we can write In can be determined from the relationship ex-
ou pressed in Chapter A3, namely,

(7 )
A13.5 Method 3. Stresses from Moments, Section Prop-
erties and Distances Referred to any Pair of
The minus Signs have been placed before each Rectangular Axes through the Centroid of the
term in order to give a negative value for crb section.
when we have a positive bending moment, or Mxp The fiber stresseS can be found without
is the moment of a couple acting about Ap-X p' resort to pr tnc i pa.L axes or to the neutral
positive when it produces compression in the axis.
upper right hand quadrant. Myp is the moment
Equation (4) can be written:
about the Yp-Y p axiS, and is also positive when
it produces compression in the upper right hand MX = k cos ¢ Ix - k sin ¢ I XY ----(11)
quadrant.
where Ix :; / ya da and Ixy :; / xyda , and Mx =
BENDING STRESS EQUATION FOR SYMMETRICAL M cos 9.
BEAM SECTIONS In like manner,
Since symmetrical axes are principal axes 11y : - k sin ~ ly + k cos ~ I XY - - - - - (12 I
(tei~ / xyda ~ 0), the bending stress equation
tor bending about the symmetrical XX and YY axes SolVing equations (11) and (12) for sin ¢ and
1s obviously, cos 0 and substituting their values in equation
(3), we obtain the following expreSSion for the
cr = _ MxY _ Myx - - - - - - - - - - - - - (7a) fiber stress crb: -
b Ix Iy
(11yl x - 11Xl xy) x _ (11xl y - 11yl xy l Y _ (13)
A13.4 Method 2. Stresses by use of Neutral Axis for
Given Plane of Loading.
IXl y - li:r Ixl y - I~y

The direction of the neutral axis NN, mea- For Simplification, let
sured from the XX p principal axis is given by
dividing equation (61 by (51. K. : lxy/(Ixly - I\y)

Tan 0" = - (8) K. : ly/(IXl y - I "\y)

K, ~
Ix/(IXl y - I\y)

SUbstituting these values in Equation (13) : -

crb = - (K,11y - K~Mx) x - (KaMx - Kl~Y)Y - (141

In Method 2, equation (8) was used to find

The negative Sign arises from the :act that
Q is meaSUT3d from one principal axis and ~ is
measured in the same directicn :rom the other
principal axis.
Since equation (8) gives us the location of
the neutral axis for a particular plane of load-
Ing,the stress at any point can be found by re-
SOlving the external moment into a plane perpen-
dicular to the neut~al axis ~-N and using the
the positi0n cf the r.eutral axis for a given
plane of loading. ~he loca~ion o~ the neutral
axis can also be r ound r e Ia t Ive to any pair of
rectangular centroidal axes X and Y as follows: -
Since the stress at any point on the neutral
axis must be zero, we can write frcm equaticn
(141 that: -
(K Sl'1y - K1Mx)X = - (KJ1x - KaI":ylY for all
points located on the neutral ax:s. ?rom ?lg.
A13.1 tan ~ ~ -y •
x
Thus tan 0 = _ (KJ1y - KlMX) - - - - - - - (15)
(K,I'lx - K. 11yl
 A13.4 BEAM BENDING STRESSES

It frequently happens that t~e plane of t~e Al3.8 Illustrative Problems. Example Problem l.
bending moment coincides with e~ther the X-X or Fig. A13.3 shows
the Y-Y axiS, thus making either Mx or ~y equal a unsvnme
J~~~
t r i ca I one f, P y " 6000#
to zero. In this case, equation (15) can be cell box beam wi th i ;1 1 .
simplified. For example, it My = 0 four corner r lange / ,/ IPx:1600~i
member's a, o , c and d. I - ,1/
~' I I
I /

tan o = Ixy _ ------(16)

Let it be required to
det ermme the bending S? 'I
I 1/' I
ii
I

I,. I

1/1
J'
'I I, I - - - - - - - o - t i !
axial stress in the N ,1/ 1/
and It l1x = 0 / II II
._-
Ix ____ - - - - - - - - - - - - (17)
four corner members
due to the loads ?x
and Py acting 50"
1/ II
/.10",'
L // "

,;/ ,
il/'
tan 0
I xy from the section
abed.
a" II T .5°" ," ! I
,yf'l
A13.6 Advantages and Disadvantages of the Three In this exa.. n ple b I: JIT
I I
Methods. solution the s nee t 12" I' at!
Method 2 (bending about the neutral axis
tor a given plane of loading) no doubt gives a
ccnnec t.i ng the corner
members w::.ll be con -
xtL d ( -LX'
80' ,#1/.'-------,-16"''''- -"-:I:, .40"
better picture of the true action of the beam atder-ed inarrect ive in' ' 1

relative to its bending as a Whole. The point bending. The stresses Fig. Al3. 3
at maximum fiber stress is easily determined by will be determined by
placing a scale perpendicular to the neutral each ot the three me-
axis and moving it along the neutral axis to thods as presented in this chapter.
tind the point on the beam section farthese away
tram the neutral axis. In airplane design, SOLUTION
there are many design conditions, which cr~ge
the direction of the plane of loading, thus, The first step common to all three meth~ds
several neutral axes must be computed for each is the calculation of the moments of inertia
beam eectlon, which is a disadvantage as com- about the centroidal X and Y axes. Table A13.1
pared to the other two methods. gives the detailed calculations. The proper-
In determining the shears and moments on ties are first calculated about the reference
airplane structures, it is common practice to axes x'x r and ylyt and then transferred to the
resolve air and landing forces parallel to the parallel centroidal axes.
airplane XYZ axes and these results can be used
directly in method 3, whereas method 1 rsquires TABLE A13.l
a further resolution with respect to the prin- Area
cipal axes. Methods 1 and 3 are more Widely Mem. nA" y' x' Ay' Ax' Ay,2 Ax,2 I Ax'y'
used than method 2. 256.0 -192
a 1.0 12 -16 12 -16 144.0
Since bending moments about one prinCipal
axis produces no bending about the other prin-
b 0.5 8 0 4 00 32.0 0 I 0
c 0.8 0 -16 0 -12.8 0 204.8 0
cipal axiS, the principal axes are convenient d 0.4 0 0 0 0 0 0 0
axes to use when calculating internal shear flow Totallll I 2.7 16 .28.8 176.0 ! 460.8 -192
distribution.
A13.7 De!lec:tions.
Calculation of Centroid of Section: -
The deflection can be :ound by using the ZAx'
beam section properties about the neutral axis x =
lA = -2:7
28.8 ~
- 10.567"
tor the given plane of loading and the bending
moment resolved in a plane no~l to the neutral lAy' 12..- = 5.926'"
axis. The deflection can also be :ound by re- Y= lA = 2,7
solving the bending ~oment into the two prin-
Cipal planes and then using the properties about Ix ::: 176 2.7 x 5.926~ = 81.18
the principal axes. The resultant deflection is
the vector sum at the deflections in the direc- I y = 460.8 - 2.7 x 10.667~ = 153.58
tion ot the principal planes.
I XY = - 192 - (2.7 x - 10.667 x 5.926) = - 21.33

- \,,:~'.-->,~. _;_-, -/::.--~_~)_:34· ~~~::"-~-""-"_!'><:';~~::i.~_':',,~" ~"~::",";:-"1~',;-,; :'.,~,: ,~.: c~. :.•._I;,~.,,;·=.--c:· ~"'":: :)r"'~~ .:':~.~-J'~.~T;.-:"/:~~::~'1:.'~'i:~i,.'!._
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A13.5

Solution by Method 1 (bending about principal axes) (Positive moments produce compression in upper
The angle 0 between the x axis and the right band quadrant I.
principal axes is given by the equation,
Calculation at Stresses
2 I xy 2 (-21.331 - 42.66
tan 2 0 = ,,------'''i_ = 153.58 - 81.18· 72.40 = _ Mxp Yp Myp xp
I y - Ix
"D - - - . - - -
"lCP ryp
= - .589
Stringer a
2 .0 = 30° - 30' , 0_15°=15'
Xp =- 6.74, Yp = 4.45
I x p = Ix cos~ 0 + Iy 5in~ .0 - 2 I xy sin ¢ cos 0 hence
268700 x 4.4.5 - 156200 (- 6.74)
= 81.18 x .9648 a + 153.58 x .2636 a - (- 21.33 x °b = - 75.38 159.34

- .2636 x .9648)2 15900 - 6600 22500i/ln ••

= 75.56 + 10.67 - 10.85 = 75.38 in.~ Stringer b

I y p = Ix sin a .0 + I y cos a 0 + 2 I xy sin 0 cos .0 Xp . 9 . 75 W , yp.4.82

• 81.18 x .2636 a + 153.58 x .9647 a + 10.85 - 268700 x 4.82 - 156200 x 9.75

75.38 159.34
159.34 In. ~
I' IYp
; - 17180 + 9570 7610i/ln ••

I 1 Fig. AI3.3. Stringer c

! f-6.7·l_~ a I 9.75,,--..j
Xp 'II -

- 268700 x - 7.12
3.58, Yp 'II _ 7.12

- 156200 (- 3.58)
I /
Xp -..i.,. __
"10'

L'D.667" b
O'b 'II 75.38 159.34

X - .",... - -".1,-= ~ \- - - - X • 25400 - 3520 • 21880

:: II ·''iEl5'-t '.82 f
! 200 " 5.926 and similarly for stringer d, 0b = 21900.
c I ~ d .;--1""
3 I I 11.80 ~
.58
Ypt- y I y'
Solution by MethOd 2 (Neutral Axis Method)
Let g' • angle between yy axis and plane
Fig. A13.3a shows the location of the principal at loading.
axes, as well as the perpendluclar distances
from each corner member to each at the principal
axes. These distances are readily determined tan Q' =-300000
80000 = - .2667
from elementary trigonometry; hence
The external bending moment about the X and
Y axes equal J Q~ = 14° - 56' and Q = 14° - 56· + 15° - 15' ,.

~~ = 6000 x 50 : 300,000w~

My = - 1600 x 50 = - 80,OOOw~
Let a = angle be~Neen neutral axis NN and
the Xp axlS~
These ~oments will be resolved into bending
moments about the x p and yp principal axes. IXp tan Q - 75.38 (- 0.5816)
tan a • - ~'--,~-
159.34 - .275
Mx P • 300,000 x cos 15° - IS' - 80000 x
hence a = 15° - 24' (see Fig. A13.3b).
sin ISO - IS r = 289700 - 21000 = 268700 w• Since the angle be~Neen the X axis and the
neutral axis is only 9' , we can say
i1y =- 300000 x sin 15° - 15' 80000 x IN = Ix = 81.18.
p ResolVing the external bending moments
normal to neutral axis, we obtain
cos ISO 15' = - 79000 - 77200 = _ 156200 wi
A13.6 BEAM BENDffiG STRESSES

~ = 300000 X .9999 + 80000 x sin Oe - g' = K. = Iy/(Ixl y - I' xy) =

153.58/12016 = ,012794
300200"#. The bending stress ~t any point is
a
K~ - Ix/(IxI y - I XY) - 81.18/12016 .00674
given by, ::z

Stringer a Stringer a

x=-5.333", y = 6.074"
Yn = 6.074 + 5.33 x .0025 = 6.087"
300 200 x 6.087 _ 22500i/ln. 5I 0b - - [.00674 (- 80000) + .00177 x 300000J x -
81.15

Stringer b [.012794 x 300000 - .00177 (- 80000)J y

Yn = 2.055" hence 0b = (8) x - (3697)y;

- 300 200 x 2.05 7570 hence crb = 8 x - 5.33 - 3697 x 6.074 = - 22450
~
:II _
°b 81.18
Stringer b
Stringer c
x = 10.667, Y = 2.074
Yn = - 5.92
0b = 8 x 10.657 - (3697) 2.074 = 85 -7660 = -7575
Ob =- 300 200 (- 5.92)
81.18 = 21850
Stringer c
Stringer d x .. - 5.333, y=-S.926
Y = - 5.95 5.333 - (3697 x - 5.926)
300 200 (- 5.95)
°b = 81.18 = 22000 42 + 21900 - 21862

Stringer d
y x e 10.667, Y =- 5.926
Q' : 14°581
~::z 15015' 0b = 8 x 10.667 - (3697 X - 5.926) =
Q:30011'
plane of .I:0ading
~ 85 + 21900 = 21985
X 9' 52' N
-15°15' NOTE: The stresses Ob by the three methods
were calculated by 10" slide rUle. ~ence the
"p small discrepancy betNeen the results for tte
three methods.
Yp Fig. A13.3b
Y
Error in Stresses Due to Assumption that Section
Bends About X and Y Centroidal Axes Due to Mx
Solution by Method 3 (Method Using Properties and My.
About X and Y Axes)
(Equation 7a)
Mx =300000, My = - 80000

Ix .. 81.18, I y = 153.58, Stringer a

IXY = - 21.33 y = 6.074, x '; - 5.33

- 21.33 300000 x 6.074 - 80000 x - 5.33

= 81.18 x 153.58 _ 21.33 ll = °b = - 81.18 l53.S8 = - 25190

- 21.33 Stringer b
12016 = - .00177
y z 2.07, x = 10.567
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A13.7

Stringer b - contJd. Mx • 14260 x 50 • 713.000'~

300000 x 2.07 - 80000 x 10.667 I'ly = - 760 x 50 • - 38000'#

8l.18 153.58 • - 2110
These moments are resolved into bending
In like mar~er for stringers c and d moments about the prinCipal axes, as tallows: -
O'b
c
= 19130 and O'b
c
= 27440 Mx = 713000 x cos 0 - 38000 x sin ;1 = 700,OOO~~
P
Comparing these results with the previous I'lY
p
=- 71300 x sin 0 - 38000 x cos 0 '-140000'#
results it is noticed that considerable error
exists. Under these eroneous stresses the in- From equation (7), the general formula for 0b is:
ternal resisting ~oment does not equal the ex-
ternal bending moments about the X and Y axis.
Example Pr.ablem 2.
Fig. A13.4 shows a portion ot a cantilever
2-cell stressed skin wing box beam. In this ex- Stress on Stringer 1:
ample, the beam section is considered constant,
and the section is identical to that used in yp = 1.85 11 and Xp = - 17.86 11 (scaled from fUll
size drawing).

(700000 x l. 85) - 140000 x (- 17.86)

I
14260*
181.17 437
I I I I: II I'"\ ,I
11 I 1 I I I I
'I
11 II 11 7150 - 5700 12850~/sq. in.
I II 11761l1]1 I 1 ' I 11 "
I I I: II , I I I II ,1 11 Stress on Stringer 9:

"1
I! i II I I 1 I I II [I II
I! I' I I I I I I II I' II 50"
I; I: II I I I I I II II II Yp = 9.04", xp • 14.24"
',J..I II II! I I i 11 II II
III r- 4-, I1 I I I ~ I 1I II II
! III
1
, I-
II
-+-r -J.1r..-l-
1- Ir+- --+ i-I -r-t r- T1
II I I : I l i t I ,I II "b -- (700000 x ,).04) - 140000 x 14.24
111 181.1/ 437
III: II I I a
I'I[ II 3 4
;11 2 4560 ::. - 30340t/sq. in.
= - 34,900 +
I
X-.-_I Stress on Stringer 12:
B

700000 x (- 6.80) - 140000 x (- 8.22; =

Fig. A13.4 181.17 437

• 26200 + 2630 - + 23570i/sq. in.

?lg. A3.13 of Chapter A3 tor which the section SQlytion by Method. 2 (Neutral Axis Method)
properties were computed and are as follows: - In fig. Al3.5 let Q' be the angle between
the Y-Y axis and the Plane of the resultant
Ix = 186.6 In. I y = 431. 7 tn ." = 36.41 in ." bending moment. Resultant bending mement,
;1 = eO - 16.25" Ix • 18l.2 437
p M • v713000 3 + 38000 3 • 714600" lb.
The resultant air load on the wing outboard
of section ABC is 14260~ acting up in the Y di-
rection and 760i acting tcrNard in the X direc-
tan g' • - 38000
713000
.- .0533, hence 9' =- 3° _ 3'

tion, and the location of these resultant loads Let Q equal the angle between the plane of
is 50 from section ABC (Fig. A13.4). the resultant moment and the Yp axis.
;he bending stress intensity at the cen-
troids or stringers number (1), (9) end (12) Then a = 9' + 0 = 3° 3' + 8° - 16' 11<;1 - 19'
will be calculated uSing all t~llee methods.
From equation (8), the angle be~Neen the Xp axis
Solution by Method 1 (Bending about Principal Axes) and the neutral axis ~ Q, and
The bending moment at section ABC about the
X and Y axes is, I xp tan 9 181.17 x (- .200)
-tan a
I yp 437 =
Al3.S BEAM BENDING STRESSES

Y YO - 36.41
I;;JXY) = 186.46 x 431.7 - 36.41;;J
Plane of
Resultant
~',.
gl-!
f
Moment- I "
= - .00046
\
\
\ I '
36.41 ; J

= 431.7 = .C0545
793000

= 186.46 .0'02355
793000

Stress on Stringer 1:
Fig. Al3.5
Y1 = 4.39", Xl = - 1""'.41"
~ .083. Hence, a :;; 4 G - 45'

Fig. Al3.5 shows the relative positions of

the neutral axiS, principal axes, and plane of
loading. crb, = [.002355 X (- 38000) - (- .00C46 x (13000)J
The component of the external resultant
~~ment about the neutral axis N equals: - (-17.41 )

MN = 714060 x sin 83 G - 26' = 709350 in. lb. - [.00545 x 713000 - (- .00046 x - 3EOOO)J 4.39

IN = I x p cos~ a + I yp sin a a (238.5)(- 17.41) - (3868) 4.39

= 191.17 x .9966 a + 437 x .0878 01
183.37 tn ." 4150 - 17000 = - 12850#/in.;;J
Stress on Stringer 1: Stress on Str1~ger 9:

Yn~ = (Yp + x p tan a) cos a = (1.85 + 17.86 x Yg = 6.S9", Xg = l5.39"

.083) .9966 = 3.32'

0b~ = - [238.5J 15.39 = [3868J 6.39 :;;-30320;/in • .Ii
Mn Yn = - 709350 x 3.32 =_ 12850*/sq. In.
crb
,= - ---
In 183.37 Stress on Stringer 12:
Stress on Stringer 9: Y1a = - 5.55",
Yn = (9.04 - 14.24 x .083) .9966 = 7.84'
• crb .. =- [238.5J (- 9.11) - [3868J (- 5.55) =
• 23620i/ln.·

NOTE: In the three solutions, the distance

Stress on Stringer 12: from the axis in question to the stringers 1,
9, and 12 have been taken to the centroid of
Yn1.ll = (6.80 + 8.22 x .083) .9966 = 6.10 each stringer unit. Thus, the stresses ob-
tained are average axial stresses on the
0b = - 709~~.~7- 610 = 23600t/sq. in. stringers. If the ~imum stress is desired,
1a the arm should refer to the most remote part of
the stringer or the skin surface.
Solution by Method 3
JWproximate ...Method.
Mx = 713000"1, My = 38000'~ It is sometimes erroneously assumed that
the external bending moments Mx and ~y produce
Ix = 186.46 I y = 431. 7 Ixy = - 36.41 bending about the X and Y axes as though they
were neutral axes. To show the error of this
The constants K1 J K~, and Ka are first deter- assumption, the stresses will be computed for
tatnec - stringers 1 and 9.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A13.9

Stringer 1: ab

-TAr
.-
I
Yl = 4 + .39 = 4.39 r-TB 3"
2.64
-----, -L
I
Xl = - 33.15 + 15.74 - - 17.41
2" --t
...<--cB·2;64 1 3"
a b = - MxY CA' l
Ix
=- 713000 x 4.39 1- 38000 x - 17.41) = Fig. A13.8 Fig. A13.9
ab 186.46 431.7 stress intensity at bottom edge of portion A is
- 18330#:!i n :l .
ah 12 . 25 / 3 ) = 0.75 Ob.

Stringer 9: TA = CAr = Average stress times area

Y. a 6.89, Xg = 15.39
= - 713000 x 6.89 (- 38000 x 15.39)
ah 186.46 431. 7 The distance from the neutral axis to the
centroid of the trapizoidal stress loading on
- 25000#! l n :l . portion A is 2.64 inches. In like manner,

Example Problem 3.
TB = CB' = (0 +2 Ob) 3 x 0.25 = 0.375Ob
The previous example problems were solved
by substituting in the bending stress ~quatlons.
The student should solve bending strs's prob- The arm to TB is 0.667 x 3 = 2 inches.
lems by equating the internal r-es ts t Lig moment
at a beam section to the external bel.ding mom- For equilibrium of the beam free body, take
ent at the same section. To 111ustrlte Fig. moments about point (0) and equate to zero.
A13.6 shows a simply supported 10adfd beam. The
shear and bending moment diagram for the given ZMo = - 500 x 48 + 2.54 x TA + 2.64 CA' + 2 TB +

beam loading is also shown. Fig. A13.7 shews

the beam section which is constant along the 2 eB ' = a
span.
500# 1000 300 SUbstituting values as found for TA, T3 etc,
t .6°1 j 1'-2" -1
- 500 x 48 2.64 x 1.15 O'b + 2.64 x 1.15 O'b
A
T
+
1:1'
48" +-- 120"--r36'~
1110 690 + 2 x 0.375 crb + 2 x 0.375 crb = a
610';-_--, 8"

5~t---300 1 hence O'b = 24000

7.58 = 3170 pe f ,

Solution using Bending Stress Formula.

Fig. A13.7
°b = 11-I-e where 11 = 3"
- 24000
e = Moment
I = Neutral ofaxis.
inertia about

Calculat10n of r ,
For Portions A and A'

:or 'lie b Portions b, h",

1
I - 12 x 0.25 x 6' = 4.50 in" •

ITotal = 22.76 in" .

A13 10 BEAM BENDING STRESSES

henc e ternal force action from this nethod of solu-

tion. In solv:ng nonhomogeneous beams a~d
beams stressed above the elastic l~mit stress,
this ~ethod of solution often proves necessary
or advantageous because no simple beam bending
which checks first salut~on.
stress for.nula can be derived.
In Fig. Al3.1Z, let oQ be the intehsity on
Example Problem 4. the most remote fiber, or 3.09" above the neu-
Fig. A13.10 shows a loaded beam and Fig. tral axis. Table A Shows the calculations of
Al3.1l shows the cross-section of the bearn at the total stress on each of tte portions of the
section a-a'. aet ormme the magnitude of the cross-section and their noment about the neutra:
maxl~um bending stress at secti1n a-a' ~~der the
aXiS, all in te~ of the ·~own stress, cr~.
given beam load~~g. The beam section is unsym-
~etrical about the ~orizontal centroidal axis. TABLE A
Simple calculations locate the neutral axis as
I 2 3 4 5 6
shown in Fig. A13.ll.

-
12*/in.
Portion Area Average Total y : arm Resisting
' 1·1' I I II A Stress Load to N. A. moment
i~l~
I
5#/in. See in
, (2) (4) x (5)
Fig. terms x (3)
---I
n
I--- 100"---1. 30"-;:-'00" A13.13 of c b
-11-r-
I

Fig. A13. 10
J
RA
1
He i
-4-
-th-2"

1/2 _-<=
__
Jl-li rI I

2
1.045

1. 00
-.337crb

-.757
-.353crb

-.757
1. 39

2.35
- .490 O'b
~1. 780
I 'I I I 3 - 3' 1.00 -.837 ·.837 2.62 -2.190

1-27.2 in. 4 N- -I-f- N 3~'6" 4 1. 09 .350 .387 -1. 44 -0.557 i

3t 5 - 5' 0.56 .578 .323 .1. 93 I -0.622 ,I
I

(~t~I 2.9;..
6 1. 50 . 821crb 1. 2370b -2.57 -3.180 Cb

11 44
t=-3 Totals 6.19 O. OOOOb -8.82 O"b
~
1
Fi•. AI3.11 -.--

- , - I-- 2" -1 Explanation of Table A.

?ig. Al3.l2 shows a tree body of the por-
tion af the beam to the ri~~t of section a-a.
The bending stress intensity diagram is shown by
the horizontal arrows acting on the beam face at
section a-a. Fig. A13.l3 shows the cross-sec-
tion of the beam at sectien a-a.
The general procedure will be to determine
the total bending stress load on each portion,
1 to 6, of the cross-section and then the mom-
ent of each of these loads about the neutral
axiS, the summation of which must equal the ex-
ternal bending moment.
•• I I
_N~
2 91"
; ing moment of 36670"1.
Fig.AI3.13
,I
a - 36670
Fig. A13.12
Column 3 gives the average stress en each
of the 6 portions. For example, the stress on
portion (1) varies from zero at t he neutral
2.09
axis to 3.09 crb = .574 crb at the upper edge of
(1). ThUS, tne average stress _ .674 ab + 0 =
2
.337 abo On part 3 or 3' J the stress on the
lower edge· .674 crb and a b on the upper edge.
Then, the average stress = (1 + 2. 674 ) crb =
.337 abo Col~ 5 1s the distance from the
neutral axis to the centroid of the load on
each of the blockS. For portion (1) the stress
• • • I ,
pattern 1s triangular, and y = 2/3 x 2.09 =
\ 1.39". On portions (2), (3), (5) and (6) the
bending st~ess dlstr1b~tion 1s trapezoidal, and
_
the arm 1s to the centroid of this trapeZOid.
The total internal res~sting ~oment of
- 8.82 at ~rom Table A equals the external bend-
ThUS,
Cib = 8.82 = - 4l60#/in. a.

This method of solution involves ~ore cal- For equil18rium, the total co~pressive
culations than that required in SUbstituting in stresses on the cross-section of the beam ~ust
the bending stress fo~ula, however, the student be equal to the total t ens i Le stresses, or l.H
should obtain a better understanding of the in- must equal zero. Col~ 4 of Table A gives the
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A13.11

R
total load on each portion of the cross-section 1 1

T~
and the total of this column is zero. -- 21"- --..j 2"1-
T
The uending stress on the lower fiber of

,:. r
4.1
the cross-section 1s directly proportional to
3
the distance from the neutral axis or, ab1 ower = I Maple 3.33"

LI1j I

11.,l-NI·A.
hence
SOLUTION USING BENDING STRESS FOR:-mA.

- 36670 x 3.09
I N•A• = 27.2 in." 1Fig. Al3.14
rr I
1"
! ]" I

I
~teel
Fig. A13. 15
2.73"

22.3 x 1.5 = 33.4"

Fig. Al3. 16
16

crb (upper fiber) = 27.2 = - 4160 psi. Transformed Section

into Equivalent Spruce
which checks the above solution.
ant materials is: -
A13.9 Bending stresses in Beams with Non-Homogeneous
Sections, Stresses within the Elastic Ranges. Espruce = 1,300,000 psi.
In general, beams are usually made of one = 1,600,000 psi.
material, but special- cases often arise Where E"ap1e
two different materials are used to form a beam = 29,000,000 psi.
section. For example, a commerical airplane in Estee1
its lifetime often undergoes a number of changes
or modifications such as the addition of ad- SOLUTION:
ditional installations, fixed equipment, larger
engines, etc. Th1s increase in airplane weight The first step in the solution is to
increases the structural stresses and it often transform the reinforced beam section of Fig.
becomes necessary to strengthen various struct- A13.l5 into an equivalent beam sectIon composed
ural ~embers at the critical stress points. of the same material throughout. This 1s possi-
Since space limitations are usually critical, it ble, because the modulus of elasticity of each
often necessitates that the reinforcing ~terlal material gives us the measure of stiffness for
be stt:fer than ~he original material. Years that material. In this solution the reinforced
ago, when spruce wood was a common ~terial for beam will be transformed into a spruce beam
wing beams, it was normal practice to use reln- section as illustrated in Fig. Al3.l6.
~orcements at critical stress points of stifter
wood ~terial such as maple in order to cut dawn
the size of the reinforcements. In aluminum
Emp1e = 1:....6~0~0,:..~0~0~0 = 1. 23
0
Iepruc e 1,300,000
alloy beams, the use of heat-treated alloy steel
reinforcements are otten used because steel is
29000,000/10500,000 or 2.76 times stiffer than Esteel = 29,000,000 = 22.3
alu:ninum alloy. Espruce 1,300,000

Thus to transform the maple reinforcing strips

SOLUTION BY ~ OF TRANSFOFJ1ED
SEAM SECTION.
To illustrate how the stresses in a com-
posite beam can be ~ete~ined, ~NO example prob-
lems will be presented.
Example Problem 5.
Fig. Al3.1~ shows the original beam section
which is entirely spruce wood. ?ig. A13.l5
shows the reinforced or ~odlfied beam section.
~NO ears of maple wood have been added to :he
upper part as shown and a steel strap h~? been
added to the bottom face of the beam. The prob-
lem is to find the stress at the top and bottom
points on the beam section when the ~eam section
is subjected to an external Jend~ng ~oment of
60000 in. lbs. about a horizontal axis which
causes compression in the upper beam portion.
The modulus of elast:city (E) for the 3 dif~er-
in Fig. A13.l5 into spruce we increase the
width of each strip to 1.23 x 0.5 = 0.615 inches
as shown in Fig. A13.16. Likewise; to trans-
:O~ ehe steel :ein!orcing strip into spruce, we
~ke the Width equal to 22.3 x 1.5 = 33.4 inches
as shown in Fig. A13.16.
This transforned equivalent section 1s now
handled like any homogeneous beam section which
is stressed within the elastic :lm1t of the
~terials. ~he usual calculation Nould locate
the neutral axis as shown ar.d the ~oment of in-
ertia of the transformed section about ~he
neutral axis would give a value of 51.30 in.~.
Bending stress at upper edge of beam section: -
- 60000 x 3.33
51.30
=- 3900 psi.
Referring to Fig. A13.15, this stresS would be
A13.I2 BEAM BENDING STRESSES

the stress in the spruce sect:on. Since the ~e­ alloy. ~he original area of 0.1 sq. in. 9ach
intorcing strips are maple, the stress at the have been ~ultiplied by these st1f~ness ratio
top edge of these maple strips would be 1.23 values.
times (- 3900) = - 4880 pSi.
The bending stress at the lower edge of the
transformed beam section of Fig. A13.l6 would be: T
4.65
_
60000 x (- 2.73) -_ 3200

+-
i
°bt = 51.30 ' ps.

The stress in the steel reinforcing stra? Fig. A13. 18

thus equals 22.3 times 3200 = 71500 psi. Transformed Beam
Since all these stresses are below the y = 5.35" Section into Magnesium
elastic limit stress for the 3 materials the Alloy.
beam bending stress formula as used is
app.l Icabj e ,

Example Problem 6. The solution for the beam section of rig.

Fig. A13.17 shows an unsymmetrical beam A13.18 1s the same as for any other unsym-
section composed of tour stringers, a,b,c and metrical homogeneous beam section.
d ot equal area each and connected by a thin The first step is to locate the centroid
web. The web will be neglected in this example of this section and determine the moments of
problem. Each ot the stringers is made from inertia of this section about centroidal X and
different material as indicated on Fig. A13.l7. Y axes.
The beam section is subjected to the bending
moment Mx and My as indicated. Let it be re- Z Ay .446 x 10 + 0.1615 x 10
=
ZA =
y = 5.35"
quired to determine the stress and total load on 1.1365
each stringer in reSisting these applied extern-
al bending moments. ZAX x 6 + .431 x 4
I<---- 6" - - j
x = ZA = 0.1615 1.1385 = 2.365"

r
steel 10 • .10 • •
Ix • 0.6075 x 4.65 5:35 = 28.27 in. •
ra "",':rues. ,1
+ 0.531 x

I Steel • My'" 5000". I y • 0.546 x 2.365

•
+
•
0.165 x 3.635 + 0.431 x

10" x--L. M:x:'" 10000"*

1.635
•
= 6.34 in. •

1 MagneSium
c .1 Alloy .10
1 4
" ---j
Alum.Alloy
Area of Each
stringer :II) sq. in.
I xy = 0.446 (- 2.365) (4.65)
0.1615 x 3.635 x 4.65
0.10 (- 2.365)(- 5.35)
0.431 x 1.635 (- 5.35)
- - 4.90
= 2.73
= 1.27
- - 3.77
TOTAL - - 4.67 in. <6 =
Fig. A13.17
sowrrON; The bending stresses will be calculated
by USing method 3 at Art. A13.S.
Since the 4 stringers are made of different
materials we will tranSform all the materials From Equation (14) Art. A13.5 --
into an equivalent beam sect ron with all 4
stringers being magnesium alloy.

Emg. = 6,500,000 Esteel = 29 = 4.46 - 4.67

~ 6.5 28.27 x 6.34 - 4.67* =
Esteel =29,000,000
~.s. 25
Esta1n.steel= 28,000,000 Emg.
= 6.5 = 4.31 -1-~' 6 = - 0.0296
0' • 47

Salum.allOY = 10,500,000 Iy
~1\..ill1. = 10.5 = 1.615 Ixl y -
E",ag. 6.5
Ix = 28.27 = 0 1797
Using the ratio of stiffness values as indicated Ixl y - 157.4 •
above gives the transformed beam section of F1g.
A13.18 where all ~ter1al 1s now magnesi'~ 10,000 in. lb. My • 5000 in. lb.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A13.l3

Consider Stringer (c): - to see if they equal Zero.

x = 2.365, Y - 5.35 z p := 6 + 1167 162 1010 = 1 check.

Substituting in ab equation Forthermore the moments of the stringer loads
about the X and Y axes Toust equal the applied
Gc =- [0.1797 x 5000 - (- 0.0296)(- 100 00 )Jx external bending moments.
Take moments about a Y axis through
stringers (a) and (cl.
- [(0.0403)(- 100,00 ) - (- .0296 x 5000 )J Y
6 x 162 - 4 x 1010 ... 5000
(898 296 ) x - [(- 403 ) + 148 ] Y =- 12 In. lb.

a c • - 602 x + 255" y - - (18) Take moments about X axis through stringers

(e) and (0),

For stringer (c) x =- 2.365, y =- 5.3S Z Mx = 1162 x 10 - 162 x 10 - 10000

hence
=Oln.lb.
Gc = - 602 (- 2.365) + 255 (- 5.35)
(The calculations in this example being
= 1423 1360 = 63 ps t . done on a slide rule can not provide exact
checks) •
The total toad Pc in stringer (c) thus equals
0c (Area) = 63 x 0.1 = 6 Ibs. tension. A13.10 Bending Stresses of Homogeneous Beams
stressed above the Elastic Limit Stress Range.
Stringir (a): In structural airplane design, the applied
loads on the airplane must be taken by the
x = - 2.365, y = 4.65 structure without suffering permanent strain
which means the stresses should fall within the
tram equation (18) elastic range. The airplane structural deSign
loads which in general equal the applied loads
0a 602 (- 2.365) + 255 x 4.65 = 2613 pSi. times a factor of safety of 1.5 must be taken
by the structure without collapse or rupture
Pa = 2613 x 0.446 - 1167. lb. tension. with no restriction on permanent strain. Many
airplane structural beams will not fail until
the stresses are considerably above the elastiC
Since the true area of this stringer is 0.1 stress range for the beam material.
square the stress in this steel str~nger equals Since the stres~-strain relationship in
1167 /0.1 = l1670 psi. tension.
the inelastic range 1s not linear and also
since the stress-strain curve for a material in
Stringer (o ) : the inelastic range is not the same under ten-
sIle and compressive stresses (See Fig. A13.19),
x = 3.635, y = 4.65 the beam bending stress for~ulas as previously
derived do not apply since they were based on a
ab - 602 x 3.635 + 255 x 4.65 = - 1006 PSi. linear variation of stresS to strain. Experi-
~ental tests hcwever. have shown that even when
Pb = - 1006 x 0.1615 = - 162 1bs. stressed in the inelastic r~~ge, that plane
sections before bending remain plain after
True stress in stainless steel stringer after bending, thus strain deformation is still
= - 162 /0.1 = - 1628 pSi. linear which fact simplifies the ~roblem si~ce
the stress corresponding to a given strain can
Stringer (d): be found from a st~ess-strai~ curve for the
beam material.
x = 1.635, Y = - 5.35 A general method of approach to solving
beams trat are stressed above the elastic range
OJ =-"
602 x 1.635 + 255 t- 5.35) = - 2345 psi. can best be explained by the solution of a
problem.
x 0.431 1010 Ibs.
Example Problem 7.
- 1010 ?ortlon (al of Fig. A13.20 shows a solid
True stresS = 0.1 = - 10100 PSi. round bar made rr-on, 24ST aluminum alloy ma-
terial. Fig. A13.l9 shows a stress-strain
To check the results, check total str:~ger lo~ds curve for this mater~al. Let it be assumed
A13.14 BEAM BENDrNG STRESSES

that the ~imu~ failing ccmpressive stress
occurs at a s~raln of 0.01 in. per inch. The
problem 1s to deterwlne the ultimate reslstln~
moment developed by this round bar and then
compare the result with that obtained by uSin~
the beam bending stress for~ula based en linear
variation o~ stress to strain.
,--
g i '!.l .r
A13.19 shows that :or a given strai~ ~~ tr.e ~n­
elastic range the resulting tensile stress 1s
~1gher than the ~esultlng compressive stress,
and :urthe~ore from internal eqUilibrium the
total tensile stress on the cross-section must
equal the total compressive stress.
The problem as stated assumed that a
compressive unit strain of 0.01 caused failure.
Fig. b thus shows the strain picture on the
beam just before failure since plane sections
J remain plane after bending in the ~nelastic

~11_~.,; . _.. _.~.! -J-..-"'--.i,.-,- ;l.···-

range. Table A13.2 gives the detailed calcu-
. lations for determining the internal reslstl~g
30:~ "'i··'··'j· -., . :'...:. i- moment developed under the given strain con-
dition.
c,~· t' 1
.1···· ---;._-
;!!l.:..,..~§._ ·····_.····T··---;---··
i / . . ·~i- TABLE A13.2
',ri.'it '.·.,.···'... ! I 'i / I I 1 2 3 4 , 6 7

IrLT:~ i··',,·· I '! ~_ . ,. Strip

No.
Strip
Area y e
Unit
stress IF = oA
Res.
Moment I
.. ! · I : 1·/.: ...
~ ., ...._- "A" a 1M = Fr
wIO IX ,,, :; F'iu-. Al.tn
1 .058 0.935 • 00867 53000 3075 2760
2 .102 0.840 .00773 52500 5350 4300
3 .135 0.75 .00685 52100 7025 5040
4 .153 0.65 .00591 51500 7870 4820
s .165 0.55 .00494 51000 8410 4310
I 6
7
.180
.185
0.45
0.35
.00398
.00302
43000
33200
7740
6140
3200
1920
8 .195 0.25 . 00205 22800 4450 940
9 .197 0.i5 .00108 12500 2460 280
10 .200 0.05 .00012 3200 640 10
11 .200 -0.05 -.00084 - 7250 -1450 130
12 .197 -0.15 -.00181 _17800 _3510 660
13 .195 -0.25 -.00276 -29500 -5750 1650
14 .185 -0.35 -.00374 -35500 -6560 2540
15 .180 -0.45 -.00470 -40000 _7200 3510
(Q)
TCHSIO ...
Fig. A13.20
Te1oI$IL£
5Tltasa
16
17
.165
.153
-0.55
-0.65
-.00566
-.00663
-43000
-44800 .:~~~g 4170
4710
16 .135 -0.75 •• 00759 -46000 -6210 4880
19
20
.102
.058
-0.84
-0.935
-.00846
-.00937
-47200
-48000
I -4810
-2780
4210
2690
Total 3.140 I 740 56735
.. Col. 1 Rod diVided. into 20 strips . 1" thick:.
3 Y = distance from centerline to strip c. g.
(b \C.,,,. ~ (e) Col.

~.
Col. 4 e = strain at midpoint", (y - .0375)/103.75 !

Col. 5 Unit stress for e: strain from Fig. A13.19

co......,.,,·
IJ~JIf""
SOLUTION:
Since the stress-atrain diagram in tension
is different from that in compression (See Fig.
A13.19) the neutral axis will not coincide with
the centroidal axis ot the round bar rsgardless
of the fact that it 1s a symmetrical shape.
Thus. the method of solution is a trial and error
one since the location of the neutral axis can
not be solved for directly. In Fig. b of Fig.
A13.20 the neutral axis has been assumed 0.0375
inches above the centerline axis of the bar. It
Nas assumed toward the tension side because ob-
servation of the stress-strain curve of Fig.
Col. 6 Total stress on strip.
Col. 7 Moment about neutral axis. r=(y-.0375J [
Sf.au
The summation of column (6) should be
zero. Since ~he discrepancy is 740 lbs.~ it
~eans that the assumed POSition for the neutral
axis is a little too high, however the dis-
crepancy is ~egliglble. The total internal
r es ts t tng mcment is 56735 in. lbs. (Col. 7).
It we take a maximum unit c cmpress tve
strain of 0.01 we find the corresponding st~ess
from F'ig.A13.l9 to be 48500 ps t • I f this
stress is used as the failing stress in the
bea~ ~ormula M - cr I we obtain,
c
M = 48500 x 0.785 = 38000 In. lbs.
(0.785 = cI of round bar = ,
nT~/4)
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
Thus using the ~ame failing stress at :he
far extreme fiber the beam for~ula based :n
linear stress-strain relct~onship gives an ul-
timate bending strength of 38000 as compared to
a true strength of 56735 or only 67 percent as
::J.llch.
Fig. c of Pig. Al3.l0 shows the true stress
:istribution on the cross-section, which ex-
plains why the reSisting ~oment 1s higher than
Nhen a triangular distribution is used.
The problem ot the ultimata bending
strength of struct~al shapes is discussed in
ietail in Volk~e II.
A13.11 Curved Beams. Stresses Within the Elastic
Range.
The equations derived in the preVious art-
icles of this cr~pter Wer9 fer beams that were
straight. Thus in ?ig. ~13.2l, the element or
length (L) used in the derivation was constant
over the depth of the beam. The strain (6L/L)
was therefore directly ~roportior~l to 6L which
had a linear variation.
In a curved beam, the assumption that plane
sections remain plane after bending still
applies, however the beam se~ent of a curved
beam cannot have squal width over the depth of
the beam because of the curvature as illustrated
in Fig. A13.22, or in other words the length of
the segment is greater on the outside edge (L )
than on the inside edge (Ll). Thus in calcu-
I-
4L--! f-=-i
~-I--"ivl
~ I I,
Fig.A13. 21
Straight Beam
Stress
FIg. Al3. 22 Distribution
Curved Beam
lating the strain distri~ution over the beam
depth the change in length ~L at a point must be
divided by the segment Width at that point.
Thus even though plane sections remain plane
the strain distr1bution over the section will
not be linear. '!'he width of the segment at any
point is directly proportional to the radius of
curvature of the segment and thus the strain at
a point on the segment is inversely to the
radius ot curvature. This gives a hyperbolic
type of strain distribution as illustrated in
:ig. A13.22, and if the strains are within the
elastic limit the stress distribution will be
similar. The development of a beam :ormula
based on a hyperbolic stress distribution is
A13.15
given in most books on advanced engineering
mechanics and will not be repeated ~ere.
It is convenient however to express the
influence of the beam curvature in the for.n of
a correction factor K by which the stresses ob-
tained by the beam formula for straight beams
can be multiplied to obtain the tr~e stresses
for the curved beaIT~. Thus for a curved beam
the maximum stress can be calculated from the
equation
cr = K M c - - - - - (19)
I
Table A13.3 gives the value of K for
various beam section shapes and be&~ radius of
curvatures. The table shows that far only
rather sharp curvatures is the correction ap-
preciable. In general for airplane fuselage
rings on frames the curvature influence can be
neglected. However there are often fittings
and mechanical structural units in airplane
construction whose parts involve enough curva-
ture to make the influence on the stress of
primary importance. ~he concentration of
stress on t~0 inst1e Bd~e of a curved unit in
bending may In r Luence the rat tgue strength of
unit considerably, thus a consideration of the
possible influence of curvature should be a
regular ;art of design procedure.
In the inelastic or plastiC stress range,
the tn rjuence of beam curvature should be
consider~bly less since the stiffness ot a
material in the inelastic range is much less
than in the elastic stress range and changes
rather slowly as the stress increases.
A13. i2 Problems.
(1) Fig. A13.23 shows the cross-section of a
single cell beam with 12 stringers.
Assume the walls and webs are ineffective
in bend1ng. Calculate load in each
stringer by use of beam formula. Also
calculate stringer loads by equating in-
ternal resisting ~ament to external bend-
ing ~oment. Bach stringer area is same
and equals 0.1 sq. in. applied bending
moment Mx =-
100,000 in. lb.
, I I
L5"-r5"--.l....5"-S"_·_' 5.. -1..5.. -'
I, I
" _'
I _I _ I'
d
bee f
(2) Same as problem (1) but change external
bending moment to My = 200,000 in.lb.
(3) Fig. A13.24 shows a beam section with 4
stringers. Assume web and walls in-
ef!ectlve in bending. Stringer areas
A13.15 BEAM BENDlliG STRESSES

TABLE A13.3 TABLE A13.3 (continued)

VALUES OF K FOR USE ill THE VALUES OF K FOR USE IN THE

BEAM FORMULA o : ~ BEAM FORMULA a : Mc
I I

SECTION
R
C
FACTOR K
IN"",E
FmER
vUTSID,t,
FmER
.'
R SECTION R
c
FACTOR K
lliSIDE
FIBER
OUTSIDE
FffiER
'-R 1.2 3.41 0.54 0.224
I 1.2 3.55 0.67 0.409

Pi
1.4 2.40 0.50 0.151 1.4 2.48 0.72 0.292
1.6 1. 96 0.65 0.108 1.6 2.07 0.76 0.224
I.a 1. 75 0.68 0.084 1.8 1. 83 0.78 0.178
2.0 1. 62 0.71 o, 069 2.0 1. 69 0.80 0.144
2.0 1. 33 0.79 0.030 3.0 1. 38 0.86 0.867

~I
4.0 I. 22 0.84 0.016 4.0 1.26 0.89 0.038
6.0 1.14 0.89 0.0070 6.0 1. 15 0.92 0.018
6.0 1; 10 0.91 0.0039 8.0 1. 10 0.94
10. a 1. 08 0.93 0.0025 10.0 1. 08 0.95 ~·~~~lO
1.2 2.89 0.57 u•• o, 1.2 2.52 0.67 0.408

dJl
1.4 2.13 0.63 0.204 1.4 1. 90 0.71 0.285
~tI--41 l' f- I,
I
1.6 1. 79 0.67 0.149 1.6 1. 63 0.75 0.208
1.8 1. 63 0.70 0.112 I.a 1. 50 0.77 0.160
I
3T~
I 2.0 1. 52 0.73 0.090 2.0 1.41 0.79 0.127
3.0 1. 30 0.81 0.041
'L_: I 3.0 1. 23 0.86 0.058
I-R-j 4.0
6.0
1. 20
1.12
0.85
0.90
0.021
0.0093
.L
r...:-
,
'
R ----oj
4.0
6.0
1. 16 0.89 0.030
1.10 0.92 0.013
0.92 0.0052 I
6.0 1.09 8.0 1. 07 0.94
10.0 1.07 0.94 0.0033 10.0 1. 05 0.95 g.g~:~
1.2 3.01 0.04 2.37 o. 73
ib--.j I 1.4
1.6
2.18
1.87
0.60
0.65
~:~~
0.168
1.2
1.4
1.6
1. 79
1. 56
0.77
0.453
0.319

TtIJT
0.79 0.236
1.8 1. 69 0.58 0.128 1.8 1.44 0.81 O. 183

1 1i
2.0 1.58 0.71 0.102 2.0 1. 36 0.83 0.147
3.0 1.33 0.80 0.046 3.0 1.19 0.88 0.067
II 4.0 1. 23 0.84 0.024 4.0 1.13 0.91 0.036
6.0 1.13 0.88 O.all 8.0 1.08 0.94 0.016
~R--I 8.0
10.0
1.10
1. 08
0.91
0.91
0.0060
0.0039
8.0
10.0
1. 06
1. 05
0.95
0.96 g. ~~~~
1.2 3.09 0.58 0.336 1.2 3.28 0.58 0.269
1.4 2.2:5 0.62 0.229 1.4 2.31 0.64 0.182
-oj I
t=:JTIi
1-3b 1.6 1. 91 0.66 0.168 1.6 1.89 0.68 0.134
1.8 1. 73 0.70 0.128 1.8 1. 70 0.71 0.104
2.0 1.61 0.73 0.102 2.0 1. 57 0.73 0.083
b 1 I 3.0
4.0
1. 37
1.26
0.81
0.86
0.046
0.024
3.0
4.0
1. 31
1.21
0.81
0.85
0.038
0.020
--401---1 6.0 1.17 0.91 0.001 8.0 1. 13 0.90 0.0087
r-
R 8.0 1.13 0.94 0.0060 8.0 1. 10 0.92
g. g~~;
r 5b --I
10. a
1.2
1.4
1.11
3.14
2.29
0.95
0.5.
0.54
0.0039
O.,5.
0.243 r-
Trent I
41
-! I
10.0
1.2
1.4
1. 07
3.55
2.48
0.93
0.67
0.72
0.409
0.292

~e::Jl
1.8 1.93 0.62 0.1'19 1.6 2.07 0.76 0.224
I. a 1.'14 0.65 0.138 2
, 1. 8, 1. 83 0.78 0.178

, 1- ! I
2.0
3.0
1. 51
1. 34
0.68
0.76
0.110
0.050 4~
I I
t,
t/2
t
2.0
3.0
1. 69
1. 38
0.80
0.86
0.144
0.067
4.0 1. 24 0.82 0.028 4.0 1. 26 0.89 0.038
1.1 I'
t;-I 6.0
8.0
10.0
1.15
1.12
1.12
0.87
0.91
0.93
0.012
0.0050
0.0039
f j... C
I- R-1
~ I 6.0
8.0
10.0
1. 15
1. 10
1. 08
0.92
0.94
0.95
0.018
0.010
0.0065
1.2 3.63 0.58 0.41.
i<it.<!1- I 1.4
1.6
2.54
2.14
0.63
0.67
0.299
0.229
• e equals distance from centroidal axis to neutral axis.
T
~'=ill ., 1 I,
......JCf-
, 1.8
2.0
3.0
4.0
1.89
1. 73
1.41
1. 29
0.70
0.72
0.79
0.83
0.183
0.149
0.069
0.040
References: Wilson and Quereau. "A Simple Method ot
Determining stresses in Curved Flexural
Members"
"Advanced Strength of Materials", by Seely.
f.- R~ 6.0 1. 18 0.88 0.018
8.0 1.13 0.91 0.010
10.0 1.10 0.92 0.0055
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A13. 17

s~own in () on fi~Jre. ExterTIal applied (7) Fig. A13.27 shows 3 dif~erent Je~ sections.
Je~ding ~cments are; They are wade af aluminun alloy whose
stress-strain G~agram is the sa~e as t~at
Mx= -500,000 in. lb. a~d My = 200,000 in. lb. plotted in Fig. AI3.19. Determine the
Find stress on all four stringers by all ultimate internal resisting moment it the
three ~ethods which were explained in this naximum compressive strain is limited to
chapter. 0.01 in./in. Consider that upper portion
is in compression. Compare the results
1.2)
a obtained with formula M = 0bIlc, where crb =
compreSSive stress when unit strain is
(0.5)
"i -,--
5"
0.01.

(1.0) (0.4) '-..L

I • +-C dI Fig. Al3. 24 1-- 1" ---1
r" 12"---, 1-1"-1 1-1/21 :{y4
(4) The Zee section shown in Fig. A13.25 is
sUbjected to bending moments of Mx = 500
= ED 01 lJ !

1 1/4"
in.lb. and My 2000 in. lb. Find bending
stresses at pOints a,b,c,d.
I
-n'-
I-- 1"--1
Fig. Al3. 27 8 fu4
i

'~
314 .L
Lx
"L 1-;;'
L 15 c

... d
-Jl/4j-
--..L
Fig. A13. 25 500*
c
50011

2" - - - 1

(5 ) Fig. A13.26 shows a beam section composed

of three different materials. Find the
stress at top
and bottom
Magnesium
~1/8 '
l
paints on
beam due to a Alum. I 1/8 0.1" 0.1"
bending moment Alloy I
--I 1-0.4'4 ~
.... *" 5" -l ~.15"

~
MX= 80,000 In. lb. T
Et"'".ag.= 6.5x10 s 1/1/8 .4;~~O.I"
Esteel = 29 x lOS
Steel
-=-
r "T" 2'•..., 1/16 -L~ '0.1" L
0.6"

Sec. 1 Sec. 2. Sec. 3

EAlum.= 10.5x10<!l Fig. A1J. 26

(6 ) Fig. Ai3.26 shows (8) Fig. A13.28 shows a curved beam, carrying
a cross-section of a two 500 lb. loads. Find bending stresses
wood beam composed at points C and C", when beam cross-section
at 3 kinds of wood is made 3 different ways is indicated by
labeled A, 3 and CJ sections 1, 2 and 3. Use Table A13.3.
glued to~ther to
for.n a composite
beam. If the beam
is subjected to a
bending moment
Mx = 75000 in.lb.,
find intensity of
bending stress at Fig.AI3.26
tap edge of beam.
Also find total end load on ?ortions 3
and C.
A = Spruce. E = 1,300,000 ?si
3 = Maple. E = 1,500,000 psi
C ::: Fir. E = 1,600,000 psi
A13.18 BEAM BENDING STRESSES

DOUGLAS DC-8 AIRPLANE. Over-all view of the test wing section representing center wing section of DC-8.

c -_.I ___
a".

A ctoae-cp view of wing test section showing details of wing ribs, stringers, etc.
 CHAPTER A 14
BENDING SHEAR STRESSES - SOLID AND OPEN SECTIONS
SHEAR CENTER

Al4. 1 Introduction.
In Chapter A6, the shear stresses in a Load -idxi-
Fig. A14.2
member sUbjected to pure torsional forces were D' F'
, ,

¥
considered in detail. In Chapter A13, the sub- _ _ _ ryo __
N. ""----' ~l_'_,o -X
ject of bending stresses in a beam subjected to
pure bending was considered in considerable de- I "Co0l;
dA • ---"-'- Ct " , ,
tail. In practical structures however, i t , ,
seldom r~ppens that pure ~endlng forces (coup- Beam seC~ion t!===±=:!==,:===j
o F
les) are the loading forces on the bea~. The Fig. Al4.!
usual case 1s that bending ~oments on a beam cr c -jdxt"- cr '
are due to a transfer of external shear forces. ~
N'A'~'
JA~-:r::r
Thus bending of a beam usually involves both
bending (longitudinal tension and compression
stresses) and shear stresses.

'~lL/ 0 ~J,j~
C1tf
The same assumptions that were made in
Chapter A13 in deriving the bending stress
equations are l1kp.wlse used in deriving the
equations for flexural shear stresses. With
AU,
crt at
F\.
crt crt
c. .

flexural shear stresses existing, the assumption Fig. Al4.3 Fig. A14.4
that plane sections remain plane atter bending
1s not completely true. since the shearing is as indicated. The stress crt is greater than
strains cause the beam sections to slightly cr~because the bending moment due to the given
warp out at their plane when the beam bends. beam loading is greater at beam section DO' than
This warping action is usually referred to by at FF'. Now consider that this beam portion dx
the term nshear lag~. However, except in cases is further cut as indicated by the notch DeEF in
of beams with wide thin flanges, the error in- Fig. Al4.l, and this segment is shown in Fig.
troduced by neglecting shearing strains is quite Al4.4 as a free body with the torees as indicated.
s~all and therefore neglected in deriVing the Let crt = ~lmum tensile stress at a distance
basic flexural shear stress formula. The prob- c from the neutral axrs ,
lem of shear lag influence is considered in Then the stress at a distance y tram neutral
other chapters. axis is cry = at y/c.
The total load on an element of area dA of
AB.2 Shear Center. the beam cross-section (see Fig. Al4.2) thus
T~en a beam bends without tWisting, due to
equals C1ct / y dA •
some external load system, shearing stresses are
set up on the cross sections of the beam. The Now, referrin~ to Fig. Al4.4, the total
centroid of this internal shear force system is tensile load on each face of this segment will be
often referred to as the shear center for the calculated.
particular section. The reSUltant external
shear load at this section must pass through
the shear center of the section if twist of the Total load on face CD = °ctJC Y dA - - - - - (1)
section is to be prevented. ThUS, if the shear Yo
center is known, it 15 POSSible to represent the
external lead influences by eNO systems, one
that causes flexure and the other which causes
only t'Nist.
Total load en face FE = crt
c
.r Yo
c
y dA (2 )

AB.3 Derivation of Formula for Flexural Shear Stress.
2ig. A14.1 shows a loaded Simply supported
beam. :dhen the beam bends downward due to the
given loading. the beam portion above the neu-
tral axis is placed in compreSSion and that be-
low the neutral axis in tension. Consider a
short portion dX of the beam at ~oints OF on the
beam and treat it as a free body as shown in
Fig. Al4.3. The variation of tensile and com-
pressive stress on each face of the beam portion
Al4.l
From Chapter A13, the equation for flexural
stresS cr was derived, namely crt = Me/I. Let M
equal the bending moment at beam section DO' and
Ml tr~t at beam section FF' and let I and I' the
moment of the inertia of the cross-sectional area
about the neutral axis at these same beam sections
respectively. Then substituting value of crt in
equations (1) and (2) we can 'NTite,
Total load on face CD = ~JCy dA
- - - - - (3)
Yo
A14.2 BENDING SHEAR STRESSES SOUND AND OPEN SECTIONS SHEAR CENTER

Total load on race FE = l' M'JC ydA (4 )

Yo

Now let ~ b dx equal the shearing force on

race CE at the segment in Fig. Al4.4 where ~
equals the shearing stress and b dx the Shearing
area. For equilibrium of the segment, the total
forces parallel to x-x axis must be zero. It
the beam is at unitorm cross-section, which is
the case in our problem, then I = If and c and
Yo are the same in both equations (3) and (4).
Then the resultant horizontal force on the
sides at the segment equals the ditference of
the values in (3) and (4). or Fig. A14.5

Resultant horizontal load· M-I M' fC ydA. neutral axis are given on the tigure.
Yo SOLUTION: -

For equilibrium of segment in x-x direction, Shear stress at neutral axis x-x
Z Fx =- -M-
1 -M'
- JC ydA + ~ b dx = 0
y, "t"x_x • "0'
V J3.09
oY d.A

M M' JC dA
hence ~ • I ; dx - - - - - - - - - - (5)
y, Table A shows the calculation of' the term

However M ---ax
M' dll
= a:x=V=the external shear on
the beam section ..
3 . 09
[ , Y dA.

v
hence "t" ::: IE JC ydA (6 ) TABLE A. REFER TO FIG. a.
y,
PORTION Area d A Y ydA

It Is Important to nots that equation (6) 1 2.09 % 0.5 , 1. 045 1. 045 1. 096
applies only to beams ot unitorm section (con- 2 3 x 0.5 , 1. 50 2.34 3.510
stant moment at inertia). In airplane wing 3 0.5 x 0.5 , 0.25 2.84 0.710
structures the common case is f'or beams to vary 3' 0.5 x 0.5 :; 0.25 2.84 0.710
in cross-section or moment at inertia, and it
this variation is considerable, equation (6) SUM 6.026
should not be used and resort should be made to

~
equations (3) and (4). This tact Is Illustrated
in example problem 2. This matter at variable 2 3' ~}"h~
,
cross-sections is discussed later in this ,

chapter. Fig. a 1 Fig. b

., JY
! I

x-c-. -lyx x----.:

, I,

A14.4 Example Problems. Symmetrical Sections.

External Shear Loads Act Thru Shear Center. hence "t" ~ 850 x 6.026 ; 377 psi.
27.2 x 0.5
Example Problem 1.
Fig. A!4.5 shows the cross section at a Calculation ot shear stress at point 1-1, : -
beam symmetrical about the 1-1 axis. Assume
that a beam with this cross-section is Subjected
~o a loading which produces a shear load in the
1:'1_1
V f3.09y dA
= IxYb •
Y direction = to vy = 850 lb. Its location is
through the shear center of' the section which 2.09
lies on the centroidal y axis ot the beam sec- Fig. b shows the eftective areas in this
tion due to the symmetry ot the section about intregration. thus in Table AJ we leave out
this axis. Let it be reqUired to determine the portion 1, hence I y dA = 4.93.
shearing stress at the neutral axis x-x and at
pOints 1-1 and 2-2 ot the cross-section as
850 x 4.93
shown in Fig. Al4.5. The neutral axis location Substituting, ~l-l = 27.2 x 0.5 = 308 psi.
and moment at inertia of' the section about the
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A14.3

This Shearing stress is at a sect~on just SOLUTION: -

adjacent to portion (2). If i t was taken just
adjacent to portion (1), then the width b in the Bending Moments: -
eq~tion would be 3 inches instead of 0.5 inch 11132 = -600 x 132 = - 79,200"i
and the shearing stress would be 0.5/3 times that
shown above, or in other words the shearing 11175 = -600 x 175 = -l05,OOO"i
stress changes abruptly when the shear area 11218 = -600 x 218 = -130,500"*
cr~ges abruptly.
Table A14.1 shows the results of calculating the
Shearing stress at point 2-2 en cross-section: - bending stresses at 3 points on each side of the
Fig. c shows effective area, thus in Table neutral axis for the 3 stations. For example
A, the portions (1) and (2) are omitted and for station 132
3 . 09

1 y d.A equals 1.42.

2.59

850 x 1.42
"MAX. =~ (I1 Ie) =,- (79200 x 4)
38.67·

(tension at top edge and compression at lower

='_ 8180 pSi.

Substituting, ~2-2 = 27.2 x 1.0 = 44.4 psi. edges) .

(The width b = 0.5 + 0.5 for the two portions For a paint 1 inch from either edge of the
3,3' i. beam
The Shearing stresses as calculated act in the
plane of the beam cross-section in the Y di- a = +- (79200 x 3) = ± 6135
38.67
psi.
rection and also with the same intensity paralle
to the Z axis which is normal to the beam TAJ:ILE AB.1
section. z
Example Problem 2. VARIABLE 110l1Th'l' OF INERTIA.
Fig. A14.6 shows a cantilever beam loaded On Top tnt lI\t I
With a single load of 600 lb. at the end and BenQ1ng or 1" From 'I 2" From 1
Moment Bottom Top or Top or i p~,rt~~n ~,rt:~D i ~rt~~11
acting through the centroid of the beam cross- M fiber Bottom Bottom' A B r 'C
section. The beam section is constant between y-4",y-3" 1,"2,,1
stations a and 132, then it tapers uniformly to ,

the sections shown for stations 175 and 218. 132 79,200"f =.8180 :t6 U 5 +4090 7157i1 i H12' I 1023* I
;~:~ : ~~~: i
I 175 1I 105, OOO"fl .!oU80 +8135 +4090 1023 ,
The shear stress distribution on the beam cross-
section at stations 175 and 218 will be deter- i I
Z1. 130,800'" =.8180 +6135 t4090 1023 i

:nined.
From the results in Table Al4.l, it should be
'-43"~43"'- +--- 1 3 2 " - - - -
Fixed
!!-__;-__'-_-,
1- I -

~~i-----;;c:-~O;::+-;-;;;;-;;;:r-;o.,-------;;2
f"0. noticed that the change in moment of inertia be-
tween the three stations is directly porportlonal
to the change in bending moment, hence the same
Sta. '218 Sta:175 sci.. 132 Stl.O value for the bending stresses for all three sta-
tions. Columns 6, 7 and 8 give the total bending

~.::1
stress load on portions A, B and C of the three
cross-sections (see Fig. A14.6). These values
equal the average stress on the portions times
B the area of the portion; for example, for station
132, the load on portion A:

load = 6180 +
2
6135
x 1 x 1 7157 Jil, and for

portion C:

sts, 218 sta. 175 Sta. 132 load = 40902 + a x 2 x .25 - 1023*
I : 64.0 in. 4 I ::; 51. 33 in. 4 I: 38.67 in. 4
Fig. A14.7 shows the tension and compressive
stresses acting on a portion of the beam between

._lS4
 .
A14 4 BENDING SHEAR STRESSES SOUND AND OPEN SECTIONS SHEAR CENTER

stations 175 and 218. Fig. A14.8 shows the re- shear stresses between stations l32 and 175
sUIting horizontal shear stress pattern result- would be the same, since the change in bending
tng trom the loads in Fig. A14.7. For example, moment and moment of inertia have been ~de the
it we take a section along the beam 1" from the same as between stations 175 and 218.
tap or bottom edge ot the beam and treat this Figs. A14.9 and A14.10 show the shear
portion as a free body as shown in Fig. A14.12
applying ZH = 0, stress patterns i f the formula -r = V ~~ dA be
used for each station. The discrepancy is con-
ZH =- 7157+7157+'t'x43xl.O= 0, hence 't'" =0 slderable as the equation does nat apply to
beams of varying section.
~
613 -~.~
7157. X - 7157 * To illustrate the calculation by the shear
stress formula. the shear stress will be calcu-
09 ~'---B IB - -- _10224*
lated at the neutral axis for the beam sect~on

r
1023*
1- 1023• at station 175.
--- --- --- -

~a
.!.Qill. I--- 1023* V
't'=Tb oYdA

18180
Iv ._-B
··AI
-
I IB ---10224*
Ik:!--7157f
where f> cIA =1 x 1 x 3.5 -+ 2 x 1 x 2.5 + 2 x
~

;;
.;
Fig. Al4.7
-
~
~

~
0.25 x 1 • 9.0

hencs. -r = 600 x 9.0 = 420 ps t , as compared

Fig. A14. 8 Fig. Al4.10 Fig.A14.11 51.33 x 0.21;
to the true shear stress of 477 in FIg. A14.8.

~
409
II 20'
TABLE A14.2
,--J ('4'2
~
49:6 MAXIMUM SHEAR STRESS FOR SIMPLE SECTIONS
477*/0" 1r431 !.--420
LOCatiOD of
to.~·
1
C ross section Max. Shear Stress
Max. Shear Stress

~ ~
to 59.5
I I
True
Shear
Pattern
Sta 218 Sta 175
Shear Stresses
By Equation (6)
N-B-N I 3V
l' = 2A e• 0 I

43"
I
I I II
0 -il-1/4"
7157

-- --
15336
F sx43x.25
7157
10224 N-ON 4V
't' = 3A
I
II
I
e• 0

Fig. A14.13 ,

D7~
43"
~7 N-~N II -r = 4V(
I
3A 1'"
Dd)
na .. d. e• 0 ,

"'1"1- ====~--
F sx43xl
I
i
Fig. A14.12

Stmllarly, treating the portion between the

edge or the beam and a point 2" tram the edge as I
tN-E\i-N -r = 2A
3V
e='6
d

a tree body diagram as shown in Fig. A14.l3,

eL -'
!

ZH =- 7157 + 7157 ... 15336 + 10224 + " x 43 x

•25 = 0, therefore -r 477 psi • = 0
.t"-.~-~ l'
9V
= 8A e=8"
d

Obviously, the shear stress on portion C is con-

(j~e
stant, since the end load on this portion at I
both stations Is the same, or 1023*.
Fig. Al4.8 shows the general shape at the
N---- N 4V
't = 3A
I,
I
e • 0
shear stress d1stribution on the beam section at
any paint between stations 175 and 218. The I
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
AB.5 Maximum Shear Stresses for Simple Cross-sections.
Table Al4.2 gives the value of the maximum
shear stress on a few simple sections and where
it occurs on the cross section, (e is distance
from neutral axis to point of maximum shear
stress). V equals the shear load normal to the
neutral axis and it acts along the centerline
axiS, thus no twisting on the section. A is the
total cross-sectional area. The maximum shear
stress is given in terms of the average shear
stress which equals VIA'
AB. 6 Derivation of Flex\J.ral Shear Flow EquaUon.
Symmetrical Beam section.
To emphasize further the fundamental re-
lationships, a second derivation of the equation
for shear stress distribution will be presented.
Fig. A14.14 shows a portion of a cantilever beam
carrying a load P at the tree end as shown. _
This load is so located as to cause the beam to
bend in the XZ plane without twist about a Y
axis. The problem is to derive relationShips
which will give the magnitUde and sense of the
shear flow distribution on the cross-section at
the beam.
Fig. A14.15 shows a free body of a small
portion of the beam cut out from the upper
flange of the beam at points (a b) in Fig.
A14.14. Under the given external load P it is
obvious that the upper half' of the beam is sub-
jected to compressive stresses. In Fig. Al4.15,
Cy is larger than Cy since the cantilever bend-
ing moment is greater at station Y' •
Su. y ;-I""'_X-;--;
r,
&===~~~;E!J~~ St2.y /
Sta.y· .L
dy '~b
\Y
Sla.f' 1iJ1}--~d
x-
AI C'
Y nitude and sense for qy at any desired paint on
Fig. A14.14 Fig. A14.16 Fig.A14.15
The free edge of beam flange forms the
right side face of the element in Fig. A14.15
and thus the shear flow force on this face is
zero as indicated. The shear flow forces on the
1nternal or cut faces 1n Ibs. per inch are qy
and qx as indicated. Since the sense of these
shear torces is unknown, they will be assumed as
acting in the positive direction. (See Fig.
A14.l6 for positive sense of forces acting par-
allel to each of the coordinate axes XYZ.
Now consider the equilibrium of forces in
the Y direction for the element in Fig. A14.15
ZFy=o,or
A14.5
qYb dy "'" (C y - Cy) = 0, hence
qy =- ( Cy' - CY)
b dy
But (C y- Cy) =!1X I~ !1x J: z dA (See Art. A14.3)
hence qYb = - : t Ja
bz
dA
however;' = Vz , the external shear in the Z
direction.
hence qy = _ v z
b Ix Ja
b
z dA r - (7)
Equation (7) gives the Change in shear flow force
qy between points (a) and (b) and since in figure
A14.15 the value of qy at (al 15 zero because of
a tree surface, the value of qYb in equation (7)
is the true shear flow force in lbs. per inch at
point ("b). The stUdent should realize that
equation '(7) gives the shear flow q in the Y di-
rection. The minus sign in equation (7) means
that the pOSitive sense as assumed by the arrow-
head on qYb in Fig. A14~15 1s incorrect or should
be reversed.
The initial problem was to determine the
shear flow force system in the plane of the beam
cross-section or the ZX plane. From elementary
engineering mechanics, we know that if a shearing
stress occurs on one plane at a point in a body,
a Shearing stress ot the same intensity exists on
planes at right angles to the first plane, or in
general at a point,
(8 )
Before the shear flow in other planes is
completely defined its sense (pOSitive or nega-
tive) must be :<nown. Equation (7) gives the mag-
the cross-section. The question of the sense at
the associated qx and qz is easily determined trom
an observation involving equilibrium at moments.
This tact will be explained by referring to a
number of tree body diagrams.
Fig. A14.17 shows a free body of a small
element cut from the beam in Fig. A14.14 at point
(a) on the cross-section. The forces on this
free body are the compressive forces C~ and C on
the front and rear faces and the shear forces on
the various faces as indicated. The right side
face of the element Is a tree surface and thus q
on this face is zero.
The shear flow qy on the left side face is
calculated from equation (7) namely
 A14.6 bENDmG SHEAR STRESSES. SOUND AND OPEN SECTIONS. SHEAR CENTEH.

back side and through a point en the line of

C action of c' and C, the shear flow qz on the
r-7C~ I;>.,,I.-.!:......
;
I front face must act downward in order to balance
•, the moment due to qv .
-@ dy'l:r I q_# a
Fig. Al4.22 shows the results tor a free
CIt':';' I 1>:>'" I," x
.....t..
'-e-~=iI~ body of the lower flange plus a small web por-
c'
Flg.A14.21 "
,~di- r-:""~ tion. qy is POSitive from equation and thus qz
must be downward for mament equilibrium.
From the results obtained for these 6 dif-
C b t. a ferent locations, a simple rule can be stated
C' Fig.A14.l8 Flg.A14.l7 C'
relative to sense of shear flows in the plane of

""'~=01
'.t@;r
~
00(>:"1
I,

d--
I

qx •
--e
q
Z

qx
"
~
the cross-section, namely: -
If the calculated shear flow is directed toward the
boundary line between the two intersecting planes of the par-
ticular free body, then the shear flow on the other plane is
also directed toward the common boundary line, and con-
versely directed away if the calculated shear flow is directed
away.
Fig.A14.22
r Fig.Al4.19 T' Fig. A14.14 shows the sense of the shear
Fig.A14.20 flow pattern on the beam section as determined
for the given external loading.

A14.7 Shear Stresses and Shear Center for Beam Sections

For the given beam loading in Fig. Al4.l4 the with One Axis of Symmetry.
load P is up, theretore Vz has a positive sign. Example Problem. CHANNEL SECTION.
For ~~y portion at area (A) ot the cross-section Fig. Al4.23 shows a cantilever beam with
the distance z in the above equation is there- channel shaped cress-section carrying a 100 lb.
fore POSitive. Therefore in substituting in the downward load as shown. The problem is to de-
above equation qy cames out negative for any termine the lateral position of this load so that
point above neutral axis, and likewise for any the beam will bend without ~Nist. This posit:on
point on beam section below the neutral axis the will coincide With the lateral pOSition ot the
distance z would have a negative Sign and qy centroid of the shear flow system on the beam
would came out positive. Therefore the sense of cross-section-which holds the external load in
the calculated shear flow qy on the 13ft side of eqUilibrium Without twisting of the beam section.
the element in Fig. Al4.l7 is negative or as The cantilever beam has been cut at a section
indicated by the arrow on the force vector. Now abc d (Fig. A14.23) which is tar enough rr-on the
to find the sense of the shear flow qx on the fixed end or the beam (not Shown) so that the
front face take moments about a Z axis acting in etfects of beam end restraint against section
the plane of the rear race and through pOint (0) warping can be neglected. In Fig. A14.23, the
internal forces holding the beam in eqUilibrium
Which is on the line of action of the forces C'
and C. Only two forces have moments, namely the are sketched in. They consist of a longitudinal
side shear force qy dy and the front tace shear stress system of tension and compreSSion and
variable shear flow system in the plane of the
force (qx!2)dx. It is obvious by observation cross-section. In this problem we are only con-
that qx on front tace must act to the lett as sidered with the internal reSisting shear flow
shown it the moment is to equal zero. The total system.
shear force on tront face is (qx/2)dx because For solution of thiS problem the moment of
shear flow at right edge 1s zero and it varies inertia Ix must be known. If calculated it
linearly to qx at left edge at front face. would be Ix • 0.2667 in·.
Figs. Al4.l8, 19 and 20 show free bodies ot
elements taken at the other three corners of the SOLUTION: - From equation (7)
beam section whiCh are labeled 0, d, e in FIg.
Al4.l4. For all tree surfaces q is zero. The
sense of qy as before 15 given by the equation
as explained betore, hence qy is negative in
.- Vz Z z
Ix
A - - (8)

Fig. A14.18 and POSitive in tl~~es 19 and 20.

A Simple consideration of moment eqUilibrium as we know that the shear stress Is zero at a free
explained tor Fig. A14.l7 gives the sense of the edge, thus the solution at equation (8) is
shear flow Qx as shown in the 3 figures. started at either paints (a) or (d).
The shear Vz is - 100 lb. Thus equation
Fig. A14.21 shows a free body of the entire (8) for our problem reduces to,
upper beam flange and a short portion of the
beam web. qy from the equation is negative and - 100 __
this acts as shown ~n the figure. Now if we qy =- 0.2667 Z z A 375 Z z A - - - - - - - (9)
take moments about a X axis in the plane of the

s.u
 j."::'--';::;;;'_"""i.;H~_

""= .""

ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AB.7

100 lb. boundary line as shewn in Fig. Al4.24.

Cammon sense tells us that the resisting
Free end shear flow qz on the channel web must be directed
~b-,
upward because it is the only force system that
q: 563
can balance the 100 lb. load as far as ZFz = 0
is concerned.
In general the shear floW is continuous
around the section and only reverses when it
passes through zero which only happens in closed
tubular sections. In general, it is possible in
most cases by observation only, to determine the
Shear Flow Diagram sense of the shear flow at some one point on the
c Fig. A14. 25 beam cross-section. The shear flow being like a
"-,.---,.. flow of liqUid will continue in the same general
direction along the center line at the parts that
make up the beam section.
The small arrows on the beam section at
Fig. AI4.23 show the sense of the shear flow
pattern over the beam section. In Fig. Al4.25,
Fig. A14.23 Fig. A14. 24 C. Qy the shear tlow values as calculated at the var-
ious paints are plotted to fonn a shear flow
diagram for the beam section. Between points a
We will start at point (a) in solving equa- and b or d and c, the arm z in equation (9) is
tion (9) and proceed around the section. constant and thus q varies linearly as plotted.
Point (a) q
Ya = a (free surface) Between b and 0 or 0 and c the arm z changes and
the area is also a function of z, thus q varies
Point (b) Z = - 1, A = area between (a) and (b). as za or parabol1c as plotted.
The initial problem was to locate the
qy =qy +375Z~ZA centroid of this final shear flow system which
b a is generally referred to as the shear center.
q = a + 375 (- 1)(1 x 0.1) • - 37.5 lb/in. In Fig. A14.25, Qab' Qbc and Qed represent the
Yb resultant of the shear flow force system on
these three portions of the beam cross-section.
Point (0) on X axis. Each torce 15 equal in magm tude to the area of
the shear flow diagram tor the particular beam
qy
a
- 37.5 + 375 Zb z A section portion. Hence,

= 18.75
= - 37.5 + 375 (- 0.5)(1 x 0.1) = - 56.3 lb./in.
~b = 1 x 37 .5/2
. lb.

'"be • 2 x 37.5 + (56.3 - 37.5) x 2 x 2/3 = 100 lb.

Point (c).
«Cd = 1 x 37.5/2 = 18.75 lb.
qy = -56.3+375 (0.5)(1 x 0.1) =-37.51b./in.
c The resultant R at these three shear forces
will now be determined.
Point (d).
ZFz=lOOlb., Z Fx = 18.75 - 18.75 = a
qy =-37.5+375 (1) (1 x 0.1) = 0 (tree surface)
d Hence
~e know that. the intensity of shear flow in the
ZX plane at any point equals that in the Y di-
R :: (z F z " + Z Fx 2) J. / a = (100" + 0)J./2 = 100 lb.
rection at the same paint. Thus qx or qz equal
The ~oment of the resultant about any point such
the qy values above. The sense of the qx and qz as (b) in Fig. Al4.17 must equal the moment of
shear flows must be known before they are com- the shear flow forc~ system about point (b).
pletely defined or known. Fig. Al4.24 shows a Let e be distance from point b to line of act10n
tree bOdy at a small element at point (a) on the at the resultant R.
end of the lower beam flange. For any point be-
low the X centroidal axis equation (9) will give Hence R e :: Z 11 b (of shear now system) or
a minus sign ror Gy. Thus in Fig. Al4.24 qy
acts as shown, or directed toward the boundarJ 100 e = 18.75 x 2 or e 0.375 inches.
line be~Neen the side face and the front face.
Then by the simple rule as given in the previous Thus the centroid of the internal shear re-
article qx is also directed toward this cammon Sisting force system lies on a vertical line
A14 8 BENDING SHEAR STRESSES SOUND AND OPEN SECTIONS SHEAR CENTER.

0.375 inches to left of point b as shown in Fig. Method 3. The k ~ethod.

AI4.17. For bending about the centroidal Z axis
without twist the resultant of the internal
shear flow system would obviouSly, due to sym-
metry of section about X centroidal axiS, lie on z z A (14
the X axis, hence shear center for the given
channel section is at point 0' in Fig. Al4.l7. EXPJ1PLE DROBLE1 USING THE T:ffiEE DIFFERENl'
The external load of 100 lb. would have to be =HODS.
located 0.375 inches to the left of the center-
line of the channel web It bending of the Fig. A14.26 shows a Zee Section subjected
channel without twist 15 to occur. to a 10,000 lb. shear load acting through the
shear center of the section and in the direction
AB.8 Shear Stresses for Unsymmetrical Beam Sections. as shown. The problem will be to calculate the
In chapter Al3, which dealt with bending shear flow qy at two pOints on the beam section
stresses in beams, three methods were presented namely paints b and c as indicated on the figure.
tor determining the bending stresses in beams The shear flow at these ~NO places will be cal-
with unsymmetrical beam sections. The bending culated by all 3 methods.
stress equations for these three methods will be Since all 3 methods require the use at
repeated here: - beam section properties and since the direction
ot either the principal axes or the neutral
Method 1. The Principal Axis rretnoo . axis are unknown the z
tirst step in the
a _ solution regardless ot i I
"4
(9) which method is used
Is to calculate the .1"
~dl ,tl
1b~
section properties 2
Method 2. The Neutral Axis Method. about centroidal X and
Z axes. Table A14.3 x-- -c r-r-x 3"
gives the calculations.

A~I"l
a = - - - - - - - - - - - - - - - (10) The section has been
divided into 4 portions
labeled 1, 2, 3 and 4.
,Method 3. The Method USing section properties V = 10000 lb. . 4'J: ..L-
about centroidal Z and X axes. For k :1"
brevity this method will be called
the k method. Fig.AB.26

a = - (k. Hz - k, !'lx) x - (k. !'lx - k, Hz) z (11) TAi:lLE A14.3

where, Portion M"

IA)
Mm
, Arm
x
1
Ax> ",'
i Ax" ! " I 'x
I i
1 0.10 1. 4'5 ·0. 4"J •. 06525 I.
.21025 I .02025 i .00838 000017
2 10.14 0.70 0 : 0 .068,601 0 . 000117. 02287
a '0.141 ·0.701 0 0 .06860 0 1·000117.02~87 ,
4 0.10 .1.451 0.4:)1·.06525 .21025 .020251.008381.000017 I

Z ·.1305 .55770 .040:)0 .01690 .04577 ,

k Iz
• = Ix I z - I'
xz IX' = Z A + Z i x = .6035 in",
Zl1

In referring back to the derivations ot Iz = Z A Xlii + Z i z = .0574

equations (5), (6) and (7) the above equations I xz = Z A xz = - .1305
(9), (10) and (11) can be written in terms ot
beam external shears instead ot external bending (Note: In Table A14.3 t x and i z are the mom-
moments as tallows: - ents ot inertia of each portion about
Method 1. The Principal Axis Method. its awn centroidal axis).
VZp VXp
SOLUTICN BY PRINCIPAL AXES METHOD. (Method 1)
q y : _ - l : zp A - - l: x p A - - - - - - (12)
I xp I zp Let ~ be angle between Principal axes and
the X and Z axes. From chapter A13,
The Neutral Axis Method.
tan 2 ~ = 2 I xz
I z - Ix
l: Zn A - - - - - - - - - - - - - (13)
= 2 (- 0.1305) _
0.0574 _ 0.6035 - 0.477S
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES Al4.9

hence 2 ¢ = 2S 32.2'
0
- or $ = 12° _ 46.1' Now SUbstituting in eqUation (15)
sin ¢ = 0.2210 and cos ¢ = 0.97527 .
The moments of inertia about the principal axes qb = - 0.~16 (1 x 0.1 x 1.5136)
can now be calculated.
Ix = Ix cos· ¢+ I z sin· ~ - 2 I xz sin ¢ cos ~ 6792
- 0.02782 (1 x 0.1 ) ( - 0.1184) = - 1756 + 2890
p
,
= 0.6035 x .97527 + .0574 x .2210 - 2 (-0.1305)
--'
= 1134 Ib/in.
x .97527 x .2210 = .63316 in'. Calculation of shear flow at point (c).
I z := Ix sin' ¢ + I z cos' ~ + 2 I xz sin ~ cos~
p
,
= 0.6035 x 0.2210 + .0574 x ~ + 2 (-0.1305)
. For portion (2) area A = 1.4 x 0.1
zp =
.70 x .97527 0.6825 in.
xp = .70 x .2210 = 0.1547
=
0.14

x .97527 x 0.2210 = 0.02782 in~. The shear flow at paint (c) equals the
The equation for shear flow q is, shear flow at point (b) plus the effect of the
portion (2) between points (b) and (c)., hence
- - (15) 7348
~ = 1134 - 0.63316 (0.14 x 0.6825) -

In Fig. A14.26 the external shear load is

10000 lbs. acting in a direction as shown. Re-
0.~2~~2 (O.lA x 0.1547) = 1134 - 1109 - 5285 =
solVing this shear load lnto z and x components,
we obtain, - 5260 Ib./in.

Vz • 10000 x cos 30° • 8667 lb. The shear stresses at these two points (b)
and (c) would equal q/t = 1134/0.1 and - 5260/0.1
Vx = 10000 x sin 30 0
• 5000 lb. or 11340 psi and - 52600 psi respectively,
Resolving these z and x components further SOLUTION BY NEUTRAL AXIS METHOD. (Method 2)
into components along the principal axes we In thiS solution n~
obtain,
it is necessary to find \ i
= 8667 x .97527 - 50000 x .2210 =7348 lb. the neutral axis for
the given external load- : .O~~6" \1
iag. In Fig. A14.28,
=8667 x .2210 + 5000 x .97527 = 6792 lb. the angle Q is the angle zn: .186

Calculation at shear flow at point (b). (See

between the plane of
loading and the zp pr tn-
I
~ """."\."'
Fig. AI4.26). Cipal axt e , and this .g
Fig. Al4.27 shows the pasi~ion of the angle Ii) equals :30 e + 12 0 .l12
0 . 4.61
Fig. A14.28
principal axes as calculated. The shear flow - 46" = 42° - 46". 15Q.,2S' '
at the free edge of the upper portion (1) is 'pn
Let a equal angle between xp
zero. For the shear flow at point (b), the principal axis and neutral axis n-n.
area to be used in the summations Z zoA and From chapter A13, we find,
Z xpA Is the area of element (1). The arms zp
Ix tan Q
and x p can be calculated by Simple trigonometry. tan a = - -"-,p,---
Fig. A14.27 shows the value of these distances,
namely xp = - 0.1184 and zp = 1.5136.
= _ 0.63316 x 0.9245 = - 21.052
0.02782

Whence. a = - 87" - 17'< (See Fig. Al4.28 for

location of neutral axls.

sin a = 0.9989 , cos a = 0.04742

In = Ix cos 2 a + I z sin 2 a, substituting,

p p
Fig. A14.27

("""~
( '.) r
A14.l0 BENDING SHEAR STRESSES. SOUND AND OPEN SECTIONS. SHEAR CENTER.

In:: 0.63'316 x .04742

•
+ .02782 x ~
• = .02919 q = - [34.25 x 5000 - (- 7.406 x 8667)J Z xA -

The component or the given external shear load [3.257 x 8667 - (- 7.406 x 5000)J L z A, whence
normal to the neutral axis n-n equals
Vn • 10000 x sin 45° - 29' • 7130 lb. q - - 235438 Z x A - 65258 Z z A

Shear !low at point (b)

From equation (13)
For portion (1), x = - 0.45 in., z : :; 1.45 in.
q : - Vn Z Zn A
A :::; 1 x 0.1 :::; .1
In
SUbstituting,
Shear tlow at point (b): -
The distance from the neutral axis to the qb =- 235438 (- 0.45) 0.1 - 65258 x 1.45 x 0.1
centroid or portion (1) equal Zn • - 0.0466 in.
hence = 10597 - 9462 = 1135 lb./in.

7130 Point (c)

qb = - :029I§ (1 x 0.1) (- 0.0466) = 1135 1b/in.
For portion (2), x:::; 0, Z :::; 0.7, A :::; 1.4 x .1
= .14
Shear r10w at point (c): - Zn = 0.1868 In.
Clc = 1135 - 235438 (0) .14 - 65258 x 0.7 x 0.1
Clc =1135 - .~~9 (1.4 x 0.1) .1868 = 1135 - 0 - 6395 = - 5260 lb./in.

= 1135 - 6390 =- 5255 1b./in. General Comments.

SOLl1l'ION BY l'lE:l'HOD 3 - (The k Method).
In this method, anl) the section properties
about the centroldal X and Z axes are needed.
These properties as previously calculated in
Table A14.3 are,
Ix =0.5035, Iz :: 0.0574,
The shear tlow equation tram sq. (14) IS,
q = - (k. Vx - k, Vz ) Z xA - (k. Vz - k, Vx )
Z z A - -
The author prefers solution method number 3
slnce it avoids the calculation of additional
angles and section properties as required 1n
methods 1 and 2. Furthermore, in calculating the
shears and moments on the airplane Wing, fuse-
lage and other major sturctural units it Is con-
venient to refer these shears and moments to the
conventional X Y Z axes, and thus these values
I", =- 0.1305 can be used in method 3 without further resolu-
tion. From an investigation of many airplane
stress analysts reports, it appears that the en-
gineers at most airplane companies preter to use
method 3.
- - --- - - - - - (16) A14.9 Beams with Constant Shear Flow Webs.
F1g. A14.29 shows a z

~ITe
beam composed ot heavy
tlange members and a curved
thin web. For bending
=- •01762
0.1305
- - 7.406 about the X-X axiS, the

Iz 0.0574
web on the compreSSive,
side ot the beam absorbs
x--:-~
~ i I
.
1<0 = Ix Iz - Irl = .01762 =3.257 very little compressive
stress, since buckling ot q ' I ,,~

the web will take place .......Flange

Ix 0.6035 under low stresses, par- Z Fig. A14.29
1<0 = Ix I z - I xz• =0.01762 = 34.25 tlcularly when the curvature of the web Is
small. on the tension Side, the web will be
Resolving the external shear load ot more eftective, but if the flange areas are
10,000 into x and z components, we obtain, relatively large, the proportion of the total
bending tensile stress carried by the web is
Vx = 10000 sin 30' = 5000 lb. small as compared to that carried by the tension
flange. Thus for beams composed of indiVidual
Vz = 10000 cos 30 0 :: 8667 lb. flange members connected by thin webs it is
orten assumed that the flanges develop the en-
Substituting values of Vx ' Vz and k values in tire longitudinal bending resistance w~ic~
equation (16) we obtain - therefore means that the shear flow is constant
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A14.ll

over a particular web. In other words in the EXA!1PLE PROBLEM - RESULTANT OF A CONSTANr FLOW
FORCE SYSTEM.
shear flow equation q : Vz I z d A, if the area
Ix Fig. A14.30 shows a constant flow force
of the web is neglected then q is constant be- system thru points ABC o E with q • 10 lb.
tween flange members. per inch. The resultant or this force system
is required.
RESULTANr OF CONSTANT SHEAR FLOW FORCE SYSTEm
Fig. A14.Z9 shows a beam assumed to be
car-rying a downward shear load (not shown) and
to cause, bending about axis x-x without twist.
Assuming the two flanges develop the entire
bending reSistance, the shear flow q is con-
stant on the web and acts upward along the web
to balance the assumed external downward load.
The resultant of this reSisting shear flow
force system will give the lateral position of
the shear center for this beam sectlon. The
problem then is to find the reSUltant ot the SOLUTION: -
shear flow system.
Let q = load per inch along web (constant). Draw clOSing line between the beginning
and end points ot ror-ce system. (line AE). The
Let R = resultant of the q torce system. length h ot thiS closing line is ZO incheS.
From elementary mechancis, From eq, (17) R = q h = 10 x 20 = 200 lb.
The direction of the resultant is parallel
R =.jz qx -+- Z qz ,where qx and qyare the x to line AE or horizontal In this problem. To
tlnd the location of R take moments about any
and y components of the q forces along the web. point such as (0). D~ lines trom point (0) to
SiLce q is constant, Z qx is zero, hence, points A and E. The enclosed area (A) equals
(17) s x ic- s x io- .5nxS· -+-lOx 50 = 189.3 sq. in.

Equation (17) states that the magnitude ot From eq, (19)

the resultant of a constant flow torce system e • 2 :a Z x zlg9.3 • 18.93 In.
is equal to the shear flow q times the straight hA
line distance between the two endS of the shear
flow system. Fig. A14.30 showS the resultant at ZOO lb.
Since Z qx is zero, the direction of the acting at a distance e from (0) and parallel to
resultant 15 parallel to the straight line line AE.
JOining the ends of the web.
The location of the resultant torce is A14.10 Example Problems for Beams with Constant Shear
found by using the principle at moments, namely, Flows Between Flange Members.
that the moment or the resultant about any
pOint must equal the moment or the original EXAMPLE PROBLEM 1. Beam Section Symmetrical
force system about the same point. In Fig. About One Axis.
Al4.29 assume point (0) as a moment center. Fig. A14.31 showS an open beam section com-
ThenRe=qLr posed ot 8 flange members connected by thin
butR~qh z
hence e _ q L r _ L r (18)
-(f1l-T

In equation (18) the term L r is equal to

~Nice the area (A), where area (A) Is the en-
closed area tormed by drawing straight lines x
trom moment center (0) to the ends of the shear
flow force system. Thus
2 A (19)
e:"11 - - - -

The shear center thus 11es at a distance e

to the left of point (0), and the external shear
load would have to act through this point if Fig. AB.31
~Nistlng were to be eliminated.
 A14.12 BENDING SHEAR STRESSES. SOUND AND OPEN SECTIONS. SHEAR CENTER.

sheet to form the webs and walls. The flange in order to indicate at what point the shear
members are numbered a to h and the areas of flow ~ is being calculated.
each are given on the figure. It will be as-
sumed that the webs and walls develop no bending Qab = - 2.5 Za z A = - 2.5 x 5 x 0.1 =
resistance and thus the shear flow between ad-
jacent flange members will be constant. The = 1.25 Ib./in.
problem is to determine the shear center for the
beam section. The first letter of the subscript refers
to the flange member where the shear flow q is
SOLUTION: - being calculated and the second letter indicates
8n which adjacent side of the particular flange
Since the beam section is symmetrical member. Hence qab means the shear flow at
about the X axis, the centroidal X and Z axes flange (a) but on the side toward (b).
are also principal axes, since the product of
inertia I xz is zero. Qab = qba = - 1.25 (Since no additional flange
The vertical pOSition of the beam section area is added, and thus
centroid due to symmetry is midway between the shear flow is constant on
upper and lower flanges. sheet abo
To find the horizontal pOSition of the
centroid; take moments of the flange areas 'lbc = qba - 2.5 "b Z A
about the left end or line be: - - 1.25 - 2.~ x 5 x 0.4 - 6.25 lb./ln.
x= lo A x = 0.4 x 15 + 0.2 x 10 ? 0.2 x 5 = - 6.25
ZA 1.6 Sco - 2.5 !c z A
= 5.625 In. = - 6.25 - 2.5 x (-5xO.4)= - 1.25

The moments of inertia tor the section about

= 'led =- 1.25
the centroidal x and y axes are: - 'ldc 2.5"d z A 1.25 - 2.5 (- 5 x 0.1)
= 0
qed = 'lde = 0

I z = 0.8 x 5.625
2
? 0.2 x 0.625
•
? 0.2 x
qer = 0 - 2.5 "e z A =o - 2.5 (- 5 x 0.1) = 1.25
• qre = qer .. 1.25
4.375 ? 0.4 x 9.375- = 64.4 In~.
qrg = 1.2: - 2.5 "r z A = 1.25 - 2.5
HORIZONTAL PCSITION OF SHEAR CENTER: - t: 5 x 0.2) = 3.75
The horizontal pOSition of the shear center
will coincide with the centroid ot the shear
%r = qrg = 3.75
flow system due to bending about axis xx with- qgh = 3.75 - 2.5 "g z A = 3.75-2.5x5 x 0.2 =
out ~Nist. For simplicity, to elimir~te large 1.25
decimal values tor shear flow values an ex-
ternal shear load Vz =
100 lb. will be assumed
qhg = qgh = 1.25
and the internal reSisting shear flow system qha = 1.25 - 2.5 loh z A = 1.25 2.5 x 5 x 0.1 =
will be calculated for this external loading. 0 (checks free edge at h).
From equation (8)
The Sign or sense of each shear flow is for
the shear flow in the y direction as explained
SUbstituting values at Vz and in the derivations of the shear flow equations.
Ix The procedure now is to determine the sense of
the shear flow in the plane of the cross-section
100 " z A 2.5 " z A or in the xz plane. It is only necessary to
40
determine this sense at the beginning point, that
We could start the solution at either at is in sheet panel abo The surest way to deter-
two points (a) or h since these points are free mine this sense 1s to draw a simple tree body
edges and thus qy is zero. In this solution, sketch at flange member (a) as illustrated in
Fig. Al4.3l. The shear flow on the cut face is
we will start at the free edge at paint (a) and qy(ab) = - 1.25 and
go counterclockwise around the beam section.
The area or each flange member has been concen- this value 1s shown
trated at a point coinciding with the centroid ~n the free body. By
of each flange area. In solving for the q val- s1mnle rule given at
ues the subscript y will be omitted, and sub- the end of Art. Al4.6,
scripts USing ~~e flange letters will be used the shear flow in the
plane of the cross- Fig. A14.31

----

"'" : ' .~. <: ~ -v- ;,. ::.. '~ .~_:~'-_ ~ ~·f,;~~-;'.. ":<h:~2.:-._;~ '".:: ~ :. ,>:. ~:, :·.L" ~~: ~ '..<~~' ~ :', ;-: ," .~,,;)~":':' ~.~ ;: ~.,~ ~ ~ '~ :,:::,. ,~.;:.;. '~·~.r . ~~? ·5J-:~~. . t';-· ~
 ANALYSIS A D DESIGN OF FLIGHT VEHICLE STRUCTURES A14.13

section is also directed toward the cammon We will again start at the free edge ad-
boundary line and thus qx(ab) has a sense as jacent to flange (a) where qy - O.
shown in Fig. Al4.3l. The sense of the shear
flow on the cross-section will now continue in
~b =- 1.55 Za x A =- 1.55 (- 0.625 x 0.1) =
this direction until the sign changes in the 0.0971 Ib./in.
origianal calculation, which means therefore the
shear flow sense will reverse. Fig. A14.32 qba = 'cab = 0.0971
shows a plot of the shear flow pattern with the qbc = 0.0971 - 1.55 Zb x A = 0.0971 - 1.55
sense indicated by the arrow heads.
(- 5.625 x 0.4) = 3.592
'leb = qbc = 3.592
'led = 3.592 - 1.55 Zc x A = 3.592 - 1.55 x
5.625 x 0.4) = 7.087
'ldc = 7.087
'lde =7.087 1.55 (- 0.625 x 0.1) = 7.184
qed = 7.184
qet = 7.184 - 1.55 x 4.375 x 0.1 = 6.504

Fig. A14.32
qte = 6.504
The results will be checked to see it
qrg = 6.504 - 1.55 x 9.375 x 0.2 =3.589
static equilibrium exists relative to Z Fx and qgt = 3.589
Z Fz O. 'lgh ·3.589 - 1.55 x 9.375 x 0.2 - 0.674
Z F z = 100 (ext. load) - x 6.25 - 10 x 3.75 +
~~
Fig. Al4.33 shows the plotted shear flow
1.25 x 0.5 x 4 - 1.25 x 0.5 x 4 = 0 (check). results. The signs of the calculated shear
flows are for shear flows in the Y direction.
Z Fx = - 5 x 1.25 + 5 x 1.25 - 5 x 1.25 + 5 x Simple Consideration at a free body of flange
1.25 =0 (check). member (a) will ~ive the sign or sense of the
shear flow in the plane of the beam section.
The shear flow force system in Fig. A14.32 Thus in Fig. Al4.34 qx must act as shown when
causes the section to bend about axis xx with- qy is positive.
out twist. The resultant of this system is 100 .674
.0971
lb. acting down in the Z direction. The posi-
----
tion at this resultant will thus locate the
lateral pOSition of the shear center.
Equating the moments of the shear flow
system. about some point such as (c) to the mo-
i b a

3.589 11

~~4
ment of the resultant about the same point we
obtain: -
100 e = 10 x 3.75 x 15 - 1.25 x 0.5 x 2 x 5 -
1.25 x 0.5 x 2 x 10 + 1.25 x 0.5 x 2 x 15 e=6.43"

hence e = 562.5/100 = 5.625 inches.

Fig. AU.33
=--=,,---'L
Thus the shear center lies on a vertical line
5.625 inches to right of line be.
I
CALCULATION OF VERTICAL POSITION OF SHEAR
I I I

~
CENTER.
-----For convenience as betore, we will assume I
• 1".,...l I
a shear load Vx = 100 lb. and compute the re- Ox _5.625 -I
sisting shear flow system to reSist this load I
in bending about axis ZZ Without twist. The a t
resultant at this shear flow system will give
the vertical location ot the shear center. The
shear flow equation 1s,
Fig. A14.34 1
S. C.
6.43
•
q = _ Vx Z x A ::I' _ ~Z X A - 1.55 Z x A Fig. A14.35
y Iz 64.4
A14.14 BENDING SHEAR STRESSES. SOUND AND OPEN SECTIONS. SHEAR CENTER.

Checking to see it Z FZ = 0 and Z Fx = 0: - Calculation of Ix, I z and I xz --

Z Fz = 3.592 x 10 - 3.589 x 10 • .5 x .0971 - .5 Ix • Z A z~ = .5 (5.333 a + 2.667 8 ) + 1
x .0971 5 x .674 - .5 x .674 ... .5 x .0971 - .5
1" (5.333 ... 6.667 )
8
90.667 8

x .674 - .5 x 6.504 + .5 x 7.087 = 0 (check) I z = Z A x~ = 1 (10.667 ) + 2 (5.333 8 ) = 170.667

8

I xz = Z A xz = 1 x 6.667 (- 5.333) + 1 (- 5.333)

Z Fx = 100 + 5 x .0971 + 5 x .674 - 5 (7.087 +
(- 5.333) + 0.5 x 10.667 x 2.067 + 0.5 x 10.667
7.184 + 6.504) = 0 (check)
(- 5.333) =- 21.333
The resultant R of the internal shear flow
system is a horizontal force of 100 lb. acting The constants k~, k 8 and k 3 are now deter.nlned --
toward the left. To find the location of the
resultant take moments about a point 0.5 inch I xz - 21.333
below paint (c). Iz - 90.667 x 170.667 - 21.333 a

Re=ZIl - 21.333
15019 .00142
100 e • 5 (.0971 + .674) 11 + 10 x 3.589 x 15 +
.5 x .0971 x 5 - (.5 x .674 x 10) + .5 x .674 x Iz 170.667
= 0.01136
15 + .5 x 6.504 x 15 + .5 x .674 x 10 - (.5 x Ix I z - 15019

.0971 x 5) Ix 90.667
0.006037
100 e = 643 Iz - I xz
' 15019
• = 643/100 • 6.43 inches Resolving the given shear load of 141.14 into z
and x components, we obtain,
Fig. A14.35 shows the resulting shear center
location for the given beam section. Vz • 141.14 x sin 45° 100 lb.
EXAMPLE PROBLEM 2. unsvnmetrr ica). Beam Section. Vx • 141.14 x cos 45° ~ 100 lb.
Fig. Al4.3S shows a tour 1'lange beam seC-
t ron, The areas at each flange are shown ad- From equation (14) -
jacent; to tlange. The external shear load V
equals 141.14 lb. and acts in a direction as qy = - (k 3 Vx - k 1 Vz ) Z x A - (k. Vz - k 1 Vx )
shown. The problem is to tind the line ot Z z A
action ot V so that section will bend Without SUbstituting
twisting.

qy = - [.006037 x 100) - (- .00142 x 100J Z x A

I Load: 141.14

a 1.0 I Z It-450 [.01136 x 100 - (- .00142 x 100)J Z z A

~
to
-!,,:__ +=- '!,:5 --r
I d
whence
HI! a"
0.5....L
b c qy = - 0.7457 Z x A - 1.278 Z z A
H,a"-j Fig. A14.36

We will start at flange member (a) where qy 15

SOLUTION: - zero on the tree edge side of the member.

used.
To solve this problem, method (3) will be ~b =- 0.7457 X 1 (- 5.333) - 1.278 x 1 x 6.667
To locate centroida1 x and z axes: - - - 4.544
= ~b = - 4.544
=
1 x 12 + 0.5 x 8
3 =5.333 in. = - 4.544 - 0.7457 x 1 (- 5.333) - 1.27S x
1 (- 5.333) = 6.249 1b./in.
= (0.5 + 0.5) 16 = 5.333 in.
3 = %c = 6.249
 A L 5 A E51 o v A14.15

qed = 6.249 - 0.7457 x 0.5 x 10.667 - 1.278 x A14. 11 Shear Center Location By Using Neutral Axis
Method.
0.5 (- 5.333) = 5.680
In a beam subjected to bending there is a
~c = qed = 5.680 definite neutral axis position tor each differ-
~a = 5.680 - 0.7457 x .5 x 10.667 - 1.278 x 0.5 ent external plane of loading on the beam. The
shear flow equation with respect to the neutral
x 2.667 = 5.680 - 3.977 - 1.704 = axis is,
o (checks free edge at d where qy
must be zero.)
qy =_ Vn ~ Zy,A (20)
In
Fig. A14.37 shows the resulting shear flaw re- where, Vn = Shear resolved normal to neutral axis
sisting pattern. The sense of the shear flow in In Moment ot inertia about neutral axis =
0"",
Zn = Distance to neutral axis
I
a

liR'
V= 141.14

141.14
d
U~
Ox
a
Ftg.A14.37a
In finding the shear center location ot an
unsymmetrical section, it is convenient to as-
sume that the Z and X axes are neutral axis and
q =5.680 find the shear flow system for bending about
each axis by equation (20). The resultant ot
b qe 6.249 c each of these shear flow force systems will pass
through the shear center, thus the intersection
Fig. A14.37
ot these two reSUltant forces will locate the
shear center.
the plane or the cross-section 1s determined in
web at flange member (a) by the simple tree Example Problem
body diagram or stringer (a) in Fig. A14.37a
Check Z Fx and Z Fz to see if each equals The same beam section as used in the
100. previous article (see Fig. A14.36) will be used
to lllustrate the neutral axis method.
6.249 x 16 = 99.99 (CheckS Vx = 100)
Fig. A14.38 shows the section With the
" Fz =- 12 x 4.544 - 8 x 5.68 =- 99.94 (checks cerrtr-cfda.L axis drawn in, The X axis will now
Vz=lOO).
The resultant of the internal resisting
~!
~
a I d
Y 100i + lOO~
~~.47
shear flow system equals • 141.14 --l. _ _ _ -x
lb.
To locate this resultant we use the prin-
5.333
I
1*1ll66
b I C b c
Ciple ot moments. Taking pOint (b) as a moment
center,
Fig. A14.38 Fig. A14.39
141.14 e = 8 x 5.68 x 16 - 4.544 x 30 n be assumed as the neutral axis for an external
212 plane of loading as yet unknown. T~e 1'1111 further
hence e = 141.14 = 1.50 inch. assume that when this unknown external loading
is resolved normal to the X neutral aXis, that
Therefore external load must act at a distance it 1'1111 give a value of 100 lb., or Vz = 100.
e = 1.50" tram (b) as shown in Fig. A14.37. The From the previous article Ix = 90.667.
load so located will pass thrti shear center of
section. To obtain the shear center location, Since the X axis has been assumed as the
another loading on the beam can be assumed, and neutral axis, equat10n (20) can be written
where the line ot action of the resultant of the
resisting shear flow system intersects the re- qy _;!:z. Z ZA, hence,
lx
=
sultant as fo~~d above would locate the shear
center as a Single point. It the shear center
location is desired it 1s convenient to assume Qab 9~~~67 x 6.667 x 1.0 -7.35 lb./in.
=- =
a unit Vz and Vx acting separately and find the
horizontal and vertical locations of the shear 'Ibc 7.35 - 9~~~67 x (-5.333)1 = -1.47
=-
center from the 2 separate shear flow force
systems. qed = -1.47 - 9~~~67 (-5.333) 0.5 1.47 =
Fig. A14.39 shows the resulting shear flow
values.
A14.16 BENDmG SHEAR STRESSES. SOUND AND OPEN SECTIONS. SHEAR CENTER.

ZFx =1.47 x 16 =23.52 lb. Take moments about (b) and let (e) equal
distance to "esultant R.
ZFz = 7.35x12+8xL47 =- 99.96 lb. (check)
100.8e = 3.125x1'tx5 a+3.13xSx16
R =..; 100· + 23.52 = 103 li
lb.
e = 753/100.8 = 7.47 In.
tan Q = 23.52/100 = ,2352
Fig. AI4.40 shows the position of this re-
hence Q -= 13° - 16' l0o-tS R = 103 sultant torce. T~ere it intersects the ?revlous
resultant force gives the shear center location.
=
Let e distance fram 23.52
A14.12 Problems
resultant R to point b.
(1) Fig, Al4.42 ~1+1+1"1
shows the cross-

It
Equating moments ot b b -.-
sectlo~ ot a wood
resultant about (b) to that of shear flow
system about (b), beam glued to-
gether on lines i a a
lO3e = -7.35 X 6ft a + 1.47 x 8 x 16 a-a and a-b. The
beam Is subjected 4"
e =- 644
103 = -6.25 In.
to a vertical
shear V : 2400 I
Fig. A14.40 shows the location of the
resultant. We know the shear center 11es on
lb. Determine
shearing stress
on sections a-a
r-f""'4c----,
i
-L
r i---.L
the line ot action of this resultant. Thus we and a-b. Find
maximum shearing
Fig. A14.42
stress on beam
section.
"~~
~.

6.25 ·3.13 (2) Fig. A14.43 1-1"1

Fig. A14.43

shows a Zee sec-

,~.'Lr---a"';~~---'
,:;;'7>::~-Ra1D0.8
Shear
Center
Fig. A14.40
Fig. AH.41
tion loaded by a
1000 lb. load
act1ng through
the shear center
T 1r
b-~-b
l / 16

is. c.
~1-1"-1
as shown. Find -l2'I'
must obtain another resultant force which the shear stress L-J'l / I
passes through the shear center before we can at sections a-a 1000 lh. ' I¥-,-
det1nitely locate the shear center. Theretore
we will now assume that the Z centro1dal axis
and b-b by three .
different methods.
is a neutral axis and that a resolution at the
external load system gives a shear Vx = 100 lb.

qy =- ~Z XA, Iz = 170.667
Qab * - 17~~~67x (-5.333)1 =3.).25 lb./ln.
I"
_ 100
qbc - 3.).25 -170.66 (-5.333)1 = 6.25 lb./1

Qed = 6.25-1~g~66 (0.5)(10.667) =3.13

Fig. A14.41 shows the· shear tlow results.
ZFx =-6.25 x 16 =-100 lb. Fig. A14.47 a
1.0 .4 d
=6 -6
ZFz =-8x3.13+12x3.125 = 12.5 lb.

R =v' looa + 10.Zli ;: 100.8 R=100.8 bLO .4c

tan Q = 112
0Z
= .125 12.5~ r---1O" -----j Fig. A14.46

(4) Determine the shear center location tor the

beam sections in Figs. A14.47 and A14.48. As-
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AB.l7

sume flange members develop entire bending

q=lOi1/in.
stress resistance.
• b

~
,...

10"

( 10"
I
1-+
I 10"
_+_

l~o,~/1
I

r- "-i -.r 20

Ll50"~
Fig.A14.51
FIg.A14.52

(7) In Fig. Al4.5l, the shell structure is sub-

(5) Determine the shear center for the beam
section at Fig. A14.49. Assume only the 8
stringers as being effective in bending. Area
of stringers (a) and (b) =
2 sq. in. each.
All other stringers 1 sq. in. each.
=
jected to a torsional moment M 50,000 in. lb.
The shell skin shown dashed 15 cut out, thus the
torsional moment IS resisted by the constant
shear tlow on the two curved sheet elements ac,
and bd. Determine the value at the shear tlow.
(8) Determine the moment at the constant flow
torce system in Fig. A14.52 about paint {Ole
Also tind the reSUltant at this torce system.

r--l2-J-" 8 t10""
1 sq. in.
c 1.5°" 1. 0 0.8 1. T
R1I7.5 .5 sq. in.
1~' 2a'
I
I liS
1.2 o. T>" 0.8 1.0 1
.8 sq. in.
20'
Fig. AB. 53
.1 !-7+-15"-+8-1
Fig. Ai4. 54
1----19" I .5 sq. in. {g} In Fig. AI4.53, the tour stringers a, b, c
Fig.Al4.50 and d have the same area. Assume the webs in-
effective in resisting bending stresses. De-
termine the distance (e) to product bending
about the horizontal axis without twist.
(6) Determine the shear center tor the un-
symmetrical beam section of Fig. A14.50. As- (10) For the wing cell beam section In Fig.
sume sheet connecting the four stringers as A14.54, determine the location ot the shear
ineffective. Areas of stringers shown on Fig. center. Assume webs and 'Malls ineffective in
bending.
A14.18 BENDING SHEAR STRESSES. SOUND AND OPEN SECTIONS. SHEAR CENTER.

STRUCTURAL TESTING IS AN IMPORTANT PHASE OF STRUCTURAL DESIGN.


SHEAR FLOW IN CLOSED THIN - WALLED SECTIONS
Al5.! Introduction. The wing, fuselage and em-
pennage structure ot modern aircraft Is essen-
tially a single or multiple cellular beam With
thin webs and walls. The design ot such
structures involves the consideration ot the
distribution ot the internal resisting shear
stresses. This chapter introduces the student
to the general problems of shear flow distri-
bution. Chapter A14 should be covered betore
taking up this chapter.
AlS.2 Single Cell Beam. Symmetrical About One Axis.
All Material Effective in Resisting Bending
Stresses.
Fig. Al5.l shows a single cell rectangular
beam carrying the· load of 100 lb. as shown.
The problem Is to find the internal resisting
shear tlow pattern at section abed.
10""
I
I
I qo = O.
I •
L _
/ .05
/ "
.05
1---'20" --~
-~-
.05
Fig. A15-1
Solution 1
Due to symmetry ot material the X cen-
troldal axis lIes at the mid-height of the
beam. The shear flow equation requires the
value of I X I the moment ot inertia of the
section about the X axis.
Ix = {zX.15XIO;'+Z[30X.05X5
From Chapter A14, the equation tor shear
flow Is 1
qy • -~
x
Z z.A - - - - -
This equation gives the change in shear
flow between the limits of the sunnaatn on,
CHAPTER A 15
open sections we could start the summation at a
tree surface Where q would be zero, thus the
summation to any other point would give the true
shear flow qy. In a closed cell there Is no
tree end, therefore the value of qy Is unknown
for any point.
gquatn on (1) g1ves the shear d1str1but10n
tor bending about the X axis lit thout twist.
The general procedure Is to a~ a value or
the shear flow qy at same point ~d then find
the shear tlow pattern tor bending without
twist under the given external load. The cen-
troid of this internal shear flow system will
be the location where the external shear load.
should act for bending without twist. Since
the given external shear would have a moment
about this centroid, this unbalanced moment
must be made zero by adding a constant shear
flaw system to the cell.
To illustrate we will assume qy to be zero
at point 0 on the web ad.
The term ~= 100/62.5 = 1.6
x
'lao = -1.6 Z~zA = -1.6 x2.5 x 5 x 0 .1 = -2 1b/1n.
qba = -2-1.6 Z~zA = _2_1.6x5x20x .05 = -10
x-f
z
qo'b
Cleo'
= -10-1.6 L:~I zA=-10-1.6x2.5x5x .05
=-11- 1.6 z~, zA =-11-1.6 x (-2.5)5 x
=
-11
.05 =r -10
'led 3 -10 -1.6 z~ zA = -10 -1.6(-5)20 x .05
= -2
qod = -z-1.6 z~ zA =-2-1.6(-2.5)5xO.1=0
Fig. A15.2 shows a plot at the shear flow
reSUlts. on the vertical web the increase in
shear is parabolic since the area varies
directly with distance z.
l1
] =62.5 in ....
The intensity at qx and qz In the plane at
the cross-section Is equal to the values of q
found above which are In the y direction. The
sense of qx and qz 15 determined as explained
(1) In detail in Art. A14.6 or Chapter A14.
In
Al5.!
16~'
./.
A15.2 SHEAR FLOW IN CLOSED THIN4WALLED SECTIONS, SHEAR CENTER.

Adding this constant shear flow to t~at of

2 H/in. 10 JI/in. Fig. A15.2 we obtain the tinal shear tlow
pattern of A15.3.

(11 .66 If/in.

IIh~'-".~ q =11 i1hn.
-8.34 i/in.

2 >/In. '--!44~-
10 H/in.

1. 66 Jii/in.
Flg. A15-2 -6.34 ft/in.

Fig. A15.2 also shows the resultant shear Fig. A15-3

flow force on each of the four walls of the It this constant shear flow at -8.34 was
cell. The resultant shear force cn each not added then the external load of 100 lb.
equals the area of the shear flow diagram on would have to be re-Iocated so that it passed
each position. through the centroid of the shear flow pattern
in A15.2. To find this centroid location, we
For example, can equation the moment at the internal shear
flow pattern about some point to the moment at
~b =( 2 ; 10 )20 = 120 lb. the resultant at the system about the same
point. -

'<da = ~ x 2 x 10 = 6.67 lb. gesuj.tant R ~ V ZFx '"+ Z'i'y i:l: V 100" + 0

Qbc = 10 x 10 + 0.667 x 1 x 10
The internal shear flow force system as
given in Fig. Al5.2 will now be checked forw
equilibrium with given external shear loading
at 100 lb. as shown in Fig. A15.l.
ZFz = 100 (external) + 6.67 - 106.67
=
0 (check)
ZFx = 120 - 120 =0 (check)
Equilibrium at moments must also be
satisfied. Take moments ot all forces about
poirrt d. The external load has no moment
about d.
~ =120 x 10 + 106.67 x 20
; 3344 in. lb. clockwise
Thus we have an unbalanced moment which
must be made zero it we are to have equil-
ibrium.
The unbalanced shear tlow of 3334 in .Tb ,
can be balanced by adding a constant shear
in a qounter-clockwise direction around the
call. The value ot this balancing shear flow
would equal,
q =- ~ = 2~0 = -8.34
(A equals area at call = 10
= 100 acting down.
=106.7
Take moments about paint d. Let x equal
distance to resultant. \
RiC = l:!1.:l (internal system)
100 x=120 x 10 + 106.67 x 20
hence x = 33.33 in. Thus the externalload
would have to be moved 33.33 inches to right it
the pattern at A15.2 would hold it in eqUilibrium.
Since we assume q = zero at point 0, this means
an open cell with the tree end at 0 would bend
without twist it the external load was moved the
distance x.
The stUdent should work this same problem
by assuming q at some other point is zero in-
stead ot point 0 as assumed in the above solu-
tion.
Solution No.2. Shear Center Method.
In this solution, we determine the centroid
at the internal shear flow system for bending of
the closed section about axis X without twist.
This point is called the shear center. The
external shear load can then be resolved into a
shear forcs acting through the shear center plus
a torsional moment about the shear center.
We start the solution, exactly as in solu-
lb/in.
tion 1, by assuming the shear tlow q a at =
paint O. In a sense we are cutting the cell at
o and making it an open section. The resulting
x 20) shear flow is as given in A15.2. This open
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
section will bend without twist if the ex-
ternal shear load acts through the shear
center of the open section.
The closed section will be assumed to
bend without twist, and the resulting shear
flow pattern will be determined.
The equation for angular twist Q (See
Chapter A6) per inch length of beam is,
Q =_1_ L: qL where L equals the length
2AG t'
of a web or wall.
or 2AQ =~ Z ~L. The right hand s ide or
this equation represents the total shearing
strain around the cell which must be zero for
no twist of cell. Since G is constant, we
can assume it as unity as only relative
values of strain are needed in the solution.
Thus the total shearing strain 6 around cell
is proportional to,
o =Z i:' - - - - - - - - - - - - - - -
Using the values of q from Fig. A15.2
and SUbstituting in (2),
A
'..I
=• ..s!o
'" t
=[2 x 5
~ x .1
• (10 • 2) 20
2
+ (n - 10) (.667 x 5)1 2
.05 :J
Since the section and shear flow pattern
is symmetrical about the X axiS, the substi-
tution above is written for one-halt at the
cell and the results multiplied by 2. 'dhen
the shear flow q varies over a portion the
average shear flow is used in the above
substitution.
If the cell is not to twist the relative
VNlst of 7000 must be cancelled by adding a
constant shear flow q around the cell to give
a total shear strain at -7000.
The shearing strain for a constant q
equals,
0= Z qL
t
=qLS
[9.1
+...§.2. +
.05
_ 7000_
hence q - n06 - -0.36 1b/in.
Fig. AlS.4 shows the resultant shear
flow pattern if the constant shear tlow ot
-6.36 is added to the shear flow pattern ot
A15.2.
A15.3
6.36 If/in. ,. Q '" 23.8*
.64 Him.
~~4~: s. C,. - - - ..,i+III'rl;1'
8-Q=23· ..l
4.36 */in.
Fig. A15-4
The location of the resultant of the shear
tlow force system of Fig. Al5.4 will locate the
horizontal pOSition of the shear center. Due
to symmetry of the section about the X axiS,
the vertical pOSition at the shear center will
be on the X axis, because for bending about the
Z axiS, the shear tlow would be symmetrical and
thus, the resultant would coincide with the X
axis.
(2)
Fig. A15.4 also shows the resultant shear
load Q on each portion at the cell wall which
equals the area at the shear flow diagram for
various portions as shown.
10 x 5
.05 + .05
ZFZ = - 56.9 - 43.1 = - 100, which balances
the external load of 100 lb.
=7000 =
ZFx a by observation. Take moments about
point 0, the intersection at axis XX and side ad.
- _~ _ 43.1x20+ (-23.8+16.58)10_
x - ZFz - 100 -
7.90 in.
Hence the shear center lies on the X axis, 7.9 ft
from side ad.
The moment ot the external load of 100 lb.
about shear center equals 100 x 7.9 = 790 in.
lb. clockwise. The moment or the internal shear
flow ot Fig. A15.4 is zero thus we have an un-
balanced moment ot 790. Therefore tor eqUil-
ibrium. of moments we must add a constant shear
flow q aro~~d cell to develop -790 in.lb. or
.JLl
.O~
z =-7000 =- 1.98 lb/in.
Adding this constant shear flow to that of
Fig. A15.4, we obtain the tinal shear flow
pattern which will be identical to that in
Fig. A15.3 or the results at solution 1.
AlS.3 Single Cell - 2 Flange Beam. Constant Shear Flow Webs.
As discussed in Art. A14.9, the common
aircraft cellular beam 1s made up of thin sheet
AlS.4 SHEAR FLOW IN CLOSED
wallS and webs which are stiffened by members
usually referred to as flange members. A
simp11fying common assumption is to assume
that the flange members alone develop the
resistance to the bending moment. This
assumption therefore means that the shear flow
ts constant between flange members.
Fig. A15.5 shows a Single cell beam With
two flanges at A and 8. Find the internal
shear flow force system when the beam carries
the external load of 100 lb. as shown.
;----',"'1
f--IO,J V-lOO#
, 100
I AIS.4 Shear CeRter of Single Cell - Two Flange Beam.
,
I
.~ .J:: 1/2"
.--r.... z or the beam as given in Fig. AlS.S. In other
i IO"~
,",b ~
A Fig. Al5-6
~/2"
Fig. A15-6
It is assumed that the two flanges
develop the entire bending stress reSistance.
This means that shear flow is constant on each
web. Let q1 and q. be the constant shear flow
as shown in Fig. Al5.5. The sense or direction
at these shear tlows will be assumed as indi-
cated by the arrow heads in the figure.
The torce system. in the plane at the
craBs-section has 2 unknowns, namely q~ and
q. and thus we can solve for q~ and q. by
Simple statics.
To find q. take moments about point A.
~A • 100 x 10 - q. (2 x 80.52) =0
hence q•• IOCO/Ib 1.04 = 6.21 lb/in.
The Sign of q. cames out plUS, thus the
assumed sense at q. as shown in Fig. Al5.5 is
correct.
In the above equation of moments about
paint A, the moment at the constant flow
system. q. about A, equals Q.. times double the
enclosed area tormed by lines running tram A
to the·ends at the web which carries the shear
rlow q.. In this case the area is the area
or the cell, or 80.52 sq. tn.
To tind the r-enatntng unknown q1. we use
the eqUilibrium equation,
ZFz = 0 =- 100 - 10 x 5.21 • 10q.
THIN-WALLED SECTIONS. SHEAR CENTER.
=
hence q~ 15.21 Ib/in. (Sign cones out positive,
hence assumed sense of q~ 15 correct).
Fig. A15.7 shows a plot of the internal
shear tlow resisting system.
A single cell beam having only two flanges
can resist only external loads which are parallel
to the line AB, and thus a two flange box beam.
is not used very often in aircraft structure.
When the bending moment in a plane at right
angles to line AB 15 small, the resistance of
the curved panel to compressive bending stresse~
may be sufficient to resist such external bend-
ing mcments and thus be satisfactory.
6.21
Let it be required to tind the shear center
wor-ds where would the external load have to be
placed so that the beam would bend without
tw1st.
2.05 7.95
q
q -
constant q
Fig. A15-8 Fig. Al5-9 Fig. AlS-lO
Fig. A15.8 shows the resulting shear flow
system in reSisting the 100 lb. external load
acting as shown in Fig. A1S.5. This shear flow
system will cause the cell to twist. Therefore
we add a constant shear flow q to the cell to
prOduce zero twist (Fig. AlS.9). The centroid
of the combined shear flow system will then
locate the lateral location of the shear center.
To tind q we must write an expression
which measures the twist when SUbjected to the
shear flows ot Figs. A15.8 and 9 and equate the
result to zero, and then solve tor the one un-
known q.
A=.QL-
v u T- o. (Clockw1se Q 1s pos1t1ve)
Substituting, using the shear flows of Figs.
A15.8 and A15.9,
=0
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
o = _ 16.21 x 10 6.21 x 24.28 + 1Qi
.04 .025 .04
+ 24.28q = a
.025
hence q : 8.26 lb. in.
Adding this constant shear flow to that
or Fig. Al5.8, we obtain the shear flow at
Fig. Al5.10.
The lateral position of the shear center
is given by the location at the resultant at
the shear flow system at Fig. A15.10.
The resultant R ::. ./ IF + IF
' x ,I Z After finding the shear center location,
ZFx = a
ZFz = 10 x 7.95 + 2.05 + 10 = 100 lb.
Theretore R = 100 lb.
Equate moments or R about point A to
moment at shear tlow system. about; A.
Re = l:MA
100e =2.05 (2 x 80.52)
e = 3.30 in.
Thus shear center lies 3.30 inches to
lef't or line AB. (See Fig. A15 .10) .
Thus it the given external load at 100
lb. acts through the shear center, it will
produce the shear flow system of Fig. Al5.l0.
However it acts 13.3 to right at shear center,
hence it produces a clockwise moment at 100
x 13.3 =1330 in. lb. on the cell.
equilibrium, this moment must be balanced by
a constant resisting shear rlow around cell
which wt i i produce a moment or -1330.
- 1330
The required q = 2 x 80.52 = - 8.26 lb. in.
which if added to the shear rlow system or
A15.10 will give the true shear floW system
or A15.8.
Thus having the shear center location,
the external load system can be broken down
into a load through the shear center plus a
moment about the shear center. The shear
flow due to each is then added to give the
true resisting shear flow.
It should be noticed that the web or skin
thickness dOes not influence the magnitude or
the shear flow system in a Single cell beam.
A change in thickness, however. effects the
unit shearing stress and therefore the shear-
ing strain and thus tn computing angular twist
AIS. S
of the cell, the web and wall thickness does
influence the amount of twist for a given
torsional load. In the shear center solution,
it 1s known what portion of the shear flOw 1s
due to torque or pure tw1st, and also that due
to bending Without twist, which fact 1s some-
times of importance.
TORSIONAL DEFLECTION OF CELL
The angular twist as gl ven by the f1nal
shear flow pattern of Fig. Al5.8 equals
2QAG =- IqL/t, whence
2QAG = -6.21x24.28_16.21Xl0= -10082 __ (3)
.025 .04
i
we found that the external load had a moment
of 1330 in. lb. about shear center, which was
resisted by a constant shear flow at - 8.26
lb/In. The angular twist under this pure
torque shear tlow should therefore give the
same result as equation (3) above.
2QAG = - 8.26 x 24.28 _ 8.26 x 10. _ I0082
.025 .04
which checks the result or equation (3).
A1S.5 Slngle Cell-Three Flange Beam. Constant Shear
Flow Webs.
100 lb.
-Is" 1--25 lb.
Fig. Ai5-12
For
c'+_+~
Area Cell
• 128. 54 sq. in. Fig. Ai5. ia
. Fig. A15.11
Fig. A154ll shows a sIngle cell beam with
three flange members, A, B and C, carrying the
external load as shown. A three flange box i t
the flanges are not located in a straight lfne
can take bending in any direction and therefore
Is otten used 1n deSign because at its simpli-
city.
For such a structure, there are six un-
knowns, namely, the axial load in each stringer
and the shear flow q in each of the three sheet
panels that make up the cell. For a space
structure, we have six static equations ot
equilibrium, thus a three f Lange single cell
 Al5.6 SHEAR FLOW IN CLOSED THIN-WALLED SECTIONS, SHEAR CENT ER.

beam can be solved by statics if we assume

100
that the three flange members develop all the
bending stress resistance. thus producing .j.5" 1_
constant shear flow webs. ,...- -.::r A
Fig. A15.12 shows the crosS section ABC.
The tr~ee unknown resisting shear flows have
been assumed with a positive sign. (Clockwise
tlow is positive shear flow). These three
unknown shear tlows can be determined by
C .- - 'lac

Fig. AlS-14
I
( I "be 'lab,
-' B
+
Fig. A15-15 Fig. A15-16

statics.

To find ~a take moments about point B The three unknown resisting shear flows
and equate to zero. will be assumed with the sense as i~dicated by
the arrow heads.
ms 100 x 5 - 25 x 10 + C1ca (128.54 x 2) o
To find Qac take ~aments about 9
~
hence 'lea - - 257.08 =- 0.972 1b!ln.
2MB =100 x 5 - ~c (128.54 x 2) = 0
To tlnd qab take ZPz = 0
Qac = 1.945 lo/in.
ZFz = 100 - 10 x 0.972 - 10~b = 0
ZPx =- 15 x 1.945 + 15Qbc = 0
~b = 9.13 1b!ln.
qbc = 1.945
To tlnd qbc take EFx = 0
ZFz =100 - 10 x 1.945 - lO~b =0
ZPx - - 15 x 0.972 - 25 - 15qbc =0 ~b = 8.055 1b!ln.
hence qbc =- 2.639 1b!ln.
The algebraiC signs ot the unknown Q value
The signs of Qca and qbc came out negative, all come out POSitive. thus the assumed direction
hence the sense of the shear flow on these cell of shear flows in Fig. A1S.l4 is correct.
wall portions is opposite to that assumed in
Fig. A15.l2. The resulting shear flow pattern To make the cell VNlst zero. we must add a
is plotted in Fig. Al5.13. constant shear flew q to the cell (see Fig.
A15.15). The relative twist under the shear
The student should realize the thicknesS flow of Figs. 14 and 15 will be equated to zero.
or the wall elements does not influence the
shear flow distribution if we assume the three Z qLt =_ 1.945 x 20.71 1.945x15 + 8.055x10
flanges develop the entire resistance to the .03 .025 .04
bending moment. + 20.71q + lOq + l5q
.03 .04 .025
=a
AlS.6 Shear Center 01 Single Cell-Three Flange Beam.
Constant Shear Flow Webs. T~ence, q = 0.322 Ib/in. with sense as
assumed in Fig. A1S.lS. Adding this constant
Let it be required to determine the shear shear flow to that of Fig. A15.l4 we obtain the
center location for the beam in Fig. AlS.ll. shear flow system ot Fig. A15.16. The resultant
The shear center is a point on the beam cross- R of this shear flow system is obviously - 100
section through which the resultant external lb., since the external load was 100 lb. The
shear must act it the cell is to bend without location of this resultant R will therefore
twist. locate the horizontal pOSition of the shear
center. Equate moment of resultant R about
The shear center location ~ill be deter- point 8 to the moment of the shear flow system
mined in two steps, first its horizontal about B, Whence.
location and then its verti~al location.
100e =1.623 (128.54 x 2)
C~lculation of horizontal location:-
or e = 417/100 = 4.17 in. from line AB.
We will assume any vertical shear load, (Fig. A15.16)
as the example, the same vertical shear as used
in the prOblem of Art. Al5.4, namely, a 100 Calculation of Vertical Position of Shear Center
lb. load acting five inches from A, as
illustrated in the following Fig. A15.l4. A convenient ~orizontal shear load will be

.
~ '~ ~~~ ;~~~'~ ",~' ;: :',~,,/ :/~ _~'< :", \ ",:~:,;)',:,:;:c~:1 :~~: <-~~-~;. "'., ._:~~ ':'s- ~>:~: ,:L<~;.; <~_~ "'-:~," ~- I:,~~ :.,~.<'~ ~ ~-;':~-/,,",:r;~;'6:"~~--'~"~:
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
Al5.7

assumed acting on the cell. Since we used a
25 lb. load in the example problem of Art. 15.5,
we will assume the same load in this solution.
Fig. AlS.l7 shows the loading and the assumed
directions at the three unknown shear ~lows.
A 2S
...
.1!l4\~
A1S.7 Single Cell-Multiple Flange-One Axis of
Symmetry.
Fig. A15.20 shows a Single cell beam With
8 flange ~embersJ carrying a 100 lb. shear load.
The resisting shear flow system will be calcu-
lated.

C
~-t:J
'lea

qc
Sab

B
+ q

q
q _
R:l2.1. T

1.01
e
I
~

B
~
S"
.40" .10" .10" .20"
z

Fig. A15-17 Fig. A15-18 Fig. A15-19

-T-"'IIb----r ---il" g

10"
Solving for the three unknown shear flows
in Fig. A15.17. __L = 'l-d =i:. !_
OMB = - 25 x 10 + qca (12S.54 x 2) • 0

qca = 0.972 1b/in.

;: f.-S"+S"+S"~
Fig. A15-20
ZF x =- 25 + .972 x 15 + 15qcb = 0
~b = 0.695 The beam section which is symmetrical about
the X axis 1s identical to the beam section
ZFz = 10 x 0.972 - 10~b =0 relative to flange material which was used in
example problem 1 of Art. AI4.10.
~b = 0.972
SOLUTION:-
A constant shear flow q is n0W added to
cell to make twi5t zero (Fig. Al5.l8). Assuming the 8 flanges develop all the
bending stress reSistance, the shear flow will
Writing ZqL/t for both loadings and therefore be constant between flanges. Since
equating to zero:- the beam section is a closed one the value of
the shear flow q at any point i s unknown, Thus
Z L/" - 0.972x20.71 0.695x15 + 0.972 x 10 we will imagine the top cover cut between flange
q ... - .03 - .025 .04
members a and h J thus ~klng q zero in this
+ 20.71q + lOq + l5q := 0 panel due to the free end at the cut. ',,/e now
.03 .04 .025 find the internal reSisting shear flow system
for bending at this open section about axis
Whence, q = - 0.324 Ib/in. x-x under a external shear load Vz 100 lb. =
Adding this constant shear flow to that of The calculations would be exactly like
Fig. AlS.l7 we obtain the values in Fig. those in example problem 1 of Art. A14.10 and
A15.19. will not be repeated here. Flg~ AlS.21 shows
the plotted results as recopied from Fi~. A14.32.
R (the resultant) = 25 lb.
I J 100 lb.
Equating moment of resultant about 8 to 1 7 S
' "1
~oment ot shear flow system about B,

Re =OMS cut ,-1.25-- I

25. = 0.648 (128.54 x 2)

Therefore e = 6.65 inches.

Thus shear center lies 6.55 inches above B,
and 4.17 inches to left of B as previously
found.
Fig. A15-21
A15.8 SHEAR FLOW IN CLOSED THIN-WALLED SECTIONS. SHEAR CENTER.

It ZFx and ZFz are considered tor equil-

ibrium or external and internal loads, they
will be found to equal zero.

To check ZMy =
zero take moments about
same point such as C.

ZMe : - 100 x 7.5 + 10 x 3.75 x 15 - 1.25 x 5

1.25 x 10 + 1.25 x 15 = - 187.5 In.lb.

Therefore to make ZMc =

0, a constant
shear flew q equal to M/2A : (187.5/2 x 11 x 15)
= Oe57 Ib./in. is required. Adding this con-
stant shear flow to that in Fig. A15.21, we
obtain the tinal shear flow pattern at Fig.
A15.22. This r fna L pattern is not much Fig. A15-23
di!rerentrram that or Fig. A15.21, the
reason being that the location at the imaginary
cut to make q equal zero ,was not tar from the Solution 1. Using Section Properties and Ex-
true tact, since the final q in thiS panel was ternal Shears with Reference to
only 0.57. It we had started the solution by Centroidal Axes (The K Method).
assuming the web be cut or qbc = 0, then the
correction constant flow that would be needed In Art. A13.8 the calculations for this
to satisfy ZMy =
0 would have come out q = beam sectton gave -
-5488, since this ts the tinal q in web be.
Since the shear flow which is a load on the
cell wall intluences the required thickness ot
Ix = 81.18, Iz =153.58, Ixz =- 21.33
sheet required, it is good practice to try to Fig. Al5.Z3a shows the location of x and z cen-
place the imaginary cut at a point where the trotdal axes.
shear is near zero, so that preliminary z
estimates in routine deSign relative to shell
thickness required will be based on shear flow 1.0 I
values that are near the tinal values.

IJ:.' I .'.0 _I

.--1
.80-1 ! 10.667
d

.1 .•
"1
_--=:b+-_ 2.074

0
928

1.82 5.333 I

Fig. AlS-23a

In Art. A14.8 ot Chapter A14, the method

ot solution was referred to as the K method.
The shear flow equation (see Eq. 14 of Art.
Al4.8) Is,
1.82
.68 1.82
qy = -(k, Vx-k, VzI J:xA -(k, Vz-Ie, Vx ) ZzA
Fig. Al5-22
-21.33
A1S.8 SIngle Cell - Unsymmetrical - Multiple Flange. 81.18 x 153.58 - 21.33' =
Example Problem 1 -21.~3
12016
=- . 001775
Fig•.U5.23 shows a 4 flange unsym-
metrical Single cell beam carrying two external k .. = Iz/IxI z Ixz' =153.58/12016 = .01279
loads as shown. This beam is identical to the
one used in example problem 1 ot Art. A13.8 k , : Ix/IxI z Ixz' = 81.18/12016 : .00674
which dealt wtth bending stresses.
For the given beam loading the external
shear loads at sectton abed are,

Vz = 6000 lb., Vx =- 1600 lb.

 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A15.9

Substituting in the equation tor qy as Adding this volume of q to those in Fig.

given above,
qy = - [.00674 (-1500) - (-.00177 x 6000UZxA
- ['01279 x 6000 - (-.00177)(-1500UZzA
whence
qy =0.16 ZxA - 73.91 ZzA - - -
In using equation (4) to compute the
shear flow pattern we will imagine top panel
ab cut, thus making the shear flow qy = zero
in this panel. Subt. in (4) - - - -
A15.24 we obtain the final shear flow reSisting
pattern in Fig. Al5.25.
Solution 2. PrinCipal Axes Method
The shear flow system can of course be
found by referring section properties and ex-
ternal shear loads to the prinCipal axes of the
beam section. The equation for shear flow 15
(see Eq. 16 of Chapter A14),
V V
qy = - ~ Z zpA - ~ xpA - - - - - (6)
xp zp
(The subscript p refers to principal axes.)

~c =0 + .16 x 1 (-6.333) - 73.91 x 1 x

The section properties about the principal
axes were computed tor this same beam section
6.074 =-449.78 Ib/in. on page A13.5 at Chapter A13. The values are:-

Qed =- 449.78 + .16 x .8 (-6.333) - 73.91 Ixp =76.38, I zp =169.34

x .8 x (-6.926) = - 100.05 Fig. AIS.26 which was alSo taken from page
A13.5 shows the location ot the principal axes
= - 100.05 + .16 x .4 x 10.667 - 73.91 and the distances tram the tour flange members
x .4 (-6.926) = 76.80 to the principal axes.
qba = 76.80 + .16 x .6 x 10.667 - 73.91
x .5 x 2,074 =0
Fig. A15.24 shows the plotted shear flow
results. This pattern satisfies ~Fz 0 and =
ZFx = O. To check eqUilibrium of moments

6000f =V z

~~8'j
cut
I b 1600=
Iq=449 .78 I - Vx

L.J,~o~-J q=75.8

Fig. Al5-26

Fig. A15-24 Fig. A15-25

Before substitution in equation (5) can be
made, the given shear loads Vz = 6000, and Vx =
-1600 must be resolved normal to the principal
about a y axis. Assume a y axis going through axes.
paint d.
= 6000 cos 15°-15' -1600 sin 150_15'
~ = 6000 x 8 - 1600 x 8 - 449.78 x 12 ': 5367.9 lb.
x 16 = - 61160 in. lb. Vx = -6000 sin 15 0-15'-1600 cos 15 0_15'
, ~ -3121.8 lb.
Thus for equilibrium a moment of plus
51160 in.lb. is reqUired. This is prOduced Vz
B P --
nenc s -1- 71.21 -19.69
by adding a constant shear flow q around the xp
cell 'NallS, where
Subt . in equation (5) J

q =J1.
2A = 261160
x 160 = 169 . 88 Ib .in
•
qy = -71.21 Z ZpA + 19.59 Z xpA - - - -(6)
(A = area of cell = 160)

! t f-
 Al5.10 SHEAR FLOW IN CLOSED THIN-WALLED SECTIONS. SHEAR CENTER.

Assume qy =0 in top panel abo

Solution. The K method of solution will
q,c =-71.21 x 4.15 x 1 + 19.59 x (-6.74) be used.
x 1 = -448.92 k = I xz = 36.41
:0. IxIz-Ixz" lS6.5x431.7-36.4l'"
Qed =-448.92 - 71.21 (-7.12).8 + 19.59
= 36.41 '" •0004598
(-3.58).8 =- 99.41 79185

'ldb =- 99.41 71.21 (-2.90).4 + 19.59 k • Iz = 431.7 := .005452

=
x 11.80 x .4 75.05 It/ln. • IxIz-Ixz II 79185

qba =75.65 - 71.21 (4.82).5 + 19.59

ka
=
-
Ix = 186.5 ~ .002355
IxIz-Ixz a ~
x 9.75 x .5 = 0 (check)

These shear flows are practically the
same as obtained in solution no. I as recorded
in Fig. A15.24. Discrepancies are due to
slide rule accuracy.
For equilibrium of moments, take moments
about (b).
Z!1b '" 6000 x 8 - 1600 x 8 - 448.92 x 12
x 16 = -50993 In.lb.
A constant shear flow q around cell must
be added to produce 50993 in. lb. tor equil-
ibrium. This balancL~g shear flow is,
Which Is the same as in solution no. 1.
Example Problem 2
qy = (k~VX-k.V2)ZXA - (k.Vz-k 1VX ) ZzA
Substituting in above equation,
qy ~ -( .002355 x 400 - .0004598 x 1000)l:xA
-( .005452 x 1000 - .0004598 x 400)ZzA
qy • -0.4822 ;:xA -5.268 ZzA - - - - - (7)
Since the value of the shear flow is un-
known at any poi~t on the cell walls, it will
be assumed that the cell wall is cut between
flange members 1 and 10, thus making q zero on
the sheet panel numbered (1-10). Then using
equation (7) the shear flaw is calculated in
going clockwise around the panel. Columns 1 to
10 of Table AlS.l show the calculations in
solVing equation (7). FOF explanation on how
to determine sense of shear flows qy in Columns
9, 10, and 11, review Art. A14.6 of Chapter
A14.

Fig. Al5.27 illustrates a typical· single

cell wing beam With multiple flange members. The shear flow values in column 11 Would
The external shear load on this beam section be the results if the external loads as given
is Vz =
1000 and Vx =
400 located as shown. were so located as to act through the centroid
of this shear flow force system. Since they
The internal Shear flow resisting pattern will
be calculated. do not we will solve for the unbalanced moment
on the beam section about paint (0) the cen-
ThiS beam section is the same as that troid of the beam cross-section. The moment of
used in example problem 5 of Chapter A3, where the shear flow force on a sheet panel be~Neen
the calculations of the section properties any two adjacent flange members 1s equal in
were made. magnitude to q times double the enclosed area
formed by drawing lines from the moment center
The results were: (0) and the ends of the particular sheet panel.
Fig. A15.28 illustrates this explanation. The
Ix = 186.5, I z = 431.7 J Ixz = 36.41 value m in column 12 of the table lIsts the
double areas of these various triangular areas.
V7.:1000~ 7.
2"
_ 9 Taking moments of all forces both external
Vx:4OOf 5 6 7 8 and internal about point (0),
p-
LX
Z!1o '" 1000 x 2 + 400 x s +17123 = 20323 in. lb.
X
10 (17123 equals summation of column 13)
1._ 15 161
Thus for eqUilibrium a negative moment of -20323
z Fig. AlS-27 is needed. This moment is prOVided by adding a
 -e , 4

ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AlS.11

Table A15 1

1 2 3 4 5 6 7 8 9 10 11 12 13 14

Area Arm Arm qy = ~O. 4822 q =q

= -5.268 .~co19 m Final
Member A Z ZA X XA ZZA ZXA (ZXAI
qy
(ZZAI
• sq. in. qxzm q'
col 10) qxz -20.6

0 0 0 0 0 0 -20.6
1 O. 14 4.396 0.615 -17.41 _2.437
2 0.14 .446 0.902 -13.54 _1.896
0.615 -2.437 1. 175 - 3.240 2.065 55.2 114 -18. 53
3 0.38 7.396 2.810 -- 1. 517 -4.333
9.11 ~3. 462 4.327 -7.795
2.089 --22.794
7.991 5.902 44.2 261 ~ 14.70

4 0.17 7.766 1. 320 5.44 _0.925

5.647 _8.720
3.758
4.205 -29.748
19.036
25.543
32.0
38.2
OS
976
- I.
4.94
5 0.17 7.946 1.351 - 0.86 -0.146 6.988 -8.866
4.275 -36.865 32.590 33.0 1075 11. 99
6 0.17 7.896 1. 342 3.14 0.534
8.340 -8.332 4.017 -43.935 39.908 33.2 1325 19.31
7 0.17 7. 96 1.308 7.14 1. 214
9.648 -7.118 3.432 ~50. 825 47.393 40.2 1905 26.79
0.17 7.296 1.240 11.74 1.996
10.888 -5.722 2.470 -57.358 54.888 32.6 1789 34.29
0.29 6.896 2. 000 15.39 4.463
12.88 -0. 59 0.318 - 7.894 67.576 251. 8 1702 46.98
1 0.3 -8.411 -2.964 15.39 5.386
9.924 4.727 -2.279 -52.280 54.559 46 2509 33.16
15 0.31 .8.224 -2.587 9.64 2.988
7.327 7.715 -3.720 -38.862 42.582 48.2 2052 21. 98
14 0.31 -7.734 -2.395 3.32 1. 629
4.982 8.744 -4.216 -26.245 30.44 46.4 1413 9.84
13
12
0.31
0.28
-7.004
-5.554
-2.171
-1. 554
- 2.96 -0.917
- 9.11 -2.551
2. 13 7. 27 -3.7 4 -14. 19 18.59 47.6 885 - 2.01
1. 25 .276 -2.544 - 6.632 9.176 36.6 336 -11.43
11 0.17 -4.504 -0.766 -13.54 -2.302
0.493 2.974 -I. 434 - 2.597 4.031 42.6 172 -1 .57
1 0.17 -2.904 -0.493 -17.51 _2.976
0 0 0 0 -20.
1
Z 17123

constant negative shear flow around cell where AlS.9 Two Cell-Multiple Flange Beam. Symmetrical
magnitude equals About One Axis.

_1'l_20323_ Fig. A15.30 shows a two cell cantilever

q - 2A - 2 X 493 - -20,6 lb./1n. beam with 10 flange stringers. The cross-sec-
tion is constant. Let it be required to deter-
(493 =area of cell) mine the internal shear flow in reststing the
1000 lb. load acting as shown. For simplifi-
Adding this constant shear flow to that in cation, the top and bottom sheet covering and
column 11, we obtain the final shear flow in the three vertical webs will be considered in-
column 14. Fig. A15.29 shows true shear flow effective in taking bending flexural loads.
pattern. Since the beam section is symmetrical about the
X axis, the beam will bend about this axis in
reststing the given external load. The moment
of inertia of the section about the X axis
equals 250 in."

.11 .12 ,13 ,14 .15

Fig. A1S-28

34.

-.
1199 19.

~
5 - 6 7 8

Values from col. 14

Table A1S.1
Fig. A1S-30

Solution 1 (Without use of shear center)

1.98
33.16
The internal shear flow is statically in-
Fig. A1S-29 determinate to the second degree, stnce the

_~It-,~
 AIS.12 SHEAR FLOW IN CLOSED THIN4WALLED SECTIONS. SHEAR CENTER.

shear flow at ~y pOint in each cell is ~~known. zerQ. Thus 'the shear flow in the ver-t ica t web
Therefore, to make the flexural shear flow cc ' Is dete~lned by the stringer c alone,
statically determinate, ~ value for the shear na:nely
flow q in each cell will be assumed at some
paint, and the flexural shear flow for each c'
'lcc ' =-4Z c zA=-4xlX5 =-201b.lin.
cell will then be calculated, consistent with
the assumed conditions. These resulti~g static We can now continue around cell (2) starting
shear flow systems Will, in general, produce a with stringer en where we were ,revlously
different total shearing st~ln around the stopped. _
perimeter of each cell, or in other words,
01'
produce a different cell twist. Since full ~Idr = qb' c '+ 'lee' -4Z,zA
continuity exists between cells, this condition c
= 0 2 0 - 4 x (-5) 1 = 0
cannot eXist, and therefore an unknown constant
shear flow of q:l. in cell (1) and c, in cell (2)
must be added to make the ~Nist of both cellS 4 x (-5) 0.5 = 10 lb./in.
identical. This fact gives us the basis for
one equation and the other equation necessary qe ' e = 10 4 x (-5) 1 =30 lb./in.
tor the Solution of the two unknowns q:l. and qa
is given by the requirement of eqUilibrium,
namely, that the moment of the external and
the internal shear forces about ~~y point in =
Qdc 10 - 4- x 5 x 0.5 ~ 0, which checks
the plane of the cross section must equal zero. the assumed value of q = 0 in panel cd.
In Fig. Al5.31 the flexural shear flow has
been assumed as zero just to the left of The shear flows in cell (2) could of
stringer c in cell (1) and just right of course been found by starting in ~anel cd
stringer c in cell (2). The balance of the where the shear has been ~ssl1med zero and
flexural shear system consistent With this proceeding clockwise around cell as for exa~ple
assumption is calculated as follows:
Clde =a - 4- x 5 x .5 = -10 lb./in.

lOf/in.
Qee' = -10 -4xSxl::a -30 Ib./in.
c
qe'd' = -30 4x (-·5) 1 =-10 lb./tn.
50>/iD. --l'0:" -30./1n. '1i' c' .. -10 - 4x (-5) 0.5 :: 0 lb./ln.
The magnitude of the results are the same
as previously calculated but the Signs are
opposite. As emphasized previously the shear
tlow calculated together with its Sign Is in
Fig. A15-31
the y direction or qy. The direction of the
The general shear flow equation is, shear flow along the cell walls 1n the xz
plane can be determined by drawing Simple free
1000 ZzA = -4ZzA body diagrams as illustrated in Cr~pter A14 but
qy
V
=---". ZzA = -
it Is simpler to use the automatic rule of
Ix 250
Art. A14.11. To illustrate, the solution
Cell (1). Starting in panel cb where the started in panel cb and proceeded counter-
shear has been assumed zero and proceding clockwise around cell (1). By the rule this
counter-clockwise around cell means that the shear flows in ~he xz plane will
have the same sign as qy. Since the signs of
qCb =zero (assumed) qy were negative, the d1r9ction of ~he shear
tlows in the xz plane .vjLl, be counter-cjoctcvtsa
%a = -4ZzA = 0-4x5xO.S = -10 lb./in. around cell (1).
Saa' = -10-4xSx2 =-so lb./in. To obtain the shear flow in the vertical
web eel it was necessary to start at c and go
~rb' = -50-4x (-5) x2;:: -10 lb./in. toward c!. T~iS direction Is counter-cloc~~lse
with respect to cell (2) or clockN!Se with res-
qb'c' • -lO-4x (-5) xO.5 = 0 lb./in. pect to cell (1). qy for this panel from the
equation was found to be -20. If we consider
We cannot proceed beyond stringer o' be- web cc n as part of cell (1) then the direction
cause there are two connecting webs with un- in solVing the s~atlon ZZA was :loskwise wi:h
known shear flows. ~e can get around this respect to cell (1) and by our ~Jle the shear
difficulty by going back to stringer c, where flow in the Z direction will have the 0P9bs1te
the shear flow on each side of c was ass~~ed sign to qy or plus 20, which ~eRns clockwise

" -~,.~. ,s.'~:c .:':_: __ .,~.~:_" .. ~ . . . . . '~:.,,; :<,:,',/: ,~,:~~._. -~~:. '_. ':~, '.,=;,_-. -v-: ''': __-'.'''\.' ':;"-•.:.;-,,,.•..-',- -",".;:.:-~~i;:;
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A15.13

Fig. (Al Since there is continuity between cells, 0:1, =

Q. • Also since area at each cell is the same,
Fig. (Bl
A, = A,. Equating (1) end (2),

l533q:1, - l583q• - 10840 = 0 - - - - - (3)

One other equation is necessary to solve tor

unknowns q~ and q. , and it is given by the
moments at the external and internal shear
forces about any point in the plane at the cross
section, which must equal zero tor equilibrium.

Take moments. about point (b) at the shear

tlow system ot Figs. A15.3l and A15.32 and also
the external shear load or 1000 lb. J which in
this case has no moment about our assumed
moment center.
ZI1tJ:=I -50xlOx5+ZQxlOx5+l0x30x15+Z00q 1.
Fig. A15.32 illustrates the unknown con-
+ ZOOq. = 0
stant shear flow systems q:1, and q. which must
act on cells (1) and (2) respectively to pro-
duce the same cell twist when added to the hence, 200q1. +200q• +3000:: a - - - - - - - (4)
snear flow system at Fig. Al5.31. The sense SolVing equations (3) and (4) tor q:1, and q., we
at q:1, and q. has been assumed clockwise or obtain
positive in each ce~l.
q, =-4.07 lb./in. q, =-10.80 lb./ln.

The tinal or true internal shear flow sys-

tam then equals r.hat at Fig. A15.3l plus that at
Flg. A15.32 when q, =
-4.07 end q, -10.80 lb./ =
in., which gives the shear flow diagram or Fig.
Al5.33.
14.07 10.8
Fig. Al5-32 .8

The equation tor the angular twist Q per

unit length at beam is,
L
2AGQ • Zq:;:
-. 8
(Values ln lb./in.)
using the shear tlow values in Figs. Al5.3l
and 32, the angular twist or each cell will Fig. Ai5-33
be calculated by SUbstituting in the above
equation. Solution Z (By use of shear center)
For cell (1) In this solution, we find the flexural
shear flow for bending about axis X-X Without
2.A. GQ :-10x 5 _ 50 x 10 _lOX 5 +20xlO +~+ twist. The centroid of this internal shear
1.:1, .03 .05 -.-m- .03 .os: system locates the shear center. The moment of
the external shear load about the shear center
3 x lag, _l~,
.w . . produces pure torsion on the 2 cell beam. Thus,
adding the shear due to this pure torsion to
- 6670 - - - - - - - - - - - - - - - (1) that of pure bending, we obtaIn the final re-
Sisting internal shear tlow.
For cell (2)
In bending about axis X-X without twist,
the shearing strain tor each cell as given by
equations (1) and (2) must equal zero. Hence;

1200 q, -333q, - 6670 =a - - (5 )

4170 - - - - - - - - - - - - - - - - - - - (2) -333q :1, + 1250q• + 4170 =a - (6)

 A15.14 SHEAR FLOW IN CLOSED THIN· WALLED SECTIONS. SHEAR CENTER"

Solving equations (5) and (6) for q s and qa' the reSisting internal torque must therefore
we obtain qa = - 2.0 lb./in. q~ : 5.00 lb./in. equal -3600. Therefore,
Therefore, taking these values of ql and qa
in Fig. A15.32 and adding the results to that qt(l) = .00254(-36001 = -9.17 lb./:n.
of Fig. Al5.31, we obtain the shear flow
pattern of Fig. Al5.34 which is the shear flow SolVing for qt(2)
system for bending without twist about X axis.
The centroid of this shear system locates the
shear cent er ,
_ lr-ac"A, +
qt (2) - z[
31190000
auA] T
In Fig. Al5.34,
+ 10 + -lQ) 100 + .!9. x 20;1
ZV = 0 = -10 x 45 - 10 x 28 - 10 x 27 =- 1000
lb., which checks the external shear of 1000
= 21 Q19..
(.03 .03 .05)
31190000
,03 ~Im= 00245T
.oj' .
lb. ZH = 0 by observation of Fig. A15.34.
hence: qt(2) = .00245x··3600· -8.85 lb/in.
v- 1000,*
8 Therefore, it we add to the shear flow sys-
-5
tem ot Fig. A15.34, a constant shear flow of
-9.17 lb./in. to cell (1) and -8.85 lb./in. to
cell (2), we will obtain the true internal re-
28 Sisting shear flow of Fig. A15.35, which checks
-4
solution 1, any discrepancy being due to slide
rule accuracy.
-5 5 (Values in lb./in.)

Fig. A15-34

To find the horizontal position or the centroid

ot the shear flow in Fig. Al5.34 take ~oments
about point"a:

l:Ma = 10x27xlO +10x2Sx20

+ 5 x 8 x 10 - 5 x 2 x 10 = 8600 in.lb.
8600 =.
h ence , -x = lOO6" 8 6' t 0 th e rlgn-...
. of web aa ' .
Fig. A15-35
Torsional Detlections
The external shear load ot 1000 lb. acts 5 ft to
the right of aa' , and therefore causes a moment The angular ~Nlst of each cell is the same.
about the shear center equal to (8.6 - 5.0) The value of the angUlar ~Nlst Q per anit length
1000 = 3600 in. lb. To resist this torsional ot the beam can be founo USing the shear flow
moment, a constant torsional shear flow qt(l) pattern of Fig. A15.35 which is the true ,result-
and qt(2) must act on cells (1) and (2) ant shear flow, or the pure torsional shear flows
respectively. or qt(l) • -9.17 lb./in. and qt(2) = -8.S5 lb./in.
may be used it desired. '
The values of qt(l) and qt(2) can be found
by USing equations (16) and (17} at Art. AS.ll The results Will, of course, be the same.
of Chapter AS. Thus For example:

For cell (1) due to qt(l) = -9.17 Ib./fr..

and qt(2) = -8.S5 lb./in.

2GGA - ZqL _ (9.17xlO) 3+9.17X10 8.S5xlO

1 -'t- ( .03) .05 - .03
= BOOO
Cell (2)
2GGA • ZqL = (8.85X10)3 8.85xlO 9.17xlO
, t I .03 I + .04 - .03
= 8000
Since the external torque equals 3600 in.lb.,

-- .-

~ .""':c~}~:i~:,~;.':~::~,;;~'~~~~~i{fS~l\~\~-(:.~~;~,,~~~;~~k~·:~!;-',}?,o:~:0~~~;~~F;-~:~;~~:';~?,}~ -:'.. :~:~:~::~::~: ~~/~ :~~~:~.':,~~~:-f ::~~1:;~:.~~;': ~:~~~?{E~:~

 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AlS.15

Cell (1) Final stresses - Fig. A15.35
200A _ZgL_(14.17x5)2. (4.17X5)2.54.17X10
• t - ( .03 ) ( .03) .05
26.68x10 = 8000
.03
Cell (2) Final stresses
200A _ZgL_(1O.85x5)2. (0.85x5)2.26.68 x 10
• t ( .03 ) ( .03) .03
-19.15 x 10
• 04
= 8000
A15.10 Three Cell - Multiple Flange Beam. Symmetrical
About One Axis.
Fig. A15.36 shows a 3-cell box beam sub-
jected to an external shear load of 1000 Ibs.
as shown. The section is symmetrical about
axis XX. The area of each stringer is shown
in parenthesiS at each stringer point. The
internal shear flow system which resists the
external load of 1000 lbs. will be calculated
assuming that the webs and walls take no
bending loads, or, the stringers are the only
effective material in bending. The moment of
inertia about the XX axis of effec.lve ~terial
equals 250 in ~ (Note: this beam section is
cell the same, since it any ~Nisting takes
place, all cells must sutfer the same amount.
Furthermore, for eqUilibrium, the moment of
the internal shear flow system plus the moment
of the external shear load must equal zero.
For bending about axis XX, the flexural
shear flow will be assumed as zero at a point
just to the left of stringer a in cell (3) and
just to the left and right of-stringer c in
cells (1) and (2) respectively. one might con-
sider the cells as cut at these three points •
Fig. A15.37 shows the flexural shear flow under
these assumptions. Since the leading edge cell
(3) has no stringers and the covering Is con-
sidered ineffective in bending, the shear flow
will be zero on the leading edge portion since
the shear flow was assumed zero just to the left
of stringer a. The resulting flexural shear
flow tor the 3 cell section will therefore be
identical to Fig. Al5.3l and the calculations
for the flexural shear flow will be identical to
those in Art. Al5.7.
l~
-lOi/in.

I'
I
10 t/in.

---I
j~
identical to the two cell"beam ot Fig. A15.30
plus the leading edge cell (3). I I 30./in.j
I ~ 50.JI/in. I 20t/in.

I I I~
-
-,
I l:\
laflin. 10tt/in.

025 Fig. A15-37

.05 .03 .04
+"--- _._-------- x Fig. A15.38 shows- the unknown constant
10"
shear flows q~, q~, and qa which must be added
(3) (II (2) to the flexural shear flow of Fig. A15.37 to
make the twist Q of each cell the same. The

L 1(2.01
3'

1(·5) I
5"---1-5"----1- 5" -
b' .03 c'
1(1.0)

Fig. A15-36
d'
1(.5)
5"(-1-5"--1
.03
I(La)
e' sense of each has been assumed pOSitive in each
cell.

r - ---r
If <1 I
"I <3If I
Solution No.1 (Without use of shear center) I
(21
The system is statically indeter.ninate, to I
the third degree, since the value of the shear
flow q at any paint in each cell is unknown. I

The value of the shear tlow will be as- Fig. A15-38

sumed at a point in each cell and the flexural
shear tlow tor bending about the XX axis will The angular twist Q for each cell equals Q =
be determined consistent with this assumption. I
A constant unknown shear flow q J,' q a' and q , 2AG Z q Lit
for cells (1) and (2) and (3) respectively
will be added to the static flexural shear Using the values of q in Figs. Al:.37 and
flow so as to ~ke the angular twist Q of each Al5.38, the 'ffilue 6G will be computed for each
cell.
AIS.16 SHEAR FLOW IN CLOSED THIN-WALLED SECTIONS. SHEAR CENTER.

Cell (3)
-12.12/F/in. -2.12*/in. -7 2.91 ill/in.
2QA G = ZqL
e t
I'
ZQ x39.4G = 10 x 50 ... 15.71 q ~ ... 10~ _ 1031. f
.0S ~ -;-os- .00 22.91 I/in. ~';

or Q1 =10.5q.-2.55q.+127 - - - - - - - - (1) ( 1)
-37.620/ln.
(2)
-24.97 if/in.
,

Cell (1)

2QA G = ZqL
• t I in.
-12.12*/in. -2.12li"/ in. _7.0911/in.
2Qx10OG = -50x10_2(10X5) +10x20+3(10Q.)
.05 .03 .03 .03 Fig. A15-39

... 10q1. _ 10q~_10qll hence Q:} = 6q1.-q3-1. 67 Q a A15.11 Shear Flow in Beam with Multiple Cells. Method of
-:05 --:ob:-cr;-
Successive Approximations.
- 33.34 - - - - - - - - - - (2)
The general trend in airplane structural
Cell (2) design appears to be to the use of a relatively
large number of cells. There are various
2QA G =ZqL reasons for this trend some of which are:
• t (1) USing multiple interior webS, the detri-
2Q x 100 x G =_ 20.03x 10 + 2 (10.03x 5) + 10.040
x 30 + mental effect at shear defo~tion on bending
stress distribution is decreased; (2) the fail
sate characteristic at the wing is increased
3x 10 Xqa ... 10q,_lOq; hence QJ = 6.25Q. _ because the wing is-made statically indeter-
.63 .04.03 minate to a high degree and thus failure of
1.67q1. ... 20.83 - - - - (3) individual units due to fatigue or shell fire
can take place without greatly decr-eas tng the
Taking moments at the internal shear flow over-all ultlmate strength of the wing;
systems at Fig. Al5.37 and A15.38 and the ex- (3) the ultimate compressive strength of wing
ternal load of 1000 lbs. about stringer a and tlange units is usually increased because
equating to zero:- - column action 1s prevented by the multiple webs
which attach to flange members.
l:Ma = 10 x20 x 10 + lOx 30x 20 - 5 x 1000
In Chapter AS, Art. A6.l3, the method at
+ 78.6Qe'" 200q1.'" 200Q. = 0 successive approximatton was presented by de-
= 3000+7S.5Q.+200Q. +20Oq,=0 - - (4,) termining the t: eetsutng shear flow system when
a multiple cell beam was subjected to a pure
Solving equations (1) (2) (3) and (4,) tor the torsional moment. This method at approach has
unknown q1.' qll' qe and 00 we obtain: now been extended to determine the resisting
shear r Low when the beam is s ub jec t to flexural
q. = - 2.12 lb./in. bending without ~Nist·. USing these ~NO methods
the shear flow in a beam with a relatively large
Q. = - 7.09 1b./in. number of cells can be determined rather rapidly
Q. = -14,.5
1b./in. as compared to the usual method of solving a
number of equations.
Q1 = - 19.9
PHYSICAL EXPLANATION OF THE =HOD
Adding these constant shear tlows to the flex-
ural shear tlow of Fig. A15.37, we obtain the Fig. AlS.40 shows a 3-cell beam carrylng
true internal resisting shear flow as shown in
and external shear load V acting through the
Fig. A15.39.
shear center of the beam section but as yet un-
known in location. In other wordS, the beam
bends about the symmetrical axis X-X without
twist. The problem is to determine the internal
reSisting shear flow system for bending without

• ''The Analysis of Shear Distribution lor Multi-Cell Beams

in Flexure by Means of Successive Numerical Approx:i~
mations." By D. R. SAMSON. Journal of the Royal
Aeronautical Society, Feb. 1954.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A15.17

twist. In this example, it is assumed that the
bending moment is resisted entirely by the
flange members as represented by the small
circles on the figure, which means that the
shear flow will be constant between the flange
members.
The first step in the solution is to make
the structure statically deter.ninate relative
to shear flow stresses for bending Without
twist. In Fig. A15.41 imagine each cell cut
at points a, band c as shown. For the given
shear load V, the static shear flow qs can be
calculated, assuming the modified section bends
about axis X with nO twist. Fig. A15.41 shows
the general shape of this static shear flow
pattern.
-x
~ll Cell Cell
(1)- - (2) - -(3) - -X
% -q;-
a b c Fig. A15-41
e-,
--- 'Is
Now consider each cell as a separate cell.
The static shear flow qs acting on each cell
will cause each cell to ~Nist. Since zero
~Nist is necessary a constant shear :lo~ q~ to
cell (I), q~ to cell (2). and q~ to cell (3)
must be added as shown in Fig. A15.42. and the
magnitudes of such value as to make the ~Nist
of each cell zero. However. the cells are
actually not separate but have a common web be-
tween adjacent cellS, thus the shear flow q~
acts on web 2-1 which is part of cell (1). and
thus causes cell (1) to twist. Likewise cell
(3) is twisted by q~ and cell (2) by both q~
and q~. Therefore to cancel this additional
cell twist. we must add additional constant
shear flows q:, q; and q~ as shown in Fig.
A15.43, and considering each cell separate
again. However, since the cells are not separ-
ate these additional shear flows effect the
twist of adjacent cellS through the common web.
As before this disturbance in cell VHist is
Fig.A15-40 again cancelled or made zero by adding further
closing shear flows q~. q:, q:
as shown in
Fig. AI5.44. This procedure is repeated until
the clOSing shear flows become negligible. In
general the converging of this system is qUite
rapid and only a few cycles are necessary to
give the desired accuracy of results.
The total closing shear flows q:.. q" and
q, are then equal to -

= q~ + q~ + q~ +
-
".~--t ---
1- -I
'Is
q.
q. = q~ + q: + q: +
-

:<I'
I hi q. = q~ + q; + q: +
\
1"2
~
(21 I
___t The t 1nal shear flow on any pane I then
'>':! equals, (See Fig. A15.45)
'Is
". q = Qs + q:. + q" + qs - - - - - - - - - (1)

The centr01d of this final shear flow

system locates the shear center of the section,
relative to bending about the X axis.

DERIVATION OF EQUATIONS FOR USE IN SUCCESSIVE

APPROXIMATION METHOD

Fig. AI5.46 shows cell (2) or the 3-oe11

beam shown in Fig. A15.45. qs is the static
shear tlow and ql' ~
and ~ are the re-
dundant or unknown
shear flows. Since (1) r ~ "2
--
"s
-II
cell (2) does not twist
under these shear flows (2) q11
we can write in general,
II
-- II
l~ = 0 - - - - - - (1 ) --
".
Substituting the var-i-, Fig. A15-46
ous shear flows on cell
(2) in Fig. A15.46 into eq, (1).
A1S.18 ANALYSIS
L L L L _
Z q - - q Z -... qJ..z -"'q3 Z-- 0 - - - - (2)
" s t a" t :1._ t 3_t
The SUbscript (a) on the summation symbol
tmplies summation completely around cell (2)
whereas the subscripts :1.-8 and 3-a implies
summation only along webs 1-2 or 3-2 respect-
ively. L is the length of a sheet panel and t
its thickness.
The first term in equation (3) represents
the proportion or the static shear flow qs
which must act as a constant shear flow around
cell (2) to cancel the twist due to qs' The
resulting value of this first term will be
given the term q~.
The second and third terms in (3) repre-
sent the constant closing shear flows required
in cell (2) to cancel the twist of cell (2)
due to the influence of qJ. and q~ in the ad-
jacent cells acting on the c~~on webs between
the cells. The ratio in equation (3) before
q. will be referred to as the carryover
influence factor fram cell (1) on cell (2) and
will be given the symbol C1.-'" and the ratio
before q~ in equation (3), the carryover
influence factor from cell (3) to cell (2) and
it 1"1111 be given the symbol C:s_:a' Thus
equation (3) can now be written as,
As explained above, q~ is the value of the
necessary closing shear flow tor zero twist
when the adjacent cell shear flows are zero.
Hence first approximations to the final shear
flows in each cell can be taken as neglecting
the effect of adjacent cells, or in other words
each cell is considered separate. Hence the
first approximations are,
q, - q~
q. = q~ { (5)
q. = q~
By SUbstituting (5) in (4) a second ap-
proximation for q~ is obtained, namely,
qll = q~ + C._ q~ ... C:!-11 q~ - - - - - - (6)
q~ + q: - - - -
where q: is the correction added to the first
approxi~tion. In a similar ~nner corrections
q~ and q; are made to the approximations for
q:l. and q3 T~erefore as a third estL~ate for
q3' these ~urther corrections should be added
AND DESIGN OF AIRPLANE STRUCTURES
and thus equation (7) becomes,
Thus by repeating the above procedure, a
power series of the carryover influence factor
is obtained. In general the convergency is
rapid and only a relatively few cycles or oper-
ations are needed for sufficient accuracy for
final shear flows. A solution at a problem
wll1 now be given to show how the necessary
operations form a very sim?le routine.
- - - (3) A1S. 12 Example Problem Solution. p r-cblem No. 1.
Fig. A15.47 shows a cellular beam with
five cells. The flange areas and the. web and
wall thicknesses are labeled on the figure.
The prOblem will be to determine the internal
shear flow pattern when resisting an external
shear load of Vz = 1000 Ibs. Without twist of
the beam. Having determined this shear flow
system the shear center location tallows as a
Simple matter.
Fig. A15.48 shows the ass~ed static
condition tor determining the shear flow system
in carrying a Vz load of 1000 lb. without ~Nist.
The static condition is that all webs except
the right end web have been imagined cut as
indicated thus making the shear flow qs at
these points zero.
In this example problem it will be assumed
that the flange members develop all the bending
stress resistance, which assumption makes the
shear flow constant between adjacent flange
(4 ) members.
The total top flange area equals 5.5 in. II ,
and also the total bottom flange area. Due to
symmetry the centroidal X axis lies at the
mid-depth point.
Hence, Ix = (5.5x5~)2 =275 in."
0_
~
= -1;;"
Vz ZzA -- 1000 ZzA = -3.636
- 275 l:zA
Starting at the lower left hand corner,
the static shear flow Qs will be computed going
counter-clockwise around beam.
Qab =- 3.636(-5)2 = 36.36 lb./ln.
qbc = 36.36 -3.636(-5)1 =54.55 lb./ln.
qed =54.55-3.636(-5)0.5 =63.64
Continuing in like manner around the beam, the
(7 )
values of qs as shown in Fig. Al5.48 would be
obtained.
The solution from this point onward is
made in table form as shown in Table A15.2
which should be located oelow a drawing ot the
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A15.19

=(~L
cellular beam as illustrated, and the numbers
in the Table should be lined up with respect =.§£9. = .2336
to the cells as indicated. C a-l. L: L 856
.1:
The solution as presented in Table A15.2
Is carried out in 17 simple steps. The first We are now ready to start the solution
step as given in row 1 of the Table is to proper by successive apprOXimations. In row 5
compute for each cell the value for L: ca J. of the Table, the first apprOXimation is to
t assume a value q' added to each cell which will
where qs is the static shear flow on each cancel the twist due to the static shear flow
sheet panel of a cell; L the length of the when the cell webs are not cut, but each cell
panel and t its thickness. Values for qs are is considered separate or independent of the
taken from Fig. A15.48. other. This constant closing shear flow ql
equals,
For example, tor CellI L
zqgt"
ql = Z.b. •
The minus Sign is necessary
Z qg ~ = 2(36.36 x 10) .~ = 18180 t

The sign is positive because Qs Is posi- because the twist under the static shear flow
tive. (Clockwise shear flow on a cell is must be canceled. The values for q' are re-
positive.) Row 1 in the Table shows the values corded in row 5 of the Table.
as calculated tor the 5 cells.
For example for cell (1),
The second step as indicated in row 2 of ,- 18180 - _ 21.238
the Table 1s to calculate the value of the q--856-
expression L: Lit for each cell.
For cell (2),
For example, for CellI,
qI 3 _---gsa-
27275 -_ _ 28 . 71
z...h.=...lQ.... (10)2. 10 = 856
1. t .064 .04 .05
Steps 6 to 13 as recorded 1n rows 6 to 13
For cell 2, at the Table are identical in operation, namely,
the carryover influence tram one cell to the
adjacent cell because of the cammon web between
z...h.= 10.2(10) • .12.. = 950 the cells. As a closing shear flow 1s added to
• t -:05 ~ .04
each cell to make the cell twist zero when they
The third step as indicated in row 3 is to are considered separate, this result is contin-
calculate the value for the Lit of the common ually disturbed because at the crnnmon web.
web be~Neen two adjacent cells. Gradually these corrective shear flows become
smaller and smaller until the cells reach their
For example, for web bb' between cells true state and possess zero twist. In the
(1) and (2),
Table, arrows have been used for two cycles
to help clarity the operations.
f-\tL) lo_
=..l£..
.05
=200 For example in row 6, the carryover shear
flow from cell (1) to cell (2) is,
The fourth step is to determine the
carryover factor from one cell to the adjacent - 6.700 x .2105 =- 1.414
~ell. The results are recorded in row 4 of
From cell (2) to cell (1), the carryover value
the T2.ble.
1s
Referring to equation (3) for general
expla~ation. the carryover factor from cell (-4.480 - 8.330) 0.2336 =- 3.000
(1) to cell (2) is,
From cell (2) to cell (3),

200
(-4.480 - 8.330) 0.250 =- 3.215
Cl._<;J .. 950 3 .2105

and the carryover factor from cell (2) to

cell (1) equals,
A15.20 SHEAR FLOW IN CLOSED THIN~WALLED SECTIONS. SHEAR CENT ER.

, b' c d' 5 ,
Fig. A15-47
.04 .04 .04 .03 " .03
Flange and Cell
Web Data 10" X .064 (I) .05 (2) .04 (3) .04 (4) .03 (5) .03 x

.04 .04 .04 .0' .03

.5 c .5d .5 e
5 Cells at 10" == 50" -------------'

Fig. A15~48
Assumed Static
Condition ~or
~'

t,
36.36

Cell
(I)
r ut
54.55

(2)
c'

ut
63.64

(3)
d'

ut
72.73

r ut
81. 82

(5)
f '

roo]
,
,

,
1
Shear Flow qs 0 0 0 ( 4)
To
a 36.36 54.55 c 63.64 d 72 73 e 81.82
Table AIS. 2

Row OPERATION

I. Z q" L!t for each cell 18180 I 27275 I 31820 I 48487 I 87880
2. Z L!t for each cell 856 950 1000 I 1250 I 1333
3. L! t of cell web 200 250 250 333
4. Carn OVer Factor (C) .2105 I .2336 .250 I . 2633 . 200 I .250 .2501.266
5. 1st Approx, q' == -EQs Lit!!. Lit ~21. 238...... _-29.7" ~31. 82 .... -38.79_ ~65. 910

6. q" == Cq' (Carry over) - 6.700~-4.480 -8.38~7.170 -9.70~6.364 -17.560~- 9.700

7. q" : Cq" (Carry over) - 3. 000- ...... -1. 414 -4.430'" ....3. 216 ~5. 975 ......- 3. 372 -2. 580~ ..... - 5.975
8. ETC (Carry over) - 1. 362 -0.631 -2.518 -1. 460 -1. 489 -1. 837 -1. 592 1. 489 -
9. Carry over -
0.736 -0.287 ~O. 777 _0.787 _0.872 -0.590 ~O. 396 0.872 -
10.
11.
Carry over
Carry over
- 0.248
- O. 138
-0.155 -0.436 _0.266 _0.246 -0.332 -0.232
-0.052 -0.135 -0.148 ~0.141 ~0.lO2 -0. 065
0.246
0.141
-
-
12. Carry over -
0.044 -0.029 -O~076 _0.047 -0.042 ~0.O58 ~O. 037 - 0.045
13.
14.
Carry over
q.q'~q".q" •• ---)
-
0.003
-33.47_
-0. 009 -0.003 ~O. 003 ~O.OO3 -0.002 ~O. 012
-52.52 .... ~63. 39 .... -73.92_ -84.38
0.003 -
15. 1st Reiteration -12.27 - -. -7.05 I -16.690><:13.13 1-18.4a>-<:12.68 1-22."'><--'-16.46
16. q' from row 5 -21. 238 -28.71 -31. 82 -38.79 ~65. 91
17. q -33.51..... -52~ 45 ....... -63.43 .......... ~73. 91 __ .......~84. 39
18. 2nd Reiteration 12.23'- ---7.05 1-16.70>-<:13.10 1-18.'&"-12.69 I -22.'4~-16.'6
19. q' from row 5 -21~ 238 ~28. 71 -31. 82 _38.79 ~65.91
20. q -33.'7 -52.46 -63.40 _73.92 ~84. 39

Fig. A15-49
Closing Shear
1---1 r~ --t~----tj----tfl---
Flows to Make I II II II II
Twist of Each
Cell Equal Zero
f q'33.51 It q=52.45 II q=63.43 II q=73.91 It q.64.39

I If
----~~-- - tl--- tj-- ---
l 2.65
! 2.10 0.27 1. 18 2.57
Fig. A15-50
f1
Final Shear Flows.
(Fig. A15.48 plus Fig. A15.49)

CALCUL'.TION OF SHEAR C&\'rER LOCATION

f""
t 2.85
1
I
16 94
.

2.10
, 0.27 t
10 46
.

1. 18
10.48

2.57
15.61

In Fig. A15.47 let x

= distance tram lett end ot bea~ to shear center. Taking moments about
upper left corner of the shear flow forces in Figs. A15.48 and A15.49 and equating to 1000 X.
lOOoX = 10(36.36+54.55+53.64+72.73+81.82 )10 ... 100 x 10 x 50 - 2 x 100(33.51 + 52.45 .;. 63~43 + 73.91
= 19472 or
.;. 84.391, hence leOoX 19.47 tnchee , x=
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
In row 13 of the Table, the carryover
values are 50 small that the process Is termin-
ated. The final constant shear flow that must
be added to each cell to cancel the twist due
to the static shear flow equals the algebraic
sum of the values from the beginning row 5 to
row 14. The results are shown in row 14 of the
Table.
The results in row 14 are obtained after a
·considerable number of multiplications and
additions of numbers, thus it is easy to make a
numerical mistake. To check whether any
appreciable mistakes have been made, we take
the values in row 14 and consider these values
of constant shear flow in each cell as that
causing zero twist if cells ~~e separate. Then
bringing the cells together, through the common
webs causes a disturbance in tHist and this is
made zero by the carryover values. This step
in the Table is referred to as a reiteration
and is indicated in row 15. Then adding the
values in row 15 to the initial approximation
ql in row 5, which value is repeated in row 16,
we obtain the final value of q in row 17. The
values in row 17 are practically the same
magnitude as in row 14, thus no appreCiable
mistakes have been made. If the difference
was appreCiable, then a second, and if needed,
even a greater number of reiterations should
be carried out. In the Table a second reiter-
ation is shown in rows 18, 19, 20 and the
results in row 20 are practically the same as
in row 17.
It will be assumed that the solution was
stopped after first iteration, and thus the
valueS in row 17 are the constant shear flows
that must be added to the static shear flows
to produce bending without twist. Fig. AlS.49
shows these final closing constant shear flows.
Adding these values to those in Fig. A15.48 we
obtain the final shear tlows in Fig. AI5.49.
The lateral location of the shear center
for this given 5 cell beam coincides with the
centroid of the shear flow force system in
Fig. Al5.50. The calculations for locating
the shear center are given below Fig. Al5.50.
Solution 2 of Problem 1
In solution 1, the assumed static condi-
tion involved cutting all vertical webs except
the right end web. Thus the static beam
section became an open channel section and the
resulting static shear flows must obviously be
far different t~ the final true shear flow
values, since the webs always carry the greater
shear flows in bending without twist. This
fact is indicated by the relatively large num-
ber of steps required in Table A15.2 to reach
a state where successive corrections were small
enough to give a desired accuracy of final re-
sult. Thus it is logical to assume a static
A15.21
condition where the static shear flows in the
webs should be much closer to the final values
and thus hasten the convergency in the succes-
sive approximation procedure.
Thus in Fig. A15.51, we have assumed the
top panel in each cell as cut to give the static
condition. The static shear flow is now con-
fined to the vertical webs and zero values for
top and bottom sheet. Table A15.3 shows the
calculations tor carrying out the successive
approximations and needs no further explanation.
It should be noticed that atter the first ap-
proximation was made in row 5, only three carry
over cycles were needed in rows 6, 7 and 8 to
obtain the same degree at accuracy as required
in 8 cycles in Table A15.2 for solution 1. Fig.
A15.52 shows the final shear tlows which equal
the constant shear flows in each cell fram row
9 of Table added to the static shear flows in
Fig. A15.51. These values check the results of
solution 1 as given in Fig. A15.S0, within slide
rule accuracy. In Table A15.3 no reiteration
steps were given. The student should make it a
practice to use such checks.
A15.13 Example Problem 2.
All Material EUective in Bending Resistance.
The general trend in supersonic wing struc-
tural design is toward. a large number ot cells
and relatively thick skins, thus in general, all
cross-sectional material or the wing is effective
in resisting bending stresses and thUS the shear
flow va!ies in intenSity along the walls and
webs at the beam cells. Fig. Al5.53 shows a ten
cell beam with web and wall thicknesses as
shown. It will be assumed that all beam mater-
ial is effective in bending. The shear flow
resisting system for bending about the horizontal
axis without ~Hist will be determined. The cen-
troid of this system will then locate the shear
center.
Fig. A15.54 shows the static condition that
has been assumed. namely, that the upper sheet
panel in each cell has been imagined cut at its
midpoint, thus making the static shear flow zero
at these points. The static shear flow values
Qs are shown on Fig. A15.54. To explain how
they were calculated, a sample calculation will
be given.
The moment of inertia at the entire cross-
section about the horizontal centroidal axis lS,
Ix for top and bottom skins,
= (50 x 0.125 x 2.5')2 .. 78.0
Ix of all webs = (0.912x5~)/12 = 9.5
Total Ix ::; 87.5 in"
For convenience an external shear load
Vz ::; 8750 lb. will be assumed ac~ing on this
beam section.
Hence, q =-~ZzA
Ix =_875
87.~
9 ZzA =-100 ~zA
I ~
. .)
,. .......1
AIS.22 SHEAR FLOW IN CLOSED THIN_WALLED SECTIONS. SHEAR CENTER.

Now consider Fig. A15.55 which shows a Starting at point (al in cell (1) where the
sketch or cell (1) plUS halt or cell (2). As shear rlow is zero and going counter-clockwise
previously explained the upper cell panels were around the cell. the static shear flows are
assumed cut at their midpoints (a) and (ml. as follows: -
Solution II out
0 0 I 0 '~ 0 0 0
I
out out out out

Fig. Al5-S1
Assumed Static
Condition foE" 36.36 (1) 18.18 (2) 9 09 (3) 9.09 (4) 10.09
01 (5) 18.18
Shear Flow eu and
ReIlU1t1ng lis Values.

0 0
1 I
Table A1S. 3

Row OPERATION

1. E qs Lit for each cell -2045 I -1362 I 0 I 758 I 3031

2. Z t for each cell "8 950 1000 1 1250 I 1333
3. L t of cell web 200 250 250 333
4. Car over factor C .2105 \ .2336 .250 I. 2633 .200\ .250 .250 .266
-z
5. 1st approx. ,= qg L/t/Z Lit 2.389 1.434 0 _0.606 -2.274
6. "_C I (carry over) 0.334 0.503 0 O. 358 _.152 0 -.605 -0.152
7. ", =C .. (carry over)

..". .... ---

0.118 0.071 .055 0.126 _.152 0.041 -.041 _0.152
8. ETC car over O. 029 0.025 _.007 0.031 0 -0.005 -.040 0
9. 0 " 2.870 2. 081 0.211 -1.256 _2.578

Fig. AlS-52
F1na.l Shear Flows
2.87 2.081 0.211
I 1.206
2.5 S

(Row 9 Plus Ftg. A15-S1)

Compllu'e Results with
Solution I (See Flg.AlS-SO)
33.49 18.96 10.98 10.55 f 10.41 15.60

2.87 2.0S1 0.211 1.256 f 2.578

Example Problem 2 Ten Cell Beam - All Material Effective in Bending.

Top Skin - -. 125 Inches Thick

Fig. Al5-53
.
g (1)
.
~

~
(2)
..
~

~
(3)
.=
~ (4)
.
g (')
..
i!: (') -
=
~
(7) ;;;
= (S) "~
~ (9) "0
~ (10) 8T
5"

-
Bottom Skin - • 125 Inches Thick I--- 5"--1
1
-- -100 - - -1" - -
78 7So 78 7S o 78 78 078 7S o -
7878 0 78 78 0 7878 0 7S 7S o
-1" - -15"
j
78 a 7878 a 78

- ;7~5~f
_7 8 -rse -m '-.156f- .,..156

Fig. A15-54 _9
" "•.. 185. 185.4 IS5.4 18S.4f 181.3 181.3 178.5
_7
-0-
_78
-0- -0- -7878
0_
_n78 -~m
_0 _ _ 0_

-7878 .78m
_ 0 _

-7878
_0 _

.78m
_0 _0

Table AIS. 4

2;0, Lit 2200 0 0 0 0 U90 1350 0 1 -4760

\ 0
ZL/t cell 211.4 186.4 IS6.ol 186:4 186.4 211.3 195 219 227. '7203.6
Lit web 53.2 53.2 53.2 53.2 53.2 61. 8 61. 8 69.5 69.5
C .285 • 252 .285 .28S .285 .285 .285 .28S .285 .285 .304.316 .292 .304 .317 .329 .306 .317
q' =-EQsL/t/l: Lit -10.42 0 0 0 0 -6.92 0 _5.63 0 20.91
C.O. 0 ..2.980 0 0 1.97 0 -2.1 1.71 00 _1.7 .63 0
C.O. ~.75 0 -4).860 ..0.57 0 ".5 6-1.20 0 -1.11 1.60 0 1.48
C.O. 0 ..(l.21 .25 ".2 0 ".1 ".1 ..0.50 0 .(1.53 0.15 00 0.1 0.47 0
C.O. .. Q 11 .00 -0.1 3..o.07..oJ.0 ..0.19..0.070 ".1 90.l! 0 ~ll 0.20 0 0.19
C.O. ..0.00 ..0.03 ..0.08 ..0.0 ..{l.O7.{1.0 ..0.02 ..0.07 ..0.08 -'J.O a ..0.09 .03 00 0.0 0.06 0
q _11.28 _3. 53 -1. 29 -1.14 -2.8S -9.00 -<l.2S -5. as 5.56 22.58
.
Reiterate

I,
..0.89
-10.42
-3.21..{).36 -1.01-0.32 -Q.38..o.81 ..0.3 -2.56..0.8 -1.34 -2.74 1.53 -1.241.S3 -1.6 7.15
0
_3.57
0 0 0 -6.92 0 _5.63 0
1.70
20.91
-11.31 -1. 33 -1.17 _2.8S -9.07 -4.27 -5.04 5.55 22.61
·89 31 ee.es .JH.57 74.H .79.33 76.67 ·7".17
- ". .
-
15.12_87.07 ~8.93 ·~2.Z7 13.73_~3.04 7~9
- .,- -
83.55 _55.39 [00,61

109.3 t 17'1'.66 183.16 185.24 187.11 191.59 176.SU 182.07 167.9\ 181.H 120.51

Fig. Al5-S7.
- ."
Flnal Shear Flow Values.
--
ee.es ·81.' 14.4' .79.33
-
76.87.79.17
- -
78.33 .so.se - - --
rs.ra .a7 ,~7 5R.93 ·aZ.27 .3.13 _d3.04 ;2.96.72.45 - - -
U.S5 _55.39 100.61

(Note: Shear Flow at Ends of Webs Equal Sum of Shear Flows in Adjacent Sk1n Panels. )
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A1S.23

Procaading to (g)
cut

TXT
b a h
qg = qhg+100Z~ = 156+100 (1.25x2.5x
h
.094) = 156 +29.4 = 185.4 1b./in.
5" c - - (1) - - - - g J - - ( 2 ) - - rZA
qrg = 185.4+10OZ = 185.4+100 (-1.25)

11 Qa =0
d
Fig.Al5-55
g
(2.5 x .094) = 156 1b./in.

qrk = qrg - qra • -156 + 78 = -78

qkr =-78-100Zl<zA =-78-100 (-2.5)(2.5x

t
bZA
qb =_100Z =-100x2.5x2.5xO.125
a 0.125) =-78+78 = 0
= -78 1b./in.
Fig. A15.56 shows a plot at these calcu-
qc = qb-100ZcZA =-78-100 (1.25x2.5
lated values. The arrows give the sense ot the
b shear flows.
8
x 0.064) =-78-20 =-98 1b./in. ~7~8~)IDjjjJiIi;;;~~~~~~~iiiijjiii;;=
=-98-zdu 78
qd
c
0.064) = -98 + 20 = -78
=-98-100 (-1.25)(2.5x
I Fig.
98 (2) A15-56
qa =_78_Z aZA =-78-100 (-2.5)(2.5 x
d
0.125) = -78+78 = 0
j 78
qra • O_ZrZA = 0-100 (-2.5)(2.5 x 0.125)
a
=78 Fig. A15.54 shows the calculated static
At point (f) there are two other connecting shear flow values for the entire 10 cells. The
"falues are recorded at the ends of each sheet
sheet panels so we cannot proceed past this
panel and at the midpoints of each sheet panel.
joint in calcUlating the shear flow in the two Clockwise shear flow in a cell is positive shear
connecting sheets at (f).
flow. Since an interior web is part of two ad-
jacent cells, the Sign of the shear flow on
Thus we go back to point (a) and go
vertical webs is referred to the left hand cell
clockHlse, in order to determine whether sense is positive
or negative.
Qa = 0
Having determined the static shear flows
qha = 0+100zhzA = .0+100 (2.5x2.5xO.125)
a which will be referred to as qs' we can now
start the operations Table Al5.4. The first
= 78 1b ./in. horizontal row gives the calculations of the
twist at each c ell under the static shear flows,
With two other webs intersecting at joint
(f) the shear flow summation cannot continue which 1s relatively measured by the term
past (r), hence we go to paint (m) in cell (2) ZQs ~ for each cell.
where shear flow is zero due to the assumed
cut at point 1m).
With all mater1al effective in bending the
shear flow varies along each sheet. Fig. A15.56
~ = 0 shows this variation on the sheet panels of cell
qhm = 0_100Z
hZA
=-100(2.5x2.5xO.125)
(1). The term Zqs Lit is nothing more than the
zn area of the shear flow dia~am on each sheet
divided by the sheet thickness. To illustrate,
= -78 consider cell (1) in Fig. A15.56.
Now at joint (h) we have the shear flow at 78
Upper sheet panel: -
magnitude on each top panel, thus the shear
flow in the vertical web at (h) equals the sum
at these two shear flows or 156. Zq ...b. = -(0 + 78) ....§.&... +(0 + 78) 2.5 = 0
s t 2 0.125 2 0.125
Al5.24 SHEAR FLOW IN CLOSED
For lower sheet panel: -
ZQs ~ = a (Same figure as for upper sheet)
For lett hand vertical web: -
Treating the shear flow diagram as a
rectangle with height 78 and a parabola with
height 98 -78 = 20,
l:<ls.k=-78XS _(20xSXO.667) =-7142
t 0.064 0.064
For right hand web or cell (1),
The shear flow diagram is likewise made
up at a rectangle and a parabola.
zn_J..= lS6XS+ (186 S-lS6)S 0.667 -9342
'" t 0.094 • x 0.094 -
Therefore for entire cell (1)
Z% ; = -7142 + 9342 = 2200, whi ch is the
value in row (1) or Table A1S.4 under cell (1).
The results for the other nine cells as calcu-
lated in a similar manner are recorded in row
I ot the Table. The procedure as followed in
the remaining rows or Table 4 is the same as
explained in detail for example problem 1. In
Table Al5.4 only one reiteration is carried
through as the values in the bottom or last
row are practically the same as arrived at atter
the firth carryover cycle. Adding the con-
stant shear tlow values in the last row in the
Table to the static shear tlow values in Fig.
liS.54 we obtain the tinal shear flow values
of Fig. Al5.57. The resultant of this shear
flow pattern is a force at 8750 acting dawn
in the Z direction. Its location would be
through the centroid at the shear flow force
syatem.. Let X equal distance from upper lett
hand corner of beam to line of action at shear
flow resultant~force.
Taking moments of shear flow torce system
of Fig. Al5.54 plus the constant shear closing
values in each cell as given in the last row
or Table Al5.4 and equating to 8750 i, we
obtain;
Due to uniform static shear flow on each
Web: -
M = (lS6 xS) (S + 10 + lS + 20+ 25 + 30+ 35 beams where this is not the case, the shear flow
+40+45) +78x5x50 = 175000
Due to parabolic static shear flow in each web:-
M = (29.4xSxO.667)(S+10+1S+20+2S)
+ 25.3x5xO.667(30+35) +22.5x5xO.667
(40+45) +20xSx 0.S67xSO = 22520
Due to constant clOSing shear flows as
THIN-WALLED SECTIONS. SHEAR CENTER.
listed in last row in Table A15.4: -
M = +(-11.31-3.S7 -1.33 -1.17 - 2.88 - 9.07
-4.27-S.04+S.SS+22.51) 2x25 =-S20
Total Moment =
175000 + 22520 - 520 = 197000
hence X: = 197000/8750 = 22.5 in , , which equals
the distance from left end of beam to shear
center location.
Referring to the final shear flow values
in Table A15.57, it will be noticed that the
final results are not much di:ferent from the
assumed static shear flows with the possible
exception of the two end webs. If we had
assumed all the webs cut except one to form the
static condition, then Table AlS.4 would have
required several times as many carr-y-over- cycles
to obtain the same accuracy of final results.
Al5. 14 Use of Successive Approximation Method for Multiple
Cell Beams when SUbjected to Combined Bending and
Torsional Loads.
The internal shear flow reSisting force
system for a beam subjected to bending and twist-
ing loads at the same time is carried out in VNO
distinct steps and the ~esults are added to
given the true final shear flow system. First,
the shear flow resisting system 1s found for
being without vnist as was explained in this
chapter. The results of this first step locates
the shear center. The external load system- is
then transferred to the sheat center, which
normally would produce a torsional moment about
the shear center. The internal resisting shear
flow system to balance this torsional moment 1s
then handled by the successive approximation
method as explained and illustrated in detail
in Art. A6.13 or Chapter A6.
A15.15 Shear Flow in Cellular Beams with Variable Moment
01 Inertia.
The previous part of this chapter dealt
entirely with beams of constant moment of iner-
tia along the flange direction. In airplane
wing and fuselage structures, the common case
is a beam at non-uniform section in the flange
direction. In cases where the change of the
crosB-sectional areas is fairly well distributed
between the various flange members which make up
the beam cross section, the shear flow results
as given by the solution for beams ot constant
moment of inertia are not much in error. For
results may be considerably different from the
actual shear flows. This fact will be illus-
trated later by the solution of a few example
problems •
AlS. 16 The Determination of the Fle:mral Shear Flow
Distribution by Considering the Changes in Flange
Loads. (The 6,P Method. )
Fig. A1S.58 shows a Single cell distributed
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
A15.25

flange beam. Consider the beam acts as a canti- flange members, see Fig. A15.60. Again ZFy =0,
lever be~~ with the bending moment existing at
section (A) being greater than that existing at whence ~P3 + ~P~ + qyd =°
secti~n (B) and that the bending moment produces
compression on the upper surface. By the use of _ (6P,. 6P.)
orqy-- d
the flexural stress equations, the bending
stress on each stringer can be found, which if
multiplied by the stringer area gives the Therefore starting at any place where the
stringer axial load. Thus at beam section (B), value of qy is known, the change in the average
let p~, P~, P3 , etc. represent the axial loads shear flow to some other section equals
due to a bending moment M. The external bending
- - (1)

It the summation 1s started where qy is zero

then equation (1) will give the true average
shear flow qy.

Fig. A15.6l shows sheet panel (3,3 1 ,4,4! )

isolated as a tree body, Taking moments about
corner 4' and equating to zero for equilibrium,
l:M =d(6P,)b_ bd=O
~l d qx

Whence, qx = ~P3/d - - - - - - - - - -(2)

Thus for rectangular sheet panels between
flange members the shear tlow qx or qz equals
the average shear qy.
Flg.A15-58 The same rules as previously presented to
determine the sense ot qx or qz atter haVing qy
can be used and will not be repeated here.

To show that equation (1) reduces to the

shear flow equation preViously derived and used,
consider a beam with constant cross-section and
take a beam length d = I inch. Then,

I\I'l = Vzd = Vz(l) = Vz

ilP =~ZA =..:!.z...ZA (Where A = area of
Fig.A15-59 Fig.A15-60 Fig.A15-61 Ix Ix stringer) .

moment at section (A) is M + ~, hence the From equation (1) qy = Z6P

stringer axial loads at section (A) will equal
P~ + ~p~, P a + ~Pa' P 3 + ~P3' €tc. These
stringer axial loads are shown on Fig. A15.58.
Imagine the upper sheet panel 2, 2 1 , 3,
3' is cut along line (a-a). Furthermore con-
sider stringer number (3) cut out and shown as
a free bOdy in Fig. A15.59. Let qy be the
average shear flow per inch over the distance
d on the sheet edge bb. It has been assumed
pOSitive relative to sense along yaxis.
For eqUilibrium of this free body,
ZF y = 0, hence ~P3 + qyd = 0
whence qy = - ~P3/d
For a free body inclucir~ two str:ngers or
SUbstituting value of ~P found above,
qy =_:!.z..
Ix
ZZA - - - - - (3) which Is the
shear flow equation preViously derived for beams
with constant moment of inertia.
AlS.17 Example Problem to Compare Results in Using
Equations (1) and (3).
Fig. Al5.52 shows a square single cell beam
with six flange stringers. Between points Band
C, the beam has a constant flange section which
is shown in Section 8-B. ~he numerals beSide
each stringer represent the area at the stringer.
Between points B and D, the flange material
tapers unifon4ly with the flange material at
Doint A as indicated in Section A-A. It should
be noticed that the increase in flange area 1s
Al5.26 SHEAR FLOW IN CLOSED THIN-WALLED SECTIONS SHEAR CENTER

t!A
~A
'i
B

lID with the shear flow system of F1g. A15.63, which

therefore 1s the tinal shear flow system for
this method of solution.

~ 20" -+- B

30" ----I
10--ro" ~ bl"J"
Ole Fig. AlS-83 Fig. A15-84
~-l c' I b' ...1.. Solution No.2. Considering AP Loads in Flange
. k" +."~ 11-5"+5"..,1 Stringers. (£guation 1)
SECTION A-A SECTIO;," B-B TO "'-C Bending moment at section AA = 1000 x 50 ::: 50000
I x=SxS 2x2=250in4 Ix=3x5 x2=150in
in.lb.
Fig. Al5-62 Bending moment at section BB = 1000 x 30 ::: 30000
obtained by increasing only corner stringers b in.lb.
and c. The shear flow on section A-A will be
computed using equation (3) which applies only Considering Section 8-B:
to beams with constant section and also by Bending stress intensity at midpOint at
equation (1) which applies to beams with vary- stringers by the tlexural formula:
ing moment ot inertia.
- M.:, _ 30000 x 5 - 1000 pst.
"b-I 150 -
Solution 1. USing Shear Flow Equation for x-
Beams at Constant Cross-8ection. Axial load in each at the stringers a, b, and
(Equation 3) c = 1000 x 1 = 1000 lb.

Since q at any pOint on the cell 15 un- ConSidering Section A-A:

known, it will be assumed that the upper sur-
tace on Section A-A 15 cut through the midpoint _ liz = 50000 x 5 - 1000 pai ,
at r Lange stringer (a), thus making the shear
crb -IX 250 -
flow qy equal to zero on this free surface. Axial load in stringer (a) = 1000xl = 1000 lb.
One-halt at stringer (a) thus acts with each
side at the top surface. In this solution the Axlalload in stringer (b) or (c) ::: 1000x2:::
webs and walls will be. assumed ineffective in 2000 lb.
resisting bending stresses, thus the shear flow
is constant between adjacent stringers. These resulting axial loads are shown act-
1ng on the portion between points A and 8 in Fig.
Starting at midpoint of stringer (a) and A15.65, which equals the results as shown in Fig.
going counter-clockwise around cell, A15.66.

~ =0 (assumed cut) ,., # '"

~ .... ~
=-~ l:ZA =- f~oO x 5 x 0.5 = -10 lb/in I / /

d'~h,4
/ J'" I";""'~ /Iw'
'lac
'leA = -10 e • b
1000
'tcc ' • -10-25QX5x2=- 50 Ib/in. l ~V '~~ .. -
Proceeding around the cell the balance ot the
shear flow ccufd be ca.Lcu.Ia'ced , but due to sym-
/ .!'I;;~I ;§Or ":lIb'
metry enough values have been found for the ~ Fig. Al5-8!i ~ Fig. Al5-66
shear flow to draw the complete shear flow piC-
ture tor bending about the X axis when it is Having tound the AP flange loads over a
assume~ that one-halt ot the area of stringer length (d) of 20 w , the shear flow can be com-
(a) acts With each adjacent web. Fig. A15.63 puted by equation (1).
shows the resulting shear flow diagram. The
resultant at this shear tlow pattern is a It will be assumed that one-halt at the 4P
1000 lb. torce in the Z direction and its load in stringer (a) will tlow to each adjacent
location through the midpoint at the box since wet. However, there 1s no AP load in stringer
the tlow 1s symmetrical. The external load at
1000 lb. also acts through the midpoint or the
(a) hence Qab Qac = =
O. Then from equation (1),
cell hence the external load 1s in eqUilibrium qbb' = O-Z~= O-~= - 50 lb./in.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AlS.27

t.Pb' -1000
qb I a I == -50 + Z -d- :: -50 - 20 == 0

Due to symmetry the lett side at cell would

give the same results. The results are plotted
in Fig. AlS.64. Since the increase in section
moment ot inertia between beam points 8 and A is
increasing at the same rate as the external
bending moment, the average shear of 50 Ib./in.
is constant between the ~NO beam section A and
8. Comparing Figs. 63 and 64, we tind the
first method gives a shear flow ot 10 lb./in. Fig. A15-70
in the top and bottom webS, whereas actually it
is zero. This seems reasonable since the entire
increase of flange area was placed in stringers
band c.
Example Problem No.2

The same beam as in Problem 1 will be used

except that the cross-sections at beam points B Fig.A15-71
and A are as shown in Fig. Al5.67. The increase
in flange area between beam points 8 and A has whence q:l, = Vb a
been placed entirely in stringer (a) which b:
Changes from 1 sq. in. at 8 to 3 sq. in. at A.
But V = qgb a , whence
3
•
1
b c a q, =(~:)"q. - - - - - - - - - - - - - (4)
Section Ix = Ix' Section Fig. AlS-57
A-A 250in 150 in4 B-B From Fig. A15.71
c' a' b' c' .' b'
1 1 1
'la=.:!...=~'
b. 0.
hence q. = b %
6•
The results at using equations (3) and Substituting value of q in (4}
(1) relative to the shear flow pattern are
given in Figs. A1S.68 and A15.69 respectively.
(The stUdent should check these results.) It
q, =0:) % - - - -- - -- - (5)

should be noticed that the true shear flow is
greater in the top and bottom skin than that
given by equation (1) which applies only for
beams of constant cross-section.
Thus having the shear flow on the stringer
edge of the sheet panel, the shear flow on the
large end of the tapered panel can be found by
equation (5).
A15.19 Example Problem of Shear Flow in Tapered
Multiple Flange Single Cell Beam..

" Fig. A15.72 shows a tapered single cell

beam with 6 spanwise stringers or flange members.
The beam 15 loaded by a 1000 lb. load located as
Fig. A15-58 Fig. A15-59 shown. Assuming the webs ineffective in bending
AlS.18 Shear Flow in Tapered Sheet Panel.
the internal reSisting shear flow pattern will
be de'termtned ,
Major aircraft structural units such as
the Wing, fuselage, etc., are tapered in both In this solution the shear flow at Station
plan form. and depth and therefore the sheet 120 will be determined by considering the ~p
panels between flange members usually are flange loads over a length of 30 n or between
tapered in width. Fig. A15.70 shows a canti- Stations 90 and 120.
lever beam tapered in depth and carrying a
load V at its end. The flange reactions at Consider section at Station 120:-
the left end have been found by statics. A Me 1000 x 120 x 5
free body diagram of tr.e web is shown in Fig. Bending stress O'b =-1- == 450 1333.33
psi.
AlS.7l. Take moments about paint (0) and
equate to zero.
The horizontal component of the axial load
ZMo =(~~) b. - q,b,d = 0 in a stringer equals O'oA (where A = area of the
stringer) •
Al5.28 SHEAR FLOW IN CLOSED THIN~WALLED SECTIONS. SHEAR CENTER.

Since panel ab is tapered in width from 6"

Fig. A15~72 at station 120 to 5.5" at station so, the shear
f-ll"-1 flow ~(ab) at station 120 can be found from
I 25 I
T3.:,; a Il" c 1.75
equation (5).

8.5"
L , a'
Ix=271
b' c
Sec. SU. 90 qX(ab) = 24.87x5.5/6 =- 23.2 1b./in.
3.25 2.5 75
qY(bc) = 24.87-471/30 = - 40.57
1000f
qX(bC) =- 40.57x5.5/6 = - 37.2
40.57-196.6/30 = - 47.12

qZ(cc') = 47.12x8.5/10 = 4O.01b./in.

Since the 6P loadS are the same on the

lower stringers but tension the shear flow cal-
culations if continued would give the same
values as found on the top surface. Fig. A15.74
shows the shear flow pattern on station 120.
Pa =4 X 1333.33 =5333.33 lb.
IN PLANE FORCES PRODUCED BY INCLINATION
Pt = 3 x 1333.33 = 4000 lb.
OF FLANGE MEMBERS
Pc =2 x 1333.33 = 2666.7 lb.
Since the box tapers in depth and width, the
Consider beam section at Station 90:- flange stringers are not normal to section 120,
thus X and Z force components are produced on
- 1000x90x4.25 - 1411.5 pSI. section by the stringer loadS.
,,~ - 271 -
hence stringer loads are, These in plane force components are:-
Pa =3.25 x 1411.5 = 4587.4 For stringers a and a' ,
Pt = 2.50x 1411,5 = 3528.7
Pc = 1. 75 x 1411.5 = 2470.1
Px =5333.33 x 2/120 = 88.9 lb.
Pz = 5333.33 x 3/120 = 133.3 lb.
The change in axial load IU' in the
stringers between stations 90 and 120 equals For stringers b and b l ,

the difference between the above loads, whence

<\Pa =748, <1Ft =471, <\Pc = 196.6
Px = 4OGOx 0/120 = 0
Pz = 400G x 3/12G = 100 lb.
FIg. Al5.73 shows these ap loads acting
as wIng portion. Since shear flow Is unknown For stringers c and c' ,
at any point, assume q equals zero in web aa' .
The average shear flow in each sheet panel Px = 2666.7 x 2/120 = 44.4 lb.
=
over a length d 30 inches.can now be calcu- Pz = 2666.7 x 3/120 = 66.7 lb.
lated by USing equation (1).
Fig. A1S.74 shows these in plane force
qy = 'la' a - -Z<\P
d- Qar a • 0 (assumed) components due to the flange axial loads.
10",,*
133.3 1 100 66.7
- 24.87 Ib ./in.
889
.- 1,3"1
a -
I
b - - -<; -44.4
f

1
10"
23.2 37.2

40.0~
t
j
Fig. A15-74

J 88.9

133.3
'23.2 b' 37.2
8" 1
lOa
8"
-
;
j -
66.7
44.4

Fig. A15-73
/ I
746 471
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A15.29

The forces in Fig. AIS.74 will be checked
for equilibrium.
ZFz = 1000 - 266.6 -200 -133.4 -10 x 40 = 0
ZFx = 0 by observation.
Take moments about stringer (a),
l:Ma = -3 x 1000 + 200 x 6 + 133.4x 12 + 10 x 40
x 12+(+88.9 - 44.4)10 + 23.2 x 6 x 10
+37.2x6xlO = 8670 in.lb.
Taking moments about point a of the forces
in Fig. Al5.74 but replaCing the shear flows in
the top and bottom panels by the values found
above, we would obtain an unbalanced moment at
7970 in. lb. instead of 8670 preViously found.
The correcting shear flow would then be g =
-7970/240 =
-33.2 instead of -36.1 as previously
rcuno , The tinal shear flow pattern would be
as shown in Fig. AIS.76, which values should be
compared to those in Fig. A15.75.
i'.4 2.1

For equilibrium a 33.2 Fig. AI5-76

moment of -8670 is 15.4 2.1
needed, which is
provided by a constant 12.9 1.1
shear flow q =
M/2A = A15. 20 Problems
-8670/240 = -36.1 lb./in. 36.1 3.9
Adding this constant
shear flow to that of 12.9 1.1
Flg. A15.74 gives the 1000 r"..j 2000 lb.
final shear tlow as
shown in FIg. A15.75.
Fig. Al5--75
!-,,,·tp r- 12 j'-t-- 12;'-I
Trr===;.0;;'3:=:'=;i 1 ·.u04 cf4
Solution No. 2 10''' ~,05 _ ...072 1,2''" - .072 • _ -.04
..L [.03 L I
This same beam and loading will now be .2U''-----1 \03
solved using the shear flow equation derived Fig. AI5-77 Fig. A15-78
for beams of constant cross-section.
(1) Determine the resisting shear flow pattern
Since the stringers are not perpendicular for the loaded Single cell beam as shown in
to the beam cross-section they have a z force Fig. AIS.77. Assume load P = zero in this
component 'Nhich thus assists in carrying the problem. Assume all material eftect1ve in bend-
external shear load in the z direction. These ing. Make two solutions, one or them involving
Pz components at station 120 have been calcu- the use at the shear center.
lated in the other. solution.
(2) 8ame as problem (1) but add load P = 1000 lb.
ZPz =- (2xl33.3+2xlOO+2x66.7)=-600 lb
(3) FIg. A15.78 shows an unsymmetrical Single
Total Vz (ext.mal) = 1000 lb. cell beam loaded as shown. Assume all material
effective in bending. Determine resisting shear

'TD
Let VZ(net) be shear load to be taken by flow diagram..
cell walls, 400 lb.

VZ(net) = Vz - ZPz = 1000 - 600 a 400 lb.

5 L." 1.
Calculation ot static shear flow assuming
q in sheet panel aa" is zero.
I
=
10" .04 .03
Fig.
A15-80

ZZA =~x5X4=17.8 -.L B

(.4 sq.in.
Fig.AI5-79
450 lb.in.

This corresoondS to value ot 23.2 in (4) Flgs. Al5.79 and AlS.eO show two loaded
previous solution (see Fig. A15.74). single cell - 2 flange beams. Assume the flanges
develop all the bending resistance. Determine
qY(bC) = 17.8+~gX5X3 = 31.1 lb./in. the shear flow reSisting pattern by two solu-
tions, namely, without and With use or shear
As canpared to 37.2 in Fig. AlS.74. center.
qy(CC') = 31.1+~gx5x2 = 40.0 lb.ln. (5) Fig. AIB.Sl shows a Single cell - 3 flange
beam SUbjected to loads as shown. Assume the 3
which is the same as in Fig. Al5.74.

-j r
)
TC
 A15.30 SHEAR FLOW IN CLOSED THIN-WALLED SECTIONS. SHEAR CENTER.

flanges develop the entire bending resistance. (11) Fig. A15.84 shows a ~ultiple flange-ctrcu-
Determine the internal shear flow reststing lar beam section. Find the reststing shear
system. flow pattern when carrying the external shear
300 lb. 60"" load of 5000 lb. as located in figure. Assume
100 lb. 1,1

r~.02 f
45°
10"
Q
A
.0' '.02
cell skin ineffective in reSisting beam bending
stresses •
t1000 lb.
l-_~
.8....., I"
I
bl--
'\_1"
B
l -F-
.03
( . ' \ 500 lb • 0.5
tV= 3000 lb.
.1
10'Y
r 81< "'"1

rU
.04 I
Fig. A15~81 Fig. Al5-82
.03
.3 .1 .1 .1
(6) Find shear center location for beam in
Fig. AlS.81 if the 3 flanges provide the entire [-'0'1'0''"1 I 4 panels a 5" = 20"
bending resistance. Fig. A15-85 Fig. A15-86
(7) Find the internal resisting shear flow (12) Determine the shear flow resisting system
pattern tor the 3 flange-single cell beam of for the beam section of Fig. A15.85. The 6
Fig. AlS.82. Assume webs or walls ineffective flanges r~ve areas at 0.2 sq. in. each. Skin
in bending. is .032 thickness. Assume skin ineffective in
bending.
(8) Determine shear center location for beam
at Fig. A15.82. Webs and walls are ineffective
in bending.

.3
20
1 00* .4
(13) Find the shear flow resisting system for
the unsymmetrical beam section in Fig. A15.8B.
Flange areae and skin thicknesses are given on
figure. Assume skin ineffective in bending .

Top Skin .03 (14) neterrune shear center location for beam
.04[ section in Fig. A15.SB.
Fig. A16-a3 .035 10"
Bottom Skin .03 T f 500 lb.
.3
f--- 8 panels at 5"

(9) Fig. AlS.83 shows a mUltiple flange-single

=30"
-1
,4
T8"
...L
.06"
.05

.05
.03

1--12"--+-8" -I
[:303

.03
.04 $02 I

cell beam section. Find resisting shear flow

system ir webs and skin are ineffective in
resisting bending stresses. All skin flange
members have area of 0.1 sq. in. each.
(10) Find shear center location for beam
section in Fig. A15.83, it webs and skin are
ineffective in bending.
Fig. A15-87 FIg. A15-88
(15) Fig. A15.S7 shows a 2 cell beam section.
Consider all material effective in reSisting
bending stresses. For the given beam loading
determine the internal reSisting shear flew
system.

(16) Find shear center location for beam section

In Fig. A15.87, All material effective in bend-
Ing.
(17) For the 2 flange-2 cell beam in Fig. A15.88,
R 320" All flange areas deter.nine the reSisting shear flow pattern when
5" = 0.1 sq. in.
;---t 5" each. Skin
beam section is loaded as shown. Webs are in-
x-.--- -- - - ;---t 5" thickness .035 :II
effective in bending.

7" (18) Fig. A15.89 shows a 4 cell beam section

~---t with 6 flange members. Assume walls and webs
ineffective in bending.

(a) As a first problem assume trat the left

FIg. A15-84 and right curved sheet panels are removed,
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AlS.31

i 8" 1 1000
lb.
flow equaticn for
lat! .SCI" .5C1" ;-12.5''1 beams of const~~t
I 04 Cell Cell
. Cell Cell 03
section. Campare
the results. All six
12" . (I) .05 (2) .04 (3) .04 (4) . flange members have
.L .03 .03
Fig. A1S-92 0.2 sq. in. area
each at section A-A
and taken uniformly
1 - 1 2 " + 8"--j to 0.1 sq. in. at
end c-c,
FIg. A15-89
1000f
leaving a 2 cell beam section. Find the 'S
resisting shear tlow system under the et r5",5"1
,r 1,
gl ven loading.
j ~iH
-r-r- ,
(b) Now add the lett curved sheet to form -=---120'~'- - - - '
cell (1) thus giving a 3 cell beam. Find
shear flow system.
(22) Add two interior webs to the beam of Fig.
(0) Now add the right curved sheet to A15.9Z, connecting flange members a-a~ and b_b n ,
form cell (4), thus giving a 4 cell beam thus making it a 3 cell beam. Find the shear
section. Find the shear flow system. flow resisting pattern at section A-A by the
liP method.

(19) Fig. A15.90 shows a 10 cell multiple Fig. Al5-93

flange beam section. Area ot each ot the 22
flange members equal 0.3 sq. in. Assume webs
and skin ineffective in resisting bending
stresses. Find internal resisting shear flow
system for 2000 lb. shear load acting through
shear center of beam section. Find location
at shear center. For solution use method ot
successive approximation.
(23) Fig. A15.93 shows a ciycular Single cell
beam with 8 flange members. The area of each
flange member 1s 0.1 sq. in. throughout the beam
length. For the given 400 lb. externaj loading
determine the resist1ng shear flow pattern at
section A-A ~s1~g the ~p method over a distance
of 25 inches between sections A-A and B-B.
Assume cell wall ineffective in resisting bend-

-x- f·1 ~
S;,-.I!R1S!3}0G 10 .1 -x-
ing stresses.

-~c:-=-::-;-;::;""---I
\ 0.1
7 cells a 6" : 42"
.08

Fig. A15-91

(20) Fir.a the shear center for the 7 cell beam

section of Fig. AIS.91 for bending about XX
axis without ~Nist. All beam material effect-
ive in ~ending. For solution use method of
successive apprOximation.
(21) Fig. A15.9Z shows a Single ce11-6 flar.ge
tapered beam carrying a 1000 lb. load as shown.
Calculate resisting shear flow pattern at
section A-A by two methods. (1) By ~p method
over a distance of 20 inches between sections
A-A and 8-3, and (2) By USing general shear

-'
i:t:L
'I
AlS.32 SHEAR FLOW IN CLOSED THIN-WALLED SECTIONS. SHEAR CENTER

.:1,f!;1

~£!W.~~~::'~'~ ~
;~'i
DC-a Fuselage Main Cabin Framing Structure
;;r:;:~s.!:.~, "':~~' -.~. ,~:.,,·~;~-,;i Also Window Belt Treatment
~·.;..t. " ~ :.,. :~r~;::·;-l~J;i;;;.iMf·:;:;·S!:~
•

DC-a Fuselage Lower Cargo Compt.

Basic Structure

DOUGLAS DC-a JET AIRLINER. BASIC FUSELAGE CONSTRUCTION

&
 CHAPTER A·16
MEMBRANE STRESSES IN PRESSURE VESSELS
ALFRED F. SCHMITT

A16. 1 Introduction. at adjacent meridian curves and a pair of ad-

The structural designer IS orten called Jacent parallels.
upon to develop a vessel which Is to contain a
fluid under pressure. Occasionally the design
of such a vessel 1s not critical trom either a
weight or shape standpoint and almost any suit-
ably strong sealed vessel will suffice. More
often, the strength, weight and torm ot such a
unit are closely prescribed and rigidly con-
trolled. Thus, the pressurized cabin or a
modern aircraft Is a sealed pressure vessel con-
taining an atnosphere at near sea level pres-
sures and whose functional requirements include:
1 - the transmIttal ct heavy loads from
the tail surfaces and trom internal dead loads,
11 - the necessity for nonstructural cut-
outs for doors and windows, •
111 - an efticient shape tram both the aero- Fig. A16. 1
dynamiC and space utilization points at view,
tv - a minimum or weight. The radii Rm and Rt shown on the figure
are found by erecting local normals to the sur-
structurally, the most efficient torm at tace at the element at its corners. Rm is the
pressure vessel is one in which the lateral radius of curvature at the meridian curve: it
pressures are supported by tensile stresses may be either positive (inward pointing), nega-
alone in the curved walls at the vessel. Ex- tive (outward pointing as in Fig. AI6-l), or in-
amples at such Shapes are those assumed by finIte (at inflection points or straight-line
pressurized rubber balloons 8-~d canvas tire meridian segments). Rt is the radius at curva-
hoses and by the tree surface at a drop at ture of the section normal to the meridian
'Hater (in which the surface tension forces pro- curve. For Simple forms of pressure vessel Rt
Vide the support). The walls of these vessels is always positive; all radii Rt point inward
have zero bending stlftnesses and hence have the and intersect the axis 0-0, although not gen-
prooerties ot a membrane. The stresses develop- erally normal to 0-0 (see radius Rt erected trom
ed, lying wholly in tangential directions at pOint C or Fig. A16-1).
each point, are called membrane stresses. Fig. Al6-2 is a detail at the surface ele-
In shells of technical tmportance, the ment. The forces per unit length· in the merid-
walls do, of course. have same bending stiffness ional and tangential directions are denoted by
and hence may carry some transverse loadings by
flexural stresses. Indeed, the boundary con-
Nm
and Nt, respectively. Shear stresses are
urt rone imposed on the, shell may be SUC~l as to
necessitate some localized bending near edges I"·
• "
=.--
and seams. An efficient pressure vessel design R
•
is one in which the configuration minimizes these
nejar-tures trom a tpue membrane stress system,
i.e., nimniaas the degree of local bending
stresses induced.
',-
d ••
[J -',
I"• \ ".
AlS.2 Membrane Equations of Equilibrium: Shells of
Revolution Under Rotationally Symmetric Pressure
Loadings.
',- fftUtl1t - - It
Consider the equilibrium of a differential
element cut from the srEll of revolution of Fig.
• Fig. Al6.2

A16-1. (The figure is drawn to resemble a

familiar folding paper Christmas bell, since
such an Object may aid in Visualization.) The "
• Hereafter referred to as "stresses" although their
element is cut out by the intersection at a pair units are pounds per inch rather than psi.

Al6.!

i ../r-,
i -.;- '/
Al6.2 MEMBRANE STRESS:c.S IN PRESSURE VESSELS
absent due to the symmetry of the problem. The
included angles between the pairs of nerldlo~~l
and tangential rcr-ces are :is!ll/Hm and dSt/:Rt,
respectively.
Summing forces normal to the dlf:erential
element, one has
or,
p - - - - - - - - - - - - - - - -
Here p is the Internal pressure, positive out-
ward. Note that the shell wall thickness does
not enter eq , (1). The pressure p may vary in
the meridional direction but is constant in the
tangential direction by hypothesis (rotational
symmetry assumed).
Eq. (1) is one equation containing two un-
knowns. Another equation may be obtained by the
condition of equilibrium ot a portion of the
shell above or below a parallel Circle. ThUS in
Fig. A16-3, the pressures acting dmvnward on the
lower portion at the shell are equilibrated by
the upward vertical components at the meridional
stresses, Nm.
Fig. Al6.3
Summing torces vertically
p rt (~ sin ~)' = 2 rt (~ sin ~)
Solving,
Nm = P Rt - - - - - - - - - - -
2
Eqs. (1) and (2) determine completely the
membrane stress state in the rotationally sym-
I'letric shell problem: the problem is thus seen
to be statically determinate.
We note that eq. (2) should not be used in
cases at hydrostatic pressure loadings. The
basic cqncept is that of shell eqUilibrium, and
consequently tor this class of ~roblema the
manner of shell support must be considered.
Thus, in the tank ot Fig. A16-4a, the upper cy-
lindrical portion requires no meridional stress-
es sL~ce the load is reacted at the supporting
ring O~. (In these analyses the structural
weight, wnich always requires some stresses for
its support, Is neglected.) In the lower hem-
• •
Fig. Ale.4a Fig. A16.4b
ispherical portion meridional strssses are re-
quired as shown in Fig. A16-4b. Hence, in this
class of problems it is best to derive the nec-
essary Second equilibrium equation (correspond-
ing to eq. 2) by considering the individual
characteristics of the struct~e.
Al6.3 Applications to Simple Pressure Vessels.
Example Problem 1
Determine the membrane stresses 1~ a cylin-
drical pressure vessel of circular cross section
(radiUS Ro), haVing hemispherical ends, i! the
internal gas pressure is p. Also find the
greatest combined normal strese.
Fig. Al6.5
SOLUTION:
In the hemispherical ends Nm = Nt by
symmetry and, at course J Rm = Rt = R o•
I'm sin ~ Hence eq. (1) is SUfficient to determine the
stresses in this portion ot the structure. One
has
(2)
Nm P R.
Stress = t = 2t
In the cylindrical portion the radii are
Rm = 00 (the curve at Fig. A16-5 is the meridian
curve and this is a straight line fer
the cylindrical portion).
Eq. (1) becomes
I'm.
00
•

ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A16.3

Nt = P RQ

R, 2.7 Ho P
Stress =Nt =p Nb = 2
t t
Therefore
The meridional stress in the cylindrical portion
Is found from sq. (2); Nb cos ~ = 2.7:, p (.92.)

p R,
2
Stress =P R o Fig. A16-6c)
2t
Since shear stresses are absent everywhere
in the meridional and tangential directions)
these are the directions of principle stresses.
Hence the greatest combined normal stress Is
identically equal to the greatest meridional or
tangential stress as just computed. It Is seen
to be
"max = p R,;t
~ple Problem 2.
.~ important problem in pressurized cabin
design concerns the Shape ot the end bulkheads.
;fflhl1e hemispherical bulkheads (such as used !n
Example Problem 1) are highly desirable from a
stress standpoint, such fQ~ are uneconomical
as regards space utl11zat~on. on the other ex-
treme, a flat bulkhead, while ?roviding far more
useful volume, cannot resist the pressure load-
ing by membrane stresses and hence 15 structur-
ally inefficient. A compromise configuration is
that shown in Fig. Al6-6, in which the bulkhead
is a spherical surface of low curvature,
(~dished head~) supporting the pressure loading
by membrane stresses. A reinforcing ring,
placed at the seam, resists the radial component
of these s-tr-eeaes . Problem: find the com-
pressive load acting in the reinforcing ring.
= 1.25 Ro P
Finally, the compressive ring load, F, is (ret.
2F = r:1.25 R, p sin 9 R, d 9 = 2.5 R~ P
In computing reinforcing ring stresses
tram this result it is necessary to include
some effective skin frmn the adjacent shell
walls when the ring cross sectional area is
figured.
.
Example Problem 3.
Another form of bulkhead used to close a
circular pressure cylinder is elliptical in
section as shown in Fig. Al6-7. Such a bulk-
head shape provides tangential meridional forces
at the seam (requiring no reintorcing ring as in
the last example) and yet is reasonably s!flcient
as regards space utilization. Problem: de-
termine the membrane stresses in such a bulk-
head.
Fig. A16.7 1..

Solution: In cartesian coordinates the equation

of the bulkhead meridian is

From the calculus, the radius of curvature of

this ~eridlan curve is

Solution; as shewn in the explOded view, Fig.

A16-6b, the radial components of the Julkhead
stresses are resisted by the reinforcing ring,
the cylinder wall bei~g presumed to after no
resistance to concentrated transverse :orces. The radius Rt 1s found ~ost readily by ob-
serving that it is normal to a tangent to the
For this case 9 ~ arc sin 2:7 ; 21.7°. meridian curve (see figure). After finding the

J""'~
,~
,1,
A16.4 MEMBRANE STRESSES ill PRESSURE VESSELS

slope of the tangent (angle $), one computes SOLUTION:

Rt = X/Sin~. The results are
From sq. (1) the "hoop" loadings in the
upper and lower cylindrical lobes are

The meridional stress ts found from eq, (2). Summing forces horizontally at the floor joint:
Thus,
Nt = P (RU cos al, + RL cos a.~) •
_ P Rt _ p ("a:..'..;.Y,-'-,--:,--,,--,-,--
Nm--2--~- Assume all stresses are equal and are given by
s.
Substituting the expreSSions for Rm, Rt and Nm
into eq, (1) one r inds Su = s = P Ru/tu

SL = S = P RLitL
Sf = S = f:t (RU cos al, + RL cos a, )

or particular interest are the stresses at Letting the weight denSity of the material
the seam. Here y = a and x = a. One tinds be wJ the weight per unit length (axially) along
the cabin is (w times the developed length of
walls and floor).
Nm=~
w = ',V 12 tuRu (n - a,) + 2tLRL (n - a.)
Nt=pa(l-~) + 2 tt RU sin a:l. } •

This last result is import~~ since it in- Solving for the various thicknesses tram above
dicates that compressive tangential stresses are and substituting, one finds (to obtain a result
possible it a> b VZ. As will be seen below, symmetric in appearance use was made of the
such a situation is undersirable because at the tact that RU sin a:l. = RL sin aa)
large resultant difference in radial expansions
between the cylinder and bulkhead (the bulkhead
actually contracts radially it Nt is negative) W =Z; 1RU [n - (1:l. + ~ sin 2 a:l.) +
producing high seconcary bendIng stresses in the
vessel walls.
R~ (n - a, + ~ sin 2 a,) f
Example Problem 4. Since one may show that the cross sectional ar-ea 01 this
Determine the weight per unit length or the fuselage Is
double-cyllnder tuselage cross section as a
function ot the internal pressure J allowable A z
,r.
RU\'1t - a1 +
1
'2 sin ),1\1t
2 U:l. t RL • 1. )
aa + '2 Inn 2 a ••
stress and the geometric parameters ot Fig.
AlS-8. For structural efficiency it is desired an important consequence 01 this calculation is that the ratio
of shell weight to shell volume 1s
to maintain equal membrane stresses in the skin
and raccr-, W. 2wp
A --.-
and Is therefore independent of the combinations of Re, RL,
U:l. and a. used. The designer 1s thus free to choose these
shape parameters 80 as to satisfy other requirements.

Example Problem 5.
Determine the membrane stresses in a
conical vessel of height h and halt apex angle
Fig. A16.8 a. The cone forms the bottom of a large vessel
f1lled With a liquid of speCif1c weight w and
havtng a head of l1quid H above the cone. The
compt.ete..unt t is supported from above.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AlG.S

these secondary stress peakS, ar~~ing that local

yielding of the material will level them out.
__J-r _--
~--
However, such stress peakS may prove to be of
great importance in cases of ~epeated loadings
, H wherein fatigue failure g; considered likel?
To point up the major weakness in the mem-
brane analYSiS one need only compute the radial
displacements In the two different elements that
make up the pressure vessel of Fig. Al6-5, viz.,
Fig. Al~. 9 the cylinder and the hemispherical bulkhead. By
Hooke's law, the tangential strain in the cyl-
inder is

SOLUTION:
(IJ. Poisson's =
ratio, = .3 for
alum1num)
The meridian cur-ve or the cone has a
radius of curvature Rm » cc ano , at any point a
=~(l-~)=
P R,
distance y down from the top at the cone, the .S5
radius Rt is Et 2 Et

(n - y) tan a By integration the radial displacement of any

R=
t cos a point on the cylinder is seen to be (ref. FIg.
A1B-10) •
Then from eq. 1, at any level y

N - ~ - w (H + y) (h " y) tan a
t - p - cos a.

To tind the meridional stresses Nm in the ,

cone, the equilibrium of a segment of height P R,
Fig. Al6.10 .85 - -
h - Y is considered (Fig. Al6-9b). Summing Et
torces vertically,
AdjaCent to tne seam, the tangential strain
Nm ~ 2 n (h - y) tan a cos a = in the hemt spner tca), bulkhead is

n (h _ y)& tan a a • w (H + y) + P R,
EtSLKllD = 2Et (1 - ,.) = .35
+ w ~ (h - y)' tan' a • {h - y)
Hence, by integration
Solving P R,
= .35 "'"'Et""
tan a
""m ="wG - (h - y) [ H + y + -1 (h - y) ]
cos-a. 3 Thus the cylinder tends to expand more than
the bulkhead - a situation prevented by the seam
bevHeen these elements. It follows then trzt
Al6.4 Displacements, Boundary Conditions and Local the seam experiences a transverse shearing action
Bending in Thin Walled Shells.
It is appraoriate at this point to examine as indicated in Fig. A16-ll. These shear forces
some of the raresaing illustrative cases to de- in turn produce bending moments in the sr.ell
termine Whether or not the membrane stresses wall as shown on the figure.
computed gave ss.t i srac tcr i tv accurate measures
at the shell stresses. Anticipating the answer,
we state trat, while the membrane analysis will 6'olkhd. •
~oTl
give the primary streSS system in a shell-like
pressure vessel, a careful (and often lengthy) T
analYSis of induced bendi~ caused by boundary
effects will reveal localized ,secondary stress
peakS. In static strength analyses of properly
designed- vessels it is the practice to neglect
~ 1'ill-ijjj-,=.....=..;:...::.1
- various codes and standards give proportions ol common Fig. Al6.l!
vessels which will correctly limit secondary stresses.
See for example reference (1).

1("
1_, L
A16.6 MEMBRANE STRESSES ill PRESSURE VESSELS
While it Is not our purpose here to take up shell bending
in detail, some indication of the character and magnitude of
these bending stresses should be available to place them in
proper perspective. The most striking thing about t.hese wall
moments Is that they are quickly damped out, becoming neg-
ligibly sma.!!Jdown to l';b of their maximum value) at a distance
of about 4~ from the seam. Thus. for an instance. In a
circular cylindrical shell of 40" radius and .065" will thick-
ness, these moments are so damped at 6.5" from the seam.
The next important consideration is an appreciation of
the magnitude of these secondary bending stresses. For the
case of the pressure vessel of Fig. A16-5, the meridional
stresses are increased about 3~ at the point of maximum
moment, while the tangential stresses are increased only
about 3t. Fo rtunateiy, in this class of vessel, the tangential
stresses are the ones designed by (they are twice as great as
the meridional stresses) and hence the secondary stresses
have little importance for this case (see Chap. 11, pp. 389~~22
of reference 2). In other configurations one 15 not aipys so
tortunate, and detailed analysis may be required. (see refer-
ences 3, 4, 5 and 6).
The situation at the seam of the above
vessel is typical at many seams or boundaries
where elements are jOined which would experience
ditterent expansions it loaded separately.
.~ong such seams and boundaries are those:
i-where the meridional cur\~ture changes
abruptly. It changes from Rm = Ro
to Rm = 00 at the seam in Fig. A16-5.
i1 - Where a sudden change in d1rection of
the meridian curve occurs. In Example
Problem 5, above, considerable shell
wall bending would be induced near the
seam. In fact, a reinforcing ring
would probably have to be added at the
seam as was done in Example Problem 2,
above.*
l1i - at which structural members of differ-
ent stitfnesses and different loadings
join. In Example Problem 2, the cyl-
inder tends to ex?and the most, the
bulkhead quite a bit less and the re-
inforcing ring, being loaded in com-
pression, tends to contract. Other
seams and/or boundaries ot this type
are those where an abrupt change in
shell wall thickness occurs (addition
at a doubler) or where a shell is fast-
ened into a foundation.
Good deSign tends to minimize the magnitude
of the secondary bending stresses by avoiding
combinations at elements which would have highly
incompatible distortions. ThUS, the analysiS ot
~ple Problem 3 shews that if one closes a
circular pressure cylinder w1th an elliptical
bulkhead in which a = 2 b, compressive tangen-
tial stresses would develop in the buliL~ead. In
such a case the bulkhead would tend to contract
radially While the ~in cylinder would tend to
expand as always. Thus, the shear and induced
moment at the seam would be aggravated, produc-
ing (as it happens) a tangential max~um stress
.. certain details of the design of such reinforcing rings are
given in the codes and standards.
13% abav o t~e membrane stress (as against only
3Sb above for the nemtspber tcz,; bu.lkhead ) (ref-
erence 2, p , 410). For this type of bulkhead,
boiler codes sometimes pemit a ratio of alb as
high as 2.6, however.
A16.4 Special Problems in Pressurized Cabin Stress
Analysis.
Because at functional req~irements over
and above those of a sL~ple pressure vessel,
the pressurized cabin shell of an airplane ~~s
a number of stress ar~lysis problems peculiar
to its confi~uration. Several of the more
general of these will be considered ~ere.
DISTRIBUTION OF STRESSES BEr"EEN Sr:ELL
AND STRINGERS.
To stabilize the shell wall in transmitt-
ing hea'ry tail loads tr~ough the fuselage,
longitudinal stringers ars added. T~ese
same stri~gers will also help to carry the
meridional pressure loads. The skin and
stringers must, of course, have equal st~ains
in the longituditudinal. directions but, beca~se
the skin 1s in a ~No-dimensional state of stress,
they cannot rave equal longitudinal stresses:
hence the follOWing analysis.
Let the ~eridional (longitudinal) stresses
in the skin and stringers be ~ and sL, respect-
ively. St will be the tangential (hoop) stress
in the skin. From eq. (1) we again have
p R,
St = - t -
If N is the total number at stringers, each of
cross sectional area AL, then eqUilibrium longi-
tudinally requires
The condition of equal longitudinal strain in
the skin and stringers yields
where ~ is Poisson's ratio (= .3 ~or al~1num).
Solving these three equations one !inds
p R,
-t-
pR. (1 + 2 ~ a) p s, (1 + .6 a)
":n = 2t (I + al = 2t (1 + al
p Ro (1-2u)_ p R, .~
sL = 2t (1 + a 1 - 2t (l + a)
Where a = NAL/2n Rot is the ratio of total
stringer area to skin area. A little s'tudy will
show that t(l + a) is a sort or ft et f ect i v6 shell
wall thickness": i t is the result of taking all
the cross sectional area (skin plus st~ingers)
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
and it uniformly around the peri-
distrib~ting
~eter. On this basis, the results are a little
disappointing: the stringers are carrying only
40% of the stress one ~ight expect if the net
longitudinal load (p n R~) were distributed
evenly over the entire cross sectional area
(2 n Rot (1 + al).
The meridional skin stresses are reduced by
the factor (1 + .6 a)/(l + a) from what they
would be Without the stringers. For structures
of usual proportions this decrease may amount to
20 to 30% but clearly can never exceed 40%. In-
asmuch as the bending stresses due to tail loads
will be superposed on these pressure membrane
stresses, the reduction is certainly beneficial·.
INTERACT I ON BET.'EEN RINGS AND SHELL.
Because of the neceSSity for transmitting
various concentrated loads fram within the cabin
and from the wings and tail to the main she.ll
a.~d because it is also necessary to provide some
lateral restraint which will stabilize the
stringers and skin against an overall instabilit
failure, the pressurized fuselage ot an airplane
contains a considerable number of rings and
frames distributed along the length of the shell.
These rings are seldom, if ever, spaced closely
enough such that they can be considered effect-
ive in carrying a part of the hoop stresses (in
the way the stringers were effective in carrying
part of the meridional stress). Rather, they
act more like widely spaced restraining bands
having the effect shown exaggerated in Fig.
A16-12.
1 tion if an excessive weight penalty is to be
Fig. A16.12 Restraining rings along a press-
urized tank. The action is representative of a
fuselage with widely spaced rings inside.
It is obvious trat the rings in this case
will produce secondary bending stresses in the
skin and hence may have a detrimental effect on
the Simple membrane stress system. Equally
~ul are the tensile loadings developed in
the rivets joining the skin and rings. Detailed
analyses which will permit quantitative evalu-
• If one looks at the problem from the point of view of a
stiffened shell, loaded priIDarily by bending and shear loads
from the tail, an which the pressure membrane loadings are
to be superposed, an interesting effect appears. Because
the internal pressure tends to stabilize the curved skin
panels on the compression side, the effective width of skin
acting with the stringers is increased. The section proper-
ties of the cross section may then change in such a way as
to produce little or no variation in the maximum tiber
stresses. Indeed, the maximum tensile stresses may actu-
ally be reduced by the addition of the internal pressure load-
ingl (see reference 7).
A16.7
ations of these effects in a speCific case are
to be found in references 2 (pp. 395-406), 6
and 8.
One proposed solution to the ring-shell
interaction problem is the floating skin.
BaSically, the idea is to reduce the radial
stiffness of the connection be~Neen the shell
and the rings 50 as to allow the shell to ex-
pand freely under the pressure loading. The
connection between ring and skin must still re-
tain its shear stiffness so that ring loads may
be transferred to the shell wall by tangential
shear flows. Fig. A16-13 shows the basic idea
of the radially flexible connection. Many
Fig. A16.13 Ring skin cross
section showing the action of
a radially flexible connection.
variations of this type or "mount~ suggests
themselves, some of which may nave mer:!.t for
other reasons. For instance, the transmittal
of wind and other Vibration noise into the
cabin of a high speed transport is a problem
which might be treated simultaneously by the
proper choice of connection be~Neen the ring
and the shell.
DOORS AND I,JINDOWS.
The various cutouts in the shell of a
pressurized cabin require special ccnsidera-
avoided.
Consider the panel removed from the pres-
surized cylinder of Fig. A16-14a. Following a
-------
_____ -~,"" pA
1d__-.,::-_-_-_-:.-_-_1r
. _ ~ _ _
11
1
Ii \
(a)
~
I
Fig. A16.14 (b) t
/
common practice in dealing with c~tou~s, we
determine what forces the panel-to-be-removect
applies to the main structure around the border
of the cutout, and then superpose a set of
equal but opposite, self-equilibrating stresses
to cancel these. The c~tout border is then un-
stressed and the panel may be removed without
disturbing the new stress s}~tem in the ~in
structure.
I <:"l .I
j .' j
 Al6.8 MEMBRANE STRESSES IN PRESSURE VESSELS

In examining the figure to dete~ine wr~t Fig. A15-16 shews the typical cutout
sort of canceling stress system must be supplied, structural arrangement. ~hile analytical
we see that the tangential hoop stresses border- approaches have been tried, it is probably
ing the cutout cannot be canceled by a self- safe to say trAt the ~rue elastic stress dis-
eqUilibrating set since they have a radial com- tribution in such a configuration cannot be
ponent. However, the radial component of these computed. The necessity for avoiding high
stresses will actually be supplied by the door intensity stress concentrations (With their
or window pressing oU~Nard against its fra~e. attencant fatigue likelihood) makes empirical
Hence, it is only the component of the hoop inforTation mast useful in such cases. On
stresses along a chord which need to be canceled the other hand, a simple rational analYSiS,
(Fig. A16-14b)·. based on principles outlined above, will very
The llmmediate problem bec~mes one of de- likely suffice tor a static strength check and
Signing a structure to effectively support a set tar most design purposes. (,:"150 see reference
of uniformly distributed self-equilibrating 8, pp. 16-23).
stresses acting in the plane of the chord con- The above discussion has concentrated
necting the upper and lower edges of the opening attention on the problems of carrying the hoop
(Fig. A16-15a). stresses around a cutout. The longitUdinal
pressure stresses, while being s~Aller them-
selves, are intensified by bending stresses
fram the tail loads. Hence, the longitUdinal
stresses across the cutout ~y make this con-
dition (or the combination) most severe.

Fig. Al6.15

All that appears necessary to support the

stress system is to provide horizontal headers
at the top and bottam of the cutout, Which, as
beams, will carry the loads across to the Sides
of the frames where the loads cancel (Fig. Al6-
l5b). For cutouts of usual sizes in pressurized
fuselages, the stress system to be supported in
this manner is quite large and it proves un- Fig. Al6.16 Structural arrangement
economical to design a single horizontal frame around a cutout. Most or all of the
member of sufficient bending stiffness to resist shaded skin area would probably be
them. Instead, the shell wall itself is em- doubled.
ployed to help carry these loads across. The
skin is used to form a beam of considerable
depth, the skin being the web ot this beam, with
the horizontal frame member and one or more LARGE DEFLECTIONS OF ?I...AJ."JE PoA...r.~LS; "Q,UILTING".
longitudinals forming the beam flanges (Fig. Al6- The use ot flat skin panels in a pressur-
15c) • ized fuselage cannot always be avoided. Since
Because of the heavy shear flows and direct the thin skin has little bending stiffness, it
stresses developed, the skin is usually doubled cannot support the lateral pressure as a be~~
in this region. Additional stringers may also ("plate", more correctly) end hence must deflect
be added to relieve the stresses. The rings to develop some tensile membrane stresses which
bordering the cutout (and forming part of the will then carry the loading. The resultant
frame) are extended same distance above and be- bulges of the rectangular skin panels between
low the cutout proper (unless they coincide with their bordering stiffeners give a "qUilted" ap-
a reguXar ring location, in which case they pearance to the surface.
carryall the way around). Even in the case of curved skin panels
quilting will occur: if the internal stiffening
.. Clearly one of the design requirements will be to make the framework (transverse rings and frames and
frame sufficiently stiff in bending against radia.l forces so longitudinal stringers) is relatively rigid and
that the door or window can bear up evenly against the is everywhere tightly fastened to the skin,
frame. then each skin panel is restrained along its

,. ~ '. ~ .: - ~ '"0 .. ; ... ; .-:" "-_·' ... ~.F , ••' • , • , .,'~. ~ .. ' , "' • • -•• -,_ ~~, " ~, • -~., 0 _. \ ••
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A16.9

four sides (borders) against the radial expan- REFERENCES

sion normally associated with the shell membrane
stresses. The result is a sort of three-dimen- (1 ) API-ASME* Unified Pressure Vessel Code
sional-case of the behavior depicted in Fig. 1951 Edition, e t , seq. - -
Al6-1Z* •
From a structural viewpoint, the unfortun- (2 ) Timoshenko, S. "Tl1eory of Plates and
ate aspects at qUilting lie in the high concen- Shells"
tration of stresses occurring near the panel McGraw-Hill, N.Y., 1940
edges and in the tensile loadings on the rivets Watts, G. and Lang, H., Stresses in a
which join shell to stiffeners. The aerodynam- Pressure Vessel With a Conical Head,-
ic characteristics of a qUilted surface are Trans. ASME, vaL 74,-1952, pp. 315-324.
highly undesirable in a high performance air-
plane; "hence again quilting is to be avoided. (4) - - - - - - - - - - , Stresses in a
Computations of stresses in quilted panels, Pressure Vessel With a Flat Head Closure,
inasmuch as they involve large, nonlinear de- Trans. ASME, vol:-74,-l952, pp. 1083-1090.
flections, are difficult. An additional (and (5 ) - - - - - - - - - - - , Stresses in a
quite necessary) complication is that of having Pressure Vessel With a Hemispherical-Head,
to introduce the stiffness properties of the Trans. AS'ME, voL 75,-1953, pp. 83-89-.-
bordering members. The reader is referred to
Chapter A.l7 for a further discussion ot the (6 ) Roark, R. J. wFormulas tar Stress and
problem. A Simplified approach, indicative ot Strain lt , McGraw-Hill, N. Y. &:d Edition,
trends, is given there along with further ref- 1954.
erences to the literature. (7) Howland W. and Beed, C. Tests Of Pres-
surized Cabin Structures, Journ. Aero.
Al6.5 Shells of RevolutiOn Under Unsymmetrical Loadings. Sci. vo1:-a:-NOV. 1940.
Problems in which the shell af revolution
experiences unsymmetrical loadings are not un-
ccrmcn in a ircraft structural analys is • The
nose of a fuselage, the external fuel tank and
the protruding radome are shells of revolution Designs by Krafft
which may be loaded unsymmetrically by external Ehricke of Convair
aer-ocynentc pressures. Again, the same external for Space Travel
fuel tank shell receives an unsymmetric internal
hydrostatic pressure load tram th~ weight at
tuel directed normal to the shell axis.
Because of the unsymmetry at the problem,
membrane shear stresses are now present and so
the analystm"ust solve not two, but three equa-
tions in three unknowns (Nm, Nt and Ns). More-
over, these become differential rather than
algebraic equations.
Because the derivation at the differential
equations of equilibrium. is rather Leng'thy , and
because their general solution cannot be written
(rather, only specitic solutions for certain
cases may be be found), no details are repro-
duced here. The reader is referred to pp. 373-
379 of reference 2 tor the derivation of the
equations ~~d for an example problem.

Outer Space Vehicles will Present Many New Problems to

• One deSign which reauces quilting in the curved skin. the Aeronautical Structures Engineer.
fastens rings and frames to the inner surface of "hat" section
stringers only. Thus the ring is not directly fastened to the
skin which is therefore not continuously restrained around
each ring circumference. The result re a modified floating • American Petroleum Institute - American Sodety ol
skin. Mechanical Engineers.
A16.10 MEMBRANE STRESSES IN PRESSURE VESSELS

Douglas OC~8 Under Ccaatructton and Assembly. Fuselage is Pressurized to Permit High Altitude Flight.

~:,
"

Section of DC~8 Fuselage being Lowered into Hydrostatic Test Tank.

 CHAPTER A-17
BENDING OF PLATES
ALFRED F. SCHMITT

At7.1 Introduction.
It was seen in the last. chapter that thin with another SUCh, we write the equations down
curved shells can resist lateral loadings by by a direct appeal to past experience and
neans of tensile-compressive membrane stresses. intuition.
As w1ll ~e seen later, thin flat Sheets, by de- Fig. A17-l shows the differential element
flecting enough to provide both the necessary of a thin, initially flat plate, acted upon by
curvature and stretch, may also develop mem- bending moments (per unit length) Xx and My about
brane stresses to support lateral loads. In the axes parallel to the y and x directions respect-
ar~lysls of these situations no bending strength ively. Sets of tNisting couples Mxy (= - Myx)
1s presumed in the sheet (membrane theory). also act on the element.
In contrast to the membrane, the plate 1s
a bvo-dlmenslonal counter?art of the Jearn, in r
which transverse loads are resisted by flexural
and shear stresses, with no direct stresses in
its middle plane (neutral surface).
The skin may also be classified as either a
plate or a ~e~brane depending upon the magnitude
of transverse deflections under loads. Trans-
verse deflections of plates are small in compar-
ison with the plates' thicknesses - on the order
ot a tenth of the thickness. On the other hand,
the transverse deflections of a membrane will be
. /
yx
r
Fig. A17.l
•

on the order of ten tL~es its thickness.·
Unfortunately for the engineers' attempt at an orderly
cataloging of croclems, most aircraft skins fall between the
above two extremes and hence behave as plates having some
membrane stresses.
Plate bending investigations ~Ave for a
longttme been important in aircraft st~uctural
analyses in their relation to sheet buckling
problems. Recently they have assumed new im-
portance With the introduction of thick skinned
construction and still more recently With the
use at very thin low aspect ratio wings and
control surfaces which behave much like large
plates, or even are plates in same cases.
It is the purpose of this crapter to pre-
sent briefly the classic plate formulas and
same applicattons. Appropriate references are
cited in lieu of an exhaustive treatise, which
could hardly be presented in one chapter (or
even one volume) as witness the voluminous
literature on the SUbject.
A17 • .2 Plate Bending Equations··.
Technical literature in this field abounds
with many excellent and elegant derivations of
the plate bending equations (references 1 and 2,
tor inStance). Rather than labor the subject
As in the case of a beam. the cu~rature in
the x, z plane, a a.N/aX.:l, is proportional to the
moment Mx applied. The constant of proportion-
ality is l/EI, the reCiprocal of the bending
stiffness. For a unit Width of beam I = t 3 / 12 .
In the case Of a plate, due to the Poisson
erfect, the moment My also produces a (negative)
curvature in the x, z plane. ThUS, altogether,
With both moments acting, one has
Where ~ Is poisson's ratio (about .3 for alumin-
um) • Likewise, the curvature in the y, z plane
Is
These two equations are usually rearranged to
give the moments in terms of the curvature.
They are written
aX
i" ay" -
w) (1 )

• As will be seen later, the presence or absence of mem- My =D ( aaXy.:lw + " OX

e- a x.ll
w) __ (2 )
brane stresses is not wholly depenctent upon the magnitude
of deflections, but is also determined by the form of ce-
flection surface assumed by the sheet (in turn dependent upon Where D • ,t'/12 (1 - ~X)
the shape of boundary and loading).
The twist of the element, aaw/axay
.. the assumptions implicit in the following analysis are (= a:aw/ayax) is the change in x-directlon-
spelled out in detail in Art. AI7.5, below. Slope per unit distance in the y-dlrectlon (and
A17.1
 A17.2 BENDING OF PLATES

(and visa versa)···. It ~s proportional to the In a similar manner, a ~oment summation about
~Nistlng couple Mxv ' A careful analysis (see the x axis yields
references 1 and 2) gives ~he relation as
Q =al1y+~ (6)
l1x y = D (1 - ~) axdy
a' w (3 ) y ay a x

(Equations (4) and (5) correspond to V = dM/dx

Equations (1), (2) and (3) relate the applied in beam theory).
bending and twisting couples to the distortion One t1nal equation is obtal~ed by summing
of the plate in ~uch the same way as does forces in the z direction on the element:
M = EI d~Y/dx~ :or a beam.
While a few highly instructive problems may
be solved with these equations (see reference 1,
q = aa ~x + aa ;y _ _ _ _ _ _ _ _ _ _ _ _ _ (6)
pp. 45-49 and reference 2, pp. 111-113), they
are of little technical importance. Hence we Equations (4), (5) and (6) prOVide tr~ee
move on to consider bending due to lateral loads. additional equations in the three additional
Fig. A17-2 shows the same plate element as quantities Qx, Qy and q. The plate problem is
in Fig. A17-l, but with the addition of internal thus completely defined.
shear forces Qx and Qy (corresponding to the ~V" To summarize, we tabulate below the quan-
of beam theory) and a distributed transverse tities and equations obtained above. For com-
pressure load q (pSi). With the presence of parison, the corresponding items from the
these shears, the bending and twisting moments engineering theory at beams are also listed.
now vary along the plate as indicated in Fig.
A17-2a. (For clarity, the several systems of
forces on the plate element were separated into CLABS IT'" PLATS T!:Eoo.Y
BEAl1
the two figures of Fig. Al7-2. They do, of THEORY
course, all act stmultaneously on the single Coordinates x r x
element) •

1i /,'..
~ ~/
~4Z
:?-"'" '~~41C~.,
K~.4IC~
»'
..-&P::'I:I":;!';~'
V---..#
~L
::;w
Geomstry

Structural Bending
Characteristic Stiffness
De!lections
a'"'w
Distortions axa ,

0- 12
o-W
~

E:t~
•
aa.,Of
• dxdy

11 - \-'0 ")
I ~
y

ax
EI

,,(
"p
tIC.,
:2
(a)
~ ,)
(b)
' ..... Loa<Ungo
ccceree
oar,
t Latera q
x' ,
M

Fig. An. 2. The differenti3ls are increments which "Hooke's l1oment- !1x = D ('"
ax-").l. ,·W)
ay:r
should be written more precisely as, for instance, La.' Distortion
4Q r • (aQr/eY)4r. Relation a aw oa.,Of) ~ day
My = 0 ( ii? ..... ax" 11"::.I dX'"
The next relations are obtained by summing Mxy =0 (1 - ;..) axdy
,,,
moments in turn about the x and y axes. For ex-
ample, we visualize the two loading sets of Fig.
A17-2 acting s11nUltaneously on the single ele- Equilibrium Moments v = d i'!
ment, and sum moments about the y axis. cx

IIx ely + (llyx + d lIyx) dx + (Q.x + d Q.x) dx dy =

Forces
(11x + d l1x) ely + l1yx dx

DiViding by dX dy and discarding the term at Finally, one verJ l.m.portant equation is
higher order gives obtained by eliminating all internal forces
(Xx, My, Mxy, Qx, Qy) between the above six
equations. The result (whfcn the student
shoUld obtain by himself as an exerCise) is a
or, relation between the lateral loading q and the
deflections w·:
rv.: = a l'lx + a l1x y _ (4)
-sx ax a y (7

••• If W', the deflection function, is a continuous function of

x and y (as it must be, at course, I.n any technically im-
portant plate problem) then at each point aaw!axdy : • the corresponding equation for a simple beam is
a"vlaydx, as is proven in the calculus. q/EI ; d"Y!dx".
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AI7.3

The plate bending problem is thus reduced 1

to an integration of eq , (7). For a gfven
lateral loading q (x, y), a deflection function
A.,lID = D'lmn
nt. (:n'a n')'
~ -+- ~
b
w (x , y) 15 sought which satisfies both eq, (7)
and the soecifled boundary conditions. Once Hence the reqUired deflection surface (~~d the
found, w ·(x, y) can be entered into eqs , (1) to solution to the prob1em-j has the equation
(5) to determine the interr~l forces and stress-
es.
w = ~(m,l n')' sin ~ sin -
nbn -y - (10)
Al7.3 An illustrative Plate Bending Analysis. n n ~-+-~ a
Assume a lateral loading applied to a rec-
tangular plate having all edges Simply supported The maximum deflection is seen to occur
(hinged). The coordinates are chosen as in Fig. where the trigonometric functions have values
A17-3. With foreknowledge of the general use- at unity and q is also a maximum.
fulness of the result, we assume a sinusoidal If eq, (iO) is substituted into eqs . (1),
loading of the form (2) and (3) one obtains

q _ Qmn Sin m ~ x sin n ~ y

m, n = 1, 2, 3,
(8)
Mx =- n
• .
(a'""" b'~)' (~'a'-+ "b'r.') s i nm-an-x sin -o-
;lL +
r: ;,: y

My ;:
cmn
- n' (",? +~)' o .n")
m'
7'"" b' Sin~ sin n a y
a --,---
em (1 - :..L) mn mn x n
it y
l1XY =n' ("'+n')'
~ ~
ss cos a cos - b -

Fig. Ai7. 3 Sinusoidal loading on a In a Similar manner the transverse Shears

rectangular plate. Sections through
the loading shown for m=3, n=2.
To find the resulting deflected shape of
the plate we try a solution of the form
may be found from eqs , (4) and (5).
With such results as these the plates'
stresses may be determined as desired. For
example, the maximum direct bending stresses
are seen to occur where the shear stresses (due
to Mxy) are zero. Thus

w = .!._____ 5 in In n: x sin n n: y (9 ) sx (_ MC) = ::1xt 6 i'!x

-tmn a b 1 21 t'
and hence
where ~ is the unknown deflection amplitude.
This trial deflection function is known to sat-
isfy the boundary conditions on the plate since
at x = a J a and at y.:: 0, b we have - - - - - - - - (11)

w= 0 (zero deflection at the supported

edges) The reader having a familiarity with Fourier ~
methods will recognize immediately that the above analysis
a' w (zero moment at the hinged provides the key to the solution of the problem of any general
ax" edges; see eqs. land 2) loading q (x, y) on the same plate. SUch an applicatiOn is
made by determining the proper combination of sinUSOidal
It remains only to find the value of Amn which pressure terms (each of the form of eq. oj such that their sum
will closely represent the desired loading. The sum of the
will satisfy eq, (7). Substituting (8) and (9) corresponding deflection functions (each of the form of eq, 10)
into (7) one obtains gives the desired solution. Details of this type of analysts are
to be found in reference 1 On PI'. 113-176 and 199-256.
In common with all problems which are
rorennet ec in terms of a partial differential

= 'lmn/D ~- the uniqueness of solutions to the differential equation of

or the form of eq, (7) is a classical proof appearing in num-
erous advanced texts on mathematics and matbemaucal
pbystca. Since the equation is known to have a unique
solution, then any solution fOWldlor it is the one ana
• the common factor sin m ~ x sin ~ has been dlvided out. I only correct solution.
Al7.4 HENDING OF STRESSES

equation. the solution of t~e plate Je~jin~ prob- TABLE A17.1

lem depends strengly up0n the Joundary conditions
(both the shape of the boundary and :~e :J?es of Stress and Deflection Coefficients for a. UniIormly
support provided t~ere). ~~e above exanple ~y Loaded Rectangular Plate Having Various Edge
Conditions
be said to rave been deceptively easy jecause of
both the simple shape of the bCw4dary and th~ I
I Long Sides ! Short Sides I,
type of support. Plate ,racIsms whersin the All Sides Pinned, Pinned, All Sides
plate planfor.n is not a si~ple geometric f~gure I Pinned Short Sides Long Sldes I C'lamp ed

must be solved by numerical ~9ans. 's to tte i Clampea

I Clamped.
I

type of support, a f~ll discussion of boundary b/. c ~ a ~ ! a I ~ , J ! ~

conditions fer ~lates is to be found in refer- .420 II .0209 ! .420;.0138
• I
.3078
1.0 1.0443 . 2874 ! . 0209
ence 1. ;p. 59-95. 1.2 .0616 I .3756 I .0340 i • 522! .0243 .462 .0188 .3834
A17.4 Compilations of Results for Plate Bending Problems. 1.4 .0770 .-1518, .0502 I .600 i .0262 . .;86 .0226 I .4356
Fortunately for the practicing engineer, it 1.6 .09061.5172 ! .06581 .654 ! .0273 I .500 .0251 I .4680
1s not necessary to perform analytiC computations 1.' .1017 r • 5688 ! . 0799T. 690 .0279 .502 i .0267 .4872
as discussed above for the great Tzjority of 2.0 .11061.6102 I .0987 i .714 1.0284 1.504! .0277 I .4974 I
practical plate problems. Problems of the type 3.01.13361.7134: .128!.750 I - - I - -
illustrated above, plus the ~iad variations 4.0 I .1400 I . '7-110 ! - , - - - I - -
pOSSible, became very fashionable exercises 5.01.1416 .7476 I - - I - - - -
<XI l . 1422 ! . 7500 i . 1422 i . 750 .0284 .498 .0284 .498
amongst mathematicians following the discovery
by LaGrange of eq, (7) in the year 1811. The Rectangular Plates Under Various Loadings
results of mBJlY researchers' labors have Jesn
compiled in various torms for handy reference. • S. r ixcenenko , IITheory of Plates and Shells II ,
A common and Impor-tant case is that of a pp. 113-176, 199-256.
uniformly loaded rectangular ~late (Fig. A17-4). • J. P. Den Hartog, IIAdvanced Strength of
The major engineering results are the values of Materi2ls~, pp. 132-134.
the maximum deflections ar.d the max~um stresses
developed. These may be put in the form (a is • R. J. Roark, ~Formulas for Stress and
the length of the short side); Strain", pp. 202-207.
Circular Plates Under Various Loadin~s
~ = a ~
E t ' - - - - - - - - - - - - - - - (12)
(same t.ta-ee references, in crd~r)

q a'
"!lAX = jl to - - - - - - - - - - - - - - - (13) • pp. 55-84, 257-287.

• -p. 128-132.

where the coefficients a and ~ are given in

Table Al7.l tor the tour most cammon edge
• pp. 194-201. 209-211.

conditions. A17.5 Deflection Limitations in Plate Analyses.

Fig. A17.4
Similar presentations may be made for many
dozens of other cases. With the ready availa-
bility of comprehenSive catalaglngs of these
problems in references devoted to the purpose,
there appears to be little virtue in duplicatlon
here. Hence the follOWing list of selected
references is presented. Additional references
are to be r ounc in turn within these works. 'Ae
note that, because of the linearity of the plate
bending problem, superpositton of solutions is
POSSible to extend even further the usefuL"1ess
of these extensive listings.
In the intrOductory remarks of this chap-
ter it was stated that a plate may be distin-
guished from a membrane by the sma.Ll, order of'
its deflections (on the order of a few tenths
of its thickness). We will re-examine this
statement here to show that this is ~ot so
much a definition as it is a~ acc~acy limita-
tion imposed by one of the assumptions ~de in
the plate analYSiS.
There are several fa~iliar assuaptlons
from beam theory which, of course. carryover
here, inasmuch as the plate analysis resembles
the beam analysis rather closely. These "beam
theory assumptions ll are:
- elastic stresses only are presumed,
ii - small slopes (so that a~w/ox~ and
a~/ay~ are·good approximations to
the curvatures),
iil - at least one transverSe dimension
(length or width) be large compared to
the thickness so trAt shear deflections
may be neglected.

&
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A17.5

However, the beam theory assumptions do not

place a very severe restriction on the ma-gTIitude
at detleclfrons permitted. Deflections of sev-
eral times the plate thickness would be per-
missible i: these were the only restricting
assumptions.
In deriving the plate bending equ~tions it
<-----t-
•
was assumed t~~t GO stresses acted in the middle (0)
(neutral) plane of the plate (no membrane
stresses). ThUS, in s~ing forceS to derive
sq. (6), no membrane stresses were present to
help support the lateral load. NoW in the sol- Fig. Al7-5b shows the desired one-dimen-
utions to the great ~jority of all plate bend- sional problem which now resembles a loaded
ing problems (solved as in Art. AI7.3), the de- cable. The differential equation of equilib-
flection surface solution found is a non-de- rium is obtained by summing vertical forces on
velopable surface, i.e., a surface which cannot the element of Fig. A17-5c (draw with all
be formed fram a flat sheet without same strech- quantities; loads, deflections, slopes and
ing of the sheets' middle surface-. But, if curvatures shown positive). One obtains
appreciable middle surface strains must occur,
then Large middle surface stresses will reSUlt,
invalidating the assumption upon which eq. (6)
was derived.
Thus, practically all loaded plates deform
into surfaces which induce some middie surface or
stresses. It is the necessity for holdins down
the magnitUde of these very powerful ~iddle
surface stretching forces that results in the - - - - - - - - - - - - - - - (14)
more severe rule-or-thumb restriction that elate
bending formulae apply accurately only to prob- where s is the membrane stress in psi.
lems in which deflections are a few tenths of Eq. (14) is the differential equation of
the plates' thickness. a parabola. Its solution is
A17.6 Membrane Action in Very Thin Plates.
There is still another source of middle W=2~Xt (a-x) - - - - - - - - - - - - - (15)
surface strains in plates: this is the re-
straint against in-plane movements offered by The (as yet) unknown stress in eq. (15)
the edge supports. ~bile not important in prob- can be found by computing the change in length
lems wherein deflections are limited in accord- or the strip as it deflects. This ~stretchn 1s
ance with the restriction of the last article, given by the difference be~Neen the curled arc
such rest~aint does assume great linportance in length and the original strai&~t length (a).
the case of large deflections of very thin Thus:
plates which support a major share of the load
by membrane action. It is, in fact, useful to
consider the lliniting case of the flat membrane
which cannot support any of the lateral load by
bending stresses and hence has to deflect and
stretch to develop both the necessary curvatures
and membrane stresses.
The two-dDnensional membrane problem is a
= J:JdW' • dx' - a

nonlinear one whose solution has proven to be

very difficult. Rather than attempt to treat
the complete problem, we can study a simplified
version whose solution retains the desired
general features. The one-dimensional analYSiS
= [/1. (:/ dx - a

of a narrow (unit ',vidth) strip will be treated.

This strip is cut tram an originally flat mem-
brane whose extent in the y-direction 1s very
great (Fig. A17-5a).
- The cone and cylinder are examples of developable sur-
faces, the sphere is a nondevelopable one. It is a familiar
experience that the skin of an orange cannot be developed in-
to a flat sheet without tearing.
Since the slope ctw/dX is small compared with
unity, we use the binomial theorem to 'Nrite
dw) ' l' /' -_1 +
( 1+ ( dX) 1(dW)'
2 dx
.. here "ds" is the differentia! arc length of the calculus
and has no kinship with the s which denotes the membrane
stress throughout the remainder of the analysis.
A17 6 BENDING OF PLATES

Hence at the middle of the long side of ~he panel.

6 = J} ~ (:YJ + dx - a
We note that the limiting case, alb = 0, cor-
responds to the one-dimensional case analyzed
earlier .. Unfor'tunaue Ly, an extraporat i cn of
these ~No-dimensional results to that limit
does not show agreement with the one-1imen-

= ~ n:r dx
sianal result. Presumably the jiScrepancy way
be traced to the excessive influence of inac-
curacies in the assumed deflection Shape Jf the
~embrane as used in the approxDnate two-dL~en­

Substituting through the use of eq. (15) and in- sional solutions.
tegrating we find Experimental results reported in reference
4 show good agreement with the theory ~or
gJi a 3 square panels in the elastic range.
O=Z4s't'
TABLE A17.2
Now by elementary considerations I
Membrane Stress and Deflection Coefficients i
o ; s a
E alb 1.0 1.5 2.0 2.5 3.0 I 4.0 I 5.0 i

n, .318 .228 .1. .125 .10 .068! .052[

Equating these last and solVing we find I
n2 .356 .37 .336 .304 .272 I .23 ! .205!

s = .347 [ E ( qT a) 'J'/'- - - - - - - - - - - - (16)

It eq. (16) is sucet trut ec into eq, (15)
one gets for the maximum detlection (x : ~ )
q a)'/'
"'MAx • .360 a ( Et - - - - - - - - - - (17)
Equations (16) and (17) display the essen-
tial nonlinearity of the problem, the stress and
the detlection both varying as fractional ex-
ponents ot the lateral pressure q,
Solutions of the complete two-dimensional
nonlinear membrane problem have been carried
out·, the results being expressed in forms iden-
tical with those obtained above for the one-
dimensional problem, ViZ.,
- - - - - - - - - (18)
'I' - - - - - - - - - - (19)
Here ftaft is the length at the long side of the
rectangular membrane and n~ and n e are given in
Table A17.2 as functions of the panel aspect
ratio alb.
The maximum ~embrane stress (sMAX) occurs
Al7.7 Large Deflections in Plates·-.
In the preVious articles of this chapter
the results of analyses were outlined for the
~NO extreme cases of sheet panels under lateral
loads. At one extreme, sheets Whose bending
stiffness is great relative to the loads applied
(and which therefore deflect only slightly) may
be analyzed satisfactorily by the plate bending
solutions. At the other extreme, very thin
Sheets, under lateral loads great enough to
cause large deflections, ~y be treated as ~em­
branes whose bending stiffness is ignored.
As it happens, the most efficient plating
designs generally fall between these two ex-
tremes. on the one hand, if the designer is
to take advantage of the presence of the in-
terior stiffening structure (rings, bulkheads,
stringers, etc .}, which is' usually present tor
other reasons anyway, then it is not necessary
to make the skin so heavy as to behave like a
~pure~ plate. On the other ~~d, if the skin
is made so thin as to necessitate supporting
all pressure loads by stretching and developing
membrane stresses, then permanent deformation
reSUlts, producing ~quilting~ or ~'Nashboardingn.
The exact analysis of the two-dimensional
plate which undergoes large detlections and
thereby supports the lateral loading partly by
its bending resistance and partly by membrane
action is very involved. A one-dimensional

.. The discussion to follow will be concerned primarily with

• The work of Henky and F01'pl is summarized 10 reference 3,
PI'. 258~290 and in reference 4. The partial differential
equation solved is given in reference 1 on p. 344 (eq. 202)
and. the approximate method 01 solution usually employed is
sketched out on pp. 345, 346 of this same reference. The
reader who would compare presentations amongst these ref-
erences should note the differences in the definitions 01 the
plate dimensioning symbols "a" and ''b''.
problems dealing with the support 01 a uniform pressure
load on a fiat skin panel. It may, therefore, help the
reader to fix his ideas if he visualizes the discussion as
applied to the problems of analysis of a single rectangular
sKin panel taken between the stringers and bulkheads 01 a
seaplane hull bottom. EqUally useful is the picture of the
very nearly fiat panel between rings and stringers in the
slightly curved side of a large pressurized fuselage.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A17.7

analysis, parallel to that at Art. Al7.6, is to The approximate large-detlection method

be found in reference 1, pp. 4-10. A more
elaborate two-dimenSional analysis is shown on
pp. 347-350 of this same reference.
An approximate solution of the large de-
flection plate problem can be obtained by adding
together the flat plate and membrane solutions
in the following way:
Solve sq. (12), the plate bending relation,
for q; call it q' ,
q' _w.•• _Et'
··niv.
- a a6
outlined above ~~s serious Shortcomings insofar
as the prediction of stresses is concerned.
For simply supported edges the maximum combined
stresses are known to occur at the panel mid-
point. Fig. A17-7 shows plots ot these stresses
for a square panel as predicted by the approxi-
mate method (substituting q' and q" into eqs.
(13) and (14) respectively and cross plotting
With the aid at Fig. A17-6-). Also shown are
the maxunum stresses computed by the exact
large-detlection theory (reterence 5).
l<> .X&o~

approx.
----
- -
Now solve eq. (18), the membrane relation, tor
q; call it q",
'0

..
o•

10
a
' ./'
,,.-'
:.--
e---

The sum of these two pressures gives the total /' V

lateral pressure, called Simply, q.

q = q' + q"
o
o
/
-:.-
100 4 1'0 '00

1 Et 3 1 Et :5
Fig. A17. 7 Large·deflection theories· mid-
q = -a - - b - '''l1AX +,--,-.."..
a40 (_)6 n, a
Ww.x - - - - (20)
panel stresses; simply supported square panel.
a
Because ot the obVious desirability at
Eq. (20), we see,. is based upon summing the in- using the results of the more exact theory, s~e
dividual stlttnesses at t~e two extreme be- or these are presented in Table A17.3. The
havior mechanisms by which a flat sheet can treatment at additional cases (other types ot
support a lateral load. No interaction between edge support) may be found in reference 6, pp.
stress systems is assumed and, since the system 221, 222.
1s nonlinear, the result can be an approXimation
only.
Eq. (20) is best rewritten as TAaLI Al1'.'

Lup D<Ia.-. ~ p_ C-mcl-.

,.
(lIILIIGnD P....... l..oui ('ll ~, """'..... ~I

q "
E t 40
= U"'MAX)(~)'
a \~ b
+ 1
~
(WMAX\'
t 7
_- - - - ( 2 1 ) ...
I
- I
or Slr_
c
_It I 'L' •
,430 1'810
"... .. n
I.U
,.
P-..C.,.rtb:1

, ,.n
.. t/D4
no
,.~
,n
,... - ....-
I.U '.n
L. ..• ..• I ...
Fig. A17-6 shows eq. (21) plotted for a i ,,~/1\~ 3.110 $.10 1.1'0 lD.ao ,u 14.:10 ,1$.10 11.011 II. III
square plata USing values at a. and TIl as taken I ~/nal 0.10
,..
•... ........
I 1.10 1.10
, .ee '-" 10. :10
from Tables A17.1 and Al7.2. Also plotted are
the results of an exact analysis (reterence 5). 11 ta.- _It I .IM ·_11.24 1.44 1.1'11 I.M '.M "
As may be seen, sq. (21) is somewhat conserva- I 1 ,U l'.~
--
'ab / gJ i 4." 1.11 jlO.)O 14• • 1'.4
11•• 0
tive inasmuch as it gives a deflection which is
too large for a given pressure.
II
_.'" • 1-
L
.,
y 11E11 1 1. 2I
' ..
'M' ..""'..II
1.40 4.1' '.11
.n.... •0lIl_01. ..........
_ _ 01.
1.'1 1.10

..
II.U 10.10 10.110 L1.»

"0 J. lOt&l . n l l • .. . 'v.

II
.xac~

&pprox._
---
'00

"0
VI A17.8 Considerations in the Applicationa of Large-De-
neeuce Plate and Membrane Analyses.
Before concluding this chapter it 1s
~ pertinent to note several serious omissions in
the developments outlined above with regard to

»:
100
their application to flat pressure-panel
analyses within a ship hull or fuselage. The

o
o
~-
0.' vlt
-'
1.0
.~~i.~:!..~
l.~
-
2.0 • but 1181ng n2 • . 260 in eq, (19), This value gives the
stresaes at the center of a square panel whereas n2 •
Fig. A17.6 Deflections at the midpotnt 01 a simply .356 in Table A17. 2 is for stresses at the panel edge.
supported square panel by two large.deflectian
theories.

,c<
,O~
A17.8 BENDwa OF PLATES
large-deflecti~n plate and the ~embr~ne analysGs
were dev al cped for applications where the plate
bending analysis appeared inadequate. However,
these analyses the~elves presumed condit10ns
seldom encountered in practice.
FIRST, the analyses assume unyielding sup-
ports on the boundaries of the sheet panel. In
practice, the skin is stretched acrOBS an elas-
tic framework ot stringers and bulkheads. It
follows, therefore, that the heavy membrane
tensile forces developed during large deflec-
tions will cause the supports to deflect towards
each other thereby increasing the plate de-
flection and relieving some of the stresses.
A simple one-dimensional analysis for a
membrane strip having elastic edge supports
(parallel to the analySiS of Art. A17.6), shows
errors an the order of 25 per cent are likely if
the framework elasticity is neglected (referenc~
7). At this writing no two-dimensional treat-
ment of this problem is known to the writer.
SECOND, it is seldam that the analyst has
to cheek a panel for lateral pressure loads
alone. Most Often, the entire ~fleld~ of panels
on the framework of stringers and bulkheads must
Simultaneously transmit in-plane loadings from
the tail load bending stresses and the cabin
pressurization stresses.
Inasmuch as the large-deflection plate and
membrane analyses are nonlinear, it tollows that
correct stresses cannot be found by a straight
superposition. The magnitude of the error in-
troduced by such a procedure is difficult to
estimate in the absence of an exact analysiS. A
one-dlmensional analysiS, parallel to that of
Art. A17.6, but With elastic supports and axial
load, is given in reference 7. These results,
which indicate the effect of the axial load to
be quite important, may be used as a guide in
lieu of
more complete two-dimensional studies.
The interested reader is referred to the orig-
Inal work for details.
RE...J;lERENCES
1. Timoshenko, S. "Theory of Plates and Shells",
McGraw-Hill, N. Y., 1940.
2. Den Hartog, J. P. ~Adv2.nCed Strength of
Materials~, McGraw-Hill, N. Y., 1952.
3. Sechler, E. and Dunn. L. "Airplane Str~c­
tural Analysis and Design", John Wiley.
N. Y., 1942.
4. Heubert, M. and Sommer, A., Rectangular Shell
Plating Under Uniformly Distributed r.yd~Q­
static Pressure. NACA TM 965.
5. (selected large-deflection plate references).
a) Moness, E. Flat Plates Under Pressure,
Journ. Aero ..Sc1., .§, Sept. 1936.
b) Ramberg, W., McPherson, A. and Levy, S.,
Nonnal Pressure Tests of Rectangular
Plates, NACA TR 748, 1942.
c) Levy, S. Square Plate With Clamped Zd~es
Under Normal Pressure ?rOducin~ Lar~
5'e"f'Iec~ NACA TR 740, 1942. --------
d) Levy, S. Bending of Rectangular Plates
With Large Deflections, NACA TR 737, 1942.
e) Chl-Teh Wang, Nonlinear Large Deflectlor.
Boundary-value Problems or Rectangular
Plates,
--- NACA TN 1425.. . 1948.
f) Chi-Teh ~ang. Bending at Rectangul~r
Plates With Large ner'iecurons , NACA "i'N
1462, 1948.
6. Roark, R. J. ~Formulas for Stress and Strain~
McGraw-Hill, N. Y. 3~ Edition, 1954.
7. Flugge, W. Stress Problems in Pressurized
Cab Ina , NACA TN 2612, 1952.-
 CHAPTER AlB
THEORY OF THE INSTABIUTY OF COLUMNS AND THIN SHEETS
(BY DR. GEORGE UANIS)

PART 1
ELASTIC AND INELASTIC INSTABIUTY OF COLUMNS
Ala. 1 Introduction, Thus the total bending moment at section z
will be:-
Part 1 of this chapter will be confined
to the theoretical treatment of the instability M= Pu + Q~Z on upper portion - - - - (d)
of a perfect elastic column and an imperfect
elastic column. The column Is the Simplest of
the various types ot structural elements that 11. Pu + Q(l-a) (l-z) on lower portion (e)
1
are. subject to the phenomenon of lr~tabl11ty.
The theory as developed tor 001umns forms the From mechanfc s of simple bending, we have
basts tor the study of the instability of thin the deflection equation,
plates, which subject Is treated in Part 2.
d"u 11
AlS.2 Combined Bending and compreeeton of Columns. --a = - - - - - - (1)
dz Ely
Consider a column with one end simply Thus the deflection u(Z) ot the column
supported and the other end hinged (Fig. AlS.l) IS,
under the simultaneous action of a compreSSive
load p and a transverse load Q. Without the Qaz
load P the bending moment due to Q would be:- , (0 ~ z ~ 1 - a)
1
p
d"u (1-a) (l-z)
El dZ" = - Pu - Q 1
I
I (1 - a ~ z ~ 1) - - - (2)
a I
If we introduce the notation,
.ll:....-Jf+--
I
Q

I - (3)
I
\ The general solution of eq, (2) is:
\
\ Qaz
U ... C:L cos Kz +. C. sin Kz - PI '
--\u (0 ~ z 1 - a) - - - - - - - - - (4a)
o (I-a) (l-z)
u = C;5 cos Kz + c, sin Kz - 't i '
A
~!¥-
(l-a::z::l) - - - - (4b)
p
Fig. AlB.l Where C~J Ca, C3 J and C40 ~re constants of
integration to be deternined from bo~~dary
On the lower portion or the column conci tions.
_ caz
11, - -1- ----
-- ---- - -- (a)
z
For eqe . (4), s tnce u = 0 tor z = a and
= 1, it follows that:
On the upper portion
C~ = 0 and C3 = - c. tan K1 - - - - - (!)
M~ :: ~ (l-z ) - - (b)
At z = (1 - a) the two portions of the
Due to the deflection u(z), the axial deflection curve given by (4a) and (4b)
load P contributes to the bending moment by respectively must have the s~e deflection and
the amount:- slope. From these two conditions we determine
Cll and C•.
M~ = Pu - - - - - - - - - - - - - - (c)
Al .1
AlS.2 THEORY OF THE INSTABILITY OF COLUMNS AND THIN SHEETS

c,= ~ sin Ka C =_q sin K(I-a) (g) the bOdy would cause only infinitesimal Changes
PK sin Kl' to PK tan K1 in the displacements and the body recovers 1f
the added loads are removed. When the dis-
Substituting (r l and (g) into eq, (2), placements are continuously increased With
we obtain: little or no further increment at loads, the
system is unstable. If the body will remain
Q. sin Ka ca in the displaced position after the removal
u = PK sin K.l sin Kz - PI z ot the disturbance, the body is said to be in
neutral eqUilibrium. Havir~ these definitions,
(05 Z S 1 - a) - - - - - - - - - - -rsa ) we will not investigate the behavior of the
column betore and atter the critical load !s
u =~ sin K(I-a) s1 K(I-z) q(l-a) (l-z)
reached.
PK sin Kl n PI
(l - a S z s 1) - - - - - - -(5b) p p

It load Q is applied at the middle of the

GA(~
s "-...:;;J~__ - -
_/
column the maximum deflection is:-
q tan 19,. Fig. AlS.2
2 -----(6)
2PK Ass~e, as shown in Fig. A18.2, that a
Simply sunported colwmn loaded by an axial load
It is obvious that for ~ = ~ , P is bent by a small disturbance. If the
deflection IS u, the bending noment due to P
is Pli.
tan ~I _ _ . Thus the maximum ceriecta on of From basic mechanics, we know that,
column becomes infinite for K1 ~ n and from £1
R = -M , whence
eq, (3).
EI
n'E1 If = -Pu - - - - - - - - - - - - -(sa)
Per = - - - - - - - - - - - - - - (7)
I'
The exact expreSSion for the curvatuve of
Equation (7) is an Important equation the neutral axis i5:-
derived first by Euler. It gives the critical
compressive load which causes infinite detlect- 1 d9
R = dS ' where s is the arc length at the
ion in a column and it specifies the ultimate
strength or a column in compresSion. deformed axiS, and Q the angle between the
tangent to the curve and the z axis. ThUS,
It is obvious from eq, (7) that Euler's
critical load is independent of the magnitude
ot the transverse load Q. It seems, therefore, EI.@+Pu=O - -(8b)
ds
that even in the absence of the transverse
load Q, the maximum deflection becomes infinite Differentiating (8b) with respect to s
under the action ot only a compress iva load as
.g1ven by eq , (7). and since ~~ = sin Q, we obtain:

AlS. 3 Elastic Stability of a Column.

E1 ~~9 + P stn 9 = a - - - - - - - - (Bc )
The above conclusion as to the critical Multiplying (ec) by dQ and noting that:-
load was based on purely mathematical reasoning.
We have found a critical value of a compressive d'9 d9 -_ lis
d9 d (d9)
load which causes infinite deflection. d5T dB' an d 1rrcegr-a
- t1 ng

Far more important, however, is an in-

vestigation at the stability of a column which EIj ~ d (dQ) ... P sin QdQ = C, or
ds d5
should ~e based on phySical arguments. The
question arises as to what hanpens before the EI (d9)' _ P cos 9 =c
load P reaches its critical value as given 2" cs -(8d)
by eq. (7) and also how the column behaves if
this critical value is exceeded.
Since at end A, g =- a and M = E1 d9
ds
An elastic system is called stable under we find that C = - Peas Q.
given loads When infinitesimal loads added to
 =

ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A1B.3

Now let k ~~ - - - - - - - - - (8e) - - - - - - - - - - - - (8n)

Then sq. (ad) becomes,

n~
2P
------------ . (80)
1 d9 J~
V2 ds = -k V cos Q- cos a 1 or
Let us now write the bending moment M ~
PO at the middle paint in non-dlmentional form:

m=..L.~=~ (8p)
Per 1 1t
The total length ot the column in the
deflected shape 1s given by:- Since p = sin %is a function at a so is the

= _Ll d9 - - - (8h)
elliptic integral K and the ratios P~r and i
-a 12 k ..; cos Q _ COS <l
calculated tram equations (sa) and (80).

or 1 = fa
-a
d9
----(81) Thus P~r is a function or t calculated by
2k/ stn" ~ -sln"9/2 means at tables giVing elliptic integrals.
Thus m can be plotted against 0/1 as shown in
Denoting sin ~ by P and introducing a new FIg. A18.3.
variable 0 :-
011
sin ~= p sin ¢
O~/l
Equation (81) then becomes,
n
l=:f:-iz d9 =2K - - - - (8J)
1<: a / I-p"' sina 0 k

where Is called the Fig. A1S.3

complete elliptic integral of the first kind, " ' - - - - - - - - - - - - - - - - - - . ; ,.. m

and i t can be found in tables. It a and
therefore p Is very small, 'then p3 sin It III can Let us now examine the stability at
be neglected in equation (8j) and then various eqUilibrium configurations. Assume
that a load P~ is acting on the column and
n the column has a certain maximum deflection a
1 = f 1a 2 d¢ =~ =n j Ei where P~ does not correspond to O.

The non-dimensional maximum bending

whence P ;: n a:Er _ _ ____ _ ___ __ (8k) moment is:-
l'
:n' • _p'_ (_I
6
Pcr i- - - - - - - - - - - - (Bg)
The deflection at the midpoint of column
15:- 9 = 0, du = ds sin El, and trom Eq . (Sf)
The m~ versus Oil curves are straight lines.
u(z= .1)=0'.1... La sIn Qd9 (81) The column is in equilibrium if m = m~, or in
2 2k 0 lalna.!. _ atn" 9 other wordS, it the m~(O/l) curve intersects
2 2 the m(O/l) curve.
or in terms of .0:- We see from fig. Al8.3 that these curves
intersect at the origin only i f p lo <. Per. The
column, therefore, has only one pOSSible
- - (8m)
equilibrium form, tor example, that for which
all - 0, which is the straight form. When
From equation (8j) and equatton (8k), we obtain p~ ~ Per, there are two points of intersection,
AIS.4 THEORY OF THE INSTABILITY OF COLUMNS AND THIN SHEETS
one when 0/1 = 0 and the other (point A) tor
which 6/1 ~ O. The column thus bAS two
possible equilibrium forms, one straight and
one bent.
Let us now assume that at 6 = 0, the
column Is displaced by a small disturbance
and acquires a deflection O~. For pl C Per'
we see trom Fig. 3 that m s- m1.. Thus p1. Is
not sutficient to maintain the column in
eqUilibrium in the bent form and i t will spring
back to its straight rorm . Thus for p.l. ... Per'
the straight form Is stable.
It p1. > Per' then m1. > m. Thus m1. will
bend the column still further. This means
that if p1. - PeT' the straight form of
equilibrium Is unstable. The column will
continue to bend until m1. becomes equal to m
(point A in Fig. 3). It the column Is dis-
placed further tram A, the deflection becomes
larger than O. and m _ mJ. at the new position.
The column will spring back to pOint A. Point
A is therefore stable.
At P .: Per, the mJ. versus 0/1 Une is
tangent to the m curve at the origin. There-
fore, ror an infinitesimal disturbance, the
column will remain in equllibrium at the
dIsplaced POSition since for such small
disturbances mJ. remains equal to m. The column
is therefore in neutral equilibrium.
AlB. 4 The Failure of Columns by CompreSSion.
In diSCUSSing the ~tability ot a column
in the previous section, it was shown that
below the critical Euler load (Eq. 7), the
straight form is stable, above the Per the
bent torm is stable and at Pcr the eqUilibrium
is neutral. By plotting the curve PIPer versus
0/1 as shown in Fig. AIS.4, we observe the
following behaVior.
(a) Below PIPer = 1, there Is only one
eqUilibrium position, 0/1 • O.
PIPer
P;a./Per
B
1.0 .~
o. I 0.2 0.3
Fig. AlS.4
(b) At PIPer = 1, or at point (A), a bifurcation
at equilibr!um occurs and the column starts
to acquire two POSSible neighbor pOSitions
of equilibrium, the straight and the bent.
(c) Above PIPcr = 1, the column has two pOSSible
equilibrium positions 0/1 0 and ~ /1 .,. o.
Thus as far as initiation of instability is
concerned, the Euler load as given by Eq. 7 can
be considered as the critical load. The question
arises Whether this load has a practical use
for design purposes. A logical design criterion
is obviously the maximum load which a column can
sustain. We observe from FIg. A18.4 that the
load P increases tor increaSing displacement 0.'
This behavior is due to the development of large
deflections due to bending. However J over a
considerable range of deflections oJ the P -+ 6
curve is practically horizontal (for instance,
between points A and B the ratio 0/1 varies
trom zero to - 0.4). For such large deflections
for which the column load does not change
practically, i t is obvious that the column
ceases to fUnction properly. Therefore, trom
this point of view J the Euler load can be con-
sidered that which characterizes the maximum
strength ot the column.
The riSing part ot the curve BD holds as
long as the material behaves elastically. At
some paint OJ however, inside the almost flat
portion ot the c~rvr. the inner fibers of the
column acquire maximum stress equal to the
yield stress. If we carry out an elastlc-
plastiC analYSiS of the subsequent behavior J
we observe that the curve drops almost
1mm.ediately. Again this maximum load PB is
very near the Euler load. For deSign purposes,
therefore, the Euler load J which is a buckling
load, is a very good approximation to the
Ultimate load which the column can sustain.
Another argument will contirm the above
conclusion. In discussing the DuCkling of
columns in the pr-evf ous paragraphs , we have
assumed that the column 15 initIally straight,
centrally loaded and made of
homo~eneous material. Actual
columns, however J are imperfect
r
due to initial crookedness (for
instance, due to unaVOidable
tolerances in their manutacture),
due to slight load eccentricities
and due to lack of complete
hamogenity. Therefore, a certain
amount of bending Is always
present even for small loads.
Let us now examine the
behavior of such Initially im-
perfect columns by assuming a
certain initial deflection uo
Fig. ot the column axis (see Fig.
AlS.5 Ala.S). For small deflections,
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AlB.S

the change of curvature due to subsequent approaches the horizontal line PIPer = 1
bending (after loading) is:- asymptotically. ThiS curve, however, is valid
for small deflections for which the approxl-
mation:-

In the differential equation of deflection
one can prove that llR is th8 change of
curvature which for an initial straight column
coincides with the curvature itself. Thus in
the present case, where the bending moment is
Pu, the equation of deflection becomes:-
is valld.
By a treatment similar to that in the
previous paragraph, we will find that for
large deflections the load deflection curve
raises after the paint I (curve FIH). Due to
the onset of plasticity, the actual curve

.- k'u - - - - - - - - (9)
drops at the point I~ (curve FII~Hl). The
failing load at I~ can be either greater or
smaller than Per' but it is usually very near
to it.
Let us express uo in Fournier series:-
~ nuz In the above discussions we have shown
Z 3 sin--
uo .. , - - - -(10) that for all practical purposes the Euler
n 1
buckling load can be considered as the ulti-
Substituting (10) in (9), we find the • mata load which a real or practical column
can sustain. Besides its closeness to the
solution which satisfies the boundary actual ult~te load, the critical load can
conditions (u = 0 for z .. 0, Z = 1) is:- be eaSily calculated from equation (7) with-
~ nz out the necessity of carrying out a lengthy
u ::.1: on sin T - - - - - - - (lla) calculation which will include the initial
n=l imperfections and plastiC effects.
It should be noted, however, that the
where on • ,,-_:;-,,:- - - - - - - - - - (llb) buckling load ~iven by equation (7) is valid
when the uniform stress due to a compreSSive
load (0 = PIA, where A is cross-sectional
area) is below y1eld stress. It a is above
the yield stress, the theory of plasticity
predicts another value for the buckling load.
The deflection of the column at the center Referring now to equations (11) we tind:-
15:-
Pn = n~Pcr (Per tram equation 7)
- -(12)
~n
If we plot the deflection versus the load
we obtain the curve (Fig. AlB.B), which

piPer
Thus as P approaches Per' we see that

c etc.

I'
1.0 A - - ~- -
, G · x -- v
'":Il8. , '·I
_..,;6'i';-,,;_
- PIPer
<,
-;
'H~ In a buckling test we measure a: Omax
- 6, where 6 is the initial deflection at the
middle point. Thus:-

o ;;: 6:nax - 6. • ""O'L-_ and,

Per
h-::!-----------"1:"/I P
1

o
Per p - 6 = 6:l,
Fig. AIS.6 - - - - - - - - - - (13)

113
AlB.8 THEORY OF THE INSTABILITY OF COLUMNS AND THIN SHEETS

If in a buckling test we plot 6/P versus 0 1 0 1

0, we can obtain the critical load experi-
~entallY without kn~Nlng the initial deflection a 1 0 0
sin kl cos kl 0
1 1
6:1,' It is simply the inverse of the slope of
this curve. - sin kl - cos kl a 0

AlB. 5 Buckling Loads of Columns with vartcas End

COnditi0118. or sin kl =0 or kl = nn (n = 1, 2, 3 - -I

From the conclusions reached in the whence (lea)

previous discussion, we can consider the
buckling problem as an instability prOblem of
an initially straight column. Thus we assume Thus for Pn equal to the right hand side
a certain deflected position near the straight of equation (16a), we have a possible rorra at
configuration as another possible equilibrium equilibrium of the bent form. The smallest
form and seek the loads under which the non- value of Pn occurs at n = 1, and this 1s the
straight form is possible. Furthermore, only buckling load:
a small deflection analysts is necessa~J.
The general differential equation of
Per- = -r
nllEI
- - - (16b)

bending-buckling 18:-
The buckling load for other end conditions
can be derived in Similar manner.
- - - - - - (141
INELASTIC COLUMN STR~IGTH
and the general solution is:-
AlB.6 Inelastic Buckling. Introduction.
U3 c~ sin kz + C. cos kz + C3 z + c. - - (15)
Euler's theory of buckling 1s valid as
-The coefficients C~, C., Ca and C. long as the stress in the column nowhere ex-
depend on the conditions at the end supports. ceeds the elastic limit of the column material.
The various end conditions are:- We have seen that the analysis tor perfect and
imperfect elastic columns leads to the same
Free end: -
d'u
dz lll = ° d'u
dZF = ° result, namely, equation (7).
The case of the inelastic buckling, that
d'u
Pin end: - u =0
dz'" = ° Is, instability under axial load exceeding the
elastic limit stress, presents some difficulties.
As we will see, the perfec~ column analysis
Fixed end:- u ~ ° du
dz • 0
leads to a different expression !ar the critical
load than for the perfect column. This is due
to the tact that 1n the plastic stress range,
Thus we have 4 end conditions. These the material behaves differently under loading
give systems at four linear homogenous and unloading, as illustrated in Fig. AIS.?
equations. A trivial solution of these is Let us now examine the two cases:
the zero solution. For the buckling state,
however, C1 , Ca , C3 , C. are not all zero.
The condition at non-zero solution of the --:::::: Loading
above system is that the determinant at the
coefficients of C1 , C., C3 and C. is equal to I
zero. From this equation, we calcUlate the
buckling load. o
/ /~nJoading
•• /

For example, in a simply suoported beam,

d'u
(u = 0, dZllI = 0, at both ends), the end
-
•
;; /
/

conditions furnish give Deformation E:

Fig. AlB.7
c, + C. = 0 C. =0
AIB.7 Perfect Column. Reduced Modulus Theory.

Let us assume tr~t the perfect column 1s

For buckling we ~ust r~ve:- first compressed uniformly up to the stress 0.
To study the critical value, ocr, ot 0 for wh1ch
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AlS.7

the column becomes unstable we assume:- 15 8 11 = E 15 Ell - - - - - - - - - - - (19)

(1 ) That the displacements are small so that
the relation between the radius of
curvature R and the deflection u of the
elastic exrs is,
- - - - - - - - - - - (17a)
The distribution of the compressive (-)
stress and the tenSile (+) stress due to
bending is shown in Fig. Ala.sa. The stress
becomes zero on line (a a1.), which is at a
distance e from the centroid c. For
eqUilibrium of stresses on the cross-section
we have,

(2) Plane sections remain plane, therefore _ ~~ OS1d.A + ~all OS.dA =0 - - - - (20)
the change of straIn due to bending at a
distance h on the plane of bending IS, and for eqUilibrium tor moments,
h dllu
0< = if = h dz' - - - - - - - - - - (17b)

(3) The stress-strain relation tallows the Due to the linear distribution at stress,
Simple tension curve for the material. we have:-
(4) The plane of bending is a plane ot 00,
symmetry of the cross-section.
os, =-h
a, >
- - - - - - - - - - - (22)
Assume now a column with the cross-section Ocr.
as shown in Fig. Ala.sa be compressed in the os. =-h
a. •
IntrOducIng now (17b), (18), (19) and
(22) in equation (20), we obtain,

-Et ~, + EQ. = 0 - - - - - - - - - (23)

where,
Q~ = ~a1 h1. dA , Q,1I = ~a. h.dA (24)

are the moments of the cross-sectional areas

Fig. AlS. Sb
,'.,
!I
Flg. AlS. Sa
plastic stress range and that the compressive
stress prior to instability·be-a. To consider
the condition of buckling, let the column be
slightly deflected transversely. The stress
on one side of the column will then increase
due to the bending following the stress-strain
curve, while on the other side the stress will
decrease and will therefore follow the un-
loading elastic line shown in Fig. Ala.ab.
For small changes of the stress, on the
first side, the variation of stress Is related
to the variation of strain by:-
6 81 = Et (a) Oe1. - - - -
where Et (aJ Is the slope of the stress-straIn
curve at stress cr. On the second side the
Changes will follow the elastic relation,
that 1S,
to the right and left at lIne a a·.
From eq, (21) we Obtain,
- - - - - - (25)
where,
E= - - - - - - - - - (26)
E is the so-called reduced modulus, and
I1. and III being the moment of inertia of the
two sides.
We observe that the position of the
neutral axIs in terms of the axial stress is
gtven by eq, (24), While the buckj.tng eq. (25)
is Similar to the e Laat tc buckling eq. (14).
However, the value of K here is not given by
eq , (3), but by,
- - - (18)
K = ~ - - - - - - - - - - - - - (27)
Therefore all the results of the preVious
analysis will be valid for the case of In-
elastic buckling. For instance, ror a SImply
AlS.8 THEORY OF THE INSTABILITY OF COLUMNS AND THIN SHEETS
supported column according to eq, (7), we will
have,
(28)
Since E is a function of Ocr given by eq. (28)
and the value of Et at the unknown Ocr, the
calculation of the critical stress requires a
trial and error sireultaneous solution of
equations (23), (26) and (28).
AlS. S IInperfect Column. Tangent·Modulus Theory.
The tangent-modulus theory, originally
proposed by Engesser (Ref. 1), is based on the
assumption that at the critical state, no
stress reversal takes place, and the critical
stress, therefore, is determined only by the
tangent modulus Et. This theory was abandoned
early since according to the preViOUS dis-
cussion with the classical definition of
instability (perfect column, bifUrcation of
equilibrium) strain reversal does take place.
In recent years, hOwever, this tangent modulus
theory has been proved useful.
Under the assumption at ho strain reversal
both sides of the cross-section in Fig. AlB.Ba,
will be characterized by the same linear stress
distribution, corresponding to the tangent-
modulus Et. Thus the buckling equation will be,
EtI ~:~ .;. Pu =0 - - - - - - - - - - (29)
and the critical stress for sunply supported
end conditions becomes,
crt = n'E;;! __ - (30)
AI'
Since I .. + I a ~ I and J:" ~ Et, it tallows
from (26)
l: ~ Et and err ~ crt - - - - - - - - - (31)
The critical stress crt, therefore pre-
dicted by the tangent-modulus theory 1s
smaller than or from the reduced modulus
theory. Although tor perfect columns, the
assumption or no strain reversal is in
contradiction to the material behaVior in the
plastic range, most expertments have given
results more closely to the results by the
tangent-modulus theory.
To resolve this controversy, Shanley
(Ref. 2), proposed the following explanation.
For Simplicity, let a ~No-flange buckled
column be formed by two rigid legs (see Fig.
AlB.9) jOined in the middle by a plastic
hinge. Assume that this column starts to
buckle as soon as crt is reached. By consider-
d
1/2
b
1/2
Fig. AlS.9
ing the effect of a load P above the critical
load Pt corresponding to tangent-modUlUS,
Shanley proves the relation,
~
P = Pt 1 + 1
1+-r) - - - - - - (32)
~+
2d 1--r
where
It must be emphaSized that the buckled con-
figuration Is a stable one similar to that
considered in ~he refined Euler's theo~J.
Shanley has recognized the fact. tnat such a
stable configuration ~y exist after exceeding
the tangent modulus load.
It R = PIPt , Shanley found that the
relation between the variation of stress due
to bending and the compreSSive strain 6t
corresponding to at is:-
Concave side:
et
2 (R-l)
- - - - (33)
Convex side:
In Fig. A18.10, OE3,/Et and OEa/Et are
plotted against R tor ~ = 0.75. We observe
that while the strain on the concave side
increases very rapidly and reaches an infinite
value at the reduced-modulus load, the strain
on the convex side decreases initially verJ
slowly. Due to this picture we can conceive
that in a real column, which ras initial im-
perfections, the compressive strain will
increase more rapidly. Furthermore, the
rapid increase of 6E~ will cause a fast re-
duction of Et. The column, therefore, loses
its usefulness after the tangent-modulus has
been slightly increased. Thus the tangent-
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A18.9

,
~
~
. •
,
\
~;"";----'~j.-+--h,.4-+-+-l----l-i--+­

f-+-+-l'::::=f7~~k-+-l-+--+-I-+-
;,
II

I· -I-i--+--+-+-,.4--+~-l----l-i--+-

" \
;. .
• • ~

f'oI._"""". _
• m _
su_ ~
RAno (1..+) I. ,~ * • m ~

".~
HS.T..tOi _ _ .... (0.- i-. oti_l.
, A.n nl .... /(11 L_
• r-, t'l~
FIg. AlB. 11

,
•
...... Modulu ".a •

..,
GfU,T £.. TUTS
. . ,.. t18 1'UI1

FIg. AlB.lO

mOdulus, even though it does not actually ~ r-, ~. l~~

def1ne an unstable configuration, it repre-
sents the lower limit ot a spectrum ot possible I~ ~ .,...
buckled contigurations, the upper limit ot • <~
f':::: ~
which is the reduced modulus load which !'I..
corresponds to intinite deflections. .......
r-, t--.
Thus to summarize, sut't1cient experi-
mental results are available to show that the a ~ • • • ~ ~ ~ • • ~ ~

tailing stress at a column in the inelastic / ..1••• _

sU:l8UlIU1l1ATlO,(l.Jf)
...... _ IrSoT."'p IIHi'j•. bf IHii., t, S i•. J.
range can be round by replacing E by the
tangent modulus Et in Euler's equation, or, FIg. AlB.l2

Ocr ..
Figs. AlB.ll and 12 show how experimental
results check the strength as given by the
Euler equation USing the tangentwmodulus Et.
References:
- (34)
Ret. 1. Engesser F., Schweezeriche Bauer
Zeltung. Vol. 26, p. 24, 1895.
Ret. 2. Shanley, F.R., Inelastic Column
Theory, Jour. Aeronautical SCiences,
1947, p. 261.

..... _.
A18.10 THEORY OF THE INSTABILITY OF COLUMNS AND THIN SHEETS

PART 2

THEORY OF THE ELASTIC INSTABILITY OF THIN SHEETS

AIS.9 Introduction. curvature at the middle surface can be found

aw oW
approximately by omitting powers at -a
x
,-.-,
oy
Thin sheets represent a very common and
llnportant structural element in aerospace as compared to unity, as it has been done for
structures since the major units ot such the curvature at beams. Thus tne curvature
structures are covered with thin sheet panels. at the deflected middle surface in planes
Since compressive stresses cannot be eliminated parallel to xz and yz planes respectively are:
in aerospace structures, it is unportant to
know what stress intensities will cause thin
sheet panels to buckle. Equations tor the
1
Rx
=_ aa:
ax - - - - - - - (la)
buckling at thin sheet panels under various
load systems and boundary c ondt t ions ha. va Another quantity used in the problem of plates
been dert ved many years ago and are readily Is the so-called twist of the middle surface
available to design engineers. Part C ot this given by:
book takes up the use at the many buckling
equations in the practical design ot thin
sheet structural elements. The purpose at - - - - - - - - - - - - - - - (lb)
this chapter is to introduce the student to
the theory at thin plate instability or how The strains can now be expressed by means of
these buckling equations so widely in use by curvatures and twist at the middle surface.
design engineers were derived. For a broad In the case at pure bending of prismatic bar
comprehensive treatment at the subject at a rigorous solution was obtained by assuming
instability at structural elements, the student that cross-sections ot the bars remain plane
should reter to some at the reterences as atter bending and rotate so as to remain
listed at the end at this chapter. perpendicular to th' deflected neutral axis.
Combination at such bending in two perpen-
Al8. 10 Pure Bending of Thin Plates. dicular directions brings us to pure bending
ot plates.
To dart ve the theory ct Instab1l1 ty or
thin plates, we must" t1rst derive the theory
or the pure bending at thin plates.
In the tolloWing the analysis will be
contined to small detormations. Let x, y be
the- middle plane or the plate betore bending
occurs and Z be tpe axis normal to that plane.
,-- X Points at the x,
y plane undergo y z
small displace- Fig. 2a
ments, w In the
z-directlon,
Which will be
y
reterred as the
detlection at
the plate. The
z slope at the
middle-Burrace A
in the x-. and
y-directions
. ,,
i '

dr
r------ x
atter bending are
,
-d{jX-. ---- f+-l..,.
_ aw aw y
, ix - ax' iy = ay z
y , Fig. 2b
For small de- Let Fig. (2a) represent a thin rectangular
Fig. 1 flections, the Plate loaded by unifOrmly distributed bending
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AIS.11

Mx , My per unit length at its edges. These a.llw 2z

moments are considered pOSitive when they are exy = as axay = Rxy - - - - - - - - - - - (3)
directed as shown in the fIgure, I.e. when
they produce compression in the upper surface Since we treat the problem ot plates aS,a plane
at the plate and tension In the lower. Let stress problem. we find by means at Hook's law
also (Fig. 2b) be a rectangular element cut
out at the plate with sides dx , dy, t , The Ez I V
thickness t Is cons fder-ed very small compared "x = (I_Va) (RX + Ry) •
with the other dimensions. ObViOUSly the
stress conditions at the edges at all such Ez a.llw a"'w
elements will be identical to that ot Fig. 23. (l-i7i (ax' + V ay') - - - - - - - (4a)
Assume now that the lateral sides at the
element remain plane during bending and rotate Ez I 11
about the axes so as to remain nonmal to the "z = (I-V') (Ry + R
x) =
detlected middle surtace. ' Due to symmetry
the middle surface does not undergo any
extension and It is therefore the neutral
surface.
These normal stresses are linearly distributed
hom the geometry at the above described over the plate. thickneSS. Their resUltants
torm ot deformation the displacements in the must be equal to Mx and My respectively:
x, y, Z directions can be found as tollows:
J h/2 "x z dydz =MxdY
A point B on the middle sur-race has been -h/2
displaced to B~ by W in the z-direction. An
element at surface dzdy has rotated by an
angle equal to the slope at the deflected
J h/2 cry z dxdz =MydX
-h/2
middle surtace in the direction so as to
remain normal to the middle surface. See Substituting tram (4) we rtnd:
Fig. 3.

B M
x
= -D (a.llw + V a.ll,;) (Sa)
ax.ll oY

My==-D (o"'w
ay.ll + 7J o·w)
ax.:! - - - - - - ( 5b )

EW"3
where D = 12(I-V.) = the tlexural rigidity or
1
the plate. It besides the tlexural moments :M:x ,
My, there are uniformly distributed twisting
z moments Mxy and Myx along the sides of the
FIg. 3 plate of Fig. 2a, these must be equal to the
resultant at distributed shear torces crxy, 0xz
This angle tor small displacements is obviously along the sides at the element at F1g. 2b.
equal to ~;. Thus the horizontal displacement From eq, (3) we obtain:
Ux in the x-direction at a pOint at distance Z
trom. the middle surface is; 2Gz _ a·w
"xy = cryx. Rxy - 2Gz axay ~ - - - - (4<1)
aw
ux ~ - Z ax (The Sign - indicates
negative displacement tor positive z ) ,
MxydX • J h/2 "xyz dXdz , MyzdY =
-h/2
In a similar manner we tind the displacement
in the y-df.r-ec t fon , The complete displacement J h/2 oyxz dydz, Mxy =M -
yx-
system is; -h/2

aw aw D(l-lI) aaw (5c)

Ux· -zax' Uy. -Zay' Uz =-w(x,y) - - (2) -nay

The corresponding strains are: Equations (5) give the moments per unit length
tor pure bending and twisting at a plate.
a.ll·N Z
Ey =- z ayr == R
y
,
Ala. 12 THEORY OF THE rNSTABrLITY OF COLUMNS AND THIN SHEETS

A1S. 11 The Differential Equation of the Deflection Surface. a!'!xy aMy

- - dxdy - - - cxdy + Qy cxdy
ax ay =0 or
To develop thg theory at small deflections
ot thin plates we ~ake one more assumption.
a!1xy a!1y
At the boundary at the plate we assume that
its edges are free to move in the plane ot the a;- - ay + Qy = 0 - - - (6b)
plate.
From the moment equilibrium in the y-direction
we find:
al1xy aMx
--+ ---
ay ax Qx = 0 - - (6c)

By elImIr4ting Qx, Qy :rom 6 a,b,c, we tind

the equilibrium relation between the moments:
a.llMx ya:ilM a"!1 xy
-ax:il
-+- -
ay:il- 2 axay = - q (7 I

To represent this equation in terms of the

deflections w of the plate, we make the
assumption that the expreSSion (5) derived for
pure bending holds approxtmately also in the
case of laterally loaded plates. This
Fig. 4 assumption is equivalent to neglecting the
eftect on bending of the Shearing ~orces and
Thus the reactive forces at the edges due to the compressive stress cr~. This is an ex-
the supports are normal to the plate. With ter~icn or the engineerir~ theory of bending
these assumptions we can neglect ~~y strain of bea~. As in the case of beams it gives
in the middle plane during bending. good approximation for bending or plates under
transverse loads.
Let us consider. Fig.4, an element dxdy
at the middle plane. Along its edge the Introducing equation (4) into (7') we find:
moments MXJ My, Mxy are distributed. These
are the resultants of the bending and twisting aflw a"w
stresses distributed linearly along the --+ --+ - (8)
thickness of the plate (see eqs. 4 and 5). ax' ayfl
If the plate is loaded by external forceS
normal to the middle plane in addition to the The problem of bending of plates is thus
above moments there are ve~ical shearing reduced to integrating eq. (8) tor w. The
torces QXJ Qy, acting en the sides of the corresponding shearing forces in terms of th=
element ot Fig. 4. displacements are found from eqs • 4- and 6b
and c:
. h/2 aMxy a!1x a a.llw e-w
=f
O'xzdr J Qy
-h/2
"yzdz - (6)
=--+--= - D
ay ax ax
(o-x.ll + 3y :il ) ---(9al
_ aMy a ae w a.llw
Let Q be the transverse load per unit area
Qy - ay
~- - D- - (9b)
acting normally to the upper tace of the ax - ay (ax:il + 3ya)
plate. ConSidering the force equilibri~ in
the z-direction ot the element of Fig. 4 we The above analYSiS is sufficIent to seek
find: solutions ot specific problems. The general
procedure is to find approximate solution ot
aQx aQy the fourth order differential equation (8)
- - dxdy + - - dydx + qdxdy ::; 0
ax ay or = whiCh satisties the given boundary displacement
and force conditions.
aQx aQy
-- + -
ax ay + q =0 - - - - - (6a)
Ala.12 Strain Energy in Pure Bending of Plates.

In evaluating the strain energy of a thin

Taking the equilibrium ot the moments acting plate we shall ignore the contribution ot the
in the X-direction we obtain: shear strains which are generally small for
small detlections.
 ANALYSIS AND DESIGN OF
z Substituting !1x ' My, M~ in terms ot the dis-
Fig. 5a
I
I a
I aw
y
Mxy \ 2
\ Mxy
\ loads in the plane of the plate which are
Fig. 5b
The strain energy stored in a plate
element is obtained by calculating the work
done by the moments !1x dy , and M~ on the
element during bending. Since ~he sides ot
the element remain plane during bending the
work done by Mxdy is obtained by taking halt
the product a: :1xdY and the relative angle
of rotation of the two sides at the element.
Since the curvat~re in the x-direction is
a.a w
- ax a , the relative angle of rotation of the
sides 1 and 2 of distance dx will be - - , - dx.
ax
Thus the work due to Mxdy is:
I aaw
dVx = -"2 Mx axlil dxdy
Similarly the work due to MydX is:
I aa w
dVy - -"2 My axa dxdy
The twisting moment 1'1xydy also does work
against rotation of the element about the
x-axiS. The relative angle or rotation of the
two secti ons 1, 2 is oovtoustv axay dx.
the work done by MxydY is:
1 aa w
dVycy = 2' Mycy axay dxdy
and the work due to Myxdx = MxydX is:
(It is noted that the twist does not affect
the work prOduced by the bending moments,
neither the bending affect the work produced
by the torsional moments).
Thus the total strain energy per unit
volume at the plate is;
FLIGHT VEHICLE STRUCTURES A1S.13
K dVy + dVxy + dVyx
<lxdy =
sw a2 w
ax+a x.a dx
placement tram eqs. (5) we find:
i;!+~dx - - - - -(11)
oy a xay
!oJ This expreSSion will be modified later when we
will consider the superposition of compressive
related to the problem of plate buckling.
AlB. 13 Bending of Rectangular Plates.
The general ditterential equation tor
bending plates was given in section C2.3
(eq. 8). Two very usetul methods at solution
have been widely used, namely, the Fourier
Series MethOd and the Energy Method. Both
methods will be developed in the tollowing
for rectangular plates and various edge-
supporting conditions. The edge support
conditions are classified as tallows:
a'w
a) Built-in edge or Fixed: The deflection
along the built-in side is zero and the tangent
plane to the deflected middle surface is
----Cal horizontal. Thus if for instance the x-axis
coincides with the bUilt-in edge these con-
ditions are:
(wl y • 0 = 0 - - - -(l2a)
- (b)
b) Stmply supported edge: The deflection
along the simply-supported side is zero and the
bending moment parallel to this side IS also
zero. Thus if the plate is sUnply ·supported
a'w Thus along the x-axt s we have;
- - - - - - - (c)
c) Free edge: The bending moment,
twisting moment and shear torce along the tree
side is zero. Thus if the free side coincides
With the straight line x = a L we have:
- - - (d)
(<<x)
x=a
=0
However, as was proved by Ki rcnorr two boundary
conditions are only necessarJ to find a unique
solution of the bending problem. He has shown
tr~t the two last equations at the above
conditions can be replaced by one condition.
Al8 14 THEORY OF THE INSTABILITY OF COLUMNS AND THIN SHEETS

These boundary conditions are satisfied ~f we

Expressing the condition (MX)x=a = 0 in terms

take:

w= z
.. .. l: ("
'"1IlIl
sin ~ sin nny
a b
---(14a)
ot w we find the final tonn of the boundary :n=l n=l
conditions along the tree edge:
By substituting in sq. (18) with q given by
eq . (13) we find:

- - - - - - - - (12c)
z z Cmn
m=l n=l
In the following solutions tor various ~
mrtx nny
edge conditions will be developed. I: amn stn-a:- sin-- - - - - - (14b)
b
n=l
1. Simply supported rectangular plates
This relation is identity it:
Lat a plate with sides a and b and axes
x, y, as shown in Fig. 6, be simply supported
around the whole periphery and loaded by a
distributed load q = f (x,y).

Two methods ot solutions will be and thus:

developed:
1 ..
a) Navier solution by means ot double W = n"'D Z z sin mnx sin ;lny
a b
Fourier Series: m=l n=l
- - - - - (140)
We can always express t (x,y) in the form
of a double trigonometric (Fourier) series: In the case of a load qo uniformly distributed
over the whole surtace we have:
q = !(x,y) = l:
~

l: "mIl ern =
a- ern "F"
nrty
-(13a)
f (x,y) = qo = const.
m=l m=l

where: amn =ab4 fc,a .{,b t(x,y) mrtx nny

sinTstnodXdY
"__
~WJ.
=4q,
ab
fa f b s tn ~ • ern nTty dxdy
0 0 a b
=
169
- - - - - - - - -(13b) ~----- (15a)

-" where m, n are both odd integers. If either

x
or both are even amn = 0 and substituting in

H-'- t::
I--
I'
(140 ) :
169 0 0.:> anx nrty

I F
00
s ina- sin --0--
W=n"D Z Z
. m=l,3,S n=l,3,S
mn {m
aa
a
+£)'
b~

y <\

ffiII1Jtq Fig. •
with maximum oernecnton at the center,
- - - - - (15b)

~ ~-l
The boundary conditions are: OZn"D Z Z (-1) 2
m=l,3,S n=l,3,S
W = 0 , Mx ; a at x = a, x =a
W = 0 , My ; 0 at y = 0, y = b, or:
-----(15c)
(1) W =0 (2) ~; = 0 at x = 0, x =a This 1s a rapidly converging series and a
satisfactory approximation is obtained by
(3)11=0 taking only the first term. For a square
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AlB.15

plate this approximation becomes;

, .T ~
b/2 1'--' ~
0.0454 qoa (for V : 0.3)
Ell'
t
b/2
f----t...-.~x
Fig. 7
which is by 2-1/2% is error with the exact
solution. 1
The expressions for bending and twisting
moments are not so quickly convergent. To y
tmprove the solution another series solution
can be developed as follows: where for symmetry m = 1,3,5 •
b) Levy alternate single series solution: This equation can be satisfied for all values
of x it:
The method will be developed for uniform
load qo • const. Levy suggested a solution in m"j("
the rorm: + - - Ym ::: o - - - - - -(1gb)
a'
W: -
Z
m-1
Ym (y) sin "a TtX
- - - - - - - (16) The general solution or (1gb) is:

Ym (y) = qa" [",- cos h ~

a + 8m. ~
a sin h ~
a
where Ym is a function of y only. Each term. D'~
ot the series satisfies the boundary conditions
a'w + ('1_ sin h mny + ri.. ~cos h mn
y]
w = 0, ax'" ::: a t x = a. It remains to
0 '"'!U a '"'Ina a
determine Ym so as to satisfy the remaining
-(20)
aaw = 0 at y = b.
two oouncar-y .cono i t i one '1'1 = 0, ay"
Since the deflectIon is symmetrIc wIth
respect to the x-axis it tallows that em =
A turther Simplification can be made it
we take the solution in the fonn Om : O. Thus:

W=WJ.+W a - - - - - - - - - - - - (17a)

where
(L cos h rnn:y + lL mny sm h mny) sin ~
- - (17b) "'m a '-ma a a

is the deflection of a very long strip with or

the long side in the x-direction loaded by a
uniform load qo supported at the short sides
x = 0, x = a, and free at the two long Sides.
w: q~" ; (-b + Am cos h ~y + Em ~
Since (17b) satisfies the differential y)
equation and the boundary conditions at x ::: 0, sin h n: sin ~
X = a, the prOblem is solved it we tind the
solution of: =
where m 1,3,5 . • Substituting this

--.
a"W a
ax'
.--, 0
a"w a
ay' (18)
expression into the boundary condItions:

w : 0,
a·w
ay": 0 for g : -
+ b
"2 we tind:
With w a in the form of (16) and satisfying
together with wJ. of eq. (17b) the boundary 4
a'w • b Am cos h am+am Em sin h Cm,+lleme = 0
conditions w = 0, ayr::: 0 at y = - 2' (see - -(21a)
Fig. 7). (An + 2Bm) cos h Om + am 8m sin h am = 0
Substituting (16) into (18) we obtain:
where am = mrtb/2a

mton" ID1tX From these equations we find;

' - _ Ym) .srn - _ : 0
a' a
- -(lea)
 AlB. 16 THEORY OF THE INS:rABILITY OF COLUMNS AND THIN SHEETS

2(~ tanhOm + 2)
Am =
nllm" cos ham (V ' = qa::l + (l-V)qa'n' Z
~

ux}mox 8
m=1,3,S
2
8m = - - - - - - - - (21b )
n"m ll cos ham '( 2V
m Am -l-V 8m )
Thus:

2:... [1- ('m tanham·2) (Mylmax = Va-

na
- (l_V)qa'in::l
II 0<:1
Z
m" 2 cos h rJ:n n.=1,3,5

cos h 20mY + am . 2 y sin h 2a mY ] - - (2l!)

b 2 cosham b b
c) Solution by means of the principle
sin mrnc - - -(210) of Virtual work
a
From the discussion of the method (a) we
The maximum deflection occurs at the middle can represent w in double Fourier Series:
X -_ 2"'
a y - O·
- .
m-l w= Z
nrtx nrty
---(22a)
Z Cmn sin - . - sin -b-
Wmax
__ 4qa' H) 2 ~_am tan ham. 2) mal n=l
~'D m=1,3,5 mil 2 cos ham
The coefficients Cmn may be considered as the
co-ordinate defining the deflection surface.
- - - - - - - -(21d) A virtual displacement will have the fo~:
The summation of the first series of terms
corresponds to the solution of the middle of ~..
'-'....
= '0....
'-'''mIl' S
1n =
-a. sl n nny
"F" - - - (22b)
a uniformly loaded strip, sq. (17b). Thus;
OQ m-l
5 qa" 4qa' Z (-1)-2- The strain energy Vi can be found by substi-
tutlng (22a) Into eq, (6). Atter a few
Wmax :I 384 0 - 1"(-3D m=1,3,5 fu" algebraic manipulations we find:
am tan h am'• 2
2 cos ham - - - - - - - -(21e) Z C!nIl1l
ne L
This series converges very rapidly. Taking a
square plate, alb = 1, we find from (21a): (23)

e.t.c. Let us now examine the deflection or the plate

at Fig. 6 with a concentrated load P at the
_ 5 qa 40 4qa. 40 point with co-ordinates x Y = n, The :II { ,

- 384""ll - - - (0.68562 - 0.00025· .... ) increment of strain energy due to the incre-
~'D ment ot the deflection by:
40
- qa
= 0.00406 D
Ow
= ... mroc nny
,-,Cmn sin -a- sin "F" - - - - - - (24a)

We observe that only the !1rst term of the

series in (2le) need to be taken into con- is found trom (23)
sideration. n 40 ab m' n"::1
OV I =-4-Cmn (. . . . b") 0 Ctnn - - - - (24b)
The bending moments are found by substi-
tuting (210) Into eqs , (5). The msxlmum
The increment of the work of the lead Pis:
bending moments at x = t 'y =0 are:
ow = P0Cmn sin m! sin n~n - - - - - _ (240)
From OVi - OW = 0 we obtain:

_1i 411·• •••··--.1I&1II• • • • • • • •111•••••••••••••••••••••••••ill. . . .Lil

.!!!
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES Al8.l7
.

,rr'ab "
[--4- -mn P sin ~§ . sin n~~ ] Let us consider the case of a uniform load q.
we write the deflection in the form:

6Cmn = a - - - - - - - - ~(26a)

Since 6C~ is arbitrary its coefficient must Where w~ is the deflection of a simply
be zero. Thus: supported strip of length, a, which for the
system of axes of Fig. 8 can be written (see
4P sin a3.. • sin E!!!l
b
Levy's method in previous section):
Cmn = ll
mnx
n"abD (m.a ,. n ), sln- ,
artJT a-

_ 4P and w2 is represented by the series:

TrJ _~ Z
ee
mol
w2 = L. Ym (y) sin ~ - - - -(260)
m-=1,3,5 a
mrtx nrty
sin ---a- . sin -b- - - - - - -(24d)
where w. being a solution of
This serieS converges rapidly.
plate (alb = 1):
For a square a"w2
ax' .--.a"Wa
aY' = 0

1 is found as in the previous section:

Z Z - - - -(240)
m=l n=l (m2 + n 2)2
Ym = qaDt. (Am cos h~ + Em mrry sin h mny +
By takir~ only the first four terms we find a a a
sin h ~,. n __ 2!!!!. h mny)
Fa'
Wmax:! 0.01121-
em
a -m a cos a _ -(26d)
D-
It is obvious that the two first boundary
"Nhich is 3-1/2% less than the correct value. conditions are identically satisfied by w =
w~ ,. wa • The coefficients Am, Bm, Gm, Om
(2) Rectangular plate with ~NO opposite must be determined so as to satiSfy the last
edges Simply supported, the third edge tree tour boundary conditions. USing the conditions
and the fourth edge built-in or simply (25b) we obtain:
supported.

+.-
L _ 4,
,--,..-----r.... x .nm - - 1"(lImli , em=-Dm
By the conditions (25c) we find:

1 .. 0_'_.
1\''''.
!3"1I1(1-1I) co.sIl·%+211coShB.,.-!l(l-!l)~ !ln~a.,..(l-/l·)
:3+11)(1-111 cOSh·s..:" n·l/)·a.,· .. (."1/). ___ _ (25s)
;, 0_'_. (3+111 (l-ll)stnh a., COShs,.·V(l"!I)stn~B.t-II(l-lI)a" CO!h~-(l-II)'3:l
n"",' (3+JlI(l_lIlcos~'3,,+fl_I/)' s,.··(1+11)·

y
Fig.
Assume that the edges x ~ a and x = a are
simply supDorted, the edge y = b free and the
edge y = 0 bUilt-in (Fig. 8). In such a case
the boundary conditions are:
a
Substituting Am, Em, em and Dm in eq, (2ed) we
find the deflection. The ~imum deflection
occ~r~ at the middle of the free edge.
AlS. 14 Combined Bending and Tension or Compression of

w -= 0,
a'w =
aX""" 0, for x = 0, x = a (a)
Thin Plates.

In developing the differential equations

aw
w = O'ay=o, tor y 0 (b) of equilibrium in previous pages, it was
assumed that the plate is bent by transverse
a w 2
a'w -- (25 ) loads normal to the plate and the deflections
(ax"". ay.a) - 0, were so small that the stretching ot the middle
plane can be neglected. It we consider now the
ta'W
ay' .2(1-V) 1.:!!...]
Clray = ° for y = b (0 ) :ase where only edge loads are active coplanar
AlS. 18 THEORY OF THE INSTABILITY OF COLUMNS AND THIN SHEETS

with the middle surface (Fi~. 9) we h~ve a r-------;.. x

p lane stress problem. I f we assume that the
stresses are ~niformly distributed over the Nx
dx
thickness a~d denote by Nx' Ny, N~J' Nyx the
resultant force ot these stresses per ~~it
length of linear element in the x and ow
Y-directions (~ig. 9) it 15 obvious thRt: ax
N + oW"dx
Nx .: hox ' Ny .: hoy x ax
- - - - - - - - (27) y
Nxy :;:: Nyx .: hoxy = hoyx Fig. lOa

Consider an element of the ~iddle surface

.-.".-_=-_--'<__.... x
dx. dY (Fig. loa). The conditions af force
eqUilibrium in the x, y-direc~1ons are ~~ven
by eq, (28), Consider now the pr-ojec t i on ot
the stresses Nx, Ny, NXY in the z-directlan:
(a) Projection of Nx: From FIg. loa it
follows t~at the resultant :Jrojectlon is:

dx)(~+a2w rjxl
x +awx
N a aw
- I X Y ax +
(N
ax ax a;{"l ,
y
and neglecting terms of higher order:
Fig. 9

(N aaw + aNx aWl

The equations of equilibrium in the absence 'xaxa ax 'ax ----tal
at body forces can be written now in terms of
these generalized stresses Nx , Ny, N~j by (b) Projection of Ny: By similar argument
substituting frem eq. (27 ) to the equations we find that this ~rojection 1S equal to:
of equilibrium.
2
(N a w + aNy aw)

--.
aNx
ax
aNxy
ay --.
aNy
- 0 , ay
aNxy
ax - o - - - - - (28)
Yay' ay ay
-------(b)

On the other hand if the plate is loaded by ,0:;,

transverse loads the stresses give rise to
pure bending and tWisting moments only, The
equations of eqUilibrium for the latter have
been given in before (see eqs. 6, 7, 8). If dy
both transverse loads and coplanar edge loadS
are acting simultaneously, then tor small
vertical deflections the state of stress is
the superposition ot the stresses due to
1
Nx, Ny, Nxy and MX J My, I1xy. For large
vertical deflection af the plate, however,
Nxy
-=
a Zttd
there Is interaction of the coplanar stresses y Y
I
a.~d the deflections. These stresses give rise
to additional bending moments due to the non-
zero lever arm of the edge loads fram the Fig. lOb
deflected middle surface. as 1n the case of Projection of Nxy and Nyx:
(c) From Fig. lOb
beams. When the edge loads are ca~pressive we find:
this additional moments might cause instability
and failure of the plate due to excessive
vertl~al deflections,

In thiS chapter the problem of instability and neglecting terms of higher order:
of pla~es will be examined.
~
2

the edge loads are compressive and

'~en
'N aw +
\ xy axay
aWl dxd
ax· ay ....y - - (c)
give rise to additional bending moments eq. (8)
must be mOdified. SimilarlY we find the projection of NyX = Nxy:
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AIS.19

- - - - - - - (d) ViN 1 II (NXEx+NyEy+Nxyd"xy)

=2 0 0 0
dxdy =

Thus in eq . (6a) the terms given by (a), (b), 2~1 I (N; +N; -2!INxNy + 2(1+V)Nxy') '- - (a)
(e) and (d) should be added (divided of course
by dxdy}:
During bending due to transverse loads
or/and due to buckling we assume that the edge
loads and consequently Nx , Ny, Nxy, remain
constant. Its variation is thus zero and we
do not consider it in the following. Let us
apply now the transverse load that produces
bending. (~e can also consider bending due to
aut due to the equation of equilibrium (28) the other transverse distUrbance, which is the
two terms inside the parentheses in (e) are case of bUCkling). It u, v, are the displace-
zero. Thus: ments at the middle surface due to the coplanar
loads (which are assumed constant across the
ao.x
dX
+ 0iY
y
+ q aaw
+Nxax a +
thickness) and w the bending deflection of the
plate it can be shown that the strains are:
aaw aaw
Nyayr + 2Nxy axay :: a ------(29) = au + 1.. (a w ) .
ex ax 2 ax
Eq , (29) replaces eq , (ea ) when edge loads are =~+ 1:. ~).
present. Eqs. (6b) and (6c) are, however, still Ey ay 2 ay
valid since they ex~ress moment eqUilibrium _ au av aw oW
of the element dxdy ln wnlen ~ne con~rl0U~~0n
ot Nx, Ny, Nxy is zero. Thus eliminating Qx ,
O'J:Y - ay + ax + ax . ay - - - - - - (b)

Qy between (6b), (6e) and (29) we find:

Let us apply now bending with constant coplanar
stresses. Due tn stretching of the middle
surface the energy Is;

1 aa w aa'N aaw II (NxEX+ NyEy+Nxy oxy) dxdy - - - - (0)

o (q+)Jx ax 3 +Ny ay" +2Nxy axay) - - - -(30)
Introducing (b) into (c) and adding the strain
energy due to bending, eq , (11), we find the
Eq. (30) replaces eq , (128) when edge loads total change of strain energy due to bending
are present and the deflections are large so which Is:
that instability ~ight occur.
The distribution of the coplanar stresses
Nx, Ny, Nxy can be found trom eqs , (28) by
solving the plane stress problem. In the
following the above theory will be applied to
rectangular pla~es.
AIS. 15 Strain Energy of Plates Due to Edge
Compression and Bending.

The energy expression for pure bending,

eq. (II), must be complemented to include the
contribution of the edge coplanar loads.
Assume that first the edge loads are applied.
Obviously the strains due to the stresses
NX1 NY1 Nxy are:
Here u, v are the additional coplanar displace-
ments after bending has started. It can be
shown by integrating by parts that the first
integral is the work done during bending by the
edge loads. For instance taking a rectangular
plate this integral becomes;

fa afa b [Nx
au oV au av]
ax + Ny ay+ NX"J (ay +ax) dxdy =

- - - - - - - - - - -(31) fa b (I Nxul~ + INxyvl~ ) 1y + faa (I Nyvj ~ + 1NXYU! ~)dx-

-orarblU(aNx+~) +V(aNY+~)l
The strain energy is: dxdy - - (e)
t> ax oy ay ax

I.J C
AlB.20 THEORY OF THE INSTABILITY OF COLUMNS AND THIN SHEETS
Obviously the first two integrals represent
the work done by the edge loads while the
second integral 1s zero due to the equilibrium
equations (28). Thus the work of the edge
loads is:
IItI -_ II [ Nx ai{+Ny
au ,av au + av]
ay +Nxy (ay ax) dxdy- - (r)
We assume now that for small deflec~ions
the stretching at the middle surface of the
plate is negligible. (This 15 the so-called
1nextensional theory at plates). In this case
by zeroing the strains 1n eq, (b) and substi-
tuting in (f) we find;
"\'IN 1 II (a w) .
= -2 [ Nx ax +N
j, Y
(-)'
ay +2Nxy -
In the strain energy expression, eq. (a)
the first two terms cancel each other and the
strain energy is due only to bending:
li
Vi = ID!I f("lI W+ aayliJ
W' lII _ The energy method can also be used in investi-
2 lax:!
' w a'w
2 (l-lI) [ aax'· a'"
ay" - (axay)'
In the absence ot transverse loads the work
of external torces is Simply due to the edge
loads:
ExpreSSions (32b) and (c) will be used in
solVing the buckling problem by means ot the
principle ot Virtual work.
AlB.16 Buckling of Rectangular Plates with Vartous Edge
Loads and Support Conditions.
General discussion
In calculating critical values of edge
loads for which the tlat torm ot eqUilibrium
becomes unstable and the plate begins to
buckle, the same methods and corresponding
reasonings as tor compressed bars will be
employed.
The critical values can be obtained by
assuming that the plate has a slight initial
curvat~re or a small transverse load. These
values ot the edge loads for which the lateral
detlection w becomes infinite are the critical
values (see Part 1 for similar treatment in
columns) .
Another way of investigating such
instability is to assume trat the plate buckles
due,to a certain external disturbance and then
to calculate the edge loads for which such a
buckled conf1guratio~ (ceflections different
from zero) is possible. It was found in ~he
case of column that this latter solution
(Suler's solution) approaches asynptotlcally
the first at the l1~it where the deflections
become extreillely large but for even small
deflections the edge load acquires a value
very near the Euler's critical value. The
latter teChnique is mathematically more
convenient and it gives ~or plates also a ~ery
good estimate of their compresSive strength.
In the following we shall use this latter
approach by assuming a plate with edge loads
and no transverse load. Sq. (30) beco~es in
this case;
aw ew awl a"w +2 a"w ." a'w - 1 (N a'w +N a"w +21N ;;.'l.w
ax , ay
- dxd y ax" ax 3ay2' ay" - D ! X axr
I y if xy axay
u ,
i
(32a) (33)
By solving eq. (33) 'Ne will find trat the
assumed buckling made is possible (w ~ 0) for
certain defi~ite values of the edge loads, the
smallest of which deter.nines the critical lead.
gating buckling prOblems. In this ~ethod we
]1 dxdy - - (32b)
assume that the plate is i~itially under the
plane stress conditions due to the edge loads
and the stress distribution is assumed as known.
We then consider the buckled state as a possible
configuration of equilibrium. T~e change of the
work is given by eq. (32a). We inter,ret here
w as a Virtual displacement though we do not use
the variation symbol O. Thus the increment of
- - - - - - - - (320) work ow is ~iven by (32a) and the increment of
strain energy 6Vl is given by eq. (32b). If
6~Vl for every possible shape of buckling the
flat equilibrium 1s stable. If 6W~Vi for a
certain shape at buckling then the flat :on-
figuration is unstable and the plate will buckle
under any load above the critl:al load. It
OW = OVi' the eqUilibrium is neutral and fro~
this equation we find the critical load. The
critical load therefore 1s found :rom the
equation:
1 II [ N (aw.
- '2 x ax) + Ny (aw •
ay) + 2Nxy
sw s
ax· ay dxdy =
D ! a'w +ay')'
zlll(ax"
a'w [a·w
-2(1-11) ax'
a'w a'w Jl
'ay" - (axa/j,dxdY
- - - - - - - - (34)
Here w is a certain assumed deflection which
satisfies the boundary conditions (Virtual
deflection) .
Ala.17 Buckling of Simply Supported Rectangular Plates
Uniformly Compressed in One Direction.
Let a plate of sides a and b (Fig. 11)
Simply supported around its periphery be com-
pressed by load Nx uniformly distributed
along the sides x = a and x = a. From the
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A1S.2l
x
a .... a+c+e+
I b + d + f + - (bl
;....... I:::::::i
~ 1 ~ Nx
ThiS fraction has some intermediate value
between the maximum and the minimum of the
tractions (a). It follows that i f we wish to
make the fraction (b), which is Similar to the
Fig. 11
fraction of eq. 35d, a minimum, we ~ust take
only one term in the numerator and the corres-
ponding term in the denominator. Thus to make
obvious solution of the corresponding plane the fraction of eq. (35d) minimum, we must put
stress problem we find that the state of all the parameters Cu , Cu , Cn •••• except
stress is everywhere a simple compression one, zero. This is eqUivalent to assuming that
equal to Nx (the load at the periphery). the buckling configuration is of simple
sinusoidal form in both directions, i.e.
The deflection surface of a simply mrtx nrty
supported plate when bending takes place have Wmn = Cmn sin a- .
sin 0-. The minimum
been found preViOusly (see eq. 14a). Its expreSSion tor Cmn obtained by dropping all the
general expreSSion can be written in a double terms except Cmn becomes:
series rora:
= - - (350)
f - - - (35a)
~

W= Z Z Cmn s in '"':' atn

m=l n=l It is obvious that the smallest value of Nx is
The increment of strain energy found by obtained by taking n = 1. This means that the
plate buckles always in such a way that there
sUbstituting (35a) in the right-hand side of can be several half-waves in the direction or
(34) Is:
compression but only are half-wave in the
perpendicular direction. Thus ror- n = 1, eq ,
:: fl· ab D c;' Z (35e) becomes:
8 m=l n=l

The increment of work done by the external

forceS is found by SUbstituting (35a) into the
left-hand side of eq . (34) and Nx .,. const , The value of m (in other words the number of
Nx :: Nxy ::::r O. Thus: half-waves) which makes this critical value
the smallest pOSSible depenrts on the ratio
alb and can be found as follows:
6'. =~NX [arb
o ."
(~Wtd.."'{dY =
oX
= = Let us express (36a) in the form:
n'b
ea Nx Z
m=l n=l
Z m' Col..,' -- ---- (350 )
(N )
x cr = k -n'D
ba
- - - - - - - - - - (36b I
From the equality OW ~ OVi' solving for Nx ' we a,
obtain: where k is a numerical factor depending on (bl.
From (368 I and (36b) we have:
= =L: Cmn a (? " + ~a)a
naaaD Z a
b 1 aa
m=l n=l k = -aa (m + -:II. _)a - - - - - - - - - (36c)
Nx =
ee
--- (35d) ba
ce
Z Z m' C""
,
It we plot k against ~ for various values of the
m=l n-l
integer ~ = 1,2,3, .•..• we obta1n the curves
Here Cmn is arbitrary. ~e are interested, of Fig. (12). From these curves the critical
however, to find that values of Cmn which make load factor k and the corresponding number of
Nx minimum. To that effect we use the follow- half-waves can readily be deter.nined. It is
ing mathematical rea~oning: only necessarJ to take the corresponcing point
(;~) as the axis of abc1ssas and to choose that
Imagine a series of fractions:
curve which ~ives the smallest k. In Fig. 12
a 0 e _ - - (al the portion of· the various curves which give
15" 'd .. , . the critical values or k are shown by full
lines. The transition from m to (m + 1) half-
If we add the numerators and the denominators waves occurs at the intersection of the two
we obtain the fraction: corresponding lines. Fromeq. (36c) we find:
 Ala.22 THEORY OF THE INSTABILITY OF COLUMNS AND THIN SHEETS

•
7
6
5
\\
\
'm'"
"\
\

A:
n=3\

~I:
,\
'
"mO<'

¥
\ x," ~ /
~,,"
".-
,

4 , , ,
3 ,
, ,,, ,;
2 , ,
I ,,
, I, ,
o , Fig. 13
o '16 ~

3/12 4
alb
_ (3'79.)
Fig. 12

znb a (m.+l)b
-+-= ---+ a~d by introducing t~e pa~~et~r:
a JIb a or:
nOD
00
• =-.-
a h - - - - - - - - - - - - (37'))
~=vm(m+l) ---(36d)
b
',.Je obtain:
Thus the transition from one to two half-waves 3
occur for: a ... '18. a
0e (n - n~) - - (37.:::)

~ =j 1 (1+1) =f i Taking any integer :n a~d n the corres~oncing

deflection s~rface of the buckled )late i8 given
rr-om t'NO to three for by:

\-.'mn :: ('-tnn 8 I n -a-·

mnx
s ~..n -b-
nrty (38)
~ =j 2 (2+1) = 16
and the cOY7espondlng Ox, cry are g1ven by (37c)
and so on. which 1s a straight line in the diagram cry, Ox
(Fig. 14). By plotting such lines for various
Tne number or half-waves increases with the pairs of m and n we !inc the region of stability
ratio alb ann for very long plates m Is veFy and the critical combination of ox, 0y which is
large. on the periphery of the polygon for.ned by the
full lines of Fi~. 14.

Thl S means that a very long plate buck.Lee in

half-waves the lengths of which approach the
/
width ot the plate. The buckled plate Is /
subdivided into squares. ' ...
\ I ...............

The critical value or the compression '-0

\~
stress Is: ,-
I
I
= (Nx 1cr = --=:.:.:..:=-:-,. t"
---(36el I
/
h 12 (l-v"1 b" I

(t : thickness) Fig. 14

AIB.18 Buckling 01 Simply Supported Rectangular Plate A18. 19 Buckling of Simply Supported Rectangular Plate
Compressed in Two Perpendicular Directions. Under Combined Bending and Compression.

Lat (Fig. 13) NX 1 Ny the uniformly distri- Let us consider a Simply supported
buted edge compressions. Using the same as rectangular plate (Fig. 13). Along the sides
before expression for the deflections (eq. 35a) x :: 0, X = a there are linearly distributed
and ap~lYlng the energy equation (33) with edge loads ~lven by the equation:
Nx, Ny - constants (whiCh is the sol~tion of
the corresponding plane stress problem) we find:
Nx := No (l - A{) - - - - - - - - - (39a;
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A1S.23

which is a combination of pure bending and rle examine for each value of m the solutions of
~ure compression. Let us take the deflection the system (38f). Starting from m 1 and =
a 6 a i n in the foron: denoting:
oo
w = ""
.z Cmn ~
sm TJtX
-,-. 1 nny
!5 n -b- - - - (39b) Ocr =
(No)cr
- - - - - - - - - " (39g)
m=l h

No we obtain from (39f):

I
I
a
C'ln [(1 + n:l~) - Ocr ::~ (1-})]-
b
1 allih =co CUni
al-crcrn"D Z (n:ol_jlll)a-------- (39h)
1

These are homogeneous equations in a~~, a~a •..

y etc. The system possesses a non-zero solution
Fig. 15 (which indicates the possibility of buckling of
the plate) if the determinant of eq. (39h) is
zero. From this condition an equation is
Substituting in the right-hand side of obtained for Ocr' Obviously the system (39h)
eq • (34) wefind the variation of strain
is of infinite number of equations (n = 1, 2,
energy: 3 .•• ). A sufficient approximation is
obtained by taking a large but finite number
or terms and find the solution of the rrm te
determinant (USing for example digital
computers). Thus a curve of ecr versus aID is
while the increment of work 1s: obtained for m • 1 like that or Fig. 12.
Repeating the same calculation for m = 2, 3
;w=.lrarbN (l_,L)(aw). dxdY= etc., we find similar curves of two, three, etc.
2 Jo Jo 0 A b ax
:'I
~
4
I
~
b
2
Z .z
OCIaa
c
TIlI1
:1 m:1:1 1
n _~ Z
ar--
OC

b
:1:1
m
~
1'[
halt-wave lengths. The regions of the curves
with minimum ordinates define the region or
stabU1 ty as In F1g. (12).
m=l
AlS. 20 Inelastic Buckling of Thin Sheets
"=
£. z Cmn:l- - - Z
Bb' ""
[4 !1=1 1'[:01 n=l The problem of the inelastic buckling of
thin sheets has been extensively studied by
various authors. The main difticulty in such
Equating (39c) and (39d) and solvir.g for No studies is in reference to the stress-strain
we find:
relations at plasticity under complex states
at stress. Many controversial discussions
have appeared in literature without resolving
the theoretical difficulties. For this reason
we will not develop the theory of inelastic
buckling in this chapter. Some ot the better
references on this subject are listed below.
- - - - (39.)
Cnapt er- C4 presents the plasticity
"!he coefficients now Cmu are so adjusted that correction factors to use in calculating the
(Nol er becomes minimum. 8y taking the inelastic buckling strength of thin sheets.
derivative of expression (38e) with respect
to each coef~icient emn and equatir~ these to AlB.21 References.
zero we f:"nd:
(1) Bleich,?: Buckling Strength of Metal
Structures. Book by McGraw-Hill.
(2) Stowell, E.Z., A Unl~ied TheoFJ of Plastic
Buckling of Columns and Plates. NACA
Report S9S, 1948.
(3) Gerard and Becker: Handbook Of Structural
Scabi11ty. nACA T.N. 3781, 1957.
(4) Gerard: Introduction to Structural
Stability Theory. Book by ~cGraw-Rll1
Co., 1962.

'r)
........ '

"·-~~~f.
" ':-
:;:,:;0""
A18.24

Courtesy of The Boeing Company, Seattle. Washington

This multiplEl exposure photograph of a Boeing supersonic transport model shows the variable-sweep wing
in three configurations: forward for takeoff and landing, swept part way back for transonic flight. and swept
completely back as an arrow wing for 1800-mtie-an whour supersonic cruise.

SPECIFICATIONS (Basic Design)

Gross Weight 430, 000 pounds

Payload. 150 passengers
lWge 4, 030 miles
Takeoff Distance (Max. gross weight) 6, 000 feet
Landing Distance 5, 800 feet
Cruising Speed I, 800 miles an hour (Mach 2.7)
Takeoff Speed 175 miles an hour
Approach Speed 136 miles an hour
Wing Span (forward position) 173 feet, 4 inches
(aft poSition) / 86 feet, 4 inches
Length 203 feet, 10 inches
Height 48 feet, 4 inches
 CHAPTER A19
INTRODUCTION TO WING STRESS AL'\[ALYSIS

A19,l TYPical Wing Structural Arrangement
For aerodyr.a~lc reasons, the wins cross-
section must have a streamlined shape commonly
refeTT9d ~o as an airfoil section. The aero-
dJ~~~lc ~orces in flight chan~e in ~gn1tude,
~irection and location. Likewise in the various
landing o8erations t~e l~ads ctange in ~agnl­
tUde, direction and lecation, th~s the required
3tructure must be one that can efficiently
relatively thin. In general the wins st~uc~ural
flange arrange~ent can be classified ~nto two
types; (1) the concentrated :lange type where
flange ~aterial is connected directly to in-
ter~al webs and (2) the distributed flan~e ~J?e
where str1ng~rs are a~tached to s£in between
internal webs.
A9.3 shows several structural a~an~e­
F'~g.

resist loads causing combined tension, com- ~ents for wing cross-sections for superscr.lc
pression, bending and torsion. To provide aircraft. SupersoniC airfoil shapes are
torsional resistance, a portlon of the airfoil relativeLY thin compared to subsonic aircraft.
surface can be covered with a metal skin and
then adding one or ~ore internal metal webs to
produce a single closed cell or a ~ultiple cell

l
wing cross-sect~on. The external skin surface Fig. d
which 15 re Lat tve Ly t ntn for subsonic a i rcrarr
is ef:icient for resisting torsior~l shear
s~resses and tension, but quite inefficient in
------- 1
::-8sisting ccmpressive st~8sses due to 8end1ng
of wing. To provide strength efficiency, span-
wise st~f:ening ·~~its commonly referr~d to as Fig. e
:lange stri~gers are attached to the inside of
the surface skin. ~o hold the skin surface to
airfoil Shape and to provide a ~edium for
transferring surface air pressures to the
cellular Jearn st~cture, chordwlse fo~ers and 0 0 0 0 0 ... ----
ribs are added. To transfer large concentrated
loads into the cellular beam structure, heavy Fig. t /--II
ri~s, cOIT~lonly referred to as bulkheadS, are ~ , i _____ ]
,L
used. " ,

Figs. A19.l and AlS.2 illustrate typical

structural ar-rangement s of wtng cross-sections
~or sU8sonic airc~ft. The surface Skin is
/
Fig. g
F'ig. a

c r--------~
"'----------
L c ,
Fig. b '!'"
,
,
-l F 1::;. h
~
'---.. ,
T
1 T!.,-
)" ----

Fig. c

Fig. A19.1
I F'1,~ •
\::] ::1 L

,
,
I r'
eJ
Concentrated ?ls.ncc8 ":ne of ,';in;:: 8ea:n.
Deshed line represents sec cndar-y s t.ruc tur-e . In Fig. A19.2 C:)rr:Y:lon :'YP8S of '.~·1 ng Seem Flange
aany cases this por t t on ~E :ab::-ic c over-ed, Ar~ange:ner:t.

A19.1

--.
j:.1.7
Al9,2 ANALYSIS OF WING STRUCTURES
~o withstand the h1~h surf~ce preSsures a~d :0
obtain sufficient strength ~uch ~hicker ~i~g
skins are usually necessary. Modern ~il:ing
machineS pe~it ta~ering of skin ~h~C:~1es2es.
To obtain ~ore rlanse ~ater~al intesral flange
~~its are ~achined on the thick s~in as 111~s­
trated in Fi~. k.
?lg. j
< ,
][ I[ j[ I[
Fig. k
<c
Light Weight Core
Fig. 1
Fig.A19.3 Wing Sections - super-soni c Aircraft
In a cantilever wing, the wing bending
moments decrease rapidly sparmise frem the
maximum values at the fuselage support points.
Thus thick skin construction must be rapidly
tapered to thin skin for wei~~t effiCiency, but
thinner skin decreases allowable compreSsive
stresses. To promote better effiCiency sand-
wich construction can be used in outer vortion
01' wing (Fig. 1). A light weight sandwich core
is glued to thin skin and thus the thin skin is
capable of reSisting high ccmpressive stresses
since the core prevents Sheet from buckling.
Al9.2 Some Factors Which Influence Wing Structural
Arrangements
(1) Light Weight: ~
The structural deSigner al'H.aYs strives for
the minimum weight which is practical 1'ram.a
production and cost standpoint. The higher the
ultimate allowable stresses, tee lighter the
str~ctures. The concentrated flange ~ype of
wing structures as illustrated Fi~. (a, b and c)
at Fig. A19.l per.nits high allowable compressive
flange stresses si~ce the fl~ge members are
stabilized by both web and covering sheet, thus
eliminating col~~ action, which permits design
stresses aporoaching the crippl~ng str~ss of the
flange ~emb~rs. Since the flang~ nembers are
few in number, the size or thickness required is
relatively large, thus giving a high cri~pling
stress. On the other hand, this type of desi~n
does not develop the effectiveneSS of the ~etal
covering on the compressive si~e, ~hlch WllSt be
balanced against the saving in the weight of thel
r range tnember-s ,
In the distributed type of flan~e arrange-
merrt (Fig. ,:"19.2), a cor-t t on c< t:-,e::::::,',',::":::g on
the c0mpression si~e is ~a~e effec:172 since ::
1s attaChed to closely spac8c stringers cr SQrT~­
ga t t ons . However , the r Iange s tr tnger s cc t.veen
cell nebs are Supported July at rib or j~lkr.ead
points ~~d thus suf:er CJl~~ action no~,al to
the cell coveri::g. ThiS factor reduces the
flan~e allcwable c8mpressive stresses since it
is not practical to space wing rl~s less ~han
12 to 18 times the flange stringe~ :e~th. Th~sJ
i~ ~here were no other ~ontr2l11n; =~ctcrs, or.e
could easily ~2ke calc~lations to dste~lr.e
which of the above would prov~ li~htest. In
general, if the tor-s i cna L ror-c es cr; t.ne .v tng
are sraaLl , thus recu ir-ma cn::'y a m;n c ove r tng ,
the concentrated flange ty;e of Nln3 st~~cture
should prOve the li~htest.
In general, the flange ~aterial should be
placed to sive the largest moment 0: inertia in
the Z direction, which ~eans ~n general that the
flan~e ~ater1al should se placed bet~een the 15
and 50 per cent 01' ':Iing chord r r'om the leading
edge.
The secondary or ~istrij~ting structure aft
of the st~~ctural box bea~ should be ~de as
light as pOSSible and thus in ;eneral the farther
forMara the rear closing web at the box be~, the
lighter the wing structure as a whole.
In the Layout of the main spanwise flange
members bends Or changes in direction should be
avoided as added weigh: is required in spl~cing
or in transverse stiff~ners whl~h are ~ecessary
~o change the direction of ~he load in ~he r:ange
members. If flan~e members must be s01iced, care
should be ~ken not to splice them in the region
of a max imum cross-section. Pur-thermcr'e , in
general, the smaller ~he ~umber of fit~ings, the
lighter ~he structure.
(2) Wing-Fuselage Attachment:
If the airplane 1s 0= the low ~nng or the
high wing type, the entire ~ing structure can
co~tinue in the way of the 2i:-plane body. How-
ever, in t~e ~id-wing tj~e or se~i-low wing
type, limitations may prevent extending the
entire wing through the :uselage, and some ot
the shear webs as well as the wing covering
nust be terminated at the side of the ~Qselage.
If a distributed flange type of cell structure
were used, the axial load in the ~lange string-
ers would ~ave to be transferred to the T.embers
extending through the r use Lage . '1'0 cr-evice fer
this transfer of large loads requires struct~l
weight &~d thus a concentrated fl~nge t:~e of
tox str~cture X:~ht prove the DRst type of
structur-e.
(3) Cut-ours in Wing Surface:
I The ideal arranse~ent where continUity of
structure is aa i nta mec oyer t ne en t i r-e sur-race
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
of the structural box 1s sel~om obtainec in ac-
t~al air?lane cesigr. due to cut-outs in the
Ning su~~ace ~or such items as retractable
landing gears, ~il compartments and bomb and
gun bays. r: the distributed ~~ange type of
box beam is used, they are interrupted at each
cut-out, which requires that means nust be
?rovided for drifting th~ flange loads around
the opening, an arrangement wntcti adds weight
because conservative overlapping assumptions
are usually made in the stress analysis. The
additional str~cture and riveting to provide
for the transfer of flange load around large
O?enings adds c8nslcerably to the production
cost.
?or landing gears as well as many other
installations, the «ing cut-outs are confined to
the Lower- surface, thus a structural arrangement
as illustrated in Fig. A19.4 is quite common.
The upper surface is of the distributed flange
type whereas the lower flange ~terial is con-
centrated at the two lower corners of the box.
In the normal flying conditions, the lower sur-
face is in tension and thus cell sheet covering
bet~een the cut-outs is equally effective in
bending if shear lag infl~ence is discounted.
For negative accelerated flying conditions, the
lower surface is in compression thus sheet cov-
ering between corner flanges would be ineffec-
tive in bending. However, since the load fac-
tors in" these flight conditions are approximate-
ly one half the normal flight load factors, this
tneffectiveness of the lower sheet in bending is
usually not c~itical.
Cut-outs 11~ewise destroy the continuity of
intermediate interior shear webs of such
sections as illustrated in Figs. (c and i)3 and
the shear load in these interrupted webs must
be transferred around the opening by special
~ulkh3ads on eac~ side of the cut-out, which
means extra weight.
Fig. A19.4
I~ ~ny cases, c~t-outs in the leading edge
are necessary due to ~ower ?lant installaticns,
landing gearxells, etc. Furthe~ore, in many
airplanes, it is desirable to ~ake the leading
edge ~crtion remova~le for inspection of the
many Ins ta Ha t Ions cvnf ch occupy thiS space in
the oor-t.tcn of the wtng near the fuselage. If
such is t~e case, ~hen an interior Neb should
be lecated near t~e ~ront Jf the wing section.
Ins~ectio~ 2QCrS ~8r the central oor~~cn of
box beam s t ruc t ure ar-e usua Ll.y Ioc a t ec on
Al9.3
the lower side of the Wi~g. They are usually
fastened to two spanwise stringers With screws
and the re~ovable ~anels are ef:ective i~ re-
s Let Ing bending and shear load. (See Fig.A19.5)
Removable panel for
Fig. Al9. 5 assembly and inspection
purposes.
Cutouts in the wing structural box destroys
the continuity of the torsional resistance of
the cell and thus special consideration must be
given to carrying torsional forces around the
cut-out. This special problem is discussed
later.
(4) Folding-Wings:
For certain airplanes, particularly Carrier
based Naval airplanes 3 it is necessary that pro-
viSion be made to fold the outer wing panels up-
ward. T~is dictates definite hinge pOints be-
~Neen the outer and center wing panels. If a
distributed flange ty~e of structure is used,
the flange forces must be gathered and trans-
ferred to the fitting points, thus a compromise
solution consisting of a ~ll number of span-
wise members is cammon practice.
(5) Wing Flutter Prevention:
With the high speeds now obtained by modern
airplanes, carefUl attention to wing flutter
prevention must be given in the structural lay-
out and design of the Wing. In general, the
critical flutter speed depends to a great ex-
tent on the torsional rigidity of the wing.
'Ahen the mass center of gravity moves aft of the
25 per cent of chord paint, the critical flutter
speed decreases, thus it is important to keep
weight of the wing fon'ard. At high speeds
where ncompressibility" effects became ~portant,
the torsional forces on :he wing are increased,
which necessitates extra skin thickness or a
larger cell. DeSigning :or flutter prevention
is a highly specialized problem.
(6) Ease and Cost of Production:
The airplane industry is a mass production
industry and therefore the structural layout of
the wing must take into account production
nethods. The general tendency at this tL~e is
to deSign the wing and bOdy structure, so that
s~b-assenblles of the various parts can be made,
which are finally brou~~t together to form the
~iQal assemblY af the wing pa~el. To make this
~rocess ef~!cient requ1res car9t~1 consideration
1~ the ieta:ls and layout ot the win~ structure.
?hotogra?h A1S.6 illustrates the sUb-asse~bly
~ ------ -,
<_C'--
A19.4 ANALYSIS OF WING STRUCTURES

.- .
.'

, ~ \
::1')1v .,\, / \

Designing To Facilitate Production.

Photographs by courtesy of North American Aviation Co.

 ANALYSIS AND DESIGN OF
break~down of ~he struct~ral parts of an air-
plane of a leading airplane company. Fabri-
cation and asse~bly 0: these ur.its ~ermits the
installation of ~uch eGui~ent jefcre assembly
of the units to tr.e final assembly.
A19.3 Wing Strength Requirements
~NO major strength require~ents must be
satisfied in the structural design of a wing.
They are: - (1) Ur.der the applied or limit
loads, no part of the structure mus~ ~
stressed beyond the yield stress of the mater-
ial. In general, the yield stress is that
stress which causes a permanent strain of
0.002 inches per inch. The terms applied or
l1mit refer to the same loadS, which are the
maximum loads that the airplane should en-
counter during i~s lifetime of operations.
(2) The structure sr~ll also carry Design Loads
without rupture or collapse or in other words
failure. The magnitude of the Design Loads
equals the Applied Loads tines a factor of
safety (F.S.). In general, the factor of
safety for aircraft is 1.5, thus the structure
must withstand 1.5 times the applied loads
without failure. In miSSiles, since no human
passengers are i~volved, the factor of safety
is less and appears at this ti~e to range
between 1.15 to 1.25.
Aircraft factors of safety are rather low
compared to other ~ields ot structural design,
chiefly because weight saving is so important
i~ obtaining a useful transportation vehicle
relative to useful load and performance.
Since safety is the ,aramount design require-
ment, the correctness of the theoret1cal design
XllSt be checked by extensive static and dynamic
tests co verify whether the structure will
carry the design loads without failure.
A19.4 Wing Stress Analysis Methods
In ~any of the previous chapters of this
book, internal forces were calculated for bo~h
statically determinate and statically inde-
ter.ninate structures. The internal loads in a
statically deter.ninate str~cture can be found
by ~~e use of the static equilibrium eq~tions
alone. The over-all structural arr~ngement of
members is necessary, but the size or shape of
no individual ~ember is required. In other
wordS, design consists of finding Inter~~l
loads and then supplying a member to carry this
load safely and efficiently. In a statically
indeterminate structure, addltlopzl equations
beyond :he static equilibrium equations are
necesscTJ' ~o find all the 1r.terr.al stresses.
The additional equations are supplied tr7.n a
consideration of structural distortions, which
~eans that the size a~d shape and ~i~d of
~aterlal for members of ~he structure ~ust 8e
known barer-a internal s tr-easea can be deter-
~ined. ~hls fact means a trial and er~0r method
FLIGHT VEHICLE STRUCTURES A19.5
is necessary. A~othe~ l~portant distinction is
that a statically dete~inate wing structure
nas just e ncugh member-s to produce stability
and if ene ~ember is removed or ~ails, the
ent1re st::-ucture will usually tail, whereas a
statically indeterminate structure has one or
more additional ~embers t~~n are necessary ~or
static stability and thus some members could
~ail Without causing the entire strJcture to
collapse. In other words, the structure has
a fail safe characteristic in that a redistri-
bution of internal stress can take place it
some members are over loaded. In general,
statically indeterminate structures can be
deSigned lighter ar.d with smaller overall de-
flections.
?"!:S'I'HODS OF STRESS ANALYSIS :;'OR STATICALLY
INDE"I'E:R..:."1INATS ":HNG STP.l'CTITRES
TWo general methods are commonly used,
namely,
(1) FleT~al beam theory with sL~pli~yir~
assumptions.
(2) SolVing for red~"1dant forces and stresses
by applying the principles of the elastic
theo~J by various methods such as virtual
work, strain ener~J, etc.
The second method is no doubt more accu-
rate since less assumptions are necessary. A
wing structure ccmposed of several cells and
many spanwlse stringers is a many degree re-
dundant structure. Before the development of
high speed computing ~chinery, so-called
rigorous ~ethods were not usable because the
computing requirements were ~Pcssible or
entirely ilnpract1cal. However, present day
computer facilities have changed the situation
and rigorous methods are now being ~ore and
more used !n aircra~t structural analysis. Art.
A7.9 and AS.IO in Chapters A7 and A8 present
~trix methods for finding deflections and
stresses to oe used with computer faCilities.
A19.5 Example Problem 1. 3-Flange - Single Cell
Wing.
Fig. AIS.7 shows a portion of a cant~lever
Wing. ~o prOvide torsional strength a single
closed cell (4) is fomed by the interior web
AB and ~he ~etal skin cover for~ard of this
web. Thin sheet 1s relatively weak in resist-
ir~ compressive stresses thus 3 flange stri~gers
A, B and C are added to develop efficient ber-d-
ing resistance. The structure to the rear of
suar AB is referred to as secondary str~cture
and 20nsists of thin metal or tabric covering
attached to chordwise wing ribs. The air load
on this Dortion ~s therefore carr~ed [orNard
by the r~bs to the single cell beam.
A wing 1s subjected to many flight condi-
A19.6 ANALYSIS OF WING STRUCTURES

tions. The engineers who cal:ulate the ~ppl:ed

loads on the wing usually refer the resulting 1-7"-=22"
shears and moments to a set o~ conver-ient x, Y -1/2
and z axes. Fig. A19.7 shows the location of
these reference axes. The job o~ the stress
engineer 1s to ~rovide structure to resist
these loads safely and efficiently. ~he ge~eral
VZ"1/IOO~MY"
'5

~60.500
procedure is to find the stresses or loads 1~
all parts of the cell cross-section at several Vr700 i Mx"'l,lOO,OOO
stations along the spanwise direction and tram ~10"""""
these loads or stresses proportion the required Fig. A19. 8
I Mz",aoooo
areas, thicknesses and shapes.
SOL~ION.
In this example, the internal leads will
be calculated for only one section, n&~ely, ASSUI1?T'IONS: - It will be assumed tnat the
that at Station 240. It will be ~ssumed that 3 flange stringers A, Band C develop the 3r.tire
the design critical loads from the critical resistance to the bending ~oments about ~he Z
flight condition are as follows. and X axes. FJr skin ~~cer compression t~~s
assumption 1s nearly correct since the skin will
Mx = 1,100,000 in. lb. Vz = 11,500 lb. buckle under relatively low stress. Since sheet
can take tensile stresses, this assumption is
l1z = 80,000 in. lb. Vx = 700 lb. conservative. However, since the thin sheet
My = 460,500 in. I'D. cover must resist the shear stresses we will
~ake this conservative a3s~~.~tion. ~he main
Fig. Al9.8 shows these shears and ~oments advantage of this approximate assumption is
referred to the reference axes with origin at that it ~kes the structure statical:y deter-
point (0). Moments are represented by vectors minate.
with double arrow head. The sense at the
moment tallows the right hand rule. Fig. Alg.a shows ~~e wing c~t at Station
240. The wlliL~own forces are the three axial
7.27" loads in the stringers A, Band C and :he three
t.J
3. ii' ,2. 33 shear flows Qab' ~c and qbc on the three sheet
,t
-=::::±':::J)
r-..L~" panels, ~king a total of 6 unknowns and since
I T there are 6 static equilibrium equations avail-
STA.O - 32"-"""t able for a space structure, the struc~ure is
statically determlr~te.

Since ~he size and shape of flange ~e~bers

A, B and C are unknown, we guess their centroid
locations as indicated by the dots in Fig.A19.8.
The axial load in each at the 3 stringers has
been replaced by its x, y and z components as
shown on the figure. The external applied loads
-a:re given at the reference origin (0) as shown
in Fig. A19.S.

~e now apply the equations of 8Guil~bri~~

to tind the 6 '~owns .
•
;< To find Cy take ~coents about Z axis t~~ough

,.."' and 8,
PC!.:J.ts A

I
l:l1z(ab) - - 22 Cy + 60000 = 0, whence
Cy = 3636 lb.

The result CJmes out with a .plus sign thu~

STA.240
indicating that the assumed sense of tension

r= was correct.

X ax:s
-
To find By take moments about ~
'thr-u (A).

lMX(A) = - 11 By + 1,lCO,OOO + 3636 x

Fig. A19.7
0.125 =0
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A19 7
-9065
whence, By = 100,043 lb. tension as whence qbc = """'6'65:: - 13.66 Ib./in. and
assumed. having the sense as aSSQ~ed.
To find Ay take ZFy =0 In the above moment equation the moment of
the shear flow qbc about point (a ') equals qbc
ZFy =- 100,043 - 3636 + Ay = 0, times twice the area of the cellar 665.
whence, Ay =
103679 lb. and com-
To tind Qac take ZFx =0
pression as assumed.
Since the d~rection of the 3 stringers is
ZFx =- 2879 + 2584 + 316 - 22 Qac - 22 X
13.66 - 700 = 0
known, we can find the X and Z components of
the stringer loads by simple geometry. whence Qac = 35.45 lb./in. with sense as
assumed.
The r, x, and z length components of the
three stringers from the dimensions given in To tiM 'lab take ZFz =0
Fig. A19.7 are found to be,
ZFz =- 1296 - 12.'10 + 45 + 35.45 x 0.5 -
11.5 x 13.66 + 11500 - 11 'lab =0
Member s x z

A 240 6.2 3
whence, Qab = 805 lb./in.

B 240 6.2 3 The loads on the stringers and sheet panels

c 240
The force components are therefore: -
Az = 103679 x 3/240 = 1296 lb.
6.2/240 = 2579 lb.
Ax = 103679 x
Bz = 100043 x 3/240 = 1250 lb.
Ex = 100043 x 6.2/240 = 2584 lb.
Cz = 3636 x 3/240 = 45 lb.
Cx • 3636 x 20.93/240 =316
Fig. A19.9 shows these forces applied in
the plane of the cross-section at Station 240,
together with the unknown shear flows and the
external forces acting in the plane of the
cross-sectlon.
have now been determined. The axial load in the
20.93 3 stringers is practically equal to their y com-
ponent since axial load equals their y torce
component div!ded by the cosine of a small
angle. The stress engineer would tind s1m1lar
stresses at a number ot stations along the spa.~.
These 6 stresses are generally reterred to as
primary stresses. Usually in most structures
there are secondary stress effects which must
be considered betore tinal member sizes can be
determined. For example, internal webs at a
box tYJ)e beam are designed usually as semi-
tenSion tield beams. Tension tield beam theory
shows that the flange members are Subjected to
lb.
additIonal stresses besides the primary stresses
as tound above. The subject at secondary
stresses and the strength deSign ot members
and their connections to carrJ given stress
loads is taken up in detail in Volume II.
A19.6 E::o::unple Problem 2. Metal Covered Wing With Single
External Brace Strut.

1
1296
"
---22'~'....".0:-'111-
fo!
, In Chapter AZ, the stress analYSiS ot an
2Jil 3181 externally braced fabric covered monopl~~e wing
267,-+ A 'lac 2161 C 6 was considered. To provide sufficient torsion-
45 / al strength and rigidity, two external brace
struts were necessary. However, 1t a wing 15
%c metal covered, a single external brace strut
can be used, since ~he closed cellar cells
for,ned by the ~etal sheet covering and the in-
Fig. A19 9 ternal webs provide the torsional resistance
and the wing can be des fgned as a S1.mply sup-
ported beam with cantilever overhang. An
To find qbc take moments about point (a') excellent example of this type ot wing struc-
ture is the Cessna aircraft Model 180 as shown
:Mar = + 2679 X .5 - 2584 X 11.5 - 316 in the photograph. An excellent airplane
x 0.375 - 45 x 22 + 700 x 12 relative to pertormance, ease of ~nutacture
- 11500 X 39 + 460,500 - coo and maintenance.
(665) =0
To introduce the student to the general
approach Of stress ar41YZing such a wing struc-
A19. B ANALYSIS OF WING STRUCTURES

ture , a Liznl t ed -nscuss i c- -"i':h a few ca LcuLa-. Fig. A19.1D and AIG.Il shows the wl~g
~ions ~111 be ~rese~ted. d1mensions ane genera: str~c:~~3.1 :ayout of a
monoplane wing wl~h o~e external brace s:rlt.
~ The wing panel is attached ':0 ~~sela~e by
:§ Single ;in fl':tlngs at points A and 8 with ~1n
axes parallel to X axis. The ~ating 1~~8 ot
4fii\ the :lttlngs at paint A are ~ace snug :1t but

-- those at B with sane gap, thus ~ag reaction

of wing loads on fuselage is resisted entirely
at fitting A. Since the fittings at A and B
cannot resist moments about x axiS, it 1s
necessary to add &~ external brace strut DC to
make structure stable. The p&~el str~c:ure
consists of a main spar ACE anG a rear spar 3F.
Tte entire ?anel is covered with metal skin
~-;.;~ .'~>;~'-~ forNard of the rear spar.
Cessna Aircrart Model 180
Metal Covered Wing with A SL~pllfled air load has been assumed as
One External Strut shown in Fig. A19.12, ~~mely, a unifc~ load
w = 30.27 Ib./in. of span acti~g at the 30 ~er
cent of chord point. ~ben this resultant load
is resolved into z ~nd x :o~~or-ents the results
-t-tr-
27"
a- are Wz = 30 lb./in. and Wx = ~ lo./in. as Shown
....L
15"
la- - Front Spar
-'C-- in Fig. A19.l2.
T The general physical ac:ion of the wing
l--- BO" 90" C structure in carry~ng these al~ loads can Je
-.---- ~ considered as 3 rather distinct actter-s,

t' L.
namely, (1) The front spar ACE resists the
bendl~g moments and shea~s due to load w . ,
z
Strut C. y (2) The skin and webs of the two-cell tube
resists all mrr~ents about y axis or broad:y
speaking tors~onal moments, (3) the entire ~anel
cross-sect ton resists the bending moment and
Fig. A19.l0 flexural she£r due to drag load wx, with the
top and bottJm skin acting as webs and the t~o
spars as th~ flanges of this box beam.

General Calculations; -
z The unknown external reactions (see Fig.
AI9.ll) are Ay, Az, Ax, By, Bz and DC., or a
total of 6. Since 5 static equations of
equilibrium are available, the reactions are
statically deter.r.lnate. Reaction DC is also
the load in brace strut CC.
30.27 To :lnd reaction DC take moments about x

"/m·4t~m.
axis through pOints A, 3
~_J ZMx(AB) = (- 170 x 30 x 170/2) + DC(60/99.4)60
-til- =0
Fig. A19.l2 whence, DC = 8979 lb. (T::e Sign cernes out
plus so the sense assumed in Fig. A:9.11 was
correct.) The load in the strut 1s therefore
8979 lb. ~ensicn.

To find B., take moments about a z. exts

thrcugh point rAJ.

~~(A) =- (4 x 170 x 170/2) + 27 By =0

Fig. A19.11 whence, By = 2141 lb. ac t Ing wt tb sense
aesumed ,
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A19.9

To find Ay take ZFy =0 points A and B. This couple force equals the
moment divided by distance between spars or
ZFy =2141 - 8979 (80/99.4) + Ay =0 15300/27 = 566.7 lb.

whence Ay = 5085 lb. Front spar (ACF) loads due to Wz =30 lb/in:
To tind Bz take ~oments about y axis w'" 30 lb./ln.
through (A). t t t t t lltttt
=+ (30 x 170 x 3) - 27 Bz =0 ===c y-,-"...c:::= __
lliy(A)
7226/ I
whence 8 z = 567 lb. acting as assumed. Cz.:t5329
~~. • 59
To tind Az take ZFz = 0 80 Flg. A19.14

ZFz =170
+ Az
x 30 - 567 - (8979 x 59/99.4)
=0
80"
+- 90" --i
whence Az = 796 lb. actlr~ as assumed. Fig. A19.l4 shows free body of front spar
ACE. To rind strut load DC take moments about
The tittings at A and 8 should be designed (A) •
to take the reactions at these pOints as found
above. The external strut DC and its end lMA =(- 30 x 170 x 170/2) + 60 (DC x 80/99.4)
fittings ~t carry the tension load ot 8979 lb. = 0

The next step is to find the stresses and Whence DC =

8979 lb. Tension which checkS
loads on the structural parts of the wtng panel. value previously found.
Since the spar ACE ~~ resist the bending The y and z components of the strut re-
~oments about x axis the airloads in Fig. action at C will then be,
A19.12 are ~oved to the spar centerline as
shown in Fig. A19.13. Cy = 8919 (80/99.4) = 7226 lb.

30#/in. Cz = 8979 (59/99.4) = 5329 lb.

.,....,=:::-J~3.30.90". These values are indicated on Fig. Al9.l4.
<:-Il!"/ln~
Flg. A19.13
To find Ay take ZFy =0
ZFy : - 7226 + Ay =0, hence Ay =7226 lb.
The torsional moment or 90 in. lb. per
inch ot span is resisted by the cellular tube In finding the reaction Ay previouslY, the value
made up of two cells (1) and (2). In many was 5085 lb. The difference is due to the drag
designs the leading edge cell Is neglected in bending moment which tends to put a tension load
resistir~ the torsional moments due to many on front spar and compression on the rear spar.
cutouts, etc., thus cell one could be assumed wx"'4*/in.
to provide the entire torsional shear resist- lffftlfttffff
ance and the shear flow tor this case would be BY-B Rear Spar
q = Mv/2A where My equals the torsional moment
at a ~iven panel section and A the enclosed Front S r
area of cell (1). It both cells were con- C
sidered ef~ective then the sheet thicknesS is
necessary before solution for shear flow can Fig. A19. 15
be computed. Refer to Chapter A6 for computing
torsional shear flow in ~ultiple cell tubes. Fig. Al9.15 shows the air drag load of 4
lb./in. The bending moment on panel at a
The maxfmum torsional moment would be at distance y tram the wing tip equals Wx (y) (y/2)
the fuselage end of the wing panel and its = 4y ·/2 =
2y·. The axial load Py in e I trier
~gnitude would be 170 x 90 = 15300 in. lb. spar at any distance y from tip thus equals
Since the top and bottom skin is not attached bending moment divided by arm ot 27" or Py =
to fuselage, this torsional moment must be Zy'iJ/ 27 = .07 4y a. The axial load at points A
thrown off on a rib at the end ot the panel and B thus equal .074 x 170· = 2141 lb. (tens ton
and this rib in turn transfers this moment in in front spar and compreSSion in rear spar).
terms or a couple reaction on the spars at Thus each spar is subjected to an axial load
A19.l0 ANALYSIS OF WING STRUCTURES
increasing from zero at tip to 2141 lb. at
fuselage attachment ,oints and varying as y~.
At point C on from spar the axial tension load
would be .074 (90') = 600 lb.
The design of the tront spar between
points C and E would be nothir~ more tillL~ a
cantilever beam subjected to a bendir~ torces
plus an axial tensile load plus a torsional
shear flow. The design of the spar between
points C and A is far more complicated since
we have appreciable secondary bending moments
to determine, which must be added to the pri-
~ry bending moments. Fig. Al9.lsa shows a
free body or spar portion AC.
w::; 30
fltltt~
~Mc::;30x90x45
5085-<:- -. - - - -- '
1 Fig. Al9.15a
j The stress analYSiS of this wing would consist
The lateral load of 30 lb./in. ~ends the
beam upward, thus the axial loads at A and C
will have a moment ar.n due to beam deflection
which moments a&e referred to as secondary
moments. To find deflections the beam moment
of inertia must be known, thus the design of
this beam portion would fall in the trial and
error procedure. Articles A5.23 to 28 of
Chapter AS explain and illustrate solution of
problems involving beam-column action and such
a procedure Would have to be used in actually
deSigning this beam portion.
The rear spar SF receives two load systems,
namely a varying axial load of zero at F to 2141
lb. at B and the web of this spar receives a
shear load rrcr.m the torsional moment. The
rear spar is not SUbjected to bending momen~s.
In Fig. A9.l0 the secondary structure
consisting or chordwise ribs and spanwise light
stringers riveted to skin are not shown. This
secondary structure is necessary to hold wing
contour Shape and transfer air pressures to
the box structure. This secondary structure
is discussed in Chapter A2l. The broad SUbject
at designing a member- or structure to withstand
stresses sarely and effiCiently 1s considered
in detail 1n later chapters.
A19.7 Single Spar. Cantilever Wing -
Metal Covered
A Single spar cantilever wing with metal
covering is often used particularly in light
commercial or private pilot aircraft.
Suppose in the single spar externally
braced wing of Fig. A19.Il, that the external
brace strut DC was removed. Obviously the w1~g
would be unstable as it would rotate about hinge
fittings at ~oints A and S. To make the
structure stable the Single pin fitting at (A)
would ~3ve t8 be replaced by two fltti~gs, one
on the upper flange ane the other an :he lower
flange in order to be able to resist a ~x
moment. Fig. A19.l6 shaNs this modification.
The fitting at B could remain as before, a single
pin fitting.
Fig. Al9.16
of the spar AE reslstl~g all the Mx mcments and
the Vz shears and acting as a cantilever beam.
The torsional moment about a y axis coinciclng
With spar AE would be resisted by shear stresses
in the cellula~ tubes tanned by the skin ane the
spar webs. The d~g bending and shear fo~ces
would be resisted by the bean whcse flar~es are
the tront and rear spars ant the web being the
top and bottcrm skin.
A19.8 Stress Analysis of Thin Skin - Multiple Stringer
Cantilever Wing. Introduction and Assumptions.
The most common ~ype of wir~ con5t~uctlon
is the ~ultiple stringer type as illustrated by
the six illustrative cross-sections in Fig.
A19.2. A structure with ~any stringers and
sheet panels is statically indeter.ninate to
many degrees with respect to inte~al stresses.
Fortunately, structural tests of complete wing
structures show that the Simple beam theory
gives stresses which check fairly well with
measured stresses it the wing span is several
times the wing chord, that sweep back is minor
and wing is free of ~jor cutouts and discon-
tinUities. Thus it is common procedure to
analyze and design a wing overall by the beam
theorJ and then investigate those portions of
the wing where the beam theory ~y be in err~r
by USing more rigorous analysiS methods such as
these explained and illustrated in Art. AB.IO
of Chapter AB.
ASSUMPTIO~S - BSAM ~~£ORY
In this crapter the wing bending and shear
stresses will ~e calculated USing :he un-
symmetrical beam theory. The two ~in assump-
tions in this theory are: -
(1) Transverse sections of the 8eam originally
plane before ben=ing remal~ plane a~ter be~di~g
of beam, This assumption means that Longftudf -.
nal strain varies cirectly as the jlstance from
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
the neutral axis or strain yariation is linear.
(2) The longitudinal stress distribution is
directly proportional to strain and therefore
trom assumption (1) is also linear. This
assumption actually ~eans that each longitudinal
element acts as it it were separate tram every
other element ~~d that Hook's law holdS,
namely, that the stress-strain curve 15 linear.
Assumption (1) neglects strain due to shear
stresses in skin, which influence is commonly
referred to as shear lag effect. Shear lag
effects are usually not ilnportant except near
major cut-outs or other major discontinuities
and also locations where large concentrated
external forces are applied.
Assumption (2) is usually not correct if
elastic and inelastic bUCkling of skin and
stringers occur before failure of win~. In
applying the beam theory to practical wings,
the error at this assumption is corrected by
use ot a so-called effective section which is
discussed later.
A19.9 Physical Action of Wing Section in Resisting External
Bending Forces from Zero to Failing Load.
Fig. Al9.l7 shows a common type of wing
cross-section structural arrangement generally
referred to as the distributed flange type.
skin
a b
'-~vr--~
web c(stringers) web
Fig. A19.17
The corner members (a) and (b) are conSiderably
larger in area than the stringers (c). The
skin is relatively thin. Now assume the wing
is subjected to gradually increasing bending
forces which place the upper portion of this
Wing section in compression and the bottom
portion in tension. Under small loading the
compressive stresses in the top surface will
be small and the stress will be directly ?ro-
portional to s~rain and the beam formula
0c = Mxz/l x will apply and the moment of
inertia Ix will include all of the cross-
section material. As the external load is in-
creased the compressive stresses on the thin
sheets starts to buckle the sheet panels and
further resistance decreases rapidly as further
strain continues, or in other words, stress is
not directly proportional to strain when sheet
buckles. Buckling of the skin panels however
does not cause the beam to fail as the stringers
and corner members are lowly stressed compared
to their failing stress. ~he stringers (c)
are only supported transversely at wing rib
points and thus the stringers ac~ as columus and
fail by elastic or inelastic bending. The
A19.11
corner flange members (a) and (b) are stabilized
in two directions by the skin and webs and
usually fail by local crippling.
Now continuing the loading of the wing
after the skin has buckled, the stringers and
corner members will continue to take additional
compreSSive stress. Since the ultimate strength
of the stringers is less than that ot the cor-
ner members, the stringers (c) will start to
bend elastically or inelastically and will take
practically no ~Jrther stress as additional
strain takes place. The corner members still
have considerable additional strength and thus
additional external loading can be applied until
finally the ultimate strength ot the corner
members is reached and then complete failure of
the top portion of the beam section takes place.
Therefore, the true stress - strain relationship
does not follow Hook's Law when such a structure
is loaded to failure.
In the above discussion the stringers (c)
were considered to hold their ultimate buckling
load during considerable additional axial
strain. This can be verified experimentally
by testing practical columns. Practical col-
umns are not perfect relative to straightness,
uniformity of material, etc. Fig. Al9.l8
shows the load versus lateral deflection ot
column midpoint as a column is loaded to failure
and fails by elastic bending. Fig. Al9.l9 shows
similar results when the failure Is inelastic
bending.
z
Fig. A19.18 Fig. A19.19
La
6 '" Central Deflection
B '" Central Deflection
The test results show that a compreSSion
member which fails in bending, normally con-
tinues to carry approximately the ~aximum load
under considerable additional axial defo~tion.
Thus in the beam section of Fig. Al9.l? when
the stringers (c) reach their ulti~ate load,
failure of the beam does not tallow since cor-
ner ~embers (a) and (b) still have remaining
strength.
A19.10 Ultimate Strength Design Requirement
The strength design reqUirements are: -
(1 ) Under the applied or limit loads, no
structural member shaIIibe stressed above
the material yIeld point, or in other
words, there must be no permanent de-
:crmation or deflection of any ?art of the
structure.
 A19.12 ANALYSIS OF WING STRUCTURES

(2 ) Under the ~ loads which equal the

applied loads ti~es a factor of safety,
no failure of the structure shall occur.
The usual factor ot safety for conventior41
aircraft is 1.5, or the structure must , . , , I
carry loads 50 percent greater than the , I ,
I I, •,
I
I
[
applied loads wi~hout failure. The higher
-l. f" .- -!W, i';'.,- r,L
I

~I';' kf-.
the stress at failure ot a member the less -ow,
material required and therefore the less I -
structural weight. The stress engineer 1. it- '\
thus tries to deSign members which fail in
the inelastic zone. Fig. A19.21

\ stress equal to the stringer stress. l~ Fig.

The bending stress beam ~ormula crc = MclI A19.2l ~N is the et:ective sheet width ~o go
does not take care of this non-linear stress- with each stri~er. The total stress on the
strain action and thus some modification of the effective Widths carrying a uniform stress eq~al
moment or inertia of the beam cross-section is to the stringer stress equals the total load on
necessary it the ultimate strength of a wing the sheet panels carrying the actual varying
section is to be computed fairly accurately. stress distribution. The equation for effective
The stress engineer usually solves this problem width is usually written in the ~or.n,
by using a modified cross-section, usually
referred to as the effective cross-section. • 2w = kt (E!crstl.•
A19.11 Effective Section at Failing Load. a widely used value for K = 1.9, whence
In order to use the beam formula which
assumes linear stress-strain relation, 2W = 1.90 t (E!crst)-• - - - - - - - - - (1)
corrections to take care of skin ~uckling and
stringer buckling must be fOr.TIulated. The 0st =stress in stringer
effectiveness of the skin panels will be con-
sidered first. Therefore if we know the stress in the
stringer we can find the width of sheet to use
When a compreSSive load 1s applied to a With the s~ringer to obtain an effective section
sheet stringer cambir~tion, the thin sheet to take care ot the sheet buckling influence.
buckles at rather low stress. For further
loading the compressive stress various aver Effective Factor for Buckler Stringe~s.

the panel width as illustrated in Fig. Al9.20.

The stress in the sheet at the stringer at~ach­ Consider the beam section in Fig. Al9.l7.
ment line is the same as in the stringer since If we take a stringer (c) and attach a piece of
it cannot buckle and therefore suffers the same sheet to it equal to ZW, the effective width
strain as the stringer. Be~Neen the stringers and test it in compresSion and brace it in a
the sheet stress decreases as shown by the plane parallel to the Sheet, the resulting test
dashed line in Fig. Al9.20. This variable stress versus strain Shortening curve (c) of
stress condition is difficult to handle so the Fig. Al9.22 will result. ~he length of the test
stress engineer makes a convenient substitution specimen would equal the rib spaci~g in the
by replacing the act~l sheet with its variable wing. The corner members (a) or (b) in Fig.
stress by a width ot sheet carrying a uniform A19.l7 being stabilized in two directions will
tail by local crippling, thus it a short Piece
Stress Distribution on
Stiffener
tn, ;;,>,
7"'-."'>J
f stress Distribution on
Shee_~t
VI"
_
r of this member is tested to failure 1n com-
preSSion the test curve (A) in Fig. Al9.22 is
IK III ~ obtained. C~le (C) shows that the strir.ger
/:1' I \ 'I I',
'I"
II
I; , : '- holds apprQx~tely its maximum load ror a
1 /
,' I , I ',
1 1\
', I!• I ''
, 1\, - 'Ii 'I I
-,__i.Ost
.L considerable strain periOd. C~ve (A) shows
'r ' , ,
\-r' / 1 - ,
I
, ,
, -'lac that for the same unit strain member (a) can
take considerable higher stress. It we take
H· j,!-I-H ~~ Ii %'-1 1( a unit strain ot .006, the poir.t ?t which the
~imum stress of 47000 is obtained in member
I~ 1--»,
~
, - IiI (a) (see paint (3) on Curve A) then the stress
±
,
'T at the same strain for ~ember (C) will be 38000
l-r . I~ (see pomt (2) on Curve C) .

Fig. A19.20

., The buckling of sheets 14 taken up In detail in

Volume II.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
Fig. Al9.22
For stress analysis procedure using the
beam formula, we assume a linear stress vari-
ation fram zero to 47000 psi. Since stringers
(c) can only take 38000 pSi at the same strain,
or in other words, stringer (c) is less effect-
ive than members (a) and (b). The effective-
ness factor for stringer (c) ~quals its ulti-
mate strength Givided by the ultimate strength
of member (a) or 38000/47000 • • 808.
Al9.12 Example PrOblem
The wing section in Fig. AI9.23 is sub-
jected to a design bending moment about the x
axis of 500,000 in.lb., acting in· a direction
to put the upper portion in compression. The
problem is to determine the margin of safety
for this design bending moment. The material
is 2424 aluminum alloy.
f- 6" ___ 6" - - ' - - 6"
I L04 I I
6" -----l
~ 7, 8 and 9. The areas of the lower corner
(1)
T
10"
- 10--.05
\ (10)
....L.
L .035
Fig. A19.23
SOLUTION
The beam !o~ula :cr bending stress at any
point 15 crb : Mxz/I x. To solve this equation
we must have the effective moment of i~ertia
of the beam cross-section. The bottom surface
being in tension under the given design ~er.ding
moment is entirely effective, however the top
surface has a variable effectiveness since :he
skin, stringers and corner tlange ~embers ha7e
different ult~te :ailing stresses.
From equation (1) the effeotive width of
Al9.13
sheet to use with each rivet line depends on
the stress in the stringer to which it is
attached. The failing stress for the stringer
will be used, which means that the failing
stress of the stringers and corner flanges must
be- ~~own betore the effective width can be
found. For this example the zee stringers have
been selected of such a size as to give an
ultimate column failing stress of 38000 psi,
and the corner flange members have been ~de of
such size and shape as to give a failing stress
of 47000 psi. These failing stresses can be
computed by theory and methods as given in
Volume II. The sizes have purposely been
selected to given strengths represented by the
test curves of Fig. A19.22.
The effective width with each rivet line
from equation (1) would be,
For zee stringer 2W
, = 1.90 x .04 (10.5 x
108/38000)a = 1.25 inches
Thus the area of effective skin: 1.25 x
.04 = .05 in. a • The area of the zee stringer
is 0.135 sq. in. which added to the effective
skin area gives 0.185 sq. in. which value is
entered in Column (2) Of Table I opposite zee
stringers numbered 2, 3 and 4 in Fig. A19.23.
The same procedure is done for the c orner- mem-
bers land 5 With the resulting effective areas
as given in Table I. On the bottom side which
is in tension all material is effeotive. The
skin width equal to one-half the distance to
adjaCent stringers is assumed to act with each
stringer. Taking the area of the angle section
as 0.11 and adding skin area equal to 6 x .035
= .21 or a total of 0.32 sq. in. which value is
shown in Column 2 of Table I opposite stringers
members plus bottom skin and web skin would
come out as recorded in the Table.
The next step is to correct tor stringer
effectiveness in compression. The failing
stress fer the zee stringer was given as 38000·
and fer the corner member-s 47000 psi. The
effectiveness factor for the zee stringer thus
equals 37000/47000 = .808. This factor is re-
corded in Col~ 3 of Table I. for the corner
members 1 and 5 and all the ~ension ~embers the
factor is of course unity. The balance of
Table I gives the calculation or the effective
moment of inertia ZAz~ about the x r.eutral axis.
T~e compreSSive stress intensity at the
centroid of the zee stringers ~hus equals,
crb = Mxz/I x : 500,000 x 5.57/59.30 = 46600 pSi.
The true stress equals the effectiveness
factor times this stress: .808 x 46600 =
37400 psi. The failing stress equals 38000
hence margin at safety = (38000/37400) - 1 =
.01 or one percent.
:: n
~ "
c:
A 19 14 ANALYSIS OF WING STRUCTURES

TABLE Al9.1 The stress analysis of this wing would

1 2 3 4 5 6 7 I 6 show the resulting bend:;'ng and shear stresses
for a number 0: spanwt se stations for the cr-t t i
Stringer cal c.esign load condi t i ons , In this example
Stringer jEllect-
Slringe
Area
EUect- <Ve
sorut i cn the bending longit"J.d.1nal stresses 'Nill
Plus z' Az' z::;:z'-l. Az' be deter:r.lned on cross-sections at two stat i ons
Number Effective iveness Area
Factor IA) namely, stations 20 and 47.5, and the shear
SIdn
stresses w111 be deter:nlned for station 20. In
1 0.37 1.0 0.370 4.50 1.664 5.47 11.05 this example problem, the lead!.ng edge cell wil 1
2 O. 185 0.808 0.149 4.60 0.685 5.57 I 4.62 be cons leered tne r r ec t Ive as well as any atr-uc-
3 0.185 0.80S" 0.149 4.60 0.685 5.57 ! 4.62
ture to rear of rear beam, hence s-cruct;.u-e Is a
one cell beam with ~ultiple stringers. A second
4 0.lS5 0.808 O. 149 4.601 0.685 5.57 4.62 s o.Lutt on including t he leading edge col., to :0r-a
5 0.370 1.0 0.370 4.50 1.664 5.47 11.05 a two cell beam w111 also be presented.
6 0.417 1.0 0.417 -4.60 -1.920 -3.63 5.50
7 0.320 1.0 0.320 -4.63 -1.480 ~3. 66 4.28 A.i'{:'.LYSIS FOR BEND IN3- Lm;-G I ':'!J:)1:;).L STPLSSES
6 O. 320 1.0 0.320 -4.63 -1.480 -3. 66 4.28
9 0.320 1.0 0.320 -4.63 -1.480 -3.66 4.28 LongI tuc inaj stresses (tension or coro-
pression) are produced by ext.erna; rcrces nor:r:.a 1
10 0.417 1.0 0.417 -4.63 -1.920 -3.63 I 5.50
to the cross-section and by bending ~oments
Z 2.981 -2.897 Ix::;: 59.80 about x and z axes. The stress equations are:
z' ::;: distance from It x axis to ceatrotd of stringer area
"n = PIA - - - - - - - - - - - - - - - (2 )
z ::;::Z Az' /:ZA::;: -2.897/2.981", ·.97 in.
where "n = longitudinal stress .
on corner members (1 ) and (5 ) the cam-
press ive stress =
500,000 x 5.47/59.80 =
P = treid
external
of
load acting ":hraugh cen-
effective w~ng cross-sectle n
45000 ps t , With the failing stress being
47000 the margin of safety Is 2 percent. A = effective area of cross-section.
Now suppose we would have omitted con- For any given flange ~ember 'Ni~h ar-ea (a)
sideration of the stringer effectiveness factor the load Pn on the member would be.
and omitted column (3 ) of Table J. Carrying
through the calculations at Table I, the value
of Z would be -0.76" and Ix would be 63.08.
Pn = ana ---- - - - - - - - - - - - (3 )
The stress intensity on the zee stringers T~e stresses due to bending acments are
would then be 500000 x 5.36/63.08 42500 ps t , = rr-cn Chapter A13. Art. A13.5: -
Since the failing stress for stringer is 38000
the aarg m of safety would be (38000/42500)
= a negative 10.5 percent. The previous result
1 - "t = -(K Mz-Kl. Mx)x - (K. T1x -x, Mz) z- (4 )
3

"b = bending stress with plus being ten-

was a plus 1 percent, thus a difference of 11.5 where
percent in the results. sion
The purpose of this Simple example problem
was to emphasize to the student that failure of
K, = I xz/( Ix Iz - I xz ,..)
real aircraft stiffened skin structures occurs K. = Iz/(I x I z - Jxz 2)
u.~der non-linear stress-strain conditions and
the elastic t heory must be modified d fairly .. J(, = I x/( Ix Iz - Ixz'" )
accurate ea t imatea of the failing strength of
a compOSite s'tz-uc tur-e is to be obtained. The normal component or ~he axial load in
A19.13 Bending and Shear stress Analysis of Tapered _ a rLange member equals "b a) where (a) is the
Multiple Stringer Cantilever Wing. Unsymmetrical area of the flange member. Since the angle be-
Beam Method. tween the normal to the beam section and the
In general cantilever Nings are taperec. in centr-oiuaj axis of a stringer is generally qUit e
both depth and planform. Fig. A19.24 i uue- small. the c t rrcr-ence between the cosine o~ a
trates a typical structural layout of the outer small angle and unity is negllg~ble and thus
wing panel for a sma.Ll. airplane. The structure the nonal component can be considered as the

spanwt sa ~
"
.
consists of a front and rear beam (spar) with
o't...r .l.ng~~
Tayering in cross-sectiorzl
~
C'':'''''c!

material is obtal~ed
o. . .'
bet .....e'-'•• the ... no hceams , I
3efcre equations 2, 3 and 4 can je solved.
axial load in the stringer.

by decreasing size of ~embers by cuttir~ off the effective cross-sect~cn area ~us~ be known
portions of the spanw1se stringers and corr.er as well as :he ~oments of ine::-tia and ~roduct
flanges and decreasing the skin and web thick- of ~nertla about x and z c9~troidal axes.
ness.
 ftJ

ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A19.15

WING BODY SECTION PLAN VIEW LOOKING INBOARD FROM TIP

I 2. 3 4
6
7

--r-

/4 "ii2.
"

/3 /2
" 10 q 8

-
UPPER SURFACE STRINGER AND SKIN ARRANGEMENT
<47·5" '00 70
:ITA FIO'-----"F----'T'---'T'-----''t'---'"IL--q.''!.-2.1
'.0 'PO 120

f-- I
I I I I

::lJ1
II
, -\1 Skin thickness
)-."4 i .03.2,. \ ,0z.5 between beams

UPPER SURFACE STRINGER AND SKIN ARRANGEMENT

'0 ..
, uo

Fig. A19. 24 Structural Layout of Tapered Cantilever Wing

A19.16 ANALYSIS OF WING STRUCTURES

1/ -\
rOt: '""-'-I -to- 1. "---i
-' ~
I T Upper Front Flange .050
, 1:
.b4
r--- 2 1/4----..-;

area of angles", . 508

area of skin "'. 090
Upper Stringer.
No. 2 & 6 (t "'.064) T Upper Rear Flange

area stringer'" . 257 1 area of angles", . 508

area of web "'.050
.648
area skin '" . 096
-:353 1 area of skin
area of web
"'.090
",.032
No.3,4 & 5 (t '" . 050) . 630
area stringer '" . 205
area skin '" . 096
.04 upper skin . 301

Z' 7

~--L_
•~I----------""='T'-'-,r--,j <f"'----,:;-;;:;:;
- - - - - .. o X Centroidal Axis

-~
~ .c

N ~
z '"••
14 C!

8
.032 lower skin

I
,- Lower Surface Stringer
area of bulb angke e • 11
7/8 area skin '" . 16
I -:27
5"

Lower Rear Flange

2

Ir 3.75
1
area of angles '" . 508
area of skin "'.120
area of web "'. 064
. 692 Lower Front Flange
area of angles '" . 508
r2
area of skin "'. 160
area of web "'. 100
:768
-
,
s
Fig. A19. 25 WING SECTION AT STATION 20. EFFECTIVE AREAS.

Z'
I
. 032 upper skin

.•• ,\\.o-c.
0\ ~e'"
7

•o _r
["".0.. , X' reference axis
~:----=====-~r======-=,,========...::;~===..::"::=======-=..:::j::=i
.- X Centroidai Axis

14
13
12
10_j 9 8
Z' Z "'- .025 lower skin
Fig. A19.26 wrsc SECTION STATION 47. 5
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A19.17

Fig. A19.25 shows the cross section at Substituting in equation (4)

Station 20 divided into 14 longitudir~l units
numbered 1 to 14. Since the external load con- crb =- [.00098 x - 285000 - (-.0002125 x
dition to be used places the top surface in
compression, the skin will buckle and thus we 1300000lJ x - [.004378 x 1,300,000 -
use the effective width procedure to obtain the (-.0002125 x - 285000lJ z .
skin ?ortlon to act with each stringer. Fig.
Al9.25 shows the effective skin which 1s used whence, crb = 3.3 x - 5639 z- -----(5)
with each flange member to give the total area
of numbers (1) to (7). The skin on the bottom Column 12 of Table Al9.2 gives the results
surface being in tension is all effective and of this equation for values of x and z in
Fig. A19.25 shows the skin area used with each columns 10 and 11. Mu1ti~lying these bending
bottom flange member. stresses by the stringer areas, the stringer
loads are given in column 13. The sum of the
The next factor to decide is the stringer loads 1n this column should be zero since total
effectiveness as discussed and explained in the tension must equal total compression on a sec-
previous example problem. For the structure at tion in bending.
Fig. Al9.25 we will assume that the compressive
failing stress of the stringers is the same as Stresses at station 47.5:
that for the corner members, thus we will have
no correction factor to take care of the Ix =157.4, Iz = 700, Ixz =-35.4 (Table
situation of flange members having different U9.3 )
ultimate strengths.
K, = 35.4/(157.4 x 700 - 35.4') =
Table A19.2, columns 1 to 11, and the -35.4/108950 =-.000324
calculations below the table give the calcula-
tions for determining the section properties K, = 700/108950 = .00643
at Station 20, namely A, Ix, I z and I xz•
Table Al9.3 gives the same ~or wing section at K, = 157.4/108750 = .001447
Station 47.5. The areas 1n column (2) are less
sir-ce sizes have Changed between Stations 20 crb =- [.001447 x -215000 - (-.000324 x
and 47.5.
1,000,000)J x - [(.00643 x 1,000,000
Calculation of longitudinal stiess due to (-.00324 x -2150001J z
Mx and My bendi~g moments: -
whence
The deSign bending ~oments will be assumed
and are as follows: - crb = -14.5 x -6360 z
Station 20 t'lx = 1,300,000 in .n., Column 12 of Table A19.3 gives the results
of this equation and column 13, the total
t'lz =- 285,000 in.1b.
stringer loads at station 47.5.
Station 47.5 ~
"x = 1,OCO,000 in.lb. The stresses 1n column 12 of each table
!'!z = - 215,000 tnv Ib , would be canpared to the failing stress of the
flar~e members to obtaIn the margin of safety.
The moments about the y axes are not
needed in the bending stress analysis but are ANALYSIS FOR SHEAR STRESSES IN ',4EBS A.."ID SKIN
~eeded in the shear analysts which will be made
later. ~he shear flow distribution will be cal-
c~lated by using the change in axial load in
To solve equation (4) the constants K., the stringers between stations 20 and 47.5, a
Ka and ~ must be known. method c~~only referred to as the AP ~ethod.
For explanation of thiS ~ethcd, refer to Art.
For Station 20 from Table A19.2, Ix = A15.l6 of Chapter Al5.
230.3, I z = 1030 and Ixz - - 50, whence -
The shear flow in the y direction at a
Kl. = Ixz/(I x I z - Ixz'i) = -50/(230.3 x 10.30 point n of the cell wall equals,
- 50') = -50/235500 = -.0002125 n oP
Z- - - - - - - - - - - - (6)
K, = I z/235500 = 1030/235500 = .004378 a d

K3 = I x/235500 = 230.3/235500 = .00098 where qo 1s a known value of shear flow at some

point 0 and the second term is the change in
shear flow between paints 0 and n.
A19.18 ANALYSIS OF WING STRUCTURES

TABLE A19. 2
SEC1'10N PROPERTIES ABOUT CENTROIDo\L X AND Z AlC4S

.in~ Section at St~t1on 20 (Co~press10n On ~pper ~urface)

1 , 3
• 5 II 6 7 8 8 10 I II 12
!
13 !
I, Flange
Mo.
Are"
A Z' AZ'
,
AZ·2 X'
! AX' AX';': AX'Z' Z ~ Z' - z! X ~
.' - x I "b
"
= "'b , I
!
,
I 3z
1 .6<18
.353
5.50
5.90
3.56
2.08
19.61
1':::.30
-0.10
5.65
_0.06
1. 99
0
11.30
_0.36
11.78
6.68
7.08
-17.4
-11.65
_37660
_39940 i _24390
_14070
!;
.300 I 5.85 1, 76 10.28 1l.:W 3.36 37.60 19. 6 7 1 7.03
I -
- 6.10 _39620 ! -11890
i • .300 5.55 I 1.66 9.25 16.85 5.05 85.00 28.10 6.73
,
0.45 -37950 ! _11370 I

o
6
.300 5.05 I 1.52
1, 55
7.65
6.33
22.40
28.00
6.72
9.89
I 150.50
277. 00 34, 00
43.50 I 6.23
5.58
I ::;.10
iO.70
-35115
! -31450
, _10510
-11090
7
.3511 4.40
2. ;.!4 7.95 35.72 23.15 826.00 80.00 4.73 18.42 I _26560 ! _16700
8
.630 3.55
.692 _8.40 -5.81 48.90 ! 35.72 24.70 I 881.00 -208.00 -7.22 I 13.42 .;0750 28230
,
8
10
.2701-8.50
.270 -8.50 i
I _2.30
_2.30
19.55
19.55
27.80
22.60
1.50
6.10
208.00
138. 00
-- 52.00
64. 00
I
I _7.32
-7. J2
!
10. SO
5.30
41280
41250 I
11180
11170
'i.i. .160 -8.30 _1.33 11, 05 16.70 2.67 44.60 - 22.20 i -7.12 !
I - 0.60 40100 I 64)0

12 .27
,
-8.00
I _2.16 17.30 10.90 2.94 32.00 - 23.50
- 11.35
-G.82
I -_11.7
6.40 38400 10390
13
1< ."
.768 I
-7.50
_6.50
_2.02
_5.00
15.20
32.50 -
5.60
(LID
L51
.08 !
8.45
0 0.50
_6.32
-5.32 I _17.40
35570
29950 I!
9600
23020

i 1:. 5 8 4 _6.56 238.0 I I 96.44 i 2700 _163.8 , i

0.00

,· _6.56 • _1.176
Ge ne r a l Nates:"

I I
·
5.584

96~44 ;. 11.3
5" l' ig. A19.25 for Section at Station 20.
Rete e enc e axes X'X' and 2'2' are assumed
Properties are calculated with respect to th"se
shown.
"
238 _ 5.~84 ~ 1.176~ axes and transferred to the c e nt r ci da I X c nd
230.3 in.';
Z axes.
2700 ~ 5.584 ~ 17.3 2 1030 in.';

I~z • -163.8 ~ 5.584 x-l.17~ x 17.3 • ~50 "'b * ~.3 X _ 5639 Z

TABLE A19. 3
SEC1'IOK PBOPERTIES ABOUT CIDfTROlnAL X AND· Z AXES

llinf:" Section a t Station 47.5 (Col:lpression on upper S~rtace) I

!
1 a 3
• 5 G 7 s 9 10 II 12
!
13

1"lnn;.;o Area
AZ'~ AX· 2 ,. - . _ 1r ,
)(0. A Z' AZ'
I
X· AX' AX''Z'
" Z' X X· "b • "'"b .J,

,
1 • .;7G 5.(0
I
2.66 14.95 - 0.10 - 0.05 0 - 0.27 6.21 -15.60 _3927:; ~18700
I
!,
I 3
.318
.2136
~.83
~.75 I
1. 86
1, 53
10.80
8.80
4,40
10.00
1. 40
2.56
6.16
26.60
8.15
15.30
6.44
6.36
,, _11, 10
- 5.50
-40790
-4025 ,
_12990
_10710 ,

I, •56 I
.266 1, 44 -38400 I _10200
.266
5.'Z
4.90 1, 30
7.81
6.40
15.60
~1. :::0
4.15
5.64
64.75
1::0.00
22.40
27.60
6.03
5.51 i 0.10
5.70 _3~075
-
9320
1
.318 4.30 r 1, 37 I 5.88 I :':6.70 8.50 226. :;0 35.60 4.91 !, 11.20 -31360 ! _10000
i
i 7 .476 I _7.10
3.60 1.71 6.17
i 31.70 15.10 478.00 54.40 4.21 16.20 -27030 , _12880
!
8 .35'::
i -7.40
-3.93 27.90 31.70 17.51 :iS5.00 -124.50 I -6.';9 16.20
,I 41050 I 22770

I
8
10 !,
.23:;
.':;35 -7.40 I
_1.74
1. 74 I
1:::.90
12.90
:i!5.50
21. 30
I, 6.23
5.00
165.00
107.00
-
-
';6.10
37.18
_6.79
_6.79
I
11.00
5.80 I
43090
43180 I
10140
10160 i
II .1:.:5 I -7.20 ! -0.90 6.48 15.10 1.89 l8.50 - 13.60 -5.59 - 0.40 i .;1900 I 5240 ,
1 c~ · ;:3<: ;_e.. 90 _1.62 11.20 9.60 :':.::6 :0:1.70 - 15.GO -6.29 _55.90 I 40085 I 9410
13 . 23:; ,i _6. 2S _1.4 • 9.20
-
4.50
I -
1. 06 4.7S - 6.61 _5.84 _11, 00 i 36060 I 8470
,
"1 , 50~ -::'.40 I

i
_3.27 17.70 0.10 0.06 0 0.33 _4.79
,,
_15.50 30725 18600

I I - 79.0
..
4.61 _2.80 1~9. 1 71. 3 1804 0 !
-~'. aO/4. 61 . _0.51" I
r . 71.3/4.61 . 15.50"
General Notes:
See Fig. A19.26 for section at Station 4 •. 5

· -- 4.61 :t.61 2 . Reference axes X'X' .nd 2'Z' are assumed

. " shown.

.
Ix IS!!. 1 157.4 i
I,
·. -.9 -1304 4.61 15.5': . 700
°b -14.5 X - 6360 z

I" 4.61 x -.61

• 15.5 - 35.4 I
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A19 19

~p equals the cbAnge in stringer ~~1al load areas are referred as ~ values (See Fig. A19.27)
over a distance d in the y direction. Column 6 of Table A19.4 records these double
areas which were obtained by use of a planimeter
Since the cell in our problem is closed the Column (7) gives the moment of each shear flow
value qo at any point is unknown. we assume it about the e.g. and the total of this col~
zero on web 1-14 by imagining that the web 1s gives the moment about the e.g. of the complete
cut as shown in Fig. A19.27. Equation (6) thus shear flow system of Fig. Al9.27 or a val~e of
reduces to, 256060 in.lb.

The double areas (m) can be found approxi-

- - - - - - - (7) mately as follows:

Columns 1 to 5 of Table A19.4 show the solution

of equation (7). The shear flow values of qy
The moment of
the shear flow q on 1--- x,
x,
----1 _~~
I
in column 5 are plotted on the cell wall in
= =
the web (2-3) about
2 -, . 3T

I F
~
Fig. A19.27, remembering that qy qx qz. For point 0 equals q
rules giving direction of qx and qz refer to times twice the area
Chapter Al4, Art. A14.6. (A 1. + A a ) . In most
cases, the area Aa , A. z,
245.9 288.8 ", , 374.6
414.2 can be neglected.
206.'
I - - -;. 4 5, - - -6--- 553.2
l 1
By Simple
-rj geometry, the area
z I A1.= 1/2 rx, Za - X",Z;s)' The moment of the
x -W area qz~ shear flow q on web (2-3) thus equals q (X;, Z.

T~rea A\;·.· ' - X a Z;s)' Since values of X and Z for all

I " :1 flange pOints with reference to section (e.g.)
'L_~3£,-£ 11 10. - .Qx_L_ are given in the Table A19.2, it is unnecessary
to use the planimeter except for regions of
I'O.T .~; - - _ . - - I sharp curvature.

Fig. Al9.27 !10l1ENr OF EXTERNAL LOADS

ABOUT e.g. OF STATION 20
TABLE A19.4
As stated before the engineers in the
Calculation of Flexural Shear Flow Assuming q 2 'Zero in applied loads calculation group supply the shears
Web Between Flange Members (1) and. (14).
and moments at various spanwrae stations. ve
(Shear values are average between stations 20 and. 47. 5) will assume that these loads are: Vz = 12000
1 a 3
• 5 e 7 a 9
=
lb., Vx +2700 lb., My =-
390,000 Lm Ib . The
location or the reference y axis used by the
p p
Flange at Station at Station 2,p
q. m q' • loads group will be assumed as located at point
No.
au 47.5
LP
-27.5 Z 2f.5 sq.
in.
mq
" q .. ql
o in FIg. A19.28 relative to cross-section at
Station 20.
1 ~24390JII -18700 206.8
a 206.8 48.4 10000 -328 -121. 2

:
-14070 -12990 39.1
3 -11890 -10110 42.9 245.9 43.8 10780 -328 - E12. I

•s VZ'12r~q:
-11370 -10200 42.5 288.8 42.0 12120 -328 - 39.2
331.3 42.6 14100 -328
!~.~.
-10510 - 9320 43.3 3.3
e ~11090 -10000 39.6 374.6 43.6 16340 -328 46.6
414.2 42.0 17500 ·3213 86.2
7
a
-16700
28230
·12880
22770
139.0
~198. 7
SSJ.2 197.0 108900 -328 225.2 v x-- 27001~ .01--L-
11.8 -
s _ 37.8 354.5 47.4 161300 -328 26.5
My=39~33.3
11180 10140
10 11170 10150 _ 36. 7 315.7 37.2 11800 -328 - 11.3
11 5430 5240 ~ 43.2
280.0 40.8 11420 -328 - 48.0
11 10390 S410 • 35.15 236.8 42.2 10000 -328 - 91. 2
13 9600 8470 - 41. 1 201. 2 38.0 7650 -328 -126.8 Fig. A19.28
L60.1 54.0 8650 -328 -167.9
, 14
1
23020
I
18600 -160.3
0 204.0 0 -328 -328
Therefore moment ot external loads about
Total mq . 256060 c vg, is,

For equilibrium af all the forces ln the ZMc.g. =12000 x 33.3 - 2700 x 11.8 - 390000
plane of the c~oss-section ~y ~ust equal zero. = 41800 In.lb.
ror convenience we will select a morr.ent y axis
through the c.g. of the c~oss-section. The Moments Produced by Inclination 01 Flange Loads With
~oment of the shear flow q on any sheet element Beam Section.
equal q times double the area of the triangle
formed by joining the e.g. with lines going to Since the flange members in general are not
each end of the sheet element. These double normal to the beam secttons, the flange loads
 A 19 20 ANALYSIS OF WING STRUCTURES

have components In the Z and X direct1ons. Due to external Loads = 4.1800 in. lb.
Columns (4 ) and (7) of tree Table A19.5 g1ve the
values of these In ;Jlane c:::mponents. The slopes T;;'e~ the total unbalancec :::o:n.e~t = J.C"'~
• ...-•.A) +
dx/dy between stations 20 and 47.5 are f'olU1d by 256060 + 4l8CO =
302795 Lnc Ib ,
scaling from Fig. A19.24. F1g. Al9.29 shows
these 1nduced In plane forces as fOW1d in Table For equt Lfbr tum, this ::!lust be baIancec oy
A19.5. a constant shear- flow q,.

TABL£ A19. 5 hence

302796
[n Plane Moments About Section e. g. Produced by
Components o{ Flange Loads
In Plane
q, = 2A
!1 -- 2 X 461.5 -- 328 lb ./in.
Station 20
(Note: 461.5 = total area of cell )
1 2 a 4 I 5 s 7 8

Flange dx
Px = I M e. g• d,
Fz " Me. g, = The shear stresses q, are listed In Column
(8 ) of Table A19.4.
p
No. dY dY
p dx I = Px z dy p.!!!..
dy
·(P z Xl
The final or resultant shear flow qr at an y
I -24390 -.046 1120 7480 .025 ·610 -10600
2 -14070 0 0 0 .022 -310 • 3610
pOint therefore equals
3 ·11890 0 0 0 .022 -262 · 1600
4 -11370 0 0 0 .022 ·250 · 112
qr = q + q,
,
5 -10510
-11090
0
0
0
0
0
0
.022
.025
-231
-278
1180
2970
7 ·16700 ,021 - 350 -1660 .025 -417 7770 The resulting values are given 1n Column 9
,
B 28230
11180
,021
0
594
0
-4300
0
.018
. 017
-508
·190
9350
2000 of Table A19.4. r1g • A19.30 illustrates tr..e
10 11170 0 Q 0 .015 -167 885 :-esults graphioally.
11 6430 0 0 Q .015
"
.-177 · 57
12
12
14
10390
9600
23020
0
0
-.046
0
0
·1060
,
56-10
Q
Q
,017
,018
.018
·173
-416
• 1130
• 2020
• 7250 -_.- _.
1
/ /

2 3 4 -5- ,--, /

~~ II
7160 • 2224

7.~ •
+,.g.
m
NOTES:
--Column (2); P Ircm Table Al9. 2
Column (5) and (8): Values of Z and X are found iII
0' '/
Colwnns 10, 1 I of Table Al9. 2
---
lli
'(L//
13 12
'-_.-
11 10
---8
9

810 417 Fig. A19.30 Final shear flow diagram. For val-
I

t·
3 ues see Column 9 of Table Al9.4.

r"
1 I' 15 16
310 282 250 231 278 Having deter:nined the shear flows, the
shear stress on any panel would be qjt. I:l
+,.g. checking the sheet for strength In shear and
106

I
14 ~3 lil 1,1 .
10 9 8
I
594
combined shear and tension, interaction rela-
tianShips are necessary. The strength deSign
418 1~3 1;7 9
18
d7 1~0 508
of sheet panels under combined stresses Is
covered in considerable detail i::1 Voli.J:l8 II.
A19,14 Bending and Shear Stress Analysis of 2-Cell
Fig. A19.29 In plane forces prOduced by flange Multiple stringer Tapered Cantilever
axial loads. Wing.

The moments of these in plane components about A two-cell beam IS also quite common In
the section e.g. are given in Columns (5 ) and wing structural design. A 'DNa-cell structure
(8) of Table Al9.5. In general, these moments In bending and torsion Is statically indeter-
are not large. minate to the second degree since the shear
flow at any one pOint In each cell Is unknown,
Total I'!oments of All Forces About Section .:.¥; . However, due to continuity between cells the
at atat i cn 20: angular twist of each cell must be the same,
which gives the additional equation necessarJ
Due to flanges = 7160 - 2224 4936 in.lb. = for solving a ~No-cell beam as compared to the
(Ref. Table Al9.5) single cell analysis.
Due to assumed static shea:- :low 256060 = Example Prcb Ieta
1n.lb. (Ref. Table A19.4)
To avoid re'Oet1t1on of similar t e calcu-

£ SKU
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A19.21

lations as was used in the previous single cell The first 7 columns of this table are the same
problem, the bending and shear stresses will be as in Table A19.4, since no stringers have been
dete~lined for the same structure as in the added to cell (1), and the shear q is assuoed
previous example except that the leading edge zero In cell (1).
cell is considered effec:lve, th~s ~king a
2-cell structure. Since there are no spanwise ~o ~e the twist of each cell the same and
st~ingers in the leading edge, very little skin also to make the summation of all torsional
on the compreSSive side will be effective. On forces zero w111 require two unknown constant
the tension side, the leading edge s~in would shear flOWS, q~ in cell (1) and qa in cell (2).
be effective in resisting bending axial loads Thus VNQ equations will be written, namely:
and thus the moment of inertia would be slightly
different from that found in example problem 1. (8 )
Since this pr0blem is only for the pur,ose of
illustrating the use of the equations, the Zl1c .g , = 0 (9 )
leading edge s~in Nill be neglected in computing
the bending flexural stresses. ~lth this The VNist Q per unit length of a cell
assumption, the bending stresses and flange equals
loads at stations 20 and 47.5 are the same as
for the previous prob Lem. (See values in ccj.cnn _ 1
12 and 13 of Tables A19.2 and A19.3.) Q - 2AG 1:qL/t uo:
Shear Flow Calculations: - The modulus of rigidity G w1l1 be assumed con-
stant and thus will be omitted.
To compute the static shear flow, each cell
is assuned cut at one point as shown in Fig. Consider cell (1): (Refer to Table A13.6,
A1S.31, and thus the shear flow is zero at C01UITL~S 10 ~~d 11)
points (a) and (b).
Fig. A19.31
Area of cell (1) =23.5 sq. in.

G L/~

·~r
2 3 4 Q, ; _1 q Lit. Z q, Lit. 1: q.
5 2A,

[
6

~83.5 14 Area
Cell (2)
:z461. 5 sq. in. Q, ;
2 x1 83• 5 ~ + 1278 q~ - 230 qJ, whence
9
1.3 12
'.' 10
Q1 =7.65 q~ - 1.378 q.- - - - - - - - (a)

For cell (2):

Table A19.6 (Column 5) giyes the va.Iue of
the static shear r icw under- these assumptions. Ar-ea of cell (2) :: 461.3 sq. in.

TABLE A19. 6
SHEAR FLOW 2 CELl. - -
VULTIPLE FUNCJi TAPERED BEAM
(Aver:li:C Sh.. nr Flo.. Between Stations 20 and 47.5)

1 ! , I
3
,
, • I , 6 7
• 9 10 11 12 13 14

Flant:e i
<0.
P :P .l I." ),,~P_ • , LI<
~(~t c'(gl
q (Lit)
Ce~~
'r ,
i at St;... i :J.t_Sta·l_ '; p.
! cO , 4 •• 5 , .1.0> 1-7.;, "' (i;. ) I I ~(~) (2)
'1 '2
Cl-KIl-KI2
,
1 1- 24390 , _18100 206.8
206. sl
48.4 10000 5.15 .04 1<4 29800 0 _317 -1l0.2
2
,
3
I -14070
i -11890
.11310
I
_12990
_10710
_10200
39.1
4:.\.9
245.9 43.S
288.8 4:;:.8
10780
12120
5.50
5.50
.04
.04
137
137
33700
39600
0
0
-317
31 7
I -- 11.1
28.2
42.5 -
_317
331.3 42.6 14100 5 • .50 .04 137 45400 0 H.3
1

5 _10510 - eaac ~3.3

374.6 ~3.5 1634lJ 5.50 .04 137 I .51400 0 _317 I 57.6
6
r
-11090
-16700
_10000
_128S0
39.6
139.0
4.14.2 42.0
553.2 197.0
17500 5.50
108900 12.25
.04
.032
137
382 I 56700
211500
I 0
0
I -317
_311
:
I
97.2
236.2
a ::::8:<30 22170 -t9a.7
_ 31.8 35~. 51 ~7.~ 16S00 I '-' .032 r vz 61000 0 -317 I 31 • .5
s 11180 10140 , 316.71 37.2 11800 5.0 .032 156 49400 0 _317 : 0
10 11170 10160 _ 36,7
2S0.01 40.8 11420 5 •• .032 181 50650 0 -311 I - J7
11
12
6430
10390
5240
9410
_.43.2
- 35. 6 236.81 42.2 10000
I .032 ! 1S1
I 42900 0 _311 -_115.8
80.2
13
14
9600
~3020
i
:
8470
18600
... 41.1
160.3
201. 2 J8.0
160.11 .54.0
'""0
8650
5.'
5.3
'-'
.032 !
.032 ,
166
172
33400
21500
0
0
_311
-317 _1.56.9
1
, o 204 0 11.15 .051 ! 2" 230 0 0 _62.8 -311 _2.54.2
1
Lead
cd!>:e
eee
,
i
! I
!
0 J7l 33.5 .032 i 1048 i 0
,I -62.8 -62.' 1

,
,
, I
,
S~ I I 256060 1218 2469 I 0 133050

MOTES: ~ length of web she~t betweea flange s.mbers. Cal. ~ .. (Cal. 2 + Col. 3)/27.5
t web thickness.
A19.22 ANALYSIS OF WING STRUCTURES

~1.5 ~733050 q~
A19.15 Bending Strength of Thick Skin - Wing Section
2 X - 230 + 2469
?igs. i and k o! Fig. AIG.3 ill~s~rate
approxi~ate srzpes for air~oils c~ supersonic
G. = 794 - .2495 q, + 2.678 o, - It)
aircraft. Such e u-ro i ;s nave r eiat ive iv low
thlc~~ess ratios and since supersonic ~l:itary
For continuity O~ ~ust equal O~, hence aircraft have comparatively h:gh wl~g loaclngs,
equatlr~ (al and (b):
it is necessary to go to thick skin in order to
resist the wing bending ~cments efficiently.
7.899 q~ - 4.056 qa - 794 = 0 - (c)
The ultinate compressive stress of s~ch struc-
tures can be made rather uni rcrn and o c curr-t ng
For equilibrium the summaticn of all mom- at stresses considerably above t~e yield ~oint
ents in the ,lane of the cross-section abo~t the
of the material. Since struc:ures wust ca~ry
section (e.g.) ~ust be equal to zero, or the design loads without failure, it is neces-
sary to be able to calCUlate the ultimate bend-
Z ric.g. = 0 ing resistance of such 3. wtng section i t t::e
~argin of safety is :0 be given ~or various
The ~~ent of the external loads about the load conditions.
section e.g. is the same as in previOus problem.
= 41800 in.lb. The question of the ultimate bending
!1 external forces resistance of beam sectlor~ that ~al1 at stresses
beyond the elastic stress Yar~e is treated in
The induced moment due to the ~n plane
Article A13.l0 and example problem 7 of Chapter
components of the flange axial loads is likewise A13 and should be studied again betore ~ro­
the same as in previous problem (see Table ceeding with the following example ,~oblem.
A19.5).
A19.16 Example Problem
Mdue to flange loads =4936 in.lb.
~a illustrate the ;roced~re of Art. A13.l0,
The torsional moment due to ~he static a portiO~ o~ a thick =kin Ni~g section as
shear flow rr ce Column (7) or Table Al9.6 equals 1l1L.:.strated In F'fg , A19.34 'N111 be cons Lder'ed ,
256060 in. lb. The torque due to the unknown
constant shear flows at q1 and q~ 1s equal to
twice the enclosed area ot each cell times the
6" 6'" 11.-3/ S'"
shear flow in that cell, whence
l~
3" ~l2
3 ~ 1","" T
3 -l13/8"
+
8 2 a
due to q::,. and Gil) = 2x83.5 q~ -+-2 x 461.6
l'l( 1 '"
2-
qll = 167 q:l, -+- 923 qll .t.. 2
---x

Therefore ZMc.g. = 167 q~ + 923 qll . .

302796 =0
Solving equations (c) and (d), we obtain,
q1 = -62.8 lb./in., qll = -317 Ib./in.
These values are listed in columns (12) and (13)
ot Table A19.6. The final or resultant shear
flow qr on any sheet panel equals the sum at
q .+ q~ + qll' The results are shown in cojunn
(14) of Table A19.6. Fig. Al9.32 shows the
potted shear flow pattern. Comparing this
figure With Fig. A19.30 shows the et!ect of
adding the leading edge cell to 'tne Single cell
or the previous problem.
3
l-iS
3
16 I--fa-
- - (d)
C'--J ,..-, '=='
L "
Fig. A19.34
For s~pliclty the section r~s been symmetrical
about the x-x axis. The mater~al is aluminum
alloy. In this problem the ~aterial stress-
strain curves will be assumed the s~e ih bot::
te!'lsion and compression. The ?rOblem 1s to
determine the ~gin of safety for ~~is beam
section when subjected to a j eaf gn bending
IDoment Mx = 1,850,000 in.lb.
SOLUTION:
Since it is desirable to use the beam
to~ula ab = ~xZ/lx, it is necessary to obtain
a mod::ied beam section to correct ~or the !'lon-
linear stress-strain relationship since the
give structure will tail ~nder stresses in the
inelastic zone. The =axilnUffi compressive stress
at sur~ace of beam will be ass~ed at 500CO pSi.
~ls value could be calculated from a conslder-
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A19.23

ation of crippling and column strength at the
sti:fened Skin, a subject treated later.
Curve (A) of Fig. A19.35a is a portion of
the compressive stress-strain diagram of the
aluminum alloy material from zero to 50,000 PSi.
Due to symmetry about the x-x axiS, we
need only to consider one halt or the beam
section. We divide the upper half of the beam
section into a horizontal strips, each 3/8 inch
thick. Each beam portion along these horizontal
strips can be placed together to form the areas
labeled (1) to (8) in Fig. b or Fig. A19.35.
Since plane sections remain plane after bending
in both slastic and inelastic stress zones,
Fig. c shows the beam section strain picture.
Fig. A19.35
The values in colljffiTI (3) of Table A1S.7
represents the true compressive stress at the
~idpoint ot a strip area when the beam 1s
resisting its maximum or failing bending'nament.
The values in column (4) represent the com-
preSSive stress at the midpoint ot the strip
areas it the bending stress is linear and vary-
ing trom zero at neutral axis to 50000 psi at
edge of beam section (Curve B of Fig. A19.35a).
To illustrate, consider strip area number
(2) in Fig. b ot A19.35. ~oject a horizontal
dashed line from midpOint of this strip until it
intersects curves A and B at paints (a) and (b)
respectively. From these intersection paints
project downward to read values of 48000 and
40600 pSi respectively.
In using the linear beam formula, the
stress intensity on strip (2) would be 40600
but actually it is 48000. The ratio between
the two 1s given the symbol K. Thus to modify
the linear stress to make it equal to the
nonlinear stress we increase the true strip
areas by the factor K, giving the results ot
co Lumn (6).

The modified moment at inertia (column 8)

equals Ix = 115.52.
The design bending moment was 1,850,000
in.lb.
Consider point at ~idpoint of strip (1)

Table A19.1 Shows the calculations for

Z = 2.8125 Inches
obtaining the ~odifled moment of i~ertia of the
cross-section to use with the linear beam 0b =Mx Z/I x = (1,850,000 x 2.8125)/115.52
formula. = 45100 psi.
TABLE Al9. 7
This stress is based on the modified strip
1 , 3 4 5 6 7 6 areas. The true stress Ob(t) on strip 1 thus
Area K= Arm
equals Kab = 1.049 x 45100 =
47400. The allow-
Portion ac/ac' Ae-KA Z AeZ' able stress at failure equals 49200 from column
A Oc Oc'
(3) or Table A19.7. Hence margin ot safety =
1 4.500 49200 46900 1.049 4.71 2.8125 37.30 (49200/47400) - 1 =
.04 or 4 percent.
a 2.437 48000 40600 1. 182 2.88 2.4375 17.14
3 0.211 46000 34400 1. 340 0.282 2.0625 1. 20 The margin of safety for points other on the
beam section will likewise be 4 ~ercent. For ex-
4 0.211 43200 28100 1. 539 0.324 1. 6875 0.92
ample at mtcpo trrt of strip (3), Z ::: 2.0625. K:::l.34
5 0.211 39200 21840 1. 790 0.378 1. 3125 0.65
6 0.211 33200 15620 2.121 0.449 10.9375 O. 38 whence O-:t ::: [(1,850,000 x 2.0625)/115.52J1.34-
7 0.211 20500 9360 2.190 0.461 10. 5625 O. 15 = 44400.
8 0.211 7000 3120 2.240 0.473 0.1875 0.02
whence ~rg1n ot safety (46000/44400 )
Z 57.76
- 1 = .04
Ix --'
= 115. 52
The moment of inertia Without modifying
Col. (2) = Area of Strip
Col. (3) Stress at midpOint of strip as read from Curve (A) the strip areas would come out to be Ix = 104.42
Col. (4) Stress at midpctnt of strip as read from Curve (B) hence the stress at midpoint at strip (1) would
Col. (5) Nonlinear Correction Factor K = crcia c' calculate to be crt ::: (1,850,000 x 2.8125)/104.42
Col. (6) Modi1ied Area = Ae = KA
Col. (7) Arm rrom Neutral Axis to Midpoint of Strip = 49900. The allowable str~ss ror linear stress
Col. (8) Modified section moment of inertia variation would ~e 46900 from column (4) of
Table. Hence margin of safety would be
A19.24 ANALYSIS OF
(46900/49900) - 1 = -.06. Tte elas:ic theory
thus gives a ~a~gin of safety 10 percen~ less
than the strength given when t~ue st~ess-strair.
or non-linear relationship is used.
If the s&~e comparison was ~de for bendl~g
about Z axis of this same beam section the
difference would be consi1erably more than 10
percent as more beam area is acting in the
region of greater cescr-epancy between cur-ves A
and B.
Al9.l7 Application to Practical Wing Section
A practical wing sec~lon involves these
facts: - (1) The section is ur~ynmetrical;
(2) external load planes change their direction
under different flight conditions; (3) the
material stress-strain curves are different in
tension and compression in ~he inelastic range.
Since the stress analyst 3ust deter.nine
critical margins of safety for ~y conditions,
it would be convenient to have an interaction
curve involving Mx and Mz bending moments which
would cause failure of the wlr~ section. This
interaction curle could be obtained as follows:-
(1) Choose a neutral axis direction and its
location.
(2) Assuming that plane sections remain plane,
and taking the ::lax11ll.um sera in as tnat
causir~ tailure ot the compressive flar.ge,
use the stress-strain cu~re to dete~ine
the longitudinal stress and then the
internal load on each element of the c~oss­
section. A check on the locat:on of the
assumed neutral axis is that the total
compression on cross-section must equal
total tenslon. Since the location was
assumed or guessed, the neutral axls must
be moved parallel to itselt to another
location and repeated until the above
check ls obtained.
(3) Find the internal resisting :noment about
the neutral axis and an axis normal to the
neutral axes. Resolve these moments into
moments about x and z axe'S or Mx and Mz•
These resulting values of Mx and are Mz
bendlr~ moments which acting together will
cause failure of the wing in bending.
(4) Repeat ste~s 1, 2 and 3 for several other
directions for a ~eutral axls which results
will glve additional combinat10ns of
and Mz moments to cause wing failure. Thus
an interaction curve involving values of
Mi and Mz which cause failure of wing in
bending is obtained and thus the margin of
safety tor ~~y design condition ls readily
obta Inab Le ,
WING STR lie T URES
A19.l8 Shear Lag Influences
I~ the beam ~heory, the aSSk~~tiQ~ ~s ~ade
tha~ plane sections ~errzi~ plane after be~cin~.
In a beam ~nvolving sheet and stringer ~anels,
this asswm~ticn ~eans that the sheet panels
have inf:n1te shearing rig:d~ty, whlct o~ course
is not tr~e as shearing stresses produce shear-
~ng strains. The effect of sheet ~anel shear
strair.s is to cause Some str1~gers to ~esist
less axial load than those calculated oy Jeam
theory. T~is cecreased eifect:veness of
stringers is referred to as "sneer ::'2.g~ effect,
since some stringers te~d to lag bac~ f~8m the
pos~tion they would take if ~lane sec!ions r8-
~in plane after bencing.
In general, the shear lag effect ~n sheet-
stringer structures is not appreciable exce,t
for the following sltuat~ons: -
(1) Cutouts which cause one or more stringers
to be discontinued.
(2) Large abrupt changes in ex~ernal load
applications .
(3) Abrupt changes ~n str~nger areas.
In Chapters A7 and AS, strains due to
Shearing stresses were considered ~n solving
for d~stortions and stresses in str~ctures in-
volVing sheet-st~1nger construction. Even ~n
these so-called rigorous methods, s~plify1ng
assw~ptions nust be made as for example, shear
stress is constant over a pa~icular sheet
panel and estimates of the modulus of rigidity
for sheet panels under a varying state of
buckling must be made. The number of stringers
and sheet panels in a normal wing is la~e,
thus the structure 1s statically indeteroinate
to ~ny degrees and solutions necess~tate the
use of high speed computors. Before such
~~lyses can be made, the size and thickness of
each structural part must be kno.vn, thus rapid
approx1wate ~ethods of stress analysis are
desirable in obtaining accurate prellminar;
sizes to use in the ~ore r~gorous elastic
analysis.
To illustrate the shear lag problem in its
Simplest state, consider the three stringer-
sheet panel unit or Fig. A19.36. The three
stringers are supported rigidly at B &id equal
loads P are applied to the two edge str1ngers
labeled (1) at point (A). The center strir~er
MX (2) has zero ax1al load at (A), but as end B is
approached, the sheet panels transfer some of
the load P to the center stringer by shear
st~esses in the sheet. At the support paints B
the transfer of load trom side stringers to
center stringer is such as to make the load in
all three str1ngers approx~ately equal or
equal to 2P/3.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A19.25

B ( 1) A equals 3b. At beam section 2-2 stringers (5)

2P/3 -
Sheet
2P!3
B '(2)
Sheet
2P/3 . . -
B (1)
Fig. A19.36
The theoretical load in center stringer
can be calculated by methods of Chapter A7 and
AS, and the results would ~ive the solid curve
of Fig. A19.37. To simplify the solution, it
is common practice to ass~~e the load distri-
bution in the center stringe~ to varJ according
to the dashed curve in Fig. A19.37 which
indicates that in a distance 3b, the load 2P
1s equalized between the three stringers.
2P/31-- - - -- - -"
---.
--
p and (7) have become effective since they inter-
J
I
sect triangle at points (a) on section 2-2. At
section 3-3 point c, stringer (6) becomes fully
x*- effective.
b
-L- p
A To handle shear lag etfect in a practical
wing problem another column Would be inserted
in Table A19.l between columns (3) and (4) to
take care of the shear lag effect. The shear
lag effectiveness factor which we will call R
would equal the effectiveness obtained trom a
triangle such as illustrated in Fig. A19.3S.
For example, the shear lag factor R at
beam section 1-1 in Fig. A19.38 would be zero
for stringers (5), (6) and (7) and one for all
other stringers. At beam section 2-2 stringers
(5) and (7) have a factor R = 1.0 since they
are tully effective at paints (a). Stringer
(6) is only 50 percent effective since section
Theory
2-2 is halfway from section 1-1 to paint (c),
thus R = 0.5 for stringer (6). At beam section

~
' " Approximate 3-3, stringer (6) becomes fully effective and
Load ill '<, ~ thus R = 1.0 for all stringers. The tinal
I'
I
Stringer modified stringer area (A) in column (4) of
(2)
0';;- --'- Table A19.l would then equal the true stringer
area plus its effective skin times the factors
B ----3b ~A KR. The procedure tram this point would be the
same as discussed before. Thus shear lag ap-
Fig. Al9.37 proximations can be handled quite easily by
~odifylng the strir.ger areas. USing these
A19.l9 Application of Shear Lag Approximation to modified stringer areas. the true total loads
Wing with Cut-Out. in the stringers are obtained. The true stresses
equal these loads divided by the true stringer
?ig. A19.38 shows the top ot a multiple area, not the mOdified area.
stringer wing wh1ch includes a cut-out in the
surface. The str1~gers (5), (6) and (7) must A19. 20 Approximate Shear Lag Effect in Beam Regions
be discont1nued through the cut-out region. where Large Concentrated Loads are Applied.
3 2 Wing and fuselage structures are otten re-
I I
D :: :~ I I /Cut-out
II
quired to resist large concentrated torces as
for example power plant reactions, landing gear
reactions, etc. To illustrate, Fig. A19.39
'r--! I
I
r represents a landing condition, with vertical
<, load. The wing is a box beam with 7 stringers
-'H,h ~I
a1.-/
1.-/ <,
'
c ----... .-/ Top Surface A
,--' "I <,
b
l ~ <, Effective-
-, 1-3b_ f---3b -.I ness Tri-
-, an 1e
I
:-- I , b -;1
-, '-'
1 I
3 2

Fig. Al9.38
II I j I ,
~ Dead wt.
inertia forces
It 1s assumed that the ef:ectiveness of
these 3 inter~~pted stringers 1s given by the
triangles in the figure. At beam section 1-1 ~
~ Bending Mom.
these strir~ers have zero end load. The
stringer load is then assumed to increase IV~
linearly to full effectiveness when it inter-
sects the 51des at this triangle whose height Fig. A19.39
A19.26 ANALYSIS OF WING STRUCTURES

I A
and flange nemoera . Fig. (a) shows the bending
moment diagram due to ~he landing gear react~cn 6"
T
cr-----......
C B
alone. The internal resistance to :his ~ending 20"~
moment cannot be uni:orm on a beam section

Tn
adjacent to section A-A because of the shear
strain in the sheet panels or Nhat is called
shear lag effect. To epproximate this stringer

';
effectiveness, a shear lag trian~le of length
3b 1s assumed, and the same procedure as
discussed in the ~revlous article on cut-outs
is used in finding the longitudinal stresses.
ISO" J I I
It should be understood that the bending ~oments I I

l ! s
due to the distributed forces on the wing such
as air loads and dead weight inertia loads are I I
not included in the shear lag considerations,
only the forces tr~t are applied at concentrated I
points on the structure and must be distributed
into the beam. A side load on the gear or -I 9" !- 21" I 60"
power plant would produce a localized couple -r/ Cell
plUS an axial force besides a shear force as in
Fig. AI9.39. The resistance to this couple and l ~-.a:.t,,-II_..EO.. -=====-
axial force would likewise be based on the
effectiveness triangle in Fig. AI9.39. Fig. A19.41

A19.21 Approximation of Shear Lag Effect for Sudden
Change in Stringer- Area
Stringers of one size are otten spliced to
stringers of smaller size thus creating ~ dis-
continuity because of the SUdden change in
stringer area.
Fig. A19.40 shows the stringer arrangement
in a typical Sheet-stringer wing. Stringer B
1s spliced at paint indicated. The stringer
area A. is decreased SUddenly by s~licing into
a stringer with less area A1 •
Top Surface of Wing
•
I Single
-3b 3b"'"
Fig. A19.40
To approximate the shear lag effect,
assume the area of stri~er 8 at splice paint
to be the average area of the two sides or
intensity acting upward in the z direct10n
and 0.25 lb./in.a average Intiens i ty acting
rearward in the x direction. The center of
pressure for z forces is on the 25 percent
ot chord line measured f~om the leading edge
edge and at mid-height of spar AB tor the
x air forces. Assume the 3 stringers A,
8, C develop the en~ire resistance to ex-
ternal bending moments. Find axial ~oads
1n stringers A, 8, C and the shear :low in
the 3 sheet panels of cell (1) at wing
stations located 50~, 100~ and 150" :ram
wing tip. Consider structure to rear of
cell (1) as only carrying a1rloads tcnvard
to cell (1) and not resisting'wing torsion
or bendi:Ig.
Fig. Al9.42
Pin
FIttings
'c
I

(A 1 + A.)/2. This average area is then assumed I

to taper to As and A1 at a distance 3b tram the
splice point. "The shear lag effectiveness
factor R will therefore be greater than 1.0 on
the side toward the smaller stringer A, and
less than one on the side toward the stringer
with tfie greater area As, since the average
area was used tor the splice point.
A19.22 Problems
(1) Fig. A19.41 shows a cantilever, 3 stringer,
single cell wing. It is subjected to a
distributed airload or 2 lb./in.~ average
S I
(2) Fig. A19.42 shows a monoplane wing w1th
one external brace strut. ~he wing is
fastened to !uselage by Single pins at
points (a) and (b). ':'he r i t t tng at (b) 1s
designed to take of: drag reaction. The
alrloads are Wz = 40 Ib./in. of wing span,
with center of pressure at 30 percent of
chord tram leading edge and acting upward,
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A19.27

and Wx = 5 lb./in., acting to rear and Table A gives the str~nger areas at sta-
located at mid-depth of wir~. Find re- tions 0 and 150. Assume stringers have
actions at points (a), (b) and (d). Find linear variation in area be~Neen these two
axial loads on front and rear spars. Find stations. Use 3Dt as effective skin with
primary bending moments on front spar. compression stringers.
Find shear flow on webs and walls. Neilect
structure forward of front spar and rear- Find axial loads in stringers at stations
ward of rear spar. 150 and 130 and detennlne shear flow
system at station 150.
(3) Fig. A19.43 shows a portion of a single
cell - multiple stringer cantilever wing. (4) Same as problem (3) but add an internal
T~e external air loads are: web of .04 thickness connecting stringers
(3) and (8).
Wz = 100 lb./in. acting upward and whose
center at pressure is along a y axis coin- (5) Same as problem. (4) but add a leading edge
ciding with stringer (3). cell with radius equal to one-half the
front spar depth. Take skin thickness as
Wx = 6 lb./in. aC4ing to rear and located .04 inches.
at mid-depth of wing.

TABLE A

'D' 6
J...
f-3• 3 +-" . 3j
-12
5
..i

1 4
<1-- 12 -+4 i
I I
I I
I

,
I
I

I I

I
I
I

,.032

T
~-J
T
1 ~-.050
12" .032~~ 10"

C 025
1
~5"" 52 t 5 +5j 150" I
STA.O
STA.150

Fig. A19.43
 , - ,- _ .' -'.0"- - - .-. . '~"'.::1

. ... . -. . ' . . . . .., ,',. " ' , . . .. ,'t,,,

~.t.l.

A19.28 ANALYSIS OF WING STRUCTURES

North American Aviation FJ3 "Fury Jet". View Shows Bottom Wing Skin.
Note Integral Construction of Skin and Stringers.

North American Aviation F-1OO "Super sabre" Wing. View Shows End
of Outer Wing Panel. Note Thick Skin.
,....

Douglas DC-a Wing. View ShOWing Lower Surface of Outer Wing Panel
Between Center and Rear Spar.
 CHAPTER A20
INTRODUCTION TO FUSElAGE STRESS ANALYSIS

A20.1 General. In general the pur-pose of an air- to Chapter A5, where ex&~ple calculations of
plane 1s to trans?ort a commercial payload or a fuselage shears and moments are presentee.
military useful load. The commercial payload
of a ~odern airliner may be 100 or ~ore pass- The basic fuselage structure 1s essentially
engers and their baggage. These passengers a single cell thin walled tube with many trans-
must be ~ransported safely and comfortably. verse fr&~es or rin~s and longitudinal st~ingers
For example, an airliner flies at high altitudes to prOVide a combined structure which can absorb
where temperatures may be far below zero and and transmit the many concentrated and distri-
where the air denSity 1s such as not to sustain buted applied forces safely and efficiently.
human life. These facts mean that the body The fuselage is essentially a beam structure
which car~les the passengers must be heated, subjected to bending, torsional and axial
ventilated and pressurized to prOVide the forces. The ideal fuselage structure would be
necessary safety. Air travel ~ust be acceptaJle one rree of cut-outs or dt soont.tnut ties, how-
to the passengers," thus the airplane body must ever a practical fuselage must have ~y cut-
shield the passengers from excessive noise and outs. Fig. (a) shows the basic interior fuse-
vibration. and further.nore efficient, restful lage structure of a small airplane with skin
and attractive furnishings must be provided to removed. It consists of transverse frames and
~ke travel enroute comfortable and enjoyable. longitudinal stringers. Photographs 1, 2 and 3
The portion of the airplane which houses the illustrate fuselage construction of late model
passengers on payload is referred to as the large aircraft.
fuselage. Fuselages vary greatly in size and
confi~uration. For example, the ~uselage of a
supersonic military airplane ~y house only one
passenger, the pilot, the remainder of the
fuselage interior Spdce being used to house the
,ower plant, to prOVide retracting space for
landing gear, and to house the many mechanical
and electronic installations which are ~ecessary
to :ly the airplane and carry out the various
operations for whlcn the airplane was designed
to acccmplish. Many groups of engineers with
various backgrounds of training and experience
are therefore concerned with the deSign of the
fuselage. The structures engineer plays a very
L~portant part because he is responsible :or
the strength, rigidity and light weight of the
fuselage structure. Fig. (a)

A20. 2 Loads. Basic Structure.

The wing, being the lifting bOdy is sub-

jected to large distributed surface air forces,
whereas the fuselage 1s subjected to relatively
small surface air forces. The fuSelage is sub-
jected to large concentrated forces such as the
wing reactions, landing gear reactions, empen-
nage reactions, etc. In addition the fuselage
houses ~y items of various sizes and weights
which therefore SUbject the fusela~e to large
mer-t ra forces. In addition, because of high
altitUde flight, the fuselage ~ust withstand
inter~al ~ressures. and to ~andle these i~ternal
pressures efficiently requires a circular cross-
section or a combination of circular elements.
The student should refer to Chapter A4 for PHaro. NO.1
further ciscussion of loads on aircraft and also Fuselage Construction of Fairchild F-27 Transport

A20.1
A20.2 FUSELAGE STRESS ANALYSIS

PHOTO. NO.2
View LOOking Inside of Rear Portion of Fuse lage
of Beechcraft TWin- Bonanza Airplane.

PHOTO. NO.3
Fuselage Construction of Boeing
707 Jet Airliner.

(FOR GENERAL DETAILS OF DOUGLAS DC-a FUSELAGE CONSTRUCTION sEE PAGE A15. 32)
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
A20.3 Stress AnalySiS Methods. Effective Cross-
Section.
It is common practice to use the Si~plified
be~ theory 1~ calculating the stresses in the
skin and stringers of a ~uselage structure. If
the fuselage is pressurized, the stresses in the
skin due to this lnterr~l pressure ~ust be added
to the stresses which resist the flight loads.
In wings the skin in the ~iddle region of the
airfoil 1s relatively flat and thus the skin is
usually considered as made up of flat sheet
panels. In fuselages, however, the skin is
curved and curved sheet panels have a higher
critical compressive buckling stress tr~n flat
panels of the same size and thickness. In
small airplanes, the radius of curvature of the
fuselage skin is relatively small and thus the
additional buckling strength due to this curva-
ture may be appreciable. A simple procedure ot
approximately including the effect of sheet
curvature will now be explained.
Fig. A20.l illustrates a distributed
stringer type of fuselage section. Assume that
external loads are applied which produce bending ling stress ocr to the bending stress crt exist-
of the be~ about the Y axis with compression on ing at that point for bending of the fuselage
the ~pper portion of the cell.
',"
y y
_ - J L N. A.
Fig. A20.1
UP to the pOint of buckling of the curved sheet
between the skin stringers, all the material in
the beam section can be considered fully affec-
tive and the bending stresses can be computed
by the general flexure formula CO ~ Mz/I y ,
where I y is the centroldal ~oment of inertia of
the entire section. Wnen a bending compression
stress is reached which causes the c~red sheet
between stringers to buckle, are-arrangement
takes place in the stress distr~bution on the
section as a whole. Theory as well as experi-
mental results indicate that the ultlmat) com-
pressive strength of a curved sheet with edge
stringers can be approximated by the following
c~o assumptiOns.
A20.3
(1) A small width of sheet w l on each side of
the attachment line of skin to stringer ~s
considered as carrying the same compressive
stress as the stringer, as was discussed in
Chapter A19. These effective sheet widths
""l are shown as the blackened elements ad-
jacent to the stringers on the compressive
side in Fig. A20.1.
(2) The remainder of the curved sheet between
stringers, namely, b-(w l + w.) carries a
maxtmum compressive stress Ocr = .3 E t/r.
This value for ocr is conservative. E is
the modulUS of elasticity of the skin
material, t the skin thiCkness, and r the
radius of curvature of the skin. These
curved sheet elements are shown by the
hatched skin lengths in Fig. A20.l.
Since the thin curved skin between the
stringers normally buckles under a compreSSive
stress far below the buckling strength of the
stringers, the curved sheet is treated as an
element with varying effective thickness which
dependS on the ratio of the curved sheet buck-
section. Hence the effective sheet thickness
for the c~led sheet panels can be written,
- - - - (1)
or an effective area can be written
----(2)
where b l 15 the width at curved sheet between
the effective sheet widths w1 , w., etc. (See
F1g. A20.1).
To illustrate this approach in obtaining
the effective cross-section of a fuselage
section, an ex~ple problam will be presented.
The example problem will be broadened to some
extent for the purpose of intrOducing the stu-
dent to deSign procedure.
A20.4 Example Problem.
Let it be required to determine the stringer
arrangement for the approximate elliptical
Shaped :uselage section shown in Fig. A20.2.
The following data will be assumed: -
Design bending moment about y axis = 160,000
in. lb. (pr-oduc tng compression on upper portion).
Zee stringers, one inch deep and with an area
equal to 0.12 sq. in. shall be used.
The ultimate compressive strength of the zee
stri~ger plus its effective skin and a ler~th
equal to fuselage frw~e spacing is assumed to
be 32000 psi. T~e skin thickness is .032 and
all material is (2024) al~inum alloy with E =
lO,3CO ,OCO psi. The ~uselage stringers are to
be symmetrical about section center lines.
A20.4 FUSELAGE STRESS ANALYSIS

'II = 1.7x .032'/10,300,000/32000 = .975 ~:1.

wld~h of .975/.032 = 30.5 sheet

~.75 Which eqlials a
t~icknesses. Since the bending stress decreases

...;'
x'" 4
3 2 1
to zero as the neutral axis is lpproached, ~~d
since the curved sheet Jetween the Z stringers
can carry loads up to its b~~kling stre~g~h, a
;Z 6
5 preliminary value of e r r ec t ive width 'Ii =
4:0t
will be ass~ed acting with each stringer. Thus
" total area of st~inger plus effective skin
"t: 8
7

Height", 50"
equals 0.12 ... 40 x .032:01 = 0.16 sq. 1::1. The
number of st~ingers required is theref)re 2.00/

-+,0
"

e- 11
9 Width '" 30"
y' -'-......-a--lo-=~...J.---+------+-
;::=F=-::-::...:=--=::j:=N~.~A.Trial2
Z", .135"
15
16
1 32000 pS1. The procedure f~om this ;oint is
18
19
20
21 2223
Solution:
The first thing to do Is to determine ap-
proximately how many Z stringers will be re-
quired so that a section can be obtained to
work with. Since the internal resisting ~om­
ent must equal the external bending moment, one
can guess at the internal resisting couple in
terms ot total compressive flange stress and an
effective internal couple arm.
For elliptical and circular sections with
distributed flange ~terial, the approxi~te
effective resistir~ ar.n of the internal couple
can be taken as 0.75 times the height ~, and
the average tension or compressive stress as
2/3 the maximum stress. ~hus equating the
external bending ~ament to the internal resist-
ing moment an approximate total area Ac for the
compreSSive side of the fuselage section can be
obtained.
=: Ac (.67 0b)(.75 h), Whence,
= 160000
.75x50x .67 x 32000
Part of this total area is pr~vided by the
effective skin area. The ef~ect~ve width to
use with each rivet line equals w = Ct ,;E/crst.
We will take C = 1.7 which is a commonly used
va Lue ,
0.16 = 12.
Fig. A20.2 shows how the stringers were
placed to give 12 stringers on the top hal:.
y' Since the skin on the lower half is in tension
and therefore fully effective, the neutral ~xis
wtll fall below the center lir.e and thus the
N. A. Trial 1
two stringers on the center line will be con-
sidered as part of the requi~ed 12 stringers.
A t~selage cross-section has ~OW been obtai~ed.
The desired :inal result is that the naximum
Fig. A20.2 compressive stress will be near but nat Over
still a trial and error ~rocess since the
Symmetrical about effective sheet on the compressive siGe depends
the Z axis
on the ma~itude of the compressive bending
stress which in turn 1s influenced by the ~ount
Location of stringers or effective sheet ~~d the buck11ng load carried
below y' - Y' same as by the curved sheet.
above Y' - Y'.
Usin~ the preliminary stringer ~rrangement
ot Fig. A20.2, Tables A20.1 and A20.2 give the
calculation of the e~fect~ve ~ornent of inertia
of the section about the horizontal neutral
axiS, Table A20.1 deals with the stringers and
the effective sheet elements and Table A20.2
deals with the curved buckled sheet elements.
In the trial No. 1, ~he following assump-
tions are made:
(1) A width of 30 thicknesses of skin act
with each stringer en the upper or ccmpress1ve
side.
(2) The area of the curved sheet betwee~
the effective sheet widths as fOW1d in (1) is
modified to give ~~ eftective area by ~ulti­
plying by a K ~actor of 0cr/ab. where a cr 1s
the buckling compressive stress and 00 is the
bending stress at the center of the curved sheet
element assuming 32000 at the extreme upper
fiber of the beam section and zero at the hori-
zontal center line, With linear variation in
between these points.
• 2.0 sq In
... NOTE: Since the entire skin on ~he lower half
t s e r rec t tve , a mor-e logical aasumpt t on '.... cu.Ld
be to guess at the l~cation of the neutral axis
and use a variation of ab between the neutral
axis and the extreme fiber. ~hiS approaCh will
not be used in this example.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A20.5

TABLE A20.1

TRIAL NO. 1 TRIAL '0. 2

1 2 3 • 5
! , 7
• 9 10 11 12

-.
E:<planatory Notes for Table Ala.l

.-.
•, •
;1
, ::: 0 :;~
'N
N =~.o ,. • . I TRIAL NO. 1

••
Area I
az· 2 , ~, •0 ,- .,
- ·.,-- ·..-.. ••.. •• " I "'
••
•• ., I
,z2
Col. 1 For DUlllbering of sUff-
iiiei='S'iDd sh.ee't elellMlDts, .ee

.. '....'. '. .
-, a
aq.i:n.
A~
z'
."
, . >. Fig. xao.a.
.~
• ," b• • Col. 2 Stiffener area .12 •
" 30 • • 032 x .032 .18
re. s'ti!feners
, ,
I •"
. :1e .15
• lS
.15
24..21

22'°1
18.2
3.631
3.30
2.73
87.'
72.6 I 25.38
27.58 _30020 1.0
_27700
4.9.71 21.58 -23550
1.03
1.12 I
.032
.033
.036
.152
.153
.156
4..19
3.88
3.36
115.4.
98.5
72.4.
2.
' 10 • Below
the centerline each

010
• .15
.15
13.3
6.9
1, 99
1. 04
26.4 16.68 _18200
7.2 10.28 -11200
1.28
1.66
.04.1
.053
.161
.173
2.69
1.78
44.7
18.3
s'titfener co~idel"ed
aC'ting sepal"ate1y. The '"
.:: 12
~ 13
.12 I ° I ° ° 3.38
2.3 -0.18
3680 - 2.88 • 092 .212
.226
0.72
-_0.42
.04
2.'
0. T
entire skin be'tween
conllider_
." I
.224
.1201
-3.2
_6.9
-0.72
-0.83 5.7 _3.52 c •
••
.120 1.5
s'tifteners
ed as uui't. , "
::: 15
-16
.,7
.224.
.12
.160
-10.1
-13.3
-15.8
_2.28
_1.60
_2.53
23.0 _6.72
21. 3 -9.92
40.0 ! 12.42
••
• ·,•
•
.226
.12
.160
-1, 52
_L19
_L99
10.2
11.8
24.7
CoL J All z' m
~ e' to horizontal cea-,
,~

terlille aJtis.
llIeaaur-

.1. .12 -18.2 _2.19 39.81-14.·82 " .12 _1.78 26.4

; 19
'>120
':21
::: 22
.16
.12 .20.31
_22.0
.16 _23.7
.12 -24.2
-3.24
_2.64
_3.80
.. 2.90
65.81 16.92
58.0i-18.62
90.0 20 32
70.2 1_20.82
•
~

~
·• -"••
~
.16
.12
.18
.12
_2.70
_2.24
_3.25
-2.50
45.6
41.6
66.0
52.1
TRIAL NO. 2
~ z s dis'tallce 'cc neutral
aXis ,.
found fro. l"l!-
suits of Trial No. 1.
.23 .088 24.91 _2. 19 1 54.5 -21. 52 • • .088 _1.89 4.0.7 Col. 7
Col. 8
~h • 1600000 z/1470
~ effective widtb based
aUK. 2.4.9 -12.23 714..3j 2.627 _2.86 .73 .n S'treS8 1111 Col. 7

TABLE A20.2

TRIAL NO TRIAL NO 2
1
;

-.••.."•,
I 2 3
• 5
. .-••• 6
·
7 9 10 11 12 13

- .- • •- • ..• --"...
,
• •-
"~ 15 16
·•
17 18

...- "., ·-,". -.

_b N
~ <,
, • "'
•• " .N
...
, .- ..
e
, • -
;l
.
<, Ofl 0 .
·
aa ' az,2 oz .,2
" r •! z'
"
-t • ... •• •• •
<.

--. -•
0 'b

-•••• ~
•
, • •
oOM"
• • <
0

• N

_30800
>"
ON
• >< ,

.021 .sa 16.7

l;~:;: -25800
1 2.27 .032 11 -9000 -31900 .282 .020 50 12.4 .292 12.,.
3 4.G4 .032 11 -9000 -30300 .296 .038 ".91°.
23.70.90 21. a _29500 .305 3.99 .039 1.06 28.7
"~ 5 .032 _4510 -26000 .173 I .022 20.30.45 9.1 23.68 .174 3.93 I
.022 0.52 12.3
•c
: i
".04
4.04 .032 ""ae _2600
_2600
_20200
_12900
.128 .016 15.8 0.25 3.9 19.18
10.1 0.40 •• 0
_20900
_14700
.124
.177
13.80
5.53
.01510.29 5.5
.031 0.42 5.7
3.2 0.42 1.3 1 i1~:::
6.04 .032 .201 .039
•• 11 6.52 .032 ae -2600 _ 4100 .63 .131 - 7200 .362 4..73 .0155 0.36 2.'
j TOfALS .266 2.92 52.0 TOTALS .183 3.24 71.3

Exp taoato rv Notes to r Table Azu.z Explanatory Notes for Table Azu.z
I Trial xo.
1 Trial xe, 2
Col. 1, 2, 3.
Col. 5
Col. 6
• (see Fig. Ala.l for mean1ng of terms)
.E • 10,300,000 for alW11inwa alloys
Cb val"ieS as :II. s'tl"a1ght line fl"Olll 32000
top of cell to Z8l"O at centel"11n8
=
distance frOlll c81l'tro1d of element to
., Col. 12

Col. 13
Col. 14
z - distance to ~eu'tl"ll.l ax1.
as found ill results of Trial
110• 1
Ob ; 1600000 z/1470
based on s'tress (1) of Col. 13
Col. 9 z'
centerline ax1s of cell
i

Results of Trial !l'o. :!. Results of Trial !fo. 2

Considering resuLts of both Tables and multiplying by l since Total effective al"ea (2.627 + .183)2 • 5.62
only one half of cell was considered:
~az. (_0I.86 + 3.24.)2.0.76
ia total effec'tive al"ea • (2.4.9 + .266)2
Z ... 76/5.62 • • 135" above N.A. of trial 1,
_ 5.51 sq. in.
rNA .. 2(673 + 71.3) - 5.62 Jt .135' ~ 14.89
£,az, • (_12.23 + 2.92)2. _18.82

i ~Zaz'/Za .. _18.62/5.51 .. -3.38"

2• 6
INA • 2(714.3 ... 52) _ 5.51 x 3.38 14.70 in.
A20.6 FUSELAGE STRESS ANALYSIS

Due to the s~~etry of the section, Tables

A20.l and A20.2 ~~'le c~lculatior.s for o~ly one- q~f ,
.., ....
.

haIr of the material, ttus the results are

~ultiplied by two. General explanatory ~otes
are given below each table.
The results of trial No. I give a neutral
axis 3.38 n below the center line and a moment of
inertia of 1470 in.'. In Trial No.2, the ef-
fective sheet widths are based on the ~oment of
inertia of 1470. The results of trial No.2
give a moment of inertia of 1489 in.' with a
neutral axis .135 n above the first location. If
a third trial were used, making use of the 14S9
moment of inertia, the change would ce quite
small since the effect of a small change in
stress on the effective sheet width is negligi-
ble.
The compressive stress on stringer No.2
using the resulting moment of inertia and
neutral axis location, therefore beccmes
"b = l'lyZ/I Y = 160000 X 27 •45/1459
The allowable stress was 32000, hence the
margin ot safety is (32000/29500) - 1 = .OS or
eight percent. If a s~aller ~rgin of safety
was desired some ~aterlal would be eli~inated
and the calculations of Tables A20.l and A20.2
would be repeated.
Calculation of Shear stress in Skin at Neutral
Axis
The equation for the shear flow q
point on the skin is,
q. qo - Vz zaz - -- - - - - - -
Iy
Due to symmetry of cross-section about the Z
axis the shear flow qo 1s zero at a point on
the center line Z axis. The summation of the
term az between a point on the Z axis and the
neutral axis is given in Table A20.3. The
values of areas (a) and arms (z) are taken
trom Tables A20.l and A20.2.
= 0+ i~& 19.57 = .0132 Vz It./~n.
T~e shear stress ~ = q/t .0132 Vz / . 032 = .413
Vz
The average shear stress on ~~e secticn would
be 'Tav' = Vz/ 2ht = Vz/2x30x .032 = .312 Vz.
Thus for this sr2pe 0: crass-sectien and
stringer a~angement the ~axi~um shear stress
is .413/.312 times the average shear s:ress or
apprOXimately 4/3 times as large.
The procedure as given above is quite con-
servative relative to the true or actual margin
of safety, because a linear variation at stress
with strain has been assumed and failure of the
section is assumed to occur when the most remote
stringer reaches its ultimate compressive stress.
= 29500 Actually in a static test of a fuselage to
~si destruction, the f~selage section as a whole
will not collapse when one stri~ger ~uckles,
but will continue to take increasing ~oad ~til
other stringers have reached their ~ltixate
strength, Furthermore, in a typ i ca I ruse Iage
struc~ure, stringers of various sizes, sha~es
a~d Lherefore different compressive strengths
are used. ~~d thus to obtain a better ~easure
of the ultimate strength of a ~uselag8 section,
modifications in stress procedures are ~ace to
measure stringer effectiveness. T~is subject
waS discussed in some ceta11 in Arts. :1 and 12
~t same of Chapt er- A19. ':.'0 illustrate s'tz-tnger- effect-
iveness in fuselage bending stress analYSiS, a
simple example problem will be ,resented.
(3)
A20.5 illtimate Bending Strength of Fuselage Section.
Example Calculation.
Fig. A20.3 shows the cross-section of a
circular fuselage. The Z stringers are arranged
s~etricallY with respect to the center line
Z and X axes.
Three sizes at Z stringers are ~sed as
~llustrated in Fig. A20.4 and are labeled S.,
Sa and 8 3 , These symbols are usee. on Fig.

B1e_llt xo,
TABLE A20. 3

.. A20.3 to indicate where each type of stringer

is used. The strlr~ers en each side cf the
section are numbered 1 to 13 as shown on Fig.
(l)
'2l
.021(28.28
.152(21.58
_
_
.13)
.13) · 0.59
<1.17 A20.3. Fig. A20.5 shows a plot of the stress-
(J)
(4)

(&,
(5)
.039(27.08
.153(25.38
.022(23.68
_
_
_
.13)
.U)
.13)
· 1.05
3.86
0.52
strain curve ~or the three stringer t~?es loaded
in compression and with a col~, ler~th equal to
.156(::1.58 _ .13) 3.3<
(7) .015(19.18 _ .13) 0.28
the fuselage frame spacing. Fig. A20.5 also
'8l
(9)
• (la)
.161(16.68
.031(13. <18
_"
_
.13)
.13) · 2.66
0.<11
shoWS a tension stress-strain clagram for the
material which 1s aluminum alloy (2024). The

..
.173(l0.28 • .13) 1.75
(U)
(12)
.055(6.58
.212(3.38
-- .13)
Total
.13) · 0.35
0.69
19 67
ultimate bending strength will be calculated
~or bending which places the upper portion in
compression.
Substituting in equation (3)
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A20.7

Since the location of the neutral axis is

Skin =- .032 3
2
S, S,
- - -
.006
unknown, a location will be assumed, namely, 7
inches below the center line axis as shown in
4 S. Fig. AZO.3. The entire calculations for deter-
S, ~lning the effective moment of inertia cab best
5
be done in Table form, as shown in Table A20.4.
T ~
~'
Due to Sj~etry about the Z axis only one-half
of the structure need be considered since the
CL AxiS"" results can be multiplied by two.
7"
X -, X Column (1) lists the stringer numbers
~medY relative to location and Column (2) according
Neutral Axis
S, to types Sl' Se and 8 3 , Column (3) gives the
9 stringer area. On the tension or lower slde of
10 the section, the skin 15 all effective and that
S, area of skin halfway to each adjacent stringer
,,( '8' 11 ~ is assumed to act with stringers numbered 9 to
IQ": Y 12
13
- - - -• 0036 13 and this skin area 15 recorded in Column (5) •
Fig. a an the compressive side the skin is only
Strain partially effective. The effective width w tor
Fig. A20.3 Diagram each stringer rivet line depends on the stringer
stress. We will take the effective width w •
1.9t vE7OST. The effective area Ae will then
13 equal wt. These effective skin areas are re-
. .!! r- 16 .., 9 corded in Column (5). In solVing this equation
r:~I"" -'Lno?l~ the stringer stress crST has been taken as -36500

l'~~ Stringer 81.

B C~ [t 1: psi on stringer number (1), and then varying
linearly to zero at the neutral axis as indi-
cated in Column (4). This assumption Is not
Stringer SII Stringer S 3 true but accurate enough to obtain effective
Area =.135 Area = .18 Area = .08 skin areas. To illustrate, consider stringer
Fig. A20.4
number (1). The effective area As equals wt- =
1.9 x ,032"(10,300,000/36500)i- = ,032 sq. In,
Column (6) gives the sum ot the stringer and
effective skin areas or As + Ae.
In this example problem, the effectiveness
of the curved sheet panels be~~een the sheet
effective widths will be neglected since its in-
fluence is small. It could be included as il-
lustrated in the preViOUS example problem.
Column (7) lists the distances tram the assumed
neutral axis to the centroid ot each stringer-
skin unit. We now assume that plane sections
remain plane or a linear strain variation.
Referring to Fig. A20.S. it is noticed that when
a unit strain of .006 15 obtained in stringer
S1. type the compressive stress 1s 36500, which
represents its ultimate stress. Stringer (1)
is of Sl type and Is located tartherest from
the neutral axis. SUb. Fig. (a) of Fig. A20.3
shows the strain diagram with .006 at stringer
(1) and varying as a straight line to zero at
the neutral ~~is. Column (8) of Table A20.4
records the unit strain at each stringer cen-
troid. The true stress at each stringer pOint
due to :hese strain values is read from the
curves on Fig. A20.5 and recorded in column (9)
of the Table. It should be noted that stringer
(3) although closer to the neutral axis than
stringer (1) carries a higher stress than
stringer (1). This is pOSSible because when
stringer (1) reaches Its ~aximum-stress, it
bends but continues to hold the same stress with
A20.8 FUSELAGE STRESS ANALYSTS

TABLE A20. 4

1 2 3 4 5 6 7 8 9 10 11 12 13
Effect- Total I
stringer Stringer Linear Arm Unit Stress
tve Skin Stringer Strain K=o/cr1. KAZ' KAZ'
Area Stress Z' ATE. KA (Z = Z'I
Ar-ea Area
No. Type A, a'b A
(in) E a
Ae
1 S, 0.135 -36500 .032 .167 35. 7 -.00600 -36500 1. 00 . 167 5.96 213 I
I
2 S, 0.135 -34700 .034 .169 33.8 -.0057 -36500 1. 05 .178 6.00 203

3 S. 0.180 -31000 .036 .216 30.3 -.0051 I -39100 1. 26 .272 8.25 250 I

4 S, 0.135 -26600 .039 .174 26.0 -.0048 -36000 I 1. 35

I .235 6. 10 159 I

I
5 S, 0.080 -20500 .044 .124 20. 1 ·.0034 -31500 1. 54 . 191 3.85 I 77 I
I I
6 S, 0.080 -13400 .054 .134 13.7 -.0023 -24000 1. 79 .240
I 3.30 45 I

7 S, 0.080 -7150 .074 .154 7. a -.0012 -12500 1. 75 .270 I 1. 89 13

8 S, 0.080 0 a . 080 10.0 a a 1.0 .080 i a a
,

9 S, O. 080 6130 .216 .296 - 6.0 .0010 10000 1. 63 .4831 -2.88 17

10 S, 0.080 12280 .216 .296 -12.0 .0020 20500 1. 67 I .494 i
I
- 5. 91
I 71 I
I
I I
11 S, 0.080 16800 .216 .296 -16. 5 .0028 30000 1. 78 .526 -8.69 144
I
12 S, 0.080 20400 .216 .296 -20.0 .0034 35000 I 1.71 , .506 -10. 10 202
13 S, 0.080 21700 .216 .296 -21. 2 .0036 38000 1. 75 , .517 1- 10. 93 232

Z 4. 159 -3. 16 1626

increasing strain, but stringer (3) which has
not reached its maxi~um strength of 39000
continues to take increasing load.
Since we wish to use the beam formula CO =
MyZ/I x in computing stresses, we must modify the
stringer areas to give a linear stress variation
since the formula is based on a linear stress
variation. The stringer modification factor K
equals the ratio of the true stress in column
(9 ) of Table to linear stress value in column
(4) or K = a/a 1.. The results are recorded in
column (10) • The modified stringer areas are
then equal to KA and are recorded in column
(ll) • Column (12) gives the first moment of
the modified areas about the assumed neutral
axiS, giving a total value ot -3.16.
The distance Z fram the assumed neutral
axis to the true neutral axis is thus,
Z=Z KAZ' (ZKA
_ -3.16 _
- 4.159 - - 0.76"
TIle true N.A. would fall about .70 inches
below assumed pOSition. The effect on total
sum at Column (13) would be negligible, thus
Table A20.4 will not be revised.
Column (13) gives the calculaticn of the
er.rect r ve moment of inertia with Z 1. being equal
to Z. The effective moment of inertia is there
tore twi ce the sum of Column (13) or 3252.
Calculation of Ultimate Resisting M~ent.
The maximum stress at the most remote
stringer which is number (1-i' is 36500. From
the beam formula,
Mx ="b1x/Z
= (36500 x 3252)/30.7 + 0.7
= 3,260,000 in. lb.
This bending strength when compared to any des1 gn
bending moment about the X axis Nould give the
margin of safety relative to benci~g strength.
If the moment of inertia had been computed
without regard to non-linear stress variation,
or in other wordS, using K equal 1 for all
stringers the neutral axis would have come out
4.9 inches below the centerline axis and the
moment of inertia would have calculated to be
2382 in. • The reSisting moment developed
would then be (36500 x 2382 )/33. 6 2,600,000 =
in.lbs. Thus the true strength is 25 percent
greater than the strength for linear stress
variation. This result explains why such
structures test overstrength if deSigned on
linear stress variation basis.
Atter stringer stresses are obtained
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A20.9

uSing the modified areas of Table A20.4, the
true stringer areas ~ust be used ~o tind the
true stringer loads, which must be used in the
shear flow analysis.
A20.5 Shear Flow Analysis for Fuselage Structures
The shear flow analysis can be made once
the effective cross-sections of the fuselage
are obtained. The procedure is the same as was
illustrated tor Wing structures in Chapter A19.
To illustrate, two example problems will be
presented.
I y = (15:1 X .Lx z ) + (13.86 0 1 1 + 1 0 . 6 1 : 1 + 5 . 7 4 011).1
x 4 = 180 in.4.
Table A20.5 gives the necessary calculations
for determining the flange bending stresses and
the net total shear load to be taken by the cell
skin. Since the cell 1s tapered, the stringers
have a z component, thus the stringer axial
loads help resist the external shear load. The
summation of column (8) ot Table AZO.5 gives
-333.4 lb. for a summation for halt the fuselage
section.

Example Problem 1. Symmetrical Tapered Section. Hence, net web shear at station 0 equals:

Fig. A20.6 shows a portion of a tapered VWeb == Vext • + Vtlange = 2000 + (2 x -333.4)
circular shaped fuselage structure that might
be representative of the rear portion of a -= 1333.2 lb.
fuselage for a small airplane. Since this
example is only for the purpose of illustrating The results in this particular problem show
shear flow analysis, it will be assumed that that at station 0 the flange stringer system re-
the 16 stringers are the only effective mater- sists one third of the external shear load. At
ial. In an actual stress analysiS, the effect- station 150 the web system will resist the en-
ive cross-section would have to be used as tire external shear load at 2000 lb. since the
illustrated in previOUS articles A20.3 to A20.5. load in the stringers is zero.

The problem will be to deternine the In actual deSign the net Neb shear should
stringer stresses and the skin shear flow stress be used since in many cases it will decrease
system at Station (0) under a given load system the sheet thiCkness reqUired one or more gauges.
at Station (150) as shown in Fig. A20.S. Calculation of Flexural Shear Flow.
Solution No. 1 - Solution by Considering Beam Properties
at Only One Section. VZ(web) Z a Z 1333.2
q = qo - ly
= qo - 180 Z az
If the Change in longitUdinal stringer or
flange material is fairly unitor.n this method = qo - 7.40 Z az -- (A)--
can be used nith little error i~ the resulting
shear tlow stresses. Due to symmetry ot the sectIon about the Z
axiS, the flexural shear tlow in the web at the
Moment of inertia of section at station (0) center line is zero. Therefore, qo will be taken
about centroidal Y axis: as zero and the summation in equation (A) will

Area of all Stringers :E • 1" 2000t

Frame spacing '" 15"
16

15 1
\
14 _stiffeners

13

12

11

10 S
Section at Station
Section at Station 0
"

STA.O
150"
Length not to scale
.\
STA.150
150

Fig. A20.5
A20.10 FUSELAGE STRESS ANALYSIS

TABLE A20. 5 q3~ = -7.06 + 23.66 ~ 16.60

2
• 6 7 8 9
q~8 = -7.06 + 27.91 = 20.85 etc.
sue- Arm rea O'b '" Px=°b
a dz
ener -1667z (3)(') Fig. A20.7 shews the results i~ graphical
No.
z a
(paf] lba.
dX tor,n, on the left side o~ the section the shears
are of the same Sign and t~erefore aGd together.
1 15 ·.05 -25000 -1250 -.0333 0 -41. 6 o
2 13.86 .10 -23100 -2310 -.0308 .0127 -71.3 -29.4
3 10.61 .10 -17700 -1770 -.0236-.023 -41. 7 -41. 7
•5 5.7
o
.10
. 10
- 9550
0
- 955
0
-.0127-.030
o -.033
-12.1 -29.5
0 o
5 - 5.7 . 10 9550 955 .0127-.030 -12.1 29.5
7 ~10.61 .10 17700 1770 .0236 -.023 -41. 7 41. 7
8 -13.8 .10 23100 2310 .0308 -.012 -71. 3 29.4
9 -15 ".05 25000 1250 .0333 0 -41. 6 o
Shear taken by stringers", -333.4 0

·1/2 of stringer one is assumed acting with each hail of

cell.

NOTES:
Col. 4 ab'" -Mz/l y '" -2000 x 150 x z/lBO = -1667z

Col. 5 Total x component of load in stringer member.

For practical purposes, it equals axial load in
stringers since cosine of a small angle is
practically one.

Col. 6 The slope of the stringers in the z and y

directions can be calculated from the dimen-
sions at the two end sections and the length of
the cell. (see Fig. A20.6)
Col. 8 The in plane components of the stringer axial
&: 9 loads at stancn O.
start with stringer (1)
Fig. A20.7
Solution No.2. Shear Flow by Change in Stringer Loads
Between Adjacent Stations. t.ll' Method.
The shear ~low will be calculated by C8n-
siderlng the change in the ~x~a: load in the
lor.gitudlnal str1~gers between fuselage sec-
tions at stations (0) and (30).

Fig. A20.8 shews the bea~ section at sta-

qu = 0- 7.40 x .05 x 15 = - 5.55 lb./in. tion 30, the st~inger areas jeing the Saffi8 as
at station 0, but the sec t t cn as a wno Le is
qUi = - 5.55 - 7.40 x .1 x 13.86· - 15.80
smaller due to the taper of t~e cell.
qa~ = - 15.80 - 7.40 x .1 x 10.61 =- 23.66 1 2
area of each
qu =- 23.66 - 7.40 x .1 x 5.74 =- 27.91 stringer"" . 10

The torsional ~oment T about the centroid

ot the section at station (0) equals 5 x 2000 ~
10000 in. lb. (clockwise Nhen looking toward
station 150). Due to the symmetry of t~e sec-
tton at station 0, the in-plane components ot
the stringer loads produce zero ~oment about
the section centroid.
Table A20.6 gives the calculations tor the
For eqUilibrium a const~~t inte~~l shear tle~Jral shear system. ~he procedure ~s the
flow q 1. is necessary to make Z Mx :: 0 same as illustrated :or wing struct~res in
Chapter A19.
2A =-=--=""'
a x n "x""'-Ls
="7« = -7.06 lb ./~n.
-10000
q ,. =-1 Comparing the r-esu.Lt s ot column 13 wi th the
flexural shear flow as ~ound by sol~tion No.1,
Adding the torsional shear- f Low q l.. ':0 t.ne we find the second sol~ticn gives a ~axi~w~
flexural shear flow Q, the ~ollowlng results shear flow of 28.95 lb./~n. aga~~st ~ value cf
are obtained: 27.21 for the first soluticn. T~e first ~ethod
deaLs wt th the pr-cper-t i es at only ov:e section
qu = -7.06 + 5.55 =-1.51 Ib./1:1. and this cannot include the etfect~ of change
q." = -7.06 .,.. 15.80 = 8.74 in ~anent of inertia on the shear ~low. The
 \
ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AZO.11

TABLE A20. 6

1 2 3 4 5 6 7 6 9 10 11 12 13

Sta. 0 Sta. 30 Sta. 0 S13. 30 Strtnzer Load ~P Panel Flexural Shear

Area Area Sta. a Sta. 30 crb'" °b-;; Sta, 0 sta. JU 30 Taper Flow
Stringer dPK
No. a a
sq. in. sq. in.
Arm Arm
,z
-16672 -16222 P x '"
(psi. ) (pet. ) "b a
P x ' - (Col. 8 - Col. 9) Corr.
"b a 30 Factor
ao q=Z~ 30
K lb. lin.

1 .05 . os 15. 00 14.00 ·25000 -21300 -1250 -1085 6.17 .935 5. 76

S. 76
2 .10 .10 13.86 12.93 -23100 -19700 -2310 -1970 11. 33 .935 10.60
16.36
3 .10 .10 10.61 9.90 -17700 -15050 ·1770 -1505 8.83 .935 8.25
24.61
4 .10 .10 5.74 5.36 - 9550 - 8155 - 955 - 815 4.65 .935 4.34
28.95
5 .10 .10 0 0 o· 0 0 0 0 0 0
28.95
5 .10 .10 - 5.74 - 5.36 9550 8155 955 815 - 4.65 .935 - 4.34 24.61
7 .10 .10 -10.61 - 9.90 17700 15050 1770 1505 - 8.83 .935 - 8.25 16.38
8 .10 .10 -13.86 -12.93 23100 19700 2310 1970 ·1l.33 .935 -10.60
5.76
9 .05 .05 -15.00 -14.0 25000 21300 1250 1065 - 6. 17 .935 - 5.76

NOTES:

Col. 6 O'b = -2000 x 150 z/180 '" -1067 z

Col. 7 Cib ~ -2000 x 120 z/1S7. 2 = -1522 z

Col. 10 Change in axial load in each stringer between stations 0 and 30 divided by distance between Stations. This
result represents the average shear flow induced by the loading up of each stringer between stations 0 and 30.

Col. 11 The width of a skin panel at Station 0 is 5.88 inches and 5. 5 inches at Station 30. The shear flow on the edge
of the panels at Station 0 equals (5.5/5. a8) AP/30. (See Art. AlS. 18 of Chapter A15 for explanation). This
refinement is usually neglected and the average values as given in Col. 10 are used which are conservative.

Col. 13 Due to symmetry of structure, the shear flow is zero on z axis. Thus shear flow at any station equals the
progressive summation of the shear flow values in Col. 12.

second method is recommended tor practical will assume the stringers are the only effective
analysis procedure. material. In actual design practice the effect-
iveness of the skin and each stringer would have
Since the section is symmetrical, there are to be considered as explained in Articles A20.4
no moments i~Guced by the in-plane components of and 5.
the stringer forces at station O.
The problem will be to determine the
The torsional sh~ar flow forces are the stringer stresses and the skin shear flow val-
same as in solution ~ethod No. 1 and these are ues at station (0) due to the given external
added to the values ot column 13 of Table A20.6 loads ot ?z = 4000 lb •• ?y = 1000 lb. and Px =
and give a pattern similar to ~ig. A20.7. 1500 acting at station (150) as .snown in Ff.g ,
A20.9.
A20.7 Example Problem. Tapered Circular Fuselage
with Unsymmetrical Stringer Areas. SOLUTION:
Fuselage cross-sections are seldom all sym- Since Ne choose to use the ~ method in
metlical relative to stringer and skin areas finding the shear flow system at station (0).
because the practical fuselage has cut-outs such we will find the stringer loads at two stations,
as dcors, etC. To illustrate the unsymmetrical namely, station (0) and station (30). The first
case a simplified case will be presented. steo is to find the moment of inertia of each
fuselage section about centroidal Z and y axes
Fig. A20.9 shows a portion of a ~apered and the product of inertia about these axes.
~~se1ag9. The stringer a~eas are such as to Table A20.7 (Col~~s 1 to 11) gives the calcU-
make the cross-sections unsymmetrical ~elatlve lations of ~he section properties for station
to oending material. Again tor Simplicity, we

(l
. ~.>~
I (J
-~
A20.12 FUSELAGE STRESS ANALYSIS

b
1- 8 .4 8 --I,

.1
Fig. A20.9
d
(.1)
.1)1 12"
'i
,c
8.48

a (.6) 1

Cep-erUpe

d 150
(.2)
(.2)
.2)
• L- x i

150" ,I Section at Sta. 150

Z' Z Sta. 0
The skin stringers are located symmetrical:y with res~ect ~8 :he
Section at Station 0 centerline axes, however' the stringer areas as g tven 1::1.( ) or.
the figure are not symmetr~cal With these axes. It is assumed in this ;roblem that
the stringers taper uniformly between the values as given for station Gane 150.
The cell would of course have interior transverse ~ra~es which are not shown on the
figure.

TABLE A20. 7

Section Properties at sta. 0 Total Stringer Loads at Sta. 0

1 2 3 4 5 8 7 a 9 10 11 12 13 14

Stringer Area Arm Arm ,,' az,2 .y' ay,2 az'y'

•• yo O'c",F/Za P s ",a(Ob-od
"b
No. a
" y' z' - z y' - y '" -1500/3.40 • (2) (12 • 13)

a .80 10.5 -12.00 6.30 66.20 -7.20 86.50 -75.50 11.36 -14.46 -15080 -441 -9312
b .10 18.98 - 8.48 1. 90 36.00 -0.85 7.20 -16.10 19.84 -10.94 -21960 -441 -2239
c
d
.10
.10
22.50
18.98
0
8.48
2.25
1. 90
50.80
36.00
-0
0.85
0
7.20
0 23.36
16.10 19.84
- 2.46
6.02
-22600
-16742
-441
-441
-2304
-1719
e .80 10.5 12.00 8.40 88.10 9.60 115.10 100.80 11. 36 9.54 - 7692 -441 -6506
f .80 -10. 5 12.00 -8.40 88.10 9.60 115. 10 -100.80 - 9.65 9. 54 11958 -441 9216
•h .20 -18.98
-22.50
8.48 -3.80
-4.50
72.00 1. 70 14.40 -32.30 -18.13 6.02 18798 -441
-4'11
3673
3,0,
.20 0 1~1. ,0 0 -21.60
i;;~~
u 0 - 2.4'
i .20 -18.98 - 8.48 -3.80 72.00 -1. 70 14.40 32.30 -18.13 -10.94 -441 2634
j .30 -10.50 -12.00 -3. 15 33. 10 -3.60 43.25 37.80 - 9. 65 -14.46 4592 -441 1246

Sum 3.40 -2.90 643.9 8.40 403.2 - 37.70 -1500

Reference Axes Z' and y' are taken as the centerline axes. General Notes:
(see FIg. A20.9)
Col. 12 0b'" 307. Or - 936. Lz
Location of centroid and transfer of properties to
centroidal axes. Col. 14 . Since the total tensile stresses equal to I
total compressive stresses in bending,
Z '" -2.90/3.40 '" -. 855" the sum of Col. 14 should equal the ex-
ternal applied normal load.
y '" 8.40/3.40 '" 2.46"

Iy '" 643.9 - 3.40 x 855 2 '" 641. 4

Iz '" 403. 2 - 3.40 x 2.46 2", 382.6

I%y'" -37.7 - 3.40 x 2. 46 x -.855", -30.55

 _.~

ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A20.13

TABLE A20.B

Section Properties at Sta. 30 Total Stringer Loads at Sta. 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14

z: y: °c"'F;Za .Ps =
Stringer Area Arm Arm az' az,2 ay' ay,2 az'y' °b
No. a z· y' a' - Z y' - y =-1500/2.98 a(Ob + Gel

a .50 9.80 -11.2 4.90 48.1 -5.60 62.8 -54.9 10.90 -13.31 -14800 -503 -7651
b .10 17.72 - 7.92 1. 77 31. 4 -0.79 6.3 -14. a 18.82 -10.03 -21090 -503 -2159
c .10 21. 00 0 2.10 44.1 0 0 0 22.10 - 2.11 -21447 -503 -2195
d . 10 17.72 7.92 1. 77 31. 4 0.79 6.3 14.0 18.82 5.81 -15647 -503 -1615
e .66 9.8 11.20 6.47 63.2 7.40 83.0 72.5 10.90 9.09 - 7088 ·503 -5011
f .66 - 9.8 11. 20 -6.47 63.2 7.40 83.0 -72.5 - 8.7 9.09 -11282 -503 7100
•
h
.20
.20
-17.72
-21. 00
7.92
0
-3.54
-4.20
62.8
,".2
1. 58
0
12.6
0
-28.0
0
-18.60
-19.90
5.81
- '.11
-17583
17::143
-503
-503
3416
,.""
i .20 -17.72 - 7.92 -3.54 62.8 -1. 58 12.6 28.0 -16.60 -10.02 12135 -503 2327
j .26 - 9.a -11.20 -2 55 25.0 - 2. 92 32.6 28.6 - a.70 -13.31 3570 -503 797

Sum 2.98 -3.29 520.2 6.28 299.2 -26.3 -1500

~
N Z'
~.:ic.
Reference Z' and Y' axes are taken as the centerline axes.
b I:~ 1. 10 d
Z = -3.29/2.98 = -1.10" .10 .1~.?
7.9 2'[ 2'
Y= 6.28/2.98 = 2.11 a
.sot- Z .66 e Section at
Station 30
x 1. 10 2 ::516. 6
9.a z'"
ly = 520.2 - 2.98
If.~l=-) i ~ l:;.:,{
Iz = 299.2 - 2.98 x 2. 112 :: 286.0 ;1.- tr y - y
,~l. Z2.11
I zy = -26.3 - 2.98 x 2.11 x -1.10:: -19.4 j
.26 • ll.~~ I
.20 .20 Fig.A20.10
•
(0) and the similar columns of ~ab1e A20.8 gives My = 4000 x 120 + 1500 X 8.10 = 492150 in. lb.
the calculations for station (30).
Mz = -1000x120+1500x2.11 = 116820 tn Lb , s

Before the bending and shear stresses can Pn = -1500 lb., Vz = 4000 lb., Vy = -1000
be calcalated, the external bending moments, lb.
shears and normal forces at stations (0) and (30)
must be known. Calculation of Bencting Stresses.

Station (0):
At station (0): -

~he bending moment about y neutral axis at

station (0) equals, where
My =Pz (150) + Px (7.85)
K1. = Iyz/(IyI Z - Iyz:il)
= 4-000x150+1500x7.85 = 511800 tn.J,b , K" = Iz/(Iy!z I y z .)
~z • Py (150) - Px (2.46) K~ = Iy/{Iyl z - I y Z Ji)
= -1000 x 150 + 1500 x 2.4-6 = -14-6310 in .Tb , SUbstituting values from Tables A20.7 and
A20.S:
The shears at station (0) are Vz = P z =
4000 Ib •. and Vy =
?y = -1000 lb. K1. = -30.55/(641.4 x 382.5 -30.55.:1) =
The normal load Pn at atat i'on (0) referred. = -30.55/244670 = -.0001248
to cen~oid of section eq~ls ZPx = -1500 lb.
K. = 382.6/244670 = .00156
In a similar ~a~~er, the values at station
(30) are. (see Fig. A20.10) K~ = 641.4/244670 = .00262
A20 14 FUSELAGE STRESS ANALYSIS

Substituting K values in equation for ab: ~cments in :he all ~n~e~nal anc ex-
)l~ne, Q~
ternal ~orces ~us~ be zero. Col~~ (7) of Table
r.00262 x-1453l0 - (-.0001248 x A20.9 ~iJes :~e ~oment of ~he flex~al shear
about tn; s potnt , (See notes and 3"ig. be Low
61.1800)J y - [00156 x 5E200 -
~ '" expjanat i on. )
':'able ""'''''.~
~

(-.0001248 x -145310)J z
TABLE A20. 9
whence SHEAR FLOW CALCULATIONS

crb = 307.0 y -936.1 z (pl~s crb is tension) 1 2 3 4 5 6 7 8 9

Station (30): q e
Stringer Px at P x at .6.Px Z~ m mq ql
No. Sta. 0 su. 30 30
K• =-19.4/(516.6 x 286 - 19.4') ,JU sqc m,

= -19.4/147620 = -.0001315 a -9312

-2239
-7651
-2159
55.37 55.37150.04 83001-52 3.37
b 2.67 58.04202.46118201-52 6.04
c -23 4
K, = 286/147620 = .001936 d -1719
- 1
-1615 347 61.67202.4612500 -52 9.67
-6506 -5011 49'83 65.14 150.04 9780 -52 13.14
e
K" = 516.6/147620 = .0035 I 9216 ...7100 -70' 53
_ 8' -7
114.97 252.0 29000 -52 62.97
44.44 150.04 6660 -52 - 7.56
I 3673 3416
ab =- [0035 c -116800 - (-.0001315 x h 3809 3489 -10'~" 35.87202.46 7270 -52 -16.13
i 2634 2327 -lO'2~ 25.20202.461 5100 -52 -25.8
492150] y - [001936 x 492150- j 1246 797 -14'97 14.97 150.04 2240 -52 -37.03
. 0 252.0 0 -521-52.0
a
(-.0001315 x 116820I] z Sum 92670 I

whence NOTES;

Col. (2) and (3) from tables

"b = 344.3y- 937.7z
A20.7 and A20. 8
Col. (4) .6.Px :: • [PXSt.a. 0
Column (12) in Tables A20.7 and A20.8 gives 'th..e
results of solving the equations for ab' - P~ta. 3~
Since an external load of 1500 lb. is act- CoL (6) m :: double areas
ing nor.nal to the sections and through the (see Fig. a).
Col. (7) mq '" moment of
section centroids, an axial compressive stress shear flow q on
crc is pr-oduced on the sections. (See CnLtzane each web element
13). The total load Ps in each strlr~er equals about 0' (Fig. a)
the area of the stringer times the ccmbined
bending and axial stresses. (See column 14 of
each table).
Calculation 01 Flexural Shear Flow q, Moments Due to In Plane Components of Stringer Loads.
Table A20.9 gives the necessary calcula-
tions to determine the shear flow at station (0) Since the stringers are not normal to the
based on the change in stringer loads between section at station (0), the stringers have ~n­
stations (0) and (30). The correction of the plane 'components which :nay produce a eoment
average shear due to the taper ln ~he skin about the lntersection of the s~etrical axes
panelS as was done in example problem (1), which has been selected as a ~oment center.
Table A20.6, column (11), is omitted in this Table A20.l0 gives the calculations for the in-
solution since it tends toward the conser~;ative plane components and their moments about paint
0' .
side. Since the ef:ective cross sec~ion is un-
symmetrical, the value of the flexural shear Moment of External Load System About Point (O').
flow qat any point is unknown thus a value rcr
q at some point 1s assumed. In Table A20.9 The 1000 lb. load at station (l50) acting
the shear flo''i q in the web a j is assumed zero. in the Y direction ~as a noment arm of 7 n about
Column (5) gives the results at other points the pOint Or of station (0).
under this assumption.
Hence external moment = 1000 x 7 = 7000
Moment of Shear Flow about Intersection of in. lb.
Centerline Axes
Therefore the total mcment about the
For equilibrium in the pl~~e of the cross assumed moment center 0' =
section at station (0), the summation of the
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A20.15
~.
TABLE A20. 10

MOMENTS DUE TO IN PLANE COMPONENTS

OF STRINGER LOADS

2 3 4 5 6 7 6

Py = Mo' Pz = M o '"
Stringer Px dv p2Y. dz Px..!!I.
No, (lbs. ) OX xdx P y z' dx dx P z y'

a -9312 .0026 248 2600 ·.0233 217 -2600

b -2239 .0188 42 798 -.0421 94 796
c -2304 0 0 0 -,050 115 0 y.
d -1719 -.0188 32 608 -.0421 72 610
e -6506 -.0266 173 -1820 -.0233 152 1820
f 9216 -.0266 245 -2570 .0233 214 2560
3673 -.0188 69 -1310 .0421 154 1307
h 38 9 0 0 0 .050 190 0
2634 .0188 49 930 .0421 111 941
1246 .0266 33 346 .0233 29 348
Sum -1634 1612

NOTES:
---Col. (2) from Table A20. 7
Col. (3) equals the slope Fig. A20.11
of stringers
inyandz
directions. Shear flow distribution.
(see Fig. A20. 9)
Col. (5) Values of z' where large concentrated loads are applied can
and (8) and y' from be determined by the procedure given in Articles
Table A20. 7. 18 to 20 of Chapter Al9. A more rigorous
Fig. b shows a~~lysiS can be made by the application of the
the P y and Pz basic theory as ~iven in Chapter AS.
components from
Cols. (4) and (7).
The problem of shell stresses due to in-
Total moment about O' '" ternal pressures 1s presented in Chapter A16.
-1634 + 1612 '" -22". The strength deSign or the fuselage skin in-
volves a question of combined stresses. The
broad Droblem of the strength design of struc-
92670 due to shear flow q tural elements and their connections under all
-22 due to in plane components or types of stress conditions 15 covered in Volume
stringers. II.
7000 due to the external loads.
A20.9 Problems.
Total= 99648 in. lb.
(1 )
Therefore :or equilibrium a moment of
-99646 is required which can be provided by a ~'S'
~.

constant shear flow q:l. around the cell, hence

I
20"
= T =-:-",9,,96;;,48;;_
i- = -52

+
q, 2A lb./tn. (957
.. 2 x 957
enclosed area or cell)
,
1/
i,
Skin .035

value or q Is entered in column 8 or

~is
Table A20.9. The ~esult1p-g shear flow 1~ any
web ~ortiQn qr equals ~he algebraic sum of q
and ql.' (See Col. 9, Table A20.9). F'~g.
A20.11 shows the results in graphical form.
Fig. A20. 12
A20.8 Discontinuities - Shear Lag - Pressurization
Stresses - Combined Stresses.
Fig. A20.12 shows the ~ross-sectlon of a
A practical :uselage has ~ny cut-outs. circular fuselage. All stringers have same
The approximate ef~ect of these discontinuities a~ea. naTlely 0.12 sq. in. Skin thickness is
as well as the shear lag er!ect at sections .035 inches. Stringers are 1 inch in depth.
A20.16 FUSELAGE STRESS ANALYSIS

All ~terial is aluminum alloy. E = 10,500,000 <-.-.....--'---'-=

-~-=-
psi. The ultimate compressive str~ngth of
stringer plus its effective skin is 35000 pai' l
For e r.rec t tve sheet width use w = l.9t (E/OST )"i".
For buckling strength of curved panels use
ocr = .3 Et/r. Determine the Ulttmate bending
moment that the fuselage section will develop
for bending about horizontal neutral axis. Use
linear stress distribution. Follow procedure
as given in example problem in Art. A20.4.
(2 )

Fig. A20.13 shows

the cross-section of a
rectangular fuselage.
The dots represent
stringer locations.

.~
Three types of string-
J' S, ers are used, namelY,
.Ok S l' S.lI and 5:5' Fig .
"" S, AZO.14 shows the
ult~te compressive
s, stress-strain curve
/032 for each ot the three
stringer types and
S:.. s~ s, also the tension
Fig.A20.13 stress-strain curve
ot the material.
Determine the Ultimate bending resistance (4) Sa~e as Problem (3) but change area of
ot the P~elage section about the horizontal stringer no. (2) to 0.3 sq. in., thus makfng an
neutral axis it the maximum unit compressive l~symmetrical section.
strain is limited to .008. Reter to Art. A20.5
for method ot solution. A20. 10 Secondary Stresses in Fuseiage stringers and Rings.

ADDITIONAL DATA. Area s'tz-tnger- S l :: .12 sq .fn, ;
S.lI :: .25 sq.in.; Sa :: .08 sq.i~. E = 10,500,000
pat ••
(3) Fig. A20.1S shows a tapered circular
fuselage with 8 stringers. The area of each
stringer is 0.1 sq. in. A$sume stringers develop
entire bending rBsistance. Find the axial load
in stringers at station (110) due to Pz and Px
loads at station (0). Also find shear flow
system at station 110 USing ~P method. Use
properties at station (90) IN OBTAINING AVERAGE
SHEAR FLOwE.
7 tension field action in skin panels. It is
.110
Fig. A20.15
The stresses that are found in ~he
stringers or longerons at a typical fuselage by
use of the modified beam theory or by the more
rigorous theory of Chapter AS, are referred to
as primary stresses. Because of the necessity
ot weight saving, most fuselage structures are
designed to permit skin buoxj.tng , which means
ttat shear loads in the skin are carried by
diagonal semi-tension field action. This
diagonal tension in the skin panels produces
additional stresses in the stringers and also
in the fuselage rings. These resulting stresses
are referred to as secondary stresses and must
be properly added to the primary stresses in
the strength desi~ of the indiVidual stringer
or ring. Chapter CII covers the subject at
these secondary stresses due to diagonal semi-
suggested to the stUdent that atter st~dying
Chapters Al9 and A20, that Chapters ClO and Cll
be referred to in order to obtain a complete
stress picture for skin covered structures.
 CHAPTER A21
LOADS AND STRESSES ON RIBS AND FRAMES

A21. 1 Introduction. For eer-ocvnamt c reasons the
wing c~ntou~ in the chord direction must be
~~lntaln~d ~lt~c~t appr~clable distortion.
Unless t~e ~ing sk1n 1s qUite thick, spanwlse
str~nge~s ~ust be ~ttached to the skin in order
to Inc~ease :he bencins efficiency of the wing.
Therefore to hold the Skin-stringer wing surface
to c~ntour sha~e and also to 11~it the length
of strln~ers to an efficient column compressive
strength, 1~t2~~~1 support or brace units are
reqUired. ~hese str~ctural units are referred
to as wtnz ribs. The ribs also have anot ner-
major pur~cse, na~elYl to act as a transfer or
distribution unit. All the loads applied to
tne '.'ling are reacted at the wtng sucportlng
pOints, thus these a~pll~d loads must be trans-
rarre~ into the Nlng cellular stru:ture ~om-
posee o~ skin, strin~ers, spars, etc., and then
reacted at the Wing support ?o!nts. The aJvlled
loads nay be only the distributed surface ai~­
leads which require relat~vely light internal
ribs to ~ravide th~s carry throush or tr~nsfer
~equlr9ment, to ~ther rug~ed or heavy ribs
which ~ust absorb and transmit large concen-
trated applied loads such as those from landing
gear reactions, power ?lant reac~lons anc. ~~se­
Iage reactions. In between these two extreaes
of applied load ~a~nit~des are such loads as
reactions at sup~orting points for ailerons,
flaps, leading edge high li:t uni~s and the
many internal ceaa .vet ght leads such as fuel
and military anmament and other installations.
Th~s ribs can 'ffiry from a very light structure
loads into the fusela~e shell such as those
from landin~ gear reactions, wing reactions,
tail reactions, power slant reactions; etc.
The dead wei~ht of all the payload and fixed
equipment inside the fuselaroe must be carried
to frames by other structure such as the
fusela~e floor system and then transmitted to
the fuselage shell structure. Since the dead
weight must be multiplied by the design accel-
eration factors, these internal loads become
qUite large in ~gnitude.
Another important purpose or action of ribs
and frames is to redistribute the shear at dis-
continuities and practical wings and fuselages
contain many cut-outs and openings and thus
, discontinuities in the basic structural layout.
A21. 2 Types of Wing Rib Construction.
Figs. A2l.1 to 6 illustrate the co~non
types of win~ construction. Fig. 1 illustrates
a sheet metal channel fer a leading edge 3
Skin
flange
.
OiO Beam
Web

which serves ~rlmarlly as a former to a hea'TY

str~cture which must receive and transfer loads
involving thous~~ds of poundS.

Since the airplane control surfaces (verti-

cal and horizontal stabilizer, etc.) are nothing
more than small size wings, internal ribs are
li~ewise needed in these structures.

The skirr-stringer CJnstruction which fOrTIS

~he shell of the I~sela~e likewise needs in-
ternal forming units to hold ~he ~uselage
cross-section to contour shape, to limit the
col~un length of the stringers and to act as Fig. A21. 2
transfer agents ot internal and externally
applied loads. Since a fusela7e ~us~ usually
r~ve clear internal space to house the ?ayload
such as passengers in a commercial trans2ort,
theSe internal fuselage units which are usua:ly
referred to as frames are of the aDen or ring
type. Fuselage frames vary in size and strength
from very light for.ner type to rugged heavy
types which nust transfer large concentrated Fig. A21. 3
A21. 1
A21. 2 LOADS AND STRESSES ON RIBS AND
Fig. A21. 4
o 0 0
o o Removable
o Leading Edge
Portion.
O~~~~~~~
o 0 000
Fig. A21. 6
stringer, single spar~ single cell wing
structure. The rib 1s riveted, or soot-welded,
or glued to the skin along it bo~~dary. Fig.
2 shows the same leading edge cell but with
spanwise corrugations on the top skin and
stringers on the bottom. On the top the rib
flange rests below the corrugations, whereas
the stringers on the bottom pass through cut-
outs in the rib. Fig. 3 illustrates the gen-
eral type at sheet metal rib that can be
qUiCkly made by use at large presses and rubber
dies. Figs. 4 and 5 illustrate rib types for
middle portion ot wing section. The rib
flanges may rest below stringers or be notched
~or allowing stringers to pass through. Ribs
that are SUbjected to considerable torsional
torces in the ?lane of the rib should have some
shear ties to the skin. For ribs that rest
belaw~trlngers thiS shear tie can be made by
a rew sheet metal angle clips as illustrated in
Fig. 5. Fig. A21.7 shows an artist's drawing
at the wing structure or the Beechcraft Bonanza
commercial airplane. It should be noticed that
various types and shapes of ribs and formers
are required in airplane design. Photographs
A2l.l to 3 illustrate t~~ical rib construction
in various type aircraft, both large and smaLl ,
Since ribs compose an appreciable part at the
wing structural weight, it is important that
they be made as light as safety permits and
also be'efficient relative to cost of fabrica-
tion and assembly. Rib development and deSign
involves considerable static testing to verify
and assist the theoretical analysts and deSign.
A21. 3 Distribution of Concentrated Loads to Thin
Sheet Panels.
In Art. AZl.l it was brought out that ribs
FRAMES
.,( Wing Skin
/ Shear Tie
,-
o
- ....
-
Ribs in 3 Spar Wing.
Fig. A21. 5
were used to transmit exter~al loads into the
wing cellular beam structure. Concentrated
external loads must be distributed to the rib
before the rib can tra~sf~r the load to ~he
wing beam st~~cture. In other wordS, a con-
centrated load applied cirectly to the edg~ 0:
a thin sheet would cause sheet to buckle or
cripple under the localized stress. Thus a
str~ctural element ~sually called a web stiff-
ener or a 'Neb r lange Ls fastened to the 'Neb and
the concentrated load goes into t~e stl:fe~er
which in turn transfers the load to the web.
To ~et the load into the stiffener usually re-
quires an end fitting. In general the c isur i ,
buted air loads on the wing surface are usually
of such ~gnitude that the loads can be distri-
buted to rib web by direct bearing of flange
normal to edge of rib web without causing local
buckling, thus st1f~eners are usually not
needed to transfer air pressures to wing ribs.
EXAMPLE PROBLEM ILLUSTRA.TING ':'RJ.NSFER OF CON-
CENTRATSD LOAD TO SHSST PANEL.
Fig. AZl.8 shows a cantilever beam com-
posed at 2 flanges and a web. A concentrated
load of 1000 lb. is applied at paint (A) in the
direction shown. Another concentrated load of
1000 lb. is applied at paint (E) as shown.
To distribute the load ot 1000 lb. at (A)J
a horizontal stiffener (AB) and a vertical
stiffener (CAD) are added as sho'ND. A t1ttlng
would be reqUired at (A) which would be attached
to both stiffeners. The horizontal component
of the 1000 lb. load which equals 300 Ib is
taken by the sti~fener (AB) anc the ver~ cal
component which equals 600 lb. is taken y the
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A21. 3

d b
- L-------'"'

Fig. A21. 7 General Structural Details of Wing for Beechcraft "Bonanza" Commercial Airplane.

vertical stl:fener CD. The vertical load at E Figs. A2l.9 and A2l.l0 show free bodies of that
would be transferred to stiffener SF through portion including web panels (1) and (2) and
fitting at E. The problem is to find the shear stiffeners CAD and AS and ~he external load at
tlows in the web panels, the stiffener loads (A). In rig. A21.9 the shear flows q J, and q ,
and the beam flange loads.
s,
10- 10"-+-'0"--f-- 10"-r'O"i I t~q, --- D

q':~'1I
G ~lange F D
D 9" q t (1)
'f
(I) :/'ti!fener 1 B ~800 B A

TW-
9"
qll~
/
, Web (3) /'
, 3" q I (21 'e IY (2) 1,600 lb.
/ - !3"
f- ~800 lb. E---C
, II (2)
l~ 1000 lb.
s,
10"
I

---l
,"-x q.
/H -, flan g e E 6001 b. Fig. A21. 10
r
1000 lb.
Fig. A21. 9
Fig. A21. 8

on the top and bottom edges respective~y have

SOLUTION; It Nill be ass~ned that the bea~ been as.sumed 'filth the sense as ShC'NTl. Ta:-clr.g
~langes develop th8 en~~re resls~ance to bea~ mcments about ~olnt E,
bending ~oments, thus shear flow is constant
on a 'Neb pane L, mE = 800 x 3 - 12 x lOq J, = 0, whence q ~ =
20 lb./ln.
The shear flews on .veb pane Ls (I} and (2)
computed t.r-eat tng each c omccnerrt Gf
-1'/111 be Zrx = 800 - 20 x 10 -lOqa = 0, whence qll =
the 1000 lb. load as acting separately and the 50 Ib./in.
results added to give the final shear :low.

()1_~

:v"'0 "C
A21. " AND STRESSES ON RIBS AND FRAMES

PHOTO. A21.1 Type of Wing Ribs Used in Cessna 180 Model Airplane, a " Place Commercial Airplane .

..
• .. ~
,
..........

~ • '''#o>..-e~

'j, •,.
'~'--- •
.~~ v

>
J.

PHOTO. A21. 2 Rib Type Used in Outer Panel-Fuel Tank Section- of Douglas DC-8 Commercial Jet Airliner.
,

,..,
;

.,-
.,.,. -

PHOTO. A21. 3
, "
Rib Construction and Arrangement in High Speed, Swept Wing, Fighter Type of Aircraft.
North American Aviation - Navy Fury - Jet Airplane.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A21. 5

Reterring to Fig. A21.10, Solving equations (1) and (2) gives,

ZME = 600 x 10 - 12 x 10q .. = 0, whence q .. = q. = 70 lb./in., q , = 10 lb./in., which

50 lb. in. checks first solution.

ZFx = -50 x 10 ... lOqa = 0, whence qa = The shear flow qa in web panel (3) is ob-
50 lb./in. tained by considering stiftener EBF as a free
body, see Fig. A21.16.
Combining the two shear flows for the two loads,
ZFY :: 12q;s'" 10 x 3 - 9 x 70
q~ = 20 + 50 =70 Ib./in. -1000 = 0
F
50 = 10 Ib./in.
T II
t Iq
whence, qa :: 133.33 lb./in.
Fig. A21.11 shows the results. Fig. 9"
q,
, ",70 The shear flow q~
A21.12 shows stiffener AB as a tree body, and
Fig. A2l.l3 the axial load diagram on stiffener
AB, ~hich comes directly from Fig. A2l.11 by
1 B
,\0
could also be found by
treating entire beam to
starting at one end and adding the shear flows.
3" Cl~ t:. right of section through
panel (3). For this free
T E
body,
1000
BOO#( tension) Fig. A21.16 ZFy =-600 - 1000 + 12q, =0
B --------=1 Fig. A21. 13
whence, q~ = 133.33
q.. =70
B =qa - = 10
-=::"A Fig. A2l.17 shows diagram of axial load in
stiffener EF as determined from Fig. A21.16 by
starting at one end and adding up the forces to
Fig. A21. 12 any section.

Fig. A21.l4 shows a free body of the After the web shear
F tlows have been detennlned

~
vertical stiffener CAD. and Fig. A21.l5 the i g' A2 1. 1 7
axial load diagram for the stiffener. the axial loads in the
beam flanges follow as
B _ 570 lb. the algebraic sum of the
o shear flows. Fig. AZI.18
Tq ,
A~
I E 1000 lb. shows the shear flows
9" ='0, (tension)
+_A'10 630> (tension)
along each beam flange
as previously found. The
qa 600
3" I
upper and lower beam flange loads are indicated
I =10.
...L.. C C by the diagrams adjacent to each flange •
Fig. A21. 14 Fig. A21. 15
4700
700 lb.
Tension
'rhe shear q. and qa could of course
~lows
be dete~ined USing both components of forces
at (A) acting Simultaneously. For exa~ple,
consider tree body in Fig. A21.ISa.
; 133.33 ---
133.33 13333 70
/

~
J- -r -;
t-- 10"-1 _800
q I 133.33 133.3=L 133.33 10
'1000
- 600

,
100* tension
<'

~ig'A2L18
Fig. A21. lSa
L CIa t'IBx ~ 600
&800<1
lb.
-+
3" 3900* compr essron
E- - - C .....L...
q.
In this example problem the apolled extern-
ZFx :: 800 - 10q. (1 ) al load at point (A) was acting in the plane at
the beam web, thus eNO stiffeners were suffi-
ZFy = -600 + 9q. (2) cient to take care of its two components. Often
A21. 6 LOADS AND STRESSES ON .RIBS AND FRAMES

loads are applied Nhich have t~~ee recta~gular

comDonents. In this case, the structure should
be arranged so that line of action of applied
force acts at inter5ect~on of two webs as
illustrated in Fig. A2l.l9 where a load P is
applied at point (0) and its co~ponents Pz , Py
Fig. A21. 20

tLx y
s,
with q known the shear and bending ~cment
Fig. A21. 19 at variouS sections along the rib can be dete~­
~ined. For example, consider the section at 9-3
in Fig. A21.20. Fig. A21.2l shows a free body
and Px are distributed to the web panels by of the portion rorNard of this section.
us ing three stiffeners S J.' S a and S,3 inter-
secting at (0). The bending ~oment at section 8-B equals:
In cases where a load must be applied MB =2qA 1. where A 1. is the area a f the
normal to the Neb panel, the stiffener must be shaded portion.
designed strong enough or tranSfer the load in
bending to adjacent webs. Let Fx equal the horizontal component of
the flange load at this section.
In this chapter, the webs are assumed to
resist pure shear along their bouncarles. In
most practical thin web structures, the webs
will buckle ~~der the compreSSive stresses due
to shear stresses and thus produce tensile
field stresses in addition to the shear
stresses. The subject or tension field beams Fig. A21. 21
is discussed in detail in Volume II. In gen-
eral the additional stresses due to tension
fIeld action can be superimposed on those
found tor the non-buckling case as explained
in this chapter.

A21. 4 Rib for Single Cell 2 Flange Beam.

Fig. A2l.20 illustrates a rib in a 2-

flange single cell leading edge type of beam.
Assume that the air-load on the trailing edge
portion (not shown in the figure) produces a
couple reaction p and a shear reaction R as
shown. These loads are distributed to the cell
walls by the rib which is fastened continuously
to the cell walls. Let q = shear flow per
inch on rib pertmeter which is necessary to
hold rib in eqUilibrium under the given loads
P and R.
The true upper flange load Fu = Fx/cOS Q1.
and the lower flange load equals FL =
Fx/cos Q~.
The vertical shear on the rib web at 3-8
equals the vertical component of the shear flow
q minus the vertical components of the flange
loadS. Hence

Taking ~aments about some paint such as (1)

at all ~orces in the plane at the rib:
Illustrative Problem
ZM1. = -Ph + ZAq =a The rib in the leading edge port:on of the
wing as illustrated in Fig. A2l.22 ~ill be
hence analyzed.
q =Ph/2A. (A = enclosed area or celli A distributed external load dS shown will
be assumed •
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A21. 7

Solution: 's
w = 8/F/in. I F
The total air load aft of beam = 8 x 4-0/2 =
160 lb. The a~. to its e.g. location fr8~ the
b=a~ equals 40/3 = l3.33 •
fl
Hence the react~ons
-rf- 30°

Iv:e:_l_
at the beam flange paints due to the loads on Area of Shaded
the trailing eGge ~ortion equals: 7.67" Portion = 15.4 sq.tn.

P = 160 x 13.33/10 = 213.2 lb. (See Fig.

A21.23 ) Fig.A21. 24
30°
Shear reaction Vr = 160 lb. I
Is
F

A21. 5 Stresses in Rib for 3 Stringer Single

w = 8 Ib./in. Cell Beam.

Fig. A21.25 shryNS a rib that fits into a

single cell beam with 3 stringers labeled (a),
(b) and (c). An external load is applied at
point (a) whose components are 5000 and 3000
------ Ibs. as shown. Additional reactions from a
trailing ed~e rib are shown at points (b) and
10" (c). A vertical stiffener ad is necessary to
~===dll.. distribute the load of 5000 lb. at (a). The
following values will be determined; -
-""1--- 10" 40"
I
Fig. A21. 22 (1) Rib web shear loads on each side of stiff-
ener ad.
w '" 8 lb./in.
(2 ) Rib flange load at section ad.
1~ ~! It! t ttl f ;~ I p= 213. 2 lb.
(3 ) Rib flange and web load at section just

~\11
'"' ~
Area
A '" 139.3 sq. in.
j
I
(=L
r
10"
to lett of line be.

Fig.A21.25
/' t./
» > -20~ ~/i
I
~
"'=:: _ _ _ (l)
160
P =213. 2 lb.
d£:------f1 5°O I
- I 1" I
Fig. A21. 23 Area:=160"'A2 t 1~2 I
~qcb ,/.r
9.3" /

Let q be the cor.stant flew reaction of the web • J/-

cell skin on the rib perimeter which is neces- ~"=,=:_~-lJ...=;_",,,,;e="'_="'~b
at 3000 lb. qba
':1.- 500
sary to hold the rib in equilibrium under the
applied air loads. 5000

Take moments about some point such as the

I--- 15·' --1
lower flange (1).
SOLUTION. It will be assumed that the 3 string-
Z!1 = -213.2xIO+8x15x7.5+2x139.3 q ers develop the entire Wing beam bendlr~ resist-
=0 ance, thus the wing shear flow is constant be-
tween the stringers. The wing rib Is riveted
whence, q = 1232/278.6 = 4.42 lb./in. to the wing skin and thus the edge forces on
the rib boundary will be assumed to be the same
~ith the applied forces on the rib known, as the shear flow distribution. In othar wordS,
the shears and bending moments at various the three shear flows Qadc' qba and qcb hold the
sections as desired can be calculated. For external loads in equilibrium. The sense of
example, consider a section B-B, 2.5" from the these 3 unknown shear flows will be assumed as
leading edge. Fig. A21.24. shown in Fig. A21.25.

Bending moment at section B-B = To find Gacc' take moments about point (b)
8x2.5x1.25+4.42x2xI5.4 = 161 Inv Ib ,
A21. 8 LOADS AND STRESSES ON RIBS AND FRAMES

ZMe = -2(A1. + A.) Qadc .;. 5000 x 15 - 500 To find web shear
~d take ZFz = 0
=0
x 11.5
=-2(60 + 160) QadC + 75000 - 5750 =0 215B-Ib-t
d

74.611,
T'

9.3"
= 5000 - 9.3 x 74.6
- 9.3 q~d =0
QadC = 157.3 Ib./in. I I"ad
whence,
assumed.
with sense as
2034.....J i.- c,l
3000 whence, %d =463 1b./
'C
5000 in.
To tind qcb take ZFz = 0
Fig.A21. 27
To find r Lange
ZFz = 5000 + 200 - 157.3 x 11.5 - 11.5 load C' take ZFx 0, =
<lob =0 conSidering joint (a) as a tree body,

whence, Qeb = 295 lb./in.

ZFx =2034+3000-C! = 0, whence, C! = 5034 lb.
At joint (d) T' obviously equals 2158 lb. The
To tlnd qba take zrx =0 sti:!ener ad carries a compressive load of 5000
lb. at its (a) end and decreases unifOrmly by
ZFx = -500 + 3000 + 500 - 157.3 x 15 the amount equal to the two shear flows or
463 + 74.6 = 537.6 lb./ln.
-15 qba 0 =
The results obtained by conSidering Fig.
whence, qba =42.7 lb./in. AZl.27 could also be obtained by treating the
entire rib portion to the lett ot a section
With these supporting skin torces on the
just to right ot stiffener ad, as showr. in Fig.
rib boundary, the rib is now in equilibrium
and thus the web shears and tlange loads can A21.28.
be determined. Consider as a tree body that
To find rib flange load T' take moments
portion or the rib Just to the lett at the about point (a).
stiffener ad centerline as shown in Fig. A2l.26.
ZMa = (16/17) T' x9.3-157.3x60x2 =0
T
whence, T' = 2158 lb.

I~T'
I
I' 9.3'"

l~
~""';=="'t3000 C'
To tind tlange load T take moments about 5000
point (a), Fig. A21. 28

ZMa = (16/17)T x 9.3 - 157.3 x 60 x 2 = 0

To tind flange load C' take ZFx = 0
whence, T =2158 lb.
=-e' +3000 + 2158(16/17)
ZFx =0
To tind flange load C take ZFx = 0
ZFx = =
2158 (16/17) - C 0, whence C 2034 lb. = whence, C' = 5034 lb.

To tind wab shear Qad take ZFz = 0 To rind %d take ZFz = 0

ZFz = 2158 (6/17) - 157.3 x 9.3 + 9.3 Qad = 0 ZFz = 2158(6/17) - 157.3 x 9.3 + 5000 - 9.3
=
whence, Qad 74.6.1b./in. %d =0
T6 tind the shear in the web just to right
of stiffener ad, consider the tree body tormed
whence, ~d =463 Ib./in.
by cutting through the rib on each side at the The above values are the same as previously
stiftener attachment line as shown in Fig. obtained.
AZl.27. The torces as found above are shown
on this tree body. The rib flange loads and web shear will be
calculated for a section just to lett of line
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A21 9

cb. Fig. AZl.29 shows the tree body for the

rib to left ot this section. 6000

I Fig. A21.30

t- ~:.~.
a skin
8500 b
t

~t ]- Fg~-
6'1'50 ~ Web 8.40"
7500 CQ I

iI
d , .....i...-4500#
z 30" skin c
L. 5000 2~0
Fig.A21. 29
SOL1JrION:
To tind tlange load C take moments about The total shear load on the wing in the Z
pOint (b). direction equals Vz -6000 - 5000 + 2000 = =
-9000 lb. and Vx =-8500 + 7500 - 4000 + 4500 =
l:IIb =-157.3x2 (160+60) + 5000 x 15 --500 lb.
-11.5 C = 0 The boundary torces on the rib will be
equal to the shear flow torce system on the
whence, C = 500 lb. cell walls due to the given external torce
system.
To tind tlange load T take ZFx = a
ZFx = 3000-157.3x15-15x42.7-500+T = 0 From Chapter Al4, page Al4.8, equation (14)
the expreSSion tor shear flow is,
whence, T = 500 lb.
qy = -(K15Vx-K1.Vz) .r.xA- (K.Vz -KJ, 11x)
To find ~b take ZFz = 0 ZZA - - - - - - - - - - - - - - - (1)
IT z =5000 -157.3 x 11.5 -11.5 q~b =0 The constants K depend on the section
properties at the wing cross-section. Table
whence, qbb = 278 lb./in. A2l.l gives the calculation of the moment ot
inertia and product ot inertia about centraidal
The ~bove results could have been obtained Z and X. axes. In this example the 4 stringers
with less numerical work by considering the a, b, c and d have been considered as the entire
forces to right of section cb in Fig. AZI.29. effective material in resisting wing bending
stresses.
A21. 6 Stress Analysts of Rib for Single Cell
Multiple Stringer Wing. TABLE A.21.1

When there are more than three spanwise

1 , 3 , , • , • • 10 11

stringers in a wing, there are rour or more F1an<e

No.
Am
A Z' X' AZO AZO' AX' AX" Z.ZO- Z XaX:-X
panels in the cell walls, thus the reactions at AX' "
the cell walls upon the rib boundary cannot be a '.00 5.50 0 U.OO 60.5 0 0 0 6.38 -U.8
round by statics as was possible in the 3 b 3.55 30 4.43 15.72 37.50 U25 132.9 4. oil 18. :z
I. '"
stringer case of the previous example problem.
e 1.15 -8.40 30 - 9.68 82.20 34. SO 1035 ·289.8 ·7.54 18. :z
Fig. AZl.30 illustrates a wing section con- d 1. 70 ·8.50 o -U.04 71. 90 0 0 0 ·5.64 ·U.8
sisting at tour spanwise flange members. The Z 8.10 - 5.2'7 230.3 72.00 2160 -1511.9
concentrated loads acting at the four corners
of the box might be representative or reactions
tram the engine mount or nacelle structure and z =ZAZ'/ZA =-5.27/6.10 = _.865 ft

the reactions tram a rib which supports the

wing rlap. These loads must be distributed X =ZAX' /ZA = 72.0/6.10 = 11.8'
into the walls or the wing box beam which neces- Centraidal X and z moments of inertia:
sitates a rib. Serare the rib can be designed,
the bending and shear forces on the rib must be Ix =230.3 - 6.10 x .865- = 225.8
determined. The calculations which follow
illustrate a method at procedure. Iz = 2160 - 6.10 x 11.8' = 1310
Ixz = -156.9 - 610 x -.865 x 11.8 =-94.7
 A21. 10 LOADS AND STRESSES ON RIBS AND FRAMES

~lth the Ning section ~ro~ert1es ~no'Nn, Z1'!r. • ,:>.

~
_ = -11000 x 11.8 - 3500 x 6 .36 - 7500 x
the constp-nts K can be calc~lated. 5.65 - ';'000 x 4.41- 4500 x ? .54 - 2000 x 18.2 - ~x

KJ. = Ixz/(Ixl z - Ix(2) 2x90-74Bx2xlC8.7-461x2xG9 =-6~84CO 1:1.10.

= -94.7/(225.8x1310-94.7') = Fer equ11ibrlum Z~.g. must equal zero,
-94.7/286700 = -.00033
therefore a constant ~low shear qJ. acting around
the rib perimeter t s necessary wn i ch wtLl. er-e-
Ke = Iz/286700 = 1310/256~00 = .00456 duce a moment of 648400 in.lb.
K, = l x/286700 = 225.8/286700 = .000786
q, = 2·;~.... 25~"'_E",.~,,_0",0,=
= ~2x368.5 = 880 Ib./:n.
SUbstituting in equation (1),
(Note: 368.5 = total area of cell)
qy = - G000786 (-500) - (-.00033) (-90008
ZxA - E00456 (-9000) - (-.00033)(-500) Adding this shear flow ~o that of Fig.
A21.31, ~~e resulting force sj~tem at F:g.
ZzA A21.32 1s obtained. The reactions of the beam
cell walls on the rib have now been cete~ined
whence,
and the bending mo~ents and shears on the rib
can new be calcul~ted.
qy =3.363 ZxA + 41.205 ZzA - - - - - (2)
Since the shear tlow at any point on the 6000
cell walls is unknown, it will be assumed zero
on web ad, or imagine the web is cut as shown
8509..- ~ __ q-=-.i3~
in Fig. A21.31. The static shear flows can Ia - - ---i9 00
now be round. il t "l,
'lab = 3.363 (-11.8)(2.0) + 41.205 x5.36
7500
=
~1
Id
Fig.A21. 32
I '"
t:
x 2 =444 lb./in.
-\I=:'·-'-=;-=q"'."'.'"',"'9=_=,"",=-_=,"Ci'- _ 4500
qbc = 444 + 3.363 x 18.2 x 1.25 + 41.205 5000
i
2000
x 4.41 x 1.Z5 = 748 Ib ./in.
Fig. A21. 32
~d =748+3.363xI8.ZxL15+41.205 To illustrate, consider the rib section
(-7.54)(1.15) =461 Ib./ln. B-B which passes through the e.g. or the beam
section. ~lg. A21.33 shows a ~ree body of the
These shear tlows are plotted on Fig. bulkhead portion to the left ot section 8-B.
A21.31. Refer to Chapter A14 regarding sense
of shear flows.

rooo Fig. A21. 31

8500' a -9..=444
I '"""~r "~ - - _b "00
6.36
'= Ax --".0,
Area=900..
, ' ,
~I .-
I~
4.4,
-F -p--
5.~4
~~ ea~.~
- ro.• ,~ ~ea= lOB. 7
'/'//)"/;-...,..
t; r
: ~ 754
7500 d ~~~e~ .. 99 ', '~ t~
I -I~q ;;"4BUb. /...- - tC . .00
5000 2000 Fig. A21. 33 Fig. A21. 34
I-- 11. B I lB.2 ---1
I Total Cell
Area", 368. 5°" Moments at section B-B will be referred to
the point (0);

The moments ot the forces in the plane of ZMo = -11000 x U.8 - 3500 x 5.36 - 7500 x 5 .64 + 436
the rib will now be calculated:
x 2 x 36 + 880 x 2 x 70.8 + 419 x 2 x 38.3 =
Taking moments about the e.g. at the beam -38200 1n.lb.
cross section (See Fig. AZl.31):
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A21. 11

The resultant external shear force along
the section B-B equals the summation of the z
components of all the forces.
'f = ZFz = -11000 + 12 x 3eO - 436 x 0.36 + 419
x 0.96 = -195 lb.
The resultant load normal to the section
8-8 equals the summation of the force compon-
ents in the x direct1on.
H = ZFx = -8500 + 7500 + (436 - 419) 11.8
= -800 lb.
Fig. A21.34 shows these resultant forces
referred to Doint (0) at the cross-section. If
we assume that the rib flanges develoD the
entire resistance to normal sta-esses., we can
find flange ~oads by simple statics.
To find upper flange load Fu take moments
about lower flange point.
the boundary forces on a rib located adjacent
to a c~t-out is to find the applied shear flows
in the wing on two sections, one on each side
of the rib. Then the algebraic sum of these
two shear flows will give the rib boundary
forces. Aith the boundary forces known the rib
web and flange stresses can be found as pre-
viOUSly illustrated. The procedure can best
be illustrated by example problems.
A21. 8 Example Problem. Wing with Cut-Out
Subjected to Torsion.
Fig. A21.35 shows a rectangular Single
cell wing be~ with tour stringers or flanges
located at the four corners. The upper surface
skin is discontinued in the center bay (2).
The wing is subjected to a torsional moment ot
STA.O
T 'I _l~a ,,
4Q'i: -
c,
I IRib,lF' "000,,
~ 'I , AI

zrI

whence,
= 12.6
~u
Fu-38200-800x6.6

= 3443 lb. tension

To find FL ase ZF x - 0
Z~X
=0

= 3443 - 800 - FL = 0, whence FL =

/ «:»:
STA. 50..1
"
;::.-
STA.

~
4.~o
'
II,I!
21)0/

jJ'
1'7/
J, Rib

-----.1-:1--4000
Rib (A) /,1
~~~;:c:~-
moved in
lb. Bay (2)

~ ~ I ! . I
2643 lb. compression. ,q r Yr--- ---1.-
ST~O ' ;, ,
The shear flew on web equals V/l2.6 =
195/12.6 = 15.5 lb./in. This result neglects ::;1 (Rib ~o in. lb. Fig.A21.35
effect of fl~~ges not being no~l to section
8-8, which inclination is negligible in this I\ 40" ----i
case.
80000 in.lb. at Station (70) and a couple force
I~the entire cross-section of rib is at Station (50) as shown in Fig. A2l.35. The
ef!ective in bending, then the web thickness problem will be to determine the applied torces
and flange sizes of the rib would be needed on rib (A).
to obtain the section ~oment of inertia which
Is necessary in the beam equation for bending SOLUTION:
stresses. The forces at (0) would then be
referred to neutral axis of section before The applied shear flow on the cell walls
bending and shear 3tresses on the rib section will be found for two cross-sections of the
could be calculated. wing, one on each side at r tb (A).

To obtain a complete ~icture of the web
and flange forces, seve~a1 sections along the
~ib span should be analyzed as illustrated for
section B-B.
A21. 7 Rib Loads Due to Discontinuities in Wing
Skin Covering.
As referred to before, rib~ i~ acdition to
transmitting external loads to wing cell
structure are also a means of re-distrlbuting
the shear forces at a ~iscontinuity, the ~ost
comnon discontinuity being a cut-out in or-e ar
acr-e of tne Nebs or 'Nalls of the .vt ng beam
cross section. ~he usual ~rocedure in f1nding
In bay (1) the torsional moment M is 80000
in.lb. The applied shear flew en a cross-
section of t~e wing in bay (1) thus equals,
q = ul'! = 2 x30000
10 x 4,0
= 100 lb .ym •
This shear flow system is shown on Fig ..
..;21.36 wntc a is a rree body of rib (A). In
bay (Z), since the top skin is removed. the
tcrsional moment must be taken by the tront and
rear vertical webs, since any shear flow in the
bottom skin could ~ot be balanced.
The torsional ~oment in bay (2) is,

9 c, .»
-"..:-"<.. C
A21. 12 LOADS AND STRESSES ON RIBS AND FRAMES

Bay (2)
!,1.q'",300 ..."..Rib A Bay II) ---rl----- L. 30"-----!--_ Bay (3)
t~;;:::;;;:::;~~~;;::;;;1lt-4000#
l ,~~-_ ..9. ':....lQ..o _ -"'1 t q':300 T
h .110" P-.i~;::='=,=,=,=,=,=,===~
p Vl3000 Fig.a

·~:t=;:~~===:=]lrv:iooo
q:: 100 '1
t I lUGl+ J -4000Jil
I _p
q:: 10

Fig. A21. 36 VL I ~ --------===::J Fig. b

M:: T e------ Beam Bending Moment Diagram
V.3000!....... _
11 = 80000 + 4000 X 10 = 120000 tn.lb.
I Beam Shear Diagram Fig. c
The total shear load an each vertical web
thus equals 120000/40 = 3COO lb., which ~ives q:: 300
a shear flow q' = 3000/10 = 300 Ib./in. on each - --- - - - -:1
web. ThiS applied shear is shown on the tree Fig.d
body of rib (A) in Fig. A21.36. On the left
end of the rib a shear flow of 100 is acting

Fig. A21. 37
q:: 100 flange loads at the bea~ ends are P = 45000/10
:. 4500 lb. (See Fig. a).

The deflection of the rear bea~ would be

up and on the other side a shear flow of 300
Is acting down, thus the rib web must take the
difference or 200 acting down. On the right end
of the rib the load on the rib web is 200 Ib./
in. up. The loads on the top and bottom
flanges at the rib Is obViously 100 Ib./in.
Fig. A2l.37 shows the loads applied to the rib
boundary when the torsion in bay (1) and the
external couple force is transferred to the
cross-section at bay (2).
the reverse of ~ig. a. and thus all forces
would also be r~versed.
Fig. A21.39 shows bay (1) 0:: the wf ng as a
free bedy acted upon by the :lange loads due to
bend:ng of the beams in bay (2). These interr.al
flan~e forces from bay (2) must be held in
equilibrium oy ~he internal stresses ~n the ad-
jacent wl~g structure of bay (1).
~ '-'lr@
I:! ..
/ ;-Rib (Ai / :i
".

ADDITIONAL EFFECTS DUE TO DIFFERENTIAL BENDING

OF BEAMS IN BAY (2).

The torsion In bay (1) and the external

couple force are thrown off as couple torce on
the front and rear beams of middle bay (2),
with the total shear load on each beam being
3000 lb. as previously calculated. These beam Fig. A21. 39
shear loads must be transmitted to bay (3) and
thus cause bending of the bea~s in bay (2). According to the well known ~rinciple of
Since each beam is attached to relative rigid mechanics formulated by Saint Ve~~at, the
box structures at each end, namely bays (1) stresses resulting from such an internal force
and (3), the beams tend to bend with no rotation system will be negligible at a 1istance from
of their ends. If we neglect the deflections the ror-ces . ~hls distance in case of a cut-out
of these end box structures, we can assume that is usually assumed as approximately equal to
the beams bend with no rotation of their ends ~he width at the cut-out, or in g~neral ~o ~he
or each beam is fixed ended. Fig. A2l.38a width of the adjacent wing bay. Thus in Fig.
t LLus'tr-a'tee the deflection of the front beam in A2l.39 the flange loads 0: 4500 lbs. each are
bay (2) under the assumption of no end rotation. assumed to be dissipated at a unifJr.TI rata for
The beam elastic curve has a point of inflection a distance of 20 inches. Thus the shear flow
at the span midpoint. Figs. 38b, c show the created by each stringer load which equals the
beams bending moment and shear diagrams. change in axial load per inch in the strip~er
In bay (II equals 4500/20 = 225 lb.
The end moments are M = VL/2 = 3000 x 30/2
= 45000 in.lb. Assuming the b~a~ flanges Fig. A21.40 ShONS a segment 1 inch Nide
develop the entire bending resistance the beam ~ut trom wing bay (1) with the ~p load in each
; '-.",,~,''''' -,~. <-

-=
ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A2L 13-

:lange member. ~o !ind the shear ~lcw on ~~e bay (1) as shown. The areas of corner stringers
cross-section the frent web is tirst assumed a, 0, c and dare shown in ( ) adjacent to each
cut, and thus the static shear floW qs = Z6P stringer.
from cut face where q~ is zero. Fig. A21.40
shows th1s static shear flow.
B800 lb.
No Skin on
Bottom of r-- 15,,--t
Bay (2).

Fig. A21. 40 Rib

~O
T
1"

For equilibrium of the cross-section, the

moment of the forces in the ;lane of the cross-
section must equal zero. Taking moments about
lower left hand corner of the qs force ~ystem,

i1 = 225x40xlO = 90000 tnv Ib , For equili-

brium a moment of -90000 1s necessary .. There-
fore a constant shear flow system q must be
added to develop a moment of -90000. Thus
q ~ M!2A ~ (-90000/2 x 10 x 40) • -112.5 lb./ln.
Adding this shear flow to that for qs in Fig.
A21.40 gives the final values in Fig. A21.41. 12" Fig. A21. 43
This shear flow system represents the stress Rib
c

112.5
I· 30"
='" - -

Fig. A21. 41 The mi"dle bay (2) has no skin on the

112. 5
system caused on c~oss-sectlon of bay (1) due
to the differential bending of the beams in
bay (2). This shear flow system must therefore
be resisted by rib (A) as it ~ust terminate at
end ot bay (l). Therefore the shear flows in
Fig. A21.41 are applied bounda~y loads to
rib (A) and these T,Ust be added to the rib
loads in Fig. A21.37 to give the final rib loads
of F~g. A21.42. ~lth the final rib loads
botto~ surface, or in other words, the middle
bay has a channel cross-section, which fact
often happens in practical ~ing design as for
exa~ple a space or well tor a retractable land-
ing gear. The problem will be to find the
shear flow in bays (1) and (2) and the boundary
Loads on rib (A) between bays (1) and (2).
Solution No.1
This method of solution will make use of
the shear center location for bay (2) in order
to obtain the true torsional moment on bay (2).
~ith this torsional moment known, the ~rocedure

312.5 tl ____
- - -I'
212. 5
-4000*
t312.5
'_4000*
is similar to the previous example involVing
win~ torsion only.

'.-'e wf Ll. first c~lc~late the shear flow in

212.5 Wing bay (1). ~ig. A21.44 shows the c~oss-
Fig. A21. 42
section.

:-t
aBOO/!'
known, the rib fl~nge and web stres5es can be z
found as previously explained. L. -,-
ho:-,-'5"
a
1
I
(0.5)
AZl. 9 Example Problem. Wing with Cut-Out Subjected 6" b 4,,1
to Bending and Torsional Loads. I ~

TO l!--~------""ro:sr
2400 4'"

-r- 6"
* ...L:.-
~ . . g. A
-. " s ..ows
21 .~ " a pertl on 0 f 3. 4 str 1nger-1 3" d
single cell cantilever beam composed of 3 bays _ I (0.5
~o~ed by the four ribs. The loads on the
structure consist of loads applied to end of
~x-~
~10" z 30" ----i Fig. A2l. 44

o ,' . . /
'--"--I
A21. 14 LOADS AND STRESSES ON RIBS AND FRAMES

The section moments of i~=rtia are ~2eded ~herefore

we ::~s~~~r jay (2) in l~s ~:~e
in calculating shear flows. c8~d~tlon w1t~ bot:om skin ~~movec. r1;.
A21.?7 shows t~~ cross-section of bay (2).
Ix == (j x s e x z ) + (O.5x";''i x2) = ae 1:1:
8800
I 1
X = 'i:ki:/ZA = (1 x30)/3 = 10 tn , I- 15"---1

Iz::l (2xIO Il ) + (lx20 1l ) = 600 In~

I ~-.._~"'=-:,~
2400 12"H 'lad
Vz ::I 8800 lb., Vx == 2400 lb. 3::r-ll
~ "d
c
qy:ll -~ ZzA -*ZxA , substituting Fig. A2l. 47

':'he 'thr-ee shear- :'l'J',':sC3.r. be cc t cr-atncc by

qy = - 100 l:zA - 4 l:xA (a)
statics.
Since the shear flow is unkno~n at any
paint on cell, we will assume frent web (ad) as !F x ::: 2400 - 30 Qab = 0, whence ceo = SO
cut or carrying zero shear.
z:-''(J. = 2400 x 3 - 3600 x 15 - SO x 30 x 12 + qbc
~c = -100 (-6)(1) -4 (-10)(1) • 640 lb./in (8x30) = 0, whence Gbc· 640

Qcb = 640 -100 (-4) (0.5) - 4 (20) 0.5 = 600 ZFz = 8300 - 9 x 640 + 2 x 50 - 12 Sac == C
whence Qad 320
qba = 800-100 (4)(0.5) -4 (20) 0.5 = 560 :=

Fig. A21.48 shows the ~esults. ~~1s shear

Fig. A21.45 shows these static shear flows. :low system 1s t~e final or tr~e shear on bay
.~8800 (2) •
I- 15'---,
-toJ.l:: !~._
a- - J§.Q ~i'::.-
bl'll
n°lr d
c
640

----640- Fig. A2l. 48

Fig. A21. 45
Since we have a. channel or ooen 'Ni::g C:"05S-
To this shear flow, a constant shear flow section in bay (2), any torsional moment on
this bay :nust be transmitted by ji::erent~al
must be added to make Ll1 = O. Take moments bendin~ of the front and rear beams. ~o obtain
about point (d).
the torsional :noment an bay (2), the shear
center location must; be xncwn ,
ZMa = -8800 x 15 + 2400 x 3 + 560 x 30 x 12 + 800
x a x 30 = 268800 in. lb., or -266800 ~orizontal location of shear cent3r: -
Assume the section bends about centroidal X axis
is reqUired for eqUilibrium, hence the reqUired
without twist under a Vz load of 8800 lb.
constant shear flow q = -M/2A = -268800/2 x 300
= -448. Adding this shear flow to that of Fig. q Vz
= - 1; ZzA , or q =_ 100 ZzA
A21.45, we obtain the shear flow of Fig. A2l.46.

Qcb =- 100 (-4)(0.5) = 200

Qta = 200 100 (4)(0.5) =0
'lad =0 - 100 (6)(1) = -EOO
192 Fig. A21.49 shows the shear flow results
Fig. A2l. 46 !or bending about x-x wit~out twist. 7he line
of action of the resultar.t of :his shear flow
force syste~ locates the horizontal ~csitlcn of
This shear flow 5yste~ would J8 the shear flow t~e shear center.
system for all 3 bays if the bottom skin in bay
(2) was not re~oved. 2e~ovlng the bottcm skin X .. (200x8x30)/8800 = 5.45 in.
in bay (2) ',vill :nodif':l tneee shear flows of
:<'i -. ,1.21.+6.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A21. 15

l~~.~
ential bending at beams in bay (2), thus equal
the beam shear times halt the span at bay (2)
600 diVided by the beam depth.
+d R::8800
For tront beam P = 3841 x 12.5/12 = 4000 lb.
Fig. A21. 49
For rear beam P = 3841 x 12.518 = 6000 Lb ,
Vertical position of shear center: -
Fig. A21.52 shows these.flange loads
Assume section bends about centroidal z applied to bay (1). These loads are dissipated
axis without twist under a load of Vx = 2400 lb.

q = _.'!x.
Iz
ZXA =_ 4

'<cb. -4x20xO.5
ZXA

=- 40 lb./ln.
~
7°:--- 1 _

..L / Bay (1)

"1/
qba:: -40-4x20xO.5 = - 80

<lad = -80 - 4 (-10) 1 =-40 Rib A 1'1.

IIC!L--~ Fig. All. 52
Fig. A21.50 shows the shear flow results.

~
/
unifOrmly 1n bay (1) over a distance of 30
inches, or the shear flow per inch prOduced by
c these flange loads equals tJ' :: P/30, whence
Fig. A21. 50

The vertical distance Z trom point (a) to

<lPa = <lPd = 4000/30 = 133.3 and <lPo = <lPc
the line of action of the resultant which = 6000/30 = 200 lb.
locates tne vertical location ot shear center
IS, Fig. A21.53 shows an element or bay (1)
ana inch wide with these dP loads. The shear
Z = l:!1a/2400 = (40x8x30)/2400 = 4 In. flow q assuming the front web cut equals Zt.P.
The resulting static shear flows which equals
Z<lP Is shown In Fig. A21.53.
Fig. AZl.5l shows the shear center location
and the external loads. The moment about the
shear center which equals the torsion on the
wing bay (2) equals,

Ms • c • = 8800 x 9.55 - 2400 x 13 = -115240

In. lb. Fig. A21. 53

e.c. 8800
.-::r-l i,.9..,,~
2400
~{H- k'" lb Fig.A21. 51
The moment of this shear flow system about
point (d) = 133.33x30x12-66.7x8x30 = 31980.
c For ZM = 0, we need a constant shear floW q =
- 31980/2 x 300 = -53.3 lb./ln. Adding this
constant shear flow to that or Fig. AZl.53
Thi~ torsional moment must be resisted by
gives the shear flow system at FIg. A21.54.
front and rear beams. Hence shear load on These results represent the effect on bay (1)
each beam = 115240/30 =3841 lb.

As in the previous example problem in-

volving torSion, the beams in bay (2) will be
assumed to bend without rotation ot their ends,
I
53.3.
--~
80

Fig. A21.54
f 120
I
I
'01.3 j
_
-- ,192

Fig.A21.55
• 232
I I
or in other words the bending moment at mid-
point of bay 1s zero. The flange loads at '0 ..- 272-
paints a, b, c ~~d d on bay (1) from the d1fter-
A21.16 LOADS AND STRESSES ON RIBS AND FRAMES

of removing the botto~ skin tn bay (2). Adding
the shear flo~s of Fi~. A21.5~ to ~h1se c:' 5"1 ~.
A21'~J ~e obtain :~= final 3~ea~ ~lcws in ':lay
(1) as sho~n in fiS. A2l.55.
aOmJD.;;RY LOi..OS 0;' 813 (A)
192 with sense ~s shown. Since this s~lrr is
nt satng »e reverse this shear' r lc.v and rmc tne
reSisting shear flows )n t~e ~ther ~~:se siees
of tr:e bay crOss-sectiJn. Fi~. ~21.5S shows
the sec t i on, with the 3 unknown shear ~1:J'I'./s
Qab, qbc and Qa1'

T~e boundary leads on rib (A) ~ill equal

the ~1f:erence b0tween the shear flows in ~ays
(1) and (2). FiS. A21.56 shows ~ free Jody af
1\_ '1a.2-" 1;': _
rib (A) with the shears :lcws obtai~ed f~am
figs. A21.55 and ;"21.48.

To r Ind qat use L:F x = 0, 192 x 30 - 30 c-a::, =0

whence, ~b = 192
llt:t = -30 x 192 x 12 + 8 Qbc x 30 = 0
whence, qbc = 288
3F z = 4 x 192 - 8 x 288 + 12 'lao =0
whence, Gad = 128
The resulting applied boundary forces to AddIng the shear flows of Fig. A21.58 to
the rib equal the alGebraic sum of ~he shear those of Fig. A21.46 sives the final shear flows
flows on each side of the rib which ~ives the in bay (2i as shown in Fig. A21.59. ~r.ese re-
values in Fig. A2l.57.
272
--80

181.3r~~~ 272
I: 408 320 ;LI_ _- - ; ; - - - ] : 640
o Fig. A21. 59
Fig. A21. 57
, sults check the results in Fig. A21.48 obtained
With the rib boundarJ loads known, the in solution method (1).
stresses in the rib can be found as previously
illustrated in this chapter.

Solution No.2

This method of solution first finds the

shear flow in all bays assuming bottcm skin is
not removed in center bay (2). This gives a
shear flow in the bottom skin. However, the
skin in bay (8) is actually removed so a
corrective set at shear flows on bay ,(2) along
th~ boundary lines at the bottom skin ~ust be
applied to eliminate the shear flows found in
the bottom skin. The problem then consIsts at
finding the influence at these corrective shear
flows upon the shear flows as found for bays
(1) and (2) when bcttom skin in bay (2) 'NaS
not reJloved. I;
l' 192
Fig. A21. 60
-------
The f~rst step is to find the shear flows
in all bays assumi~~ bottom skin in bay (2) is
not removed. The calc~lations would be exactly ?1~. A21.60 shows the corrective shear
like those in solution (1) and the shear ~low flows of Fig. A21.58 applied :c bay (2). On ~~e
in all bays would be those in ?Ig. A2l.46. The bottom sAin the correc~ive shear :lcw is sho~TI
bottom skin in Fig. A2l.46 has a shear flow of on the bouncary of the cut-ou~. These shear
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A21. 17

flows cause di~~erential bending of the front

and rear beams in bay (2). If we make the
ass~~~tion that the beam end sutfer no rotation,
the be~ding ~oment is zero at niQPOint of the
bay and thus th~ fl~nge loads at ~oints a, b,
c and d of bay (1) equal the algebraic s~ of
the shear ~lows on each side of a flange times
half the span of bay (2) or 12.5 inches. Thus
:rQ~ FiS. A21.60, Frame in Center
Portion of Fuselage
Pa = (192 + 128)12.5 = 4000 lb. compression for "Vought" FaU
Airplane
Pb = (288 + 192)12.5 = 6000 lb. tenRion PHOTO.A21.4

Pc = (288 + 192)12.5 = 6000 compression

Pd = (128 + 192)12.5 = 4000 tension

Referring to Fig. A21.52, we find that the

P values above are the same as the P values
obtained by solution (1). Thus the remainder
of solution (2) Nould be identical to that in
solution u i, and therefore the Calculations
wi:l not be repeated here.
A21. 10 Fuselage Frames

Frames in a fuselage serve the same pur-

pose as ribs in Wing structures. Ribs are
usually of beam or truss construction and can
be stress analyzed fairly accurately by statics.
Fuselage !ra~es however, are at the closed
ring type of structure and are therefore static
~lly indeterminate relative to internal
stresses. Once the applied loadS on a frame
are known the internal stresses can be found
by the application of the elastic theorJ as
covered in Chapters A8, A9, AlO and All. The
loads on fuselage frames due to discontinuities
in the fuselage structure, 3uch as those due to
windows and doors, can be approximately de-
termined by the procedures preViously presented
for wing rros .
The photographs on page 32 of Chapter Al5
show some of the frame construction of the
Douglas DC-8 airliner. Other pictures of
fuselage construction are given ln Chapter A20.
Photographs A21.4 and 5 illustrate typical
fra~e construction and arrangement.
A21. 11 Supporting Boundary Forces on
Fuselage Frames.
'When ext$rnal concentrated loads are
applied to a ~~sela~e fra~e through a suitable
fitting or connection, the frame is held in
equilibrium by reacting fuselage skin forces
Nhich are usually transferred to the frame
boundary by rivets which fasten fusela~e skin
to fr~~e. Since the fuselage shell is usually
stress analyzed by the beam theory, it is
therefore consistent to deter.nine the distri-
bution of the supporting skin forces by the
same theory.
PHOTO. A21. 5
Fuselage-Wing Portion
of "Martin" 404 Transport
A21. 12 Calculation of Frame Boundary Supporting Forces.
Ex~ple Problem 1
Fig. A21.61 illustrates a cross-section of
a circular fuselage. Two concentrated loads of
2000 lb. each are applied to the fuselage frame
at the polnts lndicated. The problem is to
determine the reactlng shear flow forces in the
fUselage skin whlch will balance the two ex-
ternally apolled loads. This tusela~e section
~1ght be considered as ~he aft portion of a
~edium size fUsela~e and the loads are due to
air loads on ~he horizontal tail surfaces. To
make the numerical calculations short the
fuselage stringer arrangement has been assumed
symmetrr t ca L,
Solution:
In this solution the fuselage skin resist-
ing forces will be assumed to vary according to
the general beam theory. The general flexural
shear flow equation ~or bending about the Y
axis is,
A21. 18 LOADS AND STRESSES ON RIBS AND FRAMES

= -31.82 - 6 .275 x .30 x ::"3.5 = -57.22

q =-?-Y ZzA , where Vz = 4000 lb.
:= -57.22 -5.275 x .loS x 10 = -86.62
The moment of inertia Iy of the fuselage
cross-section is required. In this 5 tmoj i r i ed
illustration, the area ot each stringer plus its
q
.-. = -66.62-6.2?5x.l.Sx.5 = -71.32
Due to symmetry of effective c~css-5ectlon,
effective skin will be taken as .15 sq. In. The
student should of course realize after stUdying the shear flow is s~~etrical about the Y axis.
Chapters Ale and A20 that the true effective
area should be used on the compr'esa tve side and As a check on the above Nor~, the s:~a­
that the skin on the tension side of the fuse- tian of the z corn~onents of the shear flew on
lage 1s entirely effective. These facts would each skin Danel between the stringers shaule
equal the external load of 4000 lb.
tend to make the effective c~oss-sectlon unsym-
~etrica1 about the Y axis. Since the only pur-
pose ot this illustrative solution is to show ZF of skin qhear flow equals the vertical
how the trame loads are oalanced, the section
prOject~d length of each ~anel times the shear
flow q on that panel, or
being assumed as ~etrica1 which will greatly
decrease the amount of calculations required.
ZFz:=li G-.4x16.S7+2.7x31.82+3.5X57.22
+ 5xSS.62+3x71.32] 4 = 4000 lb.

Double angles at
3,3',9,9'.
200~O 11.5 ,e-i
1 I'
2000

11. 5 ' J
Fig. A2l.62 shows the frame with its
balanced load system. The internal stresses
(check)

2 , bulb angle
stringers can' now be found ~y the methods of Chapters
AS to All.

2000 2000
!
t
6' Y

11 11'
Fig. A2l. 61

Moment of inertia of fuselage section

about Y axis which is the neutral axis under
our slmplltled assumptions.

Iy = .15 (17 .S- + 1S.2- + 13.5- + 13.5- + 10- +

5-)4': 637 in,'
Fig, A2l. 62
Due to symmetry ot effective section and
external loading, the shear flow in the fuselage
skin on the z axis or between stringers 1 and 1 Example Problem 2. Uns~etr1cal Vertical Loading
and lIar 11 wtll be zero. Thus starting with
stringer (1) the shear flow in the skIn restst- In certain conditions in flying and land-
ing the external loads of 4000 lb. can be ing, uns)m~etrical concentrated loads are a~?lied
written around the circumference at the section. to the fuselage or r.ull structure. For ex~~ple,
Fig. A21.63 shows the same section an~ fr&ne as
was used in Problem 1. Due to an uns~metrical
q = Vz ZzA
Iy
=_ 4000
637
ZzA =_ 6.275 ZzA load on the horizontal ~ail, the reactions from
the tail on the fuselage are as illustrated in
q .-. = -S.275x .rs x iv.s = -lS.S7 Ib./in. the ~lgure. The total load ~n the z cirect10n
1s still ?COO lb. but the loads a~e not sym-
~etrlcal about the z axis. For analYs~s pur-
q .-, = -16.S7-6.275x .lSxl6.2:= -31.82
poses, consider tha loads as ~ransierred to the
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A21. 19

e.g. of the section as indicated in Fi 0 . A21.64.

The ~o~ent of the two ioaGs about the c.g. =
1500 x 11.5 - 2500 x 11.5 = -U500 in. lb. The
I-- 8" -4- 8" -+- 8" ~f"0

I'~
..
sh3ar load Vz = 4000 produces the same shear h f
(3)
flo~ pattern as Fig. A21.62. To balance ~he /
(1) e 500
moment of -11500, a constant shear r Low q :I. /
/
(2) b'
around the fra~e is necessary. / web
(4)

- M- 11500 a = 5 .6-
Q-2i-2xnxlS lb • /1 n ,
/
;;.
t'
I
0 /

600
I" 500
I" 1000
(A = area of fuselage cross-sectton)
Fig. A21. 66
Adding this constant force system to that
ot Fig. A2l.62 J gives the final boundary (2)
supporting forces on the frame as illustrated
in Fig. A2l.65. The elastic stress analysis
+500
of the t~~e can now proceed.
2000
1500 2500 c b

l -~.I.- ~--II Rib Fig. A21. 67

a
!
" 2000
i I fX-1 10"
·1
I. \
\,
Fig. A2l.67 shows a wing rib inserted in a
Fig. A21. 63 Fig.A21. 64 3 flange single cell wing.beam J which is sub-
jected to the external loads as shown.

(1\ Flnd rlb flange loads ,t (e) and (d).

\2500 (2) Find rib web shear flow on each side
at stiffener cd ,
(3) Find rib flange and web loads at
section 5" to left of line abo

(3 )

l OOa
t

~
d
web
1
(2) Fig. A21. 68
web
-'

I
r
7 1000
600

to
t
500
b

I
I

?ig. A21.68 shows a 3 stringer single cell

A21. 13 Problems.
wing beam. A rib is inserted to distribute the
(1) Fig. A2l.66 shows a cantilever Jearn loaded concentrated loads as ShONTI.
as shown. Find the shear flow ~n each of the
4 web panels. Draw axial load dia~ram for (1) ?lnj shear ~lcws in rib web ~anel (1)
each ot the vertical web sti~:eners and also (2) and (3).
the horizontal stif:ener be. Plot axial load
diagram tor beam :la~e ~embers as obtained (2) Find rib flange loads at sections de
~r~m web shear flews. and ab ,
A21. 20 LOADS AND STRESSES ON RIBS AND FRAMES

(4) 1000 Fig. A21.69 (8) Sa~e as preblem (7) but with top skin re-
shows a 2 stringer, moved instead of lower ski~.
2 cell wing beam.
A ~ib is inserted (9) S&~e as (5) but with read spar web removed
to transfer 1000 lb. instead of bottom skin.
load to beam struc-
ture. (10) Same as problem (7) but with rear spar 'Neb
Fig. A21. 59 b removed instead of bottom skin.
r---10"-j Find shear flow
in rib web in each (11) In Fig.
cell aojacsnt to AZl.7l the external
line abo Also rib bulkhead leads P
flange loads ad- and P2 equal 4000
jacent to pOints lb. each and ?~
(a) and (b). equals zero. The
fuselage stringer I \.~"
~terial consists
_ JL'_
,
of tour omega.
T
~71:
sections with an
area of .25 sq, is-
in. each. Deter-
4' ~ " • 1
I
mine the skin re-

~ :-/(----=--
Sisting forces or.
the bulkhead in
ba lanc tng the
iOi': .... I RibB above loads. Fig. A21. 71
~ ~ 0(;' :1 Neglect any effec-
4' '"r,,'_JL.- -:P -
-/1_1_ .J-S:=;-'(~ P,
tive skin in this
problem.

(12) Same as' problem (11) but ~ke p~ = ~OOO

'+" r--
oS' __
".f(
a
IjRibA
I
I
1

-'-P,
Corner stringer Areas:-
and Pa = 6000.

(13)

(14)
Same as problem (12) 'but add

In a water
p~ .- 30CO lb.

:: j r\
c {a} • I sq. in.

I
landing condition

-l d
'--,'-r-- P2
r-
(h) •. 8 sq. in.
(c) "'.5 sq. in.
(d) '" 0.4 sq. in.
the hull frame
of FIg. A2L 72
L
I
I s"-1 b

ta:LJ Fig. A21. 70

is SUbjected to a
nor.nal bottom
pressure at 200
lb. per in.
The area ot the
bUlb angle e"',
I

40"
I
~,

(5) Fig. A21.70 shows 3 bays of a cantilever

stringers is
.11 sq. in.
"" I
4-
I
single cell, 4 stringer wing beam. The bottom each and they d
skin in bay (2) is removed. Find the shear are 7/8 in. deep. I 8.5"
flows in all bays and boundary loads on ribs The area 0r the I
(A) and (B) when the external wing loads are as Z stringers is e:.-.-.
,
24" _
follows: T = 56000 in.lb., P 1 = 0, P a = 0, .18 sq. in. each
p ~ = 2000 lb., p .. = 2000 Lb , , Po = O. and the depth 1.5 Fig. A21. 72
in. The area of
(6) Same as problem (5) but upper skin in bay the stringers a,
(2) is removed instead of the lower skin. b, c , d and e ~s
.20 sq. in. each. Neglecting any effective
(7) Same as problem (5) but with the following skin determine the skin .ras i st mg torces on
external loads. the frame in balanCing the bottom water pres-
sures.
T = 56000
in.lb., P,=50001b.,
(15) Same as problem (l4) but consider that
Pa = 2000 lb., p~ P, = =
0 and P, = 1000 lbj ths ~ater pressure is only acting on one side
at the bottom of the frame.
 CHAPTER A22
ANALYSIS OF SPECIAL WING PROBLEMS
ALFRED F. SCHMIIT
A22. 1 Introduction.
In a previous chapter (A19) analyses of
wing beams were carried out using the engineer-
ing theory of bending and rational ~odlflcatlons
thereot. As discussed there, wing configura-
tions which depart radically fram the usual
conception ot a "beam" present the engineer
with the choice of making apprOXimate and/or
empirical corrections to bea~ theory, or of
following a complete ar~lytlc treatnent of the
structure.
This chapter illustrates the latter ap-
proach to several syeclal problems associated
with aircraft wing structures, viz.,·
Art. A22.2 - stresses around a panel cutout
Art. A22.3 - shear lag problem
Art. A22.4 - cutout in a box beam
Art. A22.5 - swept wing box be~
Aside from presenting one analytic treatment
of these problems, a discussion is given of the
physical nature of each phenomenon. An under-
standi~g of the nature of the problem 1s of
prtne importance, since no one ar~lytic
technique can be all-powerful in the solution
of stress problems. The analyst must exercise
jUdgment and ingenuity in approaching each new
situation.
In this chapter all analyses are made
using the matrix formulation of the Method of
Dummy Unit Loads (Chapters A7, AS), a famili-
arity with which is assumed.
Such problems as those listed above are
too unwieldy to be studied here in great detail;
hence no attempt at exhaustive analyses has
been made. To bring into relief the main
features of each problem, the structure selected
tor analysis is one Rhich is Si~ple in con-
struction and so loaded as to exhibit clearly
the phenomenon under study. Many practical
detailS, such as the effects of sheet wrinkling,
rivet and fitting "give", stress concentrations,
etc., have been Side-stepped so as not to be-
cloud the objective. Further, the problems of
idealization of the original structure, into
• One other important special problem - the so-called "bend-
ing stresses due to torsion" - is not treated here specifically.
As indicated in Chapter AS, the general box beam analysis
presented there encompasses this problem (Example Prob.
15, p, p. AS. 24 through AS. 27).
A22.1
the one finally analyzed, are treated only
lightly; additional references are cited where
appropriate.
The analyses shown are strictly applicable
to (reasonably) thin-skinned Wings only, where-
in the "constant shear flaw" assumptions are
valid, viz.
i-the sheet carries shear stresses only
1i - normal (direct) stresses are carried
1n the flanges (spar caps and stringers) with
effective areas of skin lumped in. In all
cases handled here the "skin was assumed tully
effective (stresses below skin buckling stress -
see Art. AI9.Il, Chapter A19).
To enhance the usefulness of these problems,
all the structures chosen for analYSiS were
taken from referenced NACA (National Advisory
Committee for Aeronautics) publications Wherein
the reader may find detailed discussions of the
problems, other methods of analysis and data
obtained from tests upon the specimens. Where
available, these data nave been-used herein for
comparison.
A22. 2 Stresses Around a Panel Cutout
"Cutouts in wings and fuselages constitute
one of the most troublesome problems confront-
ing the aircraft deSigner. Because the stress
concentrations caused by cutouts are localized,
a number of valuable partial solutions at the
problem can be obtained by analyzing the be-
havior, under load, of simple skin-stringer
pane IS" (l).*
Thus, in the case of a wing beam with a
panel cutout of the upper surface (Fig. A22.l),
it would be feasible to analyze the section
!mmediately around the cutout as a flat sheet-
stringer panel under the action of axial
stringer loads and edge shears (coming from the
spar webs). The axial stringer forces could be
computed with sufficient accuracy by the en-
gineering theo~J of bending (E.T.B.) since
these are removed SUfficiently far fram the
cutout pr-ope,- , The edge shear nows are readily
computed by those elementary conSiderations
which give the spar-web shear flows •
.. Numbers in parentheses refer to the bibliography at the
end of the chapter.
A22.2 ANALYSIS OF SPECIAL WING PROBLEMS

J I

*2 / w3

Ft!C. A22.1
Fig. A22.2 Idealization of the half panel by
use of substitute stringers
The sheet-stringer panel may, in general,
contain a large number of longitudinal ele- Fig. A22.3 gives the geometry of the ideal-
ments (stringers). The labor involved in ized panel.
treating this multi-element structure in detail A:,'" .703 in
2

is prOhibitive, and thus an appropriate ideal-

ization must be made. First, it is likely that .-
~
,-.r~--n------------
I
A ,= .212 in 2

the panel may be considered to be symmetric

about a longitudinal axiS, so that only the
halt-panel need be handled. Second, the com-
+ ~
A,= 1.045 in

in
2

.2
plex, multi-stringer structure is replaced by +U F=H==;,------ AR'" 0.25
one ~ving but three stringers. As indicated ,__ __lL__ t:, '" 0.0331 in
in Fig. A22.2, these stringers are:
~
~
~
-'--
t2 '" 0.0331 in
:-b. -ba ... :L
*1, a substitute stringer having for its
area all the effective area of the fully con- o~ _0 f---, b:, ,.5.96 in
A:l., Alii A 3
tinuous members to one side of the "combing - t:, - t a ba =7.56 In
stringer" (the stringer bordering the cutout)
and placed at the centroid of the area of Fig. A22.3 L:15.0 In
material for which it substitutes. The stress
which this stringer develops is then the SOLUTION:
average stress tor the ~terial it replaces.
Fig. A22.4 is an exploded view of the half-
'2, the combing stringer, being simply the panel Showing the placement and numbering of the
main continuous stringer bordering the cutout. internal generalized torces (Art. A7.9, Chapter
A7) and the external loading. Note that the
'3, another substitute stringer, this applied axial stresses were assumed to be con-
one replacing allot the effective material stant chordwise, giving stringer leads ~ro­
made discontinuous by the cutout. It is located portional to the stringer areas; their sum is
at the centroid of the material it replaces, P JiI , one ot external loads.
and its stress is the average stress for this
same material." The applied edge shear flOWS, coming fr~
the spar web, were assumed constant spanwise,
The sheet thicknesses used are the saQe as from a constant shear load. Other load
as those of the actual structure ..... distributions may be tandled by allowing these
applied shears to vary trom ~anel to panel.
For very extreme load variations additional
.. An alternate id.ealization, in which stringers iH and #3 are transverse members could be inserted to create
located along the lines AB and CD, respectively (Fig. A22.2), more spanwise panels allowing a better fit to
was used in Reference (2) for a box beam loaded in torsion.
the spar shear variation. The applied shear
.... When the longitudJ.na1 members themselves contribute to flows were considered as the other external
the shear stillness of the cover (as is the case for "hat" section load and deSignated P~.
stringers riveted to the skin so as to form small closed
cross sections), an effective thickness must be used. This ?anels on the centerline have zero shear
point is discussed in Reference (3). In this source, however,
the increase in shear stiffness is accounted for, not by
cue to symmetrJ (Fig. A22.3).
increasing skin thickness, but by decreasing the panel width •
an equivalent procedure.

--_._-_
-- ..
"-'--'--
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A22.3

The hal: ,anel was twice indete~inate.

Member loads qll and qa were selected as redun-
ql, -r"" 359 (P 2+3PJ.L)
dants. With these set equal to zero, successive
qu=·108 (Pa +3PJ.L) r apOlications of loads PJ. and Pa were made.
ql.lI "',533 (P a +3Pl. L)
Also, suCCeSS1?8 applications of unit values of
qll and qa were made. The results:

~~ 1 2 6 8

1 4.755 .316 -1.68 - .840

2 9.010 .316 - .840 -1.68
3 -19.76 -1.316 1.68 .840
4 -39.51 -1.316 .840 1.68
5 a - .268 1.00 a
6 a a 1.00 a
Fig. A22.4 7 -12.09 - .268 a 1.00
8 a a a 1.00
9 - 1.317 - .0450 .1l2 .056
Experience has shown that for symmetric 10 - 1.317 a - .056 .056
panels Symmetrically loaded it is satisfactory
to consider transverse members to be rigid in II .711 .0450 - .056 - .112
their own planes (4). Thus, in this problem, 12 a .0355 a a
member flexibllities for forces qll' qll' q, and
qa may be taken to be zero. In the actual
13 - 1.599 - .035q a a
NACA test specimen with which results are to 14 a .359 a a
be compared, those transverse members bordering
the cutout appeared to have been heavily re- 15 a .108 0 a
inforced (to an extent unknown to the writer). 16 a .533 a a
Hence it is logical to take their stittnesses
as great. 17 16.16 .359 a a
18 4.86 .108 a a
Member fleXibility coefficients were
collected in matrix form as below using the 19 23.98 .359 a a
for.nulas of' Chapter A7. Note that the co-
efflvlents for subscripts 5, 6, 7 and 8 were
set equal to zero (rigid transverse members).
=C ij • MEMBER FLEXIBILITY COEFFICIENTS (E 1) (Voidl denote zeros)

I.
;a

~ 1 2 3 • 5 6 7 6 • 10 11 12 13 l' 15 16 17 16

1 14.22 3.56 3.56

2 3.56 14.22 3.56
3 47.17 11.79 11. 7
•5 11.79 47,17 11.
6
7
8
•
10
7 024
7 024
11 7 024
12
13 8 0
14 3.56 '1.11
15 11. 7 23.58
16 4.785

I.
17
18
3.56
11. 79

The "off dtagonal" values have negative sign because the sense of those internal generalized forces haVing subscripts (14). (15).
7.11
23.58
'.7

(17) and (18) was taken OPPOsite to that used in the derivation in Art. A7.10. A change in sense requires a. change in sign In
off-diagonal coefficients only.)
A22 4. ANALYSIS 0 F SPECIAL WING ,
PROBLEMS

,
The following matrices Nere rorned:

[-ern]
Per eq. (17) of Chapter A8:

= C"5206
-6633
-262.'0
-262.8
d('"
.93 •'
~", .•

J' I 1111'c?1 Y 7
>i'
."
~. I
, .
.~

"'1,( 11/ J(IJfl

5; '5~15/'5~5 l
Per sq. (18) of Chapt.e r AS:

rc r s]
= [392.1
293.5
293·U
392.1 II i .5
i
The inverse or thIs last was found; CALCULATED MEASURED
Fig. A22.5 Comparison between calculated and
~rs -~=10-[5.802
-4.343
-4.343J measured stresses (psi) on the
5.802 hall-panel.

Finally, per eq. (23) of Chapter AS, the

unit load stress distribution was,
inserted in theOCij matrix (in place of the
zeroes used above) were

~ 1 2

1 -10.93 -.65
~ 5 6 7 8

5 18.03 5.04
~i~
2 -18.34 -.65
.. - 3 - 4.08 -.351
= 5 5.04 10.08
4 -11.66 -.351 7 18.03 5.04
5 1.40 .115 8 5.04 10.08
6 1.40 .383 The problem. was solVed retracing the same
7 3.79 .115 steps as before, but us Ing rthe mOd1fIedf=i~
matrix, to yield the stresses:
8 15.88 .383
[G1m] = 9 - .270 .019 X 1 2
10 - .510 0 1 -11.11 -.621
11 - 1.15 -.019 2 -16.83 -.621
12 0 .0355 3 - 3.90 -.379
13 - 1.60 -.0355 4 -13.17 -.379
14 0 .359 5 2.14 .104
15 0 .108 6 2.1'4 .372
16 0 .533 7 2.52 .104

~ImJ FLEXIBLE
17 16.16 .359 8 14.61 .372
=
18 4.86 .108
RIBS
9 - .26 .017
19 23.98 .359 10 - .62 0
The above analytical results are compa.ed ~ 11 - 1.05 -.017
with NACA test data (1) in Fig. A22.5 for the 12 0 .036
loading P II = 1, P:l. = O. Agreement is seen to
be good. 13 - 1.60 -.036
14 0 .359
EFFECT OF RIB FLEXIBILITY:
15 0 .108
To investigate the influence of rib 16 0 .533
fleXibility, the problem was reworked assuming
aluminum rib caps, of constant area A s .25 in., 17 16.11 .359
as the transverse members bordering the cutout. 18 4.86 .108
The appropriate member flexibility coefficients
19 23.98 .3<i9
 ANALYSIS AND DESIGN OF
Comparison of this result with the pre-
vious one for the rigid ribs reveals that the
most important effect of rib flexibility was
to increase the concentration of stresses in
the combing stringer bordering the cutout. It
should be noted, however, that for this sym-
metric panel, the use of a very fleXible rib
as compared With a rigid rib led to stress in-
creases of the order of only 10% in the combing
stringer. Thus, the "rule of thumb" that trans-
verse tlexibilities may be neglected in sum-
metric panelS is re-affirmed.
A22.3 Shear Lag Analysis of Box Beams
"The bending stresses in box beams do not
always conform very Closely to the predictions
of the engineering theory of bending. The
deviations fram the theory are caused chiefly
by the shear deformations in the cover of the
box that constitutes the flange of the beam.
The problem of analyzing these deviations from
the engineering theory of bending has become
known as the shear lag problem, a term that is
convenient thou~~ not ,very descriptive." (3)
Fig. A22.6 illustrates the basic problem.
The beam cover sheet is loaded along the edges
by shear flows fram the spar webs. These shear
flows are resisted by axial forces developed
in the longitudinal members (spar caps and
stringers). According to elementarJ consider-
ations, the stringer stresses should be uniform
chordwise at any given beam station ("ele-
mentary theoryff in the figure). Actually, the
central stringers tend to "lag behind" the
others in picking up the load because the
,r, '
I
, in the top cover tapering linearly from root to
:"" 1--"
--Elementary
Theory
- - -Actual stress
Fig. A22.6
intermediate sheet, which transfers the loads
in from the edges, is not perfectly rigid in
FLIGHT VEHICLE STRU.CTURES A22. s
shear. The action may be comprehended readily
by visualizing an extreme case: a large degree
of ~lagff would occur if the load transferring
skin were made of a highly flexible material
such as a plastic sheet or even rubber. In such
a case the inside stringers would be out of
action almost entirely! With the inside stringer
stresses lagging, the outSide stringers and spar
caps must carry an over-stress to maintain
equilibrium (~actual~ in the figure).
Fig. A22.7 shows the beam analyzed herein.
-j3@8"1-
12"
-+
-L
6 '5"
A ...3497
Flg. A22.7
The beam is an idealization of one tested
by the NACA and reported in reference (3). Note
that the beam has no lower cover sheet and that
it is symmetric about a vertical axis. Trans-
verse bulkheads are located at stations 12", 24~,
36 ~ and 48 ~ from the root.
The actual beam spectmen had three more
stringers than shown in the idealized structure,
these being located one each midway between the
pairs or longitudlnals shown on the beam cover
of Fig. A22.7. In the idealizing process, these
extra stringer areas were divided equally be-
~Neen adjacent longitudinals. The stringer
areas shown are the effective areas, with those
tip. All skin was considered efrective in
carrying direct stresses.
Same detailed discussions of the techniques
of idealization of practical beams are given in
references (5) and (sa).
SOLUTION:
To permit the handling at the calculation
in a limited space, it was elected to analyze
9 ,..--,
'0.J ..)
A22.6 ANALYSIS OF SPECIAL WING PROBLEMS

the structure for a single transverse (vertical) to zero, the stress distribution due ~o a one-
tip load symmetrically placed. In that case, half pound load at the tip ( a unit load
because or symmetry, it was necessary to treat divided equally between beam halves) was readily
only one-half of the structure. In addition, computed.
no shear flows could appear in the middle .07692
panels. Further, it is known that t~e influence
of rib fleXibility on shear lag is slight for 0
symmetric systems, so that the ribs were con- -.9230
sidered rigid in their own planes; hence no
generalized forces were needed on the ribs to .9220
describe their strain energies. 0
Fig. A22.8 shows the placement and number- .07692
ing ot the generalized forces on the halt-be~~. 0
-1.8461
1.8461
0
~lmJ = .07692
0
-2.769
2.769
0
.07692
0
-3.692
3.692
0
fleXibility coefficients were
Memb~r \.
computed with the fonnulas of Chapter A7 and
arranged in a matrix. Next, the unit redundant stress distribu-
tion was computed. Fig. A22.9 illustrates a
The shear flows q., q, q:La and q u were typical calculation, showing the stresses in the
selected as redundants. Setting these equal t t p bay for qll = 1, q"l' = q1.ll = qu = O.

OC\ , MEMBER FLEXIBILITY COEFFICIENTS (E -= 1)

I}; 1 2 3 4 5 6 7 6 9 10 11 12 13 14 15 16 17 18 19 20

1 2'71
2 16 6"10
3 1 .74 2.684
4 0.0 4. 724
5 17.99 4.202
6 3271
7 16 640
9 2.684 8.939 1. 785
8 4.724 17.84 4.255
1 4.202 15.78 3.708
11 32 1
12 16 640
1 1. 785 7.142 I. 785
14 4.255 16.084 3.840
1 3.708 13.89 3.324
16 3271
17 16 540
16 1.785 3.571
19 3.840 7.518
20 3.324 6.439
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A22.7

rrj = r 175
-1848
-1260
- 460.7}

16,810 9,536
3'~~
r- =
[,210
16,810
9,536
3,041
32,160
9,536
3,041
9,536
25,030
3,041
3,041
3,041
18,650

The inverse at the latter was formed:

- ,1m
~
' 285 -L505 - ,656
r:". -~ = 10-' -1.505
L.:: s.J - ,656
4.219
-1.000
-1.000
4.687
- .279
- .495
Fig. A22.9 Application of a sell-equilibrating unit - .183 - .279 - .495 5.519
redwl.dant stress qll.
and, finally, per eq. (23) of Chapter AS,

The complete redundant stresses were:· .07692

.03453
X 2 7 12 17 - .9230
.5086
.4144
1 a a a a .07692
2 1 a a a .03136
-1.846
3 a a a a 1.055
4, -12 a a a .7907
.07692
5 12 a a a .02401
6 a a a a -2.7691
1.690
7 a 1 a a 1.079
a a a a a .07692
.01005
9 -12 -12 a a -3.692
= 10 12 12 a a 2.493
1.199
11 a a a a
12 a a 1 a Fig. A22.10 shows the above computed
stresses and those reported by the MACA as ob-
13 a a a a tained by test. Agreement is seen to be qUite
14 -12 -12 -12 a good. 4.589- ,.1.S85
15 12 12 I 12 a ~ 1. SSS

16 a a a a CALCULATED
17 a a a 1 3.416
a 2.364
18 a a a
1.278
19 -12 -12 -12 -12
20 12 12 12 12 5, 1

The following matrix products were formed

(per eqa , (17), (18) of Chapter AS):

• This is an obvious place in which to use combinations of MEASURED

redundants to decrease the structural coupling (reference
Chapter AS, pp. AS.29, 30). (Recommended as an exercise
lor the student.) Fig. A22. 10 Comparison between calculated and measured
stresses (psi) in a box beam
A22.8 ANALYSIS OF SPECIAL WING PROBLEMS

A22.4 Stress Analysis of a Box Beam With a Cutout.
In Article A22.2, one technique was em-
ployed for computing the stresses around a cut-
out. In that analysis the effect of the cutout
was presumed to have been localized about the
cutout region; consequently, the problem was
treated by isolating the affected panel. QUite
often, when the cutout is placed well inboard
on the wing, its influence on the root stresses
is appreciable. Therefore, it is desirable to
be able to consider the overall ;roblem of the
box beam with a cutout for such cases.
nThe most convenient and the most rapid
method of analyzing structures with cutouts is
the indirect, or inverse, method. The analysis
by the indirect method is made in two steps.
First, the structure is analyzed for the basic
condition that exists before the cutout is
made. The results of this basic analysis are
used to calculate the internal forces that
exist along the boundary of the proposed cutout.
External forces equal and OPPosite to these
internal forces are then introduced; these
external forces reduce the stresses to zero
along the boundary of the proposed cutout, and
consequently the cutout can new be made without
disturbing the stresses. n (3)-
T~~ calculat10n :or ste~ (2) above, will
new be carried out oy conslder1ng a unit shear
flew, q12 = 1, applied ~hile all other loads
are zero. Under this loading con~iticn the
relative displacements at rec~~dant c~ts 2, 7
and 17 are equal to =<2,1'1' 0<::", 1aand OO:::l",,~a
respectively, where these coef:!cle~ts have
been computed pr'evtoi.rs Ly 1.:1 Ar t i c Ie ";22.3.
To restore continuity (reduce the relative
displacements to zero) th~ee red~~dant ~0r:es
are applied, one eaCh, at the cuts 2, 7 ~nQ 17.
The appropriate equations spec:~y1ng cont~nu1ty
are
Note that the~e equations say si~?ly that
the deflections at the reduncant cuts due to
the (unkncwn ) r'ecundant forces must be equal
and oppoat t e to the deflections due to q::.:a = 1.
All coefficients in the above equation
were computed in Art. A22.3. SpeCifically,

It is desired to ~odi!y the calculation of

the preViOUS article (A22.3, nShear Lag
' ,210
16,810
16,810
32,160
3,04;]
3,041
{q,}
c, .:: -
99,536
'036}
{3,041
[:
Analysis") to allow for the presence at two 3,041 3.041 18,550 q;:,-;o

cutouts symmetrically placed. Panel nq.~n was

removed while the Single tip load remained. The a~Jve matrix is the [ocrsJ of Art.
Since the unit remained symmetric, the data A22.3, with the nq u TOWn and nq u column" re-
tram the previous analysis, in which the trans- moved.
verse rib stitfnesses were taken to be
infinitely great, should still yield satis- The equations were salved·· to give the
tactory results. values of the redundants for a unit applied
load, q.2 = 1, as
SOLUTION:
.1399
The calculation was accomplished in three
steps: .2134
= 1.00
1) The stress distribution was found in
the "basic structural! (no cutout). This work - .1055
was carried out in Article A22.3 where it was
found that q .. = .024011bs./inch. The q12 force 1s included in the above for
later convenience. Then the complete stress
2) A stress distribution was found for distribution, due to applying a unit shear flow
an napplied load" ot q u = l. Such a loading q•• = 1, was
has zero ~xternal resultant.
3) .02401 times this last stress distri-
bution was subtracted from the ~irst •

• The procedure described here is quite generally useful for •• See Appendix for a method of "extracting" the inverse of
studies of the effect of remoVing one or more members; this matrix from that previously found for the complete
such might be required for an analySiS of the effects at
structural damage. ~rJ matrix.

_,_ .• _ - - , . . ' _ C-=~, ~._-

-
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A22.9

where the row matriX ~12,mJ is simply the w12

r ow" or@im]BASIC

o COMPARISON WITH TEST DATA:

- .1399
o Reference 3 reports test data for the case
1.579 analyzed, the stringer stresses being plotted
-1.579 ~elow in Fig, A22.ll.
o
- .2134 In the actual test specilnen a stringer
o passing through the cutout was severed, it hav-
4.240 ing zero stress at stations 12 and 24, there-
-4.240 fore. However, during the idealization process
o discussed in Art. A22.3 (tor the beam without
1.00 cutout) the area for this stringer was placed
o partly with the combing stringer and partly
-7.760 with the spar cap. In the same 'NaY, some
7.760 effective sheet trom the midportlon at the
o panel, now made discontinuous by the cutout,
- .1055 was added also to the spar cap and combing
o stringer.
-6.494
6.494 It follows then that the full idealized
areas at the combing stringer and spar cap
( [glrJ taken from Art. A22.3) ShOUld not be used in figuring the stresses at
stations 12 and 24 (produced by forces q,J
q~o, q14' q1e)' ~ith these areas reduced by
To obtain the stresses in the loaded bea~ the appropriate SUbtractions, the stresses
wi~h the cutout, 0.02401 ~.nes the above were computed and are plotted in Fig. A22.l1.
w
stresses were SUbtracted from the wbasic Agreement with test data is seen to be qUite
stress distribution of Art. A22,3, as pre- satisfactory.
viously eX~lained.
.07692
.03789
- .9230
.4683 CALCULATED
.4547
.07692
.03648
-1.8461
.9536
l = ks l
{ GImjcUTOUT( -.02401{G l: .8925
L'nfBAS IC in.:;J .07692
o
-2.7691
1.8763
.8927 MEASURED
.07692
.01258
-3.692 Fig. A22. 11 Comparison between calculated and measured
2.649 stresses (psi) in a box beam with cutouts.
1,037

Note that q1~ is now zero and the cutout
panel may be "lifted out w•
In the case of a structure under a variety
of external loadings (m = 1, 2, 3 ---), the
~ore general equation, carresponding to the
above, i s
IG l = IG l _fG} L G ,,,,,J
L imJcU'l'OUT li;nJBASIC 'liJl.ll
A22.5 Analysis of a Swept Box Beam.
WExperi~ental investigations of swept box
beams have shown that the stresses and distor-
tions in a swept wing can be appreciably dit~er­
ent in charac~er from those that would exist if
the ~oot were ~o~l to the wing axis. The
principle ef~ect of sweepback on the stresses
occurs under bending loads and consists in a
concent~atlon ot bending stress and vertical
A22.10 ANALYSIS OF SPECIAL WING PROBLEMS

shear in the rear spar near the fuselage. With
regard to distortions, the effect of sweep is
to produce some twist under loads that would
produce only bending at an unswept WIng and
some bending under loads that would produce
only twist of an unswept wir~.~ (6)
In the following example a swept box beam
is analyzed by the matrix methods of Chapter AS
and, in particular, by the specific techniques
of reference (7). The method accounts for the
interaction between the swept cover panels and
the longitudinal members. It is this action
that is responsible for the distinctive
structural craracteristics of the swept box
been ,
Again, we emphaSize that the method used
here 15 strictly applicable to thln-sklnne~
Wings at beam-like proportions only. Consider-
ing the wide variety at structural layouts
which may be employed in swept w1ng configura-
tions, a comprehensive treatment carillot be given If loads were to r4ve been a~pl:ed inboard of
here. An excellent review at methods better
adapted to thick-skinned construction and to
"plate-like" (very thin, Wide) wings, may be
found in reference (8). One method of ~n~lyz­
ing such wings is given in Chapter AZ3.
THE STRUCTURE:
The structure shown is F:g. A22.12 is a~
idealization at the NACA test beam of refer-
ences (6) and (9), in which a Single substitute
stringer has been employed along the cover
sheet ta allow for the ant:tl~ated shear lag
effect. The figure shows only one-half or tte
complete unit, which was built symnetrically
about the axis corresponding to the langitudln~:
axis at the airplane.
Only tip loads were to be applied (at
points A and B). The outer section of the be~
was assumed to carry stresses which could be
calculated reasonably well by the engineeri~g
theory at bending (E.T.B.). For this pur,ose
it was jUdged satisfactory to consider the
outer 66~ at the Je~ as a Single bay (A-B-D-CJ.
the tip, it would :iave been necessary to con-
steer additional bay dtvf s i ons between .;'-3 and
C-D (that is, insert additional ribs at statlcGs

f
/"-.

52"

.050 .050

--i.e H

,..A-.539 t,,-
Section C-C
.050 .050 (Steel) :;--t=. 125

66"

.078 t-,
A A
t·l +
J A=l. 966 A=I.430

A 1) (2) B_1 .078 -f " c

Section B-B
, r-'~,
:> .L
~..d5" .[.-15" ~
,

Flg. A22.12 Fig. A22.13

 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A22.11

of load application). Rib C-D was located at
one of the actual rib locations in the NACA
test specimen and was assumed rigid in its own
;>lane.
The choice of bay C-D-F-E as a single bay
was somewhat arbitrary. For improved accuracy,
additional ribs inboard of C-D could have been
used. Note that any ribs placed inboard or
point F will produce triangular skin panels in
the cover sheets. Examples of treatments for
such panels may be found in references {9},
(10) and (11).
Rib E-F was considered flexible in its own
plane, it being known that the flexibility of a
rib is important at a locati0~ where a structure
changes direction.* Note that this rib was
made at steel in the test speCimen.
Effective areas of longitudlnals as shown
in Fig. A22.12 were computed by considering all
of the skin to be effective. The spar cap
areas are equal to the sum of the areas of the
angle member at the cap location, plus one-
half of the effective area of material be~Neen
the cap and the substitute stringer (this area
i~cludes several stringers as well as Skin)
plus one-sixth of the attached spar web area!*
The substitute stringer area was collected in
like ma~~er from the half-panelS to either side.
affects total deflections only and can be
omitted in a stress analyses where ceriect icns
ar-e not sought.
Since only symmetric loadings were con-
sidered in this analYSis no shear was trans-
mitted by the carry-through bay and hence no
shear flows were shown in that portion.
Sets at additional axial forces (q18
through q1e:) were applied to the ends or the
flanges and stringers adjacent to the Obliquely
cut ends of the cover sheet panels in bay
C-D-F-E. These forces are necessary to account
for the interaction between the swept covers
and the longitudlnals. As shown in Fig. A22.14,
the pure shear flow on the oblique edge is ob-
tained by superposing onto the panel a zero-
resultant system consisting of a uniform tensile
stress of intenSity 2q plus a pair of concen-
trated balancing loadS. The balancing loads
must be contributed by the bordering longitudl-
nals and hence react on these as tensile loads
(Fig. A22.l4C). The balancing loads applied to
the stringers are shown dashed since they are
internal forces within the bay and are not to
be entered into the equilibrium equations for
the structure.
P=qL

The method used in calculating the

effective areas of the rib caps (E-F) is given
in detail in ~efersnce (9), from which the
va lue used here was taken. The "carry-through
2J
P"'qL
(a) (b) (c)
bay" cover sheet thickness is equal to tnat·
used on the specimen (.050") plus a weighted Fig. A22.14 Showing how the uniform. shear stress on
increase to allow for the presence of splice an oblique panel end (b) is created by
plates along the plane of symmetry (see refer- superposition of a uniform tensile stress
plus two balancing forces (a). The
ence 9). balanCing forces react as tensile loads on
the bordering longitudin:aJ.s (c).
INTERNAL GENERALIZED FORCES:

Fig. A22.13 shows the choice and numbering From an energy Viewpoint, these dashed
of the generalized forces. forces account for the additional strain energy
stored by the axial components or shear flows
The beam was rigidly supported at points in the non-rectangular panelS. This energy is
E, F and at the two corresponding points on the stored in the cover panels themselves (and is
other beam half. These might correspond to the accounted for in this manner since the longl-
:uselage ring attach points in an airplane. tudinals contain the effective area irom the
The vertical end caps on rib E-F were considered cover sheet·** ) and in the longitudinals Which
rigid ~xially, so that no fleXibility co- react against these components.
efficients were associated with the reactions
qal5 and qae:' Flexibility of these members "Dashed loads 11 are applied to the longi-
tUdlnals adjalsnt to any obliquely cut panel
end. Similar dashed loads would be applied to
• The effect of neglecting this rib's flexibility is demon-
strated later in this example.
-** This much of the energy could be accounted for in another
** The factor of 1/6 is used so that the effective area con- fashion by modifying the member flexib1llty coeUlcient for
tribution of the web results in a structure having the same the sheet panel. See Reference 12. where this was done.
moment of inertia about a horizontal ax:I.s as the original. However. that referenc e incorrectly neglects the additional
Some of the problems of idealization are discussed in energy stored in the longitud1nals, as was demonstrated in
reference (2), p. 16. Reference 13.

,_. - ~ ---

.~i~;~';~.~~:~~2~;~~~'~~;~~1k:·;(~~~~~:{;>i~~\;~~~;:~;~~~;~f);:~,~·~·:;,~;.:~.';~: . ~. '. {yZJ!;::~7 ~.~{~~

A22.12 ANALYSIS OF SPECIAL WING PROBLEMS

the outboard ends of the panels in bay C-D-F-E
if they too were cut obliquely. Such ~anel
configurations arise often in swept wing con-
struction having ribs parallel to the air-
stream. Formulae for ~ore general quadrilateral
panelS are given in Reference -
'!'HE STRESS DISTR!BUT I O,:S
For the symmet.r-t c rcadmcs considered
here the structur e was indeterminate only two
times since the auter bay was assumed to be
determinate by the S.T.B.
Stresses in outer bay by S.T.B.:
Flange stres ses at rib C-D (for both
P1 = 1 and Pg = 1 )
Stresses in inner bays:
According to the discussi~n of Art. A8.12,
Chapter AS, the deter.ninate stress d~stributicn,
§tmJ, may be any stress distribution in 8qU~:'­
ibrlum With the applied loads, and preferably
one close to the final true stress distribution.
The magnitUde of the redundant fcrces is re-
duced by use ~f ~ satisfactory estiillate of the
true stresses.
The stresses in ~he two ~nner bays were
determined for both gim and glr si~ultaneously.
Since this inner portion of the structure is
two times indeterminate we can estimate two
loads. For this purpose the two flange loads
q 111 and q lo4 were written as

II .. 66"* qu = q~11 -+ q~11

I = .
88.57 i n •., C = 3.5" qh = q~ .... q~.
IlC
to -1'" 2 .608 psi where the (Single) pri~ed values are approxi-
~te values determined by the £.T.B. and the

Therefore, double primed values are the unknown corrections

(the redundants). Using My/I at stations 66"
qa = q, = 2 .608 x 1.121 = 2.924 lbs. and lIS" from the tip gave "
q. = 2.608 x 1.373 = 3.581 Ibs. q~11= 5.228 p~ -+ 5.228 PI1
For a unt t transverse load at the shear qi. = 3.899 P~ -+ 3.899 Pl1
center (midpoint , because of symmetry)
The eq~illbrium equations for the elements
of the structure were written next by summing
Qa =~=
Ibs
2x66 • 02713 Trl forces and moments.

2.924
Joint F
E'% qb= .O2713+~
q., + q •• = 0
... 07143 Ibs q• .3 + qlla = 1.414 q16
Til
= 5.513 P~ + 5.513 PlI + 1.414 q~"
The unt t load was shifted 15" to either side
by application a r a torque, T = 15 in.lbs. Joint E
The unt rcra shear flow superposed was
q.. + q •• = 0

q=~=2X
15
210 = •
03571 Ibs
in
q•• q.. = 1.414 q..
= 7.392 P, + 7.392 p. + 1.414 q~2

Finally, su perposition gave the stresses m about E-F

tor the outer ba yas
q.. + qu + q•• = 11. 918 P, + 8.888 p.

I1Z 1 2

1 .1071 .03572
2 .06284 - .00858
= 3 .00858 - .06284
4 .03572 .1071
... This is a rather crude way to estimate these loads and is
5 2.924 2.924 used here only for simplicity. The analyst is generally better
advised to exercise a little more ingenuity in making these
6 3.581 3.581 estimates, even to the extent of being guided by other swept
7 2.924 wing solutions.
2.924
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A22.13

E 1 0 0 0 -1 0 0 q.
~ 1 -.7115 0 0 0 0 0 q.
G
• G
•
"

01'
0 0 21.2 0 21.2 0 1

~j " "'j q""

l~ iN==.
F ,;
~
~
~.
0 0 1,682 1 0 0 0 q" =
~
0 0 0 1 1 0 0 q ..
C H ~

q. q" 0 -21,2 0 0 -21.2 1 0 q"

Shear Flows around the non- 0 0 0 0 0 -1 1 q ..
rectangular panels. (Check
by summing moments about
E, G and F.) - .1403 - .2111 o o
.04431 .04431 .01923 o
Cap EC
2.756 2.756 o .707
qe - .7115 qa = .04431 P:l, + .04431 Pa + = .04432 .04432 o .04545
.01923 q~a .2831 .3540 o o
Rib Vertical at E -3.696 -3.696 -.707 o
-5.466 -2.436 .707 .707
q. - q .. =- .1403 Ii - .211 p.

Cap EX}
Atter inverting the matrix ot coefficients
on the lett hand side at this last equation,
and multiplying through thereby, the stresses
were obtained as
Joint G
c, .1006 .0295 .00676 -.00676
.707 q:l,::) q •• =0 q. .0791 - .0208 -.01753 _.00950
.707 q1,,) q .. + q .. = 0
q" .0013 - .0411 .00401 .0230
Cap GF q" = .0422 .1135 -.00675 .00675
q.. .2409 .2409 .00676 -.00676
qll1, 3.087 .964 -.9355 -.3444
Rib Vertical at F q.. -2.379 -1.473 -.2284 .3625

The complete determinate and unlt-

Cap DF redundant stress distributions, USing the re-
sults up to this paint, were therefore:
ql.l + 1.662 q1,Q = .04432 (p1, + P II )

+ .04545 q~40
The fi~st fIve of the above equations
were readily solved by substitution, y~elding:

q" = s., = 2.756 P, ,.. 2.756 P, + .707 q~40

-qllo = q.. = 3.696 P, + 3.696 P, + .707 q~ "
q.. = 5.466 p" + 2.436 P, .707 q"u
- .707 q~ ..

Also, from equll1brl~ of joint G:

The remaining seven equations were

arranged tor solution thus:

- ~!
-437-
~'"'"
 A22 14 ANALYSIS OF SPECIAL WING PROBLEMS

52.3 1-
q.. = = 2
~Pn~
X (- ) q , = -12.5 q,
1"'( 1 2 12" 14" X '25
• - x
.0 :::I

1 .1071 .03572 0 0
= -.989 P lo • .260 Fa + .2191 q~a • .1168 q~
•
2 .06284 - .00858 0 0 q.. = ~Pclt = 73.5
2 X 125.8 '5
X ~ X I-I qe = -17.5 q,
3 .00858 - .06284 0 0
4 .03572 .1071 0 a
'"' -1.384 Plo • .360 r, • .3068 q;a • .1662 q"
Similarly, for the panel HGFD;
5 2.924 2.924 0 0
6 3.581 3.581 0 0 qlo7 • -.0146 Plo •• .773
.460 P a -.0449 q~a - .2576 q~

7 2.924 2.924 0 0 q l.8 = -.0244 Pl, ?a -.0754 q~a - .4324 q~

The following matrix were rornec:

8 .1006 .0295 .00676 - .00676 (c r • eqs . 17, 18, 23 of Chapter AS)
produc~s

9 .0791 - .0208 - .01753 - .00950

10 .0013 - .0411 .00401 .0230 [<><r,J = [ -173.7
17.70 250.2J
- 50.43
11 .0422 .1135 - .00675 .00675
[-=rJ = [114.5 42';IJ
12 5.228 5.228 1.0 0 42.96 83.04
1m:gl~ 13 7.729 3.444 -1.0 -1.0
C=rsj =[ - 0.01082 - .00556~
14 3.899 3.899 0 1.0 .005563 0.01483
15 - .989 .260 .2191 .1188
Finally, the true stress dis~ributlon was found
16 1.384 .360 .3068 .1662 as,
17 - .0146 .460 - .0449 - .2576 X 1 2

18 .0244 .773 - .0754 - .4324 1 .1071 .03572

19 .2409 .2409 .00676 - .OC676 2 .06284 - .00858
20 3.696 -3.696 - .707 0 3 .00858 - .06284
21 3.087 .964 - .9355 - .3444 4 .03572 .1071
22 2.379 -1.473 - .2284 .3625 5 2.924 2.924
23 2.756 2.756 0 .707 6 3.581 3.581
24 3.696 3.696 .707 0 7 2.924 2.924
25 5.466 2.436 - .707 - .707 8 .0747 - .0051
26 2.756 2.756 0 .707 9 .0740 .0112

DASHED LOAD CALCULATIONS:

10 .0581 - .0039
11 .0681 .1481
In the above matrix, loads qu' ql, 5' q.. 12 4.070 2.241
and ql,a were obtained tram q. and ql,O following
formulas given in reference 7 for general
quadrilateral panels. The equations applicable
~1mJ = 13 5.213 4.292
14 6.573 6.038
to a parallel-sided panel are:
15 - .925 - .140

/', 16 -1.295 - .201

-
v: 0)'
d 17 .6514 .0431
Pc "'c:;a qL x 2
18 -1.093 .073
c
.P~T
P D", c + d qL x 2 19 .2150 .2060
pdq
r L Balancmg Load 20 -2.877 -1.584
l),.;... --..L (= D~shed. Load)
ForJP1i1ae from Ref. 7 21 3.249 3.022
22 -1.145 - .015
For the panel CEnH; c = 52.3", c = 73.5" , 23 4.646 4.268
c t d = 125.8", L ~ l5~. Hence 24 2. 877 -
1.:::184
25 4.394 3.036
26 4.652 4.268

..iiiiiiiiiiiiiiiiiiiiiiiiii
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A22.15

.,
MATRlX OF MEMBER FLEXIBIUTY COEFflCIENT-'· (E = 1) {FormulaB from Chapter A7:lnd Re£erenee 7)

1 , 3 4 5 , 7
• s 10 11
" ia 14 15
" 17
" 19 ac 21 ! 22 " ,.
" "
15.400

1 2 60
1 40
70.170 1 .46 15.46
50.0 8.984 8.984 8.984
52.33 6.542 6.542
112 130
• 400
46000
,
5 0
15.46 30.92 30.9Z
.984 17.97 17.97 17.97
.42 .08 13.08
15.46 30.92 30.92
8.984 17.97 17 97 17.97
8.984
6.542 ." 13.08
17.97 7.97
13.08
2130
9.046 4. 2
, 4 '" 04 4.523
4.523 9.046
s.
4
•
19.58

CO!lPARISON WITH TEST DATA: requirement. More details concerning the test-
ing teChniques and method at data presentation
5.2 would probably resolve this conflict. Both test

/
~.6

2~
3.8
Root

2.6
2.6

-,
118)......
2.8

3.2
f\

2.'
2.'
~.6
2.'
and calculated values clearly exhibit the
characteristic build-up at stresses in the rear
spar of a swept Wing.
RIB E-G-F RIGID:

As a matter at interest, it was decided to

/ (66) investigate the effect upon stresses when rib
E-G-F is taken to be rigid. Such a calculation
is readily achieved by putting the member
Leading ........... flexibility coefficients for the rib equal to
Edge
't' zero .
CALCULATED
/
Station (0) ...-
MEASURED Thus, in the matrix those co-
etticients with Subscripts 19, 20, 21, 22 and
E: IjJ
Fig. A22.15 Comparison Between Calculated and 23 were set equal to zero and the complete
Measured stresses (psi) in a. Swept
Box Beam. calculation was repeated.
The results, tor spar cap loads at the wing
Fig. A22.15 shows a comparison between root (a most sensitive point), were:
the calculated stringer stresses and those
measured by the NACA as reported in Reference 6.
The stresses showa are for a unit tip load,
centrally placed (p. = p. = 1/21b.). ~ 1 2

Considering the limItations on the analysiS

~l~RIGID
RIB
= 12 3.824 1.792
the agreement Is generally satistactorj. Thus, 14 6.993 6.610
the discrepancy in the MY/I stresses at station
65 may be attributed to a lack ot precise A comparison with the fleXible rib ca1cu-
knowledge of the test parameters. The calcu- lations follows:
lated stresses in the leading edge spar between
stations 65 and 118 cannot duplicate the ex- SPAR CAP
LOADING FORCE FLEX. RIB RIGID RIB % DIFF.
perimental variation since only a single bay

I was employed in this region in the idealization.

The fact tr~t the calculated root stresses run
consistently above the test values is 1ifficult
Simple
Bending;
Flo ::P~=l
q"
q,.
6.311
12.61
5.616.
13.60
11
8
to explain. Inasmuch as the calculated stresses
satisty equilibrium, the test values, all being Torsion; q" 1.829 2.032 11
lower, would seem to defy this runcamerrta; Plo=-P~=l q,. 0.535 0.383 28 ,.
)--. t'?
,x:: /) !
A22.16 ANALYSIS OF SPECIAL
ConSidering that rib SGF was relatively
rigid to begin with - being nade of heavy
gage st3s1 - it may be seen that neglect of
the flexibility of a corresponding all-al;~inum
rib could lead to serious arrors.
REFERENCES
(1) Kuhn, P., Duoer-g , J. E. , and Diskin, J. H.,
Stresses Arolli,d Rectan~ular Cut-Outs in
Skin-8tringer Panels Under ft~lal Load - II,
NACA 1rJR w6B (ARq 3J02), Oct. 1943
(2) Rosecrans, R., A Method for Calculating
Stresses in Torsion-Box Covers with Cut-
outs, NACA TN 2290, Feb. 1951
(3) Kuhn, P., and Chiarito, P. T., Shear L3.g
in Box Beams: Methods of Analysis and
Experimental Investigations, NACA TR 739,
1942
(4) Haft, N. J., and Libby, P. A" Recommenda-
tions tor Numerical Solutions of Reinforced
Panel and Fuselage-Ring Problems, NACA TR
934, 1949
(5) Kuhn, P., Approximate Stress Analysis of
Mu1ti-3tringer Beams with Shear Defor.natlan
of the Flanges, NACA TR 636, 1938
(sa) Kuhn, P., Deformation Analysis of ~lng
Structures, NACA TN 1361, July 1947
(6) Zender, G., and Llbove, C., Stress and
Distortion Measurements in a 45° Sweot Box
Beam Subjected to Bending and Torsion,
NACA TN 1525 ..
WING PROBLEMS
(7) weme , L. B., and Lans tng , 'A •• A Metccd ~Gr
Reducing the AnalYSiS ~! Com~lex Retun1~nt
Str~ctures to a Routine Procec~re, J8urnal
Q: Aero. SCi., 19, Oct. 1952
(S) 'williams, M. L., A Review of Certain
Analysis Meth~ds for Swe~t Wing Structures,
Journ. of Aero. SCi., 19, p. 61~, lS52
(9) ~e1denfels, R., Zender, G., and Llbove, C.,
Stress ~d Distortion Ar~lysis of a Swe~t
Box Beam Having Bulkheads Perpe~c~cular ~o
the Spars, NACA TN 2232 . •
(10) Bispl1nghoff, R., and Lang, A., An In-
vestigation of Defo~tlons and Stresses
in Sweptback and Tapered ~ings with Dis-
continuities, Mass. Inst. of Tech. Rept.,
July, 1949
(See also, Journ. of Aero. ScL. ..la, p , 705,
1951)
(11) Denke, :., The Matric Solution of Certain
Non-Linear Problems in Structural Analysis,
.jour-n, of Aero. se i .• ~' 1956
(12) Levy, S., Compirtat Ion of Influence Co-
efficients z'or Aircratt Struc-:ures witt'..
Discontinuities and Sweepback , Journ. of
Aero. SC1., 14, Oct. 1947
(13) Is Lfnger , J. S., Stress Analysis and
Stress Measurements for a Swept Back Wing
Having ~ibs P~ra11el to the Airstream,
McDonnell Aircraft Corp. Report 1127, April
1949.
i
 CHAPTER A23
STRUCTURAL ANALYSIS OF A DELTA WING
ANALYSIS BY THE "METHOD OF DISPLACEMENTS"
ALFRED F. SCHMlTI'
A23. 1 Introduction
The pur~ose at this chapter Is two-fold:
first, to present a discussion and an example
of techniques for handling the structural
analysts of highly redundant low-aspect ratio
wings (typified by the delta wing) and second,
to illustrate the nDlsplacement Method" of
structural analysis (l)~J a methOd fundamentally
different from those of Least Work or Dummy-
Un! t Loads.
The very low aspect ratio, thin wing IS a
structural configuration of relatively recent
origin. It Is a bUilt-up structure of ribs,
spars and a cover sheet but yet 1s so thin and
so highly redundant that its structural
characteristics are actually closer to those of
a tapered flat plate than ~o a beam. In a beam
the prlmarJ stresses are longitUdinal; indeed,
one of the basic assumptions of beam theory is
that transverse stresses are negligible and
that cross sections remain undetormed. For the
very low aspect ratio wing however, chordwise
stresses and deformations are of great im-
portance. The degree of redundancy of these
low aspect ratio wings is very high because of
the multiple pathS by which a load ~y be car-
ried over the gridwork at spars and ribs.
In the wing beam problem examples of
Chapters AS and A22 redundant internal loads
were determined by use of the Method of Dummy-
Unit Loads - a variant of the Least Work
theorem. ~nl1e the ~ethod is perfectly general,
it becomes increasingly difficult to obtain
satisfactory accuracy as the degree of redun-
dancy of the system increases. To some extent,
accuracy may be retained by skillful choice of
redundants and through the use of carefully
chosen determi~te stress distributions (see
Art. A8.12, Chapter AS). A high degree of
engineering jUd~ent nust be exercised. how-
~ver. SVen so, it has been found difficult to
apply successfully these Least ~ork ~ethods to
the low aspect ratio ~ulti-5par wing.
Several authors (2), (3), (4) have pre-
sented methods of aralysis for the highly re-
dundarrt wing based upon use of the "Displace-
ment MethOd" of analySiS. These references
differ primarily in the techniques advocated
Numbers in parentneeee refer to bibliograpny at end
of chapter.
A23.1
for the breakdown and idealization of the
structure into primary elements. In the follow-
ing example, a method following that proposed
by Levy (2), is applied to an idealized delta
wing structure. The choice of this method for
presentation is primarily for pedagogical
reasons - it being the least detailed and con-
sequently the easiest to grasp conceptually.
The reader who 1s interested in actual appli-
cation is recommended to Reference (4) for
techniques which are probably better able to
handle practical problems because of their
greater generality. fleXibility and growth
potential.
A23.2 Basis of the "Method 01 Displacements"
The Method of Displacements draws its
name from the fact that displacements rather
than forces are dealt with as the independent
variables. In earlier chapters the relation
between structural detlections and applied loads
'NaB written through the use of the fleXibility
influence coefficients as (see Eqs. 24, 26,
Chapter A7)
- (1)
where Om 1s the detlection at paint m and Pn is
the external load at point n. If one forms the
inverse of ~mnJ
written as
'
this equation may be re-
{P n}=Gmn-~{~}= ~ronJ {~} - - - (2)
where [KmnJ =. [Amn-:LJ is called the st1!tness
matrix.
~nile the fleXibility influence coefficients
give the deflections per unit load, the stiff-
ness TIatr~x lves the loads ger ~nit deflection.
Thus, anyone column (say the :n"') of Knm
gives the values of the forces (reactions)
developed at all numbered points ot the structure
when ~he corresponding point (paint m) is de-
flected a unit ~~ount, all other points held
fixed.
ConSicer, for illustration, the grid-Ilke
structure of Fig. A23.1 into ~hich a delta wing
structure has been idealized. Note that the
torSional stiffness of the Shear-carrying Cover
sheet has been accounted for by the presence of
 A23.2 STRUCTURAL
quadrilateral "torque boxes" connecting ~olnts
ot intersection of the spars and ribs. This
/ Rib
Fig. A23.1 Idealized Delta Wing Structure
idealization is discussed in more detail in the
example. Assume that the points of inter-
section have been numbered as in Fig. A23.2.
Fig. A23.2 Reactions Developed at a Net of
Points when a Single Point is
Displaced.
It one point, such as point 6, is dis-
placed a unit amount, the other points remain-
ing fixed, reactions are developed at these
various points as shown. These reactions may
inclUde moments as well as forces, should
rotations have been taken as cer-ement dis-
placements in addition to the translations (see
points 5 and 7 of Fig. AZ3.2). The values of
reactions so developed would for.n the 6th
column at [Knmland the 6th row, since [KnmJ is
always symmetric.
-
, ~ :,_~. _~ . ~'._" ··<'~.'-~~'.'k"". >_~,>,: ••_ .~._':
ANALYSIS OF A DELTA WING
are written for ~he individual elements of the
st~ucture, they ~re Simply ~dQed to g~ve the
stiffness ~trix of the composite. Finally,
the resulting stif~r.ess ~trix ~zy be inv8:ted
to yield the matrix of ~lexibility influence
coefficients (Eq. 2 "in r-everse"}.
We note here, ~s an aSide, ~hat t~e Monent
Distribution Method of Chapter All, the Slope
Detlec~ion Method of Chapter A12 and the~od
of Successive Corrections of Cr~pters A6 and
A15 are all ex~~ples ot the "Displacement
Method" o~ structural analysis. In each of
these methods the displacements are taken as
the independent variables and these are adjusted
to achieve equilibrium of the loaded str~cture.
The nadjUBtment n may appear as a systematic
relaxation or artificial constraints (~oment
Distribution and Successive Corrections) or it
may be dane mathematically !~ one stroke by the
solution or a set at simultaneous equations
(Slope Deflection Method). The latter approach -
solut10n of a set of simultaneous equations - 1s
essentially that :'ol1owed herein, the "so Iut t on"
being effected by ~trix inversion.
It 15 to be said of the r.ethcd ot Displace-
ments that it 1s complementary to the Least ~ork
method in that it is better suited to the
handling of the highly redundant pr-ob.Iem , While
Least Work is better suited to problems of tew
redundancies.
A23.3 A Delta Wing Example Problem
The idealized structure of Fig. A23.l will
be analyzed to determine
(a) the influence coefficients
(b) the internal stresses as a function at
the applied loadS.
The grid points are numbered as in Fig.
A23.2. Note that the numbers increase to the
rear and outboard.
IDEALIZATION:
In the structure un~er consideration mem-
ber 1-2-3-4 lies on the airplane centerline.
The bending stiffness or this member includes
that or the half-fuselage plus one-half of the
"carr/-through structure." More detailed
representations or the structure in this region
are generally deSirable, the overSimplified
model used here being employed to limit the
amount at data to be handled. Some techniques
which may be applied in idealizing the structure
in this region are given in Reference (5).
Idealization, particularly with regard to
effective skin areas, has been discussed in
some detail by Levy (2). The cc~plexities of
. -~-'.. :'--;e/'--~':~'·'>--·' ,c . . ·,·.{::·.1.. 'T'... ~,_.<.;_.":~:_,_:...,.,,::..:;-.:::
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A23.3

this phase of the problem are too great to per- Beam Element Prooerties
mit an expansive treatnent here.· Briefly,
Levy's recommendations are; (Moments of inertia, in (inches"), are
assumed to varJ linearly between numbered
(a) all skin may be considered effective points)
between spanwlse spars when computing the cap
areas of such spars. This assumption is sub- li2 Fuselage Beam (1-2-3-4) Spar 2-5
ject to modification, of course, if spanwise
stresses are anticipated which will buckle the
3@48"· -48"--
skin.
(b) for streamwise ribs an effective width
==
1= 2000
"II
2700
-- !
3400
I
2200
of O.362L, where L is the rib length, may be
taken as acting with each rib cap (Fig. A23.3a).
(e) for the leading edge spar an effective Rib 5-6-7 8par 3-6-8
width of skin of .181 of the span between span-
wise spars is taken as acting with the cap of
each such spar segment (Fig. A23.3b) - 48" - - 48"- _48"_48"_
, '

1 r ! !
I = 9.24 20.79 3.59 47.23 40.82 12.33

Rib 8-9 Spar 4-7-9-10

_ _·_3@48"
-
'V'"
I (b)
-
[
1= 6.23
48"-

!
2.87 5.09
[
4.78
£
3.82
1 !
0.28
Fig. A23.3 Effective Widths of Cover Sheet for
Rib and Oblique Spar Caps (After
Levy, Ref. 2)
Leading Edge Spar (1-5-8-10)
The properties of the structure (Fig.
-----3@87.8~'l_-----
A23.1) after idealization, are summarized as
fol1 ....S: f ,
I '" 16.02 12.29 8.30 O. 26
Net Point Beam Depth
(see Fig. A23.2) (inches)
1 - - - - - 5.56 SIGN CONVENTION Al'Jll NGrATION:
2 - 9.12
The Sign convention and notation adopted
3 7.80 in conjunction with the grid numbering scheme
4 - - - - 3.16 of Fig. A23.2 is as follows:
5 4.88
Forces:
6 7.26
Pm - transverse torce at jOint m, ~ UP
7 - 3.06
8 - 4.02
Mm - moment at joint ~ acting about a
pitching axis, + NOSE UP
2.74
9 -
Nm - moment at joint m acting about a
10 - 0.72 rolling axis, + ~IGHT ~ING UP
All cover skin t = .051" Qm - moment at joint m acting in the plane
of an obl1que beam member (a :nember
neither parallel nor normal to the
st'reamwtse direction), + CLOCKWISE
',/HEN THE MEMBER IS VIE'';ED ',-lITH ITS
• A rational, systematic means of treating cover sheet JOINT NUMBERS INCREASINJ LEFl' TO
panels is given in Reference (4). RIGHT •
 A23.4 STRUCTURAL
Displacements:
6m - transverse displace~ent
+UP
~ - rotation of JOi~t m about a pitching
axis, + NOSE UP
~ - rotation of joint m about a rolling
axiS, + RIGHT ~ING UP
Tm - rotation of JOint m in the ~lane of an
oblique member, + IN DIRECTION OF +
Qm (see page .'123.31
For a special purpose it will be convenient
later to introduce another set of displacements
to be called "relative displacements." ~nere
any confusion can exist, the displacements de-
fined above, which may be visualized as being
measured with respect to a set or reference axes
fixed in space, will be referred to as "absolute
displacements. "
~iththe above sign convention, any beam
element (spar. rib or oblique member) which is
viewed so that its joint numbers increase from
left to right, will have positive forces and
displacements taken in the same sense as every
other member. This paint is illustrated in
Fig • .'123.4.
(a)
1 Rib 5-6-7
.p.
, Spar 4-7-9-10
ip..,
,
" Q.
'Po
Fig. A23. 4 illustrating the Homogeniety
of the Sign Convention
CHOICE OF PERTINENT FORCES AND DISPLACEMENTS:
Before beginning the analysis proper. the
analyst must decide upon which torces and dis-
placements are to be considered pertinent. One
will, at necessity, have to consider trar~verse
forces (and translatorJ displacements) at all
net points (ten paints in the example). In
addition, rolling moments, N, (and rotations ~)
must b~ considered along the airplane centerline
on all spars carrying across the centerline.
T~erever three or more beam-like members inter-
sect (e.g. paints 5 and 8 of Fig. A23.2), their
bending stiffnesses will react against each
other and hence couples in two planes (M, N)
must be considered at these points.·
ANALYSIS OF A DELTA WING
~herev~r addi:l)~al knowl~d~e about the
effects of various loadin;s are required. C8~es­
of jOint n, po~dlng forces or ccup~es should be ~dded. T~us,
the influence of additlop~l forces at paints
lnte~ediate to the grid intersections, may be
accounted fer by the adjitlon of approprla:e
extra forces along beam spans. Couples may be
applied at any pOints where deflection Slopes
are required, e.g., the streamise slope at the
trailing edge ~lght be needed in an aero-elas~ic
analysis, in which case c oupLee :-': .., ?1.,., M II and
I"l:::., 0 (c r • Fig. ;'23.2) wouj,c oe emp.l.oyed , The
effect of couples from a~x~llary aeradyPZOic
surfaces and/or actuators ~y be desired, in
which case appropriate additions may be made.
In the present exa~~le, ~ecause of space
limitations, only the ~inimum num~er of torces
and corresponding displacs~ents are conslderec,
viz:
P~ through Pl 0 - forces on net paints
N~ through N~ - rolling couples on through-
spars at the alr~l~a center-
line
M:l.' N 8 , Me' N e, M!I - pf t.chtng and rolling
couples at points at Intar-
sect~on of three be~~ elements··
A23.4 Calculation of Element Sillfness Matrices
The task of computing the stiffness matrix
for any one element at a configuration is a
relatively straight-forward structural problem.
(b)
This problem may be either a statically deter-
minate or a redundant one depending upon the
geometry at the element anq upon the nunber of
pertinent forces and displacements associated
with it. If i t is a redundant problem. it 1sa
small one by comparison with the overall struc-
ture of which it is an element. One might say
that a feature 0:' the Displacement Method of
AnalYSis 1s that it "~kes (many) little ones
out of bIg ones"!
For instance, looking at Spar 2-5 of the
present problem (Fig. A23.5a), we see t~~t
there are four forces (Teactions) aC~lng on it
• The implication here is that beam torsional stiffnesses are
not considered in such interactions. This assumption is
probably quite satisfactory in general, bending stiffnesses
being much greater than torsional stiffnesses 'for most
beam elements. On occasion, a beam will have consider-
able torsional stiffness and it will then be necessary to
account for it. The leading edge spar might be such a
beam if it replaces the "D" nose section of the Wing in
the tdealization.
.. Couples Q:a" Q 0 and Q!I' in the plane of the oblique Lead-
ing Edge spar, are included hereby, since any such
couples may be resolved into M:a,. N::.,; Mo. N e and Me. N!I
components. See "Transformation of Matrices for
Oblique Beams". below.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A23.5

element stiffness matrices involves first f1nd-

la) (b)
Fig. A23.5
to be related to the tour pertinent displace-
ments at its ends. That is, we seek the ralatl0
between the loads (reactions) Nand P and the
displacements 1
and~. The first column of the
above matrix relates the four loads to the
rotation iaat the left end with ~l! = s , = I l l !
= O. The situation is shown in Fig. A23.5b
where it 1s seen to be a fa~11iar single re-
dundant problem. ThUS, to obtain the complete
stiffness ~trlx above requires the solution ot
four Simple redundant problems, one for each
column at the matrix.
ir~ their flexibility influence coefficients and
t~en invArting these ~atrice5. By way of illus-
tration, the calculation of the influence co-
efficients is carried out in detail below tor
spar 4-7-9-10. (wnile the calculation is
ef:ected here by the Least Work matrix method,
any of numerous other me thods (Moment Area,
Elastic Weights, etc.) may be employed at this
stage, Should the analyst see fit.)
Fig. A23.4b shows the forces applied to
spar 4-7-9-10, corresponding to those considered
pertinent for the wing (see above).
For the umuediate calculation of influence
coefficients, forces p. and p~o will be con-
sidered as fixed end reactions: deflections
thus computed will be measured relative to the
beam element ends. To distinguish the deflec-
tions thus obtained they will be referred to as
relative deflections and will be denoted by the
lower case Greek letters 6 and ~.
Fig. A23.7 shows the internal generalized
forces for this spar.

For Rib 8-9, on the other hand, we seek the

(.'9-",
q,
==~" s ,
10
relation
Fig. A23.7 Internal Generalized Forces Used
in Influence CoeUicient Calculation
for Spar 4-7-9-10

Member fleXibility coeff1cients for the

as may be seen from Fig. A23.5, which depicts tapered beam segnents are computed next using
the calculation of the first column of this the formulas of Art. A7.l0 of C~~pter A7.
matrix, th9se are statically determinate calcu- Collected in matriX form these became (E = 1)
lations.
3.213 1.623 0 0
1.623 6.841 1.864 0
[ail] = 0 1.864 8.55 4.923
0 0 4.923 21.54
Fig. A23.6
Next, unit values of the loads N., p~ and
Pa are applied succeSSively to obtain the unit
stress distribut1on:
Obviously, one can proceed in the above
fashion to work through all the elements of the
s~ructure solVing for the stiffness matrix of
each by application of these Simple arguments. I~ N, P, p.
3ecause a large number of beam stiffness cal- 1 1 0 0

~imJ =
culations of the above type appear in this 2 .567 -16 -8
prOblem, an alternate (but equivalent) pro-
cedure was developed for systematically treat- 3 .333 -8 -16
ing these members. This procedure 1s detailed
4 0 0 0
in the follOWing section.

Beam Element Influence 80efficients Finally, the following matriX triple pro-
duct is formed to give the i~:luence coefficients
The ~ethod em~layed to ~btain the beam
(ct. Eq. 24, Chapter A7).
A23.6 STRUCTURAL ANALYSIS OF A DELTA WING

Co~plete 3ea~ Eleffient Stlf~ness Matrices

The stiffnsss matr~x given above r91~tes

= [1~~::0
-119.9
-141.6
27715
2566
- 1 .J.O:I.::'
0 __
2 ". . ;::.;::.
3104
'J the 1eflections relati,e to the beam element
ends to the loads a?plied to the bean element,
end reactions exclUded. It 1s desired next to
obtain the "compjete" beam element stiffness
matrix in which ~bsol~te beam deflect~cns are
related to all the loads applied to the bean
o~J written out as an equation; element incl~ding the end reactions.

For Spar 4-7-9-10 Consider deflections ~irst. It is desired

to transform from the relative deflections of

_1l9'J~{N.}
{
~ .} llo.20 -141.6 Fig. A23.3a to the absol~te deflecticns of ?ig.
A23.8b.
5~ = -141.6 2776 2566 ?~ -(3)

6. -119.9 2566 3104 P; 6 m+ r

-==-
Note again that for this bea~ element the
flexibility influence coefficients just com-
---Lm-r

(al
L
f

puted are for relative deflections (note

symbols used). The end transverse deflections
(A~, A~o) and the end transverse reactions
(p. J P~o) do not appear i~ these results.
Beam Element Stiffness Matrices (Relative (bJ DATUM
Deflections) Fig. A23. S Geometry of la) Relative and (b) Absolute
Displacements
The next step involves finding the beam

cient matrix.
be of the form
.
element stiffness matrices tor relative deflec-
tions by invert~ng the above influence coeffi-
-- The results after inverting will
From the geometry

+
'c'!ll+s - ~
L
c~ the figure, one r2S

6m+r = -
L - kn+r
L

_ iL + iI,n+. - iI,n
<Im+. - p+s L
where ~kJ may be called the Wrelative stiffness
matrix. Specifically, for Spar 4-7-9-10
For comparison with the matrix form to follow,
N .) [,3554
~:~~~:~~{::)
.02303 these equations are rewritten as
r , = .02303 .003013
{P" - .005306 -.001600 00143~
• 6. -T
1 +
T
1
~+s

I...m+r - L I...m+r
• We note that the inlluence coeUicient matrix to be inverted o",+r = + D.:n+r ~+s
does not appear to be "well conditioned" in the sense of Art. L L
AS. 12 (see p. AB.28). The situation is more apparent than
real, however, and arises because of a difference in units 1 1
among the coefficients of Eq. (3). Thus, the appearance is <Im+. = - L +
L ",,+.
readily altered by "scaling" (factoring out appropriate con-
stants). For instance, we may write
+ ~+.

( ~:::: -::::: -::::~

::IuJ. (~o'P_} In matrix form these equations are written

31.0~
so as to prOVide a transformation fram absolute
6./,0 l1.96 25.66 lOP. to relative displacements. In matrix f0~ the
index (subscript) is understood to inc~ease
Scaling in this fashion does not actually enhance the con- monotonically down the column, and inasmuch as
dition of the matrix (which is basically related to the size
of the determinant of the matrix) but it does permit one to
the grid numbering scheme gives this same
assay the problem better. property to the joint numbers on an element one
has:
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A23.7

where
- - - - - - - - - - (5)

~ Nm~
Pi:!

!'1m WITH END

REACTIONS
= [s]
~
-p~ !Nm'Im1 --
TA1THOUT END
REACTIONS
(6)

r J where
1
1 -L"
J
I l2:----~ : T1 0
------1-----7------
r I
1 ! ~------:£J 0
0
r
r
r !.m-L
r
I
r
I
r
I
I
Q
r
I
t: ~lm - L)/.!J
-----------,--
I J
t
[TJ = I
r
I
I
-L- I
I
I
I
[I] I
I
I
-!.m
L
I
r
I
,
I
o
J
I
J
I
I
J
I
I
0
I
I
0
r r
I
I
I
J
r
0 [sJ = I
I
I
J
[I] I
I
I
I
---- -- ----1-- ----- J r
I
I
I
I
I
I
I
1 1
0 -T ~---~ I
I T 1 a
I
r t
I
0
I
__ ~ 1. _

Here [1J denotes a unit matrix.- The matrices T1 I

1 L lm/L J I
I 1
T
1 I
are shown ~partitioned~, those portions which o :[2;-~: 1
transform the end rotations being shown separ-
ated frcm those which transform the transverse
deflections. The above transformation matrix 1s [sJ
not square, having two more rows than columns,
The transfor.nation matrix C:TJ in Eq. (5) is these extra rows yielding the end reactions.
not square, haVing two more columns than rows to
allow for the added two deflections, one at each PrOVided the Sign convention is observed
beam end. as originally adopted, and provided the grid
numbering scheme 15 such that joint numbers
Consider next the loads. The applied loads increase to the rear and outboard, then the
may be transformed, using the equations of above transformation matrices will apply equally
statics to yield the end reactions. well to ribs, spars and oblique beams:

~m+r replace N by M or Q, ~ by Q or ~··and ~ by 9 or T.

M --Lm+r-
IIJ- M m+ s
For the case of beam elements haVing couples
I .;;==~~~~~~d
, , applied along the beam (e.g. the L.E. spar,
- L ----- Fig. A23~ simple modit1cations of these
P m+ s
transformation matrices are necessary.
Fig. A23.9
The relative stiffness equation, Eq. (4),
From Fig. A23.9 one has
}'-. 1 'L
= -T-T m~r Pmer- (L - Lm.+r) - "L+ s

Pm+ s = -?-- t Z
a-r P!l1+r • Lm+r + N."L·+ s
is the relation be~Neen loads and relative
deflections at a beam element. Substituting
from eqs , (5) and (6), into (4), one' obtains
In matrix form, the general expression
which introduces the end reactions in te~s ot
the other applied loads is
~i = [sJ [kJ[TJ
WITH END
tll ----- ABSOLUTl:
(7)

REAC~IONS DEFLECT IONS

• We have used L .J and { } to denote row and column
matrices, respectively (see Appendix A). ThUS, on 't is introduced here as a symbol for the relative rotation
l..P - - - - - 2.J
is a row of zeroes. of a joint in the plane of an oblique mem~
A23 8 STRUCTURAL ANALYSIS OF A DELTA WING

Eq, (7) is the relation between absolute de- For Spar 2-5
flections and the co~plete loading on the ele-

n[= L~'~n
ment. ve let -.1218 ,1218
Pa -.1218 .004208 -.004208 -.08054 2:.",
PI - .1218 -.004208 .004208 .02054 ~~
be the "complete" ele~ent stiffness ~trix.
.K" 1.555 -.08054 .0605? 2.317 ~"
Note that [SJ is the transpose of G'J, as
For Spar 3-6-8
it must be if [KJ Is to be s~~etric.
To continue the illustration, the and
[sJ matrices for Spar 4-7-9-10 are now written.
For this member the rotation at station 10 Is
not considered and the couple at that end is
zero, hence the last row and column of [TJ and
[TJ .'

p.
p.
P,
:;
.
.I " ~~
-.03423

-.2534
.1055
_.02::..29
-.02423
.00275';'
_.::;0-'1.1:2
.:J01:X:S
.01725
.:265
_.00.:;.1:2

. '''''',,-<>-
J,.....

_. ::;03066
_.05.03...
_.02228
.J013~3

_ ,'O:J3CC'6
.OOl~C8

. :)3908
OJ '~1"''''''

':;':;: t::
.,,3,,0~

1.33
I·

~.

1.

[sJ are omitted in writing them from the general

For 1/2 Fuselage Beam
expressions.
M,
c-
158.5 -4.312 5.580 -1.533 .2652
- 8.

n
Equation (5) written for Spar 4-7-9-10:
P, -4.312 .1557 -.2385 .09993 -.017l8 6 ,
~, P, = 5.560 -.2385 .4343 -.2701 .07429 -,
-,

0,. -
[_oo,~
0 _.667
a
1.0
a
a -.333
OOOO~ il,
il,
.,
D -1.533 .09993 -.2701 .0:::;............... -.1027
" ........ .04560
I c.~

,,,
t

P, .2652 -.01713 . 07429 -.J:v':::'! ~

6. 0 -.333 a 1.0 -.667 /, , ~

-
/'" For Rib 5-6-7

rJ 0
Equation (6) , written for Spar 4-7-9-10:
'~::~rl
3 -.G£253 .0275
r- a- .0,536:0 _.00:2:>. .v,"'" v _. I
M. La a '. '" -.02253

·:o:;;~ :: J
? .0276
:'
_ .OC12C4 .CClS3Z o ·c·' I '

P, -.006945 - .667 - .333 ? -.005066" .::J,3570 _.c,c:s

P, = a
P, a
1.0
a
a
1.0
{::) For Rib 8-9
P" .006945 - .333 - .667
- - -.006910
The matrix [KJ is now formed per eq . (a) • M') = 63322
P!l -.006910
For Spar 4-7-9-10: • (?g .006910

-.01605 ,0230'3 _.C:::::306

"~l~ i. 1 For Leading Sdge Spar ......

rJ IT'
p. -.01606 .0~,;g52 -.001637 .0,6252 .0.:65
?
p.
= .02303
-.006306
-.001637
.0,:5252
.ooaoia
-.00160::J
_.001600
.OOH38
.C~2238

-.0~46~
:: j I
.3>178 -.01846 I .C1:;'~ J I .-.l~O I'~ c I
p,. -.001622 .0.:65 0,.:22:3:0 _ .::J,'i-B3Z .::J,2229 :: 0
P, -.01946 .C,~';'90 -.O,d901 ::J -.':;1756) c I
p. ,01".;..s -.0,.54,,,0 .c.saes -.0,394:1 .0030<34 -.01.0.5" o!
In the manner of the above illustrative P.
case one goes through the calculations for the .';'1&0 -.01766 .OC3~24 .21...07 l.~';' I .3C16 ,0 i
remaining beam elements to obtain:
::J J - .01259 • JO~ 1001 •.3016 I --q, . cc:;:;sz I
o

• The notation . ON means there are N zeroes following the .. Because of its Size, this matrix relation (and some to
decimal point and preceeding the first significant figure, follow) is written in a condensed tabular form. The correct
e. g••• 0.. 276 '" . 0000276. interpretation should be obvious.
 M

ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A23.9

~ransfo~~tiJn of Matrices for Oblique 8ea~s: Next, consider resolution or the in-plane
T~e complete stiffness matrix of any beam
bea~ element cQuyles. Counles Q~, Q.8 and Qa
element not either ~arallel to or nor.nal to the have components in the pitching and rolling
strea~ise direction, such as the leading edge
directions. ~or exa~ple, at joint 5
spar in the present example, requires an addi-
tional modification to yield expressions relat-
I1 8 = sin J\.. Q.. .707 Q.. =
and
ing couples and rotations about pitching and
rolling axes, rather than in the plane of the Na = cos A Q,s = .707 Q,s
member. Such a modification is readily made.
Similar relations for the other couples along
In the present example, a pitching rotation this member lead to the matrix transformation
of joint 5 (9 s) results in a rotation of jOint 5 for this element's loads:
in the plane of the leading edge spar given by
~~ = sin A 9, • •707 9" 11, .707 0 0 0 0 0 0
where A- is the sweep angle of the spar. Like- N, .707 a 0 0 0 0 0
wise, a rolling rotation of j~::nt 5 (~,) has a Q,
rotatlor~l component in the plane of the leading
P, 0 1 0 0 0 0 0
edge spar given by P, 0 0 1 0 0 0 0 P,
T:
= cos .'- 9, = .707 ~, P, 0 0 0 1 0 0 0
P,
Tten, when joint 5 experiences both pitching and 11, = 0 0 0 0 .707 0 0 P,
rolling rotations, the total rotation in the Q,
plane of the leading edge spar is N, 0 0 0 0 .707 0 0
11, 0 0 0 0 0 .707 0 Q,
rn
J. II = .7079 11 + .707 ~,
N, 0 0 0 0 0 .707 0 T"
This last equation, and s~ilar ones for joints
1 and 8, when put in matrix fOIiJ, yield the P" 0 0 0 0 0 0 1
~atrix transformation for the displacements.

, The above two matrix transformations - one

.707 .707 0 C 0 0 0 0 i 9, for displacements and one for loads - (and note
r:: 1 o
0
c 0
c 0
0
0
0
0
o
0 Q
t
,
-,
e,
that they are the transpose of each other) are
now applied to the complete stiffness matrix

~l
0 1 0 C 0 C 0 equation for the leading edge spar as given in
,, 0 Q .7C17 .707 0 0 0 ", tabular form in the last sectIon. The opera-
". o
0 c 0 0 0 o
c
•707 .tsn 0
'. t10n involves premultiplying by one transforma-
, .i,o) l- 0 0 0 0 0 1 I. tion matrix and postmultiplying by the other.
The result is:
"
I-
'"::'0

.
Leading Edg~ . SDar cv
namDlete _.,
Stl~Pness ~atr1x . . ,
(~rans~ormed)

9, ~, ", Il, Il, 9, ~, 9, ~, Il"

M~ .4487 .4487 -.01376 .01376 0 .2094 .2094- 0 0 0
N, .4487 .4487 -.01376 .01376 J .2094 .2094 0 0 0
P, -.01376 -.01376 .03;:)4-90 - . ° 3 5490 0 -.01248 -.01248 0 0 0
p. .01376 .01376 -.0 30490 .0 39428 -.0 33941 .002534 .002534 -.00890 -.00890 0
P, 0 0 0 -.0 33941 .0~4422 .009947 .009947 .006469 .006469 -.04,4998
M, .2094 .2094- -.01248 .002534- .009947 .7168 .7168 .1507 .1507 0
N, .2094 .2094 -.01248 .002534- .009947 .7168 .7168 .1507 .1507 0
11, 0 0 0 -.00890 .006469 .1507 .1507 .3889 .3889 .002391
N, 0 0 0 -.00890 .006469 .1507 .1507 .3889 .3889 .002391
P" 0 0 0 0 - .0, 4998 0 0 .002391 .002391 .0,4998
A23.10 STRUCTURAL ANALYSIS OF A DELTA WING

Torque Box Stiffness Matrices ~ = P~~brn

u,;
Lnbm +
Ln
Lmbn
In the case of the torque boxes, the stiff-
ness matrices may be computed directly (they Since it is only the deflection of one
are dete~inate). The following approximate corner (say, point m) which 1s to be related to
analysIS is suggested as being satisfactory for the corner reactions, the box is now rotated
most torque boxes. A ~ore detailed analysis about the effective root axis to reduce on to
may be In order for boxes with extr~me geome- zero. The resulting total detlection of point
tries. An alternate method is presented in Jl is
Reference (2) and an appropriate discussion may
be found in Reference (6). A completely differ- ., _,,-+Lm 5
:.wTOTAL - 4Il 1:;- n
ent treatment at the cover skin is given in
Reference (4).
= P:nLm (l.nbm. + Lmb n) It + "-- I
Consider the quadrilateral box of FIg. GJ'""" \" Ln n '-'TIl

A23.10
Re-solvlng this, we write
(r~ Ln - - - : GJ Lo
T ' '<n) Pm = Lm(bn + bml (t-nom + Lmbnl Cn>rCTAL
"'T L / 2 - - L / 2 ' --,-
_ _ _ .. _ _ .Iil - +bn
This last expression relates the load at
I
~s,
t~Axia
:
\-----.lbm
(m)
point m to the detlection at that point, with
the ather three corners undetlected. 7he re-
L...... \ ....... I
actions developed at these Jther corners are
L
,........, w" '
now found in terns of Pm USing the equations of
Equivalent Representative statics (s~ation ot forces and summation ot
, Root Section moments- about two axes). The r-esuj t , expressed
in matriX fo~ is
Fig. A.23. 10

For purposes or this immediate analYSiS, the

box 15 considered cantilevered from one end,
such as r-s. An elastic axis (e.a.) is
assumed to exist midway between the long sides
and an "etfective root" 15 employed. The torque
on the box is
where

A, = Lnbm + ~bn +
Ln(b r +
bs (Lm - Lnl
os)
An (assumed) average rate of twist, Q, Is com-
puted approximately by using the GJ at a repre- -(Lnbm + Lmbnl + br (lm - Lnl
sentative section halt way along the box.* A.
" Ln (or + b s )
Q = O~ = Pmbm O~ Pnbn It now the corner pOints have absolute
displacements Om, dn, dr and tis, one can find
Then the deflection On and ~ are given by from the geometry of the unit that the total
relative displacement of point m may be writ~en
= QLnbn = l'mbm OJ- Pnbn -nbn
On I.
Cn>rarAL = L 1 A,

Om =QLmbm = Pmbm dJ- Pnbn 1

~"'1Il
h ..

By ~ing moments about the effective root, Pn

15 found in terms at Pm. and then eliminated from
the above equations to give When these last three equations are com-
Lnbm Lmbn bined to relate the four corner reactions to
On = PmLnbn +
the absolute corner displacements, the torque
OJ Ln
box stiffness ~trix 1s seen to be t~e square,
symmetric matrix,
* Calculation of. thisGJ is discussed later. The symbol "s''
used. in these few equationa bears no relation to that appear-
ing in the main body of the method.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A23.11

For box 6-7-8-;)

= (Lnbm+l<nbn)
GJ Ln/Lm 1 ~ ~ A, ~ GJE=l = 34.3
(bn+bm) -L,n/Ln
Finally, the stiffness matrices for the
A, ----(9) three boxes at the example become: (c r .eq, 9)
A,
For box 2-3-5-6
In the above derivation the torsional
stiffness OJ at a representative section is 1 -1 -1 1
used. The stiffness is obtained fram Bredt1s
equation tor the twist of a single cell thin- -1 1 1 -1
[KJ = 0.000898
walled tube (Eq. 18, Chapter A6) -1 1 1 -1
1 -1 -1 1
g _
-~
T
Ie.: For box 3-4-6-7
fram which, by definition
1 -1 -1 1

GJ = [KJ =0.000482
-1
_1
1 1 -1
1 1 -1
1 -1 -1 1
Here A is the area enclosed within the tube
cross section by the median line of the tube For box 6-7-8-9
wall and the integration is carried out around
the tube perimeter (index s gives distance along 1 -1 -1. 1
the perimeter). For the torque boxes en-
countered in delta wings it is probably satis- -1 1 1 -1
factory to neglect the ds/t contribution from [KJ = 0.000310 1 1 -1
-1
the vertical webs, it being small compared with
the corresponding contribution from the cover 1 -1 -1 1
sheets.
A23.5 Complete Wing Stiffness Matrix
In the example wing three boxes (2-3-5-6,
3-4-6-7 and 6-7-8-9) are to be used. These The stiffness matrix tor the composite
boxes are each 48 inches square in plan and wing now may be obtained by forming the sum or
have average depthS (assumed here to oe the the complete stiffness matrices tor the beam
unlfor.n depthS ot the representative sections) elements and the torque boxes. For this pur-
at 7.26~, 5.32" and 4.27~, respectively. Fig. pose a large matrix table Is laid out and en-
A23.ll shows the assumed representative cross tries fram the individual stiffness matrices
section of box 2-3-5-6 and its OJ calculation. are transferred into the appropriate locations.
Wherever multiple entries occur in a box these
are summed.
ie
,
l_ t

48"
= . 051 ·,
_ Betore the large wing mata-tx is laid out
it is necessary to observe first that the
A: 7.26 x 48 =348 in .. matriX which would be obtained as just indicated
Note: would be singular, i.e., its determinant would
f t:= 2 x. 051 " 1. 88 x 10
~ ~ ~ E be zero, and hence it could not be inverted to
G := 2:6
yield the flexibility influence coefficients
4(348)8. 1 . .. (see Appendix A). This condition arises due to
GJE=I" 1. 88 x 10~x 2.6" 99.3 lb.fn. the tact that the equations represented (19 in
(ds/t contribution of vertical webs neglected) number in the example problem) are not inde-
pendent; three of these equations can be ob-
Fig. A23. 11 Calculation of GJ for the Representative tained as linear combinations ot the others.
Section of Box 2-3-5·8
That there are three such interdependencies
In similar fashion one finds: may be seen from the existence ot the three
equations of statics which may be applied to
For box 3-4-6-7 the Wing (summation at no~l forces and sum-
mation at moments in two vertical planes):
hence three ot the reactions expressed by the
structural equations in the matrix may be found
 A23.12 STRUCTURAL
tram the others by the equations of statics.
To remove the "Singularity" rr-cm the stiffness
matrix it is only necessary to droD out three
equations - achieved by removing :hree rows
and corresponding columns (so as to retain a
symmetric stiffness matrix).
The act ot removing the three equations
selected is also equivalent to assuming the
corresponding deflections to be zero. In this
way a reference base for the deflections 1s
also established. The choice of reference base
is somewhat arbitrary, but, following a sug-
gestion of Williams (7), a triangular base will
be employed as shown in Fig. A23.l2.
5' '5·
.~
Fig. A23.12 Deflection Reference
Base
Here the deflections at paints land 5 are zero,
fixing the reference triangle, since the point
corresponding to 5 in the other halt at the wing
(say, 5'), will have zero deflection also due to
symmetry. The third conditi In is applied to
point 2: pOint 2 will be assumed to have zero
rolling rotation (~~ = 0) for symmetric wing
loadings and to have zero transverse deflection
(6 8 = 0) for antisymmetric loadings.-
Hence the following equations (rows and
columns) are to be omitted trom the wing stIff-
ness matrix:
Pu
Po'
N.. ~,
",
", }
tor symmetric loadings
P"
P .,
P"
",)",",for antisymmetric loadings
The 16 x 16 wing st t rfness matrices thus
obtained can now be inverted as they are non-
- The antisymmetric loading pattern is one wherein the wing
is loaded equally, but in opposite sense, on corresponding
points o( the two wing halves. Any general loading may be
resolved into the sum or one symmetric and one anti.
symmetric loading.
.L ;~~,\"~<~;~;i'/::,;;9~~;~~; g·~r:~/~~:·;:;;t%;i?r:.~~L{~~~;~~i:::~~~;~~<::;~~~V~} /',l~?T:~JL~;~~~j~
ANALYSIS OF A DELTA WING
singular. However, if this ste) ~s carr~ed au:
one fines that :he resultar.t inf:~e~ce co-
efficients are those for- onjv t.ne "':e.l.::'-wl:-,:
acting alone and supported by ccr.s:r~in:s as
ass~ed above. To account for the ~resence o~
the other ::al! of the vnng , i t is necessar-y to
speCify additional geo~etr~c con1:tlons along
the airplane centerline. This ste? 15 accon-
plished by assumir~ the following deflect lens
zero (ellmir~ting their corres~ondlng rows and
columns from the matrix):
for symmetric loadings (zero
lateral slope or rolling
rotation along the alr,lane
centerline)
for antisymmetric loadings (zero
transverse deflections and pitch-
ing rotations along the airplane
centerline)
It will be seen that in both cases an
addItional 3 equations are eliminated from the
16 x l6 matrIces reducing theo to 13 x 13'5.
(In.general, there is no reason why the
matrIces :or the s~etric and antis~etric
cases have to be the sarne size. A rotation
i~, for instance, would be zero in one case
only. )
~rltten below 1s the 13 x 13 wlr~ stiff-
ness matrix for ~he ant1syremetric case
(~~ = ~~ = ~~ = ~~ = ~& = 9~ = 0). As explained
earlier, each entry therein is the sum o~ all
corresponding stlffnesses for all elements
~eeting at the point. A typical multiple entry
occurs for joint 0 - (row P /I' column 6/1)' These
comprise:
0.007178
0.001832
from spar 3-6-8
from rib 5-6-7
0.000310 from box 6-7-8-9
0.000898
0.000482
0.01070
from
from
':'OTAL
box 2-3-5-6
box 3-4-6-7
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A23.l3

[:.::nnJ ' . . ing Stiffness Matrix, Ar,tisymmetrlc Case (Note: Voids denote zeroes)

, I,
", '. ",
'. '" " " I I, j, I. I. I.
P, .01070 -.GOl<;ZO -.003376 .O~310 .02760 ! .1065 -.05634-
p. .001420 .004066 .0~310 _.00191 .0~2238 -.0~066i .02303
p. .003376 .G~310 .002604 -.0~4537 -.0~4998 .009947!-.o,441 -.02228 .009947 .045.:::>5
p. .0,310 -.00191 -.0,4537 .001892 -.0~4632 .006910 -.005306
,

."
0
.0,2238 -.0~~998 -.0~4632 -.0~2729 .002391 -.001662
,
.002391
M, .02760 - .OOS06E .009947 1.555 .1507 .2094- .7168 .1507
M. -.0~441 .006910 .002391 .1507 .7211 .1507 .3889
N. .2094- .4487 .2094
N, 4.302 1.5s.:;
N, .1065 -.02223 3.256 -.2334
N. .02303 _.005306 -.001662 .3554
N, .009947 .7168 .1507 .2094 1.555 3.0338 .1507
N, .05634 .04555 .002391 .1507 .3889 -.2834 .1507 1.719

m::: '.WG FLEXIBILITY 11ATRIX:

The Wing flexibility influence coefficients are now found by forming the inverse of the above
s~ifrness matrix (see Sq. 2). An automatic digital computing machine is the essential tool for
this step. The details of the procedure and techniques employed in forming such inverses are not
of concern here and only the result 1s presented. It is assumed that this phase at the work can
be handled with dispatch by experienced computer personnel.·

~AmnJ ' Wing In!luence Coefficients, Antlsymmetrlc Case

..
0 P, p. P, P ~o n. M, N, N, N, N. N. N.
73';'.3 se8.7 1386 :,';64- 2"'19 -17.14 -14.S9 7.750 _.113...0 -15.77 -2.3.52 .5092 -14.22
"
.>, 988.7 2'~03 2118 4030 537<5 -16.32 -38.23 a.610 .04SS1 -19.96 _70.~ -.1342 -24.2.2

'. 1386 2118 4001 4751 7692 -36.09 -25.10 17.57 •.5620 -23.72 -30.20 -1.555 -66.22

'.,
-~Q
15G4
Z"'lS
~30

53,6
4"'61
7692
5-558 13,610
13,510 2.5,670 -;:>a.69
-~a.09 -32.32
-131.0
1 18 • 01

27.75
.1::.26
.2836
-34.28
-~7 .C4
-6:5.70
-26.56
-.5052
-.7S<06
-67.49
-123.3
-17.14 -18.32 -36.09 -33.09 -58.69 1.<25 .09551 _.~77g .07'273 .3438 -.201.2 .4227
"•• -IL5!? -33.231-25.10 -32.32 -1.'31.0 .09.551 .2701 -.00806':;' .02523
.0.'303
.2976 .6358 _.07823 -.2044
t 7.'7"60 8.510 17 .57 13.01 2"'.75 -.47-;;) -.00a06... 2.,513 .~73a -.1527 -.1592 -.131: _.2198
I, -. ::'~40 .0';"':':>1 .,')620
.1825 .2836 _.0-2'73 .02~25 .04-38 .2S54 .80834-7 .0)9086 -.17.;.;3 -.01744

t _1."3.7"7 -19.96 -2.3.72 -34.28 4 7.04 .;503 .2[;76 -.1.':27 .00834'::' .;:---'"
.~, .'::614- .02309 I .1::;28
I ~. -2.3.;:;2 _70.:'0 -30.':00 _65.70 -25.56 .-34:38 .6338 -.1592 'OJ sees .-~61'" '3.271 1-. ':]0251-+ .)1...20
I, .--~O;2 -.1342 -l.~55 - .50.'52 -.-846 -.2012 -.072.23 - .1311 -.17.;.3 - .-::'<:309 -.0025H . ...a21
". ~~ ,
.V"...:>.::....

I, -H-.22 -24.22 _'36.22 -'57.49 -123.3 ..;.2"- -.2Q44 -.2198 -.017;;.4 . 1~2S -.014,,0 (.04,82... 2.079

" In spite of the exercise of great care and much ingenuity, problems of such great size and such a nature will arise occasion-
ally as to defy satisfactory inversion. For these problems such clever physical concepts as "block" solutions of portions of
the structure are available. The interested reader is referred to a discussion of Ref. 4 and to techniques of block and group
relaxations in the literature on Relaxation Methods (e. g. Ref. 8).
A23.14 STRUCTURAL ANALYSIS OF A DELTA WING

A23. 6 Wing Internal Stresses
following t~e calcul~tion of the Ning
influence coefficients, a relationShip is avail-
able between the a~plied wing loads (assumed
given) a~d the wing qefo~tlJns. In a sJ1nbolic
fashion,
Equation (10) provides t~e cesired relation
between ~xternal loads applied tc the wing and
t~e forces acting on a speCi~lc eleme~t. These
~crces on the element might be locked upon as
those necessary to hold the element in the de-
flected srape conforming to that of the wing.
Of course, with these element forces known the
finer details of the stress distribution within
the element are readily found by standard
techniques.

Note that the ~atrix [KJEL~\~ in Sq. (10)

will have to be "blown up" to an accr-opr-tata
Earlier in the analysts, defor.nations of a size be Io re it may be pre:nultiplied:mto [;I,.lln:G'
structural el3ment (beam or box) were related to
the forces acting on that element by the This enlargement is accomp11shed by the in-
(complete) element s~lffness ~trix: serticn of columns of zeroes at each column
location corresponding to a wing defo~atlon
which does not affect the speci~lc element under

~~}
consideration.

For illustratlcu, the indicated steps are

EL=rr now carried through for spar 4-7-d-lO. Refer-
ring to previous work we find for tt.e element
Now since the deformations of an element matriX:
must conform to those of the wing,

.8mlm }
--- ---
={.1m}
8m
~. .::., o.
". ,
, "

b~ EL=rr ~ W!m
II,

p.
.3554

-.0160(·
-.016013

.O~~952
.0230:3 -.005306 -. ;:·01562

-.001537 .0~52.52 .0,155

Hence, the previous equations may be combined [KJ: , p. .023C .3 -.001637 .003013 -.J01600 .0.2238
immediately to y1eld
P, - .oor .306 .0.5252 -.C016CO .001438 -.:J.4632

P., -.u....J,662 a . --
• • ':""Q I •0• 2-->8
;,;:... -.0.';'632 • J~2229
.
.

Expanding (and rearranging to agree in numerical order with the wing matrix),

A, AT A, A. A" e, 9, ~. ~. ~, ~. ~, ~.
N. 0 .0230:3 0 -.005306 -.001662 0 C 0 0 0 .3554 0 0
p. 0 -.001637 0 .°36252 .0,,165 a 0 0 0 0 -.01606 a I 0
rKJ
J - E p. 0 .003013 0 -.001600 .°3223 8 0 0 0 a 0 .0230:3 0 0
L £LL'"1'!ENT
p. 0 -.001600 0 .001438
-'°3 4632 0 0 0 0 0 - .005306 0 a
P" a '°3 2238 a -'°3 4632 '°3 2229 0 0 0 0 0 -.001662 0 0
 L

ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES A23.1!i

Then multiplying out with [~~WING per Eq. (10l, we get the five forces en spar 4-7-9-10 (see

Fig. A23.4-b) in ter,ns of the 13 wing applied loads as:

p.
P,
P,
p.
-.C~367 -.0017'7 .o, 794 .0,304 .002694 .0.1736 .0.3834 .0.2706 0 -.0.8n 1.00 0 -.0.555 p ..
N.}
p.
p.
-.0304
.06294
-.1945
.3735
•1212 .2190
,2098 -.'0191
.4985
-.2915
-.0,320 -.001257 .0,180
.0,5408 .001850
.0.24-9
-.0,330 -.0.625
.00145 -.02696 -.0.687 -.004:34
M.
-.00299 .03148 M,
.0,1705 .007071
p. -.00375 -.1650 .05466 .3787 _.9174- -.0,107 .0.975 .0,1133 .0.486 .001451 _.00281 -.0,135 -.001107 N,
p. .002905 -.01364 .03347 -.07,,65 .7092 -.0,111 -.0,682 .0.348 -.0.1202 -.0.5815 -.001665 .0.3325 -.001618 ,N,
:::u:s:1. N.
N.
N,
N,
WHIG

Each column ot torces in the above matrix
must satisty the equations of statics on the
element. A variety of checks on the accuracy
of such a result are thus available.
The student will find it instructive to
study carefully this last result to observe in
what manner a load applied to the wing appears
on this spar element. For instance, one sees
that of the load p~ ~ 1 applied to the wing,
.3735 goes onto spar 4-7-9-10. Examination ot
Fig. A23.l reveals that the remainder, 1.00 -
.3735 = .6265 must be taken up by rib 5-6-7 and
the torque boxes 3-4-7-6 and 6-7-8-9.
REFERENC!l3
4) Turner, l1. J., Clough, R. 1,./., Martin, H. C.,
and Topp, L. J., Stiffness and Deflection
AnalysiS of Complex Structures, Journal or the
Aeronautical Structures, 23, September, 1956.
5) Kroll, W." Effect ot Rib Flexl'bili ty on the
Vibration Modes or a Delta Wing Aircraft, Insti-
tute or Aeronautical Sciences, ?reprint No. 585,
1956,
6) Wooley, Ruth, Check at Method ror Computing
Influence Coefficients of Delta and Other Wings,
National Bureau of Standards Report 3655, 1954.
(Avallab1e as ASTIA No. AD40866).

1) Argyris, J. H. and Kelsey, S., Energy

Theorems and Struc~ural Analysis, Aircraft
Sr.gineering, Oct. 1954, et. seq.
2) Levy, S., Structural Analysis and Influence
Coef:icients for Delta ~ings, Journal ot the
Aeronautical SCiences, ~, 1953.
3) Schuerch, H., Delta Wing DeSign Analysis,
Society or Automotive Engineers National Aero-
nautic Meeting Preprint No. 141, September IG53.
7) Williams, D., Recent Develonments in the
structural Approach to Aeroelastlc Problems,
Journal or the Royal Aeronautical Society, 58,
1954 (see also, Aircraft Engineering, 25, 1954)
8) Southwell, R.V., "Relaxation MethodS in
Englneerlr.~ Science", OXford 1940.
A23.16 STRUCTURAL ANALYSIS OF A DELTA WING

AIRCRAFT WITH "DELTA" WING SHAPES

CONVAIR
SUPERSONIC F-I02 INTERCEPTOR

CONVAIR
SUPERSONIC B-58 BOMBER

",,-"

~;::~~~~~,]t4?~f~"'",.
~".~ ~~

COMMERCIAL TRANSPORT FOR THE 1970s? -- This is an artist's conception of a passenger transport of the future -- cruis-
ing at three to five times the speed of sound at 60,000 feet or higher. This is one of hundreds of configurations considered by
Convair Division of General Dynamics Corporation at San Diego, Calif., in its studies of supersonic airliners.

·.-~·c ..-_,. ,-.,'~' ~

 PART B
FLIGHT VEHICLE MATERIALS AND THEIR PROPERTIES
CHAPTER BI
BASIC PRINCIPLES AND DEFINITIONS
Bl.1 Introduction.
The flight vehicle structures engineer
faces a major design requirement ot a high
degree at structural integrity against tai lure ,
but with as light structural weight as possible.
Structural faIlure in flight vehicles can
otten prove serious relative to loss ot lIte
and the vehicle. However, experience has
shown that i t a flight vehicle, whether
military or commercial in type, Is to be
satisfactory tram a payload and performance
standpoint, major ettort must be made to save
structural weight, that Is, to el1minate all
structural weight not required to insure
against failure.
Since a flight vehicle Is sut~ected to
various types ot loading such as static,
dynamiC and repeated, which may act under a
wide range or temperature conditions, it is
necessary that the structures engineer have a
broad knOWledge of the behaVior of materials
under loading it sate and etticient structures
are to be obtained. This Part B provides
information for the structures design engineer
relative to the behavior at the most common
tlight vehicle materials under load, and the .
various other conditions encountered in tlight
environment, such as elevated temperatures.
Bl.2 FaUure of Structures.
A flight vehicle like any other machine,
is deSigned to do a certain job satiSfactorily.
It any structural unit of the vehicle sutters
effects which in turn effects in some manner
the satiSfactory performance of the vehicle,
the unit can be considered as haVing tailed.
Failure at a structural unit is theretore a
rather broad term. For example, tailure at a
structural unit may be due to too high a stress
or load causing a complete fracture at the unit
while the vehicle is in flight. It this unit
should happen to be one such as a wing beam or
a major wing fitting, the failure is a serious
one as it usually involves loss of the airplane
~nd loss of lIfe. Likewise the COllapse of a
trut in landing gear structure during a land-
1&' operation at 'the airplane can be a very
erious failure. Failure of a structural unit
:,ay be due to fatigue and since fatigue rat iure
,LS ot the fracture type Without warning indl-
,,' cations of impending failure, it can also prove
, to be a very serious type or failure.
BL 1
Failure in a ditferent manner can resUlt
from a structural unit being too rfext tn e , and
this fleXibility might influence aerodynamic
torces SUfticiently as to prOduce unsatisfac-
tory vehicle flying characteristics. In some
cases this fleXibility may not be serious
relative to the loss of the vehicle, but it is
still a degree at failure because changes must
be made in the structure to provide a satis-
factory operating vehicle. In some cases,
excessive distortion such as the torsional
tw1st of the wing can be very serious as this
excessive deflection can lead to a build-up
of aerodynamic-dynamic torces to cause flutter
or violent vibration which can cause failure
involVing the loss of the airplane.
To illustrate another desgree of failure
of a structural unit, consider a wing bUilt in
ruel tank. The stiftened sheet units which
make up the tank are also a part or the wing
structure. In general these sheet units are
des tgned not to wrinkle or buckle under air-
plane operating conditions in order to insure
against leakage at tuel around riveted con-
nections. Therefore if portions of the tank
walls do wrinkle in operation resulting in
tuel leakage, which in turn require repair or
modification at the structure, we can say
failure has occurred since the tank tailed to
do its Job satisfactory, and involves the item
ot extra expense to make satistactory. To
illustrate rurther, the flight vehicle is
equipped With many installations, such as the
control systems for the control surfaces, the
power plant control system etic , , which involves
many structural units. In many cases exces-
sive elastic or inelastic deformation at a
unit can cause unsatisfactory operation ot
the system, hence the unit can be considered
as having tailed although it may not be a
serious failure relatiVe to causing the loss
of the vehicle. Thus the aero-astro
structures engineer is concerned with and
responsible for preventing many degrees ot
failure at structural units which make up the
flight vehicle and its installations and
obviously the greater his knowledge regarding
the behavior of materials, the greater his
chances of avoiding troubles trom the many
degrees of structural failure.
Bl.3 General Types of Loading.
Failure of a structural member is in-
fluenced by the manner in which the load is
81.2 BEHAVIOR OF MATERIALS AND THEIR PROPERTIES
applied. Relative to the length or time in
applying the load to a member, two broad
classitications appear logical, namely,
(1) Static loading and (2) Dynamic loading.
For purposes or explanation and general
discussion, these two broad classitications
will be further broken down as tollows:-
Cont i nuous Loading.
Gradually or slowly applied
Static Loading loading.
Repeated gradually applied
{ example, it a welght ot magnitUde W is
loading.
nnpact or rapidly applied
Dynamic Loading loading.
{ Repeated impact loading.
Statio Continuous Loading. A continuous load
ts a load that reaatne on the member ror a long
period or time. The most common example is the
dead weight or the member or the structure 1t-
selt. When an airplane becomes airborne, the
weight ot the wing and its contents 1s a con-
tinuous load on the Wing. A tank subjected to
an internal pressure tor a considerable period
of time is a continuous load. Since a contin-
uous load is applied tor a long time, 1t 1s a
type ot loading that provides tavorable con-
ditions tor creep, a term to be explained later.
Far airplanes, continuous loadings are usually
associated With other loads acting simultan-
eously.
Static Gradually or Slowly Applied Loads. A
static gradually applied load Is one that
slowly buildS up or increases to its maxunum
value without causing appreciable shock or
vibration.. The time ot loading may be a
matter at aeccncs or even hours. The stresses
in the member increases as the load is in-
creased and remains constant when the load
becomes constant. As an example, an airplane
which 1s climbing with a pressurized tuselage,
the internal pressure loading on the fuselage
structure 1s gradually increasing as the
difference in air pressure between the inside
and outside ot the tuselage gradually increases
as the airplane climbs to higher altitudes.
Static Repeated Gradually Applied Loads. It a
gradually applied load Is applied a large num-
ber of times to a member It 15 referred to as a
repeated Load, The load may be ot such nature
as to repeat a cycle causing the stress in the
member to go to a maximum value and then back
to zero stress, or tram a maximum tenslle
stress to a max~ compreSSive stress, etc.
The situation envolving repeated loading Is
important because it can cause tallure under
a stress in a member which would be perfectly
sate, it the load was applied only once or a
small number or times. Repeated loads usually
cause tailure by fracturing without warning,
thus repeated loads are important In design of
structures.
Dynamic or Impact Loadlng. A dynamlc or impact
loading when applied to a member produces
appreclable shock or Vibration. To produce
such action, the load must be applied tar more
rapid than in a static loadlng. This rapid
application at the load causes the stresses
In the member to be momentarily greater than
it the same magnltude or load was applied
statically, that 1s slowly applied. For
gradually placed on the end of a cantilever
beam, the beam will bend and gradually reach
a maximum end deflection. However i: thls
same weight of magnitUde W 1s dropped on the
end at the beam from even such a small helght
as one toot, the maximum end detlection will
be several times that under the same statlc
load W. The beam will vibrate and finally
come to rest with the same end detlection as
under the static load W. In bringing the
dynamic load to rest, the beam must absorb
energy equal to the change in potentlal energy
or the falling load W, and thus dynamic loads
are otten reterred to as energy loads.
From the basic laws ot PhySlcs, torce
equalS mass times acceleration (F = Ma) and
acceleration equals time rate at change at
velocity. Thus it the velocity or a bOdy
such as an airplane or missile Is changed In
magnitUde, or the directlon of the velocity
of the vehicle 15 changed, the vehicle 1s
accelerated which means torces are applied to
the vehicle. In severe flight alrplane
maneuvers like pulling out ot a dive rr-on
high speedS or In striking a severe trans-
verse air gust when flying at high speed; or
in landing the airplane on ground or water,
the torces acting externally on the airplane
are applled rather rapldly and are classed
as dynamlc loads. Chapter A4 discusses the
subject ot airplane loads relative to whether
they can be classed as static or dynamic and
how they are treated relative to deslgn at
aircraft structures.
B1. 4 The Static Tension Stress-Strain Diagram.
The intormation tor plotting a tenslon
stress-strain diagram ot a material is ob-
talned by loading a test specimen In axial
tension and measuring the load with corres-
ponding elongation over a given length, as
the speCimen 1s loaded statically (gradually
applied) tram zero to the falling load. To
standardize results standard size test
specimens are specified by the (ASTM) Amerlcan
Society For Testing Materials. The speed of
the testing machine cross-head should not
exceed 1/16 Inch per Inch ot gage length per
minute up to the yield point at the material
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
and it should not exceed 1/2 inch per inch of
gage length per minute from the yield point to
the rupturing pOint of the material. The
instrument for measuring the elongation must
be calibrated to read 0.0002 inches or less.
The information given by the tension stress-
strain diagram is needed by the engineer since
it is needed in strength design, rigidity
design, energy absorption, quality control and
many other uses.
Fig. 81.1 shows typical tensile stress-
strain diagrams of materials that fall in three
broad classifications. In the study of such
diagrama- various facts and relationships have
been noted relative to behavior of materials
and standard terms and symbols have been pro-
vided for this basic important information.
These terms will be explained briefly.
Modulus of Elasticity (El. The mechanical
property that defines resistance of a material
in the elastic range is called stiffness and
tor ductile materials is measured by the value
termed modulus of Elasticity, and, designated
by the capital letter E. Referring to Fig.
Bl.l, it is noticed that the first part of all
three diagrams is a straight line, which indi-
cates a constant ratio between stress and
strain over this range. The numerical value or
this ratio is referred to as the modulus ot
El.eas't t c.t ty (E). E is.stherefore the slope of
the im tial "st'ra fght portion --or--the stress-
strain diagram and its numerical value is
obtained by dividing stress in poundS per
square inch by a strain which is non-dimensiona
or E :: tiE, and thus E has the same units as.
stress, namely poundS per square inch.
The clad aluminun alloys rave two E values
as indicated in the lower diagram of Fig. Bl.l.
The initial modulus is the same as for other
aluminum alloys, but holds only up to the pro-
portional limit stress ot the sott pure
aluminum coating ~terial. Immediately above
this paint there is a short transition stage
and the material then exhibits a secondary
modulus of Elasticity up to the proportional
limit stress of the stronger core material.
This second modulus is the slope of the second
straight line in ~he diagram. Both modulus
values are based on a stress using the gross
area wntch includes both c are and covering
material.
Tensile Proportional Limit StresS. (Ftb)' The
proportional 11mit stress is that stress which
exists when the stress strain curve departs
tram the initial straight line portion by a
unit strain of 0.0001. In general the pro-
portional limit stress gives a practical
dividing line between the elastic and lnelastic
range of the ~terial. Th~ modulus of
elastiCity 1s considered constant up to the
Bl.3
_--7--
Ultimate Tensile Stress
Proportional Limit
(a) Material Having a Definite
Yield Point (such as some Steels)
Strain ~ Inches Per Inch
______-:::::-:=z.--. Ultimate Tensile stress
Proportional Limit (b) Matertals not Having a
Definite Yield Point (such as
Aluminum AllOYS, MagneSiUm,
and Some steels)
.002 Strain - Inches Per Inch
I
Ultimate Tensile
~ Stress
,
••
•
~
(c) Clad Aluminum Alloys
strain - Inches Per rece
Fig. S1. 1
proportional limit stress.
TenSile Yield StreSS (Fty)' In referring
to the upper diagram in Fig. 81.1, we find
that some materials show a erarp break at a
stress considerably below the ultimate stress
and that the material elongates considerably
with little or no increase in load. The
stress at which this takes place is called
the yield point or yield stress. However many
materials and most flight vehicle materials do
not show this sharp break, but yield more
gradually as illustrated in the middle diagram
of Fig. Bl.l, and thus there is no definite
yield point as described above. Since
permanent deformations ot any appreciable
amount are undesirable in most s~ructures or
machines, it is normal practice to adopt an
arbitrary amount of permanent strain that is
considered admissible for deSign purposes.
Test authorities have established this value
of permanent strain or set as 0.002 and the
stress which e xi st.ec to cause this permanent
strain when released from the material is
called the yield stress. Fig. Bl.I shows how
B1.4 BEHAVIOR OF MATERIALS AND THEIR PROPE RTIES

it is determined graphically by drawing a line
fram the 0.002 point parallel to the straight
portion of the stress-strain curve, and where
this line intersects the stress-strain curve
represents the yield strength or y1eld stress.
Figs. Bl.3 and B1.4 compare the shapes of
the tension stress-strain curves for some
cammon aircratt materials.
Bl.5 The Static Compression Stress-Strain Diagram.

Ult1mate Tensile Stress (Ftu)' The ultimate
tensile stress is that stress under the maxi-
mum load carried by the test specimen. It
should be realized that the stresses are based
on the original cross-sect1onal area or the
test speclmen without regard to the lateral
contraction of the specimen during the test,
thus the actual or true stresses are greater
than those plotted in the conventional stress-
strain curve. Fig. B1.2 shows the general
relationship between actual and the apparent
stress as plotted in stress-strain curves.
The dit!erence is not appreciable until the
higher regions at the plastic range are
reached.
.,....-------...,
~ ACtualStre~/
it~
..
~j
Appar...
Stress
~ '---::-,..-,::--,---- Fig. Bl.2
g
-"
0 short or restrained against lateral buckling,
Because satety and light structural weight
are so tmportant in flight vehicle structural
design, the engineer must consider the entire
stress-strain picture through both the tension
and compressive stress range. This is due to
the fact that buckling, oath primary and local,
is a common type at failure in rll6ht vehicle
structures and failure may occur under stresses
in either the elastic or plastic range. In
general the shape or the stress-strain curve
as it departs away fram the initial straight
line portion, is ditterent under compressive
stresses than when under tensile stresses.
Furthermore, the various tlight vehicle
materials have dirferent Shapes far the region
of the stress-strain curve adjacent to the
straight portion. Since light structural
weight is so tmpartant, considerable etfort is
made in design to develop high allowable
compressive stresses, and in many flight
vehicle structural units, these allowable
ulttmate design campressive stresses tall in
the inelastic or plastic zone.
Fig. Bl.5 shows a comparison or the
stress-strain curves In tension and compres-
sion tor four widely used aluminum alloys.
Below the proportional limit stress the
modulus at elastiCity is the same under both
tension and campresslve stresses. The Yield
stress in compreSSion Is determined in the
same manner as explained for tension.
Compressive Ulttmate Stress (Fcu)' Under a
static tension stress, the ultimate tensile
stresS of a member made trom a given material
is not influenced appreciably by the shape of
the cross-section or the length of the member,
however under a compreSSive stress the
Ultimate compreSsive strength at a member Is
greatly intluenced by both cross-sectional,
shape and length at the member. Any member,
unress very short and ccecect , tends to
buckle laterally as a whole or to buckle
laterally or cripple locally when under
compreSSive stress. It a member Is qUite

-"
:;: 180 then tailure for some materials such as stone,
wood and a rew metals will be by detinite
.
53
,;
120 tracture, thus giVing a definite value tor
the ult1mate compreSSive stress. Most air-
craft materials are so-ductile that no fracture
80
•• 1s encountered in compress1on, but the materIal
53" 40
yieldS and swells out so that the increaSing
cross-sectional area tends to carry increaSing
load. It Is therefore practically impOSSible
to .select a value at the UltImate cccpresetve
stress or ductile materials Without having
Fig. Bl.4

-
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES Bl 5

'0
:!.4S-T3 SH[ET 24S-T4 [XT"USIONS
/
.0
TllleM"!!:$$ .. Q,:2~O

.L X-CO"'''~l!~OH\
-IN
.0
TIUCIl.Jrl(SS" 0.250-11'1.
,/ "
, ""oTiNSION 1'[1'1'10"",,\ f, - - - - -
.7',,

-/1. _. 'I ,,,

.0 '0

-~~
;_1 I
~-T!NSIO!'!:_. "- ......... I~sslo"
-- ,,,
W'COt,lPIl£SSION
'0 I

'0 •• -::s
• I I'-. ,,
·•
'0
/ /
'0 '0 Fig. Bl.6
'0
/ /
/ '°11 The slope of this tangent gives the local rate
0'0 e • • • ot change of stress with strain. The secant
I! ;( 101 IN/IN E X I~ PSI
'0 00
• x
a • • •
10 1 IN/IN E X 10-· PSI
'0 modulus Es Is deter.mlned by drawing a secant
(atr-at ght line) tram the origin to the point
.0,- I COMPllU:U01. in question. This modulus measures the ratio
.'N~
'l'ENSIO'" I I
between stress and actual strain. Curves
,of-....
I·COMPIIl(SSlOIll .0
w-ttNIIOH
r-- t-- ..J po which show how the tangent modulus varies
~~ .. -t("'IO
m wt th stress are referred to as tangent modulus
curves. Fig. 81.5 illustrates such curves
'0f-
I ~~ .0
/
"- tor tour different aluminum alloys. It should
.0
•<
/ ~ '0
;;
• / be noted that the tangent modulus Is the same
as the modulus or elasticity in the elastic

··•• / V
<0 • range and gets smaller in magnitude as the
· I '0 stress gets higher in the plastic r.ange.
30
·V
'0f-
~
'0 · II B1.7 Elastic ~ Jnela.stic Aettoll.

I
'0 It a member is subjected to a certain
'0 L f.'LCt. ...a 7SS-U SHEET
.....D "LATt / stress, the member undergoes a certain strain.
It this strain v.anlshes upon the removal or
TH'CII:7SS 0jIS-O'jS.'l"
'0/ 755-T. !XTI'USIONS the stress, the action is called elastic.

•
a
x 10'
• • •
I"""
'0
E x 10-- PSI
0
0 z
nUCIlNIE:SS.. 3-tN

x_(',.._
Fig. Bl. IS
sane arbitrary measure or crlteron. For
wrought materials It is normally assumed that
Fcu equals Ftu. For brittle materialS, that
are relatively weak In tension, an Fcu higher
• • •
Generally_speaking, tor practical purposes,
• 1I to. INn" e x 10'"
'0
PSI " a material Is considered elastic under stresses
Up to the proportional l1mi t stress as
preViously detined. Fig. 81.7 illustrates
elastic action. However, it when the stress
is removed, a residual strain remains, the
action Is generally reteITed to as inelastic
or plastic. Fig. BI.S illustrates inelastic
action.

t::- i~
than Ftu can be obtained by compressIve tests Elastic Action
or short compact speCimens and this ulttrnate
compressive stress is generally reterred to
as the block compressive stress. ••
•" Permanent
S1.6 Tangent Modulus. Secant MOdulus. In Q)

straiD
Modern structural theory tor calculating Sir.... "t j- straiD
the compressive strength ot structural members
Fig. Bl. 7 Fig. Bl.8
as covered in detaIl in other chapters of this
book, makes use of two additional terms or
values which measure the stiffneSS ot a member Bl.6 Ductility.
when the compressive stresses in the member
tall in the inelastic range. These terms are The term ductility tram an engineering
tangent modulus or elasticity (Et) and secant standpoint indicates a large capacity or a
modulus ot elastic! ty (Es). These two modi- material for inelastic (plastic) deformation
fications of the modulus ot elastiCity (E) in tension or shear without rupture, as
apply in the plastic range and are illustrated contrasted with the term brittleness which
in Fig. 81.6. The tangent modulUS Et Is indicates little capacity for plastiC de-
determined by drawing a tangent to the stress- formation without failure. From. a phys t ca.L
strain diagram at the pOint under consideration. standpOint, ductility is a term which measures
Bl.6 BEHAVIOR OF MATERIALS AND THEIR PROPERTIES
the ability of a material to be dra'Nn into a
wire or tube or to be forged or die cast.
Ductility is usually ~easured by the percentage
elongation of a tensile test specimen arter
failure, tor a specified gage length, and 1s
usually an accurate enough value to compare
materials.
La - L ,)
Percent elongation = ( L Q 100
where LQ =original gage length and La = gage
length after fracture. In referring to
ductility in terms of percent alangation, it
is Unportant that the gage length be stated,
since the percent elongation will vary Nith
gage length, because a large part of the total
strain occurs in the necked down portion of the
gage length just before fracture.
SI.9 Capaci.ty to Absorb Energy. Resilience. Toughness.
Resilience. The capacity of a material to
absorb energy in the elastic range is referred
to as its resilience. For measure of
resilience we have the te~ modulus of
reSilience, which is defined as the maximum
amount of energy per unit volume which can be
stored in the material by stressing it and
then c~pletely recovered when the stress is
removed. The maximum stress for elastic
action for computing the modulus of resilience
is usually taken as the proportional limit
stress. Therefore for a unit volume of
material (1 cu. in.) the work done in stressing
a ~terial up to its proportional limit stress
would equal the average stress f' p / 2 times the
elongation (eo) in one inch. If we let U
represent mOdUlus at reSilience, then
u • Gp) Ep ---------- -----
But e p = tplE, hence
U =(t;) (t,:) =!p"/2E
Under a condition of axial loading, the modulus
at resilience can be found as the area under
the stress-strain curve up to the proportional
limit streSS. Thus in Fig. 81.9, the area OAB
represents the energy absorbed in stressing
the material tram zero to the proportional
limit stress.
High reSilience is desired in members
SUbjected to shOCk, such as springs. From
equation (1), a high value ot resilience is
obtained when the proportional 11mit stress is
stress-strain
Curve"\. " G
~I
••• fp
A
;;"
~'"
~
~
:>
dE
= measure of
I
ductlli tv.
E C H I K
Unit Strain (e)
Fig. Bl.9
high and the strain at this stress is high,
or from equation (2), when the proportional
limit stress is high and modulus of elastici~J
is low.
In Fig. 81.9 if the stress is released
trom point D in the plastic range, the recovery
diagram will be approximately a straight line
D& parallel to AO, and the area CDE represents
the energy released, and often referred to as
hyper-elastic resilience.
Toughness. Tougr~ess of a material can be
defined as its ability to absorb energy when
stressed in the plastic range. Sir-ce the
term energy is involved, another definition
would be the capacity of a material for
reSisting fracture under a dynamic load.
Toughness is usually measured by the term
mOdulus of Toughness which is the amount of
strain energy absorbed per unit volume when
stressed to the ultimate strength value.
In Fig. 81.9, let t equal the ~verage
stress over the unit strain distance d~ from
F to G. Then work done per unit volume in
stressing F to G 1s fd~ which is represented
(1) by the area FGRI. The total work done in
stressing to the ultimate stress f u would
then equal I ok fde, which 1s the area under
the entire stress-strain curve up to the
(2)
Ultimate stress point, or the area 0 A J K 0
in Fig. 81.9 and the units are in. lb. per
cu. inch. Strictly speaklng it should not
include the elastic resilience or the ener~J
absorbed in the elastic ~~ge, but since this
area is small compared to the area under the
curve in the plastic range it 1s usually
included in tougr~ess ~easurements.
It should be noted trat the capacity at
a ~ember for resisting an axially aDplied
dynamiC load is increased by increasing the
length at a member, because the volume is
increased directly with length. However, the
ultimate strength remair.s the same since it
is a function of cross-sectional area and not
of volume of the material.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
Toughness is desirable characteristic
when designing to resist impact or dynamic
loads as it gives a reserve strength or factor
of sarety against failure by fracture when
over-loading in actual use should happen to
cause the member to be stressed rairly high
into the plastiC zone.
The property at ductility helps to produce
toughness, but does not alone control toughness
as illustrated in Fig. Bl.lO, which shows the
s given Yield point stress can be readily ob-
Unit strain
Fig. sr.ro
stress-strain curves for three ditferent
materials. Material (A) is strong but brittle,
Whereas material (C) is weak and ductile, and
material (B) represents average strength and
ductility. However, all three materials have
the same moduius or toughness since the areas
under all three curves is the same.
B1. 10 Poisson's Ratto.
When a material is stressed, it will
derorm in the direction at the stress and also
at right angles to it. For axial loading and
for stress below the proportional ltmlt stress,
the ratio at the unit strains at right angles
to the stress, to the unit strain in the
direction of the stress is called poisson's
ratio. It is determined by direct measurement
in a tensile or compressive test at a speCimen,
and is approximately equal to 0.3 tor steel
and 0.33 tor non-terrous materials. In many
structures there are members which are Subjected
to stresses in more than one direction, say
along all three coordinate axes. Poisson's
ratio is used to determine the resultant stress
and detormation in the various directiOns.
pOisson's Ratio in Plastic Range. Information
is somewhat limited as to Poisson's ratio in
the plastic range and particularly during the
transition range tram elastic to plastiC action.
For the assumption ot a plastically
incompressible isotropic SOlid, poisson's ratio
assumes a value or 0.5. oeram and Weldrorm.
tound tram their research, that the transition
or poisson's ratio rrmn the elastic value or
around 1/3 to 1/2 in the plastic ra.'1ge is
gradual and is most pronounced in the yield
pOint region at the stress-atrain curve.
81.7
BI. 11 Construction of a Stress-strain Curve Through a
Given Yield Stress by USing a Known Test stress-
Strain Curve.
Materials in general are produced to-
satisfy certain guaranteed minimum strength
properties such as yield stress, ultimata
tensile stress, etc. Thus in the deSign ot
important members in a composite structure,
the minimum guaranteed properties must be used
to provide the required degree at safety. In
general, most materials will give properties
slightly above that guaranteed values, thus
the test stress-strain curve or purchased
material cannot be used for design purposes.
The stress-strain curve passing through a
tained for a test stress-strain curve as
tallows ;
In Fig. Bl.ll, the heavy curve 0 8 C
represents a known typical stress-atrain curve
tor the given material. Let the minimum
guaranteed yield stress be the value as shown
at point A, using the 0.2 percent method.
Then proceed as tallows:
(1) Draw a straight llne through potnt 0 and
paint A which will intersect the typical
curve at point B. Point B may be above
or below the typical curve.
(2) Locate any other point on the typical
curve such as point C, and draw line
rr-ca 0 through C.
(3) Locate point D on line 0 C by the follow-
ing rat.t o r-,
OD=~XOC
(4) Repeat step 3 to obtain a number or
pOints as shawn by dots on Fig. BI.II,
and draw smooth curve through these points
to obtain desireu stress-strain curve.
t ( . . - . . I. . . .
Fig. Bt.11
B1.12 Non-Dimensional stress-strain Curves.
The structural designer is constantly
cOnfronted with the design or structural units
which fall by inelastic instability. The
 BI.8 BEHAVIOR OF MATERIALS AND THEIR PROPERTIES
solution ot such problems requires information
given by the compressive stress-strain curve.
Since flight vehicles make use of many differ-
ent materials, and each material usually has
ma.ny different states of manufacture which
give different mechanical properties, the
question at time required to obtain certain
design intormation from stress-strain curves
becomes important. For example, in the
aluminum alloys alone there are about 100
different alloys, and when elevated temperatures
at various time exposures are added, the number
at stress-strain curves required is turther
greatly increased.
Fortunately, this time consuming work was
greatly lessened when Ramsberg and Osgood
(Ref. 1) proposed an equation to describe the
stress-strain curve in the yield range. Their
proposed equation specifies the stress-strain
curve by the use of three parameters, the
aonnua at elasticity E, the secant yield
stress F O. '1" which is taken as the Una at
slope 0.7E drawn trom origin (see Fig. Bl.12),
and a parameter n which describes the Shape
of the stress-atrain curve in the yield region.
In order to evaluate the term. n, another stress
Yo •• is needed. which is the intersection of
th9 curve by a line or slope of O.SSE through
tho origin (s•• Fig. Bl.12).
O.8SE O.7E
E/l Stress-Strain
c.....
_ ¥o-cJ .....Jl... FIg. 81.12
).......... ~ The Ramaberg and Osgood proposed three
parameter representation of stress-strain
relations in the inelastic range ta r-,
.. Ee t 3(
Fo.'I'=FO'.~~Fo.?
t )n. --- - - - - -
The sQuation for n Is,
n = 1 =. log. (17/7)/log.(F •••IF •••• )
Fig. Bl.13 Is a plot of .quation (4). The
quantities EeIF-o . ? are non-dimensional and may
be used in determining the non-dtnenslonal
curves of Fig. Bl.14. E. n, and F 0.'1' must be
mown to use these curves 'n obtaining values
... ~t..-
on the stress-strain curve. Table 81.1 gives
the values of F o . ,., to.ee. n , etc .• for many
flight vehicle materials. Notice that the
shape parameter varies widely tor materials,
being as low as 4 and as high as 90.
HI. 13 Influence of Temperature on Mater1al Properties.
Before the advent of ~~e supersonic air-
plane or the long range miSSile. the aero-
nautical structures engineer could design the
airframe ot aircraft using the normal static
mechanical propertles ot materials, sInce the
temperature rise encountered by such aircratt
had practically no effect on the material
strength properties. The development of the
turbine Jet and rocket Jet power plants pro-
vided the means ot opening up the whole new
fIeld of supersonic and space flight. The
flight environmental conditions were now
greatly expanded. the major change being that
aerodynamic heating caused by high speeds in
the atmosphere caused surface temperatures
on the airframe which would greatly etfect
the normal static material strength properties
and thus temperature and time became important
in the structural deSign ot certaIn types of
tlight vehicles.
Bl. i4 Creep of Materials.
It Is obvious tbat temperature can weaken
a material because If the temperature is high
enough the material will melt or tlow and thus
have no load carrying capec t tv as a structural
member. When a stressed member Is subjected
to temperature, it undergoes a change at Shape
in addition to that of the well known thermal
expanSion. The tenn creep is used to describe
thiS general intluence at temperature and time
on a stressed material. Creep is detined in
general as the progressive, relatively slow
change in shape under stress when subjected
to an elevated temperature. A simple illustra-
tion ot creep Is a person standing on a
bituminous road surface on a very hot summer
day. The longer he stands on the same spot
the deeper the shoe soles settle into the road
surface, whereas on a cold winter day the same
time ot standing on one spot would produce no
noticeable penetration of the road surface.
(3)
High temperature. whe..l used in reference
to creep, has ditferent temperature values for
different materials for the same amount of
creep. For example, mercury, which melts at
-38OF. may creep a certain amount at -75 0F,
(4) whereas tungsten, whf en me 1 ts at 6170OF, :nay
not creep as much at 200QoF as the mercury
uncer -75OF. All materials cr••p under-
conditions Of temperature, stress and time ot
stress application. The simplest manner in
which to obtain the etrects at creep Is to
study its ettect on the static stress-strain
diagram tor the material •
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES Bl.9

Fig. B1. 13

50

.0 ""'-
30
r-,
25
-,
20
-,
~
15
-, toe. (17fT)
11-1+
~. (FO.7"'0'.86)
10
••
1
• <,
•
5
I'....
<,
• r--,
3
t"-
•
1. 5

1. 0
1.01 1.02 1.03 1.D41.051.08 1.08 1.1 1.15 1.2 1.3 L'" 1.51.8 L8 %.0
"0. T/FO• as

Fig. Bl.14

1.'
,
1.' ,I
»<
, ~ /"

f:?
1.0 "

0.' h -a

~~
f
FQ.";
0.'

~~ ,
0 .•

1/ 1
0.'

o
1/0.' 0.' o•• 0.' 1.0 1.' 1•• l.S '.0

C / c'"
-::..- t' ~,
 ,- .,:

Bl.10 BEHAVIOR OF MATERIALS AND THEIR PROPERTIES

Table 81. 1 Values of Ftu> 1""",. Eo:. 1"0.7. F 0.85, 'to or v arlOUS !Iota
. rer ia1s
,
Unci ., Roo m &. El evat odTempera ures (F rom R .. t 6)

MATERIAL
Temp.
E"". Temp. I
,
e, I"tu".
-. 'r I" :Y:' ., E, 1"0.7 I 1"0.as ,

i ,,i '"
OF ~. n
H,. I koSi 106pl!l i I kosi I kosi . I
STAINLESS STEEL
AISI 301 114 Hard Sheet
Tr:uuJ'ferse Compresllion
1/, I, RT zs 125 ao 27. a 73 I, sa 5. ,

L.olJiitudirlal Compresllion 1/2 I RT as 125 43 26.0 28.2 aa s. 2

A1SI JOI 1/2 Hard Sheet 1/, ! RT 15 150 118
TJ"Innene Compression 1/2 400 na
no
106.5
27.0
23.2
116.5
106.5 '"" U
a.e
S. a
1/'
1/' i '00
1000
107.5
ee
20.9 106.5 96.5

Loq:ltlldlnal Compression 1/,

1/,
! RT
400
15 "'
150
na
ee
53.3
16.2
26.0
22.4
94.5

"
45.5
83.5
J7
as
I
6.0
4.4
4.7
1/2 '00 no 52.8 ac. 1 44 31 3. s
1/, 1000 ae 45.2 15.6 40 30.5 4.3
AlSI 301 3/4 Hard Sheet 1/, RT 12 <7, 180 27.0 163.5 151. 5 13.2
Trana'l'eMie Compresaton 1/,
1/'
<00
600 I '"
138 '"
138
24.1
22.4
153 I 142.5
140
13.2
I
Lofli1,tu4lna1 Compresllion
11'
,/2
1000
RT 12
11J
175
ua
78
16.9
26.0
'"to
127 I 1Jl
61.5
11.2
19.2
7.5
I
1/,
1/'
400
60Q '"
"a
71
70.3
23.3
21. 5
65
65.5 I se
56.5
I '.6
e. a
1/, 1000 na 59.3 18.2 ss 46 5. S
AISJ 301 Full &rei Sbeet 11' RT 6 165 179 27.0 <83 172 <8
Tranuerse COl'nprepwll. 1/,
1/,
.00
600
168
15'
"a
"a
25.1
23. a
17.
<7,
18. I ie
18
Lou¢tudln.a1 Comprnllwn
11'
1/,
1000
RT a
131
185 Ii(
130
65.
21.6
26.0
141. 5
77.5
'"ea
135.5 21.5
5.2
1/, 400 "a 60. a 24.2 74 59.5 5
11'
1/'
'00
1000 '"
131
79.9
66.3
22.9
20.6
74
ea
sa
42.5 , •••s
3.
l'l"'" PH B&r II. Fol"l1nP 1/, RT e 165 27.5 168 <80 J4
1/, .00
700
180,
'"as
146 '"
105.5
25.3 "7 129
er
I <8
n
17.7 PH (THtOSO) Sheet, Strtp II. Plate.
1/'
1/'
1/,
1000
RT 180
62.6
23.1
21.2
29.0
'"60 sa
145 II 7. 1
7.4
t '" .010 tD .1251D.. 1/, .00
'"H.ea
18'
H. 27.11 '"n7
'146 s. a

17·7 PH (RlI9S01 Sheet:, Strip II. Plate,

11'
11'
700
"00 I
ue
61.5
24.9
20.3 sa ".'"
47
I 8.4
e
1/' RT 196
t ... 010 tD • 125 In.
JlO
""45 29.0
"" 16.4
19.9DL (AMS 5526) 6 19-9DX
(AMS 55381. Sheet, StZ'tp &: Plate I, 1/, RT JO ss 29.0 36.5 32 7.5
19-QDL (AMS 5527) C. 19_QDX 1/, RT 90
(AMS 5539) Shnt, Strip &: Pl.a.te
PRl!l~7Vo (THl0501 Sheet &: Strip. 1/2 RT I "e '"
190 170 I
29.0

28.0
65
171
74

134
7.2
22.5
t ... 020to.187ir1.
PHU-7Mc (RH950) Sheet &: Strip,
t '" . 020 tD .187 in.
11' RT II 4 aae '00 28.0 are 189 '-'
i
L.OW CARBON &: AL.LOY STEELS ,
AJSI 1023 1I 1025 Tube. Sheet" Bar,
Cold Flniab.ed RT aa ee ae 29.0 32. '7 31. 5 J4
AJSl4130 NQrmaJUed, t ~ .188 In. 1/2 RT J3 90 70 29.0 61. 5 53 e.•

AISl4130, 4HO, 4340 Heat Treated

1/,
11'
1/,
1/'
'00
900
1000
RT aa
I
..
a,
58

125
I,
61. s
48.2
30.8
113
27.3
23. a
40.8
29.0
ss
40
aa
III
48
32.5
aa
102
I
7.3
5.'
4.7
10.9 ,
1/'
11'
500
.60
113
SO I 98.3
58.9
27.3
23.2
se
66.5
SO
61. 5
10.9
I
AISI 4130, 4140. 4340 Heat Treated
1/,
1/'
1000
RT 18.5
34
150
49. '7
14'
20.8
29.0
45.5
145
41
140
"ass.a I
11' '00 135 27.3 rae as I
1/' 550 105 '"88.5 23.2 58 '"
63.5 18. 5
AISI 4130. 4140, 4340 Heat Truted.
1/'
11'
11'
1/2
1/,
1000
RT
500
650
100'
,
I
15
76
180

'"
'"sa
63.8
17'
15'
109.3
20.6
29.0
27.3
23.2
20.6
"
17'
15'
109.4
57
17'
153
10'
.
10.9
eo
aa
s.e
AIB14130. 4140. 4340 Heat Treated 11'
1/,
RT
500 I 13.5 200
180
"
19' 29.0
27.3
"
19'
58
19' JO
17' 172..5
1/, '50 140 23.2
16' .46
as II
A Y
11' 1000 r 104 '"87.1 , 20.6
121.5
57
117
83 19
I
A-2M (AMS m!5Aj Sheet, ?Iate
I,
."..., 11'
1/2
RT
600
1000
15 140

'"sa
95
88.4
29.0
24.4
93
57
er
57
ai
14
13.5
11'
1/, I 1400
115 81. '7
50.3
19.8
14.2 50
ts
47
12• .5
15. 3
K-MONEL Sbeet, Age Hardened
MONEL Sheet:. Cold Rolled '" Annealed
1/,
1/2 I RT
RT
15
35
12'
70
JO
ae
Z6.0
26.0
ee
20
ea
17
13. 5
'.4
.
tNCONEL-X 1/, RT 20 155 10' 31.0 10. 100 23.5
1/2 .00 162 95.6 28.9 94 89 17
11' '00 141 90. a 28.4 a8.6 18. 5
1/, 1<:00 10. aa 23.2 78.8 ar
"
I
I, I

i I,
I
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES m. n
1'lOble al 1 Valu
" .
of Fw I''Y' Ee, FO 7 FO 85
()fL. ~L
n, [or V;u1.01lB MaterwlI Under Roo m &: Ele v aled l'om", r:;ltl1rell (FrQ m Ret 6j (C0 nUnu ed)

MATERIAL
Temp.
""''"'.. Temp.
'F ,
o. F"" FCY'
"
I 106pSl -. .
Fo• 7. F O•8S
I ,
ALUMINUM ALLOYS "" "" Ul'"
'"
2014~T8 ~Naiona 2 RT 7 60 ea 10.7 53 50.3 18.5
t ::!' O. 499 in. 2 300 42.5 10.2
2
2
420
600
"28to 21 3.2
4.1.5
20.5 "
19.5
,. S
"25s.,
8.0 7.' S••
1/2 300 Sl 4:1.5 10.2 44.0 42.5 25
1/2 420 31 28 9.2 ae 25.2 29
2014-1'6 Forpngs 2 RT 7 ea sa 10.7 52. :1 SO 20
t ::!' 4 In. 2
2
2
1/2
1/2
300
'50
600
300
..0
sa
29
10
53
32
41
22
43
7••

25.5
10.2
9.2
7. ,
10.2
9.2
40.5

,
21.5

42.5
280
.. 38.5
20
3.0
"21.
19
12. e
3.2
15.8
"6
2024-1'3 Sheet & Plate, 2 RT 12 65.'~Q 10.7 .a-c 36.
.
*"
11.' ,
Heat Treated. t :: . 250 In.. -: 2 300. 10.3 35.7 • 33.5 15 ...

I
I 2024-1'4 Sheet &. Plate.
2
2
2 RT
SOO
700
12
"
28

36
7.•
8. ,
8. ,
1.7
2U
6.2
31!.7 •
22.8
s.s
34. s.,
10.9 j
8.2 '
ia e
Heat Treated, t ~ O. SO In. 2 300 34
2
2
'00
700
24
7
10.3
8.' """
22
30.5
21
14. e
10.2
2024-1'3 Clad Sheet &. Plate,
Heat Treated, t = .020 to .062 In.
2
2
RT
300
12 60 27
34
6.'
10.7
,0. 3
60
35.7·-
25
"
25
30.3
18.5
12
11

... .. .
2 SOO 24.5 B., 21.7 20 7.9
2024-1'6 Clad Sheet &: Plate,
Heat Treate<l, t ? O. 063 In.
2
2
2
700
RT
300
8 ea
B.S 6. ,
10.7
10.3
'.8
44.3
•••
40.7
l}5
11
2 SOO 22 8.' 31.5 28 8.3
2 700 6 8.' 7.0 6.0 6.8
2024-1'8 Clad Sheet &. Plate, 2 RT 8 60 47 10.7 47 43 10.6
Heat Treated, t < O. 063 In. 2 300 43.2 10.3 42.3 2L7 10.8
2 SOO 31 8.' 29.s 28 7.8
2 700 6 6.' '.00 '.0 U
2024-1'81 Clad Sheet, Heat Treated, 2 RT s ee
tcO.064 In. 2 300 " "50.s 10.7
10.3 51. :I
51.8
46.5
11.2
io
2
a
6061-1'6 Sheet, Heat Treated &: Aced, 1/2 RT 42 ae
t<O.25In. 1/2 300 " "
29.5
10.t 34 31
1/2 '50 ,11.5
3••
B•• "
19.3
28
17.7
28
10.9
7075-1'6 Bare Sheet &. Plate,
t~0.501n.
1/2
2
a
RT
BOO
300
7 76 .....
54
7.5

.. ,
7.0
10.5
6.6
to. .
55.8
6.2
B3
52.5
15. :I
9.2
15.8

7075-~ Extrusions,
t~0.251n.
a
2
1/2
2
2
...
425
'00
RT
300
7 73
25.5
8
30
70
54
8.1
'.2

.. ,
B.l
10.5
25.4-
7.2
34 .s
72
22••
'.2
32. e

"54.'
12.1
3.7
re
12.8
58.5 11'
2 '50 22•• 7.8 21.3 IL. 7.2
2 .00 8 '.3
,..s
6•• '.3 3.2
7075-1'8 Die ForginjJ:',
t ~2 ill.
1/2
2
2
2
2
"0
RT
, .
300

600
7 71
25
"
47.S
18.5
7.0
7.B
10.5
8••
7.B
'.3
53
47.8
17.3
'.0
.
"
.Ll
te
3.7
B.8
15.2
1e.e
12
19
1/2 22
7Q75- 1'6 HaDd Forginp.
Area :s 16 sq. In.
2
2
RT
"0
300 • 72 B3
51.6
7.8
10.5
••• "
63.8
~1.2
22
61.5
60
10.9
25
21.5
2 '50 7.6 20.3 13.7
2'"
"'.0
7075- re Clad Sheet &. Plate,
t ~O. 50 In.
2
1/2
2
2
a
RT
600
.5O
300
'50
B 70 .7.6
24
60
:l0.5
'.3
7••
10.5
•••
7.8
....
6.0
se.e
54
19.7
25.3
er.e
51.7
17.5
•••
19.5
19.5
20
<6
2 600 7.7 '.3 7.7 S.s 3.6
7079_ TlS HaPd Fof"iinP.
</2
</2 RT
"0
, 67
23
59
7••
to. 5
27.2
59.5
.'"
25.3
57.5
12.4
,.
23
t ~ 6.0 In. 1/2
1/2
300
'600
W ,
67
21
•••
7.8
4S.5
20 12
</2 7.0 '.3 e., 3.s '.0
MACNES M .... LLO
AZ61A Extrusions,
t~0.249il1.
RT • ae 14 6.3 12.9 12. J is
HK31A-0 Sheet </2 RT , 12 30 12 6. S 10 8. , 6
a.s e.s •.s
I
t • 0.016 to O. 250 in. 1/2 300 20 11.1 6.16
</2 .00 is '.3 4.94 7.s '.6 ••a
</2 '00 10 '.9 3.77 3.3 I .• 2.2
KK31A-1i24 Sheet, 1/2 RT
• 34 19 6.s 17.3 14.6 6.2
t-0.250ill.

THMn ~eaJ.~
</2
1/2
</2
300
.00
.00 I
22
17
11
17.7
14. a
7.a
L2
•••
3.8
IL 6
13. 1
'.7
12.6
10.5
S.2
I
I
e.,
,
'.9
..
Plate &. Strip
n-ao-w Annealed
&. Sheet, t ~.
Sheet,

Bar
167 ill.
1000

i/a
1/2
</2
RT
RT
.00
600
10
10
120

130
10.
sa
~
I '"
123
se
... s
15.5

16.0
14.1
t3.0
119.5

127
67
85.5
I 102
124.5
83
aa
I
."
13.7

22·

I
</2
</2
800
1000
67
70 t
". ,
SO .s
11.8
'.7
80.5
er
I

I
-n
59.5
21. 5

"

0,......
~
.....---
Bl.12 BEHAVIOR OF MATERIALS AND THEIR PROPERTIES

Fig. Bl.lS (curves A and B) show the B1.15. The General Creep Pattern.
stress-strain curve for a material. Curve (A)
is tor a low elevated temperature condition A typical manner of plotting creep-rupture
and curve (B) that ot a high elevated test data is illustrated in Fig. Bl.1S. For
temperature condition. The results were ob- metals tested at high value of stress or
tained by a normal testing ~chine procedure temperature, three stages in the creep-time
requiring a short time test period, hence the relation can be observed as shown in Fig. Bl.lS.
results can be considered as independent ot The initial stage, otten called the stage ot
time. primary creep, includes the elastic deformation

--
and that region where the rate ot creep de-
~;":'t p',,- pc ! tormation decreases rather rapidly with time,
~!:'!-:;,

I:';",:
'.rr
"r:~":::'
'';::~~~,
::p- ~r-
.
i
I'
I
' J..- i

;
...
,~, ...~l..._-L_
.... .... __..... ---
...f --
HMr'-....
--
which no doubt Indicates an influence ot strain
hardening. The second stage, otten referred
to as the secondary creep stage, represents a
//c ., .-
,,~, 'c_
-'0'
, j ;

...««:
0

:~;i .- -7' f
:.::.;,z-_~_ ; stage where the rate ot strain has decreased
W,.~~ ~
~:
-. , .-- I. DIQlIt1'-'tWE TIJeOlIl
to a constant value (except tor high stress)
. i--·- tor a conSiderable time period, and this stage
~~~~i~ """~ ,<" ,. , H' ! represents the periOd ot minimum creep rate.
.
~~.
I

- ~\ The third stage, otten called the tertiary

?f.
. ':L " i, l . .Ii 0
I . creep stage, represents the period where the
'. t, f i,:
.

..
,
.~ ,

~~ ~.?

,."., ..
t"
...• ~,
e"
~;"

, / r-
-. ,- I
,
' .

....
. , reduction in cross-sectional area leads to a
higher stress, a greater creep rate and tinally
rupture •
f.'.! L". .' II I' :" FO-" !&~
~-::~-. . , Transition Point. The inflection ~oint between
.....
"

r.S
i;.'!c ,•.

--
..•. r..
"y ',"
,i'.~, ?:".i' , - the constant creep rate ot the second stage and
J
the 1ncreasing rate ot the third stage is
""""'- reterred to as the transition paint. Failure
Fig. Bl. 15 (Ref. 2) - Effect of temperature and time on the generally occurs in a relatively short time
strength cba.raetertst1cs of metals. attar the tranSition point. Transition points
may not occur at very low stresses and ~y
The tigure shows that the higher also not be definable at very high stresses.
temperature (curve B) reduces the ultimate Minimum creep rate is that indicated in the
strength, yield strength 'and modulus ot second stage, where the creep rate is practi-
elasticity or the material as compared to cally constant.
curve (A) which is a test at a lower temper-
ature.
Initial j Second Stage Third
EFFEX:T OF TIllE (Approx. Constant) Stage
stage :
Rupture
I Point
It temperature and stress are ot such
combination as to prOduce appreciable creep, ,,
I
I

then t1me becomes an important errect , For

example, in Fig. Bl.lS, it the materIal at low
J
,,
temperature (curve A) Is stressed to a value P
within the elastic limit ot the material, and ,,,
I
.-
this stress Is maintained tor a considerable ,
, , Transition Point
time periOd, very little creep, it any, will
be detected, and when the stresS Is removed, ~ Slope"" Creep Rate
the material will practically resume its Creep Intercept
original dimensions. However, it the materIal
Is stressed to the same value P but under a
high temperature condition, creep will occur Time - firs.
Just as long as the stress P remains applied Fig. Bl.16 Typical creep-rupture curve.
and. the stress-strain curve will take a Shape
like curve (e) In Fig. Bl.15. This time Fig. 81.17 shows a plot ot cree-time
dependent strain will follow a characteristic curves tor a material at constant temperature
pattern. The material will never return to and several stress levels. It is noticed
its original Shape atter creep has taken place that increasing stress changes the creep-time
regardless it the stress is removed. It both relationship considerably.
stress and high temperature continue, rupture
produced by creep will tinally occur. ai. 16 Stress-Time Design Charts.

In many structural deSign problems i~­

volV1ng elevated temperatures such as power
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES B1. 13

1
i
o

Fig. St. 19 (Ref. 3) • Effect of exposure at elevated temper-

atures on the room-temperature tensile yield strength (Fty)

_. of 7079-T6 aluminum alloy (band forgings).

Fig. Bl.17 (Ref. 2) Creep curves for a material at constant

temperature and various stress levels showing the character-
istic stages of creep behavior.
iI
plants, the critical design tact or is not a
strength but the permiSSible amount at creep
that can be allowed to still permit the
structure or machine to tunction or operate
satiSfactorily. Extensive tests are usually
necessary to provide reliable creep design
informatIon and such test intormation IS often
recorded in the form as illustrated in Fig.
Bl.1B.
-'
Fig. SI.20 (Ref. 3) - Effect of temperature on the tenstle
yield strength (Fty) of 5 Cr-Mo-V aircraft steels.

, , strasses at a given temperature when heated

:m
, II I
, I d', :' , II! " , , tram zero up to 2000F per second under constant
, load conditions tran when loaded attar the
I, , ,
'II I " ,
II11 , IIII material has been 1/2 hour or longer at
i: I , I' II , , I I II II constant temperature. Increasing the temper-
atura rate tram 2000 F to 20000F per second or
II II , III i II I i
--- - i '1111 more results in only a small increase in
strength. Figs. Bl.2l and Bl.22 (Ret. 4)
show the etfect ot temperatura rates up to
Fig. BI. 18 (Ref. 2) - Stress-time design chart at a single lOoOF per second upon the yield and Ultimate
constant temperature for selecting limiting' stress values.

SI. 17 Effect of Time of Exposure.

",~-l--+--+-+--+-+--+----1
In general materials can be roughly
classified into those which time ot exposure
..--
to elevated temperatures has great influence
on the mechanical properties and those where
such exposure produces relatively small etfect.
This general tact IS illustrated in Figs.
81.19 and 81.20. The yield strength ot the
aluminum alloy In Fig. 81.19 IS tar more
- ~

"'f---+-+--+-+-~~+--+-

influenced by time or exposure to elevated

temperature than the steel alloy as shown in
Fig. Bl.20.

S1. 18 Effect 01 Rapid Rate of Heating i

SupersoniC aircraft and missiles are
subjected to rapid aerodynamic heating. The Fig. S1.21 - Tensile yield stress of 2024-T3 aluminum
results of tests indIcate. that in general alloy for temperature rates from O.20F to lOOop per
metals can withstand substantially higher second and of stress-strain tests for l/2-hour exposure.
Bl.14 BEHAVIOR OF MATERIALS AND THEIR PROPERTIES
~'C-I- , - - - , - - - - , - - , - , - - - - ,
I
Fig. Bl.22 - TensUe rupture stress of 2024-T3 aluminum
alloy for temperature rates from O. ~F to l00 0F per second
and ultimate tensile stress of stress·strain tests for 1/2-
hour exposure.
strength respectively of aluminum alloy as
compared to values when loaded after the
material has been exposed 1/2 hour at constant
temperature.
B1. 19 General Effect of Low Temperatures Upon Matertal
Properties.
The development or the missile and the
space vehicle brought another factor into the
ever increasing number ot environmental con-
ditions that etfect structural design, namely,
extremely low tetnperatures. For example, in
space the shady side of the night vehicle is
subjected to very low temperatures. Missiles
carry rueta and oxidizers such as liqUid
hydrogen and oxygen which boil at -423 and
_297oF respectively. In general, low
temperatures increase the strength and stiff-
ness at materials. This etfect tends to
decrease the ductility or the material or, in
other WOrdS, produce brittleness, a property
that is not desirable in structures because
of the possibility ot a catastrophiC failure.
In general, the hexagonal closely packed
crystalline structures are best suIted for
giVing the best service under low temperatures.
The most important at such materials are
aluminum, titanium, and nickel-base alloys.
Fig. 81.23 shows the effect of both elevated
and low temperatures 0n the ultimate tensile
strength ot 2014-T6 aluminum alloy under
various exposure times.
B1.20 fatigue of Materials.
DeSigning structures to provide satety
against what is called fatigue failure is one
of the most important and dlfttcult problema
rac ing the structural designer at flight
Fig. 1.23 (Ref. 3) - Effect of Temperature on tntimate
strength (Ftu) of 20l4-T 6 Aluminum Alloy.
vehicles. Fatigue failure is failure due to
being stressed a number of times. For example,
a beam may be designed to sately and efficient-
ly carry a deSign static load and it will carry
this static load indefinitely without failure.
However, if this load is repeated a large
enough number of times, it will fail under thiS
static deSign load. The higher the beam stress
under ~he static deSign lead, the less the
number of repeated loadings to cause failure.
To date no adequate theory has been
developed to clearly explain the fatigue
failure of materials. Fatigue failure appears
to begin with a crack starting at a point of
weakness in the material and progreSSing along
crystal boundaries. A microscopiC examination
of metals indicates there are many small
crackS scattered throughout a material. Under
the action ot repeated stress these small
cracks open and close during the stress cycle.
The cracks cause higher stress to exist at
the base ot the crack as compared to the
stress it there were no crack. Under this
repeated concentration of stress, the cracks
will gradually extend across the section or
the member and finally causing complete
failure of the member.
Fatigue testing consists of 3 types:-
(1) the testing of material crystals, (2) the
testing of small structural test specimens,
and (3) the testing or complete compOSite
structures. A tremendous amount or test
information is available for the second type
of tes~1ng. More and ~ore attention is be1~~
given to the third type of testing. For
example, a complete airplane wing or fuselage
is often subjected to elaborate fatigue
testing in order to insure the safe deSign
lite of the airplane.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES B1. 15

The presence or cracks initially is not magnitUde at stresses in aircraft structures Is

necessary to start a fatigue or progressive
failure as irregularities such as slag
inclusions, surface scratches, pitting, etc.,
can cause corrosive action to start, thus
supplying the condition tor the promotion at
cracks and the resultant progressive failure.
The strength of ferrous metals under
repeated stresses is otten reterred to as the
endurance or fatigue limit. The endurance
limit stress 1s the stress trat can be repeated
an infinite number of times without causing
fracture of the material. Non-ferrous
materials such as the aluminum alloys do not
have an endurance limit as defined above but
continue to weaken as the stress cycles are
increased. Due to this fact and also since
the required service life of structures and
machines vary greatly, it is customary to
refer to the strength under repeated stresses
as endurance or fatigue strength instead at
endurance limit. Thus the fatigue strength
is the maximum stress that can be repeated
tor a specified number of cycles without
producing failure at the structural unit.
The results of testing a speCimen under
repeated stresses such as tension, compression,
bending, etc., is often plotted in a torm
which is referred to as the S-N (stress versus
cycles) diagram, as illustrated in Fig. 81.24.
discussed brietly in Crapter A4. The limited
discussion which tollows will deal only with
the effect or impact loads on the behavior of
materials.
IMPACT TSSTING METHODS
There are in general two types of tests to
determine the behavior of materials under
impact loads. The usual tmpact test which has
been conducted tor many years Is referred to as
the notched bar test and consists of subjecting
notched specimens to axial, bending and torsional
loads by the well known Charpy or Izod impact
testing machines. In both of these machines
an tapact load is applied to the specimen by
SWinging a weight W trom a certain vertical
height (h) to strike and rupture the notched
specimen and then stopping at a vertical
height (ti I). The energy expended in rupturing
the specimen is then equal approximately to
(Wh-Wh I). This type or test 1s primarily used
tor studying the influence ot metallurgical
variables.
The other type ot impact testing is made
on unnotched speCimens and the general purpose
is to obtain the stress-strain dia~ at
materials under impact load or the load-
distortion diagram of a structural member or
composite structure as the unit is completely
fractured under an impact load.

\ Bl.22 Examples of Some Results of. Impact Testing

.-. 11::t 10"

!
'Jl 10 X 10"
-, of Mater1als.

Figs. 81.25, 26 and 27 show the resUlts

•
•
g
'"
9 x 10•

8 x 10 •
'"r--... at impact tests upon the stress-strain curve
as compared to the static stress-strain
diagram (Re t . 5).

•
7 x 10 10 •
! I I j
10" 10"
Number of Cycles for Failure, logH-
10 'r looj : :
I '
i
, ,
, ,i
I :r
"'
•
, II i
'
04 f.'jsec
i
A '.±t.'
50
Fig. B1.2. i~i
~' I~ 1 : i

The problem of fatigue design ot aircraft ?isol

§ ~0'
u i
: ::rstar,c
,

airframes is covered in Chapter Cl3 at this _ 40' , I i f I ,

book. ~ :1 I:
III , I
~o:?O I I
Bl.21 Effect of Impact Loading on Material Properties. ,
! I '
12 . .s 20
.~4 28 .J2
An impact load when applied to a structure Unit ~on9aIiQn.lnlin.

produces appreciable shock or Vibration. To

prOduce such action, the load must be applied
rapidly, that is, in a short interval at time. Fig. Bl.25 • S'ress-strain curves, 24ST aluminum alloy.
The efrect at linpact loads differs trom that
at static loads in that impact loads appreciably
ettect the magnitude at the stresses produced
in a member and also the resistance properties
or behavior of the material under load. The
importance and etfect at dynamiC loads on the
 .
Bl 16 BEHAVIOR OF MATERIALS AND THEIR PROPERTIES

, ,
'I'at:e B1.2 snows additional impact testing
I I I , ,
I I
so I I Slotic v '" /5C f:'sC'cl
results as compared to the static test results.
~40 Ih
c- , , , ]V '" co rl/s9Ci I , i Table B1. 2
~30
,
I , I 'I
, , , I , Comparison of Strengths, Ductility and
I
,i
,
§ , , I , I
I Energy Absorption Under Impact
, ,
·••
'.20

~ 101
I
I
, I
,
,
, ,
,

I
,
i
, I
I I
I and Static Loads. xxx

Ratio of Impact to Static Value For,

I I I I I , , ,
Material
0 IJ2 04 C6 :J8 ,0 '2 /4 16 /3 Yield Maximum Elong- Energy
I;IlI elongotion, in./in. Point Load ation

SAE No. 6l4()X 1. 020 2.205 2.865

Fig. Bl. 26 - Stress-strain curves, Dow metal X. SAE No. 101 1.328 1. 285 1. 418 1. 376
SAE No. 101 1. 996 1. 397 0.992 0.946
18-8 Alia 1.224 1. 212 0.682 0.600
17ST Alum. Al. 1.29 94 1.
Brass 1. 387 1. 142 1. 163 1. 203
Aluminum 1. 522 1. 323 1. 628 1.6
Copper 1. 1. 390 1. 27 .
240
"
x Oil quenched from 162QOF.
I I ! I I I ! I XX Cold-rolled.
I I I I ! , i :c:a From "Stress-Strain Relations Under Tension and

,
i!OO Impact Loading" by D. S. Clark & G. Datwyler,
'/ ~ 80 frl$f1C I
Proe. A.S. T. M. 1938. Vol. 38.
--<.C:-t-- : ,
,,
s" I
I
I I
I
I
I !:;/~1I,1 r\ "-.'
I I I i \ , ,I I
REFERENCES
~I I'/ ~
.,-, !
I
I

I
,
I I I I ,
JSO fI/s.cr"-:- I

!
I :
,,
i
Ret. 1. Ramberg & Osgood. Description of
Stres~-Strain Curves of 3 Parameters.
NACA, Tech. Note 902.
"
~
,
I
I I ,
,
,
:

I
I
I
I Ret. 2. T1me and Temperature Greml1ns of
, Dest·.llctlon. By L. A. 'rerxorvicn,
'" i I
,
, I
Correll Aero. Lab. Research Trends.
I I I i I I I I I I i Se"'; • 1956.
0 ce 04 06 co .,0 /2 .14
(.Inil ef0l'l90hen. '''',lln. Ret. 3. Military Handbook (~IL-HDBK-5)
Aug. 1962.
Ref. 4. NACA Technical Note 3462.
Fig. Bl.27 - Stress-strain curves, SAE 6140. drawn lO20oP. Ret. 5. NACA Technical Note 868.
Ref. 6. From Structures Manual, Convair
Astronautics.
 CHAPTER B2
MECHANICAL AND PHYSICAL PROPERTIES OF METALUC
MATERIALS FOR FUGHT VEHICLE STRUCTURES

General Explanation. It would require several hundred pages to list the properties or the many
materials used in flight vehicle structur~l design. The metallic materials presented in this
chapter are those most widely used and should be SUfficient for the use of the student in his
structural analysis and design problems. All Tables and Charts in this chapter are taken tram
the government publication '~111tary Handbook, MIL-HDBK-5, August, 1962. Metallic Materials and
Elements for Flight Vehicle Structures~. This publication 1s for sale by the Supt. or Public
Doc~~entsJ Washington 25, D.C. The properties given in the various tables are for a static
loading condition under room temperature. The effect of temperature upon the machan1cal
properties 1s given in the various graphs.

AISI ALLOY STEELS

Table 82. 1 (AIBI) Alloy Steels

I
Alloy .... AISI 4130. 8630, I AISI 4130. 4140. 4340. i
4140 AISI
and 8735 8630,8735, and 8740 4340 4340
8740
I I
Form ....... ........... . Sheet, strip, All wrought forms
plate. tubing
I
Condition .. .. . N Heat treated (quenched and tempered) to
I obtain F I .. indicated

Thickness or diameter, in.. >0.187

, :<0.187
'j

I
Basis .. ....... . ..... . .... . . ... ... . ..... .. . .. . .. ... ........ . .. .. ... ........I (0)

Mechanical properties:
F h • ksi.. ... 96 90 125 150 180 200 260
F ,,,' ksi. .. .......... 75 70 103 132 I, 163 176 217
F.", ksi .. ... . ..
' 75 70 113 145 179 198 242
F ,.., kai . . . . ...... ... I 55 55 82 95 109 119 149
F h ... ksi:
(eID~1.5) . .. . I .. ...... 194 219 250 272 347
(ejD=2.0) ....
, 140 140 251 287 326 355 440
.!
Ph". ksi
(e(D=1.5) . .. . i . ... 151 189 230 255 312
(eID~2.0) . .. ..
,
... I, 180 218 I 256 280 346
.,,I
, i
" ,
e, percent. ~ L 10
See table See table 2.3.1.l(c)
2.3.1.l(b) •T 3

E, 103 psi. .
i---'---------'------
i 29.0
E., 103 psi. 29.0
,
G. 10e psi ..... -! 11.0
--------1-------------------
Physical properties: 1

oM, Ibiin.3. 0.283.

C, Btu, (lb) IF) .. 0.114 (at 32° F).
K, Btul[(hr) (W) (F),'ftj. 22.0 (at 32° F).
""' lO-~ in ..·in./F. 9.3 (0° to 200° F).

82.1

: :t ..
-"'.. - ~ '....J
B2.2 MECHANICAL AND PHYSICAL PROPERTIES OF METALLIC MATERIALS FOR FLIGHT VEHICLE STRUCTURES
A:SI ~~LOY S~2SLS (Cc~t.)

'.,~~~=._._._.

,~140
-
e
~
E
,
Slrenq1h Of temperature
E~posure up to
' --'

1/2 hr.
o
~120

E
o
o
cr ,

-0 \00 , ,., 4QO 600 ecc 1000 '200 1400 ,600

" r."'(l.rof~"',F

'"
-e Fig. 82.2. Effect of temperature on the compressive
•
o , E. yield strength (Fey) of heat-treated AISI alloy steels.
'" 80 roo

•I ..
e

oe 60
,

FiT Ft. I .
~
0

.'
..
eo
,
!
40 ,
,

, ,
,
,
" 600 ecc 1000
Temperal"llI, F
IlOO I-lOO 600

, , Fig. 82.3. Effect of temperature on the ultimate

shear strength (Fsu) of heat-treated AISI alloy steels.
20 ,
,

·, .cc

o -400
o
,

400 800 1200

,i
F Temperature 1
~
a
Fig. 82. 1. Effect of temperature on Ftuo Fty. and E of
AISI alloy steel ·•
j

~
l
$00 soc 1000 ;lOO 1400 1600
T_ral......, F

Fig. 82.4. Effect of temperature on the ultimate

bearing strength (FOnJ of heat-treated AISI alloy steels.

i. --=-=-..o.
~
0 . ~==SE=~~~l~~~~~
._;-;;:.:.~=

,J
ec
"
" '0 400 600 ecc '000

r''''p''QfUI'e, F
'200 1400 1600

Fig. B2. 5. Effect of temperature on the bearing

yield strength (F'bry) of neat- treated AISI alloy steels.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES B2.3

5 Cr-!'lo-V AIRCRA?!' STEEL

Table B2. 2 Design Mechanical and Physical Properties of 5Cr- Mo- V Aircraft Steel

Alloy ...... ... . ...... . 5Cr-Mo-V aircraft steel.

Form ........... · . .... ... All wrought forms.
Condition ...... ... Heat treated to obtain the F II. indicated. Fig. 92.8 Effect of temperature on
Section size .. .. .. ..... Up to 12 in. diam, or equivalent. the tensile and. compressive modulus
(E and Eel of 5 Cr- Mo- V aircraft
steels.
Basis ....... , .. ... . . ... ... (0) (0) (0)

I
Mechanical properties: I
F h • ksi ...... ........ 240 260 280
I
FIll. ksi .... ..... ......I 200 220 240
F. II ! ksi ..... · . . . . . . . . I 220 240 280
,
F .... ksi ...... · . . .... 145 155 170
F~ .... ksi:
(e/D-1.5) ..
(eID~2.0) . . . ... .
F ••II , ksi:
(e/D-1.5) ..
... .... .. .... .
400

. ......... ... .
.. . ...... . ... .
435

.. . . ...... ... .
....

..
. . ....
465

. . .....
-,
Fig. 32. 9 Effect 01 temperature on
(e/D-2.0) . . . 315 340 365 the compressive yield strength (Fey)
01 5 Cr-Mo- V aircraft steels.
e, percent:
Bar, in 4D ... .... . 9 8 7
Sheet, in 2 in. Cb} . 5 5 4 !
Sheet. in 1 in ...... 8 7 5 i.
1
E. 10' psi .. .. . ...... 30.0 •.;
E •• 10 8 psi .. ' 30.0
G, 10' psi. . . .... 11.0

Physical properties:
«I.Ibjin. 3 . . . . .
I 0.281. Fig. B2. 10 Effect of temperature on
C, Btul(lb)(F) . 0.11(') (32" Fl. the ultimate shear strength (Fsa) of.
K, Btul[(hr) (It') (F)lltl. 16.6 (400° to 1,100" F). 5 Cr-Mo-V aircraft steels.
a, 10-~ in./in./F ..... 7.1 (80 0 to 800 0 F); 7.4 (80 0 - 1 . 2 0 0 0 F).

• Minimum properties e:q:le1:ted when heat created as recceimended in

section 2..U .0.
II
~ For sheet thickne~ greater than 0.0li0ineb.
• Calcn.lated value.
•
.1
!
l
-,
Ftg. H2. 11 Effect of. temperature on
the ultimate bearing strength (Fbrn)
01 5 Cr- Mo- V aircraft steels.

j
I
•
.'

-,
Fig. 92. 6 Effect of temperature on
-,
Fig. 92.7 Effect 01 temperature on Fig. H2. 12 EUeet of temperature on
the bearing yield strength lFbry} of 5
the ultimate tensile strength (Ftu) of the tensile yield strength {Fty} of. 5
cr- Mo- V aircraft steels. Cr- Mo- V aircraft steels. ce- Mo- V aircraft steels.
82.4 MECHANICAL AND PHYSICAL PROPERTIES OF METALLIC MATERL\LS FOR FLIGHT VEHICLE STRUCTURES

17-7 PH STAI~1.~S STSEL

Table B2. 3 Design Mechanical and Physical Properties of 17-1 PH Stainless Steel

Alloy. 17-7 PH

Form .. Sheet, strip, and plate(") Bars and Forgings(b)

Condition. TH 1050 RH 950 TH 1050 ! RH 950

Thickness or diameter, in .. 0.005 to

1 0.005 to 0.1875 to 6 and under
0.500 0.1874 0.500
(---
Basis . s S (,)

Mechanical properties-
Ft•• ksi . 180 210 200 170 i 185
F,•• ksi . 150 190 180 140 ! 150
F ••• ksi . 158 200 189 147 ' 158
F ••, ksi. . 117 136 130 111 120
F~ ••• ksi:
(eID-1.5) . 2'97 346 380 280 305
(e/D~2.0) . 360 420 400 340 370
F~.." ksi:
(eID-1.5) . 225 I 285 270 210 225
(elD~2.0) . 247 I 313 297 231 247
8, percent .... See table 2.7.2.1(b) 6 5

E, 10' psi. . 29.0

E •• 100 psi . 30.0
G, 10° psi. . 11.0

Physical properties:
tal. lb/in.~ . 0.276.
C, Btu/Clh) (F) . 0.11. (~)
K, Btullihr) (ft')(FJ/flj. 9.75 (at 300 0 F).
a, 10-4 in.lin./F. 6.3 (70 0 to 600 e F) for TH 1050.
6.8 (70 0 to 600 0 F) for RH 950 .

• Test d.1reetJon longitudina.l for widths less than 9 tn.: transverse for short transverse (tbtcknesal direction.
widths 9 in. and over. • Vendors guaranteed minimums for F,., Fro, and e.
• Test dlneUon longitudinal; these properties ncr eppucabre to the • Calculated value.

I
j
J
J
•
~
•

Fig. B2. 13. Effect of temperature on the ultimate Fig. B2. 14. Effect of temperature on the tensile yield
tensile strength (Fm) of 17-7 PH (THI050) stainless steel. strength (Ftyl of 17-7 PH (TH1050) stainless sreet.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES 82.5

17-7 PH STAINLESS STE~L (Cant.)

roc
''''

.....
.~
! ..
l

Fig. 82. 15. Effect 01 temperature on the compresl!Iive Fig. 82. 18. Effect 01 temperature on the ultimate
yield strength (Fey) 01 17-7 PH (THI050) stainless steel. bearing strength {FbruJ of 17-7 PH (THI050) stainless steel.

IJ
•
.:

Fig. B2. 16. Effect of temperature on the ultimate Fig. B2. 19. EHed of temperature on the bearing yield
shear strength (Fsn) of 17-7 PH (THI050) stainless steel. atrength (i'bry) 0117-7 PH (THI050) stainless steel.

IJ
•,J
1
•
! e &CO 100 IOOCI 1200 l..co I.cG
l
-.'
Fig. 82. 20. EUect of temperature on the tensile and
compressive modulus (E and Ed of 17-7 PH (THI0SO)
stainless steel.
 •
82.5 MECHANICAL AND PHYSICAL PROPERTIES OF METALLIC MATERIALS FOR FLIGHT VEIITCLE STRUCTURES

17-4 PH STAINL~S S?S2L

Table 82.4 Design Mechanical and Prryaical Properties of 17-4 PH Stainless Steel

,
Alloy, . .... 17--4 PH
i
l
Form .. Plate Bars and forf;ings(b)
Fig. 82.23 Effect of temperature on
Condition .. ... H 900 H 900 the compressive yield strength (Fey)
of 17-4 PH (H900) stainless steer.

Thickness or diameter, in.... 8 and under

i
Basis .............. ...... S ('J S !
I
Meehanical properties:
F, •• ksi ... ... . .. . .... 190 190 i
F ,... ksi .......... .... 170 170 l
F~ •• ksi ........ .... 178 178
F••• ksi ........... ... 123 123
F~n. ksi: Fig. 82.24 Effect of temperature on
313 313 the ultimate shear strength (F au) of
e'/D-1.5)··.···· . 17-4 PH (HSOO) stainless steel.
e'/D-2.0) ........ 380 380
F~'''J kei,

II
e'/D-1.5) ...... 255 255
e'/D=2.0) ...... 280 280
e, percent:
In 2 in............ 10
,
104.D .......... ... 10
)
l
l
E, 10 psi ..
8 29.0
E •• 10' psi . 30.0
G, 10' psi . 11.0 Fig. 82.25 Effect of temperature on
the ultimate bearing strength (FOru)
of 17-4 PH (H900) stainless steel.
PhysioJ properties:
w, lb/in.3 .•.. 0.282.
C, Btu/(1b)(F) 0.11 (32" to 212" F).
K, Btu/[(hr)(ft') (FJlrtl. 10.3 (at 300" F); 11.2 (at 500" F); 13.1 (at 900" F). iI
a, 10- 1 in. lin. IF. 6.0 (70" to 200" F); 6.1 (70 0 to 400 0 F); 6.5 (70 0 to
900 0 F).
,
"
;
• Vendon guaranteed minimums (or F, •• F, .. and (. l
• Test direetion longitud.in.a.l; these properties not apptlceble to the shon transverse (thiclme5S) direction.

Fig. 82.26 Effect of temperature on

the bearing yield strength (Fbry) of
17-4 PH (H900) stainless steel.

iI
"
l
l

Fig. B2.21 Effect of temperature on Fig. 82.22 Effect of temperature on Fig. 82. 27 Effect of temperature on
the ultimate tensile strength (Ftu) of the tensile yield strength (Fty) of the tensile and compressive modulus
17-4 PH (H900) stainless steel. 17-4 PH (WOO) stainless steel. (E and E c > of 17-4 PH (H900)
stainless steel.
 A::fALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES 82.7

~~-360 S7AINL~S STSSL

Table 82.5 Design Mechanical and Physical Properties of AM-350 Stainless Steel

Alloy, AM-350

Form .. Sheet and strip" Sheet and strip"

Condition .. DA seT
Thickness, in ... 0.187 and under 0.187 and under

Basis. s s
Fig. 82.30 Effect of temperature on
the compressive yield strength. (F cyl of
Mechanical properties: AM~350 stainless steel (double-aged).
F,~, ksi . 165 185
F!I/, ksi . 135 150
F CI/' ksi . 142 158
F.~, ksi. . 107 120
F"r~. ksi:
(eiD = 1.5) .
(eiD=2.0) ...
272
330
305
370
j
F"rl/' ksi: I
(e/D=1.5) . 202 225
-s
(eiD=2.0) , 223 247 I
e; percent. 10 10
!
l
E, 10 psi.
6 29.0
Fig. 82.31 Effeet of temperature on
E" 10 5 psi., 30.0 the tensile and compressive modulus
G, 106 psi. 11.0 (E and Ed of AM~350 stainless steel
(double-aged).

Physical properties:
w, lb/in. 3 . 0.282.
C, Btu!(lb) (F) . 0.12 (32 0 to 212 0 F).
K, BtUl[(hrj(W) (FHtj .. 8.4 (at 100 0 F); 11.7 (at 800 0 F). j
a, 1O-~ in.iin./F . 6.3 (70 0 to 212" F); 7.2 (70" to 932 0 F).
!
I
• Test direction longitudinal for widths less than 9 m.; transverse for widths 9 in. and over. •
.'
!
l

e'tg , 82.32 Effect of temperature on

the ultimate tensile strength (Ftu) of
AM-350 stainless steel (SCT).

j
!
l

Fig. 82.28 Effect of temperature on Fig. 82.29 Effect of temperature on ~ 19. 82.33 Effect of temperature on
the ultimate tensile strength (Ftu) of the tensile yield strength (Fty) of the tensile yield strength (Fty) of
AM-350 stainless steel (double-aged). A11-350 stainless steel (double-aged). AM-350 stainless steel (SeT).

'.~ ~-,1 c-:

:'. . :- t.
B2.8 MECHANICAL AND PHYSICAL PROPERTIES OF METALLIC MATERIALS FOR FLIGHT VEHICLE STRUCTURES

AM-350 STAINLZSS S!EEL (COilt.)

I
!
•~ ""
J •~ '"
J J '"
•.r ;
'"
1 "'
• ~
! l e
l e = sec ecc
"""F
"'"TtrroerQrun, "00 «cc ,"00

Fig. B2.34 Effect of temperature on the tensile and Fig. B2.36 Effect of temperature on the compressive
compressive modulus (E and Ed of AM-350 etatnless yield strength (F cy) of AM-350 stainless steel (SCT).
steel (SeT).

!J j
;
J ;
E J
•
! ~
l l

Fig. B2.35 Effect of temperature on the bearing FIg. B2. 37 Effect of temperature on the ultimate
yield strength (Fbry) of AM·350 stainless steel bearing strength (Fbru) of AM-350 stainless steel
(SCT). (SCT).

FIg. B2.38 Effect of temperature on the ultimate

shear strength (F su) of AM-350 stainless steel
(SCT). .
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES B2.9

AISI 301 STAINLESS STEEL

Table 82.6 Design Mechanical and Physical Properties of AISI 301 Stainless Steel

Alloy ....... .... . . .. . .... AISI 301 •

Form ........... .. ... Plate ~, sheet, and strip

,
Condition ..... .. . . ...... Annealed h -hard H hard ~ hard Full hard
I I I
Basis ........ ..... . . .... S S S S S

Mechanical properties:
• F I ,.. ksi:
L .... ....... .. . 75 125 150 175 185
T ... . . ...... . ... 75 125 150 175 185
Fill' ksi:
L ...... · . ...... .. 30 75 110 135 140
T ... ..... ....... 30 75 110 135 140
F.", ksi:
L ........... ... . 35 43 58 76 85
T .. . . .... . . .... 35 80 118 160 179
F .., kai . .. .. . ..... 40 67.5 80 95 100
F br . , ksi:
('/D-1.5) ...... ........ .. . ....... ......... .. . . ...... . . . .......
(,/D=2.0) . . ' . .. . 150 250 300 350 370
F~.". ksi:
(,/D=1.5) .. . . ... .. . ......... ...... ....... . .. . . . . . .. .. .
(,/D-2.0) . .. . 50 140 200 240 270
e, percent ... ...... . . ( .) (.) ( .) (.) ( .)
E, lOG psi:
L .... ... . . .... 29.0 27.0 26.0 26.0 26.0
T ..... . .. 29.0 28.0 28.0 28.0 28.0
E., lOG psi:
L .... .. . ......... 28.0 26.0 26.0 26.0 26.0
T .. ....... . ..... 28.0 27.0 27.0 27.0 27.0
G. 10' psi. .. ...... ... 12.5 12.0 11.5 11. 0 11.0

Physical properties:
Cot, Ib/in.3.... 0.286.
C. Btu/(lb) (F) . .. . 0.108 (at 32' F) .
K, Btu/[(h.) (ft')(F)/ft). 7.74 (at 32' F).
a, 10- 6 in./in./F ... 9.2 (700 to 2000 F).

_ Properties rcr annealed condition allIo applicable to anneated AlSI

302,303.304.321. and :W-,
• Only annealed condition applicable to plate.
• ,see table 2.8.1.1 (b).

:-!OTE.-Yield strength, partiCUlarly in compn!5IIion. and modulus of

el.aaticity in the longitudinal direction llIay be raised appreciably by
thermal stres;s-relieving treatment In the range 500- to 800" F.
" - -' -. - ' - -' '-. - ' '. - _- ", , -;, -~, r~f;1
-~&J

82.10 MECHANICAL AND PHYSICAL PROPERTIES OF METALLIC MATERrALS FOR FLIGHT VEHICLE STRUCTURES

; AISI 301 S7AIN~SSS STEEL (Cout.;

j
i

,J
Fig. 82.39 Effect of temperature on
ii Fig. 82.48 Effect of temperature on
the ultimate tensile strength (Ftu) of the compressive yield strength (Fey)
AISI 301 (half-hard) stainless steel. .' of AISI 301 (full-hard) stainless steel.
J
l

!i Fig. 82.44 Effect of temperature on

!I
the bearing yield strength (Fbry) of
AISI 301 (half-hard) stainless steel. •
.'
,
J
1
!

Fig. 82.40 Effect of temperature on

the tensile yield strength (Fty) of
j Fig. 82.49 Effect of temperature on
the ultimate shear strength (F su) of
AISI 301 (half-hard) stainless steel.
i AISI 301 (full-hard) stainless steel.

.
J
J
1
J
J,
Fig. 82.45 Effect of temperature on J
the tensile and compressive modulus
(E and Ed of AISI 301 (half-hard)
stainless steel.

Fig. 82.41 Effect of temperature on Fig. 82. 50 Effect of temperature on

the compressive yield strength (Fey) the ultimate bearing strength (Fbru)
of AmI 301 (half-hard) stainless steel. of AISI 301 (full-hard) stainless steel.

ii
ii •
~
Ii
s• ,J s
,J ,!
Fig. 82.46 Effect of temperature on
the ultimate tensile strength (F tu ) of
Fig. 82.42 Effect of temperature on AISI 301 (full-hard) stainless steel. Fig. 82. 51 Effect of temperature on
the ultimate shear strength (Fsu) of the bearing yield strength (F'bry) of
AISI 301 (half-hard) stainless steel. AISI 301 (full-hard) statntess steel.

ii
•
J
,
J

Fig. 82.43 Effect of temperature on Fig. 82.47 Effect of temperature on Fig. 82. 52 Effect of temperature on
the ultimate bearing. strength (Fbru) the tensile yield strength (Fty) of AISI the tensile and compressive modulus
of AISI 301 (half-hard) stainless steel. 301 (full-hard) stainless steel. (E and Ed of AISI 301 (full-hard)
stainless steel.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES 82. 11

2014 ALUMINUM ALLOYS (SHEST & PLAT~J SXTRUSIONS, fORGINGS)

NOTE: Values in (A) colwmns are nin:~un guaranteed values. Values in (B)
col~ will be met or exceede~ by 90 percent of material supplied.

Table 82.7 Design Mechanical and Physical Properties of 2014 Aluminum Alloy (Sheet and Plate)

Alloy. 2014

Form. Sheet and plate

Condition. -T6l1

1.501- 2.001- 3.001-/1

Thickness, in. 0.02<HJ.039 0.040-0.499 0.500-1.000 1.001-1.500 2.000
I 3.000 4.000

Basis. A B A B A B A B A A A
I - --- - - - --- - - - - - -
Mechanical properties:
Flit, ksi
L. 65 67 68 70 68 70 67 68 65 63 59
T 64 66 67 69 67 69 67 6B 65 63 59
ST . 58 54
FIll, ksi
L. 58 60 60 62 60 62 59 62 59 .'57 55
T. 57 59 59 61 59 61 59 62 59 57 55
ST 53 51
Feu, ksi
D. 60 62 60 62 61 64 61 59 57
T. 61 63 61 63 61 64 61 59 57
ST 59 57
F, ..,ksi. 41 42 41 42 41 41 40 39 37
Fbr.., ksi c
(../])=1.5). 102 105 102 105 101 102 98 89 88
(e/D=2.0) 129 133 129 133 127 129 124 129 112
Fbrr/' f.:lliC
(r/D=1.5) 84 87 84 87 83 87 83 80 ii
(el D =2.0) 96 99 96 99 94 99 94 91 88
e, percent
L.
7' 6
8
8
6
6
..
,6 "3 4
2
3
1.5
ST 1
E, 10· psi. 10.5
E c, 10' psi. 10.7
G, 10. psi -.l.()

Physical properties:
1oI,lb/in. l 0.101
C, Btu!Ob)(FI 0.23 (at 212°F)
K, Btu/[lh,)(ft')(FI!ftj. !)O (at i7°F)
<J:, 10-- in./in./F 12.5 (68° to 212°F)
B2.12 MECHANICAL AND PHYSICAL PROPERTIES OF METALLIC MATERIALS FOR FLIGHT VElllCLE STRUCTURES

2014- ALt.n'1I.\'tl1 ALLOY:::; (SHEET & PL.;\TE. ~TRUSIOIlS. ?ORGINGS) (cont.)

Table 82.8 Design Mechanical and Physical Properties of Clad 2014 Aluminum Alloy (Clad Sheet and Plate)

Alloy. Clad 2014

Form ... Sheet and plate

Condition. ... .. ... _T6 cJ

Thickness, in. .. . .... . ........ <0.039 0.040-0.499 0.500-1.000 1.001-1.500

Bui8.. ............. . ....... A B A B A B A B

Mechanical prcperdee:
PtMo Jt;.n
L .. ..... .. .. ......... 64 64 65 67 65 67 64 66
T ....... ............. 63 63 64 66 64 66 M 65
ST. . . ... .. . ...... ... .. .. ..
F trh /en
L .. ... .. . ... ...... . ... 56 56 58 60 58 60 57 60
T .. .... . .. ........... 55 55 57 59 57 59 57 60 .'
ST .. .. . .. . .. . ........ .. .. ..
p~ k.ri
L .. ........... ...... . 56 56 58 60 58 60 59 62
T .. .. .. ........ .. . . .. 57 57 59 61 59 61 59 62
ST .... ... .... .. . . ... ..
Ft. .. len .. .. . ...... . ... 39 39 39 40 39 40 39 40
P"".,/eno
(./D-1.5) .. 96 96 98 101 98 101 96 98
(./D-2.0). .... 122 122 124 127 124 127 122 124
p.........
(./D-l.5). ... .. ... 78 78 81 84 81 84 80 84
(./D-2.0). ......... ... 90 90 93 96 93 96 91 96
e, percent
L .... . .. . ............. 7 8 6 6
T ....... .. .......... . 7 8 6 4
E, 10' pn.... .. . ... . ..... .
10.5
Ee, 10' pai. ... ....... .. ....
10.7
G, 10' "... .... . ..... . .. . ...
4.0
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES B2.13

201~ ALUMINUM ALLOYS (SHZET & PLATE, EX1RUSIONS, FORGINGS) (Cant.)

Table B2. 9 Design Mechanical and Phystcal Properties of 2014 Aluminum Alloy (Extrusions)

Alloy. 2014

Form. Extruded rod, bar, and shapes

Heat u-ted
andapd
by"'"

Condition. -T6e -T82

Ceose-secncnel area, in.s <25 >25, <32 Up to 32

Thickness, in. 1I ....... 0.125-0.499 0.500-0.749 0.750-1.499 1.500--2.999 3.llOO--l.499 ;;::0.750 0.125"'-'99

Basis .. ......... A B A B A B A B A B A A

Mechanical properties:
F tvo kri
L. 60 61 64 68 68 73 68 73 68 73 68 eo
T ................ eo 61 64 67 63 66 61 63 58 61 se se
F'I1I, kn
L. ...... 53 57 58 62 eo 6.5 eo 6.5 eo 65 58 53
T .. ... ... . .. . 53 57 55 59 54 58 52 55 49 53 47 47
PCfI1lui
L. . ........... ...... 55 59 eo 64 62 67 62 67 62 67 .. 53
T. 53 56 58' 62 57 61 57 61 57 61 .. 48
F N. kn.. ......... 35 35 37 39 39 42 39 42 39 42 39 35
FM, hi
(,ID-1.5). 90 92 96 102 88 95 88 95 88 95 .. ..
(,ID-2.0). 114 116 122 129 109 117 109 117 109 117 ..
P/wv.Jcn
(,ID-L5). ....... 74 80 81 87 78 85 78 85 78 85
(,ID-2,O) .. .... . 85 91 93 99 84 91 84 91 84 91 ..
I, percent
L .. ... .. .... 7 7 .. 7 .. 7 .. 7 .. 6 7
Td .. .. ... . ... . 5 5 2 2 .. 1 1
E, 10- ]Mi.. 10.5
E e, 10- pn . . ..... 10.7
G, 10- pn.. 4.0

., '-I ,~

" ..;.. \
-,.,..-, ~
B2.14 MECHANICAL AND PHYSICAL PROPERTIES OF METALLIC MATERIALS FOR FLIGHT VElfiCLE STRUCTURES

2014- AUNINUM ALLOYS (SHE:S7 & PU.TS, EXTRtTSIOl-JS J FC?Gn;oS) (Cont • )

Table 82.10 Design Mechanical and Physical Properties of 2014 Aluminum Alloy (Forgings)

Alloy 2014

Die Hand forgings I Hand forg:ngs

Korm. forgings Length <:3 times width Length> 3 times width
, I
Condition. -T4
I -T6

Thickness, ill .... .... <4 inches <6 inches

>16, >36, >144, >16, >36, > 144,

Ceoee-eeeuonal a.rea., in. l <:16 <:36 <:144 <:256 <16 <:36 <144 <:256

ilasis.. ........ A A A A A A A A A A
---
Mechanical properties:
Ftv, k&i
L. ..... .... . . .. 55 65 65 65 62 60 65 65 62 60
T . .. ... ..... .... 52 62 63 6. 59 58 63 63 59 58
ST ..... 60 60 56 55 60 60 56 55
FIV,bi.
L. .... .. . ....... . . .. 30 55 55 53 53 "0
0- 55 53 53 52
T. ....... 28 52 55 53 52 50 55 53 52 50
S7' .. . . .. .. .... 55 53 52 50 55 53 52 50
F CVo in
L. ...... .. 30 55 55 53 53 .')2 55 53 .j3 52
T•. .. ..... .... 28 52 55 53 52 50 55 53
ST ... ..
[l ft> m .. .... 34 39 40 40 38 37 40 4Q
F",.,.,k.ri
(t:/D-1.5). .... 91 91 87 84 91 91 87 84
(,ID-2.0). 117 117 112 108 117 117 112 108
F'In-v,1csi.
(,ID-1.5). 77 74 74 73 77 74 74 73
(t:/D-2.0). 88 85 85 83 88 85 ss 8.
e, percent
L. .... . ....... 11 7 10 9 7 5 10 9 7 5
T. ............ 3 6 5 3 2 4 3 2.5 .,
ST. 3 2 1 1 2 2 1 1
E, 10' ]Mi. 10.5
E e• 10' psi. 10.7
G, 10' psi .. 4.0
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES B2.15

EFFECT OF TE:i'!PER.~TURE ON 2014 ALUMINUM ALLOYS

"'J

"'0
-
~ -
- ~
- ~
-
,ec - - -
- ~
~

- ~--
- STUtC.'t
(J.'osvu
n

"
TC"~(.UII.[
r e '0 000 .. .-
- - -
- ---
- -- -- -- -. -
- ~ ~

-- -
--~
0:=
--
- -

...,...... -=-==-
- - - - --
'/2
'0
'00 - --
- -
'000
'0 000
...-

--
-- - -
- -
- - --
- -- - -
'"
-,==:== --- - -
~=
-- -- - - ---
o '''00 -lOO 0 200 )00
-'00 '00 JOO '00 500 lO
T04"E:FlATtRE. .".

Fig. B2.53 Effect of temperature on the ultimate strength (Ftu) of 2014·T6 aluminum alloy (bare and clad sheet 0.020-0.039
in. thick; bare and clad plate 1. 501-4. 000 in, thick; rolled bar, rod and shapes; hand and die forgings; extruded bar, rod and
shapes O. 125-0.749 in, thick with cross-sectional area ~ 25 sq. in.).

,'" ,
-, -.
-'':L -~
- '.......--..
: -'-'-t+.....-
':1. __ ~~

I - ."-"._;"
~_~_tE:'
,,:~- - 'J ,.,

'''' '-- -
•.
-+-l-~.; Snuu"
:-7 AI H .. ~[tUvU
,,.... El'OSIII[ v~ 10 10 000 u

'00

80

=
-..;::: "
,
- --'/2'0 .., ......
zw - tc '000 ..
--~.,
'00
V 40 000

•
" -
~

.--..., - ,.--
,---'
-
I ·:.::2E?~-
:~ :-- ~-'-.,~~: rr;
-"-
-
~

--- -
- c= t=;:-=
-lOO -200 -'00 o 200 JOO '00 500 lO 700
•
Fig. 82.54 Effect of temperature on the tensile yield strength (Fty) of 2014-T6 aluminum alloy (bare and clad plate
3.001-4. 000 in. thick; rolled bar, rod and shapes; hand and die forgings; extruded bar, rod and shapes 0.125-0.499
in. thick with cross-sectional area a 25 sq. in.).
B2. HI MECHANICAL AND PHYSICAL PROPERTIES OF METALLIC MATERIALS FOR FLIGHT VEHICLE STRUCTURES

EFFECT OF TEMPERATtTRE ON 2014 ALFMINlR1 ALLOYS (Cont.)

j
I
.'
,!
Fig. 82.55 Effect of temperature on
the compressive yield strength (Fey)
of 2014-T6 aluminum alloy (all
iI Fig. 82. 58 Effect of temperature on
the bearing yield strength (F'bry) of
2014-T6 aluminum alloy (all products
products except thick extrusions). except thick e..x trastcns).
J
!
,r"
I

Il .. Fig. 8.2.57 Effect of temperature on

the ultimate bearing strength (Fbru)
J
of 2014·T6 aluminum alloy (all
products except thick extrusions).
! ,!
I
---_-,.... _""'-
Fig. 82. 56 Effect of temperature on Fig. B2. 59 Effect of temperature on
the ultimate ebeer strength (Fsu ) of the tensile and compressive modulus
2014-T6 aluminum alloy (all products (E and Ed of 2014 and 2017 aluminum
except tbick extrusions. alloys.

2024 ALUMlNUl1. ALLOY (BARE SHEE:I' & PLATE, EXTRUSIONS, BAR, ROD & '"lIRE)

Table B2. 11 Deatgn Mechanical and PhySical Properties of 2024 Aluminum Alloy (Sheet and Plate)

Alloy ..... 2024

Fol"ID ••. ....... Sheet end pl.te Coiled. !heet
Heat t~t.ld
Condition. Heat treated by uaer- Heat treated- ..no rolled HelOttree.ted
-T42 -1'3 T< -T36 _T<

ThieJmes (in.}.
0 : = 0.501 1.001- 2.(lOI- 0.010- 0.250- 0.501- I 1.001-1 2.001-
O.OZQ-O.500 0.012-Q.128
<0.250 0.500 1.000 2.000 3.000
sw... A A
0.249 0.500
"""AT ,.\ B A B ------12...
1.000 2.000 3.000

Meo:h&nica..t properUe8:
A A A
B_~_+.~- _A_I_8_ ------
A 8

lI,., .en:
L .. se ee es ra
62

""
62 60 68
" 63
" eo " 61 70 62 66

.." ... " ..

T 60 ss
62 62 56
" " 66 62 67 63 56 69 71 62 66

.
lily> .q,.'
L. 4() 38 .18 36 36 sr ... ... " ... 46 .. 60 ea 4() 41

I"C'It
T
.en
L.
T.
4()

4()
38

.18
36

38
36

38
38

36 4() 42' 1'

4()

36
..
43

..
4()

38 42
4()

42
4() 52

49

..,
54

51
se
4()

4()
'I

41
. 4() 36 38 36 38 46
" 42 38 43 .7 42 56
... 4() 41

..'" '""" "" " , .

P_Us:. J7 38 J7 36 43 41 36 'I 36 36 J7 4()
136 :
p.... /t;JA.c
(eID_I.5) 03 96 03 00 102 101 102 I 92 96 105 108 03

.
(eID_2.0) 118 122 118 II. 196 129 12. 121 120 129 118 122 133 137 118 128
P~bi·
(eID_l.5).
(eID_Z.O)
pell:ent
56

"
53
61
53
81
53
61
53
er
sa n
19 62 "
14
69
78 ;: I:; eo ee
ea
"
. .
96
88
100
se
"
st
ee
• I
f,

s 10'
T
,.. '" 12 8
'" • '" 12 8 ,_: (d) , i', i"
10.5
Eo, 10' pn'. 10.7
G, 10' pM. '.0
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES B2.17

2024 ALU11IN1JM ALLOY (BARE SHEEr & PLATE, E:XTRUSI ONS, BAR, ROD & '.IRE) (Cant.)

Table B2. 12 Design Mechanical and Physical Properties of 2024 Aluminum Alloy (Extrusions)

Alloy.................... · . 2024

Form........... ........... Ertruded bars, rods, and shapes

Heat Heat
treated treated
Heat treated by user· cold
worked
and aged

Condition ..... .. ........... -T4· -T42 -T81

Thickness, in ... ............ 0:050- 0.251}- 0.501}- 0.750- 1.500- 3.000:;; All thick- :<0.250
0.249 0.499 0.749 1.499 2.999 ness

:;;::2:5, :;;::32
·
Cross-sectional area, in. l .... :<25 =<32
I
Basis ........ ........ ... . .. A B A B A B A B A B A B A A
I-

MechanicaJ. properties:
Fl•. k&i
L ... .... .. . ...... 57 61 60 62 60 62 65 70 70 74 70 74 57 64
T ... 57 61 60 62 60 62 ;8 61 ;4 57 ;0 ;3 SO ..........
F~. k&i
L .. ........... .. 42 47 44. 47 44 47 46 54 ;2 ;4 52 64 38 56
T. ...... 42 46 43 46 42 45 41 44 38 41 36 39 36 ........ ..
F••• k" ·..... . ...
L. .. . ....... 38 41 39 42 39 42 44 ;2 SO ;2 SO' ;2 38
T ....... .. . 38 41 39 42 39 42 42 48 42 44 42 44 38 ..........
F••. kJli ... ...... . 30 32 32 33 32 33 34 38 38 40 38 40 30 · .. . . . . . . .
F m. kai
('ID~U) . ... ... .. 85 91 85 91 85 91 85 91 85 91 85 91 85 ..........
(e(D=2.0l. .. . 108 114 108 114 108 114 108 114 108 114 108 114 108 .. ........
F..... lcai
(,/D=U). .. . ;9 66 60 66 60 66 61 66 62 66 62 66 53 ..........
(,/D~2.0) . ..... 67 75 69 75 69 75 71 75 73 75 73 75 61 .. ........
e, percent
L .. ....... ... 12 12 12 ... . 10 10 . .. 10 12 ..... .. . ..
To .... .. 6 ... . 6 .. 5 .. . 2 . ... 2 ... . .. ...... · . . .. . . . . .
E, 10- pai ... . . .. ..... 10.5
E e. 10' pai.... 1O.i
G, lOt psi. 4.0
82.18 MECHANICAL AND PHYSICAL PROPERTIES OF METALLIC MATERIALS FOR FLIGHT VEHICLE STRUCTURES

2024 ALUI1INl.J11 ALLOY (3 ..e....p .E SHE2I' "'= ?L{'~S, ~T:\USIC:NS, SAH, ROD & '~'::\.:::) (Corrt . )

Table B2. 13 Design Mechanical and Physical Properties of 2024 Aluminum Alloy (Bar, Rod, and Wire);
Rolled, Drawn or Cold Finished; Rolled Tubing

Alley. 1 mu
I--------~----­
Form .. .. 1 Bar. rod and wire; rolled. drawn or cold- finished Tubing
, - - - - - - - - - - - - - - - - -----,.------
Heat 1 Heat
Condition .. I -T4 or -T351 Heat treated I
treated
by
i treated
cold
user' worked

~36
1 1 :

I -T3 : i,nd aged

Cress-sectional area, in. -T42 -T81
. I ,
I I I 1 ' I ' : i--- ---
Thickness. in .. I Up to 11.001- 12.001-~ 3.001-~ 4.001-: 5.001-! 6.001-:0.018 to 0.5001 0.018-
11.000 ! 2.000 '3.000 14.000 5.000 4 1 6 . 0 0 0 4 1 6.500 a , 0.500

Bas;,,· .. ·········· .. ····l-A-I~-A-!-A-!-A-i-A-I-A-' A ! B i A A

-1'-'--'1'-:-1 I !
-

I '1
: - - 1
'-
Mechanical properties: I ' '1 I
r.; ksi 1

L. 62 i 82 62 62 I 62! 62 1 62 64 70 I 62 68
T ... 61 59 57 55 i 54 521· 1 j.
r.; ksi
Li 40 ~ ~ ~ ~ ~I ~ a' 46 1 40 ' 60
T. '1 40 39 38 37 37 36 . I
i
1

F~II' ksi
L .. .. .. .. "I 32 32 32 32 32 32
1
42 46 40
T... '1' . . . . .I.
F .... ksi.. .i 37 37 37 37 37 37 37 1 39 42 39
r,.•. ksi. I
(e/D-1.5) " . " .. 93 93 93 93 93 93 I
93 105 96
(erD-2.0) " . '1 118 118 118 118 118 118 118 133 122
F br•• ksi ,
1

(eID=1.5). L 56 , 56 56 56 56 56 I 59 , 64 58
(e(D=2.0)...... ' 64 I 64 64 64 64 64 !. 67 ,
74 64
e, percent
10'1 101
1

L. " " 10 10 10 10 10
T. " ". 10 ,
1
8 6 4 2 2
1
1°'1 !

E, lOS psi ... 10.5

E<, IDS psi .. 10.7
G. IDs psi. . 4.0

Physical properties:
J
0.100
w,
C, Ib(in.
Btu/ilb) IF). '.1'.
0.23 (at 212" F)
K. Btul{(hr) (ft 2HF)/ftj 70 (at 77° F) •
a, 10-s in.iin./F. 12.6 (68 0 to 212 0 F)
 Table 82.14 Design Mechanical and Physical Properties of Clad 2024 Aluminum Alloy (Sheet, Plate, and Coiled Sheet)
,

Alloy .... - '"

Clad lI):lt

"·orlll ", , -_ ..... - Slwct llnd Illate Coiled shed 811eet and Illato
-~-
lIellt
lleat trented IIeat trellted • lIellt treated b)" user' lIellt treated lIellt treated, cold worked,llIloJ aged
end fvlled rreared "d
IIl:ed
----- >-
Ooudntcn. -1'3 -'fJ6 -T< -1'42 -T< -TO -Tltt -T'" Z
--

o.Hlo-j 0,1163- o 250- 0, !Jon - I Olll- a 001- 0.010-

OOOJ- O/lW'/I;,()llJ-j 1,250- ~500- IOlJl- 2.001- O,flJIt- "003- 0.010-1°.00:1- <0,003 0.0,»-IO·063-IO,2ro-l >-
'fl<ickness,lo._ ...... .... ....... 0,11fi2 O.:UII 0.400 1.000 ' 2,000- a.on- 0,062
O.2dl 0002 0400 0.&10- 0,400 1.000' 2.000- 3.000' 0.062 0.128 0.062 0.2411 ~0,0li3 0.002 02411 0.49Q O. SOlI- n r-
Width, III. - ... - ... ..... .....
-----
--
- - - - - - - - -, ' - - - - - - - - - -
--- - - - - - - - -
-
<"
-- -- -- -- -- -- <0' r "''"
B " ~

'">-
Oasis .. ----.- A A
" " " "
A A A A A A A A A A A A
-- A - " "
" " -- A
" " " -- " "A A A A
" " - -" - -
A A A
-- -- -'- - - - -
..
-- -- -- -- -- -- --~
--, -- -- N
0

., ... ". ,-" "" "" "'" .,.

M,·eh"llkall·rol~·"II"s N
Z
10'.. ,1.'11
t.
r
f,1 l;~ Ii:! Ii.' m
m
Ill, 1\, Ii'l
s se "
" ." " " as" ae" "
fl,
" " f,~
., I~I li.~ m
., (" 5~ I ea Jij, IJ~'
62
" " "' " "'" ""' "'" 62
" '"
6~
'"
rs rn
rs ;n u
Ii'l
6:1 "
~ "'"
"
.
.~'t iii li~ .,~ li.~ f~\ Illi 1\' li2
" "
10', •• '"i
t. ,,' " "
" m ',- "
,,- '"' ".
~l; S.~
.,
., , '" "" "" " '"'" " " "" " "'" "" ae.~~ ~li 41;
. 36
" ". " '" .."
36 a, st
'" sc" " ea
" 0, eu ee es
" es fll;
H
'"
~ cZ
~

:6 '" " es
ae fll; III
T :1" I~ S2 IJ~ ali ati a,

. " '"'" " 0'

'" " "
" " " .." ae
f', •. hj

,'" '" " " " a, '" ae '" " '" as sa "- O
31;

'" a, "3. " '" " :I."'" ,,' "" " r,; '"" " "'" '" "Ii," "
J.
r
a, J:, a,
n
u;
"
Jij,
~'l

ae ali " as
a,
" '" '" " W
C
" '" "a:, " "" "" "' " " " "" '" " I, " "" " '"
ta C-

.,
."
,.·•• ,hj ,
" 3, a; a,
" a,
" J.~
'" '" " " " " a~ 3\1
~
'" " " .
:l~ :I~ 4~

,." " "

,

"· ••• ,bi 4 ."

(e/Dj_U
'" !IJ 11,\
'" "
!'.~ U~I
" "' "' " " " "' '"
1111 WI lUI 101 ll~
" Hl:l" lin" IIf," I 12U" '""' " Ul> Ill) 12,
ll~ JlXI 100 1Il610'J Ul,) ItIl, 11I3 l~ Ul t<
Il~ 131 IJ~ la~1 LJa la~ lal 133

.,
~
11M In I:?
'" .. l'i
(fiD)-2d 12, 1:t1 12, lJl 12ll12~ 116121 112 JIll lOll 110 112 111I iii

.
1I~ l~" 1211<'11 12,~
f'" •. .\.Ii" '" Ii, ., tJ
Cl
", ...'"
l" " '" " '" lUI lU"TU;" """ " "
(f!D)_U ll~
'" " "
" "" ""' "'"'. "" " "" ,II 6. ~:I'" " '" ""
"~:I ", '"'" '" "" " "'" " ", 1:/111
III
,:I "
Iii
"' "
T'
(,'DJ~2.U 102 1Il4
1O~r:~;
,1\ 511

, "
~

,,,-, ,::1 ' , , , ,, a , .,"jr '"

11 ,0 OJ <:
f. 1~'rl"'JlI

'/' , . 4 ...
'" " " "
(t). ,
'" "
,
"J • (oJ
E, lUlllli
"Ii IO~
'"
~

"'''H~I " I
n t<
K,.,·
,-:" lilli'll
lJ.~

I '" I " I 11I.0

I I
os IU.U
I I
os lUll
j" I
\Ilfl 0
H
f:;
'"
...'"
I'rl III"
"
I "' I 1 1 1 1'",I, ,1;:1 "' I "...n'"
Ul
K"" ,-
11..
I 1II,2
I I II.• 11I.2 IIU! !I.' 1Il,2 11... 111.2
l'i
fI, 1I1',,,j tJ
l'h)"~kolIIlIOI'''rlk~'
Il.UWI 0.100 "'"
101,11>11,,'
C,ll1u/(161(IO 1f,2;l ('11 2120
711 (at ijO I")
t') 0.23 (.. t :mQ F)
86 (at 77" F) '"'"
K, 11lu:IChr)(ft '/(F)I
III 10-' III.flll./t'.·. '"
12." (.,~o 10 2120 F) 12.6 (611" to 212" ".J
-
• For the ~lH'~S n'lloved {(·lnl_·r-TaM lI'I'lI<' Ollt}·), nil \'oIlut'$ .tiI" Ill.· -1'4 l"'IlI"'f al.pl)" IO'ltl, Ih.' hlilsllLud, liS II ruunt! h·.st 11..·... ml'n l~ n·'lulr.·" lor lt'stil'lt', The Yllhh'S j!:IYl'lI hHe lor thickm'ljS 0..)11/)
elrel'lloli ul ~' •• wldeh lIIay tN.. som"I\'h,,1 IO\\"'r 'nei, ,uul Im'al"f hay" ~"n DlljUbll·,1 10 r"l'n·....·nt II,,· U~"ruj!:U "fOl,,'Hh's across 111,· ...liull' l.-'ctiou,
." Ileul Irrull',1 hy Ullef" rd,'rs 10 lilt n',l!("riILJ sohlllllll Ill'ut 1""all',1 Ii)" Ih., lllit'f r"jl:ar'!ll'ss of Iho hll'luolinj!: 11l.·I,ItIlI,I'nll.
"flor telll''''f of Ihe rtlJtlerlal. ~ 5,·" Tuhl" a.I.I.I.I .
• fllledft''llUon mllllnmms lor rllI,l nll,l"rlal U.Jij.llllt'lIlhkk lind h"al'I"r lin' lor lIll' j-ur., IlIat.'rllll, ',1'10,-"IolIl:IIlIon vuhlt's 1m'IIlI'lit- .... IUn"'~ Ilrt' r",uu'ln Til!>'" 3.2,3.1l (t'l
.82.20 MECHANICAL AND PHYSICAL PROPERTmS OF METALLIC MATERIALS FOR FLIGHT VEmCLE STRUCTURES

EFFECT OF TE!'PERATURE ON 2024 ALtmINUM ALLOYS

!I I
,
I
~ .•'
! ! •
I I

Fl.g• .82.60 Elfect of temperature on
the ultimate tensile strength (Ftu) of
2024-T3 and 2024-T4 aluminum alloy
(all products except extrusions).
Fig. 82. 64 Effect of exposure at
elevated temperatures on the room-
temperature ultimate tensile strength
(Ftu) of 2024·T3 and 2024-T4 aluminum
a.1loy (all products except thick
extrusions).
Fig. 82.68 Effect of temperature on
the ultimate bearing strength (FOro)
of clad 2024-T3 and clad 2024-T4
aluminum alloy (sheet).

II
!
I

FIg. 82.61 EIfect ot temperature on Fig. 82.65 Effect of temperature on Fig. 82.69 Effect of temperature on
the ultimate tensile strength. (Ftu) of the compressive yield strength (F cy) the bearing yield strength (Fbry) of
2024-T3 and 2024·T4 aluminum alloy of clad 2024-T3 and clad 2024-T4 clad 2024-T3 and etas 2024-T4
(extrusiOns). aluminum alloy (sheet). aluminum alloy (sheet).

!I iI
,•
!
'-....- - 1
__ 100 __
f _ ",,,oODQO'"

Fig. 82. 62 Effect of temperature on
the tensile yield strength (Fty) of
2024·T3 and 2024-T4 aluminum alloy
(all products except extrusions).
'-'
Fig. 82.66 Effect of exposure at Fig. 82.70 Effect of temperature on
elevated temperatures on the room- the tensile and compressive modulus
temperature tensile yield strength (E and Ed of 2024 aluminum alloy.
(Fty) of 2024-T3 and 2024-T4
aluminum alloy (all products except
thick extrusions).

j
!
I

FIg. 82. 63 Effect of temperature on Fig. 82. 67 Effect of temperature on Fi.g. 82.71 Effect of temperature on
the tensile yield strength (Fty) of the ultimate shear strength (F su) of the elongation of 202C-T3 and 2024-T4
2024-T3 and 20Z4-T4 aluminum alloy clad 2024-T3 and clad 2024-T4: alumtnum alloy (all products except
(extrusions). aluminum alloy (sheet). thick extrusions).

;or-_
 Table 82. 15 De.sign Mechallical and Physical proper-t ins of 7075 Aluminum Alloy (Sheet and Plate)

Alloy. ....... . ........... 7075

Form. Sllcet end plate
Condition T6"
Thickul-'I:I8, in. ... 0.015-0.039 0.040-0.249 0.250--0.500 0.501-1.000 1.001-2.000 2.001-2,500 2.501-3.000 3.001 3.500 3.501-4.000 :-
B B B B B 'z
- - - - - - - - -A- -B- -A- - -
B
Basis. ....... A A A A A A A
Mechanical properties:
~-
-- ---- -- :-
r-

., "''"-
/o'jll.' klJi ~
L. 76 78 17 76 17 76 76 82 78 80 73 75 70 72 68 66
T 76 78 17 76 17 76 17 80 17 70 73 75 70 72 68 66 '3en
ST .. 67 69 64 66 64 62 :-
Z
/0"/1' hi >-
1> 66 69 67 70 67 60 66 72 68 7l 62 65 60 62 58 56 t: "
7'
3T
It'e/l, klSi
65 68 66 66 66 68 66 66 66 66 62
61
65
63
60
56
62
60
58
56
56
64 "~
H "
-
t'l
~

C>
I.. 67 70 68 7l 60 7l 66 72 68 7l 65 67 63 65 62 60 >- z
T
37'.
70 73 7l H 69 71 66 72

40I
68
..
71 65
64
67
66
63
62
65
63
62
62
60
60
b
Sl ....
0

r,« ksi
r-; bi"
46 47 46 47 '6 47 47 46 47 43 45 '\ 43 49 30
~ r-
(,jD ~ 1.5). 114 117 116 119 . 108 110 110 115 109 112 102 105 68 10\ 65 6'
[>J

(,/D~2.0)
P,,'1I' kISt"
144 148 146 150 139 142 142 147 140 144 13\ 135 126 130 \22 110 '""
(e/ D = 1.5) 62 67 64 68 87 00 00 64 88 62 81 78 81 75 73
5:
(,/D~2.0) 106 110 107 H2 100 104 104 108 102 106 63
8'
67 00 03 87 8'
'"
rn
e, percent ~
L. .... 7 .. 8 8 .. 6 5 .. 5 .. 5 .. 5 tJ
T 7 8 8 6 .. 4 3 3 3 2
37' . ... .. .. .. .. .. .. .. \ 1 1 '"
"U
E, 10' pili. 10.3 5:-3
e; 10- psi. ........ 10.5 to
0, 10' llsi .... _.. 3.9
Physical properties:
w,lb/in. '. 0.101
C. B'u/(lb)W). 0.23 (at 212°F)
K. B'u/«I,,)(f")(F)/f,1 76 (at 77"F)
II, 10-' in./in 'I". 12.0 (68" to 212"F) ..

• For tIuo .u-. nII_"

&aaper -1'''1, aD
til. p~pLlon ot , ••. .A.ppUcable F •• ...a lu
r.u"Jnu,. (ff&.) Di..",,~ of
u- tor Ua. -Tt
to80_:
Ute
e-pu
''''
.pplF "Ua

(A - ' _ , )
1.001-1.500
1.101 -a 000
"8M T..L.. '.1.1.1.1.
L
L
..
0.15"'1.000 L ..
i
II Rl
I
,
Table B2. 16 Des ign Mechanka.l and Physical Properties of Clad 7075 Aluminum Alloy (Sheet and Plate)
w
w

3:
Alloy. .- ..... 'o"
~
Clad 7075
Form ...
Coudiuou
Sheet anti plate
T6' ,.
n
r-
0.015 0.040-- 0.063- 0.188- 0.250- 0.600- 1.001- 2.501
2.001 3.001 3.501
~
'I'hioknesa, in. . . ..... . ... 0.039 0.062 O.IM7 0.249 0.499 1.000' 2.000' 3.000b
'
2.500~ 3.500' 4.000"
Beeia. A u n n n ---
A A
-- c-- -- - - -- -- - - - -A A n A n A n A n A n A -o
~-
-- -- -- A 0 '"
~
Mechanical properties: -o
Pill, k8i~L c»
70 73 72 74 73 75 75
T. 70 73 72 74 73 75 75
77
77
75
75
77
77
77
75
79
77
76
75
78
77
71
71
73
73
68
68
70
70
66
66
64
64
>-
r' ,.r-o
~

SOl' .. .. 67 69 64 66 64 62 ~
PIli' bi-L. 61 64 63 65 64 66 65
H -o
67 65 67 68 66 69
66
ill
,..r-~ '"'"
60 62 58 60 56
sr
T. 60 63 62 64 63
..
65 64 66 64 66 64 66 64 67 60
61
62
63
58
59
60
60
56
56
"
54
54
/I'CII. bi-·L. 62 65 64 66 65 67 66 68 66 60 66 69 66 69 62 64 61 62 60 58 r- ;'j
T 64 67 66 68 67 69 68 70 66 60 66 69 66 69 62 64 61 62 60 58
0 OJ
sr
1" ..., ksi c . 42 44
.. 64 66 62 63 62 60 "',.rn 0'"
i'fn", ksi--(el D = 1.5) 105 110 lOB
43 44
III
44
110
45
112
45
112
46
116
45
105
46
108
46
108
47
\11
45
\06
46
109
42
09
43
102
40
95
41
08
30
02
38
90
c-,tn 3:'"
(e/D=2.0). 133 139 137 \41 139 142 142 146 135 130 139 142 137 140 128 131 122 \26 119 115
f:> ...,.'"
1<'/)'1/, k:siC-(e/ D = 1.5). 85 90 88 91 92
(elf) =2.0). 08 102 101 104
90
102 106
91
104
94
107
84
98
87
100
86
90
88
102
86 90 78 81 75 78 73 70 "~ r-
r-
e. pcrcent-c-L 7 .. 8 8 8 8 6
99 104 90 93 87 90 84 81 u n
5 5 5 5 en
...OJ~
7' . 7 8 ..
81'
8 8 8 6 4 3 3 3 2
'"
L-'
1 1 1
..
~-J
.

E, 10' psi -Pri.

~

l
Sec. 111.:1 10.3 H1.:l
-u r-
s; 10' lJ8i-l'ri. U.5 !I.8 10.0
~: '"
a, 10" }lsi
::il~C .. 10.5
!1.7
10.5
IIU)
I tJ.5
." '"0
L';

'"r-"
It).:!
Physical properties: -~
n
0
w,lb/ill. 1
C, Btu/(1L)(F)
K,lItu/{(hr)(ft»(l")/ft]
Ie, 10-' in./ill./l<'
..
0.101
0.23 (at 212"1<')
7li (al 77° F)
~

" "q
<:
OJ
12.U (tiB" ttl 212"1")
a 10'01' the ..tr"", reli"ved !.emper -T66J, aU ...alu~ lor the -Tll
II Tb....e valuell except In lhe ST direction have been adju.ted to include the IntlueD"",
"'
o
r-
"I;""ption 01 ,.'~ •. Alllllic.aLle ,..'~. valu"" ar.. q 'ollow.:
Thoi(/,;lI"" (i-n.J l>iu"li..... 01 ted
of the 1 1/2% per .Ide nomInal claddlDlJ thickneae. '"
...'"
0.260-· 0.499
0.600-2.000
2.001-2.600
2.601-8.000
L
I.
I.
L
"'ell (A tl4l"e.,
ea
eo
. "See TaLIe 8.1.1.1.1.

n"
ci
-s
" ""
OJ
'"
 Table B2. 17 Des igu Mechanical ~U1d phystcat pr-oper-ties of 7075 Aluminum Alloy (Extrusions)

----~---------- - - - - - - - - - - - - - ; : ; = - -
7075
Alloy - - - - - - ---------._---------
Furm . Extrusions (rod, harH, /Iud shnpes}
Condltioll
. -_'1~"c:;. ==-----
- -
Cro:;:;-bl~d,i(Jllulllrcu, iu. 2 <20 ._- <20 I
>2t1, <32 . <32 >-
"' -- --0
Z
4,5UO~5.0
ThidlJu~HS ~, in. !J 0,250- O.4ml 0.5000.749 O.750-1.4!l9
- --
1.500-2 .H!)!)
-
3.000-4AWJ 0 >-
...e-
--0
A II A II A II A II A II .. A
---_ II A II en
Habb.
--- .•.. - - ~~- -_._-------- ----._.- - --- - - - _._-_.- ---- - --- ---- ---- - - - - - - 00
".
Meehani!",ll pruplJrtie::;: r- ~
00
p ..., k,;j §
L 78 82 81 85 81 85 81 85 81 85 81 84 78 81 78 81 H
>-
l' 76 78 77 70 73 75 72 74 eo es 62 U4 62 tH UO U2
~ Z
"
F,~, kl>i ".
us
L
1'.
70
U4
7 -l.
U·7
73
UU
77
U8
72
U3
7U
66
72
U2
7U
U4
73
5U
7U
58
71
54
74
5U
70
54
73
5U 53
71
55
r::
0
"'"
00

r.; ksi 7U 72 75 72 75 U8 72
"' ~

o
z
71 75 74 78 73 77 72
1'.
L.
71 75 74 78 73 77 73 75 60 72 U5 U8 !:1
45 45 47 45 4U 43 H 43 H
;J 0
45 47 47 47 c:
/t'n, ksi
F~, .., k,;i
43 45
" en
H
'"
(1:/»= 1.5). 101 JO 7 07 102 U7 102 !J7 102 07 102 07 101 04 07 lH 07 0
Z
en
'"
r-
~

(1:/»=2.0) . 125 13I 130 t3li 130 13U 130 l3l1 130 136 122 12ti 117 125 117 J 25 o
0:
F ~'"' J,;,;j ~I

(t'/J)= 1.5). 01 oU 80 85 70
101
84
IOU
70
101
M
JOI
70 84
IOU
78
02
81
96
77
01
80
U5
75
88
78
02
&J
0 "'
e,
(el J)= 2.~) ,
pl"J'(~I"IIL
U8 104 102 108 106 H
~,

8
''""
0:
L 7 8 7 8 7 8 7 8 7 8 7 8 U U H

4 3 I I I I o
1'. 5 5 - _._,,-- - r-
- - - - - - '-'-'-'-'-'-'-'--'-' ----
B, loa -psi .. 10.3 ~ '"
H~, I (Ja IMi, . 10.5 00
on
--(}!. _!~~~~--'-'-''-'-''-'-':.'-'.'-- ~- - - - - - - - - -
Phyaieul Ilr()II.~rlil"::;:
----
3.9 0
"
"''c:"
0,101 o
w, Ih/iu,3 .. s:
'" "''"
H
0, HLlI/(Ib)(J<'). 0,23 (lit 212°]<') c:
1\., HI u/I{hr} (fLl) (F)/fLl 76 (lit 77° [") "l
~.~.~~~~_~0)_;'_'-I_F '_'_''_'_''_'_'". .~"--- 12.9 (68° to ZI2°F)
~~~~~~~'---
n1d11~n Ii".) Aria ('f, in.l
_
l'irldlon 0/ 1'1" F..CA oalu~l)
'" 00

• I"ur tI ... stress rcll"""l! 1"rHl'n~ -'j'li.'t1O ..Ill! -'1'6:i1l. ull vuhir~ Jar the -'1'6 temper llflVly, wllh
llo~ """'1'1["10 of V.... I. A1JI,lkllbh~ I.... \'llllll'~ erettstco uetow: l.lillll---2.llllU <'" I. 10
ee
}.,. (A ""lUI,) 3.00lH.tll1l <32 I.
Thitt,u.t (f".) Au.. (,q. h•.) mUd/ono/lnl
10 t.WO-5.j)OO :<32 I. 0'
<u.2lU <20 I.
I .'or eltmslon~ with olllsllllhHIII: leila, Llle load clirrylnlllil>lIlty 0'
such ICl:stilllill he dt'II'OIl!l,C<\
O.25IlU.HU
07W·I,tUll
:<20
<20
J.
I. "
11 011 the bll.!ils ot the 'lro,wrlle:l ot ure llj'I'Wl'rllllo eclurun (:()(resl>ou<lhl~ 10 Ihe Irg Ihlckllt,s.:J
B2.24 MECHANICAL AND PHYSICAL PROPERTIES OF METALLIC MATERIALS FOR FLIGHT VEHICLE STRUCTURES

7075 ALUMINUM ALLOY (EXTRUSIONS, FORGINGS, BAR, ROD, WIRE) (Cont.)

Table B2. 18 Design Mechanical and Physical Properties of 7075 Aluminum Alloy
(Hand Forgings and Die Forgings)

Alloy. .. ... . . . . , . . 7075

Form... . ......... H&.ad-fo~ stock, length Hand-forged stock, length Di.

<3 times width > 3 times width forgings

Condition. ............... _T64 -T5

Thickness, in ....... <3

Croee-eectioD1l &rea, m. l •••

<16 >16, <36 >36, <144 <16 >16, <36 >36, <144

Iluis ... ............. .. . A A A A A A A

Mechanical properties:
Ft»o .bi
L .. . . . . . . .. .. . · . 75 73 71 75 73 71 75
T ........ ....... 75 71 69 73 71 69 71
ST. , ... .. . .... · . 72 68 66 70 68 66
r,., .en:
L .. ....... ..... 64 61 60 63 60 59 es
T ....... . . ..... 63 60 58 61 59 57 62
ST. ... . .. .. ..... 63 60 58 61 59 57
F~Ui
L .. .. . ......... 64 61 60 63 60 59 6.i
T ... . . .. ......
' 63 60 58 61 59 57 58
ST. . . ... . .. .. ..
F... Ui. ........ . ... 45 44 43 45 44 43 45
" .... Iui
(.ID-l.5). ... 97 95 ss 97 95 85
(.ID-2.0) . ...... 13.5 124 114 135 124 114
" .... hi
(.ID-I.S). .... · . 90 79 78 88 78 77
(e/D-2.01. ..... 96 91 90 94 90 88
t, percent
L. ... .. ...... 9
.,7 4
",., 7 , 7
T. ...... .. . 4 2 3 2 3
ST .. .. . ........ 2 2 1 I 2 I
E, 10- pIi .. ....... 10.3
Eo 10- pri . . . .. . . . .. 10.5
G, 10- pn. ....... 3.9
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES B2.25

7075 ALUMINUM ALLOY (SXTRL~IONS, ~ORGINGS, BAR, ROD, ~IRE) (Cant.)

Table B2. 19 Design Mechanical and Phys icnl Properties of 7075 Aluminum Alloy
(Bar, Rod, Wire and Shapes; Rolled, Drawn or Cold-Finished)

Alloy ........... ...... 7075

Form. . . ........... Bar, rod, wire and shapes, rolled. drawn or cold-finished

Condition ..... ...... -T6 or -T651

I
Thickness. in. ... Up to I
L001 I 2.001 3.001
LOOO d I 2.000 d 3.000 d 4.000 d
I

Basis ... .. . ......... A A A A

Mechanical properties:
F l .. , ksi:
L ...... . . .... 77 77 77 77
LT .. . . ... 77 75 72 69
r.; ksi:
L ... .. . ..... 66 66 66 66
LT .......... 66 66 63 60
F••, ksi:
L ............ 64
LT . . . . . . . . . . . ...... ...... .......
64 I
..... ...... ......
64 64
. ... . ..
F.... :':~i. .... ... 46 46 46 46
F hn .. ksi:
(,/D= 1.5) .... 100 100 100 100
(,/D=2.0) .... 123 123 123 123
F~ ..., ksi:
('ID=I.S) .... 86 86 86 86
(,/D=2.0) .... 92 92 92 92
e. percent:
L .. .. .. .. .. . '7 7 7 7
LT . .. .. . .. . . 4 3 2 1

E, 10e psi. .. ... 10.3

•
E., 10 pet . 10.5
G. loe psi . 3.9
82.26 MECHANICAL AND PHYSICAL PROPERTIES OF METALLIC MA.TERIALS FOR FLIGHT VEmCLE STRUCTURES

j
I
Fig. 82.75 Effect of exposure at
elevated temperatures on the r-oom-
temperature tensne yield strength
!I
(Fty) of 7075-T6 aluminum alloy
(all prccucts}. .'
,! • ,!
Fig. B2.72 Effect of temperature on Fig. 82.79 Effect of temperature on
the ultimate tensile strength (Fm) of
7075-T6 aluminum alloy (all products) J the bearing yield strength (Fbryl of
7075-T6 aluminum alloy (all products).
I

,!
j
Fig. 82.76 Effect of temperature on
the compressive yield strength (Fey)
of 7075-T6 aluminum alloy (all products).
,
a
..: .0

,~
,
,-",--.'
"'" 0.00

,f
FIg. B2.73 Effect of temperature on Fig. 82. 80 Effect of temperature on
the tensile yield strength (Fty) of the tensile and compressive modulus
7075~T6 aluminum alloy (all products). (E and Ee) of 7075-T6 aluminum alloy.
I
.•' •
!

Fig. 82.77 Effect of temperature on 1

the ultimate shear strength (F su) of
7075-T6 aluminum alloy (all products). ,
!

Fig. B2.74 Effect of exposure at Ftg. 82. 81 Effect of exposure at

elevated temperatures on the room- elevated temperatures on the elongation
temperature ultimate tensile strength of 7075-T6 aluminum alloy (all products
(Ftu) of 7075-T6 aluminum alloy (all except thick extruatons}.
products). -,

Fig. 82.78 Effect of temperature on

the ultimate bearing strength (Fbru)
of 7075-T6 aluminum alloy (all products).
 Table B2.20 Design Mechanical and Physical Properties of AZ31B Magues tum Alloy (Sheel and Plate)

Alloy . AZ31B

Form .... Sheet und plate

-0 ~H26
Condition ...•.

Thickness (in.) . 0.016- 0.061- 0.250- 0.501- 0.016- 0.250-- 0.375- 0.501- 1.001- 0.250- 0.375- 0.439- 0.501- 0.751- 1.001- 1.501-
0.060 0.249 0.500 2.000 0.249 0.374 0.600 1.000 2.000 0.374 0.438 0.600 0.760 1.000 1.600 2.000
I-~-I-~-I-- --I--,---f---I-- - - - - - - - - - - - - - - - -
,.
Z
A B
BnsiB A B A B
------- -- ---------------- -------------- --- '"
Mecha.nical properties: ''""
~
1",.., kai. ~

L . 32 36 32 36 32 30 39 40 38 37 36 34 39 38 38 37 37 35 35 c
Z
T . ...... 40 39 38 37 36 40 39 39 38 38 36 36
(.~:~;/':!i;
'{,': 18
19 16 19 16 ~ 30 26
''29
24
27
22
26
20
23
27
30
26
29
26
29
26
28
23
26
22
25
21
24
r.; kai. 18 17 16 16 14
L 12 13 12 10 10 24 26 20 16 13 10 22 21
T- .
F kai .. .. . . 17 23 17 23 17 18 26 18 18 . 18 18 18
F_ klli.
(,/D~ 1.6)... 60 62 60 62 60 68 60 66 64 58 66 56 .
(,/D~2.0)........... 60 61 60 61 60 68 70 66 63 ... 68 65 66
1"6"" kai.
(,/D~1.6).....
(,/D~2.0)
e, percent
29
29
30
30
29
29
31
31
27
27
43
43
44
44
38
'38
34
34
. 40
40
39
39
36
36 . .
~

L
T
. ............
.
12 18 12 19 12 10 6 8
8 .... 10
8 8
10 10
6.~
8 8
10
6
8
6
8
6
8
6
8
6
8
6
8
6
8 .'"
c;
o
c:
6.6
2.4 ''""
~

Physlcel properties:
w. Ib/in.' . 0.0639
C. BTU/(Ib)(F) . 0.25 (a.t 78 0 F)·
K. BTU/(I,,)(lL')(FJ/fl . 66 1212 to 672" F)
_. 10·' in./in./F . 14 (6f; to 212 0 F)
• Ihthll.tL..... • Trll.llsverllll F •• tdlowtlbles ere 6fJUlli Io or cruUt than tb, 1lJbIttud.lna1 POI .UO....blu.

(C')
".;
iJ\,
B2.28 MECHANICAL AND PHYSICAL PROPERTIES OF METALLIC MATERIALS FOR FLIGHT VEmCLE STRUCTURES

AZ31B ~AG}ITSIUM ALLOY (SH~~ & PLATE) (Cont.)

j /
.I
.,
1 ..

Fig. 82.86 Effect of temperature on the ultimate shear

strength (F s u) of AZ3IB-H24 magnesium alloy.
• • •
r:=I~o''''
Fig. B2.82 Typlcal stress-strain and tangent. modulus
eurves for AZ31B-O magnesium alloy at room
temperature (longitudinal).
I.I
•
.'
!
j l

.I
• Fig. 82. 87 Effect of temperature on the ultimate bear-
ing strength (Fbrn) of AZ31B~H24 magnesium alloy.
!
l

Ftg. B2.83 EHeet of temperature on the ultimate

tensile strength (F tu ) of AZ31B-H24 magnesium alloy.

j
.I
.!-. Fig. 82.8a Effect of temperature on the bearing yield
strength (Fbry) of AZ318-H24 magnesium alloy.

Fig. 62.84 Effect of temperature on the tensile yield

strength (FtyJ of AZ31B-H24 magnesium alloy.

I
.I
•
!
l

Fig. 82.89 Typical stress-strain and tangent-modulus

Fig. B2. 85 Effect of temperature on the compressive curves for AZ31B~H24 magnesium alloy at room
yield strength (Fey) of AZ31B-H24 magnesium alloy. temperature.
 Table 82.21 Design Mechanical and Physical Properties of HK31A Magnesium Alloy (Sheet, Plate and Sand Castings)

. ... HK31A
Alloy . .

Form . . ... . ...

~~--

Sheet· and plate • Sand

castings·
..
Z
r-
~
Condition. .... .. . ... -0 -1124 -TO
OJ
~
"'
w
~

Thick ness, in. . . . . ... . ..... . 0.01ti-

0.250
0.261-
0.1iOO
0.501-
1.000
1.001-
3.000
0.016--
0.126
0.126-
0.250
0.261-
0.1iOO
0.501-
1.000
1.001-
2.000
2.001-
3.000 ... . . ..... ~ .
W

Z
"~ t:l

A B A B (.) t:l
Buala . . . . . ............ A B ~

- §
,. '"
W
~

o
Mechentce! properties: f: z
PI., kai . ..... ... . .....
F,u. kai .....
30
18
32
19
30
l'
30
16
29
14
34
20
30
28
34
23
36
24
34
21
34
18
33
16
33
15
27
13
...
0
0

1'~u, k"i . ... .. . . . ...... 12 13 10 10 10 20 21 19 22 18 13 11 11 13 to '"

...... .. . . ..... Bi
r.;ksi. .. . ... . ... . . ...
1'..... ksi
("/D~I.5). . . .... .
22

43
22

40
22

43 . .......
22 . .......

..... ...
23

49
23

52
23

49
23

51
23

49 . ....
23 .. . . ....

.. . ....... ..... . . .. ... . .

tl -'"
r-
c
("/D~2.0) . .. . . . .. .. . 51 54 51 ...... . . . ....... 67 00 57 58 57 . .... .. . ........ ..... . .. . .... ..... '0
r ''""
'"'" ee
F••u, ksi <:
("/D~ 1.5) .... .. ' . .. . 24 25 21 ...... ........ 34 35 33 30 31 ........ . ...... ........ . .. . .....
........ .. . .. ... ... . . ... . . . . . , . , ... . . .. . ." .......
("JD~2.0) .. .. . .. . . . 24 25 21 34 35 33 30 31
"" 'r-"
co
~

0
e, percent. . ... " . 12 20 12 12 12 4 0 4 8 4 10 10 8 4 ~
E, 10° pai . . . . ........ . . 6.5 0 '"
w
. . ..... •. 5 1.;
'"Z "''0""
EI, 10' pai.
G, to' p.ti. .. . ... . ... . ... 2. 4 H

g
"''""
Physical properuea:
w, lb./in.'. .... . . ..... 0,0647
C, BTlJ/Ob.llF) ... , ... 0.25 (320 to 2120 F)
K, BTU/(h,.)(ft.')(F) ft. 60.0 (at 68° Ji') '"
W

«: to-I In./io./I<'. ..... . ..... 15 (68 0 to 392 0 F)

• f'rop"rU~s tor 'I.heet lIud pllliu IlJ'(l lako:!D plUlIllello lile direction of rollin.. Tl'anIVCraoproPOrtIall • Mechanl(llli proponlea are ba.sod UJ!'llD tho ruaranteod te.D.1lI1o propertlea from IIllplmllely-(:ll4t
are equal to or Irellter than tho Irmll;IUldlulll propertloa. tlll& ban. 1'he IIWIchanlcal propenle, Ilf ban cut from eNUnl' U1aY be .. klw as 16 porocnl of the
• RekreliOO III10uld be lIllOdll to tllll q>eclll(l roqulremt:lIl1 of tho procur1P, or oarllt\cllllu, qeDey tabulated values .
with rellUd to tho USil of lhe abovo value.lll tho delll,u eaatlDp.ct'

I
• J
82.30 MECHANICAL AND PHYSICAL PROPERTIES OF METALLIC MATERIALS FOR FLIGIIT VEmCLE STRUCTURES

HK3lA l1AGNESI11!1 ALLOY (SHEET, ?LATE & SAND CASTI)iGS) (Corrt . )

"'" ·
•f '"'" •,te '"
so
~ •
• ~ •s- ~

.'
• '" , '"
0

~ ,
0

,(X) zoo XlO -00 sec 600 roo ItOO

T _ '..... F

Fig. 82.90 Effect of temperature on the ultimate Fig. B2.93 Effect of exposure at elevated
tensile strength (Fta) of HK31A~H24 magneeuim temperatures on the room-temperature tensile
alloy. yield strength {Ftyl of HK31A~H24 magnes iurn
alloy.

[00 zoo 300 -00 ecc 600 roo ItOO

T~t_.F'

Fig. B2.91 Effect of temperature on the tensile Fig. B2.94 Effect of temperature on the ultimate
yield strength (Fty) of HK31A-a24 magnesium tensile strength (Fta) of HK31A~T6 magnesium
alloy. alloy (sand casting).

i
•
~
•

Fig. B2.92 Effect of exposure at elevated Fig. B2.95 Effect of temperature on the tensile
temperatures on the room-temperature ultimate yield strength (Ftyl of HK31A-T6 magnes ium
tensile strength (F tu ) of HK31A- H24 magnesium alloy (sand casting).
alloy.

Ll
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES 82.31

AZ61A, AZ63A, AZ80A M.;GNESIUM (EXTRUSIONS, FORGINGS, CASTINGS)

Table 82.22 Design Mechanical and Physical Properties of AZ61Aa Magnesium Alloy
(Extz-ustcns and Forgings)

Alloy. AZ6IA

Extruded bar, rod, and Extruded Extruded

Form. solid shapes Hollow shapes Tubes ForginC

Condition. ........ ........ ' -F

Thickness, in .... <:0.249 0.250-2.499 0.02lHl.750

Besis ... ... . . .... ·.

Mechanical properties:
FhI , kn
L. 38 39 36 36 38
T. .. ..
r.; hi
L .. 21 24 16 16 22
T .. ..
F¢11. k.ri
L. ..... ..... I' I' 11 11 I'..
T. . . . .. ... .......
F,," m. 19 19 ·. 19
FIIrv, k$i
(,/D -1.5). ......... '5 45 50
(,jD-2.0). 55 55 ·. 60
F.... kM
(,j D -1.S). . .. · . 28 28 28
(,jD-2.0). 32 32 ·. 32
e, percent ..
E, 10- psi.
8 9 I 7 7 6
6.3
Ee, 10' p"':. ·. 6.3
G, 10' p.ri. 2.4

Physical properties:
w,lb/in. t . 0.1)647
C, Btuj(lb)(F). 0.25 (at is''F) Ir
K, Btuj[Cb,)(fto)(F)/ftj. 46 (212" to 5i2"F)
~, 10-' ir-../in./F 14 (65" to 212"F)

<J PToperti~s '-or extruded b&l'S, rods. shapes, tubes. and fol'ltings are during fabrication,
tll.ll:~n oaralI~1 to the dlrecttcn of ~xtru.sion or maximum ffil'tAI new ~ Estimated.
B2.32 MECHANICAL AND PHYSICAL PROPERTmS OF METALLIC MATERIALS FOR FLIGHT VEffiCLE STRUCTURES

AZ61.A , AZ63A, AZ80A MAGNESIU11 (:::XTRVSIONS, FORGINGS, CASTINGS) (Cont , )

Table B2.23 Destgn Mechanical and Physical Properties of AZ80Aa Magnesium Alley
(Extrusions and Forgings)

Alley. .U80.'

Form. Extruded bars, rods, and solid shapes Forgings

I
i

Condition. -F -T5 -F I -T5

Thickness, in .... .... ,., ... <'0.249 0.2.10- 1.5~ <'0.249 0.250- 1.500-
1.499 2.499 1.499 2.499

Huia... ........ .. ............

Meehaniea1 properties:
F ta • kn
L .. . ........ ... ....... 43 43 43 47 48 48 42 42
T. ............... . . ... ..
Fly, k.ri
L .. ............ . .. .... 28 28 28 30 33 33 2" 28
T ... • • • • . .. . .
• • • ' •••• 4 •
·. .. ..
JI~ kn
L .. . . ... .. . . .. . . .. . . . .. 17 17 .. 28 27 18 25
T ....... ....... . .. .. ·. ..
p... hi. ......... ....... 19 19 19 20 20 20 20 ,0
PIw.,kn
(.ID-1.5). ... ...... ' 48 48 '8 50
(.ID-2.0). ........ 56 56 56 .. 70
F.... kri
(.ID-1.5) .. ......... , 36 36 36 42
(.ID-2.0) .. ........ 40 40 40 ·. ..
e, percent...... ..... ........ 9 8 6 4 4 4 5 2
E, 10' pIi......... .... 6.5
B" 10' pn. ..... ... . ..... 6.5
G, 10 1 p:ft... ........... .. . 2.4
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES B2.33

AZ6lA, AZ63A, AZSOA MAGNESIUM (E:CTRUSIONS, FORGINGS, CASTINGS) (Cant.)

.. _~
Table B2. 24 Material Specifications for AZ63A
Magnesium Alloy
'''- A_r,n,il. 1I
e.com_ _ Il
-,train
-ttrQIn
C. ren ..l. lan~l' ~ ..l_
o'C_'~lft04IIl",
Specification Type of product

QQ-M-56 .. Sand' castings

QQ-M-OO .... Permanent-mold cutU1gs

s.
TABU: 4.2.3.0<"). Denp MccA4nical aM PhyftctU
P-ropmiu of AZS3A Mag7Ulnum- Allov (elUting_)

Alloy. . .... AZ63A

Sand and permanent- '. •

5M*, QOOl in.A\.
•
Form. mold castings Torw,.nl MadIll.... tf psI

Fig. 82.96 Typical stress-strain and. tangent-modulus

Condition. ... -F -T4 -T5 -T6 curves for AZ63A-T4 map,estum alloy (sand casting)
---I - - --- --- at room temperature.

Thickness (in.) .. ..
'-'"
A.r.....II. ttnn-,traln
Basi8~ .. ...... . ...... e.c~" ...-~n
C.Ttnlil, ~ lr'OClul...
Do~"",~

Mechanical properties:
F tur lui .... . . .. . . ... 24 34 24 34
Phi. k..n.. .. . - ........ 10 10 11 16 1
Ft:fI, Jeri .. ... . .. . ." . 10 10 11 16 §
F,.." kai ... 16 17 19
F/wv., kn
(,/D-L5). 36 36 50
!
(,/D-2.0) .. 50 50 .. 65
Fl1rT/.kn
(,/D-L5). 28 32 36
(,/D-2.0) . ..... 30 36 45

'.
t, percent .. 4 7 2 3
E, 10- pn.. 6.5
Ee• 10- pn... 6.5 • •
$traWl,QOOI inA\
•
G, 10· pori. 2.4 TGftqenf Mocl"~ 10· ""

Fig. B2.97 Typical stress-strain a.nd tangent-modulus

curves for AZ63A-F magnesium alloy (sand casting) ar
room temperature.
Table B2.25 Design Mechanical and Physical Properties <xl
of 8Mn TItanium Alloy

j ..
~
w

Alloy 8Mn
J ~o
J J
Form Sheet, plate, llud strip
• ~n
>-
! ! t-

Coudiuon AlIllealed
1 1
1000 1200 1400 1600 ~
'~"
Fig. 82,101 Effect of temperature on the ultimate 00
Thickllt'::lll, ill Fig. 82,98 Effect of temperature Oil the ultimate ~

tensile strength (Ftu) of BMn annealed tttantum shear strength (Fau) of 8 Mil annealed titanium o
>-
alloy. allOy.
~ r-
'o
ljasis A o-J
H '"
0
'00 o-J
~ ''"..."
s,- 'i'l"
H
Mccllll.uieal properties:
1",.., k~i
L I~O r- '"
0
r 120
r-
0 '"
;:
1"'11' bi
I- flO
'"
tn
...'>-"
r r-
110 ~ t-
1"~II' kei ~ n
L
~...
110 200 400 600 lIOO lCOO 1200 1400 1600 200 400 600 eoc ICOO 1200 1400 1600

T 110 1.,,,p.''''u<t. f 'U

p..., k~i 8'
T'~f"'_,f
r
o-J
M

~
Fig, 82.99 Effect of temperatur-e Oil the tensile Fig. 2.102 Effect of temperatur-e all uie ultimate to
Ftm.. /.:~i
yield strength (Fty) of 8 Mn annealed titanium hear-lug st rength (to'bru) of 8 Mil annealed titanium
(e/l) = 1.5)
(el D =2.0)
110 alloy. alloy. "'co r-'"
n.; k.~i "3
'"0
(el/) = 1.5) 130 '" '"
H
'0
,00
'"
e,
(el D = 2.0)
Iwr"l~nt 10
!
j eo
I,
I
!ii::~
r-
Cl
E, 10 0 pei
Re , 10" JI$i
.. 15.5
ro.u
J !
< eo
,Ijjijjjilf
' ,,:I ~ IJ '
;'j
'"
M
G, HI" JI$i J •
.:
, m
n
• I
eo .@!JW
J r-

\I'Iii .nun.,.
,:!,!I!:; M
Physical properties: " eo ..•' ...
'"
,S
,~"

w,lll/ill.' 0.171
lillirii '"
C, Btu/(Ib)(r').
1\,IHu/{(hr)(ft')(I<')/ftI
0.118 (ilL 68"1<')
e.a
1000 1200 1400 1600
0
0 '00 ,00 "00 "00 "00 ...o"ct
a, 10- 0 ill./iu./F. 4.8 (ilL 2IXjOJ")
Fig. 82.100 Effect of temperature 011 the Fig. 2.103 Effect of temperature Oil the tensile '"
M

compressive yield strength (l<'cyl of 8 Mu and compressive modulus (~ and Ed of 8Mn '"
annealed trtanium alloy. tltaniutu alloy.

,/
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES B2.35

6Al-4V TlTANIUl1 ALLOY (BAR & SHEEr)

Table 82.26 Des tgn Mechanical and Physical Properties

of 6Al-4V Titanium Alloy

Alloy. ...... . . .. .... . . .... 6AHV

Form.... ......... . ..... Bar

I Rheet.

Condition.. ... . . .... Annealed.

Thieknese , in. .... <1.5 <:0.181

«l -"I(lQ 000.:10 tWO
r_••••• t•• ,

IlaoUo ........ ... , . .... A A Fig. B2.104 Effect of temperature on Ftuo Fty and E of
6Al-4V annealed titanium alloy (sheet and bar)
-
~Ieebanical pecperties:
fi't.. hi
L. ... 130 130
T. .. .... ... . .... .. 130
F'll, hi
L .. .. .... . ... 120 120
T. ...... . .... .. 120
r.; kJti
L.. .... 126 126
T. ..... . ... . ..... 126
1"... ksi. ....... . .... 80 7.
p"",, k.si
(,!D-1.5I. 196 191 Fig. 82. 105 Effect of temperature on the compressive
(,!D-2.0) ..... .. 248 yield strength (Fey) of 6Al-4V annealed titanium alloy
244 (sheet and bar)•
p.... kn
(,I D -1.51. ... 174 163
(,ID-2.0) .. .... 205 198
t, percent .. ........... 10 10
E, 10- pn
L .. .... . .. ... 16.0 15.4
T. . . ..... .. ..... .. 16.4
E(, 10- pn
L .... .. . ... . .. 16.4 16.0
T . . ..... . .... 16.9
G, 10- pai.. . . ..... .... 6.2 ..

Physical properties:
lIf, Ib/in.'. .......
C, Btu/llb)(F) ...
K, Btu/[(h,)(lt')(FI/ltj.
0.160
0.13S (at. 68°F)
3.8 (at 63°F)
alloy (sheet and bar).
-'
Fig. B2. 106 Effect of temperature on the ultimate
bearing strength (Fbru) of 6Al~4V annealed titanium

II, 10-' in./in./F.. 4.6 (at ZOOOF)

Jj
•

Fig. B2. 107 Effect of temperature on the Ultimate shear

strength (Fsu.l of 6Al-4V annealed titanium alloy (sheet
and bar).

~1'::t
:><... ~"---
B-2. 36 MECHANlCAL AND PHYSICAL PROPERTIES OF METALLIC MATERIALS FOR FLIGHT VEHICLE STRUCTURES

INCONEL X NICKEL ALLOY (SHEE'r')

Table 82.27 Design Mechanical and Physical Properties

of Inconel X Nickel Alloy

Alloy. Inconel X

Form. Sheet

Condition. . ...... Precipitation heat-treated Fig. 82. 108 Effect of temperature on

the ultimate tensile strength (Ftu) of
precipitation heat treated Inconel X
Thickne8!, in .......... . . nickel alloy.

Basis.. .... . .... .... . ...

~Ieehanical properties:
P,... hi
L. ...... . ..... 155
T. . . . .. . . ..... 155
FtJIokri Fig. 82. 109 Effect of temperature an
L ... ... . . . .. . .. the tensile yield strength (Fty) of
100
precipitation heat treated Inconel X
T. ....... . 100 nickel alloy•
fi'~/IJ hi
L ... ..... 105
T. ... 105
F,.. ksi..... lOS
Fbm. kin
(t>ID-1.5).
(,fD-2.0). .. 286
F1)r", kri
(t'/D-1.5).
(,/D-2.0) 186 Fig. 82. 110 Effect of temperature an
e, peeeent. 20 the compressive yield strength (Fcyl of
E, 10- pri.. ... . ....... 31.0 precipitation heat treated Inconel X
E~ 10-pn. .... 31.0 nickel allay.
G, 10- pn.. ........ .

Phy&ical properties:
w, lb/in.e. 0.304
C, Blu/ObHF). 0.109
K, Blu/[(lu-)(fl')(F)/ft] 8.7 (80° to 212°F)
z, 10.- in./in./F ... 6.4 (100° to 200°F)
'_f
Fig. 82. III Effect of temperature an
the ultimate shear strength (F su) of
precipitation heat treated Inconel X
nickel alloy.

'-'
Fig. 82. 112 Effect of temperature on
the tensile modulus (E) of Inconel X
-,
Fig. 82. 113 Effect of temperature on
the ultimate bearing strength (FOru)
,-,
Fig. 82. 114 Effect of temperature an
the bearing yield strength (FbryJ of
nickel alloy. of precipitation heat treated Inconel X precipitation heat treated Inconel X
nickel alloy. nickel alloy.
 PART C
PRACTICAL STRENGTH ANALYSIS &
DESIGN OF STRUCTURAL COMPONENTS
CHAPTER C1
COMBINED STRESSES. THEORY OF YIELD AND ULTIMATE FAILURE.
Ct. 1 Uniform Stress Condition ~l ~l
..+
~llD[J \4~O
Aircraft structures are subjected to many

l~l
types ot external loadings. These loads otten
cause axial, bending and shearing stresses
acting simultaneously. It structures are to be ,+
deSigned satisfactorily, combined stress re-
lationships must be known. Although in practi-
cal structures uniform stress distribution 1s
- ~l
g;:.~~
o'
0
Fig. Ct. 3 Fig. Ct.4
not common, still sufficient accuracy tor
design practice 1s provided by using the stress stresses.
relationships based on uniform stress assump-
tions. In deriving these stress relationships, Fig. Cl.4 shows a tree body atter the
the Greek letter sigma (a) will represent a block has been cut along a diagonal section.
stress intenSity normal to the surface and thus
a tensile or compressive stress and the Greek For eqUilibrium the sum of the forces along
letter tau (~) will represent a stress intenSity the x-x axis equals zero.
parallel to the surface and thus a shearing
stress.
0(1) + 2 (~l cos 45 0) =0
cos 45
Cl.2 Shearing Stresses on Planes at Right Angles.
1 cos 45 0) cos 450 _
hence, 0 =2 ('t"
i - -r (2 )

2 1 Therefore when a point in a body is sub-

jected to pure shear stresses or intensity 't",
B normal stresses or the same intensity as the
shear stresses are produced on a· plane at 45 0
't"zlbr't"z A

2 1 wit~ the shearing planas.

Cl.4 Principal Stresses

Fig. Cl. 1 Fig. C1.2
Fi3. Cl.l shows a circular solid shatt
subjected to a torsional moment. The portion
(A) at the shatt exerts a shearing stress 't"z on
section (1-1) and portion (B) exerts a resist-
ing shearing stress 't"z on section (2-2). Fig.
Cl.2 illustrates a differential CUbe cut from
shatt between sections (1-1) and (2-2). For
equilIbrIum a reSIsting couple must exist on
top and bottom face at cube. Taking moments
about lower left edge of CUbe:
~x dxdy (dz ) - ~z dzdy (ox ) =0
hence, 't"x ='t"z - - - - - - - - - - -
Thus If a shearing unit stress occurs on
one plane at a point in a bOdy, a shearir~ unit
stress ~f same intenSity exists on planes at
right angles to the first plane.
C1. 3 Simple Shear Produces Tensile and Compressive
Stresses.
Fig. Cl.3 shows an el~mentary bloc~ of
unit dimensions SUbjected to pure shearing
For a body Subjected to any combination ot
stresses 3 mutually perpendicular planes can be
found on which the shear stresses are zero. The
normal stresses on these planes of zero shear
stress are referred to as principal stresses.
Cl. 5 Shearing Stresses Resulting From Principal
Stresses.
In Fig. Cl.5 the differential block is
subjected to tensile principal stresses ax and
Oz and zero principal stress a. The block is
cut along a di~gona~ section glvlng the free
body of F1~. C1.6. The stresses on the diagonal
section have been resolved into stress compon-
(1) ents parallel and normal to the section as Shown.
For eqUilibrium the summation of the stresses
along the axes (1-1) and (2-2) must equal zero.
ZF ~-1. =0
an dudy - ax dzdy cos Q - Oz dydx sin Q =0,
Whence an = ax dzdy COS Q + az dyctx sin Q
dudy dudY
e1.l
C1.2 COMBINED STRESSES. THEORY OF YIELD AND ULTIMATE FAILURE.

ct. 6 Combined Stress Equations

0" z dydx
Fig. Cl.7 shows a different1~1 block sub-
jected to normal stresses on two planes at r~g~t
angles to each other and with shearing forces on
the same planes. The maximum no~~l and shear-
ing unit stresses will be determined.

Fig. Cl.8 shows a free bodY dia3ram of a

portion cut by a diagonal plane at angle Q as
shown.

z dydx

I
/
/
2, I
I • <Judy
, I
'I
, ~

g "
:. "'o&~
/
/ ~
---"-,--
/ /
, , / ,,/ ,,
2 Fig, CL 7

a z dydx I
Ftg. cr.s
an dudy
dzdy _ dydx _
But dudy - cos Q and dudy - e in Q, whence
an = Ox cosa'1 + Oz s1n ag - - - - - - - - - (2a)
,,
Q/ / / '!- ~;;;"--
The normal stress an at a po1nt 1s always // / -, --
/ I
less than the ~tmum princ1pal stress Ox or
o"z at the po irrt ,

ZF._. = 0 Fig. Cl.B

ax dzdy sin 9 + o"z dydx cos Q =0

For eqUilibrium the sum ot the forces in the z
But dzdy
dUdy
= cos Q and dydx
dudy
= sin Q and x directions must equal zero.

ZFx = 0
hence,
'" =Oz sin Q cos Q - ax cos Q sin Q
an dudy cos dudy sin Q - ax dzdy - 't'xzdxdy
'" = (Oz
or, Q -+ '!
- ox) sin Q cos Q
=0 ----- --------(4)
or,
' " = (1/2) (az - ax) sin 2 Q, where Oz Is
ZFz = 0
maximum principal stress and ax Is :nin!::!l.lm
principal stress.
an dudy sin
Q - ~ dudy cos Q a z dydx ~xzdzdY
Since sin 2 Q is maximum when Q :: 45 0 , =0---- - - - - - - - -(5)

'" = (amax - amin )/2 - - - - - - (3 ) By dividing each equation by du and noting that

Stated in WOrdS, the maximum value of the shear- dzdy

dudy -
=cos Q and dydx
dudy
= sin Q
,
we obtain:
ing unit stress at a paint in a stressed body
1s one-halt the algebra1c differences of the
maximum and minimum principal unit stresses. (an-Cx) cos g + ('" - "'xz) sin Q = 0 - - - -(6)
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES Cl.3

(an - azl sin g - (T - TxZ) cos g = a - - -(7) Cl. 7 Mohr's Circle for Determination of PrinCipal
Stresses.
The maximum normal stress an will be maximum It is sometimes convenient to solve
when Q equals such angle Q' as to make ~ = zero. graphically tor the principal stresses and the
=
Thus if ~ = a and Q Q' in equations (6) and maximum shear stress. Mohr's circle furnishes
a graphical solution. (Fig. C1.9a). In the
(7), we obtain,
Mohr method, two rectangular axes x and are z
(an ax l cos g' - 'txz sin g' =a - - - - -(8)
chosen to represent the normal and Shearing
stresses respectively. Taking point a as the
origin layoff to scale the normal stresses ax
(an azl sin g' TxZ cos g' =a - - - - -.(9)
and Oz equal to OB and OA r-espeot t vely. It ten-
In equations (8) and (9) an represents- £he sion, they are laid ott to right of point a and
principal stress. Dividing one equation by to the lett if compression. From B the shear
another to eliminate Q' , stress ~xz is laid ott parallel to Oz and with
the sense of the shear stress on the face DC of
Fig. Cl.9b, thus locating pOint C. ~ith point
an-ax: "t'xz whence, E the midpoint of AS as the center and with
"xa on - Oz radius EC describe a circle cutting OB at F and

On
=ax + Oz + ~ '(ax
2 --V ZOZ )
.+'txz·- - (10)
G. AD will equal Be and will represent the
shear on face AS of Fib. b. It can be proven
that OF and OG are the principal stresses Omax.
and Omin respectively and EC 15 the maximum
shear stress "t'max.' The prinCipal stresses occur
In equation (10), tensile normal stress is on planes that are parallel to CF and CG. (See
plus and compression minus. For maximum an use Figs. c and d). The maximum shear stress occurs
plus Sign before radical and minus sign tor on two sections parallel to CH and CI where HEI
minimum. on. is perpendicular to OB. Ifox should equal zero
then a would coincide with A.
To tind the plane or the principal stress-
es, the value of Q' may be solved for trom
equations (8) and (9), which gives: °n(max)

tan 2 Q' = 2 ~xz ----(11)

ax - Oz
Q' 1s measured :rom the plane at the
largest normal stress ax or oz. The direction
of rotation of Of trom this plane is best de-
termined by inspection. Thus if only the a. I
shearing stresses 'txz were acting, the maximum. LTc 0 I
principal stress would be one or the 45 0 planes, B
J1~
I
the particular 45 0 plane being eaSily deter- (e)
I
mined by inspection or the sense ot the shear , I
stresses. Further.nore it only the largest o~,..,o1') D /
normal stress were acting it would be the maxi- '/
mum principal stress and Q' would equal zero.
Thus 1t both a and 't act, the plane at the
principal stress will be be~Neen the plane on
(d)

TcZT
T
J;foX: R

.-Lax-az
,---
which °acts and the 45 0 plane. As stated 2
(zj

betore cr refers to either ax or Oz whichever a.

is the largest. Z Fig. Cl.9
MaXimum Value or Shearing Stress. (~max.)
Cl. 8 Components of Stress From PrinCipal Stresses by
Mohr's Circle.
The maximum value or 1:' tram equation (3)
equals, In certain problems the prinCipal stresses
may be known as in Fig. Cl.9 and it 1s desired
"max, = (an (max. ) - an("ln.) )/2 ---(12) to tind the stress components on other planes
deSignated by angle Q. In Fig. Cl.lI the axes
Substituting the maximum and minDnum values at x and z represent the normal and shear stresses
an tram (10) in (12), we obtain maxDnUD. shear- respectively. The principal stresses are laid
ing stress as tallows: orf to scale on ox giving points D and E respec-
tively. Construct a circle with A the midpoInt
=• at DE and with diameter ED. Draw angle CAB equal
C1.4 COMBINED STRESSES. THEORY OF YIELD AND ULTIMATE FAILURE.

to 2Q. It can be proven that 08 represents the 't xz= 5000

normal stress on the plane defc of Fig. CI.IO,
and CB represents the shear stress 't 'on this
plane.

b
Fig. Cl.12
z
0x-!-- ! / -r-·o-x ,'-
. ,J~
"g ,
--- g
I

Fig. C1.10
I
0-.
crz. ax are principal
stresses.

z
On on plane (defc) ---j I
I
Fig. C1. 13
.___.. I
I
0n(min) :;.
Ox • 10000
+ - an(m a x)=120 7
-jLJ
: -2070 Z
A D
B
I Substituting values,
I
I I an = 1000~ + 0 ~ V(1000~ + O)~ . . 5000" = 5000
---+1OX20,~ ... 7070 hence, 0n(max.) : 5000 ... 7070 : 12070 pSi
On(mln.) = 5000 7070 = - 2070 pSI

Fig. Cl.l1 ""'X. = (1/2) (On (max. ) On(mln.)) (ReLEq.l2)

= (1/2)(12070 - (-2070)) = 7070 psi

Cl. 9 EDmple Problems.
'tmax.can also be computed by equation (13),
Example Problem 1.
whence,
The maximum normal and shear stresses will be
determined tor the block loaded as Shawn in
Fig. Cl.12.
""'X. = • vieOOO~ . •
0)" • 5000' - - 7070 psi

The graphical solution making use of MOhr's

circle is shown in Fig. CI.13. From reference
tan 21;11'1=Ox2 -'tXZOz = 10000
2X5000=1
- 0
axes x and z thru paint 0, the given no~al
stresses Ox = 10000 is laid ott to scale on ox hence, Q~ =22.5°.
and toward the right giving point B. From B the
shear stress 'txz = 5000 is laid oft parallel to Example Problem 2.
oz to locate point C. With E the midpoint or
OB as the center at the circle and with radius T~e maximum normal and shear stresses will
EC a circle is drawn which cuts the Ox axis at be determined for the block loaded as shown in
F and G. The maximum and minimum. principal Fig. Cl.14.
stresses are then equal to of and oG which ~ O"z = 20000
equals.12070 and -2070 respectively. The maxi- ~ I
mum shear stress equals EC or 7070. II ' 't'xz '" 12000

Algebraic Solution: From eo , (10),

Ox' 10000 ~lDl~ Ox = 10000

- ~ 't'xz = 12000
I Fig. C1. 14
O'z .. 20000

--
--------------_.._--.. I. _-~

ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES ci. 5

Fig. Cl.15 shows the graphical solution oz)a~ + (OXOy + oyaz + 0xoz
3
0 - (ox + Oy +
using MOhr's circle. From ,otnt 0, Ox = 10000 2

and O'z = -20000 are lald off equal to 08 and OA -r YZ _Yx:y2)a - (OXOyOz + 2YyzYxzYXY - 0XYyz-;J
respect 1 ve Ly .. TXZ equal to 12000 is lald off - OyYxz• - OzYXya =0 - - - - - - - - - ~ -(14)
?arallel to OZ at A locating C. rlith ~ the
mtdpo tn t of AS as the center of a circle of
radius EC a circle 1s drawn which cuts the ox
axis at F and D. The maximum no~al and shear
stresses are indicated on the rlgura.

Fig. cr.is

/
y Fig. Cl. 16

--.i..
Q''''180~15'''' _ Fig. Cl.l? shows the principal stress
o B F system which replaces the system of Fig. Cl.16.
It can be shown that the maximum shear stress
~. is one of the following values.

I I -r:nax.=::t(OL all)

- - - - - -(15)
~ ~ ;:::to",,",]
+ 1
or r: max. = -"2 CO. - 0'3)

+ 1
I " •• = -"2 (0, - OL)
"n(mini1420~
cr t max,
ron(max)' -2420
The planes an which these shear stresses
act are indicated by the dashed lines in Fig.
Algebraic Solution Cl.IS J namelYJ adhp., bdge and dcef. The
largest or the shear stresses in equations (15)
depends on the magnl~ude and Signs or the
=ax + Oz
2
- az)" + -rxz ..
2 principal stresses J rememberin~ that tension
is plus and compression is ~inus when making
the SUbstitution in equations (15).
= 10000 - 20000 :!: .,j(1000o-~-20000)) ', 12000'
2

.:~
5000 t 19200 b C

hence, On(max. ) = -5000 19200 = -24200 pet

°n(mln. ) = -5000 + 19200 =
14200
q, '"
= :t 2 x 12000 Fig. Cl. 17
't"max. 19200, tan 2Q'
1000 - (-20000) = .8
Q' = ISO _ 50' C1. 11 Principal Strains

Cl. 10 Triaxial or Three Dimensional Stresses
For bodies which are stressed in three
directions, the state or stress can be defined
completely by the six stress components as
illustrated in FIg. Cl.16. USing the same
procedure as was carried out for a two-di~en­
sional stress system, it can be shown that there
are three prinCipal stresses O'LJ a. and 0.,
whose values are the three roots of 0 in the
following cubic equation.
The strains under combined stresses O3,':'"e
usually expressed as strains in the direction
or the prIncipal stresses. Consider a c?se or
Simple tension as illustrated in Fig. Cl.l9.
The stress 0 1 causes a lengthening unit strain
e in the direction or the stress OLJ and a
Shortening unit strain ~' in a direction at
right angles to the stress OL.
The ratio of e' to e is called Poisson's
ratio and is usually given the symbol~. thuS,
IJ. = ~I /e
Cl. 6 COMBINED STRESSES. THEORY OF YIELD AND ULTIMATE FAILURE.

E tcr, The strain energy can be expressed in

2.-1.-- --~ terms of stress by substituting values ot E in
:• ' terms of G from equations (17) into equation
I, ',

,
cr, ,

)-
-i--cr, (19), which gives,
,
E' , E' /

~
/
2" ,:-- 2"
cr,
/
Fig. Cl. 20 U =..l.. + 0:+ 0: - 2iJ.(01.0.... G,GIS + 0,,01.)
:",-r--" lcr, 2E
01. Fig. Cl. 19 - - - - - - - - - - - - - - - - - -(20)

Since e = 04/E, we obtain, For a two dimensional stress system,

O"~ ~ 0 and equation (20) becomes
- - - ~ - - ~(16)

Now consider the cubical element in Fig.

CI.20 sUbjected to the three principal stresses
cr~, cr. and G e , all being tension. The total C1. 13 Structural Design Philosophy. Limit and Ultimate
unit strain e 4 in the direction or stress cr~ Loads. Factors of Safety. Margin of Safety.
will be expressed. ObViously, 01. tends to
stretch the element in the direction or o~ The basic philosophy governing the
whereas stresses a. and a~ tend to shorten the structural design of a flight vehicle is to
element in the direction or a~, hence, develop an adequate 11ght weight structure that
will permit the vehicle to accomplish the
f:l. -Et - ~ - ~ , whence operations or missions that were established
as deSign requirements. The jOb of a
cammercial airliner is to carry passengers and
e, -T - pa. + oe)
'=-
cargo fram place to place at the lowest cost.
To carry out this job a certain amount ot
and Similarly for e. and e~,
flight and ground maneuvering is required and
-----(17)
the loads due to these maneuvers must be
cr. - E'O'e
u. ) carried safely and efficiently by the structure.
CSII 'Z E ... 0'1.
A military fighter airplane must be maneuvered
in flight far more severely to accomplish its
desired job as compared to the commercial air-
liner, thus the flight acceleration factors
For a two-dimensional stress system, that for the military tighter airplane will be
is, etr-essee acting in one plane, as = 0 and considerably higher than that of the airliner.
the principal strains become, In other wordS, every type of flight vehicle
will undergo a different loan enVironment,
1 which may be repeated frequently or infre-
E, = - (0 - flO. ) quently during the lite of the vehicle. The
E '
load environment may involve many factors such
E. = t (cr. - flO,) -----(18)
as flight maneuvering loadS, air gust loads,
take off and landing loadS, repeated loads,
1
high and low temperature conditions, etc.
E. :; E (a~ + cr. )
Limit LoadS. Limit loads are the calculated
Equations 17 and 18 give the strains when maximum loads which may be SUbjected to the
all the prinCipal stresses are tensile stresses. flight vehicle in carrying out the job it is
For crnnpressive prinCipal stresses use a minus designed to accomplish during its lite time
sign when substituting the principal stresses of use. The term limit was no doubt chosen
in the equations. because avery flight vehicle is limited
relative to the extent of its operatiOns. A
cr, 12 Elasttc strain Energy flight vehicle could easily be deSigned for
loads greater than the limit loadS, but such
The strain energy in the elast1c range for extra strength which is not necessary ror
the unit CUbe in Fig. Cl.20 when subjected to safety would only increase the weight at the
combined stresses is equal to the work done by structure and decrease the cammercial or
the three gradually applied principal stresses military payload or in general be detr1~ental
Ol" cr." and O"IS. These stresses produce strains
to the design.
equal to el.' e. and Ee and thus the work done
per unit volume equals the strain energy. Thus Factor or Safety, Factor Of safety can be
if U equalS the strain energy, we obtain, defined as the ratio considered in structural
design of the strength of the structure to
U • ¥ + ~ + 0,,2';) - - - - - - -(19) the maximum calculated operational loadS,
that is, the limit loads.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
Yield Factor at Safety. This term is detined
as the ratio of the yield strength of the
structure to the limit load.
Ultimate Factor of Safety. This term is
defined as the ratio of the ultimate strength
at the structure to the limit load.
Yield Load. This term Is defined as the limit
load multiplied by the yield factor of satety.
Ultimate Load. This term can be detined as the
limit load multiplied by the ultimate factor
of safety. This resulting load is often
referred to by engineers as the design load,
Nhich is misleading because the flight vehicle
structure must be designed to satisty both
yield and ultimate failure and either one may
be critical.
Yield Margin or Safety. This term usually
expressed in percent represents the additional
yield strength at the structure over that
strength required to carry the limit loadS.
Yield Margin or Safety = Yield Strength _ 1
L1m1t Load
Ultimate Margin or Sarety. This term usually
expressed in percent represents the additional
ultimate strength of the structure over that
strength required to carry the ultimate loads.
_ Ultimate Strength
Ultimate Margin at Satety - Ultimate Load
Cl. 14 Required Strength of Flight Structures.
Under Limit Loads:-
The flight vehicle structure shall be
,designed to have surrtc ient strength to carry
simultaneously the limit loads and other
accompanying enVironmental phenomena tor each
design condition without undergoing excessive
elastic or plastic deformation. Since most
materials have no definite yield stress, it is
crnmnon practice to use the unit stress where a
.002 inches per inch permanent set eXists as
the yield strength of the material, and in
general this yield strength stress can be used
as the maximum stress under the limit loads
unless definitely otherNise specified.
Under Ultimate Loads:-
The flight vehicle structure shall be
designed to withstand simultaneously the
ultimate loads and other accompanyIng environ-
mental phenomena without failure. In general
no factor of safety is applied to the environ-
mental phenomena but only to the limit loads.
Failure of a Structure:-
This term in general refers to a state or
condition of the structure which renders it
Cl. 7
incapable of performing its required function.
Failure may be due to rupture or cOllapse or
due to excessive deflection or distortion.
Cl. 15 Determination of the Ultimate Strength of a Structural
Member Under a Combined Load System. Stress
Rat1o·lnteraction Curve Method.
Since the structural designer or flight
vehicles must insure that the ultimate loads
can be carried by the structure without failure,
it is necessary that reliable methods be USed
to determine the ultimate strength of a
structure. Structural theory as developed to
date is in general SUfficiently developed to
accurately determine the ultimate strength ot
a structural member under a single type at
loading, such as axial tension or ccmpression,
pure bending or pure torsion. However, many
of the members which compose the structure at
a flight vehicle are SUbjected Simultaneously
to various combinations at axial, bending and
torsional load systems and thus a method must
be available to determine the ultimate strength
ot a structure under combined load systems.
A strictly theoretical approach appears too
difficult for solution since failure may be
due to overall elastic or inelastic buckling,
or the local elastic or inelastic Instability.
The most satisfactory method developed to
date is the so-called stress ratio, inter-
action curve methOd, originally developed and
presented by Shanley. In this method the
1 stress conditions on the structure are repre-
sented by stress ratios, which can be con-
sidered as non-dimentional coefficients
denoting the fraction of the allowable stress
or strength for the member which can be
developed under the given conditions ot CQm-
bined loading.
For a single Simple stress, tt~ stress
ratio can be expressed as,
R = stress ratio • f -(22)
where r is the applied stress and F the
allowable stress. The margin of satety in
terms or the stress ratio R can be written,
1
M,S·:il-l. O - - - - - - - - - -(23)
Load ratios can be used instead of stress
ratios and is often more convenient.
For example ror axial loading,
R • P/P a , where P = applIed axial load
and Fa the allowable load.
;, o ~-;.
",,-..•.1 ...;
ci. 8 COMBINED STRESSES. THEORY OF YIELD AND ULTIMATE FAILURE.
For pure bending,
R = MlMa , where ~ = applied bending
moment and Ma the allowable bending ~oment.
For pure torsion,
R TITa, where T is applied torsional
moment and Ta the allowable torsional moment.
For combined loadings the general
conditions for failure are expressed by
Shanley as follows:-
R~ + R; + R~ + ----- = 1.0 - -(24)
In this above expression, R1 , R. and R3
could reter to compression, bending and shear
and the exponents x , s, and z give the
relationship tor combIned stresses. The
equation states that the ra i ture of a
structural member under a combined loading will
result Only when the sum ot the stress ratios
Is equal to or greater than 1.0.
For some ot the stmpler combined load
systems, the exponents of the stress ratios in
equation (24) can be determined by the various
well known theories of yield and ra tjur-e that
have been developed. However, in many cases
or combined loading and tor particular types
or structures the exponents In equation (24)
must be determined by making actual failure
tests of combined load systems.
SInce the stress ratio method 'NaS pre-
sented by Shanley many years ago, much testing
has been done and as a result reliable inter-
actiop equations with known exponents have
been obtained for many types or structural
members under the various combined load
systems. In a number ot the tollowing chapters,
the interaction equations which apply will be
used in determining the ulttmate strength
deSign of structural members.
C1. 16 Determination of Yield Strength of a. Structural
Member Under a Combined. Load System.
As explained in Art. Cl.14, the flight
vehicle structure must carry the ltmit loads
Without yielding1 which in general means the
Yield strength ot the material cannot be
exceeded when the structure is subjected to
the limit loads. In some parts of a flight
vehicle structure tnvoLvtng compact unf t or
pressure vessels, biaxial or triaxial stress
condit~ons are often prOduced and It is
necessary to determine whether any yielding
will occur lUlder such combined stress action
when carrying the limit loads. For cases
Where no elastic instability occurs, the
follOWing well known theories of failure have
been developed.
1. ~3X1mum Principal Stress Theo~y
2. Maximum Shearing Stress TheorJ
3. Maximum Strain Theory
4. Total Strain Energy Theory
5. Strain Energy of Distortion Theory
6. Octahedral Shear Stress Theory
The reader may review the explanation and
derivation ot these 6 theories by referring
to such books as listed at the end of this
chapter.
Test results indicate that the yield
strength at a point in a stressed structure
is ~ore accurately defined by theories 5 and
6 followed in turn by theory 2. Since
theories 5 and 6 give the same result, they
mi~ht be considered as the same general theory.
In this chapter we will only give the resulting
equations as derived by theory 6, since
theories 5 and 6 appear to be the theories
used in tlight vehIcle structural deSign.
Cl. 17 The Octa..hedral Shear Stress Theory.
Since this theory gives the same results
as the well known energy of distortion method
it is often referred to as the Equivalent
Stress Theory. The octahedral shear stress
theory may be stated as follows:- In elastic
action at any point in a bOdy under combined
stress action begins only when the octahedral
shearing stress becomes equal to 0.47 fe'
where f e is the tensile elastiC strength of
the material as determined from a standard
tension test. Since the elastic tensile
strength is somewhat indefInite, it Is common
practice to use the engineering yield strength
Ft y• In this theory it is assumed that the
tensile and compressive yield strengths are
the same. -
Figs. Cl.21 and Cl.22 illustrate the
conditions ot equilibrium involving the octa-
hedral shear stress. In Fig. Cl.21, the cube
Is subjected to the 3 principal stresses as
shown. A tetrahedron is cut irom the cube
and shown in Fig. Cl.22. Three ot the sides
of this tetrahedron are parallel to the
3
3
I I
I, - ,,
,
-+--1, I,
,- 2 ==--2
Fig. Cl. 21 Fig. ci. 22
-n~-
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STH CTURES C1.9

prinCipal axes, while the n~rmal to the fourth For a triaxial stress system J
side makes equal angles with ~he principal
axes. The octahedral shear and normal stresses
are the resulting stresses on the fourth side.
The equation for the value of the normal
octahedral stress 15, For a biaxial stress system, f y• fS yz J
t s yx = 0
_ 1
foct-3'(f1.+fa+f,,) ------(25) j' = ,; f~ + f~ - fxt z + 3f sxz
ll
- - -(32)

The equation for the octahedral shear C1. 18 Example Problem 1.

stress 15,
Now the octahedral shear stress is 0.47
of the normal stress.
Let r be the effectlve axial stress In
unlaxial tension or compression which results
in the given octahedral shear stress.
_ 3
f = t s oct / 0 . 47 =v'2 f Soct
Therefore multiplying Eq. (26) by
3/~ we obtain for a condition of principal
triaxial stresses,
- 1 I a a a
r=/2 v(t,-f,) +(t,-t,) +(f,-t.) - - - -(28)
Let F equal the allowable tensile or
compressive stress. If the yield strength is
being dete~ined then
I1a.rgin of safety M.S.=
A cylindrical stiffened thin sheet fuse-
lage is fabricated from 2024 aluminum alloy
sheet which has a tensile yield stress
Fty = 40000. Find the yield margin of safety
under the following limit load conditions.
(1) A limit bending moment produces a bending
stress ot 37000 pSi (tension) at top
point at fuselage section. The tlexural
shear stress is zero at this point.
( (2) Same as condition (1) but pressurization
- - - - - 27) at fuselage produces a circumferential
tension stress of 8600 psi and a longi-
tudinal tension stress ot 4300 psi.
(3) Same as condition (2) but a yaWing
maneuver of airplane prOduces a limit
torsional shearing stress of 8000 pSi
in fuselage sk~n.
SOLUTION: Condition (1)
This is a uniaxial stress condition for
point being considered.
F
-=- - 1 -(29) Fty 40000
t Yield M.S. = r;;--1= 37000 - 1 = .08

For a biaxial stress system taking f:s = OJ SOLUTION: Condition (2)

we obtain,
-f = v/ f'1. + f'II •.1,], t II There are no flexural shear stresses at
- - - - - - - -(30) the fuselage point being considered. Since
It is often more convenient to use the x J y no torsion 1s being applied
and z component of stresses instead of the to fuselage no torSional
prinCipal stresses. Fig. Cl.23 illustrates shear stresses exist. The
the various component stresses. stress system at the paint
be1r~ considered 1s thus a
f, biaxial stress system and
flo and f 21 are principal
stresses.
f1. 37000 + 4300 : 41300 ps i
is
t II = 8600 pat
~-+-_ix
From equation (30),
is xy

Fig. Cl. 23 = J 41300 21

+ 8600a - 41300 x 8600
Cl.IO COMBINED STRESSES. THEORY OF YIELD AND ULTIMATE FAIT..URE.

whence r = 37700 psi The vessel wallIs to be stressed to the

yield stress at 42000, thus t = 42000.

vhence
SOLUTION: Condition (3) (42000) •
Since a torsional
shear stress has now f SoIvtng , p = 970 psi.

tx-ottx
been added, the new ,z f s_=8000
! ......... ---
stresS is still two PROBLEMS
dimentional, however
the given tension (1) The combined stress loading at a point
stresses are nat } f~ =:8000 in a structure is as to11ows:- f z =
principal stresses f
z
-xa -1000, t x =
-2500, t s =
2000. Determine
due to the addition the magnitUde and direction ot the
at the torsional shear principal stresses. Determine the
stress. maximum shearing stress. Solve both
analytically and graphically.
!x=41300 psI. !z=8600 ps f , !s=8000 ps i ,
(2) Same as Problem 1, but change t z to 4000
Instead at
tinding the principal stresses and t x to -3000 and t s to 2500.
and USing Eq. (30), we will use the t x and t z
stresses and use Eq. (32) (3) A solid circular shatt is subjected to a
limit bending moment at 122000 inch
poundS and a torsional moment of 250,000
inch pounds. It di~eter is 4 inches
and the yield tensile stress is 42,000,
=" 41300 1I+ 8600·- 41300 x 8600 + 3 x 8COO·
what is yield Margin or Satety.
? = 40200 pst • x.a, =~
40200
- 1 =-.01 (4) A thin walled cylinder ot diameter 6
inches is subjected to an axial tensile
Thus yield is indicated since M.S. Is
load ot 15,000 pounds, and a torsional
negative. moment or 12,000 inch pounds , l..lhat
should be the wall thickness it the
Example Problem 2.
permissible yield stress is 30,000 psi.

A cylindrical pressure vessel is 100 (5) A closed end cylindrical vessel is 15

Inches in diameter and 1 inch thick. The inches in diameter and a wall thickness
vessel Is made ot ·steel with Fty ; 42000 pSi. ot 0.25 inches. The vessel 1S SUbjected
Determine the internal pressure that will to an internal pressure ot 10,000 pSi,
produce yielding. and a tensile load at 22,000 pounds. If
the yield tensile stress or the material
SOLUTION: This applied stress system 15 1s 75,000 p 1, what torsional moment can
biaxial with no rteccuraj, or torsional shear. be added without caUSing yield.

Let: p equal internal pressure t, REFERENCES: -

t = wall thickness = 1 in.

-Q-t,
I
d = diameter = 100w Nadat , Theory or r'tcw and Fracture or Solids.

t. =ct reuererenetar stress due Ttmoshenko, Strength ot Materials.

to pressure p
FreUdenthal, The Inelastic BehaVior at
t • =pd and t :I.-it
-pd Engineering Materials and Structures •
2t
Marin, J., Engineering Materials.
From Eq. 30
Seely & Smith, Advanced Mechanics of Materials.
r=/t:+t:-
 CHAPTER C2
STRENGTH OF COLUMNS WITH STABLE CROSS·SECTIONS

C2. 1 Methods of Column Fatlure. Column Equations. The range AS in Fig. C2.1 Is tor a range
or Lip values of below 20 to 25, and repre-
In Chapter Ale, the theory ot the elastic sents a range where failure Is due to plastic
and inelastic instability of the column was crusntng of the column. In other words, the
presented. The equations tram Chapter Ala column Is too short to buckle or bow under
tor a pin end support condition are:- end load but crushes under the high stresses.
This column range or stresses 1s usually
For elastic primary failure, referred to as the block compression strength.
nil::; A column, however, may fail by local
Fe = --"--=-- - - - - - - - - - (1) bUCkling or crippling due to distortion of
(Lip) ,
the column cross-section in its own plane.
For inelastiC primary failure, The horizontal dashed line in Fig. C2.1
represents the condition where the prtmary
column strength is limited by the local weak-
Fe = - - - - - - - - - - - - (2) ness. This line moves up or down aCCOrding
to the value or the local weakness. The
deten:nination ot the column strength when
Where Fe = compressive unit st;~ss at failure 1s due to local weakness is covered
failure = PIA stress. in another chapter.
E = Young's modulus C2.2 Column End Restraint. Fixity Coefficients.
Et = tangent modulus Column Effective Length.

L = column length
The column strength is influenced by the
p = radius or gyration of cross- end support restraint against rotation and by
section any lateral supports between the column ends.
The letter C is commonly used to indicate the
Fig. C2.1 shows a typical plot or Fe end fixity coeffiCient, and C = 1.0 for zero
versus Lip. It the column dimensions are such end restraint against rotation, which can be
as to cause it to fail in range CD in Fig. prOduced mechanically by a pin or ball and
C2.1, the prtmary failure Is due to elastic socket end support !1ttlng. Thus including
instability and equation (1) holds. This the end restraint effect equations (1) and (2)
range of Lip values is ofteL referred to by can be written,
engineers as the long col~~ range.
A B c D
FC = , Fe = CTt.lEt
- - - - - (3)
(Lti> )" (Lip)"
I I I I
Local Crippling Limit Let L' = effective length of the column
which equals the length between inflection
points or the deflected column under load.
Fe
Then L' = Live (4 )

Thus equation (3) can be written as,

L'/~

Fig. C2.1 Fe = Fe -- - - - (5)

(L'/p)' (L'/p)'
The range Be represents the range of L~
values where failure 1s due to inelastic in- If we let P = tailing or critical load,
stability of the column as a whole and equation equation (5) can be written as equatton (6)
(2) applies. This range Be 1s often re~erred by realizing that p =
FeA and p ~. =
to as the short column range. naSI naEtI
p =- -
(L' )'
p =- -
(L' )' - - - - - -
(6)
C2.1
C2 2 STRENGTH OF COLUMNS WITH STABLE CROSS-SECTIONS

Fig. C2.2 shows the deflected colu~~ TABLE C2.1

curve under the load P for various end and
lateral support conditions. The effective 17-7 PH(THI050) Stainless Steel Sheet
lengths L1 and the end f~xlty coefficients Ftu "" 180, 000 E:: 29 X 105
Fey:: 160,000
are also listed.

20,000 29 x 10 5 119. 56
30,000 29 x 10 5 97.62
40,000 29 xlO 5 84.54
50,000 29 x 10 " 75.62
60, 000 29 x 10 e 69.03
70,000 29 xl0'" 63.91
80,000 29 x 10 " 59.78
89, 600 29 x 10 os 56.49
96,000 27.55xlO" 53. 19
102,000 26. 10 x 10 " 50.22
107,400 24.65 x 10 5 47.57
112,200 23.20 x 10 e 45.15
117,000 21. 75 x 10 " 42.81
121, 500 20.30 x 10 e 40.58
125, 800 18.85 x 10 e 38.43
Fig. C2.2 130,000 17.40x10" 36.32
134, 000 15.95xl0 5 34.25
C2.3 Design Column Curves for Various Materials. 137,700 14.50 x 10 e 32.22
147,000 11.60 x 10 " 27.89
For routine design purposes it 1s con- 158,100 8.70 x 10 e 23.29
venient to have column curves ot allowable 167, 600 5.80 x 10 " 18.47
tailing column stress Fe versus the effective 173, aoo 2.90x10" 12.82
slenderness ratio Il/p. In equation (5) we
will assume values ot Fe' then find the This equation is plotted in Fig. C2.16.
tangent modulus Et corresponding to this For a given material, n,.F o . ? and S ~ust be
stress and then solve for the term L' /p. known. Then assuming values of F, we can
Tabla C2.1 shows the calculations for find corresponding values of Et/E from Fig.
17.7 PH (TRIOSO) stainless steel sheet at C2.16. For values of Z, Fa.? and n refer to
roam temperature. The reSUlts are then plotted Table Bl.1 in Chapter Bl.
in F1g. C2.7 to give the column strength curve.
Similar data was calCUlated tor the material C2.5 Non-Dimensional Con... .m Curves.
under certain exposure time to different
elevated temperatures and the results are also QUite useful non-dimentional column
plotted in Fig. C2.7. Figs. C2.3 to C2.l5 curves have been derived by Cozzone and
gi ve column curves tor other materials under Melcon (See Ret. 3).
various temperature conditions. Use ot these
curves will be made in example problems later The E:uler column equation is
in this chapter. The horizontal dashed line F = n a£t / (L1 / p ) a , which can be written,
is the compressive yIeld stress. Values
above these cut-ott lines should be sub- (L'/p)'
stantiated by tests.
F n'
C2.4 Tangent Modulus Et from Ramberg-Osgood
Equation. The problem there~ore resolves ~tself
into obtaining and expreSSion fOT Et/F from
The baste Ramberg-Osgood relationship the non-di~ensional relationship. To do this
tor Et is given as tallows: (See Ref. 1) multioly both sides of equation (~) by Fa ?/F
and equate to B a. •
1
- - - - - - - (7)
(;~~ ~. 9 1
=-;:;---;;---;:;----:-
-L.- + ~ n (-E....-)n
Fa.? 7
= B'
Fa.?
- - (8)

Et = tangent modulus ot elastiCity Fig. C2.l7 shows a pla~ of this equation

as taken trom Ref. 3, and shows FIFo.? versus
E = modulus ot elastiCity B for various values ot n.
For definition ot other terms see Article The shape of the knee of the stress-
B1.12 ot Chapter 51. strain curve is given by the Shape parameter
n and the abscissa B incorporates the
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES C2.3

120 RoT.
R. T. Flg. C.4

Fig. C1• .1
--~
110 S!JOGF'
.!lee1 AJS[ 4130,' ~l4O. 4340
-1t.U-Tl'eamlnFiia-~ t:zsOOO'"
_Sl<!eI, AlSt 4130, H40; 43olO T_lIIp. Blip'-up to 1/2 II&. •
t"'" O. las
roe -- .-------,
Temp. E:::qlo....re up to 11: IIR.
-F ...QOOO<HRo~. 1---.
m
."
·-"!'":':""..'1~---

P';7 .. -,‫סס‬oo (It. T:)

~ ~-_.--~-' -.-"-",..""-,-,,,--,-
- I 0I)lI0'P -:-

l()ll(l9F··

" 10 "~20 30 40 50 80 70 80 90 100

L'IP
10 -~--,--~-----~._------------.-
Figure C2. 4

180 "---= ---R;~ -.-:----:- - - . -

Figure C2. 3

I~O-----R;-'l';
rie, C%.~
Steel Als:l 4130. 4140. 4340
--.--- H:iU'1'.i'nted ToFi;u;- "-15'‫סס‬OO

1:10'- .500S'E,-__ c__ ' -'-'-.!·-~;?i4~~'-~- 130":"--~

120--'---- I:O~---'---~
-~~""85Q01f­

110 -------.--- liD .:....-.-.-;~

100 ,_ -.C- ._-._ _ _ _ '-_._"

aecsn
so

50'

.
10-20
--_._~_.

W 20 30 ~ ~ ~ W 80 90 100 30 40'- ~o-- eo

r:/p t'! P

Figure C2. 5 Figure C2. 5

:, ,-f "
"",,,~'--:>K~
C2.4 STRENGTH OF COLUM!'iS WITH STABLE CROSS-SECTIONS

1B0;.,.....,.....~·-· ._~,---

170-----
RoT.
'" Ro-T.--'

Fig. C2. l'

11_1 P!f(TH1050\5ta1ll.IeS& 9:.... 1
'so F~; C2.8
11......Pll stainl.....a steel
.~ Strip&! P!:Ite" (Bars'" Forg1D.gsl

150 ---- •
L 010 _ 0; 125 In.
:r"mp., Exp', up to 1/2: HR.
F - "iaooco .-
'" .. -Temp.' E:qI. ~p to 1/2 fiR.
Pm'" ISOOOO
tu
_Fer -. 182000
'" Fey-'" 1ll5QOO

140·

130 '"
'"
110 ,.

100· .-

~ sc
,"
80!--'---
"
70·

''::~~'''':',.i-.:_~ :;,
40 ;....:.: . !-

"...---.:'-',---o,,-,.,..
20
...
~,----,----.-

:~_~:::~~;"""--'----:~-_.----,.--.-'-,.-
,

.•..,.........
. ".- ...... _ ..- '.---' -
~.~i-';'- ~¥---,. "-~'';;;::'''~''-i'--''~'-i''''''-'' '. ;--,-,-,,--~o~ 50- 50 ec 'go" 'ioo
10203040 "L'/f' ., eo 90
'"
L>/f'

Figure C2. 7 Figure C2. 8

R. T;" Flg. C2.10

---~---· ..---01U'r!i;,_TlJ A!wn; Alloy
Bue,~ .,Pllle, t ~ O.W Ia..
_._ P'tq-- 16000.
Fey-'" 6'lllQO;
'-----:---:-T~ Ezp;'.-lBB:

".' .. . "'- .,
;,-,... __ ._..,._.:,._ .
' . " "

:.~:_:. ~-"':_:::.....c.:i._
'c' . ", ... " ,., " . . ."",.
t>. :,:;, .' "" ,
...:..-i------'-•.c.:::;..:::.':.:'::.:.:,:....:,-,.c..:.:.:.,-;-.:..,_
50 ,~ .

. . . . . . ., .
._-----"----'----,-.._---"--.- .-
,

,. ,".-,'

----
·~--'i'7---.--.-----'~----.,..;-··,--'·

"-'---
, '._~-

10 ~ 30 40 50 • 70 80 90
-
100
L'IP
ac - - - - - - - _ . -_._---'-- Figure C2. 10

10--rD""30~·--~60--70 ae 90 100
Vip
Figure C2. 9
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES C2.5

,
R. T.

-
f':Ilt. C2,l1
70'l5·Te Alum. Alloy BJltl''l8iOll.
t ~ O.2Ull.
Temp. EJql. • Z HR.
.. FIi", C'.12
__ 'lO'r.!.·Te_Alum. Allay Clad ~
I:! a.50!D.
1'e~. EXp • • a 1m:

Fro'" 75000_ eo Fta--'lOOOQ--------- -

FC'U .. 7‫סס‬oo r-<:i.. 64QOO-

eo '00'"

30
-'-""45Q:offi ilal,·:'-""-··
20'" -.--.--

20 -----_ .._--_._."

=:-:::::;:::~
iu ........

~ W ~ ~ ~ w ro ~ ~

L'IP
10 W ~ ~ 50 W ro ~ -~6 100
Figure C2. 12
L'IP

Figure C2. 11

Fl(; C%.IS
'l0'r5-:t'll Ahzm. All>:lT {Ole Fcc;mgat
. "-'-...
---r-~Z'!li:"----~·-'· .-.-
TflIIP---'-.'Z BR.--
__,)'__,"_71O!lO ....,. __.._
--- Fcr--saooo
"<--

40:_

30

\0 20 W ~ ~ ro ro SO ~

L'II"
Figure C2. 13
C2.6 STRENGTH OF COLUMNS WITH STABLE CROSS-SECTION

Fig. C2.1. 130-· R. T,

·-7l119_~A1W:IIo- A1lo7'iI:ul<t F«tinf Fig. ca. 15
I .:t&.O lIL.- Tl-&AJ~4V T1tan1um Alloy
-
Temp. exp •• 1/2 HR.
------"Iii &"67000 '" au '" Sheet
t ~ O. 187
Annealed
In.
__ ":~1 '::"~ ~ _ n, TtmlVc Exp;; sl!2lJR:.
F hl·13oooo
Ft:T~ lUOOO
'"'
"

40 --- --------

------- -------- 20 -------- ----------- .----.---

'" ---_._-------~~--'--------

~3040501lO108090

L'IP ,-.-,-"20--30---40----50-60-70-80-- 90
L'!P
Figure C2. 14
Figure C2. 15

-:::;1. 00 1. 00
i'-. -c -,... r-, , I"\··n- 0 i I
z
o
n
l--.'
r-, <,
t--.
f---- <, ~"
t--.
.:'-( ""
0 I"::- .,0
\ .'0 ..n"-I i ,1
1 1 I

s-
1'- 1 1
\'r .90 .90
I'--.. t---- r-, 1'- i-, h" 1".\.9 I
I,
,
! i
,I I I
"

~
..; .80
<,
l--. 1"- 't-. r-, 7':> l\ l\ \ I 1 1 ,
.80
1'--- r-,
,
<, ~\ l\ ~\\
I \ , , 1 I

0
.o-f'- ~ k 3 R: ,\ ,\ : ,
,
,,
I I .70
I 1'.5p:;; 1:::-\- i\\ ~\\
"

,
I -r-,
1
,
, ,
,
,
,
, I
.e 0 I ! t"'" ~ \'\: 1
, I
I I I · .0
!
,
, 1 !"":ile.\\
i, ,
, I 1 i ,
,

E, , ,, W- t-- r--: , 1 1 I !
E .5 0 ,I , n ~ 2. 9 .50
1 !~ I'-
, ,
I ,,
t--.. 1
,

,
i
I
I
\'-- ",N
I I I I-----:....J. , I : i I
.4 0 I" ,Ii:'- ,
~I
1
.40
1 1 ! i I ,
1 i-----. '-1 1

.I
.3 0
i
E,
I 1
, 1 I 1
,
,
! I~
[:\ - ,
1'--.'
'i-..... 1
i 'i---J.
~:
I --j
.30
,
,

E =1 .! n (..,..L..tn- ~ ,
I i 1 ,"-. 1 I .1
, ,

,
+7 F o ... 1 1\\ ~. <, ]"'--..., N 1
,

.a 0 , , , I · 'C
~ ~""'-. 'i----: I, I~
I I I I I I I
I I
I Fig. C'. I. Dimensi.onless tangent modulus stress curves. I
, ~~ 1--..---"""" -s~ I , · ,0
. 10
! I 1 I i
1 1 \\ 'R~~~ ,---"l
-!~ ~Il ,<f ~
, 1
I
,
i : , I 1 I I
o .10.20.30.40 -
.so .60 .70 .80 .90 1.001.10 1.201.30 1.40 1.50
F/F" ... Fig. C2.16 (Ref. NACA T. N.902) (1:: sis].
 ,.
Z
,.
r-

-,.''"'""
1i . 0 . i Z
tl
tl

-
l'l
0.9
\d'p·l~ c'"

~IJr
Z
0 .•
0
'"
Fe
0.7
ilt '"r-
-
-F- 0.8 ,"
0.7
I,
.
I'.···
;"
o

.'"
-,.
>oj
0.5

-'"
l'l
i
.70 o
r-
i i, l'l

i I
'"
>oj

l'! 'c":
0.1
.'. ! "c:
>oj

0.2 0.4 3.2 3.4 '"

l 'l

'" II
Fig. C2.17
Ee il
 C2.8 STRENGTH OF COLUMNS WITH STABLE CROSS· SECTIONS

particular ~ropertles of the material EO.7/F. Thus in Fig. C2.l8, ~he deflection of
paint (0) away from tangent at nidpoi~t c
Inser-t Ing va Iue of F • Tt~t/(L' /p)a in equals unity in our assumed conditions and i~
equation (3), also equals the first ~o~ent of :he area of
the M/E! diagra~ between (0) and (C) about (0).
Et ~ ( , , . ,'\ B' (Fig. C2.I9).
( n" Et/( L '/.0 ) a) -E-J= The value of the ordinate for M/EI diagr~n

or B =.!c
n
j ,',., (£)
E \ P
----(9)
at any point x rr-cn 0 is :1 sin 1"'! LX
The total area under ~he curve is,
The use or the curves in Fig. C2.17 will
be illustrated later 1n the example problem
sotut i ons .
P
area = -,I .!
.L

"
sin rXdx II [- * cos

C2.5 Strength of Columns With Variable Cross-Section

or Moment of Inertia.
= EI
P
-I[~ (- IIJ[- *(IIJl· ir (*+~)
To save weight 1n a built up column or hence area -_2PL 'o_PL
tEL and haL ar- "a - --:::;-;-
" ru"
forged column, the member 1s tapered or Is
made ~lth a non-uniform cross-section. To The center of gravity or the ~alf area is
find the ulttmate strength or such columns,
it Is usually necessary to use a trial and AX=/xda
error method. The general method of solution
involving a consideration of column deflection PL- oL/2 rt
Nill be illustrated for a case of a long (~Ol) x ~. /
= ':'.1. x sin L x ex ,
column with unlfor.n cross-section. "oW
o
Fig. C2.18 shows a pin ended column in a L/2
L -
deflected neutral equilibrium position when
carrj1ng the ultimate or critical load P.
-x
n =/
o
x sin "
LX dx

Assume that the shape of the deflected column

follows a sine curve relationship with the Integrating this Simple expression and solving
deflection at midpoint eqU?~l to unity (see for x we obta~n:
Fig. C2.IBI.
x = Lin
The ~t'Qn of the deflected column
-CfL"'n x '\
curve is y =\T . ; It ? is the end load, the Taking moments about point (0) of the M/E1
~-.--I diagram between 0 and C about 0:
bending moment at any point = M = Py = P
'1 sin az, hence,
" L
L'
By the well known "mament area" principle
(see Chapter A7; Art. A7.l4), the deflection which is the Euler equation, and ~hus the
ot a point (A) on the elastic curve away from assumed sine curve was :he proper one ~ur th~
a tangent to elastic curle (8) equals the first deflected elastic curve of the column.
moment of th~ M/EI diagram between (A) and (B)
about (A). Suppose that the elastic curve of de-

'
flected column had been assumed as a parabola
with unit deflection at ancpomt . Ff g . C2.20
shows the M/EI diag~~.

r
p

!
SIll.
C.UItVe:
The area or one-half the diagra~ =
1 2 P _ PL
2' ('3 LX,I) - 3EI
; [-~'I
L. I- C
Taking moments about 0 ot the area between 0

L
'~ Fig. C2.18 Fig. C2.19
and C;
5 5PL"
I ~ Yo --I=..£!:..
I 3EI 16 L = 48EI bence ,
'U
P
 ANALYSIS AND DESIGN OF
48EI 9.6 EI
P -= 5L a = L' which co~pares with P =
9. 9EI/L· of the Euler
equation or an error of
3 percent.
p are below the proportia~.al limit stress of the
Fig. C2.20
We can now apply the same procedure to
a colunn with non-unifonn cross-section. The
steps in this procedure for a column
symmetr'ical about the center point are as
follows:-
(1) Assume a sine curve tor the deflected
column with unit deflection a center
paint.
(2) Plot a moment of inertia (I) curve for
column cross-section.
(3) Find the bending moment curve due to end
load P times the lateral deflection.
(4) Divide these moment 'falues by the EI
values to obtain MlEI curve. The modulus
of elasticity E is considered constant.
(5) Find the deflected column curve due to
this M/EI loading.
(6) Campare the shape of the derived column
deflection curve with that originally
assumed as a sine curve. This can be
done by multiplying the computed de-
flections by a factor that makes the
center deflection equal to unity. Since
the assumed sine curve is not the true
column deflection curve, the computed
deflection will di~fer somewhat fram the
sine curve.
(7) With the computed deflection curve,
modified to give unity at center point,
repeat steps 3, 4, 5 and 6. The results
this time will show derived deflection
curve still closer to the assumed
deflection curve.
(8) To obtain th~ desired accuracy, the pro-
cedure in step (7) will usually have to
be repeated again.
FLIGHT VEHICLE STRUCTURES C2.9
(9) Solve for load P by writing and expression
tor the cerj.ec t t on at the center point
which equals unity. This is done by
using the moment area principle as was
done in the previous example problem in-
volving a column with uniform section.
In the above outlined procedure, E has
been assumed constant or, in other words, the
column failure is elastic or failing stresses
material. The practical problem usually
involves a slenderness ratio where failure is
due to inelastic bending and thus E Is not
constant. For this case, a trial and error
method of solution in necessary using the
tangent modulus of elasticity which varies
with stress in the inelastic stress range.
C2.6 Design Column CUrves for Columns with Non-
Uniform Cross-Section.
Figs. C2.2l and C2.22 give curves for
rapid solutIon of two types of stepped columns.
Figs. C2.23 and C2.24 gives curves tor the
rapid solution at two forma of tapered columns.
Use at these curves will be illustrated later
in this chapter.
C2. 7 Column Fixity Coefficients c for Use with Columna
with Elastic Side Restraints and Known End
Bending Restraint.
Figs. C2.25 and C2.26 give curves for
finding fiXity coefficient c for columns with
one and two elastic lateral restraints and
Fig. C2.27 gives curves for finding c when
restraining moments at column ends are known.
Use at these various curves wili be
illustrated later.
C2.8 Selection of Mater1al8 for Elevated Temperature
Conditions.
Light weight is an Unportant requirement
in aerospace structural design. For columns
that fail in the inelastic range of stresses,
a comparison of the FCY/W ratio or materials
gives a fairly good picture of the efficiency
of compreSSion members when SUbjected to
elevated temperature conditions. In this
ratio Fey is the yield stress at the particular
temperature and w is the weight per CU. inch
of the material. Fig. C2.28 shows a plot of
Fc /w for temperature ranges up to 600 0 F.
With 1/2 hour time exposure tor several im-
portant aerospace materials.
C2. 9 Example Problems.
PROBLEM 1.
Fig. C2.29 shows a forged (I) section
member 30 inches long, which is to be used as
 i
I

CIUTICAL WAOI:I-N()N·UHI~R"COLUMNS CRITICAL W"lllH<lOtHJNlfORM COLUWNll

Fig. C2.21 bJubl.lit!lPl'!'1 . PIA IlDded

Ca" fUl, (1:1:1,

'." .(j' ..... I' I H
f:-. -.I.-',---I.-.-j
.. I'
i. :11'1'".";;1;;':1"1

~~llJ
, '"' • l' I"
i " " :; ~ _. ..j .'
, .'
I~ ,i
t' . I t
>

•
1 • 't (
,I
"
••
I:il :ri, ,
" I
,
, ..
! ,n
III I, I
,
,
, .,.I , ,
,I ,I'!
'-1-, ,
,
,
. ..
.T

!iii ". ;;1

'j:
• , ..
"'>,j....
II. .. . .. .. .. '":J ..

CIUTICAL 1.o"OO-NON-lItllf'ORM ~ CRITICAL LOADS-NON-UNIFORM COLUMNli

...'"
COI"tonl T~,c1",e •• - "".pu 1" PloQ-rorn, Solid Priam. Tap"rloK in W\4UI """ ~
Solid eyliliden Taperlll( to Coo... ..... _ a(~)!
Fig. C2.24 >-
III
r-
"'
o
'o"
'"'"
I

~
.. "'...'"o
i..
II
I
I
'
,
.1
-
o
z
r.lIo '"
, i
•, I 'I '
. . .. I '.1

.. ...
• T !
I .. 1: .. ..
• ,
! I
II II
Fig. C2.23
(QlJlnla

Reference - N. American Avaiation Structures Manual

 SIMPLY SUPPORTED COLUMNS WITH ONE ELASTIC RESTRAINT
4.0.
3.8 q •
3.6 : 140
4.0
3.. K' '"
120
3.2
3.8 "
3.0 100
2.8
2.6 80
C 2.4
2.2
~....,
Iq
' 80
,.
Z

e"
2,0
1.8 40 i
"''"
1.8
1.' 20 ,.J ~

o
Z

i'if
2.8 ' ""j ,
1.2 . . . . .
! I ..,o
1.0 0.1 0.2 O.S 0.6
Fig. C2.25 x/L C 2. 8 1

2) , i I
';,l,'
8 I'
2.2,ill
I: i'

7
I
2.0 .' :

6 1.8
c
1.6

I'
2; 1.0
i i
1I I 1.' ,
o 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 o 10 20
Fig. C2.26 xlL Ref. Convair Structure Manu.al .l±.!o Fig. C2.2'l
EI

(j
C2.12 STRENGTH OF COLUMNS WITH STABLE CROSS~SECTIONS

Fig. C2.28 Solution:

(1) AISI Steel. F tu= 180, 000
Since the column may fail by bending about
(2) 17-7PH. stainless steel, F tu = 2l 0, 000 either the X or Y axes, the col~~ strength for
(3) 7075-76 Alum. Alloy bending about each of these axes Nill be cal~u­
(4) AZ31 B Magnesium Alloy lated. Since the column strength is a fUnction
of the radius of gyration of the cross-sectien,
(5) 6AL-4V Titanium Alloy the first step in the solution will be the cal-
cujat.Lon of Ix and I y, from which o x and Py can
be found.

Calculating Ix: In Fig. C2.30 the section will

be first considered a solid rectangle 205 x
2.75 and then the properties of portions (1)
a.~d (2) will be subtracted.

Ix (rectangle) ::: l~ x 205 X 2075 3 ::: 4.32

Portions (1) and (2)
----.;....;.:.........;.-_._._--
. ._-

Ix :I -l~ x 1.5 x 1.25 3 - 4( .625 x .25 x 1.292>l) :::

-1.29
(1 0 or (2) about its x centraidal axis is
negl1gi ble )
Ix = 4032 - 1.29 ::: 3.03 -in ;"

100 200 300 400

TEMP. of
500 600
Px =15. , Area A ::: 205x2.75-2x 075xl.25-
A
4x 025 x .625) = 4.375 sq.
a compression member. Find the ultimate
in.
strength of the member i t made fram the
following materials and subjected to the given Px = V3.03/4.375 = .83 in.
temperature and time conditions.
Calculation at I y:
Case 1. ~J1terial7079-T6 Alum. Alloy hand
forging and roam temperature. I y (solid) = 121 x 2.75 x 2.5' = 3.58
Case 2. Same as Case 1, but Subjected 1/2 Portion (1) = -(1.25 x .75x .875')2 -(1.25 x
hour to a temperature of 3000F. .75'/12)2 = - 1.52
Case 3. same as Case 2, but tor 600oF. Portion (2) = -(.25x .625x .833")4-4(.25x
1.25'/36) = - .488
Case 4. Material 17-4 PH stainless steel,
hand forging at room temperature. Iy = 3.58 - 1.52 - .488 = 1.58 in.'

Fig. C2.29 Py = V 1.5814.375 = .60'

Column strength is conSiderably influenced

[ ••
by the end restraint conditions. For failure
by bending about the x-x axis, the end restraint
against rotation is zero as the single fitting
bolt has an axis parallel to the x-x axis and
,Y thus c the fiXity coefficient is 1. For-
failure by bending about the y-y axis we have
end restraint which will depend on the rigidity
at the bolt and the adjacent fittir~ and
structure. For this example problem, this
restraint will be such as to make the end
fixity coefficient c = loS.

Fig. C2.30
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES C2.13

?or failure about x-x axiS,
L' : L/VC: 30/VT: 30, L'/Px: 30/.83
: 36
For failure about y-y axis,
L' : 30N'"T.5: 24.6, L' /p y : 24.6/ .60
: 41
Thus we make use ot the curves in Fig.
C2.17.
Case 1. Material 7079-T6 Alum. Alloy forging.
Table 81.1 of Chapter 81 s~rlzes certain
material properties. The properties needed
to use Fig. C2.l7 are the shape factor n, the
moduls Ec and ~he stress FQ.~o Referring to
Table Bl.l, we find that n = 26, Ec :
10,500,000 and Fo.~ = 59,500.

Therefore failure is critical for bending The horizontal scale in Fig. C2.17 in-
about y-y axiS, with trip = 41. volves the parameter,

Case 1. The material is 7079-T6 Alum. Alloy

hand forging. Fig. C2.l4 gives the failing (L' /p)
stress Fe for this material plotted against
the t'/.o ratio. Thus uefng L r /.0 = 41 and the Subs t t tut tng i-.
room temperature curve, we read Fe = 50500 psi.
Thus the falling load if P = FcA : 50500 x
4.375 = 220,000 lbs.

Case 2. USing the 300 0F curve in Fig. C2.14 Using Fig. C2.17 with 1.01 on bottom
for the same L'/.o value, we read Fc = 40,400, scale and projecting vertically upward to
=
and thus P 40,400 x 4.375 =
177,000. n =26 curve and then horizontal to scale at
left side of chart we read FcIFo.'l'_ = .842.
Case 3. USing ~he 600 0r curve, Fc reads 6100
and thus P : 6100 x 4.375 = 26700 lbs. Thus Then Fc : 59,500 x .842 = 50,100, as
subjecting this member to a temperature of compared to 50,500 in the previous solution
600 0F for 1/2 hour reduces its strength tram USing Fig. C2.14.
220,000 to 26,700 Ibs., which means that
Alum. Alloy is a poor material for carrying Case 2. From Table B1.1 for this material
loads under such temperatures since the subjected to a temperature of 60QoF for 1/2
reduction in strength is quite large. hour, we tind n = 29, Fc • 9,400,000 and
:ase 4. ~~terial 17-4 PH stainless steel Fo.'T =
46,500.
forging. Fig. C2.8 gives the column curves
for this material. For L' /.0 =
41 and using Then B : ~ j 9,~O:goOO (41) : .917
the roam temperature curve we read Fc =
135,200 and thus P = 135,200 x 4.375 =
591,000 rbs • From Fig. C2.17 tor B = .917 and n = 29,
we read FcIFoo'T = .88, thus Fc = 46,500 x .88
C2.10 Solution Without Using Column Curves. = 40,900 as compared to 40,400 in the previous
solution.
vnen primary bending rat iure occurs at
stresses above the proportional limit stress, EXAMPLE PROBLEM 2.
the failing stress is given by equation (5)
which is, Fig. C2.3l shows an extrUded (I) section.
A member composed of this section is 32 inches
Fe = n'Et/(L'/p)' long. The member is
braced laterally in the I_lt~l
Since Et 15 the tangent modulus of x direction, thus
elastiCity, it varies with Fc ' and thus the failure will occur by
relation at Et to Fc ~ust be known betore the
equation can be solved. To plot column curves
bending about x-x axis.
The ~ember 1s ?in ended 1'2 x _
it - - -1t
for all materials in their ~y manufactured
forms plus the various temperature conditions
would require several hundred individual
column charts. The use or such curves can be
and thus c = 1. The
material is 7075-T6
Extrusion. The problem
is to tind the failing
1
,
,
,
"
1
-1--
avoided if we know several values or parameters stress Fc under room
regarding the material as presented by temperature conditions. Figo C2.31
Ramsburg and Osgood and expanded by Cozzone
and Melcon (see Arts. C2.4 and C2.5) for use This (I) section correspondS to Section
in colunh~ design. 15 in Table A3.15 in Chapter A3. Reference
C2.14 STRENGTH OF COLUMNS WITH STABLE CROSS-SECTIONS
to this table gives,
A = .594 sq. in. Px = .618
L' = L, stn c = 1, L'/px
Fig. C2.1l gives the column curves for
this material. For L'/p = 51.7 and room
temperature we read Fc = 38,500 pSi.
Solution by using Fig. C2.17,
From Table Bl.l tor this material we tind
n == 16.6, Be :. 10,500,000 and F 7 = 72,000.
Then, B =~ / 10,~~~~go (51.7)
From Fig. C2.17 tor 8 = 1.36 and n = 16.6, we
read FC/F •. 7 = .537, hence Fc :: .537 x 72,000
= 38,600.
Consider the member is subjected to a
temperature at 4500F for 1/2 hour.
From Fig4 C2.11, Fc =21400
Us1ng F1g. 02.17:-
From Table 81.1, n :: 8.8,
and Fe/F 0 . 7 = 29,000.
B =1. j 2!l,000 (51.7)
n 7,800,000
From Fig. C2.17 we find FC/F 0
Then Fc :: .74 x 29,000 :: 21,450 psi
A very cammon aluminum alloy ln aircraft
construction is 2014-T6 extrusions. Let it
be required to determine the allowable stress
Fe for our member when ~de ot this material.
Since we have not presented column curves
for this material, we will use Flg. C2.l7.
From Table Bl.1, tor our material, we
find n ~ 18.5, Ec = 10,700,000 and Fa 7 =
53,000 . calculation will show is true as explained
Th n B
e
=.1n j 53,000
10,700,000 (51. 7) = 1.16"
From Fig. Cl.17 tor B 1.16 and n = = 18.5,
we read Fc/F o • 7 = .71, hence Fc = .71 x
53,000 =
37,600.
The result shows that the 2014-T6 materlal
gave a failing stress of 37,600 as compared to
38,900 tor the 7075-T6 material which has a
Fcy of 70,000 as compared to Fcy = 53,000 for
the 2014-T6 material. The reaSon for the
7075 material not showlng much higher column
failing stress Fc over that for the 2014 alloy
is due to the fact that the stress existing
under a L'/p value of 51.7 ls near the pro-
portional limit stress or Et is not much
dlfferent than Ec• the elastic modulus.
= 32/.618 = 51.7
To illustrate a situatlon where the 7075
material becomes more efficlent in co~parlson
to the 2014 alloy, let us assume that our
~ember has a rigid connection at its end whict
wlll develop an end restraint equivalent to a
fiXity coefficient c 2. =
Then L' = 3ziV2""= 22.6 and L' /p =
Q • 22.6/.618 =
36.7.
~ 1.36 For the 7075 material fram Fig. C2.1l, Fc =
58,300
For the 2014 materlal, we use Fig. C2.l7
_ 1 j 53,000 _
8 - n 10,700,000 (36.7) - .823.
From Fig. C2.17 tor B = .823 and n =
pSi 18.5, we read Fc /F 0 . 7 = .87, when Fc = .87 x
53,000 = 46,100 as compared to 58,300 for the
7075 material, thus 7075 materlal would permit
lighter weight of required structural materlal.
Ec = 7,800,000 The stude~t should realize that if the
stress range Is such as to make Et =
Ec' then
the bending failure is elastic lnstead of
= 1.00 inelastic and equation (5), USing Young's
modulus of elasticlty Ec' can be solved
= .74 directly Without resort to column curves or
a consideration of Et, since Et is equal to
• 7
Ec ·
The student should reallze tr~t equation
(5) ls for strength under primary column
failure due to bending as a whole and not due
to local bUCkling or crippling of the member
or by twisting tailure4 The SUbjec~ of
column design when local failure is involved
is covered in a later chapter.
In example problem 2, we have assumed
that local cripplir.g ls not critical, which
and covered in a later chapter.
C2.11 strength of Stepped Column.
The use of curves in Fig. C2.22 will be
illustrated by the solution for the strength
of two stepped columns in order to lllustrate
both elastic and inelastic tallure at such
columns.
Case 1. Elastic failure.
Flg. C2.32 shows a double stepped pin
ended column. The member is machined from a
1 inch dlameter extruded rod made from
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES C2.15

7075-T6 material. The problem is to find the This is a relatively short column so the
maximum compreSSive load this member will failing stress should fall in the inelastio
carry. range where E is not constant, therefore the
solution is a trial and error procedure. we
Portion 2 Portion 1 will base our first ~~ess or trial on an
i ~3/4" Dia. j +1"Dia. j Portion 2
J.
averagb Lip value.

p for portion 1 is 0.25 inches

p for portion 2 is 0.1875

Fig. C2.32 Average p = (6 x .25 + 6 x 0.1875)/12

= 0.22
PORTION 1 PORTION 2
Then Lip = 12/0.22 = 54.5, use 55.
= .7854 in. II All= .4418
= .0491 in .... I. = .0155 Fig. C2.11 is a column curve for 7075-T6
= 10,500,000 Ec = 10,500,000 Alum. Alloy extruded material. With LIP =
55, we read allowable stress Fc = 33,500 psi.
From Fig. C2.22) Pcr = B(EI,) . This is Therefore
L'
the Euler equation for failure under elastic P = Fc A : 33,500 x 0.7854 = 26,300 lb.
bending. It the ratio aiL equals 1 or a
uniform section, B becomes n~ or 10 as shown f l : 33,500 and til = 26,3001.4418
in Fig. C2.22. The curves in Fig. C2.22 = 59,500.
apply only to elastic failure. Since the
member in Fig. C2.32 is rather slender we The stress t in portion 2 is above the
will ass~e the failure is elastic and then proportional llmit stress so a plasticity
check this assumption. correction must be made in USing the curves
in Fig. 02.22.
Ell 10,500,000 x .0491
El a = 10,500,000 x .0155 = 3.17,
aIL = 30/60 Referring to Table 81.1 in Chapter Bl,
= 0.5 we find the following values for 7076-T6
extrusions:- n = 16.6, F o • T = 72,000,
From Fig. C2.22 for aiL = 0.5 and EIl/EI. Eo = 10,500,000.
=3.17 we read B 7.0 =
The tangent modulus Et will be found for
'...' hence, Per =7 x 10,500,000 x .0491 = the stresses f 1 and f ll •
60 •
For Portion 1, fl/F o . ? = 33,500/72,000
1000 lb. = .466

The stresses in each portion are, Referring to Fig. C2.16 and using 0.465
and n = 16.6, we read Et/E = 1.0, thus Et =
t, = 1000/0.7854 = 1280 psi E and thus no plastiCity correction for
Portion L
t. = 1000/.4418 = 2270 psi
For Portion 2, f./F o . ? =59,500/72,000
These compressive stresses are beloN the = .826
proportional limit stress of the material so
Ec 15 constant and our solution is correct. From Fig. C2.l6, we obtain Et/E = .675
whence, ~t = .675 x 10,500,000 = 7,090,000.
Case 2. Inelastic Failure.
EL 10,500,000 x .0491
The column has been shortened to the El a = 7,090,000 x .0155 = 4.7
dimensions as shown in Fig. C2.33. The
diameters and material remain the same as in
Case 1.
From Fig. C2.22 tor aiL = .5, we obtain
Portion 1
8 = 5.6.
J

Then Pcr = BEll. 5.6 X 10,500,000 x .0491

= ----'------'----
r-""3" -+--- .:6" _b.3·'1 L' 144
I·t. L=12" ------>1 : 20,000 lb.
Fig. C2.33
C2.16 STRENGTH OF COLUMNS WITH STABLE CROSS-SECTIONS

OUr guessed strength was 25,300 lb. Our C 1-1/4 -.058 A 1-l/4·. 058 B 1-1/4 •. 058 D
guessed strength and calculated strength must <
be the same so we must try again. I
Trial 2. Assume a critical load P ~
23500 lb. I
t, = 235001.7854 29900 * E F
I+-- 30" - _......-
)1;
t. 23500/.4418 ~ 53100 Fig. C2.34

Portion 1. f /F o • T = 29900/72000 = .415 The member AS is welded to th~ee adjacent

tubes at joints (A) and (B). Since ~hese
From Fig. C2.16 ~or n = 16.6, we read tubes are the s~~e at joints (A) anj (a), ~he
!:tIE = 1.0. fixity at the ends (A) and (B) of member AS
is the sane.
Portion 2. fa/F o • T = 53100/72000 = .738
Referring to Ff g , C2.27, the tern ~ is
From Fig. C2.16, EtlE = .90, whence detined as the bending restraint coe!:icien~­
Et = .90 x 10,500,000 = 9,450,000. spring constant expressed as inch pO~dS ~er
radian. In Fig. A,
EI, 10,500,000 x .0491 the moment M re-
quf red to rotate E1 (constant) 8
EI a 9,450,000 x .0155
end (A) through 1
From Fig. C2.22 tor ail = .5, we read radian when tar end
B = 6.62. is fixed is 4EI/L. Fig. A
For derivation of
x 10,500,000 x .0491 = this value refer to Art. All.4 of Cr~pter
The Per = BEll.
6.62
All. It the far end (3) is pinned in
L' 144 (Fig. A), a moment M = 3EI/L will rotate end
(A) through one radian. To je slightly
23,650 lb.
conservative, we will assume the far ends of
This practically checks the assumed value, members coming into joints (A) and (B) as
thus the answer is between 23,500 and 23,650 pinned. Thus ~ = 3EI/L. The s~~ of ~ ~
and it further accuracy 1s desired another 3EI/L will be camputed for the 3 members
trial should be carried through. which form the support ot member- AB at end
(A) •
The other types ot columns With non-
uniform cross-sections as shown in Figs. M:ember AC:- I = .03867, IlL = .001289
C2.21, C2.Z3 and C2.24 are solved in a Similar Member AE:- I =
.02775, IlL .00071 =
manner. These charts are to be used only with Member AF:- I =
.02402, IlL .000962 =
pin ended columns. The end fiXity coefficient
c for tapered columns is not the S~.e as for ~ = Z 3EI/L
uniform section columns.
= 3( .001289 + .00071 + .000962) 29,000,000
C2.12 Column Strength With Known End Restraining
Moment.
I..l. ::: 258,000

.Fig. C2.27 shows curves tor ~lndlng the
end fiXity coefficient c for two conditions
of known end bending restraint.
To illustrate the use or these curres,
a simple problem will be solved.
Fig. C2.34 shows a 3-bay welded steel
tubular truss. The problem 1s to deter.nine
the allowable cQmpressive streSS for member
AB. This strength is influenced by the fixity
eXisting at ends A and E. The diameter and
wall thickness of each tube in the truss is
shown on the figure. The material is AlSI
steel, Ft u : 90,000, Fty = 70,000, E =
29,000,000.
In Fig. C2.27 we need term I..l. L/~I. The
L/El refers to ~ember AB. Thus ~ LIEl =
(258,000 x 30)/29,000,000 x .0367 = 7.28.
~e use the upper curve in Fig. C2.27
since restraint at both ends of member AB
is the same. Ttus tor I..l. LIEl = 7.28, we
read end fiXity coefficient c = 2.58.
Then L' = Live = 30;'; 2.58 = 18.5.
o for member = .422 inches.
L'lp = 18.6/.422 = 44.0
From column curve in F1g. C2.3, we read
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES C2.17

allo·Nable failing st~ess to be Fe = 55,200 C2. 14 Problems.

psi.
(1) 6061-T6 Al~lnum Alloy sheet, heat-treated
If the far ends of the connecting members and aged has the following ~r~perties:
were assumed r txed instead of Pinned, then
~ = 4EI/L, or we ca.~ multiply previous value (a) Under room temperature:- F O • 7 =
of 7.28 by 4/3, which gives 9.7 whiCh, used in 35,000 pst, Ee = 10,100,000 pst, and
rig. C2.27, gives c = 2.80. n = 31

L! /p = 30/v'2':'8 x .422 = 42.5. Then from (b) For 1/2 exposure at 300 oF:- F O • 7 =
Fig. C2.3, Fe = 56,600 psi. Since the far 29,000, Eo : 9,500,000 and n = 26.
ends are less than fixed, the assumption that
far ends are pinned gives fairly accurate For the above two cases (a) and (b),
results. determine Et (tangent modulus values) n-om Fig.
C2.l6 and then calculate and plot column curves
In a truss structure all members are for these 2 material conditions.
carrying axial loads and axial loads effect
the ability of ~embers to resist rotation of (2) Fig. C2.36 shows
their ends. Art. A11.l2 of Chapter All the cross-section ,"t
explains how to take account of the effect of a compression i
of axial load upon the stiffness of a member member. Calculate
as required in calculating the end restraint the failing com-
coefficient ~. pressive load under
the following
C2. 13 Columns With Elastic Lateral Supports. cases:-

Figs. C2.25 ~~d C2.26 provide curves for Case 1. L = 25 inches.

finding the end fixity coefficient c to take Material AISI Steel 4140, I-z-l
)'1
care of elastic lateral supports at pOints Ft u = 180,000. Take end
~idway between the column ends. fiXity coefficient c = 1 Fig. C2.36
for bending about x-x
To illustrate the use of these charts, a axis and 1.5 about axis y-y.
round bar 0.5 inches in diameter and 24 inches
long is braced laterally as shown in Fig. (3) Same as Problem (2) but member is sub_
C2.35. The bar is made of jected to a temperature of 850°F tor
AISI Steel, heat treated to 1/2 hour.
Ftu = 125,000. The spring
constant for the lateral (4) ~NO extruded channel sect tons identical
support is 775 lbs. per to Section No. 50 in Table A3.1l in
inch. ~ L=24 Chapter A3, are riveted back to back
and used as a column member. If member
Moment of Inertia of 1/2
:rod = .003068, ±1 is 26 inches long and end fixity is
C = 1 and material is 7075-T6 extrusion,
what is the tailing compressive load.
Radius of Gyration =
.125 inches. Fig. C2.35 If member is fastened rigidly to adjacent
structure which ?rovides a fiXity c 2, =
775 X 24;5 what will be the failing load.
120
29,000,000 x .003068
(5) Consider the column in proble~ (4) is
?rom Fig. C2.25 for x/L = 10/24 = .416 ~de from 2014-T6 Aluminum Alloy extrusion.
and q =120, we find c 2.92. = Find tailing load.

Then L' : L/ve: 24/1"2."92: 14.08 The pin anced single stepped column as
shown in Fig. C2.37 is made Qf AISI-4130
L' /0 : 14.08/0.125 = 113 no~allze1 steel, Ftu = 90,000, Fcy =
70,000. Determine the ~tmum compressive
Fe
3
E
= (L'n /p)3
3
:: n x 29,000,000 = 22,500 load member will carry.
(113)3 t:>-psi 1" Sq. ~r

If the stress is above the proportional

l~lt stress for the naterial, then the trial ~ a = 21" '_1'_9_"=1
and er~or approach ~us~ be used as i1lust~ated f.---- L " 30" --.I
in the problem dealing with a tapered column. Fig. C2.37
C2.18 STRENGTH OF COLUMNS WITH STABLE CROSS-SECTIONS

(7) Same as Problem (5) but member Is exposed
1/2 hour to a temperature of SOOoF.
(8) Same as Problem (5) but change dimension
(a) to 10 inches, and L to 14.28 inches.
Find the failing compressive load for
the doubly stepped column in Fig. C2.38
if member is made from 7079-T6 hand
forging.
(11) The cylindrical tapered member in Fig.
C2.39 15 used as a compression member.
If member Is made :Tom AISI Steel 4140,
Ftu = 125,000, what Is the failing load.
L
-+- ae 12 -----oj<- b·9
30"
:I
Fig. C2.39

(12) Same as Problem (7) but change dimensions

to a = 6", b ::; 4.0", L = 14 inches.
References:
FIg. C2.38 (1) NACA Technical Note 902.
(2) Non-dimensional Burkllng Curves, by
(10) Same as Problem (7) but Change dimensions Cozzone Z Melcon, Jr. of Aero. Sciences,
=
a. 6 ft , b 4", L = 14 ft • October, 1946.
(3) Chart from Lockheed Aircraft Structures
Manual.
 CHAPTER C3
YIELD AND ULTIMATE STRENGTH IN BENDING

C3. 1 Introduction. compreSSion. In this example solution, we will

find the internal resisting moment when we
Members subjected to bending alone or in limit the unit strain at the extreme edge on
combination with axial and torsional loads are the compreSSive side of the beam section to
quite co~on in flight vehicle structures. 0.010. Now plane sections remain plane after
The lImit design loads on a structural member bending in both elastiC and inelastic stress
must be carried without permanent distortion conditions when member is in pure bending.
and the ultimate deSign loads must be carried We will guess the neutral axis as located
without rupture or failure. The well known 0.0375 inches above the centroidal axis as
bending stress e~~tlon tb = Me/I, assumes a shown in Fig. C3.1b. Having assumed the
linear variation ot stress with strain or, in max1mum unit strain as .010, we can draw the
other words, the equation holds tor stresses strain diagram at Fig. C3.lb. We now divide
below the proportional lImit stress or, in the cross-section in Fig. C3.1a into 20
general, the elastic range. Failure ot a narrow horizontal strips. Having the strain
~e~ber in bending, unless there Is local curve in Fig. C3.lb, we can tind the unit
weakness, does not occur at stresses in the
elastic range but occurs at stresses in the
inelastic range. Since the Ultimate strength
of a member is needed to compare against the
ultimate deSign load to be carried, a theory
or procedure Is necessary which will accurately
determine the ultimate and yield strength of a
~ember in bending.

C3.2 Basic Approach to Finding the Bending strength

of Members.

The problem is to determine the internal

reSisting moment of a beam section when sub-
jected to stresses which fall in the inelastic
range of stresses. This stress can be taken
as the ultimate tensile or compreSSive stress
ot the ~terlal or limited to some stress or
deformation in the inelastic range. To obtain
the true internal resisting moment, we ~ust
know how the normal tension and compreSSive
stress varies over the cross-section. The
stress-strain c~-ve far the material provides
the sa~rce for ob~aining the true stress
pic~ure. I: a nate rial has a different shape
in the tensile and compressive inelastic .01 .008 .006 .004 .002 0 .002 .004 .006 .008
zones, the neu~ral axes does not coincide with
the centroid~l axis, thus adding some difficult
to an analySis 4ethod. The analysis procedure (al Fig. C3. 1
for deter.Jining the true internal resisting TeNSION
~oment is best explained by an exa~ple
solution.
C3. 3 Bendlng Strength of a Solid Round Bar.

~ig. C3.1a shows the cross-section of a

round solid bar ~ade of alQninum alloy. The (e)
stress-strain curve u~ to a unit strain of
.010 in. per inch is given in Fig. C3.2. Note
that the shape of the curve in the inelastic
zone 15 not the saue for both tension and UHIT

C3.1 "'AI"

~ - ----.--- -
-7'r:.~ ~. ~,: .~;_:' ';. ;" ,/;~. ;.~:,~~-,> ~~~,~:,;~' . ~. :</:/~<:.~~£~ :,~/'.:,'~ ~.;.~~:::-;.- _":. :1~.:i: ~= ';-"1,(" ~. ~,~" ~ ,:~>~ -' ; ..-, ~<~~~~~~:~~~:~:~~;t~~.?~f- :;,~~;~.
 C3 2 YIELD AND ULTIMATE STRENGTH IN BENDING

strain at the -ncoo i nt of each str-t». .c t th the
strain known on each strip, the stress existing
can be !ound by use of the stress-strain curve
in Fig. C3.2. The total loae on each str1;J
then equals the stress times the area of the
strip. The internal resisting ~o=ent then
equals the s~tlon of the load on each
strip times the distance from the strip to
the neutral axis.
Table C3.1 shows the detail ~alculatlons.
If the neutral axis has been selected in the
correct position, the values in col~~ (6)
ot the table should add up to zero since total
tension must equal the t~tal compression on
the beam cross-section. The small discrepancy
ot 740 poundS in the summation of column (6)
1s not enough to change the location of the
neutral axis or the total internal resisting
moment appreciably. Column (7) gives the
total internal resisting moment as 56735 in.
Ibs. when the strain is limited to the .010
strain as previously discussed. The stress
at this strain tro~ Fig. C3.2 is 49000 psi.
USing this stress in the well known beam
=
fonnula M rr/e , we obtain M: = 49000 x 0.785
= 38450, which is much less than the true
value or 56735.
TABLE C3.l
1 2 3 4
Strlp
Slrlp
No.
Area y E
.. "A"
.058 0.935
1 .00867
2 .102 0.840 .00773
3 .135 0.75 .00685
4 .153 0.65 .00591
5 .165 0.55 .00494
given bending ~oment, and al~J ~J~ teter~lnir.g
t~e true internal resisting ~:~e;.t Jf a bea~
section, structural deSign 2~gi~eers ~ake ~se
o~ a fictitious failing ben~ir.s stress ?o'
wtn cn is re re n-ec to as a cccujus 0:' ructu-e
stress in pUTS bendin~, T~3n the ultimate
bending ~om3nt that can be developed jy a
given beam cross-secticn ~~d a ~lven ~aterial
is M = FbI/C. DeSign cu:ves fer finding ro'
the modulus ot rupture, are given later in
this chapter.
Since there are ~ny flight vehicle
materials and all kinds ot SD2Des used in
structural members, the basic approach for
solution as illustrated in Table C3.1 becomes
very time consuming. DeSign engineers always
search for simplified methods which give
sufficient accuracy. Thus Cozzone (Ref. 1)
has developed a Simplified procedure for
finding the modulus ot yield or rupturing
bending stress Fb' The method is Nicely used
in t~e aerospace industry in structural
deSign.
C3.4 The Cozzone Simplified Procedure.
The Cozzone method in its si~ple5t fern
assunes a symmetrical rectangUlar bea~ section
and the same shape of the stress-strain curle
5 5 7 in both tension and compression. ?1;. C3.3
represents the true bending stress '~riat1cn
Unit Res. over the bea~ cross-section when failure
Stress F :crA Moment occurs. Cozzone now replaces this tru~ curve
CO M= Fr by a trapizoidal stress variation as sh0wn :n
3075 2760
Fig. C3.4. The stress f o is a fictitious
53000
52500 5350 4300 stress which is assumed to exist at the neutral
52100 7025 5040 axis or at zero strain.
51500 7870 4820
51000 8410 4310 Fig. C3.3 Fig. C3.4 1_ Im _!

..[]IbT
6 .180 0.45 .00398 43000 7740 3200 m
[tMAX=fml~-oI~fb'"
7
8
9
10
.185
.195
.197
.200
0.35
0.25
0.15
0.05
.00302
.00205
.00108
.00012
33200
22800
12500
3200
6140
4450
2460
640
1920
945
280
10
[2
-- --- N·A.lO
IMAX=f

f
o
r i ;
L
11 .200 -0.05 -.00084 - 7250 -1450 130
12 .197 -0.15 -.00181 ~ 17800 -3510 680
13 .195 -0.25 -.00276 -29500 -5750 1650
14 .185 -0.35 -.00374 -3'5500 -6560 2540 fMAX fMAX N. A. ""-
-40000 -7200 3510 True Assumed Assumed
15 .180 -0.45 -.00470 Stress Stress Stress
16 .165 ~0.55 -', 00566 -43000 -7100 4170
17 .153 -0.65 -.00663 -44800 -6850 4710
Fig. C3.6
18 .135 -0.75 -.00759 -46000 -6210 4880
19 .102 -0.84 -.00846 -47200 -4810 4210 f MAX[ ---;-- ~
20 .058 -0.935 -.00937 -48000 -2780 2690
Total 3.140 740 56735 ~ I .....-// I
Col.
Col.
1
3
Rod divided into 20 strips. 1" thick.
y ,. distance from centerline to strip c. g.
-
•
OS
?
--- /
~Assumed Stress- I
I

Col. 4 E ,. strain at midpotnt '" (y ~ .0375)/103.75. Strain Curve I

Col. 5 Unit stress for E strain from Fig. C3. 2.
Col. 6 Total stress on strip. i
Col. 7 Moment about neutral axis. r = (y - .0375).
True Stress-Strain II

Curve

Since it is desirable to use the bean Strain e

!'ormula In !1nding bending stresses due to a Fig. C3. 5

-- -
_" .:~~).r,[.:~':'-~:'-"":{-~~;,~~;~'-~'.~1,:~~;'~·(~·~~~~~::":~:f~~,·L~'::··+;·~'~~;''''''5~';·'~:i··;t''':~;';~~~I.:--.~",,:~- -,. ';"~+ -~;/~~,', r-"-:"...~., '."~,'. ~,~v
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES C3.3

The value of f O is determined by ~king

the r~quirement that the internal ~o~ent of
the t~~e stress system must equal the moment
of the assumed tra~9zoidal stress system
Fig. C3.7

-I- - ,
,
V-
k Factor for Some Typical Shapes

,.
I.

u
,.,
,
Z~

which results from the assumed stress-strain -K- ./' -0-

curve as stown in rig. C3.5. Fig. C3.6 shows
the trapez~ldal stress pattern drawn to a ~~~- .,
r.s -0- ,
-0- ..,
larger scale and showing only one half of the
sy~~trial ba~~ section. Thp, trapeZOidal
stress pattern has been divided into a

--_ .-
rectangle (r) and a triangle (~) as shown in
tne r tgur-e , _.-1-
......
Let, Mb = total lnternal resisting moment.
internal m·y,nent developed by
-
_--
.._........
.. _-_.. -
,,,,._-

.._-
mr h',1
portion (r). ','rrrr-rrrrrrrrrrrrrrrrl
mb = internal mom~nt develooed by
portion (b).

. '
.. . r--
,

Sin:e fb varies linearly from zero to tb' .

"
.'J

the stress is elastic and thus the beam

equation holdS, or
I
mb : fbIlc for entire bea~ section. The
stress variation on porticn (r; is constant or
•.. 00 ..•
rectangular, thus
F i g . § Section Fa.ctor K for r, Cl, and C Sections (Ref. 2)

C3. 5 Deetgn Curves for Finding Modulus of

T~en m~ = ro2~ for entire bea~ section Rupture (>D).

But fb = f m - f o (from Fig. C3.6 ) The modulus or rupture Fb may be a yield

modulus, that is, in equation (3) the value at
Thus, lib = (fm-f a) 1..
c
2f oQ, or f m is equal to the ~~eld stress of the material.
It ~y also be the ultimate modulus of rupture,
Moe in which case the value at t m in equation (3)
I
2(1
f m • f o (I/e - 1) ---- --- (1) equals the ultimate strength or the material.
The modulus at rupture may be limited to a stress
between the yield and ultimate stress of the
- 2~
Let k -VC ----(2) material because at local cripoling or by
excessive distortion. Regardless of what value
is used for f m in equation (3), the corres-
k is a beam section shape factor.
ponding value at fa must be known before the
value or Fb can be determined. Figs. C3.9 to
Let Fb = Mbe/I, then from equation (1) C3.23 give strain curves tor various material
and the corresponding fa curve. The use or
Fb = f!D. + fa (k - 1) - - - - (3 )
these two curves per.nit the determination at
Fb if the k shape factor for the particular
Fb is a fictitious Mc/I stress or the
be~ section being considered is knO\VTI. In
mQ~ulus of rJpture for a particular cross-
deriving the valueS of fa, the following
sec~lon at a ~iven ~im~~ stress level. assum,tions are made.
The values at k vary between 1 and 2.e. (1) The stress-strain curve is assumed
If calculatea value of ~ is greater than 2
use 2.0. Fig. C3.7 shows the valu~ of the the same in tension and compression.
sh~De factor k for several typical shapes.
Ff g . C3.8 snows curves for the r-atn d
(2) The neutral axis is assumed to coin-
cide with the centroidal axis.
detenoination of the k factor for 3 c~on
beam sections. (3) Duri~g bending plane sections remain
plane.

(4) The cross-secticn is not subject to

local or torsiQnal instability.
C3.4 YIELD AND ULTIMATE STRENGTH IN BENDING

(5) Beam-column, curvature and shear lag
effects are considered negligible.
C3. 6 General Accuracy of Method.
(1) It 1s exact for a rectangular section
under pure bending with moment vector parallel
to a principal axis.
(2) For double s~etric sections under
pure bending and moment vector parallel to a
prinCipal axis, the accuracy should be within
5 percent ..
(3) Single symmetric sections will vary
tram practically exact to definitely uncon-
servative (moment vector normal to axis ot
symetry) .
(4) For sections subject to combined
bending and axial load, the results will vary
tram practically exact to conservative.
(5) For unsym:netr1cal bending, with and
without axial load, the results will vary from
practically exact to conserlative.
C3. 7 Example Problems 1n Finding Bending strength.
EXAMPLE PROBLEM 1.
to give the value of fa. The value of ~o fro:n.
this chart o~eration ~ives f o 2 29000. T!':er.
from equation (3)
FbYield = 65000 + 29000 (1.5 - 1) = 79500 p~i.
Then yield bending moment = Myp = Fb1/c
Thus Myp = 79500 x .0938 = 7460 In. lb.
For finding the ultimate resisting
bending, we use Ftu which is 75000 as the
value of f m in equation (3). Ag~in going to
Fig. C3.17 to stress of 75000 on stress-st~ain
curve and the vertically dOem to to curve, we
obtain f o = 70500.
Then F b (Ul t )· 75000 + 70500(1.5-1)= 110250 pSi
Then MUlt ': Fbllc = 110250x.0938 = 10370 in.lb.
Let us assume that 1s 1s desired to limit
the strain in the extreme tiber ~o .03 lnc~es
per inch. What would be the bending ~oment
developed under this limitation.
From Fig. C3.17 for a unit strain of .03
the corresponding stress from the stress
strain curve is 74700 and the f o stress 1s
61200.

A rectangular bean section is 0.25 inches Then Fb = 74700 + 61200 (1.5-1) = 105300
wide and 1.5 inches deep. What yield and
ultimate bending moment will the section Then M = FbIle = 105300 x .0938 = 9900 In. lb.
develop when made trem 7075-T6 extruded
alum1nu:n alloy. EXAMPLE PROBLEM 2.

Solution: The modulus ot bending stress is The symmetrical I beam section in Fig.
given by equation (3), (a) 1s subjected to an ultimate design pure
bending moment M': 14000 in. lb. ~h~t Is
Fb = t m + to (k - 1) - - - - - - (3) the margin of safety if
the beam is made' of
k _ 2Q. - 2 x 0.. 75 x .25 x .375 x = .1408 magnes i urn forging AZ6lA
.0938 which has Ft u = 38000
1.375 125
. . ..fI*.
2!JEl
- DC - (.l.) .25 x 1.5'(1/.75)
12 and Fty = 22000.
k = 1.50 Ix = 1~ x 1.375 x 2 3 _
1 .125
The value of k could also be found in Fig. 1
C3.7. 12 x 1.25 x 1. 75 3 Fig. (a)

Material is 7075-T6 aluminum alloy. From = .916 - .558 = .358.

Fig. C3.17, Ftu = 75000, Fty = 65000.
Ix/c = .358/1 = 0.358
To tind the yield bending strength, the
value at t m in equation (3), the maximum Q = 1.375 x .125 x .9375 + .875 X .125 X
stress.permitted on the most remote fiber is .4375 = 0.209
65000, the yield stress of the material. To
find to, we go to Fig. C3.17 and find the k =~ = 2 X .209 1.17
pof rrt on the stress-strain curve that Ix/e 0.358
corresponds to a stress of 65000. This paint
1s proJected vertically downward to intersect k could have been ~rom Fig. C3.8 when
the curve f o• This paint is then proJected using m = .125/1.375 = .091, and n = .125/
to the streSS scale at the edge of the chart 1.75 = .0715.
 ANALYSIS AND DESIGN OF FLIGHT VEHICt.E STRUCTURES C3.5

From Fig. C3.l9, for t m = Ftu = 38000 we
tind in projecting vertically downward to to
curve gives r 0 = 23700. Then subt. In
equatlon (3)
Fb = 38000 + 23700 (1.17 - 1) = 40770
Mult = FbI/c = 40770 x .358 : 15000 in. lb.
llar.ln of Safety = M"M1t _ 1 = 15000 _ 1
• 14000
Assume we desire the stress inte~sity at
a point 0.5 inches trom neutral axis it full
bendIng strength was developed. From Fig. /
C3.l9, the unit strain when stress Is 38000
15 0.35. Then sInce plane sectIons remaIn
plane atter bending, the unit straIn at· point
.50 lnch from neutral axls 15 (.5/1) (.035) =
.0175. From Fig. C3.l9, the stress existing
at thIs strain is f = 31000 psi. A linear
variation ot stress as used in the flexural
equation would give halt the maxImum stress
= -'-(J7 or 38000/2 = 19000 psi as agaInst the true
stress at 31000.
.
50

e ..
A

10

0.112 0.0"
e ""lA
0.01 ,."
0.02 0.0" 0.011 0.08 0.10 0.12

""'lA

.. ~-=~@i:j~i4io:~£"" '-';:'::'':-o:;::~
,::-:2:~t;-'·:·~!;~:",,,~:f~~i;~'I~:;::t··· -,:··":':;:"-'=if::~~,,.='000::'::..:
E~;_:,~·.-'i:;g~":;if~?::;~:~:". ,~::.:'.::: :~'f:..- -==....:=::-::. ~-:-:d~'::i::2:J.~l:~""
'; ...~.:l
"s s
- .,:"-"01 .:..~.::-t:~~;E~E:~ii~E';ii,--'=-'::E~-§§:"o;§f~¥':'~~"=r_~"'::;,":''':-;-':
;~:" ,0-::::;i~:Ol;';:.~::f0~g'~;::·:~'-:.,~q~,i~~E"5'fE2f~ ~~f.#~",~'o-'
,." '.N 0.011 0.08 0.10 0.11 o 0.01 0.02 0.03 0.04 0.05
, ""lA e 10/. .

r=:.::~8:~'f++~~~i~::;~~;::r-~:tn::':J=~:':~:~ttt~ ·"~,T:;:§j~~·t:r ~tF1

_._._-----.!~~~~j:_~'E:i~".,.-

._....:~,~_.__ .i_ic~L~~!_~:,Al.1ar~~,PIaIe'~
. ..:..r:so-~ : ""::"""'"'- _
-c-=---c,---,.--.-:---:,:-~;~,.,.~==~~= . . ·-'-,=Cjii.f~
-----------_._.~-

0.02 0.0 1 0.08

e ""lA
0.08 0.10 0.12
0.0' 0.0" 0.01 0.08
, ""lA
0.10 0.12

Curves tor finding Fb' Fb = tm + f o (k - 1).

C3.6 YIELD AND ULTIMATE STRENGTH IN BENDING

;::~'.~·i·:~~"''';:,:b;~,".t<'':\~.« ,:.,: ;, j
10fE:o:"'~-' -':,~::~::::.~~~.:.:.=~~.~~_.

0.01 0.02 0.02 0.04

....
0.05 0.055
'""~.-

-.-.
o.~.;·~~.c~t~-=~;~
0.01 u.oa 0.03 0.04 0.05
e WID W"

~~-,-oo;:~...,..:.;,:.:~'----'---,---o

~;U!imi",,' Hljnd'''''''''hlliri
. :';.;." ..::.:,. :':;.u.•. Sl$'lD'-;'-"'; ., ~ ~I '
-~--...,.~--,-,---c-;;- -" ,,'~'" .""";",-,.~;-~''"~'~'-'~---.

-::~~~~-~,::::·.:~=~t~;;~=~
-,,--. -

0.01 O.OS 0.03 0." 0.0$

, WID 0.01 0.02

rJ~~;~L;'7:,':--;'2:~;':";---.
,'+riL.~~J,',_
.... . ..,
,<0
- -"-"~'?:"-~--?"'=--"-----"-".- t1Iia.-~k;'. r·~-
--" --- .....---.- ----~." ~

';
".-
.
J
<0

0.02 0.03 0.008 O.012 0.018 0.020 0.024 0.028 0.032 0.035
, W .. I':. \lDcbel/lnl:/ll

Curves tor finding Foo

 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES C3.7

portion on each side ot the neutral axis ot the

---- _·_--_·~-:::--~:.:;::.:_·;:;:;_·_-C-= = ~ = = =
._.. _-----_._~",-,-.:....:.
entire section.
.. --_.--"...:..: Label the beam portion below the neutral
axis as (1) and that above by (2) •
__ .-~._,---=--

.....::...-'=~ __ _ 'C--,_·-., ;,
Portion 1. Fig. (d) shows
how the lower portion (1) ~ 1-0.1"
!.' is made a symmetrical T ," , T
section about neutral axes 278,,1,1.39
by adding the dashed
portion. The internal -~r-u''-{--N.A.
1.39
bending resistance will be _ J:. \
tound tor this entire
Fig. (d)
section in Fig. (d). one
halt ot this amount will
then be the true moment developed by portion (1).
E:CAMPLE PROBLEM 3. Unsymmetrical Section.
I =121 bh ~ _ 1
- 12 X 0.1 X 2.78
a
= 0.178
Fig. (b) shows a tee beam section,
symmetrical about the vertical axis. It the
material is 17-4 PH stainless steel, what
1,/c, = 0.178/1.39 = .128
ultimate bending moment will be developed it Q, = 1.39 X 0.1 X .695 = .0965
bottom portion is the tension tlange.
2Q., = .193
--<..1_1-1/2 ---+I Ell"'. 0154
---- = 0.193 = 1 5
i Lo. 0.128 •
I"
IN.A. --
0"
----1------
iO.610
From Fig. 03.22, Ftu = 180000 which equals
y=1.39 to trom curve = 156000
1____ ___l
o. 1"~ I....
.~
E1. =.03
Then Fb, =t m + to (k - 1)
Fig. (b) Fig. (e) = 180000 + 156000 (1.5 - 1) =258000
M • (Fb, 1Jc, It =258000 X .12B X
The neutral axis will tirst be determined.
0.5 = 16550 in. lb.

The factor 1/2 is due to the tact that

portion (1) is only one halt the beam section
Fig. (c) shows the unit strain picture. in Fig. (dl.
The lower edge at the be~~ section is strained
to the maximum value at .035 as shown on the Portion 2. Fig. (e) shows the developed
stress-strain curve in Fig. C3.31. Since symmetrical section tor the upper portion (2)
plane sections remain plane the unit strain at the beam section.
E2 at the upper edge ot section is E. = .035 x
.61/1.39 = .0154. , ~l.s--->I e =.0154
Solution 1
o·0.tT'lC
+ -1
N.A.~, - - - - 1.22 ~S[:~~
Equation (3) was'derived tor a symmetrical o. -61
--M-I"5"4-
I I
section about the neutral axis. The equation .3- C::==_::-:::::l _
involves finding the resisting ~oment Fig. (e)
Fig. (I)
developed by one halt the beam section and
~ultiplying by 2. This is permissible since
the unit strain at both top and bottom edges The unit strain picture Is shown in Fig.(f)
is the same. In this solution we will continue
to use equation (3). To do this it Is
necessary to ~ke symmetrical sections for the III = l~ x 1.5 x 1.22 3 - l~ x 1.4 x 1.02a = .104

3Cf.-
C3.8 YIELD AND ULTIMATE STRENGTH IN BENDING

I./e, = .104/.61 = .1704 Laic a = .0514/.61 = .0842,

Qa :: .61 X .1 X .305 ? 1.4 X 0.1 X 0.56 ;. .097 QII :: 0.61 x 0.1 x .305 + 1.4 x .1 x .561

2'~ iii = 0.194

= .0971

k, = 0.194/.1704 = 1.14 k. = .0971/.0842 = 1.15

The stress for a unit strain of .0154 As explained in solution (1), for e =
fram the stress strain curve in Fig. C3.22 is .0154, f m :: 172000 and to :: 129400.
172000, and to :: 129400.
Fbll = 172000 + 129400 (1.15-1) 191400
Then Fb, = 172000 + 129400 (1.14-1) = 190100 M, 191400 x .0842 = 16100
M, =~ (Fb,,1,/e,) = 190100 x .1704 = 16200 Mtota1 =M 1 M3 = 16620
+ + 16100 =
32720 In. lb.
Total resisting moment:: Ml. ? Mil :: 16550 ?
16200 • 32750 In. Ibs. EXA11PLE PROBL~ 4. Fig. C3.24 shows an un-
sym~etrical I beam section. The material is
Solution 2. 7079-T6 aluminum alloy die forging. The
upper portion 15 in bending compreSSion. It
Instead or making each portion a will be assumed that the compressive crippling
s~etrical section as was done in solution (1) stress for the outstanding upper legs of the
and dividing the results by two, we will tind section is ,65000 psi. (The theory and method
the internal bending resistance at each portion ot calculating crippling compressive strength
as is when bending about the neutral axis ot is given in another chapter.) ihe ultimate
the entire section. Equation (3) now becomes deSign bendIng moment is 16500 in. lb. Find
tor each portion at beam section, M.S.

Fb =t m + to (k - 1) - - - 3' 1---1.5-01 E=.009

where kl. = I~/'

, e, ,

The section modulus at each Dortion refers

to neutral axis otentire beam section. Fig.
I -N.A.-

(g) shows lower portion (1).

= .0895 1] ~:~~ e ~::. 0129

Fig. C3.25
b_! 1+-0.1
I,/e, = .OB95/1.39 = .0645
FIg. (g) Solution:
QJ,:: 1.39 x .1 x .695
= .0965
kl. = Q./IJ,/cl. = .0965/.0645 :: 1.50
The mqximum compressive stress permitted
From equation (3 1 ) , Fb = 180000 + 156000 is 65000 ps i . From Fig. C3.23, USing the
(1.50 - 1) 258000 = stress-strain curve, we obtain a unit strain
ot .008 for this stress. The unit strain at
MJ, = Fbl. X Il./cJ, =258000 x .0645 :: 16620 the bottom edge of sectIon is .009 (1.1781
0.822) :: .0129 as shown in Fig. C3.25. From
Fig. (h)ShoNs uDper Fig. C3.Z3 this strain causes a stress ot
portion (2) 67500 pet .

Tr-;f~i
15
r, :: ~ X 1.5 x .61:5_ 1 1- • --+1 Upper portion: (See
t~ 161=c2 Fig. I)
',' 0.1
~ x 1.4 x .51:5 N.A_
_1 x 1 .~-
0.822 :.i"
Fig. (hi I NA -'3 x . 822' - '3
1
= .0514
-LN.A.-- - --N.A.
xL4 X 0.722 3 = .102 Fig. (0
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES C3.9

I/e = .102/.822 = .124 section will develop in bending about x

axes and Mx the design bending.
Q = 0.822 x 0.1 x 0.411 + 1.4 x 0.1 x 0.772
::: .1418 (2) Using MY 1 carry out the sane procedure for
bending about y axes and tind,
k = .1418/.124 = 1.142
r~ax = 65000. From Fig. C3.23, fa = 30000 Then .the moment ratio Rb tor combined
Then fb = 65000 + 30000 (1.142-1) ~ 69260 bending is.

m. = fbI/e = 69260 x .124 " 8580 in. lb. Rb = Rbx + Rby

Lower Portion. Then margin or safety M.S. =R~ - 1.

INA = ~x

- 31 x
0.75 x 1.178"

0.65 X 1.078
3
I!_N.A.
1.178
C3. 9 Section With One Axis of Symmetry. with Moment
Vector not Parallel to Either Axis.

Since the symmetrical axis Is a principal

INA = .137, I/c = .137/ 1 1 axis, the procedure in this case Is the same
as for the double aymnetrlc case.
1.178 " .1164 "0.75-; To. i
Fig. 0) = Rb x + Rb y
k = .1425/.1164 " 1.222

f max : 67500 II.S. ".l.._1

Rb
From Fig. C3.23, f o ::: 44500 C3. 10 Unsymmetrical Section with No Axis 01 Symmetry.

tb :: 67500 + 44500 (1.222-1) :: 77:!·70 psi

Fig. C3.27
M
!D. ~ = .1164 x 77370 = 9000 in. lb. shows an unsym-
metrical section
Total internal allowable resisting subjected to the
mccent > m1. + m, applied moment
vector I1.
or Ma = 8580 + 9000 = 17580 in. lb.
For this Fig. C3.27
II 16500 case the pro-
Rb " 11a " I7580 = 0.94 (Load ratio) cedure 15 as
follows :
= R~ - 1 = 0.
194
II.S. - 1 = .06 (1) Determine
principal axes
C3.8 Complex Bending. Symmetrical Section. Moment location by equation,
Vector Not Parallel to Princi.pal Axis.
2Ixy
tan 2Q '=
Fig. C3.26 shows a y Iy-I x
double sym~etrlc section.
The x and y axes are M where x and yare centroldal axes, Ix
therefore principal
axes.

The following pro-

--. and I y are mo~ents of inertia about
these axes and Ixy the ?roduct of
inertia.

cedure can be followed I (2) Resolve the given moment M into compon-
which IS qUite con- I Fig. C3. 26 ents Mx p and Myp '
ssrvat I ve . Y
(3) Follow the sa~e procedure as betore.
(1) Resolve the given ~oment into cry,nponents
about x and y axes, or Mx and My. (4) The stress rat10 Rb = Rbxp + Rbyp
(2) USing Mx follow the procedure as given
in the exanple problems and find Rbx =
MxlMax' where Max is the internal moment
C3.10 YIELD AND ULTIMATE STRENGTH IN BENDING
C3. 11 Alternate More Exact Method for Complex
Bending.
A beam sectIon when resl~ting a p~e
external bending moment bends about an axis
that is called the neutral axiS. No matter
what the shape of the beam cross-section for
any given external moment, there is an axis
about which bending takes place. The general
case involves an unsymmetrical beam crass-
section and material which has different
stress-strain curves in compression and
tension in the inelastic range. The neutral
axis therefore does not pass through the
centroid of the cross-section and thus the
method at solution is a trial and error
approach. The solution procedure is outlined
in Chapter A19, Article Al9.l7, and there tore
will not be repeated here. Also the chapter
dealing with the design at beams with non-
buckling webs explains and illustrates .how
the ultimate bending resistance at an entire
beam section is determined.
C3.12 StrengtbUnder Combined Bending and
Flexural Shear.
The previous part at this chapter has
dealt with the determination of the strength
at a beam section in pure bending. The usual
beam deSign problem involves tlexural shear
with bending. In tlnding the true internal
resisting moment, the Cozzone simplified
method derives a trapezoidal bending stress
distribution which will produce the same
internal resisting moment as the true internal
bending stress system. A triangular stress
system is then derived which will also give
the true bending moment.
Now the equation for flexural shear stress
tor a triangular bending stresS distribution
Is
- - -
Thus to use equation (A) for a
trapizaidal bending stress, a correction factor
(C) must be applied or equation (A) becomes
t s -- ££S.
It
To illustrate how the correction factor
(C) can be determined, the I beam section used
in example prable~ (2) will be used.
We will assume the ulti~ate design moment
at 14000 in. lbs. is produced by a load of
600 Ibs. acting on a cantilever beam at a
point 23.30 inches from the fixed end of the
bea~. Thus the be~ section at the support
is subjected to bending moment of 14000 and
a shear V = 600 Ibs. The problem is to find
the margin of safety under this combined
loading.
For pure bending only the stress ratio is
- M -- 15000
Rb-I1 14000 = ' 933 (the va.iue
1 crt u'5000 i s
a
obtained from example problem 2).
The stress ratiO in shear is Rs = fs/F au ,
where f s is the flexural shear stress and Fsu
the ultimate shear stress of the material. -
The problem therefore is to find the value of
ts'
The equivalent trapeZOidal and tr iangiuar-
bending stress distribution will be determined
tor the deSign bending moment of 14000 in. lbs.
For a triangular stress variation,
Fb =Mell = 1400010.358 =39150 pSi.
From example problem (2) the shape factor
k was 1.17. On Fig. C3.l9, the curve for
k = 1.17 has been plotted. Starting with the
Fb stress of 39150 at the left scale, ~un
horizontal to an intersection with the k =
1.17 curve, the prOjecting vertically down-
ward to intersections with the stress- strain
curve and the fa curve to give 35800 and
19700 for fa' The stress results are shown
graphically in Fig. ~~.28a and Fig. C3.28b.
t 11-- 'i--.{ 19700--+1 I 1.... 16100 /+39150..
J _:""==Tt~](
-I ~8 ....[ I.-
1/2 Beam Section 970b
Fig. C3. 28a Fig. C3. 28b
Case 1 Case 2
The flexural shear stress is a function
- (A)
at the rate of change at the bending stress.
Thus we can obtain a shear correction factor
C by comparing the bending stresses in the
two ~tress distribution diagram.
The shear stress 1s maximum at the
neutral axis in this particular problBffi.
- - (B)
The total normal force on the cross-section
of beam above the neutral axis equals the
stress times the area.
For simplification, the beaL, section will
be divided into the two portions labeled (a)
and (b).
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES C3.11

For Case 1:- the cross-section will be determined. The

correction factor tor this paint is found by
Load on portion (a) is, comparing the loads on portIon (a) tor Case 1
and Case 2 stress patterns.
19700 x 1.375 x 0.125 = 3380 lb.
Case 1. Load on (a) : 3380 + 2600 = 5980
16100 + 14100 "" 1.375 x 0.125
2 = 2600 lb. Case 2. Load on (a) = 6300.

Portion (b) Hence, C = 5980/6300 "" .95

19700 x 0.875 x 0.125 = 2155 C3. 13 Strength Under Combtned Bending Flexural Shear
and Axial Compresston.
14100 x 0.5 x 0.875 x 0.125 = 771
The SUbject of the ultimate strength
Total Force 8906 Ibs , design under combined loads is treated in
detail in a later chapter.
For Case 2:-
A conservative interactIon equation tor
Portion (a) combined bending, shear and axial load is,

39150 ; 34250 x 1.375 x 0.125 = 6300

or M.S. = .; (Ra + Rb)'

Portion (b)

1874 res, Rb inclUdes effect of secondary bend-

34250 x 0.5 x 0.875 x 0.125 = 8174 ing moment due to axial load ttmes de-
flection.
Thus the correction factor Is,
C3. 14 Further Values of fm and fa.
C • 8906/8174 = 1.09
Table C3.2 gives the yield and ultimate
Then f s at neutral axis Is, values at f m and to tor a number of other
materials common to the aerospace field of
t Cvq - 1.09 x 600 x 0.209 = 3060 1 structures. The yield and ultDnate modulus
s "" It - 0.358 x 0.125 ps • in bending Fb Is found by substituting in
the equation.
The ultimate shear stress for this
particular magnesium material Is 19000 psi.
(See chapter on material properties.)
TABLE C3. 2 Values at (m <U1d fa for FilId.tnr Fb
1'8 3060
Stress Ratio R8 ::: F ::: 19000 ::: 0.161 Fb'"flll+to(k-l) Yield Ultl.....
sw
M:U:ertal S... fCl"Fty Co tm'"Ftu Co
The interaction equation for combined ZOI4-T8 AJ.Ua:l. AJ.. 01. Forgtnp t s 4 in. 52 18.8 ea 57.5
bending and flexural shear Is 6061-TS Alum. AI. Sheet to! 0.02 17.5 40.5
7075-T8 Alum. Ai. me Forgings
7079-T6 Alum. AI. Hand Forgln(s(L)
e ~ 2.0
t s 6. a
"esea 25.5
33.5
52
ts
71
70.0
67.0
Rb + R~ = 1 A. Z81A MlIcnesluCl AI.:i::rtru.. (Long)
lDCIA_O M.ap.elli\l.ln AI. Sheet
t:50.25
.016-0.25
21
16
T••
•. c
38
30
15.7
22.6
1 ZK60A Magnesium AI. Forg1ngs (Lolli) ae T.5 52 30.8
, AISt Alloy Steel (Normal1zedj ""0.188 TO 43.2 so 83.1
Whence, Margin at Safety =
h; -" r e O. las
'""
AISI Alloy Steel (Normalized) 47.1 86. a
+ R~ AISI Alloy seeet (Heat Tre"ted)
, "
rca ISL!l !l8.0
I
1
AlSI Alloy Steel (Heat Treated)
AISt Alloy Steel (Heat Treated)
AISI Al10y Steel (Heat T~ated)
'"
16'
66.0
53.0
65.0
16'
16'
146. a
172.0
192. a
or, M.S. = /.9'&:/'+ .161" - 1 = .05 tT-7PH Stainless Steel
lT5
ISO 62.0
5$.0
'"aas
16' 161. a
160. a
PHIS-7 MQ {RH9501 Stainless Steel
n,
200

Thus the effect of the flexural shear

TI-8MN Titanium Alloy 36.0
'" 116. a

stress was to reduce the margin or safety of PROBLEMS

.07 in pure bending to .05 in the combined
stress action. (1) A round tube 1-1/2 inches in diameter
has a wall thickness of .095 inches.
AS further calculation at the shear It is made of aluminum alloy whose
correction factor C, its valu8 for the shear stress-strain CU~le Is shown in Fig.
stress at the upper edge of portion (b) of C3.2. If the maximum unit strain in
C3.12 YIELD AND ULTIMATE STRENGTH IN BENDING

compression is limited to .008, what Fig. 1

bending resisting moment will the section --*.I+-- 1 --+1
develop. (Note, since stress-strain
curve has different shape in tension and j"'t:.: O. 1
compression, neutral axis does not

t
coincide with center line axiS, thus use
trial and error method. ...... 0.1

(2) Same as Problem (1) but use a square tube

1
with 1-1/4 inch outside dimension and .1<4- 14"-l
.081 inch wall thickness.

Use the Cozzona method for solving the t~l

8 I
tollowing problems. -'+-16
-Jol 1 I+-
(3) Find the ultimate bending moment that 2
Fig. 4 Fig. 5 Fig. 6
each ot the following beam sections will
develop when bending about the principal
axis and made trom each of the following (4) A simply supported beam has a span of 24
materials. inches. Depth 0: bea~ limited to 2 inches.
It must carry an ultimate load of 4000
(a) 7075-T6 Alum.1num. Alloy Extrusion. Ibs. located at midpoint of beam. ~aterial
Ftu = 75000, Ft y = 65000. is 7075-T6 aluminum alloy extrusion.
Design an r shaped section to carry this
(b) Ti-6Al-4V Titanium Alloy. load. Neglect areas ot corner fillets
Ftu = 130,000, Fty Z 120,000. that would be used in extruded Shapes.
(c) ArBI Alloy Steel, heat treated. REFERENCES •
Ftu =
150,000, Fty = 132,000.
Ret. 1. Bending Strength in Plastic Range. By
F. P. Cozzone, J. Aeronautical Sci.,
l1a.y, 1943.
Ret. 2. Vought Structures Ma.'1.ual.
 .O.T, _ .,_~ • • ~.~
CHAPTER C4
STRENGTH AND DESIGN OF ROUND, STREAMLINE, OVAL
AND SQUARE TUBING IN TENSION, COMPRESSION, BENDING,
TORSION AND COMBINED LOADINGS.
C4. 1 Introduction.
Berare the advent or the stressed-skin
structure tor ai rc ra.rt , or during the period
when tuselage and wing were rabric covered,
round, oval and square tUbing were used in
designing the major structure or the fuselage
and wing. It the wing and tail were externally
braced, streamline tubing was used. The
development or the metal covered structure
el1minated the use or tubing in fuselage and
wing design, however, tubing continued to be
used for landing gear structure, engine mounts,
control systems, fixed equit:m,ent such as
passenger seats, etc.
With the opening or the space age, tubing
as a structural unit in space vehicles is
again being widely used because drag in space
is not an important factor. Round tub mg is
the best shape tor transmitting torsional
torces and thus Widely used in control systems.
Round and square 'tub Ing permit Simple con-
nection or end fitting design. The metals
industry has made available a large number at
diameter and wall thicknesses and thus the
structural designer has a large number ot sizes
to select rroa.
C4.2 Design for Tension.
In general the strength deSign require-
ments are that the l1m1t loads must be carried
without exceeding the tensile yield stress
(Fty) of the-material and the ultimate deSign
load which is equal to the limit load times a
factor of safety must be carried without
failure, which means the tensile stress cannot
exceed the (Ftu) Of the material. In general
for aircraft, the factor of safety is 1.5.
For unmanned missiles and space vehicles the
factor of safety ~y be as low as 1.2. Since
the ratio of Ftu to Fty for ~terials varies
Widely, sometimes the yield under limit loads
is more critical than failure under the
ultimate deSign loadS, thus the stUdent should
always be sure he has the critical situation.
Since elevated temperature and time of
exposure effect the yield and ultimate strength
of materials, the problem of material selection
relative to light weight becomes an important
design factor. Fig. C4.l shows a plot of the
Ftu/w ratio versus elevated temperatures up
to 800 degrees, where w 1s the density of the
material. The tUbe materials are AISI alloy
C4.1
steel and 2024 aluminum alloy. Observation or
Fig. C4.l shows that for temperatures below
3500F, aluminum alloy is lighter unless the
steel is heat-treated to Ftu • 180,000 or
above. Above 350 oF, the ult1mate strength of
aluminum alloy falls Off rapidly, but steel
continues its rather uniform decrease in
tensile strength.
The graph explains why aluminum alloy
cannot be used entirely for the surrace of
supersonic airplanes flying at speedS around
2000 miles per hour, as aerOdynamic heating
would produce surface temperatures in the
region where the strength Of aluminum alloy
decreases rapidly.
to :''':':;'::-,i.=:,:c:,:::,~:~::''-:; ,L:::::: ,:h .:~~:.: :;c::;,ltt;~~,~'~::r-:':;:~~3:-:_'-~:. =j
:~:~¥~~~;;~~->~C~;2~fu~;;i~
700
g 600 j;~~~~~~-"~'"~;;~=:~~~~~~~~i
~ : f~:~g~~~~;~:'~~2-~~~;:_;~:~;~rf~~~~
2 ~~"c:"_j':::.::::3·':=:-=-::: .. -. . ' ·'··-'"""E--- --. :-:-"'':;:';:;'1
~ f"'"-''' ,",",.."",=, -.~.''''''',~,.,,~'''''
300 ~~~~~~~~i~~ii:~.~iE~~;~~~z_ ;~=t=~~=-:_ i_ ._~=:--o:
: : ~~~~tiii~~-f~i~iiti
100 200 300 400 500 600 700
Temp. OF. 1/2 Hr~ Exposure
600
FIlJ. C4.1
Practically every structural tubular
member in a flight vehicle structure must be
fastened or attached to another adjaCent
member. The connection can be made by USing
same sort at end fitting which is fastened to
the tube by rivets or bolts or by welding. It
rivets or bolts are used, holes are cut in the
truce walls which means the tube is weakened
since tube area is cut away, thus net area on
any tUbe cross-section must be used in calcu-
lating the tube tensile strength. If welding
is used in the connection, the welding heat
causes grain growth in the tube material
adjacent to the weld area, which decreases the
tUbe strength, thus a welding correction tact or
must be used in calculating the ultimate tensile
strength. DeSigning for tension will be
illustrated later in this chapter.
 STRENGTH &: DESIGN OF ROUND, STREAMLmE, OVAL AND SQUARE TUBING
Ct.2 IN TENSION, COMPRESSION, BENDING, TORSION AND COMBINED LOADlliGS.
C4.3 Design for Compression.
The strength of members with stable cross-
sections when acting as columns can be cal-
culated by Euler's equation it the bending
failure Is elastic, or E 1s constant (eq. C4.1)
and tor inelastic bending faIlure, Euler's
equation ~lth the tangent modulus Et replacing
E (eq. C4.2) checks test results. (The student
should refer to Chapters AlB and C2 for theory
on column strength.)
(C4.1)
(L'/p)"
Fe' (L'/p)" - - - - - - - - (C4.2)
U Is the effective length and equals
L/~ where c is the column end fiXity
coerr icrent •
Long -and Short Columns
For many years the prOblem or SUbject or
inelastic column strength or !allure was
treated almost entirely trom a consideration
ot test results. That is, sufficient tests
were made to establish the shape of the failing
stress curve in the region where the failure
was at stresses above the proportional limit
stress of the material. MathematIcal curves
were then derived to tit the test results.
Engineers referred to the .columns which tailed
by inelastic bending as short columns, and
thus reterred to the equations that tit the
test data as short column equations. The
columns that ~ed by elastic bending were
then referred to as long columns.· The test
CUI"le tor long columns would follow the Suler
column equation (C4.l) and thus tests were
not necessary to establiSh allowable falling
stresses in the so-called long column range.
Thus over the years short column equations
based on test results have been presented by
official government agenCies for use in
structural design. The official publication
tor the aerospace tield is the Military
Handbook IUL-HDBK (Rer. 1).
C4. 4 Column Formulas for Round Steel Tubes.
From (Ret. 1) the basic short column
equations are:-
Fe' Fcc [l-F e o (L'/p)"/4n'E] - - - - - (C4.3)
Fe Fcc { 1- 0.3C27 [(L' /p )/n
V EIF e o] r,oj --- (C4.4)
Where Fc o is the colunn yield stress
(upper limit at column stress for primary
failure). It can be determined tram test
results by extending the short column curve to
a point corresponding to zero length, ignoring
any tendency of the test curve to rise rapidly
tor very short lengths where failure is by
block compression. Table C4.l shows the
resulting short and long column equations
after values of Fco and E have been substituted
in equations C4.3 and C4.4 and E in the Euler
equation. The column headed transitional
L'/p, represents the value at L' /p where failure
change trom inelastic to elastic failure or, in
other wordS, it is the dividing paint bevneen
the so-called long and short column range.
Thus if the equations are used, the Lr/p value
must be known in order to select the proper
equation.
C4.5 Column Formulas for Aluminum Alloy Tubes.
Fram (Ret. 1), the basic short column
equations for aluminum alloys are:-
Fe • Fe o [1 - 0.385 (L' /p)/nV E/F e o] (C4.5 )
Fe = Fco [1 0.333 (L'/p)/nV E/Feo] (C4.6)
Fe • Fc o [1 - 0.272 (L'/p)/nV E/F e o] (C4.7)
For long columns:-
Fe = nOE/(L' /p)" ------ - (C4.8)
The equatior~ for determining Fco are
given in Table 04.2 (from Ref. 1). The table
also indicates which ot the three short column
equations to use for the various aluminum alloy
ma.ter-ta.Ls ,
To illustrate the use of Table C4.2, the
column formula for 2024-T3 aluminum alloy
tUbing will be derlved:-
From Chapter 82, we r ind the following
strength properties for 2024-T3 tubing,
Ftu a 54000, Fey. 4200C
From Table C4.2, the equation tor Fco is,
Fco ~ Fey (1 + ,; Fey/lOCO SUbstituting,
Fe o • 42000 (1 + V42/1000 • 42000 (1 + .205)
= 50600
From Table C4.2, we note that short
column equation C4.5 applies.
SUbstituting in this equation,
Fe = 50600 [1-0.385(L'/p/nV 10,5oo,000/50600J
or Fe = 50600 - 431 L'/p - - - - - - - - (04.9)
C4.6 Column Formulas for MagneSium Alloys.
From (Ref. 1) the following short column
equations for various ~gnesium alloy materials
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES C4.3
Table C4. I Column Formulas for Round Steel Tubes

Short columns (a)

Transi-
Material Ftu. its! Fty. Itsi Fcc, Itsi tiona! (C) Long columns (b) Local
Column formula (b) Basic failure
equat~on
L'/a

1025 .. , •....•...... 55 36 36 36,000-1. 172(L'/p)3 ", C4.3 122 266 x lO'/(L'/p)' ·... (d)
4130 , •... , ..... , .. ,
Heat-treated e alloy
95 75(0) 79.5 79,500-51. 9(L'/p) J,. 8,. C4.o4 91 286 x 10o!l/(L'/p).11 ·... (d)
steel. , , ...•• ' ...
Heat-treated alloy
125 103 113 113,000-11. 15(L'/p)' .• Co4.3 73 288 x 10'/(L'/p)' ·... (d)
steel, ..•... '" .•
Heat-treated alloy
150 132 145 145,000-18. 36(L'/p)" .. C4.3 83 286 x lOo!I/(L'/p).II ·... r')
steel ......•..••. 180 183 179 179,000-27. 95(L'/p) a •• C4.3 58 286 x 10o!l/(L'/p).11 ·... (d)
a Equatfun C4. 1 may be used in the short column range if E d Not necessary to investigate for local lnBtabUity when
is replaced by Et obtained from the combined stress- O/t 50.
strain curves for the material. e This value is applicable when the material is furnished
b L'/p '" L/py'C: L'/p shall not exceed 150 without specific in condition N (MIL·T-6736) but the yield strength is
authority from the procuring or certificating agency. reduced when normalized subsequent to welding to 60
c Transitional L/p is that above which columns are 'tong" Itsi.
and below which they are "short." These are approxtmate
values.

are given in Tables C4.3 and C4.4. or inelastic instability of the column as a
whole. As the slenderness ration L'/p gets
Table C4. 3 Column Formulas for Magnesium- smaller, the Fc stress increases. Now it the
Alloy Extruded Open Shapesa
diameter of the tube is relatively large and
GENERAL FORMULA the wall thickness relatively small or, in
other WOrdS, it the diameter/thIckness (D/t)
K(Fey) • ratio 15 large, tailure will result by local
Fe' (L',jJ)m crippling or crushing ot the tube wall and
this local tailing stress is usually repre-
(Stress values are in lesi)
sented by the symbol Fcc. The values or Fcc
in general have been determined by tests (see
Alloy K a m Max. Fe design charts for Fcc versus D/t ratio).
MIA .• , ,.. .
. '
AZ3IB, AZ61A, AZ80A.
. . 180
2,900
1/2 1.0
1.5
0.90 Fey C4.9 Design Column Cha.rtl!I•
1/4 Fc~
AZ80A-T6, ZK60A-T5 . 3,300 1/4 1.5 O. 6 Fey In design, column strength charts are a
a Formulas given above are for members that do not fail by great time-saver as compared to substituting
local buckling. In the various column equations, thus a number
at column charts are presented in this chapter
to facilitate the strength check at columns
Table C4. 4 Column Formula for AZ31B-H24 and the strength deSign of columns. Fig.
Magnesium-Alloy Sheet C4.2 is a chart of L'/p versus Fc for heat
(1.05 Fcy)'(L',p ) , treated round alloy steel tUbIng. Fig. C4.3
Fc'" 1.05 Fcy- Is a similar type or chart for aluminum alloy
4 n"E round tubing. Fig, C4.4 gives column charts
Max. Fc .. F cy for magneSium alloy materials. All three
charts are taken tram (Ret. 1). Figs. C4.5
and C4.6 represent a turther S1mplication ror
the design at steel and aluminum round tubing.
C4.7 Short Column Equations for Other Materials.
C4.10 Section Properties of Round Tubing.
For other metals tor Which short column
equations are not available, the use of Euler's Table C4.3 gives the section properties
equation, USing the tangent modulus St can be of round tubing. A tube is deSignated by
used (eq , C4.2). Refer to Chapter C2 for giving its outside wall diameter (0) and its
intornatlon on how to construct column strength wall thickness (t). Thus a 2-1/4 - .058 means
curves using this equation. a tube with 2-1/4 inch outside diameter and a
wall thickness in inches ot .058. Since a tube
C4.8 Column Failure Due to Local Failure. is symmetrIcal about any axis, the polar moment
or inertia, which Is needed in torsion problems,
The equations as presented give the equal twice the rectangular moment of inertia
allowable stress due to failure by bending of as given in Table C4.3. For weIght comparison,
the column as a whole and the action is elastic the weight at steel and aluminum tubing is
 STRENGTH & DESIGN OF ROUND, STREAMLlliE. OVAL AND SQUARE TUBrnG
C4.4 IN TENSION. COMPRESSION, BENDING. TORSION AND COMBINED LOADrnGS.

Table C4. 2 Column Formulas for Aluminum Alloys

Alloy and Temper Product F•• Short Transitional Long

Columns L'!p Columna

2014-T'J, T", T451 Sheet and Plate.;

202....T3. T351, T36, T4, Rolled Rod, Bar and
T42 Shapes;
Drawn Tube

5052-All Tempera
6083--All TempeJ'l!l
~AUTempera
54M-All Tempera AU Products F•• (1 +..,; F,./lOOO) Equation 1.732."/ElF.. Equation
MM-All Tempe" C4.5 C4.8
6Ofl1-T4, T"Sl, T4510,
T4511

All Cut Alloys and Sand and Permanent

Tempen Mold CaatinlJ8

2OU-T3310, T3S11, T4, Enru.sions

T42

2014-T6, T651
~T8,T81,T86,~1 Sheet and Plate*j
7075-T6, T6S1 Rolled Rod, Bar and F •• H +../ F .,/1333) Equation 1.346• ../E / F.. Equation
7173-T6, T6S1 Sha.pes; Drawn C4.6 C4.8
Tube

6081-T8, Teal Sheet and P1.te·

201.....T8, T651, T6510,

T6SlI, T6S2
2024-T8, TOI, TaStO,
2024-T6, TOI, T85tO,
T8611, T862
7075-T8, T8610, nm, Eztruaiona, Forgings
T862
70'1l1-T6, TastO, T6SlI,
T6S2
7178-T6, T6S10, T6S11

Rolled Rod, Bar and F ,.(1+ ,; F .,/2000) Equation 1.22. ."/E/F.. Equation
Shapes; C4.7 C4.8
6Ofl1-T6, T65l, T8StO, Drawn Tube;
T6S11 Extrusions

X202O-T6, T651 Sheet and Plate.

'"IDdudes eJad u weUu tNIre sheet and plaC.l. Equation C4. 8 may be used. In tbe 'hort uolumn rllZI.P If E' Is
TtmsttioDal L'I_ Is that above _hleh the columns .,. "1onC'" llnd l'l!plaeed by £, obtained from the eompreulva StrlSSoStraincurve for the
beknr _b1ch t.be1 ..... ''short''. matllr1&l.
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES C4.5

given in tfie last two columns of the ta~le. Table C4.4

C4.11 Some General Facts in Tubing Design. Tension Allowables Near Welds in Steel Tubing (X-4130)

Type of Weld Normalized Welded after HT HT after

1. For a given area, the larger the tube Tube Welded or Norm. after Weld Welding
diameter, the greater the column strength
if failure due to local crippling is not *Tapered Welds
critical. of 3()O or Less .947 Ftu 90,000 psi .90 Ftu
All others .841 Ftu 80,000 psi .80 Ftu
2. The higher the D/t ratio of tube the lower
the crippling or local failure strength. * Note: Gussets or plate inserts considered 00 "taper"
with ¥. .
3. If columns fall within the long column ** For(X-4i30) Special, comparable, values to the FtuJ
category, the use of higher strength alloy equal to 90, 000 and 50, 000, are stresses 94, 500 and
steel or aluminum alloy will not increase 84, 100 pst, respectively.
strength at column since E is practically Ret. Anc-5
constant tor all Chrome-moly steel alloys
and l1kewtse for all aluminum alloys.
Failure is due to elastic buckling of the C4. 13 IDustrative Problema in Strength t:hecktng
column as a whole and is therefore a and Design of Round Steel Tubes as Columns
function only at I, L' and E. and Tension Members.

4. The col~ end restraint eftects the neces- PROBLEM 1

sary tube size. Consult the design require-
ments of the Army, NaVYJ and C.A.A. in this
matter. In general with welded steel
tubular trusses a coefficient of C = 2 is
permissible except tor engine mount and
Nacelle structures. For trusses with
riveted joints a value of not over 1.5 is
generally permissible.
5. The student should realize that practical
limitations such as clearance requirements
may determine the diameter ot the tube
instead of strength-weight considerations.
Thus design can consist of checking the
tubes available under the given
restrictions.
C4. 12 Effect of Welding of steel Tubes Upon the Tension
and Column Strength.
Si~ce welding etfects the grain structure
ot the tube material adjacent to the weld,
tests show the strength of the material
adjacent to the weld is decreased as compared
to the unwelded material. If a tapered weld
1s used J the etfect of the weld is decreased.
Table C4.4 shows the allowable stresses in
tension to use when tension loads are carried.
In short cOlumns, the prunary column
failing stress may be greater than the local
crippling strength ot the tube adjacent to
the weld at the end of the tube. This local
tail1ng stress Jue to welding 1s reterred to
as the weld cut-otf stress and the column
compressive stress F c should not exceed this
value. This cut-off weld stress is shown by
the horizontal Itnes in Fig. C4.2 and C4.5.
Tube sIze: 1-1/2 - .058, Length L : 30 In
End tix1ty coefficient C = 1.
Material:- Alloy steel J Ftu • 95000. The
tube is welded at ends.
Ult~te design loads are:- P -14,500 =
lbs. compression, and P ~ 18500 lbs. tension.
Required the Margin ot Satety (M.S.)
Solution: The ccnpreeatve (M.S.) will be
determined first. As the Simplest solution,
we can use the column curves in Fig. C4.5.
For a length ot 30 and C = 1 J trom the upper
right chart we project upward to the inter-
section with the 1-1/2 diameter tube and then
horizontally to the lett hand scale to read
the column strength or 14800 Ibs. which we
will call the allowable tailing Pa •
P 14800
M.S. = Pa - 1 • I4500 - 1 = .02
The tube strength could also be tound by
using Fig. C4.2 as tallows:
L' : L/V-C- : 30/v-r- : 30
L'/p : 301.5102 : 58.7. p 15 round rrom
Table C4.3 as well as the tube area 0.2628
sq. in. Using 58.7 tor L'/p on lower scale
and projecting upward to the Ftu = 95000
curve, which is the lower curve, and then
horizontally to left hand scale we read
Fe =
56500 psr ,
Whence, Pa = FcA = 56500 x .2628 =
14850 lb.
The solution obViously could be made by
SUbstituting in the short column equation for
steel haVing Ftu = 95OO0 J or
 STRENGTH & DESIGN OF ROUND, STREAMLINE, OVAL Ai"ID SQUARE TUBING
C4.6 IN TENSION, COMPRESSION, BENDING, TORSION AND COMBINED LOADlliGS.

"0
... \ --
rcce at Weld Cutoffs
I
"" 1\ FlU' ISO k$1
II
I Heot Treated After Weldin
Heat Treated Cutoff,ksi Reference
10 Flu,ksi
160
i \., 180
150
144
120
REF. 1
150
\
125 95 r-
Normalized
100 76.6
Cutoff Meattreofed
95 67.5 I-
ofter welding
1\ Ftu~ISOksi
"" \~ .0 '0

-.... ...
1

____ Unwelded COlulTV'l

•
110
1\ tu~125 ksi
ollowobles
·
"
;;;•
\ e
E,
a JO

·
u

eo
\1 a0
a• 20

"
~u
0

• - \ r e l d cutatf cne- heat

I"- treot 4130
°0 20 .,
,,\ { ~,
Ftys7S kSl,
Ftu=95 kSl Fig. C4.3 2024-1'3 and 6081-T6
'\ round aluminum alloy tubing.

. '\
-,
<, ....
0 20 60 80 I00 I20 4
10 I60 eo
I '00 220
L'Ip

Fig. C4.2 Allowable column stress for heat-treated.

alloy-steel round tubing.

50 I I
~
• I I
Euler curve, E = 6.5 .lIOe psi I----
.5 40
AZ80A·T5 and ZK60A-T5 I
••
~ 30 _ 0 96 Fey
in
VV e.lfruded shapes (except hollow)
I I
AZ80A·F e.ltruded sncpes
I' +-
;
c
E
.2 20
o
~V .I' ~~Round tubes,
(except hollow I
I I
Fey = 10 ksi
I I
U'

"""'~ I
u.."
10 I i
.
Fey !
o I I I

o 20 40 60 80 100 120 140 180

LIP

Fig. C4.4 Allowable column stresses for magnesium-alloy columns.

 ~
~ 10

.
o
~ 8
;;i
~ 6

1;'"
z 4
."'''"" 2
~
"6
(.J
(:=2
C=l

"'~
is 20
.'"
0
0 20 40 6{) 80
~ 16 COLUMN LENGTH - INCHES

~
[g
.
g 12 38

.'"
ez ~ 32

."'''""
z
8
..
'" 28
0
:Ii
"0
,.J ~ 24
(.J [g
C=2
C=1 0
0
20 40 80 80 .'"
0
20

i'l
COLUMN LENGTH - INCHES
.'"
o 18
z
z
"0 50 '".'"
.
0
'" "'
z
:Ii
12

"'Z
Q 40 "
,.J
0 8
< (.J
"'
" 30
.
g 4

'c.". 20 C=2
z C=l
."'
''""
z 10
20 40 60
COLUMN LENGTH - INCHES
80 100

:Ii
"6 Fig. C4.5
(.J STRENGTH OF CHROME MOLY ROUND STEEL TUBES
0
C=2 FTY '" 75, 000 PSI FTU • 95, 000 PSI
C=l 0 40 80 120 160
COLUMN LENGTH - mCHES
 STRENGTH & DESIGN OF ROUND, STREAMLmE, OVAL AND SQUARE rusmo
C4.8 IN TENSION, COMPRESSION, BENDINCi, TORSION AND COMBlliED LOADINGS.

7000 U .r .- 'Ftt;;;q4;6--:-

;iil;~ii~~~~~Ft. -' !'f'~~CiO~'~'?~.~'Ci"'~'irEi'~~!-:-'

I:r.l t':: .- - '--~-I'
.... --. 20~TJ ~.T!JBn-:::····
S
....
6000 .. :;--.r-ru:'" a ~:.= ~"! Uop& PIlI
~ :
--l:;~;~jJLI~~:~·L
I 5000 ..:::.t' >-- ·:':':';':~.~--"';~:lOi.:ii.hti~;:.- '-'::-::1- -

~ . _.=-~~:l::::~=::.~r~~-:~~­
'" 4000
0:4
.J
•. ~_.

.-':~~:~_"';-:_~~~~f=;!~ IlIiz;; .
'"
<
~ 3000
':':_:i:::;:~':;:::.t.r:~; . *'::::::::;
~-':::""'l··.":- -:-...:,o:.-:~_--: ';::
;:::~?t·':;-:~l::~:-4~J
'" ~: ~;:~ ~:i~~-~~~-~5
:.·::.-:::::'=.~f~~§
~+-:-=-"':-~::t;~:- e,:=...-; -=-·t:=:J
~1000
'" -_~~ __.~. ::-:::":'---:L:=~:~!=~:f~NW-~*~
- -=::::~I
o
120304050 60 20 40 60 80 100 120
COLUMN LENGTH· U'lCHES (C .. 1) COLUMN LENGTH - INCHES (C '" Xl
35000
14000
~
Q
S 12000
..
Q
1.. ,
30000

Q 25000
~'OOOO ~
.J

..
.J

.J 8000
~ 20000
'"<
'"
~ ~ 15000
'"<'"
.
8000
..
.J
<
~ '0000

i... 4000
;:
.J
e 5000
'"'" 2000

0 20 40 60 80 100 120
COLUMN LENGTH - INCHES (C'" 1)
35000
~t:.~:~
~
Q
30000

.''""
Q

25000
.;-wi::~-F~*~

~ ~. ~~~~g~_::

~ 20000 ~ t:J;=IiG'E~" .:.'

; 15000 '

..
'"
<
Eo< 10000
I

~
~
:l 5000
lii" - c:
: :.1__

a 20 40 60 80 100 120
COLUMN LENGTH· INCHES {C .. 1}

,--.-. ,~""'--._; --- ---~ :-::--._---:--~ - ..:......_- ~ -~~ ~~--:::-:-~ -= ---;-=:- -=--"-------
 /'
Wilight

b
Wlltght
LIt/iOO i_. Lb/IOO iu.
Dla.

1/'
lla ••

.022
A
.OU1.
.P
.0.10
I
.000103
I/Y
.oooan
D/'
11.3. ...••• ...•••
StOilI Dural Dia.

1 1/8
Go••
.035
A
.20:13
.P
••501
I
• OU.5
I/Y
.0113•
D/.
53 ••0
et••l
5.73 ........
Dural

.028 •01953 .0781 .OOOU2 .000878 '.33 .0.1 .2811

.....
."61
.'~1
.11120
.1367'1
.12500
.14511
38.25
32.30
7.811
1.38 3.311

....
.058 .3311
.nu•

........
3/' .02' .03063 .1231 .000442 .003...6 13.3. • '8- .31 .eee • 3616 .16168 2'.80 10.47 3.7"
.035 .03731 .120' .000548 .003tU 10.12 1.08 .3. .083 .4873 .&342 .18787 .200110 22.'0 13.25 •. 73
.08018 .116. .000882 ,003836 1.85 1.43 .'1

••• •
.....7' •••
.0.1 .3003 .8800 .14288 .14288 0&0.80
.04.162 .167:1 .001180 .004 ...1 11.0 1.17 .06a .353' .8a61 .18816 .18896 30&.0&6 10.03* 3.5S-
1/'
.eas .0:1.113 .11141 .001390 .006659 14.2' 1.0&6- .0611 .3961 .614.5 .U5a .1851'" 30.75 11.11 '.00 ;>
•0.0 .0""3 .1618 .00178' . DOn..... 10.20 1.'6 .eaa .....18 ••783 .2300 .2301 2.... 10 14.18 5.06
Z
.0811 .68611 .n..... .2588 .2588 21.011 18.11 5.'18
'j' .02'
.035
.05252
.08.81
.2113
.2090
.002346
.002833
.007603
.009065
22.30
11.85
1.48
1 ••• •
.M
.e8·
."
2 1/4 .048 .3388
. s....
.7'1.3 .2052 .1824 0&5.80 '.5' 3."3
.'"
;>
e-

-
.04' .08867 .204" .003704, .011852 12.77 2.51 .051 .17153 .2.01 • 2134 31,'0 1l.SO· ".011.

'j'
.058
,028
.10331
.08351
.2018
.25U
.004185
.OOCl ..5
.0134.211
.011053
10.11
28.80
21•• 3
1,80 ...
1.05

....
.065
.0.3
.095
.4... 21
.5851
....32
• 172a
.7887
.762•
.2685
.3322
.3741
.236•
.29:13
,33211
34.80
27,15
23.70
12.84
1'.01
18.22
4.52:
5.72
8.61
>oj
g• '"
,0311
.OU
.07882
.107.1
.2631
.2485
.00503.
.008161
.OU.21
.01718a
21 ....a
16.30
2.23·
3.06 1.0' ' a 1/3
....
.0"1 .3773
.4450
.8887
.8636
.2834
.3318
.2287
.2655
61.00
43.10
10.88
12.60
3.83
4.50
•n ;>Z
,..
.. .,"'""
.0118 .1260. • aUII .001601 .02:031 12••'" 3.51 l.a.
.065 .13961 .a433 .001278 .oa201 11.63 3.'6 1.42 .oal .4.7:1 .8813 .3681 ,2960 3•••11 14.0._ 11,03* w
.083 .6302 .U60 .4801 .3888 30.10 11.86 8.38
.7' ....16 • aD. 3..

..'.," "-
7/. .038 .07"61 .2988 .006S81 .016289 31.:13 2.11 • Ot5 .7178 .8&0' • 6191 a8.30 1.27
.035 .0.238 .2913 .008181 .018853 26.00 2.82· ••4·
.0". .12716 .a925 .010882 .024&1 11.86 3.60 1.2' a 3/4 .04' .4168 .8661 • 3783 .215• 68.10 11.71 4.20 0
.058
.066
• 14881
.18Ml
.28'•
.28615
.01248"
.013853
.02853
.03UI
16.10
IS."
4.2a
•• 88
1.11
1."
....
.058
.0811
• 4901S
• 6"'83
.8964
.......
.06211
....34
.4446
..au
.'18•
.3233
.3698
.4501
47.40
42.30
33.111
13 ••0
lIS.IO.
19.70.
4 ••8
6.55-
7."
Z

1 ,036
,0...
,068
.10811
.U640
.34l"
.3387
.012388
• 018594
.019111
.02"n
.0331•
.03822
38.68
20.40
11.20 ....
3.01. 1.01.
4.15 1.48
I.U
•
.095

....
.058
.7124
.6361
••393
1.0403
.8.01
.11802
.50"
.3868
:18.85
n.70
2:1.48
18.18
8~O3

5.42 - ...
'.", ...0
>oj

-'"
.17164 .3337
.085 .190.3 .3314 .020870 .OC193 16.S8 1.41 1.93 .0811 .19'3
.7808
1.0380
1.031a
.6451
.8091
....305
•15398
4'.20
36.16
18.•6
21.66.
6.08
1.70- '" r-
1 l/B .035 .11.85 .3858 .01782 .03188 32.10 3.40· 1.2:1 .0.$ .8870 1.021. .'1118 .810" 31.61 24.66 8.78
...0 o
.049 .16564 .3808 .0:u.02 .04270 a2 ••1S 4.8'· 1.88 .120 1.0U7 1.0181 1.1278 .7ISII 25.00 30.18 11.00
.058 .184.2 .3780 .02176 .04833 11.40 6.11 1.31 >oj
.085 .21860 .31156 .03062 .01421 11.30 '.14 a.lIo 3 1/... .058
.066
.5818
.8504
1.1:187
1.1283
.1410
.8261
.41580
.11077
66.io
60.00
18 ••7
18.40
Ii. 8.
8.11 0 '" .,""
" -
1 1/4 .036 .13360 .4287 .oa481 .03.... 3!i.10 3.7'· 1.35· .083 .la51 1.laOI 1.0381 .837. 38.16 23.3" 1.35*
" '"
,049
.058
.066
.1848'
.2172
. 2420
.«250
.42lf
.411'
.03339
.03887
.G4280
.063...2
.06187
.08818
211.60
n.n
19.2a ....
6.23· 1••7·
6.111 a.20
2.46
31/2
.095

....
.120
.'418
1.1100
.1014
1.1180
1.1074
1.2141
1.1721
1 ....472
1.0348
.7211
.1808
.6114
34. aD
27.10
63.80
28.18
33.43
18,81:1
26,20
8.62
11 ••5

7."
Z

>oj .,
n
r-

-"'" '"
1 3/' ,035 .1473 .4738 .03309 .04814 31.26 •• 17 1.41 .083 .tilt 1.20U 1.3012 .7t36 42.20 9.01 d
.04.9 .2041 • 4691 .0449a .08534 28.05 6.78* 2.01 .095 1.0182 I,aO«3 1.473. .8ua 38.85 :18.70• 10. as.
.058
,065
.24.00
.2875
.4681
.4838
.011213
.06113
.011183
.08341
23.70
al.15
8.80
1.68
a.43
2.70
.120 l.a742 1,1868 1.8220 1.0"11 21.15 3.,00 la.89 '"
z >oj
3 3/4 .086 .16:15 1.3031 1.a171 •••14 67 ••0
o.
21.30 7.80 d
1 1/2 .035
.04.
.0118
.0811
.1'11
.2234
• 2628
.2.30
.6181
.5132
.1510a
.11071
.04324
.058111
.08841
.076:18
.06181
.07141
.011:11
.1007.
4a.1O
SO.80
alS.86
13.01
4.118
'.32·
7.ta
•• 30
....
1."
2. ae-

2 ••'
.083
.095
.120
.168:1
1.0908
1.3885
1.2.88
l.2.n
1. a841
1.8080
1.8228
2.21185
.151.
.8722
1.2036
46.20
38.60
31.2i6
:17.
30.84.
3'.70
'.81
11. Q.tIt
13.1.2
n
>oj
d
.'",
.0.3 .36.5 .6018 .08305 .124.07 18.08 10.47 3.14
• .065
.083
.8035
1.0214
1.3116
1.3852
I.U67
1.8517
.1779
.119'
81.60
...a.20
22,75
ai, 16
8,12
10.3a
1 5/8 .0311
.0.9
.1741
.3-&26
.5622
.111175
.05528
.015..0
.08803
.0.2:19
46.40
33.16
4.96 1.77
' ••7· a.48
.081S
.120
1.1615
1 ....21
1.3810
1.312i1l
2.2228
a.711U
1.1114
1.377'
42.10
33.33
33.n
41 ...0_
11.78
14 ••Olt
'"
,058 .a8511 .1164'" .08118 .lQ801 a8.00 •.ot 2.81
.065 .3188 .51120 .0.707 .11948 :16.00 '.06 3.:13 . 1" 1.1321 1.4551 3.873a 1.1408 31.16 48.10• 17.6S1t
• 1/...

1 3/4
.'"
.031
...021
.11185
.un
.606D
.119811
.08938
.147151
.079a7
18••1
60.00
11 ...0
6.32
•• 08
1.n
4 1/2 .15' a.1281 1.63'. 1.0282 2.2347 28.80 '0.40* n.lla-
.049 .2618 .8017 .09...78 .10832 35.70 7.42* 2.611· ... 3/.. .1" a ••••• 1.'143 1.0a13 2 ••583 26.a5 18.25_ 27.201t
.018 .3083 .6986 .11048 .1:162'" 30.20 8.73* 3.12
.0611
.083
.3441
.43..7
.11982
.1801
.12230
111138
.13.71 28.80
111.10
'.n
1:1.33
3.48
4.40
• U IIUIIJ)lIlD ",81.0
.11.:11'
 STRENGTH &: DESIGN OF ROUND, STREAMLINE, OVAL A..~ SQUARE TUBING
C4. 10 lli TENSION COMPRESSION BENDllia TORSION AND COMBINED LOADINGS

Fe =79500 - 51.9 (L' /p)." L' = L/ve = 40/'1'2 = 28.4

= 79500 - 51.9 (58.7)·"
The short column equation applies since,
as shown in Table C4.1, the transitional L'/p
is 91 and the value for our tube is 58.7.
TenSile Strength
Since the tube is welded, the tube
material adjacent to the weld is weakened.
The weld correction values are given in Table
C4.4. We will assume a weld other ttan
tapered. Let Fa : allowable or failing tensile
strength of tUbe.
= 56500 psi In Fig. C4.5 we can use L ::: 40 and c = 2
scale at bottom of chart or use c 1 scale=
and L' = 28.4. Reading the chart we obtain
Pa ::: 9200 lbs. Thus the c = 2 fixity
increased the strength of the tube trom 6000
to 9200.
Case 4. Same as Case 3 but heat treated to
Ftu ::: 150,000 after welding.
L'/p = 28.4/.425 = 66.8
From Fig. C4.2 USing 150,000 curve, we
read Fc = 63000, whence

Pa ~ Ftu (weld factor) (area at tube)
= 95000 x .841 x .2628 = 21000
!I.S. = (Pa / p )- l = (21000/18500) -
0.13, thus compression 1s critical.
PROBLErl 2
Case 1. Tube size 1-1/4 - .049, L = 40 in.
c = 1
Material: Alloy steel, Ftu = 95000
Find ulttmate compressive load it will
caITy •
Pa = FcA = 53000 x .1849 = 11650 lb.
lbs. In this case heat treating produced
additional strength, whereas in Case 2 it did
1 = not. The reason for this Is that failure
occurs in the inelastic stress range and heat
treating raises the material properties in the
inelastic range. The end fiXity changed the
column tram a so-called long column to a short
column.
The strength could be found also by sub-
stituting in the short column equal for
150,000 steel as given in Table C4.1,

Solution: From Fig. C4.5, Pa = 6000 lbs. Fc = 145000 - 18.36 (L'/p)·

Case 2. It tube was heat treated to Ft u = = 145OCO - 18.36 (66.8)' = 63000 psi.
150,000, what compressive load would
it carry. PROSLEM 3
Solution: Fig. C4.5 cannot be used since Case 1. Tube size 2 - .065, L = 24, c = 1.5
Ft u 2 150,000, thus we will use Fig. C4.2. Material Ftu = 95000. Welded at ends.
L' = Live = 40NT = 4. From Table C4.3, Ultimate deSign load' 25000 lbs.
p = .425 and area (A) .1849. = What is M.S.

L'/p 40/.425 = =
94. From Fig. C4.2,
using the 150,000 curve, we find Fe = 32500.
Then Pa = =
FeA 32500 x .1849 = 6000 lb. Thus
heat treating the tube rrcm 95000 to 150,000
for Ftu did not increase the column strength.
For a L' Jp =a 94, it is a long column and
failure is elastic and E is constant.
The strength could also be calculated by
Euler's equation from Table C4.1.
Fe =286,000,000/(L'/p)'
= 286 ,000,000/(94) a
same as preViously calculated.------
Solution: L' = L/ve = 24/~ = 19.7
From Fig. C4.5 for L = 19.7 on c = 1
scale, we project upward to the 2 inch tube
and note that it intersects the horizontal
weld cut-off line which gives an allowable
column load at lett scale of Pa = 26700 lb.
Failure in this case is local cripplIng
adjacent to welds at the tube ends.
M.S. = Pa/p = 26700/25000 - 1 = .[17
Case 2. Assume tUbe 1s heat treated to Ftu
- 32500 pe i , the = 125000 atter welding. wbat Is tube
strength.

Case 3·. Same as Case 1, but assume tube Is L'/p = 19.7/.6845 = 28.8
welded to several other tUbes at its
end and that the end fixity developed USing Fig. C4.2 with L'/p =
28.8 and pro-
is c • 2. jecting up to 125000 curve, we again note that
horizontal weld cut-oft line is intersected

~ - --
,~,>,. ',~-'. <, t ~:: -,'~~'.'~'>.~"_\~":~~~~~~'~'~:.>~.~~'~~'_'>.~~" ;,:.':;,~.,,~ ~:">~::';'~_:~~,~~,:"'f ':-~~_. :~ ~ .. _·,:~~::~,,~f~:~/J:~~(:
------- .. ~ - . -~-- ---------

ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES C4.1l

giving Fe = 95000, whence Pa = 95000 x .3951
= 37500 Iba ,
If the tube had not been welded at ends
the dashed part ot the column curve could have
been used, thus giving additional strength.
PR08LEM 4
Fig. C4.7 shows a steel tubular engine
mount structure for a 1050 H.P. radial engine.
The ultimate design tension and compressive
load in each member as determined tram a stress
analysts for the various flying and landing
conditions are shown in ( ) adjacent to each
member. The true length L ot each member Is
also shown. Using chrome-moly steel tUbes,
Ftu = 95000, select tube sizes for the given
loads. It Is cammon practice to assume the
column end fixity c =
1 tor engine mount
members, since the mount Is subjected to
considerable vibration and the true rigidity
given by the engine mount ring is difficult
to accurately determine.
The results show that 1-1/4 - .035 is the
lightest. Since there is danger in welding
.035 thickness to the other heavier tube gauges
particularly the engine ~ount ring which is
usually relatively heavy tor this size engine,
a ~ln~um tUbe thickness or .049 will be used,
hence the 1-1/8 - .049 tube will be selected.
Consider Member (2)
Design loads 11650 tension and 4250
compression. Since the tension load appears
critical, the tUbe will be designed for the
tension load and then checked for the
compressive load. The Ftu of the material
equals 95000 psi. Since the engine mount in a
welded structure, the strength at the tube ad-
jacent to the end welds must be reduced to .841
x 95000 ~ 80000 psi (see Table C4.4).
Hence tUbe area required = 11650/80000 =
0.146 aq, in. From. Table C4.3, which gives
the section properties of round tUbes, we
select the following sizes:

1- .049, Area ... .145, 11.S• • (.146/.146) -1 =0

1- .058, Area = .172, M.S. = (.172/.146) -1 ....19
1-li8 .049, Area = .166, M.S•• •14
1-3/8 •035, Area = .147, M.S• = 0
To obtain a reasonable margin at safety, the
1-1/8 - .049 will be selected.
Front View
Fig. C4.7 Many structural deSigners prefer to have
large margins of safety on engine nount members
Consider member (3). Ultimate deSign load = as considerable trouble has been encountered in
- 9250. Referring to the column charts or the fanura or cracking at engine mount members.
Fig. C4.5, we find for C :: 1 and L :: 31.4 the
fOllowing tUbe sizes for a strength near The strength or the 1-1/8 - .049 tube as a
- 9250 Iba . column for length = 31.4 and c := 1 equaLs -6700
lbs. tram Fig. C4.5 which gives a margin of
sarety or (6700/4250) - 1 = .58 on the maximum
1-1/2- .035, Pa = 8850 (weak). (Note-Pa = compressive load. The student should select a
allowable load).
tUbe size for member (l).
1-3/8 - .049, Pa - - 10350,weight::a5.78 lb./lOO",
C4.14 illustrative Problems USing Aluminum Alloy
M.S. = (10350/9250) - 1 = .12 and Magnesium Round Tubes as Columns and
Tension Members.
1-1/4- .058, Pa = -10000, weight = 6.15, M.S.
= 10000/9250) - 1 = .08 In general alloy steel round tUbes must
be heat treated to around 180,000 to 200,000
Thus use 1-3/8 - .049 since it is the lightest ulttmate tension strength before they can cam-
as well as the strongest. pare favorably with aluminum round tubes on a
material weight basis. However, aluminum
Consider member (4), Load =- 5470, L = 30, alloy as used for tubes cannot be welded
c 1= satisfactorily and ti:us in a truss structure
the end connections involving riveted and
Fr'om Fig. C4.5: bolted connections add weight and design
difficulties as compared to welded connections
1-1/8- .049, Pa = -7100, 'Nt.= 4.68, M.S.: .30 in steel trusses.
1-1/4- .035, Fa '= -6500, 'Nt.= 3.78, M.S.= .19
1" - .058, Fa:: -6000, wt .» 4.86, M.S.= .10
 STRENGTH &: DESIGN OF ROUND, STREAMLmE, OVAL A..lIffi SQUARE TUBING
C4.12 IN TENSION, COMPRESSION, BENDlliG, TORSION AND COMBINED LOADlliGS.
PROBLEII 1
Case 1. Tube size 1 - .049 round.
L = 24, c = 1. Material 2024-T3
Find failing compressive load.
solution: The column curves in Fig. C4.6 are
slightly conservative because the equation used
was slightly different tram the equation now
epec i ri ec In (Ref. 1).
=
Use L 24, we read for 1-049 tUbe a
tailing load ot 2800 lb.
As a second solution, we will use Fig.
C4..3. L' = L/ve = 24./v'"""L • 24.. L' /P =
241.3367 = 71.3. From Fig. C4.3, we read
Fc = 20000. Then Pa =
FcA =
20000 x .1464. =
2930 lb.
The answer could be obtained by substi-
tUting in equation C4.9,
Fc = 50600 4.31 L'/p
= 50600 4.31 (71.3) = 19900
Pa • 19900 x •1464. = ~ lb.
A column may also tail by local crushing
or crippling ot the tube wall, thus the
crushing stress Fcc should be determined to
see it it is less than the pr~ry bending
tailing stress tor the column.
For our tube the diameter over thickness
ratio D/t = =
1.0/.04.9 20.4.0. Values or D/t
are given in Table C4.3.
• Solution: From Table C4.4 tor 2-1/2 - .065
Referring to the small chart in the upper
right hand corner ot Fig. C4.3, we tind for a
D/t ot 20.4. that Fcc = 4.7500 pSi. Since this
stress is greater than the bending failing
column stress at 20,000, it is not critical.
Case 2. same as Case 1 but use c = 1.5 and
change material to 6061-T6 aluminum
alloy.
L' = Llve= 24./v'"T.'S = 19.7, L' /p = 19.7/.3367 Thus Fc Is critical and Pa = 46500 x .4972 =
.: 58.5
From Fig. C4..3, Fc = 22S00
Whence Pa = 22500 x .14$4 :: 3300 lb.
Fcc tor D/t = 20.4. from Fig. C4..3 • 38500 (not
crttical).
Case 3. Same as Case 2 but change material to
magnesium alloy, Fcy :: 10,000.
For L'/p = 58.5 and using lower curve on
Fig. C4.4, we read Fe = 7600. Then Pa = FcA -=
7600 x .1464. = 1110 lb.
C4.15 Strength of Streamline Tubing.
If a round tUbe is exposed to the air-
stream, the air drag 15 about 15 times greater
than if it were given a streamlined shape, thus
streamline tubes are used when the member is
exposed to the airstream.
Streamline tubes are drawn from ro~~d
tubes. In designating a streamline tUbe, the
round tube from which it was made is used and
then the fineness ratio is also given. The
fineness ratio is the ratio LID, which dimen-
sions are shown in Fig. C4.8. The ~ost common
fineness ratio used is 2.5 to 1. Table C4.4
shows the section properties of streamline
tUbing having a fineness ratio of 2.5 to 1.
Figs. C4.9 and C4.l0 give curves for finding
the column failing stress F c and the local
crushing stress Fcc.
Fig. C4.8
B '" . 944L R '" • 19D
o D '" . ~714d L :: 2. 5D
d =- Equiv. Round Dia.
PROBLEM 1. Case 1.
A streamline tube made tram a basic round
tube of 2-1/2 - .065 size has a fineness ratio
at 2.5 to 1. The length L is 30 in. T~~e
c • 1. ~~terial is alloy steel Fty = 75000.
Find the ultimate compressive load the member
will carry.
size we find the following section propertles:-
Area (A) =
.4972, P (major axiS) =
.5137 In.
= =
Then L' = L/V-C- 30/y-r- 30, and L'/p =
30/.5137 = S8.5
D/t value for tube:: 2.5/.065 :: 38.5
From Fig. C4.9 tor L'/p = 58.5 and D/t =
38.5, we read Fe = 46500 psi. For D/t = 38.5
and reading from small chart in upper right
hand corner at Fig. C4.9, we read Fcc = 66500.
23000 lb •
Case 2. Same as Case 1 but change material
to 2024-T6 aluminum alloy.
For this material we use Fig. C4.l0.
For L'/p =
58.5, we read Fc = 26000 psi.
For D/t = 38.5, we read Fcc = 37500 (not
critical). Thus Pa = 26000 x .4972 = 12900 lb.
_ C4. 16 strength of Oval and Square Shaped Tubes
in Compression.
Tables C4.6 and C4.7 give the section
properties for square and oval Shaped tubes
respectively. For the design of these Shaped
 ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES C4.13
Streamline ·TABLE C4.4 SECTION PROPERTIES OF STREAMLINE TUBING (Fineness Ratio 2. 5 to 1)

1~ .O\S 20 1.0114 4286 07116 167} oem ~

z
Mat<><

14116 con
,
Minot

J~
•
Minot
~"i¥
RO""G
0.0
Wail
Dec'! Ca,. Mo;ot

s.rm
A...
Mil\Ot "0
,m
..,~
z •
MalO'" M-o.,ot
,
Minot •
Miaot

"'.
~, IS \.4185 .10t8
0021 OIOl .14lS a ..
049 la .1079 }668
"'" "" ~,
" "50
."',
"51 .nu
1> on i.isco
"" ""u
.
10 .3140 0018 0111 .11S; 0118 0195 Hll 1611
sooc "'''
049
osa
18
11
1272
.1489
4113
.5061
OO}1
0041
0148 1703
0168 .1671 "" ,W,
'50'
0181
nIS
·H9S "" a
''''
7178
.m
.1815
"
."
,
ms
.cse
20
ta
I.H8S
, ...
.1Oti!
"'"
.m
.~Bl5
00"
00",
~,
0151
0100
0118
.10211
1912
1919
'm!
01)9
0278
0151
0145 .,..
ocr .,,,
.oco 1J( 049
". "
18 3.706S 1.)714
....
"" ,L""
.490'
1.414 .J'"
1'9)

1Yi .O}S 20 .",

1116
.,,, .con 0196 1291 .01S? .OJl9.,,,
.~,
.oe "
"" "
..nc n
""
"" ,.... "M
1·364
24n
1'765

"''''..
1.1170
"" "" coer 0253 2140 0341
"'" .'" 4S73 a
.1914

". ""
- "" '.m ""
~, }.l11
02n 1105 ~,
.." OOM
"'''
~,

" usn
2165 .7]~9 OIOl
00..
0320 1181
"'" "''' 4'"5
'" "
~,
11 4.04~~ 1.1141
""
'''' "'''
,m
1.0}7 2311

"" "''''
0246 2559 .o;SS .'51'\04
,." " ''' wn 1.~86

. "'"
lJ.( 0}5 20 714l

"a"
.083
ollS 1509

''"' ,." .w"

0116 .)119
~,

". ""
.18049
1111
.14ll)
618~

.US
OIl)
0145
Ol11
0406
14TI
241\ .0618
""
.0714 ""
.no
", »,
"'''
1.08S1
\.}959
''''
3.691
4.146
.3925
.4814
,,,
~,

"
Hi

Hi
015
~,

". """
~,

on
20 U'\04} 78S7 ien
,~,

aacc
.167S
..""""
8158
.""
.0118
.01P
.0181
.019$
0300 .UlS
0400
.~I
.0'50'1
.cns
.1m
1745
..719
i096
"'" ''''
~" oeec

""
cts

"'" "'''
~n om
""
,.,.
""
.S581
""os,
m
.nc
""
"a"
,.
4.3818 1.8S71

"'"
,,, .
.5816
S1~8

1.1800
Lm
2.111
50m
},Wl
4,011
""
.2961

""
"50
50"
20 1.0m 8S71 1611
"" "" "'" '"
1.5186 H6} .61~1
~,

.ota "
"" ""
.oe
"
U"
.1617
19H
50"
7S93
.m '''"
"" ""
1.1106
0239

.0316
.048} .304S
.OSSS
.0611
.on7 .29n
.)014
.1989
"'"
"'"
.nu
.08IS

''''' "'''
'''''
"",
6161
.61)1
.,,,
.6108
~,

"" ""a
""
16 4.noo 2.lXlOO .7014
89LO
1,0162.
2.38'
3-019
).4S'
.m,"
.~1}4

.. .....
.nc 1.1141 4.})1
,."" """
'" ".
om 1.64l4 m.
Hi 0)5
.oe
". ""
20 1.191)
"" '''' ""
"" ",.
814S
"
0116
051}
})6S
3314 """
use "'"
''''' "'"
. ."
.1416
1B~~
"'" on8 0448 3181
",n
.1l14 .1\6~9 Hi .065 16 '.0570 1.1418
"" USB

...,,,
" .oa
"" ""
~,

.m ,.,
1.08}
1.361 .0410
0118 m6
088} .3191
m'
".,
124S
,~,

1732
6)74
.",
.cs
.110
"na "" ,m
L"""
1.168S
l.I,1
}.708
4.6S1
~"
".
0976 .3149
" 1.~51

.", ,"
. !•• 1.7641

"''' "'50
..
116}
.on
",U ""
.0498

..,
Hi 20 1.]600 1.<:'000 )Ii}}
~,

". "rr
1MB
,""
"50 0611
ens
}583
'M' Ill}
nl1
.1116 O6S
.os}
16
14 "'''
1.0114
''''' "'""
1.170 50 .08106
.3551
lSlS
1~91
.176S
, "'"
.14U .116} .095 I} 1.16n

"""" "" .4341 1.478 0511 1042 3461 1191 .1101 .nc 11 L..,.,
L...,
"" " 49}9 1.619 'm .11'\04 HI8
"" ."''' ,.,,,
.7061 .156
.188
~
H. 1.14S1
ms Ion 6818 0515 .}902 ua
"''' ,.,,, i: .'"
"" '''"
Iii 10 1.S18S 1.0714
~, 1!11 .0411 .0719 }SSI .168S .1296 .06' '.1.~1l LUllS

'" "rt .HI1

,..e
1.116
"'~'"" 0900 lBI5
'''' .lSlS .TIll
""
."".uc n
L"'"
1.1<\00
"" "
""
.oss "" .-1613
1.1~7
\.jBiI
L"" "'"
OHI 0993 .J~
.1114 .l119
.1HS .l6S7
.2186
""'
.mll "'''
.2719
.7690
7614
.," ,.,.a 1."10
,."'"
5111
"" "" .'M 1.J930
1 O}S
.""
10 1.6970 1.142.8 1161 7]4S
urn
0)16
"'''
"'" ""
50"
.1076

""
8113

'"'
H M683 1.5113 USH

aeea ""
OSlO
". "" "'" .cs 1.;141
~,

""
.m """
lH9
j9H

,."
""
1.203
I.Hl ""
~"
r.sn ''"'
1.699

"'"
.10}4
1141
119B
.1SS6
,~

.Hl'
.}7+4 ""
.1731
.I!m
nc
!2111
8117
.SIS6
.8115
. ,.'"
''''
.,"
,us ,."
51 1.6~ll
2.1111
1.'~403
1.9411
".,
2)4 049
.osa
18

""
J-O}4} I.l.8S7
...,
.l388
J""
1.l~1
\.]~S
'm
"'''
.1144
.1l19 .14SO
.<S"
.1111 ,os.
9Jt5

sm
""
"" ""
.l8l'
.,,'" LSI7
"'"
~,

.oe
" 1.911 1161 ,," 50" .9ll)

'" "
.uc u
.6431

"''''
11116
1.7}0
.1198
.15)0
.2020
2411
,."
.,." ""H70

"""
.42"

30
I
60 40
I ~50"
r-, 50

I I i"
20 40 so 80 00
,OJ' I
30

OtBoSlC rour'lCl tube dlornefe~

II! I
°020~60ao »c
=~
i
00
L'/p . ," Fig. C4.'9

10 _ ALLOW ABLE COLUMN AND CRUSHING STRESSES

. CHROME MOLYBDENUM STREAMLINE TUBING. ~,
Fig. C4.10 Streamline 2024-T3 Tubing
F ty = 75000 PSI ~

o~ .
o 20 40 60 80 100 120 14<l
Lip
 STRENGTH & DESIGN OF ROUND, STREAMLiNE, OVAL AND SQUARE TUBING
C4.14 iN TENSION, COMPRESSION, BENDING, TORSION AND COMBffiED LOADINGS.

Squares TABLE C4.6 SECTION PROPERTIES OF STREAMLINE TUBING

-,
~.

... G••

," .035
'"ra 6549 oes
OW, 0881

'"
'"
'"
..
"uzo ~~

"
~"
=
om
coos
oon
oo~
~,
c..
".
.ce,
.
"" t.cu
9O'l,
1.195
.1051
"OJ
.1203
.10101

"'"
. ".
-
re oou Ute
~, ~u
"'''
~,
.081
O'l, 1} 1.i17
1436
,,.,
.1608
1915
"'.
'" "'''
."" u

.. cots 2.136 .2586

~w
"''''
,~

.on
"
"
w ~" "''' .c-s
"'''' "'" '" "'"is
1.065 .1451 .1658
~,

"""
oon ",. .oss
.
,~
~,
"'"
.,"" oon al41
.. "
1.25'
1.402
.1691
.1877
.1932
.1146
,."
-
00" 1.774
""O'l' .23U