THEORY ». OF ELASTICITY By 8. TIMOSHENKO And J. N. GOODIER Profesvors of Hngincormng Mechanics Stanford Univoreity GEES Bp ‘os: \@ a, Sioeri Maga 001281 ae "A CENTRALE cwneaci srovrmurans fi smi Wb Fe: aes Tao i Bie a Valu! oil? 24. peu Duise naw vo yonk ToRoNvo — _ MeGRAW-HILLL BOOK COMPANY, Ise, 1951THBORY UF ELASTICITY Copyright, 1954, by the United Engineering Trustees, Inc. Copyright, 1951, by the MeCiraw- Till Book Company, Inc. Printed in Uwe Thited Btateo of America, AUl rights rewrved. This book, for parte thereof, may nol. le reprwduced in any form without permission of the publisbers. ‘att rm MAPLE PRESS COMPANY, TORK, PALPREFACE TO THE SECOND EDITION ‘The many developments and clarifications in the theory of elasticity and its applications which have occurred since the first edition was written are reflected in numerous additions and emendations in the present edition, ‘The arrangement of the book remains the same for the most part. . ‘The treatments of the photoelastic method, two-dimensional problems in curvilinear coordinates, and thermal stress have been rewritten and enlarged into separate new chapters which present many methods and solutions not given in the former edition. An appendix on the method of finite differences and its applications, including the relaxation method, has been added. New articles and paragraphs incorporated in the other chapters deal with the theory of the strain gauge rosette, gravity stresses, Saint-Venant’s principle, the components of rotation, the reciprocal theorem, general solutions, the approximate character of the plane stress solutions, center of twist and center of shear, torsional stress concentration at fillets, the approximate treat- ment of slender (¢g., solid airfoil) sections in torsion and bending, and the circular cylinder with a band of pressure. Problems for the student have been added covering the text as far as the end of the chapter on torsion. It isa pleasure to make grateful acknowledment of the many helpful suggestions which have been contributed by readers of the book. 8. TrwosHENKo J, N. Goovrer Pato Auto, Carr, February, 1991PREFACE TO THE FIRST EDITION During recent years the theory of elasticity has found euusiderahle application in the solution of engineering problems. ‘There are many cases in which the elementary methods of strength of materials are juncerate to furnish satisfactory information egurding stress distri- bution in engineering structures, and recourse must be made to the more powerful methods of the theory of elasticity. The elementary theory is insufficient to give information regarding Incal stresses near the loads and ueur the supports af beams. It fails also in the cases when the stress distribution in bodies, all the dimensious of which are of the same order, has to be investigated. ‘I'he stresses in rollors ‘nd in balls of bewrings ean he found only by using the methods of the theory of elasticity. Tho clementary theory gives no means of investigating stresses in regiuus of sharp variation in cross scetion of beams or slufts. It is known that at reentrant corners 2 high slress concentration occurs and as & result of this cracks are likely to start at euch corners, espeviully if the structure is submitted to a reversal of stresses. The majority of fractutes of machine parts in service can be attributed to such cracks. ‘During recent years considerable progress has been made in solving such practically important problems. In cases where 2 rigurous solii~ tion cannot be readily obtained, approximate methods have been developed. In sume cases solutions have boon obtained by using experimental methods. As an example of this lke photoelastic method of solving two-dimensiousl problems of elasticity may bo mentioned. The phatorlastic equipment may be found now xb universities and also in many industrial research Iuboratories. ‘Ihe resulta of photoslastie experiments lave proved especially uocful in studying variuus cases of stress concentration at points of sharp Variation of cross-sectional dimensions and at sharp fillets of reentrant ecurners, Without any doubt these results have considerably influ- enced {he modern design of machine parts and helped in many cuses to improve the construction by eliminating weuk spots from which cracks may start. Another example of the succcssful application of experiments in the golution of clasticity problems is the soup.filn methad for deter mining stresses in torsion and bending of prismatioal bars. The viiviii PREFACE 10 THE FiRST FDITION difficult problems of the solution of partial differential equations with tgiven boundary conditions are replaced in this ease by measurements ff slopes and deflections of a properly stretched and loaded soap film. ‘The experiments show that in this way not only a visual picture of ‘the stress distribution but also the necessary information regarding magnitude of stresses can be obtained with an accuracy sufficient for practical application. ‘Again, the electrical analogy which gives a means of investigating torsional stresses in shafts of variable diameter at the fillets and grooves is interesting, ‘The analogy between the problem of bending Of plates and the two-dimensional problem of elasticity has also been successfully applied in the solution of important engineering problems. Tn the preparation of this book the intention was to give to engi- neers, in a simple form, the necessary fundamental knowledge of the theory of elasticity. Tt was also intended to bring together solutions of special problems which may be of practical importance and to Geseribe approximate and experimental methods of the solution of elasticity problems. Having in mind practical applications of the theory of elasticity, matters of more theoretical interest and those which havo not at present any direct applications in engineering have been omitted in favor of the discussion of specific cases. Only by studying such cases ‘with all the details and by comparing the results of exact investigations With the approximate solutions usually given in the elementary books on strength of materials can a designer acquire a thorough under~ standing of stress distribution in engineering structures, and learn to tise, to his advantage, the more rigorous methods of stress analysis. Ta the discussion of special problems in most cases the method of direct determination of stresses and the use of the compatibility equations in terms of stress components has been applied. | This method is more familiar to engineers who aro usually interested in the magnitude of stresses. By a suitable introduction of stress functions thir method is also often simpler than that in which equations of equilibrium in terms of displacements are used. ‘In many cases the energy method of solution of elasticity problems bas been used. In this way the integration of differential equations is replaced by the investigation of minimum conditions of certain inte- grals. Using Ritz’s method this problem of variational caleulus is Reduced to a simple problem of finding a minimum of a function. Tn this manner useful approximate solutions can be obtained in many practically important cases,PREFACE TO THE FIRST EDITION ix To simplify the presentation, the book begins with the discussion of two-dimensional problome and only later, when the reuder has familiar ined himself with Ube vuriuus methods ased in the solution of problems of the (heory of elasticity, are thrcc-dimensional problems discussed. ‘The portions of the book that, although uf practical importance, are such that they cun be umilted curing the first reading are put in emall type. The reader may retun to the study of euch problems after finishing with the most essential portions of the Louk. ‘The mathematical derivations are put in an elementary form and usually du act recinire more mathematical knowledge than is given in eugineering schools, In the cases of more complicated problems all necessary explanations aud intermediate calonlations are given 90 that the reader can follow without difficulty through all the deriva- uns. Only in a fow easoe aro final results given without eunplete derivations, ‘Then the uecessury references to the papers in which the derivations cun be fond are always given. Tn mimeraus footnotes reforonoes to papers and buoks on (he heory of elasticity which may be of pructival importance are given. Thess references muy be of interest to engineers who wish to study some special problems in more detail. They give also a picture of the modern devclopmont of the theory of eluslicity and may he of some ‘use to graduate sludents who are planning to take their worl: in this field. In the preparation of the book the contents of previous book (“Theory of Elasticity,” vol. J, St. Petersburg, Hussia, 1914) on Uke same subject, which represented a couruc of lectures on the theory of elasticity given in several Russian engineering schovls, were used to a large extent, The author was avsisled in his wrk hy Dr. L. H. Donnell and Dr. J... Gandier, who read over the complete manuscript and tu whom ho ia indebted for many corrections and suyyestious. ‘The anthor tukes this opportunity lo thunk also l'rof. (i. H. MacCullough, Dr. E. f. Weibel, Prof. M. Sadowsky, and Mr. D. H. Young, who assisted in the final preparation of the book by roading sume porlivus of the uanuscript. He ie indebted also to Mr. L. 8. Veenstra for the prep- aration of drawings und to Mrs. J. 1). Webster for the typing of the Manuscript. 8. TorossenKka Uxivenstrr or Micmaan December, 1989CONTENTS pnevace v1 umm SECOND EDIOM. 66 fey Ppopsox wo wr, Pinar Komnos. + vit ‘Noramion. Lee xvii Cuarren 1, INTRODUCTION Elnaticity « Stress . | Notation for Boren and treooca + ‘Components of Stress Components of Strain. |. Tooke's Lav. . Problems oc ee viene dl 1 2 8, 4. Onarmen 2, PLANE STRESS AND PLANE STRAIN 7 Phu Stress te u 8, Plane Sirnin 2. bees tae 2 y, Strossata Pink 33 10, Strain at a Point . W 11, Measurement of Surface Sirius 0 et 19 12. Constrnction of Mohr Strain Circle for Strain Hosctte, . a 1, Different Tquations of Baquiibrium bee OE 14, Boundary Conditions... . te 22 16, Compatibility Equations. ©... 6 6 ss : 1 16, Stress Function... - see 26 Problema.) .. om Cuapren 3, TWO-DIMENSIONAL PROBLEMS IN RUULANGULAR COORDINATES 17. Solution by Polynomials... . - bee cy 18, Saint-Vennnt’s Principle... . + + be 33 |. Determination of Displacements... . . 34 1W, Bending of a Cantilever Tanded ab the End . - 36 21. Bending of a Boam by Uniform Toad . . a9 22. Other Cases of Continuously Loaded Rea “4 23. Solulivn of Uke ‘Two-dimensional Problem Series. - 16 24, Other Applications uf Rou cy Probleme... - 53 INSTITU Tit POLITEHN. | TIMISOARA Fant 1+ tan 7 tans we find Ait re, 20+ rE B “ ‘Thus tho relation between shearing siruin aud shearing stress is defincd by the couslauls J and ». Often the notation : E TTR ® is used. Then Eq. (4) becomes =f VG ‘Vhe constant G, defined by (5), is called Uke modulus of elasticity in shear or the modulus of riyidily Tf shearing stresses act on the sides of an clement, as shown in Fig. 3, the distortion of the angle between any two coordinate axes depends only on shearing-stress components parallel lo these axes and we obtain aaedm edn mide The elongutious (3) and the distortions (6) are independent of cach other, Hence the general case of strain, produced by three normal and three shearing compononts of stress, can be obtained Ly superposition on the three elongations given by Es. (8) are superposed three shear- ing strains yiven by Eys. (6). Fiqnations (4) and (6) give the camponcnts of strain as functions of the components of stregs. Sometimes the components of slress expressed as functions of the components of slrain are needed. ‘These cau be obtained us fulluws, Adding equations (3) together and using the notations ese tbate Onratate m We obtain the folluwing relation hetween the volume expansion ¢ and the sum of normal stresses: en ize ®)10 PHBURY OF ELASTICITY In the case of @ uniform hydrostatie pressure of the amount p we have vou NP and a, (8) gives _ 8 -2)p . oF which reprevenis the relation betwee unit, volume expansion ¢ nd Iiydrostatio pressure p. The quanlily £/( — 2») is exlled the modulus of volume expansion. Lising notations (7) and solving Eqs. (8) fru, va, on, we find vE RE G4 idem TT Hes vE B . wn aes Tite - weg Ene nr a ae or using the notation ti = 7FHna=% (0) and Fry. (5), these become os — M+ Whey ay — de + 2G ay) 4 De +t Dies Protiems 4 Show thal gs (1) continue t Tul ifthe clement of Fig, 4 is in motion and hha an nngular acceleration lke 2 rigid body. 4 Sapmose an clastic mulrrinl contains a lange minutes of evenly distributed wa ee partir, no thak a magnatic Ald ents on any clement ds dy Os ee are dy dz nnna an ax parallel Uwe axis, What moliGowon will be needed in Eqs (1)? Ging somervasons why the formule (2) will be vaid for small nicains omy Fee rele ayer is unndwichelhehwcen two porfeety rir yates, to which 1 ie hombal. The layer 18 eomprserl Lxtween tho plate, Une comroroseive, SA Hie” gappeaing thas the attachment to the piss prevento late, strain aa svrpleely rd the mpsuent Young's mods Gs yu/a) an cars of Rand». Fearne raany tines 1 ifthe material of tie layer is nearly neon yevsible BY hydrostatic presnure oe prove ta: Fa, (8) follows from Fas. (12), (10), and (8)CHAPTER 2 PLANE STRESS AND PLANE STRAIN 7. Plane Stress. If a thin plate is loaded by forces applied at the Luundary, parallel to the plane uf the plale and distributed unifurmly over the thickness (Fig. 8), the stress components 7, Tas, Ty: ALE ZeTO On both faces of the plate, and it may be assumed, tentatively, that they ure zero also within Ube pluie, The state of stress is then specified by vay yy Tay ONLY, and is called plane stress. It may alsn he assumed that y Y Fra. 8. these three components are independent of 2, i.e, they do not vary (urough the thickness, They are thon funotions of z and y only. 8. Plane Strain. A si:ilur simplification is possible at the other extreme when the dimension of the hody in the z-direction is very Iurge. Téa long cylindrical or prismatical body is loaded by forces which are perpendicular to the longitudinal cloments and do not vary along the length, it may be ussumed Uhul ull cross sevtions are in the same condi- tion, Tt is simplest to suppose at first tha’ the end sectiuns are euu- {ined between fixed smooth rigid planes, so that displacement. in the axial direction is prevented. The effect of removing these will be examined later. Since Usere ix uy axial displacement at the ends, and, by symmetry, at; the mid-sention, it may he assumed that the same holds at every eross section. ‘There are muny important problome of this kind—a rotaining wall with lateral pressre (Wig. 9), a culvert ur Guumel (By, 10), w eylindsiewl tube with internal pressure, a cylindrical roller compressed by forces in n TITUTUL POUITEHN. TiMISQAKRA © BIBLIOTECA CCNTRALA13 THEORY OF ELASTICITY ‘a diametral plane as in a roller pearing (Fig. 11). In each case of conrse tho loading must nul vary along the length, Since condilions are the sume at, all crogs scetions, it is sufficient Lo consider only a slice hetwecn two sections uuil distance apart. The components u and vol the displacement arc functions of «and y Lut are independent of tho Fan He Jungitnlinal coordinate . Sinee the lougiludinal displacement w is zero, Eqs. (2) give av, aw met oy ou tw te = Eten (a) -@e a“ Gre ‘The Jongilncinal normal otreee ¢, ean bu found in terms of ge and ow by means of Hooke’s law, Eys. (8). Since ¢, = 0 wo find or — ou %) =O or os — ¥(04 + oy) @) "These normal slrexsex act over the cross sections, inclu the endo, here they represent forces required to maintain the plane strain, und provided by the: fixed smooth rigid planes.PLAND STRESS AND PLANE STRAIN 18 By Eqs. (a) and (6), the stress components r-, and ry, are cro, and, by Eq. (b), ¢, can be found from¢, and¢,. Thus the plane strain prob- lem, like the plane stress problem, reduces to the determination of v2 oy, ond ry 08 functions of z and y only. 9. Stress al a Puinl. Kuuwing the sirevs components 6,, 6), tay ab any point of a plate in a condition of plane stress or plane strain, the strese acting on any plane through this point perpendicular to the plate and inclined bo the x- aud yeuxes cau be calculated from the equations of statics. Let 0 be a point of the stressed plate and suppose the stress components ¢,, oy, Ty aro known (Pig. 12). To find the stress for any plane throngh the z-axis and inclined to tho @ and y-axes, wo take a plane BE parallel to it, wba sual distance from Q, so that this latter plane together with the coordinate planes euis out from the plate a very small triangular prism OBC. Since the stresses vary continuously over the volume of the hody the stress acting on the plane BC will approach the stress on the parallel plane through O as the element is made smaller. In diseussing the conclitions of equilibrium of the small triangular priam, the body force can be neglected aa a small quantity of a higher order (page 4). Likewise, if the element is very small, we can neglect the variation of the siresses over Uhe sides wud assume thut the sbresses are uniformly distributed. The forces acting on the triangular prism can therefore be determined by multiplying the stress components by the areus of the sides. Let N be the direction of he normal to the plane BU, and denote the cosines af the angles between the normal V and the axes # and y by Fie. 12. cosNz=1, cosNy=m ‘Chen, if A denotes the area of the side BC of the element, the areas of the ather two sides are AL and Am If we denote by ¥ and ¥ the components of stress acting on the side XC, the equations of equilibrium of the prismatical element give K — lug tomy ¥ = moy + tem ue Thus the components of stress on any plane defined by the directionCy YHBORY OF PLASTICITY casines band m can wanily be calculated from Fas. (13), provided the fhrve romponents of stress ¢., 6, Tay ab the point O ure known. Totting @ bi the angle between the sorvsal A and the a-axis, su thn ) = cosq and m = sin a, (he normal and shearing eamponenta of stress on the plane BU are (from Eys. 12) a — Keova + ¥ sina ~ uz cost « +o, sin? oe 2ray sin ey eon pm Fcos a — FX sin x = rey(oo8® a — sis? a) + (ey — 22) sin COB @ 3) Lémay be seen Uni the angle @ ean be eluasen in such a manner “hut the shenring streas r becomes eeynal to zero. For this vase. we have ta(eus! wc sin? a) + (@, — a) sin a cos. ~ 0 or . 1 sino cos 4 ] ta ate mare Ta (id) From this equation two perpendicular dirootions ean he found for ‘which the sheuring stress ig necro, These directions are called principal direstions and tho corresponding normal stresses principal strevee. Tf the principal directions are Luken as the = and yeaxes, try 18 2er0 and Liga, (J3) are simplified to re COS? ce + ny Sin & 8) ain Qa(oy — 64) (3) ‘Tho variation of the stress components «and r, 08 We vary the angle 1, can be casily represented graphically by suaking,a diagram in which oo pane taken ap coordinates. or each plane there will correspond ‘x point on this diagram, the coordinates uf whieh represent the valnes of ¢ sud r for this plane, Figure 11 represents such a diagram. For the planos perpewlienlar to tho principal directions wo obtain points A and R with abseiseas o, and oy, respectively. ‘Now it can be proved that the stress componente for any plsur 20 with on angle « (Tig. 12) will he represented by cour dinates of a point on (he circle having AB us adiameter. To find this point it is unly necessary TO mews from the ‘point Aim the same direchion a3 & jo mexsured in Fig, 12 an are sub tending an angle equal to 2a. TI /) is the point oblwined in this man- ner, then, from the Ggnre, ow obey em OF = 001 COR = Sato DF = CN sin 2a = ¥(¢, — 2) sin Bot “Thin prapical mothod Ss die us O» Mohr, Ziitngemicer, 1069, p. 113. Son) ky hig *"Teshmieche Merhnik,” 24 eda, 1014, { 4 g93 2a = a, cost a + 0, sin?PLANT STRESS AND PLANE STRAIN 15 Comparing with Iiqs. (13') it is econ that the coordinules of point: 12 sive the numorieal values of stress components on the plano BC at te fngle a. To buing into coincidence the sign of the shearing component wvo take + positive in the upward direction (Fig. 18) and consider shear Jy stresses as positive when Uhey give » comple in the clockwise direc- tion, as on Une sides he and ad of the clement abed (Fig. 136), Shearing tresses of opposite direction, as on the sides wh and de of the element, ure considered ay negulive.? ‘Ag the plane HC rotates about an axis perpendicular Lo the cy-plane (Fiz. 12) in the clockwise directiuu, snd varies from U to 7/2, the i Fra. 18, point D in Fig. 13 moves from A tu fi, so that the lower half circle determines Une sitess variation for all valuco of « within these limits, ‘The upper half of the circle gives stresses for #/2 < a < 7. Prolonging the radius CD to the point #; (I'ig. 18), t,, tating the uugle x + 2c, inslead of 2a, the stresses on the plane perpendicular ty BU (Wig. 12) are obtained. This shows that the shearing stresses on two perpendicular planes are numerically equal as previously proved. As for normal stresses, we see from the figure that OF, + OF = 20C, the sum of the normal stresses over two porpendiculur eruss ser= tions remains constant when the angle a changes. ‘The maximum shuuriug stress is given in the diagram (Fig. 13) by the maximum ordinate. at the e1rele, é.c., is equal to the radius of (he circle. Henec fom. ~ a7 (in) Tt acts ou Une plaue for which @ = +/4, i¢., on the plane biseoting the anglo between the two principal stresses. “This rule is used only in Um comstruction of Mobr’a circle, thorwiso tho File given on p. 3 holds.16 TREORY OF ELASTICITY ‘The diagram can be used also in the case when one or both principal stresses are negative (compression). It is only necessary to change the sign of the abscissa for compressive stress. In this manner Fig. la represents the case when both principal stresses are negative and Fig. 14) the ease of pure shear. (a) @) Fi. 14, From Figs. 13 and 14 itis seen that the stress at a point ean be resolved into two ports: One, uniform tension or compression, the magnitude of which ie given by the Prvoiaaa of the center of the circle; and the other, pure shear, the magnitude of siiueh is given by the radius of the circle. When several plane stress distributions veessuperposed, the uniform tensions or compressions can be added together rg ra ra algebraically. The pure shears must be added together by taking into account ‘he directions of the planes on which they are acting. Tt ean be shown that, if we superpose two systems of pure shear whose planes of maximum shear make 9» anele of 2 with each other, the resulting system will be another ease of pure shear; Fea caaple, Fig. 15 represents the determination of stress on any plane defined by a, produced by two pure shears of magnitude +, and rz acting one on the planesPLANE STRESS AND PLANE STRAIN Ww 23 au] ya (Fig. 15a) and the other on the planes inclined to xz and ye by the anele 0 (Fig. 150). In Fig, 16a the coordinates of point D represent the shear and normal stress un plane CB produced by the first system, while the coordinate of D: (Hig. ‘ob) gives the stresses on this plane for the second aystem, Adding UD and UD; geomelrically we obtain OG, the resultant stress on the plane due to both systems, ‘tho coordinates of Q giving us the shear and normal stress. Note that the magni- tude of OG does not depend upon «. Hence, as the result of the superposition of two sheurs, we obtain a Mohr circle for pure shear, the magnitude of which is sgivon by OG, the planes of maximum shear being inclined to the xz and ye planes by fan angle e«jual half the angle GOD. A diagram, euch as shown in Fig. 13, can bo used also for determining principal siressey if Ube sbress compo nenlsng, oy, ty for any two perpendicular planes (Fig. 12) are known, We begin In such u case with the plotting af the two points D and Dy, representing stress con- ditions on the two coordinate planes (Fig. 16). In this manner the diameter DD; of thecireleis obtained. Construct- ing the circle, the prinvipal stresseso, said oy ure obtained from the intersection of the cirele with the abscissa axis, From the figure we find Oi Frag, 16, (18) =0C + op a ete VES) tee F Ya te a AS The maximum shearing stress is given by the radius of the circle, 40, a = OC —CD= a -4 mom) = Ve an Yb ret a7) Tn this manner all necessary features of the stress distribution at a point ean be obtained if only the three stross components gx Oy; Tay Te THEW 10, Strain at a Point. When the strain components ex, ¢, You At & point are known, the unit clongation for any direction, and the decrease of a right angle—ihe shearing strain—of any orientation al he point can be found. A line element 1) (Wig. 17a) between the points (x,y), (© + da, u + dy) io translated, stretched (or contracted) and rotated into the line element P'Q’ when Uhe deformation veeurs. The dix18 YHEORY OF BLASYICITY placement components of # are #, ave those of Q are du yy 4 oH 0 ® ut aot dy, 94 ae + 5 it P'Q’ in Fig. 17a io now trunsinted so that P’ is brought back to P, his in the position PY of Fig. 17%, und GX, RQ” represent the come ponents of the diplacement of @ relative lo P. "Thus uy, 4 oe 8 ag QR = Feet ay RQ = ale t ay (@) ‘he components af this relative displacement QS, $Q", normal to PQ" and along PQ", cum he found from these as QS = —QR sin @+ RQ" 008 6, AQ" = QR coe 6+ RQ" sin # (6) ignoring, the small angle QPS in comparison with 0. Since the short. Ime QS may Ve identified with un are of o circle will canter P, SQ” xy) Coe Pmt a lay Re lerazyedy) fp a 2 ‘q” ta) a’ « nt ives the otrotch of PQ. ‘The unit elonyetinn of PQ", denoted hy en it SQ’/PQ. Using (b) and («) we have ‘aude, Sulu) y gig g (onde , av o( Sey 4 ame cede | ay dey ds" oy as, = 2 cost au 4 2) ai sat = cet Bt 5 2) an 0 e080 + Fh 0 or es = cg 008 OE Yay Sin # 008 0 + 6 sin? O © svhich giver the unit elongation for uny direction 0, ‘The angle ve through which PQ in rotated bs QS/PQ. Thus from (1) and (a), . ofdude ) ou wy ‘ay die, ou dy vno(’ B+ ia) + 28 ON Ge ae | ayds or cage ‘av au OM gine sews o+(m- 2) sin ot 0 ayaa @PLANE STRESS AND PLANE STRALN 19 ‘The lino element PT at right angles to PQ makes an angle # | (x/2) with the z-direction, and its rotation vo47 is therefore given by (d) when + (x/2) is substituted for 9, Since cos [6+ (x/2)] = — sin @, ain [0+ (#/2)] — cos 6, we find mM , ‘ae au 3th cog? Port - Zante (& 2) in 9 coc @ wore ‘The shear strain 7» for the directions PQ, PT is Yo — Vout, sv ne BB) (oat 6 ~ sink 8) + & 20) 9 sin 0 coo 8 dye = Hm (vos? @ — win? @) + (ey — #2) sin 0 cos 0 fy Comparing (c) and (f) with (18), we observe that they may be obtained from (18) by replacing » by ey, + by ye/2, a by er, ay BY ty, Tey DY Yex/2y aud a by 0. Consequently for cach deduction made from (18) ws bu wand r, there is 4 corresponding deduction frum (e) and (f) as to eg and "yo/2. Thus there are lwo values of 0, differing by 90 deg., for which is vero. ‘They are given by v7 . — tan 20 @-4& The corresponding strains «9 are principal siruins, A Mohr circle diagram anulogous Lo Fig. 18 ur I'ig. 16 may be drawn, the ordinates representing -y4/'2 and the abscissas ¢. The prineipal etraing ¢, ¢2 will be the algebraically groatcst and loast values of ¢¢ ax x funclion of 6. ‘The greatest value of y¢/2 will be represented by the radins of the circle, Thus the greatest shearing strain 7» as. is given by Yemen 11, Measurement of Surface Strains. ‘I'he strains, or unit clonga- tions, an a. surface are usually most conveniently measured by means of electrie-resistance strain gauges. The simplest form of such # gauge isa short length of wire insulated from and ylued tv the surface. When Stretching vecus the resistance ul Lhe wire is increased, and the strain ean thus be measured electrically. ‘The effect is usually magnified by louping the wires backward and forward several times, to form several ange leuylls connected in series, The wire is glued hetween two tabs of paper, and the assembly gined to the surface. ‘The usc of these gauges is simple when the principal directions are nn in the “Llandbook of Lixperimental20 THEORY OF ELASTICITY known. One gauge is placed along each principal direction and direct measurements of ¢, «2 obtained. ‘The principal stresses o1, oe may then be calculated from Hooke’s law, Eqs. (), with os = oy oy = %% ¢, = 0, the last holding on the assumption that there is no stress acting Ga the surface to which the gauges are attached, ‘Then (= or = Eat ve), (L- or = Ble + ves) ‘When the principal directions are not known in advance, three meas- urements are needed, Thus the state of strain is completely deter- mined if ¢,, ¢ ra can be measured. But since the strain gauges meas- « = a “e i ae a a ps ey (a) fe) Fis. 18, ure extensions, and not shearing strain directly, it is convenient to measure the unit elongations in three directions at the point. Such a set of gauges is called a “strain rosette.” The Mobr circle can be drawn by the simple construetion* given in Art, 12, and the prineipal ‘Strains can then be read off. ‘The three gauges are represented by the ‘three full lines in Fig. 18a. ‘The broken line represents the (unknown) direction of the larger principal strain «, from which the direction of the first gauge is obtained by a clockwise rotation 4. ‘If the 2- and y-directions for Eqs. (¢) and (f) of Art. 10 had been taken as the principal directions, ez would be ¢:, ¢ would be ex, and 1 would be zero. ‘The equations would then be ee = e.cost 0+ eqaint 6, dys = — (4 — &) sin 0 cos 0 where 0 is the angle measured from the direetio of a. ‘These may be written Ha +e) + Ha — @) 008 28, ve = —He — «) sin 20 and these values are represented by the point P on the circle in Fig. 18. If @ takes the value $, P corresponds to the point A on the circle in Fig. Glenn Murphy, J. Applied Mechanice (Trans, A,S.M.E.), vol. 12, p. 8-209, 1945; N. J. Hoff, sbid,PLANE STRESS AND PLANE STRAIN 21 18d, the angular displacement from Ube e-axiy being 26. The ubseivsa of (his paint isey, which is known, If @ takes the value ih + ar, / moves. to B, through a further angle AFB = 2a, and the abscissa is the known value case. If 6 takes the value ¢ + @ + 8, P moves on to C, through afurther angle BFC = 28, yud Une abeciase is cassie 'The problem is to draw the circle when these three abscissas and the two angles a, 6 arc known. 12. Construction of Mohr Strain Circle for Strain Rosette. A tem- porary horizontal euxis is drawn horizontully from any origin 0", Fig. 18), and the three measnred strains e4, ee) 4 taiei¢ Iniel off along it Verticals are drawn through these points. Selecting any point D on the vertical through case, lines DA, DC are drawn at angles « and 6 to the vertical at D as shown, io meet the other two verticals at 4 and C, The circle drawn through D, A, and (is the required circle, Its center W is determined by the intersection of the perpendicular bisectors of CD, DA, The points representing the throe gauge directions arc A, B,andC, The angle AFB, being twice the angle ADB at the cireum- fereuce, is 2a, und BFC iv 28. Thus A, H, Care at the required ungulur intervals round the cirale, and have the required abscissas. ‘The ee axis can now be drawn as OF, x and the distances from 0 to the intersections with ihe eiree give nea. ‘The angle 2, is the angle of FA below this axis. y 13. Differential Equations of Equilibrium. We now consider “*y/ the equilibrium of usuull recluu- (Jp gular black of edges 4, &, and unity (Fig. 19). The stresses ‘acting on the faces 1, 2, 3, 4, and their positive directions are in- dicated in the figure. On ae= count of the variation of stress throughout the material, the value of, for instance, , is nok quite the sume for faco 1 as for face 3, The symbols a», oy, tw refer to the Point x, y, (ae mid-point of the rectangle in Fig. 19. The values at the Mid-points of Ue fwes are deuoled by (6,)s, (¢,)o, ete. Since the faves are very small, the corresponding forces are obtained by multiplying these values by the arcas of the faces on which they act." * Mure precise considerations would introduce terms of higher order which ‘Vanish in the final limiting proocas,22 THEORY OF ELASTICITY ‘The body force on the block, which was neglected as a small quantity of higher order in considering the equilibrium of the triangular prism of Fig. 12, must be taken into consideration, because it is of the same order of magnitude as the terms due to the variations of the stress components which are now under consideration. If X, Y¥ denote the components of body force per unit volume, the equation of equilibrium for forces in the x-direction is (a)sk ~ (a)ak + (rey)ah — (roy)sh + Xhk = 0 or, dividing by Ak, a) = (ox)a 4 (mda ~ Cot ye a at Sa et x m0 If now the block is taken smaller and smaller, i.c., h +0, k 0, the limit of [(¢2)1 — (¢2)s]/h is 60,/A2 by the definition of such a derivative. Similarly [(ray)a — (r2y)a]/k becomes dt/dy. The equation of equi- librium for forces in the y-direction is obtained in the same manner. Thus oo, ore yy Bt getx= (18) Oey , Otey ay baz FY =O In practical applications the weight of the body is usually the only body force. Then, taking the y-axis downward and denoting by p the mass per unit volume of the body, Eqs. (18) become (19) ‘These are the differential equations of equilibrium for two-dimensional problems. 14. Boundary Conditions. Equations (18) or (19) must be satisfied at all points throughout the volume of the body. The stress compo- nents vary over the volume of the plate, and when we arrive at the boundary they must be such as to be in equilibrium with the external forces on the boundary of the plate, so that external forces may be regarded as a continuation of the internal stress distribution. ‘These conditions of equilibrium at the boundary can be obtained from Eqs. (12), Taking the small triangular prism OBC (Fig. 12), so that the side BC coincides with the boundary of the plate, as shown in Fig. 20,PLANE STRESS AND PLANE STRAIN 23 and denoting by X and ¥ the components of the surface forces per unit ares at this point of the houndary, we have E = lee ry ¥ = mo, + lew in which / and m are the direction cosines of the normal NV to the boundary. In the particular case of a rectangular plate the coordinate axes are usually taken purullel Lo the sides of (he plule aud the boundary eondi- tivus (20) ean be simplified. ‘aking, for instance, a side of the plate parallel to the 2-axis we have for this part of the boundary the normal N parallel to the y-axis; hence | ~ 0 and (20) m= +1. Equatious (20) then become > ~ R= tm, F= toy 2 Ilere the positive sign shonld be taken if the | NS normal N has the positive dircotion of the Y y-axis and the negative sign for the opposite ¥ Fra. 20, direction uf NV. Ibis seen from this that at the benndary the stress components hecome equal to the components of the surface forees per unit area of the boundary. 15, Compatibility Equations. The problem of the theory of élas- ticity usually is to determine Lhe slate of slress in a body submilted lo the action of given forces. In the case of a two-dimensional problem it is necessary to solve the differential equations of equilibrium (18), and the solution must be such as to satisfy the boundary conditions (20). These equations, derived by application of the equations of statics fur absolutely rigid budies, and containing three stress compu- nents og, oy, Tey, Are not sutticient for the determination of these compo- nents. The problem is a statically indeterminate one, and in order to obtain the solution tho elastic deformation of the body must also be considered. ‘The mathewulival formulation of the eondition for compatibility of stress distribution with the existence of continuous functions u, v, 1 defining tho deformation will be obtained from Eqs. (2). In the ease of two-dimensional problems only three strain components need be considered, namely, a or a ae ay ay an ae en (@) These three strain components are expressed by two functions u and v; honec they cannot be taken arbitrarily, and there exists a certain rela-24 THEORY OF ELASTICITY tion between the strain components which can easily be obtained from (@. Differentiating the first of the Eqs. (a) twice with respect to v, the second twice with respect to 2, and the third once with respect to z and once with respect to y, we find He, Oy _ He Gy t Gel aay ep This differential relation, called the condition of compatibility, must be satisfied by the strain components to secure the existence of functions wand v connected with the strain components by Eqs. (a). By using Hooke’s law, [Eqs. (3)], the condition (21) can be transformed into a relation between the components of stress, Tn the ease of plane stress distribution (Art, 7), Eqs. (8) reduce to epee), val) 22) ya = ry = OEY, (23) Substituting in Eq, (21), we find a # Prey Kem + Ze-m=04Fe & ‘This equation can be written in a different form by using the equations of equilibrium. For the case when the weight of the body is the only body force, differentiating the first of Eqs. (19) with respect to z and the second with respect to y and adding them, we find Substituting in Eq. (6), the compatibility equation in terms of stress components becomes a (é + 5) (+0) =0 ( Proceeding in the same manner with the general equations of equilib rium (18) we find x, aY > ont a) (25) B+ Bera -atrn( In the case of plane strain (Art. 8), we have os = Woe + 0)PLANE STRESS AND PLANE STRAIN 25 and from Hooke’s law (Eqs. 3), we find fem pd Me HLL edd 1 (28) ey = Bl — vay — 1 + va] ta" aa i “ Tey (37) Substituting in Eq. (21), and using, as before, the equations of equilib- rium (19), we find that the compatibility equation (24) holds also for plane strain, For the gencral case of body forecs we obtain from Eqs. (21) and (18) the eomputibility equation in the following form: a 1 ex , a¥ (% + yt @iae-,1,(% r) (28) ‘The equations of equilibrium (18) or (19) together with the boundary conditions (20) and one of the ahave compatibility equations give us a system of equations which is usually oufficient for the complete deter- mination of the stress distribution in a two-dimensional problem,t ‘The partionlar cases in which certain additional considerations ure nooessary will be discussed latcr (page 117). It ia interesting tio note that in Uke case of constant body forces the equations determining stress distribution do not contain the elastic constants of the smalterial. Hence the stress distribution is the same for all isotropic materials, pro- vidled Lhe equations are sufficient for the complete determination of the stresses. ‘The conclusion is of practical imporlance: we shull yee laler that in the case of transparcnt materials, such as glass ar xylonite, it is possible ty determine stresses by an optical method using polarized light (page 131). rom the above discussion il is evident that experi- mental reaulte obtaincd with a transparent material in most. cases can he applied immediately to any other material, such os atecl. Tt shomld be noted alsy Ural in Ue ease of coustunt body forees the compatibility equation (24) holds both for the casa of plane stress aud for the cuse of plane strain. Heneo the stresa distribution is the same in these two eases, provided the shupe of Lhe boundary and the external forces are the same.* ¥In plane otraas there arc compatibility avulilivus ublwer fact violatod by our assumptions. It is shown in method of the present chapter gives good ap * This etatcment may require modification wlwn the plate or eylinder has holes, for thon the problem can be correcily solved nly ly eonsidering the displace Monts as well aa the stresses. Bee Aut. 99, xr in thi he26 THEORY OF ELASTICITY 16. Stress Function, It has been shown that # solution of two- dimensional problems rednees to the integration of the differential equations of equilibrium together with the compatibility equation and the boundary conditions. I we begi + Ube cuse when the weight ‘of the body is the only body force, the equations to be satialicd arc (sec igs. 19 and 24) an 1 ory a on * (@) pt ge log <0 ay * oe a ot (2+ SB) ete -0 ® ‘To these equations the houndary eunditions (20) should be added. The usual method of solving those equations is by introducing a new function, called the stress funriion.t Ax is eusily checked, Eqs. (a) are satisfied by taking any function ¢ of z and y and putting the following expressions for the stress companents _ oe -™ — 8%. oe By OM ays Oh tw = BG 89) In this manner we ean get a variety of solutions of tho equations of equilibrium (a), The truc solution of the problem is that which satis fies also the eumputibility equation (b). Substituting expressions (29) for the stress components into Eq. (6) we tind thal. the stress funeLion wb monst satisly Une equation a6 4 ate at Frag + 3 . =0 (au) ‘Thns the sulution of « two-dimensional problem, when the weight at the body is the only body force, rerhiees lu Gnding w sulution of Eq. (30) which satisfies Lhe boundary conditions (20) of the problem. In the following chapters this meth! uf sulution will Le applied to several examples of pruclieal interest. Let us now consider a more general case af Luly farsi sul wave that these oreee have a potential. ‘then the components X and ¥ in Rip. (18) ure given ‘by the oquations +'This funetion was introduced in tho solution of two dimensional problems by G.D. Airy, Brit. Aswe. Advancement Sei, Rept, 1862, and io oometimes called the Airy stress functin.PLANE STRESS AND PLANE STRAIN at bcs in which ¥ is the potential function. Equations (18) becomo a Bray Fe tao soy thao -yi%, gape 2s 7p te oH VT VHS rey hy (aly in which ¢ is the stress function. Gubstituting expressions (31) in the compati- bility equation (25) for plane stress distribution, we find where (Oe An analogous equation can be obtained for the case of plano strain. ‘When the body forec is simply the weight, the potential Y is —egy. In this case the right-hand side of Liq. (84) reduces to zero. Ly taking the solution ¥ — 0 of (32), or of (80), we find the stress distribution from (31), or (29), ay 32 ant (a) vs 00th oy = wo, yw 0 (a) ‘a5 a possible tate of stress duc to gravity, ‘This is a state of hydrostatic pressure qv in two dimensions, with zero stress at y = 0. It can exist in a plate or cylinder of any shape provided the corresponding boundary forces arr ayplivl. Chruniier- ing a boundary element as in Fy. 12, Pys. (18) show (Tonk Uhere must be x normal pressure pyy om the boundary, and sero shear stress, If the plate or cylinder is to Iw supported in some other manner we have to superpose 2 boundary normal tension pgy and the new supporting foress, The two together will bo in equilibrium, und the determination of their effects ie a problem of boundary forecs only, without body foroee.t Problems 4. Show Ghat Eqs. (12) remain valid when the element of Fig, 12 has accolerntion. 2. Find graphleally the principal strains and their dircctions trom rosette ‘wesurements BK cng LB KU, ring = O95 TOPE Hn yr far where « = 6 = 45°. “Lhis problem, and the yrumeul eww of w potential V uch that the right-hand Bide of Liq. (92) vanish, Inve heen discussed by M. Biot, J. Applied Mechantes (Lrans, A.8.M-B.), 1985, ps AnsTHEORY UF ELASTICITY 3. Show that the Hino elements at the paint 1, y which have the maximum and sninimoum, rotation wre {howe in Uh Le ges peauicular discos @ determined by oe ay fam aw tense (5 )/ GEIS) 4. ‘Tho otreceos in a rotating disk (of unit thicknoss) ean be regarded as duo to contrifugal oreo as body force in a stationary disk. Show that this body force is Aerivable from tha potential V = —$pu8(a + yf) wl pin Ue ews, le Ue angulir velocily uf rolation Galivut the origin). &. A disk with its axis horizontal has the gravity stress represented by Eas. (d) of Art. 16. Make a sketch showing tho boundary forces which cupport its ‘weight. Show by another sketch tho auxiliary problei of boundary foroee which must bo eolved when tho weight is ontirely supported by the zesetion of a horlaontal surface on which the disk stands 6. Accylinder with ils xin hurizintal Ins Ue gravity shies represenbal by Rus @ of Ari. 16. Tis ends are confined between smocth fixed rigid planes which suiulain the condition of plane slrain. Scotch the forces acting on its surface, including the ends. ‘T. Using the strosa atrain relations, and Eqs. (a) of Art. 19 in the equations of ‘equilibrium (18), show that in the absence of body forees the displacements In problems of plane stress must. sntisfy Bu eu lee and a companion equation. 8. The figure represents a “tooth” on a plato in a state of plane stress in the plane of the paper. The faces of the tooth (the two strnight ines) ar: fe fram force. Prove thik then ix nn stevss ball sh (ND. The anal Une ewan far rewsabra,CHAPTER $ TWO-DIMENSIONAL PROBLEMS IN RECTANGULAR COORDINATES 17, Solution by Polynomials. It lias been shown that the solution of (wo-dimensional problems, when body forces are absent or are con~ stant, is reduced to the integration of (he differential equation a ae, a aa t ara t aye =o (@) having regard to boundary conditions (20). In the ease of long reclungular strips, solutions of bq. (a) in the form of polynomials are of interest, By taking polynomials of varius degrees, and suitably ad= justing their cocfficients, a number of practically importuil prublems ean be solved.! Boginning with a polynomial of Ue second degree e-Gattbayt sy w ” Pre 21 which evidently satisfies Eq. (« mo — 0, we find from Bys. (20), puting #on 7 = a, All threo strees components are constant (aroughoul the body, i, the stress frnetion (b) represents a combmation of uniform tensions or compressions? in two perpendicular direclious aud # unifurun shear ‘The farces on the houndaries must equal the stresses at thesc points us discuscod on pago 23; in tho case of a rectangular plate will sides parallel Us the evurdinate axes these foraes are shown in Fig. 21. +A, Mosnagor, Contpt. vont, Vol. 18%, p. 14/2, 1901, See also A. Timpe, 7 Math, Physit, vol. 52, p. 348, 1905. 2 This dopends on the sign of coefficients ay and by, ‘The dinertinns sf xlrrscest imiontad in Fig, 21 are thas enrrespnling tar peosiive valinw of a, ha, 8 2930 VHEORY UF BLASTICITY ‘Let us consider now # sires function in the form of a polynomial of the third degree: 60 = gat + Boyt Gov + te © ‘This also satisties Hq. (a). Using I find (29) and pmtting py = 0, we Set = can bday any Oy — jg et hay i tay oa ay he — cay For a rectangular plate, taken as in Fig. 22, assuming all ooofficionts except dz equal to zero, we obtain pure bending. If only coefficient. as iy different from zero, we obtain pure bending by normal stresses applicd to the sides y = Leot the plate. If coetficient.bs or cs is taken Fra, 22, Tra. 23. Gifferont from zero, we obtain not only normal but also shearing stresses acting on the sides of the plule, Figure 23 represents, for instance, the case in which all cocflicicnta, except bs in function (¢), are eqnal io zero, ‘The directions of stresses indicated are for bs positive. Along the sides y = te we have uniformly distributed tensile and cumpressive slesses, respectively, and shearing stresses proportional toz, Ontheeidos = Lwehave only the constant shearing stress —hy, and there are nu slresses ucling on the side x = 0, An analogous stress distribution is obtained if cocificient cs is taken different from zero. Tn (aking Ue stress function in the form of polynomials of the second and third degrees we are completely free in choosing the magnitudes of the coefficients, since Ey. (a) is sutisfied whatever values they may have. In the eave of polynomials of higher degrees Eq. (a) is satisficd only if certain relations belween Uhe vuellicieuts ure sutisGed, Taking,TWO-DIMBNSIONAL PROBLEMS 3h for instance, the stress function in the form of @ polynomial of the fourth degree, ben 75 and substituting it into Eq. (a), we find that the cquation is satiaficd only if pay) a ety!) 4-3 2 hi ty dy a rg" veh gy aut | @ = —Qc +a) Tho stress components in this case are Pes os = FG = eat + dary ~ (2eu + aay? Poe _ gas ° Oy = ee = at + Bary + cay be be . To “aS = — bat — aogey — 4 ye Coofficionts ay, . . . , dy in these expressions are arbitrary, and by suitably adjusting them we obtain various conditions of loading, af a rectangular plate. For instance, taking all coefficients except d, equal. to nero, we find an my, y= 0 ty oe @ Acouming ds positive, the forees acting on the rectangular plate shown Fig. 24 und producing (he stresves (e) ure us given. On the longi- tudinal cides y = +¢ are uniformly distributed shearing forces; on the ends sliewring forves are distributed according to a parabolic law. ‘The shearing forces acting on the boundary of (he plate reduce to the couple! = del 4, Ld, ayers 2 det Zaett This couple balanoce tho couple pro- duced by the nermal farees along the ¥ side z ~ 1 of the plate. wae 78 Let 1s consider a stress fu fifth degree. yn in he fore of polynomial of the as ds a2 6 at gigtv | gga é- Mott ip Be ret stat The thickness of the plate is taken caual to unity. OARA neural hs32 WHEORY OF ELASTICITY Substituting in Eq. (a) we find that this equation is sulisfied if 0g — —(2cy + 3a5) Se ~ —Hbs + 2ds) ‘The corresponding slress components are: Sth davty ~ Beg + Bngdny? — Bag + Bdey* -3 aged + heaty $ eeng? + & ey ai 4 natty — dary? + 3 (Bes + Ras)y* Again evefficients as, » iy are arbitrary, and in adjusting them we obtain solutions for various loading conditions of a plate. Taking, t x He T Hater e~ $e =] ite(ten$e2) | isons |_,__| fp te? J 7 7) (6) Fra. 98. for instance, all evefficients, exeeph ds, equal ta sera we find oa = ds(aty — $y") uy — ey (9) Toy — Oy? ‘Lhe normal forces are uniformly distributed along the longitudinal sides of the plate (Fig, 25a). Along the side x = 7, the normal forces consist of two parts, one following a linear law and the other following the law of a cubic parabola. ‘The shearing forces are proportional to x an the Inngitndinal sides of the plate and follow a parabolic law along tho side 2 ~ 1, ‘The distribution of those etresses is shown in Fig. 25). Since Bq, (a) is a linear differential equation, it may be concluded that a cum of soveral solutions of thio equation is also a golution, Wo can superpuse (he elementary sulntions eonsidered in this article and in this manner arrive at new solutions of practical interest. Several examples of he application of (his method of supe considered.TWO-DIMENSIONAL PROBLEMS 83 1B. Saint-Venant’s Principle. In the previous article several cases were discussed in which cxact solutions for rectangular plates were obluined by taking very simple forms for the stross function ¢. In each case all the equations of elasticity are satisfied, bul the solutions are exact only if the surface forece aro distributed in the manner given. In the case of pnre hending, for instance (Fig. 22), the bending moment must be produced by tensions and compressions on the ends, these: densions and eumpressions being proportional to the distance from the neutral axis, ‘The fastening of the end, #f any, must be suel ay nol lo interfere with distortion of the plane of the end. If the above condi- tions are not fulfilled, ie, the lemlug moment is applied in some different manner or the constraint is such that it imposes other forees om the end section, the solution given in Art, 17 is no longer an exact solution of the problem. ‘The practical ntility of the sulution however js not limited to such 9 specialized ease, It can be applicd with sutfi- cient accuracy to cases of bering in which (he conditions at the ends are not rigorously sutisficd. Such an extension in the application of the solution is usually based on the so-called principle of Saint Vonant, ‘This principle states that if the forces acting on » sunall porliow of the surface of an elastic body are replaced by another statically equivalent system of forces ucling on the sume portion of the surface, this redis- tribution of loading produces substantial changes in the: stresses Iucally Lut hus a negligible effect on the stresses at distances which are large in comparison with the linear dimeusious uf the surface on which the forces are changed, For instance, in the case of pure bending of a teclangular strip (Fig, 22) the cross-sectional dimensions of which arc ‘small in comparison with its length, the manner of application of the external bending moment affects the stress distribution only in the Vicinity of the ends and is of nu cousequeuce for distant cross sections, at which the stress distribution will be practically as given hy the solu- tion lo which Fig, 22 refers, The same is true in the case of usiul leusivu. Ouly near the loaded end does the etroas distribution depend on the manner of applying the tensile: furce, aud in cross sections at a dietanee from the cnd the stresocs are practically nnifermly disiribuled, Sume examples illus- trating this statement and chowing how rapidly the stress distribution becomes practically uniform will be discussed later (see page 52), “This principle was nluded étrangers, vol, 14, 1 dieoussed later (see p. the famous memolr on torsion in Mém, savant tw the principle of consorvation of enorgy it34 THEORY OF ELASTICITY 19. Determination of Displacements When the components of strose are found fru the previons equations, the components of strain can be obtained by using Hooke’s law, Eqs. (3) and (6). Then the displacements «wl v can be obtained from the equations mi wy ar ay @) The integration uf these eqnations in each particular ease docs not present any difficulty, and wo shall have several examples of their application. Tl may he seen at once that the strain components (a) remain unchanged if we add to wand o the liner functions msatly mse be w in which a, b, and ¢ ure conslunts, ‘This means that the displacements are not entirely determined by the stresses and strains. Ou the dis- placements due Lu the internal strains a displacement like that of a rigid body can be superposed. The constants a and cin Eys. (0) repre- sent a translalury motion of the body and the constant 6 is a small angle of rotation of the rigid body about the z-axis. Tt has been shown (sev page 2) that, in the case of constant body forces the stress distribution ie the came for plane stress distribution ur plane strain, The displucements however are different for these two problems, since in the ease of plane stress distribution the eumpouests of strain, entering into Eys. (0), are given by equations ee ee and in the case of plane strain he slraia components are: Gea Blea = Ho, +o = GIG = Hoe = (1 + al 1 fo Blew — Woe $00] = lll = Ary = ol + Dud 1 Ya ~ Gta Tt io cacily verified that these equations ca be ohtained fram the pre- ceding set fur plane stress by replacing Z in the latter by #/(1 — »*), andr by v/(1 — »), These substitutions leuve(?, which is 4/20. 1 »), uuehnaged. ‘The integration of Eqs. (a) will be shown later in dix cuscing particular problems,TWO-DIMENSIONAL PROBLEMS 3 20, Rending of a Cantilever Loaded at the End. Consider a cantl- lever having a narrow rectangular cross action of unit width bent by a Fores P applied at the end (Mig. 28). The upper and lower edges are free from load, and shearing forees, having a resultant P, arc dis tributed along the end x = 0. ‘These conditions can be sulisfied by a proper combination of pure shear, with the stresses (e) of Art. V7 repre —f sented in Fig. 24, Superposing the pre shear ray — —b2 on the stresses (@), we find ,=0 op @ ty = a Gy Fra. 26, ‘To have the longitudinal cides y = +e frcc from forces we must have from which ‘To catisfy the condition on the loaded end the sum of the shearing forves distributed over Uhis eud aust be equal to P. Hence! (e- sat) ay =P _3F de from which bs Substituting these values of da and ha in Vigs. (a) we find ~82(,_-¥ ve 46 é Noting that $e 1s the moment of inertia I of the cross section of the cantilever, we have an", mo > ) ry ee yy ios I2 * The minus sign before the integral follows from tho rule for the agn of shearing sino Siren cnn ends SD put i upward ew 3)i 36 THEORY OF ELASTICITY ‘This coincides completely with the elementary solution as given in books on the strength of materials, It should be noted that this solution represents an exact solution only if the shearing forees on the ‘ends are distributed according to the same parabolic law as the shear- ing stress r= and the intensity of the normal forees at the built-in end is proportional to y. If the forces at the ends are distributed in any other manner, the stress distribution () is not a correct solution for the ends of the cantilever, but, by virtue of Saint-Venant’s principle, it can be considered satisfactory for cross sections at a considerable distance from the ends. Let us consider now the displacement corresponding to the stresses, (®). Applying Hooke’s law we find =e = Me _ Pty on oF oy Eo Er (© a ty Pg we Et Game) @ ‘The procedure for obtaining the components u and v of the displace- ment consists in integrating Bqs. (c) and (d). By integration of Eqs. (©) we find »Pay* we PY eM, 0 = Shp the) in which f(y) and f(z) are as yet unknown functions of y only and 2 only. Substituting these values of w and v in Eq. (d) we find Pat, dfty) Py? 5 diz) ~ prt dy + 2er +" de ~ - (@—y) BEIT" de ae # In this equation some terms are functions of z only, some are functions of y only, and one is independent of both z and'y. Denoting these groups by F(z), Gy), K, we have Fa) =- P2408 = UW 4 Py _ Put OW) = "ay + oer ~ 2G _Pe 216 K and the equation may be written F(x) + Gy) = K Such an equation means that F(z) must be some constant d and Gy) some constant e, Otherwise F(x) and G(y) would vary with 2 and y, respectively, and by varying 2 alone, or y alone, the equality would beTWO-DIMENSIONAL PROBLEMS 37 violated. ‘Thus e+a= Te © and Px? Yo) _ _ Pv, Py geit “ay = ~ get ae t Functions f(y) and f(z) are then = Py 4 Py! fy) = pet t+eeteuts Sie) = Bt doth Substituting in the expressions for u and » we find _ Paty wPy? , Py® ET Var ugtuts ee Pe ath ® ser + cart e+ u ‘The constants d, e, g, h may now be determined from Eq. (e) and from the three conditions of constraint which are necessary to prevent the beam from moving as a rigid body in the ry-plane. Assume that the point A, the centroid of the end cross section, is fixed, ‘Then u and 9 are zero for x = I, y = 0, and we find from Eqs. (), PR g =0, h=— SEI ~ dl The deflection curve is obtained by substituting y = 0 into the second of Eqs. (g). ‘Then Or = FE ‘v0 6EI 6ET For determining the constant d in this equation we must use the third condition of constraint, eliminating the possibility of rotation of the beam in the zy-plane about the fixed point A. This constraint can be realized in various ways. Let us consider two cases: (1) When an ele- ment of the axis of the beam is fixed at the end A. Then the condition of constraint is ‘av (ess =o ® (2) When a vertical element of the cross section at the point A is fixed, -dd-2) a)38 THEORY OF ELASTICLLY ‘Then the condition of constraint is In the firnt cue we obtain from Eq. (h) PE o= ~ oi and from Ea. (0) wo find Substituting all the constants in Eqs. (9), we Paty Py? Py? (a5 Pot set — oer t oe | \oni~ 2G)" wPryt Pat Pee PE im) aar | eer 2n1 + 3ET "The ecnation of the deflection ourve is 1 Phe, PR ren ~ Fe Oe + @ which gives for the deflection at the louded end (« = 0) the value PP/SEI_ This enincides with the value usually derived in elementary books on the strength of materials. ‘Ty illustrate the distortion of cros3 sections produced by shearing stresses let us consider the displucement u at the fixed end (c =). Por this end we have from Eqs. (m), oly’ | Py _ Poy Wea= gay + gig ~ BIT au) Py? Py _ Pet (2)... = +6 ~ BG @ au __8P aufi=0 “aa ‘The shape af the cross scotion after distortion is as shown in Fig, 27a. ‘Dus to the shearing sess ray = —P/4e at the point A, an clement of the cross section at A rotates in the zy-plane about Ube puiul, A through an angle 3P/4cG in (he clurkwise direction. If a vertical clement of the cross sectiou is fixed at A (Hig. 276).TWO-DIMENSIONAL PROBLEMS 39 joslend of « horizontal element. of the axis, we find from condition (2) and tho firet of Eqs. (@) fr °™ ORT and from Iq. (¢) we find a — Ph _ Pe BEI 3G Substituting in the second of Figs. (g) we find Pot Pir, PR, Pet yo ~ Gar ~ ger t gett are! — © Comparing, this with Viq. (n) it can he conchided that, due to rotation { ‘ (a Pra. 27, of the end of the axis at A (Fig. 276), the doflostions of the axis of the cantilever are inereused by the quantity Pe 3P aa ¢-*) =a¢-4) ‘This is the so-called effect of shearing force on the deflection of Une lem. An practice, at the built-in end we have conditions different from those shown in Fig, 27, The fixed section is usually not free lo distort and the distrihtion of forces af this end is different from that: given by Eqg. (6). Solution (0) is, however, satisfactory for compara tively long cantilevers at considerable distances from the terminals. Q1. Bending of a Beam by Uniform Load. Lot o beam of narrow Tectangular eros section of unit width, supported at the ends, be benteo THEORY OF HLASTICITY by a uniformly distribuied Inad of intensity g, as shown in Fig. 28 ‘The conditions ot tho uppor and lower edges of the beam ure: Cau) Cie = 9, rey e — (@) ‘The conditions at the ends x — | éare f rw dy — Fat, fie-au - "The last two of qs. (6) state that thero is no longitudinal foree and no bending couple applied ul the ents of the beam. All the conditions (a) and (b) can be satisfied by combining certain solutions in the form ihe tart Pee 1 a “ i— oudy=0 {b) Y fal to @ Tro, £8, of polynomials as obtained in Art. 17, We begin with solution (0) illustrated by Fig. 23. To remove the lensile siresex slong the side y= cand the shearing streaces along the cides y = -+e we auperpose a simple compression ¢y = ag from solution (B), Art. 17, snd (he stresses oy = by and ty = —ber in Fig, 23. Tn this manner wo find ds(xty — 3y") tdsy" + boy + ae () —dery® — be From the vonditious (u) we Gud Ta det — hy = 0 ido | b,0 | a2 - 0 bie? — bye + 1 — —@ from which 40 Substituting in Eqe. (¢) and noting that 2c'/3 is equal to the moment of inertia J of the rectangular cross-sectional arca of unit width, wo find‘TWO-DIMENSIONAL PROBLEMS 41 It can easily be checked that these stress components satisfy not only conditions (a) on the longitudinal sides but also the first two conditions (®) at the ends. To make the couples at the ends of the beam vanish we superpose on solution (d) a pure bending, oz = diy, oy = ty = 0, shown in Fig, 22, and determine the constant ds from the condition at a= l [ican =f [-35e - 2) +au]vav from which Hence, finally, (33) The first term in this expression represents the stresses given by the usual elementary theory of bending, and the second term gives the necessary correction. This correction does not depend on « and is small in comparison with the maximum bending stress, provided tha span of the beam is large in comparison with its depth. For such beams the elementary theory of bending gives a sufficiently accurate value for the stresses a;. It should be noted that expression (32) is an exact solution only if at the ends c = +1 the normal forces are dis- tributed according to the law 73 2 2 x -14aGe -2ev) i.e. if the normal forces at the ends are the same aso: for = +/ from Eq. (83). These forces have a resultant force and a resultant couple equal to zero, Hence, from Saint-Venant’s principle we can conclude12 THRORY OF ELASTICITY that their effeels un the stresses at considerable distances from the ends, say at distaneos larger than the depth of lhe beam, can be neglecled. Sulution (35) at such points is therefore accurate enough for the ease when there are no forves X. Tle discrepancy between the exact solution (83) and the approxi- mate solution, given by the first term of (88), is due lo the: fack that in deriving the approximate solution t is assumed that the longitudinal fibers of tho beam are in a condition of simple tension. Fram solution (d) it can he seen that there are compressive stresses 6, between the fibers, These stresses are responsible for the yn represented by the second term of solution (33). The distribution of the com- pressive stresses ¢, over the depth of the beau is shown in Wig. 28¢+ ‘The distriluitiun of shearing stress rm, given by the third of Ege. (d), over 9 crocs scetion of the beam coincides with Unt given by the nsnal elementary theory. When tho boam is loaded by its own woight, insteail uf tlie distrituted load g, tho eolution must be modified ky putting, = Zaye in (88) and the last two of Has. (@, and adding the stresses a 2-00 — 9), ty BO te) For the stress distribution (c) can be obtained from Ege, (20) by taking = boolea? + ¥8) and therefore represents a posable stato of strocs due to weight and boundary forces. On the upper cdge y ~ —¢ we have ¢, — 2pge, and on the lower ely uv =6, ee —0, ‘Thus whon the stresses (¢) are added to the previnun sailition, swith ¢ — pga, tho etroee on both horizontal edges is zero, ul the Wout en he beama ‘eonciota only of ite own weight. ‘The displacements u and pcan be calculated by the method indi- cated in the previous article, Agsuming that at the eenlruid uf the middle cross section (« = 0, y — 0) the horizontal displaccment is zcro and the vertical displacement is equal to the deflection 4, we Lud, usiny solutions (d) und (88), 2 2 2 [(- -s)rte(Ge om) ' a(lv-evt3e)| oa 1 yt ~ of 5 SF ghey + of Pye all q [tet _t 1, 1 wah [Me- go few (ui grea] es It can be yeeu from the expression for u that the neutral surfaco of the ve t ssUWO DIMENSIONAL PROBLEMS 43 ‘beam is not st the centerline. Due to the compressive stress (na = = tho center line has tensile strain 19/22, and we find (me — Fe From the expression for » wo find the equation of tho deflestion curve, () ooh | -H-peet (it deex| in Assuming that the deflection is zero at the ends ( — +1) of the center line, we find =i Bed o Smelt eR tS a) ‘The factor before the brackets is the deflection which is derived by the elementary analysis, assuming, that crass sections of the heam remain plane during bending. The second term in the brackets repreconts the correctiva ususlly culled the effect of shearing force. By diffcrentiating Eq. (f) for the deficction curve twice with respect, to 2, we find the following expression for the curvature: B)-aPaten(ed] crx EI It will he seen that the curvature is nul exactly proportional ty the bending moment! ¢(? — z%)/2, ‘The additional term in the brackets represents (he aevessary correelion to Ube usual elementury formula. Amore general investigation of the curvature of beams shows" that the correction term given in expression (35) can also be used for any case of continnonsly varying intensity of Inad. ‘The effeck of shearing force ‘on the deflection in the caso of a concentrated load will be discussed later (paye 107). An elementary derivation of Ue effort af Une shusring fore of the deflection curves of Vesuies Tnos bevy coud by: Rank ast Grashof! in Grsmauy. ‘Taking Uh ‘This was pointed out first by K. Pearson, Quavt. J. Math., vol. 24, p. 60, 1889. #800 paper by I. vy. Kérmén, Abhandl, aerodymam, Inst, Tech. Hochschule, Aachen, vol. 7, p. 8, 1927. * Rankine, “Applicd Mechanics,” 14th ed., p. 344, 1895, ¢ Grashof, “Elastisitat und Feotigheit,” Yd ed., 1878,“ THEORY OF BLASTICINY /2eG), whore Q is tho shoaring foreo, riven Ity the derivative of the above sheaing strain with respect to 2, which gives (y/20). ‘The enrrectad expression, for the curvature by elementary analysis then becomes Poe 39 afm, ty? oe [Fe teu +] Comparing this with expression (Af), it is san (lid the elemeniary solution gives an exaggerated value’ for the correction. ‘Phe correction tcrm in expression (85) for the curvature cannul, Lu: altrilutedd ‘to the shearing forec alone, It 1s produced partially by the compressive stuessea 0. ‘Theee strosees aro not uniformly distributed over the depth of the beam. The lateral expansion in the z-diroction produecd by these stresecs diminishes from the ‘top to the bottom of the beam, and in thie way a roversed curvature (convex "ipwards) in produced. This curvature together with tho font of shonnag force toounts for the eorection term in Ey 3). 22. Uther Cases of Continuously Loaded Reams, By increusing the degree of polynomials roprocenting colutions of the two-dimensional y problem (Art. 17), we may obtain solu- Jf | tions of bending problems with various = iypes of continuously varying load.* —— By taking, for instance, a solution m the a e-pey € torus of « polynomial of the sixth depree —— ond combining it with the previons selutious of Art. 17, we may obtain the stresses in a vertical cantilever loaded by hydrostatic pressure, az shown in Fig. 29. In this manner it can he: shown that all conditions on the longitudinal sides of the cantilever are satistied by the following system of stresses: _7y aey* 4 wa (ane Sess) g oy a= -B+0(4- -k (a) Fin, 28, tu Eo = 99) — Scot — wi + howe — 9) Alere gis the weight of unit volume of the fluid, so that the intensity of (he Joud at a depth z is gz, The shearing foree and the bending moment at the same depth are gx"/2 and qz*/6, respectively. It is 2A Vnller ayproxinion ie given by elementary sirtinaenergy snnilernd Bee 8. Taumbiewka, “Blrength uf Mutants, 2d wl, vol. 1, p. 208, * Bee pays by npr, loc. ela; W. Te Ongonnd, oT. esearch Natl Dur. Standards, ‘vol. 28, p. 159, 1942.TWO-DIMENSIONAL PROBLEMS 45 evident that the first terms in the expressions for o. and ry are the values of the stresses calculated by the usual elementary formulas. On the top end of the beam (r = 0) the normal stress is zero. The shearing stress is -v) + Soee— Although these stresses are different from zero, they are very small all over the eross section and their resultant is zero, so that the condition approaches that of an end free from external forces. By adding to oz in Figs. (a) the term —q,z, in which q1 is the weight of unit volume of the material of the cantilever, the effect of the weight, of the beam on the stress distribution is taken into account. It has been proposed! to use the solution obtained in this way for ealeulating the stresses in masonry dams of rectangular cross section. It should be noted that this solution does not satisfy the conditions at the bottom of the dam. Solution (a) is exact if, at the bottom, forces are acting which are distributed in the same manner as ¢~ and rz in solution (a). In an actual case the bottom of the dam is connected with the founda tion, and the conditions are different from those represented by this solution. From Saint-Venant’s principle it can be stated that the eficct of the constraint at the bottom is negligible at large distances from the bottom, but in the case of a masonry dam the cross-sectional dimension 2c is usually not small in comparison with the height ! and this effect cannot be neglected.? By taking for the stress function a polynomial of the seventh degree the stresses in a beam loaded by a parabolically distributed load may be obtained. Jn the general case of a continuous distribution of load q, Fig. 30, the stressoa a any cross section at a considerable distance from the ends, say at a distance larger than the depth of the beam, can be approximately celeulated from the following equations:* + M. Levy, Compl. rend., vol. 126, p. 1235, 1808, * The problem of stresses in masonry dams is of great practical interest and has ‘been discussed by various authors. See K. Pearson, On Some Disregarded Points jn the Stability of Masonry Dams, Drapers’ Co. Research Mems., 1904; K. Pearson. and C. Pollard, An Experimental Study of the Stresses in Masonry Dams, Drapers’ Co. Research Mems., 1907, See also papers by L. F. Richardson, Trans. Roy. Soc. (London) series A, vol. 210, p. 307, 1910; and 8. D. ‘Carothers, Proc. Roy. Soc. Edinburgh, vol. 33, p. 202, 1918. I. Muller, Publications du laboratoire de photo- Aasticitt, Zixich, 1930, Fillunger, Oesterr.’ Wochechr. offentl. Baudienst, 1913, Holt, 45. K. Wolf, Sitaber. Akad, Wiss. Wien, vol. 123, 1914. sagt Seems Athena, aerodynam. Inet, Pech. Hochechule, Aachen, vol. 1, p. My0 THEORY OF BLASTICITY My ve wn +05 ~T8 64, (42 we $er(-3) (a6) ia whioh Af and @ are the bending moment and shearing farce eebntead in Uw usu way nud q ix the intensity of load at the 1 x cvoss section under consideration. These equations agree with those proviourly ob- ‘Yainod for a uniformly loaded beam (see Ar. 21), Fra. 20, edge (y = +0) of the beara, the expressions for [hw slirsees are obtained from Eqs. (26) by uperposmg a uniform tondile strose, oy — a and wfee-£ oe) su het) 28. Solntion of he Two-dimensional Problem in the Form of a Fourier Series. ‘Th lias een shown that if the load is continuously distributed along the length of a rectangular beam of narrow cross goction a stross funetion in the form of a poly- nomial may bo used in certain simple eases. If the low! is discontinunme, x iret junction in the form of a trigonometric series whould Tue usea.# ‘The eqyuntion for the stress function, 24 yy Og 8 gat aaa tae 8 ta) may bo eatiofied by taking the function ¢ In the form enon yy) ® in which min nn integer and f(y) a function of yonly, Substituting (b) into Hq. (a) ‘and using the notation ms /I = «, we find the following equation for determining. Si: ify) — Baty) + FTY) = 0 ® 1 The first application af trigonumedsie weries ia the solution of beam problems was given hy M. ©. Rilidve in » thesis, Bur divers caa de la ticxion doo pricmos reulaugles, Dordeaux, 1889. Bee also his paper in Compt. rond,, vol. 125, pp. 402- 404 and 1190-1199. "Further progreaa in tho application of this solution wis role by LN. G, Filon, Path, Y'rans,, series A, vol, 201, p. 63,1903. Several jmatioular examploo wero worked out by F. Bleleh, Raungeniaur, wil. 4, » 256, 1999,TWO-DIMENSIONAL PROBLEMS aa The general integral of this linear differential equation with constant coofficients is SQ) = Cr eush oy + Cri hoy + Cay eos ay ob Cay sinh ay ‘rhe stress fonction then is $= sin ae(U, eoah ay + Usamh av | Uwveosbav | Uweinbuy) — (d) nul Uhe corresponding stress components ate on -% = sin at (Cia? cosh ay + Cat sinh ay + Csa(? sinh oy 1 ay cosh atv) [Coal cosh ay + wy sinh ay) 28 _atsin ads cunl ay + Caninh ay + Cry cosh ay + Crysinb ay) (@) ia toy = — EE = ~ ur 008 uslCru sink uy + Cor cosh ay + Colours ry ory + ay sinh ay) + Cy(sinh ay + ay cosh ay)] Lat us consider 2 partleular ease of a rectangular beam supported at tho ends and subjected along the upper and lower edges to the action of continuously Tr 41 =f aa | z f a Fin, at, \Tealsibuted vertical forces of the intensity A cin ae ond B cin ax, respectively. Figure 31 shows the ease when a = 4a/I and indicates also the positive values of 4 aul B, ‘The stress distribution for this ease can be obtained from solution (¢). ‘The conslunts of integration Cy, .. . , Cemay be determined from the conditions on the upper aml lnwer elges of the ben, y = te. These conditions are: Por y + +6, an For y = —2 ty; oy = ~Acinas Substituting thoco values in the third of Ege. (0), we find roc sinh uc + Ora cosh ax: + Ca(conh oe + at sink ac) + Cusinh ae + at cosh ae) « v uh am) = Gafsink ac + or newh ae) = 0 Cre oh ae 1 Ona cosh uc + Caloosh ac + aaR THEURY OF ELASTICITY from which yh C4 = 08 wed ge + at stl a ce sinh we @ Ce Crs ae F ae cook ad Using, te couitions on the sides y= stoin the seeond of qs. (¢), we find aC, 00th ac + Cy alah ot + Cyreamh ac + Cut sinh we), B r(C. cosh ae — Cyninls a — Cac cosh ac + Que sinh wc) = A Ry aviing nnd subtracting these equations and using Eqa. (9), we finde (4; = AEE. sit + a0 cosh we a sinh Bar ok De dee 3 Re snl ay sa we @ ae Bae ny A=B,_acosh ae Cy Se" inh Bao — a A+B. asinh a ‘a? Binh 2a0 + 2a ‘Substituting, in Fags. (}, we find the following oxpreccions for the stress components: Gen {ee 20sh an — sinh oe) cosh ay — au sin et sinh ae a= (A+B) ee ie 1 (azsiuh we — cosh we) ah ay — ay coth oy £080 265 ag sin ax a ‘inh Yao — Tae ca = hg my eat ih at)-008 of = . 4 = my esas + coh ae) eh my sh oA * = la 1 By oe Mh = oth 2 hn my Ath pn ra “Theso stresses satisfy the cmulilivus shown in Hig. 31 along tho sides y ~ ke. ‘Av the ends af thw insun 2 = Q aud 2 — /, the atreaseo 2, uro nero and only shearing stress tay ix present, ‘This stress is represented by two terms [see Eqs (H)] ‘The fist Worm, proportional to A FB, roproeente stresses which, for the upper wud lawer halves of the cnd eroce'ecetion, aro of the same magnitude lub of opposite sigu, ‘The reeultant of thove etrosnas over the ent ix aor. ‘The second terms proportional to A — B, has resultants at thw eis of the beam which-maintan Squibbrivin with the loads applies to Live yngilaiual aides (y =e). "If these loads are the sume fur lnil sides, coeficient A ia equal to B, and the reactive forces at hie euilk vanish: Let us consider this’ partienlar ease more fi Altai, qssrming, tn ie leugth of the beam io largo in comparison. with its depth, Fe te veanud of Eqs. (#) the normal atroeees ¢, over the middly ple y = 0 vf he beau axe oTWO-DIMENSIONAL PRORLEMS 49 For long beams ar, equsl to mav/l is small, provided the number of waves m is not Jarge. Then, substituting in (Dy 1 (ot st ae = ae EP | oot ae 1 +O 4 GO + sud neglecting small quantitico of higher order than (ac) we find op one BE wake + Sy Rast Ee Me " ‘The constant terms As and Do represent a uniform loading of the beam, which was dweussed in Art, 21, Stresyea produced by termo contaming sin (mre/t) aro obtained by summing up colutions (k). ‘Tho strecece produced by terms containing ens (ma2/l) axe easily obtained from (2) by exchanging sin a for ens a nnd vine ‘vers, nel hy changing the ign of rap To ilustrale dhe applicalion of this general method of stress calculation in reetangular plates, let us consider the case shown in Tig. 33. For this ease of Symmotrieat loading the termo with oim (mea/!) vanih from oxproexions (m) and the eooffaionte 4, and .4, oro obtained in tho usual manner; ” 1928; or Dyesly, “Fourier ‘Wouriee Series and Hound * For Fourier serins se Oxguna, Advanced Calculus, Senco and Spherical Harmonies,” 1909; or Churchill sry Value Probloma,” 1041,50 THEORY OF ELASTICITY ‘The torms A and B, represent a uniform ga/t. The stress preuluined hy Ue brigon ives are olin Try sing solutions (&), eechanging sin ar for cot uz in this solution t and changing the aizn of rav. 1 x Let us sonsider the middle plane y — 0, on ¢ v Volussef neay/2P 25 THe 3 a sero.) s Y¥ Values of an/e thece fs aly Une sual stiess oy. By using Une seen of Bays. (8) we Gl ‘This stress was cvalunted by Filon’ for an infinitely long strip when the dimension @ ig very small (t.04 concentrated force P — Za). ‘The reeulte of thi calculation aP sf . 0 ey | at are shown in Fig. 34, It will be seen that o, diminishes very rapidly with 2 Ain. value z/e = 1.35, it becomes zero and is then replneed hy tenia. Pilon dinrunen, “LN. G, Vilon, Prana, How. Soc, (London), ocricd Ay vol. BOL, p. 67, 1003. Tho ame problem waa dizousged aloo by A. Timpe, Z. Math. Physit, vol. 55, p. 149, 1001; G. Meemer, Vergleichende spannungsoptisehe Untersuchungen - Dissertation, Gottingen, 1029; F, Seewald, Abkand!. cerlyusm. Tust, Tesh. Hochschule, Aachen, vol. 7, p. 11, 3097; unl H. Bay, TnyrnicasArchiv, vol. 3, p 485, 1982. n approxima: selution uf the same problem was given by M. Pigeaud, Compt. rewd., vol. 161, p. 673, 1915, The investigation ef the problom in the case of a rectangular plate of finite length wae made by J. N. Goodior, J. Applied Mechaviea (Irons, A.S.M.E.), vol. 84, no. 18, p. 173, 1992.TWO DIMENSIONAL PROBLEMS BL he forces P sure dgpbuced ene wills reget ts ing atrenses over the cross section nn in this case is of practical intercst and is shown in Fig. 96, J4 may be seen that for small ‘values of the ratio 0/e this distribution docs not resemble the parabolic distribution ivon by the elementary theory, and thut there are very large atreasce at the top ‘and bottom of the beam while the middlo portion of the beam is practically free from shearing stresses Tn the problew of Fig. 84 there will hy symmetry Ve my slisur stress. aml vattical displacement at the middle Tine wv —0, The upper half therefore corre- sponds to an cloatie layer resting on o rigid emooth bavo.! Lot us oontider now another extremo case when the depth of the plate 2c is Pp x Hye. 87, targe in comparison with the length 2 (Fig. 37). We shall use this case to show that the distribution of stresses over cross sections rapidly approaches uniformity 1s the distance from the point of application of the foreeo P inerenoce. By uoing the nocond of Kigs. (k) with eos as instond of sin aa and expressions (n) for soot elonte Aw’, equal to Bn’, we find a h egysinh ae wo 4g Saino, (neck aes obras ogy ony cosas (p) t m sinh ar + 2ae in which ga = P/2, If is emall in comparison with ¢, ae is a large number and It van he neglected in comparison with sinh ae, We can also put oc = aml ow = fe For eroso ocotions at a large distance from the middle of the plate we can write +The rough base ts considored by K. Marguorno, Ingonieur-avehiv, vol. 2, p. 108, 1981, nul » Mexibile hmt inextensible layer embedded in the elastic roerin, x ene of interest in soil unechauics, Uy M. A. Divt, Pyrics, vul. 0, p. 807, 1986,bu sinh ay = cosh ay = heer Sulnliluting Use in By. tp), we find HBORY OF HEASTICITY The — visnot very amall, say ¢ — y > 1/2, this series converges very rapidly xt itis only nooessary to take a few terms in exlvuluting ay. ‘Plhen wenn lake ia xd _ min iu EO _ ine aud putting Zag — Py we tind For y = ¢ —1, for instance, PP(rtt ogre M4 oe (ES one St wt) ..) ‘The first three terms of the scrics aro suficient to give good aeeurary and thw stress distribution in ac chown in Fig. $86. In the same fnew Ih nlress distribu~ tions for ¢—y = 1/2 and cy ~ BI are nbn givent Te is evident that at a diataneo from the end equal to tho width of the etrip the stress distribution is practically uniform, which confirme the coneluston usually mute on the basis of SainteVennnt’x pri For a long strip uc ou bu Fig. 97 the o« streswee will be tansmitted through the Jk 22 of the plate with litle change, pro- ‘vided the rate of variation along the edge is not too rapid, 'Iho streases of the present solution will, however, require some eorrer= tion on this account, especially near Uber y= te. Asolution of Ue problem of Fig. 3T wilh «=, by a different method,* In yractically uniform compreesive stress over ths middlo horizontal soction, in agrocmont with Fig. 38, The stresses in the vicinity of the points of application of the loads P will be discumwal later (seo page 85), Fro. aR 1 See paper by F. Rleteh, toe. eit 25.N. Gonier, Trans, A.8,3f.B., vol. 64, p. 178, 182,YWO DIMENSIONAL PROBLEMS 8 24. Other Applications of Fourter Serles. Gravity Loading. ‘Tho probleme cousidered in Art. 29 coucenueal a single “quar” Lor 22. ‘The xahutions, howev ‘can cqually well be regarded as representing perialio states uf slcems iu long stuips parallel to th # ax, sinee a Fourier aeries represents a periodic function, A con- tinuous beam consisting of a soquence of oqual spans similarly loaded will have such periodic wtress distribution if the end conditions are appropriate. Tf, as fn certain reinforved-conerei# binker eonstmetions, the beam is essentially a wall supported at points whist distance xysirk is eompuurnhbe with (ine depth (Fig, 89), useful results can be obtained by the present method.! The elementary beau (eRe Fi thoory is not adequate. A uniformly dietributed load ¢: on tho lower cdge, sup. ported by upward reactions uniformly distributed in widths 2 at intervals 1, presents a sperial ease covered hy Eqs. (m) of Art. 23. If the load g, 4s applied tou the upper ehge this mesedy recesary lo sul (he sires disteibution de to yl and opposite unifonnly distributed pressures yy ou both upyt an] lower edges If the load is the weieht of the beam itself the resulting body-dorce problem may ft onoe be reduced to an edge-toad problem, ‘he simple stress distribution a ym pay beh te 0 tatioties the cquations of cquilibrium and compatibility (19) and (24), It clearly represents eupport by uniformly distributed pressuro ago on the lower odge in Fig. 39. The condition that c, is zero at the lower edge, except at the supports Gof width 28), ts satisfied by adding this stress distribution to that represented by Fig. 89 whew i ix replied by 2pye, aad Ue xireo ince ts g and gs withont body 39, Problems 1, Investigate what problem of plane stress is solved by the stress function (4, 2) 4 | o-E(a-#) +50 & Investigale wha problesn is snlvest yy basa t avaa—m ‘unl to the region included in y ~ 0,y — d,# — 0, on the side « positive, | Problems of this kind are discussed, with references, in the book “Die Btat Eisenbotonbau,” by K. Boyer, 4d ed., p. 723, 1084; see also H. Crnemer, Engenieur= Archiv, vol. 7, p. 325, 1936, | INSTHTUL POLES TIMISOARA \ gypuioreca GENTRALAa4 THBORY OF ELASTICITY 2. Bow tht oly - 3 Vint — ae ek [e( my eee) Lal -s9)] is 9 stress frmetion, and Gud what problem it solves when applied to the region imcluded in y = Le, + = 0, on the side 2 positive. 4 ‘The stress function as(iy-@ wwe o=e(ha 9-4 is proposed as giving the solution for a eantilever (y = tr, 0-< <2) haul by tmniform shiner wlan, Une lawor onlgs, Che upper edge and the end z ~ U being freo fun Inad. In what reapects is this solution imperfect? Compare tho oxpros sions for the atreaaea with thooo obtamnablo from olomentary tenion and bending formulao. 8. Tho beam of Mig. 28 is landed hy upper edge, Fil exprmainnn for Ue displacement components w and 0. subs an expression for the change of the (originally unit) thiclmese, 6. The cantilever of Hig. 26, motead of having a narrow reetangular eroes #6e- tion, has a wide reotangular eross sootion, and is maintained im plane strain by puitablo forees along the vertical slées. The load is P por rit witlth on tlw en. Tustity the statement lint the slrween an, ay, roy ut (lie samne as those found iw Ark, 20, Find nu unpre for the stress vs, and sketch its distribution along the cides of the cantilever, Write down expression for the displacemont oom ponents wand ¢ whon a horigontal clement of the axie is fined at + ~ 1. tv sigh iolou UT La d 7. Show that Uf ¥ is 2 plane barmania fimeiinm, ie, ik vation dh Tayilace or oy Bt am? then the functions 2¥, u¥, (2% |v!) will cach satinfy Bq. (a) of Art. 17, and fo ean be used as otro funetions. 8. Show that (Ue Rev + Cyr +b year) sie ‘Derive sciies expressions for the stresses in a semi-infinite plate, v > 0, due to normal pressure on the straight edge (y = 0) having tho distribution im shen fi bean ‘Show that the atreos es at a point on the edge iso compression equal to the applied preseure at that point. Assume that the stress tends to disappear as y hers large. 8. Bho Mo (9) the wlresse pi Prob. 8 alisly Ey. @) vf Aut. 10. ny Raps. (@) of Ark. 28 mu @) the atzeses inCHAPTER 4 TWO-DIMENSIONAL PROBLEMS IN POLAR COORDINATES 25. Gencral Equations in Polar Coordinates. 1n discussing stresses in circular riugs und disks, curved buns of uarrow reclungiilar cross seo- tion with o ciroular axis, cte., it is advantagoous to use polar coordi- nates. ‘The posilion of w point ty (he middle plane of » plate bs then defined by tho dietance from the origin © (Fig. 40) and by the angle @ between rand a certain axis (x fixed in the plane. ‘Let us now consider the equilibrium of a smull element 1234 cut out irom the plate by the radial sections 04, U2, normal to the plate, and by two cylindrical surfaces 8, 1, uorunad to the plate, ‘Tho normal stress component in the radial direction is denoted by op, the normal component in the circumferential direetion by os, and the shearing-stress component by r74, each sysabol sepresent= ing stress at tho point r, &, which is tho mid-point /? of the element. On acon cof the variation of stress the values at the mid-points of the sides 1, ¥, 3, 4 are not quite: the swine ws Uhe values o,, a0, 7a, end are denoted by (¢-)1, ote., in Fig. 40, ‘The radii of the sides 3, 1 are denoted by r1, ri. ‘The radial force an hw side 1 iye,a7s dé which may Le written (6,7): 49, and similarly tho radial force on aide 8 is — (ors dy, ‘The normal force on side 2 has a component, along the vad iirough P of ~(os)e(r1 — 74) sin (de/2), which may be replaced by —(a2 dr (d8/2), The corresponding component from side 4 is (oes dr (0/2). The shearing furves on sides 2 and 4 give [(r0)s — (rable, Summing np forees in Use rulial direction, including body force R ver unit volume in the radial direction, wo obtain the cquation of equikbrum ),d9 = (oie dd — (dade odaar 1 [Cree (renal dr + Hr dO dr = 0 Es36 THEORY OF ELASTICITY Dividing by dr do this becomes Ts — (ora _ 1 (r9)4 Gers =e Bian) + ond + FOTO Ry my ar 3 a If tho dimonaions of the element are now Lukeu sinaller aud sunaller, lo the limit zero, the first term of this cquation is in the limit é(s-r)/ér. The second becomes oo, and the third 87,0/86. ‘The equation of equi- librinm in the tangential direction may be derived in the same manner. The two equations take the final form Sane Sime uot Rag én Lt Bret y Boh " rat arte 7 ‘These equations take the pluce of Eys. (18) when we solve two dimensional problems by means of polar coordinates. When the body force R is zero they ure sutisfied by putting Lab La ror! Rae a= @s) ma ~ 236 _ 1 #8 _ _ a (Las! 8 F880 7 Or 86 or \r 8, where #h is the stress funetion as a finction of r and @ ‘Vhis of course may be verified by direct substitution. A derivation of (38) is included in whut fullows. ¢ To yield a possible stress diotribution, thio function must encure that, the condition of eompulibility is sulisfied. Iu Curlesian coordinates (see page 26) this condition is Te S42 @ # a +TWO-DIMENSIONAL PROBLEMS IN POLAR COORDINATES 67 ‘Using these, and considering a8 a function of r and #, we find 4 _ a6 ar, 8690 _ 86.4 dd br or ant aba or 7 ag 0? ‘To got the cecond derivative with reapeot to 2, it ie only necessary to repeut (he above operation; heave (Sema Lane 5) (eave - snd) a 8g sin cus , Bpsin?@ , . Apnin Acumd = Gp ee ae tee 22g sin? @ te «) In the same manner we find ae Panto 4 22% 3° sin #003 @ , 96 ost @ ay® Bea 28% sin # cos @ , 9*6 cos? # « 2 ta Adding togothor (2) and (e), we obtain Pb rb - ap lad la oat! oy 7 art |r art ae @% Using the identity _(% 1 #\ (es _ ve axt tara t ay (+ a)B+8 and Eq. (@), the compatibility equation (a) in polar coordinates ‘vecomes a a\ (ae 1 ag 3g aT Te =) (G+ part Pap) =? 9) From various solutions of this partial differential equation we obtain solubinns of Lweedimensional prob pular coordinates for various boundary conditions. Sovoral oxamples of such problems will be dis- enssed in this chapter, Tho firct and sosond of the expressions (28) follow from Eqs. (b) and (¢). If ‘wo choose any point in the plate, and Ict tho 2 axia pass through it, we have # — 0, and o,, ¢, ate the eame, for this partioular point, ae ¢,, es. ‘Thue from (¢), putting a=0,58 YHEORY OF ELASTICITY ‘yhio oxpreseion continues to reprewnl a» whatever the orientation of the axis. ‘We find almilarly fram (), putting @ — 0, awe, — (28), = Bt 1 NG aes = oF and the third oxpression of (38) ean Iw: oft for —8%9/az ay analogous tn (h) nud (.}. 26. Stress Distribution Symmetrical aboul au Axis. Tf the stress distribution is syunuetrical with respcot to tho axis through O per- pendicular to the zy-plane (Fig. 40), Use siress camponents do not dopend on @ anl are finctions of r only. From symmetry it follows also that tho shearing stress 7,4 musi vanish. ‘Then onfy the first of the Ls equations of equilibrium (37) remains, and we buve Oo, 4 Se = Bett hm (40) wel likewise by finding the expression If the body force R ig zero, we may mse the stress function 4. ‘When this function depends only on r, the equation of compatibility (89) becomes: a 1d\fde , 1d¢' (424 (+i dip | 2dr art parr ‘Yhis is an ordinary differeutixl equation, which can be reduced to a lineur differential equation with constant coefficients by intradueing a new variable / such that r= «, Tn this manner the general solution of Fy, (41) can easily be obtained. ‘This solution ines faur constants of integration, which must le determined from the boundary conditions, By «nbstitution it can be ehocked that ¢@ = Alogr + Artlogr + Ur + D (42) is the general sulution. ‘Lhe solutions of all problems of syyuunelrical stress distribution and no body forces ean he obtained from this, ‘The corresponding stress components from Eqs. 38 are 1a A em ate eat BQ + Blog r) + 20° a= 28 2 A+ Hist dloen) 2¢ (48) If there iy uy hole ab Il inigin of coordinates, constants A and 1 van- ish, since otherwise the stress components (43) become infinite whemTWO-DIMENSIONAL PROBLEMS IN POLAR COORDINATRS 59 7 =0. Tlence, for a plate withont a hole at the origin and with no body forces, only one case of stress distribution symmetrical with respent to the axis may exist, namely that when a- = o» = constant and the plato ia in a condition of uniform tension or uniform compres- sion in all directions in its plane. If there is a hole at the origin, other solutions than uniform tension vr compression can he derived from expressions (433). ‘Taking J as sero, for instance, Eqs. 43 beeome a ~At2¢ nS (4) a- 442 ‘This solution may bo adapted to represent the stress distribution ina ligllow cylinder submitted ta tmiform pressure on the inner and outer surfaces: (Fig. 41). Let @ and 6 denote the inner and outer radii of the cylinder, and ps and p, the nniform internal and external pressures, ‘Then the boundary conditions aro: (iene = Dy (ar) = Pe (a) Substituting in the first of Eqs. (44), we ubluin the following equations todetermine A and (/: from which Substituting these in Eqs. (44) the following expressions for the stress, ‘components are obtained: — “bm =p) 1, par pb o- Ga@ ‘at wo us) _ ob. — p) 1, pa? — pat ° 1 eg pt of dighurements. Sar p. OA 4, “Taaguus sue In théarie . . sat fs dhe tay Tou + Proof Unie Bouse vara requires eo “The solution of thie probe i de PAlasticité,” Panis, 1852.60, THEORY OF ELASTICITY Tt is intoresting to note that the sum o, ++ 99 is constant thremgh the thickness of the wall of the cylinder, Tonco the stresses ¢, and os pro- dure « uniform extension or contraction in the direction of the axis of the cylinder, and erase sections perpendicular to this axis remain plane. ‘Hence the deformation produced by the stresses (45) in on eloment of the cylinder eut out by wer adjacent cross sections docs not interfere with the deformutiun uf the uvighluring ele mente, and it is justifiable to consider the ele- ment in the eendition of plane stress as we did in the above discussion. In the particular case when p, = U and the cylinder is submitted to internal prosoure only, Eqs. 45 give Paw A, (46) ‘These cquations show that ¢, is ulways x cumpressive stress and op a Leunile stress, ‘The lattcr is greatost at the inner surface of the cylin- dor, where pea? + 0%) Game = ae an (eran is wlways numerioally greater than the intemal precsure and approaches this quantity as b increases, so Ua il ean never he reducer) below ps, however much material is added on the outside, Various applications of Eqs. (46) and (47) in anachine design are nsmally dine cued in elementary books on the strength of materials.+ Tho corresponding problem fur a eylinder with an eccentric hore is considered in Art. 66. It wae solved by G. B, Joffery.? If the radius of the bore is a und Lhut uf the external surface b, and if the distance between their conters ie ¢, the maximum stress, when the cylinder iy under an internal pressure p;, is the tangential stress at the internal surface af tho thinnest part, if e < $a, and is of the magnitude 2G | at — Bao — 6) o-% | eS j Ate = 0, thie coincides with Eq. (47). 4 See, for instance, 8. ‘Timoshenko, “Slamglh uf Materials,” vol. 2, p. 236, 1041. "Prune. Ray. Sux. (lank), seaies A, vol. 9B1, p. 285, 1UYL. Soo also Brit, Assoc. Advancement Sei, Repls., 1921,TWO-DIMENSIONAL PROBLEMS IN POLAR COORDINATES 61 27. Pure Bending of Curved Bars. Let us consider a curved bar with @ constant narrow rectangular cross section’ and a circular axis bent in the plane of curvature by couples M applied at the ends (Fig, 42). ‘The bending moment in this ease is constant along the length of the bar and it is natural to expect that the stress distribution is the same in all radial cross sections, and that the solution of the problem can therefore be obtained by using expression (42). Denoting by a and b the inner and the outer radii of the boundary and taking the width of the rectangular cross section as unity, the boundary conditions are @ o = Oforr =aandr=6 @) [Perdr=0, — fowdr = —M (@) 8) re = 0 at the boundary Condition (1) means that the convex and coneave boundaries of the bar are free from normal forces; condition (2) indicates that the normal stresses at the ends give rise to the couple M only, and condition (3) indicates that there are no tangential forces applied at the boundary. Using the first of Eqs. (43) with (1) of the boundary conditions (a) we obtain A+ BC + 210g a) + 20 4 ® ft BU + 2 log d) +20 ' From the general discussion of the two-dimensional problem, Art. 15, it follows ‘that the solution obtained below holds also for another extreme case when the dimension of the cross section perpendicular to the plane of curvature is very large, ‘, for instance, in the ease of a tunnel vault (see Fig. 10), if the load is the same along the length of the tunnel.62 THEORY OF BLASTIOINY From (2) of conditions (a) we find Poort [ota or subsliluting for g its expression (42), we find -0 [¢ + Bib + % log b) + oan] - [4 + Bu + 2u lng a) + ava =0 © Comparing thie with (b), it is ey Lo ser that: (c) is satisfied, and the forces ul. the ends are reducible to a couple, provided conditions (J) are: cotisfied. ‘To have the leuding couple equal to Mt, the condition [oc ‘must be fulfilled. We have “oe ea [ oe ke aud wuling that om account of () late -9 [ae = or substituting oxpression (42) for (d) we find from (d), Alog? + BO log — oF loge) + CU 0%) - M ‘This cquation, together with the lwu Eys. (2), completely determines ihe eonstants A, B, C and we find = atigt pa — ee — ary N a N M n Cc ir [P? — a? + 2(67 lag b — a? log a)] whoro for simplicity we have pul N = (? — a}? — dats? (ioe " )TWO-DIMENSIONAL PROBLEMS IN POLAR COORDINATES 63 Substituting the values (7) of the constants into the expressions (43) for the wivews components, we find — 4 (ot oe bs tog? tat log & Sy log 2 ob bt log 5 + a toe) aay (- Mies + wre + ot og + — ot) (48) ‘This gives the stress distribution satisfying all the boundary eunditionst (a) for pure bending and represents the exact solution of the problem, provided the distribution of the normal forecs at the ends is that given by the second of Eqs. (48), If the forces giving the bending couple ace distributed over (he ends of the bar in sume other snanuer, Une sbrews distribntion at the ends will he difterent from that of the solution (48). But on the basis of Saint-Venant’s principle it can be concluded that the deviations from solution (48) are very small and may be neglected at large distances from the ends, say at distances greater than the depth of the bur. It is of practical interest to compare solution (48) with the elemen- tary solutions usually given in books on the. strength of materials. If the depth of the bar, 6 — a, is small in comparison with the radiue of the central axis, (6 + a)/2, the same stress distribution as for straight Dans iy usually assumed, If Uhis depth iy uol sunall iL iy usual in prac tice to assume that eross sections of the har remain plane during the bending, from which it can be shown that the distribution of the nor mal strcases o# over any cross scetions follows a hyperbolic law? In ull cases the maximum and minimum values of the stress og can be pre- sented in Whe form (a) ry, 1BRT The pmper, publistued iu Russias, ren ex, nnd the same problem way volved lator by M. , vol. 108, 1889, wl vol. 132, 1901) and by L, Prandtl, Soe A. Foppl, ‘“Vorlesungen aber toch- nische Mechanik,” vol. 5, p. 72, 1007; aleo A. Timpo, Z. Math, Phyeik, vol. 52, P. 348, 1005, * This approximate theory was developed by IL. Résal, Ann. mines, p. 617, 1862, and by E. Winkler, Zivdinyenirur, vol. 4, p. 282, 1858, se als his Ieok “Die Lebre von der Tastivitit- unl Fesligheit,” Chap. 18, Prog, 1867. FurUer develup- menkof Uw theory was snade by F. Grashof, “Blastisitat und Festigkeit,” p. 281, 1878, und by K, Pearson, “History of the Theory of Elasticity,” vol. 2, pt. 1, p. 122, 1893,64 YHEORY OF BLAGTICITY *the following table gives the values of the numerienl factor m ealou- Juled| hy the two elementary methods, referred to above, andl ly. the Converter m uv Ta. (h) Linear xiress| Hyperbolic stress istsibution diatribution| ‘Exget solution 3 oat | 479.08, —6L.27 | 478.05, —61.95 6.000 | | 7.72, — 4 ans | +7 vo 500 | + 2.285, — 1.095 | + = 1.130 enact formula (49).’ It ean be seen from this table that the elementary coltstion based on the hypothesis of plane cross sertions gives very accurale results, » a 7 8 dl Yl a wine at pechal antl 1668) 2 0 G2 ol ow rE oe 8 Values of 2 a zh La Max (1.0707 ae r ey BOY ot UW ft 13. rs 16 LT 1B 19 Valves of & Tre, 13. Te will be shown later that, in the ease of puro bending, the ross vweliona gotually do rewxin plang, and the diserepancy between the Momentury and the exact solutions comes from the fuel that m. tho Slementary eolution the sireso component 2, is neglected and it is The reaults are taken fem te doctorate thes, Uwlv. Michieany 1031, of Y. DilleviezTWO-DIMENSIONAL PRORLEMS IN POLAR COORDINATES 65 assumed that longitudinal fibers of the bent bar arc in simple tension ‘er compression, From the first of Eqs. (48) il can he shown that the atress ¢, is alwaya positive for the direction of bending shown in Fig. 42. The same ean be concluded at once from the direction uf siresses ag acting on the ele- ments n —n in Wig. 42. The corresponding tangential forces give resultants in the radial direction tending Wo separule longituclinal fibers and producing tensile stress in the radial direction, This stross increases toward the neutral surface and becomes a maximum wear Usis surface, This. maximum.is ulways much smaller than (ce)ax- For instance, for b/a = 1.3, (a+)nee. = 0.060(¢s)maz.3 for B/a ~ 2, (6,)uas U.1B8(o)mae$ for b/A = 3, (0, ua. 1.198 (o6)mes. In Wig, 42 the dis tribution of ¢e undo, for b/a = Zia given. From this figure wo ree that the point of maximum stress o, fs somewhat displaced from the seul axis in the direction of the ceuler of enrvature. 28. Strain Components in Polar Coordinates. In considering the displaccmont in polar courdinales lel us denute by u and » the compo- Tre, nents of the displuceueut in.the radial and tangential dircetione, respec- tively. . If w is the radial displacement of the side ad of the element abed (Fig, 44), the radial displacement of the side he is u + (@u/ér) dr. ‘Then the unit elongation of the clement abed in the radial direction is ou -® (49) As for tho strain in the tangential direction it, should be observed that it depends nol only on the displacement v but also on the radial dis- placement: w. -Assuming, for instance, that the puinl¥.o and d of fhe element abed (Fig. 44) have only Uhe radial displacement u, the new length of the are ad i8 {r+ u) dé and the tangential strain is therefore (+a —rde ou vide r ‘The difference in the tangential displacement of the sides ab and cd of [meat POLITERNE TimtSQARA |66 THRORY OF ELASTICITY the element abed io (80/00) do, and the tungerttial strain due to the dis- plicement » is accordingly 8¢/r 48, The total tangential strain is Unust (eur Considering now the shearing strain, let a’b'¢d’ be the position of the element abed after deformation (Fig. 44). The angle between the duection ad and a’d' 19 due to tho radial displacement w nud is eal te au/r 98. Ta Ue same manner the angle betwoon a’b’ and ab is equal to dv/ar, Tt chould be noted that only part of this angle (shaded in Uhe fignre) contributes to the shearing strain and the other purl, xual to 0/7, represents le angnlar displacement duc to rotation of the cle- ment abed ne a rigid body about the axis thrmgh O. Ience the total clings in the angle dab, reprosonting the shearing strain, is au, a ie 5 0 oat apr (61) Substituting now the expressions for the strain components (49), (50), (61) into the equations of Huoke’s Iaw,? ple ved er lor — 7) 2) YW = wwe can obtain sullicient ecnations for determining # and ». 29, Displacements for Symmetrical Stress Distributions. Sub ing, in the first of Hgs, (4) the stress components frum Figs 42, wo find Aya )A tn PAH 4 a0 — Boge + (1-3) B+ 20 vel By integr wn |- eaa *)Br log r — BCL ++ vr we obsiain +2cu — or] 4919 © {Pho eyrabol eg was usod with » diferent meaning ia Art, 10, it is assumed hors nik ws lave to do with plane etross, ie, that there imest a, perpeuivuls W Ure plane of the plato (G40 p11).TWO-DIMENSIONAL PROBLEMS TN POLAR COORDINATES 67 in which /(6) is a function of 0 only. Krom the second of Eas. (62), we find, by using Eq. 60, x _ 4Br a a) from which, by integration, o = ht f $0) d8 + Ful) ®) where fi(r) isa function ofr only. Substituting (a) and (6) in Eq. (51) ‘and noting that 7 is zero since 7,0 is zero, we find. 1M HOLY fg an- p00 eo from which fir) = Fr, f(#) = Hain @+ K cos 6 (a) where #', H, and K are constants to be determined from the con- ditions of constraint of the curved bar or ring. Substituting expres- sious (d) iuty Eys. (a) and (6), we find the following expressions for the displacements. «=3[- GAO $901 — Br logr — BU + 9)r + 2C — »| +Hesin @+4+ Keos@ (52) = ABO Bet Hen — Kin ig which the values of conslants 4, 12, and (! for each partienlar case should be substituted. Consider, for instance, pure bending. Taking the controid of the cross section from which @ is measured (Fig, 42) aud also un element of (he nulius ul this point, as rigidly fixed, the condi- tions of constraint are »=0, ® = otoro =O adr =r ott Applying these to expressions (53), we obtain the following equations for calculating the couslants of integruliou F, H, saul 2 Rquntion (¢) is satisfied only when f f() da iv tukon from (i) wittant. wn aaldie Live ronstont.8 THBORY OF ELASTICITY 3[- aes + 2(1 — ») Bro log ro — B(L + »)re $260 - om] +x =0 Fro+H =0 F=-0 From this it follows that F = H = 0, and for the displacement » we obtain 1Br8 _ K sino os ‘This means that the displacement of any cross section consists of a translatory displacement —K sin 6, the same for all points in the cross section, and of a rotation of the cross section by the angle 4B9/E about the center of curvature O (Fig, 42). We see that cross sections remain plane in pure bending as is usually assumed in the elementary theory of the bending of curved bars. In discussing the symmetrical stress distribution in a full ring (page 59) the constant B in the general solution (43) was taken as zero, and in this manner we arrived at a solution of Lamé’s f problem. Now, after obtaining expressions (58) for “isa\\, displacements, we see what is implied by taking B ' as zero. B contributes to the displacement v the term 4Br0/E, This term is not single valued, as it changes when we increase @ by 2r, i.c,, if we arrive at a given point after making a complete circle round. the ring, Such a many-valued expression for a dis- placement is physically impossible in a full ring, and so, for this ease, we must take B = 0 in the general solution (43). ‘A full ring is an example of a multiply-connected body, é.e., a body such that some sections can be cut clear across without dividing the body into two parts. In determining the stresses in such bodies we ‘asually arrive at the conclusion that the boundary conditions referring to the stresses are not sufficient to determine completely the stress dis- tribution, and additional equations, representing the conditions that the displacements should be single valued, must be considered (see page 118). The physical meaning of many-valued solutions can be explained by considering the initial stresses possible in a multiply-connected body. If a portion of the ring between two adjacent cross sections is cut out (Fig. 45), and the ends of the ring are joined again by welding or other Fre. 45,TWO-DIMENSIONAL PROBLEMS IN POLAR COORDINATES 69 means, a ring with initial stresses is obtained, i.e., there are stresses in the ring when external forces are absent. If a is the small angle meas- uring the portion of the ring which was cut out, the tangential dis- placement necessary to bring the ends of the ring together is bear © ‘The same displacement, obtained from Eq, (64) by putting @ = 2, is v= an ABE o Brom (¢) and (f) we find ak pat © ‘The constant B, entering into the many-valued term for the displace- ment (54) has now a definite value depending on the way in which the initial stresses were produced in the ring. Substituting (g) into Eqs. (f) of Art. 27 (see page 62), we find that the bending moment necessary to bring the ends of the ring together (Fig. 45) is we ee OT se ( ) oF ae = ay From this the initial stresses in the ring ean easily be calculated by using the solution (48) for pure bending. 30. Rotating Disks. The stress distribution in rotating circular disks is of great practical importance.’ If the thiekness of the disk is small in comparison with its radius, the variation of radial and tan- gontial stresses over the thickness can be neglected? and the problem can be easily solved.* If the thickness of the disk is constant Eq. (40) can be applied, and itis only necessary to put the body force equal to the inertia force. ‘Then (a) R= po'r @ +A complete discussion of this problem and the bibliography of the subject ean be found in the well-known book by A. Stodola, “Dampf- und Gas-Turbinen,” 6th ed., pp. 312 and 889, 1024. * An exact solution of the problem for a disk having the shape of a flat ellipsoid of revolution was obtained by C. Chree, see Proe, Roy. Soc. (Landon), vol. $8, p. 39, 1805. Tt shows that the difference between the maximum and the minimum stress at the axis of revolution is only 5 per cent of the maximum stress in a uniform disk with thickness one-eighth of its diameter. * A more detailed discussion of the problem will be given Inter (see Art. 119). “The weight of the disk is neglected,77 VHKORY OF ELASTICITY where ¢ is the mass per unil volume of the material of the disk and w the augulur velocity of the disk. Equation (40) can then be written in the form Spire) set utr = ® "This equation ia catiefied if we desive the stress components from a stress function F in the following manner: toa F, = E+ pot? @ ‘The strain components in the case of symmetry are, from Eqs. (49) and (50), du. = gat oar Fliminating u betwoon these equations, we fiud dep ecebr =0 (a Substituting for the strain compunents their expressions in terms of the stress components, (5), and using Eqs. (0), we find that Une stress function F should satisfy the following ecmation: ee EF 8+ em a) © Tk can be verified by substitution that the yeneral solution of this equation is Fm Cr $ Ch 3B pes o unl fram Mqs. (¢) we find on Ca TE putt ) oat ‘The integration constants C unl C1 are determined from the boundary egnditions. For a eolid disk we must tuke C's = 0 since otherwise the stresses (9) become infinite nt the center. The constant C is determined from the condition at the periphery (r = b) of the disk. Lf thero are no forcesTWO-DIMENSIONAL PRODLEMS IN POLAR COORDINATES 71 applied there, we have (des = 0-3 E* puttt = 0 from which BY yt and the slress components, from Eqs, (9), are a= See putt — rf) 5) BY yy 143 0p = SEP ptt — LAB pus ‘These stresses are greatest a1. he center of Uhe disk, where oe 0 = SEF pat (56) In the case of a disk with # nireudur Jule of radius u at the eenter, the constants of integration in Eqs. (¢) arc obtained {rom the cond at the inner and onter honndavies. Tf there ure uo forces acting on these boundaries, we have des = 0 th) (lle = from which we find thal. BEE pokey, a BL Substituting in Ey. (), v atl waft n(n 8) . , 7) oboe (e gee ge Lt, ‘We find the maximum radial stress al + = Gb, whero (dane, ~ 2 putt — a)? (58) ‘Tho maximum tangential stress is at the inner hounilary, where B+ fd nn = SEP ( +H) (60) Tt will be seen thai this sires is larger than (7,)12 THEORY OF HLASTICITY ‘When the radius a of Us: hole approaches acro, tho maximum tan- ential stress approaches a value twice ws yreat.as that for a golid diste (66); ie, ly making a small circular hole at the center of » solid rotating disk we doubly she maximum stress, ‘This phenomenon of stress concendration at @ hole will be discuswal Inter (see page 78). Assuming that the stresses do not vary over the thickness of the disk, the method of analysis developed above for disks of constant thickness ran be extended aloo to disks of variable Uiicknexs. This the thickness of the disk, varying with radius r, the equation of equilibrium of such an clement as shown in Fig. 40 is EZ hirvg) — he + hawt? — 0 ® ‘Thin equation is satisfiod by putting ne where F is again a stress function. ing, these expressions for the stress components into the com- -y equation (d) we arrive at the following equation for the stress ae anys — 208 (0F _ yp) — =F (8 + pute ra. e *) 0 ® In this manner the problem of finding the stress distribution in a disk ‘of vuriuble thickness is reduced to the solution of Ey. (f). Tn the particular case where Uw thickness varies according to the equation h-cr (m in which C 4s 9 constuul aml m any number, Eq, @ oan eaally be integrated.’ ‘The goncral solution bas the form Fes mrt Art | Brt _ 8+ rwte (en + Sn 1 8) while « and @ are the rools uf Uhe quadratic equation in which atone ton —1=0 ond A and B are integration cimstants which are determined from the boundary conditions. This case was invectigntod by Stodola, foe. eff Sr whw TL, Holzer, 4. gon Purbinorvesen, 1V19.TWO-DIMENSIONAL PROBLEMS IN POLAR COORDINATES 73 ‘A good approximation to the actual shapes of rotating disks can be obtained by dividing the disk into parts and fitting approximately to each part a curve of the type (m).! ‘The case of a conical disk has been discussed by several authors.? Very often the calculations are made by dividing the disk into parts and considering each part as a disk of constant thiekness.? “31, Bending of a Curved Bar by a Force at the End.‘ We begin with the simple case shown in Fig. 46, A bar > of a narrow rectangular cross section and Og——s % with a circular axis is constrained at the lower : Be | end and bent by aforce P applied at the upper end in the radial direction. The bending moment at any cross section mn is propor tional to sin @, and the normal stress os, Ss according to elementary theory of the bending of curved bars, is proportional to the bending moment. Assuming that this holds also for the exact solution, an assumption which the results will justify, we find from the second of Eqs. (38) that the stress function ¢, satisfying the equation Fro. 48, #10, 18) (8% 16, 18%9' ant tar 43) tart rae) = 9 @ should be proportional to sin @. Taking $= f(r) sin 6 ) and substituting in Eq. (a), we find that f(r) must satisfy the following ordinary differential equation: +See M. Griibler, VD. vol. 50, p. 585, 1906. +See A. Fischer, Z. oesterr. Ing. Arch. Vereins, vol. 74, p. 46, 1922; HM. Martin, Engineering, vol. 115, p. 1, 1928; B. Hodkinson, Engineering, vol. 116, p. 274, 1928; K, E. Bisshopp, J. Applied Mechanics (Trans. A.S.M.E.), vol. 11, p. AL, 1944, * This method was developed by M. Donath; soe his book, “Die Berechnung rotierender Scheiben und Ringe,” Berlin, 1912. It is described in English by 1H. Hearle in Engineering, vol. 106, p. 121, 1018, A further devclopment of the method was given by R. Grammel, Dinglers.Polylech. J. vol. 338, p. 217, 1923. The case when material docs not follow Hooke’s law was investigated by M. Grobler, V.D.I., vol. 41, p. 860, 1897, and vol. 44, p. 1157, 1900. See also H. Schlechtweg, Z. angew. Math. Mech., vol. 11, p. 17, 1981, and IngenieurArchit vol. 2, p. 213, 1981. +H. Golovin, loc. ot, eee INsTITUTUL POLITENNEC |, Tintis OAaRA BIBLIOTECA CENTRALA“u THBORY OF KLASTICITY @ id _1\ (Mla _ fy. a! pdr years wn) 79 @ This equation ean be transformed into 2 lineur differential equation ith constant coefficients (sr page 58), and its general solution is fit) < AP +B q +r + Dr log r @ in which A, H, 0, and D are constants of integration, which arc deter- mined from the boundary conditions, Substituting sululion (7) in expression (6) for tho streas funetiva, and using the gencral formulas (38), we Gil the: following expressions for the stress eormpanents: Lag, 10% _ _ 2B, DP). we BEE MS = (nar oF +2) sin 0 oy = Sha (Har tH 0) sino (60) a flay QR m nee 2 (18 (er -B4F cos # From Ue conditions that the outer and inuer luuncaries of the curved bar (Fig. 46) are free frum external forces, wo require that oa ta = Oforr = aund r= b or, from Egg. (60), ada - 28 42 mo BoD © 24d - Fe +5 = 0 ‘The last coudition is that the cum of the shearing forces cover the upper end of the Lur should equal the foree P. Taking the width of the eross section as unity—or P us Ue load per unit thickness of the plate—we obluin for 0 — 0, ° ba (ag [ore -- [EGR in a = [ars Ho Diog si =P or age 09 + 2 ESP — Dog? “ From Eqs. (¢) and (f) we find P Pa" A=2F, bee p=-Fwin @ 2NTWO-DIMENSIONAL PROBLEMS IN POLAR COORDINATES 78 in which Naat Bt (+09 log? Substituting the values (g) of the constants of integration in Eqs. (60), we obtain the expressions for the stress components. For the upper end of the bar, @ = 0, we find a=0 1 -f [r+ Leto] ® For the lower end, 9 = x/2, te = 0 ® on Flor wr +093] ‘The expressions (60) constitute an exact solution of the problem only when the forces at the ends of the curved bar are distributed in the 20) a} +} Fra. 47. manner given by Eqs. (4) and (k). For any other distribution of forces the stress distribution near the ends will be different from that given by solution (60), but at larger distances this solution will be valid by Saint-Venant’s principle. Calculations show that the simple theory, based on the assumption that cross sections.remain plane dur- ing bending, again gives very satisfactory results. In Fig. 47 the distribution of the shearing stress zy» over the eross section @ = 0 (for the cases 6 = 3a, 2a, and 1.34) is shown. The abscissas are the radial distances from the inner boundary (r = a). ‘The ordinates represent numerical factors with which we multiply the average shearing stress P/(b — a) to get the shearing stress at the16 THEORY OF ELASTICITY point in question. A value 1.5 for this factor gives the maximum shearing stress as calculated from the parabolic distribution for rectangular straight beams. From the figures it may be seen that the distribution of shearing stresses approaches the parabolic distribution when the depth of the cross section is small. For such proportions as are usual in arches and vaults the parabolic distribution of shearing slress, as in straight rectangular bare, can be assumed with sufficient accuracy. Let us consider now the displacements produced by the force P (Fig. 46). By using Eqs. (49) to (52), and substituting for the stress components the expressions (60), we find du _ sin 6] , 2B D err [zara - 9 - 2a tn +2a-»| vo ere oO au a _e = yt GO F From the first of these equations we obtain by integration i [ara = 3) +30 +9) + DU ~ ») log ‘| + $06) (m) where f(0) is a function of @ only. Substituting (m) in the second of Eqs. () together with the expression for ey and integrating, we find o= - 9! [ane + Ra |») = Dog r(t = ») +pa-»| - fro d6+F(r) (n) in which F(7) is a function of r only. Substituting now (m) and (n) in the third of Eqs. (2) we arrive at the equation AD cos @ ‘c fro 40+ s'@) + 1P'@) — FW = ‘This equation is satisfied by putting FQ) = Hr, 100) = —*Pocoso+Ksino+Lcos0 (0) in which H, K, and L are arbitrary constants, to be determined from the conditions of constraint, ‘The components of displacements, fromTWO-DIMENSIONAL PROBLEMS IN POLAK VOURDINATES 77 (m) and (n), are then = = Poco 9+ 828| 1 — 5) eye +40 ~ ant +r | Ksin 0+ Leos 0 (@) 2D pin p 2088 are Bute) v= Basin a S58! A + sys + BOD P+ Ze? DO = tog 4 cos 0 + K cos 0 ~ Lsin 0+ Hr ‘The radial sle(lectivn of Uae upper end of the bar is obtained by putting @ — O in the expression for u, which gives (uaa = 1 ) ‘The constant E ix ublained from the condition at the built-in end (Fig. 46), For # = 3/2 we have v = 0; du/dr — 0, hence, from Uhe second of Fys. (y), Ds H-Q L-® ) ‘The deflection of the upper end is, therefore, using (g), abr Pala +) (u)one = FF . a (a1) E\ (a? b4) | (a +6) log 5 Tho application of this formula will be given later. When b appronches a, and the depth of the curved bus, h = 6 — a, {a small in comparison with a, we can uee tho expression Substiluting in (61) and neglecting small terms of higher order, we obtain: 3raeP (an = - Se? which coincides with the elementary formula for this ease." By taking the stress function in the form = Ilr) cos @ Seo 8. ‘Timoshenko, “Btrength of Maurinly,” vol. 2, Art 18, 10417B THEORY OF ELASTICITY and proceeding as above, we get a solution for the case when a vertical force and a couple are applied to the upper end of the bar (Fig. 46). Subtracting from this solution the stresses produced by the couple (see ‘Art, 27), the stresses due to a vertical foree applied at the upper end of the bar remain. Having the solutions for a horizontal and for a vertical load, the solution for any inclined force can be obtained by superposition. In the above discussion it was always assumed that Eqs. (¢) are satisfied and that the circular boundaries of the bar are free from forces. By taking the expressions in (¢) different from zero, we obtain the case when normal and tangential forces proportional to sin @ and cos are distributed over circular boundaries of the bar. Combining ‘such solutions with the solutions previously obtained for pure bending and for bending by a force applied at the end we can approach the condition of loading of a vault covered with sand or soil. 82, The Effect of Circular Holes on Stress Distributions in Plates. Figure 48 represents a plate submitted to a uniform tension of magni- pittittt ly oT Tay WT Fro. 48, tude Sin thez-direction, If a small circular hole is made in the middle of the plate, the stress distribution in the neighborhood of the hole will bbe changed, but we can conclude from Saint-Venant’s principle that the change is negligible at distances which are large compared with a, the radius of the hole, Consider the portion of the plate within a concentric circle of radius », large in comparison with a, ‘The stresses at the radius b are effec- tively the same as in the plate without the hole and are therefore given by (ena = S cos? 6 = 48(1 + cos 26) @ (ros)rab = — $5 sin 26 1 Several examples of this kind were diseussed by Golovin, lo, cit, and Ribitre, doe, eis, and Compl. rend, vol. 132, p. 315, 1901.TWU-DIMENSIONAL PROBLEMS IN POLAR COORDINATES 79 ‘These forces, acting around the outeide of the ring having the inner and outer radii 7 = a und r = b, give » stress distributiun within the ring which we may regard as consisting of two paris. ‘The first is due to the constant component 4S of the normal forces. ‘The stresses if prorhices can be caleulated by means of Eqs, (45) (page 59). The remaining part, consisting of the normal forces 4S cos 26, together with the sheuriuy forces — $Y sin 26, produces sliesses which way be derived from a stress function of the form — f(r) cos 26 ®) ‘Substituting this into the compatibility equation a la 1@ ap Lag Lag (B+th+ Ah) (Fe t+ 328) —0 we find the following ordinary differential equation to determine f(r): @ oid af ldf af\_ (G+ 1d -4)(B+iZ-4 ° ‘The general solution is Sir) = Ar? + Brt + ch +D "The stress function is therefore o~ (44 Bet ch +p) oon a8 @ aul (he corresponding stress components, from Eqs. (38), are 1p 10% (4, 60 wD) = oat Ra TO (ra + 85 +22) con 20 , on = 88 = (24 + 12574 + 8) com 20 @ a (1 ae! 6C _ 2D) = 2 (888) (a4 +e #622) on The constants of integration are now to be determined from conditions {a) for the onter houndary aud from the condition Unel Uke edge of the hole is free from external forces. ‘These, conditions give80 THEORY OF ELASTICITY oc, 4D_ 1 2a 4+ 0+ = 38 24 +004 Pao ate +60 _ 2D 2A + 6Bb — — Fe 6C 2D 2A + 6Bat — 27 — => @ Solving these equations and putting a/b = 0, i.c., assuming an infinitely large plate, we obtain A=-‘ Substituting these values of constants into Eqs. (d) and adding the stresses produced by the uniform tension $$ on the outer boundary calculated from Eqs. (45) we find™ 2) Hf ris very large, o- and zr approach the values given in Eqs. (a). At the edge of the hole, r = a and we find or = 79 =0, 0 = S — 28 cos 20 Tt can be seen that ¢ is greatest when @ = 4/2 or 0 = 31/2, Z.e., at the ends m and n of the diameter perpendicular to the direction of the tension (Fig. 48). At these points (o7)mx = 38. This is the maxi- mum tensile stress and is three times the uniform stress 8, applied at the ends of the plate. ‘At the points p and g, @ is equal to 7 and 0 and we find, -8 o co that theré is a compression stress in the tangential direction at these points, ‘This solution was obtained by Prof. G. Kirsch; see V.D.L., vol. 42, 1898. Tt has been well confirmed many times by strain measurements and by the photor dlastic method (gee Chap, 5 and the books cited on p. 181).TWO-DIMENSIONAL PROBLEMS IN POLAR COORDINATES 8% For the cross section of the plate through the center of the hole and perpendicular to the z-axis, @ = 7/2, and, from Eqs. (62), 2g 0 = 0, a=5 24S 43%) Ibis evident that the effect of the hole is of a very localized character, and, as r increases, the stress c+ approaches the value $ very rapidly. ‘The distribution of this stress is shown in the figure by the shaded area. ‘The localized character of the stresses around the hole justifies the application of the solution (e), derived for an infinitely large plate, to a plate of finite width. If the width of the plate is not less than four diameters of the hole the error of the solution (62) in calculating (01)axs. does not exceed 6 per cent. Having the solution (d) for tension or «+4411 | compression in one direction, the solution for tension or compression in two perpen- m dicular directions can be easily obtained by ae ewe L. superposition, By taking, for instance, tensile stresses in two perpendicular diree- tions equal to S, we find at the boundary of the hole a tensile stress co = 28 (seo * = page 72). By taking a tensile stress S in x the z-direction (Fig, 49) and a compressive Fie 49. stress —§ in the y-direction, we obtain the case of pure shear. The tangential stresses at the boundary of the hole are, from Eqs. (62), og = S — 28 cos 26 — [S — 28 cos (20 — x)] For 6 = x/2 or 6 = 31/2, ic., at the points n and m, we find «9 = 48. For @ = 0 or 6 ==, ie., at n; and my, oo = —48. Hence, for a large plate under pure shear, the maximum tangential stress at the boundary of the hole is four times larger than the applied pure shear. The high stress concentration found at the edge of a hole is of great Practical importance. As an oxample, holes in ships’ decks may be mentioned. When the hull of a ship is bent, tension or compression is produced in the decks and there is a high stress concentration at the holes. Under the cycles of stress produced by waves, fatigue of the metal at the overstressed portions may result finally in fatigue cracks.* ‘See 8. Timoshenko, Bull. Potyteck. Inst., Kiew, 1907. We must take S equal to the load divided by the gross area of the plate. See paper by T. L. Wilson, The 8.8. Leviathan, Damage, Repairs and Strength Analysis, presented at 2 meeting of the American Society of Naval Architects and Marine Engineers, November, 1930.82 THEORY OF RLASTICITY It ie often nevemary to reduce the stress concentration ut holes, sich us aocess holca in airplane wings xnd fnselages. ‘This can be done by odding a beu! or reinforcing ring? ‘The analylicul problem has been solved by extending the mela employed for the hole, and the results have been compared with strain-gauge meusurements.” “Phe caco of a circular hole near the atraight boundary of u semi- infinile plate under tension parallel Ww this boundary (Fig. b0) wae analyzed by G. B. Jeffery.? A corrceted roeult:and s eomparinim with pholurlastic tests (ece Chap. 5) were given later by R. D. Mindlin.* Fae 0, "Tho stress at the hole, ul Uw: point n ncarcot tho edge, becomes a very urge smnltiple of the undisturbed tensile sires when mn io smell eom- parod with np. G. B. Jeffery also investigated the ease uf a nmiform, normal pressure p. acting on the boundary of the hole, This is a problem of practical jnterest. It gives the etrosees near rivet, hole while tho hot plastie rivot is being foreed ume under pressure. If the hole is very far from the straight edge tho stresses at (he lunndary of the hole, from Eqs. (46) (page 60), wre = Py If the hule is near the straight edge, the (ungential stresses are no longer constant along the Laundary of the hole, The maximum tan= gentiud slress is at the points k and J und is given hy the formula _,vte (elas. = Bi (63) 1 Ge 8. Timoshenko, J. #ranklin Inst., vol. 197, p. 50%, 10845 aloo 8. Timoshenkn, “Abreugth of Matorieds,” 2d o4., vol. 2, 1» 817. °§, Levy, A. E, McPherson, and F.C. Smith, J. Applied Mechanirs (Truns. A.SALE), vol. 15; p- 360, 104%. References to prior work may Tw foul in thir ‘paper. } Trans. Roy. Soc. (Lanlin), series A, vol. 221, p. 200, 1921. ‘Proc. Sar. Rxpl. Gtiess Anaiyois, vol. 5, p. 88, 1948TWO-DIMENSIONAL PROBLEMS IN POLAR COORDINATDS 88 ‘This stress should be compared with the tensile stress at the point on the straight edge of the plate, given by the formula feet oe (64) For d =r V3, the two expressiuus have the same magnitude. If d is greater than this the maximum stress is at the circular boundary, and if it is less, the maximum stress js at the point m. ‘The case of « plate of finite width with a sirenlar hole on the axis of sywmetry (Fig. 01) was discussed by R. C. J. Howland! He found, for instance, that when 2r = 4d, 09 = 4.38 ab the point n and oe — 0.758 ‘at the point m. ‘Phe method used in this article for + analyzing stresscs round a small cir- cular hole can be applied when the plate is subjected to pure bending,* ‘The cases of a row of circular hales in an intinite plate,*4* a row of holes in a strip, and in o semi-infinite plate,? and a ring of holes in a plate? (under all-round tension) have also been investigated. A method devised by Iengst has heen applied to the case of a hole in a. square plate’ under equal tension in both directions, and under shear* when the hole is plain or reinforced. Holntions have heen obtained for the infinite plate with a cirenlar hole when forces are applicd to the boundary of the hole,**for the corre- 1R.C, J. Howland, rans, Hou, Soe, (London), Series A, vol. 229, v.49, 1020. *&, Lusi, Phil, lag, Vebruary, 1980, p. 210; also Sei, Papere Inet, hue, Chem, Resoarch (Yokyo), vol. 8, p. 65, 1928, ‘Lhe corresponding problem for an elliptical holo was eolved earlier by K. Wolf, 3. tech. Physik, 1929; p. 160, ‘The circular hole in a atrip is discussed by R. C. J. Howland and A. C. Btevenson, Trans. Roy. Soe. (London), series A, vol. 232, p. 155, 1993, A proof of convergence of the series solutions is given hy TO. Knight, Quut. J. Muth, Qafmuid series, wol. fi, p. 255, 1934, #M, Badowsky, Z.amyrn. Muth. Meck, vol. 8, p. 107, 1928, * HO. Towwhan, Proc. Rny, Soc (Fanulim), series A, val 148, p. AT1, 1985: °K. J. Bohuls, Proc. Neer, Akt. wun Welrnsrhuppen, vl. 48, pp. 288, $41, 487, 4nd 624, 1942, val. 48, pp. 282 nnd 292, 1948. °C, B. Ling, P. &. Wang, and K. 8, Jing, Teck. Rept. No. 9, Bur. Aeronaut, Revoarch, Chengia, China, Fan, 1944, 1, B Ting und P. 8. Wang, Pech, Rept, No. 6, #bfd., June, 1943, H. Henysl, 7 angew. Math. Mech, vol. 18, p. 44, 1938. * C.K. Wang, J. Applied Mechanics (Trans, A.S.M.E,), vol. 13, p. A-77, 1046, “W.G, Rickley, Trans, Roy. Soc. (London), serles A, vol. 227, p. 383, 1928, Fin Mt84. THEORY OF ELASTICITY sponding problena of the attip," and for a row of holes parullel nnd near to tho etraight edge of a semi-infinite plate® (row of rivet holes). Tt an elliptical hole is made in an infinite plate under tension 8, with ono of the principal axes parallel to the tension, the stresces at the ends of the axis of the hole perpendicular to the direction uf the tension arc ona(ieng (65) where 2a is the axis of Une ellipse perpendicular to the tension, and 2b js the other axis, Thio and other problems concerning ellipses, hyper~ polos, and two circles aim cliscnssed in Chap. 7, whore references will be found. ‘A yory slonder hole (u/l lurge) perpendicular to the direction of the teusion canses a very high strees concentration.? ‘This explains why cracks transverse to upplied| forces tend to spread. The spreading can be stopped by drilling holes at the ends of the exuck ta eliminate the sharp curvature responsible for the high stress concentration, ‘When a hole is filled with material which is rigid or has elastic con- stanta different from (hus of the plate (plane stress) or body (plane struin) itself, we have the problem of the rigid ur elastin inclusion, “Vhis ha boen solved fur cirenlart and elliptic inclusions.’ ‘The results for ihe rigid circnlar inclusion have been confirmed hy the photoelastic method® (see Chap. 5). “The stresses given by Eqs. (62) for the problem iudiesled hy Vig. 48 are the same for plane struin ax for plane stress. In plane strain, how- ever, the axial stress 2 = wt + ue) must act on the plane ends, which are parallel wv the 2y-plane, in order to make ¢ ero. Reiuvul of these stressc3 from the ends, to arrive al the eondition of free ends, will produce further stress whieh will not be of a two-dimensional (plane stress or planc strain) character, If #7. O. Knight Phils Mag, gerbes 7, vol. 19, p. 617, 1935. 20, B, Ling and M. C. Hau, Tech, Repl No. 10, Bur, Acronaut, Research, Chengtu, Chino, February, 1948. "the problowa of a narzow sin. wx discussed by M, Badowsky, Z. angew. Math Mech., Vol. 1U, P. 17, 1930. O16. Serawa and 'G Nishinmua, Repl. Acronaut. Revearch Inst, Tokyo Tap. Unie., vol. 6, no. 28, 1981, T.N. Goodier, Urano, 4.8... vol. 88, 39 (1939) TL. H. Damell, “'Thendore von Kérinén Anniversary Volume,” p. 290, Pasa- dona, 1941 ¢W. F. Thibodeau and Lo 1p. 898, 1938, Wood, J, Research Nuil, Rur. Standards, vol, 20,TWO-DIMENSIONAL PROBLEMS IN POLAR COORDINATES 85 the hole is small in diameter compared with the thickness between the ends, the disturbance will be confined to the neighborhood of the ends, But if the diameter and the thickness are of the same order of magni- tude, the problem must be treated as essentially three-dimensional throughout. Investigations of this kind" have shown that oe remains the largest stress component and its value is very close to that given by the two-dimensional theory. ‘33. Concentrated Force at a Point of a Straight Boundary. Let us consider now @ concentrated vertical foreo P acting on a horizontal Fra, 52, straight boundary AB of an infinitely large plate (Fig. 52a). The dis- tribution of the load along the thickness of the plate is uniform, as indi- cated in Fig. 52b. The thickness of the plate is taken as unity so that P is the load per unit thickness, The distribution of stress in this ease is a very simple one* and is called a simple radial distribution. Any element C at a distance r from the point of application of the load is subjected to a simple compression in the radial direction, the radial stress being 2P cos 6 oe (66) 1A. E, Green, Trans, Roy. Soc. (London), series A, vol. 198, p. 229, 1948; E, Sternberg and M. Sadowsky, J. Applied Mechanics (Trans, A.8.M.E.), vol. 16, P. 27, 1949. * The solution of this problem was obtained by way of the three-dimensional solution of J. Boussinesq (p. 362) by Flamant, Compl. rend., vol. 114, p. 1465, 1802, Paris. ‘The extension of the solution to the case of an inclined force is due to Boussinesq, Compt. rend., vol. 114, p. 1510, 1892. See also the paper by J. H. Michell, Proc. London Math. Soc., vol. 82, p. 35, 1900. The experimental investi- sation of stress distribution, which suggested the above theoretical work, was done by Carus Wilson, Phil. Mag., vol. 82, p. 481, 1801,386 THEORY OF ELASTICITY ‘The tangential stress 09 and the shearing stress r,s are zero. Tt is easy to sce that these values of the stress components satisfy the equations of equilibrium (37) (page 56). ‘The boundary conditions are also satisfied because ¢» and 1,9 are zero along tho straight edge of the plate, which is free from external forces except at the point of application of the load (r = 0). Here, becomes infinite. ‘The resultant of the forces acting on a cylindrical surface of radius r (Fig. 528) must balance P. It is obtained by summing the vertical components o,r dé cos @ acting on each element r d@ of the sur- face. In this manner we find 2 feast 040 = P ‘To prove that solution (66) is the exact solution of the problem we must consider also the equation of compatibility (89). The above solution is derived from the stress function 3 off co cos 8-78 ¢ -2r sin @ @ ‘We can verify this by using Eqs. (88) as follows: 1a¢ 4 19% _ _ 2P cose . or For Oe - oT ee . FF <9 6") ‘which coineides with solution (66). Substituting the function (a) into Eq, (39), we can easily show that this equation is satisfied. Hence, (a) represents the true stress function and Eqs. (66') give the true stress distribution. Taking a circle of any diameter d with center on the a-axis and tan- gent to the y-axia at O (Fig.-52a), we have, for any point C of the cirele, dcos @ =r. Hence, from Eq. (66), 2P od ‘ie, the stress is the same at all points on the circle, except the point 0, the point of application of the load. Taking a horizontal plane mn at a distance a from the straight edge of the plate, the normal and shearing components of the stress on thisTWO-DIMENSIONAL PROBLEMS IN POLAR COORDINATES 87 plane ul any point Af (Fig. 52a) are caloulated from the simple com- pression in the radial direction, a, = on cost gm — COO OP aoe g 7 7a 2 sin? 0 ens” 0 (67) Jn @ eos? tay ~ v-sin 8 005 9 ~ — 2P Sin 8 008? OL =? gin dcont 0 7 za oy ~ or sin? @ = — = In Fig. 58 (he distribntion of stresses ¢+ and ray along the horizontal plune mn is represented graphically. At the point of application of the load the stress is theoretically infinitely large because a finite foree is acting on an infinitely small - A Py x oe 1 lo, js S| x J 4. Ze Ty 2 na mn ln lag bar in x Fis, 88 area, In practice, at the point of application there is ulways a certain yielding of material and as a resull of this the load will become dis- tribuled over 4 fimile area. Imagine that the portion of material which snffered a, plastic flow is cut out from the plote by a circular cylindrical surfacc of small radius as shown in Fig. 52). Then the ‘uations of elasticity can be upplied Lo the remaining pertion of the plate. An analogous solution can be obtained for a horizontal force P upplied to the straight boundary of the eomi-infinite plate (Fig. 54). ‘The stress components for this case ure oblained from the same bigs. (GH); it is only necessary Lo measure the angle 0 from the dircetion of the force, as shown in the figure. By caleulating the resultant of the forces acting on o cylindrical surface, shown in Fig. 54 by the dottedRe + PHEORY OF ELASTICITY lino, wo find - a cost odo = —P ‘This resultant balances the external furce P, and, as the stress compo- nents 7,» aud ay ut the straight edge are zero, solution (86") rulisfies the boundary conditions. Having, the solutions for vertical and horizontal concentrated forces, solutions for inclined forves ure obtained by superpsition. Resolving the inclined force / into two components, P eux « vertically and P sin Fro, 54, a horizontally (Hig. 66), the radial stress at any put (68), o= BP con con 1 P dna os(5 +°)| ar = ew (+o) (68) ‘Hence Tiqs. (66°) can be used for any direction of the furee, pravided in cach ease We measure: the angle 0 from the direction of the force. “The stress funtion (a) may be weed alco in tho ease when a coup ix acting on the wimight houndary of an infmito plate (Fig. 582). Tt is emy la see that the nlwen function for the eace when the tensile force P isnt the pint Or, aa distanes from the origin, 9 obteined from. ¢, Eq. (@), egaried fr Use inoment as a function. fof cand yinatond of rand 2, by writing y instead of yand also Hinatoad of P. ‘Tuo and the origins! stress function @ rau be corabined, and wo then obtatn the stzoee function for the Ln eal mud opposite forces applied at O and O,, inthe form: day + 0) + oe) When a is vory small, this appronches the vilus att oro at o (a) in Eq. (b), and noting (eco page 57) thet te 28 28 e009 Ba Guet ee