Chapter No.
Al
..
...
A2 A3 A4 A5 A6 A7
r
j,
;,
Statically Indeterminate Structures. Theorem of Least Work. Virtual Work. Matrix Methods. Bending Moments in Frames and Rings by Elastic Center Method. Column Analogy Method. Continuous Structures' Moment Distribution Method. Slope Deflection Method.
BEAM BENDING AND SHEAR STRESSES. MEMBRANE STRESSES. COLUMN AND PLATE INSTABILITY.
A13 A14 A15 A16 A17 A18 Bending Stresses. Bending Shear Stresses  Solid and Open Sections Shear Center. Shear Flow in Closed ThinWalled Sections. Membrane Stresses in Pressure Vessels. Bending of Plates. Theory of the Instability of Columns and Thin Sheets.
The 3Dimensional Equations of Thermoelasticity. The 2Dimensional Equations of Elasticity and Thermoelasticity. Selected Problems in Elasticity and Thermoelasticity.
;...,.....
l _ ~
....
.j
13;
Chapter No.
82
Basic Principles and Definitions. Mechanical and Physical Properties of Metallic Materials for Flight Vehicle Structures.
Combined Stresses. Theory of Yield and Ultimate Failure. Strength of Columns with Stable CrossSections. Yield and Ultimate Strength in Bending. Strength and Design of Round. Streamline, Oval and Square Tubing in Tension, Compression. Bending,
Torsion and Combined Loadings.
C12 C13
Buckling Strength of Flat Sheet in Compression, Shear, Bending and Under Combined Stress Systems. Local Buckling Stress for Composite Shapes. Crippling Strength of Composite Shapes and SheetStiffener Panels in Compression. Column Strength. Buckling Strength of Monocoque Cylinders. Buckling Strength of Curved Sheet Panels and Spherical Plates. Ultimate Strength of Stiffened Curved Sheet Structures. Design of Metal Beams. Web Shear Resistant (NonBuckling) Type. Part 1. Flat Sheet Web with Vertical Stiffeners. Part 2. Other Types of NonBuckling Webs. Diagonal SemiTension Field Design. Part 1. Beams with Flat Webs. Part 2. Curved Web Systems. Sandwich Construction and Design. Fatigue.
Appendix A
INDEX
Accelerated MoUon of Rigid Airplane. Aircraft Bolts . Aircraft Nuts Aircrait Wing Sections 
A4. 8
D1.2 D1. 2
Type, A19. I Aircraft Wing Structure Truss Type. A2.14 Air Forces on Wing . A4.4 Allowable Stresses (and Interactions) . Cll.36 Analysis of Frame with Pinned Supports . A9.l6 Angle Method C7.1 Application of Matrix Methods to Various Structures A7.23 A4.1 Applied Load Axis of Symmetry. A9.4
Beaded Webs Beam Design  Special Cases. Beam Ftxed End Moments by Method of Area Moments Beam Rivet Des Ign Beam Shear and Bending Moment Beams  Forces at a Section Beams Moment Diagrams. Beams with NonParallel Flanges Beams  Shear and Moment Diagrams aeams Statically Determinate & Indeterminate . Bending and Compression of Columns Bending Moments  Elastic Center Method. Bending of Rectangular Plates Bending Strength  Basic Approach. Bending Strength  Example Problems Bendtng Strength of Round Tube' Bending Strength  Solid Round Bar. Bending Stresses Bending Stresses  Curved Beams Bending Stresses  Elastic Range Bending Stresses  Nonhcmcgenecus Sections. Bending Stresses About prmcipal Axes . Bending of Thin Plates Bolt Bending Strength. Bolt & Lug Strength AnalysiS Methods Eolt Shear, Tension &
> >
Buckling of Stiffened Flat Sheets under Longitudinal Compression Buckling under Bending Loads Buckling under Shear Loads. Buckling under Transverse Shear Carry Over Factor casngnaao'e Theorem Centroids  Center of Gravity. Cladding Reduction Factors. Column Analogy Method. Column Curves  NonDimens tonal . Column Curves  Solution Column End Restraint. Column Formulas . Column Strength. Column Strength with Known End Restraining Moment Combined Axial and Transverse Loads  General Action COmbined Bending and Compression. Combined Bending and Flexural Shear. . Combined Bending and Tension Combined Bending and Tension or Compression of Thin Plates Combined Bending & Torsion. Combined Stress Equations . Compatability Equations. Complex Bending' Symmetrical Section. Compressive Buckling Stress for Flanged Elements . COnical Shells  Buckling Strength Constant Shear Flow Webs Constant Shear Flow Webs ~ Single Cell  2 Flange Beam. Constant Shear Flow Webs Single Cell ~ 3 Flange Beam. Continuous Structures Curved Members Continuous Structures Variable Moment of Inertia COre Shear Correction for Cladding. Corrugated Core Sandwich Failure Modes. Cozzone Procedure Creep of Ma.terials Creep Pattern . Crippling Stresses Calculations . Critical Shear Stress Crystallization Theory Cumulative Damage Theory. Curved Beams Curved Sheet Panels Buckling Stress Curved Web Systems CutOUts in Webs or Skin Panels . Deflection Limitations in Plate Analyses. . . Deflections by Elaanc Weights
C2.2 C2.l3 C2.1 C4.2 C7.21 C2.16 A5.21 C4.22 C3.l0 C4.23
A18.l7 C4.23
Deflections by Moment Areas. A7.30 Deflections for Thermal Strains . A7.17 Deflections by Virtual Work A7.9 Delta Wing Example Problem. A23.2 Design for Compression C4.2 Design Conditions and Deaign Weights .. A5.12 Design Flight Requirements for Airplane . A4.6 Design Loads . A4.1 Design for Tension . . C4. 1 Differential Equation of ALB. 12 Deflection Surface. AZD.15 Dtsconttnutttes Distribution of Loads to Sheet Panels . A2l.2 Ductility. . . B1. 5 Dummy Unit Loads A8.6 Dynamic Effect of Air Forces. M.13 Effect of AxIal Load on Moment Distribution. Effective Sheet Widths Elastic Buckling Strength of Flat Sheet in Compression. Elasttc  Inelastic Action. Elastic Latera! SUpPOrt Columns. Elastic Stability of Column Elastic Strain Energy . Elasticity and Thermoelasticity  OneDimensional Problems ... Elasticity and Thermoelasticity  TwoDimensional Equations Electric Arc Welding . . End Bay Effects. End Moments for Continuous Frameworks . Equations of Static Equilibrium Equilibrium Equations
All. 22 C7.l0
C5.1
Bt. 5
C2.l7 .'\17.2 ci, 6
A26.l
en, 9
AS.2 A5.1 AlB. 1 A9. 1 Ala. 13 <:3.1 C3.4
ci. 2
All. 10
A2.l
A24.2
cr. 15
C3.1 A13.1
A13. 15
A15.3
A15.5 All. 31 All. 15 C12.26 C7.4 e12.27 C3.2
31. 8
B1. 12
Ber:ding Strengths
D1.3
Boundary Conditions Box Beams Analysis Brazing Buckling Coefficient Buckling of Flat Panels wiui
A24.8
A22.5
D2.4
CS.l
C12.25 CS.6
C9.1 Cll.29
D3.7
ntsetmuer Feces
Buckling of Flat Sheets under Combined Loads. Buckli.ng or Rectangular Plates
A18.20
A17.4 .\7.27
Failure of Columns by Compression. A18.4 Failure Modes in Curved Honeycomb Panels. . . C12.20 FaUure of Structures 81. 1 Fatigue AnalySiS  Statistical Distri.bution ... . C13.4 Fatigue and FailSafe Design . C13.8 Fatigue of Materials B1. 14 Fatigle SN Curves . C13.13 Fillers. 03.5 Fitting Design . 01. ~ Fixed End Moments All. :3 Fjxed End Moments Due to Support Deflections All. 9 Fixi.ty cceutcterus. C2. 1 Flange Design . CIO.l Flan~e Design Stresses . ClG.2 Flange Discontinuities. ClO.7 Flange Loads Cll. a Flange Strength (Crippling) . CIO.4 Flat Sheet Web with Vertical Stiffeners CIO.l Flexural Shear Flow Distribution AlS.24 Flexural Shear Flow Symmetrical Beam Section A14.5 Flexural Shear Stress. A14.1
INDEX Continued
Static Tension StressStrain 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 Carryover 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 of Plates. . . . . . . . AlB. 12 Streamline Tubing  Strength. C4. 12 Strength Checking and Design  Problems . . . C4.5 Stren~_ ".: Round Tubes _ ..nder Combined Loadings . C4.22 sness Analysis Formulas Cll. 15 Stress Analysis of Thin Skin Multiple stringer Cantilever Wing . . . . . . . . . . A19.10 Stress Concentration Factors. C13.10 Stress Distribution & Angle of Twist for 2Cell Thin Wall Closed Section . . A.6.7 StressStrain Curve . . B1.7 StressStrain Relations . A24.6 Stresses around Panel Cutout. A22.1 Stresses in Uprights Cll.17 Stringer Systems in Diagonal Tension . . . . . . Cll.32 Structural Design Philosophy. ci, 6 Structural Fittings . . A2.2 Structural Skin Panel Details. D3.12 Structures with Curved Members . . . . . All. 29 Successive Approximation Method for Multiple Cell Beams . . . . . . . AlS.24 Symbols for Reacting
4
ToUlgent Modulus TangentModulus Theory Taxi Loads . . . Tension CHps . . . . . TensionField Beam Action. 'renetcn Field Beam Formulas Theorem of Castiglta.no . . Theorem of Complementary Energy. . . . . . Theorem of Least Work . Theorems of Virtual Work and Minimum Potential Energy Thermal Deflections by MatriX Methods . . Thermal Stresses . . Thermal Stresses . . . Thermoelasttcity  ThreeDimensional Equations. Thin Walled Shells Three Cell  Multiple Flange Beam.  Symmetrical about
One Axis . . . .
Bl.5
A1B.a C13.33 D3.2 en. 1 C11. 2 A7.5
Trusses With Multiple Redundancy . . . A8.11 Trusses With Single Redundancy . . . . AS. '7 Tubing Design Facts C4.5 TwoDimensional Problems. A.26. 5 TwoCell Multiple Flange Beam. One Axis ot Symmetry . . . . Al5.11 Type of Wing Ribs. . A2l.t tntimate Strength in Combined Bending &: F1exural Shear tnUmate Strength in Combined Compression, Bending, Flexural Shear &: Torsion. Ultimate Strength in Combined Compression, Bending &I Torsion . . . illtimate Strength in Combined Tension. Torsion and Internal Pressure p in psi. Uniform Stress Condition. Unit Analysis lor Fuselage Shears and Moments. . UnsymmetriCal Frame . . Unsymmetrical Frames or Unsymmetrical Frames using PrinCi~ Axes. . . .'Tnsymmeirical Structures " , \t, . 'Jy  Load Factor " .("rC4.25 C4.26
Ct. 24
C4.26 Cl.I
AS. IS" A9.2
AIO.4
Torsion  Circular Sections. Torsion  Effect of End Restraint . . . . . Torsion  Noncircular Sections . . . Torsion Open sections Torsion of ThinWailed Cylinder having Closed Type StiHeners . . . . . . . . . Torsion Thin Walled Sections. Torsional Moments  Beams Torsional Modulus of Rupture. Torsional Shear Flow in Multiple Cell Beams by Method of Successive Corrections . . . Torsional Shear Stresses in Multtple ThinWall Closed Section  Distribution Torsional Strength of Round
4Cell
Al9.5 A6.l
A8.16 A6.3
AS. 4
R_
M.tS A9.13
.'
AS. IS
A6.S
A5.9 C4.1'7
.....
A4.7
':"'A
:''"it
A6.10
AS.7
Tubes Tubes
.....
4
Fitting Units . . . .
A2.3
A14.2
Transmission of Power by Cylindrical Shaft. . . . TriaXial Stresses . . Truss Deflection by Method of Elastic Weights Truss Structures Trusses with Double Redundancy. . .
A7.33
A2.9
AS. 10
Wagner Equations .. Cl1.4 Web Bending &: Shear Stresses CtO.5 Web Design . . . Cll.18 Web Splices . . . . CIO. 10 Web Strength. Stable Webs. CtO.5 Webs with Round Lightening. Holes . . ClO.17 Wing Analysis Problems A19.2 Wtng Arrangements. Al9.1 Wing Effective Sectton A19.12 Wing Internal Stresses A23.14 Wing Shear and Bending AnaLysis . . . A19.14 Wing Shear and Bending Moments . . . . A5.9 Wing  Sbear Lag . A19.25 Wing Shears and Moments AS. 10 Wing Stiffness Matrix.. A23.11 Wing Strength ReqUirements A19.5 Wing Stress Analyl'lis Methods .'1.19.5 Wing  Ultimate Strength . A19.11 Work of Structures Group. Al.2
C7.2C
CHAPTER Al
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 takeoff 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,
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 understanding 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 setup that is necessarJ. I. II. Preliminary Design Section. TecrJlical (1) (2) (3) (4) (5) (5;
_~alysis
Section.
Aerodynamics Group Structures Group ~eight and 3alance Control Group Power Plant Analysis Group Materials and Processes Group Centrols AnalYSiS Group
III. Component DeSign Section. (1) (2) Structural DeSign Group (~lng. Body and Control Surfaces) Systems Design Group (All mechanical, hydraulic, electrical and ther.nal installations) Tests Section.
The ~odern commercial airliner, militarJ airplane, missile and space vehicle is a highly
Al.I
IV.
Laborato~J
Al.2
ENGINEER
(1)
(2)
(3)
(4) (5) (6) (7)
Wind Tunnel and Fluid Mechanic5 ~est Labs. Structural Test Labs. Propulsi~n Test Labs. Electronics ~est Labs. ElectroMechanical Test Labs. Weapons and Controls Test Labs. Ar~log and Digital Computer Labs.
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 swumary tables. The final results ~y 61v8 complete shear, moment and no~l forc~s =e~er=~d to a convenient set of :CY2 axes for major airc ra.r t units such as the Wing, rus eIage , e t c . THE WORK OF STRESS ANALYSIS
~\m S~R~~GTH
v.
VI.
GROUP
Flight Test Section. Engineering Field Service Section. 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 crosssectional 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~ requirements. ~he stress group ~ust consta~tly work closely with the Structural DeSign Sect:Gn in order to evolve the best structural overall arrangement. Such factors as ~ower ~lants, bUilt in fuel tanks, landing gear retracting wells, and other large cutouts 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, elevated 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
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 aircraft 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 detrimental 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 sUbgroups as tollows:(1) (2) (3)
(4)
Applied Loads Calculation Group Stress AnalySiS an~ Strength Group Dynamics AnalYSiS Group Special Projects and Research Group
GROUP
Before any part ot the structure can be finally proportioned relative to strength or rigidity, the true external loads on the aircraft 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.
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 presented 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 or danger of failure and therefore the dynamic
ANALYSIS
AND DESIGN OF
A1.3
groups have a speCial SUbgroup which are working on deSign problems that Nill be encountered in the near 1r distant future as aviation proThe major structural units of aircraft such gresses. For example, in the r.t ruc turea Group, as the wing and fuselage are not rigid bodies. this subgroup might be studying such problems 7hus when a Sharp air gust strikes a fleXible as: (1) how to calculate the thermal stresses wing in high speed flight, we have a dynamic in the wing structure at supersonic speedS; load situation and the wing Nill vibrate. The (2) how to stress analyze a new type of wing dynamicist must determine whether this vibration structure; (3) what type of body str~cture is 1s serious relative to induced stresses on the best for future space travel and what kind of wir~ structure. The dynamics group is also materials will be needed, etc. responsible for the determination of the stability and performance of miSSile and flight Chart 1 illustrates in general a typical vehicle guidar.ce and control systems. The ~keup of the Structures Section of a large dynamics group must work constantly with the aerospace company. Chart 2 lists the many various test laboratories in order to obtain items which the structures engineer must be reliable values of certain factors that are concerned with in insuring the structural necessary in many theoretical calculations. integrity of the flight vehicle. Both Charts 1 and 2 are from ChanceVought Structures THE ',jaRK OF THE SPEC IAL PROJECTS GROUP DeSign Manual and are reproduced with their permi saton. In general, all the various technical
characteristics must be changed or modified in order to insure reliable and safe operation.
U..f ,. ~EIlOE''\STlc:'4
~'IlOE''\STIC cU"", i
,.. l ,
11:'
'
lAKlIlATo'" I
~~u,
I STWCTVI~S
""
I.
, .c...."I.IC
'NO
: UN"
"Owu
, .....c:..
~~,
'H~
~'''G''
COMPUIATICH GIlOU'
:3
e.13W
ETT?
5ZWF
AI. 4
AEROSPACE STRUCTURES
ENGINEER
MATERIALS OF CONSTRUCTION
FA.STENERS WELDING BONDING ?LATE~ND BAR FORGINGS CASTINGS XTRUSIONS SHEET METAL SANDWICH ?tASTIC l,OMINAIT BEARINGS
STIFFNESS CRITERIA
FLUTTER
CONTROL SYSTEM STABILITY PANEL fLIJTTERSKIN CONTOURS CONTROL SYSTEM DEfLECTIONS
STRESS ANALYSIS
SKIN PANElS BEAM ANALYSIS STRAIN COMPATIBILITY STRAIN CONCENTRATION JOINT ANALYSIS BEARI~G ANALYSIS BULKHEAD ANALYSI S fITTING ANALYSIS i1'ERMAL STRESS ,'oIECHAN1CAL COMPONENTS D:PERIMENTAL STRESS ANALYSIS
THERMAL EFFECTS
MtCHANICAL VIBRATIONS
COMPONENT ANALYSIS
UNIT SOLUTIONS
ALLOWABLES
YIHDING FRACTURE ATlGUE WEAR, BRINELLING CREEP DEFliCTIONS Tl'ERMAl O"FECTS STIFFNESS COMBINED LOADINGS 3UCKlING
RESIDUAL STRESS HEAT TREAT CONTROL STRESS CORROSION STABILITY AT TE~PRATURE SPECIFICATION CONFORMANCE BLUE PRINT CONFORMANCE
DfFlICTION ANALYSIS
STIFFNESS
CHAPTER A2
A2.1 Introduction.
equilibrium must constantly be used by the stress analyst and structural designer 1~ obtaining 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.
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:
ZFy = a
l:Fz
l:F x
=a
=0
or
m, = o m, = o m, = o
    (2.2)
To completely Cefl~e a force, we must know its ~agnltude, direction and ~olnt of ann1ication. 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 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
A force acting in space is completely defired :: 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 sueartzeo .
EQ.UILI3RIUi1 SO.UATICNS FOR GSNERAL SPACE: urCNCOPLANAR) ?ORCE SYSTE?:
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 resultant 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,
SG.UILIBRIUM
ZI1x ::: 0,
n1z:=
 (2.3)
OF GE:N""'..2.AL
l:F x ZF y
= 0 = 0 ZF z = 0
m, = a m, = 0
~ ...
= 0
    (2.1)
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 acre 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
Co~curre~~ 7;e~ns
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 equilibri~ equations could be used.
=0 =0
ZFx
=0
2.4
ropes
SYS':'~
(':'he subsc=lpts 1, 2 and 3 refer to different locations for z axes or moment centers . )
.AC
A2.2
TRUSS STRUCTURES.
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 combinations available are,
l:F x
=0 }
= O
2.5
reacted 'Jy other ext erne ; orces , 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 establiShing the ~orcs c~aracterist1cs of dirsction and pOint of application. Ball and Socket Fitting
(The z axis or moment center locations must be other than through the paint of concurrency)
EQUILIBRIUM: OF COPLANAR P.4RALLt:L FCRCE
SYS'I'~
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~ available 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 equilibrium equations available would be: l:F y : 0 n1 z
For any space or coplanar f8rce suc~ as ? and Q acting on the bar, the line o~ action of such forces must act t..ucucn 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.
=0
or
l:l1z. : a
4I'1z
Ii
=a
2.6
~'l

II
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
For any force such as P and Q acting i~ the xy plane, the line of action of such a ~~rce
~ust
ZF
=a
or
ZI1 1 = O
2.7
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.
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 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 .
<@o========:j):Jl_
If a bar AB has si~Gle pin f~t~ings at 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.
eac~
,B
To completely define a force in space reqUires 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
A2.3
L.
y
Fh"
liillirn
//11
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 first column. Any reacting force at joint (A) ~ust be horizontal since the support at (A) is so designed to provide no vertical resistance. Cables  Tie Rods
L
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
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 direction 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.
~P
~~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 rurtner attached to a fitting portion that prevents any horizontal 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 rollerpin 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.
Ty~e
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
t
tRy
of Fitting
",4",
Lubricated Slot ! ~
b'RI=Jn'
Double RoUer
~ ',~
The above graphical symbols represent a reaction in which translation of the attachment point (b) is prevented but rotation of the 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 direction as an unknown, it is more convenient to replace the resultant reaction by two components at right angles to each other as indicated in the sketches. /1 .
~,
A2.4
TRUSS STRUCTURES.
l~eEdge
!
I
Rollers
~ FA
I
Roller.
minate, and the degree of redundancy dapends on the number of ~~nowns beyond that ~umber which can be found by the equations o~ static eqUilibrium. A structure can be statically indeterminate 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 procedure is required for design analYSiS of statically indete~inate structures.
A2. 6 Examples of Statically Determinate and Statically Indeterminate Structures.
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
rrrrrt
(bJM.,
'" 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 reaction is campletely unknown since all 3 force characteristics are unknown, namely, magnt 'tude , direction and point of application. It Is convenient 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.
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 knowing 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.
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 external reactions or internal members than is necessary for stability ot the structure under a given load system it is statically ir.deter
Fig. A2.1
In the st~~cture shown in Fig. 2.1, the 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
~~Qwn
A2.5
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 reaction is unknown in direction because 2 struts enter jOint C. Magnitude is also unknown but paint of application is known since the reaction magnf tudes , must pass through C. Thus we have 5 unknowns, E:x:ample Problem 2. namely, Re, Ra, RO, Vc and He. For a copjanar force system we have 3 equilibrium equations P available and thus the first conclusion might Pin be that we have a statically indeterminate structure to (53) = 2 degrees redundant. However, 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 with the 3 equations ot equilibrium. Thus Fig. A.2.3 drawing tree bodies of the structure to lett at Fig. A2. 2 pin E and to right or pin F and equating moments about each pin to zero we obtain 2 equations ~lg. 2.2 shows a structural frame carrying which do not includ& unknowns other than the 5 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 thereis known tor each reaction, however, the magnitore statically determinate. tude and direction of each is unknown making a total of 4 unknowns With only 3 equations of Example Problem 4. 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 .. c 0 equilibrium, then each part must likewise be Vc in equilibrium and we can cut out any portion as a f~ee bOdy and apply the equilibrium Fig. A2.6 Fig. A2.5 equations. Fig. 2.3 shows a free bOdY of the frame to left of pin at C. Taking moments Fig. 2.5 shows a beam AB which carries a about C and equating to zero gives us a fourth superstructure CED which in turn is SUbjected to the known loads P and Q. The question is 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 The external unknown reactions for the entire they have no ~oment about C because of zero structure are at points A and B. At A due to ams. As in exa~ple problem 1, the reactions the roller type of action, magnitude is the only at A and B have been rep Iac ed by H and V comunknown characteristic ot the reaction since ponents ins:ead of ~sing an angle (direction) direction and point ot application are known. At B, ~gnitude and direction are unknown but as an unknown characteristic. The struct~e is statica:ly dete~inate. paint of application 1s known, hence we have 3 unknowns, namely, RA, VB and Hg, and with 3 Exa'Ilple Problem 3. equations ot equilibrium available we can find these reactions and therefore the structure is statically dete~inate with respect to external p p p P reactions. we now investigate to see it the t E I 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 ,He fRB .J.VC body of the superst~cture as illustrated in ?ig. 2.6 and consider the internal forces that Fig. A2.4 eXisted at e and 0 as exterr~l reactions. In the actual structure the members are rigidly ~ig. 2.4 shows a straight ~ember 12 carrying a attached together at point C such as a Nelded or known load system ? and supported by 5 struts
of 3. with 3 equations of equilibrium available 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
):0
1;A
""llC'fMc "'Rn
(1)~
~In \~S::
.~
J 1 'L .
' _
'\{4x5(
A2.6
~ultiple
TRUSS STRUCTURES.
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 reaction 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 triangle CEDC. Example Problem 5
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 wtether 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 11 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 unknowns 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 22 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.
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 available. 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 reaction are unknown. At point B, due to pin only 2 unknowns, na~ely, ~gnitude and direction, 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.
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 44 and use the lower portien as a free body. The same reasoning as used above would show us we have one more unknown than the numberof 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
A2.7
the strJcture Nould be still stable and all ~ine the axial loads in the members and the reaxial stresses could be found by the actions on the spar. equations of static equilibrium without regard to their size of crosssection or the kind at Solution: The first thing to decide is whether material. Adding the second diagonal member the structure is statically determinate. From in each panel would necessitate knOWing the the figure it is observed that the wing spar is size of all truss members and the kind of supported by five struts. Due to the pins at material used before member stresSes could be each end of all struts, we r~ve five unknowns, found, as the additional equations needed must ~~elYJ the magnitude of the load in each strut. come from a consideration involving distortion DirectIon and location at each strut load is of the truss. Assume for exa~ple, that one known because of the pin at each end at the diagonal in the upper panel was lett out. We struts. We have 3 equations of equilibrium for would then be able to find the stresses in the the wing spar as a single unit support'ed by the ~embers of the_upper ?anel by statics but the 5 struts, thus two ~ore equations are necessary lower panel Would still be statically indeit the 5 unknown strut loads are to be found. terminate to 1 degree because or the double It is noticed that the wing spar includes 2 indiagonal system and thus one additional equation ternal single pin connections at points a and 0'. is necessary and would involve a consideration This establishes the fact that the moment of all of truss distortion. (The solution ot staticforces located to one side of the pin must be ally indeterminate trusses is covered in equal to zero since the single pin fitting canChapter A.B.) not resist a moment. Thus we obtain two additional equations because of the ~NO internal pin A2.7 Example Problem Solutions of Statically Determinate fittings and thus we have 5 equations to :ind 5 Coplanar Structures and Coplanar Loadings. unknowns. Although a student has taken a course in Fig. 2.15 shows a tree body of the wing sta:ics ~efore taking a beginning course in spar to the right of hinge fitting at O. aircraft structures, it is felt that a limited ~eview of )roblems involving the application 1013",(30+ 15)45 of the equations of static equilibrium is quite I 20~ 2 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 Fig. A2.15 static equilibrium as ?art ot the necessary equations in solving statically indeterminate st~uctures and since statically indeterminate In order to take moments, the distributed structures are covered in rather complete detail load on the spar bas been replaced by the reLn other c napt ers at :::115 book, only limited sultant load on each spar portion, namely, the s?ace will be given to ~roblems involVing total load on the portion acting through the statics in this chapter. centroid of the distributed load system. The strut .react.t on EA at A tas been shown in phantom ~a~ple Problem 8. as it is ~ore convenient to deal with its components YA and XA' The reaction at 0 15 unFig. A2.14 shows a much SI~plitied wing known tn ~gnltude and direction and for constructure, cor~lsting of a wing spar supported venience we will deal with its components xo 'oy lift and cabane struts wmch tie the wing and YO. The sense assumed is indicated on the 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 The sense of a force Is represented is ~ade in t~ree units, readily disassembled by graphically by an arrow head on ~he end of a vector. The correct sense is obtained from the using ern fittings at points 0 and 0 1 All solution ot the equations of equilibrium since, su~porting wing struts have Single pin fitting a force or ~oment must be given a plus or minus ~l~ts a: eac~ end. The problem is to deterSign in writing the equa~ions. Since the sense of a force or moment is unknown, 1t is assumed, 201ll/in. 401ll/in. l5#/in. if the algebraic solution of the equilibrium and " ' . ~Oi/in. ..I...,....,r;i t j " Ill' t f ! ,, equations gives a ~lus value to the magnitude r i 45"~ 82" iZO'!_60" ! Hinge then the true sense is as ass~ed, and oppOSite o' , ';0 to that assumed it the solution gives a minus A PIn A 8' B 0" sign. If the unknown forces are axial loads in Lilt Caban , +Struts truts C~ c 36" ~embers it is camoan practice to call tensile stress plus and compressive stress minus, thus Fuselage ' "30":.......i..Fig. A2.l4 E' E if we ass~.e the sense of an unknown axial ~oad 'i 8ym. 'i ~s tenSion. the solution of the equilibr!um
~ember
i
A2.8
TRUSS STRUCTURES.
equations will give a plus value for the magnitude of the unknown if the true stress is tension and a ~inus sign will indicate the assumed tension stresses should be reversed or compression, thus giving a consistency of signs. To find the unknown YA we take ~cments about point 0 and equate to zero for equilibrium
 2460 x 41  1013 x 102 82YA
~ake m~ents
about
sho~TI.
=0
Hence YA = 204000/82 = 2480 lb.. The plus sign means that the sense as assumed in the figure 1s correct. By geometry XA = 2480 x 117/66 = 4400 lb. and the load in strut ~ equals v4400 3 + 2480 3 = 5050 lb. tension or as assumed in the :lgure.
ZFX
whence, BIC 3535 lb. The ~inus sign means it acts opposite to that sho'Nn in fi~~e or is compression instead of tension. The reactions on the spar can now be deter.nined and shears, ber.ding moments and axial loads on the spar could be round. The numerical results should ~e checked for eqUilibrium or the spar as a whole by taking moments ot all forces about a dif:erent ~oment center to see i t the result "is zero. Example Problem 9.
=
= 0 = Xc
To find Xo we use the equilibrium equation  4400 = 0, whence Xc = 4400 lb. To find YO we use,
= 0,
whence
To check our results for eqUilibrium we will take moments at all torces about A to see it they equal zero.
l:11A
= 2460
On the spar portion OIA' J the reactions are obViOUSly equal to 40/30 times those :ound for portion OA since the external loading is 40 as compared to 30. Hence AIEl. 6750, XOI = 5880, YO' = 1325 Fig. 2.16 shows a tree body ot the center 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.
I
24lJ~ \, E'
12
i\~
strut
~ Brace Strut
12
12"
I
+
F
Fig.A2.17
1500",50x30
A
Shock
Strut
300
+
L
12"
12"
'':In,"
,
"
15
" 30
Flg.A2.16
C"9~
I I
L __
Fig. 2.17 shows a Simplified airplane landing gear unit with all members and loads confined to one plane. The brace struts ~e pinned at each end and the support at C is of the roller type, thus no vertical reaction can be produced by the support fitting at point C. The member at C can rotate on the roller but horizontal movement is ~revented. A knovm load ot 10,000 lb. 1s applied to axle ~nit at A. The problem is to f1nd the load in the brace struts and the reaction at C. Solution: Due to the Single pin fitting of the brace struts, the ~eact1ons
~t ~t
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 construcunknown. The roller type fitting at C fixes t1on. The stresses or loads in the members ot a truss are commonly referred to as nprimaryW the direction and point of application of the and wsecondarJn stresses. The stresses which reaction Re, leaving magnitude as the only are found under the following assumptions are unknown, Thus there are 3 unknowns Re, He and RO and with 3 equations of static equilibrium referred to as primary stresses. available, the structure is statically deter.ninate 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 inZME =  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 comes out with a minus Sign, the reaction RO has a sense OPPosite to that shown by the vector in Fig. 2.17. Since the reaction RO is colinear with the line DE because of the pin endS, the load in the brace strut DE is 16750 lb. compression. In the above moment equation about B, the reaction RO was resolved into vertical and rorizontal components at point D, and thus only the vertical component which equals (12/17) RD enters into the equation since the horizontal component has a line of action through point B and therefore no moment. He does not enter in equation as it has zerO moment about B.
To find RE take ZFv = 0
truss only at the jOints and in the plane ot the truss. Thus all loads or stresses produced in members are either axial tension or compression without bending or torsion. Those trusses produced in the truss members due to the nonfulfillment or the above assumptiOns are referred to as secondary stresses. Most steel tubular trusses are welded together at their ends and in other truss types, the ~embers are riveted or bolted together. This restraint at the joints may cause secondary stresses in some members greater than the primary stresses. Likewise it is common in actual practical deSign to apply torces to the truss members between their ends by supporting many equipment installations on these truss members. However, regardless of the magnitude at these socalled secondary loadS, it is cammon practice to first find the priwary stresses under the assumption outlined above.
GENERAL CRITERIA FOR DETERl'lINING WHEI'JlER TRUSS STRUCTURES ARE STATICALLY DEI'ERMINATE WITlf RESPECT TO INl'ERNAL STRFBSES.
ZFv
(24,/26.8) = 0
= 10000
x cos 30
+ ( 16750)(12/17) +
HE
Whence, HB = 3540 lb. Since Sign comes out plus, the sense is the same as assumed in the figure. The strut load BF is therefore 3540 lb. tension, since reaction HB is collnear wi th line SF.
To find
ZH
He
take ZH : 0
+ (
16750)
whence, Rc = 8407 lb. Result is plus and therefore assumed sense was correct.
~aments
The simplest truss that can be constructed is the triangle which has three members m and three joints j. A more elaborate truss consists of additional triangular frames, so arranged that each triangle adds one joint and two members. 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 eqUilibrium 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 .
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,
/'
,~ .
A2.10
TRUSS STRUCTURES.
500
However three i~dependent to determine the ex~er nal reactions, thus the number of equations necessary to solve :or all the loads in the members is 2j  3. Hence if the nw~Jer of truss ~embers is that given by equation (2.8) the truss is statically deter.ninate relative to the primary loads in the truss ~embers and the truss is also stable.
equa~ions a~e neceSsa~J
equations available.
I
40"
L,
L,
lL"
T
::,I(' ,
2
1
40
l
It the truss has more members than indicated by equation (2.8) the trusS is considered redund~~t and statically lnGeterminate since the member loads cannot be found in all the members by the laws of statics. Such redundar.t structures it the members are properly ?laced are stable and will support loads of any arrangement.
ANALYTICAL IlE:rHODS fOR DE:l'SRtrINI'IG
(1
L~~ ~_:)oo
',1000
Fig. A2. 19
u,~
 ,
In general there are three rather distinct methods or procedures in applying the equations of static equilibrium to finding the primary stresses in truss type structures. They are otten referred to as the method of joints, moments, and shears.
A2. 9 Method of Joints.
It the truss as a whole is i~ equilibrium then each member or joint in the tr~ss must likewise be in equilibrium. The forces in the members at a truss jOint intersect in a common point, thus the forces on each joint form a concurrentcoplanar force system. The ~ethod of joints consists in cutting out or isolating a joint as a free bOdy and applying the laws ot equilibrium for a Concurrent force system. Since only two independent equations are available for this type of system only two unknowns can exist at any jOint. ThUS the procedure is to start at the jOint Where only two unknowns exist and continue ,rogressively throughout the truss joint by joint. To illustrate the method consider the cantilever truss of Fig. A2.18. From obse~lation there are only two members With internal stresses unknown at jOint L~. Fig. A2.19 shows a free body of jOint L~. The stresses in the members L~ La and L~ U~ have been assumed as tension, as indicated by the arrows pulling away from the jOint L~.
or tension. In equation (0) the load of 1250 in L~L:;l was substituted as a minus value since it was foune to act opposite to ~hat sho~m in Fig. A2.l9. Possibly a tetter p rccedur e would be to change the sense of the ar~ow i~ ~he :ree bodY diagram for any solved ~embers Jefore writing further equilibriUW equati~ns. ~e ~ust proceed to joint L:::I instead c r joint U2, as three ~~~own nembers still exist at jOint U2 Whereas only ~NO at jo~nt La. Fig. A2.20 shows f~ee body of jOi~t La cut ou~ by section 22 (see ~lg. AS.IS). The sense of the urJL~own member s~ress L:;lU:;l has Jeen assw~ec as COIDpressic~ (pUShing toward jOint) as ~t is obViously act~ng this way to Jalance :he 500 lb. load.
500
L.
~,
~H
250
u.,
tu.
=
Fig. A2. 20
For equilibrium of jOint La, ZH and ZV 0 ZV 500 + LaUa = 0, whence J LaUa = 500 lb. Since the Sign came out plus, the assumed sense in Fig. A2.20 was correct or compreSSion.
ZH = 250  LaLl :
The static equations or equilibrium for the forces acting on joint L~ are ~~ and ZV
!V
Next consider joint Ua as a :ree jody c~t out by section 33 in Fig. A2.18 and erawn as Fig. A2.2l. The known member s~resses are showr. with their true sense as ~reviouslY found. The two ~own member stresses U 2L l and UaU l have Jeen assumed as tension.
=o.
Fig. A2. 21
=
=a
      fe )
whence, L ~U:;l 1250 1'0. Since the Sign 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 out plus, sense is s~e as assumed in figure
A2.11
ZH
=0
sense to tr2t
The algebraic sign of all unknowns came out positive, thus the assumed direction as shown an Fig. A2.22 was correct. Check results by taking
~B
&~B
Note: The student should continue with succeeding joints. In this example involving a cantilever truss it was not necessary to ~lnd the reactions, as it was ~oSSlble to select joint L 3 as a joint involving only VNO unknowns. In trusses such as illustrated in Fig. A2.22 i t 1s necessary to first find reactions R:l. or R a wni cn then provides a joint at the reaction point involving only two
~cwn
= 0
forces.
To determine the stress in member Fl , Fa and F" we cut the section 11 thru the truss (Fig. A2.22. Fig. A2.23 shows a free bOdy diagram of the portion of the truss to the left of this section.
, 'f~ _
Fig.A2.22
1400
A2. 10 Method of Moments.
a
o , 500 1
Fig. A2. 23 500
1400
~:
1
FIg. A2. 24
~;
For a coplanarnonconcurrent force system there are three equations of statics available. These three equations may 08 taken as moment equations about three different ~oints. Fig. AZ.22 shows a typical truss. Let it be reqUired to find the loads in the members F l , F a J
F", F. J
r , and F o '
Fig.A2.22
ioo""
3 2 'F"
100M 1 500#
10"
HA"'(+\"''+''.''''''!''>I
50~f
The first step in the solution is to find the reactions at pOints A and B. Due to the roller type of support at B the only unknown element of the reaction force at B 1s ~gnitude. At paint A, magnitude and direction of the reaction are l~own giving a total of three ~~cwns With three equations of statics available. For convenience the wllic~own reaction at A has been re~laced by its unknown H ~~d V components.
Takir~ L~A
The truss as a whole was in equl1ibrium therefore any portion must be in eqUilibrium. In Fig. A2.23 the internal stresses in the members F l , F~ and F~ which existed in the truss as a whole ~ow are considered external forces in holding the portion of the truss to the left of sectien 11 in e~Jil1brium in combination with the other loads and reac~ions. Since the ~em 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 equilibrium of the portion ot the truss shown. Thus the ~ortion at the truss to lett ot section 11 could be considered as a solid block as shown 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 particular member. Since there are three unknowns a moment center must be selected such that the moment of each of the two unknown stresses will have zero moment about the selected moment center, thus leaving only one illL~Own ~orce or stress to enter into the equation tor moments. For exa~ple to cetermine load F 3 i~ Fig. A2.24 we take moments about the intersection of forces F l and Fa or paint O.
momer..ts about
~oint
A,
T
Thus
500 x
II
Z~O
"3
...
= 1400
x 30  18.97 F,
=0
= 500 x 30 + 100e x 60 + 1000 x 90 30 + 500 x 120  150 VB = a Herce Vg = 1600 lb. Take ZV = 0
:= '/.
il.ence
z.v
ZH
+ 1600
= a there
momen~
Take ZH = 0
= 500
To ~ind the a~ of the force F 3 tram :he center a involves a s~~ll a~o~~t of calClllat!on. thus in general i~ 13 Simpler to resolve the ~~nown force into H and V components at a ~oi~t In ~ts 11ne Ji action such ~hat one of these ~omDonents passes thru the ~oment center and the arm of the other cJmponent can ~sually be dete~1~ed Jy inspecticn. Thus in
A2.12
TRUSS STRUCTURES.
Fig. A2.25 the ferce F 3 is resolved into its component F 3V and F3H at point 0'. 7hen taking
Fig. A2. 2S
the other two ~Jcnowns F~ and ~s lles at i~flni 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 referred to as the ~ethod cf shea~s, or the summation of all the forces nO~ial to the tNO parallel ~{nown chord members ~us~ equal zero. Since the parallel chord ~embers have no component 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
500
...."::,,,7" F ,
500 3
/IF,
Fig. A2. 28
11400
F.
Viously obtained.
= 2100
= 2215
ZV lb. as pre
= 1400
whence F. =  141 lb. (tension or OPPOSite to that assumed in the ~igure. To find the stress in illember F? , we cut section 33 in Fig. A2.22 and draw ~ free bOdy diagram of the left portion in Fig. AZ.28. Since Flo. and Fs are horizontal, the member F? must carry ~he shear on the truss on this sectior. 33, hence the name ~ethod of shears.
ZV
The load Flo. can be found by taking moments about point Ill, the intersection at torces F.
and F. (See Fig. A2.23).
ZMm =
whence, Flo. = 2800 lb. (Tension as assumed) To tinct rcrce F. by using a moment equation, we take moments about point (r) the intersection ot rorces F J, and F 3 (See Fig. A2.26). To eliminate solving for the perpendicular distance trom paint (r) to line at action of F., we resolve Fa into its n and V components at point 0 on its line at action as shown in
Fig. A2.26.
= 1400 
500  1000 F,
=0
ZMr
=  1400 x 30
+ 500
x 60
+ 60 F.V
=0
whence, FaV
Therefore F. pression
= 200
=282
lb. com
In Fig. A2.22 to tind the stress in member F. we cut the section 22 giving the free bOdy for the left portion as shown in Fig. A2.Z7. The mebhod ot moments is not sufficient to solve tor member F. because the intersection ot
Note: The student should solve this example illustrating the methods of moments and shears using as a tree body the portion of the truss to the right of the cut sections instead of the lett portion as used in these illustrative examples. In order to solve for the stresses in the members of a truss most advantageously, one usually makes use of more tha.~ one cf the above three methods, as each has its advantages for certain cases or members. It 1s important to realize that each is a method of sections and in a great many cases, such as trusses with parallel chords, the stresses can ?Tactically be found mentally without writing down equations of equilibrium. The following statements in general are true for parallel chord trusses: (1) The vertical camponent of the stress ~n the panel diagonal members equals the vertical shear (algebraic sum of external forces to one s1de of the panel) on the panel, since the chord
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.
(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.
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 To illustrate the simplicity at determining in the index triangle or  133 lb. It is negastresses in tITe members of a parallel chord tive because the horizontal component of DF truss, consider the cantilever truss of Fig. pUlls on Joint F and therefore Fa must pUSh A2.29 with supporting reactions at points A and against the joint for equilibrium. J. Considering Joint D:150
+ 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)
ZV = DFv
DC
550
133
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 proportional to the length triangles.
= O. But OF 133, and GCR 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.
ZH=GH  GF  GCR
= 300
=
=
The shear load in each panel is first written 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.
shea~
For the second panel from the tree end, the is 100 + 150 ~ 250 and for the third panel 100 + 150 + 150 = 400 Ib , , and in like manner 550 ~or fourth panel.
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 diagonals 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
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 horizontal 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.
.,/;c
,/
A2.14
TRUSS STRUCTURES.
loading, we know that ore t~l~ of the exter~al loads at jOints U3 and L 3 is supported a~ reaction R:I, and 1/2 at reac t.t cn R a , or shear in
~
33"
5"
30" 50
Length Triangle
50
Fig. A2. 30
50 50
312 + 127.5 = ~SS.S. T~ere~2re, lsad i~ U.~2 =  499.2. Si:nllarl~' 3.':: jo inc La, :"'aL 3 = 312 .,. 127.5 = 499.5. At ~oint U~, ete ~o~iz:~tal co~~onents of U:l,U~ an~ U~U3 = ~99.S ~ 52.5 = 562 which must be balanced ,:,y  562 tr; member' U aU 3 '
499.5 U
562
.9;> 1'.9 .s
0
lu,
j
i!R,
312
100
;.=
';~~
'1
The reac:icn R~ equals t~e val~e on the vertical side of our i~cex triang:e in the 91.j panel, or 375. This should oe c~ecked usi~g the truss as a whole and ta~lng ~c~ents ~jout
R,
312
p..3
jL.
100
TI:e metal covered carrt t Lever' wing 'N1t;'1 i ts better overall aerodyna~lc e~:iciency and s~f ficien~ torsion~l ri 6 ! : l t y has ~ractlcal1y replaced t~e externally braced wing excep~ for lew 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 therefore a drag tr~ss inside the wi~g is find the loads in the diagor21s, then in the necessarJ to resist loads in the drag t7USs verticalS, a~d finally in the horizontal chord direction. Figs. A2.33 and 34 shows the genmember's 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 USing double wires in each d~g truss bay, a va.Iues on the hypotenuse 0 f the index triangles. Single diagonal strut capable ot t~<ing either The sense, whether tension or campression, is tension or comprass tve loads could be rsec , deter.nlned by inspection by cutting ~ental The external brace s t rut s are stream j rne tuocs . sections thru the truss and noting the direction ot the external shear load which ~st be balanced by the vertical camponent of the diagonals. 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 or U 3L3 =  50. but
U~13V
,
center panel = (100 + 50) 1/2 = 75. The vertical shear in panel U:I,UgL1L a equals 75 pl~s the external loads at Ua and La or a total of 225 and Similarly for the end panel shear = 225 + 50 + 100 = 375. With these values kTIQwn, the other two sides of the index triangles are directly proportio~41 to the sides of the length triangles tor each panel, and the results are as sho~m in Fig. AZ.30.
If a truss is lcacec unsJ~etr~cally, t~e reactions should be ceter.r.ined ~irst, atter which the index triar.~:es can be dra'HTI, sta:ting with the end ~anels, since the ;anel shear 1s then readily calculated.
A2.12 Aircraft Wing Structure. Truss Type with Fabric or Plastic Cover
ZV =  50  U3L 3 =
a
Fig.A2.31 Piper TriPacer
For jOint U~, ZV =  50  UaL 3 v  U~L~ = 0, 75, the vertical cornpo~ent of Ua L3 rrcm index t r tang Le. . ", U ~L~ 50  75 =  125. For JOint L~, ZV =  100 + L:I,U 1 = 0,
=
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
FLIGHT
VEHICLE STRUCTURES
A2.l5
11e
Fig. A2. 33
, ,
,;
~ (0. ,
,
I
t:
\.,r:=r ".ll
~'I
~\
in all ~enbers of the lift anc drag tr~sses will be dete~,ine~. A Si~plified air loac:ng Nill be ass~~ed, as the ?ur~ose of this problem is to ;iv8 the st~dent practice in solving sta:ically determinate space t~uss str~ctur8S.
ASSL~D
AIR LCADING:
Drag wtre
AntiDrag Wire
D"gwire~ "!
Fitting \
\,i.
/
:1
(1) A constant spanwise lift load of 45 Ib/in from hin~e to strut point and then tapering to 22.5 Ib/tn at the wing tip. (2) A [erHard uniform distributed drag load of 6 lb/in. The above airloads represent a high angle of attack condition. In this condition a forward load can be placed on t~e drag truss as illustrated in Fig. A2.36. Projecting the air
:: :~' \
:1
:
V
/ I:
,
,
11'1
:+r
I
I
I,
Ir
"
"
"
Airstream
~,...,
/i
,
L ift
Fig. A2. 36
Drag
Front Beam
Rear Beam
Fig. A2. 34
~a~~le
~~e~nally
Braced Mono
litt and drag forces on the drag truss direction, the fonvard projection ~ue to the lift 1s greater than the rearNard projection due to the air drag, which dl~ference ~n our exa~ple problem has been assumed as 6 Ib/in. In a low angle of attack the load in the drag truss direction would act rearward.
SOLUTION:
A2.35 shows the structural d~mensional j~agram of an externally braced nonoplane wing. The ~~ing loads on the front and rear The wing is fabric covered between wing beams, bea~s will Je calculated as the first step in ace thus a drag truss composed of struts and the solution. ?or our flight condition, the tie rods is necessary to prov~de strength and center of pressure or t~e air~orces ~ill be assumed as shown in Fig. A2.37. r~gid1ty in the drag direction. ~he axial loads
DRAG TRUSS
=:i"':
;;
'" 1 3
Cr F.Bl
I I
[SRI
1
~
I
R.B.
24.21" Fig. A2. 37
,
1.1 22.52"
T~e
3 Dimedral
N '
Airplane I
, It ""
?. i(5lfi. 6 ,..Fuselage
Fig. A2.35
!(6)
l
running load on the front bea~ will Je 45 x 24.2/36 30.25 Ib/:n., a~d the re~alnder or 45  30.25 = 14.74 Ib/in gives the load on the ::ear beam.
 ' ,1..
I
A2 16
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 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
~~~l~g
'I"A
1/
:;;/.::B.,r
F
3'_71
"'I
/11
Drag Truss
3
R. B.
beam.
Fl"Qot Lift
Front Seam
Truss
~ember
pof nt (2)
gym.:
FB
RB
t,
114.5R ..  114.5 x 30.26 x 114.5/2  15.13 x 70.5 x 149.75  15.13 x 35.25 x 138 = O.
I ,
5.99
~.99
o o
11
hence
RJ.
Take
Bear Stru.t
1s taken normal to
'.
+
'F
15 7 9 9
j 57.49
D drag direction, S
L
~
x 114.5
side direction,
D2 32
=0
~ Vv2+
=0)
hence R~ 1295 lb. (The student should always check results by taking moments about pOint (1) to see it ZM ..
w~th
We stare the solution of joints jy sta~i~g joint (1). Free bOdy sketches cf joint (I)
Rear Beam
(W'"
14. 7#/in.
(4)
1I
R.
Ilftltll;,,/2
114. 5"
ija
R,
70. 5,,1
The rear beam has the same span dimensions but the loading is 14.74 Ib/in. Hence beam reactions R~ and R3 will be 14.74/30.25 = .4875
are sketchel below. All members are considered twof orc e r.embers or havtng pf ns at each end, thus magnt .ude is the cn.Iy unknown characteristic 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 , its influence in causlr.g loads in drag truss ~embers can then be found when the drag :russ as a whole is treated. In :he joint solution, the drag truss has been assumed parallel to drag direction which is nct quire true frc~ Fig. A2.35, but the error on member loads is negligible. JOINT 1 (Equations of ~quilibri~~)
tilnes those for front beam. hence R s = .4875 x 3770 = 1838 lb.
R~
(1)
The next step in the solution is ~he solving tor the axial loads in all the members. We will use the ~ethod of joints and consider the structure made up of trIee truss systems as illustrated at the top of the next col~. namely, a front li!t truss, a rear lift truss and a drag truss. The beams are co~on to ~oth lift and drag trusses.
Table A2.1 gives the V, 0 and S ~rojecticns or the lift truss ~embers as dete~ined from information given in Fig. A2.35. ~he true
SF
~
B
770
D,
~v
4+D
.0523
.;966 F&
.0854 SF + 0, = 0
0,
== =
l~
. (tensior.)
A2 17
RL.J(3l
~
R
(3)
VSplane
:5
:V = 1638 x
ZD
1 83 8
(drag truss
reaction on VDplane
points (2) and (4). In the design of the beam and fittings at this point, the effect of ~he actual conditions of eccentricity should of course be considered. Combined Loads on Drag Truss Adding the two load systems of Figs. A2.38 and A2.39, the total drag truss loading is obtained as shown in Fig. A2.40. The resulting member axial stresses are then solved for by the method of index stresses (Art. AZ.9). The values are indicated on the truss diagram. It is customary to ~ake one of the fittings attaching wing to fuselage incapable of tranSferring drag reaction to fUselage, so that the entire drag reaction from wing panel on fuselage 1s definitely ccnrtnec to one paint. In this example paint (2) has been assumed as point where drag is resisted. Those drag wires which would be in compreSSion are assumed out of action.
RB SR
= 1838
D, +
= 0 (4) = 0 (5)
  (6)
Solving equations 4, 5 and 6, we obtain RB =  4189 lb. (compression) SR = 4579 lb. itension) 03 = 0 Fig. A2.38 shows the reactions of the lift struts on the drag truss at Joints (1) and (3) as found above.
,J><1Xt~X
Fig. A2.38 798
~~3Gf136~i
39.5
118.5
1191 114\
231
37.5 225
58.5
281.5
4189
I
162il 11,933
254
Drag Truss Panel Point Loads Due to Air Drag Load. It was assumed that the air load components in the drag direction were 6 lb./in. of wing acting r orward, The distributed load of 6 lb./in. is replaced by concentrated loads at the panel points as shown in Fig. A2.39. Each panel point takes one halt the distributed lead to the adjacent panel point, except for the two outboard panel pOints which are affected by the overhang tip portion. Thus the outboard panel paint concentration of 254 lb. is Jetermined by taking moments about (3) of the drag load outboard of (3) as follows:
P
To
13 893
Fig. A2.40
Fuselage Reactions As a check on the work as well as to obtain reference loads on ~lselage from Wing structure, the fuselage reactions will be checked against the exterr~llY applied air loads. Table AZ.2 gives the calculations in table form.
Table A,2.2 .Point
lIeliber
D
0 1908 0 0 0
s
_13870 0
= 70.5
x 6 x 35.25/58.5
= 254
IJ.
III
Dn,
Reactio.. Jt2(React1on)
Si~pllfY the drag truss SOlution, the dra~ strut and drag wires in the inboard drag truss panel have been modif~ed to ~ntersect at hinge
RB R.. (R.eaeUon)
~/in.
,
j
6
Totals
..
F,
1191
631
9333 4579
 67 1190  33
8290 4090
"8
0 1110
.00
component = (3770 + 1295 + 1838 + 631) .9986 = 7523 lb. (error 3 lb. D component = 185 x 6 1110 lb. (error e 0) S component = (3770 + 1295 + 1838 + 631) .0523 = 394 lb. (error 6 lb.
A2.18
TRUSS STRUCTURES.
al~
The wing bea~ ~ue to the distribu~ed l~~t loads acting upon them, are also s~bj~cted to bending loads in add! t t on to the axf.a ; Loads . The wing be~s thus act as bea~cJll~s. ~he sUbject of beaT.col~lh~ action is treated in another chapter of this book. skin instead of fabriC, the drag truss can be omitted 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 a box beam subjected to combined bending ~~d axial loading.
~lng ~etal Exam~le
Lengths k Directional Components of Cabane Struts Member Front Cabane Diagonal Caba,ne Strl.\t Rear Cabane Strut Sym.
V
30 30
D
10
I s
VIL
DIL
.240 .597 .1485
'IL
.648 .538 .668
CF CD CR
12 7
30 i27150.17
6
If the
1s covered with
23.5
27 I 40.42
L ;./ vO: +
D' + "
Problem 11.
l,.jing. Fig. A2.41 shcws a high wing exterr4lly braced wing structure. The wing outer panel has been made identical to the wing panel of exa~ple prOblem 1. This outer p~~el attached to the cen ter panel by Sir~le pin fittings at points (2) and (4). Placing pins at these points ~ake the structure sta~lcally determinate, whereas if the beams were made continuous trtlo~gh all 3 panels, the reactions of the lift and cabane struts on the wing beams would be statically indeterminate since we would have a 3span continuous bea~ resting on settling supports due to str~t deformation. The fitting ?in at ~oints (2) and (4) can be made eccentric with the neut~al axis of the beams, hence very little is gained by making be&~ continuous for the purpose of decreasing the lateral beam bending mo~ents. For assembly, stowage and shipping it is convenient to build such a wing in 3 portions. If a ~ultiple bolt fitting is used as ~oints (2) and (4) to obtain a continuous beam, not much is gained because the design requirements of the various governmental agencies speCify ~hat the wing beams must also be analyzed on the assumption that a multiple bolt tltting provides only 50 percent of the full continuity.
1',
(8)
The air loads on the outer ~anel are taken identical to those in exa~ple ;roblem 1. LikewiSe the dihedral and direction ~f the lift struts SF and SR have been made the sa~e as in exauple ~roblem 1. Therefore the analysis :or the loadS in the outer p~~el Gr~g and :i~t tr~s trusses is identical to that in )roblem 1. T~e solution will be cont1~ued assumi~g the r~~,i~g 11ft load on center panel of 4.5 10./:n. and a forNard drag load of 6 Ib./in.
Sol~tion
0:
Cente~ ?~nel
90'
'8
201
R a",1650if
R a",1650.f
Fig. A2. 42
Fig. A2.42 shows the late~al loacs on the center rear beam. The loads consist of the d1str:j~teG air load and the vertical component ot t~e ~or ces exerted by outer ?anel on center panel at pin point (4). From Table A2.2 of exarr.ple ;roblem 1, this resultant V reaction eq~ls 630 +
62
692 lb.
"'.= ,"
\4)
The vertical component of the cabane reaction at joint (8) equals one hal: the tetal be~ load due to symmetry of 1cad1~g or 55 x 14.74 + 692 = 1650 lb. Solution of force system at Joint 8
1650
J...
36"
C~
CR
(1)(3)
/L:.. S
D1
C
1650
~D
ZV  1650  .7)1 CR = a
whence CR = 1650/.731
(5)
Fig. A2. 41
= 2250
lb.
(tens~on)
(6)
A2.19
'whence
eRE
ZD = De  2260 X .1485
whence De = 336 lb. drag Center rront Beam
568#
(Ref. , Table A2. 2)
=0
reaction
(4) are taken fram Tabl~ A2.2 of problem 1. The drag loac of 336 10. at (8) is ~~e to the rear cabane strut, as 1s likewise the bea~ ax1al load of  1510 at (8). The axfa I beam load of  2281 lb. at (7) 1s cue to ~eact1on of ~rc~t caba~e truss. The ,a~el ,oint loads are d~e to the given running drag lead of 6 lb./in. acti~6 forward. The reaction which holds all these drag truss loads in equilibrium is supplied by the cabane truss at pOint (7) since the ~ront and ciagonal cabane struts intersect to form a rigid triangle. ~hus the drag reaction R equals one half the total drag loads or 2634 lb. Solving the truss for the loading of Fig. A2.44 we obtain the member axial loads of Fig. A2.45.
1157
tr~ss
fo 20
71
90 Fig. A2.43
R.,; 2535.
rig. A2.43 shows the V loads on the center front beaw and the resulting V component of the cabane reaction at joint (7). Solution of force
syste~
568
568
:;: ,
17308
(11) Fig. A2.45
at Joint 7
1730B
CFB
Co
~
CF
2535
1502
VS Plane
ZV
263~
ZD
=0
ZD
Zv
kF
"'.
vn Plane
CD
=
=
=0
Co , 1058
= 2635
Tr~ss ~embers
Co
=0
Solving for CF and CD, we obtain CF   2740 lb. (compression) Co = 3310 (tenSion) adding these loads to those previously lated for 11tt loads: CF
60
calc~
Fig. A2.44 shows all the loads applied to the center p~~el drag truss. The Sand D reactions fry,n the outer ~anel at joints (2) and
Panel point Drag Load
for 6#/in.")
1157
13m.~/L\J~7 t
I
190a'
33 6 270
336
CD ,
, ,
2740 + 2635
, 
105
4368 lb.
eR
1157
fusela~e
Reactions
As a c~eck on the ~ork the ~~selage reactions 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.
n, 26341
122IT
2.rr
(211
Fig. A2.. 44
'R, 2634#
1190a
I ! ,
A2.20
TRUSS STRUCTURES. 36
Poiut 9
c=,
_105 2260 4368 9333 4579
S
_68 1510 2356
,.
e e
Total.
C.
Rear Strut CR D1a. Strut CD Front
L1~t
"
1650 2610 4205 2055 10444
sl
72
24
Strut
8290 4090
I,
S.
Rear Lift Strut
'.
o
_1502
F
16178
component
+ 65
x 45
Fig. A2.46
component =

The total side load on a vertical plane trIU centerline of airplane should equal the S component of the applied loads. ~he applied side loads =  39~ lb. (see problem 1). The air load on center panel is ver~lcal and thus has zero S component. From Table A2.3 for fusela~e reactions have ZS = 16178. From Fig. A2.45 the load in the front beam at b of airplane equals  17308 and 568 for rear beam. The horizontal component at the diagonal drag strut at joints 11 equals 216 x 45/57.6 = 169 lb. Then total S components = 16178  17308 568 + 169 =  393 lb. which checks the side component ot the applied air loads. Example Problem 12. Single SDaT Truss Plus 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 triangular 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.
~
T
36"
(lVTC I I 
/ 11
S
\
\1
/
T
Fig.46a 1
I
l.s
'\ / "'
'c! \/
D D
G
II;';
*'1
": I ::!l A I~ c,
I
;g
)'"1/
~ 1/I~l2Is:vt;>,,1
N L J ~ 7 panels @ 12" '" 84" Fig.46b  _J
ABC ABC
E E
(Q~
,=
~
c>;
SOLUTION: 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 zesu'l tarrt load of 9 ~b./in. is imagined as acting on an ima~inary bea~ located along the lIne 11. 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 calculatior.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 dered. For a run....n tng load 0: 9 Ib ./"!..n., along Ltne 11, react t cns vt Ll. be for
,.g 'JLI!
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
Fig. A2. 48
si~ple beams re~ting at ,oints 2, 3, 4, 5, etc. The distance bet~een 23 is 8 inches. The total load on this distance is 8 x 9 = 72 lb. One half or 36 lb. goes to paint (2) and the other half to paint (3). The 36 lb. at (2) is then replaced by an equivalent force system at E and 8 or (36)/3 12 lb. to 8 and (36)(2/3) 24 to E. The distance between points (3) and (4) is 8 inches and the load is 8 x 9 = 72 lb. One half of this or 36 goes to pOint (3) and this added to the previous 36 gives 72 lb. at (3). The load of 72 1s then replaced by an equivalent force system at Sand D, or (72)/3 = 24 lb. to Sand (72)(2/3) ~48 to D. The final load at 8 1s therefore 24 + 12 = 36 lb. as shown in Fig. A 2.46. Due to symmet~J of the triangle CRD, one half of the total load on the distance CD goes to points (4) and (5) or (24 x 9)/2 = 108 lb .. The distribution to D is therefore (108) (2/3) 72 and (108)/3 36 to R. Adding 72 to the previous load of 48 at D gives a total load at D = 120 lb. as shown in Fig. A2.46. The 108 lb. at po tnt (5) also gives (108)/3 = 36 to R or a total of 72 lb. at R. The student should check tee distribution to other joints as shown in Fig. A2.46.
36 Oy
= 0, Oy
To find Ay take moments about x axis 'through point N. The moment of the air loads was previously calculated as  31752, hence, ZMx =  31752 + 9 Ay 3528 lb. To find Ny take l.Fy ZFy
= 0, =a
whence, Ay =
=3528
 Ny
=0,
hence Ny
3528 lb.
The reactions are all recorded on Fig. A2.46. Solution of Truss Member Loads For simplicity, the load system on the structure will be considered separately as two load systems. One system will include only those loads acting along the line AE and the second load system will be remaining loads which act along line as. Since no bending moment can be resisted at JOint 0, the external load along spar AE will be reacted at A and N entirely or in other words, the spare alone 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. 144 144 24 72 The moment of the total air load about an x t ~xis at left end of structure equals 756 x 42 :: I 11841 1184 0 2336 31752 in. lb. The ~oment of the joint load ': I "0 i~ <\"~ :.?o~~ system in Fig. A2.46 equals (66 x 12) + (72 x IN , 36) + (72 x 60) + (56 x 84) + l?4 (24 + 48) + BOO 32 BOO 224 224 (120 x 72) + (24 x 84) = 31752 in. lb. or a check. The moment of the total air load about Fig. A2.49 line AE equals 756 x 12 9072 in. lb. The TRI.ANGUUR TRUSS SYSTE:1 moment of the distributed joint loads equals (6 + 66 + 72 + 72 + 36)36 = 9072 or a check. The load system along t~e trailing edge as causes stresses in both the spar truss ~~d the Calculation of Reactions diagcr~l truss system. ~he support fitting at paint a provides a reaction in the Z direction The structure is supported by Single pin but no reacting ~oment about the x axis. Since fittings at pOi~ts A, Nand 0, with pin axes the loads on the trailing edge 11e on a y axis parallel to x axis. It will be assumed that throu~~ 0, it is obvious that all these loads the fitting at N takes off the spar load in flow to ?oint O. Since the bending strength of Z direction. Fig. A2.46 shows the reactions the tra~llng edge member ~s negligible, the Oy, 0z, Ay, Ny, Nz To find Oz take moments about y axis along spar AEFN.
~i~"23,,,3,,6"""==i'""7<?"';:=':n,,",C"'""",,"i
A2 22
TRUSS STRUCTURES
load of 36 lb. at Jol~t S in order to be transfer~ed to pOint 0 through the dlago~al truss system must follow the path SDRC~PAO. In like ~nner the load of 72 at R to reach a must take the path RC~~PAO, etc. Calculation of Loads in Dlaeonal
Member
, y x
Tr~ss
reaction on the s;ar force on the spar <~ duces compression in truss and tension ~n Consider Joint R
but does produce ~ couple the v dir~ctlcn ~r.lch ~ro the ~0P chord of t~e s~ar the bottom c~ord.
Members:x/L
.943
z/L
.118
y/L
The load to be transferred to truss ~CJR is equal to the 72 lb. at R ~lus the 36 1J. at S wn.tcb CCII'.es to jot nt R from trcs s DRG. Hence load
~n
4.5
12 36
38.2
.314
Re =
C~
(~2 +
4.5
36
37.5
.120
.960
457 lb.
Tdhence RJ Consider Joint S truss SEF carL~ot assist in any portion of the 36 lb. lead at S because the reaction of this truss at SF would put torsion on the spar and the spar has no appreciable torsional resistance. The
trla~Jlar transterrlr~
= 457,
= 457
= 457
= 72
~
Joint Q
~BL
= (180 x G.5)(1/.118)
BP
QL
=762,
+ 66
= 762,
LP
= 762
=AO
= .943
+
DS .943 GS
=G
X~~:G !Z
Joint P Load
= 180
=246
x O.5)(l!.l18)
~ree
.....hence, DS :::  GS
Load in PA
= (246
lG40
ZFz = 305
.118 OS  .U8 GS = 0
Whence PN
= 1040
:ody.
+
Subt. DS =  GS and solving for OS, gives as = 159 lb. (tension), DS = 159 (compression) Consider Joint D
ZF x =  1040 x .943
lG22 lb.
.960 AO
= 0,
~J
=
~O
=  2G22 lb.
gi'l8S
L.
Let Ty and Tz be reactions ot diagonal truss system on spar truss at Joint D. ZFx
= 159
As pointed out ?reviously, t~e diagonal 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 the load in the diagonal truss ~embers ~eeting at a spar jOint. Let Ty equal this reaction load on the spar.
At Joint c:Ty
159 x .118  Tz = a
==
whence Tz 0, which ~eans the diagonal truss produces no Z reaction or shear load on spar truss at D.
= 287
=479
ZFy
= .314 x
whence Ty
=0
762) .314
=
100 lb.
, T Likewise at Joint c, y
=479
=326
If joint G is investigated in the same manner, the results will show that Tz a and Ty 100.
At Joint A:m
'y
=
(1040 x .314)
diagor~l
The results at joint D shows that the rear truss system produces no shear lead
A2.23
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 airborne 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 socalled 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 airplane 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.
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 aircraft. In general, the gear tor aerodynamic efficiency must be retracted into the interior of the wingJ nacelle or fuselage J thus a reliable 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 landing 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
TRUSS STRUCTURES.
Douglas DC8
Jet Airliner
A2.25
PiperApache
Piper TriPacer
Navy F4J
A2.26
TRUSS STRUCTURES.
end tor attaching the wheel and t~re. ~his cantilever beam is subjected to bending in two directions, torsion and also axial loads. Since the gear is usually made retractable, it is difficult to design a single fitting unit at the upper end of the oleo strut that will resist this combination of forces and still per.nlt movement tor a simple retracting mechanism. Furthermore, it would be difficult to provide carrythrough supporting wing or fuselage structure tor such large concentrated load systems. Thus to decrease the magnitude of the bending moments and also the bending fleXibility ot the cantilever strut and also to simplify the retracting problem and the carrythrough structural problem, it is customary to add one or two braces to the oleo strut. In general, effort is made to make the landing gear structure statically determinate by USing speCially designed fittings at member ends or at support paints in order to establish the torce characteristics of direction and paint at application. TWo example problem solutions will be presented, one dealing with a gear with a Single wheel and the other with a gear involving two wheels. Example Problem 13 Fig. A2.55 shows the projections of the landing gear configuration on the VS and VD 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 'through 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 reaction at G is unknown in magnitude only Since the pin fitting at each end of member GC fixes the direction and line of action of the reaction at G. For convenience in calculations, the reaction G Is replaced by its camponents GV and GD' For a stde load on the landing gear, the reaction In the S direction is taken off at paint F by a special designed unit.
SOUJ'I'I:JN
The supporting ~eactlons upon ~he ~ea~ at paints F, H, and G will oe calculated as a beg~nn1ng step. T~ere are six urJcnowns, na~ely FS, F'V, FO, Hy, HO and G (See Fig. AZ.56). Wlt~ 6 equations of static equilibrium available tor a space ~orce sy~tem, the reactions can be found by statics. Referring to Fig. AZ.56:To find FS take ZS Fs
+
=0
= 0,
hence FS
=a
To find react10n Gv take moments about an S axis through paints F, H. 50  24 Gv 0 Whence, GV = 6500 lb. with sense as assumed. (The wheel load of 15000 lb. has been resolved into V and D components as indicated in Fig. A2.55) . With GV known, the reaction G equals (6500) (31.8/24) 8610 lb. and similarly ~he component GD (6500)(21/24) 5690 lb.
l:l1s
=3119 x
1 , 1.6
=
"
;0
.
II
cJi:~~
J6"
<;
10063
1~3004 !5690 G
lll09
6500
\'D
Fig. A2. 56
I
a
433
Flg. A2. 57
A2.1.7
lEy
rE I
, , I" u
I
I
I
21
24"
1 <>
~ _ CCGO=31 a e G
"I"'I
31
the structure as a whole by taking moments about D and V axes through point A. =  10063 x 14 + 6500 x 8 + 11109 x 8 = 140882 + 52000 + 88882 O(check) = 5690 x 8  433 x 8  3004 x 14 = 45520  3464  42056 = 0 (cheek)
26"
L.
0
lel3u9*
TE=24952
Fv
By
ro"r
TE
:;:@BIS=3920
FSS 1@)_3920 F.
to
39,20. HO
~
H
The next step in the solution will be the calculation of the forces on the oleo strut unit. Fig. A2.58 shows a free body at the oleostrutaxle 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 reaction 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 manner as to resist torsional moments about the oleo strut axis and to provide 0, V and S force reactions but no moment reactions about 0 and S axes. The unknowns are therefore BI, CG, Es, 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 rotation of the moment is given by the right hand rule, namely, with the thumb of the right hand painting in the same direction as the arrows, the curled fingers give the sense of rotation.
T~ fInd the resisting torsional moment take moments about V axis through E. T~
Fig.A2.59
.....
ZME(V) =  3119 x 8
= 16
= 16 x 6500
= 24952
~
TE = 0, hence TE
in. lb.
sumed.
tihrough F.
Tdhence, CG = 8610 lb. This checks the value previously obtained when the reaction at G was found to be 8610. The 0 and V components or co thus equal, CGD = 8610 (21/31.8) ~ 5690 lb. CGV = 8610 (24/31.8) 6500 lb.
= 6 On =0
~bence, ~o
HD = 433 lb.
!D
=
=
?O
HO
GO  3119
=0
about
Whence, FD
To
=3004 lb.
~nence,
Hv
=11109
To find load in brace strut 8I, take moments 0 axis through po irrt z. ~E(D) =  14672 x 8 + 3 (BI) 22/24.6 + 24 (BI) 11/24.6 0 T~ence, BI = 8775 lb. and thUS, SlV = (8775)(22/24.6) = 7840 lb. BIS (6775)(11/24.6) = 3920 lb.
lb.
To tind Es take ZS = 0
'i.S
Fig. A2.57 summarizes the reactions as found. The results wl~1 be checked for equl11brlum of
Es 
3920
=0,
hence Es
=3920
~"'~,~.~
.~'"
A2.28
TRUSS STRUCTURES.
To find Eo take ZD
ZD
=0
= 5690  3119  ED
= 0,
+
hence ED
= 2571
To find EV take LV = 0
6500 : 0, hence
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 recorded on the figure. The equations of equilibrium for this tree bOdy are:1:3
Lower Cylinder
Fig.A2.60
=0 = 22
3920 + 3920 + FS
0, or FS
=0
Torque LinJ<
(l
)~;~O::::iAxle
~
l:l'lF(D)
HV  3920 x 2
7840 x 20 
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.
=0
=
0
~rob
zv   Fv ...
whence, FV = 10063
ZD   Fn + 2571 ... 433
The reaction R~ between the bvo units 0: the torque link at point (2), see Fig. A2.50, thus equals 24952/9 2773 lb.
whence, FD
=3004
lb.
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 transmit 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 doublecantilevernut cracker type ot structure. Fig. AZ.60 illustrates ~ow such a torque length could be applied to the oleo strut in our prOblem.
The reactions R 3 at the base of the link at 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 structure 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
A2.29
ZMA(D)
= 60000 x
28"
I
""' B~ace
0
4"
Fig. A2. 61
" "
r,
"
"
0
+
zv
~oDo I tV
36"
= 78070
=a
60000
40000
66666  VA
L.
I
=57142
40000
=
= = ZD = 0
To check the results take moments about Y and D axes through pOint o.
ZMo(V)
&~O(D)
=5 x 10000 + 14756 x
19 = a (check) 19
19  17386 x
+ 78070
= 20000 x 10  88596 x 19
=a
(check)
Fig. A2.63 shows a tree body of the oleostrut OE. The loads applied to the wheels at
ITE E EO
r
00
28"
17"'1 i7"'
1
9 "
Fig.A2.62
shows a space diagram with loads and reactions. The reactions at A, Band C have been replaced by their V and 0 components. To find reaction Cy take moments about an S axis through paints AB.
571~~
0
~ ~
.:l;G_  
4~~
~t
36"
'\!
I
ITE E _ESJ:'
Yi
r
L
28"
005'
ZXAB
=
VL
41/41
~OF5
0
"0"
(15000
10000) 64
l~. ~ith
+ 24 Cv = 0 sense as
c The reaction at C must rave a line of action along the line ~~, 28G CD since ~ember CD 1s pinned at 24 0 each end, thus the crag component and the load in the strut 2D tollow as a matter of geometry. Hence, Co ~ 66666 (24/25) 57142 lb. CD 66666 (36.93/28) 87900 lb. tension
t 50000"lb.
l
Fig. A2. 63
M'o"
t
50,OOO"lb.
the axle centerllnes have been transferred to point (0). Thus the total V load at (0) equals 60000 + 40000 = 100000 and the total D load equals 15000 + 10000 = 25000. The moment of these forces about V and D axes through (0) are Mo(V) = (15000  10000) 10 50000 in.lb. and Mo(D) = (60000  40000) 10 200000 in.lb. These ~oments are indicated in Fig. A2.63 by the vectors with double arrows. The sense of the ~cment 1s determined by the right hand thumb
= =
A2.30
TRUSS STRUCTURES.
AD BD
and finger
The fitting at point E is Ges~gned to resist a moment about V axis or a torsional moment on the oleo strut. It also can ~roVlde shear reactions Es and ED but no bend1ng resistance about S or D axes. The unknowns are the forces DG and the moment TE'
19 +19 ~ &50000 I
""A
IV I
32143
EB
AV~~17
A
_,.
I
t li~BV
_ B
I!
89215 \
'
77451
Fig. A2. 64
=  50000
TE
= 0,
whence
T~
=
ZMA(D)
500CO In .u ,
~oments
=200000 sa
=a
= 89215
x 2  77451 x 36 
38 BV
= 7143
Es = 0,
lb.
To r md AV take ZV
ZV
=0
=0
t~rougb
=200000
=
 100000 17 + 34 DFy :: 0
x 17  66666 x
To tind BD take
ZMV(D)
Y axis
A.
whe~ce,
DFV
=77451
=77451 lb.
28~ ~:,\l>
17
=0
To find AD take ZD = 0
ZD
=77451 = 90503
= 17388
=O.
or AD
=0
=0,
lb.
These tour reactions check the reactions obtained originally ~hen gear was treated as a tree body, thus glvln~ a numerical check on the calculations. the forces on each ?art of the gear the parts could be designed for strength and rigidity. The oleo strut would need a torsion link as disc~ssed in example problem 13 and Fig. A2.60.
~lth ~~own,
Then, lXle DG
= 89215
=89215
(17/28)
=54164 lb.
= 104190 = 0,
(32.73/2B)
= 25000
ED
=32143
x 35 + 28 ED lb.
A2.15 Problems
The results will be checked tor static equilibrium of str~t. Take moments about D axis through point (0).
l:!1o(D)
(I) For the structures numbered 1 to 10 deter~ne whether structure 15 statically determinate with respect to external reactions and internal stresses.
Pin 'A (1)
= 200000
l:!1o(S)
=32143
x 64  57142 x 36
=O(check)
I
~ ],8
:;,0 10
Pin
~D AB E in"
1 10
10
Pin
(2)
fIB
A
10
20
(3)
Fig. AZ.S4 shows a tree body at member AB with the known applied forces as found from the previous reactions on the oleo strut. The unknowns are AD, BD, A~ and By. To tind By take moments about D axIs tr~ough A.
10tr: A (4)
~10
(5)
B
A
A
'*'
*
I
;
I
B
(6)
10
10
A2.31
1000*
(2) Find the horizontal and vertical components of the reactionS on the structures illustrated in Figs. 11 to 15.
..J+V
tt4'~0 T Pin
i
52
Slde vtew
~!~
A
(I')
~104 10
10"
+D
Wing
+V
J'
,
Front vtew
+8
21:'
116'''1
(11) Pbi]:
Bi"'!>c
ho"
oj,
~O'
I
l~O
(13)
Flg.20
200f
Cable
(14)
500*
Pulley #~';=f:d~I"
100
k:cr,,,i<' ,..,..100/1
(6) Fig. 21 shows the wing structure of an externally braced monoplane. Determine the axial loads in all members or the lift and drag trusses for the following loads.
'O~'
A
'YO"
Front beam lift load 30 lb./in. (upward) Rear beam lift load 24 lb./in. (upward) Wing drag load 8 lb./in. acting art
= =
(3)
Find the axial loads in the members of the trJssed structures shown in Figs. 16 to 18 .
...J. jl0
15"~' T 30"
10
....J..
( 24
t10
(16)
12 4
124"1
1'000
36"
+
36"
+
36" ~
45"1
B'
~'~:::
...L
CF
30 dihedr
_800
10
130" +
(4)
I 2000 I 15"+15'...l
(IB)
D,D' l I 1 30'....
I . I I
.i
CD CR
:~tL
Fuselage
A.A'
Flg.21
Determine the axial loads in the :nembers of the structure in Fig. 19. The membera are pinned to supports at A, Band C.
(7 )
Irl
5 Bays @ 30"
3I
Pin Reactions
~DragTrUSSLS
t f t
.j
At t t t l t t t t
t tt t
v t
1100/1
D
60" Fuse
at A, B, C
Flg. 19
'D
.L
lage
~"'. L
B
:
I
25 lb/in.
' ,
110'\
(5 )
Fig. 20 shows a trlpod trame for hoisting a propeller for asseT.bly on engine. Find the loads in the ~rame tor a load at 1000 lb. on hoist.
Fig. 22 shows a braced ~onoplane wing. For the given air loading, tind axial loads in 11tt and drag trusS members. The drag reaction on drag truss is taken oft at point A.
A2.32
TRUSS STRUCTURES.
(8 )
(9 )
6"
....,._,A B C
20"
,
I
24"
13000 lb. Fig. 23 10000 lb.
10000
The r i tt tng at points A and B ~or the Landfng gear structure in ?ig. 23 prOVides resistance to VJ D and S :8actions and moments about D and V axes. Find the reactions at A and a and the load in member CD for given wheel loading.
L22"';
t
B
C F
the brace
Wheeo
"113'
wheel
'0"
res.c1:1.ons but only moment resistance about 1J axis. Find reactions at ~ and loads in :::embers SF and BC under given wheel loading.
15000 10000
Fig. 24
wheel
LD
member's The
. .J500'
333 wheel 2
CD~ A
crevices ::e
sistance tc
V, 0 ar:.d S
CHAPTER A3
two terms, center of gravity and moment ot inerIx =J;tJdA, I y =/XQdA, I z =/radA tia, are constantly being used. Thus, a brief 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
Then
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 distributed in its true or natural way. This general relationship can be expressed by the principle of moments, as tollows:
Lines: 
XL
ZLx, hence
Polar Moment of InerUa In Fig. A3.1, the moment of inertia I z =fr~dA about the Z axis is referred to as the polar moment of inertia and vcrueeee iv = ZVx, hence i = ZVX '=JXdV can be defined as the moment of inertia of an T Yarea 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 or inertia Is equal to the sum of the moments of If a geometrical figure 15 symmetrical with reinertia with respect to any two axes in the plane spect to a line or plane, the centroid of the tigure lIes in the given line or Plane. This is of the area, at right angles to each other and passing thru the point at intersection of the poobvious tram the fact that the moments of the parts of the figure on the opposite sides of the lar axis with the plane. line or plane are numerically equal but of opThe radius at gyration posite sign. It a figure is symmetrical to ~NO A3_5 Radius of Gyration of a solid is the distance tram the inertia axis lines or planes, the centroid of the figure lies to that point in the SOlid at whiCh, it its enat the intersection or the two lines or the two planes, and likewise, if the figure has 3 planes tire mass could be concentrated, its moment of inertia would remain the same. of symmetry, the centroid lies at the intersection of the 3 planes. ThUS,fr a dl1. =faM, where ~ is the radius at gyration A3.3 Moment of Inertia The ten moment at inSince, r"ciM = I, then I =f''' I1 or f =\ IT ertia is applied in mechanics to a number at Vl1 mathematical expressions which represents secBy analogy, in the case ot an area, ond aonents at areas, volumes and masses, such as I =f'A or =Ift fY'dA, r'dY, fr'd1'! etc.
jceas r XA
Sax, hence X
= Zax =fXdA
A A
A3.4 M01Ilent of Inertia of an Area As applied A3,.7 P&rallel Axis Theorem In Fig. A3.2 let I y be the moment of inertia of the area referred to to an area, the term ~ament of inertia has no phYSiCal sig~ificance, but represents a quantity the centroidal axis yy, 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 inwith distance x + d from y~y~. r luence of tue area i t.se Lf in determining the Than, Iy~ = (d + x ) IlidA total rotating moment of uniformly varying forces applied over an area. =JX'dA + 2dJXdA + 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
I
.c:":
>dA
Fig. A3.2
g.
TABLE 1
1 x ....,
I
fe.
I
Area d Y
= bd = bd 3 / 12 = Oct'
3"
I1._1. I _
a
2dA.
The first term,! x represents the acment of inertia of the body about its centroldal
axis yy and will be given the symbol I. The second term 1s zero because~xdA 1s zero since TJ 1s the centroldal axis of the body. The
= Ad
Thus in general,
I=I+Ad
a
12 This expression states that the amount ot 1 _ inertia of an area with respect to any axis in 2 bh' 3 3 = the plane of the area 1s equal to the moment ot "4 inertia ot the area with respect to a parallel = .236 h centroldal axis plus the product of the area and the square ot the distance bec#sen the two axes.r~_1
3 ,~
~
1
Y = h/3
11._1.
1,;1_a
3 = bh "V"
Triangle
vO
f i
r   I 'l 2
= bh 3
f ,,
Area
Parallel Axis Theorem For Masses. Ifinstead of area the mass at the body is considered, the parallel axis can be written: I = I + Md a , where M reters to the mass at
the body.
A3.7a hss Jloaents at Inertia The product of the mass or a particle and the square at its distance tram a line or plane is referred to as the moment of inertia at the mass at the particle with respect to the line or plane. Hence,
= bh
'2
x = 2/3
11.1.
h
h~
.3
4,8
Trapezoid
r blj
Area
= d(b
+ b 1.)
I = ZMra.
ed by a definite integral, the expression may be wrL tten I = J re dl1 Moments of Inertia of Airplanes. In both ilying and landing conditions the airpl~~e may be subjected to a~~lar accelerations. To determine the magnitude ot the accelerations as well as the distribution and magnitude or the ~ss inertia 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 expressed as tollows:
= 0.3
(b 2 + 4bb , +b 2) 36 (b + b, ~
Ix Iy Iz
Zwz 2 + Zl.I x
Zwz ll + Zl.I y r,wy2 + Zl.I z
A3 3
Elliptical
TABLE 1  Continued
Ring
~~l:
I
'~f
Ix
:=
y = d/2
Ix
:=
xl i~X
4:
n (a 1 3b 1
aa3b:a)
bd 3  a(d _ 2t)3
12
deb +a)3 _
20
W
6ab:ae
ex = V A}ea
Area ::: .215 a a
zaec Ix  I y
= (dtt') (b'bt)
and I:a II see equagiven for angle.
Tl1
~al1
x
X ::: .223 a
l::
Circular Fillet
usectton
r:n.n lIT:[b1
jtJI
a~r.o
Area
y :::
r~
I'
rb~=r
1 11
~
Xl
. x
==rr
I ,>,
..L
'I_it
Parabolic Fillet
x ::: 3 4:
Xl
y = .3 Y.
3
Area = nr'
Circle
2;~2
= nr'
T
4:
y ::: 2 y.
l\~l
Ring
....
;
"
la_Ii = 5 n r
'i
TABLE 2
e1l
= r
2
(roll  rill)
Solid Cire. Cyl.
Properties of SolidS
Vol.
~ .. '/~
1 ~ ~ :...l 1
"
',I
Area ::: n
s
=
n r'lL,. (r  radius)
ro '
l
11. 1
.'
:::
n (ro"  rl")
4:
.~
~
Jo.
)5:5' '" ,
'\
.'
M = W (Total wt.)
11
1
~"
\~
.
~.r.a
"2
+

Area _ n r "
I,~
= M [r a
(L f31])'4
ra~
Sermcrrcfe
y:::
tx
v
.424r
.1098 r< = .264 rrt
Vol.
11. 1 1a _ liI
=:::
:::
IT
L{r1 a
a
M(r1
~~~lx
t:x
,
q=outside
I1(r1 a
+ r aa)/2 + r a ll + La/3)/4
(R
a _
r ")
radius
 Mr a
2
y ::: 4 (R '1+ Rr + ra) approx. y:::2r 3n R+r IT
11 
:::
M (a:a
bll)/12
Ix:::.1098(R"r").283Rll!"" (Rr) R+ r
i s sm8l1
Lr Sp h ere S o lid
Sphere
Vol. = (4nr'I/3
Iabout axis =
2M r "1/5
Ellipse
I
~~ II' j
n a sb Ix = 4xf'x
3
3 )
rt
~~
=a
'2
"""""3""
13.4
MOMENTS OF INERTIA
TABLE 2 
Co~t~nued
J;tf'x
y
NAx
 re t
~2
sIn 20
4
 (lcos a
0)
M= w
g
2
TABLE 3
Lc1JT
b I
J.
Distal1Ce x
0.40992 0.40984. 0.40976 0.4988 0.40960 0.40952 0.40944 0.40936 0.4928 0.4920 0.40912 0.40905 0.40898 0.4890 0.4883 0.4877 0.40870 0.4862 0.4835 0.4822 0.40809 0.4795 0.4782 0.4770 0.40758 0.4746 0.40733 o 4721 0.4693 0.4667 0.4641 0.4616 0.40.592 0.4568 0.4545 0.4523 0.4502 0.4482 0.4462 0.4443 0.4409 0.40375 0.4343 0.4312 0.4284 o 4259 0.40233 0.4209 0.4188 0.4168 0.4128 0.4090 0.4060 0.4030 0.4000 0.3975 0.3950 0.3928 0.3848 ~. ~810 o 778 0.3750 0.3725 0.3702 0.3668 0.3636
Distance y
0.5008 0.5016 0.50240 0.5032 0.5040 0.5048 0.5056 0.5064 0.5072 0.5080 0.5088 0.5095 0.5102 0.5ll0 0.5ll7 0.5123 0.5130 0.5138 0.5145 0.5151 0.5165 0.5178 O .5191 0.5205 0.5218 0.5230 0.5242 0.5254 0.5267 0.5279 0.5301 0.5333 0.5359 0.5384 0.5408 0.5432 0.5455 0.5477 0.5498 0.51H8 0.5538 0.55.57 0.5591 0.5625 0.5657 0.5688 0.5716 0.5741 0.5767 0.5791 0.5812 0.5832 0.5872 0.5910
~
Area = 2 11 rt
11._1. = 11 rat
1.01 1.02 1. 03
1.04
pol ar = 2
11
r"t
A = n rt
y
I.::L__L.LI
Quartercircular Arc
2~2
= .6366
=
11
1._.
y
2
rat
1.05 1 06 1.07 1.08 1.09 1.10 loll 1. 12 1.13 1.140 1.15 1.16 1.17 1.18 1. 22 1. 24 1. 28 1 28 1.30 1.32
t:~~
o:;~~
= .2978 r >t
2
Area = n rt
~ r4.'1
t l
2 I
1.3<
= .6366 r
1;2
1.36 1.38
= 11
rat 4
1.<0
1.405 1, 50 1 .5.5 1. 80 1.6.5 1. 70 1.75 1.80 1." 1.90 1.95 2 00 2.10 2.20 2.30 2.40
;.~~
2.70 2.80 2.90 3.00 3.20 J 40 3..60 J.80 4.00 4.20 4.40 4.60
4.'~
Area
= art
sin a a '
in
Radians
=r
(a + sin 2(1)
'.00 5.50
~:~:;
6: ~~~~ 0.6152
0.6190 0.6222 0.6250 0.8275 0.6298 0.6332 0.63640
~.~
7.00 7.50 8.00 9.00 10.00
=r
(1008 c 0)
,l1xx=.W
= rlilt
(jcos a)
 r"t (0 I xx 2
sin 2(1)
2
A3.5
I"
I
'eight IArJaSX
Bo. Bailie fI/I
HorizOIItal lIo.ent
S
.x
1 2 3 4 5 6 7
8
9
10
Propeller Engine Group Fuselage Group .ing Group Hori. Tail Vert. Tail Tail Wheel Pront Land. Gear Pilot Iladio
o
37120 145600 94800 17760 13400 16400 34500
o o
~18
0
3200 10800 480 1040 _1000 _ 900
"
115
16.:1 240
8 26 _20 _30
33000
24000 411180
10 5
2000 500
.5480
where Ix, I y, and I z are generally referred to as the rolling, pitching and yawing moments of inertia of the airplane.
I"
Totals
3150
Example Problem 2, Determine the moment ot inertia about the horizontal centroidal axis for the area shawn 1n Fig. A3.4
= 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 equation 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 distances in inches, the moment or inertia is expressed in units of poundinches squared, which can be converted into slug feet squared by multiplying 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
w
Solution. we rrrst :C1nd t he moment of inertia about a horizontal reference axis. In this solution, 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 crosssection has been divided into the tive parts, namely, A, 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 axis = = :Av = 17.97 = 2.9l w ZA 6.182 BY parallel axis theorem, we transter the moment at inertia from axis x'x' to centroidal +. axis xx, Solution. The airplane center at gravity will be Ixx=Ix'x' Ajlt=79.476.18x2.9P.= located with respect to ~NO rectangular axes. In 27.2 in" this example, a vertical axis t~~ the centerline at the propeller will be selected as a refRadius of Gyration, Pxx = xx =lj27 .2 = 2.1" erence axis tor horizontal distances, and the A 6.18 thrust line as a reference axis for vertical distances. The general expressions to be solved
VI
are:distance to airplane cvg , from ref. axis Si5 y = Zwy distance to airplane c vg , from ZW ref. axis XX Table 4 gives the necessary calculations, whenee x = 417180 = 133.3" aft ott propeller
Zw
3150
= Zwx
Centroidal
Fig. A3.4
Fig. A3.5
5480
3150
Example Problem #3. Determine the moment at inertia of the stringer cross section shown in Fig. A3.5 about the horizontal centroidal axis. Solution. A horizontal reference axis x~xP is assumed as shown. The moment of inertia is first calculated about this axis and then transferred to the centroidal axis xx. See Table 6.
11)\
.......,F="",   7
Fig. A3.3
Z
I
, Ref. line
.3
INERTIA
" "
e
A 0.'"
Al 0.50 1.00
'" a. ,
'.25
2.875
0.375
",
_.ent of inertia of
".
13
uia.
2.75 15,1:3
..
..
, , ! ,
~
0
",
y= .375 +
4 X (.0841 3n
I
I
.~72
= .547"
:OJ.
appr ox ,
y = .375 + 2
= .375 + .172
=547"
.0< .0<
15.17 15.17
27.52
n
lex= [.1098(.0841+ .0625) + .04x .54.283x .0841 x .0625 x .04] = .002375
,
'0
.ca
2.12
.54
approx ,
0.281 1.00
D1 0.281 1.00
0.28 0.56
..
In~
.. 0.01
i5 \1116 . 182
1.50
17,97
I
Table 6
A,
0 0
179.47111.4
Problem #4. Determine the ::J,oment of inertia of the flywheel In Fig. A3.5a about axis of rotatlc n. The material 15 aluminum alley :::ast1ng (weight = 1 lb per cu. Inch. )
PartiOD I
Area
A,'
AI
lex
Ix 'x' .. lex... Ay2
.00281 .03428 .02078
O~787
Sec. AA
V/// / ~,
+ +
l'
.0.
.06~4
o .00281
3'
s~
.1933
.009
1= = lx~x'
Aj3
= .05787
in~
u~ cb .. \ w~
I
5"
r/
2"
I
10"
I
,
I
I
.1933 X
05745 = .55" .1933
IA
Fig. A3.5a
'
( .0465)' = .05745
Solution: ~he spokes ~y be treated as slender round rods and the rrn and hub as hollow cyl Ind. ers , (Refer to Table 2) Rim, weight n(R ," = = n(5"'4.l1)
3_r
a)x3x
.1
)
bd'
12
= (04
~2' 75')
4 3n
x 2
= .00281 in
.3=8.48 lb.
Portion 2
Hub,
'I"
Spokes,
r=
(.5
....
5x .54 .... 54 2
+
( .0
.54)
g
n
Icx
= .10ge x
(R a + r") t (R
r)  .283 Ra + r.llt
R + r
x .25 2 X .1 = .588# I or one spoke = ',.rL:a/12+>"j'r~ .588 x 3'1/12+ .588x2.S a = 4.10 lb. 1n 3 I for 4 spOkes = 4x 4.10 = 16.40
Approx.
lex
= .3trl,"
= .3 x .04 x.52
.
= .001688 in".
Moment or Inertia of an Airplane. ':'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
~ample
A3.7
necessary, which is available in the weight and balance esti=ate of the airplane. Table 6a shows the complete calc~lation of moments of inertia of an airplane. This table is reproduced. 'from :i.A.O.A. 'I'ecnnt ca L no te #575, ~sstimation of ~oments at inertia of airplanes from Design Data.~
t~e
axes, the prOduct of i~ertia of the weight about the reference axes 1s necessary. Column (14) gives the values about the reference axes. To transfer the product of inertia to the c.g. axes of the airplane, we make use of the parallel axis theorem. Thus
Z'NXZ C g =Z'NXZ(Rer. axes) Zwxz =48,857,5895325.3xl15.9x77.8=B39,253 lb. in.a Explanation of Table 1 To reduce all values to slug ft.a multiply Fig. A3.5b shows the reference of planes 32:l7 and axes which were selected. After the moments x.J:.... of inertia have been determined relative to 144 these axes the values about parallel axes 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 coColumn (1) of Table 6a gives breakdown of airordinate 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 angle 0 between the X and Z axes and the the c.g. of the items from the references planes principal axes is given by, or axes. tan 2 0 = 2I xz = 2 x 181 = .05998 hence >:1 = Columns (6) and (7) give the first moments of Iz Ix 90963061 the item weights about the Y' and X' reference 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= _
Zw
The last values i~ columns (8), (9) and (10) were obtal~ed by ~se of the parallel axis theorem, as follows:Zwxa about e.g. of airplane = 97,391,595  E325.3 x 115.9 2 = 26,691,595. zwa 2 == 33,252,035  5325.3 x 77.8 2 = 992,035 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 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 + 2,899,470 = 14,179,027 lb. In.''' I z = Zwy a + Z':.,rzl1l + ZIlIz = 3..0,287,522 + 26,691,595 + 5,157,186 = 42,136,303 lb. 1~.2 In order to dete~ine the principal inert:a
:
i
\\
( (
~
,
(
, ',0_. .
\:...,'
\\t"'; /~
\'~
\
"'y,
.'1}'
~~
.:
/
L'
'\
\
.
~
~
"\' ~
l
>="
Fig. A3.5b
es ,,:
A3.8
MOMENTS OF INERTIA
,,C.ll~e:r
i .
,.
elillt
,.
H.
..
Y
.
S.
,eCl't1oD. no,......C1bly Clln~ll:r oectloQ tc . Center .ecUon rlbl, eee . Flap Outer p&l1el ~ Outer paDfll Outu p&l1.1 rib' Al1nona
.....
106.6 102
204.6 121
" 11,098
24,751 141.462 3,960 10,983 18.672 12,462 5,401 31,966 11,053 55,4&' 141,319
,.
"
6,20a 11,662 4,631 1,166 6,799 10,114 5,747 1,947
. ,,'
.
~'ole
6a
"Y'
0 0
..'
10 ..
11.
12
t '
."
361.339 491,24.5 aoZ,164 48,598 164,514 274,478 158,407 55,390 176,378 110.200 69.394 5,184 15,181,
.,,
251,229 4091,245 235,844
lJ
H.
m
532,563 1,411,126 685,388 209,880 713.895 1,213,680 798,861 334,850 3,091,083 1,381,600 4,4075,384 194,400 985,520 61,600 23.800 280,000 l,048,S7<l 10,.72 131,670 152,366
Z8:3, .500
33,680
87.1 ,.,
31.4 172
17,601
0
31.4 OS.
8,423 11,731, ua 3,925 3,890,586 .35,434 9,726,484 145.800 3.139,026 53.iOO 51,800 1",000 l,417,5l4t 11,858 418,950 373.034 529,200 104,853 53.,900 347,633 49,005 83S.708 1,111.688 145,200 3.073,490
rUI.lac
oklle
to. Eng1Il.. _ _t
ruztl.~k
...
'0
ec
0
.n
'.""
'70
:l4O
a.aac
0 140
3,4140
c~l1GS
C.. bln &lid. _1lll11.bi ..111 fooc tI'O~U floor, "'tllg flllOlU 8o'Cto.. eow11Jl4 ~ 0111e fram Al'l'elC1JWJ ;;1001' 'l'a11 .cl.el parI.. eCc. Sld.e doora ae.aase door fabrio &D4 40pe Ta11 cone Co_Unr, atadOI18 13
,,~
c .0 66.5
'.0
'.S 210
," ,. s
SO
, ,.,
..
20 11
0 0
..
"
54
'.800
.3,200
16,940 2,394
15,470 106,400
0
58 '5
I,. 62'
SO
0
'.400 3.267
0 0
9,248 4l,3S2 62,234 151,875 5,160 za.aa4 114,308 7,144 83,200 62,108 108,300 604.472 128,JS5 385,378 142.376 6, '7l3.600 ea4,143 83,538 1,420,300 361,325
'0.
I._
2.431
"'.
"'"
sa ase
1.394 11.3 1,040 3 1.140 11,6:12 1.916
,. fuel .,.,to_
Slectr1aa.l Koht 011c.g
UJU!'L.,t .U .
.qu1~t
_.e
""
.. ....
332.4 115
1049.0
90.' ,
ae,c
aaa.c
.... ..  " ,.
0
,., 0
...
ie
.3 .0 91
5.se8
0
0 0
se
28,288 15,509 0
s
100 720 15,509 120 1.200 23,240
0
raa ,640
199,342 18.111 z64,l60 ::>S2,763 :1...35.400 1,383.0<36 210.7841
23,250
av.acc
Sl.
23,240
no
J3 52
as
54
see.ceo
1,809,171 3.369,600
1.142,~1
sa
0 0
.0 .3
M,617
S._
1.924 83,920
ee
253.658 9,060 177,600 1,374 .32,008 .3,4341 U13,ll84 108,160
ee
253.858
0
718,536 692.640 2.769,3&0 391.030 88,109 185,168 116,120 147,J.08 839,680 356.592 9415,840 1,996.600 l,04.,O!8 70,380
'.71.1. '.520
10
0
0
.5
SO
sa
I._
.,.
244.9941 118.699 19,201 116,032 U3,798 1.343,488 395.3541 .3.088,400 3,893,160 1,849,869
0
0
sa
.0 '1.
'.sao 3.498
5,187 13,800 8,157
I,l.OO 
3:13,858
'.060
l'U.834
534,300 331.6341 4010.842 1.0M,OOO 560,593
a.
ea.sce
0
69,300 1,.374
'"
..,
SO. 600
lsa
IS
0
SO
SI
100.7 128 '.0 US 3867 106.6 200.0 10' 200.0 205 780.0 132 75.0 Yl 195 '.0 IS' '.0 190 148.7 l7S '0 1.6 '.0 I., 34.0 1.'"
10. 0
IKP'n'
,go 81.' 7i,35O 82.424 74.8 41.41,199 .2a9,l38 67,342.573 10,283,443 aa,a63.738 1.8,000 .3.205,000 17,800 8.405,000 66,690 13.590.720 6,375 378.075 148,296 13:1,424 324,900 12.130 4,5.31,307 23.680 I,. 78.854 108,900 3,434 628,864
",""
32,556
ea.
2,890.14 28.800
38,600
...
4.980,508 31,090,714
OIl
Pllot Ob.e1"W'.r Fuel Vlry P1eto1 3.:)118 eaDd.l., Flo.. t l1rttta Cbart board., Or1!t 11ght TlreC a.1d. Life :raIt U8U1JL LOAO fOULS COflR.u::,.IO.
.....,.
.,.
..
'"
... , ."
14 14 14
72
U
0
se "
54
I, 7l.O
25.11
",.
.5
ese
. , . ...
,
10
101.
.. ...
.,
1.7&4
0
t ,
..
J3
'00
1,6ao,OOO 1,584,.300 ;',701,996 541,875 21,938 39.864 46,656 1.031,008 32,693 U .390 12,996 3046,834
172,380
0 0
0
157,560 3,675
'.000 '.000
,as
0
5.U7
0 0
5,13'1 see
0
1,890,000 3.649,000 8.603,Oao 452,625 57,038 70,6sa 143,120 3,159,051. .37,824 aQ,837 37,540 461,024
311,898
229,980
1'16,878 17,786.87:1
" .. ,
I
, , ,
,
, ,
J:c.g.
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 77.8 r71 200,000 Or~O 000 48, 0 1 33t1 IE.t\9i,5il5 10.Z67.5aa l~a,oj5 a~ ,aS3 1 10.287,5412 as. 691. see 26,691,595 , , , 292 035 1~ 287 52.3 I i , 99. S'~ 36,804,014 4,136,303 , I 1",179, I .3,662 5,784 I 7,912 , , , 51.8 76.1 , , , a
, ,
, ,
, , ,
~"'~'i"
, ,
, ,
,
,
, , , ,
, , I ,
1'x
J e.g. I a.g.
, :
:_~ :'~~'3
..,,< ....
'.~y
' ".
~.
"~':;,
t,
'):
A3.9
will be given the symbol Ixy, hence  (1) Ixy = I xydA      The unit, like that of moment of inertia, 1s expressed as inches or feet to the 4th power. Since x and y may be either positive or negative, the term Ixy may be zero or either positive or negative. ~oduct of Inertia at a Solid~ The product or inertia or a solid is the sum or the prOducts obtained by multiplying the weight or each small portion in which it may be assumed to be diVided by the product ot its distances trom two or the three coordinate planes through a given point Thus With respect to planes X and Y Ixy = I xy dW
Jxz =
I yz
A3.9
I + I z sin" 0 
Jxz sin 2
I.
(see
Art. A3.lll
I yp = I y I z = Ix sin" 0 + I z cos" .0 SUbstituting
+
I xz sin 2 .0
Ixp = 3061 x (0.9996)" + 9096 x (0.0300)'  181 x .0599 = 3056 Izp = 3061 x (0.0300)' + 9096 x (0.9996" + 181 x . 0599 = 9102 Iyp
=6680
Problems
A,3.7b
n
'"i
r Z i ""i
I =I
xz dW
yz dW
S~try.
1
I
I
'11$
Fig. A3.6
8
Fig. AJ.7
kR
.f0~
,~
It IT,
(1) Determine the moment of inertia about the horizontal centrOlda1 axis for the beam section shown in Fig. A,3~6. (2) For the section as shown in Fig. A3.7 calculate the moment or inertia about the centroidal Z and X axes.
It an area is symmetrical about two rectangular axes, the product ot inertia about these axes is zero. This tallows tram the tact that symmetrical axes are centroidal x and y axes. It an area is symmetrical about only one of two rectangular axes, the product or inertia, !xydA, is zero because tor each product xydA tor an element on one side at the axis or sYmmetry, there is an equal product at opposite Sign tor the corresponding element dA on the opPosite side or the axiS, thus making the expression !ydA equal to zero.
A3.l0 Parallel Ax18 Theore.
T,
1
T
A3.S
' 1'"
1<.'1"
Fig. A3.9
The theorem states that, "the product at inertia ot an area with respect to any pair at coplanar rectangular axes is equal to the product 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 total area tram the given pair.ot axes", Or, expressed as an equation,
Ixy F Ixy + AiY         (2) This equation is readily derivable by re(3) Dete~ine the moment or inertia about terring to Fig. A3.l0. yy and XX are centroidal the horizontal centroidal axis tor the section axes for a given area. YY and XX are para.lLeL shown in Fig. A3.8. axes passing through point O. (4) In the beam crosssection at Fig. A3.9 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 zero, since !ydA and IXdA rerer to centroldal A3..8 Product of Inertia 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
MOMENTS OF
INERTIA
1r
_+x
y
.l3.1l
By adding equations (3) and (4), ~e obtain I x1.+ 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
I X1. Y :l, =!X1Yl dA=!(xcOs~+:;sin,2j)(ycos.z;
xsln0)dA
~ /
(y:a..x a ) d.A
=Ixy
cos 2
+ ~
Therefore, IX:l,Yl is zero when unsymmetrical beam sections are very comtan 2 0 = ~            (6) mon in aircraft structure, because the airfoil Iy_I X shape is generally unsymmetrical. Thus, the 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 inI~ 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 respect 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 about which the ~oment of In~rtia of the a~ea is cosa0! ydA+sln al!x adA2sin0 cos .0 either greater or less than for any other axis I xydA paSSing thru the centroid of the area. Axes for which the product of inertia 1s Ix cos' 0 + I y sin' 0  2 Ixy sin 0 cos 0 (3) 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 axes are principal axes. I y1. = Ix sin a 0 + I y cca + 2; Ixy sin cos 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 angle as shown in Fig. A3.12 about the prinCi?al axes passing through the centroid. Solution: 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).
x:_I,.
='*"'""'O''Ir.._ x
Fig. A3.11
..
Port
1
A~'
,
'ox
Ax_
,03~1
'0,
,
r
,1" 1.205
A,
,0469
Ax
,281 .06205 .J.oli35
A"
,OO~8
Ax'
,211 ,0078 .2188
'x
,0077 ,9470 ,95.1
Iy
.3705 .500
."
,1205
fi:z
1.5U~3 e
,625 ,6 19
,7800 ,7858
.0781 ,1132
1 1 3 12 :z,,:Z2
0019
,167
~2X2J:.2S3 0026
A3 11
lex
of inertia of each portion about their own X and Y centroidar axes. Location at centroidal axes:~nd
ley
~ ~oment
of inertia and product of inertia from reference X and Y axes to parallel centroidal axes:
section has been broken down into 16 stringers as listed in co lumn 1. For the top surface, a width of 30 thic~~esses of the .032 skin is assumed to act Nith the stringers and a Nidth of 25 thicknesses or the .04 skin (see Col. 3). On the lower surface, the skin half Nay to adjacent stringers is assumed acting with each stringer, or the entire skin is effective. Col~ 4 gives the combined area of each stringer unit and is considered as concentrated at the centroid of the stringer and effective skin. All distances, x and y, columns 5 and 8, have been scaled from a large draWing.
= .157
Calculate angle between centroidal X and Y axes and principal axes through centroldal as follows: tan 2 0~2 Ix" ~2(.150
xp
_X x
:7 ::1 _
F,r.
......2 3
"i
I
I
Yp
.040 Skin 6'
jef.
5,
Ti:!f~
I y _ Ix
~~I
Fig. A3.13
12 13
.r
! ,
I
XRe f. x
~4
15
I~
2 0
~ 46 0 
40'
= 23 0
20'
Yp Y
I.
I
U
1xp=1 x cos" .0+1'1 Sin21212Ixy s m c cos a = .44 x .918 2 + .157 X .3965 2  2( .150) x
.3965

,
"
a
Area
,
I
<
,
"
Table 9
x .918
504
m'
I Y1l
Ix sin 2 .0 + I y cos 2 0 + 2I xy sin 0 cos el 2 ~ .44x .3965+ .157x .91a + 2(  .150 ) x .3965 x .918= .092 tn ;"
, " .U
,
4.00 6.05 7.00 7.37 .>0 1.50 1.30 6.90 6.50
",
o~
.. . '
"
lx,
~.,
"<
T'
I
q::'!
a .U a .30
IL 
,"'~i
I ,
fL
'+,( ;,I
...L
.U a.nc ." ." ." ." .U ." ." ra .U ." .n .n ." .n .n .n " .n ." ie " a. .30 .30
.J< ata s
4.90 _5.911 _1.40 _8.13 _8.62 8.81
03 .03 .011
1.14
I
ot
.0< .0< 0<
.~
.14 .38 17
.,
.U .U .U
0.560 2.240 J3.1.5 0.8471 .5.124 _29. 28j 2.660 L8.620 _24.8.5 1.253 9.234 21.18 9.68 I~~' 60 ~.288 1. 215 9.563 12.60 1.241 9.0:19 8.60 1.111 8. ceo _ 4.00 1. 885 12.252  0.35 _0.561 1. 8.51 _33.2:1 _0.83 4.0112 _29.28 _1.6661 9.913 24.8.5 _2.2941 16.916 _LS.10 _2. S2~ 20.481 _12.42 _2.612 22.432 _ 6.10 3.094 21.444 00.35
...
.0<
,
_4.6041 153.85 118.56 _1. 099 120.02 _24.80 _9.443 234.66 _66. 10 _3.600 18.205 _26.53 _2.8:: _2.142 _1.462 12.51 _10.61 _0.680 2.12 _ 4.69 _0.100 0.04 00.65 _5.6:13 181.96 111.611 _4.918 1411.16 24.39 _6.9581112.90 41. 40 _1l.191 108.40 42.90 _2.850 4.1.82 31.30 _1. 891 11.54 16. ;10 _0.122 O. il4 1. 08
~:::: :i~: ~~
.,
a. s
Location of centroldal axes with respect to ref. axes,_y='Z.Ay= 1.465 =  .396" 3.70 l:T
X=,ZAx =  58.238  
Fig. A3.12
Problem 5. Fig. A3.l3 shows a typical distributed flange  2 cell  wing bean section. The upper' and lower surface is stif:enec by Z and bulb angle sections. _Determine the ~oment of inertia ~f the sect:on aDout the principal axes.
~~~ple
15.74"
l:T
3.70
I
I
=186.5
2=
in
SoLut t on :
The properties of the crosssectlon depend upon the effective ~ateria1 Nhich can develop resis~ing axial stresses. The question of effective ~aterial 1s taken up in later c~~pter. Table 9 shows the calculations for the moment of inert~a about the ass~ed rectan~Jlar reference axes XX ~nd YY (see Fig. A3.13). The cross
IyIx
20=1632.5',
I xy
= 2 (36.41) =.29696
431.7186.5
.0=816.25'
1xp=Ix cos" .0+ 1y stn" ~2 1XY sm e cos e = 186.46 x .9896 2 . 4 3 1 . 7 x .1438 8 2 E36.41X .9896 x (.1438)J = 181.2 10.
A3 12
CENTROIDS ,
CENTER
OF
GRAVITY
MOMENTS
OF
INERTIA
I y p = Ix
2
e
Sections.
'0 ; Table A3. :0 'through A3.15 and Chart ...... si'le the section proper t t es of a few st:uct.. . ral shapes G8rrnOn to arrcra rt . Use of trees tab Ie s will 08 :TIade in later cha;Jters of th:!.s jook.
v.~
=437
in ...
,
'I =T ,il ..z~1 I R 'T
,
n"TI
X
,~
feJ
x'
,
I
'i
,::,/R
AS<T
T
T
X
B I
'I
r+
~T
X
~a
J ..J,_ __Ai2==:jI
I I
VR
r
I
I
y
R) . ) T ..'A
I
SECTION ELEMENTS
"
Z
U
NOMINAL DIMENSIONS
I
IDcMa
R
125
AREA
Square
Inches
I '" 'nInches
0035 .00<12 00" , 0071 0070 0084 .0109 .0090 Olla .010
Puc
I tOyy
rn '"
=t: +
j+ HI i1
. '00
.625
.eas
.750 .625
+
'"
.063 .050
"
Z
NOMINAL DIMENSIONS
A
.">REA
B
InChes 375 500
>
,
Inch 106 153 204 .207 JI5 .327 0995 .146 196 201 245 .247 ~53 362 .319 1~0 .H5 187 193 .229
:>ECTlON ELEMENT:>
:>q..are InChU
'n
0009 0020 003, U045
Iyv
I '"
"0'<
I ~v,'
I
I
00<.
.0062 0122 0146 0069 <lOS3
,I><, 0 o
"" 20'
1+ I}
cia
00,
Hl
r*"
18
~ "
'" .cso
.050
HHlf#I~
.078 .050
"', '"
,,, ".
125
a ." ,
..
.125
ra
us
, 1~5
".
.125
tl ;::ll::
f#as
~
~
~
.'w
1.000
.875
mceo '"
.018
~ .076
".
.ras
125
'"
259 .253
;0
.298
'" '"
.no
ce
.caro
.U
139
.CO
.0439 00'
~ J78
~ ~
.,
"
1.00<
sas
,875 1.115
1. 12.5
~ 076
".
.050
0<
.125
., ., .,
..
"
oe
.n
259 ~ 3
H: HH~
.750
'"
'"
".
' 125 .094
.0127 0852 0 17
,n
'"
~ 094
.050 .063
1+
",
,
.aao
0790 .09\6 115 .131 158 .11. .147
155
.oi rr
296
pt
~ .076
.125
'0 4 125
'"
Ri: te
i#.g.
*l.or
'~ us
.17
aie
.328 .139
.500
4*
.H
.358 .400
ao
li'+.\
~.
I. 000
.~.
.875
1. 250
.150
1.
2~O
.. . =ll!:
'" ~
~
~ .094
125
".
'"
'"
..
sea
H\!+.\aa
as
ae *
, ~ 1E 094
~
~ ~
.125
". '"
".
m 243 es
294
0159 0009 .0020 0025 ,}047 0055 0065 0079 .0094 0~13 .03ll4 OG26 0029 0049 0058 0083 0096 0113
.0123
~152
,;;
106 .148 144 1ST 2JT
;0
.csro
, 0132 015" 0183 0151 .OIS4 0214 (;276 0376
,n61
,
,
290
347
J~9
'"
aaa
2.
131 .158 141 .171
'00
1.12..5
.441 464
.125
.445
LOOO
I.
.125
'W
:jf
* =1t
1. 375
1.000
I. 250
..... ::$.:
~ us
.094 .094
".
.210
".
, .72
464 457
." ."
514 501 519 512 543 .562
, ,
* :j
11~ ,..g.
."
.750 .875
~ ~ .078 I : ~94
~ ~ ~ 125
.125
.125
'" ,
." m
134 139 .178
'00
,,.
2n
181 .216 .256 .328 236 279 .359
'"
558 555
.125
I. 500
I. 750
~ I. 2 0 .094 , ,c
000
"
...
,53
aae
,, " . ., s
.0511 1308 0439 0511 .1040
..
.0367
"
.549
.15 .HI
'sae "
L 125 L 000
1. 375
.,. Pi!:
I .
.125
ase
asa
01,6 041 061J7 02 .0031 805l 0060 01(12 .0117 .0135 0156 .0183 0225 0232 0269
" ,0643
0281 nB5 03:H
,
, ts " '" ,
zas
'159
'"
'"
397 .417
087 HO ;86
lH
e9a
"
"
122
H'
sa
,875 1.250
I. 125
1. 500
~~5. ,.
';0
.53 220 .267 .344 .265 .314
~ .125
~
12
125
=i so
1. 375
,,
csa
125
,500
."
.115
,,
'" ,
"
".
ras
.125
"
~ .094
~
."
,
a
393 .520
H} ss
~
"
125
2.000
'W
m
2,500
1.000
"
ate
505 .029 ,298
f:i!: ~
\.;00
,.
I2S
00.
"
.,
..
.125
."
a
m
.55 338
1. 125
'" .".
.0444 oa .2,. 0319 0441 27\ ~52 .0507 :68 .. 5 sea .33 '1616 .0495 .214 I 4Je .310 .0569 .52 0695 303 m ,66 0661 0757 .0832 I 0930 .53 123 0071 0493 .0190 .0650 ~57 .0234 'J79 .51 481 355 0698 5n 0334 ,07 .0389 0607 l52 'J484 (>994 .345 ss 1292 .70 1106 J08 .00 3 .179 0540 171 .00B i)621 517 0169 0705 547 ~68 .0198 543 :66 0818 .0242 S2S .1002 ' 359 .0403 1010 I .352 557 .0504 1246 I .346 SH 059\ .339 .1451 .532 , \710 ,52 2035 .0076 .J76 .177 ;0 I :7 0090 0939 HB ;H .0249 1236 258 .e rz 0289 1485 , 257 0414 350 004 1233
_3 225
,"
, '"
,
H '"
,""
, '"
"
,2,
"00
.955 .941
.354
A3 13
11"'7Y
~,~ w , '
I
Table A3. 12 Properties of Extruded. Aluminum Alloy Equal Leg Angles. (Ref. 1938 Alcoa Handbook) Axis XX or YY d I P 0.004 0.004 0.006 0.008 0.012 0.016 0.021 0.029 0.033 0.042 0.059 0.074 0.058 0.074 0.107 0.135 0.096 0.121 0.174 0.223 0.18 0.27 0.34 0.41 0.183 0.220 0.219 0.217 0.311 0.301 0.298 0.293 0.38 0.37 0.37 0.36 0.46 0.45 0.45 0.44 0.55 0.53 0.53 0.52 0.61 0,61 0.60 0.60
,.:
jit Y
:Ld ~ z Dimensions
'~Xit a! L
Area
R
~.ID.
Axis ZZ P 0.111 O. 142 0.141 0.141 0.199 O. 19:J 0.191 0.192 0.24 0.24 0.24 0.24 0.30 0.29 0.29 0.29 0.35 0.34 0.34 0.34 0.40 0,39 0.39 0.39
3/.
'I'
1 1 1 1
3/' 3/'
II'
II'
II'
II'
3/32 3/16 3/16 3/16 3/16 3/16 3/16 3/16 3/32 3/16 3/16 3/16 1/4
1/'
3/16 3/32 II' 3/16
, a
3
1~ 1/4 11/4 1_1/4 141/4 1_1/2 11/2 11/2 11/2 1_3/4 13/4 1_3/4 13/4
II'
.23'
339 .230
.30 .43
.56
1/.
3/32 3/16 II' 3/32 1/' 3/16
II'
1/.
3/16 1/4 5/16
.62 .81
II'
II'
1/'
1/4
.,.
1. 16
.49
.n
0.181 .0015 0.199 .0018 0.214 .0026 0.227 .0034 0.211 .0048 0.276 .0066 0.290 .0085 0.314 0124 0.34 .014 0.35 .017 0.37 .025 0.40 .032 0.40 .024 0.41 .031 0.44 .044 0.46 . 057 0.47 .039 0.47 I .050 0.50 .072 0.52 .093 0.53 .08 0.56 .11 0.58 .14 0.61 .17
T
I
~:
,.y
T..ble A3.14 I'rO,",rlies 01 UnequaL Angles
~R
,x
1.
"
"
0
NOMINAL DIMENSIONS
A
a
Incbe.s 500 625
Y
.050
A~~ R
'Square Inchel 0535 0818
AREA
SECTION EL.MENTS
i
.126 162 150
'" '" H= ", '" ", '" q '" '" ", '"
'45' " 45'
ft
.oea
.083
750 .625
~
~ ~ 125 ~ 094 ~
~ .125
""
"'''
187 222
lJIChell
.OOIl
.US
at f#
ta ~
1.000
.US
... ."
.108
aa
211 I . l40
'" .,
250
272 .320
.178
.n
f# filia
.125
.....
fi elf
c#
~
~ ra
~ .125
sec
'.300
1.250
1. 00
:.m 156
.,.
.228 156
."
322
i
.. '"
"I
'"
t"yy
.,
ro,
292
"
.361
.251 J27 .401l
'"
'~W
I ,
I
I , ,
C z
ti&,
~ ~ 156
.125 .125 ,
.15 .167 .228 239 . 249 .310 .321 .331 .300 .383
es
aa
... ....
.455 .591
1.250 1.750
I. 500
~.m
.156
y
,
asa
.506 517
'"
.017 .011)6 .01 2 .0111 .0155 .0197 .0278 .0346 .0217 .0306 0383 .0463 .0584 .0512 .0647 .0712 .0550 .0698 .0834 1078 .1144 1377
, ., ., .,. ,. na
.211
221
"
.1
217
.,
, .291
2 2 ' 310 257 305 .209 1 .400 .203 ' .393 00 .381
~,
.194 .l91 .465 .282 474 .276 ~68 .271 , csa .312 .365 .462 .360
'"
.389 382
".
'"
"
""
',"
". '"
'" s
n
l"Y
T ..ble A3.I3 PrOpertin of
I I
r:7
~
Tee~6eC<lOtl3
~0il~
~!aj ,
',
'r
X
6~
,
'"
Y I
"T
SOMINAL DIMENSIONS
Y I
050 .063
,
Inch
SECTION ELEMENTS
,~
+ '" R=j
~1.
r*" H!f.li.
"
",
or
"'V<
173 209 .209 .296 292 .290 379 .192 198 286 371 .371
Ip'IY
Inches 287
, eo
L
,125
.0'
.063
". ,
ias
375
r. 00'
L 2~0
'00
on
bO
!t: $:1. ,
.
,
;00
,;0
, ne '"
.H
, '"
aas
" "
W 0
NOMINAL DIMENSIONS
A
a
incheS
R
m
SECTION ELEMENTS
AR<A
~
0011 .0098 .0129 0158 O1B8 .0 ISS .0169 0244 0200 .0245
'
.228
,
,
,
0<,
240 204
".
. 125
.lH
.228
'" ua '" no
a
2,5
, ,"
.252
120
" azs
I. l25
.ca tea
.212
I ~L
; 23 24 ,
Ht,
I. 62~
, ,
"
'"
.67>
, , ,
750
,
,
I L 125
" 7'"
L L 315 , L .175
ias
.n,
240 .31l
.0182 .0345 .0052 0080 .225 ,0 60 .232 I J186 0301 Q354 .3H n, .0449 .180 II oJI .0228 .254 314 ,; 39 .047; .3H .0 0, a
ie ia
, ,H=
.750
L '00
'"
~ 078
07
125
'"
,
ats
I. 000
~
094 .094 .H .094
aea
"
tzra
"
H=,
II
!
aa ,
.341 .388 HI H
,
'"

l. 625
I. ,,0 l. J " l. 02 L B7:;
"
125
'"
,
,
ate
H'
I::tJ ~l.
13
250
,)94
125
",
458 50S
a cI
H
516
.431 .321
2. 125
234
o~oa
0314
,)40~
na
,,
,
".
1,8
,os
, ,
,
'" H" ,
'"
234 sa , '" 230 0512 0"; J393 062 0393 0627 'J394 )5l3 G6 9
322 H5 .Hl
.338
40~
~
~'
"
.oec
.125
ia us
'OS
l25 .156 156 125 .188
I
"
.,H
0
" 530
as
4~0 531
a ,
2.000 2.000
2.
~OO
Hs~
\ira
,,'
'" 740
H;
.65 626 ~63 464
HI
"
..,44
1183
.789 ~51
an
'"
on:
.0995 1200
4.14
'"
,,"
, ,
""
se
ss
,,"
'"' , aa
fiIT ,
"
.1104 1546 II 4
'" , ,.
L L L
';0
ua
, a
L 262
, a
coo
'" "
r#
3.500 . 000
"" Hm
J 13
l. 52 2.00 2.60 3. ~3
".
I. l88 2, ~24
a. 037
L a t , 63d
,. "
z aa
."
I,
, '"
, ss ,,
L
ea "
A3.14
CENTROIDS,
I I
,
"
i xo.
1 2
'Sect.
,
!
,
A
17/32
Dimensions
I
1
I c
I
't
'1
'2
ATOll
!
'rop.,ubM
I
xx'
e
I
in.
IlI:X ,
e lI:X
AlI:1s
src rU bou~
AlI:l.S
I rY
e rr
e'
13/16 I 1/'
, ,
1/16 .050 3/32 3/32 3/32 1/16 9/64 I 1/16 3/32 ; 3/32
.068 .113
.319 ,510
,
 ...C;....
liZ
lI2
,3665
..l.
~::[: _e_
'T.
7/'
7/. 7/'
,313 ,313
.383 ,383
,
,
,
9/16 9/16
3/4 3/4
.00215 I .114
!
' A
*
I ,
.1061
.0120
.329
.00302
,133
"
e
,
I
,
1l/8 1_5/16
3/8 3/'
.0751.075 3/32 1/16 . 3/32 13/32 3/32 3/32 1/16 3/32 I 3/32 3/32 3/16 I 1/8 3/16 1/16 I 1/16 3/32
.1514 I .0239
I
,
I .00550
,
: .182
.171
, ,
,
rF
~
,
II
I .230 I
i
.0080
7/8 7/8 1
.095 .1375
,4315: .00111
,
I
,
.4375: .00148 .500 .625 .906 1.25 1.125 .133 ,133 ,121 .201 .300 .217 .3::18 .255 .658 .757 .959 .793 ,500 .562
.75
,
. ''l'
I 1'0 _LL., ,
! 10
12 13
I ,
lI2
i .1736
."4
; .37l
I .480
I
,
11/4
,_'t\
I
11
11/16
,
l
r
. 4
C 
1_13/16 11
3/1' 1/.
.2920 .525
I
,
.677
I 2_1/2
14
15
: 2_1/4
1_3/8
1.1/8
,
1/.
illS
' 1/8
!
.581
! .950
I , .338 ,
.362
~
.295
l/S
1/16
1/16
.888 .176
I~ri~
~" ;~ ;
rr6 .L.
'C
11/16
3/4 3/4
1/16 : .131
!
.0178 .0335
,360 .446
I .687 ,
IS
1_5/8
1~5/8
I
1
5/32 5/32
,
1/16
! .168
.306 .359
17
i 1/16
! 3/32
1/8 3/32 5/32
1/16
I
,
I .0458
i
IS
18 2.
I
,
2 2
1 1/ 1 6 1_1/4 15/32
11/2
3/1'
3/16
1/4
! 3/32 I 3/32
i .474 i
I .509
1
I .812
.812 1 0
I
,
: .0252 .0475
!
,0191
I
i
' 3/32
1/'
: ,0821
i
i .479
.590 .326 .302 .281 .282 .514
.348
,1. 0
2
2_5/16
I 1/8
1/16 3/32
1/.
~ .378
.668
: .279
, , ,
.1095 .2326
i .541 i1. 0
11.156 ,
21
: 5/16
,
I 5/32 !
I
' 5/32
j":J:'
22
I
,
1_1/2 11/2
i 1l/8
2
1 1
; 1/16 , ,161
i
; ,0171
23
I
,
l/S
9/32
1/4
I
I
I
1/8
.384
I
,
,0346
I 1_3/4
13/4 I ,
2
2_9/16
I .643
I .809
I ,0599
,
.5/32
1/.
.600
, ,875 , .815
i 1. 00
I i
.750 ,750
,
I I
i
26
!2
I
i 2_1/2
1
1 1_ 1/
!
l/S
!
5/32 .050
, ,
i
! .385
i
; .800
,..A! ir~
27
11/16 13/16
3/32
3/32
, .404
, .793 ,
i
.795 : :
6625! .78251 .600 ,725 .719 .719 .531 .469 .531 I .525 .655 ,787
1 i
28
28 3. 31
3/16
I ,
,
!
i ,
5/S
3/4 3/4 3/4 '/1' 1/2
1~~/2
13/16
1/16
, .595
i .773
I .349
i
.456
I 'r' '.,J
~
,L
2 2 ,2 : 7/8 7/8
! 1/4
i
1/4
'I
3/32
3/32 ,1/16 1/16 ' 3/32
.2378 .2855
i
!
; .0282
'te'..I
132
11/4
! 3/32
.1820 .0120
.oaes
.0056
...
:
: 33 I
'A ;:I
I 34
35
.050
1/16
.099 .115
.~,
1.,,10
"
9/16
9/16
I I1
1
1/16
I i
,
.151
38
,1 1/4
.ove
1/18
3/321.170
I .00446
I
L'"
e
a ...
I I
I I I
I .227
.302 .360 .267 .265
~
31 38
3. 4.
LII~JI
, ,
,
11/16
1 _ 1/ 4 1.1/2
2 2
.100 .120 I
I .508
.825
3/4 3/4
.075 .072
.110
.1446 .1826
'
i .012
.102i l 0 2
, ,
: .249 .330
, , .757
.750
.604
i
1.00'
3/32
1.00
I .0180 I .0228
.714
.700
CHAHT A3. 1
0.35
.1
~
~
1 I 11 P:llili/lli il1mHW Ii!1~IWilf 'IWi' illi J!:..I.IIIJ I ;~I Imll! LIIIIII :'. Iii J 111m II Ii !II iiiill'11~1.ljjj1!l!!I!lIljjllHmmnn ~
jli
1.0
.. ..
~
w
.1
JIi I'
.,
.,
..
11:1Mo.3O .. ! II
1
,:1: i I 1,
.,
.'_, i:
'+,
II
I .:: I T1
...
~
~
,. ,. rZ
~
i: 'i~
:'1'
, . , I'I
;':. :1
::1.
"" ,i 1
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
Z tl
tl
~
'"
~
030
i
0
~
~
0
z " .,
0
Ii
0,15
~
~
Ii w
0
~
~
" .,
r
.,
'"
'"
~
<:
'" r0
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~
.,
0.05
'" ., '0c"t
~
I~
"
Mom. ollnarUa.:
1.0
1.0
2.0
2.5
3.0
I
.025
~.
'.
.'c.
II  IIEIGIiT OF CHANNEL
,
II; HEIGIIT OF CHANNEL
3.5
I;
V~
\1 3
CENTROIDS,
CENTER OF GRAVITY,
MOMENTS OF INERTIA
~3.1S,
t~e ;r~~ t~e
(5) ~or the ~ean Beetien in ~i~. calculate the ~oment oi inert~a ~bo~t
ci~al ~xes ~ssu~i~g :~e
fJUY
st~i~;~rs ~s
only effective
~aterial.
,.. ,
Fig. A3.14
J
"
/0_
4... IS I. , .,
l.'
Il"
8 ..;...
10
' I
...+
i
08
Fig. A3.15
~~t
I'.T L I zo'
1' t
o.?
Od~
15
Fig. A3.19
1"8 i
(1) For the section of Fig. A3.14 determine the moment of inertia about each of the ormerpal axes Xp and Zp. (2) Calculate the moment of inertia of the section in Fig. A3.l5 about the ?rincipal axes.
(6) F~~. A3.19 shows a wing 0e~~ section with a cutout on the lcwe~ surface. Jeter~ine the ~oments of inertia about the ~r1~:i~~1 axes assuming the eight stringers are the cn:y ef~2c t ive :naterial.
It ~ l : D~
I
5 spaces @5'" 35"
F,g. A3.20
Fig. A3. 17
(3) Fig. A3.l6 illustrates a box type beam section with six longitudirral stringers. Determine the ~oment of inertia of the beam section about the principal axes tor the following assumptions:(a) Assume the beam is bending upward putting the top portion in compression and the lower portion in tension. Therefore, neglect 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 compression.
(4) For the three stringer single cell boX beam sectien in Fig. A3.17, calculate the moments of inertia about the principal axes. Assume all 'Neb or wall material ineffective.
(7) Fi~. A3.20 shews a 3 cell ~ulti,le flan?e beam. !he 7 flan~e T.embers on t~e upper face of beams have ~~ area of .3 sq. in. each and those on the bottom skin 0.2 sq. in. each. The bottom skin is .03 inches in thickness. Compute the moments of inertia about the principle axes assuming that the flange :nembers and the bottom skin comprise tr.e effecti~e material.
Fig. A3.21
..
(
Fig. A3.18
r
/9 '
(8) Fig. A3.21 shows the crosssection of a small fuselage. T~e dashed line represe~ts a cutout in the structure due to a Goor. Assume 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 section about the principal ~es.
CHAPTER A4
engers safely, efficiently and comfortably over various distances between airports. on the other hand the Air Force Fighter type of aircraft has a job at shooting down enemy aircraft or protecting slower friendly aircratt. To do this job efficiently requires a far difterent contiguration as compared to the DC8 transport. since modern aircraft tly in subsonic, transFurthermore the Fighter type airplane must be sonIc and supersontc speed ranges. furthermaneuvered far mere sharply to do its required more, there Is a wide range ot Wing configurjob as compared to the DC8 in do Ing its reatIons, such as the straight tapered wing, the qurred 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 conairplane 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 body to Withstand these acceleration inertia wing and thus effect the magnitude and distribution 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 overthe wing. The theoretical cal the maneuvering accelerations for the DC8 are culation ot the alrloads on the airplane 1s too not dictated by what the human body can withlarge 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 perthe airplane are determined by a group at enformance 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 overall 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 alruse at aerodynamics, it is that phase at aerotransportation, along with uniformity and efdynamics 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 reon 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 aerodynamrcs, to insure or guarantee the per Limit or Applied Loads. The terms limit and applied refer to the tormance, stability and control at the airplane. same loads with the civil agencies (C.A.A.) A basic general overall knowledge at the USing the term limit and the military agenc i es loads on aircratt is desirable in the study at USing the te~ applied. aircraft structural theory, and hence thiS Lim1t loads are the maximum loads anticichapter attempts to give this information. In pated on the airplane during its lifetune of a later Chapter dealing with wing design, this service. Subject will be further expanded. The airplane structure shall be capable or supporting the limit loads withuut suffering A4. t Limit or Applied Loads. Design Loads. detrimental permanent deformations. At all loads Because an airplane Is des Igned to carry up to the limit loads the deformation of the out a definite job, there result many types or structure shall be such as not to intertere with aircraft relative to size, conf1guration and the sate operation of the airplane. performance. For example, a commercial transport like the Douglas DC8 is designed to do a Ultimate or DeSign Loads. jab of transporting a certain number of passThese 17NO terms are used in general to mean
A4. 1 Introduction.
Betore the structural design or an airplane can be made, the external loads acting on the airplane in flight, landing and takeott conditions must be known. The complete determination ot the air loads on an aIrplane requires a thorough theoretical knowledge or aerodynamics,
A4.1
>4
A4.2
GENERAL
LOADS
ON
AIRCRAFT
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 oV9rall 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 fabrication 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 maximum acceleration it can be SUbjected to in flight or landing. Since these are under the control of the pilot it 1s poss~ble in emergency 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. Although 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 airplane, 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:D ue to Airplane Maneuvers. (under the control of the ~ilot). (1) A1r Loads { Due to Air Gusts. (not under control of pilot).
(4) Take
ott
HOi s t i ng Airplane. , 1 1 I " Towim:: Atrul.ane. (5 ) upee a ....oa"'s ~ " 1" I. <.,.. .... ~ " { Beaching of nu_l~y~e ~~~~l~r.~ Fuselage ?reSsurlzi~g. (5)
Wei~~t
In resolving external loads :or stress analysis purposes, it is convenient to ~ave a set of reference axes. ~he rs!ererce 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 ~.~.
y
z
x
x
Z
Fig. A4.0
The term weight is that constant force, ,roportional ~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 equilibrium, namely. the weight of the airplane, ~he air forces on the complete airplane, and the power plant forces. The pilot can change this balanced 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 deaccelerate. 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
(2 ) Landing.Loads
Landi ng on Land.
(Wheel or
sk1 type).
A4.3
F :: unbalanced
tw
Fig. A4. 1
        \
Motion
~
lf an acceleration normal to at = ra, and a = n the flight path at A and directed toward the center ot rotation (0). From Newton's Law the effective forces due to these accelerations are:2 F = !".rwlf = Mv /T _
n
rw
Fig. A4.2
where w = angular velOCity at the paint A. a angular acceleration at paint A. r radius at Cllrrnture at :light path at point A. The inertia forces are equal and OPPOSite to these etfective forces as indicated in Fig. A4.3. These inertia forces can then be considered as part at the total force system on the airplane which is in equilibrium. It the velocity or the airplane along the path 15 constant, then at = zero and thus the inertia force F = 0, leaVing only the normal t inertia torce Fn It the angular acceleration Is constant, the following relationships hold.
W 
a force on each mass particle of mLa, m~a, etc., thus the total effective force is zma = Ma. I f these effective forces are reversed they are reterred to as inertia torces. The external forces and tha inertia forces therefore torm a force system in equilibrium. From basic Physics, we have the following relationships tor a motion of pure translation it the acceleration is constant:v 
(1)

(2 )
(3 )
where,
5
(6) (7 )
(8)
moved in time t , Vo = initial velocity v = tinal velocity atter time t. Inertia Forces on Rotating Rigid 80dies. A common airplane maneuver is a motion along a curved path in a plane parallel to the XZ plane of the airplane, and generally referred to as the pitching plane. A pull up trom steady flight or a pullout from a dive causes an airplane to follow a curved path. Fig. A4.3 shows an airplane following a :urved path. It at point A the velOCity is increasing along its path, the airplane 1s being subjected to two accelerations, n~elY, at' tangential to the curve at point A and equal in magnitude to
Center of Curvature
= distance
where Q
= angle
of rotation in time t.
In Fig. A4.3 the moment To or the inertia forces about the center of rotation (0) equals I'tra(r)= MTaa. The term MT a is the mass moment pf inertia of the airplane about point (0). Since an airplane has considerable pitching moment of inertia about its own center ot gravity axiS, it should be included. Thus by the parallel axis
To = loa
Q
+
where I = Mfa and I e.g = moment or inertia of airplane about Y axis through e.g. at airplane. Inertia F0rces For Pitching Rotation of Airplane about Y Axis Through e.g. airplane.
I e.g a
            
(9)
01
r I
Flight Path
an=rwlf=Vlf / ?
=E2~~MTa
AI
'.g. M? w lf = Mv a / ?
Fig. A4.3
In flight, an air ~~st may strike the horizontal tail prodUCing a tail torc~ which has a moment about the airplane c.g. In some landing conditions the ground or water forces do not pass t~xough the airplane e.g., thus producing a moment about the airplane e.g. These moments cause the airplane to rotate about the Y axis through the e.g. Therefore for this etfect alone the center or rotation in Fig. A4.3 is net at (0) but at
A4.4
GENERAL
LOADS
ON
AIRCRAFT
o. Thus F and F t n 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.
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 a1rfoil chord line. As the angle at attack is Changed the resultant air force Changes in magnitUde, direction and center of pressure location.
R
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 reterring to the Shape of the crosssection at a wing. Figs. A4.4 and A4.S illustrate the air pressure intensity diagram due to an airFig. 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 illustrates this resolution into lift and drag components.
Angle of Attack
12"
Fig. A4.o4
Fig. A4.5
stream flowing around an airfoil shape tor 'both a positive and negative angle of attack. The shape and intensity at this diagram is influenced by many factors, such as the Shape of the airfoil itself, as the thiokness to chord ratio, the camber of the top and bottom surfaces 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
~
..
I :
:
'0
1 rAUlODTN..... IC
,J
Cl1:NTEII
o.p
'UGHT DlllltCTIOlI
FLIGHT DllIECTIOIII
'!'?5?:>
Fig. A4.9
Fig. A4.8
Aerodynamic Center (a.c. ). Since an airplane tlies at ~y different angles of attack, it means that the center ot pressure Changes for the many flight design conditions. It so happens, 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.
A4.5
r/o
..,
E
tiplying factor by which the forces on the airplane in steady flight are multiplied to obtain a static system ot forces equivalent to the dynamic force system acting during the acceleration of the airplane. Fig. A4.ll illustrates
Z
&l Tail)
T_
Fig. A4.11
T
L
=engine thrust.
= total
forces in steady horizontal flight. L represents the total airplane lift (Wing plus tail). Therefore L = W. Now assume the airplane is accelerated upward along the Z axis. Fig. A4.12 shows the additional inertia torce wag/g acting downward, or opposite to the direction ot accaleration. The total airplane lift L tor the
Z
",L
"... e.g. 0
w
Fig. A4.12
hg.,
unaccelerated condition tn Fig. A4.ll must be mUltiplied by a load factor n to produce statiC a equilibrium in the ~ direction.
Thus,
x ZF ~.= 0,
0,
n L  W  ~ e., ::; a g a
=
~ +
ZMy
 '"
+ '"
Since L Hence n
te = a
An
cos
Q+T
sin ~  E
=W
::; 1
+ a~
0,
l1a
La  D b + Tc
cos
ZFa
= 0,
0, 
 '"
airplane can at course be accelerated along the X axis as well as the ~ axis. Thus in Fig. A4.l3 the magnitude ot the engine thrust T is greater than the airplane Drag D, which
..L
cos
Q +
T sin
 IL
 E
=
Im
Ma 
La  Db
+ Tc cos ~ +
Ee
=0
~.g.
r Forcee
Signs used: lMoment  C'Iockwf se is positive. Distances tram e.g. to force Plus is up and toward tail.
A4.7 Load Factors.
w
ng. A4.13
~waz g
T is greater than D
The term load factor normally given the Symbol (n) can~dettned as the numerical mu1 The bar through letter Z has no significance. ing without bar. Same mean
causes the airplane to accelerate forward. It is convenient to express the inertia force in the X direction in terms of the load factor n and the x
.Zq
A4.6
GENERAL
LOADS
ON
AIRCRAFT
,eloc~ty
:ar
~om
imi t.ec to
2.
ax
reasonable ~lide speed whic~ ~s su ficient to take care of ~easo~a8le flis~t opera~icns.
A4. 9 Gust Load Factors.
~~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
Therefore the loads on the airplane c~~ be discussed 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 ?articular 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.
,
"
be ass~ed to vary linearly wl':h :he a~gle of attack. Thus in rig. (a)J let ?Ol~t (3) represent the nomal air?lane :orce ccef:icient C7
'')..
necessary to ~inta~n level ~l~sh: velOCity V and point (s) the 'ffilue
Softer
The Civil and Military Aeronautics Authorities 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 produced 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 included 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 operations 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 consideration. 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 accumulated data obtained by instal:ing accelerometers in commercial and military aircraft and flying them in all types of weather and locations. it has been found that a gust velocity or 30 ft. per second appears sufficient. The speed or velocity of the airplane likewise effects the loads on the ai~plane. The higher the velOCity the higher the aerodyna~ic Wing ~oment. Furthermore the gust accelerations increase with airplane velocitYJ thus it is custcmary to limit the partic~lar ai~plane to
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 ranian) .
KU~
V Fig. b Fig. a
Tr4e load factor increment jue to the gust KU can then be expressed
KUVS
m 575 ',.,[
(A)
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 acceleration due to gravity.
A4,7
in C
zA
wit~
on
~V
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
(E)
Therefore the gust load factor n when airplane is flying in horizontal altitude equals
n ;: 1 ... 3KmV
I,.J/S
1s
(e)
in a vertical altitude
(D)
3KmV
 ',,/s
               
A4.10 illustration of Main Flight. Conditions. VelocityLoad 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 requirements can be represented by plotting load factor and velocity to obtain a diagram which is generally referred to as the Velocityacceleration diagram. The results of the above specification would be similar to that of Fig. A4.14. Thus, the lines AS and CD represent the restricted 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 ircjcne 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~.
would gl ve an acceleration less than the limited values given by l1nes AB and CD. The positive ,and negative gust accelerations 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 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. 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 gust lead factor, thus inVOlving only partial pay load, fuel, etc. On the dia~am, ~he points A and 3 correspond 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 respectively. Generally speaking, if the airplane is deSigned for the air lOBds produced by the velocity 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 reqUirements 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 understanding 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 airspeed indicator, that is, one that would indicate true air speed at sea level under standard atmospheric conditions. The relation between the actual air speed Va and the indicated air speed Vi is given by the equation
VI
.~
Va
A4.8
GENERAL
LOADS
ON
AIRCRAFT
I, I'
',,;,
~
It;l""
,,""'
1Cl:T
,~~.
I J"
L ;" 1,/
L [,,,
'Qi
'.__
i!~;~~~~;i:;'"lli~~,~;["!
A4.11 Special Flight Design Conditions.
E "'T i.l '1;;, ","" ,f'" t ~:~ ;~"L~ :;.t.r' : ._::: ,:::~::::~:~T:: U~~::
<: ...
'''.l'''l:;;'~
I,; D.""
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. In cases where the land1ng gear is attac~ed to wing or when the :uel and engines are carried 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 inboard of landing gear and engine attac~~.ent points.
A4. 12 Example Problems Invcivtng Accelerated Motion of Rigid Airplane.
<L
L i::: k
~ ~rr: ~~~;U ~~ 2; ;lli::" '~ii~l, "'?'. ,;S I , ~~~~~~ ~~imI; ~il:I~~ ~~;t~~~~~~~:~: '~~J:~A:EfSS:! H:' \~~;~ i~~~~ ~~i~ m~;~~~~g~~~;g~ i;~:!~mlli~~lli~tiTIHr~ li::~i :1',::[" ""W i :~I"" :!i:lFi IeJ!iijjJijjjij'iitljijrJlic;'J' Fi~:t4:' I.'
As previously explained, it is general 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 presented to illustrate this general proceaure. Example Problem 1 Fig. A4.l5 ill~strates an airplane landing on a Navy aircraft carrier and Jeing arrested by a cable pUll T en the airplane arresting tcok. It the airplaneweight is 12,000 Ibs. ar.d the airplane is given a constant acceleration of 3.5g (112.7 ft/sec~), ~lnd the hook pull T, the wheel reaction R, and the distance (d) be~Neen the line of action of the hook pUll and the airplane e.g. If the landing velocity is 60 ~.P.H. what is the stopping distance.
w" 12000 lb.
There are $anY other rli~~t conditions 'which may be critical ror certain portions at the wi~ or fuselage structure. Most airplanes are equipped with flaps, to decrease the landing 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 must be checked for all possible flap conditions within the specified requirement relative to maximUm speed at which the flaps may be operated. Generally speaking, the flap conditions wtll efrect only the wing portion inboard or the flap and it is usually only critical for the rear beam web or shear wall and for the top and bottom walls or the torsion box. This is due to the fact that the detlection flap moves the center ot pressure considerably aft thus producing more shear load on the rear shear wall as well as torsional moment on the conventional cantilever box metal beam. The a1rplane must likew1se be investigated for aileron conditions. Operation of the ailerons produce a different air lcad on each side or the airplane wing which produces an angular rolling acceleration of the airplane. Furthermore, the deflected ailerons change the magnitude and location of the airfoil characteristics, thus calculations must be carried aut to determine whether the loads in the aileron conditiOnE are more critical than those for the normal flight conditions. For angular acceleration resulting fram pitChing moments due to air gusts on the tail,
Lx
T
24"
Fig. A4. 15
Solution: On contact of the airplan~ with the arresting cable, the airplane is decelerated to the ri&~t relative to ~ig. A4.15. The motion is pure translation horizontally. The inertia ~orce 1s
l1a
=~
= (12~00)
3.5g
= 42000
lb.
The inertia torce acts opposite to the direction at acceleration, hence to the left as shown in Fig. A4.l5. The unknown forces T and R can now be solved tor by using the static equations of equilibrium.
ZFx = 42000
hence,
T cas 10 = 0
T = 42700 lb.
A4.9
hence,
~o
= 19420
lb.
hence,
ZF~
find the distance (d) take moments about c.g. at airplane, hence,
ill
= 29800
 9000
R1
=0
c.g.
19420
x 24  42700
d = 0
hence,
d = 10.9 In.
~
The velocity at end ot catapult track can be found trom. the rof towtng equation
V_Voll~2as
Landing velocity Vo
60 M.P.H. = 88 tt/sec.
Subt: hence stopping distance s = 34.4 ft. Example Problem 2 An airplane equipped w1th float is catapulted into the air tram a Navy Cruiser as illustrated in F1g, A4.l6. The catapulting torce P gives the airplane a constant horizontal acceleration 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.
or
v=
82 tt/sec. = 56 M.P.H.
Example Problem 3 Assume that the transport aIrplane as illustrated 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 vetoo1ty 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 reac 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 _ _
cg:Ma
<
A
78"
t,
R,
Fig. A4.16
Solution: The airplane is being decelerated horizontally hence the inertia torce through the airplane e.g. acts toward the tront ot the airplane. Since the braking torce is given we can solVe tor the deceleration factor by the equilibrium equation,
l:Fx
~85"j
Solution: The torces will be determined just atter the beginning of the catapUlt run, where the car velocity 1s small, and thus the lift on the airplane wing and the airplane drag can be neglected. Horizontal inertia force acting toward the airplane tail equals,
= 35000
II H g
 Ma
henes,
Max = 35000 a
or
whence
X=
35000
ZFx = 900  P
To find
2:1:
R~
27000
a, hence P
= 26100
lb.
o 
take
~oments
about point A,
v hence,
s
Vo = 2 a x s
I25' = 2 (11.27) s
=0
= 695
ft.
A4.10
GENERAL
LOADS
ON
AIRCRAFT
= AirPl~e
7.7
Lift
= 100,000
Re =
x 21
35000 x
9 +
38
Ra
o
111800  4100 14000
~ple
ZF
= 47000  100,000
R1 = 53000 lb.
= 0
Example Problem 4 The airplane in Fig. A4.lB weighs 14,000 Ib It is flying horizontally at a velocity ot 500 M.P.H. (732 ft/sec) when the pilot pulls it upward into a curved path with a radius ot curvature of 2500 tt. Assume the engine thrust and airplane drag equal, opposite and colinear with each other (not shown on Fig. A4.l8).
Flnd: (a)
(b)
(c)
Acceleration ot airplane in Z direction Wing Llft (L) and Tall (T) forces Airplane Load tactor.
Problem 5 Assume the airplane as used in example problem 4 is in the same attitude as used in that example proo tem. Now the airplane is turther maneuvered by the pilot suddenly pushing the control stick forward so as to give the airplane a pitching acceleration of 4 rad/sec ". (a) Find the inertia forces and the tail load T, assuming the lift force on the wing does not change. (b) Find the torces on the jet engine which weighs 1500 lb. and whose c.g. location is shown ln Flg. A4.19. Assume moment of inertia I y (pitching) of the airplane equals 300,000 lb. sec a in.
93700 lb.
,
 D c :. .
210"
Engine Thrust
J
14000
..
L.
Solution: 
t"'so'""",
w
Fig. A4.18
Fig. A4.19
Solution: 
V' 732 Acceleration aa = r 2500 = 214.5 tt/sac a or 214.5/32.2 = 6.67g (uj>'IaI"d). The inertia force normal to the flight path and acting down equals
Ma 6 67g = 93700 lb.  (14000) g z 
Fig. A4.19 shows a tree body at the airplane With the lift and inertia forces as found in Problem 4. The additional inertia torce due to the angular accelerat10n a = 4 rad/sec a equals,
Ia=300000x4
y
Placing this torce on the airplane through the e.g. promotes static equilibril~, hence to find tail load T takes moments about wing aerodynamic center (c.p.)
ZMc.p hence  (14000 + 93700) 8 + 210 T = 0 T = 4100 lb. (down)
which acts clockwise or counter to the direction at angular acceleration. The airplane is now in static equilibrium and to tind the tail load T take moments about airplane e.g.
ZMe.g. = 1,200,000  111800 x B  218 T T = 1400 lb.
=  4100
14000  93700 + L
o
14000
+
1400  Mas = 0
L = 111800 lb.
A4.1l
hence,
Find:
(a) The inertia forces on the air
plane. The e.g. of the engine is 50 inches aft of the airplane e.g. as shown in Fig. A4.19. The force on the engine will be its own weight at 1500 lb., and the inertia forces due to a~ and
~.
whose weight is 180 lb. and Whose location 1s shown in Fig. A4.20.
IPilot c. g.
\ 1372"1
C .I:...._7' ~
T::.
Fig. A4.20
319000 lb.
'''Y
oopooc!l=J::=,,,,_
= 12908
lb. (down)
Note if the engine had been forward of the airplane e.g., the inertia force of 778 lb. would act upward instead ot downward. In calculating the inertia force on a certain airplane item due to angular acceleration,the equation F = Mra assumes that the particular item had negligible mass moment at inertia about its own centroldal Y axis. In the case of a large item this centroidal mass moment of inertia may be appreciable and should be .included in the I y of airplane. Then to find the inertia torce tor such an item the equation F = Mra. Should be ncct r iec to be
F=(I e.g. ~)/r
The wing l1tt will be neglected in this example problem. The inertia forces on the airplane are rcrces Max and I"'.a~ and the couple I c.g. u , To find I"'A take, x
lF
x = 100,000 Max
 l1ax
=a
or hence,
= 100,000
lb.
ax =
To find
Ma~
100,000 = 11
eOO,OOO) 100,000
19
take,
Where hence
ZFs
= 300,000
 100,000  Ma a =
r = distance or ot item.
I
a~
moment of inertia at item about e.g. airplane e.g. equals 10 + Mr il where 1 0 is mass moment of inertia Of item about its own eentroidal Yaxis. F = inertia force in lbs. no~l to radius i . Example Problem 6 Fig. A4.20 shows a large transport airplane whose gross weight Is 100,000 Lb, The airplane pitChing mass ~oment of inertia I y = 40,000,000 lb. sec il in. The airplane is making a level landing with nose wheel slightly otf ground. The reaction on the rear wheels is 319,000 lb. inclined at such an angle to give a drag component of leo,Ooo lb. and a vertical component of 300,000 lb.
= mass
a iS
= 200,000 =
11
200,000) _ ( 100,000 g  2g t
To find the inertia couple I a, take moments about airplane e.g., e.g.
ZM
c.g.
=
e.g.
0.=0
c,g,
a. = 37,200,000 lb.
0.93 ran/sec>.
A4.12
x = 372"
c.g. of pilot
iL _
/'
= 0.93
are for deSign loadS, which in general are 1.5 times the applied loads. It would not be correct to saythat the Wing deflections under the applied 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 considerable portion of the wing would be stressed beyonu the elastic limit of the mater!al or into c s g , air the plastic range where the st i rrnass modulus rs plane
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 !las
= e~O)
19
= 180
The inertia torce due to the angular acceleration 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
F
12 x 40 x 0.93
17 lb.
= 163
lb.
Fig. A4.21
= 379
lb.
conSiuerably less than the modulus of elasticity, 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 Vibration of str~cture. On high speed aircraft, air gusts, flight maneuvers and landing reactions are applied quite rapidly and thus can be classed as dynamic loads. Therefore when these d~iC loads strike a ~lexible (nonrigid) 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.
The example problems of Art. A4.12 assume that tne airplane is a rigid body (eur.rers no structural der ormat tcn) On the basis of this assumption the applied loads on the airplane 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 847 airplane. The maximum upward and downward deflections shown
A4.13
Up until World ~ar II practically all airplanes were assumed as rigid bodies for structural 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 character 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 transport airplane does not have to be designed tor Sharp maneuvers producing high airplane accelerations 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 pUllUp ~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
for applying the load on the wing when striking the air gust. NACA Technical Note 2424 reports the flight test results on a twinengine 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 pUllUp maneuvers were run to give similar airplane normal accelerations. The wing had a natural frequency of 3.8 cps and the airplane speed 'HaS 250 M.P.H. Two of the conclUSions 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 pullups for all measuring 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 dynamic effect at air gusts on a large wing as determined by Bisplinghotr. The results in these tigures show that dynamic effects tend to conSiderably increase wing forces on same portions of the wing and decrease it on other portions.
000
I
Fig. A4.23
Comparative Shear
t5
:; 400 200
<
, 1\
,
<,
Dtstribution
Dynamic Analysis
   Rigid Airplane
Analysts
o
30
.2
.4
.6
.a
i'..
1.0 Fig. A4.24
~ 3 <; 2
1
/ ' ,
<,
./
1
,
1 1.5 2
Fig. A4.22
,
,
c
~ .;
e 20
10
Comparative Bending
.5
I'
.2
Moment Distribution
I"
<,
The author estimates the natural frequency of the 0.:.3 Wing to be around 10 to 15 cycles per secor.d, thus a loading time of 1 second against a tDne of 1/10 or 1/15 for halt a wing deflection cycle indicates that dynamic overstress should not be appreCiable. In general, it can be said that cynamic overstress under maneuvering loads en transport airplanes is not as great as from other conditions such as air gusts or landing . Dynamic Effect of Air Gusts. The higher the air gust velocity and the higher the airplane velocity, the less the time
,
o
3
.6
.6
i
l.0
Fig. A4.25
,
I
g.1
~
2
a
2 0
.
~ c~~~:a~~v;o~~~~i!:Ill
I
I
.. 1
ii
General Data:Wing Span "" 189 ft. GroS8 wt, "" 184000 lb. Airl.a.ne Vel. :z260 mpb.
Report on an Investigation on stresses in Aircraft Struc. tures under Dynamic Loading. M.l. T. Publication.
A4.14
GENERAL
LOADS
ON
AIRCRAFT
W: 15000 lb.
It has also been four.d ~hat landing loads 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 such loads applied to wings of large span produce 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.
(
e c.g.,
120'9; , Rl
65"
I
,,
Fig. A4.26
The advent of the turbojet and the rocket type engines has opened up a range of possible airplane airspeeds hardly dreamed of only a few years ago, and already transsonic and supersonic speed airplanes are a cammon development. From. an aerodynamtc standpoint such speeds have dictated a thin airfoil section which has thus promoted a high density wtng. Thus tor airplanes With appreciable Wing spans like Military bombers and near future jet commercial transports, which usually carry large concentrated 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 distribution or the wing structure be known. Since these factors are not known when the structural design of a wing is started, the general procedure in des tgn would be to first base the design on the assumption that the wing is a rigid body plus correction factors based on past deSign experience or available research information to approx1mately take care of the influence of the elastic wing on the airplane aerodynam1C characteristics and the build up dynwmic inertia forces. ~lth the wing thus initially deSigned by this procedure, it then can be checked by a complete dynamiC analysiS and modif~ed as the results dictate and then recalculated for the modified elastic Wing. This procedure is now practical because at the availability of high speed computors.
A4. 15 PROBLEMS.
90"
I
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 airplane 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?
Fig. A4. 27 1000 lb.
30" / '1
20"~ L
25000 lb.
W: 50000
r'oo" rr
__ c. g,
__
'LJ'
0"
g. of Crew Member
(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 forward acceleration of 3.25g. The gross ~eight of the airplane 1s 15,000 lb. (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 launching time t? (2). Assume the airplane of Pig. A4.26 is landing at 75 M.P.H. on a runway and brakes are applied to the rear wheels equal to 04 of the vertical rear wheel reaction. ~hat is the hori
(4). The jetplane in Fig. A4.28 is diving at a speed at 600 X.P.H. when ~ilot starts a 8g pUllout. Weight of airplane is 16,000 lb. Assume that engine thrust and total airplane drag are equal, opposite and colinear. (a) Find radius of flight path at start of pUllout. (b) Find inertia force in Z direction. (c) Find 11ft L and tail load T.
T
;,
~
.l
L t'. "
<i
Fig. A4.28
CHAPTER A;
AS.l Introduction.
In general, a structural ~ember that suploads perpendicular to its longltudlonal axis is ~6ferred to as a beam. The structure of aircraft provides excellent examples of beam units, such as the wing and fuselage. Very seldom do bending forces act alone on a ~jor aircraft structural unit, but are accompanied by axial and tors1onal forces. However, the bending forces and the resulting beam stresses due to bending of the beam are usually of primary importance in the design ot the beam structure.
~orts
30"
.1
25
Fig. AS. 1 10
.115"'
7S lb.
1
:i
N
~
Fig. AS. 2
" >1
a
75
g
>
. I OO . P
Fig. AS. 3
75
A beam can be considered as SUbjected to known applied loads and unknown supporting reactions. If the distribution of the applied known loads to the supporting reactions can be determined trom the conditions of static equilibrium alone, namely, the summation of forces and ~oments equal zero, then the beam is considered as a statically d~terminate beam. However, if the distribution of the known applied loads to the supporting beam reactions is influenced 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.
100 ... 1
tclM
:i
::;
H
I.
C .....
d
75
~
I.
Flg. AS. S
I OO ' P
75
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 ConsiderIng ZV = 0 In Fig. A5.3: ZV = 75  100 =  25 lb.
     
(1)
A given beam is subjected ta a certain applied known loading. The beam reactions to hold the beam in static equilibrium are then calculated by the necessary equations of static equilibrium, namely: ZV = 0, or the algebraic summation af all vertical forces equal zero.
ZH
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 aa 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 aa,
= 0, or the algebraic summation of all horizontal forces equal zero. moments equal zero.
ZM,
= 75
x 15 + 100 x 5
= 625In.lb.(2)
or an unbalanced moment of  625 tends to rotate the portion of the beam about section aa. A counteracting reststing ~oment M 625 must exist an section aa 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
unbalanced force 1 which means that the resisting moment M i in the fo~ of a couple, as shown in Fig. AS.5, or M i = Cd or Td and T must equal C to make ZH O. The tendency of the loads and reactions acting on a beam to shear or move one portion of a beam up or down relative to the adjacent portion ot the beam is called the External Vertical Shear, or commonly referred to as the beam Vertleal Shear and is represented by the te~ V. From equation (1), the Vertical Shear at any section at a beam can be defined as the algebraic sum of all the forces and reactions acting to one side of the section at which the shear is desired. It the portion of the beam to the lett of the section tends to move up relative to the right portion, the Sign of the Vertical Shear is taken as positive shear and negative if the tendency is opposite. Or in 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 negative tor down on the left and up on the right. From equation (2), the Bending Moment at any section at a beam can be detined as the algebraic sum of the moments of all the forces acting to either side of the section about the section. It this bending moment tends to produce compreSSion (shortening) ot the upper fibers and tension (stretching) ot the lower' fibers' or the beam, the bending moment 1s classed as a positive bending moment, and negative tor the reverse condition.
AS.4 Shear. and Moment Diagrams. In aircrart design, a large proportion
.,[ 5
B,'
B'I .!
I' 1 1"4'
+ 5'~5'
2 '3 RA
L.... 3'
= 1110 lb.
Fig. A5.6
Calculations for Shear Diagram: We start at ~he lett end of the beam. ConSidering a section just to the right of the 500 lb. load, or section 11, and ~onsldering the portion to the left of the section, t~e Vertical Shear at 11 = ZV =_SOO (negative, down on left.)
+
610 lb.
_ 500 lb.
~
_ 500 lb.  390 lb.  390 lb.
.,~
610 lb.
~b
3001b,
~
ZV  
~O
900 in. lb.
sao,
ot the beams are tapered in depth and section, and also carry a variable distributed load. ThUS, to design or check the various sections of such beams, it is necessary to have a complete piCture as to the value ot the vertical shear and bending moment at all sections along the beam. It these values are plotted as ordinates trom a base line, the resulting curves are referred to as Shear and Moment diagrams. A tew example Shear and Moment diagrams will be plotted, to retresh the' students kncw.l edge regarding these
diagrams.
Next, consider section 33, just to RA' ZV 500 + 1110 left side at section).
of
===
=610
(pOSitive, up on
ZV 500 + 1110 = 610 (same as at section 33). Section SS, to right of 1000 load: ZV 500 + 1110  1000 390 (down
on lett).
=
Example Problem 1. Draw a shear and bending moment diagram tor the beam shown in Fig. AS.6. Neglect the weight at the beam. In general, the tirst step is to determine the reactions. To find RB, take moments about point A. ZMA =  4 x 500 + 1000 x 5 + 300 x 13  lORS = a hence RB 690 lb. ZV 500 + RA  1000  300 + 690 = 0 hence RA 1110 lb.
Check this shear at section S5 by using the portion of the beam to the right of 55 as a free body. ZV 300 + 690 = 390, which checks (Sign of shear is minus, because ZV is up on right). Section 66, use the portion to right as a free body:
=
=
= 300
ANALYSIS
AND DESIGN OF
A5.3
V x = 270  lOX, and hence, the shear de_ creases at aconstant rate of 10 Ib./in. frem 270 at A to 180 at C. The vertical shear at section D, just to the right of load is, VD = ZVleft = 270  10 x 9  120 = 60 up, or positive.
Calculation of the Moment Diagram. Start at section 11, and consicer the forces to the left only:
ZM
= 500
=  500
x 0
=0
A
Since sections 22 and 33 are only a differential distance apart, assume a section just above RA and consider the forces on the left side only:
ZM
x 4
Ij j
9" 1'20hP
c+n
j
w=10Ib./in.
'3: I I I
j
I l
1 1
i
I
=  2000
Fig. AS.9
'n,.,._Vcs180 lb. V =60 lb.
RB=210
moment, because of tension in the top fibers). Consider the section under the 1000 in. lb. load: ZM to left =  500 x 9 + 1110 x 5 = 1050 in. lb. (positive moment, compressing the top fibers). Check by considering the forces to the right: ZM right = 300 x S  690 x 5 =  1050 in. lb. Next, consider a section over RB:
ZM right
Fig. A5.l0
A;~
Fig.A5.~.~ ~
Moment Diagram
~ 15"
16"
Shear Diagram 210 lb.
= 300
x 3
= 900
The vertical shear be~Neen points D and B, when x is the distance of any section between D and B trom A:
= 300
x 0
=0
VDB
=270
 120  lOX          
(1)
Fig. AS.S shows the plotted values. 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 clockwise. If obtained fro~ the torces to the right, it Is positive, when counterclockwise. 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
At point BJ x = 36: 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 equating equation (1) to zero and solving for x as follows: o ~ 270  120  lOx, or x ~ 1 = 15" frem
00 15
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 find when shear is zero. Fig. AS.IO shows the plctted shear diagram.
A.
1
120 x
+
9 
36RB =
O.
210
1
RA ; O.
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:
Moment Diagram: At section A just to the ri~~t of RA the bending moment, considering the the left, is zero, sinCe the arm of RA The bending moment at any section A and C, at a distance x from the lett RA, is,
reaction
(2 )
A5.4
The equation for the bending moment between D and B (x greater than 9) 1s
Mx = RAX
 P (xs ) 
wx
2""
(3 )
= 1080
150 x 
5x~
   
(4 )
Thus, the area at the shear diagram between any two points equals ::le Change in ~ending Jloment between these ~NO points. 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 270 + 180 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 triangle 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.
The general expression for the bending moment on the beam at example problem 2 is from equa t Ion (3):
Mx
Example Problem 3. Fig. A5.12 illustrates a landing gear oleo strut ADEO braced jy struts BD and CEo A landing ground load of 15000 lb. 1s applied th:cugh the wheel axle as shown. Let 1 t be required to find the axial load in all ~embers and the shear and bending moment dia~ for the oleo strut.
Therefore, the value or x that will :nake l1x a maximUm or minimum may be found trom the equation
vcr
15.0
He......e HA
But, observation or this equation indicates that the term Ra  P  wx is the shear tor the section at a distance x tram the lett reaction. Therefore, where the shear is zero, the bending moment is maximum. Thus, the shear diagram which shows where the shear is zero is a convenient medium tor locating the points of maximum bending moment.
AS.6 Relatian Between Shear and Bending Moment.
.rrx.o
B
, I";
j5.77iVB
HB
T
j
I
D
10"
i
~<f'
Fig. AS. 12
Resistance at Point
:II
~~~.at~~~~~t~~al
~
t
I
I
I
16"
I I
15000 lb.
I I
16"
V,
since the right hand portion of equation (5) is equal to the shear. Hence, dM = Vdx   (6 )
SOLUTION: 
"0"
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 V c = O. hence, Vc ~ 21550 lb. The axial load in member CE therefore equals
FLIGHT
VEHICLE STRUCTURES
AS.'
= 6660
To find HA take moments about point D, ... 10 HA = O. hence, HA
ZMD
Bracket _____
.'
Bracket
1 00 lb.
~[;,""''l500
E ,:,
II
11550
= 13340
lb.
=0,
Fig. AS. 14
hence, VB = 24550 lb. The axial load in member ED therefore equals 24550/cos 30 = 28360 lb. (compression). The reaction H B therefore equals 28360 x sin 30
14180 lb.
SOLlJrION: 
Fig. A5.13 shows the oleo strut as a free body with the reactions at A, D and E as calculated. Fig. A5.l3 also shows the axial load, vertical shear and bending moment diagrams. The bending moments due to applied loads without regard to bending'deformation of the beam are usually referred to as the prtmary bending moments. If a member carries axial loads additional bending moments will be produced due to the axial loads ttmes the lateral deflection of the beam, and these oe, :ing moments are usually referred to as seC)ndary bending aomerrts , (Arts. A2330 covers t.he calculation of secondary moments).
13340
Calculations of reactions at A and B: To find VB take moments about paint A, :MA 500 x 7  500 x 6 ... 1000 x 20 ... 1000 x sin 45 x 10 + 1000 cos 45 x 2  22 VB = O.
hence, VB 999.3 lb. (up).
=
zv = 999.3
To r Ind VA take ZV = 0,  1000  1000 sin 45 Q 500 + VA hence, VA = 1207.8 lb. (up). To r inc HB take ZH = 0, ZH =  500 ... 1000 cos 45  H B 0, hence
= O.
HB 207.l.
6660
ic= 11550
16
r.
1
133~O
!L:K:J50
I I
1O '
1
14180
~ wfl0//@~
13000 lb.
24550
With the exception of the 1000 lb. load at 45, all loads are applied to brackets which in turn are fastened to the beam. Therefore the next step is to find the reaction of the loaded brackets at the beam centerline support points. The load at E and the reaction H g at B will be also referred to beam centerline. Fig. A5.I5 (a,b,c,d) show the cantilever brackets as free bodies. The reactions at the base of these cantilevers will be determined. These reactions reversed will then be the applied loads to the beam at points C, D, F and E.
F7/?/?/'?//hi'T,;==:z:z:J~~~:am ~:84b
7500 lb.
2'j
50 ,.
~~'Moment Diagram
Fig. AS. 13 120000 133440 in. lb.
Bending
!I
8"
,orl
8"
j)
L MlY'40ci~P"'500
Fig. b
L~~'O'.1.
~~.,
EFig. d
000
,,<;:)
,,:.
,'.
F1g. AS.15
Example problem 4. Fig. A5.l4 shows a beam loaded with both transv~rse and longltudional loads. This beam loading is typical of interior beams in the airplane 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.
For bracket at C, to find He take ZH = 0, or obviously H C ~ 500 lb. In like ~er use Zv = a to tind V c = 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.'
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. Fig. A5.16
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 moment diagrams under the beam loading of Fig.
A5.l6.
500 3000
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 externally 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'
707.1
4000 0
1207.8
'R:l
Fig. AS.16
'e '
10"
t,
I Shear Dia.
a l
Axial Load Dia.
Fig. AS.20
Fig.A'.lS
~
d
707. S,,lb",,. _
b l
L_ _
'"_I__ ...:I=20=7.~l a
,
707.1
'"
~"II'
==~O. 7
[I5788.7
Shear Dia.
999.3'
Bending
A couple is two equal and opposite forces not in the same straight lir.e. Let it be assumed that the 10 in. lb. couple is made up of torces equal to 100 lb. each and an a~ Jetween them at 0.1 inch as illustrated in Fig. A5.21.
4285
3000
Moment Dia.
Fig. AS. 19
207.1
'1
1~ ~O 1"
e
lIb.
11
Fig. AS.21
U 99
Shear Dia.
The shear diagram Is determined in the same manner as explained betore. The applied external 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 desired 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 obtain 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 aoplied couple moments between C and E but not inclUding 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 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 distance 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.
The bending moment at Bright will equal that at Eleft plus the couple moment at E or 3578 + 707
~oments
The student should realize that When couple are applied to a beam it is POSSible to
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
A5.7
~ ~\.Q
I' \.
r"+5"f5"~
r     15"l R ..",100 lb. Rll""lOO Due to il, + 11500 in. lb. Due to ====:::=::=4000
.;z '" ,0 ~
P.",10*"
w; 10
Ib/ in.
R lI"'4.8
1t
t t
r2''t 10"
* *~
Due to P
R~",62
Due to Rl
p.
Due to Pli
500
........Fig. A5.22

Fig. AS. 23
support at the right end. The final bending moment curve for the true given beam then equals the sum of these separate diagrams as illustrated in the figures. STATIC MOMENT CURVES IN SOLVING STATICALLY
INDErERMlNATS STRUCTURES
The usual procedure in solving a statically structure 1s to first make the structure statically determinate by removing the necessary redundant or unknown reactions and then calculating the deflection of this assumed statically determinate structure as one step in 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
Indete~1nate
at B and thus leaVing only 3 unknown elements of the reaction at B. Fig. A5.25 shows the bending moment curves for each load acting separately on this cantilever fr~~e. Fig. A5.26 shows the tr~e bending moment as the summation of the various moment curves of Fig. A5.25. As another solution ot this fixed ended :rarne, one could assume the statically determinate modification as a frame pinned at A and pinned With rollers at B as illustrated in Fig. A5.27. This assumed stricture is statically ceterainet;e because there Pa ::rIO are only 3 unknown elements, namely the magnitude and di!!. =10 rection ot the reaction at A and the magnitude ot the re P..",10 action at B. For convenience the reaction at A is Pin~HA"'20 B resolved into two magnitudes as H and V components. The ""A"'10 B::r20 reactions VA' H A and VB can Fig. AS. 27 then be found by statics and the results are shown on Fig. A5.27. Fig. A5.28 shows the bending moment diagram on this frame due to each load or reaction acting separately, starting at A and going clockwise to B. Fig. A5.29 shows the true bending moment diagram as the summation at the separate diagrams.
=::;::::::I50Due to P a 60  60 o ?''"' .....1J.e v t>" Due to V ..J~I00
y
=::::J
50
Fig. A5. 28
'i
..1
6
'" .I
j~nu"~
Free
60  50 90
\60
~"r_,,'O
30 in.
180
lb.
FIg. AS. 25
A
Fig. AS. 29
100
...L
5"
STRAIGHT BEA11S
paints A and B are unknown in magnitude, direction and location, or each reaction has 3 unknown elements or a total at 6 unknowns for the two reactions. With 3 static equilibrium equations available, the structure is statically indeterminate to the third degree. Fig. A5.25 illustrates one ~er in which the structure can be made statically dete~nate, by freeing the end A to make a bent cantilever beam fixed
Aircraft structures present many beams which carry a varying distributed load. tlinlmum structural weight is at paramount importance in aircraft structural design thus it is desirable to have the complete bending moment diagram for the structure so that each portion of the structure can be proportioned efficiently. To decrease the amount of numerical work required in obtaining the complete shear
~.
'( i:t" ,
A5.8
~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 different ~~ltudes. Fig. A5.31 shows a free body
H1
Q
~
M1
,/
I
,
r
V,
! 
I
do;
MlC
(1)
Lr~v "OM"
(2)
Fig. AS. 32
Fig. AS. 33
Fig. AS.31
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 M 1 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 station 1 plus the algebraic sum of all forces F lying between stations 1 and 2. Again considering Fig. A5.31, the bending moment M" at station 2 can be written, MOl = M l + Vld + Fl_"a, or stated in words, the bending moment tl" at station 2 is equal to the bending moment Ml at a previous station 1, plus the Shear V at the previous station. 1 t1.mes the arm c , the distance 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.
Then from Fig. A5.33 we can 'Nrite for the resultant torces and moment at point (2) at station 2: ;
Having the resultant forces and moments for a given point on a given station, it is usually necessarJ in :inding beam stresses to resolve the forces ~nto components no~~l and parallel to the beam crosssection and also tr~sfer their location to a point on the neutral axis of the beam crosssection. For examp ;e Fig. A5.34 shews the resultant
Many structural beams carry both longitudional 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 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 is desirable to minimize the amount of numerical work in obtaining the complete shear and bending moment values. Fig. A5.32 shows a curved beam loaded with a number of different vertical loads P and horizontal loads Q. rig. A5.33 shows the beam portion 12 cut out as a treo bOdy. Hl represents the resultant horizontal force at statton I and equals the algebraic summation ot all the Q forces to the left of station 1. Vl represents the resultant vertical torce at station 1 and equals the sum of all F :orces to lett of station 1, and M equals the bending l moment about point (1) on station I due to the moments of all forces lying to the left of point 1.
sP W
v
" "J6
~~
s
Mo.:Ml wNe
Fig. AS. 36
Fig. AS. 34
Fig. AS. 3S
forces and moment at point 1 of a oe~ crosssection. They can be resolved into a normal terce N and a shear for S ~lus a moment M as l shown in Fig. A5.35 where,
A5.9
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 normally 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
Fig. AS.34
ural axis coincides with the tube centerline, or axis 11. Fig. AS.35 shows the load P being moved to the paint (0) on the tube axis 11, 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 Shearsand 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 spanwise variation C L and CO' Any particular type of airplane 1s designed to carry out a certain Job or duty and to do that Job requires a certain ~lmum airplane
velocity With the maneuvering Ilmited to certain maxtmum accelerations. These limiting accelerations are usually specified with reference to the X Y Z axes ot the airplane. Since the directions of the lift and drag forces change with angle of attack it is simpler and convenient in stress analysis to resolve all forces with reference 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 analysis 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 acting 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 The above unit load conditions are for conditions ot acceleration in translation of the
A5.I0
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 covered in a separate course in most aeronautical curricula after a student has had initial courses in aerodynamics and structures. To illustrate 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.
aer~mic
center
a.c"
xL.
I Fig.A5.38
"? Ref.
96"
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 according 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 M x or moment 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
ratio or the spanwise variation of the lift coeff~cient C L in terms ot a uniform dIstrIbution 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 location of the ~Pz load which is at the centroId ot a troplzcidal dIstrIbuted load whose end Yalues are given in Column (4). In determining these centroid locations it is convenient to use Table A3.4 of Chapter A3. The values of the shear V z and the moment M x 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 M x = 9884 in Column (12) for statIon 220 Bquals 2436, the M x moment at the preViOUS station in Column (12) plus 4908 in Column (10) which Is the shear at the preVIous statIon (230) tImes the distance 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.
Fig. A5. 39
A5.11
,
C L/_
11
12
ra
15
15
CL
II
Ratio,
Assumed Unity.
g.
7.2
7.28
605
~;:
Q,,"''S
'40
~g
.u >~
~g
~S...;
~.
"
~L
49.0
"
48.54 49.63
>.0
<2.
48.1 512.7 810.0
o
2.49 242.7 2.49 490.8 1213 4908 15052 27020 40087 53 68077 83782 99015 115579
o
60s
618
o
1787 1767 3613 7553 14093 21483 29743 38913 49013
50113
230 220
205
7."
7.89
1846
1003.5 1813.5
11. 09
8.34 8.85 9.09 3580. 9 9.33 10.06 5545.5 7480 7840 8200 8570 8940 9310 9660 10000 1 00 247683 338705 445920 570060 702949 863174 1042534 1241744 1539 4 10.31 2672.5
'940
65'iO 7390 8260 9170 10100
8 9
15.00 1.00 1 .00 1.00 15.00 15.00 15.00 1.00 1 .00 1007.0 1055.5 1104. 1153.4 1202.7 1252.2 1300.9 1349.7 1 1920.0 908.4
190
175
7 " 7.43
.4J
4538.5
160 14'
100
67.13 70.39
7.43
7. 43
10.55
6601.0 10. 80 11.04 7705.3 11. 28 11.54 11.80 12.04 10061 132880 150915 169700 189210
9
"0
72.00
73. 7. "'3
11100 11480 13270 14480 12.52 15600 13.01 16930 13.49 18200 14.05 26350
100
1.0 1.0
1.0 1.0
as
70
7. i3
12. 28 12.76 13.27 13.71 14.40 7.43 11313
ss o
.4J
12614 7.42 9. 139 4
1.0
1.0
.00
98.00
. 0
ie ,
96.00
10.00
1.0
96.00
14.40 17760 318808 19200 1875630 14.40 27848 Sum : 17760 Cheeks Total Limit Load. Assumed all HaJ! Wing.
l~en the time comes to deSign the structural makeup of a crosssection to withstand these applied shears and moments, the structural deSigner 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 another 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 landing 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 airplane bOdy 1s approached.
I _ ..,(
A5.12
The airplane body ~ust be designed to withstand 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 agencies. 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 requirements for each condition are given in the government specifications for both ~illtary and commercial airplanes. LandpLanes : Level Level Three Three landing. landing with side load. point landing. point landing with ground loop. Nose over or turn over condition. Arresting. (Usually for only Navy Carrier based airplanes) .
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 airplane 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 engineer 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 presented 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 airplane 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 procedure 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 factors. 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 ereLem, It happens to be the body of an actua... airplane and the wing used in the preViOUS example problem Nas the wing that Nent with the airplane.
($'1\"1~=r=~~~I1
hrust Line
II
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.
~ c
~
_.   
~,.L
I~==::::=== 2:T72;,,=~:.....===:j1
= o
sc.
_ .jBo!_.!..xi....   
G. (groSS wt. )
11'2"
Fig. AS. 40
:1
'
I
I
AS 13
Table A5.2 gives the Weight and Balance for the total airplane. This table 1s usually fo~ulated by the ~ei~~t and Balance Section of the engineering department and it is necessary to have this information before the airplane load analysis can be made.
esti~te
WEIGHT AND BALANCE OF AIRPLANE LESS WING GROUP AND INSTALLATIONS IN AND ON WING Ref. Axes: 1a up, z. equal &rill frnm thrust line. III distance from z. Ref . .Ax1s 5" !orwarQ at. (x propeller. .. L8 a.ft.
i
Reference
TABLE A5.2
AIRPLANE WEICHT AND BALANCE
..
~
'.m
1
N=.
PawerpiaDt grOlq) Fuselage groull Tail group Surface COlltro18 'Electrical Iylltem T1I.i1 Whee! gro~ Furn1ab.iDgs Wela:ht
w."",
y
Boriz.
"". \
Vert. (Z) arms Ill_red from thrUst line (+ II up) BorU:. (Xj arms meuu:red froID Z ms 5" (+ is aft) !ot'T.l.rd ~rop. ~
;
Vert. Mom,
~
1100 '50
19 113.5
,
er
Bori:l:.
Vert.
Vert.
""'m.
~
.um
""'m.
~
30900
39700 31550 10800 7930 10700 25520 22600 1t197oo
o
1
o
3690 1190
11'
'.m
No.
1
N=.
w. I
wt.
1100 35'
it
Boriz.
.um
19
Bortz.
Moment
~
Vert.
.um
(Z)
" ~
~
.,
e e
130
136 d55
"
"
38'
121
308 118
191
19
10
"
10
'"
10 11
12
z
s
t
e a
10
11
ra
13
Power Plant Fueelaie Gro~ WiD( Group Tail Grou~ SDrface Controls Electrical System Chu;ais Front Tall Wheel Gro~ FuruilItw1ga Radio Weipt emllty PUot Student Fuel Syllt.em Grosl weight 1
20900
311700 72750 31550 10800 7930 16450 10700
n
1
'50 11'
3150
, , ,
18
l<
1:ui00
'50
~
0
"'"'"
Plio<
em~ty
'"
,
Z
'30
 350 1100
"""~,
200 200
151
33200
19800 19700
es
"Z ". ~
~
SO,
'00
52
35
11'
4300
'"
20' '00
151
88
..
'50" 22600 I
"8900 '0200 [
37~00
"
_ 1190
'04'
Emllty ~
 '"
With \18efUlload. i
""
169700 78.7"
Z.~.2..00"
BG.O" 2.32"
"'"
2~~~
." ."
~40280
"50
of C, G. locations: .CalcuJ.ation GroSI wt. x z 376500/4300 z 98,S" a,ft of Re! .Mia z. z 40280/4300 9.4" below line i
thrust
SOLUTION:
WEIGHT At'IITJ BALANCE OF BODY ITEJ"!S.
~~IGHT
DISTRIBUTION.
Table A5.3 gives the weight and balance calculations for all items attached to fuselage or carried in t~e fuselage, except the wing and items attached to the wing as the front landing gear and the fuel. In order to obtain a close approXimation to the true shears and ~oments on the fuselage due to the dead weight inertia loadS, it is necessary t~ distribute the weights of the various items as given in Table A5.3. Fig. A5.42 shows a side view of the airplane with the center of graVity locations of the wei~~t items of Table A5.3 indicated by the (+) Signs. In the various deSign conditions, the direction of the weight inertia forces changes, thus it is convenient and customary to resolve the inertia forCes into X and Z components. ThUS, in Fig. AS.43, the weights as given in Table A5.3 are assumed acting in the Z direction through their (c.g.) locations. The loads as shown would not give a true picture as to the shears and ~oments along the fuselage, thus these loads should je distributed in a ~nner which should si~ulate t~e actual weight distribution. In ~ost weight and balance reports, the weight items are broken down into considerable ~ore detail than that shown in Table A5.3, which ~kes the wel&~t distribution ~ore evident. The ~erson maKi~g the
weight distribution should study the inboard prOfile drawing of the airplane which shows the general arrangement of all the installations and equtpment , Furthermore, he should study the overall structural arrangement as to its possible influence on fuselage weight distribution. The whole process involves considerable common sense if a good apprOXimation to the wei~~t d1s trlbutlon is to be obtained. Fortunately the large dead weight loadS, such as the power plant, tail, etc. are definitely located, thus small errors in the distribution of the minor distributed weights does not change the overall shears and moments an appreciable amount. In order to obtain reasonable accuracy, the fuselage or body is diVided into a series of stations or sections. In Fig. A5.42, the Sections selected are deSigned as stations which represent the distance from the Z reference axis. The general problem 1s to distribute the concentrated loads as shown in Fig. A5.43 into an equivalent system acting at the 7arious fuselage station pOints. Obviously, if a weight item trom Table A5.3, represents a concentrated load such as a pilot, stUdent, radio, etc., the weight can 08 distributed to adjacent station points inversely as the distance of the weight (c.g.) fram these adjacent stations. However, for a weight item such as the fuselage struct~e (Item 2 of Table AS.3) whOSe e.g. location causes it to tall be~Neen stations 80 and !20 of Fig. A5.43, it would obViously be wrong ~o distribute this weight only to the two adjaCent stations since the wel~~t of 350= 1s for ~he entire fuselage. This weight item of 350~ shou1 4 thus Je 11stributed to all station points. The contrOlling require~ent on this distribution is tr~t the moment of the distributed system about the reference axes must equal the moment of the original Height about the s~e ~~es. Fig. A5.44
A5.14
BEAMS _. SHEAR AND MOMENTS WEIGHT DISTRIBUTION TO FUSELAGE STATIONS STATION 170 200
so
80
1
120
I I
220
260
290
I
315
0 11
50
80
120
170
200
230
250
290
315
I I
I I
I
I
110
! I
Z.
011
I I
50
80
120
1 1
170
200
I
I
I
1 ~ I ....
I I
~
I
T
N
220
260
1
290
I
011
I
50
80
120
170
I
200
230
260
290
315
I
I
1r1ig....
'"
I
I
~;
i
'" ~
~M
l~ <:cl
'"
I
,,I
011
SO
80
120
1
170
I I
011
I
I
so
I
80
120
170
I
200
230
260
290
315
I
I
i
I
I
6511
j 
i! j
30# 100
" rpl
I
I I
I
90 90 .r80~ 3:0# 2':34#
I' I
i
I
Fig. AS.48.
SO
80
120
I
I
I
I
I I
170
I I
311*
200
220
260
290
I
215
I
I
7~#
I I
100
I
I
I '
o 11
50
80
120
I
170
200
230
260
290
315
I
11191
I
I
I
I
I'
I
I
I
',9' 'tf: 12e
409*
I
Fig. AS.45.
I
I
11i93
I
Fig. AS.49.
A5 15
shows how the dead weight of 350i: was distributed to the various station points considering the weights to be acting [n the Z direction. Table A5.4 shows the results of this station point weIght distribution rcr the weight item.s of Table A5.3. The values in the horizontal rows opposite each weight item shows the distribution to the various fuselage stations. The summation at the weights [n each vert tca.L column at each station point as given in the third horizontal row rzcn the bottom or the table gives the final station point weight. These weights are shown in Fig. A5.45 for weights acting in the Z direction. The moment at each total_station load about the Z axiS [s given in the second horizontal row tram the bottom of Table A5.4. The summation at the moments in this row aust equal the total WX moments of Table A5.3 or 219700"*. This check [s shown in the last vertical column of Table A5.4. The distributed system must also be distributed [n the Z or vertical direction [n such a
aanner as to have the same resultant e.g. 10cation as the original we Ight system which [s illustrated in Fig. A5.46. Fig. A5.47 illustrates how the fuselage weight distributed system as shown in F[g. A5.44 [s distributed in the vertical direction at the various station points so that the moment at this systern about the X axis [5 equal to that of the original fuselage weight or 350#. For convenience, these distributed fuselage weights can be transferred to the X axis plus a moment as shown in Fig. A5.48. Table AS.4 shows the vertical distribution at the various items at the various station points. The bottom horizontal row gives the moment about the X axes at the loads at each station point, which equals the indiVidual loads times their Z distances. The summation at the values in this horizontal row must equal the total wz aoment ot Table AS.3. This check is shown at the bottom of the last vertical column. Fig. A5.49 shows the results as gtven in Table AS.4 for the weight distribution in the X direction.
~
TABLE
,.~ i
~o.
N=,
I we, I
I
., .
u
so
urlace contro
Furni
I ElllCtrical s !item
raUl)
io
ra
Radio U i Pilot
", . , ."
..
i
,
1100
8741
,
m"
II
c I no o
. ,
,
, ,
.
,
. " , , '", ,
I
i
I
,
ire
aoc
i
i
."" ,
, ,
,
"" ,
"" ,
I
I
.'" ,
,
UOO
to
.
1.67
12$ i
,
""'.m ToW
"", "",
rs rc
..
,
, ,
,
I
I
3881 \9390
2.47 ,
."
"
te
,
11 uml
' S'l6O
, "
,
.10 I
,
au
52800
1
5. J
6.45
10 1.0
1181 19.1
491001
isace
w%..
J: z; :
= dlKtaAce
"I
a
0(
758 ,
Z;
682 1
1650 I
reterence U111 _mell III 5" fonra.rd 01 propeller. disbnce aeeve or beloW tllrust l.1D.8 or x wa.
statiol1 lro.ll;l
""
"""I ic
134200 I
'"
''''
 ""
" "'''
.",
21W100
...
.OO
5920 I
Since there are many flight and landing conditions. considerable tl.:ne can be saved i t a \mit analysis is made fer the fuselage shears. axial and bending forces. The design values in general then folloW as a sunnnation of the values in the \m[t anaLys t s t imes a ~roper multiplicat~on factor. The loads on the fuselage [n general consists ot tan loads. eng:!.ne loads, Wing reactions, landing gear reactions If attached to ruse rage and tnert i.a fore es due to the airplane acceleration which ~y be due to botn trrans Ia. tional and an~ular acceleration of the airplane. For s 1.:nplic1t.y, these loads can be resolved into cGIt::onents parallel to the Z and X axes. To illustrate the unit analysis procedure. a unit analYSiS tor our example problem will be carried out for the following unt t conditions:
acceleration or load factor tn Z acting up. acceleration or load factor [n X acting rorwarc . tail load normal to X axis acting
Unit analyses are also usua.Ll.y carried out for engine thrust and engine torque , side load on ~ll and angular acceleration, but to keep the example calculations from becoming too lengthY only the above 3 unit conditions will be carried out in detail. The others will be discussed in detail [n later paragra phs
Solution for Unit Load Factor in Z Direction.
A:) 16
act~ng
BEAMS

A5.4 or Fig. A5 . .:i5. T:'18 'Nin~ ~s attached to t:le rusejage at scations 73 ::ond 116 as s::;'~wn en ?~,. AS.50. The fittings at these points are assumed as designed to cause all the :.rag or reaction in the X ::ilrectlon to be taken of f er.t traj.y at the front fitting on station ';'"3. To place the r ....selage In equt I i ortum, t'::e wing reaction '1'1111 be calculated:
ZFx
lJ"f"S ta t ; or.
I
I
I
1
FUSELAGE SHEARS AND MOMENTS FOR ONE LOAD FACTOR IN Z DiRECTION (Dead Weight Acting Down)
sta, No.
31' :1 290
ZF z RR:
.  2555
Moment ..:oM Load or V : shear i ex : Dist. i M' Reaction : L w ! between I = VLlx (lbs.) stations (in. lbs.) (in. lbs.) w
i
I
(;, )
0 22 0 118
0
0
22 25
I
,
0
,
RR 0


 (3 )


I
30 30
550
550
, 260
230
21
0 10 0 76
 4200 I  4750 I I
I
RF
==
1780 lb., RR
= 775
161 171
171 247
lb.
200
 4830
 5130
9580
30 30
I
14710
1
Table A5.5 gives the calculations for the fuselage shears and bending moments 3.t the various station points.
sta.O 11 I I
'0
I I I
I
 
I
I , , , ,
60
120
I I I
170
I
200
230
I
260
'
290 315
I I
I
I
I I
I
I I
i
311
I
I
76
I ,
I
~
893
: 19
. 
388
307
409
 + .
21
VN
I
'
247 558
;0
 7410
I
4 36J
i
,
22i20
,
27900 I 50020
I
I
I
I
* .I
22
0 77' 1 I
3668
53888 I ,
I
.I
i:i'
,
,
I I
 6912  3493
29463
60800
1
I
64~90
I
7
I
"" I
23 '0
1281  893
39  693
34827
II 
348~
I
Fig. AS.51 shows the panel point dead weight distribution for loads acting in the X direction and att, as taken from Table A5.4 or Fig. A5.40. To place the fuselage in equllibrlum the wi::lg reactions at points (A) and (8 ) will be calculated.
(3) Up on left and down on right side of a section IS pcemve shear. (8) Tension in upper fuselage portion is negative bending moment. (I) refers to aft side of station.  refers to forward side of station. M = M at previous station in col. 6 plus 1M in col. 5
lF x
= 2555
 RH
= 0,
hence
RH
lI1A ,. 2555 x 17
I
I
19
I
I
I
I
i
I
I I
I
173
200
230
260
I
290 31'
I I I
I,'
;",
758
I i
i
I
I
I
r
662
lF z
hence
I
I
I
'\
Rp1l47.8
iRR"'1147.8
FLIGHT VEHICLE STRUCTURES TABLE A5.7 FUSELAGE SHEARS &; MOMENTS FOR UNIT HORIZONTAL TAn.. LOAD IN Z DIRECTION (Load Acting Down) 1 2 Load or Reaction w lbs. 3 4
~X
A5 17
I
,
...
So.
a ,
w" :
mX ,load. ~
'""
:w::iP":.~w" v z w:
i
,
I
s
I i
Dlr. !ZDlr.
'wa.l LGad
SlIear
I.::.M>lvt:,,:
1
sac  2029
!
t
1 315
asc : I
2l\(J i
" 11~ I
22.!
2li
'1 cI
,<0 ,<0
161
" aa
c
u
I
0
1
o:
220
!
as
30
22411 1
o'
I
j ,I
c
I
1
V , shear
: Z w
Ibs. 0 0
22g I
sta.
,  ~M ,, V~X
I ,
,
, ,
, ,
"
,I
,.,
230 : i 200
o:
10 i
01
'"
 37~
10
I
30 30 30
2029 c 1_ , 2303
315 290
272.5
"
u1
c c
1 1
,
c !
!
1
=1
0 0
25
.! 
rn
" !
,I
311 : 40IIj o
170 :
oI
I
~
120 _
115
'I
...
358
'" '"
1 49~
1
i
1
o
c
:!
:i
: I
Q
1147.6!
'I
01
,
60 .
3071
I
1
73
2555'1147.al 3861
l6Sg
L L
1
 I
0 0
12.5
,
! I
,
~_23131
su
4
0 100
17.5
i
,
ae
41320
Q
 I
100 100
,
100 100
30 30 100 100 30
 rse !
'
1214 1281
t281  a93  893
1147. a i
50 ::
u ! ,
. 
o1
, o I
_4343~ ~: I
~:~:; i ,
43462 i 27 !
a035
"
a9~ I
"I
I
1
o, I
I~
"
I I
1 :' 17 x (2555) : 4343:5 refers to alt aide 01 staJ:ion 1 _ refers to forward side of station 1 Col. te pJ,u lor ten!lion in fllHlage. i ICoL S) M M iU ~rl!Vlou. station In CoL. 9. ll1u .; Ml of col. 6 :'M2 ill 001. a
~I c : "1 "
c
~l...
I
i
3000 5000
,..
36 7
I
I !
i
400
80 73
I , ,
13521 2629
 2629 . 2629
a
0 0 0
375.6 0
The fuselage shears and :noments will be C8mputed for a unit tail load of 100 lb. on the tail acting in the Z direction, with balanc Ing reactions at tna wing attachment pOints. The center of ;Jressure en the hor:!.zontal tail is at station 277.5. Fig. AS.52 shows the fuselage loading. To r ind wing reactions at (A) and (B) : ZMA = 100 x (277.5  73)  43 RR = 0, hence RR = 475.6'" (up)
23 0 0 39
50 11
I i
0 0
0 J
CALCtruTION
MOMENTS ANIl
FLIGilT CONDITION.
,,"0., J
200
I
230
260
127~.
1
290 315
! I ,
I
I
1
I
I
I
!
I ,
,
I
1 1 1
I 1'100*
43" 
R F =375. a
RR=475.6 srA.116
Using the results in Tables AS.5, AS.6, an d the applied shears and ~oments for a g:!.'l8 n flight condition follow as a m.tter of pr'cpor'. tion and addition. To illustrate, the applied values ~or one flight condition 'Hill be given. It will be assumed that the aerodynamic calculations for this air?lane for the (H,A.A.) hfgh angle of attack condition ;ave the followtng resu Lt s , which the student will aave to accept wt nout knowrenge of how they were cotameo
. 7
Fig. A5.52
STA.73
A5.I8
Applied load factor in Z direction Applied load factor in X direction Applied tail load
~
 6.0
down
1.333
~:t.
:
~c.G.
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 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.)
c
Fig. A5. 53

...
,
,
...
,
0 0
c
c
0
c 0 c c
5220 3740  3140  5Q85
co
c c
29. 3 29.3
186.8 16 .1:1
5Q1O "
"" . ,
,
~60
e
0
200 :
110
'" .i
116 80 :
73 I
S510 11620 1_124640 ; 118ZO 1_12898~ I : 173 0 _ 86855 I 6850 17J2.0 286650 1 :290  6850 ' 17770 _ 6850 1 17770 I 2910 _313690 I 1290 47250 mo ,_314640 : 1700 H900 0 40 I 1700 '_385704 1_1710
"'"
"
192~
,,,
"
",.,
',:;i~:g: ~~:~
'" '"
"
, .,
,.,
_ 6 :l valu In '01""",,, 3 at Talli. A.5.5. 1.333 '" valu'" ill .00umn 5 01 Tabl .. .405.B. _ 1.10 '" valu" ill .00""",n 3 01 Ubi. A.5.7. eol""",,, (1) oolWIIII (2) olumll (3). 8 '" valUH III .01"""," B af 1':lQI. A.5.5. _ 1. 3;13 :l VlI.1ues In .0IW1111 9 of T:lQI.. '5.6. _ to 10 '" vaiUH ill .01"""," 6 at TaDl. AS.7. _ col""", ... {B) (7) (8). _ 1.333 '" ..uuu In .01 ....... 14) of Tabl. AS.6.
The resistance to these X ~~d Z components of the grcund react~on R is ?rOvided by ~he inertia forces of the airplane in the X ar.d Z directions Tables AS.S and AS.6 show the tuse l.age shears, moments and axial loads for inert~a loads due to one load factor in the Z and X directions respectively. Thus to obtain the fuselage forces for this given landing condition, it is only necessary to mul.t t p.Ly tne va Iues in these two tables by the ?roper factor an~ add Ute results. Thus fuselage forces due to vertical load factor ot 7 would equal 7 t1mes the values in columns (3) of Table AS.S to obtain shear ~.d 7 times column 6 to obtain bendi~g moment. Likewise the farces due to the 2.98 load factor in X direction would equal (2.98) ti~es the values in col~s (4)7 (5) and (9) of Table AS.S to obtain axial loads, shears and bending moments respectively. The final or ~r~e for~es woult be the algebraic sum of these results. Landing with
.~~gular
Acceleration
Fig. AS.53 illustrates the airplane in a level landing condition. The ground reaction is assumed to pass the center at landing gear wheel and c.g. of airplane. The fuselage Shears, moments and axial loads are recurred when the vertical ultimate load factor is 7. (Gross weight = 430ot).
SOLtlrION:
In a level UL~ding condition. i t 1s sometimes specified that the horizontal component ot the ground reaction must be a certain proportion of the vertical componeht, which causes the line of action of the grO~~d reaction ~ in rig. AS.53 to not pass tr~ough t~g e.g. of the airplane, which creates an external pitching moment on ~he airplane. This ~oment is usually balanced by the inertia forces cue to the angular acceleration produced by the ..in. balanced moment about the e.g. The shears and moments on the fuselage cue to this externa; moment could be :aund as ex~la1ned in p~t.
AS.20.
AS.20 Inertia Loads Due to Angular Acceleration.
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.
In some of the flying conditions, it is specified that the a1rplane ~ust be SUbjected to an angular acceleration as well as translational acceleration. ~is angular accele~ation of the airplane produces :nertia forces which must be calculated if the airplane is to be treated as a body in statiC
samet~es
FLIGHT
VEHICLE STRUCTURES
A5.19
equilibrium. In some cases, a tail load due to a gust on the tail is specified which produces a moment about the airplane c.g. which produces ~~ar acceleratio~ of the airplane. In certain landing conditions, the ground forces do not pass through the airplane e.g. thus producing a moment about the e.g. which for stress analysis purposes is balanced by inertia forces. Moment of Inertia of Airplane The calcUlation of the moment of inertia at 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 ...
Fz
='M:y w Xc ly
y
             
(1 )
=My w zc I
              
(2)
= 16097600
,z
" =
= .00621
w Xc
FX
= .00621 w Zc
Z wz
z A Iy
Ix
=Z wy I z =Z wy
...!. wz
...
Z A Ix
+
Z wx
Z A Iz
The last term in each at the above equations represents the moment of inertia of each weight item about its own centroldal axes parallel 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 determining 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
Where Zc and Xc are the z and x distances of wei~~t w to the airplane e.g. Columns No. 9 and 10 of Table AS.9 gives 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 acceleration 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 yawing and rolling moments.
:Z)
I
TABL.E A5.i DJo1.AlCING INERTtA F'ORCES FOR maT 100,000 IN. loB. YOWENT ABOUT Y AXlS TIlROUClll: AIRPLANE C.G. (PrI'CKlKG HOhlEHTJ 1 13j 3 1 4
$
where
!
10
5
i Arm
Arm
z.,
I..
IFlI:
1y g
i.1y
'"
Fual.
290
[4.701 IS. il IU.SO 10.40 142.50 2.1.ta 173.$0 501 :W3.50 19. 10 I 28. 1:r.2. <10.0 315  0.1 3:17.50
,UBI
i
e:
n,sl
5 i ~.nn.s! l<u.e I
 14.3 8:1.$ 1I.:l 1011 \ 1S50 20460 i 30:nO I
JlGU 51100
1.~1I
1.
3.1
2180000
~UOOO
I . '"
i
I
204100 535000
3110000
u"",",
I
,
141.3 1_20.11
hence
F
= Ty 'N
My
710 1  :n I (4l
1M Zl20
313
<17.51
1.50
i~
11091100
133000 4il5ooo
:5.5 1.1
::: l'"~:: I
51.3 U.3
1
~,
where r
1s
I
; C<l1 ........ (2J, (3), 6 C<>lumQ {$)
the weight w to the airplane e.g. It 1s convenient to t~eat the inertia force F as resolved into two components Fx and Fz hence,
T\U<8:Q
Colwall
(6)
I C<l1W1111
~ C<l1W1l11
(ill
(10)
Tabla A:i.3. z., , Z  9.4 1 9H Tabla A:i.3 for x.., X  51.5 I loeat1lla F . ooeu "lie F ,oo51l .. Sc
fram
,I,. :
e.,.
A5.20
11
50
aD ,
,
120 ,
170
, ,
230
III.
Fig. AS.62 shows t~e ~lan f8~ sf ~ ca~:~ 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
O" 10L
r~
\L~~~====:=::::,~to . _ 1:12.0"
Planform
~O"
Fig. AS. 62
1
200"
.....;
(I) Draw the shear, bending moment and ax:al load diagra~s for loaded structures 1~ Figs. 55 to BOc .
100
rIO
jX
r
I
90"
."
AS.63
" s o
'5
1"
200 101
5,0
100
~
1
30" 00 12,..,'
i 10'~soo
I10j
~in.
.l
1.
a:
(55)
I 20'~
400 SO
(56)
}2H
o~i=:I:::J'=C=IIII::t[Il ]lj'kS3 t 1 .L ,0
I
"""'""1:20"1
~1O"f400 (57)
I A
.Cable
60 j 10"
(56a)
01"
~=;r===i:==~' L
r't12"
9"
'\..600
a"
Fig. A5.63 shows plan rora of a cantilever 'Iring. The total distributed air loa~ nc~al to sur~ace is 10000 lb. The relative spanwise ~istriJut1cr. is shown. Take center of ~ressure at 24 percent o~ chord fro~ leading edge. Divi~e wing i~to 10 inch width strips and calculate VZ 1 M x and My, ane plot curves for same.
AI l' ~2
,
500 lb.
IV. F:g. A5,64 shows an externally Jraced monoplane wt ng , Take an average 'Nir.g 11ft 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 calculate anc draw the tron0 and rear beam prl~ry shears and bencing ~c~ent d1a~rams.
100
Ii
M
I
4
2.0"
60c
,e..L
c
F.B.
(II) Draw bending ~oment diagram for structures and loading in Fig. AS.51, abc
40011
Fig. A5. 64
R.B.
'o'!....M'a.
(0) 16"
100
100f
,j
4~~5
(b)
\ 10';
I~O / 100
_ _ ::\450
(c)
iii
116.. l
.L
112""",
Fig. AS. 61
FLIGHT
VEHICLE STRUCTURES
A5.21
deflections will gradually converge and the member will reach a state or equilibrium. These A beamcolumn is a member SUbjected to transverse loads or end moments plus axial loads. secondary bending moments could be found by sucThe transverse loading, or end moments, prOduces cessive steps by the various deflection principles given in Chapter A7. However, for pr i smat i c bending moments WhiCh, in turn, produce lateral 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 above successive step method. For ~embers of the axial load times this lateral deflection. variable moment of inertia, the seCOndary moments Compressive axial loadS tend to increase the will usually have to be found by successive steps. pri~ary transverse bending moments, where as If the end loads P are tension, they will tensile axial loads tend to decrease them. tend to decrease the primarY moments; tnua, in Beamcolumn members are qUite crnmnon in general, the case or axial compression is more airplane structures. For example, the bearnE of important in practical design, since buckling externally braced wing and tail surfaces are and instability enter into the problem. typical examples, the air loads producing transverse beam loads and the struts introducing axial 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 beamcolumn 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 M It 1 and Mg. 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 proremain plane after bending; that stress is proportion and need to be considered in the design portional to strain in both tension and compresat the members. sion. This chapter deals briefly on the theorJ At any point a distance x tram the beam end, of single span beamcolumn aembere , A summary the moment expression is, or equations and deSign tables is included to2  M1 ) wLx wX gether with examples of their use. The informa 11=1'1.+ (118 L xT+ 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 beamcolumn members is considered. For a comM =EI day pleted and comprehensive treatment of beamcoldx 2 therefore, differentiating equaumn theory and derivation of equations, see tion (AS.I) twice with respect to x gives Niles and Newell"Alrp~ane structures~.
A5.23 Introduction A5.24 General Action of a Member Subjected to Combined Axial and Transverse Loads.
d P dxa + EI M = w        
2M
(A5.2)
w
p_
SUbfigure a of Fig. AS.6S shows a member subjected to transverse loads wend axial compressive loads P. The transverse loads W produce a pr~ary bending distribution on the member as shown in Ffg. b. This bending will produce a tr~~sverse deflection c~ve as illustrated in Fig. c. The end loads P now produce an additional second~J bending moment due to the end load P tUnes the deflection 6 , or the bending moment diagram of Fig. d. This first seconda~J moment distribution produces the additional lateral deflection curve ot Fig. e and the end load P will again produce further bending moments due to this defiection. If the axial load is not too large, these successive
i //
w f
Fig. A5. 65
_pial
p,;1, .....,..,
Ml
I=:j
I,M2
r I
rnng Load
0<2<:;'
P
,0.,0
rt:. . . "JM~e",:
. Jd l
__L='lp (el
\ I o.ef~tl.,d .... t.,l.t s.."'><'d~1 H."'erot
Fig. A5: 66
Vp
P EI
ji"'
a:xa
F!1 = w
,
~'~~~
AS.22
BEAM  COLUMNS
solution of this differential equation gives. x x (A5.3) I1=C:l. Sin] + c , cos] + wjll where C1. and ell are constants of integration and sin x and cos x are ~~e limits at an infinite series of variable x. when x
C
1.
y=~
sin
sin j
= L,

I1
= Ma ,
J
therefore:
!"h  wj.lll
tan L
(A5.7a) The slope at the elastic curve at any paint is given by the first derivative of equation
(A5. 7al
i =
_ Mil  wja
sin L
J
(M:l.  wj.lll) cos L
+>'HX 
sin L
J
and Ga
= M:l. = I1:l.
Wjll
wj.lll.
Then,
In investigating other transverse loadings ror a single span carrying axial compression, it is round that the expression tor bending moment in the span always takes the torm:
M = C 1. S i n
x J
t(w) is a ter::n which does not include the load P or the end moments l11. and M:a. The 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 '!: span which are frequently encountered in airplane dX  0 = T cos J  J J structureS. The Table also gives equations for whence the pcfrrt of maximum bending moment; and its magnitUde. L Table AS.II 15 a table or slnes, cosines, x C D.  D1 cos tan  = " = ,"J IA5.5) and tangents tor L/.1 in radians which is more j c, L convenient to use than the usual type at trigoD1 sin 1 nometric tables. This table is based on values given in Appendix ~ of Air Corps Information The value or x must tall within x = 0 to x = L, Circular #493. The A difference have been added otherwise I11 or Ma is the maximum value. The value of the maximum span moment can be to facilitate rapid use of the tables. For single span beams, the critical value found by substituting the value trom equation of L/j is fi; that is, it the axial compressive (A5.51 in (A5.41, which gf vee load is such that the term L/j = fi, the center = D__ + wJ'          (A5.5) region of the beam will tend to deflect until the cos x combined stresses equal the tailing stress of the material. J The moment M at any point x along the span can also be written:
sin L
where
axial
n.ax
I1=D:l. [(tan
J J J 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 apis customary to locate the point of maximum span ply to a beamcolumn, because the sum of the bending moment and its value before investigatbending momenta due to the transverse loads and the axial loads acting separately are not the ing other S~l points, the value at tan !m is same as the moments when they act simultaneously. J 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
(AS.7)
AS. 27 Moments for Combinations of the Various Load Systems as Given in Table AS. I, Margins of Safety. Accuracy of Calculations.
_:.
ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES AS.23
Table AS. I
M C 1 sin..!. + Cz
j
cos~+ f(w)
C,
.~2 i
sin L
C2
f(w)
P"""'1r x _
M max
Ml
2j
P+~~j====I~P I ~L .
w ;,/ in.
T
wj2 (cos Ll)
sin L
x: .5 L
T
DZ  D cos L/i em I J
where
Tan..!.: D2  D1 cos..!:.
wj2
j
J
P+'!=~ o
I""'P
eD1 sin.,h
j
Mmax'=~twi2 cos x
x <: a, .Wjsinb
sin L
T
 Wj
o
sin.!.
)
o
o
x o a, + Wj sin a
1" tan L
T
Triangular Loading. No
End Moments
p__
..l.
tX~
(NOTE A) To obtain Max1mum Moment, compute moment at 3 or. " points in span. Draw a smooth curve thru plotted results.
w./p'"iJTi
'Fx
U ro::::r
f L
x < a,
 m
cos b sin L
T
)
(See Note A)
X)
a.  m cos..!..
tan L
m cos.!.. J
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
L jj
ill Sill
Rad1allS
~jj
t:.
Sill Lj
j
Cos Lj
f::. Cos Lj
Tall Lj
f::. Tall Lj
3.00 3.01
3.02 3.03 3.04 3.05 3.06 3.07 3.08 3.09 3.10
0.14112 0.00991
0.13121 .00 92
O. 0127
0.99262 0.99378
O. 2129
0.00993
10 6
0.01013
O. 1136
0.00994 0.10142
0.09146 0.08150 0.00997 0.07153 0.00998 0.06155 0.00999 0.05156 0.00997 0.04159 0.01000 0.03159 0.00999 0.02160 0.01000
0.09185 0.00086
0.01008 0. 817 0.00077 0.01006 0.0717 0.00066 0.01004
0.99667
0.99744 0.99810
0.08167
O. 0057
0.99867 0.00046 0.99913 0.00037 0.04162 0.0318 0.00027 0.99977 0.05184
O~
O.
0.01001
003
0.02180
.0 16
0.00160 0.01001
.
0.00160
0.01000
1.00000
0.01001 0.00841
O.
00.01841 00.02841
11.99984 9
0.99960 0.99926
0.01000
i
I
0.04845 0.00064
0.01003
0.01004
0.06852
0.06836 0.00997
0.07833
0.99766
0.00073
0.08829
0.09825
0.00093
0.99516 0.00103 0.99413
0.00995 0.10820
65
AS.28
BEAM  COLUMNS
of the bea~, or 4420 x .75 = 3315"# positive bethese values in the general expression tor M as cause it procuces compression in the top tibers. given at the top of the Table. The moment at (2) due to the cantilever overhang In a beamcolumn member, the bending momequals (20 + 10) 36 x 16 = 8640"#. Fl;. A5.:58 ents do not vary directly as the load 1s i~crea 2 ed. Thus, the student should realize that margins or safety based on direct ?roportton of mom shows the beam ?ortion between points (1) and (2) as a free body. ents to loads are incorrect and lie on the unFrom Art. AS.Z5, we have the follOWing presafe side. cise equations for a beam carrying a transverse It is recommended that four significant uniform distributed load with end compressive figures be used in computations, making use ot loads. the socalled precise equations, since the results in many cases involve small differences x L between large numbers. D~  Dl cos J         (A) tan J
AS. 28 EDmple Problems
D.. sin L
and
Example Problem #1 Fig. A5.S? illustrates a typical upper, outer panel Wing beam of a biplane. Let it be required to determine the maximum negative bending moment between paints (1) and (2), generally referred to as the maximum span moment. To obtain the true bending moments on the beam, the axial beam load as well as the end moments at (1) and (2) are necessary since they influence the deflection of the beam. Solutlon:To obtain the horizontal component Th or the lift strut load, we take moments about the
!'l
max
= ...!2>..cos x
J
w.:;
          (E)
3315"#
8640"#
= 4420# compression = 10 in~ given and assumed constant throughout the span.
=1/l.3XIoeXIO = V4420
v294I
54.23
P ';"~I~~;b~g;~g~~~~~t'~~B~~.J...
.A.
w= 2Q.IP/in.
lOll/in
Wj
"= 20 x 2941
Wood Sea
~... <;e
D1.
DII
M 1
wja
'\..!~'\.
,0' 'S:'~
lSOQ.IP=Reaction
= M..
 wjll
55505
50160
J = 54.23 = 1. 844
From Table A5.II sin
100
J = ....
26981
Fig. AS.67
D 1 sin
...... w= 2Of/in.
J
65156 53441
1.2192
Fig. AS. 68
Hence, x = .88383 x 54.23 = 48 ft , which equals the distance trom the lett end at the beam to the point at maximum span moment.
 1500 hence
Th = 442Qj1
x 100
70.75 Th = 0 Hence
= ..lL:
cos x
r ,
+
J
~
55505  :63419 58820 28,700"#
The axial compressive load induced by the lift strut at pOint (2) then equals  4420#. Taking ZH = 0 tor the load system or Fig. AS.S? gives P =  Th = 4420#. The end moment on the beam at (1) equals the end load timeB the eccentricity or the hinge :ram the neutral axis
idea as to the magnitUde at the moment, that is, the ~oment due times the lateral beam deflecbending moment at a point 48~
,,;; . r
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"# I1= (M .. !11. cos L/j) sin x/j +M cos x/j Thus the secondary bending moment equals sin L/j 1.  28700 + 18765 =  9935"# which 1s a large percentage of the primary moment. The transverse 20, OQQ!...lF==={~2='=!=2="=0=6=3=8=,,="=1=T=U=be=J~02~~000t deflection of the beam at the po1nt of max. span . L: 41. 762 ,( moment then equals  9935 := 2.25 inches upward.
 4420
Bending Moment at any Point Along Span Fig. A5.70 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, I1 = Masin x/j M:= Dl.~tan ~ . sin })+ cos + wja (Ref. Eq.
(j)
11
A5.?)
sin L!j
.99605
cos x =  .08867
tan
.s =1.2192 =value
j
Hence,
/'
E:x:ample Problem #2 Fig. A5.59 shows a simplified landing gear structure carrying a vertical load of 12000# on the axle. M.ember ABC is continuous thru B and ptnned at C. Let it be recurred to determine the bending moment at the midpoint or member BC and its lateral deflection due to the 12000# vertical design load.
This compares with a prtmarY moment of 36000/2 ~ 18000"#. The deflection at the midpoint or BC := 26066  18000 = .403 In.
20000
~ 1
~D
26"
"0:";;:86:::2"."'"
~ Ie
"112 21/2083.
Round
Tabla A5.I)
Steel
Tub"
Fig. A5.69
1 I
0
BiE,,(
AxlJ L
A
2000*
The equations as presented in this chapter assume that E is constant or in other words the stresses are Within the elastic range. In aircrart structural destgn the applied or limit loads must be taken without suffering permanent deformation, hence E 1s constant under such loads. However the aircraft structure must take the deSign loads which equal the limit loads times a factor at safety (usually 1.5) without failure. In many cases structural failure will occur under stresses in the plastiC range where the material stiffness Is less and not constant. A good approximation for ~~ effective modulus E' Is obtained as :ollows:(1) Compute Fc = PIA for the given number. (2) With this value of Fc enter the basic column curve diagram for the given material (for end fixity C = 1) and find value of L'/p corresponding to the stress Fc'
"""'"3
Solution:Solving for reactions at C by staticS, we obtaIn :he axial load in Be = 20000. The bending momen~ at 8 due to 3" eccen~ricity of the 'wheel load = 3 x 12000 := 36000"#. Fig. A5.70 shows a r~ee body of portion BC of member ABC. From Table A5.1 x '() M = C~ sin x j +, v~ cos } + ~ w
~~~~~
AS. 30
BEAM
~
COLUMNS
(4) Determine the bending ~oment at the centerline of the beamc01~~s shown in Fig. A5.74. Assume ~I 64 J O O O, 000 lb. in. sq.
(4) Then i
>
(E~ry/'
Basic column curves for various ~terials are given in another chapter of this book.
AS.30 Problems
1!
If
800<1
Fig. A5.75
500<
6000*
1
Ll.
H155"'! 35"

50"
(5) For the beamcolumn in Fig. A5.75 calculate the bending moment of the centerline of the member. Assume E = 1,300,300 psi. and I = 10 in~
Flg. AS. 71
Fig. AS.72
(1) Fig. A5.71 shows a 11/2  .065 steel tube subjected to both end and lateral loads. Determine the max~um bending moment on the tube. Campare the result with the bending moment due to the side load only. E = 29 X 10 8 ps i . I of tube = .075 in. 40 Compute lateral deflection at point of maximum bending moment. (2) The beam column member in Fig. A5.72 is made of 24ST aluminum. alloy. Calculate and plot a curve of the bending moments on the member. Also plot bending moment due to lateral loads only. E = 10.3 x 10" psi. I:::: 5.0 in ....
500.
50"
1
;,000# 18000"IF
Fig. AS. 76
(6) For the beamcolumn loading in Fig. A5.76, calculate bending moment at center point ot beam. Take S 1,200,000 psi and
10
In~
(3) Determine the maximum. bending moment for the wood wing beam and loading of Fig. A5.73. I of beam section = 17 in. E 1.3 x
10".
pr*
V
I t f !
The secondary moments in a partic~lar member due to beamcolumn~action also effect or influence the deflections in adjacent members of a continuous structure. This rat~er involved problem can be handled quite simply and rapidly by the moment distribution method as explained and illustrated in Arts. All.l2 to 15 of cnap'ter All.
s.
"000..
.
1~
10001
20007
I I I tIl I
tt
I tl l!
0'"
100001
200"     .  . "
Flg. AS. 74
CHAPTER A6
A6.1 Introduction.
Problems involving torsion are common in aircraft structures. The metal covered airplane
wing and fuselage are basically thinwalled
tUbular structures and are subjected to large torsional moments in certain flight and landing conditions. The various mechanical control systems in an airplane often contain units or various cross~sectlonal shapes which are sub1ected to torsional forces under operating conditions, hence a knowledge of torsional stresses and distortions of members Is necessary in aircraft structural design.
A6.2. Torsion of Members with Circular Cross Sections.
sumed that any radial line undergoes angular displacement only, or OB remains stral&~t when moving to OB' The unit shearing strain in a distance L equals,
Let G equal modUlus at rigidity at the and let ~ equal the unit Shearing stress at the extreme fiber on the crosssection.
~terial
Hence,
= E G = r 9 G _ _ _ _ _ _ _ _ _ _ (1)
L
The following conditions are assumed in the derivation or the equations for torsional stresses and distortions: (1) The ~ember is a circular, solid or hollow round cvi incer . (2) Sections remain circular after application of torque. Diameters remain straight atter twisting of section. (4') Material rs homogeneous, isotropic and elastic. (5) The applied loads lie in a plane or planes perpendicular to the axis of the shaft or cylinder. p
(3)
L
In Fig. A6.2 let ~p equal the unit Shearing stress on a circular strip dA at a distance p from O. 'fhen
"0 ,
='t"
__ prQG = p 9 G Lr L
The ~oment of the shearing stress on the circular strip dA about 0 the axis of the bar 1s equal to,
dM .. r pdA
= 'L
o'ig}dA
\nt.
B'
,
\
L
ao"QdA

J ,,
For equilibrium, the internal resisting moment equals the external t orai onaj moment T, and since GQ/L is a constant, we can write,
.T = Mint. =
GQ L Jr
0
Fig. A6.1
p ctA = T
GQJ
(2)
Fig. A6.2
Fig. A6.l shows a straight cylindrical bar subjected to two equal but OPPOSite torsional couples. The bar twists and each section is SUbjected to a Shearing stress. Assuming the left end as stationary relative to the rest of the bar a line AB on the surface will move to AB' under these shearing stresses and this rotation at any section will be proportional to the distance from the fixed support. It 15 asA6.1
Where J = polar moment of inertia of the sr~ft cross section and equals twice the moment of inertia about a diameter. From equation (l) GQ = ~ L r r J (3 ) Hence, T = r
or
also from equatfon (2), solv:ng for the
Q
(4 )
~Nlst
9,
= GJ TL
(9 1s measured in radians).
(5 )
C,.' vb
AB.2
TORSION
?;,obl~m
2.
done by a tWlst~n~ 8cu~le T 1~ ~ovl~g tn~ough an a~~~lar jis~l~c?~~~t :s ~qual to th3 ~roduct of the ~~gr.it~de :f the cJu~le and the angu lnr c i s cracement in rao Lans . I~ tne angular dt s pLaceasnt t s one rev jut t on , ::18 ',II'::'!:. done equals 2 n T. I: T is ex:ressed in lr.chpounds and N Is the~~ul~r ,elccity in revol~ tions per ~inute, then the horsepower tr3ns~itted by a rot~ting shaft ~y be written,
~ork
The
H, P. = 396000
2 n N T
( 0,
(11/4  .C48 in Size) suppo:,ted on :~ee ~rack3~s ~~d ~lt~ tte c:ntr:l ~od ~:tt:~g attac~~c ~c ~he tor~ue t~be above the center SUDJOrt bracket. F:nd the 8~xi~lli~ torsional shear1ns stress in :he tu~e i t the air lo~d on the ailer:n is as indicated in Fig. A6.4, and also compute the an~le sf twist of tute between horn section anc end of aileron.
hi~ge
.
i
rr
1\ :
I
where 396000 represents inch ~o~~ds of work of one horsepower fer ~ne min~te. Equation (6; may be wrttten:
T = H.P. x 396000 2nN
EXAMPLE PRCBLillS.
! I\ ::: ,I I,
63025 H.P.
N
(7)
5"
L
I
24"
24" 15"
Fig. AB.4
Problem l. Fig. A6.3 shows a conventional central sticktorque tube operating unit. For a side load of 150 Ibs. on stick grip, determine the shearing stress on aileron torque tube and the angle of ~Nist between points A and B.
SOLUTION:
airload on the sur~ace ~er.ds to rotate ~round the tC~Gue t~be. ju~ novement is prevented or crea~ed 2y a control rod attached to the torque t~be over the center supporting oracket. The total load 0n a 5t~ip of aileron cne lucre wide = 40(15 x 1/:'44) ~ 4.16 lb. Let w equal intensity of loading per inch of aileron span at tha.,leading edge point cr the aileron surface. (see ~ressure diagram in
~he
the ailer8n
SOLUTION:
Torsional moment on tube AB due to slde stick force of 150# ~ 150 x 26 = 3900 in. lb. The resistance to this torque is provided by the
150*
Fig. A6.41.
Then 3.
= 4.16
lb.
= 0.463
ilt2.'    1
...... EI~v. Control Wire
forwarG of the centerx 3 = 1.389 lb. and P a the load on aileron portion att of hinge line ~ 0.463 x 0.5 x 12 = 2.778 lb .
line at torque tube
= 0.463
~ Bearing
.J:
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 round tube equals 0.1368 in 4
The torsional ~oment per r~nlng inch or torque tube: =  1.389 x 1.5 + 2.7~8 x 4 = 9.0 in. lb. Hence, the maximum torque, which occurs at the center of the aileron. equals 9.0 x 29 = 261 in. lb.
_ Tr
1:(:nax.) 
ps t .
= 0.21
radians
0.06678 in 4
. )
or 12 degrees.
~he
torque
'n
AB.3
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
3800000
0.06678
sections.
The for~las derived in Art. A6.2 cannot be for nonc1rcular Shapes since the assumptions 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 distance from the axis of rotation the shear stress
~sed
Fig. AB.5
EP=+
Ie
:;;;::_~~I
"': max.
Fig. AB. B
Fig. AS.7
The \Shape of a~watD~a crosssection of a noncircular crosssection in torsion is needed in the analysiS by the theory of elasticity, and as a rasult only a few Shapes stich as rectangles, ;eJ:,lJ2.seS, triangles, etc., have been solved by the theoretical approach. However, a close approximation can be made experimentally for a Iznos't any shape of crosssection 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 opening which has the same shape as the crosssection 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 meaSured 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 cresssection of a bar in torsion. The membrane ar~alogy provides the follaNing 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 corresponding ~oint on the crosssection 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 crosssection 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.
~
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 surface as well as along ~he surface and this is ~ot true. The theorj of elasticity shows that the maximum shear stress occurs at the centerline 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 equations for determining the shear stresses and
To illustrate, consider a bar with a rectangular crosssection 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 crosssection by a small uniform press~re. ~qual deflection contour lines for this deflected ~embrane will take the sr~pe as illustrated 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 boundary 1s approaChed. ~lg. A6.Sa shows a section through the contour lines or the ~erlected membrane along the lines 11. 22 and 33 of Fl~.
AS.9. It is obvious that the slopes of the de
From Table A6.l it is ~otised that :or large values 0:' bit, tne values 0:: t he constants 1s 1/3, and t~us for such narrow rectangles, equations (6) and (7) reduce to,
_ _ 3 T
"MAX 
tlected surface along line 11 will be greater than along lines 22 or 33. From this we can conclude that the shear stress at any point on line 11 will be greater than the shear stress tor corresponding pOints on lines 22 and 33. The maxlmu~ slope and therefore the maximum , ,
bt:l
(5 )
_ 3 T g  bt 3 G
(9 )
,
,
,:>
,
,
~.
I __ I
! , ,
._.
\
~
I
,
,
.J
Although equations (8) and (9) have been derived for a narrow rectangular shape, they can be applied to an approx~mate analysis of shapes made up of thin rectangular members such as illustrated in Fig. A6.10. The ~cre generous the fillet or corner radius, the smaller the stress concentration at these junctions and therefore the more accuracy of these approximate formulas. Thus for a section made up of a continuous plate such as illustrated in
Fig. A6.
a
stope
Fig. A6.9
l~l
.LSlope
2~
Fig. A6.9a
3~
Fig. a b can be taken as centerline length for above type of sections
Slope
stress will occur at the ends of line 11. The slope of the detlected membrane will be zero at the center ot the membrane and at the four corners, and thus the shear stress ~t these points will be zero.
A6.6 Torsion at. Open Sections Composed of Thin Plates.
.r:
~ET b. ...
fobl.~~tl.
Fig. A6.10
Members having crosssections made up of narrow or thin rectangular elements are sometimes used in aircratt structures to carry torsional loads such as the angle, channel, and Tee Shapes. _;"'~'IrFor a bar of rectangular crosssection or width b and thickness t a mathematical elastiCity analysis gives the following equations tor maximum shearing stress and the angle of twist per un1 t length.
Fig. (a ) of Fig. A6,.lO, the width b can be taken as the total lteiigthof the cresssection. For sections such as the tee and H section in Fig. A6.l0, the polar moment of inertia J can be taken as Z bt~/ . Thus for the tee section of Fig. A6.10:
9
T ='=G3 OJ Z bt 3
a b t
(6 )
G (blt~ + bat~)
bl

3 T
radl ana  ~
(7 )
Values or a and
TABLE AS.l
CONSTANTS (lAND ,
bit 1.00 1.5Q 1.'1'S .2.00 2.50 3.00 4 S a 10 I 0 0 c, O.:W8 0.231 O.~O.2~~~O.~o.m~~~WO.3130.m 0.141 0.19 0.214 O. O. 49 O. 3 O. 1 O. .299 0.307 0.313, O. 333
'toa
For the maximum Shearing stress on leg _Tt l _ 3 T t l 3Tt l 'tb l .fb'tT=blt l S T b:lt a "  and tor the ?late bar _Tt a _ 3Tt a
J1
(10)
(
 J   bltl S + b a t a 3        
11)
It t
9
1:'
= t , = t,
3(b
then
G t
3 T
l
+ ba)
            (12)
           (13)
3 T t 1 ( b 1. + 011)
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~
Table A6.3 summarizes the for.nulas for torsional deflection and stress for a few Shapes. These for~ulas are ,based on the assumption that the crosssections are free to warp (no end restraints). Material is homogeneous and stresses are within the elastic range.
A6.8 Torsion of ThinWalled Closed Sections.
= 0.000045
Fig. AG.lla
Fig. A6.l1b
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
Q
The structure of aircraft wings, fuselages and control surfaces are essentially thinwalled 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 thlnwalled cylindrical tube which is under a pure torsional moment. There are no end restraints on the tube or in other words the tube ends and tube crosssections are free to warp out of their plane.
= G~J'
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 efficient torsional members relative to torsional deflection.
TABLE Ae.3
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.). G :: Modulus oJ. R~id1ty. K (in4) Froltl Ta.b1e.
~
Fig. A6.12 Fig. All. 13
~qds
q constant
SO
SOLID SQUARE.
SECTION
SOLID ELLlPITICAL
'tMAX :: ft'iijr (1Lt elida of minOr ana).
'T
Let ~ be the shear force intenSity at point (a) on the crosssection 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 values ~y and qby as shown in Fig. A6.l2. For 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 Since the tube sections are free to warp there can be no longitudional stresses on the tUbe wall. Considering the eqUilibrium of the segnent in the Y direction, qb;" 0, hence Qa y qby and therefore Ga = qb or in other words the shear torce intenSity around the tube wall is constant. The shear stress at any point T = qjt. It the wall thickness t Changes the shear stress
~Fy
(b)
D1
SOLID RECTANGLE.
T
K
tO.141a
CJI 2a.,
~
SOLID TRIANGLE.
ao3[~
3.3e~(1
'tM.\X
= T(3a
 1. 8bl 8a.2t)%
mtdpo1Dt of 10ll(side.
~]
6~
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.
= 0 = Qa~
A6.6
TORSION
shea~
or
~ata
~btb
= constant.
dU
CS =2 4
:5
The prOduct ~ is generally referred to as the shear flow and is given the symbol q. The name shear flow possibly came from the fact that the equation ~ = const~~t, resembles the equation of continuity of fluid flow qS = const~~t where q is the flow velocity and S the tube crosssectional area. We will now take moments of the shear flow q on the tube crosssection about some point (0). In Fig. A6.l3 the force dF on the wall element ds qds , Its arm from the assumed ooment center (0) is h. Thus the moment of dF about (0) is q cs h. However, ds times h 1s twice the area of the shaded triangle in Fig. A6.13. Hence the torsional moment dT of the force on the element ds equals,
=~ G :::;...9:.... Gt'
em : :;
but 4
:=
T" 2A
hence or
U
8A.L'IGt

cs
is the line integral ar8und the periphery of the tube. From Chapter A7 from Castigliano's theorem.
f"I _
'7
dT qnds 2qdA
aT 
au _ ~
r 4A,.T
L~ 15 J t
          
(17)
and thus for the total torque for the entire shear flow around the tube wall equals,
T = fA 2qdA and since q is constant
T = qU
since all values except t are constant. equation (17) can be wTitten,
Q
~ 4A,.IlG
        
(~8)
and since T
=2
(14)
Q=~Gfd:at
L
or
q = 2A
Where Q is angle of twist in rcdlans per lli~it length ot one ir.ch cf tube. For a tube length
(15)
Q=i<~f~s
(20)
where A is the enclosed area of the mean periphery 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
AG.9 Expressi.on for Torsional Moment in Terms of Internal Shear Flow Systems for Multiple Cell Closed Sections.
(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 crosssection and a unit length parallel to the tube axis. Under the Shearing strains the plate element to 1 0\ >0 1 <
Fig. A6".l6 shows the internal shear f Low pattern for a 2cell thlnwalled 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 outSide wall, we know that
(21)
m
T , <Is
1. t
'01
I
I
Fig.A6.14
qds~D:
Fig. AS. 15
t ,
t
\
Fig. AS. 16
n
detorma as illustrated in Fig. AS.l5, that is, the tace aa moves with respect to face 22 a distance O. The torce on edge aa equals q ds and it moves thrOU~~ a distance O.
Choose any ~oment axis such as paint (0). Referring back to Fig. A6.l3, ~e found that the moment of a constant shear force q acting along a wall length ds about a pOint (a} was e4~1 in
_ _ _ _ _ _ _ _......... iII'itMTlllil'i
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.
(0).
Let To moment ot shear flow about point Then from Fig. A6.16,
inch in the web 12 and (qe  q~) ~or web 23. For eqUilibrium, the torsional moment af the internal shear system must equal the external 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~.
~Nist
of a cell
Q=q_!~
2AG
or
(25)
Hence, To 2q,A, + 2q,A,         (23) where A1 = area ot cell (1) and A3 = area of cell (2). Therefore, the moment of the internal shear system of a ~ultlple cell tube carrying 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 t~e outside wall or tr~t cell. (Note: The web mn 18 referred to as an inside wall of either cell)
AS.10 Distribution of Torsional Shear Stresses 1n a MultipleCell ThinWalled Closed Section. Angle of Twist.
2GQ=.s.i~ A j t
ThUS, for each cell of a
ture an expression ~s can be written and equated to the constant value 2OQ. Let a l Q represent a line integral 10, and ala,
tf
~ultiple
cell struc
f ~s
a~Q
tegrals ~s for the other outside wall and interior web portions of the 3cell tube. Let clockwise direction of wall shear stresses in any cell be pOSitive in sign. Now. substituting in Equation (25), we have: cell (1)
a~o, a2~
and
Cell@
j t a
A~ [q~
=209
(26)
=2GQ
Fig. AG.17
 
 
  
 
  
 
(27)
Fig. A6.17 shows in general the internal shear flow pattern on a 3cell 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 reter to it as lying between cells 10; for the outside wall of cell 2, as 2C; and for the web between cells (1) and (2) as 12, etc. q~ = shear load per inch = ~~tl in the outside 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
Equations (24. 26, 27 and 28) are sufficient to 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 2Cell 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 ~Nocell tube (Fig. A6.18). 
AS.S
TORSION
Cell 2
(29)
     
(34)
    (30)
The summation o~ the external and internal reSisting torque must equal zero.
.040" 1 :: 25. 25"
=GTJ
                  (32)
+
where A = A1.
Aa
I
I
.050"
A6.12 &le Problems of. Torsional Stresses in MulUpleCellThinWalled Tubes. Erample 1  Torsional Stresses in Unsymmetrical TwoCellTube. Fig. A6.19
='
, .032"
Fig. A6.19 shows a typical 2cell 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 :
=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
au
=~ .04 = 335
032
Fig. A6.20 shows the resulting shear ,attern. 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 twist of tube as a whole .
= :o:;r 25.25
+ ~ + 25.3
05
=''''35 (
Solution by equating angular ~Nist of each cell. General eQuat~on 2GQ ~ ~s. Clockwise tlow at q is pOSitive. Cell 1 Subt. in general equation
tf
2Gg
105.8 q1.
r
x 1075
qa
 13.33
+ 3.165
~&iwr'
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
=4
+ +
= 89.76
[ 1735
X
] T
X
387.4 3
"2
Q=JT G
G X 89.76
length of cell.
= .001107 T = .001107
Example Problem 2.
X 83450
= 9Z.5t/ln.
Determine the torsional shear stresses in the symmetrical 2 cell section of Fig. A6.21 when subjected to a torque T. Neglect any resistance of stringers in resisting torsional moment.
SOLUTION:
~
Fig. A6.22 shows a thinwalled tubular section composed of three cells. The internal shear flow pattern will be determined in reSisting the external torque at lOo,OOO~t as shown.
Flg.A6.21
T
q.
Calculation of terms Area of cer.fa: A l = 100 Ali A Al + A3 100 200 Line integrals a =
.03"
,.03"
Iq05To'
o:l<XO'i
.031
.03
CeU@
. 03
Fig. ,A6. 22
SOLUTION:
fdtS:a 
1.0
10 + ~ .05 .03
= 866.7
= 39.3
a =
As
100
Line integrals
L~ J t
au
= .~~
= 333.3
a"
au a
.05
10
= 200
333
 20 altO :03
_ 20 a a e  .03 = 667
a;so
=.l2...= .03
.03
ZO
+.lQ.. =
.04
917
r.
= .00254D
_ (37)
.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.
a,.]
A6.10
TORSION
A6.14 Torsional Shear Flow in Multiple Cell Beams by
SUbstituting:
1.3 209 = 39 [629 q, + 200 q,  200 q,J    (38)
Cell (2)
trend in airplane wing struct~al design particularly in high speed airplanes is toward ~ultiple cell arrangement as illustrated in :1g. A6.24, n~.ely a Wing crosssection made up of a relatively large number of cells.
~he
roe =
l~O
c,
3:33 q , 
                 (39)
Cell (3)
the solution by the previous eq~tions becomes Quite labor! ous , The method of successive approximations provides a simple, rapid method :or fl~dlng the shear flow in multiple cells under pure torsion.
EXPLANATION OF
SUCG~SSI'lE COFL~ECTION ~HCD.
SUbstituting:
2i}Q
l~
917
q~J
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. Iceu~ Cell as to make GQJ. = 1. (2) (1) From equation .04 .05 "'J )0" (19) we can 'NTite, I ~s   (40)
Gg=..s..~dS
2A
A ..=89.3
'. . __...l.'~_"/'U
! 5" 1
_ Fig. a
Since practically all cellular aircraft beams have waLl, and web panels of constant thickness for each part~cular unit, the term 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,
p d:
Fig. AB. 23
= ;L :E cel l t
:2 A
(41)
234. H/in.
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.
Fig. A6.23 shows the resulting internal shear flow pattern. ~he angle of twist, if desired, can be found by substituting values of shear flows in any of the equations (38) to (40).
A6.1l

.I
I
0.212 lb./in.
=
.109#
r:. I
,.212
I~
q '" . 109
j
~
(11
I
j
(2)
Fig. b
Fig. c
(2)
= 1.
subThen
and therefore negative), in addition to the shear flow q1. = .212 of cell (1). The negative shear flow qa .109 on web 12 decreases the ~Nist of cell (1) as calculated above with the resulting value tor G91. = 0.875 instead of 1.0 as started with. Thus in order to make 091. = 1 again, we will have to add a constant shear flow q'l to cell (1) which will cancel the negative ~Nist due to q~ acting on web (12). Since we are considering only cell (1) we can compare cell wall strains instead of cell twist since in equation (41), the term 2A is constant. Thus adding a constant shear flow qi to cancel influences at qJ1l on web l?, we can vetta:
=
l.Q.
.05
15.7
.03
78.6
723 :. 0.lOS#/1n.
qi
(z
hence
cell (l)  q,
Fig. c shows the results. Now assume the two cells are joined together with the ~nterior web (12) as a common part at both cells. See Fig. d. The inq w 109 ter i or; web 15 now Subjected to a resultant shear flow at q1.  qJ1l = (.212 .109) .103t/ln. Obviously this change of shear flow on the Fig. d interior web will cause the cell twist to be different tor each cell instead of the same when the cells were considered acting separately. To verify this conclusion the twist measured by the term OQ will be com,uted for each cell.
q1.  q~
. [ .~ +....!2..J_200  842 q3 =
~
.04
.05
Cen (1),
GQ. ' 
2A, Z qt 
L _
.~;]
Thus to make GQ1. equal to 1 we must correct the shear flow in cell (1) by adding a constant shear flow equal to .237 tl~es the shear flow qJ1l in cell (2) which equals .237 x .109 = .0258~/in. Since this shear :low is in te~ of the shear flow qJ1l of the adjacent cell it will be referred to as a correction carrv over shear flow, and will consist of a carryover correction tactor times qJ1l. Thus the carryover factor tram cell (2) to cell (1) may be written as
=
Cen (2),
0.375 C.O.F. (2 to
11
Jf)
(z
'Neb (12)
~)
which equals
cell (1)
~S.3
15.7
.~;J
.237 as found above in substitution in equation (42) Now consider cell (2) in Fig. d. In
Since 91. must eq~l QJ1I if ~he cresssection is not to c~stort from its original shape, it is eVident ~hat the above shear :lows ~e not the true ones w~en the two cells act together as a unit. Now consider cell (1) in Fig. d. In brtng. ing up and attaching cell (2) the common web
bringlr.g up ane attaching cell (1), the common web (21) is subjected to a shear flow of Q1. =  O.212*/ln. (co~~terc1~c~vise as viewed from cell 2 and ~~eretore negative). This additior~l shear ~low c~~nges GQ 3 twist of cell (2) to a relative "al~e of 0.4375 instead Jf 1.0 (see preVious GGJ1I calculations). Therefore to ~ke OQ 3 equal to 1.0 again, a cor~ectlve constant shear flow q~ ~ust be added to ~ell (2) to
AS. 12
TOR S ION
cancel the twist effect of q~ = 0.212 on web (21). Therefore we can write,
q.
hence
(z ~) cell
(2) 
q,
(~) web
= q~
(C.O.F.) 1 to 2
= .0256
(21) =
.227 .00717#/in.
Fig. f shows the result1ng second set ct corrective constant shear flows for each cell. Since our corrective shear flows are q, (~) web (21) q'k '"' .00717 (43) rapidly getting = sffialler, the con(z~) cell (2) {"', . 0139 tinuation of the process dependS on \~ (il (~' SUbstituting in equation (43): the degree of accuracy we wish for ,_ (200) the tinal results. .277 q, .277 x .212 .058711!1n. Fig. f q.  q~ \723 Suppose we a dd one more set of corThus the carryover factor trom cell (1) to cell (2) in terms at q~ to make GQ~ = 1 again can rective constant shear flows q~ and q:, Us tng the carryover correction factors previOusly be written found we obtain,
q.
;:;. 
II \
I.

jI
C.O.F. (1 to 2)
(z ~) cell
() web
(21)
(2)
= 200 =
723
.277
qT
q~
Fig. e shows the constant shear flow q~ and GQ = 1 for each cell. However these corrective shear flows q'~ e , 0587 were added assuming the cells were again independent ot each other or did not have the cammon web
Fig. g shows the result s . The final or resulting cell shear flows then equal the original shear flows plus all corrective cell shear flOWS, or
:;...._ _
q'~
'" . 00385
0i,~=.OOi71j
\ (I)
t!(21)
(12).
ThUS in
Fig. e
Fig. g
1
bringing the cells together again the Interior web Is SUbjected to be resultant shear flow of q~  q~. In other wordS it we were to add the shear flows of Fig. e to those of Fig. d, we would not have GQ 1 and GQ~ equal to 1. The resulting values would be closer to 1.0 than were r OUIld tor the shear flow system or Fig. d. Considering Fig. e, we will now add a second to cells set ot correctIve shear flows q~ and (1) and (2) respectively to ~ke GQ, and GQ~ = 1 tor cells acting independently. Considering cell (1), and proceeding with same reasoning as betore~
(r inat )
q:
Fig. h shows the final results. To check the final twist at each cell the value GQ will be computed for each cell using the q values in Fig. h.
:;;..
r . 1787
I~
I'q,=
.2534
(1)
ll(~
fll I
~Fig. h
cen
(1)
+ .0747
.~~J
'"
.997
Hence
q~
=,..,.,c'"" = .0587
12
x .237
= .013911!in.
GQ. =
2 x 139.3 [
.1787
or
q~
(C.OS.)
2 to 1 times q~
Operations Table 1 arranges the calculations so that the steps jealing with the corrective shear flows can be carried out rapidly and with a minimum of thought.
A6.13
Cell 2
C.O. FactoX'
AssUIIled
.277 3
2 J
a a
237
C.O. (12) =
(~)
web 12
10
CO C.O.
T 2A 0," T
02~8
<15. zo"
00
~
.3""
~~
14 1".
z (t) cell
=
2
.05
..lQ. + ...2Q.
.40
.05
0.1
C.O. (21) =
m (~l
web 21
cell 1
10
= .~5 = .25
:05
.Explanation of Table 1 Line 1 gives the carryover factors for each cell, computed as explained before. Line 2 gives the necessary constant shear flow q in each cell to give unit rate of twist to each cell acting independently. (GQ = I). Line 3 gives the first set of constant corrective shear flows to add to each cell. The corrective q referred to as the carryover q or C.O.q in the table consists of the q in the adjacent cell times the C.O. factor of that cell. Thus .237 x .109 = .0258 is carryover from cell (2) to cell (1) and .277 x .212 = .0587 is 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 cell (1). Line 5 repeats the corrective carry over process once more. Line 6 gives the final q values which equal the original q plus all carryover q values.
Example Problem 1 (2 cells) Determine the ~nternal shear flaN system for the ~NO cell tube in Fig. A6.25 when subjected to a torque of 20,000 in. lbs.
Line 2 gives the shear flow q in each cell, when it Is assumed each cell is acting separately and is subjected to a unit rate at twist or GQ  1. The calculations for the q values are as follows: For cell (1 )
q.
Z
2A.
(~)
cell 1
2 x 100 25 40
.05
For cell (2 )
q.
Z
2A.
(~)
cell 2
r
10"
I.05
10"
.J..05
10"
I
0.10
.05
Cell 1
0.10
Cell
2
0.10
1
1
.05
Fig. A6.25
OPERlTIOMS TA8U 2
Cell 1
Cell 2 0
O.2~
,
.
CO. l"ac'toJ:"
l!IU11ed
0.' 10
2. S
.0 0
C.O
C.O
0.
10
>n.
0 11
0 10
,
11100 18870
7770 41
~able
(2):
Line 1 gives the values of the carryover factors. The values are calculated as follows: C.O. factor cell (1) to cell (2):
AG.14
TORSION
OPERATIONS TABLE 3 Cell 1 Cell 2 Call 3 .288
o.
I
,
For cell (1) T = 2 x 100 x 38.85 = 7770 in. lb. For cell (2) T = 2 x 100 x 55.5 = 11100 in. lb. Line 10 gives the sum at th9 above two val~as which equals 18870 in. lb. The original requirement of ~he problem was the shear flow system ~or a torque of 20,000"#. Therefore the required q values follow by direct proportion, whence
1
a
3
co
CO 0 C 0 C.O.
Factor
Assumed
, 6
7
23 25.30 3 88
.21
, ,.
10
T
co.
.'88
2.53 0.685
.33
.0
=I88'iO x
q~
20000
38.85
= 41.2#/ln.
= 58.9i/ln.
11
2215 9200 6020 19000 + 6020 + 2215 ~ 1 43S'"1lI Total ~ q _ to;,: T=100,000 183 112.5 '"0
"
20000 = I8870
x 55.5
Explanation of
Cell (1)
sol~tion
Cell (2 )
These values are shown in line 11 of Table 2. Check on twist of cells under fir~l q values. The relative total strain around each cell 1 boundary is given by the term. A z q tL for the cell. Thus tor cell (1)
"i =
A~
.: 144
1140
= 96 "t= 1040
A.
L
A.
L
= 56. 5
Zt::; 1055
Shear flow q for GQ 1 for each cell acting in dependently: Cell (1)
q
= 2A, =
"~
t
=
For call (2),
212
Cell (2)
q .: 2A.~
Z L t
192 1040
.1845
L z q~ =
.~~
,. 58.9
~03.01)J
Cell (3)
.107
Thus both cells have the same twist. In the above calculations q~ and q~ act clockwise in each cell, hence the shear flow on the !nter1or common Neb 15 the difference of ~~e two q values. Example Problem 2. Three cells The three cell structure in Fig. A6.26 is Subjected to an external torque of  100,000 in. lb. Determine the internal reSisting shear flow pattern.
I  12"
To avoid small numbers t~ese values of q are ~ultiplied ~y 100 and er.tered on line 2 of the table. Calculation of carry factors as given in line 1 of Table 3.
Cell (1) to (2)
"( ~)
=
cell 2
 1040 
240  .23
Cell
(2) to (1 )
C.O.
I
I
8" .032
I,
(21)
m
(3 )
web 21
!(~)cell
.04
i
12" .04 Cell 1
.0' Cell 2
I
I
.041 \
Cell (2 ) to
...i...
.04
0032
!y
'Cell 3
I
C. O.
(23)
(~)
web 23
_ 300
z(~)ce113
 1055
.234
Fig. A6.26
VEHICLE STRUCTURES
A6.15
c .0.
as given in TablA 4. Line 10 shows the correction of q values to develop a reSisting torque of 100,000n#. The ~ultlplylng factor 100,000/80630.
:5
(~)
cell 2
The balance at the solution or procedure in Table 3 is the same as explained for Problem 1 in Table (2). It should be noticed that cell (2) being between two cells receives carryover q ~lues from both adjacent cells and these ~NO values are added together before being distributed 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 8.90. Pne carry 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 multiplied 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.
40
(1600)
A6.16 Torsion of ThinWalled Cylinder HaVing Closed Type Stiffeners. The airplane thinwalled structure usually corrta tns ront tuc inat stiffeners spaced around the outer walls as illustrated in Fi~:. A6.28 and A6.29.
.025
For the open type stiffener as illustrated in Fig. A6.28, the torsional rigidity of the individual stiffeners as compared t 0 the torsional rigidity of the thinwal!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 structure, 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 overall 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 thinwalled tube and closed stiffeners are converted or transformed into a Single thinwalled tube by modifying the closed stiffeners by either one of the follOWing procedures: 
16"
(500) Cell 2
(625)
(1600)
14001
cen
Cell 3
Cell 4
Fig. A6.27
OPDATIO!(S TABLE
.oj,
Cell 1
1 2 3
Cel]. 2
20. .153 38.8 3 88 , 1< 2 00 188 0.47 022 o '0 12
cell 3
181 .078 28.4 3 as 1. 38 0.36 0.57 0 2 o O. o 02
Cell
.oj,
CO,
0 0 0
s
a
9 10
17' 22.4 1 e 1 69 1 10
",
.16", 20.5 2 22 O. 3 J
Original
~iiiener
o rs
7620 47.1
31560
22550 49.9
8900 29.41
by
65.3
oot , .tsK
I
In Fig. A6.27 the values in the rectangles represent the cell areas. The values in ( ) represent ~he Lit values for the particular wall or web. After studying example problems I and 2 one should r2ve no trouble checking the values
,by
Fig. AG.31
t:::tst ... t s K ~
A6.16
TORSION
"liner" 1n the stiffener having a thickness e = tSK dis' (See F'ig. A6.31.) The enclosed area (A) of the cell now equals the original area less that area crosslatched in Fig. A6.31. Procedure (1) slightly overestimates and procedure (2) slightly underestimates the stiffness 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 thinwalled overall cell.
t
Lateral shear, H ,  mainly ; :Partiy lateral shear and partl y torsional shear/:"""",V
fiW
Fig. A6.32
The equations derived in the previous part at this Chapter assumed that crosssections 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 crosssections. In actual practical structures restraint against thiS free warping at sections is however Often present. For example, the airplane cantilever wing from its attachment to a rather rigid fuselage structure is restrained against '~rping at the wingfuselage attachment pOint. Another example of restraint is a heavy wing balkhead such as those carrying a landing gear or power plant reaction. The flanges of these heavy balkheads Often possess considerable lateral bending stitfness, hence they tend to prevent warping ot the wing crosssection. Since only torsional forces are being considered here as being applied to the member, the stresses prOduced normal to the crosssection of the member namely, tension and compression must add up zero tor eqUilibrtum. Thus the applied torque is carried by pure torsion action at the member and part by the Iongt tuo i ona.i stresses normal to the member crosssections. The percentage ot the total torque carried by each action dependS on the dimensions and shape at the crosssection and the length ot the member. Fig. A6.32 illustrates the distortion of an open section, nanely, a channel section subjected to a pure torsional force T at its :ree end and fixed at the other or supporting end. Near the fixed end the applied torque is practically all resisted by the lateral bending at the top and bottom legs of the channel acting as short cantilever beams, thus torming the couple With n forces as illustrated in the Figure. Near the !ree end of the ~emberJ these top and bottom legs are now very long cantilever beams and thus their bending rigidity 1s small and thus the pure torsional rigidity of the Section in this region is greater than the bending rigidity at the channel legs.
Fig. A6.33 shows ~ Ibea~ s~bjected to a torSional moment T at its free e~d. The prcblem will be to deterTine ~rEt proportien of the torque T is taken by ~he flanges in bending and what proportion by pure shear, at two different sections, namely 10 inches ~nd 40 incheS trom the fixed end of the Ibeam. .,.u"","",+ Fixed
End
rL 75''1
F
'1 3
T'
FIg. A6. 35
Fig. A6.34 s~ows the torque diViding into VNO parts, namely the couple force FF ~armed by bending of the flanges of ~he Ibea~ and the pure shearln~ stress system on :he crosssection. :<'ig. A6.35 shows the twisting 0:' tne Section through a distance 5. The solution will consist in computing the angle of twist 9 under the two stress conditions and equat1nz them. Let Ta be the proportion of the total torque T carried by the fl2nges in bending forming the couple FF in Fig. A6.34. From Fig. A6.35, the angle of tNist can be written
2 Tg L"
3
naIT
..  ,.
~
~.
...
~.
ANALYSIS
AND DESIGN OF
FLIGHT VEHICLE
STRUCTURES
A6.17
Note: The deflection of a cantilever ~e~ with a load F at its end equals FL~/3EI, and I the moment of inertia of a rectangle about Its c~n ter axis tb~/12. hence Now let Tt be the ?ortlon ot the tctal torque carried by the ~e~ber in pure torsion. T~e approximate sol~tion for open sections co~posed of rectang~lar ele~ents as given in Art. A6.6, equation (12) will be used.
7he shaft rotates at constant Soeec. The difference in be!t pull on two sides of a p~lley are shown. an tne figure. Calculate the ::2...'(1~~ torsional shearing stress in the 13/4 inch shaft bet~een pulleys (1) and (2) and between
(2) and (3).
9t
Equating GS to tit
3 Tr L
Gt'(b + b w)
''iT'
2
(2) rl. 1/2 HP. motor operating at lOCO RR~ rotateS a 3/4.035 al~~lnum alloy torque tuee 30 inches long which drives t~e gear =ectan~~ tor ~p&ra~ln; a Wing flap. Dete~ine the ~~<: ~~ torslcnal st~ess in the torque sha:t ~cer fUll ~~wer and RPM. Find ~he ang~lar de:lec~ian of shaft in the 30 lr.ch ~ength. ?olar =c~ar.~ ot inertia of tube = .01 in. ~odulus ot r:gidity G = 3300,000 psi.
Tni.n.lb.
we can
3 E h
8L
i te
b3
+
2Ge(2b
b. n)
I I / I
,/
I
I
/
'f I "
3 x 10.5
10 4
3.4 x 1.75
io
= 9.65
taken
.T
Fig. AS, 36
15" :
'!'c
__ " "," e ..,.. '!!: "\._ '{:: .;.0 ee snea r :lo'iJ :::. :es:~'::::g I tne ext erna L t orque 0: scccc :n.lb. ,.;,:2: ::.:.:. wall t'u cxnees are gi '''e:: or: tno r:;,;u.:e. As "'\.::",e the tube 1s 100 In. long and :in~ :~e :c::::~3.1
~
.J_
de f t i cn , za cer 15 <:ll:"''"ll.L'1.u::l. (J= It we ccnsid.er the section 4Q inches :~c~ 32CO , COO ps 1. ) the fixed end, then L = 40 inches. Thus if 40 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 eri 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 perT '" 100,000 in. lb. cent at L = 10 inc~es fro~ support. Thus in general the e::ect of the end restraint decreaSes rapidly with increasing value of L. 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.
L e c i a t a.I L o y ,
f\
A6.18 Problems.
Fi.g. AS. 37
320*
Fig. A6.35
(1} In Fig. A6.36 ?~lley (1) is ~~e 1rivl~g PUlley and (2) and (3) are the driven pulleys,
(5) In the 3cell structure of fig. A6.37 detieratne the internal res t st tng shear ilow due to exter~.al torque of 100,000 in. lb. ?or a length of 100 inches calculate twist of cellular structure i: G is assumAd 3,800,000 psi.
A6.18
TORSION
~14'
(6) Remove the .05 interior web of Fig. A6.37 and calculate shear flow and twist. (7) Remove both interior webs of F:g. A6.37 and calculate shear flow and twist. (8) Each of ~he cellular structures in Fig. A6.38 is subjected to a torsional noment of 120,000 In.lbs. USing the ~ethod of successive approximation calculate reststing shear flow pattern.
.0'
14"i .03
.0'
05 1. .03
I I
(01
L
T
10" .04
.03
04 .03
.03
, !
I
10" .03
10" ......l
.03 .03 .03 .0' .03 I (b]
.035
.03
~035
.03
1==::::::::===
I i i I'"
'. ,
,.
The big helicopters of the future will be used in ~ny important industrial and military operat 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.
i~portant
(1) A knowledge of the loaddefo~ation characteristics of the air?lane 1s of prllnary ireportance in st~dles of the influence of structural flexibility upon airplane performance. (2) Calculations of deflections are necessary in solving for tt.e In~ernal load distributions of complex redundant structures.
Work =
L: Pi
I~l
6 Oi
in the limit as
600
Work
=Lt'J P
dO
The elastic deflection of a structure under load i s the cumuLat;'... ''j result of the strain deformation of the individual ~lements composing the structure. AS SUCh, one method of solution for the t~tal deflection ~ight in,o~ve a vectorial addition of these individual :ontriJutions. ~he involved geometry of most ?ract~cal str~ctures ~kes such an approach prohiJitively ditf~c~lt. For complex structures the more popular tec~~lques ~re analytical rather than vectorial. 7hey::ieal dt r sc t Ly 'Nith quarrt t t Les which are not themselves ceflections but from which deflections ~y be obtained by SUitable operations. ~he nethods employed herein tor deflection calculations are analytical in nature.
A7.2 Work and Strain Energy.
Fig. A7.1
curve ls a straight line whose equation 1s P = k6 and the strain ener~J is readily computed to be p' U equally, U = 2k'
A7.3 Strain Energy EXl)ressions for Various Loadings.
as defined in ~echanlcs 1s the product of rcrce t ines cr stence . It the rorce v~aries over the distance then the work 1s computed by the integral calculus. Thus the work cone by a varying force P in deforming a body an aaount 6 is ',.Jork = Pd.a and is represented by the area ander the load deformation (P6) curve as shown in Pig. A7.1. If the tefor.ned body 1s per~ectly elastic the energy stored i~ the body nay oe co~pletely rec overed , the bocy un Loac tng a Lcng the same ?6 c'Jrve :ollowee for increasing load. T~is energy is ca l Ied the elastic ~ energy q GeforT.aticn (hersa~ter ~he strain energy, for brevi~y) and is denoted by ~~e s~bol U. Thus U = Pdc. Should the cody '_be .Li near Iy
~ork
'Q
Sd5 AE 5'
(1)
Alternately.
U'
S'L
2AE   
(2 )
.r
Squations (1) and (2) are equivalent expressions ~Jr the strain energy in a Qni~o~ bar, the farner expressl~g U in terms of the defo~~tior. and the latter expressing U in~e~ of load. 7he secand form of expression is ~ore 80nvenlent for ~eneral usage and succeeding strain energy :ornulas will be ~ut in this form. ?cr a tensile bar haVing nonuni:orm properties (varying AE). or for which the axial load S varies, the strain energy is computed by the calculus. Thus the energy in a dir:erentlal element of length dx is atven by eq , (2) as
dU :: ;;'2AE Where S and AE
elastic (as are ~ost bodles jealt w1~h in structural analYSiS) t~en the loaddefo~tion
(1) Design gpec iarist, ConvairAstronautics
sex
~ength
of element.
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
U
(3)
=j
dU
=j
"2AE
~.adx
2 ~ '" JL ( n"'AE
c
1
, 2L nx)' ex . S.Ln
Example ProbleT. 1
01'5" 1.... '0" '.:; I ) (22'7\(40) '" 0.0920 in Ics , 9.87(2)(10 x 2.0 5 ) ' !
STEArN SN~GY C? FLZ:('l)ES
~~ len;th L ~~er the actien of a pure couple ~~jer~oes an angular ro:atlcn proportional :0 the couple. Thus, from elementary strength of Taterials
linea~ly
tapered
al~nin~ll
Find U.
A,
~ifo~ bea~
(Ea~ ~:=..1 f
15CO#
where the constant S1, the product of elastic modulus by sect~on ~oment of inertia. is called the ~bendlng stif:ness R Since the rotation 0 builds up linearly with M the strain energy stored is
U =
2AE =
In 2 I
h 1b ncr s
1 2:
MO
or
U =
2 Ef
M'L
14,)
Note that although P was a negative (compreSSive) load the strain energy remained positive. Example Problem 2 Find U in a uniform bar under a load _ nx( q  q , cos 2I Fig .40.7.3).
~inlng
Fo~ a jeam haVing ncn~~i~orn prcpert~es (varying El) and/or for which M varies along the beam, the strain energy 1s com;uted by the calculus. From sq. (4) the str~in energy in a beam element ot 1ifferentlal length is
q(~ m1
~WT
I.
Hence summing over the complete beam to get the total strain energy one has
1. 40
40 "
.to 2 in
J: _ lO::do 6 pd
'to.
2'
U = j dU = ~ jl1'dx 2 EI
Exa~ple
(5 )
lb/in
Problem 3 the beam of Fig. A7.4 derive the strain energy expression as a function of Mo.
For~ r;.M GEl = constant
Flg~
A7.3
c.
by
statics.
1 ! I L.J
Fig. A7.4
Solution:
= q , 2L n sin 2L
mI'
The bending ~oment diagram (Fig. A74a) was found and an analytiC expression 'Nrltten ~or v
A7.3
~=GJT

(6)
x
Fig. A7.4a
where the constant GJ, the product of the elastic modulus in shear by the cross section polar moment of inertia, 15 called the "torsional stiffness" Since the CNist ~ builds up linearly with T the strain ener~J stored 1s
Then
or
(7 )
Example Problem 4 Determine the strain energy of flexure of the beam of Fig. A7.5. Neglect shear strain energy.
varyin~
HT~7'

(8)
GEl
=Constant
z:
L
01
2"
In passing it is worth remarking that one often encounters the group symbol "GJ" in use for the torsional stiffness of a noncircular shaft or beam such as an aircraftwing. In such a case the torsional stiftness has not been computed literally as G x J, but rather asdetined by eq, (5), viz. the ratio ot torque to rate of twist. Example Problem 5 For a certain flight condition the torque on an airplane wing due to aerodynamic loading is given as shown graphically in Fig. A7.6. The variation of torsional sti:tness GJ 1s given in like manner. Find the strain energy stored.
1.
01"":~
Solution: The bending moment diagram was drawn first (Fig. A7.5a) and analytic expressions were written :cr 11.
__
Fig. A7.5a
Fig. A7.5
Inspection of the diagram revealed that the ener 5J of flexure in the right half of the beam ~ust be identical with that of the left half. Hence
U
o !_ _
Y=O
__
~_~
__
~_~_
Root
y=L Tip
Solution:
=2
A uniform circular sr~ft car~ying a torque T experiences a total CNist in a length L proportional to the torque. Thus from ele~entary strength ef ~terials
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 and were entered in the table. For convenience in ha~d11ng the numerical work all the variables were treated in ~ondi~ensional :OrTI, eq. (8) being changed as follows
I U  ~
i~' 0
.:::
GJ
;,::~ ~.~<~~
~'. ':";\':,,,.
~:~:.i.?1:
,
A7
DEFLECTIONS
OF
STRUCTURES
~~er~J ~~ ~
where
T 
TlTn , 7=y/L
GJ/GJR
strain
act~o~
of
GJ =
transverse shear loading V(l~s). lor this purpose V = ~ x dy x t and t x dy is called A, the ~eaw. cross sectional area. nance
dU
t~erefore t~e
The subscript R denotes Wrcot~ val~e (y=o). The coefficients for 31~pson's ~Jle appearing 1~ colunn (6) were taken from tte express~on for
odd n
= V'dx ZAG
and
=;'r

(r 0 + 4f J,. + 2f 2 + 4f 3 +    _ +
the enttre
U=
bea~
_ 
15 f + 3f + ~f ) 4 nz nL 4 n
J2AG
V'dx
                (10,
Sxa:nple Problem 6
TABLE A7.1
STA.
7
T
1.0
.98 .9'
(T)~
GJ
(T)'
ill
1.0
1. 26 1. 63 1. 22
m~ Coeff.
.0667 .336 216 305 070 0.0
Sc Iut t on :
1.0
1.0
.78
.8
.8
.07
2
."
a
.98 .88
.54
. 37 .24
.133 . 250
' . 200 . 0833
1.0
. 084
.35
.14
Ee"
~ 2
2 =jP o
j
. . . . . . AG CONSTANT
z=
Therefore the strain energy was
994
Then
= LTR' _. x .994
GJR
2 x
,L/2
0
Poa dx
::: Po aL 2AG
Eq , (9) may be used to comput.e the shear strain ener 5J ~n a thin sheet. The elemen~ dx x dy Is visualized as but one of many in the sheet and the total ener~J is oJtained by summing. Thus U
=~
SS7't gx
dy
        
  
(11)
~ = ~ where G is the material elastic modulus Fig. A7.6a in shear. For the displacements as shown in the sketch only the down load on the lett hand side does any work. (In general all four sides ~ove, but if the mot~on is referred to axes lying along ~NO adjacent sides of the element, as was done here, the derivation is Simplified). This load is equal to 't' x dy x t and moves an amount a' x ex, Then
dU
Here a double integration 1s required, t~e summation being carried out in both directions over the sheet. In dealing with ~hin sheet the use or the shear flow q Tt 1s often convenient so that eq , (11) rewrites
U
=2 ~
Sf q'dx dy Gt
qas
(l1a)
A very important special case occurs when a homogenous sheet or constant thickness 1s analyzed assuming q 1s constant everywhere. In this case one has
_ q II U  2Gt
=~
7Et dx dy
=2 G
7'
where sheet.
(9 )
           (12)
tdxdy 
The sketch is visualized as a side view of an element of length dx taken from a beam of total height dy.
A7.5
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).
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
j~7
T" f.bl
Limit
flo .. 0,
dU = Pda
"...JL.j a
Fig. A7.9a
Fig. A7.8
Solution: The shear flow was assumed given by Sredt's formula (Ref. Chapter A5)
q=.1.=
2A Zbe
Thus
dU = P (14)
do In words, 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.ork. The reader may find this theorem sta~quite differently in the literature on rigid body mechanics 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
1: 2Gt
T'L(b+c) 4Gb lie it
q"S
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 elements under various loadings, is readily found as a simple sum.
=0
It is next argued that if the Change in displacement dO 1s sufficiently small the load P re~Alns senSibly constant and hence
dU  d(PO)
=0
       (15)
diUpo)
=0
           (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 8alculus 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.
The quantity U  PO is called the total potent:'..al or' the system and eq, (15), resemb 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 .vorx . In structural analysiS the ~ost Dnportant ~ses of these theorems are T~de in problems concerning buckling instability ~d other nonlinear1ties. 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, exaoinatior. of the arsa above the loaddeformation cu:ve shows that increments in this area (called
(',
,.
A7.8
DEFLECTIONS
OF
STRUCTURES
the complementary energy,U), are related to the load and deformation by (see Fig. A7.G(b).
. "P
L'
r Ou
Force {Ibs .") Moment (in. Ibs , ) Torque (in. Ibs.) Pressure (los/in"') Shear Flow (lbs/in)
~force~
Area (in"')
, , ,
dU ::6dP
iiP
dU'
=6
            (16)
This is the Theorem or Complementary STIergy. Now tor the linearly elastic body a very tmportant theorem follows since (Fig. A7.9c)
dU = dU
Any generalizations of t~e ~eanings of 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) (Castigliano's T~eorem) is restricted to linearlY elastic structures. A brief exa~ple will serve tor illustration at the pOSSible pitfalls. The strain energy stored in an initiallY straight uniform column under an axial load P when deflected into a hal! sine wave is
so that
. rt x Y :: Y oSlDT; M::PY
y
, , P~U' ~ a. u
;..,
Fig. A7.9c
U .fM""" P"Yd L
0 .,
If(LY' di.)
2
2EI
4EI
dx ; Y3 n: 4L
dU = 0
dP
            
(17)
Fig. A7.10
~
P~6L~
In worne ,
~The rate of change ot strain ener~J with respect to lOad is equal to the ass0"CTat'8ddetlection".   Eq. (17) and the above quotation are statements of Castigl1ano's Theorem. For a bOdY under the simUltaneous action of several loads the theorem is writte~ so as to apply individually to each load and its associated detlection, thus:
where 0 is the end shortenir.g due to bowing. Because the deflections grow rapidly as P approaches the critical (buckling) load the problem is nonlinear. The details of the calculation of U are given with Fig. A7.l0. Now according to the Theorem of ViTtual
~ (eq. 14)
,j3
dU _ P
but Theretore
nilEI = P
=0
(17a)
P"L"
The partial derl~tive sign in eq. (17a) indicates that the increment in strain energy is due to a small change in the particular load P , all other loads held constant. i Note that by "load~ and "detlection" may be meant: Load ASsOciated Detlection
or
_ n 2 EI P 
r::r
(Euler formula for uniform co nem ) . correct resuj.t , Application at Castigliano's Theorem. eq. (17), leads to the erroneous result:
dP
dU = 0
2P6L~
dU dP
The proof of the theorem for the case of multiple ioads is
nAEI
generally formulated more rigorously. appeal to a simple diagram such as Fig. A7. 9c being less effective. se e, for example, "Theory of ElaSticitY" by S. Timoshenko.
See Art. A17.6, Chapter AI7 for detailed derivation of this equation.
A7.7
(incorrect) Moral: Do not use Castigliano's theorem for nonlinear problems. Fortunately the above restriction upon ~he use of Castigliano's theorem is not a very severe one, the majority of avery day structural problems being linear. Castigliano's theorem Is quite useful in performing deflection calculations and a variety of applications will be made in the following section~.
A7.6 Calculations of Structural Deflections by Use of Ca8tigliano's Theorem.
Then
6p
= dll dp = 43.26
x 10 P tt.
Note that in this problem the only use made of the calculus was in the differentiation. A Simple sum was used to form U, as per eq , (2). xample Problem 9 Derive Bredt's to~ula for the rate at twist of a hollow, closed, thinwalled tUbe:
ds
As the examples of Art. A7.3 have illustrated,the strain energy of a structure can be expressed as a function of the external loadings prOVided the internal load or stress distribution is calculable. Having the strain energy so expressed the deflections at points ot load application can be determined with the aid of eq. (17), Castigliano's Theorem. In the examples to follow the deflections ot a variety of statically determinate structures are computed. Methods of handling redundant structures are considered in subsequent articles. Example Problem 8 F1nd the vertical deflection at the point 01' load application of the crane of F1g. A7.ll. Cross sectional areas are given on each member. The stranded cables have effective ~odul1i of 13.5 x 10" pSi. E = 29,000,000 for other members.
U=zJ
rJ Y
This is the only energy stored, secondary etfects neglected. OVer a given cross section q 1s a constant and is given by Bredt's equation
q
Solution: Since the twist per unit length was desired the strain energy per unit length only was written. Thus, assuming y in the axial direction, no integration was made with respect to y. The integration in the remaining direction was to be carried out around the perimeter ot the tube and so the index was changed tram "x" to the ~ore ap proprtat;e "e ", Hence (compare with Example Pr'ob , 7)
~'1I5'\.l\'~C.
A'
'I
60'
'r 1i1.~"~'~:
E
.$:'
p
~~
0
u
Fig. A7,1l
= J:..
I_
2GJ'
q ads t
=~
~ cis 8AG~ t
~30'+ 60' ~
Th~ symbol ~ means the integration is carried out around the tube perimeter. Therefore
Solution: The strain energy considered here was that due to axial lo~ding in each of the four members. The load distribution was obtained fram statics and the ener 5 J calculation was made in tabular form a~ follows:
TABLE A7.2
MEMBER
I
S LaS
AE X
iO"'5
LBS
s.aL x
AE
10"
OA
AS
1. 50 P
AC
DC
40.0 50.0
63.0
4,5
0.66 P Ii
26.48 P
13.33 P 2.79 P I;:;  43.26 P
Ii
In the follOWing example the desired detlec tion is at the free end of the bar where no load is applied. A fictitious load will be added for 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 set equal to zero. This technique gives the desi~ed result in as much as the deflection 15 equal to the rate f change at stratn energy with respect to the load and such a rate exists even though the load itself be zero. Example Problem 10 Compute the axial motion at the free end at the tapered bar of Fig. A7.12.
A7.8
DEFLECTIONS
OF
STRUCTURES
eval~ate.
constant J
q 
flA'"
~
j
, Flo
x
:Q:a'nple ?roblem 11 ~ind the deflection at ,oint 8 of the beam of rig. A7.:3.
p
Fig. A7.12
EI constant
e
Solution: After addition of the ~istitious end load R the axial load frem statics was found to Je
S(x) = p.
I
 :L
L : L
A,
3"
Fig. A7.13
Solution: A fictitious load 21. NaS added at point B and the bending moment dia;ram was 1rawn in VNO parts.
0<X<:L/3
Mp
q (L  x)
t
0
S'dx AE
1
2
Po'
2A,E
I~
[Po
+ q
~
2A,E
IL 'L x)'dx
( 1 _ ,,) , 2L
~en
Betore evaluating the integrals it was observed that the steps to follow, in which U was 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 term was evaluated.
1 + 2EI
u =
Then
6
II
X + 2L
A,E
Rq (1 _
L~
r/'
,
2~I
~
L/7
V[(2P +
2)
I
.z. 2E!
11 _ L~ 2)
r
c
,
dy
,
I
3 [IP
0B=~~~
important labor savings may be had in the calculation of deflections "by Castigllano's theorem. In the strain energy integrals arising in 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. Provided certain requirements for continuity of the functions are met  and they invaria~ly are in these problems  the differentiation with respect to P may be carried out before the integration is made. The resulting integrals genAn
ir J
3yl
L/ 3 [(ZP+
e
+ ~
3y) dy
 For beams of usual high length  to  depth ratios the shear strain energy is small compared to the energy 01 flexure. Neglecting the shear energy is eqctvarent to neglecting the shear deflection contribution. (see p, A7.14)
. ._
.... 7 _
A7.9
fictitious load P L , havin~served its pur;058, was set equal to zero Jefore complet~ng the wark.
'Nere
Plan View
Q ...
PR
rL
T:!1.us
Fig. A7.14b
U = 2EI
+
_ PR)lI s in" Q Rd g
oCOS
Q +
PR(l  cos
Q)J.:lI
Rd g
Example Problem 12
Fig. A7.14 shows a cantilever round rod of diameter 0 formed in a quarter circle and acted upon by a torque To. Find the vertical movement of the free end. Solution: Fig. A7.14a shows the vector resolution of
applied torque To on
(Note use of ~RdQ~ for the length of a differential beam element instead of "dx"). Differentiating under the integral sign
Fig. A7.14
TI = + 
au
(T,  PR)R"
EI
IT /2 o sin.:ll Q \
dQ
R"
GJ
beam elements. TL (Q) = ToGas Q and the moment nl,{Q) = To sin 'g. APplication of a fictitious vertical load'? (dOwn) at the point of desired 1eflection gave the loadings shown in Fig. A7.14b.
ene
Plan View
au
= ToR
4
li
4  IT IT ) ( ~EI
A7.7 Calculation of Structural Deflections by the Method of DummyUnit 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 illsuitee 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 DummyUnit 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 obtains 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 DummyUnit Loads (Virtual Loads) I  From Cast1glianc's TO_eoram Beginning with the genera.I expression for strain energy, aq. (13) II  From the ?r1nciple of Virtual work
~nen a system J:' ~orces whose ~esultant is zero (a system in eQ~1Iibr1um) is displaced a
~ Variously called the MaxwellMohr Method, Method of Virtual Velocities, Method of Virtual Work, Method of Auxilliary
A7.10 I Cont'd.
U
DEFLECTIONS
S "dX + ~ AE 2
= ~
J!'l"u :::1
J T~f
+
OF STRUCTURES IT Cont'd.
  
etc.
IS~dX
~
small ~ount withcut dis~urbing the equl:1br1~ balar.ce, the work done is zero  obv1ously, since zero res41tant torse ~ovlng through a distance does zero work. 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 the point whose deflection is ~eslred ane the two reactions fiXing the line of referer.ce. ~d the "internal system" consisting of the axial loads acting on the truss members to produce equilibrium. T~ese latter are denoted by the synbol ~u". T~is set o~ forceS 15 considered small enough so as not to.affect t~e actual behavior of the structure during subsequent application of a real set of ~Jor loads. This unit load set is present solely fer ~thematical reasons and is called a "virtual loading" or "dummy loading". Assume now that t~e struct~e undergoes a deformation due to application of a set of real loadS, the virtual leads ftgoing along for t~ ride n Each ~ember of the structure suffers a deformation denoted by ~ ~ The Virtual loading 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 by the internal vrrtuaj loads. The work done by the external virtual forces is equal to one pound times the deflection at joint C, the reactions Rl. and Ra not moving. !~at is
J M~dX 1
,1
       etc.
<lP
'"'JP'1'P'""
i i
~M
~T
  
 etc.
Each of these 1s the "rate of change of so andso with respect to Pi" or "how much soandso changes when Pi changes a unit a~ount" OR EQUALLY, "the soandso loadln~ for a unit load PI' Thus, to compute these partial derivative terms one need only compute the internal loadings due to a unit load (the virtual load) applied at the point of desired deflection. For example, the term elM/elPt, could be computed in either ot the ~HO ways shown in Fig. A7.15a.
RATE METHOD
UNIT METHOD
Fig. A7.15a
Exter~al Virt~l Work = l' x 6 c The internal Virtual Nark is the sum over the l. = x x structure of the products of th~ ~ember vir~ual loads u by the member distortions~. That is, Likewise, d s~ Pc, where Pc is a load (real or fictitious) applied at jOint c ot Fig. A7.15b, Internal Virtual ~ork = ~ u.~ is given by the loadings for the unit load applied as shown. Then equatin~ these works, In practice the use at the WIlt load is 1 x 0c = Z ud . most convenient. Using the notation If the deformations 6 are the result of _ oM elastic strains due to real member loads S then ,m =~, 1 Do = 8L/AE for each member and one has 6 = Z u 8L v a q = aq c AE Pi'  ~ The ar~~ent given above may be extended for the unit loadings, the deflection equation QUickly to include the internal virtual work of becomes virtual bending moments (TIl), t or s i on loads (t), 6 = Sud:< + IlmdX + ~ shear loads (v), and shear flows (q) doing work 1 AE EI GJ during defo~tions due to real moments (M), (T), shear loads {Vr;ind shear flows + ~ + q qGfdY          (IS) torques (q). The general expression becomes
oM
I1 = Pl.X
m = dummy loading
(=~~J
:V
If
1.0
1.0
AE
+ ;
EI
...
GJ
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
.
ANALYSIS AND DESIGN OF FLIGHT VEHICLE STRUCTURES
'.?:'"B.'P!
A7 11
In applying eq. (18) the labor of a =efleccalculation divides conveniently into several steps: 1. Calculation of the real (actual) load distribution (S,M,T, etc.)
t~on
distribution (u,m,t, etc.) due to a unit (virtual) load applied at the potnt of desired deflection and reacted at the reference polnt(s). 111. Calculation ot flex!bl11tles,
A~ , "'1
111
"U1,"
loads
"U.'
loads
Fig. A7.16a
Fig. A7.1Bb
'.
etc.
......
AS
TABLJ: AT.3
Lbo
iv. Summation and/or Integration. Here again note the general nature of the terms "load" and "ceriece i on". (See p. A7.6) The following examples apply the method of dummyunit loads to the determination of absolute and relative deflections, both rotation and translatioo. Example Problem 13 The pinjointed truss of Fig. A7.l6 is acted upon by the external loading system shown. The ~ember loads are given on the fi~~re. Member properties are given ~n Table A7.3. Find the vertical movement of joint G and the horizontal ~ovement of jOint H.
A 10
BC
CD
,.,
aB
..
BQ
.. " .. .. ..
AIls 10 Ibo
4.1 3.074 3.0T4 5.385 5.385
AI
toslO
8.2'1'0 g.T5' g. '1'1 5.5111 5.1111 1.821 4.1128 14.3.
""Ib
10,500
2,~0
.,
a
c
II.
1.'
.....,. """""
a.. " a a a a
22.01 12.01
_21.~
~s 10 ~JCIO.
2,250
_6,~
... "
5,2llO .0. T5 1
_21.31
3."
lO.a
_.000
.3, 'liO
a
53."
111.80
50 50
3.4'
5.3G 3.0'1'4 3.074 3.0'1'4
00
SF
co
DB
.. .. ..
50
1."
3,000
_l,OOO
3.000
1000"
1000*
1000'*
a a a
a
a
fj~gl~(?'bgl~~~ l
~"
0
sods
2250
Ie 2250
10
Answers :
oG
:::
0.286"
VER
;o:"g40"
(l
N
'5250 .5250
O~OR
2000t
the joint moves to the LEFT since the unit load was drawn to the RIGHT in Fig. A7.16b). Example Problem 14 For the truss of Fig. A7.16 rlnd the following relative displacements of joints: c) the movement of joint c in the direction of a diagonal line joining c and F. d) the movement of joint G relative to a line joining points F and H.
~elatlve deflections are determined by applying unit (Virtual) loads at the points where the deflections are desired and by support ing such ~nlt load systems at the reference poir.ts of the motion. ThUS, ror solution to ,art (c) a unit load system was applied as shown in Fig. A7.16c and for the solution of
Solution: only the energy of axial loadings in the members was considered. Unit (virtual) loads were applied successively at jOints G and H as shown in Figs. A7.15a and A7.16b. All Sand u loadings were entered in Table A7.3 and the
calculat~on ~or 0
= J Sudx AE
6
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
.6
roO
.375
.375
':c:
, '..,i'
"
O..n
r
:0
It:I
Rotations, both absolute anQ "elative a~~ 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 show the unit (virtual) loadings fer par t.s (e) and (f) respectively. T~ble A7.5 completes the calculation, the real loads and ~ember ~lexib~llties (L/~) be Ing the same as fer examp.Le
dete~ined by a~ply1ng
6' 0
,. FOG
0" H
Ac
problem 13.
R=U'
.5
1jjo
.5
"us" loads
Fig. A7.16c
Fig. A7.16 d
"~~?@?1
R=l!40'll' .025
1/50*,
/\/0
1/50# R=1/40'"
. 02~ G'l/50#
TABLE A'l.4
R=1/40#
MEM. AE
...h. x las
S,b.
10,500
2, aso
'.
. ,
0 0 0
'.
 0
1".
A'r1.llcb.
AS
Be
.315 .31:1
0 0
13.11
0 0
8.23
8.23 TABLE A7.5 0 0
CD E"
FG
~xlO.s ~X10S
.'
n
0
17.61
..k.
blEM.
:I:
reWIb
GB BE
0
8,750
c o
c o
71.84
0
AE
(See Table
S,b
10,500 a. 450 2,250 5,2:10 5,2:10
0
a&lnchJ.
Il....lncb.l.
AE
A'l.3)
eo
DO
'.000
3,~0
1.0
.e"
.'20
....
0
AE
Be
6.a70 9. 7S~ 9.
7~9
. 0.5 . D25
.020
1."
.55
.55 . 73 . 73
0 0 0
_. D.5
0 0 0 0 0
.0.55
0 0 0 0
BF CG DB
2.000
1.000
.. ..
0 0
21.8
13. a 0 13. a
CD EF FG GB BE
BG
DO
.50
0
20.82 .10.41
~.591
.025 .025 i
0 0 0
5.591
2.000
_.50
e.eu
4.926 14. see 9.320 13.012 13. 012 13.012
= 65.87
Z =19.38
8, 750
0
1. 08 0 0 0 0
'.000
3.750 2,000 1,000 2,000
.OU
0
".015
0 0 0
.52
0
BF CG
,
I
0 0 0
Therefore the movement of joint c towards joint F was 0 = .06587 inches and the motion of joint G relative to a line between F and H was 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 t) the rotation of member 80 relative to member CO.
, DB
Z = 4.73
.z =0.53
~herefare
DO was 80
DG
= .00473
relat~ve
to CG was Q _ CG 3G
= .00053
radians
A7.13
a.ample fToblem IS tine the vertical eeflect:on of point C for the cantilever jea~ of ?lg. A7.17 carrying a concentrated load P at its end. Al~o find Slope of elastic ~urve at 8. Solution:
points C, DJ ~, the attachment ~oints of an aileron or a flap. The wing beam deflection bends the aileron or flap structure by applying a lcad at 0 thru aileron s~pporting bracket. To know this force the deflection of the wing beam at D relative to li~e CE ~ust be known. Solution: Origin at 8: Ii
f
with origin at 9
M =  Px
Mmdx
W
= 30
= 15
x"
(Fig. A7.17)
For virtual
load1~g
I!l.
(?ig. A7.17a)
= 0, far x cb
1 (x  b), for x)b
Px(x + b
Fig. A7.18
Hence
!1mdx
=P
Idx
= (?x'"
=
 Pbx Idx
EI ) b
rL
(x
 bxidx
1.[ x'
EI 3
bX'J L
2 b
~
,r ''1
0
Rd'
OJ x
<
20
= zero,
,
then 6
= PL
C
3/3EI
l==
1'
)'
4p
a
B
Fig. A7.17
J
=F:r 1
.... .1.
l1m dx
E1
It
c r.,l
Fig. A7.17a
k
20
50
h
50
80
15 x"
un1 t.
Fig. A7.17b
ooupl_
e
For
vi~tua1
c,
=JMmdx
2;1
= ISO; [_ 11.72
+ 6.25 + .300 
.400 +
= 0,
x cb ,
 I, x>b
?x
17.9 X lOll
EI
Hence !1mdX
(?xi (Ii ex
ex
o
J
If b
Mmdx_PfL
SI b
xcx
2EI
= 0,
= 2EI
0' ,
Therefore deflection of point 0 relative to line joining ~E is down because result comes out negative and therefore opposite to directicn of Virtual load. Example Problem 18
1''''
lb/rt
L
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 representative of a yr~ctlcal ?rOblem in aBronauttcs, in that A3 :n.1.~ht represent a rear wing beam and
a.~ple
tn
~L
'; II i i i
2
x~
x'1
wx e
m diag'l
'2
(b)
",to
'2
'oiL'
j:~~lr,1
R=l1i'
~ '\ R=l/L unit
1 R=l/L
(a)
( c)
couple
Fig. A7.19
Fig. A7.20
A7.14
DEFLECTIONS
Ol"
STRUCTURES
Find the horizontal deflection ot point C for the frame and loading of Fig. A7.l9. Also angular deflection af C with respect to line CD. Solution: Fig. b shows the static moment curve for the given loading and Fig. c the moment diagram for the virtual loading of a unit horizontal load applied at C and resisted at D.
_JMmdx   ,. EI
M
dy = J ex
and
~ = A'
V
~s
J where
cross sectional
area of beam at section and ~s ~cdulus of rigidity: and assuming that the shearing stress
V A is
cr~sssectlon.
Therefore
= Vvdx A.
~hen
tion for the shear slipS of all elements of the beam equals
= wLx
2
wx 2
:n = h
hence
_______ (a)
where V 15 the shear at any section due ~o g~ven loads. v = shear at any sect10n due to unit hypothetical load at 'tae point where tne aer iec, EI c tion is wanted and acting in the desired direction of the deflection. The reactions to the hypo1 whV' thetical unit load fix the line of reference for 12 EI the deflection. A is the crosssectional area and Es the To tind angular deflection at C apply a unit modulus of rigidity. Equation (a) is slightly imaginary couple at C with reactions at C and D. in error as the Shear:ng stress is not ~~i~orn F.ig. A7.20 shows the virtual m diagram. over the crosssection. e.g. being ~raballc ~cr a rectan~~lar section. However. ~he average shearing stress gives close results. So ~ 1. WX') ~ dx For a unito~ load of w per unit le~gth, the EI EI 2 2L J;I, , L center deflection on a s~ply supported beam 1s: 
=JL
= fL (WLx _
= ,;\
1
= ijL
(W~'  w:: ) dx
[~'  w:~ J
center =
wL;'
2f~~=2 ~ L AE s
,e
= 24 Ef
Linear Deflection of Beams Due to Shear by Virtual Work.
r
J
(wL 2
wx)
1
~
ax
AE s
wL'
SAEs
For bending deflection for a simply supported beam uniformly loaded the center deflection 1s
Generally speakingJ shear deflections in beams are small compared to those due to bending except tor comparatively short beams and theretore are usually neglected in deflection calculations. A close approximation is sometimes made by USing a modUlus of elasticity slightly less than that for bending and using the bending deflection equations. The expression for shear deflection of a beam is derived tram the same reasoning as in previous derivations. The virtual work equation tor the hypothetical unit load system for a shear detrusion dy (Fig. A7.2l) conSidering only dx elastic is 1 x O=vdy where v is shear on section due to unit hypothetical load at paint OJ and dy Is the shear detruslon at the element dx due to any given load system or any other cause.
Henae
wL'
8AE s
5wL"
24
(Er
384EI.
For
(1
= depth)
~
"'~dLdy
12
Seep. A7.4
Thus the Shearing deflection in percent of the bending deflection equals 4.1~ for a ~ ratiO ot for Ibeam sections and 1.4 percent for rectangular sections.
iz
A7.15
Example Problem 19
t ;
10011"
t '+ lO'4
'O
vA., 
501'
Unit. Load.
IF==='
torsior~l ~oment at any section due to a unit virtual couple acting at section where angle of ~N1st is desired and acting in the plane of the desired deflection. (inch Ibs/inch lb)
Example Problem
Fig. A7.22
Find the vertical deflection of free end A due to shear deformation for beam of Fig. A7.22 assuming shearing stress uniform over crosssection, and AE constant. s
Vvdx AE s V = 100' for x = 0 to 10 10 to 20 V 150 for x v = 1 for x = a to 20
"' _ I
Example Problem 20 Fig. A7.Z3 shows a cantilever landing gear strutaxle unit ABC lying in XY plane. A load of 1000* is applied to axle at point A normal to XY plane. Find the deflection of point A no~l to XY plane. Assume strut and axle are tubular and of constant section. Solution: The loading shown causes both bending and twisting of the strut axle unit. First find bending and torsional moments on axle and strut due to 1000* load.
hence
1 AEs
to ...L 0 1001 dx + AE s
fO
10
1501 dx 20 10 2500 AE s
Y
0 3
10 1 [lOOX] + = IE: s 0
~.[150X]
I /
The angle of twist of a circular shaft due to a torsional moment may be found by similar reasoning as used in previous articles for finding deflection due to bending or shear forces. The resulting expressions are: 
1/ \\ i s: ~~3'
B
36"
I
B
Fig. A7.24
,.
Fig. A7.23
Ttdx
E J
o =
JTEts~
Member AS
            (A)
Member BC
(B)
In equation (A), for translation deflections, T twisting moment at any section due to applied vNisting forces. t = torsional ~oment at any Section due to a virtual unit 1 lb. force applied at the point where deflection is wanted and applied in the direction or the desired displacement. (in lbs/Ib) Es Shearing modulus of elasticity for the material. (also nGn) J polar moment of inertia of the circular crcsseect ron. In equation (B), for rotational deflections, Q = angle in twist at any section due to the applied vNisting moments in planes per?endicular to the shaft axis. Angle in radi~~s.
Now apply a unit 1* rorc~ at A normal to xy plane as shown in F!g. A7.24 and find bending and torsional moments due to this 1* force. Member A8
m
= 1=0 =3
= x,
(r or x
=o
to 3)
Member Be
m
sin 20'
+ 1
x (for x
=0
to 36)
A7.16
DEFLECTIONS OF STRUCTURES
Subt.
6
I'!mdx +
81
JTtdx
E J
The shear :lows shown on ~he (nearly) horizontal edges of the web panels are average values. ?i6' A7.27 ~s an explOded vi~N of the beam shaNing the unit (Virtual) loads.
tr
J:
+
1 1000 X . X dx Er
f:
0
6 K1000 x 1026)
(x
1.026)J dx
+
EJ
f36 (2820)
(2.82) dx
Virtual loading. Fig. A7. 27
EI
_1_ (286200)
(16,92500)
Since both axial loads and shear flo'~ were considered, ~he form of deflection eq~tion used was
q q dxdy
Gt
I~tegrations in the ~langes were made assuming linear load variations. Such ~~ integration carried out over a uniform :lange of length L whose real load varies from 8 1 to 5 j and WhOS8 virtual lead varies :rcm ~i ~o Uj yields
EsJ
Note: A practical landing gear strut would involve a tapered or reinforced section involving a variable I and J and the integration would have to be done graphically or numerically.
Example Problem 21
For the trnnweo aluminum beam of Fig. A7.25 determine the deflection at point Gunder the loading shown. Stringer section areas are given on the figure.
A ... A=.15 B A=.15 C ,..A.15 D
J
Q
L Sudx =
AE
r I L
20"
t=.0
32
1 l
A=.08
t=.032 1
A=.08 A.05"
t=.032
..L
T IS"
The integrations in the trapezoidal sheet panels were made using the shear flows en the (nearly) horizontal sides as average values, assumed constant over the panel. with this
= qAV Gt where S 15 the panel area. The calculation was completad in Table
rF I G
+
20"
4
H+ P.j80"*
Simplification
55
qAV
q,w :ixdy
Solution: It was assumed that the webs did not buckle and carried shear only. Fig. A7.26 is an exploded view of the beam Showing the internal real loads carried as determined by statics:
3. OP_
A7.6.
c==.:===;::::~~~====
: :.:.,. p, ,
'.00 P ,_L,H P, 1.00,_:.011 .un I .,.,., ,
~ID *~i
2. 184P
3.0P..."..
.04
;~~ .O!~~~~. W I
1.201p
1. 201P
".'"
.'.'M "'':10""'.01'
'u.
".0 P
...
...:,:" ,',W")
~.r'._.
",,"'''1 i
.,.\<
I
I : 171.0 P."
a.c_r......
<>U'.
""I
: .1
The equations of statics for tapered beam webs are derived in Art. AIS.18, Ch. AIS.
CD<lHI._P'
FLIGHT VEH
LE
TR
RE
A7.17
= (271.0
2
13.4) P x 10 11
j
II5'~
L
~hr;l~Tdx
0
As noted in the "virtual work derivation" of the dummyunit 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 dummyunit 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 temperature distribution shown in Fig. A7.28. Assume material properti9s are not functions of temperature.
T
;t Idx
a
Fig. A7.29
b
Solution:
T~e axial deformation of a differential element of the u?per flange (sUbSer}pt u) was ass~~8d given by = a T ax where a was the material ~hermal coefficient of expansion. The lower flange, having received no heating underNent no expanSion.
Au
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 obtained in the flanges were:
Fig. A7.28
Solution: The thermal coefficient of expansion of the rod material was u. Hence a rod element of 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
T . dx
T~en ~he
L  x
U
(~
 "L)
~J~ o n
a T
23 idealized ~Noflange cantilever beam of ?ig. A7.2Sa undergoes rapid ~eatlng of the upper flange to a ~emperature T, uni:orm spanNise, above that of the lower flange. Deterxlne the resulting displaeemer.t of the free end.
2.X8.!l:p19 Proj19m
~he
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 surfaces shown in Fig. A7.30b.
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.
Solution:
Flg. A7.30
A7.9 Matrix Methods in Deflection Calculations. Introduction. There is much to recommend the 'lse natr rx methads'~ for the hand Ling of 'the quarrt
of Lty
An element of the beam of length Rd is shawn in Fig. A7.30ce Due to the linear temperature 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 applied at the cut surface as shown in Fig. A7.30b giving the following unit loadings around the ring. From. unit redundant couple (X)
~
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 operation is present which permits the solution of additional related proble~ by a simple expansion of the program; The notation itsel: suggests new and improved methods both of theoretical approach 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!IlYUnit Loads to matriX notation. Other appropriate references are listed in the bibliography.
=1
BASIS OF METHOD
Assume the structure to be analyzed has been idealized into a trusslike assembly of ~ods, , bars, tubes and panels (sheets) upon which are acting the external loads applied as concentrated loads Pm or Pn, each with a different numerical
my
U
R (1  cos
$)
(~)
= cos $
$
$
(.)
(b)
ua method is
 sin
e =
Then
Ju
f ru . dQ
= 
the right)
Q,)For the reader not famUia.r with the elementary arithmetic rules of matrix operations employed here, a. short appendix has been included. iL. B. Wehle Jr. and Warner LanSing, A Method for Reducing the AnalySis of Complex Redunci:Uit""""StrUCtures to a RouuneProcedure, Journ. of Aero. Sciences, 19, october
mr.
4Z
A7.l9
subscript. T~us the system of Fig. A7.3la 1s idealized into that of Fig. A7.3l~. With the above idealization an improved scheme may be employed :0 systematize the computation of deflection calculations. The following 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 numerical subscripts), is used to describe the internal stress distribution. The q's may represent axial loadS, ~oments, shears, etc. In conjunction 'Nith a set of member flexibility coefficients, 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 relate the interr~l generalized forces qi, qj to the external applied ~ , Pm or Pn With this relationship the strain energy expreSSion obtained in I, above is then transformed to give U as a function of the P's. III. Gastigliano's Theorem 1s used to compute deflections.
It is next desired to write the strain energy as a function of the q's. Continuing the illustrative example, write
1 q.j,'X'dx
1.
'
Observe that each of the integral terms in the above expression is a property of the structural element (variation of El) and of the nature of the associated generalized force (exponent on variable). Introducing the notation
= IL'Y'd Y Ela
c
=
the strain
ener~J
d y jL.E1.
c
becomes
  (19)
~
M
=
EI.
A EI,
Equation (19) is an expression for U Which could have been written lmm.ediatelY from physical considerations. Each coefficient aij is the displacement at paint i per unit change in torce at point j. This identity is eaSily seen by applying Castlgliano's Theorem to eq. (19). With this interpretation the first ter.n in eq. (19), representing the strain energy in the outer beam portion, is written by analogy to The remaining three terms, representing the energy stored ~n the lIh~er Jeam segment by qa and q~J are likewise readilY written, With proper account taken tor the cross influence at one :orce upon another (the na a3 qa q~~ term). No'te that
eq , (2) of Art. A7.3 ( U = 1 2: S'L) AE
LaY::L\~{a)
q,
~q1. q3
{I
I) q.
\
~qg
q,
q1.
(e)
ql)",
q,
(b)
JI
(d)
q1.X
O<x<L1.
qtj
M;Q3+qaY O<y<L.
q. I)
q'f
(e)
=
Fig. A7.32. Some possible choices of generalized forces. ("Relative displacements in the individual member")
= "jl         (20)
Hence
~
,.
A7.20
DEFLECTIONS
OF
STRUCTURES
The general form of strain energy expression 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 , , _I , I , 1
+qfoQ1. au
, ,
I
I
, , , ,
,
Fig. A7.32f
qNql aN' +
  
   
+ qN qN (lNN
In matrix notation
q, q.
1
o.
1
P, P.
1
r
~,
, ,
q.
q. (21)
, ,
, , '
,
, ,
qN
, ,
'NT1~tan
~~
,\';'
     (22)
The matrix
dlstr1butlon~
at
the elements are zero. In the specifiC example, eq. (21) would be 'Nritten
the generalized forces (the Q's) for a unit value of load P~, all other external loads zero.
THE STRAIN ENERGY IN TSR11S OF APPLIED LOADS .
o
o o
q,
q.
q.
If eq. (22) and its transpose are used GO subst rtuts for the q.. "s in eq. (21) one gets
2U
(23)
The problem of computing and tabulating various (lij'S is considered in detail later.
RELATING THE INTERNAL GENERALIZED FORClli TO THE EXTERNAL APPLIED LOADS
For the statically determinate structures considered in this chapter the internal forces are related to the external loads by use of the equations at static equilibrium. A set of linear equations results. ThUS, in the speCific example conSidered, it Pl and Pa are the external loads applied as in Fig. A7.32f, then by statics (reter to Fig. A7.32b)
here i and j are used intaras are ~ and n. Also [~lJ is the transpose of [G1m] I i.e. lntercr~nge of subscripts denotes transposition (see appendiX). It the matrix triple product in eq. (23) is fo~ed and defined as
Chan~eably
notat~on
In the
   
(2?)
then
A7.21
2U
i L':rJ1J) 1:1 ,I L J r,
p
        
(25)
Sq. (25) expresses the strain energy as a function of the external ~pplied loads. In the ~pecific example being used
1 r La 1L'J 1
0
plication itself yields a summation to complete the calculation. ?rcm this last discussion it is clear that eq , (26a) also may be derived by formulating the DummyUnit Load equations (Art. A7.7) in matrix notation. To illustrate the application of the ~trix methods presented thus far, a brief and elementary example 1s worked with those tools already developed. Example Problem 25 Determine the influence coefficient matrix rorxransverae forces to be applied to the uniform cantilever ot Fig. A7.33 at the three points indicated.
(25)
t
Solution:
,@
,lID
3"""" 3 e 3
I.
b, b
q.
yields
I~ ) I ~p n .f
(26 )
="q q. '
=:J~)q"
c:::::::=::=J.~ q1.
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 recollecting in catrr tx
rorn,
The matrix
~mnJ
the ext ernal points "a" ror urn t values of the loads? and 1s therefore, by definition, the
n
~trix
T~e choice and numbering at genera11zed forces are shawn on the figure. These forces were placed so that pr~vlously derived expressions for the a's could be used. The following member flexibility coefficients were computed. Note that the only nonzero coefficients ot mixed subscripts (i not equal to j) are those for loads common to an element.
It is
instr~ctive
to Nrtte eq.
(26)
out as
This expression was adopted from that developed for a transverse shear force on a cantilever beam segment in the preceding illustrative example.
L/~.dx
Q
L' =
BIEr =
aa::a
= a.....
and compare the expression with a typical term from the dumnyunlt load ~et~od equations, say SL 6=zu E: ' :n the ~~x equation (25a) ~he
A
(virt~al) TIU"
matrix correspondThe
This expreSSion is for a couple on the end of a cantilever segment (of length L/3). __
Ua "  au L/ 3
[2. i .:1 J
ing to
ydY_L a
EI  18EI
L...,. T~e
ma tr tx pr'oduct
l'J:l]
:Pn (
J
0
;ives the :nember load distrlb~t~ons due to the real applied loads, hence these are the "B" leads. Finally the operation of matrix :nulti
This expression is for the cross influence af a couple ~~d a shear load an a cantilever segment. Collecting in matrix form,
A7.22
DEFLECTIONS
OF
STRUCTURES
27
0
0 0 0
0 0 0
G!Il11]
is seen to be svmetcrc
[a ! ]
l
27
1
L'
3EI
0 0
0
6L
0 0
1 6L 1
0 0
about the main diagonal as it ~ust be: from reciprocal theorsm A .c.' nm (see eq. (20)). mn
27
1
6L
rr w
1
6L
Several ~ember ~lexibi:ity coefficients are derived below for various members ~d leadings. A more comprehensive listing 1s available in :he paper by Wehle and Lansing referenced ea:'lier.
BARS
successively applying unit leads at ~oints 1, 2 and 3 and computing the values of the q's by statics.
1
0 0 0 0
The energy in a uniform bar under varytng axial force (Fig. A7.34l is
[G1m
1 L/3
1
1 0
1 2L/3 L/3
1
0
Fig. A7.34
~ 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 point "2" only, and so forth. Finally,
= ...!c.. 3AE
and,
0
1..
'Z1
o
o
o
1
~~=3iIOl
o
a1 j
1 L/
1 0
06iL~
o o
0 0 0
a
o
1
~ u L J~ = ,q ~q = Ai: L
1
0 L/3 0
0 276"
.l. ,];1
6L ~ 3
2L L/3 0
An equally likely choice at generalized forces for the above case is shown ~n Fig. A7.34a. The strain energy is (x measured from .free end)
Fig. A7.34a
t~xee
3AE
L'
= 2AE
L'
.Y!
A7.23
In ~he case of tapered ~embers the coefflc1ents are determined by evaluating integrals of the form
o.  1 JL!(X)dx
o
U =
2~I gql
a11 =...h...
Fig. A7.34e
NOTE: The coen
Ij ]"
A"1Xi
Such a quadrature can always be made in these problema. For the linearly tapered bar the results may be obtained as functions of the end area ratios. ThUS, Wehle and Lansing give
ajj
EI
a11
=~.
= 3EI
L'
1"
11
SHEAR PANELS
icients for bars and beams are directly analogous (compare several cases) so that for tapered beams one uses the results {or tapered bars with EI in place of AB.
alj = 6A E !j 1
Fig. A7.34b
For the rectangular shear panel with a uniform shear flow Q on all edges (Fig. A7.34t) i
0jj
3AE . ~jj
1
~
',j\..
3
,
S = surface area
Fig. A7.34f
:
I
The trapezo idal shear panel (Fig. A7.34g) is treated approximately by using the average shear flow on the nonparallel sides as though it were constant throughout the sheet. Thus
S = surface area
Fig. A7. 34c::::::!:d ~e~ v~tes lin~rlY :J: ~
a11 =.....
Gt
= BEAm
Since by statics qj =
Ql ., one could use Q
Fig. A7.34g h, as j a. an alternate choice of generalized torce and
0:
Fig.
ajj 
_(h.)' S. h. lit
TORSION BAR
r1====<' 1
L
1 q,
energy Then
0.11
Then
Fig. A7.34d
L = GJ
aQl,Qj
An
alterr~te
O'u
=_L_ (= Ojl)
6EI
Example ?rob 1em 26 The tubular steel truss ot Fig. A7.35 is to be analyzed for vertical deflections at points ~ and F under several load conditions in which vertical leads are to ~e applied to all joints excepting A and D. The crass sectional areas at tube member~ are given en the figure. Set up the matrix form at expression tor the
7
A7 24
DEFLECTIONS OF
Con.tant Axial Load; 'l.i'l.j
=~
STRUCTURES
'K
1 2 3 0
0 0
3
0
1.0 0
0
!ben
0
C
Fig. A7.35
Fig. A7.35a
deflections at points E and r. Solution: The member flexibility coefficient for a unitorm bar under constant axial load is L/AE. Fig. A7.36a gives the ~umberlng scheme applied to the members and the q's (these being one and the same, since q is constant in a given member). Fig. A7.36b shows the numbering scheme adopted for the external loading points.
~i~
0
0
0
0
1.25
5
6
1.03
 .75
0
1.13
7 8
9
.848
1.55
.165
0.40
.825 .20
0.40
tsl3J3 DJsL
1 2
8
plo
Pa
t=]
rs,
o.
Il,
= Emi] [OiJ]
~Jn]
eq. (26)
257 252 257 252 389 789 252
P,
Ip,
440
pe~
Fig. A1.36a
Fig. A7.36b
=E
389 257
e..
p'l P.
J
1nche s
6,
389
789
~
2
3
a a a
1<6
P,
1 92.2
a
92.2
a
0
138
["1J] i
5
6
7
8
a a a a a a a a
a a a a
0
a a a
0 0
a a a a
55.7
a a a
0
a a a a a
185
a
0
a
0 0
a a a a
155
a a
a a a
a a a a a a a
55.7
a a
a a a a a a a a
229
The results here give the deflections of all four points. Since only the deflections of points 3 and 4 were desired the ~lrst two rows
of
tmnJ
@bl])
In the case of a Pin jointed truss, where only a single generalized force is required to describe the strain energy per member, the matrix or member fleXibility coeffic18nts 15 a diagonal matrix as above. Unit lead distributions were obtained by placing unit loads succeSSively at external loading points 1, 2, 3 and 4 (Fig. A7.36b). The results were collected in matrix form as
The matrix fo~ of equation above ~s useful in organiZing the computation of deflections for a number of different loading conditions. Thus should there be several different sets of external loads P corresponding to various loadn, ing conditions, each set is placed in column fo~ giving the loads as the rectan~~lar matriX
~nk] ,
load conditions.
tip'
A7.25
L'
(26b)
au
a"
AB;SEI
LAB~J
= L8~
a ..
GJ
a"
a ..
L;~
3EI
a"
?roblem 27
Deflections at points 1, 2, 3 and 4 of the truss of ?lg. A7.35 are Geslred for the following loading conditions:
Ccndf t t on No.
a,.
= L8~
EI
P,
2500 1200 1800
P.
2000 800 1470
1 2
3
eco
2100
1200
P,
P, ( see Fig.
450 A7.36bl 01750 1100
GJ
=1.3 EI )
1
2
0 3,' 0 0 0 0 0 0
3 0 0 '.0 0 0 0 0 0
4 0 0 0
5 0 0 0 36,0 0 0 0
,
0 0 0 0 0 36.0 0
8 0 0 0 0 0
Solution:
T~e
'.0 0 0 0 0 0 0 0
0 0 0 0 0 0 40.8 0
3 4
set up as
440 389 257 389 2500 1200 2000  800
15,500
~'"'J
~ample
1
;;
1800
1470
,
7
see
0 0 0
""
sae
0
sae
15.500
~1~ were
obtained
IT~bered
Tor the landing gear unit of SXanple Problem 20, Fig. A7.23 find the matrix of influence soef~lclents relating deflections due to 11ft and drag loads acting at point A and torque about t~e axle AB.
~
2
2
0 1.0
0
3
0 0
Sol:ltion:
st~ucture was divided into elements set of internal generalized forces ap~lied as shown in ~~g. A?37a. (Torques and noments are shown vectorlally by R.H. rules). Axial stresses were neglected in CB.
1 1.0 0
0
7he
~nd ~~e
3
4
LO
0 0 1.026
2.311
,342 3 0
0
0 0
.937
5
6
7
.342
0
0
L..
l.G
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
~he
Exa."'11ple Problem 29 T~e bea~ at example problem 21 is to be resolved by the T~trlx wethoas ?resented ~ereln.
A7.26
DEFLECTIONS" OF
STRUCTURES
Influence coefficients for points ?, G and H are to be found. Solution: Fig. A7.38 shows the choice and numbering of generalized forces.
q.
Example Problem 30 Deflections of statically indeterminate 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 distribution. The exac~ ~nternal ~orce distribution is not necessarily required in making deflection calculations inasmuch as such a calculation amounts to an integration over the structure  an operation which tends to average Fig. A7.38 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 unit loadings. elements as these were known to ue equal to In the folloWing probl~ ~he ~trix of inthose ot the upper tlange due to symmetry. fluence coefficients is dete~ined tor a sin~le Entries were made for ll in matrix form as beiJ cell, t~Ieebay 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.
",==="'_q.. .
/'" (caef!'s a 33 '" a~3 ;. a ~3/ call'td \ ' fr.two ===='~9 ~'?....;, members)
a
1
10
, I ,
e.se
=0'
1
,
10"
t,~
a a
10
aoes
17.8
....
17 8
....
''.
2500
....
8.89
.....
Unit load Yalues were obtained as in Fig. A7.27, considered to be external loading number "two". S1ml1ar diagrams were drawn for unit loads at points "one" (H) and "three" (Ft(llt'. rig.
A.
20" 1:5"
'J.
15.. J
'
72"
2 0
Fig. A7.39a
a a
1.0 .0600 11.0: .0496 2.00
~~
a a a
0
Solution: Fig. A7.39b is an exploded view of the beam showing the placement and numbering of the internal generalized forces. Note that only the upper side of the beam was numbered, the lower side being identical by symmetry. Member fleXibility coefficients were computed by the formulas ot Art. A7.l0 and entered in ~trix form as below. Note tr2t all entries tor Which there Nere corresponding leads on the lower surface of the beam were dOUbled. ay this ~eans the total strain ener~J of the beam
a
.0540 2.184
5
6 7 8
9
a a
1.0 .0545 1.00
a
.0447 3.00
A7 27
flow (q = ~) due to the torque developed in transferring the load to one side. The matriX triple product
was accounted ror. Note also that entries tor a~~, a 8 8 , a and a 1 0 1 0 were redoubled as each at these q's act on two (identical) members.
In the calculation of structural deflections there occur many steps involving simple integral properties of elementary functions. The Method of Elastic Weights (and the Area 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, relYing 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 equalto the bending moment at A due to the diagram acting as ~ distributed beam load.  Spelled out in steps; iThe ~ diagram is drawn just as it occurs due to the applied beam load ii  This dIa~ is VisualIzed as being the loading on a second beam (the conjugate beam) supported at the paints of reference for the deflection desired iii  The bending moment iro this conjugate beam is found at the station where the deflection of the original beam was desired. This bending moment is equal to the desired deflection. To prove the theorem, consider the dummyunit load (Virtual work) equation
ii
Note;
Unit load distributions were obtained tor successive applications at unit loads to points one through six (Fig. A7.39a). The internal torces Dredicted by the E.T.E, for a load thraugh the shear center (center of beam, due to symmetry) were superposed on the uniform shear
~
1
2
0.1
0.1
0.3 2
2
~L~ =
3
4
5
6 7
.0l25 .0375
0.15
0.05
0.05
a
9
10
11
.osea
.020
0023
12
1.020
'I
13
14
I .0423
3 3
.cess .csss
1 1
.=
.100 1
1
.oasa
This expression equates the exte~al virtual work done by a unit load, applied at a point de:lectlng an amount 6, to the internal Virtual work on a beam element experiencing an angular change d g = ~~. The sum (integral) of such expressions throughout a beam gives the total
15
3 3
a
2
A7. :.::8
DEFLECTIONS OF
STHUCTURES
deflection at tne ccmt (c r . eq , 103). ..~e now show tr~t the ceflecticr expression, uSin: :he above equat~on, is the SaT.9 as the bending moment expreSSion fer ~ simple beam loaded by
"an
In rig. r
co~sider
Mex
81
~3
a load on a
si~~ly
elastic weight 21 In Fig. A740, the la~cing at (a) pr ocuces the real moments of (0). Consider the ce r jeotlons ot points 3 and C due to the ar.gular
ch~nge Mdx ~
~ Md.x
supported bea~ a~c dete~ine the benG:n; 200ent at pOints band c G~e ~o ~dx acting at pOl~t c. "IT Mdx L XLdx
4EI
:2
L
BEl
in
a 0 ea~
",1 ~~emen t
1, ....L
I L
1 4
l.hLhj
' 4
I')
values of ~~e te~ bending ~o~e~ts at 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 as the influence line for illo~e~t ~t ~oi~ts b and c. Therefore deflections of a si~ple bea~ can be determined by cons Ider mg the :1 curve as an
~hese
"
~WEI diagram ~
.., !"odx
(h)
81
imaginary beam loading. ~he bending ~oment at any pOint due to this !1 load.ing equals the 1e
EI
flection of the beam
ili~der
EIpj
(c)
Mdx
=
L~4
Likewise it is easily pr)ved :~zt :he ~n~~ lar change at any section of a sl~ply supported beam is equal to the shear at that section due to ~ !'! dt azram acting as a beam Lcac . Sol
A1.13 Example Problems
(d) mb, moment diagram for unit load acting downward at point B
r16 L
f (el me" unit load at point C
Mdx
Examnle Problem 31. ~ind :he ~2rt~cal deflection and slope of paints?. and b :or be~ and loading sho'Nn in Fig. A7.41. T~e lower F~g. shows the :noment c Iagram for load? acting at center at a sl~ple beam.
lEI
r~1
If)
IP
b
4EI
Fig. A7.40
3 Mdx
4EI
lMdx
f P 2
~;P
2
L L
EI Constant
For a unit load at point b, Fig. d shows the m diagram. The value of m at the midpOint of dx (point a) = LIS. Hence
PL
16
Fig. A7.11
6"
b = ~~ . ~
16
~~
PL'
16
PL'
For deflection ot paint c, ,1raw m diagram ror a unit load at C (see Fig. e). Value or ~ on eleaerrt dx = L Hence
00
.. For srmoncrty the points A, Band C were placed at the onequarter span points. The reader may satisfy himself with the general character of the proof by substituting xA, xB and XC for the point locations and then following through the argument once again.
Deflection at point a 8Guals bending ~cm ent due to M diagram as a load di vtoec by EI. (See lower Fig. at Fig. A7.~1)
6a
PL2 ( 16
4. 
PL2 64
L) 1 11 PL3 12 EI = 768 EI
Db
= (~2
~ i~2 ~ )Elr
=
3 PL2 64 EI
1a ~~3
The angUlar c~nge ot any point equals the shear due to ~JEI diagram as a load.
Ua
= (PL2 _ PL2)2:..
16 64 E1
,
A7.29
o (Slope is horizontal cr no change from origi~al direction of beam axis. ) ?rcbl,:::n 32. De t crntne ':;;'a cer Iec t Lcn of a sixple be~~ loaded unifo~y as shown in Fig. A7.42. The bending moment exor ess i on for a unlforn loac M = wlx  ,NX2 or ~arabolic as
~x":.:nJ'"'!
I<i
,
Airplane
1
I
A
50 "  ~:)O"130"
Anercn E
B
rHear Beam
FIA
C
2"
2"
/
25li'iin.
3
shown in F'1;. A7.42a. The deflection at mfd?oint equals the Jending J10ment due to M diagram as a load.
)
15~/in.
r'
2""
wL
wz in.
lll!
'LPl ! ! If Y+
T
wL
Fig. A7.43
Fig. A7.42
.?'7A.<(..M Diagram
Solution: Due to the beam variable moment ot inertia the beam length between A and C will be divided into 10 equal strips of 10 inches each. 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
....!....wL~""Area 24
Fig. A7.42a
24 wL"
of inertia at midpoint of the strip. These elastic 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 J3am at ~o:nt B ~ill equ~l the deflection of 8 with r3spect to line jOining AC. The bending moment at C = 15 x 30 x 15 ?
10
x 15 x 10 = 8250"#
6center
=(2~ NL3 . ~  d4
c
wL3 .
~ ~)
SlI
wL4
384
sr
ccent.er
=(2~
14 'NL3  2 WL 3) Ell = 0
Slope at supports
SXP~;le P~QbleJ1
= the
reaction
1 wL3 24 EI
33.
Fig. A7.43 shews the plan ~iew of onehalf of a cantilever wing. The aileron is supported on brackets at ?oints D, E and F with selfaligning 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 t~~ee ?oints. In the desi~ of the aileron beam and similarly for cases of wihg flaps this deflection produces critical bending ~oments. Assuming that the running load distriblted to the rear beam as the wing bends as a uni~ is as shown i~ the Fig., tind the deflection of pOint 3 with respect to straight line joining ~oints A and C, which will be the deflection of with respect to line jOining D and F it bracket deflection is neglected. 7he moment or inertia of the rear beam between A and C 'varies as indicated in :he Table A7.6
Bending moment expression be~#een points C and A equals, :M: = 8250 + 60Qx + 12.5x2 , where x = 0 to 100. Table A7.6 gives the detailed calculations tor the strip elastic loads. The I values assumed are typical values tor a aluminum alloy beam carrying the given load. The modulus ot elasticity E = 10 x 106 1s constant and thus can be omitted until the final calculations. The figure below the table shows the elastic loads on the imaginary beam.
Table A7.6 Strip
'0.
I 2 J
d.
'0.
e ,
lIo_nt at mid_ point US63 20063 31063 44580 60580 79050 1000.50 123$50 149550 118150 0
I at JlIidpoint
Elastic load
rt:
...
6 7
0
10
, s
co ro .o
0
IO IO IO IO IO IO
9.5
'.5
10 10 10 10
OJ eeo cOJ
21000 30900 41450 52300 63850 6.5940 62530 61770 62320 63600
'"
so
eo
0
0>
~
co eo eo OJ .c l.D
en
0
0
en
~
0 0
0
~
'"
OJ
A7.30
DEFLECTIONS OF
STRUCTURES
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 cantilever 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 such as piping, controls, etc., ~ust be so located as not to interfere with the wing deflections between A and B. For illustrative purposes a simplified loading r~ been assumed as shown in the figure. EI has been assumed as constant whereas the practical case would involve 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.
~r400"
D
For certain types of beam problems the oethad 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~.
.,
::.
0 0
0 0
300'~1w
::,
I
EI Constant
B
0 0
'\
0 0
40"
Fig. A7.43a
E
Moment Diagram
o! in.' o 1000
,
01 ., ,
Fig. A7.43b
"" 0,
0,
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 tlctic i ous beam, Hence EIO c = 25920 x 40  25920 x 20 518000 or 6c 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 imagiClb = (A ~:n Where en Is the moment at any nary beam. at points A or B will equal numerically )B EI ' sect i cn, distant x trom 3 due the deflection of these points With respect to to unit hypothetical couole the tip points D and E and since points A and B applied at B. But ~ = ~lty actually do not move this deflection will be the at all points between Band A. movement of the tip paints with respect to the 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
A7.31
Referring to Fig. A7.44, this ex~ressian represents the area ot the M diagram between points
EI
Henceos
= (_
PL2
Z
03
L) ~
EI
= _ PL3
3EI
8 and A. Thus the first principle: "The :hange 1rr slope of the elastic line of a beam between any ~wo ~o1nts A and 8 is nw~erically equal to the area of the :1 diagram betw"een
EI
these two
Deflect~on
pol~ts.n
Principle In Fig. A7.44 dete~ine the deflection of paint B normal to tangent of elastic curve at A. In Fig. A7.44 this deflection would be vertical since tangent to elastic line at A is horizontal. From virtual work expression Os = I~ ~~ ~, where m 1s the moment at any section A dist~~ce x from 8 due to a unit hypothetical vertical load acting at B. Hence m = l.x = x :or any paint between Band A. Hence Max This EI x 8 = B eXDression ~epresants the 1st noment of the M uiagracaoout a vertical t.hru B. 'fhus the
Example Problem 36 Fig. A7.46 illustrates the same simplified wing and loadiP~ as used in example prOblem 34. Find the deflection of paint C nor~l to line joining the support points A and B. Also find the deflection of the tip paints D and E relative to support pOints A and S.
g
o
o o
~B
1 T
EI Constant E
Fig. A7. 46
h ~!,ir!'J;
~
= =
Moment
in.~
o Diagram
TijijQ
fA
r
~67"'" I ~
r
~i ~ ,~o ~4
0 Oi
Solution: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 respect to horizontal tangent at C take moments about point A". 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
~I
stant.
(area)
(arm)
1~~0
1L i
(tangent at C)
=E~
(650
~ 646) 40 x ZO
1 EI
Fig. A7.45
(518400)
= deflection of
C normal to AB.
~~am
~
PL2
__
.!LJ
3
2'
To tind the vertical deflection of the tip point D with respect to line AB. tirst rind deflection of D with ~espect to horizontal tangent at C and subtract deflection of a ~ith respect to tangent at C.
Solutlon: The ~oment diagram ror given load is triangular as shown in rig. A7.45. Since the beam is fixed at A, the elastic line at A is horizontal or slope is zero. Therefore true slope at 8 equals ar.gular c~~ge becNeen A and 8 wm cn equals area of 'nomerrt diagram between A
and B diVided by E1.
~ (respect to tangent at C)
= E; (40000 x 267 +
Hence
"8
= (PLo
1 LIZ) I
PL Z
1Z7500 576 25920 7Z0) = (10Z700, 000) (See Fig. A7.46 for areas and arms or M/EI diagram). subtracting the deflection of A with ~e spect to C as found above we obtain
ir
 2EI
l~gO
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 diVided by S1, s~nce tangent to elastic curve at A 1s norlzontal aue to ~ixed support.
A7.32 A7.15
Beam Fixed End Moments by
DEFLECTIONS OF
~e~hod
STRUCTURES
of Area
:1A
=
Pab 2 and MB
lIOllIents
L'"""
=
From the rNQ prinCiples of area ~o~ent5 ~s given in Art. A7.14, it is evident that the deflection and 510;e of the elastic curve depend on the amount of bending ~oment area and its location or its center of graVity. Fig. A7.47 shows a beam r1xed at the ends and car~Jlng a single load P as shown. T~e oend ~ng moment shown in (c) can be considered as made up or two parts, namely that for a load P acting on a simply supported beam wmcn gives the triangular diagram Hlth value Pa (La)/L tor the moment at the load ~otnt, and secondly a trapezoidal moment diagra~ of negative sign 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 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 moment area principles they are easily determined. In Fig. b the slope of elastic curve at. A and a is zero or horizontal, thus the change in slope becNeen A and B is zero. By the 1st
To find the fixed end ~owents ~or a te~~ ""itt variable moment of InertIa use the t""iI crucrana in place of :he ~oment clagraos. Problem 37 Fig. A7.48 shows a :1xended two concentrated loads. finS the moments MA and roB'
2xa~ple
je~~ carry~ng f~xedend
M9"......:.....12"~9" +.~tB
__ I ;
4t=
j 100.
200*
t:',
(a)
Ih)
Fig. A7.48
l1lO'~
15MA
~L
(e)
~~B
(b]
~lastic Curve
Fig.A7.47
,
~~T
principle of area moments, this ~ans that the algebraic sum of the moment areas be~Neen A and 8 equal zero. Hence in Fig. c
(l1A !'lbl L (Fa (L  a) 2+ L L .2
0    
~a(ta)
lEI'' '
(c}
_ (AI
Solutlon; Fl~. b shows the static ~o~ent dla~ao ass~i~g the beam simplY supported a~ A and 3. For SImpliCity in finding areas and tak~r.g moments of the moment areas the :noment Qia~~ ~~s been divided into the 4 simple shapes as ShO','iC. "he 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 .~~ known ~oments MA and Mg. ~he area of these trianglas is shown as a concentrated 10aG at t~e centroids. Since the change in slope of the elastic curve bevNeen A and B 1s zero, the area ot these moment diagrams must equal zer~, hence
5265 + 14040 + 2160 + 6885 + lSMA + 15MB:: 0
In Fi~. b the detlection or B away tram a tangent to elastic curve at A is zero, and also detlectlon ot A away fram tangent to elastIc curve at B is zero. ThUS by moment area prinCiple, the ~oment of the moment diagrams of Fig. C about points A, or B is zero. Taking moments about point A:
or
15MA + 15MB + 28350 = 0           (1 ) ~he deflection of point A away :Tom tangent to elastic curve at B is zero, therefore the ~irst moment of the moment diagrams about point A equals zero. Hence, 5265 x 6 + 15 x 14040 + 17 x 2160 + 24 x 6855 + l5CMA + 30G Me = 0 or 150 MA + 300 ME + 444600 ,. 0   (2) solving equations (1) and (2), e obtain MA =  816 in. lb. ME =  1074 in. lbs. With the end moments known, the ~etlec~ion or slope of any ~oint on :he elast~c curve between A and B can be found by use of the 2 princ~ples of area :n.oments.
~ L: 0 
              (E)
A7 33
A7.l6
Xla~tic
Ir the deflect~on of several or sll the jOints of a trussed structure are required, the method of elastic ~elghts ~y save considerable ::~e over the method of virtual work used in previous articles of this chapter. !he illethod in general consists of finding the illagnitude and location of the elastic ve i gnt for each member of a truss due to a strain from a given truss loading or condition and applying these elastic weights as concentrated loads on an imaginary beam. The bending aoment on this imaginary beam due to this elastic loading equals numerically the deflection of the given truss structure. Consider the tr~ss of dia;ram (1) of Fig. A7.49. Diagram (2) shows the deflection curve for the truss for a ~L shortening at member be, all ether members considered rigid. This deflection diagram can be determined by the virtual work ex pression 6 = u~L. Thus for deflection of jOint O. apply a uhit vertical load acting down at joint O. The stress m in bar hc due to this unit load = 2 2P = 4P. Therefore
1M c~~ ,yf1J\
,'l. 0\c
I
90 ~;.' . ;Jl
,
,/1'1\. <ltokih
I
., I
(I) g
rrr :
6 Panels @ p
Bar be
u6L _
"F" 
sr
4P6~L.L Shortening
I.!.. uLb 3r c
~lection Diagram
(2)
.: . Lbc
~~entCurve
Bar ck
6L(d+6p)
6L
f. rl,
SL, djip t
t
L"'[':'"
1M. d+2p of
Member ck (6)
"'Ela~tiC Loads on
:3
Rigid Arm
r
3r
t:. L! d+6W
6pr~ l:.L(d+6p)
The deflection at other = lower chord jOints could 3r be round in a si~ilar ~anner by placing a unit load at these joints. Diagram (2) shows the resu~ting deflection curve. 7hls diagram is ~lalnly the influence line for stress in bar be multiplied by ~Lbc' Diagram (3) shows an imaglnarJ heam loaded ~ith an elastic load 6Lbc acting along a verti4P
~iiiz:
Fig. A7.49 Table A 7.7 Equations for Elastic Weights Elastic Weight for Chord Members (See Member ab) Lower Chord Upper Chord
w=~L h
cal line thru joint O. the moment center for obtaining the stress in bar be. The beam reactions ~cr this elastic loading are also given. Diagram (4) shows the bea:n bending moment diagram due to the elastic load at point O. I~ is noticed that this moment diagram is identical to the deflection diagram for the truss as shown in
dt agram (2).
tab
&1:0""i t
elL w=I
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 corresponding to the truss ~ower chord, the bendi~g ::noment on this Imagfnar'y beam. ,/.111 equal to the true tr~ss deflection. Diagram 5. 6 and 7 of Fig. A7.49 gives a si~ilar study and the resul~s for a ~ lengthenlng of ::nember CK. The stress moment center :or this ciagonal ~ember lies at point O~ which 11es outs1~e the truss. The elastic Neight ~ at rl point 0 1 can be replaced by an equivalent system at points 0 and ~ on the imaginary beam as shown in Diagram (6). These elastic loads prOduce a bending ~oment diagr~ (Diagram 7) identical to the deflection dia~am of diagram (5). Table A7.7 gives a s~~J of the equations for the elastic .vefght.s of truss chord. and 'Neb ~embers together Nith their location and sign.
/y\/\ r
w=r
0
j,L
tab
~\t
i/NJ\I~ .~~' I aL
~
w=
J.L r
The moment center 0 of a chord member is the intersection of the other t ..... o members cut by the section used in determining the load in that member by the method of moments The Signor the elastic weight w for a. chord member is pj.ua if it tends to produce downwardder Iec t t on of its point of application. Thus for a simple truss compression in top chord or tension in botton chord produces downward or positive elastic weight
A 7 .34
tt
rmn=rry
DEFLECTIONS
OF
STRUCTURES
1000
wO 6L r ...... __ 1,_k:::.._____
/<'_:'~  ~/"</\~
,i'''l,e,
~ ~
'\
~( 000
~(.5630
~5630;D
1000
W"~:i;P.
bI
CI
d~ '2000
...<iaI>...e
i5000
a
b
Id
FIg. A
p. bw=~=2.L
bJ,Sln 9 r:, a AL "L Q=a w =~=;'1
I\&i\l\ \Q i
P=Q=T
r.
6L
fQ
~
P Q 6L r
this step are given in the figure, the stresses being w rl tten ad~acent to each member. The next step or steps is to compute the 3ember elastic weights, their location and their sense or dlrection. Tables A7.8 and A7.9 gives these ca.Lculations. Table ..1.7:8
Chord Member
E1as~ic
Weigbts
Kellber Length L
Area A
L~'
For a truas diagonal member the elastic weights P & Q have opposite signs and are assumed to be directed toward each other or away according as the member Is in compressian or tension. In t1g. a, P is greater than Q and P is located at the end at the diagonal nearest the moment center O. Downward elastic weights are plus.
TABLE J,.7. 7
AB
~i, , ,. I
ee
cd
, I ,
~ ~L IA.rJI ~:29z106 I r
1
,
i Elastic
I
l , ~~L ,
.OO~465
,.
ac
ao so ao ac ac
.2172 _5630 .2172 _7500 .2172 _5630 .U98 3120 1~~98 .1198 6870 . 1198 3120
".
, c ,
I
I
A
C
I" "
(CONTINUED)
Table A7.9
.eb Mesber
Xl.ati~ lfe~ght5.
~
P, I
d
<
Ai~l~
br d
<, " a 
on
J
PoQ'"
~L
. ,.
Bo
Cd De
. ,I
" } "
'L
w.d
'1
I
I
.il
'1
., " I :1 " I
~!
~I
II
Q 
A'
,c c,
28.25 .2421_5880 ._.0238112.75 .146 47101.0314 .... .1461_~3!>51_.01571 .. " .093 11771 .0123 " " " .09311177 .0123 " .146 _23.55 _.01.57 " " .146. 4710 .0314 " " .242 _.5380 _.0236 " "
a :12.75 . 0018~ .00246 :1 " .00123 " .000955 " C .00096.5 " .00123 " .00246 " " .00185
, ,
A
e, e,
'""?
p.'" OL
ho
Q Q=8L
d
Fig. A7.51 shows the elastic weights obtained tram Tables 8 and 9 applied to an tmagtna ry beam whose span equals that at the given truss. These elastic weights are the algebraic sum ot the elastic weights acting at each truss jorrrt
co "' 0
~
The elastic weight P acts at toot at vertical and downward it vertical Is in tension. Q. acta opposite to P at tar end at chord member cut by index section 11 used in tlnding stress in ab by method at sections.
" 13
0 0
;i
0
c:
r0
c~
g g
b
:3
0
g
0
c
;i
0 0
cOJ
r.
13 0
0
co
~
co
0 0
Fig. A7.5l
a
.17.17
1 1 1 1
B
The method at elastic weights as applied to truss detlection can be best explained by the solution at several Simple typical trusses. Example Problem 38 Fig. A7.50 shows a simply supported t~JSs symmetrically loaded. Since the axial detormationa in all the members must be tound, the tirst step is to tind the loads in all the membars due to the given loading. The results ot
Ra
J' I
I'
The detlection at any joint equals the bending moment on the imaginary beam of Fig. A7.51. Den. at J,. = (.005412 + .00185)15 = .007252 x 15 = .109" Den. at b = .109 + (.007262  .000507)15 .109 + .006755 x 15 = .209" Den. at B .209 + (.006755  .002347)15 = .209 +
.004408 x 15
.275
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 member. Table A7.l0 gives the calculation of member elastic weights. Fig. A7.53 shows the imaginary Fig. A7.54 beam loaded with the elastic weights from Table A7.l0. The detlections are equal numerically to Tabb A7.11 the bending moments on .. t his beam.
0b = .01855
x 25 = .465"
.053 (~in Bar Bb) = .518" x 50  .00387 x 25  .833" .031 (~ of ecl = .864 x 75  .00387 x 50  .00623 x
!
Ilellber
6 L
.O82~
,
15.0 17.17 20.0
... .ll ,
.00550 .00320 .00.480 .00322
Joint
....
U
25 = 1.03"
c . 091 d .091 e .083
..
.n
18
BE
BC
.056
.096 _.069 06
CD
21.16
20.0 17.17 15.00
.0738 .113
_.015 .108
~i:i~
....."
"
C
J
D
'=,
~o
~e:::,' ~
,~.,,~lI'
.061
BG
r
Q
.061
.061
.066
.066'
.061
BCD
lfeiRDer
AD BE
6L
6@ 25" 1S0"
'1
A!'1
JOin
ll
Joint
B
"
Fig. A7.52
tc
Table A7.10 Elzstic Weight Chord lteaber
AD
be
Ke~erB
H
061 061 066 .063 _.091
... =il ,
Joint
b b
.. ....
ID
El
CJ JD
.0432 9.60 .072 9.60 _.063:1 9.27 .085 10.60 .0496 10.33 10.80 _.0254 9.27 _.0585
.?~!,5
_.00450
_.00748 _.00685 _.00802 .00134 .00480 _.000115 _.00274 _. ?~?28 _.00610
A
C J
.oraa I ~0.33
G7
'.60 9.60
r
G
9.60 .00450 11.20 .00643 10.60 .00600 11.26 .007.55 ~l?~3 _. ~0126 10.93 .00454 11.26 .00891 10.60 ,00240 .00878 1~.20 9.60 .00610
r
1
I
C J
ac
3
3
CD
CO
3 30 3.
I I
c C
JOint
b
,
I
lfellber
I
I
HI,
'1
p .. 6L
"l
Joint
r2
Q .~
Fig. A7.55
"
b
i
I
.b
be eC CD
ad
Ab
I'O~~
C
0
Table A7.12
:gg~f~
hllel
PaDel Sher.r
Z Shear
.02 58
.IIloael1t
.
o
12.5 z Point
A
= = =
~ ~
ql
= 0
q
I C
=
q
I
= 0
0
=
M
q
I
= = 0
~
AD
02158
6 .01350
ileet10D
CJ
~r
N
I
E
qt1;G
Fig. A7.53
IB
YG
. . _.02
. .
~ .
04
.056 2
.07181 .07429
.8. .900
.930 .638
.508 .710
J
D I
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 with cantilever overhang on each ena.' This simFigure. Table A7.1l gives the calculation of plified truss is representative of a cantilever the elastic weights, their signs a nd points of wtng beam the ruseaage attachment points being application. Fig. A7.55 shows the imaginary
error
A7.36
DEFLECTIONS
OF
STRUCTURES
at e and The ilL deformation In each truss memoer due to the given external loading Is given on the figure. The complete truss elastic loading 'Nill be determined. vt th the elastic loading known the truss deflections from various reference lines are readily determined.
.'.
n"i
;
0
A' .M"".""",
r.:. .. . .~ ~
~,
d'
:;.. 'if' '" \~"":"''' ", <at. ~i;~,' ". ~,("p .. ~". ,;~" 0.c:. ~ ~ " tf},(~ . '01n74.~~ \072 '(', . .072 "(,"::,U74\ .U70\"~74\ .:l.."",;
Q
<
.e'
,,4
c ~~
g ,
, , ,
0
0_ ~.Oj
c' C
D'
. . :: . .. .. g s s ~ , , ,
,
Fig. A7.56
lever overhang portion of the truss r eIat t ve to support points e and e l Since the cantilever :Jor1:iOn is not fixed 2. t e since the res tra i rrt Is ceter.nined by the truss between e' and e, this fact must be taken into 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 e and e' . This elastic reaction in acting in the reverse direction Is therefore applied as a ::"oad to the imaginary beam between e and a as shown 1 Fig. A7.59.
N N
~
~,
"
"
N
~
Fig. A7.57
~.
I
0'
01
e'
,
,
~.
Table A7 .13
E'
R:.OO:l12
dL
_.070
In finding deflections this overhang elastically loaded portion 15 considered as f1xed at be .074 .0037 B a and tree at e. The bending moment at any po in t I Be .060 to .0030 0 , on this beam equals the magnitude of the 'ler~ica1 od .070 20 .0035 C , detlection at tnat point. 20 .0028 .~~6 d ~ .074 20 .0037 0 _.056 IlK Thus to tind the cer l ect i on of the truss er 20 .0028 0 of .012 20 .0036 E (Joint a) we tind the bending moment at pOint a .0""8 20 .0024 of the imaginary beam or ?lg. A7.59. Hence Ilastic Weights of Web ~ .oera deflection at a ; Z!1a (calling counter:loc'K, dL '1 P ~ I apply '2 Q .<\1, apply wise pos i t i ve ) , ; ( .01922  .00312 ) 80 + .00248 X rl at '2 70 + , joint .00333 X 60 .. . 00239 X 50 + .00468 X 40 + Jo1ut " .00234 X + .00543 X 20  .00149 X 10 ; 2.13" 30 .031 8.95 .003406 I .00346 8.95 .080 8.95 A 8.95 00895 upward. Due to symmetry of 'truss and leading o _.063 8.95 _.00702 I 8.95 .00702 B the truss we know the slope of the elastic curve , Be .075 8.95 ~.00838 B 8.95 .00838 e _.060 8.95 _.00670 c 8.95 ,00670 'C C at the center line ot the truss Is horizontal 0 r 8.95 , o;~ 8.95 .00781 ! C O~L81 zero. Thus to find the deflection of any potnt dO .0605 8.905 d 8.95 .00728 0 . 00 72 8 1 I With .076 8.95 ~. 008.50 D 8.95 .00850 0 reference to Joint f we can make use ot the o. .09' 12.0 12.0 .00792 .00792 I 0 I !f deflection principle of the moment area method. .02. 12.0 .0020 12.0 I Thus In Fig. A7.60 the vertical ce ri ect i on of a. paint for example joint a, relative to Joint t equals the moment of all elastic loads bet....een Table A7.l3 gives the calculations tor the nagnltude 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 algeN N N 0 0 0 0 0 0 0 braically the elastic weights tor each jOint t:ro~ 0, 0 0 " ~I ~I Table A7.13 the beam elastic loading as shown In , 'I " 9 , o. I, Fixed Freete 0 d C c B b A ar Fig. A7.57 Is obtained. Let it first be required to determtne the R", .00312 (From Span ee') vertical deflection at Joint r relative to the 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 calculabetween the supports e a d e' It Is only necest i ons ) Previously the deflection of t with re sary to consider the elastic weight loading bespact to e was found to be .0586" . Thus detween 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 " N N N N 0 0 0 0 0 0 0 0 0 Hence deflection at t ;  .00312 x 30 + 0 0 0 0 0 0 ~ ~ OJ " " .00232 X 15 ; .0586" (upward since >n.inus Is up) 'I , , i , I , ! ", Find the vertical deflection of the cantil E e d 0 C 0 b B A a
AD
I ,
=&
r
.0040 .0035
.oao
Apply at Joint
A
ac
20 20
..
..... .. ,. ..
""
,00":
~.0020
, ,
~I
~ ~
~
~
~ ~
~ ~
~
~
~
~
 "',
~ ~
~
~I~ ,
~I
~
~
~ ~
~ ~
~ ~
~ ~ ~
~
~
!<t
Fig. A7.50
________...;,
. ' '.:~.:~li~~.~;.,r:.~~F.,~:i!!::_
A7.37
For this problem we need only t~ consider the elastic loads betN8en pOints band d as
loads on a simple beam supported at band d (See Fig. A7.61) ~he deflection at C with respect to a line bd of the deflected truss
equals the bending moment at point c for the
(7) For the truss in Fig. A7.66 calculate the deflection of jOint C along the direction CEo E = 30.000,000 psi.
C .20,000*
S'
B
= .004743
d
8'
A
0
A
~
2
B
S'
= =
~
~=.004743
~
Fig. A7.61
~l ~l ~1
~ M N
'7'
Fig. A7.67
Rb=.004667
A1. 18
Prob 1ems
Fig. A7.62
~
45'
C
A~
O"
(8) Far the truss in Fig. A7 .67, tind the vertical and horizontal displacement of jOints C and D. Take area of all members carrying tension 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 the horizontal displacement of points C and B. E = 28.000,000 psi.
\.
3~'
~ to' 1>\)(): 3~
t A
J
0
G
H
A
+20"
B
J..
~
IB Ie
2
5000 5000
::i
::'F
,000< 30"
T
I
I I ~rr.,r.,,
1000
1000
500*
i10" T
G414~E
5060
5000
e
Fig. A7.70
D 4 @20 = 80"
(1) Find vertical and horizontal deflection of jOint B for the structure in Fig. A7.62. Area ot AS = 0.2 sq. in. and Be = 0.3. E =
10,000,000 psi.
(2) For the truss in Fig. A7.63, calculate the vertical deflection of jOint C. Use AE tor each member equal to 2 x 10 7 (3) For the truss of Fig. A7.64 determine the horizontal deflection of jOint E. Area of each truss member = 1 sq. in., E = 10,000,000 psi. (4) Determine the vertical deflection of joint E of the truss in Fig. A7.64 (5) Determine the deflection at joint D normal to a line jOining joint CE or the truss in Fig. A7.64.
b
(10) For the truss in Fig. A7.69, rind the vertical deflection of jOint D. Depth at tr~ss = 180". l..tIdth ot each panel is 180". The area of each tr~ss member is indIcated by the number on each bar in the figure. E = 30.000,000 pSi. Also calculate the angular rotation at bar DE. (11) For the truss in Fig. A7.70, calculate the vertical a.~d horizontal displacement of joints A and B. Assume the crosssectional area tor members in tension as 1 sq. in. each and those in compression as 2 sq. In. E = 10,300,000 pat , (12) For the truss in Fig. A7.70 calculate the angular rotation at member AS under the given truss loading.
1000* 5000
"
c
d
~9' ,
12'
I
I
10
' T
10'
i is'1
B iii
Fig. A7.72
Fig. A7.71
e ~a'
I
.a'
SO,OOQt
1.,
Fig. A7.65
(6) Calculate the vertical displacement of joint C tor the truss in Fig. A7.65 due to the load at joint B. tlembers a, b, c and h have areas of 20 sq. in. each. Members d , e , t , g and i have areas ot 2 sq. in. each. E = 30, 000, 000 psi.
(13) For the beam in Fig. A7.71 deter.nlne the deflection at points A and B using method of elastic 'N8ights. Also determine the slopes ot the elastic curve at these points. Take E = 1,000,000 pSi and I = 1296 In.~ (14) For the beam in Fig. A7.72 find the deflection at points A and E. Also the slope at the elastic curve at point C. Assume EI equals to 5.000.000 Ib in. sq.
A7.38 500Jl
DEFLECTIONS
OF
STRUCTURES
100,* 10".1
A ,
t=25"
5"
, 5"
"
j.<a' t ;r.~1
2"1
50"
Fig. A7.81
~A
lOOOf
'I
Fig. A7.B2
(23) In Fig. A7.81 find the vertical ~ove (15) Fig. A7.73 illustrates the airloads on 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#
lOOt 100 "'3" 1"3"
F ...
100
3 .,
F91t ;100
100*
100
9" + 9 " B
j
A
B 3"
Fig. A7.83
Fig. A7.74
Fig. A7.75
(16) For the beam or Fig. A7.74 determine the dert.ectaons at paints C and D in 'terns or EI which is constant. Also determine slopes ot the 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.
V Fig. A7.B4
L. o r a
Front View
Side View
BEMA
Fig. /07.77
(18) For the loaded beam In Fig. A7.76 deterame the value at the t Ixed end moments MA and MB. EI ts constant. Also tind the detlection at points C and D in terms at EI. (19) In Fig. A7.77 determine the magnitude at the tixed end moment I1A and the simple support RB.
p
[ aij] {qj} compute the strain energy in the' truss at Fig. A7.63 (Problem 2). The ~e~ber 11\\ ~ A '11 flexibility coefficient fer a member under Flg.A7.78 Fig.A7.79 Fig.A7.80 uniform axial load is L/AE (see Fig. A7.35a). An". U = 22.4 lb. In. (20) In Fig. A7.78 EI is constant throughout. (28) Using matrix equation (23) compu~e Calculate the vertical deflection and the anguthe strain energy in the beam at Fig. A7.71. lar rotation ot point A. Note: the choice of generalized forces should (21) For the curved beam In Fig. A7.79 rind be made so as to permit computation at the the vertical deflection and the angular rotamember fleXibility coef~iclents by ~he equations tion at pOint A. Take EI as constant. of p . A7.l9. Ans. U = 3533 lb. in. (22) For the loaded curved beam at Fig. (29) Resolve the problem of example 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 I doubles at point "2 and doubles again at "3". conatant , (HeaViest sect10n at builtin end.)
IF
(25) The cantilever beam of Fig. A7.53 is loaded normal to the plane of the paper by the two loads of lOCI each as shown. Find the deflection at point A no~l to the plane of the paper by the method of virtual work. The rectangular moment or inertia tor the t~be 1s 0.0277 in. E = 29,000,000. (26) The cantilever landing gear strut in Fig. A7.84 is SUbjected to the load or 500# in the drag direction at point A and also a torsional moment of 2000 in. lb. at A as shown. Determine the displacement at point A in the drag direction. The tube size tor portion CB is 2~ .083 and tor portion EA, 2~.065 round ~~be. Material is steel with E = 29.000,000 psi. (27) Using the matrix equation 2U
tJA
I RI
L qi I
<:' ~ ,.
 
'3 :./":
A7.39
(30) For the truss of Fig. A7 85 determine 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.
A
1.0
1.0
t:.051"
;
1.0
10"
2.0 1.5
10"
r, L
Fig. A7.85
Jrn
nl
Fig. A7.86
Answer.
[3,nJ. E
1
~ine
["
44.67 44.67
38.0 33.0
~.~
38.0
(34) Find the influence coefficients relating deflections at points land 2 of the simply supported be~ of Fig. A7 87. Use matrix methods.
33.00
(31) For the truss ot problem (30) ceterwhich of the following two loading conditions produces the greatest deflection of point 4, (All loads in pounds},
)10.
(j)
@+ ( AE:3XI~~lbS
AE:8 x 10 6 lbs
Fig. A7.87
Condition
1
p.
1000 300 500 700
p ,
800
400
400 600
Ans.
(32) Determine the matrix of influence coefficients relating drag load (positive aft), braking torque (positive nose up) and moment in the VS ~lane (positive right wing down) as applied to the free end of the gear strut assembly ot Problem 26.
7.10~ 12.~
Note to student: It will be highly instructive to rework 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.
Answer.
f2S00
I
50.7
Lo
50.7
6.97
References for
Cr~pters
A7 J AB.
the 1eflection of the load applied to the cantilever panel of F~g. A7 86. (Assume the web does not buckle). Use matrix notation . .~s. 0 = 7.94 x lO~ inches.
(33) F~nd
"Advanced Mechanics of ~~terialsn, F. Seely and J. O. Smith J 2nd Ed., John ~ileYJ New York. "Advanced Strength of Materials", J. P. Den Hartog J McGraWHill, N. Y. "Theory of tna st ic rtv", S. T1m.oShenk0 J rrccraw;
arn ,
N. Y.
A7.40
DEFLECTIONS
OF
STRUCTURES
APPLICATIO~~
Langefors, B., ~~alySis of ~las~lc Structur~5 "Introduction to ~he Study of Aircraft Vibration Matrix Transformation with Spec ~al ?,ega.r1 ~ and ~lutter", R. Scanlan and R. Rosenbaum, Sere1 Monocoque Structur~ .journ , 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 Benscoter, S. U., "The Partitioning of ~atrices !E Structural AnalySTS", .rourn. of Appl , hechs ,
Vol. 15, 1948.
Falke~~einer, H. Systematic )~alysls o~ Redundant 21astic Struct:.rres 'Jy ::ea:;s of ~aL:rix Ca:c'.!lus, .Journ , or .Aero, 3ci:: 20, ::''.753.
L. and Lansing, ~'J A Method for Reducing the Analysis of Ca~p1ex RedUr..~tructUIes tc ~ Routine Procedure, Jo~n. of Aerc. S:~.
~ehle,
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 DCa.
CHAPTER As
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 relationships between displacements must be written to permit a solution. The "Theory of Slastlclty" shows that all structures are statically indeterminate whenanalyzed in amut e detail. The engineer however, Is otten able to ~ke a number of assumptions and coarse approximations which render the problem determinate. In addition, auxiliary aids are available such as the Engineering
Theory of Bending
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 "ntimes redundant".
eOROLLARY
(MJ) and
the constantshear
flow rules of thumb (q = T/2A) (see Chaps. A5, A6 and A!3 through AIS). While these latter are certainly not laws of "statics", the engineer employs them often enough so that problems in which they are used to obtain stress distributions are often thought of as being
In an ntimes redundant structure the magnitude of the forces in n members may be assigned arbitrarily while establiShing stresses in equilibrium 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.
"cetermnace" . It is frequently the case in aircraft structural arzlysis that, in view of the requirements for efficient design, one cannot obtain 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 A7 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.
_A 20004;'
~.
c: c:
D
\.00"*
<a)
~12l
.707X
(e)
Ntto:
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 deformations, as, for example, the beamcolumn, a ~ember carrying bending and axial loads at the same time.
AB.! The Statically Indeterminate Problem.
Only the system (b) is actually required to eqUilibrate the external loads (corresponding to X = 0). Note that the system (c) has zero external resultant.
[[j Of all the pOSSible stress (foroe) distributions satiSfying static equilibrium the one correct solution is that one which results rnkinematically pOSSible strains (displacements), i.e. retains continuity of the structure.
ThUS, for example, there are an infinite number of bending moment distributions satisfying 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.
Several characteristics (and interpretations thereof) of the statically indeterminate prOblem may be pointed out. rhese character is
AS.!
AB.2
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/.
Me
undetermined by statics.
Example PrOblem A By way of illustration, the proble~ posed by Fig. A8.2 'NaS carried to completion. The bending moment was given by (x, YJ ~ measured tram the left ends of the tr~ee beam dlvls~Qns)
M
eOROLLARY
bitrarily ~hile establishing equilibrium with the external loads, relative ~ovements of the elements will result, violating continuity at n points. n zeroresultant stress (force) distributions 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.
O<;;;<L
Then
1
U' 2
f
o
M' dx
EI
'2EI
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 beam over three supports (Fig. A8.2). One ot the reactions cannot be obtained by statics.
2EI
L
+
2~I EX'(L 
g) 'ds
....El = constant
t5002Rx
&
lntegra~ 51'S!:,
(see
Fig. AS.2
oR x
au
A singly redundant beam with one reaction given an arbitrary 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 determined from statics. The strain energy U ~y then be written as a function of R x , i.e., U = r (Rx). Next form
+ _~ '1
L , fR ., \ 'X (L  0,
o
'I
"
'"'
This is the deflection at rtx due to R x o But this ~ust also be zero.Slnce the support is rigid. Hence
~. a              aR x
(1)
.. (500 +
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.........
..
AB.3
=
1500
lbs.
=
93.8 lbs.,
au ai1=
R
Example Problem E Determine the redundant fixed end moments for the beam of Fig. A8.2(a).
o =l1L J~
Fig. AB.2a A doubly redundant beam with two reactions given
two arbitrary values.
3~)
+ PL 
dx
~+2PL'\JL
3L
oX(1  3L)dx
Solution: The redundant end moments were designated as ML and MR for the left and right beam ends respectively and were taken positive as shawn. The moment equations for the two beam portions (x trom left end, y from right) were
~+2PL,\
+!lR
2 Iy
dY+ 3L
'\
~ J2~.
3L
, 3L
'\
dy
O'l1L
\dx+ Je
3L
~ + 2PL 3L
J\.
rr
3L
c
dx
11='\+
3L
O<x<L
O<Y<2L
+..l.
2EI
~   PL
Differentiati~g ander the integral Sign (see remarks on p. A78)
2 9
The Theorem of Least Work may be applied to the problem of determining redundant member forces within a statically indeterminate structure. ThUS, inanfitDn8S 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
au ax
ay =
o
0
__      (2)
50
+
au
':;=".l.BC
.~.
j
1
(lOOe
\:
+2EI~e
etc.
~lth
In words, "the rate of change of strain eneT~Y respect to the redundant forces is zero".
1 +
60
o (50,000):ld6'
Eqs , (2), like eq, (1) I are statements of the Theorem of Least ~ork. They provide n equations
2EIC D
for the n:tlmes redUndant structure. The simultaneous solution of these equations yields the desired solution of the problem. Example Problem C The can~11ever beam and cable system of Fig. A8.3(a) 1s singly redundant. Find the member loadings by use of the Least ~ork Theorem.
Obviously there was no need to ccnsider the energy in CD as its loading did not depend upon X and hence could not enter the problem. Differentiatln~ under the integral sign
au
58.3 X
~.U3
ax
so v
58.3
1x
0
50
a
ex
(b)
Fig. AS.3 A singly redundant structure with one member force given an arbitrary value (X).
Solution: The tensile load in the cable was treated as the redundant load and was glvzn the symbol X (Fig. A8.3(b)). The strain energies conSidered were those of flexure in portions AC, CD and BC and that of tension in the cab~e AB. ~ergies due to axial forces in the beam ~ort ions were considered negligible. The bending moment in Be (origin at B) Nas
= 21.44 X 101!!
ELse
Putting
AAB 0.025 in a
!'lac
= (1000 
5~~3
XJ
x
10.0 in"
gave
InAC, (origlnatA):
MAC
= 58.3 X Y
X
50
613 lts.
#
Then
The strain
ener~J
was therefore
~e
M. .',.c
[1000
vv.,,)
30
50
52:3
613
.y
520 y
AS.5
au ax
R'P
EI x
R'
J60
0
sin
Q d Q
60'
0
(1  cos g) sin g d g
""r
""
R' X f90
sin
II
g d g
6Qo
0
101
Fig. AS.3
.... (d)
 2Ei'
R'P )90
sin g d g
60'
Solution: The axial lead in the floor was taken to be the redundant (since the floor was assumed rigid, this could have been thou&~t of as a redundant floor reaction from fixed supports). The loading is shown in Fig. A8.3(d). The bending moment distribution 'NaS
11
~ locOS
II
60 0
g sin g d g
Q +
~ sin
] 60'
90 '
II
Q d Q
Evaluating,
0(g<60'
+gP
3 (R R') AETI
' R) (R rrTI:
s
S
P ccs Q
X sin Q
X sin
Therefore
U =
zir E sin Q PR
R 90'
+
2n
3P
' +R) (R EI AE
1. 50
60'
(1  cos
Q)J'
Rdg
zir
Rdg
+ X
sin
Example Problem E The portal frame of Fig. A8.3(e) is three times redundant. Set up the simultaneous equations in the redundant forces. The relative ~ending stiffnesses of the segnents are given on the figure. Solution:
1 000lf
c
2.25
2.25
1
50"
D...1
RdG
Fig. AS.3e
~E
90 '
[x sin
QJlI RdG
J60
The redundant forces selected were the bendlng'moment,the transverse 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 to applied loads and due to redundant forces
AS.6
acting individually (it being easier to compute the loadings in this fashion). The complete loading ,v,as obtalr.ed Jy superposition.
.1[50'
;0 Sln g ..
~ .. 50'1(1.
'..:>0
"""'" "we...
..
. I"
LZ5
Fig. AB.3f
50,000
Fig. AS.3g
~nte~31s
the equations
9.682V
7.459T
763.1V
484. a!'
= 111.15
x 10 3
50V
Q.6.
1
s
T s'
~
Fig. A8.3h
lT
lOOT
Fl . AB.31
MBc
 SOT (1  cos g)
MCD = 1000
Then since
',.Jhile the Theorem of Least ',.jork may be nace the basis of red~dant problem analYSiS, its direct application by the calculus, as in Art. A82.1 and A82.2, is often impractical. ~or t~e majority of problems the work is facilitated if carried out by the techniques of the Method of DummyUnit Loads. The follOWing derivation is for a doubly redundant tMISS struct~e. The extension te a more general ntimes redundant structure, in which ether loadings in addition to axial (flexure, t crst c.i and Shear) are present, is indicated later. Consi~er the doubly redundant truss of Fig. A8.4(a). It may be made statically dete~inate by "cutting" two members such as the diagonals indicated. Application of the external loads to this determinate ("cut") structure gives a load distribution, ~S", computed by satisfying static equilibrium. At this time discontinuities appear at the cuts "xn and "yn due to the strains developed.
l J Il ' dS d U = :2 ~an
aR=;j1i
au
au
au
aT
0,
x
, IP
,P,
y
..
one haS,
au
so
aM =
III
2.50
vsl
ds
S loads
) 0
Fig. AS.4a
ux
Uy loads
Fig. AB.4c
.j .
1.50
,so J .(10009'
Q.=
To compute these and subsequent displacements the Method of DummyUnit Loads may be used (Art A77). For this purpose Virtual loads are placed alternately at the x and y cuts as in Figs. A8.4(b) and (c). From the dummyunit load equations
AB.7
z S~t}
Z SuyL
(3)
For continuity these net relat~ve displacements must be zero. Equating the above expreSSions each to zero, and rearranging, gives the simultaneous equations
X
AE
llL UX + Y Z
u and u are the unitredundant stress disx y tributions as indicated in Figs. A84b, c. The subscript wow indicates these relative dlsplace~ents occur in the determinate (Wcut W) structure with the Wor i gl na l Wstress distribution. It is now desired to close up the discontinuities by application ot redundant loads X and Y to the x and y cuts, respectively, as In Fig. A84(d). Load X causes a stress distribution XUx and, likewise, Y causes a distribution YUy. The relative displacement at cut x due to redundant load X is given Fig. AB.4d by (6 is read wdisplacexx ment at x due to X W ).
AE
uxuyL AE
AE
~  z "",C AE
AE
  
(4)
ICE
Eqs. (4) are two simultaneous equations in the two unknowns X and Y. Upon solution for X and Y the true stress distribution may be computed as
(5)
For a structure which is only singly redundant, eqs. (4) and (5) are applied by setting Y = a giVing
xZ
or, Simply, SUxL
Z
ux aL = _ Z SuxL AE AE
Ai:
'L
               (4a)
Z ...L
AE
= Z XUx uyL AE
Similarly the load Y causes displacements at the cuts y and x given respectively by
e
YY
and,
            (5a)
~ Z YU y ' uyL
AE
=Y
Z uyaL
AE
Example Problem #1
and
Z YU1 llX L = Y Z uyuxL
AE
AE
Now the net relative displacement at each cut under the slmultar.eous action of the three stress systems 5, Xu and Y.u y is
x
T~a O'~g~g
30 
po.o*
1000*
.707
+ 6 + 6 = Z SUxL + X Z uxuxL xo xx xy AE AE
+ Y Z u y uxL
AE
and
Z SuyL + X Z uxuytAE
AE
+ Y Z
uyuyL AE
Fig. A8.S shows a single bay pin connected truss. The truss is statically deter.nir~te With respect to exter~l reactions, but staticaily indeterminate with respect to internal member loads, since at any joint there are 3 unknowns with only ~NO equations or statics available for a concurrent force system. The truss is therefore redundant to the first degree. The general procedure for solution is to make the truss statically deter~inate by cutting one or the ~embers;on Fig. A8.6, member bc has been selected as the redundant ~ember, and it 1s cut as shown. The member stresses S for the truss ot
A8 8
Fig. AB.6 are then deter.n1ned, the results being 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 redundant member be, and the loads in all the m.embers due to this unit load are calculated The results are recorded on the figure and also in Table AS.l under the head of u stresses. The solution for the redundant load X in the redundant aemoer be is given at the bottom of Table AS.!. The true load in any member equals the S stress plus X times its u stress.
bble U.1
~/ ~/
V'...;
Iiii'
0 1,000
0
Fig. AS. 10 u loads
Ca:c~l~tiJn
Fig. AS.9
S loads
1:1
""r
L
30 30 30 30 42.4 42.4
1 1 1 1 2 1.'
s
_1000
0 0 0
14140
u
_.107 707 _.707 _.707 1.0 1.0
';t
0 21210 0 0 0 40000
T
15
u 2L
u.
.d
d.
"
15
True Stress s + Xu
.60'
_559
3"
395 3
21.2
28.3
'55
Calculations of the deflections under load of a redundant structure are made by application of the methods of Chapter A7. Since, however, there are certain piUalls as regards symbols and also some important special techniques, the 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 Arts. AS. 10 et. eeq.}.
X true X
1~c1
."
hi
A. ~
t"r
aUlO 109.S
_5~9.
Find the horizontal ~ovement of :Joint ftd" of Example Problem 1# 1 under the action ef tne load applied there. Solution: The equation used to find the deflection is Sq. (18) of Chapter A7. ',iritten for application tv t:uss deflections it is (see Example Problem 13, p. A7.l1)
Example Pr'oo.Iem t 2 Fig. AS.S shows a Singly redundant 3member frame. Ftnd the member loadings. Member areas are shown on' the r rgure ,
z~ AE
           (A)
.4
0 ho"",
Fig. A8.8
Solution: Member ex: was selected as the redundant and. was cut in figuring the Sloads, as in Fig. AB.9. Fig. AS.IO shows the uload calculation. The table completes the calcu
Now for a deflecticn calculation the symbols '8' and "u" ;nust be carefully reinterpreted from their ~eanings in the redundant stress catcutat i cn.' For a deflection calculat1Cml the symbols of eq. (A) , above, :nean: "sLoacs " are the true loads of the redundant structure due to application of the real external leading; "uLoads " are tae loads due to a dummyuntt (virtual) load applied at the external ;Joint where the deflection is desired and in the directi on of the desired deflection.
Thus , the '8 loads" for use in ;;q. (A) are the true stresses (the sclution] of Example Problem It L
iaercn,
Mem.
AO
BO
S
0
""A
0
SuL
u 2L
""A
1059
True Load
=
s ... X
141.4 0.2
1.224
933
500
CO
1. 000
Z Z 8uL
X
The "uloads" represent add1 t i onaj infar:na tion which would, in general, appear to necessl tate another redundant stress calculation. As will be seen, such is rorcunate Iv not tr.e case. In the present pr'obIem the dummyurn t load is applied ldentically as is the lOOOJll real load and hence the uloads are simply equal to the "Sloads" properly scaled .jmvn. The following table completes the calculation.

A
ZuliL
=274.1
1':)
A8
Mem. ab
bd
S'
3.5 805 3'5 3.5 55' 855
u"
.395 .605 .395 .395 .559 .855
SuLI A
Mem.
30 30 30 30 42.4 42.4 1 1 1 1
2
SuL
L
A
S'
335.5 625.6 274.1
4,680
AO
A
3. 355x10 5 3. 128xI0 5
141. 4
.2 .2
10,980
BO
4,680 4,680 6,620 20,660 52.300
100
200
1
de ca cb ad
co
.4
o
O.227xl0 5
1.5
:. 5
= 22,700
E
=52,300 E
true loads from Exa'Ilple Problem #2 By way of Qemonatration another set of uloads, called u l , were found for this same problem, this time by cutting aember OA. The corresponding calculations follow:
si:lply l/lOOOth of the "Sloads" sfnce the dummyunit load is applied exactly as is the 10001 real load. Example Problem *2A. Find the horizontal deflection of point 0 of Example Problem 12 under application of the vertical 1000* load shown in Fig. AS.S. Solution; To compute the deflection use
rt:
0 1. 808x10 5 1. 583xlO 5 O.22Sx10 5
Su'L
The results are identical (allowing for roundof! errors). Again the symbols "S" and "u ll are to be reinter?reted for a deflection calculation as explained above in Example Problem IIA. The "ScLcada" are now the "true Loads" coraout ed in Example Problem. #2, above. The lI u_l oadS" are loads due to placir~ a dummyunit 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 u10ads in a de~lection calculation anys6t of stresses~loads) i~ static equllT'5ri'i:iinNith the dum.lilyunit 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 satisfactory set of uloads 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 dummyunit loads deflection equation, Eq. (18) of Chapter A7, was derived (refer to p. A7.l0). It will be remembered that the deflection was shown to be equal to the work done by the internal virtual loads (uloads) 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 (uloads are, in general, indeterminate since the dummyunit load is applied at an external point ot the structure. However, we recall that, iany stress df s'trfbut 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 dummyunit load as the "applied load") differs from the correct (true) distribution only by a stress distribution having zero external resultant
(~
.'18.1).
ii  a zeroresultant stress distribution moving through a set of displacements does zero work
4
AS.lO
= U STATIC
~=o
where UsTATIC 1s a uload d1st~ibut1cn cbtained from statics in a simple "cut" structure under the action of the exterr~lly applied d~~yunit load and )~=o is the zeroresultant uload system which must be superpozed to give the true uload distribution
11Z <I x "R=o = a
It follows, therefore, that any set of ulaads in static equilibrium with the externally applied dummyunit load will do the same amount of virtual work when the structure undergoes its distortion as would a "true" set of uloads computed by an indeterminate stress calculation. That is,
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).
,
Member t,
"

"y
AB
Be
30.5
30.
2~
750 0
.. ..
0
A '";" I .. ~
S\I,,1..
0 0 64,000 1250,000 21.6 0 0
..jL
A
0
027,000
0
0 0 0 0 0
1680
CD
40.5
'000
U50 0
0 0 0
eo
50 .25
..,
0
i 43.2
I
0 0 0
51.2
I I
.50
255
Q.E.D.
AB. 5 Trusses With Double Redundancy
C<
BE
ED
50 .s
40.1
, ,
o
0
1200 100 1
are
932 1015 190
I
I
30.5 750 0
Trusses with double redundancy are handled directly by Eqs. (4). By way of illustration, the structure ot Fig. A8.4, from which Eqs. (4) were derived, will be solved tor a loading P:l. 2000,* and P a = 1000#. 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 reepec t i ve Ly, x y
.. ..
AE
B'
0 0
27,000 0 0 0 0
0 62.
~
21. 6 1 0 0 0 0 0 0
>0,
30.7
,00<
0
0 0 0
I
!
15.41 105.01
I
!
40.25
SUbstituting from the table into Eqs. (4) gives (cammon fac~or of E diVided out)
I
{
Fig. AB.ll S loads
X Z ~ + Y Z uxUyL
A A
SUxL   Z! r
X Z Ux~'{L + y Z UrL
=  z S\~t
562,000 478,000
341.1 X + 36.6 Y
+
36.6 X
452.6 Y
So Ivtng ,
X;::
1550*
True Load
=S
XUx
YUy.
CTURES
A8.11
Example Problem 3 Fig. A8.l4 shows a structure composed at tour coplanar members supporting a 2000* load. With only two equations of statics available for the concurrent force system the structure, relative to loads in the members, is redundant to the second degree. Solution: Fig. AS.lS shows the assumed statically determinate structure; the two members CE and DE were taken as the redundants and were cut at poin~s x and y as shown. The member stresses 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*) y x tensile loads applied at the cut faces x and y. Table A8.3 gives the complete calculations for eqs , (4) and (5). The load in the redundant member CE was designated X and that in DE as Y.
Solving the two equations for X and Y, one obtains X = 521* and '[ = 416*. The true load in any member = S + XU + Yu which gave the values y x in the last :olumn of the table.
AS.6 Trusses With Multiple Redundancy.
By induction; eqs , (4) may be extended for application to trusses which are three or more times redundant. Thus for a triple redundancy,
 Z
SUxL
AE
SUyL AE
uxuyL
AE
+ Y Z
 Z
(6 )
X Z
and
uxuoeL
AE
+YZ
+gZ
uoe aL
AE 
 Z SUsL AE
(7)
~[
E 1200041 Fig. A8.14
True Stresses = S + Xu
+ bU
~/y
+200041 Fig. AS.I5
AS.7 Redundant Structures With Members Subjected to Loadlngs in Addition to Axial Forces.
::;qS. (6) are extended readily to cover problems in which flexural, torsional. and shear loadings occur. ThUS, for a three times redundant structure
Xa + Ya +;;a xx xy XS
~ ~,~:./
~V
ux loads
""."..::3 "T'
j/
1*
        (8) xa sx
+ Ya +:!a .y
Uy loads
Fig. AS. 16
Fig. AS. 17
=  so
X Z A + Y Z  A  =  Z A
Ux aL
uxuyL
SUxL
where
a
xx
= Z ux'L +
AE
f :n~dX + J tx'dX E1 OJ
:d Y
fP;
+
Substituting:
2350 X + 3039 '[
"0'
+
=a
yx
uxUyL =Z + AE
f mxmydX "SI
= 2,488,000.
Jtx~
Z _Y_ + AE
U
TABLE A8.3
If qXq~~dY
:ny'dX E1
+
Member
~I
~I
'x
'y
IsuxL x 10 XI0
.
i_
i7S:! 402.5
~
2150
Q
'L
AE BE CE
.286.51
.1 72 .280.5 .393.7
e o
0
.806 1. 1541
o
2.2S:!
c
2.488
280.5 576.5
..,
I
521
.00
     etc.
c
1. 00
c
0
o
0
D'
0
2446
312.5 3039
ue
iTOtal 1:
2.25:!
2.488
2350
and
".~'''~.c~:~~;;:?":!.,_
,..',..:;)~~~
A8.12
5xO
~ SUXL + AE
+
~xOX 1
JTtxox
0"
Solution: The shear flow in the 2heet panels ~as chosen as redundant. Secau2e 0: s~~~et~y :~e problem was only singly red~da~t. ?1g. AS.IS shows the Ux and qx loadings jue to the redandant 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 equation solved was (ref. eqs. (3)).
".
jj
qqxdxdy Gt
Dye
~ ySu L +
AE
J~ OX y+ E1
q<l.r
J OJ
T\ox
ax
Gt
(J
Ux
...... t..
~"ax
ax
Gt
,)
dy
+fJ
_ (J SU~E
where
jJ
OXd Y)
Gt
etc. and where 5, M, T, q are the real loads in the determinate structure; U x ' rnx' t x ' ~ are the unit (virt~l) loadings due to a unit load at cut x;
REAL LOADS
S  o in s ice s'trfngers
q  0
Ux  L  x in s1de stringers
VIRTUAL LOADS _ 2(xL) in
s,
{ ~  1.0
When evaluated, (note that the double integrals Simply reduce to a constant times/the panel area)
X
The redundant force(s) need not be an axial force but may be a moment, torque etc. Atter solution for the redundants, True Axial Forces =
S+ X u
C: ~; + 2itL)
.E. ( 2L
1 +
= 
(P1E')
+ y
+;; u.
x
       (9)
1 ME )
GtLa
etc. Example Problem 4 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'Gt
A
const
1
...
'!1o:bt:.lLb~..l;l
p L
x.
l1l
tj
"Clamped" I L
.j..
'h
Fig. A8.19
Fig.A8.I8
qx loadings
Fig. A8.19
AS. 13
Solution: !ha
bendin~
momants at
~cint
C in
~ember
621
L')
+ Y \v~ +
(L
as radundants yielding (when cut) the pinjOinted deter.ninate struct~e of Fig. AS.20. ~te virtual loadings were as shown in Figs. A8.21 anc. A8.22.
A
=
(2L/[ L')
'NaS ass~~ed
that
~ :: 100
L
g.
giving
'+
'<
B
X=1,tcu~1!L
Y=1" *
~L
Ux> m x loadings Fig. AS. 21
P
Uy, my loadings
Fig. AS. 22
.0645 PL
{ Y .456
PL
Then as usual
~
Note that in Fig. A8.22 the unit redundant loading 'NaS applied as a selfequilibrating set af unit couples. The real and virtual loadings were as follows: (member portion 3D J haVing no virtual loadings J was omitted. It could not enter the calculation.) I uy Member S M Ux !lly "'x
AB
2P$ 2P
x
+
rrue Moments
A8.8 Initial Stresses.
= !1
+ Xrn.
Ym y
0 Y2/L
BO
?y
In a red'.lndant structure In~tlal stresses are developed if J upon assemblYJ certain members must be forced into place because of lack of fit. In some situations intentional misfits are employed to obtain more favorable stress distributlons under load (~prestressing"). "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 to a distortion in the positive X direction, is a pOSitive 0Xi)J the modified condition tor continuity at the x cut would be (compare with the equations just preceding eqs. (4)).
10
After evaluation of the integrals and ~lylng t~~ough by L~ these become
~ulti
6y1
6yx + 6YY = 0
Then using the previous notations J the appropriate equations for the redundant forces ~e
X
_x _ _ ... f' L AE y ~ Ux,\L
rx:. U.
L
AE :E
= 
~ SUxL _
AE
Xi
1
 (10)
r'L
=
fUA~L
 Y1
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
Example Problem 6 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
If, as is sometLmes the case, the number of 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 compute the effect of these misalignments on the redundant cuts proper. ThUS, referring to the ~virtual work~ derivation of the DummyUnit load equations, {Chap. A7) one has
determinate structure at the X redundancycut due to initial imperfections (equivalent to initial strains) 6 throughout the structure.
x"
= 2.253
X 10 8 + .0lE X 10 8
= 2.488
as before, 1& the unit loading due to a x virtual load at cut x.Eq. (11) and similar expressions for the Y, ~, etc. cuts may be inserted in eqs , (10).
X = 985 Ibs.
Y =
57 Lbs ,
Then, as usual,
True stresses
~ple
Example Problem 8 Referring back to example problems numbers 3 and 6, assume that member BE is .025" too long. Determine the initial stresses if the other members are of proper length and no external load 1s applied. Solution:
Problem 7 Assume that in the structure of example problem 5 an angular misalignment occurred between members AB and CEO at Joint B such that the end at member AB had to be rotated 2.7. _ clockwise to fit upon assembly. Determine the moments developed Without external loads applied. Solution: The initial imperfection was 0Yl  2.7/57 3 =  .0471 radians The Sign was determined by noting that the original misalignment was in the negative direction of the redundant couple Y. The equations used from the previous problem were (noting the equations there had been multiplied by L8 x EI/L 3 ~ EI!L). .3716 X + .1526 Y ::: 0 { .1526 X + .8121 Y = .0471 E SolVing, X =  .0258 EI/L. Y = .0630 EI/L True tnitia1 stresses and moments were determined as usual.
To employ the soone equations as those of example prr~lem 3, the initial imperfections occurring at the same x and y cuts used there were computed, in this case due to the initial elongation at BE. Thus
0Xi = Z UxJl i = (1.564)( .025') = .0391' 0Yi = Z UyJl j = (1.729)(.025) = .0432'
Then, use of those previQusly computed coefficients in eqs. (10) gave, 2446 X + 2350 Y = .0391 E 2350 X
+
3039 Y = .0432 E
Y = 209 Ibs.
Yu
Stresses induced in redundant structures by thermal strains may be computed by application at methods presented above. The problem may be approached from the point ot view ot computing the
_ _ _ _ _ _ _ _ _ _IIlII.IIIl!II .. 3!!!l1"fiW
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 B;:rr determinate, the application ot the temperature distribution IS visualized. Relative displace30" x ments occur at the cuts,dencted by exT and 0yTo l , These displacements may be computed by the 40" C (permits Y DummyUnit Load method as shown in Art. A7.8 of Chap. A? After this calculation Is accompsliding Fig. AS. 23 Fig. A8.24 vertically) lished, the problem proceeds as for initial strains, Art. AS.S. Thus, the continuity condition at the cut gives (compare with the equa Solution: tions 1mmediately preceding eqs , (4) and assume tor simpliCity that the external loads are The structure was made determinate by cuts x and y as in Fig. A8.24. The unit loadings are absent, making axe = aye = 0) shown in Fig. A8.25.
~ rl
.6
ET(~;~
0xT + On + 0xy = 0 }
0yT + 0
yx
+ 0
yy
In a truss, the thermal strains produce relative displacements at the cuts given by the ~vlrtual work~ derivation of the DummyUnit Load equations (ret. Chap. A7, Arts. A7.7 and A7.8) as
exT = I
0yT =
U
" ,tSJ
vx loading
1#.6
~'~
I,.
Fig. A8.25
~o
Uy loading
a T dx }
           (12)
U: a Tdx
The thermal coefficient a was assumed constant. The calculation was set up in tabUlar form.
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 load distributions due to virtual loads at the 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, along the members as well as from ~ember to member. Then the final equations for the rial stresses in a doubly redundant truss become
X
, ,.t
lIll
'.11
sol
11.010 I'.UI
,._.Ill"
i UI 1.10.10" . 10'1
.,,,r
".
+ y
uXuyL
AE
AZ
=
+ Y
~U'L .!Ai;;
AE
JU f
x
uy
=9 a
0
T X lOll
a Tdx
 (13)
So Lvtng ,
a Tdx
= 2.01 a T x 10
II
Equations (13) may, of course, be extended tor application to structures other than ~usses. The expressions appropriate to other loadings have been developed previously (eqs. (8) et seq. in this chapter and other equations in Art. A7.3).
Y = 1.21 a T x 10
1I
True stresses are given in Table AB.4. Problem 10 upper surface of the builtin beam of Fig. A8.26 1s heated to a uniform ta~perature T.
Exa~ple
T~e
:
AS.16
Through the depth of the beam the temperature varies linearly to normal (T = 0) at the lower surface. Determine the end moments developed, neglecting axial constraint and influence of axial forces.
Then the equations corresponding to eqs. (13) were written (see also eqs. (8)).
~~
EI constant Fig. AS.26a
r
Solution: The problem was only singly redundant because of symmetry and was made det~inate by cutting the end bending restraints. Application 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
eq, 13)
....!x_...B.... + EI .81 Y
R X EI
'"' O.~ = +
: h
!!lx'dx
=0xT EI
=
aRT
"Therefore
X
L "IT =
=0
TaL/h
g
X =  TaEI
h
Note that from the last of these equations as it must because of the symmetry of ~~e ring. SolVing the first two equations
= 0,
The redundant moment compresses the upper fibers as was to be expected. Example Problem 11 Complete the problem begun in Example Problem 24 Art. A7.8, viz, that of computing the thermal stresses in a closed ring whose inner surface is uniformly heated to a temperature T above the outSide. Solution:
The ring was made determinate by cutting at the top as in Fig. A7.30(b). The unit loadings ~~d thermal deflections were dete~lned in the referenced example. The results of deflection calculations made preViously were
X = aT:I
y
= 0
A nonzero value of Y would ~roduce a varying bending ~oment which cannot Je because cf symmetry. Hence this result too, is rational.
AB. iO Redundant Problem Stress Calculations by Matrix Methods.
Orr = 2n
RaT
In the following section the Indete~inate structural problem 1s formulated in ~~rix notation. The reader is assumed to be f~iliar with the matriX applications of Art. A7.9 and the elements of matrix notation and arithmetic (see Appendix). The stress distribution of the structure is specified by a set of internal generalized ~orces, Ql' qj'(ref. Art. A7.9). Unlike the cas~
the determinate structure, these qi' qj cannot be
.. 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).
A8.17
related ianeaa at etv to the extarna I loads by the equations of stat~cs. Thus a certain subset of the qt' qj are r3dundants and are denoted by
~,
SYMBOLS  continued  the temperature (above nOrTAl) at points i , j.  the ~ember thermal distortions associated with qi J qj'
finally, the redundant forces, ~, _ qs , are computed (by satisfying continuity) the true values of all t~e qi' qj may be found by statics.
~nen,
In the notation of Arts. A8.3 et seq, the final true values of the stresses were expressed as {eq , (5) Art. A8.3)
(5)
8maOLB
G. , qJ t
'1,.'
P:nJ
sr
qs
Pn
 internal generalized forces acting on structural elements and react~ons at support points.  redundant generalized ~orces and redundant reactions  applied external loads terminate stru~turejunder application of a unit external load
P,,=l(Pn=l)
Here: [gim] (=[gjn])' is the matrix of unitload stress distributions in the determinate ("cut!t) structure found by the application of unit (Virtual) ~oads at the external loading points. The product
gtr (=gjs)  the value of qt (qj) in ~he determinate structure due to application of a unit redundant force
'I,.
G
=1
(qs
= 1)
redundant structure due to application at load P = 1 (p = 1) " n G~n (=G )  the true value of ~ (qs) for a sn unit value of applied load P = 1 (p = 1).
1"
(=0
In
[g1rJ~[gjsJ) is
ij
 member fleXibility coefficient: deflection at point i for a unit force, q1 = 1 (see Arts. A7.9, 10)'
amn
" coefficient for the de influence terminate structure: displacement at exterr~l loading point ~ for a unit applied load, P = 1.
n
(a = a ) mn nm
stress distributions found by application ot unit (virtual) forces at the red1.Uldant cuts in the determinate (!tcut Tl ) structure. Hence this 1s the matrix of ux ' u y ' etc. loads. The {~} of course, correspond to X, Y, etc. Note that the [gim] and [gtr] matrices are load distributions computed and arranged in much the sarr,e fashion as was [Gim] of Art. A7.9.
'The
a~n (=u
sm) 
infl~ence coefficient for t~e determinate structure: displacement at redundant cut r (s) fer a unit acn  i ec load, Pn = 1 (P = I),
~
(QrI:t
a~s
= a nr)
small letter It g " 1s used to indicate load distribution in the !tout" structure. By way of illustration, the r ina l result for Example ?roblem 3, Ayt. AS.5 is expressed below. FIRST, in the r cru of eq , (5): True Stresses = S + X
coefficient fer the deeraune t e (vcut") structure: displacement at red~dant cut r for a ~~it ~edundant force q = 1.
l~fluence
(u,)
+ Y (U )
(a
rs
 a
sr
A:nn
influence coefficient for complete Yedundant ctr~ct~e: ~eflect~on at externa: lO~dl~g ~oint n for a unt t applied load, ':l = L
o+
DE
(.~
:::''1
= Po. :',;::J. )
Note that within each of the sets of subscript symbols (I, j), (r, s}, (m, n) the symbols may be used interchangeably.
AS. IS
SECOND,
of externalpoint ir.fluence 80efficients in the ~cut~ struct~e: deflection at point n per ~~it load at point n.  the :r..a ntx of in~luence coeff:c~ents ~el~ting rela~iye displacements at tne "cut s" :0 external loads: dlsplaceme~t at cut r per unit lead at pOint ~.
Note that in this case [gim] consisted of only one column, inasmuch as there was only a Single external load. In Art. A7.9 the strain energy was written
(15 )
(18)
 the :::a.trix of influence coefficients relating relative displacements at the "cu~s" to ~9dundant :oads at the Wcuts": displace~ent at cut r per unit redundant force at cut s. With the above notation one may Write
where [OiJJ is the matrix of member flexibility 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)
2U
(19)
Now according to the T~eorem of Least ~ork aU/a qr = 0 for continUity. Then, differentiating eq, (19):
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 term in the ~ldd1e of the above result is correct by Qbserving that, because of ], the symmetry at
~crass praduct~
[u1J
Eq. (21) is a set of simultaneous equations for the redundant internal forces ~,qs' It :nay be compared With sq. (6) of Art. ,1.8.6, to 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
(22)
The matrix product  fa ~ gives the The various ~atrix triple products occurring L.:rs....J L..:rrrJ 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
AB.19
L /3
Liz
1
a rcr n)
{qi} = ~i~
CiJ
L/2
L
a a
V'/3
Liz
a
0
E:I
a
0
a a
Liz
1
(~~
=
The moments q~ and q. were taken as the redundants. With these set equal to zero, the internal force distributions due to application of unit values of P l and P a were determined, giving
The matrix set of~ in parentheses above, gives the internal force distribution per unit value of the external loads. It is given the symbol
(23)
ilL
~~
a a ilL a a
a a
With the applied loads set equal to zero, unit values of the redundants were applied yielding
(r =)
(2 ) (4)
so that
 ilL
(24)
Eqs. (23), (24) constitute the major result, inasmuch as they present the ~eans for computing the internal force distrib~tion in a redundant structure. Example Problem 13 The doubly redundant beam at Fig. A8.27 (a) i5 to be analyzed for the bending ~oment distribution. The beam is loaded by couples over the supports as shown.
EI constant
L
~P2
~!J
1
=
a a
11L
1
ilL
Note that redundant load qa was applied as a selfequilibrating internal couple, acting on both beam halfs.  The following matrix prOducts were formed:
~ fq,( Ii q,
..!
Pi
L
1
., L'" .,
:.1 0 0 liL
('I
L : 6E:.1
q2( II q,
Fig. AB.27
} (hi
~~
0 0 i _lit. ."
r I" 'J
!..Iz
1
Liz
0 0
0 0
L/3
L/,
L:,
~L ~L
So Lut Lbn:
The choice of internal ~eneralized forces is shown in Fig. A2.Z7 (b) .. T~e appropriate ~ember fleXibility coefficients were arranged in ~trix :or~ as (ref. Art. A7.l0 :or coefficient expressions) .
r.
!
.
J["
1
L/, 1 0 0
0 0
Liz
0
1./3
Liz 1
0 L 1 2 .. : mOl
L/2
~ ~
I' .:J
0
... ~
AB.20
~rs ~
II
_J.
_ 5EI  7[
C J
2
1 4
though the ma'trtx of member aexi'Ji2.1 ty ccefflclents was expanded to a 6 x 6, the coefficients for q, and q, were zero. ThUS,
VI/3
L/2
1
a
0
0
0 0
Next, the unit redundant load distribution was found (eq. 22).
1
~lLJ
_ L
 ;;:1
L/2
a
Liz ,
c
0
a
0
a a
0
0
La/3 L/2
0 0
 '7
=
0 0
" r!
0
0 0 0
0
0
0
r.
286
.428]
LI43
.28~
With the redundants q~ and qa set equal to zero, successive applications of llii!t external couples PJ. and Pa gave the stress distribution
0
a
0 0
[!1;]
1
0 0
a
1
0 0
_1/
0
0
= 1 L 0
0
1
0
L
2
1
0
0
286
.42
.143
.286
I/L  I/L
0 0
J
1
and, with PJ. and P a zero, successive applications of unit redundant forces q$ ~id q~ gave
0
0'\
1
=L
2
3
4
~lJ
I
=
2L
1
I
L
La
(The tabular form at presentation of the matrix Gim, above, is used here only to indicate clearly the functioning of the subscript notational scheme. In general, it should be unnecessary 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;;]
~r~
The inverse was found:
6Ei
r
,
J
sl
51
5]
l~
~5
The redundant beam problem of Fig. A8.27 is to be resolved using the redundant reactions as the unknowns. Solution: The support reactions under the loads PJ. and P a were given the symbols qe and qe, respectlvely, Positive up. These forces did not enter into the strain energy expression so that, a i
6EI
7L'
Finally,
rG l L:1m ~
r:. ,
~1~

;:'l
r;; ,
A8.21
1.286
.l25
=L
.':::;CJ..,
.'+<:.c
(COm'Cl2.r':! 'fIith s ojut i on to Example Problem 13). Problem 14 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 vertical loads applied at the four external points ~ndic2ted.
~ample
/\t:it:J%f2t\l\/V\
I
.5
1#
.5
~,,~
Fig. AB.30a
{gi3} loading
'R2
8 panels at 20" :: 160"
Fig. A8.28
the axial loads in the various ~eabers. These ~ere numbered from one to thirtyone as ShC~~l on the fl~1re. The member fle:Krbility 80efficients in this case 'Nere of
~ere
employed
coeffic1ents arf written as a column matrix below. (They 'NE::ce employed as the diagonal e lements of a square ~~rix in the matrix multiplications, but ars written here as a column to conSer.e space.) Membe~ loads qa and qo were selected as redundants. ~ith qa and Qe set ':!qual to zero (''But''), un t t loads were applied succ ess ive Iv at external loading paints one through four, t~e four stress distributionsthus fo~nd being arranged in four columns giving the matrix
~i~ (below).
The
20 20 20 20 20 20 20 40 40 40 40
40
~
1 2 3 4
0
3 0
.0
5 0
0
6
7
a
9
1.0 .5 0 0 0
0
0 .5 1.0 .5
0 0
10
11
.25 .75
. , 0
o~
a1 =~
40 40 40
22.422.4 22.4
~1~
22.4
22.4 22.4. 22.422.4
loading
fi~ure
of ~i;] 1s shown in Fig. A8.29. Next, unit forces were applied successively at the redundant cuts "three" and "rtve'' as shown in Figs. A8.30a and A8.30b. ~hese loads were arranged ~n two columns to give the ~trlx ~iJ
22.4 22.4
22.4
m" 22.4
22.~
22.422.4
0 0 1" 0 0 18 I .56 0 19 .56 0 20 0 561 21 .56 0 22 .56 .56 23 .56 .56 24 .56 .56 25 .56 .,:)6 25 v .56 27 0 .00 29 0 .00 29 0 .56 30 0 0 31 0 0
12 13 14 10 16
.25 0 0
0
_.
AS. 22
I~
1 2 3 4
0
1
1.0 .50
05 9 [ .0035
.111
.0065 .111
.0035'] .059J
.0065
.50
.50
7 8
9 10
11 12 13 14 15 16 17 18 19 20 2 22
23 24
.50 1.0
Ix
.25 .75 .50 1 2 3 4 5 '6 7
8
I
1 1.0 .529 .059 .031 .003 .001 0  .50  .764  .294.045  .017 .002 0 0 l.12 1.12  .527 .527 .527 .527
2 0 .445 .111
.059
3 0 .003
.006
,
I
.001
.003 .031 .059 .529
.059
.111 .445
.56
25 26 27
~a
9 10 11 12 13 14
29 30 31
NOTE:
 .56 .56

.56
16 17 18 19 20 21 22
23
.031
L:a.o
a. ill
136J
24 25 26 27 28 29 30 31
.006 .003 0 0 .222 .167 .085 .032 .003 .001 0 0 0 .498 .498 .622 .622 .056 .058 .058 .058 .003
.00:3
.oas
0
0
 .017 .04.5
 "'G' " " ....  . , ::
.....
1.0 0 0 .002
.':::',,7
.00 0
0
.003 .003 0 0
.001
.001
.001 .001
.031
L~'O
15.0 0
0 15.0
~.J
a1ll
13~
~.o
Example Problem 14a Fig. AS.31 shows the two bays ot a steel tubular tail fuselage truss Which is loaded by tail air loads to be resolved into t~ree :oncentrated loads applied as shown. The fuselage bulkhead at the attachpaints station (AEFK) is heavy enough so that it may be assumed to be rigid in its own ~lane. Hence, the truss may be analyzed as if cantilevered from AEFK as shown. All members are steel tUbes, their lengths and areas being tabulated below. Solution: The generalized forces were taken to be the member axial loads, these being numbered as in the table below. Member fleXibility coef:icients, L/A, (E set equal to unity for canvenience) were also tabUlated.
Next, the values of the redundant forces, tor unit values of the applied loads, were
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, another 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.
A1~"~
.f.
~ ~
,
..... B
,
__
"' "
P2
{
::I.......
'" 1.0236 P,
P
3
From Z F on Joint B.
.081759 Cl.1.  18334 Cl..  q.
q
Fig. A8.31
.1410
q1.'"
~ER
N"lJ11BER
L2NGTH
AREA
L/ A
52.8 51.0 .00."" 51.0 62.8 62.3 44.85 30.3 44.85 50.8 62.8 64.1 90.9 55.1429.62 29;52
~
RC
cli
BD CD DE ?G GH HI IJ GJ
1 2 3
4 5 6 7
5
9
J" AG
BG
3H
10 '1 12 13 145 i5
17 ,Q
;
35.47 25.43 9.20 25.43 35.47 35.47 25.34 5.00 25.34 9.20 35.47 40.4 15.0 27.57
11. '70
.565
.49~
And so forth, for the other Joints. Note that in each case the equations were arranged with the applied loads (P ) and the n redundant q's (q33, q33, q.~) grouped on the right hand side at the equal Sign. This arrangement was observed for all 21 equations, atter which the equations were placed in matrix form as
He
Dr
JJ SJ
.' J
Ie
.. ,
.165 .565 .165 .565 .630 .165 .500 .395 .395 .500 .160
.630
i,j
n
= 1,2,
= 1,2,3
.. 24
= 22,23,24
FJ
" ~.
.U
. "
20 22 23
2.:1
".
I~
sa
66.1
52."~6
(Note that there were 24 equations ~ere, the adcitiona1 truee equations being the ident::.ties
,
'Q.
.
1.06.6 lOE.S
The structure was three ti~es redundant. rra~ework of p jOints, 3p6 !ndepence~t equations ot statics nay 88 wT~tten (p. A8.10). Here, ~oweverJ stress cetails in the ~l~ne AEKF are t8 be sacrl!iCed, six equations are lost ~hereby si~ce onlY net ~orces in 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. The next steQ Nas to comnute the unit stre~s ~istri~uti~ns [glm] and [~irJ' Rather
a space
On the right hand side of the above matrix equation the ~trices are shown ~partitioned~. T~e ~irst three columnS at [OJ are the coefficLerrt s of P and the last 'three are the
n
For structures other than trusses the equilibrium equations are written for the various structural elements, equilibrium of joints alone being inappropriate.
A8.24
STATICALLY INDETERMINATE STRUCTURES
rocnc
q.
by finding the
lnver~e
of [C
Thus,
1 j]
trsJ
T~en
10
_.
[5555 .2550
.2580 l
3.C~12
'2~7~
.2775
2.3015
3.023
where
= 10
~
~
2
.18
","
2259 238
be."=,
 31A
121.~
ThUS, the ~~lt stress distributions in the 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 standardized techniques may be employed. The result in this case was
distrlbu~~on
1;\
1
1
.5415
a
2~727
, s
, " .. :~
..
.'
_2.72:7 4.0.59
.. . .,.
c .
5.;'7~
Z2
Zl
2.:!C59
~::
zrs
.:).5<> .. .
,
,
I
,, ,
~ o
0
X
1
2 3
I ,
.8412 .00017
.01,523 .00017
_7.~
..
,:;,,60
.,
.
,.
,.
.4415
 .0=
"  .aiee .5 ,5 ,
z
."'"
0
.54.
.06<:
.. ..
.&C .22
.."'
..  0
.0
4
5
..;..,5
."'~
.6529
1.404 .8443
..
 ...17
.0
2.," .. ,
1. ..35
c
0 0
r..4J.S
.3523
0 0
6 7 8 9 10
~1
. ,
~
.
, L'"
.0
.W
. 1.W9
0
.... 5
.zao
, , ... ,'."
.:...,27
,
.o se
1.102
.~O'"
1..;,01
_,.~l
'.302
0
12 13 14
15
oe o
0
1.0
The member flexibility coefficients L/ A were arranged as the diagonal elements of the matrix [u 1j ] Then, multiplying out according
16 17 18 19 20 21 22 23 24
~ar.mle
.1094
.84.50
.0973 .0494 .6638 .2350 .9185 .5116 .5120 .9193 .1061 .1434 .2023
.2660
~...,.=:: 1.296 2.203 .~38 .31.53 .1190 2.206 .E;39 3.008 i 1. 557 1.5247 i .332 .7352 I .761 .095<::'. I .096::. 1.260 1. t50 .0205 .3338 1.607 I l.S'~2
.~IV
.3143  .3896
.733
.2641
. , 
.0002
.1364
.03 4
,
 .6956
.122.3
.238 .176
2.269  .2010
~~
~r;]
[177,9 195.0
154.4
2263.0
6430'~
2273.3
Problem 15
1221
1035
114~
1224
The doubly symmetric four flange idealized box beam cf Fig. A8.32 is to oe ~n~lyzed for 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 constant in ~ac~ panel. The bea~ :s ~o~ted rl~1dly at t~e ~oot, provid1ns ~u:l restra~nt
These are the "effective areas", being the flange area plus adjacent effective cover sheet area plus one sixth of the web area. (The factor of oneeixm provides the same moment oC inertia as the distributed web area).
1035
A8.25
CT2SS
section
~ue
The ~e~ber :lexlbillty coefficients were 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)< , !
, IaI , ,
s I
, ,
. i
'
I
,
I
, ,
10
!Z
I IJ I
iaa
! ie , ",
" ace
..
,
..
,
c.
15.2.il
,.o.~
NOTE:
..
.
,
.~
.
, .,
"
::7 .5
=1
Fig. A8.32
Cover sheet shear flows q~, q7 and ql~ were selected as redundants. ~ith these set equal to zero, and with unit loads placed successively at loading points ~one~ trxough ~sixw, the
[gim]
Solution: The generalized forces employed are shown on the exp Ioded View. Fig. A3.33. No forces were shown :or the laHer surface, its ~embers and forces being equal to those of the upper surface because
matrix was obtai~ed. ~~en the cover sheets are ~cut~ the ~NO webs act independently as planeweb beam5. The details of the stress calculation for such beams are similar to those of Ixamp.Le Problem 21, Art. A7.7 and are not shown here.
of
symmetry.
In the
[u 1j ]
I~
1 ,400
matrix this fact was accounted fer by doubling the ~ember fleXibility coefficients for the members of the upper surface.
'\.'1:"
,Qle
3 4 4.
.\00
4,
.200
8 9 5.341 .00
2.571
,200
2.571
~~ \~
10
12 13 14 6.231
5.341
.08_87
,0444S 4.006
5,
.08887
2.003
4.006
,1333 2.003
IS
NOTE:
,'.~~,
q,
Fig. A8.33
ceSSively assigning unit values to the reduncants 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 selfequilibrating pair of shear flows acting one on each side of the "cut;". The ribs were considered rigid in their own planes.
?ut
q~
=1
(1: M, Z F)
A8.26
q ",0
.~E;'<S\l25
,40",25.04 .75x~=.5625
I~ .;'" ~q
Per
eo..
2.5 j
..
\q
IS"
25.04
~r~
 E
_ lOll
lX20",75
15
IX~~ 5.0
fq,=_1
Fig. AS. 34
"
~\~
'\
.=
_.ro:l56
._,
.ceeec
_.C02Z59
.~2259
q."
f q,I
q,..= 1
appendix)
=
E 10 IS
q,. 11Cj!.__.
__._...__..':;S q3 " 1
e:
Then
[3,862 2.562
 .1137
2.562
6.137
2.321
 .113J
2.521
9.499
q, = 1.0 q, = 1.0
From eq'.l111brlum of 200 rib: q. = t .5625 q. = .3125 q. = 0
~sJ
=
~rs j ~r;;]
.25) = .3125
[0699 .0699 .00598 .00598 .00016 .00016 = .03593 .03593 .03409 .03409 .00403 .0040 .01568 .01568 .01872 .01872 .01578 .0157
by hypothesiS
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 flangeloads are transmitted directly to the flange ends ot the adjacent bay (see q~, q" q .. q"
of Fig. A8.33).
~i~ I~
1 3
4
0
=
2
~i~

~1J ~~
4
1 .330
I~
1 2 3
4
2  1.00 1.00 1.00 25.04 25.04 .3125 .3125 20.87 20.87 .1387
.1387 18.79
18.79
12
6
~
5
6
~iJ
7 = 8 9 10 11
12
 1.00 1.00
l.00
0 1 12 3
 LOO
4 15
.0699 .0060 .0060 699  .0699 .0699 .0060 .330 2. 6 1.75  .150 .75 2.26 .150 .08 9 .16B .0141 .0341 .0 9  .0359 .0359 .0322 141 .0 1.w45 2.51 2.83 1.126 4  .0010 .0776 .0157  .0157 .0187  .0010 .0454 .0112 3.47 2.76 2.40 2.76 3.47 1.61
.0060  .0060
.0050 .0002 .0002 .0002 .JCO 2 .:JOC2 .00'22 .150 .0040 .00'+ o .150 I .00401.004.o .0322 .0040 . 0: 40 C341 .0040 .' 040 . .168 .0040 .004o . 22 1.16  ....22 1.5<t.5 .122 .122 .0112 .0150 .015 o .0187 .0158 .015 8 .0776 .0150 .015 o 1.61 .610 ".610 2.40 .610 .610
1.00
LOO 31. 71 31.71
13 14 15
NOTE:
The reader will observe that the result displays the ~bending stresses due to torsion~ that Is, the bUildUp of axial flange stresses near the root of a beam under torsion when the root is restrained against warp1r.g. The solution for application of a torque is readily
FLIGHT
VEHICLE STRUCTURES
AS 27
found by superposing stresses for P~ = 1 and Thus, one :inds that under this condition there 1s a root flange load of
P a = 1.
=;
44
15.0
ql.
= 3.47
 2.76
= 0.71
Ibs.
1~J
144J
388
~0525
1. 665
.0525
.0975
.888
. 0525
.888 .472
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 been det.eratned from eq . (22). The total de~lection of external loading points is then eaSily computed as the sum of, 01~, the deflection due to external loads acting on the
.0975 1.665
.028
qs}
J
.052
"cut;" structure, [amnJ {P n } (ref. eq , 16) and, TNO, the deflections due to the redundant forces acting on this same cut structure,
~,,;;] ,. ~
~'#

15.9
36.3
.052
.098
.098 .052
36.3
15.9
"J
15.9 144
eq , 17).
Thus,
Example Problem 17 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 coefficient matrix was formed as in Chapter A7, Art. A7.9, by the product.
~:nnJ
{n } ~msJ ~rsJ ~~ {q
P
= (

The matrix expression set off in parentheses above, giving as it does the deflections for unit values of the applied loadS, is the matrix of influence coefficients. Let
1243 
1522 1243 29.7 l243 1522 29.7 29.7 897.8 594.2 33.7 33.7 594.2 897.8 33.7 33.7 33.7  33.7 .8 36.8 9.7  33.7 33.7 36.8 36.8 1
29.71
so that
           (26) AS.12 Precision and Accuracy in Redundant Stress Calculations.
~ample
Problem 16 Deter.nlne the matrix of influence coefficients for the redundant truss of example problem 14.
Solution:
~roducts
~trix
Matters of prectsion are dependent upon the number of Significant figures obtained and retained in dealing with the geometry of the structure and in the care with which arithmetic oDerations are perfo~ed. In the discussion to follow it 1s assumed that all due caution 1s exercised with regard to the preCision of the work. ~~t~ers of accuracy'r4ve to 10 with the number of Significant figures :lnally obtained in the answer as influenced by the ~er of formu
A8 28
lation of the ?roblem. ~~e accuracy ~f the ~e sult may be affected by a n~~ber 0: factors, two 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 TIatrix , Cars] (eq.21).
In the matirtx
~ents
 the magnitude of the r~dunGant ~orces relative to the size of the dete~,lnate forces. These ~NO factors are concerned respectively With the left hand side and the ri~ht hand side of eq , 21, 'Viz. '
are measures o~ the str~ctural crossone red~~dant force with a~o:~er. ~he st~ength of t~is crosscoupling is de~end ent upon the choice of red~~dants made ~~ "cu't t tng" 'the s truc 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.
cou~l1n; o~
:l ;
(bl
~
(dl
Fig. AB.35
which determines the accuracy with which its inFig. AB.35(b) shows the c~oice o~ generalverse 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 cthe condition of the matrix. A weJl conditioned 2/3 1/6 0 0 matrix is more accurately inverted than a poorly conditioned one. 'rNo extreme cases are now 1/6 2/3 0 0 given tor illustration:
a) the diagonal matrix. Its ottdiagcnal elements are zero, so that it is ideally conditioned. Thus. the inverse of
C:I1J =..1. EI
0 ,0
0 0
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 and q .. = 1 gave
~
1
4
L
 3
:3
2
L L
b) a matriX all of whose elements are equal in each row. All the elements cff the main diagonal are equal to those on. rhe dete~in ant of such a ~trix is zero and nence its
~1~
0 <.
;j L
1
:3
0
3
4
A8.29
out,
laEI r,s=3,4
L'~. Lz 71 ~ 8
The crosScoupling is large. SECO~TI, suppose the moments ql and q~ had been chosen as ::edundants. 'rne "cut structurel! in this case being visualized as in Fig. A8.35(d). Application of unit redundant forces q. = 1 and qa = 1 gave
itself.
I~
1
2
1 1
a
1 1
a
I:
1
In most structures, however, the additional labor involved in seeking an ortho~onal set of redundants is not warranted. (2) In choosin~ a set of redundants which will yield a wellconditioned redundant matrix, it is best to make such "cuts" as will leave a statically determinate structure retaining as many of the characteristics of the original structure as possible. ThUS, one may consider that the structure of Fig. A8.35(d) retains more of the features of the original continuousbeam structure than 10es that of Fig. A8.35(c). (3) The degree to Which one redundant influences another (extent ot crosscoupling) can be visualized by observing how much their individual unitload diagrams "overlapH. To elaborate on this last point, refer once again to the above illustrative example. The crosscoupling of q3 with q~ may be expected to be large if their unitload diagrams are drawn as below in Fig. A8.36(a). This deduction follows eaSily if it is recalled that the dummyunit load equation for such a crosscoupling term is at the form
I:
2
1"
Large .. or
m~,
L I:
m..
tIultiply1rlg out,
J W'
:0:
Im:
EIdx
r,s=l,2 The condition of this redundant ~~trix is obviously better than that obtained with the first choice of redundants. There is less crosscoupling between the redundant forces.
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 impossible to invert a large, illconditioned matrix with the l1.mited number of significant figures available from the initiaJ. data. The following statements and rulesofthumb may be useful in the treatment of highly redundant problems.
m,la
m
z[
~
(b)
Fig. AB.36
St~dy
of
redundant choice ql' q~ (Fig. A5.35b) reveals that the crosscoupling sh~uld 88 small her~
....... since an 1rrteetra l J ':J" v .9 ~ave a contribution from Hence, the J ::I1.:n'l, ct.:.:: is
r' orm
J:nlffi' l ex EI
ex
can
(1) It is always possible to rind a set of redundants for which tIle crossco'J.pl1m; is~t a ~inimumzero, ~n tact. Theoretically t~en, by ~roper choice of redundants, tte ~a~r!x [Q ] rs ~y ~e reduced to a diagonal ~trix (idea~ly conditioned) . (la) The choice of red~~cants which gives zero cr ossc oupt inz ("orthog:Jnal funcO:ions") 1s no'\: reaai i . I!1 some special ' / found in :::ene p :;' . structures, such as r tngs anc rrames, orthogonal
the center
s~an O~ly.
sma11 er h cnan
r
0'
ex 31
or J
r ::I~
~I
Th .. us, a
~han
Visual
q,
~nspect1on
is
c...
AB.30
(4) Combinations of ~ed'~dants ~ay be ~m such a way that the ~rOduct [gir] ] does ro played to yield new '~itreduncant stress distributions which do not "overlap" as extensively exactly 'Ntat 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 [giPJ ". Then "take cmua one half of the q3 illustrative problem, the choice q~, Q~ for redundants had been made originally, leading to column plus one times ~he q~ column to give the the unitload dia~acs of Fig. A8.37(a). Inspection of the diagra~s leads one to anticipate second column of the new distribution [giO] a stror~ degree of crosscoupling and hence a new set of redundants is sought. Rather than Multiplying out in the above example return to the s'tzuc ture to choose new "cuts", combinations of the m3 and m~ diagrams are ~ looked tor ~hich will have less "overlap" and 0  2 L hence less crosscoupling. It is observed by inspection that two new 1 0  2 L stress dlstributions"'which have the desired property may be formed trom the m~, m~diagramB 1 1 by proper combination. Thus, if one halt the 1  2 m" diagram is subtracted rrcm the d'Iagrarnto fOrm one newstress distrIbUtion and one halt 1 1 the ~ diagram ~ subtracted from the m~ dia 2 gram to form the other new stress distribution, the results are as shown in Fig. A.37b. There Now form the ~trix of redundant coefficients ta. oovtous jy less "overlap" of the diagrams for these new combir.ations. for the ~ew unknowns (subscripts p, 0; [giP] =
[a
m:
C: eJ
[gj,,] )
L'
24"::' .
Fig. AB.37
New
The condition of this ~trix is ~eatly improved over that obtained :or q~, q~ alone (previously computed), ViZ
In this way two new unknowns are introducen by linear combination. In matrix notation. the old stress distributions [gir] are transformed to a new set [giP] by forming
~r~ = L~/18EI
(That the
(r , s
3, 4)
[apo] matrix
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] ,
:3
2 L L
! 3
 :3
0 1
2
L
L
! 3
1 0
~ :/J
,
1/2
the problem ~y be completed in the "0, 0 system". The appropriate equations ~re abtai~ ad from eqs . (14), (21), (23) and (25) Simply by replac1ng all "r, 5," by "~, 0". Thus
(29 )
~re
the solutions
A8.31
(30)
and where
auxiliary rules, test information, or even ~n tuitive guesswork 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 corrections to these values become the unknownrediiiidants : ) Mathematically, the magnitUdes of the redundant forces are directly dependent upon the magnitUdes of the elements in the matrices [arnJ
or [apn] on the righthand side of eqs. (21) or (30). (The righthand side of such a set at Simultaneous equations is called the nonhomogeneous part.) Thus, the relative merttsorseveral possible determinate stress distributions may be judged by forming the matrix product earn] With each and comparing results. FOR 2XAMPLE, if the doubly redundant structure or Example ?roblem 3, Art. A8.5 were formulated in matrix forn the [gir] and [a t j ] matrices would be (see Fig. A8.38 for numbering scheme)
Fig. A8.38 Generalized force numbering
apparent that if these corrections are large, then any inaccuracies in the redundants (arising frcm the difficulties inherent in accurately inverting [ur s] or [apoJ) will have an im
portant effect on the accuracy at the final a 432 0 a stress distribution. ThUS, as a matter of a 720 a a 1 general practice, It is desirable to keep the E aagnttudes of the redundant forces as small as 0 a 402 a possible. a a 312 a It follows immediately that one should use for the deter.ninate stress distribution one requiring a minimum of correction, i.e., Select .806 1.154 as the determinate stress distribution onewhich apprOXimates the true stress distribution as 1.564 1. 729 closely as posslDie. = ~1!J 1.0 a The rule giien in COnnection With the "choice of redundants " (rule 2, above) is an aid l.0 0 in ~king a gOOd selection for the determinate stress distribution. If, as suggested, a deterSeveral Possible deternlnate stress distriminate structure is obtained by making "cuts" butions will now be tried. FIRST, the which leave a system having properties similar to the original, the stress distribution obtained therein by statics should be a fair approxi stress distribution obtained by statics alone in the "cut" structure (q, = q~ 0) ~tlon to the final true stress distribution. However, it is even ~ore im~ortant to realize that the determinate stress distribution only need be in static equilibrium with the external applied loads and that it may be dete~lned with the aid of any appropriate
[gtmj
AB.32
(This is certainly a poor approximation to the final stress jistribution). Multiplying out
ExaQple P~oblem 18 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. Solution: The first step taken was ~~ examination of the unitred~~dant stress distributions :~ta~ned with ~he choice pr~vlously zade of q2, ~~ and ~12 as rsdundant.s . The VITee un i t reduncant stress distributions were represented g~aphically as in Fig. A8.39.
SECOND, a stress distribution in which loads of 0440 Ibs were guessed at for q1 and Q3' (In a t'NO times redundant structure any two forces may be assigned values arbitrarily while satisfying static equilibrium.) q~, Q~ were found by statics. Thus,
0.40 0.258 } 0.40 { 0.0672
18.79
1B.79
26.44
26.44
31. 71
31.71
= {
101
9.49}
Note that a reasonable ~ at a stress distribution resulted in nonhomogenous terms only onetenth as great as those obtained by use at the ~cut~ distribution. The magnitudes at redundants are correspondingly reduced. THIRD, as a ~tter ot interest, the true stress distribution, obtained in Example PrODlem 3, was used. The result:
Fig. AS. 39
Unitredundantforce stress distributions axial flange forces.
Inspection of' the figures showed that the following combinations of 11striJutions should give new distributions likely to have considerably less "overlapll. The nonhomogeneous terms are practically zero, as they should be. The redundant forces would be zero also.
(1)(2)
lX{gl.}~g:~~{gl,}+OX{gl'''}
(3) The demonstration here suggests a useful check upon the final result of a redundant stress calculation. After obtaining the final true stresses [Gim] (.. [Gjn]), one forms
In
~trix
was
o
and compares the result elementbydement with the matrix previously computed,
.8330
The "truematrix" elements ought to be zero, or nearly so, if [Gim] is errorfree.
AB.33
I~
1
2
1
 1.00 1.00
1.00 25.04
3774 10'
E:
.1789
'0520~
.07322 .1254
.0520
3
4
5 6
7
9
10 11 12 13
"
.0lD .0lD
. 
14 1 15
NOTE:
 1.00 1.00 1.00 29.38 29.38 1.0285 . 833D  1.0265 .026 .026
It remained to select a determinate stress distribution Which would reduce the ~gnitude of the redundants. For this purpose, the engineering theory of bending was employed to comoute stress distributions satisfying equilibrium for each application of a unit external load. The result was (refer to Example Problem 30, Art. A7.11).
~
1 2
4
.30
.10 .10 .0
 .10 .30
2.0
2.0
.10
Fig. A8.39a shows plots of the new distributions, these having greatly reduced "overlap" (compare with ~ig. A8.39).
.0125
.0423
.0125 0.15
.05 .05 1.33 1.33
.05
9 2.67 1 2.67
11 12 13
.020
.0023
31.71
31.71
14 3.00 15 3.00
2.00
2.00
.100
NOTE:
[gip]
~p~
=~
95 3495 1841 1841 0 0 ~ 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 onehalf to onetenth as large. The solution ~ay be carried to completion in the "p,o system" using the matrices [giP} Capo} [giro] and [Opn] in eqs (31) and (32).
.000017J .03124 .1254
A8.13 Thermal Stress Calculations by Matrix Methods.
was
The new redundant coefficient matrix [a ] po obtained by :n.ultiplying out per eq , (Soa ) ,
10'
E
2584
.011'32
.0000175
This mat~ix is very well conditioned, being a considerable improvement over that at Cars] round originally in Example
~oblem
The thermal stress problem is conveniently !or.nulated in matrix notation by an extension ot the teChniques presented above. ?irst, consider eq. (21), wTitten
~n
15, ViZ.,
the
form
,.Aef
<rtf .:{.J::i!:
'~:;:::;:::;,?
"
~':.)"4
A8.34
may consist of the su~ of several contributio~s should qi act on more than one member. Is desired at each of the reBecause a
rT 
[UrsJ
dundant cuts
(35 ) a ~trixform statement of the condition for { rT } = ~rJ ~iT } continuity at the redundant cuts, ViZ., the displacement 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. the unit redundant force stress distribution in To modify this equation for thermal stresses, the appropriate expressions for thermal disthe 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
Solution of eq. (36) gives the values of the redundants qs' atter which the problem ~y be completed in the usual fashion, viz,
where 0rT is the displacement at the r t h cut due to thermal straining in the determinate structure. The explicit formror this te~ will be derived below. Re~Titten,
(33) Eq. (33) is a modHied rom of eq. (21),
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
[giP] for [gir} giving as its solution the redundant forces in pn] for [Urn} etc. into eq . an indeterminate structure under the application (36 ). ot both external loads and a temperature distribution. MEMBER THE:R1lAL DISTORTIONS To der~ve an eXPliCit expression tor 6 , rT It remains to establish the forms for 6 the virtual work concept may be employed to adiT vantage. Thus, following the argument of the Thermal strains on an inrim teet inar eleftVirtual 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 indevelop. Hence only normal (as opposed to ternal virtual work done by the rt..h.redundant Shear) virtual stressesneed 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 BARB generalized torce ql due to thermal straining, The general expreSSion for the virtual 'Nork then the virtual work done by a single generaldone 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 obw = u aT dx tained by summing, giving the deflection at the r~ cut as the matrix product Where a is the ~terial thermal coefficient of expansion.
[J.
(34)
Where ~i Is the value at the qi due to a unit (vtrtua r ) load at cut r , Note that the term 6
*n iT
will be convenient later to designate by qST the solution to eq, (36) when the mechanical loads P n are zero, the
AB.35
Eeveral speCific cases are now treated. A bar under linearly varying load with linearly varying temperature is shown in Fig. A8.40. In this case
u =
8EA11S
~or a bea~ the general form or expression for the virtual work done during thermal straining by a toptobottomsurface temperature difference 6T, varying linearly over the beam depth, is (see Art. A7.8, Ex. Prob. 24).
qJ
ql
 qJ
L L
x x
qj~llql
Ti
llll1lUJ.llill T
Fig. AB.40
IYrdx
h
T  Tj
Tl  Tj
where m
= virtual
IYr
h
2f 1 + T
6
Ti+ J ql +aL 6
TIay be
zr j
This expression
aL
where
11
1Yr 1) qi
+ ct, 
(1Yr1
20T,) J q.
'
6.
jT
e a
qj~T"::1
h
= constant
Fig. AB.42
Note that variation in the cross sectional area of the bar does not affect the distortions
'IT' 'jT'
or
The alternate choice of generalized forces for the bar under varying axial load is shown in ,ig. A8.4l. By a derivation similar to that above one finds
w=
where
e ' ; (31Yr 1 6+ IYr
AjT ;;; h
aL (1Yr1 +6 20T~
wnere
Fig. AB.41
Special forms of the thermal distortion expressions for beams of varying depth may be derived readily as reqUired. 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
constant) and unifo~ T (T = Tj ) follow imi ~ediately by speCialization of the above forms. For example, for a bar under constant load qi = qj' and constant temperature, T i = Tj = T, one has 6. 1,:, = a LT.
( r~=x ~.~on~ 1
L L
Fig. A8.43
./'
I f', ...
"
AS. 36
the temperatures are equal at the left end and differ by aTo at the right end). Determine the center reactions assuming a 1s constant. Solution: In the illustrative example of Art. AS.12 this structure was analyzed by emploYing as generalized internal forces the bending moments q, and q .. over the central supports (see Fig. A8.35). The two central reactions were denoted by q~ and q~. The matrix of redundant coefficients, considering q~ and qa as redundants (the better choice, it will be recalled) was (ret. Art. A8.12)
L
2EI
solving
,}= {q s,
EI a OT,{.267} h .933
=0
...
EI a 6T,
h
1 0
0 1 1
6EI
.257
q~
2 C
1
C .933
2
c
I~
1
1 1 0
0 1 1
EI a. OT 0
h
2
3 4
933 :399/L
1.60/L
.267
 C
1
C
2
C  C
Member thermal distortions were computed. (Note that OT was negative according to the convention adopted earlier).
aL
2
X
c,
= 1.60
cates DO'>,/'N)
~lT
! 3
6
AT ,
11()
Example Problem 20 The symmetric sheetstringer panel of Fig. A8.44(a) is to be ar~lyzed for thermal stresses developed by heating the two outSide stringers to a uniform temperature T above the center stringer. Assume G O.385E.
; 
"" h.
()
(2 x ~ OT6 + ~ 0
4
T0"" (""
+ h
Solution: The panel was divided for convenience into three bays. The numbering and placing of generalized forces is shown in Fig. A8.44(b). Transverse members (ribs, not shown on Figure) were considered rigid in their own planes  a satisfactory assumption for symmetric panels. Because of symmetry only one half the panel was handled. All member fleXibility coefficients and thermal distortions were doubled where appropriate. The matrix of fleXibility coefficients was set up as (VOIDS DENOTE ZEROES)
(l
2 l "" '"  3: i
0 T~
..1. 6EI
~ ~ {q'J ~ ~ ~!
14
aL6T,
3h 2aL T 3h 0 c
q,
=()
0  1 01
C
2 L
o o
P'
AB.37
.'0
3S.
I
< , ., .2 8S.3
T:J.en
22.2 44.4
 .5
31 200
0
.0
(;
2.
0
31,200
31 2 0
0 1.0
 .5
0
~ember
thermal distortions
40 40 20
~1~
0 0  .025 .025 0
0 1.0 0 0  .025
o o o o o
o
Here, for example AlT = 2 x 2 x 10 = 40 (doubled once because ql acts on two stringer ends and doubled again to account for the other halt or the panel).
", ""'"
~r~
and
= 1
[2,2 13.8 0
86.1
1~'8J
.025 o
.025
~
",,02
1
(bl
.005
."''
"1I
"
ao
PCyp
T
1=== 60"
F=
I
=
'i.
FIg. AB.44
The structure was three times redundant. Stringer loads q~J qo and qo were selected as redun<1ants. Fig. A8.45 shows the unit redundant load Sketches for q. = 1, qo = land qo = 1.
172.2
13.8
n3lf
~
Solving,
. 1082 } '\= a ~ S .?9944 { } { ."C02
n~
0~
:::: x 10
.
108.2
99.44
A8.38
The result compares favorablY ~lth "exact" solutions made under the same assumptions (ref. NACA TN 2240) as far as the stringer loads are concerned. The shear flow result is not very satisfactory, primarily due to the use of too few "bays ~ in this analysis. Example Problem 21 The uniform fOUTflange box beam of Fig. A8.46a is to be analyzed for the thermal stresses developed upon heating one flange to a temperature T, unt rcra spanwt se , above the other three flanges.
l
L/Z
1
L/ Z
Liz
1
~IJ
1\
).
Liz
ILiZ
Liz
I,
L/2
I
Liz L/2
1
L/2
, ~:,.., \1 L
(aJ
~)
~':\~, ~'~
~,~~ q\~\~
",~,, '\~~'
:\~, ~
~'
.625
.375 83
.375
.1:5 .l25
~.I
Fig. AS. 46
\~'~ ~ ~ "\~
( 'j
.'25
[iki''' ...
.500
~'!J .
'", t
.se
.125
.37'
. 125
~ ...::.::.3
.125
.:25
.JS33
Tr.e
Solution:
i~verse
was computed to je
The beam was df vt ded into four equal bays 1.070 .06587  .5291 .2337 giving a tour times redundant problem. Four 1.368 .5291 .3613 .1018 selfequilibrating (zeroresultant) independent Gt stress distributions were taken as the unknowns, Grs::J L (0 oj  .2337  .3613 1.497  .1934 these being shown In Fig. A8.46(b). Such zero .06587 .1018  .1934 1.792 resultant strP.ss distributions are the only oneS possible in a structure having no applied loads. The matrix of member flexibility coefticMember thermal distortions (!Ii'!') were tents was tormed by collecting coefficients from 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 0 .6emc" .1667lc" heated equally, one aus't c! .500 set the corresponding ~i a LT 2 ~lU' , (:,o .. e { .1667lc" 0 .667lc" .1667ic" 8 equal to zero; this be2 .eoc cause the virtuai loads 0 .1667lc" .667',(" , .1667lc" being of opposite Sign 1 .500 adjaCent flanges, the .1;;67k~ ( Virtual Ncrk ~ust cane ,"1 ) . where k 2 = Gt / AE (b c) MUltiply1n~ out:
~IT} ~
.""".
Since externaf Ioada are not to be applied it follows from statics that the general.iaed forces for adjacent webs and cover sheets are equal, as are the loads in front and rear 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 symbol can be employed on several members whose loads are known to be equa1.
[]rLl{"IT}=~
Then the solu::ion to 3q. (36)
a T Sa
f}
i
5
w~s ~Titten
as
AS. 39
"cut" ~king it determinate, after 'Nhich the temperature distribution is applied producing thermal deflections which, when multiplied out gave
4 ' 079 } 1.938 + c) .8562 { .2413 .
{qST} =  16
Gt a L T
Ib
Finally the .conptece set at thermal stresses were (still for the case L3k 3 = 1)
4.079 2.039 L 1.938 3.008 L .8562 3.437 L .2413 3.558 L
at the external points (compare with eqs. (35) and (37) ). Simultaneously the redundant cuts will experience relative displacements. SECOND, the redundant cuts are restored to zero displacement by the application at redundant forces (thiS problem was solved in Art. A8.l3) .. The qs are given by eq. (36); they produce additional deflections at points m
The result compares favorably with an exact solution (NACA TN 2240) insofar as flange stresses are concerned, an error of less than 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.
A8.14 Thermal Deflections by Matrix Methods.
A. STATICALLY DETElIMINATE STRUCTURES
But
Therefore
The problem of the thermal deflections ot a statically determinate structure was considered earlier in Art. A7.8 in nonmatrix fon. It should be apparent from the derivation, that the matrix method presented for the calculation of redundantcut thermal deflections may be applied equally well to the problem of computing the thermal deflections of external paints of a determinate structure. Thus
(37)
or,
where point
OwT
ffi,
is the and
the~l
deflection at external
[Gmi ]
The matrix quantity in parentheses is the total strain (thermal Dlus "mechanical"). Fa:rdetinite 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 standpoint, the unit load distrioution [giml (Whose transpose is used in sq. 38) ~y be any convenient stress distribution satiSfYing atatrcsrn ~s~t of "cut" structures. One need not even use the distribution (and same
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
Bam0l[ilm]
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
Example Problem 22 Co~pute the rotation occurring at the right hand end at the beam ot Fig. A8.43. Solution: To compute the rotation a unit couple was applied (positive counterclockwise) at the right hand end. An additional generalized force, called qa, also was added at that point. Then,
, _ ct,
1.2
!.J
aL:ST o
h
 .289
It is apparent that the values of the redundant ~oments ql, qa could have been chosen arJ1trar1ly in {g1~}' above, with8ut af~ecting
"5T()
h
g 0To (3
20 T 0)_
4 a L 15 To
9h
the result.
Here, clearly,
structure visualized w1l1 lead to the same result, qa being equal to ~ity in all cases. Example Problem 23 compute the axial ~ovement of the tree end of the central stringer of the panel of Fig. A8.44. Solution:
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, equally, they are not used to describe the strain energy of the structure. Hence they are not included in writing the total str~in. (Their 6 are zero.) iT The member flexibility ccefficient matrix was
An additional generalized ~orce, qlOJ was added aXially to the free end of the central stringer (ret. Fig. A8.44b). T~en
l"n.13
88.8
! , ,
88.8 22.2
,
i
I
I
I
I
I
!2.2.2
I
I
31,2001 !31.20:> !
I
!
i
!
,
i=1,2,5 From Example Problem 19, the true thermal stress distribution was
EI a 0 h
2.2.21
I ,
31,200
44.':'
40
40 20
{ q! } =
To {.267} .933
o
!
1, 2, 5
=1
o o a o o o o
was applied.
Using the qjT as obtained in Example ?roblem 20 (With qlo = 0), the follOWing product was formed:
11.81
13.45  6.655
11.81
Then
SUbstit~tlng
9h {Omr } ts 0!J(~.
13.45 6.656
84.24
x aT
6.833
.5 928
2.402
A8.41
=1
5",
aT
L&nJ
28.19
{glm})
.50 .50 0 0 0 .025 0 0 1~
26.55 13.34
11.31 13.46
6.656 84.24
L&nl~ = L 50
Then
O:.r
~ultiplying
= 34.3 a T
6.833
.5928 2.402 It remained to find the determinate stress distribution for Qlo = 1. As the determinate distribution the
It is apparent ~rom the above result that the Simplest deter~inate ("cut") structure should be used to compute [glmJ; it is completely adequate.
CLC6URE
{gim}
stresses due to a unit load Q~o = 1 were computed in the "cut" structure obtained when q, = qll = q:s = O. (Note that this is a different choice of redundant :uts from that employed in computing the thermal stresses in Exa~ple Problem 20.) rnus
o o
1
1 1
o
o
1
and finally,
o",T = aT
LO 0 0 1
1 1000 1J
28.19
The general the~al stress problem is complicated by the fact that the material properties E, Gand a vary with temperature. The problem created thereby is primarily one of bookkeeping  co~puting the member fleXibility coefficients (U and the member thermal distortions i j) (d for a structure whose properties vary from iT) point to pOint with ths temperature. The variations of E, G and a with T Will, of course, have to be known from test data. Two additional complications, not considered here, are the lowering of the yield pOint with heating (and the attendant increased likelihood at developing inelastic strains) and the phenomenon of "creep" (the timedependent development of inelastic strains under steady loading). Should it prove necess~ry to analyze for the~l stresses under ~ore than one temperature distribution, the me~ber thermal distortion matrix { d may be gener'a Ltzed eaa i Ly into a 1T} rectangular form such as
26.55
13.34
11.81
13.46
6.656 84.24
where d
is the member thermal distortion asiR sociated with rorce qi from thermal loading condition R. The matrix [C f j ] of ~ ther.na.l
As a matter of interest, a different unit stress distribution was employed With a different choice of "cuts". If the forces q., q , and qll are set equal to zero ("cut"), app Ltca't t on of
coefficients consists of the constant coef~iclents in the d expressions preViously presented. wrile iT T would be the temperature associated with qj jR for condition R. (compare with eq. (26b),
Art. A7.111.
Note: Problems (1) through (9) below may ba worked by either the Least Work or DQ~Y Unit Load Methods. The student will be well advised to try some problems both ways tor comparison.
5000* A
(5) For the "King ~ost" truss in Fig. e, calCUlate the load in member SO. Members AS, BC and BD have area of 2 sq. in. each . ':':le continuous ~ember ADC has an area of 9.25 sq. in. and moment of inertia of 21~ in~. E 1s same for all members.
r"'"""'7i
100"
A
20"
..........
"I
jA I60
4000*
[.10'j 60
f
100" Fig. b
Fig. a
+
Fig. g
j
C,
500.
(6) In Fig. f, AS is a steel wire both 0.50 sq. in. area. The steel angle frame C3D has ~ 4 sq. in. cross section. Determine the load 1n member AB. E = 30,000,000 psi. (7) In Fig. g find the loads in the two tie rods BD and CEo lac = 72 in." ; Abd = 0.05 sq. in. Ace = 0.15 sq. in. E 1s same for all members.
1)
30"
B
I
Fig. c
IE
1000* 1000.
~.raD9ct6~" IF:oII
180"
Fig. i
(3) For the loaded truss in Fig. c, determine the axial load in all members, Values in parenthesis adjacent to members represent relative areas. E is constant tor all members.
C
so.ooos
Fig. h
'F100"I
11
Fig. d
(4) For structure in Fig. d, calculate the axial loads in all the members. Values in parenthesis adjacent to each member represent relative areas. ~ 15 constant or same for all members.
5000*
(8) For the structure in Fig. h, determine 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. Ewood = 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 continuous 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) Reso1ve Example Problem 10, p. AS.15 using as rsdundants the two restraints at one end (couple and transverse force). Solved in this way the problem is doubly redundant as no advantage 15 made of the symmetry of the st.ruc'ture, (11) Add ~NO additional members) diagonalS FE and ZC (each with areas 1.0 ina) to the truss of Fig. A7.85, Chapter A7. Find the matriX of
~
Fig.e
I
l~iI7T~ ~_ t
120"
120"
j
"'.'~'~'"":"
'.
~"._:_.,,: . . . . ' .
,.
T'
~_ "
_..
,._
A8.43
influence coefficients.
.ane ,
21.8 27.2 68.5 65.4
26.4
[Amn] = E
where A1 1s the initial imperfection associated with force qi' Refer to the argument leading to eq. (11) or Art. A8.8. (16) USing the equation of problem (15), above, resolve Example Problem 6, p. AS.l4.
(17) Using the equation or problem (15) above, resolve Example Problem 7, p. AS.14. (18) USing the matrix methods or Art. A8.13, resolve Example Problem 9, p. AS.15.
(12) Resolve 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) Resolve 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 resolve 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 Problem 20, p. A8.36. Campare the results with those obtained from the formulas developed in Example Problem 4.
(19) For the dOUbly symmetriC four flange box beam at Example Problem 15, p, A8.24, determine 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}
REFERENCES (lS) Show that the matrix equation sq. (21) 1s modified to cover the initial stress See references at the end of Chapter A7.
Douglas Dc~a airplane. Ph.otograph. showing simulated aerodynamic load being applied to main entrance door of fuselage test section.
, C f. ( 'j
A8.44
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
Al~D
RINGS
Assumptions In the derivations Which follow the distortions due to axial and Shear forces are neglected. 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 oendtng stresses on a frame crosssection. (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 rupturing stresses for the frame are above the proportional limit of the frame material. This Chapter will deal only with the theoretical analysis tor bending moments in frames ar.d ~ings by the elastic center ~ethcd. FYactical 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 seaplane illustrate both light and heavy fra~es.
In observing the inside of an airplane fuselage or seaplane hull one sees a large number 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 longitudinal 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 framing 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 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 bending forces is nec~ssary.
an~
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. Jeneral
~ethods
of Analysis:
~rinciples
There are ~ny ~ethods of applyin~ the 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 years ~n ~Q~ti~8 ~ir~l~"e :~si~n. The 7.et~od was Qri~i ~atec by ~ullerBreslau~. T~e main cifference in this ~et~od as som~ared to Tiost other methods of s~lution is that the redundant forces are assu~~i ac~ing at a s~8cial :oint called th~ elasec...a t Lcns for t ne r edunoant.s Nhic:l a re tnc ~e~d2r.t of eac~ other.
v
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 M s 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 ?ositi ve moments. ) The following deflection equations for point (A) must equal zero:Q = a Ilx = (, Ily
= a:
(angular rotation of (A) = zero) (movement of (A) in x df rect t on = 0) (movement of (A) in y direction = 0)
Fig. A9.l shows an unsymmetrical curved beam fixed at ends (A) and (8) and carrying some external loading p~, P a , etc. This structure is statically indeterminate to the third degree because the reactions at (A) and (B) have three unknown elements, namely, sagnt. tUde, direction and line of action, making a total of six unknowns with only three equations of static equilibrium available.
p.
From Chapter A7, which dealt with deflection theory, we have the following equations for the movement of point (A):(2)
Ilx=Z~~SOy
:=
(3) (4)
Z i1mds
1
P,
\
A
t
Fig. A9.1
ds
P,
I
B
In equation (2) the term m is the bending moment on a element ds due to a unit moment applied at point (A) (See Fig. A9.3). The bending moment is thus equal one or unity A~ for all ds elements C unit moment of frame.
Fig. Ag.3
..J I.
P
B
Then substituting 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
FIg. A9.2
whence,
In Fig. A9.2 the reaction at (A) has been replaced by its 3 components, namely, the forces X and Y and the moment !1 and the structure A A A is now trea~ed as a cantilever beam fixed at end (B) and carrying the redundant loads at (A) and the known external loading p~, P3' etc. Because jOint (A) is actually fixed it does not sutfer translation or rotation when structure is loaded, thus the movement of end (A) under the loading system of Fig. A9.2 must be zero. Therefore, three equations of tact can be written stating trAt the horizontal, vertical and angular deflection of ,oint (A) must equal zero.
In equation (3) the term m represents the bending moment en a element ds due to a unit load applied at point (A) and acting in the x direction, as illustrated in Fig. A9.4. The applied unit load has a positive sign as it has been assumed acting toward the right. The distance y to the ds element is a plus distance as it is
y
l~~B ,
Y
Fig. A9.4
A9.3
measured upward from axis xx through (A). However the bending ~oment on the ds element shown is negative (tension in top fibers), thus the value of m ;  (1) y = yo The minus Sign is necessary to give the correct bendir.g moment Sign. SUbstituting in Equation (3) and using M from Equation (1):
place of XA, Y A respectively. A and M The axes x and y through the point (0) are centroidal axes for the values ds/EI of the structure. This fact means that the summationsZ Er=
The
ydS
and
In equation (4) the term m represents the bending moment on a element ds due to a unit load at point (A) acting in Y direction as illustrated in Fig. A9.5. Hence, m l(x) x SUbstituting in Y ds equation (4) and USing I'! from equation (1), we lj;obtain,
These terms will be referred to as elastic moments of inertia and product ot inertia ot the trame about y and x axes through the elastic center of the ~e, and for Simplicity will be given the following symbols.
Z
x~
Fig. A9.5
"""'EI = I y ,
xds
"E'r
yiJds
= Ix
, ""
<;'
xyds = EI Ixy
Equations 5, 6 and 7 will now be rewritten USing the redundant forces at point (0).
hence,
  (7)
Equations 5, 6, 7 can now be used to solve for the redundant forces MA' X A and YA. With these values known the true bending moment at any point on structure follows from equation
(1)
          (6)
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
The term ~Z ydslEI is zero since Z yds/EI is zero, thus M o 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 element ds will be given the S~bol 0s ' that is, 05 : MgdsIEI. ~ith this symbol SUbstitution, equations 8, 9, 10 can now be rewritten as follows:(ll)
  (12)
elastic
Fig. A9.6
inelastic bracket, as shown in Fig. A9.6. Since point A sutfers no movement in the actual structure, then we can say that point (0) must undergo no movement since (0) is connected to pOint (A) by a rigid arm. Thus equations 5, 6, and 7 can be rewritten USing the reduncarrts Xo' Yo' and Ma in
Z 0s x  XoIxy + YoIy
=a
 
 (13)
Solving equations (12) and (13) tor the redundant forces Xo and Yo we obtain,
A9 4
BENDING
MOMENTS
IN
FRAMES
AND
RINGS
"'~S
IX (I
\ y ~)
"'~S
MX(~\ r::)
Table A9. 1
      (14)
TXIY
 [Z0s X
60 sY
(!t;)J
)
AB 3~:10 15
     (1,0) Be
Y e
I
y
(I _ I 'x?
TXTY
A9. 3 Equations for Structure with Symmetry About One Axis through Elastic Center.
~12 2
30
360
If the structure 1s such that either the x or y axis through the elastic center 48 a axis
CD 30=iO 15 3
Sum 32
wy2'::2250
12
120
150
Ix:
750
16800
iy=~3=0 12x3
3456
of symmetry than the Droduct of inertia Zxyds/EI = I xy = zero~ Thus making the term I xy = 0 in equations 11, 14 and 15 we obtain,
660
=z s
tds/EI
            (16)
r;Yo 
_ Zsy
             (17)
              (18)
_Zsx
r;;y
The terms i x an1 1 are the elastic ~OT.e~t of inertia of each portIon of the fra~e about its centroidal x and y axes. Si~ce I is constant over each pert ian the centroidal T.oment of inertia of each portion ~s identical to that of a rectangle about its centroidal axis. To exp'ln In for member AB: ~ '_ llO Re re rrv.ng to Fig. a,
ix
1 bh3 1 =12 = 12 I x x3 30~
A9.o1 E:m.mple Problem Solutions. Structures with at least One Axis of Symmetry.
,Exa~ple ~oblem
at ;:JOints A and
D, and carrying
a single load as
Shown. The
B!7:JC I
~'~18"_
10 lb.
30" I
liT
IX I
!
I
+=t=
= 7:': _,,0,0
<,I
A
,ds' 30' h
Rf e erring to Fig. b ,
uI '
Y
1
Fig. a
ly
.09 (negligible)
=3
The distance from the two reference axes to the elastic cer.ter can now be calculated:y
8T
!
A X  'fT11"TT]   
IY 
D   rrrtrrrr
I
I
ds=30=b
Fig. A9.7 T'ne r irst; step in the solution is to find the location of the elast1c center of the frame and the elastic moments of inertia Ix and I y .
=~ Zw
= J::N
Having the moment of inertia about axis xx we can now find its value about the centroidal axis xx ot the frame, by use of the parallel axis theorem. Ix ~ Ix  ~w(y~) = 16800  32 x 20.625~= 3188
: 660
32
= 20.625 In. =
0
0 32
oJ yl ~.!=.!= h
 I ,
U...l.
I
Fig. b
I y : I y  "w(x')
= 3456
 32(0) : 3456
ss,
The problem ~ow consists in solving eq~ tions (16), (1:) and (18) for the redundants at the elastic center, namely
..
~
~
..
A9.5
_ ~ _ Area of static Mil diagram o  ZdS!I  Total elastic weight ot structure M Moment of' static Mil diagram about
X  ZTsY 
along the neutral axis of the frame members. In Fig. A9.10 the area at the Ms/I diagram equals 0 s = 22.5 x 24/2 = 270. The centroid by siillPle calculations of this triangle would fall 10 inches from B. Fig. A9.ll now shows the irame With 1ts MslI or lts 03 load. Os is 0 s = 270 l' .. positive since I1s is B...;;:,:2 C pos rt ive . The next 1375" step is to solve the x ..:x equations for the 0: redundant at the : 20.625" elastic center. The ; I signs of the distances A 'Y D i x and y from the axes x and yare conven
o 
x ax1s
Yo Zsx Iy
Moment of static Mil diagram about y axis ==""'==::'::T;~="';=::':; Elastic moment ot inertia about y axis
Thus to solve these three equations we assume a static frame condition consistent with the given irame and loading. In general there are a nlli~ber of static conditions that can be chosen. For exanple in this problem we ~ight select one ot the statically determinate conditions 11l~strated in Fig. A9.8 cases 1 to
~ust
' fl
I
Fig. A9.1l
t i ona L,
,.
I
Thus,
= 32 (from Table A9.1 = 8.437 in. lb.
 (270)
~I ,
::T1
, 10
Xo
= ZsY = 270(9.375)
Ix
3188
=0.7939
=
lb.
0.1562 lb.
Fig. A9.12 shows these values of the redill1dants acti~g at the elastic center.
~17.
To illustrate the use at different static conditions, three solutions will be presented with each using a different static condition. Solution No.1 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 given in Fig. A9.9. The equations
75
;:8
9.375
20.625"
t F'9
8.437
):1562
B[rig.
7.5#
L06'AJ
Fig. A9.12
~~.81
J
6.06
FIg. A9.13
lo. ,e
A9.9~
Dol ,""'A , ,
'2.j~
Bn
~81
Ms
is posrnve ,
tension
on inside of frame.
The bending moments due to these redundant forces will now Je calculated.
M A
=
8.~7
 .1562 x 12
Fig. A9.10
Me
for the redundants require 0s the area of the ~s/I ~lagra~. Fig. ~9.10 shows the ~s/I curve which is obtained by divlcing the values in ~lg. A9.9 by the te~ 2 which is the moment of inertia of ~ember Be as given in the problem. Since the equations for Xo and Yo require the moment of the Ms!I diagram as a load about axes trIough the elastic cencer of frame, the area of the Ms/I diagr~n will be concentrated at the centroid of the diagra~ and along t~e centerline of the frame, or more accu~telY
= 8.437 = 9.437
=8.437
Me
NO
 .7939 x 9.375
+
.1562 x 12 :;:
14.00 In. lb.
+
.7939 x 20.625
.1562 x 12
These resulting values are plotted on Fig. A9.12 to give the bending ~oment diagram due to :he redundant forces at the elastic center.
A9.6
BENDING
MOMENTS
IN
FRAMES
AND
RINGS
Adding this bending moment diagram to the static bending diagram of Fig. A9.9 we obtain the final bending moment diagram of Fig. A9.13. The final bending moments could also be obtained by substituting directly in equation (1) ua tng subscript (0) instead of (A). ThUS,
M= M s + M o  Xoy
+
= Z0s x
ly
= [45(10)4D5X6300(12)90OX12]
3436
YOX
      (19)
a.
= 9.375,
M s
=0
=0 +
D.
liB
=30+51.56.7939X9.375+2.656(12)
fi~st
AT
~OINT
= 12,
= 20.625,
M s
=O.
Solution No.2 In this solution we will use Case 4 (See Fig. A9.8) as the assumed static condition, that is two cantilever beams with halr the exterual load or 5 lb. acting on each cantilever. Fig. A9.14 shows the static bending mcment diagram and Fig. A9.15 the Ms/I diagram.
MD
7939(20.625)
2.656 x 12
Solution }Io. 3
In this solution we will use Case 5 (Fig. A9.8) as the assumed static condition, namely a trame with 3 hinges at points A, D and ~, as illustrated in Fig. AS.16.
"5=45
"~405
45
"62515. 6~5
Fig.A9.14
1"" ~18"I
Ill<
~: 30P
I I
T'::
'
r 12"
tx
Fig. A9. 16
.08= 00
30"
Betore the bending moment clagram can be calculated the ~eacticns at A and Dare necessary. To tind VD take moments abou~ point A.
30
90
!1
10
Fig. A9.15
30
Fig. A9.l5 also shows the results or calculating the ~s value for each portion of the Ms/I diagram and its centroid location. Substituting in the equations for the redundants we obtain,
VD=2.5
HD"'l
ZIlA
= lOx6 =0
2~VD
hence, Vo = 2.5
I'1o
=Z0, Zds I
=  (45405300900) _ _ ,
3.2
 01....,6 in .u ,
=0
= 0 =10
+ 2.5 + VA
X  Z0 sY  (45405)9.375+(900300)(5.625)
o  ~ 
3188
To find HD take moments about hinge at E at all rorces on frame to right side of E and equate to zero. ZME
=0.7939
lb.
= 2.5
x 12
+ 30HD
= 0, hence nD = 1.
A9.7
The ~rame static bending noment Q~a~ram now te calc~lated and dra'NTI as shawn i~ Pig. A9.17.
~an
axes through the fr~e elastic cer.ter are then calculated as 12 ~d 0.625 inches ~espect ively.
;  E0s _  (390) ; 12.19 in. lb. Eds/I 32
Xo ; Z0sY Ix Yo
(5)
30 30
9.375 \
T
I
I
t\I
I
1[1:1 . 
,,0s'150I
I"
x = 12
...J
,y=0.625'
~:i;c:n~er
i
I
"'l50
(6)
548 3188
= 0 203
Ib
c ,
= Z~;X
;  t.;~0) = 0.1562
lb.
The final ~oments at any point can now be found by use of equation (19), namely
20
20.625
  (19)
x
Fig. AS. 17
D
Pin Pin
= 12,
= 9.375, Ms
= 30
SUbstituting Mg
= 17.76
tions)
=30+12.19(0.203)9.375+0.1562(12)
The Qoment diag~~ is l~beled in 6 parts 1 to 6 as indicated by the values in the small circles on each portion. Most of the calculations frem this ~oint onward can te done conveniently in table form as illustrated in Table A9.2.
Table A9. 2
x = 12,
Sub't , in (l9)
= 20.625,
M s
=0
Mom. A' I Os = Dia Area for gram of A Beam Por Mom. Section I tion Dia.
.
.
5,
A,y
The final complete bending moment diagram would of course 08 the s~me as dravm in Fig. A9.13 for the results of Solution No.1. Example
?~oblem
A,
_12.00
Os
_0.625 9.375 9.375 9.375 9.375 0.625 1800
93
2.
~450
3
2 2 2 2
150
. 60
15
45
30 _10.67 7.5
320 281 50 90 70
211
3
4
 6.67
.
4.00 8.00 i2.00
22.5
5
6
180 _450
 90
150
~390
720 844
93
1800
I
rectan~~lar closed supported at pOints A and B and carrying the external loads as shown. The reaction at B due to rollers is vertical. The ~rame at po~nt A is continuous through the joint but the react~on is applied through a ?in at the center of 'the j oint . The pr obLem is to d et.ermtne the bending moment; c tagran. fra~e
Sum
 540
658
In order to take mo~ents of the ~5 values in cojumn (4) of the te.oIe , the centroid of each ~ortion at the diagran must be ce~e~ined. For example, the centroid of the ~wo triangular bending morr.ent corticns marked 1 and 6 1s .667 x 30 from the lower end or 20 inches as sho'Nn in ?lg. A9.l7. Thus the 1istanceS x enc y from this 05 location to the y and x
' I n
IB
,P1"2M
w = 10*.1 in.
j
L :: 30
y 4 I' 1
.I !
,L,L I
.2401
12'
I I
1,A
"'"
1=3
x
L'24I~2f)\
D
c:
\1 B
ii'~""""7
(2)
C 3
~240
I:2
Fig. A9.18
'""*'
.1.
A\oo15~
(5'
240 240D
II c
"~~""':'O:i'";;"
: '~~:.
A9 8
BENDING
:>10YlENTS
IN
FRAMES
, AND
RINGS
<~4500
IE I I
1
:s
". 00
C~.'J:Jse c'
5':;.<:1: .::'::o..:T.e
.' 'If
,a
i
r , [E7.
(7)
,
out
A
D1
I I,
~4500
Fig. A9, 19 diagram fa, w loading
: cut iA
~1
.
9
D 3b ...
360
M,
M,
stc '::1'.:: \"5 j~:;;i '.:.; moment ::2.::I""::.;7:. CC i s 5 ':=:L::. Lm ':7'.'::' frq;:s _0 assuror cut on membe r .'B ~ "':'3~ ,:::;e c i:::"': .. :3 10 l:il...ustr?..ted 1" ::.';s. "::' .... :'9 ar;c GiJ ?':r '' J S1:::c:li:ity ::12 ,"._" metent c urve :',2.5 crown in "'c:... ree ~ar:;5, ..' '~ 83.C:". . ':::=:::5:::::'2,l!1g only one of tns ... ext.ernn l  ' t_",,,, i. :;"'3.::'.5 C , 20 S:~:'N on the s t.ruc tur e . ?L;s. A.~.::'2, .. :: 9 ::J,c;;:snt C::'::;V,:;S. these resulti::~ bend tng po!'tions :J1' tnes a benc t ng ::lO::;:.en: ,j 1 '':.:;:':o..'~.s numuerec ' to 10 2.:8 5 ['1 own tnc parerrtnes ~2 on each ;Jo:'':icn.
.
'::,8
;.~
\...
~o"',.,
... ~ . .
~
"",.,..~
,,~
.~,,
,.,~o""
0"
, ,_.,
,~
SoL;t~:n:
T~,e first step is to find the j ocat i cn of the elastic center of the rrame . Due to sy.mnetry o~ rrane about the y ex.;s , 'the elastic center 1.'111 be on a y axis 'through tt).~ middle o~ the rraae . The vert.t ce.I d~s"tance y measured from a ax i s through AD equals,
:''..18 nex"': sts'J '" ;:,'ae s:lutioD ::':'.3:'3t3 '::l~ 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.:'.::: y axes t nrougn t::e 8l3.S~~:: c ent cr. T;;'858 s i:::~l calC'llations can jest '.::e cone in tab l e "ern 3.3 i Ll.uat.ra ted in ~2.ble .'.9.3.
Table A9.3
Portion Area of M, of Ms Diagram Portion ( ) , A
1
2
....
Os ,
dist. y
dist.
X
to Y
Axis
to
Z(dIS) Y y = =
Z
os
372 33.5
= 9.67"
A 480
Axis
sx
I.
OsY
4958 25794 4958 2726
,
15
1440 3 ~
7200
i1 
rhe next step IS to deter::l.ine the elastic moment of inert ta of the rrazae a::: out x and y axes throug~ the elastic center of the f:ame.
~oment
3
4 5 6
a
15 15
5
01
7200i
 5.67
, . 9.67
7206 9000
= Ix
653.9
:
1800
17406
Members
An
k
and DC,
14.33') 2
45000 4112501
7.5
15 5 15
5
, Ix =( ~ x '3 X
14.33
7
8
2.33 540000
843751161212
 83880 i
,
326352 3064
(1'
Ix
'3
x 9.67') 2
=
= = =
2()0.9
9 10
 9.67 I  3.67
168750
16200 1
I
!
Member BC
(14.33') = (30/4) (30/2)(9.67")
2700 77700
9.67
13500 747225
26109 4645
1540.0
1402.7
Sum
Ix
3797.5
center,
,
=Iy
563
l'!:J
AD and BC.
30'
30 3
in.la.
Iy
= =
Xo
Yo
lb.
1125
= ~0'Jx Iy =
(747225)
~289
= 14.1.28
~.
:"b.
acting at
=
'y
3600
5289
Fig. A9.21 shows these redundant r,JrC2S ;;":J" t:. 8 e las c ;c cen:er. . .. ';'9.22 . . j. "'; am cue ",,0 0 ::>hONS tne bend.ing moment c Iagr cnese
~
redundant forcas. 1';:',8 calculations 'Nith rer ar, ence to fig. A9.21 are 
VEHICLE STRUCTURES
A9.9
z ~ = It
I
(36) ~ 113.1
IT
1
9 67 .
4124
,''r0
Fig. A9.21
The elastic moment of inertia about x and y axes through center paint of ring are the same for each axis and equal
Ix = I y =
nr~
= n x
18~
= 18300
M A M8
~
= 2018 = 201S
= 2018
141. '28 x 15
1.22 x 9.67
X
= 113
= 84 = 4153 = 4124
141.28
+ 141. 28
15 + 1.22
14.33
MD
= 2018
+0
,
1.22 X 9.67
Combining the bending ~oment diagrams of Figs. A9, 18, 19, 20 with Fig. A9.22 would give the true or final bending moment diagram. Example Problem 3. Circular Ring.
The next step in the solution 1s to assume a static ring condition and determine the static (Ms) diagram. In general it is good practice to try and assume a static condition such that the Mg diagram 1s symmetrical about one or 1t posSible about both x and y axes through the elastic center, thus making one or both of the redundarrt a Xo and Yo zero and thus recuc Ing considerably the amount of numerical calculation for the solution of the prOblem.
Fig. A9.23 shows a circular ring of constant crosssection sUbjected to a symmetrical loading as shown. The problem is to determine the bending moment diagram.
50 50.
In order to obtain symmetry of the Ms diagram and also the Mall diagram since I is constant, the static condition as shown in Fig. A9.24 is assumed, namely, a pin at (e) and rOllers at (f). The static bending moment at
points (a ) , (b), (c) and (d) are the same magnt tude ann. equal,
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 and :2l s Y values.
~s, ~sx
s 1s the area of the Ms/I dtagram, however 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. A9.23 Fig. A9.24
and (2).
Hence
0'
M=88"*
= Pr~(c
sin
~),
+Xo=CX:
I~~
M=177"# 1~
~
\~
since there
'
.,~
M=88"#
"
,....,_.,c,;"o:
Fig. A9.26
Solution. Due to s~etry or the ring structure the elastic center falls at the center of the ring. Since the ring ~as been ass~ed with constant crosssection, a relative value of one will be used for 1.
= 15000
5052 = 20052
Since the centroid of the Xs diagram due to symmetry about ~otn x and y axes coincides
( 11
A9.10
BENDING
MOMENTS
fra~e,
IN
FRAMES
AND
RINGS
with the center or el~stic center of the the te~s Zsx and Zsy will be zero.
Solution: The fi~2t step 15 to 1ete~lfle the elastic center of the f~ame and the elastic ~o~ents of inertia. ~able AG.4 shows the calc~latlons. A reference axis x'x' has jean selected a~ the ~icpeint of the side AB. Since a static :r&T.e condition has been selected to ~ke the M s diagram sJ:mnetrical ~bout y axes t;Iough the elastic center (see Fig. A9.2S), it is ~ot necessary to determine I y si~ce the redundant Yo will Je zero due to this sJ~T.e~ry.
Table A9. 4
SUbstituting to determine the value of the redundants at the el~stlc center we obtain, Me
= :Zd0S" ~ ~
= Z0,y
Ix
= 20052(0) = 0
18300
Yo
=0
acti~g
at the elastic center and the bending moment diagram prOduced by these forces. Adding the bending moment diagram at Fig. A9.25 which is a constant value over entire trame of 177 to the static moment diagram at Fig. A9.24 gives the tinal bending moment diagram as shown in Fig. A9.26. Example Problem 4. Hull Frame
Member
BDB'
Length cis I
w' "f
cis
'NY
IX'X' : tx .. wy2
t!: ..
n
: 49. i
AB
1.0 60.0
I.
0 0
42 42
0 0
,
,
A'B'
AC CA'
Fig. A9.27 shows a closed frame subjected to the loads as shown. The problem is to deter.nlne the bending moment diag~.
lJ
odl){)"
2. 2.
, ,
Totals
262140
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
z.~,
T A~ L'38.4
h'"''
Fig. A9.28
L ~~~~.
~ ~:_jY
_ 1 tx  12 b L h"
Let y = distance from X'X' ret. axis to centToidal elastic axis XX.
"wy y=zw=
Ix
221.2
1468
;z
6.64 in.
= Ix!
 6.64"(Zw)
The next step in the solution is to ccmputs the static moment elastic weights 03 and their centroid locaticns. In Fig. A9.28, the static frame condition assumed 15 a pin at point A and rollers at point A' , which gives the
ANALYSI~
~oment
curve as shown
1/2
t
x=
10o
38 . 4.
(4690 x 100 x")xdx 78400
sl~ered
t.ern
~3
bending moment curve will be con~ in two parts, namely (1) and (2). The represents the area of the Ms/I
= 1/2 [469~
x' . 10~ xj
784.00
38.4
21.77'
c tagran.
Thus for pottion (1) and (1' )
o
Vertical distance from line AA' to centroid 21.77 x 24/38.4 13.~". The static moment weight for Ale' 1s same as for AC, thus
=
~
~AC + ~A'C'
2 x 784.000
= 1568000
sin au
2.
30(1 0.867 
a  sin u
0.524  0.5
= 10
in.
rig. A9.29 shoWs the trame With the moment weights a located at the centroidS, together WIth the redundant forces Me and Xo at the elastic center. It makeS no dltterence where the frame t s cut to form our residual cantilever, it one or the cnt races 1s attached to elastic center and the other Is considered fixed. With the elastic properties and moment wel~~ts known the redundants can be solved for:
_(,.=.10",Cf7=0:::00~+ .=1,;;72;:;8,.,0;;0...:+...:1:::5",6",80",0'".) , 221.2
124.30'#
X :: o = ZX'sy I xx
For portion (2) of the M s diagram the area of the ~/I diagram Which equals ~s 1s
,05(.:1)
=PTrlitj
= 1,007,000
Consider member AC,
~rom free body clagrarn of bottom portfon of frame (Fig. A9.31~ the equat~on for bending moment 6000 ;4690 6000
= 98.5#
a.~d
. . . . . . . . 11
A
I

A'
C
I~
Yo 1s zero because of the symmetrical frame loading. The f1nal or true bending moment at any point equals
>0.,"'7
AC equals:
Mx :: 4690 x 100 x Q
Area of
aq\..ials,
1/2
~/I c~rle
For point C
%
o
38 . 4
= 32700
Hence Me = 32700  12430  (98.5 x 60.64)
14.300'~
= 1/2
0]38.4.
Jo
~/I
784000
fig. A9.30 shows the general shape of the true ~ra~e bending ~cment diagram.
Example Problem 5
Distance to centroid of
;"C iroJ1 A.
curve along line ?ig. A9.31 shows the general details of one
A9.12
BENDING
MOMENTS
IN
FRA:YIES
AND
RINGS
half of a SYTI'T.8t:'iC3.1 hull rreme that was used in an actual seapl~ne. The ~ai~ external load on such frames is ~he water ~ress~e cn ~he ~ul: bottom pl~ting. The hull bottom s~rtngers transfer the bottom pressure as cancentrations 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 tangential 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 sections 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 components. The sum of the vertical components should equal the vertical component of the bottom water ~ressure. Table A9.5 showS the complete calculations 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 axis. A reference horizontal ~is X'X' ~as been selected as shown. All distance recorded in the table have been obtained by scaling fram a large drawing ot the fr~~e.
~or
The moment of inertia of the frame cross sections are given in column 2 of Table /1.9,6 which have been determined from a consideration of the actual dimensions of the frame members.
The static condition assumed for computing the M s moments is a double symmetrical cantilever 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 Zsx will be zero. The calculations tor determining the static moments M s in Column 8 of Table A9.5 are not shown. The student should refer to Art. A5.9 of Chapter AS to refresh his thinking relative to bending moment calculatlor$ on curled beams. 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 columns 3, 9, 10 permits the solution for the red~~dants Me and X as shown below the table. o The final bending ~oment M at any pOint on the frame is by Simple statics equalS,
I'
I~
~~
I~
l~
Fig. A9.32
A9.13
Table A9. 5
1
0
3
Elastic Weight
w
4
Arm to Ref. Axis X'X'
w
6
w y'2
8
Static Moment
9
Moment Weight Os ::
10
11
~10
12
13
Total
Moment
y'
Arm
"
~,j
ds
~.=
I
= ds
1
'0 Axis
XX
~
es y
Xoy
MS
:: y'
26. 1 24.4 21. 4 17.3 12.0 5.9 3050 2850 2497 2025 1403 79700 69500 53300 35020 16820 17. 2 15. 5 12. 5
~ 1 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
9020 "
1 2 3 4
5
7.6
7.6
6 7 8 9 10 11 12
13
14
15
~.hine
7.6 7.6 7.6 6.4 6.4 6.4 6.4 6.4 6.4 6.4 6,4 6.4 6.4 7.0 7.0 7.0 7,0
7.0
.25
. 25
25.60
25.60 25.60 25.60 21. 30 18. 82 18. 82 18. 82 18.82
.25
. 25
 6. 3
12.6 19. 1 25. 5 32.0 38.3 44.6 50.8
151 0 I
161 323 488 543 602 720 _840 959
892
0
8.4 3. 1 3.0  8. 9
1017 4060 9320 13850 19250 27600 57500 48750 1780 1980 1470
f:
16
18
7, ,
O. 58 O. 58
19 20
Keel Totals
o. 21
O. 14
811.48
O. 39
 32  3'  24  131  9
7228
840
588 423237
1220 31. 2 2880 71. 6 15. 2 6230 159.3 286.5 21. 5 11260 18470 473.0 28.0 34.4 29020 618.0 40.9 43180 814.0 58980 1110.0 47.2 77330 1455.0 53.5 59.9 99030 1865.0 111700 63.0 64. 3 149700 86.7 67. 3 221400 128. 3 70. 3 271200 105. 8 72. 1 I 3086()0 64.9 73.7 327900 45.9 75.0 332900 I 7315.0
93.7  637 2420  6160  13220  21250  33100  52400 . 77800 111300
.. . .. .. .. .. .. . .. .. .. .. .. . .. .. .
..
9020
16650 15020 12100 8130 3000 2910  8610 14720 20830 27100 33300 39700 45600 51700 57900 61100 62200 65100 .68100 70000 71400  72600
890 6020
10710 14840 17510 18590 17650 13300 5540 4360 16610 32110 41600 78500 147300 194100 229600 247500 251280
348106
= ~Os
''w
XX
:: (7315.7 x 1000)
311.;:1
= 9020"#
x, :: Z?s y::. [
"Total.
Moment
Mo 
XoY
Columns 11 and 12 record the values of M o and Xoy for each station poIrrt . For examp.Ie , trie va Iue of Xoy for station (1) equals  (368 x 17.2) : 16650 and for stat~on
(20)
2401b
l.. 6'
,
1
6" ooi
= [968(73.7)J = 71400,
Fig. A9.44 shows the shape of the final ~cment carve as the result of the values in c ojunn 13.
A9.5 Unsymmetrical Structures.
Example prootem Solutions.
120~
I I
T I !_l_!..._x
6"
Y::5.242
:E.C.
c , rrI
~;1
I
x
..
9"
5. 032
Fig. A9.34
'1
10"
.L
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: 
(12/2)
10/1
= 31
~ra~e
= Zds/1 =
.'
I '1' ...,
A9.14
BENDING
j
MOMENTS
IN
FRAMES
AND
RINGS
me
AB to the
x = (15)0
+ 6
31
x 6
10 x 12
= 5.032
in.
Fig. A9.36 shows the centraid loc~tlQns uf the 05 values along the center line of the frame. The ~cment of these 0s values about tte x and y axes will now be calc~lated.
Z0sx
y from line Be to
elas~lc
y = 15 x
The
7.5 + ~lX 0 + 10
x 5 = 5.242 in.
Z0sY = (2160)(5.242)
(4860)(6.758)
= 70290
(21600)(2.258)
The 'ffilues of :he recunGants at t~e elastic center can now be cal~~lated USing equations
(ll), (14), (15)
J
nameIy
Ix
= (~)'
+ (15)(2.258)' + (1:)(5.2421" +
(i~)'
M o
= ;;~~7I
=  (;~620)
= 923
In. lb.
+ (10) (0.242)'
1'0
J
=606.51
2x12 2
= (15)(5.032)'
(10) (6.968)"
= 942.96
(~)(0.968)(5.242)
IxY = (15)(5.032)(2.258) +
+
68.46
(10){6.968)(0.242)
= 217.74
The next step 1s to assume some static trame condition and draw the static bending moment diagram. Fig. A9.35 shews that the frame has been assumed cut near pOint C which gives ~NO cantilever beams. The oending moment diagram in three parts for this static condition Is also shown on Fig. A9.35.
=  [Z0sx Yo
Z0sY
I y (1 _..::s:L)
(~)J
4
rxry
= L.:
132.36 lb.
942.96 (1  (606.01)(942.96)
T'i .
ISl=~2160
t '"
6"
5.242
2)
9.758
M s diagram
~1080
"I
10S3"'~60
A
.o~
1_f:Jx t U l
OS2'::
~olnt
from eq.
 XoY
+ Yox,
for exampl e ,
'1
258
600 6 750
. C.
Consider point A.
Fig. A9.35
FIg. A9.36
= 5.032, y = 5.242,
+
~s
= 1440
+
The 05 values which equal the area at the M s diagram divided by the I values of the part1culqr portion will be calCUlated.
Ma = 1440
(5.032) Point C.
=
(132.36)
0s,
~s.
x = 5.968, Y = 5.242, M s
=0
+
Me
=a
(132.36)
= 2160
21600
4860
28620
ThuS, the com~lete be~ding ~cment diagr~ could oe determined by computing several more '/alues such as ~oint D ar.d the ~xternal load points.
A9.15
Total Ix
=6451.07
= 13.~
30(0.667)' : x IS' +.lx 12
P2
= lOOt
I
x 31
_ 1 Member AD, I y 12
1 x x:r
Totally
= 10919
= 1002.9
ly'
Fig. A9,37
Solution: (30)(10)
= 253.92
= 21 (15) (1.657) =0
=1.5 30
+
31. 67 + 3.Q. 3 1
+
30
= 6!::;
~.,
,1:8
= 497.1
Member AD,
I~J
The location of the elastic axes will be determined with ~eference to assumed axes x'x' ~~d y'yl as shown on rig. A9.37.
=
Tot~l I~J
= 17G3.9
~(0)+()(151+()(301+(4f1(151 =
65.58
15'
The next step in the solution 1s to assume a static fra~e condition and draw the M s diagr~. Fig, AS.38 showS the assumed static 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.
750"
65.58
=
11.657'"
600 600
\
These distances and y locate the x and y e1ast!c axes as shown in Fig. A9.37. The elastic ~oments of inertia and ~he elastic product of inertia will now be calculated. Calculation at Ix Y barAB , x Tlxl..(,e3'''+'1'''7'')=1'7. . . .;;:~1 dem 3 1.5 ~ . "'tV ..l. .<)~. 0
. .:::. . . . .
",1111 "
Due to Pl~l,
"
c
Fig. A9.38
=D
Member CD,
Ix=~
Member Be. Ix
(10" I
= (31367}(~3.343<i)
l~ (31;367)
1 +12x 30 x
The ~ext step is to compute the '~lue of 0s for each ?ortion of the ~oment di~gran. 0 5 is the area of the !'1s / I diagram. For reference the ~ortions of the M s diagram have been labeled 1 to 4.
.t
A9.16
BENDING
~co x .. = ::000 03 = .~G 1.;J
MOMENTS
IN
FRAMES
AND
RINGS
?ortlon Portion
(1)
(2)
30
03 =Zx3e;s
~
600
31.67
~~j ~oir.t e~~als the sta:::.c M s ~lus the ~omen: due to the reduncant forces as S:lO'.\TI in !"i.;;;. ).'; ... 0. b~r.ding ~om~~: ~t o~i~:na1
The
~'t:~
o,~
2
31.67
TO t::
B
39:~8
= 5CO(120)
_500(1 ) 2
= ,3000
18.343!
!
I~
:
165
~l~C.
L.i
= 13125
 5000 :: 8125
~ :4.674
.
:
Ai
;""15 I
iD.L
11.657
167,,
\
286 240
fi'"1307
The individual 0'8 values are now concentrated at the e.g. of each of the Ms diagrams. Fig. A9.39 shows the location of the ~s values with respect to the x and y axes thro~~ the elastic center.
0s:3167
I
11"":""'15''1 .
167'4;j[~307
Fig. A9.41
Fig, A9.40
Consider
MA
~oint
A: 
M s :: 0
+
= a  124.8  4.674 x :5
M s :: 600
31.07 x 11.657
= 167
Point B.
I
,
in. ::''::1.
I I I
Ds~6000~
;
I
'....5_
I'
I
l8.
15 34 31
13.343
~
16.3~
'
!
11. 657
,
Point c.
!'Is
= 165 in.lb.
=0
+
r 15'L.........j
Z0sx
. .
i
~
0 s .:5000

Me
=0 
4.674 x 15
= 314
Ms::
0
4.574 x 15
+
1::1.1'J.
Fig. A9.39
Point D.
MD
= 6000(15)
+ 3167(5)
=105835
= 208659
= 0  124.3
31.07 x 11.657
l0sY = 6000(8.343)
+ 3167(15) +
=307
AS.6 Analysis of Frame with Pinned Supports.
in. lb.
3958(13.343) 5000(11.657)
The values of the redundants at the elastic center can now be calculated.
MO
=~ ~ Zds/I
= Z0s
(8125)
65 .58
= 124 8
n,
1b
I y  Z0s x ~)
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 points A and D instead of fixed.
x.,
Ix (1 _
;xy')
x"y
=208659 = [l0sX
(105835)(~)
10919
6451 [1 
(1753.9')
6451
The first step will be to dete~ine the elastic weight of the fr~ne, the elastic center location and the elastic moments of inertia about axes through the elastic center.
31.07 lb.
Yo
 l0sY
I
I y (1 
IXly )
208659C1753.s11 6451 ~_ 4.674 lb. (1753.9') 10919 x 6451
(;~
= [105835
10919[1 
J
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 unit moment. In this chapter this tern has been called the elastic weight of the element. 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 applied to a rigid support, the support suffers no rotation since the support is ~igid, therefore 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 moment acting on a hinge would have infinite angle change or rotation and therefore a hinge or pin possesses infinite elastic weight.
4 46 14'n
_14~!30'54 ~14.46
"F' 1 I"
L
230
~'
10 lb.
:
I
~s=270
I ,
'"
Fig. A9.44
Fn ld(~L
yJE.C. Y Fig. A9. 43
,:: oo!!J
Fig. A9.45
"
Fig. A9.46
Fig. A9.45 shows the bending moment diagram due to the redundant XO ' Adding this diagram to the original static diagram gives the final bending moment curve in Fig. A9.40.
Ag.7 Analysis of Frame with One Pinned and One Fixed Support.
Fig. A9.42
Due to symmetry of structure about the centerline y axis the elastic center will lie on this axis. Since the two hinges at A ~d B have infinite elastic weight, the centroid or elastic center of the frame will obviously lie midway becNeen A and B. Fig. A9.43 shows the elastic center 0 connected to the point A by a rigid bracket. tor frame is infinite because or the hinges at A and B.
~ds/EI
Fig. A9.47 shows the same frame and loading as in the previous example but point D is fixed instead of hinged.
n4i
I Fig. A'.
y
The elastic moment of inertia about a y axis through elastic center is infinite since the hinge supports have infinite elastic weight. Ix is calculated as follows: For Portion AB For Portion CD For Portion Be
xAi   x
The support D has zero elastic weight and the pin at A has infinite elastic weight, therefore the elastic center of the frame lies at point A. Tne total elastic wei~~t of frame 1s infinite because of pin at A.
=..lx.l.x 3 3
3 3
30'
= =
3000
The elastic moments of inertia will be calculated about x and y axes through A.
Ix
:J:..xl x 30'
= 24
3000
x ~ x 30A 10800
Ix = 16800
Fig. A9.44 shows the static frame condition assumed to obtain the M s values. The value of 0s for member Be equals the area ot the M s curve divided by I tor 8e, hence s = 4S x 24 x 1 = 270. The centroid of this s value is 10 inches from pOint B. redundant forces at the elastic center can now be solved for
'~e
= 16800 (Same as previous eXaID?le) I y = (3 x24' )+(; x ~ x 24') = 8064 30 x 30 x 12) = 7920 Ixy = (;0 x 15 x 24) + (2 24
The static frame condition will be assumed the same as in the previous example problem, hence 0s = 270 and acts lOW from B (Fig. A9.48).
/ ' " 5270
2' 2'
r~lO...i
3~'
'I
1S ' n l 11.92~
"'
rw
W1l
270 ir.f:ni ty
y
X = (270)10 = 0 = Z5 1y infinity
tl;! Xo
, 'Yo
Fig. A9. 48
' !MO
"""
~.
5038#
'i 19
28.45
~1
r, I
'a
A11.92
i. 2582*
~ 6.1'
Fig. A9. 50
Fig. A9.49
=0.482
Since the frame is ~~symmetrical, the x and y axes through the elastic center at A are not prinCipal axes, hence
lb.
CHAPTER AlO
method tr~t 1S widely uSed by engineers in cetermlnlng 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.
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
'Xo
:f),Mo
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 crosssection.
w,
x.i,
EX
A
jyo
r
ly
fx
Fig. AlO.2
:11
Fixed
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 rrce the yy and xx axes, we can wr t te
This cantilever structure wIll sutfer bending 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, X o 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 statically 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'
wnere A is t~e area of the calumn crasssection and pa and Pb the moment of the load P about the xx and yy axes respectively. If we let Pa = M x and Pb = My, the above equation can be written,
  (1)
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
~or
=:1
+ Xoy + Yox
          (2)
In Crapter A9, Art. A9.3, the equations Mo, Xo and Yo were derived. They are,
Method of Analysis due to Prof. Hardy Cross. See "The Column Analogy" Univ.lll. Eng. Expt. SU. Bull. 215.
AlD. I
AlO.2
o = Zds/I ' YO = M
 Z0sx
Iy
(4)
at any
The ter.n z0s represe~ts the area of t~e static Ms/I curve. (E r2s been assumed constant 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 elastic weight of the fr~~e and equals the s~ o~ the length of each ~ember tl~es its width which equals 1/1. Let this total ~~~~e elastic weisht be given a new 5J~bol A. In the exoressions ~or Yo a~c Xo the terms Z0s x and Z0sY represent t~e TIcnent of the static Mil curve acting as a load abcut the y and x axes respectively passing t:;rcugh the frame elastic center. Therefore let Z0s x be given a new symbol My and Z0sY a new symbol M x With these new symbJls, equation (2) can now be re'NTltten as follows: M1  p +
rectar.g~la~ ~r~~e
with
supports ~t ~oints A anj D. The bend~ng ~o~ents at ~oin:3 A, 5, C and D will Je dete~ined JY the cc~~~n a~alogy ~~t~ed. T~is ~r~~e and load~ns is identical to ex~~?le problem 1 of Art .."~.? or cater ;:.g'Nnere the sclution was ~aQe JY the elastic cen~er =e~hoG.
16"
10 lb.
45"#
18"1
I : 2 L:. 24
BI;.L_ _
:c I
10
~
~
..
L'30,,1
13
A
E;.
x
+ M{X __
 
(3)
j
7. 5
Comparing equations (1) and (3) we see they are similar. In other wordS, the indeter.ninate bending ~oments Mi in a trame are analogous to the column bearing pressures 0, hence the name nColumn ~~alogyn for the ~ethod using equaticn (1). ~ith this ~eneral explanation, the method can now be clearly explained by g~ving several example prablem solutions.
AlO.3 Fra.mes with One Axis of Symmetry.
Fig. AlD.3
2.5
Fig. AlO.4
SOLUTICN NO. i
~e first c8r.sider the ~r&~e centerline shape as shewn in Fig. AIO.3 as the crcsssection of a shart column (see rig. AlO.5). The width of each portion of the column section is equal to llEI of the ~ember crosssection. Since E is constar.t, it ~i1l be ~de w~i~y a~d the widths will then eq~al 1/1.
From the previOus discussion we can write, (1) The crosssection 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. The loading applied to the top of the analogous col~ is equal to the Ms/EI diagram, where X s 1s the statical moment in any basic dete~inate str~cture derived 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 column acts downward. The indeter.nlnate bendln~ ~oment M 1 at a given frame ~oint equals the base pressure at this same poi~t on the analogous col~14~. Thus the indete~inate moment at any paint on the frame equals (fromeq.3),
(2 )
i
I
Fig. AlD. 5
Fig. AlO.6
(3)
The first step in ~he ca:culations is to compute the area (A) of the column crosssection in Fig. AlO.S and the moreents of inertia of the column crosssection about x and y centroidal axes.
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 calcufations given in Art. A9.4 J andTable A9.l where this sa~e problem is solved by the elastic ce~ter ~ethod. These cal~ulations Nill not be re?eated here. The results Nere J Ix = 3188 3456. and 1y
Mx 270 x 9.375 2530 (positive because 8ase pressure is compressive on column portion above x axis).
My = (270x2) =540 (negative because baSe oressure is tension on column ~ortion to right" of yaxis).
Since the
fra~e
condition consiste~t with the given fra~e and loading. Fig. AIO.4 shows the condition assumed for this solution J namelYJ pinned at point A and a pin with rollers at point D. The static M s diagram is therefore as shown in Fig. AIO.4. ~e now load the column crosssection with the Ms/1 diagram as a load as shown in ?ig. AlO.S. The static mocent Sign is positive because the sGatic condition causes tension on the inside face of the frame. In the column analogy method a positive Ms/I loading is a downward or compressive load on top of cOlumn J and therefore a negative Ms/I value would be an upward or tension load on the column.
~x
=
12"J
=
20.625"
=OJ
Equation (3) requires the values of the 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 equals the area of the Ms/I diagram. For this problem the value of P from Fig. AIO.S equals J
P
Frame Point B.
l'l
i
= l2"J
9.325"
= 8.44+7.44+1.88 =17.77
l'lB = l'!sl'li = 0(17.77) =17.77 in.lb.
=22.5
x 24/2
=270
The centroid of this trianG~lar loading is 2 inches to left of yaxis. Fig. AlO.6 shows the column section with this resultant load P. now USe equation (3) to find the indeterminate moments Ml which are equal to the base pressures on the column. Bquation (3) involves bending moments M x and My and disGances x and YJ allot which must have Signs. The Signs will be determined as follows: When moment ~x produces compression on base on that ~ortion above x axis J then M x is
pos Lt.tve ,
~e
Frame ?olnt C.
= 12
= 9.325
(540)12 3456
= 8.44+7.441.88 = 14.0
Me
Frame Point D. Mi = 32
270
+
x = 12 J
3188
=20.625
3456
2530(20.625) + (540)12
A distance y is
~easured
posit~ve; ~easured
A distance x measured to right fram y axis is positive J to left negative. From Fig. AIO.6: p
= 270
Fig. A10.7 shows the final bending ~ament d1agram, which of course checks the solution by the elastic center ~ethod in Art. A9.4. The student should note that the numerical Nork in the column analogy ~ethod is practically the same as in the elasGic center method.
28.17
Fig. AIO.7
AID. 4
Solution No.2
In this solution a different static ~raoe condition will be assumec as shown in Fig. AIO.8, namely the frame is cut under the load and onehalf the 10 lb. load will be assumed as gOing to each cantilever part. Fig. AlO.9 shows the static ~oment diagram and Fig. AIO.lO the static no/I diagram with centroid locations or each ~ort1on or diagram which
= ~'s
=
12,
9.375
=12.23
Me
=M:sMi
c.
=30(12.33) =17.77in.lb.
30
90
Frame Point
x :: 12,
= 9.375
=76.0
30 Fig. AlD.9 .90
Me
4'
Fraille Point D.
= 12,
3186
Y = 20.525
(9160 )12 3456
Ml
(4)
::~ + 2531(20.525)
32
;' 99.82
MD
(1)
10
30
=Ms M1 = 90 
(99.82)
h2''1
Fig. AIO. 10 Fig. AIO. 11
Thus solution 2 checks sol~tlon 1. The student should solve this problem uSing other static conditions.
AIO.4 Unsymmetrical Frames or Rings.
are numbered (1), (2), (3) and (4). The area of each ot these portions will represent a load P1 , P a , etc. on the column in F1g. AlO.ll. Since the static moment 1s negative on each portion the load on the column section will be upward.
Pl. =10(30) =300, p. =15(6)/2 =45, P3 =45x9 =405 p. =30x30 =900
In applying 0he column analogy ~ethod to unsymmetrical frames and r1ngs, the moment of the H/EI dia~ram ~ust be taken ~bout principal axes and the ~aments of inertia ~ith resnect to principal axes. . However, as explained for the elastic center method in Chapter A9, the moments and section properties with regard to centroidal axes can be used if M x, My, Ix and I y are modified to take care of the dissymmetry of the structure. In Art. A9.2 it was shown that the redundant forces at the elastic center to ~~ s~etrical frame sections was, (see equations 11, 14, 15 of Art. A9.2).
M
= (45+405) = 2531
= 9180
My
t:ame)
111
IX

MxY + Mt{ y
POINT A.
>
12,
+
Ml :: 1650
32
2531(20.525)
3188
3456

(~)
= 36.06
AlD. 5
As
prev~ously
SOLUTION: Fig. AlO.13 shows the crosssection of the analogous column. The length of each ::nember of 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 column section properties. The total area (A) of column section equals
(15/1) + (10/1) + (12/2)
!'Ix
= M,;
(2f!)
= (1
~I
Ixl y
=31.
we
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 A :'::Ix
The calculations to determine the location 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.
y)Y
+ (Mv 
i<I y
i1X )x
  
(6
The true moment is the same as for the symmetrical section, namely,
The next step in the solution is to choose a static frame condition and determine the static (~) diagram. Fig. AlO~l4 shows the 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 portion will give the loads? on the c o Iumn,
240
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.
1440"
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.
2401b
at
, I JI 120'r:
9"
1
1=2
I' I
L'T
I
10"
1
1~40
P 1.
P II
p,
Fig. AlD. 13
11
I I
1= 1
'
l IA
'::T:7
Fig. AlD. 12
zp
AIO.6
These loads act on the centerl~ne of the frame members and t:rough the centro1~ of the geometrical moment d1agram Shapes. These centroid locations are indicated ~y the heavy dots 1n Fig. AI0.l4 and their locations are given With respect to the centroidal x and y axes. The loads Pl, Pa and 23 are now placed en the column in AIO.13, acting upward because they are negative.
we now find the ~crments M x and My wh1ch equal the moments of the loads P about the certroldal axes.
i1:x
Po1rot
C.
6.968,
Y = 5.242
M1 =923+58.37x5.242+132.36x6.S68
= 357.7
~
=Ms  Mi = 0 =
(357.7)
= 357.7
tn .lb.
FY"ame Point D.
x = 6.968, Y =4.758
?'In
= lis
0  (326)
= 326
1::.10.
The above results check the solution of this same problem by the elas~1c centeY" method in Art. AS.S of Chapter A9. The st~Gent should solve this problem by chOOSing other stat~c frame conditions.
AlO.6 Problems
~ = l1x
(*)
y
70290
x 217.74/606.51
= 25234
11' Y
(1) Deter.n1ne the bending moment dlag~am ~or the loaded structures of Figs. AIO.IS to AIO.20.
=11Y

(~) I 
= 139700 x 217.74/942.96
= (1
_ 217.74' ) 606.51 x 942.96
~ lSi 15 +151
L_45" I 3 L=30" I =2
400
400
w=lO Ib/in.
= 32258
k
= (1
ix(
x y
11
IL=30"
=2
H~'; I
= .9171
Fig. AlO.l5
II
L = 24 =6
L = 24 I =6
Ir
+:0 \Cl I I
29I
'5"~
1<> lCrJ
10
L; 24
1=4
whence,
!'1 i
i.i.
i 10
18
(7)
8+i"";
.....
:L==
II
20 5
1L==
I
20 5
Y =9.758
100
100 1/,;"
, I
400
4~",~~~
2
..~~I=
I
~4'~12'~'''' ,
= S.032,
y = 5.242
1.5
200
200
Fig. AID. 20
ME
= 11 8 ~
=1440 (1230.6)
209.4 l.n.l'tl.
(2) Solve ?rob:ems (2) and (3) at ~he end of C~~pter A9, Art. A9.9.
CHAPTER All
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 redundant 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 structure 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 moment diagram. T~e first arbitrarily assume that each span is completely restrained against rotation 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 cannot rotate. The bending moments which exist at the ends of eacri member under the assumed condition are then determined. Fig. All.2 shows the moment curves for this condition. (For calculation and formulas for fixed end moments see following 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 support 8 are not statically balanced, or speci:ically 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)
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 between the two beams in proportion to their ability 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 distributed 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 referred to as carryover moments. As shown by the obvious curvature at the elastic curves (Fig. All.4), the carryover moment is of opposite sign to the distributed moment at the rotating end. The ratio of the carryover moment to the distributed moment, referred to as the carrJover factor, depends on the physical properties of the beam and the degree of restraint of its far end. 1CarryOVer factors are discussed in a following article. For a beam at constant section and fixed at the far end, the carryover factor is 1/2). In figure AI1.4 a factor at 1/2 has been assumed which gives carryover 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 balancing moments and the ca~Jover~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.~
1.
Flxed~end ~oments:
By ~:ixed end moment 1s meant the moment which would ex1st at the ends of a member if these ends were fixed against rota~ion. stiffness Factor: The stiffness :actor of a ~ember is a value pnoport 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 2.
All.l
'."".':_'~':.""'~~:;;
'{~1~:
.111.2
THE
MOMENT
DISTRIBUTION METHOD
200
Fixed~
A
VI:
, .I.
10.,/in. . ! I"
r
0.5 when rar end is fixed. ~he letter ~ will be ~sed to designate carryover fcc~or.
15''01 L
1= Fixed
4.
B Fig. All. 1
I
,480
Distribution Factor: It a ~oment is applied at a joi~t Nr.ere :NO or more members are rigidly connected the distrlbut10n factor for each member 1s the prcport~onal part of the applied moment that ~s resisted by that member. The distribution factor for any ~ember which will be given the symbol D equals KlZK, where K equals stif~ness ~actor of a particular member and ZK equals the sum of K values for all the Joint members. The sum of the D values for any joint must equal unity.
5.
54
Sign Convention: Due to the fact that in ~ny problems where members come into a joint nom all dnect 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 moments are determined by the ~oment distribution Balancing Joint B i method using the above adopted sign convention, ! 1081162 that they should be transferred i~to the conven,./ =:::::::::... ! Carry Over Moments 811 from B to C & A tial signs before proceeding with the design of ~ ::..::::::..:c.:::...:....:c.::.._ _ the member proper.
.5881.588
I
a~pt
Fig. All. 4
Illustrations at Sign Conventions for End Xoments. Fixed end beam with Example 1 lateral loads.
i
.26~~~====~~~~~~~~~~~~~~~~/~1,
rA
I I
Adopted sign convention translation. The value ot the stittness :actor 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. carryover Factor: 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 carryover tactor is the ratio of the moment at the restrained end to that at the simply supported end. For a prismatic beam without axial load the carryover factor 15 3.
~
~ 'i
ExaI:1ll1e 2
=
All.3
1.L... _
M t~
L~
Pab
S'3M a
+
AIl..3
moment
Fig. AI1.12
Since the fixed end moments are statically facts must be obtained from the laws or continuity i~ order to solve for them. In this book the theorem of area moments will be used to illustrate the calculation of the fixed end moments as well as the other te~ which are used in the moment distribution method. (Ret. Chapter A7) The following well known principles or theorem of area moments will be used:(1) The deflection at any point WA" on the elastic curve of a beam away from a tangent to the elastic curve at another point WB" is equal to the moment a r the area of the M: diagram beEl tween the points A and 8 about point WAft. (2) The change in slape between two points ~A~ a~d ~Bft on the elastic curve at a beam 1s equal to the area of the M diagram between the El two paints ~A" and "9". The "area moment" theorems will be illustrated by the applications to the solution at a simple problem. Fig. AlI.5 shows a simply supported beam of constant ~oment of inertia and ~odulus of elasticity carrJlng a single concentrated load. Figs. All.6 and All.? show the static moment curve acd the shape of the beam elastic curve. Now assume that the ends are fix ad as shown in Fig. All.S and let the value at the fixed end moments be required. Fig. All.9 shows the shape of the final moment curves wade ~p of the static moment curve and the '~~~O~trl trapi zo i ca.; moment curve for:ned by the ur ...iknown end moments. Fig. AIl.IO shows the shape of the elastic curve, the slope at the two supports being made zero by fixity at these points.
indete~inateladditional
Since the change in slope of the elastic curve bet'.. . een ends (1) and (2) is zero, theorem (2) as applying to fixed end beams can be restated as follows. The sum of the areas of the moment diagram must equal zero. And fram theorem (1) the static moment of the areas of the moment diagram about any point must equal zero or in equation form: (1) ZM = 0          For a beam with variable moment of inertia the conditions for fixity are:jl1dx/El = 0 j!".xdx/EI = 0
==0 (2)
Figs. All.ll and AIl.12 show the static and continuity moment areas, the total area of each portion and its e.g. location. Substituting in Equation (1) M = Pab M1L MaL 0 (3)
7+7+2=
2'3+23+
~"hL
 (4)
The values of M 1 and t1a tor any value at a or b can now be found by solving equations (3) and (4) Table AlI.l gives a summary at bema fixed end moments for most at the loadings encountered in routine deSign and analYsis.
w
TABLE All. I
'1
1
I a t b :1
EI is constant
L _ _I
Fig. All. 5
r;~(
@~
~
Fa
iP I
 
b=1
..
,
;0 (5u ... 2v)
::::;.:;::::rITn , 1 I ; I L L'
L {
All.4
THE MOMENT
DISTRIBUTION
METHOD
_W
tangent to the elastic curve at poi~t A relative to a line AS is equal to the shear at A on a si:nply supported beam AB cue to t ae vr curve be::;1
t~een
<a
A and ':'hus
a acting as a load.
(ML
x 1(3) _
,~ 3EI
~
2EI
"= 2'E~I ~Jj
~
ML
x 2(3
wa
a a. 30L<I(lOL 15aL+6a)
p wa' / 20L
( 5L  4a )
Let 98 Then 1
unity ML or
3EI
Mg
= :lEI
tj~X(LXI' ~ t{WX'(LX)dX ~o
dx
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 carryover factor for a beam freely supported at its ~ar end is obviously zero. Consider the beam of Fig. All.14. uue to complete fixity at end A, the slope of the elastic curve at A is zero.
r::
stiff~ess
Factor of'
(Ref.N.A.C.A. T. N. #534)
A
or
MA
!1ll!:
2EI
1/3 + ~
2(3
=0
2EI
All.4
Carryover ?actar: 
Consider the beam of Figure All.13. By theorem (see Art. A7.l2), the slope ot a
EI is constant
L
2 Thus the carryover factor for a beam fixed at its tar end Is 1/2. Using the conventional ~om ent signs, the carry over moment is of the opposite sign as shown by the above equation. However, for our adopted sign convention inspection of the shape of the elastic curve as shown in Fig. All.14 tells us that the sign of the carrJover mament is of the same sign as the rotating moment at the ~ear end. That ls, the moment acti~g on each end of the ~ember 1s ir. the same direction, and therefore of the same sign.
= ~
~B
A Elastic Curve
""T
:;1" C'"""=+J....l
1.L 3 ML 21
Fig. All. 13
Fig. All. 14
IlJl = !1ll!:
2EI
:1A=
i1E
2
2EI 3EI
3EI 4EI
6EI
ThengB=~~=~
l2EI
Let 9B
= urn ty,
t.lten~:a
= 4EI
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 far end is 3/4 as great as one r ixec at its far end. j'urtnermore in one case the carryover factor is zero and in the other case it 15 1/2. It is t.nerer ore obvious that the values or these 1;NO ter:ns depends in part upon t::te restraint or degree of r ixat i cn of the far and of the beam.
All.S
; 2KF  2F which is the general exMA ; 3+F K(3 + F) MB pression for carryover rae tor for a degree of fixation F.
All.5a
Exa.ple ProblellLS
General Express1on~ tor Stiffness and Carryover Factor in Terms ot Fixation Factor (p) at Far End of a Beam
In the beam at Fig. AIL14 (F) fixation factor at A was unt ty since beam had been taken as conpletely fixed at A. It was found that:ME
;
To obtain a definite conception at the true mechanics of the "cross" ;nethoo, the reader is advised to follow thru the detailed solution or the follOwing simple problems. In these problems, the moment of inertia in any span has been taken as constant and all joints have been assurned to undergo no translation. Problems involving variable I and joint translation will be considered later.
.~./!Il.
4E19B
L
;
Te.ke 9B
and MA
 2 EI 9B
~A
~
s K
L
jStUfIl." hclDl' IOllltriblu1OCl Facl<ll' KI.l:K Carry Oo'er Facwr
1'lIed: End Mo",ent.
E2
~1.5 ~
~,
~,
,W. H/in.
11 11
gil
]} tt tt tl
all
W!!'J
uc nc
.00
51.51
':'hen MB; 4K MA; 2K L~kewise the results for the beam ot F~g. A1L13 give Me = 3K I1A; 0 Figs. All.lS and AII.IS show these results. Fig. All. 17 shows the general case, the r ixation factor at A being F. The difference between F'fga , AII.IS and AII.I? is that the slope at end A has changed but eB the slope at end B remains the same.
F;i~B
tlA=O
iu
.51.5
.,
.515
1157<E 0 o
c
~llS!O
57.5 _51.5
c
710.51710. 5
FI.llaJ. Moll:lefttll
ConvellUO.... 1 MOlQIIDt SIilU'
.S8~
..
8113
." '"
"".
MA=~2K
MB=4K
Fig. All. 15
F=O
Fig. All.
is
The change in :noment at end A when changing beam AIl.lS to that of All. 17 = 2K 2KF ; 2K(1F). Since 9B is xept the same value, onenarr of the moment change at end A appears at end 8 but of cppoat te sign, or Mg=4Kl/2 2K(lF) ;3KKF=K (3 + F) ;
ZI (3 +F)
F"F~
MA'= 2KF
eA' F(. 5)
MBK(3+ F)
Fig. All. 17
Thus the ge!leral expression tor the stiffness factor or a beam of constant section equals EI (3 + F') ':'he carryover factor from 9 to A
;
C
Example Problem 1 shows a twospan cont1nuous be~ with overhanging ends. '..le first assume all the joints B. C and D to be locked against rotation or the beams Be and CD are arbitrarily assumed in a fixedend condt t t on , The !irst line in the solution gives the stiffness factor of each beam. From Art. AIl.S the general expression for stiffness factor is EI(3 + F)/L. where F ; fixation factor of far end, wni en equals , since all joints have been con" steered rtxec. ThUS, stiffness factor K ; EI (3+ l)/L; 4. EI/L. Since EI/L is the same for each span Be and CD, the Stiffness factor has been wri tten as one since it 15 only the relativa values that are necessary. The cantilever SA has zero s t t r rneas , The second row gives the member distri but t on factor D at each jOint or 0 ; KlZK. ?or eXalIlple at jc ~::.t 3. the distribution factor to 3A = 0/(1 lI(c 0) ; ; . At joint r !:o. e 0) = o and to BC factor to CB ; 1/(1 + 1) ; .5 and distribution c Lfkewt se to v The thll'd row giyes the carryover factors. From Art. All.5, the carryover factor C ; 2;;'/(3 + .5 assummg the far end rrxe d F'I = 2 x 1/(3 + 1) or F = 1 for each member. The next step as shown by the 4th horizont'11 row is the calculation of the r ixed end nonants. The signs or tl:e :n.oments are according to our adopted sign convention, t~t is an internal res 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 members EA and Be. The resistance of these members to a rotat10n ot Joint B 15 proportional to their stiffness. 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 statically 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 balanced by distributing 115 between DE and DC as explained above for joint 8. As painted out in Art. All.5, when we rotate 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 carryover 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 carryover, or in other worca until all artiticial restraints have been removed and the structure rests in Its true state of equt Li brtun, To continue with the second cycle, go back to joint B and release it again tram its assumed locked state. There is no unbalance since the carry over moment tram point C was zero, thus there is nothing to distribute or carryover. Proceed to point C, releasing the joint, 'He flnd 1t balanced Wlder the carryover moments at 57.5 and 57.5. Thus the distributing balancing moments are zero. Joint D Is l1kewise in balance since the carryover moment tram C 15 zero. All joints can now be released Without any rotation since all joints are in equ1ll brium. To obtain the tinal moments we add the original raxec end moments plus all distributed balancing and carry over momen ts .
All~S
Calculate the carryover 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 given transverse member loadi:Igs or support oer rections, 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 exists balance It statically by distr~butlr:.g a counter acting moment ot oppOSite sign among 'che connecting members according to their D or distribution factors. 7. These distributed balancing moments produce, carryover moments at farend of members equal to the distributed moment t taes the carryover raetor C and of the same Sign as the dlst:oibuted moment. Record these carryover moments far ends for all distributed moments. 8. Repeat the proceedur'e of unlocking each joint, distributing, and carrying over moments until the desired precf 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 algebraic sum at the original r fxed end Jl.Oments and all distributed and carryover moments.
4.
,~
_0
~ample
.SJ/lIl..
*
t,
Problem #2
il:~
r:ac'lar
L
.
12
I'
:
geJ..s l . ,~
I
F!
q
,
.~
Q1Q~1
D~~:'IOI>~
CarrrOo.r C factor
fl.om E%ld
IOI_C.....
,I.
e .s
.~:l8!. S72
sl.s
"
I
110
I I I
,
.s.s sl.s
I
.sn!.Hs
~"
_a,
...
'I'
i
I
CarrrQru
:lod8ll\alw:lIl.t:
n Il5 72
::::><:
0 01 961_9&
X~
c _S7.S
u _72
X2~_5!'32_9
_3slo
6.110
~X2.9!24.6~
X;6.5116.5 0136
~><~
_121: 7210 12.31 _:71
car,.,Oo...
(2.3
3"" 8IIl:uI<:.....
car,.,.OvJr
~tlI
0,IZ.3~lS.ti20.6~~X:.1.1S.?::<11.JIO
7.1 10.l!.10.3 *6,1
8IIlaAc .....
01_1.7~_1.6 ::':'X~X.SZ.6
Il.] olt.3
.~318U
e.... , ..,oou
Sib SOWle ....
.~ o
I.
~r,:io
I.J!
_l.J10 _8331883
_1.611.8
~
013.9
_1.1 _I. 1 _610. tlalO. 4
7 1.
ll.oal..,..._ ..
,"
_3+413+4
erc 41610,4
Ie_Ilona! SilPUl
Ge 1:1.8 ra 1 SWIIII&ry of Proceedure
1.
.1.
.1.
.1
'1'
Example Problem 2 15 similar to prOblem 1 but two All computations should be written on, or spans have been added. ',oj'e first assume all joints adjacent to the d1agram. of the structure. locked against rotation. The stiffness factor of factor K tor each 2. Determine the stittness each span 15 :?roportional to EIIL or lIL since EI end at each member. K = (3 + F) EI/L, where F constant. The car:yover factor 15 1/2 as in is 15 the degree of fiXity at far end ot member. previous example. Fixed end moments are calculatIt all members are assumed rrxec at far end than ed as shown. Unlock jOint E, the unbalanced mo:nK 15 proportional to IlL assuming E as constant. ent 15 U5. Balance the JOint by distributing 3. Determine the distribution factor D tor each to BC and zero to EA. Proceed to joint C, 1 x 115 member at each joint of structure. D = K/l](.
1.
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 unbalanced 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 carryover 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 carryover moment of 72. Balance the joint by distributing  72 x 1 =  72 toBC 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 carryover 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 distributed and carry over ~oments. One requirement 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 = ~ t.l1cb  .5 Hlbc Kcb SUbt. values <ll'Igd 
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 remembered 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 loading it is only necessary to solve one half of the structure. Due to symmetry jOint D does not rotate 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 carry 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 factor 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 carryover factor C to B is zero since B is left free to rotate. (See Art. All.S) Example Problem #3. Simplified Solution Of Problem /12
l'IUI.
(610.4432) .5 (344 (432)J _178.644 [610.4 (768)J  .5 (883 768) 107.657.3
.~.
1.343
..., r,:H.
FlU<!,
Ratio of stiffness factors = ~ = ~ .0104 = 134 . Kcb Thus the distribution is according to the K ratios of the adjacent members. Simplifying Modifications  Examule Problem #3 The solution as given in Problem #2 represents the "Cross" method in its flli~damental and most elementary detailed form. Many modifications of the general method have been presented, in the most part for the purpose of eliminating part of the arithmetic or the n~ber of cycles
.
_
~T.S
12~
o . QHl4=ft
Oi",,.,butIOll F1cwr o , !t/!x:
lin 2
O\:N IC<:
"
,
uu c s
.;183 163
'J
ll~_
. le .IM
cu
4U
c .s
.1e8 432
Car..,_onr to C
Final mOm.nUl Con.enuon:U &lll'l_ .a831883
..
I
~
c, ..
All.8
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 Pr cb Lem ""4 sistance to rotation. 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 a and 0.5 x 176.4 = 89.2 to O. Proceed to jOint Oistrtllutioa F:Lctor Ort o and release it from its assumec locked state. CarryOver Factor The unbalanced moment is 432 + 89.2 =  342.8, n"ttd End Moments 1st we balance by distributing 342.8 be~~een DC Carry_Over which has a stif!ness of 0.139 and the support :!nll E which has an infinite stiffness or zero goes Carry_Over 3rd. Ba~ClJ'tg to DC and 342.8 goes to rigid support. ~he carCuryOver ry over moment to C from D is zero since 0.5 x <lUI BalanCing o = O. The t~na1 moments thus equals the sumC.. rryaver 5th Balancinlj: ~tions as shown which of course are equal to thE FlAal l.lomenUi results shown in Problem #2. Cmlventional Signs
a..~CUlg BIil~clng
0.>
.a831768 01115
112
0172 '35.3 Oi35.3 11209 OIU.9 ,.6.3
_<1321
"
2l6i21~
17.,111.8
12.61_16.9 _6. <11_<1. J
~.
'"
016.3 _8831883
616.
.i.
_64<4.~I~H. 3
Example Problem #4Problem 4 is similar to Problem #2 and #3, except that the support at 0 is assumed as having 50 percent fixity. Thus 50 percent of any moment at this point produces rotation of the member DC at D. In continuous wing beams, which fasten together 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 continuity be taken as 50 percent. Solution 1 of PrOblem 4 1s a detailed solution. The only change that has been made is in the stiffness factor of the support E, Nhich has been taken as equal to the beam DC, thus ~~ unbalanced moment at this poi~t is equally divided between the beam and the support. Solution 2 is a ~odi!ied solution which elimlnates considerable arithmetic. Thus it is 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 carryover 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 unbalanced moment of 115 is balanced statically by distributing 115 to BC. The carrJover 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 carryover moment of
.5 X 216 = 108 is carried over to C. Joint D is left ~~locked or in its true state of ~estraint. Joint C is now unlocked. The unbalanced ~o~ent solution #2
SxampLe
SUffn ....s
hotor K
Olst,..bIltlon Fa",o
i1c
liZ
Carry_over Fac,o.
Fixed End Mom..."" 9alanC. Jain' II
a83j1Sa
ill t 15
Carry.o..... to C
BalanCe JOIRt 0
Cury.o..... to C
Bal:lnce JOin, C
C.. rry_ove.:o 0
Elal:lnc. 0 FIn>,j Mom",,t.o
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 carryover 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 support 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 illustrating 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
VEHICLE STRUCTURES
All 9
Joint D. Since the mOment of inertia is conSince 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 soas listed in the solution table Nill be explainlution shown since beam ~s freely supported at ed in detail. 3 t.nt s joint is lett tree to rotate after reFixed End Moment For Lateral Beam Loading leasing and thus the stiffness factor of beam 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 fixedend ~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, Ml_2=L.a (5u+2V) (See Fig. (a)) it Is evident that deflecting one support rela60 tive to an adjacent support will produce moments at the ends which are assumed fixed against M2_1==L.a (5u+3v) (See Fig. (a ) I rotation. 65
All.a
..L_
v
 
Ef ,~
M ~
ll.6 ~~
I'l
ML 4EI
i~ ~ ~~
1
1 .
6 Fig. All. 18
J L
Fig. All.IS shows a fixed end beam. The support B Ls deflected a distance d relative to the support paint C. It the me:nber ls or constant
at the beam mid~oint and ME  Me. By moment area prln
T
u
=
Fig. (a]
L
L
f
CD
For Span Be:
!1bc !'leb
"
~~O~~~~:~~~~~ ~~~lP~~~i
5 L
(3
40'
60
40'
(5x3+1) = 426
(5x3+1.5)
~n.
lb.
L J1L
(3
60
40'
4El
I'lcct
60 (5 x 3.5 + 1 I
40'
V
= 507
in. lb.
mO:nent due to a transverse support settlement of d. Examole Problem #5. flected supports.
General data; 3 f/in.
?lxed End Moments Due to Sunnort Movement From Art. All.S For span Be: Deflection of B relative to C
!'!be = Me'::l
l1
= ~Eld/L.a = 5/16
inch. Hence,
It
o ": ,
jlOj 40
I!
! ; ;
rI
=6 x 10,000,000 x
ABC
t
3/ 16 "
40   40 40 o Defle<:,tion of Supp?rt!l
DI

=t""'dc :: 6 x
1::1.
;:: 234
Solution:
A iB
3/4x1~. ~ (t,i
&)
10.
ZK
011 1/'
50 426
n
112
o 390
390 234
,,.
507
161
50 _50
For si;~s of tte ~oments due to tr.ese deflections see Art All.2. Havir~ dete~ined the fixed end ~ooer.ts t~e ~eneral distributing and carrYing over process follows as indicated in the solution table. Thus at joi~t B, the unbalanced moment :: (50426 + 390) :: 14. Balance by df s trt out t ng  14 xl::  14 to Be and Zero to EA. carryover.5 x 14 ::  7 to C. Considering joint C, the unbalanced moment == (440 + 390 + 234  494  7) :: 563. Balance bYdistri but tng  563 x .571::  322 to GD and 563 x .429 ::  241 to CB. carryover.5 x 322 ::  161 to D. A~ joint D the unbalanced ~oment = (507 + 234  161) = 580. This is balanced by distributing zero to DC and 580 to the fixed support.
All.I0
THE
Check on F1nal Moments
111.8
To satisfy statics, the algebraic sum at the moments in all the oe~bers at a joint must equal zero. This requirement alone will not prove that the final moments are correct, as errors could have been made in the various distributing and carryover moments. continuity tells us that the final slope of the elastic curve at anY joint is common to all ~ambers meeting at that joint, thus for a complete check it is necessary to prove that the final moments are consistent with equal slopes for the various members. The rotation of the joints from their original assumed fixed or locked condition is due entirely to the distributed and carryover moments. The actual rotation will therefore equal the rotation of the end of a simply supported beam when subjected to end moments equal to those produced by the algebraic sum of the distributed and carryover moments. These end moments equal the final moments minus the original fixed end moments and will be referred to as AM moments. From Chapter A7, the slope at any point on a t h simply supported beam equals the shear d ue a ~ e M/EI diagrams as a loading.
All.9
End Moments tor Continuous Frame.orks 'bose )(e=bers are not in a Straight L1ne. Joint Rotat1on Only
Continuous truss frame works are quite common in aircraft construction. Welded steel tubular fuselages are composed of ~embers ~hlch maintain continuity thru the jOints due to the welding. Landi~ gears frequently :~ve ~NO members which are made continuous at their points of connection. The members of such structures usually carry high axial stresses which cause joint translation which in tur~ prOduces bending of the members since the joints are rigid. ~hese moments produce lateral deflection of the members which introduces additional secondary moments due to member axial loads times the lateral member deflections. These influences are treated in later articles. In this article to further familiarize the student with the moment distribution 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 iL/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 dec.Mab ,. termine the bending moments in the two ~embers Ell 'C.g. due to the vertical load on the axle. The probB lem has been solved using ~hree different degrees ot restrain~ at ends A and B. Joint 0 is 11 L 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 rotation under prumary bending moments. The effect ~c. g. of axial deformation and secondary moments due i to member deflections is omitted in these solu_!IL/3 tions. These factors are treated in later articles . .o.MbaL 2Er" In a practical problem the degree of restraint at points A and B would be dete~ined by 2/3 _ <IIlbaL ~ 2E1 3 the type at tittlng used and also on the r1gidity of the adjacent fuselage or Wing structure. As (&lab + .5 <IIlba) illustrated in later ex~p~es, the moment distr13EKab bution method permits the consideration at the where rigidity of the adjacent structure without adding I any diffiCUlty, While such methods as least work
AI
Kab
= L or
AB
Since the angle QA must be the same for members meetir~ at A, the general relation bet;.... 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
1S"
32"
=~
Xac
T .
N
To make this equation consistent with the assumed Sign convention, that is, the carryover moment has the same Sign as the balancing distributed moment, the above equation must be modified as follows:
Fig.
Ail. 19
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 stiffness 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 expression 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
K :II Stiffness Factor Values in ( ) are carryover factors
Fig. All. 20
For member OB
K =
are not practicable because of the large number of equations that must be solved to obtain values tor the many unknowns.
g L
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 counterclockwise 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 unbalanced 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
3+0
=a
Considering joint a in Fig. All.20 the external moment ot 18000 in.lb. is balanced by distributing  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
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 entirely by the rigid support, or no rotation takes place when joint is released from its imaginary 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 unlocking these joints which are definitely known to be free
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 supported C is 25 percent fixed and joint 0 is considered to ~intain continuity be~Neen all members 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 fixended. The relative stiffness factor K of each member is therefore proportional to IlL of each member. These K values are listed in ~ig. All.21. The distribution
All. 12
THE
MOMENT
DISTRIBUTION
METHOD
c. O. Factor Member  K. Value factor D for each ~ember at each ~oint which 2F equal K/~ is recorded in c=lon each member' ~(hF) 3 F around each jOint. Thus any balancing ~oment is distributed bet~een the joint members as per 2x1 =1 OA:l..(3+1J "'.4, 2 these d1stribution factors. The carryover ~ 20 factors for all members is 1/2. The fixed end 2xO OB.i~{3+0) :.3, 3 +0 0 moments due to external loading are computed tor the three members. For member AO, the tix1 2x.25:. 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 ZK=.83 loading, the fixed end moments are zero. In this solution the or1er of joint consideration 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 support, no rotation takes place and the balancing moment of 300 is provided entirely by the support 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 distributing  300 between the three members accordSOL tJrION .2 ing to their D values, thus  300 x .416 = Fig. All. 22  125 to OA;  300 x .168 =  50 to OC and  125 to CB. To prevent contusion it is recomby distri but.fng .75 X 25 =18. '75 to CO and the re. mended that a line be drawn under all distrimafnder of 6.25 to the support, s tnce the fix!.. ty r; buted balancing moments, thus any values above the support at C has been assumed as 25 percent _ A these lines need not be given tur~her considline is cra~m ~lde~ t~e 18.70 and the ca~ry over eration and only values below the lines need moment of 9.37 is taken over to 0. One cycle has be considered in later balanCing ot the jOi~tS. now been completed. Retur~ing to ~olnt A, ~e Immediately a:ter dlstribu~ing the moments at find 62.5 below the line. This is balanced by joint 0 the proper carry over moments should be distributing zero to OA and 62.5 to the t!..xed taken over to the far end ot each member, thus support. A line 1s drawn ~~der :he zero distri 62.5 to A,  62.5 to Band  25 to C. Joint buted ~oment to AO and the carryover woment of 8 is next considered. The unbalanced ~oment is zero is placed at O. Considering joint 0 f~r 62.5 and it is balanced by distributing 62.5 the second time the unbalanced mo~ent is 9.37 + to 80 slnce the pin support has zero sti:tness, 31.25 + 0 = 40.62 or the S~~ ot all vaiues below or no resistance to rotation. A line is erawn the col~ horizontal lines. The joi~t is balunder the 62.5 and the carryover moment ot anced by 1istributing  17 to OA ~~d 08 ~~d 31.25 is placed at O. Joint C is considered 6.62 to OC. Lines are dr~Nn under t~ese balnext. The unbal~lced ~oment o~ 25 is balanced ancing momen~s as shown in Fig. All.21 and t~e carryover mcments are taken over to the far SOL111'ION'IIl 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 distributl~g values have become quite ~ll and it SWmellB Factor K [ a is evident tha~ a ~igh degrs8 of accuracy has OA=I: 10 : .10 been obtained. The ~inal end ~oments at each 5;,lO 08= lis joint equal the algebraic sum of the values in . each COIWL~. A double line ~s placed above the DC = 225 =~ r inaa noaerrts as a df s t tnguf sht ng symbol . In ZKJointO=,24 C.O,F",ctor_1/2 the fi~~e the letters band c ~efer to balancing far All Memo. 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
Fig. All. 21
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 factors, thus making it ~ecessary to balance ~oint o only once and car~y over this final far end
All.13
of each member. The fib~e gives the calculation of the modified or actual stiff~ess and carryover factors. ~ith these !u~c,vn the solution is started as before by computing the fixed end ooments due to transverse loading on ~e~ber AO. Joints Band C are released and si~ce no fixed end moments exist, no balancing 1s required and the joints are left in their true state 0: restraint instead of locking and unloCki~ as in solution #1. Releasing joint 0 tram its imaginary fixed state the ~~balanced moment is 300 which is balanced by distributing  300 betNeen the 3 connecting members according 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 t.o B =  108.2 x 0 = 0 and 47 x .154 =  7.2 to
~oments
assumed to give 50% fiXity to these joints. In Table A11.2 a modified stiffness factor is calculated tor ~embers GI, Fl, and FH using a 50 percent fiXity at their tar ends. The last column of Table All.2 gives the s~tion of the ~ember stiffneSS factors for members intersecting at each jOint.
120* 3200"' ,.12'",
J.
l
~:/7\
Engine A
MOWlt 2400\
/\ /
/
12400".
~~~:'~d
H
'\
Problem #8 figure All.23 shows the forward portion of a fuselage side truss. Due to eccentricity of engine 2o~,t and la~ding gear members, external ~o~e~ts ~e produced on joints A, Band D as shown. Fur~he~ore lateral loads due to eGuip~er.~ installation are shown acting on ~embers BE ~nd CD. Assuming the fuselage welded joints pr ocuce rigi::i continuity of members thru ttie joint, t~e problem 1s to find the end moments in all the ~a~bers due to the_eccentric joint ~om en~s ana two lateral loads. The effect of joint tr?nslation and secondary ooments due to deflect10TIS and axial loads is to be neglected in this ex=..m.ple. So1.:J.tion: ~able All.2 gives the calculation or the stiffness factor for each truss member. The ~uselage truss aft of joints I and H tave been
Exa~ple
_1.98 ( .1.82 b
"'\ , f
,
Leuct h
L
2~"
Fig. All. 23
Landing Gear
TABLE AiL 2
1. x
1000
.
11
it1ffA
J:. J'aetar
F'',...,,j
A IS C
, 1.0240 ,0< , .0240  ,0<' 34.' .0588 AC  .049 19.25 .0588 2 .049 30.0 .0588   ., ,.. _3 .0449  34.  3 .033  ". 1_1 _1 2  .049139.5 Be 0178 o 8 1_ 8 _
cr
.049 41.
.,,240
0.582 61
1 _1 2
.S18*
3.06 1 6
S 80 4 81 6 95
D
7
D'
..
:I
,sa
4.42 3.21
6.14
1. 96
2 D.
(lSBB
ell
(I:l~141
o s 0 ,
..
(II: 1. eoutaJIt
fol' Lll _ 
>en)
~OS
~ beca~
of
f1xity)
Fig. All. 24
.,84
"
. ,
(n5
141 b .178
_187
18 b
1)i
0'.
"
ct
e , ae
b
b
.286
'' .195
0'.l15
"' ,
....
"'."' ......,..,....
:::~
"
sr s
.
__ .834
:...JJMb
na
o...ill
~b
70 ',0 b ~
140 ~
aa
. ':""""55'
"
10 . ~b
.,29 b
_ _ b 140
1)15
m"
..
n
>;.286
brot
,
~"
.\11.14
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 circles 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"#
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 appears 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 bottom members which are web members in the top ~~d bottom fUselage trusses are subjected to the l1E:B =120 x 29 .25' x 12/41.25' =725"# equipment installation transverse loads as shown. l1cD = 100 x 20' x 10/30' =  445 Let it be required to deter.nine the end bending moments in the rectangular trams due to these l10c = 100 x 10' x 20/30' = 222 transverse loads assuming full cont:nuity thru These moments are placed at the ends of the joints. members on Fig. All.23 together with the eccenSolution: tric joint moments. The process of ur~ocklng the joints, distributing and carrying over momFig. All.26 shows the solution. The dlstr1ents can now be started. In the solution as bution factors based on the member sti!fness facgiven the order of joint consideration is tors are shown inD at ends of each member. The ABCDEFG and repeat, and each joint has been bal first step 1s to compute the fixed end moments anced three t1.mes. due to transverse loads, on members AS and CD usConsider joint A:ing equations from 7able All.l. The ~gnltudes Unbalanced moment = 2400. Balance by disare 1890"# for AS and 2025~# for CD. tributing  2400 as tollows:Joint B is now released f~om its assumed To AC =  2400 x .527 =1258. Carry over fixed state. The unbalanced moment of 1890 is to C =634 balanced by distributing  1890 x .247 =  467 ~o To AB 2400 x .473 = 1132. carry over SA and the remainder of 1423 to BD. T~e carry to B =566 over moment to A =  46Sx .5 =  233. Due to symmetry ot struc~~re and loading only one t~lt Proceed to Joint B:of trame need be considered and hence these carUnbalanced moment = (566 + 3200  298) = ry over moments to A are not recorded. However, 2336. Joint 1s balanced by distributing  2336 in balanCing joint A 1t will throw over to 8 the to connecting members as tollows: same aagrrrtuee of carry over moments as thrown To BA = 2336 x .569 = 1330. carry over to over to A from B but of opposite sign since the A =  665 original fixed end moment at B is minus. Thus To Be =  2336 x .310 =  724. Carry over to 233 comes to 8 tram first balance of A as shown C  362 in the figure. The distributing moment to a of To BE = 2336 x .121 =  282. Carry over to 1423 produces a carry Qver moment of 1423 x .5= E =  141 712 at D. ~27" I 300# 300# The convenient device of drawing a line 7" 10tr10"j under all balancing moments is used to prevent contus1on in later balances of the joint. 1".035 Tube Proceed to Joint c:K= 0~~37::. 000458 Unbalanced moment = (634  362  44S) =1441 The joint 1s balanced by distributing 1441 as tollows:0:1 0 0:1 To CA:= 1441 x .44 = 635. Carry over to A = 318 36" ~ 0 ~ ::: . ,. 0 CB = 1441 x .214 = 309. carry over to B = 155 CE = 1441 x .064 = 92. Carryover to E = 46 '~I~ 0 <;gOM :<;g 11 CD =1441 x .282 =406. carryover to D = 203 This process is continued tor the ~emainder or 600# the tr