BHRBS ZAK Formulas for Stress and Strain sr emo RAYMOND J. ROARK WARREN C. YOUNGContents ONE TAREE Lis of Tables ix Preface tothe Fh ition Preface tothe iat Editon so PART ONE Definitions Definitions PART TWO Facts; Principles; Methods ‘The Behavior of Bodies under Stress Methods of Loading. Elaxiciy; Proportions of Stet and Semin Factors Afeting Elastic Properties, Lead-deformation Relation foe 3 Iiody. Plasticity, Creep and Ruptare under Longtime Loading Cfitera of Elastic Fale and of Rupture. Fatigue, Brite Fracture Stress Concentration, Eft of Frm and Seale en Strength; Rupeare Factor. Prnreing. Elaric Stability. Refrenon Principles and Analytical Methods Equations of Motion and of Equilitiam. Principle of Superposition Principle of Reciprocal Deletions. Meth of Conssent ‘Deformations (Stale Compatibility). Principles and Methods Involving Strain Energy. Dimensional Anaya Remarks an the Use| of Formula, Refereots v7 2Contents FOUR FIVE six SEVEN nicHT Experimental Methods .... Bppeee00000 EL) Measurement of Srtin. Photoclute Analysis Detection of Pati Yielding. Anslogee. Models, References Phopertes of a Plane Area or PART THREE Formulas and Examples Tension, Compression, Shear, and Combined Stress .. 73 ‘Bar under Axial Tension (or Commpresion); Common Case, Bar under “Tension (or Compreson); Special Gases. Composite Member: ‘Tromen Body under Pare Shear Stes, Cases of Direct Sheat Leading, Combined'Suess Referees. Beams; Blexure-of Straight Bars a) Suright Reams (Common Cate) Elatically Stet, Composite Beane and Bimstalie Stripe ‘Threemament Equation. Rigid frames, Beans on Elastic Foundations. Defcemation Due tothe lant of Fined Supports, Dears under Simultaneous Axi and ‘Transvre Loading, Best of Variable Section. Slrted Beams. ‘Beams of Relatively Great Depth- Beains of Relatively Great Wich, ‘Beams with Wide Flanges; Shear Lag. Beams with Very Thin Webs Beams Not Loaded in Plane of Symmetry; Fleearal Canter. Stesght ‘Uniform Beam (Common Case); Uimate Strength, Pass, or inte Stengeh, Design. References Curved Beams... Pee - 209 Bending in the'Plane ofthe Curve, Deflection of Gurved Beams of ange Reds, Cieular Ringe sod Arches Elliptical Rings Curved ‘Beans Leaded Normal to Pane of Curature. References, Torsion... Bass oeseecbooeoD50G0e.- Straight Har of Uniform Circular Section wader Pure Torsion. Bars of [Noneiculte Uniform Section under Pure Tosion, Bfet of End Constraint, Bfect of Lngiodinal Stresses, Ultimate Stength of Bare in Tonsion, Torsion of Curved Bars; Helial Spring. Miscellaneous Formulae for Cease Shas. References. Common Cate. Bending of Unifomshikness Plates with Circular Boundaries Cizculapate Deflection Due to Shear. Bimelallic = 366 Grea Pater Nonuniform Loading of Cirular Plates. Circular ‘Plates on Elsie Foundations Crcaar Pats of Variable Thickness Disk Springs. Nartow Ring under Ditibuted Tergue about Js Axi ending of Uniformsthicknes Pats with Suaight Boundaries. Efe of Large Deflection; Disptragm Stents. Plate Analysis of Plats, ‘Unimate Strength ReferensELEVEN ‘THIRTEEN FOURTEEN FIFTEEN Contents ois Columns and Other Compression Members. 44 Columns; Common Cave. Local Bucking. Strength of Lata Columns, Berens Leading nial Curvature Columns under Combined Compression ané Bending. Thin Plates with Stifener, Shore Prisms under Eecentrc Leading, References, Shells of Revolution; Pressure Vessels; Pipes 5 (Grcomstances and General Sate of Strats. Thin Shells of Revluion under Dipuibuted Loading: Producing Membrane Sass Only. Thin Shells of Revolution under Concentrated or Discontinuous Lasings Prodecing Bending and Membrane Stresses Tain Mtilemem Shell of Revolution, Thin Shells of Revelation under External Pree ‘Thick Shells of Revoltion. Pigeon Support at Intervals. Referers Bodies under Direct Bearing and Shear Stress 513 Suest Due to Preaure berwten Elstic Bodies Rivets and Riveted Joins. Miscellaneous Cases References Elastic Stabitity cee . fees ST General Considerations. Buckling of Bart: Buckling of Fat and Curved Plates. Buckling of Shel, References Dynamic and Temperature Stresses . - 564 Dynamic Loading; General Conditions. Body in a Known State of ‘Motion. Impact and Sudden Loading, Impact and Sedden Loaing ‘Approximate Formula. Remaees on Sires Due to lmpats ‘Temperature Sueser Reference Miucelanaous Tees 589 actor of Sew Coeenaion, Prope f Mate Name Indes 617 Subject Inder 617List of Tables 8 8 Pow sawn ene 8 136, Be . Shear, Moment, § Properties of Sections Formulas for Combined Stress lope, and Deflection Formulas for Beams Reaction and Deflection Formulas for Rigid Frames Ba Numerical Values for Functions Used in Table 7 Numerical Values for Denominators Used in Table 7 Shear, Moment, Slope, and Defection Formulas for Finitelength Beams on Elastic Foundations Shear, Moment, Slope, and Defction Formulas for Semiinfnite Beams on Elastic Foundations Reaction and Dejlection-Coefficiens for Beams wnder Simultancous Axial and Transverse Loading: Cantilever End Support Reaction and Deflection Coeficiente for Beams wnder Simultancous Axial and Transverse Loading: Simply Supported Ende oc Reaction and Deflection Coefficients for Beams ender Simultancous Axial and Transverse Loading: Left End Simply Spporied, Right End Fixed Reaction and Deflection Coefficients for Beam ier Simultancous Asal and Transverse Loading: Fixed Ende Shear, Moment, Slope, and Deflection Formulas for Baume under Simulta- neous Axial Compression and Traneverse Loading ‘Shear, Moment, Slops, and Deflection Formulas for Bears under Simulta: neous Axial Tension and Transverse Loading Beams Restrained against Horizontal Displacement ot the Ends Reaction and Deflection Coefficients for Tapered Beams; Moments of Inertia Vary as (I+ Ke/t)', Where m= 1.0 ‘Reaction and Deflection Cocffiients for Tapered Beams; Moments of Inertia. Vary as (I+ Ke/t)', Where m= 2.0, Reaction and Deflection Cocficients for Tapered Baars; Moments af Inertia Vary as (I+ Ke/1)", Where m= 30 ue 132 135 1m 142 48 149 130 is st 167 m 176 178List of Tabs 184. Reaction and Deflection Cocfcients fr Tapered Beams; Moments of Inertia. Vary as (1+ Ks/I)%, Where n= 40 414, Position of Flexural Centr Q for Different Sections 15. Form Factors for Beams 16, Formulas for Curved Beams Subjected to Bening in the Plane of the Curve 17, Formulas for Civeuler Ringe : 418, Reaction and Deformation Formulas for Circular Arches 19, Formulas for Curced Beams Loaded Normal tothe Plane of Curvature 20. Formulas for Torsional Deformation and Stress 21, Formulas for Torsional Properties and Strestes in Thin-salled Open Grose Sections 22. Formulas for the Angle of Taist of Uniform Thin-walled Blastic Members under Torsional Loading 28. Numerical Values for Functions Used in Table 24 26, Formulas for Flat Circular Plates of Constant Thickness 25. Shear Deletions for Flat Giveuler Plater of Constant Thickness 26. Formulas for Flat Plates with Straight Bowndaviee and Constant Thickness 27, Standard Column Formas 28. Formulas for Short Prisms Loaded Becentrically; Stress Reversal Impossible 2, Formulas for Membrane Stresses and Deformations in Thin-walled Pressure Vesrle zg BD, Shear, Moment, Slope, and Deflection Formulas for Long and Short Thin: walled Cylindrical Sells under Avisymonatric Loading 31. Formulas for Bending ond Membrane Stressee and Deformations in Thin ‘walled Pressure Vessels - 32. Formulas for Thick-waled Vessels under Internal and External Loading 88. Formulas fr Stress and Strain Due to Pressure an or between Elastic Bodies BH. Formulas for Elastic Stability of Bars, Rings and Basins 3, 36 w. 5. Formulas for Elastic Stability of Plates and Shells Naural Frequencies of Vibration for Continuous Members Factor of Stress Concentration for Elastic Stress (8) 38, Representative Properties of Some Inportant Structural Materials 180 196, 200 210 220 240 252 290 300 04 330 332 a 386 416 48 158 a8 508 516 334 550 376 390 607Preface to the Fifth Edition ‘A major change in format in the many expanded tables of formulas is the ‘most obvious modification introduced in this edition. ‘The choice of format was greatly influenced by the general availability to design engineers of high-speed digital computers and by the recent development of both hand- and desk-model scientific: digital calculators, Where space permits, some equations are presented in reduced form for use with a slide rule; but for many of the more complex loadings, the evaluation of the necessary constants and variables requires ealculator accuracy. Since trigonometric, exponential, and hyperbolic functions are now available with the necessary accuracy, the evaluation of these functions presents litle difficulty. The general use of the singularity functions defined in Chapter { has measurably reduced the space required to present the formulas in this edition, All the new formulas tabulated have been programmed for a digital computer, and numerical values of forces, moments, and deformations have been determined and checked against the assumed boundary conditions. Where space permits, representative values of coefficients are tabulated to enable the reader to verify his own programs or, by interpolation, use the tabulated coefficients directly. ‘The addition of applied deformation as a form of loading has greatly expanded the versatility of many of the tables. The applied deformations consist of externally applied concentrated angular rotations and lateral displacements to beams, plates, and shells as well as linear temperature variations through the thickness of these structural members ‘The major additions in this edition are as follows: In Chapter 7 (Beams; Flexure of Straight Bars), in addition to a substantial increase in the loadingsif Preface to the Fifth Editon considered, all stable combinations of support conditions, including free, ‘imply sapported, guided, and fixed, are considered. This has been done in ‘Table 3 for common straight beams, in Table 4 for rigid frames, in Table 7 for short beams on clastic foundations, in Table 8 for long beams on elastic foundations, and in Table 10 for beams under axial compression and trans- verse loading. ‘Table 13 has been added to give correctiow® coefficients to be applied to reactions and deformations of uniform beams to account for many types and degrees of taper in the beam dimensions. In Article 7-2 on com posite beams and bimetallic sips, examples illustrate the use of formulas for equivalent material properties and stresses. "A separate chapter on curved beams (Chapter 8) has-been included owing to the growth in information on this subject. ‘Table 17 om circular Fings now includes corrections for deflections due to hoop stress and transverse shear, a ‘more detailed set of formulas for deformations, and tabulated force, moment, land deformation coefficients for several specific load locations. For many practical problems the formulas do not need to be solved. Table 16 for Gircular arches includes most of the loadings and all of the combinations of support conditions mentioned previously for straight beams and rigid frames ‘Table 19 is an extensive'tibulation of formulas for curved beams loaded ‘normal to the plane of curvature, including the majority of the stable com- binations iof support conditions. For beams avith round cross sections or having any shape-where the polar moment of inertia is owice'the bending moment of inertia, numetical coefficients for forces, moments, and deforma tions are tabulated for several combinations.of angular spans and -losd ‘locations i “Jn Chapter9-(Torsion) Table 21 has been added to give torsional stiffness and warping constants for thin-walled open cross sections; Table 22 uses these constants in formulas to determine angles of twist and stresses in thin-walled ‘beamisunder many torsional loading conditions and most stable combinations of support conditions. In Chapter 10 (Fiat Plates) Table 24 presents extensive tables of numerical coefficients at wells formulas for both solid circular and annular plates, ‘Table 25 gives shear deflections for circular plates; Table 26 ‘for plates with straight-boundaries has been greatly expanded. Formulas for ‘strentes and deformations in bimetallic circular plates have been added in Article 10-4: Tn Chapter 12 (Shells of Revolution; Pressure Veseels; Pipes) the tables have ‘been greatly'expanded, Specifically, Table 30 has been added to cover both Jong: and'short thin-walled cylindrical shelis under axisymmetric loading within thespan rather than justat the ends, Table 31 covers bending streses In thin-walled vessels, anil the formulassfor conical shells have been completely revised to give- more accurate results for a much wider range of cone param- eters; numerical data for short conical shells are also included, Body-force Joadings-for thick-walled vessels have been iicluded in Table 32, in addition to the usual pressure loadingsPreface to the Fifth Eiltion si In Chapter 14 (Elastic Stability) there has been a substantial increase in the ‘numerical data for buckling with loads applied at more than one position on 2 colusin simultaneously. Jn Chapter 15 (Dynamic. and Temperature ‘Streses), Table 36 gives natural frequencies and nodal positions for many continuous members, including beams and flat plates. Ik is the sincere hope of the author that the modifications made in this edition are in keeping with the philosophy practiced by Raymond J. Roark in previous editions, He assumed that good engineering judgment, coupled with sufficient information about the behavior of a similarly shaped sjractural member, should enable the reader to find an adequate solution to a given problem. This book was his continuing contribution to making such informa: tion readily available, It should be understood that although many new references are included herein, there is no way 10 ensure that much excellent work has not been overlooked and omitted, ‘The author wishes to acknowledge the assistance of those persons who have carefully checked formulas in previous editions and called attention to errors and omissions, as well as those individuals, publishers, institutions, and corporations who have generously given permission to use material in this edition. Special thanks are due those who have volunteered unpublished ‘material, which has added very measurably 10 the usefulness of this book. Warren C. Youngol Preface to th Firat Edition Because they are not believed to serve the purpose ofthis book, derivations of formulas and detailed explanations, such as are appropriate in a textbook, are omitted, buca sufficient number of examples are included to ilustrate the application oi the various formulas and methods. Numerous references 10 more detailed discussions are given, but for the most part these are limited 10 sources that are generally available and no attempt has been made to compile fan exhaustive bibliography ‘That such a book as this derives almost wholly from tne work of others is selfevident, and it is the author's hope that due acknowledgment has been made of the immediate sources of all material here presented. ‘To the pub- lishers and others who have generously permitted the use of material, he wishes to express his thanks, The helpful criticieme and suggestions of his colleagues, Professors E. R. Maurer, M. O. Withey, J.B. Kommers, and K. Wendt, are gratefully acknowledged. :A consilecabie number of the cables of formulas have been published from time to time in Product Enginzering, and the ‘opportunity thus afforded for criticism and study of arrangement has been of great advantage. Finally, it should be said that, although every care has been taken to avoid errors, it would be oversanguine to hope that none had escaped detection; for any suggestions that readers may make concerning needed corrections the author will be grateful Raymond J. RoarkPART ONE DefinitionsONE Definitions ‘The definitions given here apply to the terms in question as they are used in this book. Some of these tems are defined diferently by other authors; when ths is the case, the fact is noted. When two or more terms with identical meaning are in general acceptance, they are given in the order of the present ‘writers’ preference. The references referred to by number are listed at the ‘end of this section Allowable stress (working stres): If a member is so designed that the maximum stres as calculated for the expected conditions of service is less than some certain value, the member will have a proper margin of security against damage or failure. This certain value isthe allowable ses ofthe kind and for the material and condition of service in question. The allowable stress isle than the damaging sreas because of uncertainty as to the condi- tions of service, nonuniformity of material, and inaccuracy of svess analysis The margin berween the allowable stres and the damaging stress may be reduced in proportion to the certainty with which the conditions of service are known, the intrinsic reliability of the material, the accuracy with which the stress produced by the loading can be calculated, and the degree to which failure is unattended by danger or loss. (Compare Damaging sexs; Factor of safety; Factor of wilzation’ Margin of safety. See Re. 1, 2, and Table 40.) Apparent clastic limit (useful limit point): The ses at which the rate of change of strain with respect to stress 50 pereent greater than at 2eo stress It is more definitely determinable from the stress-strain diagram than is the proportional limit, and is useful for comparing materials of the same general dlass. (Compare Elastic lint; Proportional lit; Ved point; Yield eng) Apparent stress: The stress corresponding to a given unit strain on the 34 Formulas for Stress and Strain fount ‘assumption of uniaxial clastic stress. Tt is calculated by multiplying the unit train by the modulus of elasticity and may differ from the true stress because the effect of transverse stresses is not taken into account. Bending moment: Reference is to 8 beam, assumed for convenience to be horizontal and loaded and supported by forces, all of wich lie in a vertical plane. ‘The binding moment at any section of the beam is the moment of all forces that act on the beam to the left of that section, taken about the horizontal axis ofthe section. The bending moment is positive when clockwise and negative when counterclockwise; positive bending moment therefore ‘bende the beam so that it isconcave upward, and a negative bending moment ‘bende itso that it is concave downward. ‘The manent equation is an expression for the bending moment at-any-section in-terms of x-the-éistance to that section measured from a chosen origin, usually taken at the left end of the beam, ‘Boundary conditions: As used in strength of materials, the term usually refers to the condition of stres, displacement, or slope at the ends or edges . of a member, where these conditions are apparent from the circumstances fof the problem, Thus for @ beam with fixed ends, the zero slope at each ‘ead is boundary condition; for a pierced circular plate with freely supported edges, the zero radial seess at cach edge is @ boundary condition, Bitte fracture: The tensile failure with negligible plastic deformation of fan ordinarily ductile metal. (See Art. 27.) "Bulk modulus of elasticity: The ratio of a tensile or compressive stress, triaxial and equal in all directions (eg., hydrostatic pressure}, to the relative change it produces ia volume. Central axis (centroidal axis): A central axis of an area is one that passes through the centroid; it is understood to lie in the plane of the area unless the contrary is stated, When taken normal to the plane of the area, it is called the central polar axis: Centro of an area (center of gravity ofan area): That point in the plane of the area about any axis through which the moment of the area is zero: teoineides with the centerof gravity of the arca materialized as an infinitely thin homogeneous and uniform plate. ‘Corrosion fatigue: Fatigue aggravated by corrosion, asin parts repeatedly stressed while exposed (0 salt water. "Greep: Continuous increase in deformation under constant or decreasing sizes, The term is ordinarily used with reference to the behavior of metals tunder tension at elevated temperatures. ‘The similar yielding of a macerial ‘under compressive stress usually called plastic flow, or lw. Greep at atmos- ppheric temperature due to sustained elastic sires is sometimes called aif, fr elastic dif. Another manifestation of ereep, the diminution in stress when the deformation is maintained constant, is called relaxation Damaging stress: The least unit stress of a given kind and for a given ‘material and condition pf servies that will render @ member unfit for service bbelore the end of its normal life, It may do this by producing excessive set,ome Defitioms 5 by causing creep to occur at an excesive rate, or by causing fatigue eracking, cexeenive strain hardening, or ripe Damping capaciy: The amount of work disipated into heat per unit of total stain energy present at maximum strain for a complete cycle ‘Deformation (train): Change in the form or dimensions of a body pro- duced By stress. Elogtion is often sed for tensile sin, cmpresion oF “hrtenng for compresive strain, and drain fr shear stain. Elst defomation issuch deformation as disappears on removal of stress pemanet dfrmation is such deformation ar remains on removal of sues. (Compare Se) Eecentriiy:.Aload or component ofa load normal toa given eos section ofa member scccentie with respect that section ifit doesnot aet through the centroid, ‘The perpendicular distance Srom th lie of action of the load to.either principal central axis is the econtty with respect to that axis. Elastic! Capable of sustaining ses without permanent deformation; the term is also used to denote conformity to the law of stesstrain propor tionality. An clastic stressor strain ia stress or strain within the elastic limit, “Elastic aris: The clastic axis of a beam isthe line, lengzhwise ofthe beam, long which transverse losds must be applied in order to produce bending only, with no torsion of the beam st any section, Strictly speaking, no such Tine exists except for afew conditions of loading. Usually the elastic axis is assumed tobe te ine that pastes through the elastic center of every section ‘The term is most often used with reference to an airplane wing of either the shell or mutiple-spar type. (Compare Torsional cnt; Flea ent; Elastic centr, See Ref. 4) “laste center: “Te elastic centr of.a given setion of beam is that point in the plane of the section lying midway between the flexural center and Center of twist of that section. ‘The three points may be identical and are ‘sally assamed to beso, (Compare Flennal ener: Torsional exe: Elsie exis, See Refs 4 and 5) ‘Bleatc euves The curve assumed by the axis of a normal straight beam or columa when bent by loads that do not stes it beyond the proportional limit "Elastic fnit: “The least stres chat will cause permanent Set. (Compare Proportional iit; Appane! elastic ity Vild pon; Yd swegth. See Art: 2.2 and Ref. 6) “Elstc ratio: “The ratio of the elastic limit to the ultimate strength, llipsoid of strains An elipsoid that represents the state of stain at any given point in 2 body; it has the form assemed under ses by 2 sphere Centered at the point in question (Ref. 7) ‘Plipsoid of stress: Arsllipsoid that represents the sete of stress ata given point in a body, its semiaxes are vectors representing the prneipal stresses 2 the point, and any radius voctor represents the resultant ses on @ particular plane through the point. For a condition of plane stress (one Principal sizes zero) the ellipsoid becomes the elise of sss (see Ref. 8) Endurance limit fatigue strengti): The maxiroum stress that can be6 Formulas for Stress and Strain [oie reversed an indefinitely large number of times without producing fracture of @ material (see Art. 28), Endurance ratio: Ratio of the endurance limit to the ultimate static tensile strength. ‘Endurance strength: The highest stress that 2 material-can withstand with ‘repeated application or reversal without rupture for a given number of cycles is the endurance strength of that material for that number of cycles. Unless “otherwise specified, reversed stressing is usually implied. (Compare Endurance Fini.) Energy of rupture: The work done per unit volume in producing fracture. Iris not practicable to establish a definite energy of rupture value for # given, ‘material because the result obtained depends upon the form and proportions of the test specimen and the manner of loading. “As determined by similar tests on similar specimens, the energy of rupture affords a criterion for comparing the toughness of different materials. Equivalent bending moment: A bending moment that, acting alone, would produce in a circular shaft 2 normal (tensile or compressive) stress of the same thagnitude as the maximum normal stres produced by a given bending moment and a given twisting moment acting simultaneously Equivalent avisting moment: A twisting moment that, acting alone, would produce in a circular shaft a shear stress of the same magnitude as the shear ‘stress produced by @ given twisting moment and a given bending moment acting simultaneously. Factor of afety: The ratio ofthe load that would cause failure of a member fr structure to the load thats imposed upon it in service. The term usually has this meaning; it may alio be ised to represent the ratio of breaking to service value of speed, deflection, temperature variation, or other stress producing factor against possible increase in which the factor of safety is provided as a safeguard. (Compare Allowable stzss; Margin of safely. See Table 40.) Factor of strain concentration: In the presence of stres raisers, localized peak strains are developed, The factor of strain concentration is the ratio fof the localized maximum strain at a given cross section to the nominal average rain on that eros section. ‘The nominal average strain is computed from « the average stree and a knowledge of the stress strain behavior of the mate~ rial. In a situation where all streses and strains are elastic, the factors of stress concentration and strain concentration are equal. (Compare Factor of ‘stress concentration.) Factor of stress concentration: Irregularities of form such as holes, screw threads, notches, and sharp shoulders, when present in a beam, shaft, or other ‘member subject to loading, may produce high localized stresses. This phe> ‘nomenon is called sires: concmiratim, and the'form irregularities that cause it are called stves raisers, "The ratio of the true maximum stress to the stress caloulated by the ordinary formulas. of mechanics (Hexure formula, torsion formula, etc), using the net section but ignoring'the changed distributioncm 1} Definitions 7 of stress, isthe factor of stress concentration forthe particular type of stress raiser in question. (See Art. 2.10.) Factor of stress concentration in fatique: At a specified number of loading cycles, the fatigue strength of @ given geometry depends upon the stress concentration factor and upon material properties. The fair of tess concen train in fatigue isthe ratio of the fatigue strength without a stress concen- tration to the fatigue strength with the given stress concentration. It may vary with the specified number of cycles as well as with the material. (See Niothsensitiviy retin) Factor of utilization: The ratio of the allowable stress to the ultimate strength. For cases in which stress is proportional to load, the factor of utilization is the reciprocal of the factor of safety (see Ref. 1). Fatigue: Tendency of materials to fracture under many repetitions of a stress considerably les than the ultimate stati strength atigue-strength reduction factor: Alternative term for factor of ses eomeen tration in fatigue. “Fiber stress: A term used for convenience to denote the longitudinal tensile cor compressive stres in a beam or other member subject to bending. It is sometimes used to denote this stress atthe point or points most remote from the neutral axis, But the term stress in extreme fiber i preferable for this purpose. Also, for convenience, the longitudinal elements or laments of which 2 beam may be imagined as composed are called jes. Fixed (clamped, builtin, ecastré): A condition of support at the ends of «beam or column or at the edges of a plate or shell that prevents rotation nd ransoersedsplaconet of the edge ofthe neutral surface but permits longiti- inal displacement, (Compare Guided: Held: Suporte) lesural center (shear center): With reference to & beam, the flexural center of any section is that point in the plane of the section through which a transverse load, applied at that section, must actif bending deflection only isto be produced, with no twist ofthe section. (Compare Tersional enter; Elastic center; Elastic axis. See Refs. 4 and 9.) Form factor: The term pertains to a beam section ofa given shape and means the ratio of the modulus of rupture of a beam having that particular section to the modulus of rupture of a beam otherwise similar but having 2 section adopted as standard. This standard section is usually taken as rectangular oF square; for wood its a2 by 2 in square, with edges horizontal and vertical (ce-Arts. 2.11 and 7.3). The term is also used to mean, for a given maximum fiber suess within the elastic limit, the ratio of the actual resisting moment of a wide-flanged beam to the resisting moment the beam ‘would develop ifthe ber stres were uniformly distributed across the entire ‘Width ofthe flanges, So used, the term expresesthestrength-reducing effect of shear lag Frettng fatigue (chafing fatigue): Fatigue aggravated by surface rubbing, fas in shafts with prese-fitted collars Guided: A. condition of support at the ends of a beam or column or at@ —Formutas for Stress and Strain four 1 the edge of a plate or shell that prevents rotation of the edge of the neutral surface but permits longitudinal and transverse displacement. (Compare Fisd: Held: Supported) “Held: A condition of support at the ends of @ beam or column or at the ‘edge ofa plate o shell that prevents longitudinal and transverse displacement of the edge of the neutral surface but permits rotation in the plane of bending. (Compare Fixed; Guided: Suppor) “Influence ine: Usually pertaining to a particular section of a bear, an influence line is a curve drawn so that its ordinate at any point represents the value of the reaction, vertical shear, bending moment, or deflection produced at the particular section by a unit load applied at the point where the ordinate is measured. An influence fine may be used to show the effect ‘of load position on any quantity dependent thereon, such as the stress in a given truss member, the deflection of a wuss, or the twisting moment in a shaft Toclinic: A line (in a stressed body) at all pointi on which the corre- sponding principal stresses have the same direction, ‘Lolropie: Having the same properties in all directions. In discussions pertaining to strengths of materials, isotropic usually means having the same Strength and elastic properties (médulus of elasticity, modulus of rigidity, and Poiison's ratio) in all directions. ‘Kern (kernel): Reference is to some particular section of a member. The ern is that arca in the plane of the scetion through which th line of action of & force must pais. if that fore is to produce, at all point in the given fection, thesame kind of normal stress, e, tension throughout oF compression throughout. Litders tines: See Slip tines. “Margin of safety: As used in aeronautical design, thargin of safety is the percentage by which che ultimate strength of a member exceeds the design load. The design Ind isthe applied load, oF maximum probable load, multi plied by a specified factor of safety. [The use of the terms margin of safety find design loadin this sense is practically restricted to aeronautical engineer ing (see Ref. 10) | ‘Mechanical hysteresis: ‘The dissipation of energy as heat during a stres ‘cle, which isrevealed graphically by failure ofthe descending and ascending branches of the stressstrain diagram to coincide “Menber: ny part or clement of a machine or structure such as a bear, column, oF shaft Modulus of elasticity (Young's modulus): "The rate of change of unit tensile Gr compressive atres with respect to unit tensile of compressive strain for the condition of uniasial ses within the proportional limit. For-most, but not all, materials, the modulus of elasticity is the same for tension and compression. For sonisotropic materials such as wood, it is necessary to distinguish between the moduli of elasticity in different directionsom.) Definitions 9 Modulus of resilience: “The strain energy per unit volume absorbed up to the clastic limit under the condition of uniform uniaxial sess Modulus of rigidity (modulus of elasticity in shear): The rate of change ‘of unit shear stress with respect to unit shear strain forthe condition of pure shear within the proportional limit. For nonisotropic materials such as wood, it is necessary to distinguish between the moduli of rigidity in diferent directions. ‘Modulus of rupture in bending (computed ultimate bending strengtt): The fictitious tensile ar compressive sires inthe extreme fiber ofa beam computed by the flexure equation © = Mc/I, where Mis the bending moment that causes rupture “Modulus of rupture in torsion (computed ultimate twisting strength): The fictitious shear stress at the surface ofa circular shaft computed by the torsion formula + = Ti/j, where T is the twisting moment that causes rupture “Moment of en area (frst moment of an area, statical moment of an area): With respect to an axis, the sum of the products obtained by multiplying cach clement of the atea dd by its distance from the axis y; itis therefore the quantity fz. An axis in the-plane of the area is implied. ‘Moment of inerita of an area (second moment of an area): ‘The moment ‘of in. tia of an area with respect w ap axis s the sum of the products obtained by multiplying each element of the area dA by the square ofits distance from the axis; itis therefore the-quantity f4,#. An axisin the plane ofthe area is implied if the axis is normal to that plane, the term polar moment of inertia is sed (see Chap. 5). Neutral axis: ‘The line of ero fiber stress n any given section of member subject to bending; itis the line formed'by the intersection of the neutral surface and the section ‘Neutral surface: "The longitudinal surface of zero fberstress in-z member subject t0 bending; it-contains-the neutral axis of every section. Noteh-rensitivity ratio: Used to compare sss concentration fact ky and {atigiestrngth reduction far, the noteh-sensitivity ratio is commonly defined ‘as the ratify —T)/(ly — 1). Itvaries from 0 for some sft ductile materials, to 1, for some hard:-britde materials Plastic moment; plastic hinges plastic section modulus: These terms are explained in Art. 716. ‘Plasticity: The property of sustaining appreciable (visible to the eye) permanent deformation without rupture, The term is also used 10 denote the property of yielding or flowing under steady load (Ref. 11) ‘Poison’ ratio: ‘The ratio of lateral unit strain to-longitudinal unt Strain ‘under the condition of uniform and\uniaxial longitudinal stress within the proportional limit. 49) ‘Polar mament of inertia: See Moment of inatia ofan area. Principal axes: ‘The principal axes of an area fr a given point in is plane are the two mutually perpendicular axes, passing through the point and lying10 Fermealas for Stress and Strain [ems 1 in the plane of the area, for one of which the moment of inertia is greater land for the other less thaa for any other coplanar axis passing through that point, Irthe point in question is the centroid of the area, these axes are called ‘principal central ‘axes (See Chap. 5) ‘Principal moment ofinertia: "The moment of inertia ofan area about either principal axis (see Chap. 5). ‘Principal planes; principal stresses: ‘Through any point in a stressed body ‘there pats three mutually perpendicular planes, the stress on each of which is:purely normal, tension, or compression; these are the principal planés for ‘that point. The stresses on. these planes are the principal stresses; one of them is the maximum stress at the point, and one of them is the minimum stress ‘atthe point, When one of the principal stresses is.zero, the condition is one of plane stress; when two of them are zero, the condition is one of uniaxial Product of inertia of an area: With respect to a pair of rectangular axes in its plane, the sum of the products obtained by multiplying each clement of the area dA by its coordinates with respect to those axes x and y; itis therefore the quantity dA 2y (see Chap. 5). Proof stress: Pestaining to acceptance tests of metals, specified tensile stress that must be sustained without deformation in excess of a specified ‘Proportional limit: ‘The greatest stess that a material can sustain without deviating from the law ofstres-strain proportionality. (Compare Elastic limi Apparent elastic limit; Yield pont; Yield strength. See Art, 2.2 and Ref. 6.) ‘Radius of gration: ‘The radius of gyration of an area with respect to @ given axisis the square root of the quantity obtained by dividing the moment of inertia of the area with respect to that axis by the area (see Chap. 5), Reduction of area: ‘The difference between the cross-sectional area of a tension specimen at the section of rupture-before loading and after rupture, usually expressed as a pereentage of the original area. Rupture factor: Used in reference to brittle materials, ie., materials in which failure occurs through tensile rupture rather than through excessive deformation, For a member of given form, size, and material, loaded and supported in a given manner, the rapture ficor is the ratio of the fictitious ‘maximum tensile stress at failure, as calculated by the appropriate formula for elastic stress, to the ultimate tensile strength of the material, as determined by a conventional tension test (Art. 2.11). ‘Section modulus (section factor): Pertaining to the cross section of a beam, ‘the section modulus with respect to either principal central axis is the moment of inertia with respect to that axis divided by the distance from that axis ‘0 the most remote point of the section. ‘The seetion modulus largely deter- mines the flexural strength of a beam of given material. Set (permanent set, permanent deformation, plastic strain, plastic deforma- tion): Strain remaining after removal of stres. ‘Shakedown load (stabilizing load): The maximum load that can be appliedoun} Defiitions 11 to a beam or rigid frame and on removal leave such residual moments that subsequent applications of the same or a smaller load will cause only elastic ‘Shape factor: ‘The ratio of the plastic section modulus to the elastic section modulus. ‘Shear center: See Flesural center. ‘Shear lag: On account of shear strain, the longitudinal tensile or com- ‘pressive bending stress in wide beam fianges diminishes with the distance from the web or webs, and this stres diminution is called shear dag. ‘Singularity functions: A class of functions that, when used with some caution, permit expressing in one equation what would normally be expressed in several separate equations, with boundary conditions being matched at the ends of the intervals over which the several separate expressions are valid. Singularity finctions are commonly employed in heat transfer and electro- magnetic theory at well as in the evaluation of shears, bending moments, and deformations in beams, plates, and shells. Thesingularity function most often expressed in this reference is the slep fiction, written with the bracket notation (x — a)°, which is defined as having a value of ze if'x a. The indeterminate value when x= is of no ‘consequence since herein it ie always multiplied by another function that will have a value of zero when x = a. Integrals of the step function, such as the ramp finction (x ~ a) and all others expressed as (x — a)", are defined as having a'value of zero if x a ‘Slenderness ratio: The ratio of the length of a uniform column to the least radius of gyration of the cross section. ‘Slip lines (Liiders lines): Lines that appear on the polished surface of a crystal or crystalline body that has been stressed beyond the elastic limit. ‘They represent the intersection of the surface by planes on which shear stress has produced plastic slip or gliding (see Art. 25 and Ref. 11) ‘Step function: See Singularity fiction ‘Strain: Any forced change in the dimensions of a body. A stretch is a lunsile strain: 2 shortening is a compressive strain; and an. angular distortion is ‘shear tran." The word strain is;commonly used to connote wnt strain (which see); this usage is followed in this book. Strain concentration factor: See Factor af stain concentration, ‘Strain energy (elastic energy, potential energy of deformation): Mechanical ‘nergy stored up in strested material. Stress within the elastic limit is implied; therefore, the strain energy is equal to the work done by the external forces in producing the stress and is recoverable. Strain rosette: At any point on the surface of a stressed body, strains ‘measured on each of three intersecting gage lines make possible the calcula- tion of the principal stresses. Such gage lines and the corresponding strains are called a strain roset Stress: Internal force exerted by either of two adjacent parts of a body12 Formula for Stress and Strain fawe. ‘upon the other across an imagined plane of separation. When the forces are parallel to the plane, che stres is called sear stress; when the forces are normal to the plan, the stress is called normal stress; when the normal sressis directed toward the part on which it acts iecalled compressive sres-and when i is directed away from the part on which it act, ite called lenslesress. Shear, compresive, and tensile stesves, respectively; resist the endency of the parts to mutually slide approach, or acparate under the action of applied force. Por‘bievity, the word terri often used Yo-eonnote unit sess (which see); this usage is followed in this book. ‘Stress concentration ctor: See Factor of stress cocetaton. “Sireitsolids "The solid igure formed by surfaces bounding vectors drawn atall poinw ofthe cross section ofa member and representing the unit normal streis ateach sich point. ‘The stresrsolid givesa picture of the stres dstribu- Stresestrain diagram (stress diagram): The curve obtained by-plowing tani sreses as ordinates against corresponding unit strains as abscissa. Stress trajectory ‘oostati): A Tines(in a stressed body) tangent to the direction of one of the principal streser-at every point through which it passes. ‘Supported (simply supported): condition of support atthe ends ofa bear or column orst the edge ofa plate or shell that prevents transverse displace- ment of the edge ofthe neutral surface but permits roration and longitudinal displacement. (Compare Fixed Guided Held) Syetem:. Denotes-anry member oriassemblage of memibers.such as a com- porite column, coupling, truss, or other structure. Torsional center (center of tacst,-centr of torsion, center of shear): Uf a twisting couples applied ata given section ofa straight member, that section rotates about some point in its plane. This point, which doesnot move when the member twists, ithe torsional enter ofthat section, Ttis sometimes defined as though identical with the flexural center, but the two points do not always coincide. (Compare Flewral center, BlastiesCente; Elastic Axis. See Rebs. 4 and 5.) True strain (natural strain, logarithmic strain): The integral, over the whole o'a finite extension, of each ‘infinitesimal-clongation divided by the corresponding momentary length. It isequal to.log, (1 +<), wheres is the unit normal strain as ordinarily defined (Ref. 12). ‘True stress: “For an axially loaded bar, the load divided by the corre- sponding actual cross-sectional area. It differs from the stress as ordinarily Gefined because of the change in area due to loading. Thisting-mament (torqu): At any section of a member, the moment of all forees that act on the member to the left (or Fight) of that section, taken about a polar axis through the flexural center of that section. For sections that are symmetrical about each principal central axis, che flexural center coincides with the eentoid (sce Refi. 5 and 9, timate elongation: "The percentage of permanent deformation remainingcour Definitions 13 after tensile rupture measured over an arbitrary length including the section of rupture. Ultimate strength: The ultimate strength of 2 material in tension, compres. sion, or shear, respectively, is the maximum tensile, compressive, or shear stress that the material can sustain calculated on the basis of the ultimate load and the original or unstrained dimensions, It is implied that the condition of stress represents uniaxial tension, uniaxial compression, or pure shear, as the case may be Unit strain: Unit tensile train isthe elongation per unit length; unit compres: sive strain is the shortening per wnit length; and ani shear strain is the change in angle (radians) between two lines originally at right angles to each other. Unit stress: "The amount of stress per unit of area, The unit stress (tensile, compressive, or shear) at any point on a plane is the limit, as A approaches 0, of AP/AA, where AP is the total tension, compression, or shear on an area AA that lies in the plane and includes the point. In this book, stress connotes, vont stress, in general. Vertical shear: Refers to a beam, assumed lor convenience to be horizontal and loaded and supported by forces that all lie ina vertical plane. ‘The vertical shear at any section of the beam is the vertical component of all forces that fact on the beam to the let of the section. The shear equation is an expression for the vertical sheat at any section in terms of x, the distance to that section ‘measured.from a chosen origin, sually taken at the left end of the beam. Yield point: The lowest stess at which strain increases without increase in stress. For some purposes it is important to distinguish between the ypper yield point, which is the stress at which the stress-strain diagram frst becomes, horizontal, and the lower yield point, which is the somewhat lower and almost ‘constant stress under which the metal continues to deform. Only a few materials exhibit a true yield point; for other materials the term is sometimes, used synonymously with yield strength. (Compare Yield strength; Elastic limit Apparent elastic limits Proportional limit, See Ref. 6) REFERENCES: 1. Soderberg, ©. Re; Working Strenen, ASME Paper A105, J. Appl Mech, vo. 2, no. 8, 1985 2. Unit Stem in Suruetural Material (Symposium), Tiow. dm. Soc Cin! Eng, wal. 9, 08,1927 4. ton Heydenkampf, G. S: Damping Capaciry of Materials, Pe. ASTM, vol 1, pat Th, p57, 1951 4, Ruin, P; Remark on the Elatc Axis of Shell Wings, Nat. Ack. Crm Ac, Tick Note 962, 1836. 5, Schwalbe, W. Li The Center of Totton for Angle and Channel Sections, Tim. ASME, Paper APMCSS-1, vol 5, no. 1, 1992 {6 Tentative Definitions of Terms Relating to Methods of Testing, Pos ASTM, vol. $8, pat 1p. 1815, 1985, 7, Morey, Ac "Stength of Material" Sih ed, Longmans, Geen & Co.,Ltd 1919. 8, Timathenio, S: "Theory of Blais.” Engnoerng Societies Monograph, MoGraw-t Bock Company, 1934.14 Formulas for Stress and Strain ems 1 8. Getic, A, Ay and G. 1. Taylor: The Problem of Fenureand'Ie Soliton by the Soxp ‘iim Method, Trl ep. A. Cor, Aare (Beh), Report and Memon na 399, p. 950, wn, 10, Airworthinest Requirement for Airraft, Aen, Bull. 1-A, US. Dept. of Commerce, 194, 1, Nada, A "Plasticity," Engineering Sociedes Monograph, McGraw ll Book Company, i 12 Freudenthal, A.M. “The Inlatio Behavior of Eagineaing Materials and Stractres” Jobe Wiley & Sans, Ine, 1950PART TWO Facts; Principles; MethodsTwo The Behavior of Bodies Under Stress ‘This discussion pertains to what are commonly designated as structural mate rials, that is, materials suitable for structures and members that must sustain loads without suffering damage. In this category are included most of the metals, concrete, wood, brick and tile, stone, glass, some plastics, ete. It is beyond the scope of this book to give more'than a mere statement of a few important facts concerning the behavior of a material under stress; extensive literature is available’on every phase of the subject, and the articles referred to will serve as an introduction. (For numerical values of quantities diraused in this section, the reader is referred to Table 38.) 21 Methods of loading ‘The mechanical properties of a material are usually determined by laboratory tests, and the commonly accepted values of ultimate strength, elastic limit, etc,, are those found by testing a specimen of a certain form in a certain manner. To apply results so obtained in engineering design requires an understanding of the effects of many different variables, such as form and scale, temperature and other conditions of service, and method of loading. ‘The method of loading, in particular, afécts the behavior of bodies under stress. There are an infinite number of ways in which strese may be applied to a body, but for most purposes it is sufficient to distinguish the types of loading now to be defined. 1! Short-time static loading. The load is applied so gradually that at any18 Formulas for Stress and Strain [onan 2 instant all parts are esentially in equilibrium. In testing, the load is increased progressively until failure occurs, and the total time required to produce failure is not more than a few minutes. In service, the load is increased progressively up to its maximum value, is maintained at that maximum value for oniy 2 limited time, and is not reapplied often enough to make fatigue a consideration. ‘The ultimate strength, clastic limit, yield point, yield strength, and modulus of elasticity of a material are usually determined by short-time static testing at room temperature, 2, Long-time static lading. ‘The maximum load is applied gradually and ‘maintained. In testing, it is maintained for a sufficient time to enable its probable final effect o be predicteds in service, itis maintained continuously for intermittently during the life of the structure, ‘The creep, or flow charac- teristics, of a material and its probable permanent strength are determined by long-time static testing at the temperatures prevailing under service conditions. (See Art 2.6.) 3. Repeated loading. Typically, a load or stress is applied and wholly or partially removed or reversed repeatedly. This type of loading is important if high stresses are repeated for # few cycles or if relatively lower stresses are repeated many times; it is discussed under Fatigue. (See Art. 2.8.) 4. Dynamic loading. ‘The ciecumstances are such that the rate of change ‘of momentum of the parts must be taken into account. One such condition may be that the patts are given definite avceleations corresponding to a controlled motion, such as the constant acceleration of @ part of a rotating ‘member or the repeated accelerations suffered by a portion of a connecting rod. As far as stress effects are concerned, these loadings are treated as virtually static and the inertia freer (Art 15.2) are treated exactly as though they were ordinary static loads. ‘A second type of quasistatic loading, quick static loading, can be typified by the rapid burning of a powder charge in a gun barrel. Neither the powder, gas, nor any part of the barrel acquires appreciable radial momentum; therefore equilibrium may be considered to exist at any instant and the ‘maximum stress produced in the gun barre isthe same as though the powder pressure had developed gradually In static loading and the two types of dynamic loading jst described, the lodded member ie required to resist a definite force. It is important to distin ‘guish this from impact lading, where the loaded member is usually required! to absorb a definite amount of energy, Impact loading can be divided into two general categories. In the first ‘case a relatively large slow-moving mass strikes 2 less massive beam or bar ‘and the kinetic energy of the moving mass is assumed to be converted into sirain energy in the beara. All portions of the beam and the moving mass are ‘assumed to stop moving simultaneously. The shape of the elastic axis of the deflected beam or bar is thus the same as in static loading, A special case ‘of this loading, generally called sudden loading, occurs when a mass that is : | |ae 22) The Behavior of Bodies Unier Strese 19 ‘not moving is released when in contact with a beam and falls through the distance the beam deflects. This produces approximately twice the stress and deflection that would have been produced had the mass been “eased” oto the beam (see Art. 15.4), The second case of impact loading involves the mass of the member being struck. Stess waver travel through the member during the impact and continue even after the impacting mass has rebounded (Gee Art. 15.3) ‘On consideration, it is obvious that methods of loading really differ only in degree. As the time required for the load to be applied increases, short-time static loading changes imperceptibly into long-time staticloading; impact may be produced by a body moving so slowly that the resulting stress conditions are practically the same as though equal deflection had been produced by snatic loading; the number of stress repetitions at which fatigue becomes involved is not altogether definite. Furthermore, all these methods of loading ‘may be combined or superimposed in various ways. Nonetheless, the clast- fication presented is convenient because most structural and machine parts function under loading that may be classified definitely as one of the types deseribed, 2.2 Elasticity; proportionality of stress and strain stress by mathematical analysis, it is customary to assume is elastic, isotropic, homogeneous, and infinitely divisible with- ‘out change in properties and that it conforms to Hooke’s law, which states ‘hat strain is proportional to stress. Actually, none of these assumptions is strictly true. A-structural material is usually an aggregate of crystals, fibers, cor cemented particles, the arrangement of which may be either random or systematic, When the arrangement is random, the material is essentially isotropic if the part considered is large in comparison with the constituent units; when the arrangement is systematic, the elastic properties and strength are usually different in different directions and the material is anisotropic. Again, when subdivision is carried to the point where the part under consid- eration comprises only a portion of a single ciystal, ber, or other unit, in all probability its properties will difer from those of a larger part that is ‘ap aggregate of such units. Finally, very careful experiments show that for all materials there is probably some set and some deviation from Hooke's law for any stress, however small, ‘These facts impose certain limitations upon the conventional methods of stress analysis and must often be taken into account, but formulas for stress and strain, mathematically derived and based on the assuzmptions stated, give satisfactory results for neatly all problems of engineering design. In particu» lar, Hooke's law may be regarded as practically true up to-a proportional limit, which, though often not sharply defined, ean be established for most20 Formulas for Strose and Strain [ome 2 materials with sufficient definiteness, $0, too, a fairly definite elastic limit is determinable; in most cases ic is so neatly equal to the proportional limit that no distinction need be made between the two. 23 Factors affecting elastic properties For ordinary purposes it may be assumed that the elastic properties of most metals, when stressed below a nominal proportional limit, are constant with respect to stress, unaffected by ordinary atmospheric variations of tempera ture, unaffected by prior applications of moderate stress, and independent of the rate of loading. When precise relations between stress and strain are portant, as in the design or calibration of instruments, these assumptions cannot always be macle. The fourth edition of this book (Ret. 1) discussed in detail the effects of strain rate, temperature, et, on the elastic properties of many metals and gave references for the experiments performed. The relationships between atomie and molecular structure and the elastic proper- ties are discussed in Ref. 2 ‘Wood exhibits a higher modulus of elasticity and mach higher proportional limi when tested rapidly than when tested slowly. The standard impact test con a beam indicates a fiber stess at the proportional limit approximately twice as great as that found by the standard static bending test. Absorption ‘of moisture up to the fiber saturation point greatly lowers both the modulus ‘of elasticity and the proportional limit (Ref. 4). Both concrete and cast iron have stresestrain curves more or less curved throughout, and neither has a definite proportional limit. For these materials it is customary to define E as the ratio of some definite stress (for example, the allowable stress or one-fourth the ultimate strength) to the corresponding unit stain; che quantity so determined is called the secant modulus since it represents the slope of the secant of the stresrstrain diagram drawn from ‘the origin to the point representing the stres chosen. ‘The moduli of elasticity ‘of cast iron are much more variable than those of steel, and the stronger grades are stiffer than the weaker ones. Cast ion sufers a distnet set from the frst application of even a moderate stress; but after several repetitions of that tess, the material exhibits perfect elasticity up to, but not beyond, that stress ‘The modulus of elasticity is slightly less in tension then in ‘compression (Ref. 5). ‘Concrete also shows considerable variation in modulus of elasticity, and in general its sifines increases with its strength. Like cast iron, concrete can bbe made to exhibit perfect elasticity up to a moderate stress by repeated loading up to that stress. Because of is tendency to yield under continuous loading, the modulus of elasticity indicated by long-time loading is much less than that obtained by progressive loading at ordinary speeds (Refs. 3 and 18)axe 24) The Behavior of Badies Under Stress 21 24 Load-deformation relation for a body If Hooke’s law holds for the material of which a member or structure is ‘composed, the member or structure will usually conform to a similar law of load-deformation proportionality and the defection of a beam or truss, the twisting of a shaft, the dilation of a pressure container, te,, may in most instances be assumed proportional to the magnitude of the applied load or loads ‘There are two important exceptions to this rule. One is to be found in any case where the stresses due to the loading are appreciably affected by the deformation. Examples of this are: a beam subjected to axial and wansverse loads; a flexible wire or cable held at the ends and loaded trans- versely; a thin diaphragm held at the edges and loaded normal to its plane; a ball pressed against a plate or against another ball; and a helical spring under severe extension, ‘The second exception is represented by any ease in which failure occurs through elastic instability, as in a slender (Euler) column. Here, for loads less than the critical, elastic instability playsno part and the load deformation, is linear. At the critical load the type of deformation changes, the column bending instead of merely shortening axially, and the defermation becomes indeterminate. For any load beyond the critical load, failure eccurs through ‘excessive deflection (see Art. 2.13), 2.5 Plasticity Elastic deformation represents an actual change i the distance between atoms or molecules; plastic deformation represents a permanent change in their relative postions. In crystalline material this permanent rearrangement consists largely of group displacements of the atoms in the erystal lattice brought about by slip on planes of least resistance, parts of a crystal sliding ppast one another and in some instances sufering angular displacement. In amorphous material the rearrangement appears to take place through the individual shifting from positions of equilibrium of many atoms or molecules, the cause being thermal agitation’ due to external work-and the result ap” pearing as a more or less uniform flow like that ofa viscous liquid. Tt should bbe noted that plastic deformation before rupture is much less for biaxial or Usiaxial tension than for one-way stress; for this reason metals that are ‘ordinarily ductile may prove brittle when thus stressed: ‘The laws governing plastic deformation are less amenable to mathematical statement than those assumed to govern elastic behavier, but a mathematical theory of plastic action is being'developed, Important applications are to the prediction of the ultimate strength and postbuckling behavior of struc tures and in the study of metal working,.creep, and flow (see Refs. 6.10.8522 Formulas for Stress and Strain fenar 2 2.6 Creep and rupture under Tong-time loading ‘Most materials will creep or flow to some extent and eventually fail under ‘a sustained stress Jess than the short-time ultimate strength. After a short time at load, the initial creep related to stress redistribution in the structure ‘and strain hardening ceases and the steady sa, or viseus creep, predominates ‘The viscous creep will continue-until fracture unless the load is reduced sufficiently, but itis seldom important in materials at temperatures less than 40 to 50 percent of their absolute melting temperatures, Thus, creep and long-time strength at atmospheric temperatures must sometimes be taken into ‘account in designing members of nonferrous metals and in selecting allowable stress for wood, plastics, and concrete. ‘Metals. Creep is an important consideration in high-pressure steam and distillation equipment, gas turbines, nuclear reactors, supersonic vehicles, etc Lubahn and Felgar (Ref. 8) relate creep to plasticity. Marin, Odqvist, and Finnic, in Ref. 9, give excellent surveys and list references on creep in metals ‘and structures, Conway (Refs, 10 and 11) discusses the effectiveness of various parametric equations, and Gonway and Flagella (Ref. 12) present extensive ‘creep-rupture data for the refractory metals, Odqvist (Ref. 13) discusses the theory of ereep and its application to large deformation and stability prob- lems in plates, shells, membranes, and beams and tabulates creep constants for 15 common metals and alloys. Hult (Ref. 14) also discusses creep theory ‘nd its application to many structural problems. Penny and Marriott (Ref, 15) discuss creep theories and the design of experiments to verify them. ‘They also discuss the development of several metals for increased resistance to creep at high temperatures at well at polymeric and composite materials at lower temperatures, Reference 16 it-a series of papers with extensive references coveriniy creep theory, material properties, and structural problems. ‘Plastics, ‘The literature on the behavior of the many plastics being used for structural or machine applications is too extensive to list here. Reference 117 summarizes much of the data available and gives the necessary references, Concrete. Under sustained compressive sires concrete suffers considerable plastic deformation and may flow for a very long time at stresses less than the ordinary working stress, Continuous flow has been observed over a period of 10 years, though ordinarily it ceases or becomes imperceptible within for 2 years. The rate of flow is greater for air than for water storage, greater for small than for large specimens, and for moderate stresses increases ap- proximately as the applied strese. On removal of stress, some elastic recovery ‘occurs, Concrete also shows creep under tensile stress, the early creep rate being greater than the flow rate under compression (Refs. 18 and 23). ‘Under very gradually applied loading concrete exhibity an ultimate strength considerably less than that found under short-time loading; in certain ‘compression tests it was found that increasing the time of testing from 15san 271 ‘The Behavior of Bodies Under Stress 29 to 4h decreased the unit stress at failure about 30 percent, most of this decrease occurring between the extremely quick (I of 2) and the conven tional (several minutes) testing. This indicates that the compressive stress that concrete can sustain indefinitely may be considerably less than the ultimate strength as determined by a conventional tes. On the other hand, the long-time imposition of moderate loading appears to have no harmful effect; certain tests show that after 10 years of constant loading equal to ‘one-fourth the ultimate strength, the compresive strength of conerete cylin- ders is practically the same and the modulus of elasticity is considerably sreater than for similar eylinders that were not kept under load (Ref. 22) “The modulus of rupture of plain concrete also decreases with the time of Joading, and some tests indicate thatthe long-time strength in ros-breaking may be only 55 to'75 percent of the short-time strength (Ref. 19) [Ref. 24 isa compilation of 12 papers, each with extensive references, dealing with the effect of volumetric changes on concrete structures. Design modi- fications to accommodate these volumettie changes is the main thrust of the papers Wood. Wood also yields under sustained stress; the long-time (several years) strength is about 55 percent of the short-time (several minutes) strength in bending; for direct compression parallel to the grain the corresponding ratio is about 75 percent (Ref. 4). 2.7 Criteria of elastic failure and of rupture For the purpose of this discussion itis convenient to divide metals into two classes: (1) ductile metals,in which marked plastic deformation commences ata fairly definite stress (yield point, yield strength, or possibly elastic limit) and which exhibit considcrabie ultimate elongation; and (2) bile metals, for which the beginning of plastic deformation is not clearly defined and ‘which exhibit litte ultimate elongation. Mild steels typical ofthe ist clas, and cast iron is typical of the second; an ultimate elongation of 5 percent hhas been suggested as the arbitrary dividing line between the two classes of| ‘metals.(Ref. 20), ‘A ductile metal is usually considered to have failed when it has suffered clastic file, ie, when marked plastic deformation has begun. Under simple uniaxial tension this occurs when the stres reaches a value we will denote by @,, which represents the yield strength, yield point, or elastic limitac- ‘cording to which one of these is the most satisfactory indication of elastic failure for the material in.question. ‘The question arises, when does elastic failure occur under other conditions of stress, such as compression, shear, or 4 combination of tension, compression, and shear? ‘The four theories of elastic failure that have received the widest acceptance ‘atwarious times are: (1) the maximum stress theory, which states that elastic24 Formulas for Stress and Strain fenar, 2 failure occurs when the maximum tensile stress becomes equal to 9,5 (2) the ‘maximaum sirain theory, which states that elastic failure occurs when the maxi- mum tensile strain becomes equal to 0,/2; (3) the maximum shear stress theory, which states that clastic failure occurs when the maximum shear stress becomes equal to fay; and (4) the theory of constant energy of distortion, which states that elastic failure occurs when the principal stresses ©,, 02, and 05 satisfy the equation (0, ~ 04) + (02 — a9) + (05 ~ 01)" (Of these four theories, the fourth is the one that agrees best with experimental evidence, but the third leads to results 50 nearly the same and is so much simpler in application that it is much more widely used as a basis for design (Refs. 20 and 21). ‘The criteria just discussed concern the clastic failure of material. Such failure may occur locally in 2 member and may do no real damage if the volume of. material affected is s0 small or so located. as to have only negligible influence on the form and strength of the member as a whole. Whether or not such local overstressing is significant depends upon the properties of the material and the conditions of service. Fatigue properties, resistance to ‘impact, and mechanical functioning are much more likely to be affected than_ staticstrength, and a degree of local overstressing that would constitute failure in a high-speed machine part might be of no consequence whatever in a ridge member. ‘A brittle material cannot be considered to have definitely failed until it has broken, which can occur either through a tensile facture, when the maxi- ‘mum tensile stress reaches the ultimate strength, or through what appears to be a shear fracture, when the maximum compressive stress reaches a certain value. The fracture occurs on a plane oblique to the maximum compressive stress but not, as a rule, oa the plane of maximum shear stress, and so it ‘cannot be considered to be purely a shear failure (see Ref. 21). The results of some tests on glass and Bakelite (Ref. 45) indicate that for these brittle materials either the maximum stress or the maximum strain theory affords ‘a satisfactory critcrion of rupture while neither the maximum shear stress nor the constant energy of distortion theary does. These tests also indicated that strength increases with rate of stress application and that the increase is'more marked when the location of the most stressed zone changes during ioading (pressure of a sphere on a flat surface) than when this zone is fixed (xial tension). Another failure theory that is applicable to brittle materials is the Mohr theory of faite, Brittle materials have compressive strengths greater than their tensile strengths, and therefore both 2 uniaxial tensile test and a uniaxial ‘compressive test must be run to use the Mohr theory. First we draw on a single plot both Mohr’s stress circle for the tensile test at the instant of failure and Mobr's stress circle for the compressive test at the instant of failure; thensaz 20) ‘The Behavior of Bodies Under Stress 25 we complete a failure envelope simply by drawing a pair af tangent lines to the two circles, as shown in Fig: 2.1 Failure under a complex stress situation is expected if the langest of the three Mohr circles for the given situation touches or extends outside the envelope just described. If all normal stresses are tensile, the results coincide with the maximum stress theory. For a condition where the three principal stresses ate 04,09, and 0g, as shown in Fig. 2.1, failure is being approached but will not‘take place unless the dotted circle passing through 0, and op reaches the failure envelope, ‘The acourate prediction of the breaking strength of a member composed ‘of brite metal requires a knowledge of the effect of form and scale, and these effects are expressed by the rupture factor (ee Art. 2:11), In addition, what hhas been said here concerning brittle metals applies also to any essentially isotropic brittle material, ‘Thus far, our discussion of failure has been limited to iiotmpic materials. For wood, which is distinctly anisotropic, the possiblity of failure in each of several ways and directions must be taken into account, viz: (1) by tension parallel to the grain, which causes fracture; (2) by tension transverse to the ‘grain, which causes fracture; (3) by shear parallel to the grain, which causes fracture; (4) by compression parallel to the grain, which causes gradual buckling of the fibers usually accompanied by a shear displacement on an oblique plane; (5) by compression transverse to the grain, which causes sufficient deformation to make the part unfit for service. ‘The unit stres: Producing each of these types of failure must be ascertained by suitable tests (Ret. 4). 28 Fatigue Practically all materials will break under numerous repetitions of a stress that is not as great asthe stress required to produce immediate rupture. ‘This26 Formulas for Stress and Strain foun. 2 phenomenon is known as fatigue and was formerly attributed to a supposed Ehteration in structure brought about by recarrent stressing. Metals, for instance, were said to crystallize, a belief probably arising from the facts that fatigue fractures often occur at points of weakness where the grain structure is abnormally coarse and, even in ductile material, are square and crystalline in appearance because of the absence of gross plastic flow. ‘Over the past 100 years the effects of surface concition, corrosion, temper- ature, etc, on fatigue properties have been well documented, but only in recent years has the microscopic cause of fatigue damage been attributed to gelie plastic flow in the material at the source of a fatigue crack (cack initiation) or at the tip of an existing fatigue crack (erack propagation; Ret. $3). ‘The development of extremely sensitive extensometers has permitted the separation of elastic and plastic strains when testing axially loaded specimens ‘over short gage lengths, With this instrumentation itis possible to determine ‘whether cyclic loading is accompanied by significant cyclic plastie strain and, if itis, whether the cyclic plastic strain continues at the same level, increases, for decreases. Sandor (Ref, 35) discusses this instrumentation and its use in derail Tris not feasible to reproduce here even a small portion of the fatigue data available for various engineering materials. The reader should consult mate- ‘Hals handbooks, tianufacturers literature, and design manuals such as Refs 2% to 29, 47, 49, and 58, Some of the more important factors governing fatigue behavior in general will be outlined in the following material, ‘Number of eyeles to failure. Most data concerning the number of cycles to failure are presented in the form of an S/N curve where the eyelic stress “amplitude is plotted versus the aumber of eyeles to failure. ‘This generally Ieads to a straight line plot on log paper if we account for the scatter in thedata, For the ferrous metals a lower limit exists on stress amplitude called the fatigue limit, ot endurance limit, This generally occurs ata life of from 10° 10 10" cycles of reversed stress and we assume that stresses below this limit will not cause failure regardless of the number of repetitions. With the ability to separate elastic and plastic strains accurately, there are instances when 2 plot of plastic-strain amplitudes versus Nand elastic-srain amplitudes versus N will reveal more useful information (Refs. 90, 34, 35, 71, and 72: and Morrow in Ref. 31, p. 49) Rate of loading. Up to a frequency of 10,000 ¢/min, the rate of stress repetition does not appear to affect the endurance limit of stel; for higher frequencies, the endurance limit increases somewhat, reaching a maximum ‘value at from 1200 to 1800 Hz, beyond which point there is @ decrease (Ref 37), Method of loading and size of specimen. Uniaxial stress can be produced by axial load, bending, or a combination of both. In flat-plate bending only the upper and lower surfaces are subjected to the full range of cyclic stress. In rotating bending all surface layers arc similarly stresed, but in axialane 28) ‘The Behavior of Bodies Under Stress 27 loading the entire cross section is subjected to the same average stress, Since fatigue propertiés of a material depend upon the statistical distribution of defects throughout the specimen, it is apparent that the three methods of loading will produce different results In a similar way the size of 2 bending specimen will affect the fatigue behavior while it will have litle effect on an axially loaded specimen, Several ‘empirical formulas have been proposed to represent the influence of size on 1a machine part or test specimen in bending. For steel, Moore (Ref. 69) suggests the equation (1 — 2.018) = gr(, _ 2.016 v(t 2589) = oe (1 298) where of is the endurance limit for a specimen of diameter d’ and of is the endurance limit for a specimen of diameter d”. This formula was based on test results obtained with specimens from 0.125 to .875 in in diameter and shows good agreement within that sie range, Obviously it cannot be used for predicting the endurance limit of very small specimens Heywood (Ref. 49) suggests an equation equivalent to where o, is the predicted endurance limit for a specimen of diameter d and Op is the experimentally. determined endurance limit for 2 specimen of diameter D > din this reference D = 2 in. This formula also shows reason- ably good agreement with test results up to a diameter of 2in, The few relevant test results available indicate a considerable decrease in endurance limit for very large diameters (Reft. 39 to 43). Stress concentrations. Patigue failures occur at stress levels less than those necessary to produce the gross yielding which would blunt the sharp rise in sess at a sires concentration. Tt is necessary, therefore, to apply the fatigue strengths of a smooth specimen to the peak stresses expected at the stress concentrations unless the size of the stress-concentrating notch or fillet ap proaches the grain size or the size of an anticipated defect in the material itself (sce Factor of sress concentration in ftigue ip Chap. 1 and in Art. 2.10), References 74 and 75 discuss the effect of notches on low-cyce fatigue. Surface conditions. Surface roughness constitutes a kind of stress raiser, and its effect is indicated in Table 37. Discussion of the effect of surface coatings and platings is beyond the scope ofthis book (see Refs, 49 and 67) Corrosion fatigue. Under the simultaneous action of earasion and repeated stres, che fatigue strength of most metals is drastically reduced, sometimes to a small fraction of the strength in air, and 2 true endurance limit can no longer be said to exist. Liquids and gases not ordinarily thought of as ‘specially conducive to corrosion will often have a very deleterious effect on fatigue properties, and resistance to corrosion is more important than normal28 Formulas fr Stress and Strain fenan 2 fatigue strength in determining the relative rating of different metals (Refs, 41, 42, 48, and 53). Range of stress. Stressing a ductile material beyond the elastic limit or yield point in tension will raise the elastic limit for subsequent cycles but lower the elastic limit for compression. For a more detailed description of the Bauschinger effect, see Ref. 2. The consequence of this effect on fatigue is apparent if one accepts the statement that fatigue damage is a result of cyclic plastic flow; ie, if the range of cyclic stress is reduced sufficiently, higher peak stresses can be accepted without suffering continuing damage. ‘Various empirical formulas for the endurance limit corresponding to any sgiven range of stress variation have been suggested, the most generally’ accepted of which is expressed by the Gosdnan diagram or some modification thereof. Figure 2.2 shows one method of constructing this diagram. In each cycle the stress varies from a maximum value Gg, 0 8 minimum value ute 69°04 [t-ton/out] Saotman: o6:041-1on/el] 7 ss . om % Fig 22 either of which is plus or minus according to whether itis tensile or com pressive, The mean sires is (max + Soin) fand the alienating sree is HC.aax — Sata) the addition and subtraction being algebraic. With reference to rectangular x65, 0 is measured horizontally and o, vertically. Obviously when ¢q = 0, the limiting value of, isthe endurance limit for fully reversed stress, denoted, here by ¢,, When o, = 0, the limiting value of a ie the ultimate tensile strength, denoted here by o,. Points A and B on the axes are thus located, According to the Goodman theory, the ordinate to a point on the straight line AB represents the maximum alternating stress 0, that can be imposed in conjunction with the corresponding mean stress ¢y. Any point above AB represents a stres condition that would eventually cause failure; any point below AB represents a stress condition with more or less margin of safety.sa 28) The Behavior of Bodies Under Stress 29 ‘A more conservative construction, suggested by Soderberg (Ref. 20), is to move B back to o,, the yield strength. A less conservative but sometimes preferred construction, proposed by Gerber, is to replace the straight line by the parabola ‘The Goodman diagrams described can be used forsteel and for aluminum and titanium alloys, but for cast iron many test results fall below the straight Jine AB and the lower curved line, suggested by Smith (Ref, 36), is preferred. ‘Test results for magnesium alloys also sometimes fall below the straight line. Figure 2.2 represents conditions where oy. is tensile. If ois compressive, 4, is increased; and for values of 6 less than the compression yield strength, therelationship is represented approximately by the straight line AB extended to the left with the same slope. When the mean stress and alternating stress are both torsional, is practically constant until oy exceeds the yield strength in shear; and for alternating bending combined with mean torsion, the same thing is truc. But when o,, is tensile and o, is torsional, o, diminishes as 1, increases in almost the manner represented by the Goodman line. When stress concentration is to be taken into account, the accepted practice is to apply fy (or & iF kis not known) to 0, only, not t0 2. Residual stress, Since residual stresses, whether deliberately introduced or merely left over from manufacturing processes, will influence the mean stres, their effects can be accounted for, One should be careful, however, not 0 expect the beneficial effects of a residual stress if during the expected life ‘ofa structure it will encounter overloads sufficient to change the residual-stress distribution. Sandor (Ref, 35) discusses this in detail and points out that, an occasional overload might be beneficial in some cases. Combined stress. No one of the theories of failure discussed in Art. 2.7 can be applied to all fatigue loading conditions. ‘The constant energy of distartion ‘theory seems to be conservative in most cases, however. Reference 25 gives 2 detailed description of an acceptable procedure for designing for fatigue under conditions of combined stress. The procedure described also considers the effect of mean stress on the cyclic stress range. Three criteria for failure are discussed: gross yielding, crack initiation, and cack propagation. An extensive discussion of fatigue under combined stress is found in Refs. 38, 47, and 53. ‘Stress history. very important question and one that has been given much attention is the influence of previous stressing on fatigue strength. One theory that has had considerable acceptance is the lizar damage law (Miner in Ref. 47); here the assumption is made that the damage produced by repeated stressing at any level is directly proportional to the number of cycles ‘Thus, if the number of cycles producing failure (100 percent damage) at a stress range 0, is N;, then the proportional damage produced by NV cycles of the stress is N/N, and stressing at various trex levels for various numbers ‘of cycles causes cumulative damage equal to the summation of such fractional values. Failure occurs, therefore, when 2N/N, = 1. The formula implies that the effect of a given number of cycles is the same, whether they are applied