a | Chapter 8 | DEFLECTION OF BEAMS BY INTEGRATION INTRODUCTION We saw in Sec. 44 that a prismatic beam subjected to pure bending is bent into an are of circle and that, within the elastic range, the curvature of the neutral surface may be expressed as 1_M ain (421) where M is the bending moment, E the modulus of elasticity, and I the A moment of inertia of the cross section about its neutral axis. 4 on ‘When a beam is subjected to a transverse loading, Eq. (4.21) remains valid for any given transverse section, provided that Saint-Venant's princi- ple applies. However, both the bending moment and the curvature of the ) neutral surface will vary from section to section. Denoting by x the dis- tance of the section from the left end of the beam, we write of 1 _ M(x) oF Consider, for example, a cantilever beam AB of length L. subjected to a concentrated load P at its free end A (Fig. 8.1a). We have M(x) = —Pe and, substituting into (8.1), | nes ° ° which shows tha the curvature of the neutral surface varies linearly with x Fee from zero at A, where p, itself is infinite, to —PL/EL at B, where loe| = EL/PL (Fig, 8.16), Consider now the overhanging beam AD of Fig. 8.2, which supports two concentrated loads as shown. From the free-body diagram of the beam (Fig. 8.30), we find that the reactions at the supports are Ry = 1kN and Re =5KN, respectively, and draw the corresponding bending-moment 396ee diagram (Fig, 8.3). We note from the diagram that M, and thus the curva- ture of the beam, are zero at both ends of the beam, and also at a point E located at x = 4m. Between A and E the bending moment is positive and the beam is concave upward; between E and D the bending moment is negative and the beam is concave downward (Fig. 8.3c). We also note that the largest value of the curvature (le. the smallest value of the radius of ‘curvature) occurs at the support C, where |M| is maximum. ‘From the information obtained on its curvature, we may get a fairly 00d idea of the shape of the deformed beam. However, the analysis and design of a beam usually require more precise information on the deflection and the slope ofthe beam at various points. Of particular importance is the knowledge of the maximum deflection of the beam. In this chapter we shall use Eq, (8.1) to obtain a relation between the deflection y measured at a given point Q on the axis of the beam and the distance x of that point from some fixed origin (Fig. 8.4). The relation obtained is the equation of the elastic curve, ie. the equation of the curve into which the axis of the beam is transformed under the given loading (Fig, 8.4b). 8.2, EQUATION OF THE ELASTIC CURVE We frst recall from elementary calculus that the curvature ofa plane ceurve at a point Q(x y) of the eurve may be expressed as dy ee (82) where dy/dx and d?y/ds® are the first and second derivatives of the func- tion x) represented by that curve. But, in the case of the elastic curve of, a beam, the slope dy /dr is very small, and its square is negligible compared to unity. We may write, therefore, 63) (84) The equation obiained is a second-order linear differential equation; itis the governing differential equation for the elastic curve. ‘teshouldbe noted that, inthis chapter and the next, y represents a vertical displacement, ‘while twas uted in previous chapters to represent the distance ofa given pont ina transverse ‘ection from the neutral ais of that section,998 VECHANICS OF MATERIALS ‘The product Elis known as the flexural rigidity and, if it varies along the beam, asin the case of a beam of varying depth, we must express it as 1 function of x before proceeding to integrate Eq, (8.4). However, in the cease of a prismatic beam, which is the case considered here, the flexural rigidity is constant. We may thus multiply both members of Eq, (84) by EI and integrate in x. We write ages M(x)ae + C, (85) where C, is a constant of integration. Denoting by (2) the angle, meas- | ured in radians, that the tangent at Q to the elastic curve forms with the | horizontal (Fig. 85), and reealling that this angle is very small, we have 4 ante Tho eny we Ea, (5) tho tera fora B10(e) =f" M(a)de +6, es) Pepin rca ty = f [fo mayar+ o]ae +c | | | | | | | | m=] em] Smear a ee i n= 0 va =0 = 2 (@ Simpy supported beam By =f def Miya + Ox +O 66) , 4, 2 ie where C, is a second constant, and where the first term in the right-hand a — member represents the function of x obtained by integrating twice in x the bending moment M(z). If t were not forthe fact that the constants Cand Cy are as yet undetermined, Eq, (8.6) would define the deflection of the )Overhanaing bem slope of the beam at Q. ‘The constants C, and C, are determined from the boundary conditions : or, more precisely, from the conditions imposed on the beam by its sup- ports. Limiting our analysis in tis section to statically determinate beams, ~. ie, to beams supported in such a way that the reactions at the supports hee tay be obtained by the methods of statics, we note that only thre types of s beams need to be considered here (Fig, 6): () the simply supported beam, (¢)Cantfeverbeun (©) the overhanging Beam, and (c) the cantilever beam. Fig. 8.8 Boundary conditions fr satay Inthe fist two cases, the supports consist ofa pin and bracket at A and determinate beams. of a roller at B, and require that the deflection be zero at each of these points, Letting frst x= xy, y = yq = 0 in Eq, (86), and then x = 4p, 7 = Yp = 0 in the same equation, we obtain two equations which may be solved for C and Cy. Inthe case ofthe cantilever beim (Fig, 8.60), we note that both the deflection and the slope at A must be zero. Letting x = x, Y= 44 = 0 in Eq, (66), and x =x, 0 = 0, = 0 in Eq, (85), we obtain again two equations which may be solved for C, and Cy.DEFLECTION OF BEAMS BY INTEGRATION 398 Example 8.01 "The eanilever beam AB is of uniform cross section and caries a load P at its fee end A (Fig, 87)- Determine the tcjation of the elastie eure and the deflection and slope at A. Fig. 8:7 Using the freebody diagram of the portion AC of the beam (Fig. 88), where Cis located at a distance x from end A, wwe find M Substituting for M into Eq. (84) and multiplying both members by the constant Bl, wo write ay Bat =P an Pe Integrating in x, we obtain dy 2 srl = yr +0, 68) ‘We now observe that at the fxed end B we have x = and 0 = dy/dx = 0 (Fig, 89), Substituting these values into (88) and solving for C, we have c= 4Pt which we cary back into (8): Example 8.02 “The simply supported prismatic beam AB carres a uni- formly distributed load w per unit length (Fig. 810), Detér rine the equation of the elastic curve and the maximum de- Aeetion of the beam. si = yn + ee es) Integrating both members of Eq. (8.9), we write Ely = —APr! + $l + Cy 10) But, at B we have x =L, y = 0. Substituting into (8.10), we have PPL + PL +c, C= -yrL* Carrying the value of C back into Eq, (810), we obtain the equation of the elastic curve: Ely = Ue? 4 4214 RL? P(e y ate y= doe ea ew ‘The deflection and slope at A are obtained by letting =O n Eqs (611) and (89), We find Pu PL ET a om % [x=t.0=9] y b=ny=0] oe Drawing the free-body diagram of the portion AD of the beam (Fig, 811), and taking moments about D, we find that M = juby — por? (1) Substituting for Minto Eq. (8.4) and multiplying both members of this equation by the constant Fl, we write oy OY sigh cee x4 Luts 63) Integrating tige in x, we have ee ey, ‘ ett = — hort + Tbs +6, ou Bly = —Leuxt + bebe + Cr+, (8:15) pee ae ee400 MECHANICS OF MATERIALS Observing that y = 0 at both ends ofthe beam (Fig. 8.12), we fist let x = O and y = 0 in Eq, (8.15) and obtain C, = 0, We then make x = L and y = 0 in the sme equation and write = Awl + dul! + CL C= = qui? Carrying the values of C, and C, back into Eq, (8.15), we ob- tain the equation of the elastic curve + Ls? — 1) (615) Substituting into Eq, (8.14) the value obtained for C,, we check that the slope of the beam is zero for x = L/2 and that the elastic curve has « minimum at the midpoint C ofthe bean (ig. 8.13), Letting x = 1/2 in Eq. (8.16), we have ww (_ 14 yop LY yeh) _ Slt vo= oar(- Fee e128) ‘SEIT Fig. 8.12 Fig. 8.13 ‘The maximum defection or, moré precisely, the maximum ab- solute value of the deflection, is thus Sul (lees = SE In each of the two examples considered so far, only one free-body diagram was required to determine the bending moment in the beam, As a result, a single function of x was used to represent M throughout the beam. This, however, is not generally the case, Concentrated loads, reactions at supports, or discontinuities in a distributed load will make it necessary to Aivide the beam into several portions, and to represent the bending mo- ‘ment by a different function M(x) in each of these portions of beam. Each of the unctions M(x) will then lead to a different expression for the slope {4(x) and for the deflection y(x). Since each of the expressions obtained for the deflection must contain two constants of integration, a large number of constants will have to be determined. As we shall see in the next example, the required additional boundary conditions may be obtained by observing that, while the shear and bending moment can be discontinuous at several points in a beam, the deflection and the slope of the beam cannot be discon- tinuous at any point., Example 8.03 For the prismatic beam and the loading shown (Fig. 8.14), Aetermine the slope and deflection at point D. ‘We must divide the beam into two portions, AD and DB, and determine the function y(x) which defines the elastic curve for each of these portions fal Fig. 81, From A to D(xL/4), We now draw the free-body diagram of a portion of beam AB of length x > L/4 (Fig. 8.16) and write ap L Ex —P(x-4) (621) or, recalling a (4) and rearranging terms, ig el op - 141 (622) where ys(x) isthe funetion which defines the elastic curve for portion DB of the beam, Integrating in x, we write : ie Y Bg, = 14 = — 2 miedo +c, 620 Bye dre admit sce +c 628 DEFLECTION OF BEAMS BY INTEGRATION 404 ‘Determination of the Constants of Integration. The com- ditions which must be satisfied by the constants of integration have been summarized in Fig. 8:17. At the support A, where [r= 4.4 = 6) Fig. 8.17 eadtu=va) the deflection is defined by Eq, (8.20, we must have x = 0 and 4, = 0. At the support B, where the deflection is defined by Eq, (624), we must have x = L and yz = 0. Also, the fact that there can be no sudden change in deflection or in slope at point D requires that y, = ys and 0, therefo /4, 6, = 03h Eqs (8:19) and (8.23) si aka ee as /A, gy = Yo Ep. (6.0) and (828: Pe cb MPA, ob PE aM obec, 6m) Solving these equations simultaneously, we find wee _ UPL ¢ PL . C= 13s Substituting for C, and C, into Eqs. (8.19) and (8.20), we write that, for x < 1/4, PLC, (827) be G =0,0, reLt Spe B10, = 3 pet — He (629 1 pys _ TPLE. tty, = Lees — He (6.20) esting x = L/4 in each ofthese equations, we find thatthe slope and defection t point D are, respectively, But aoe and Yo = ~ 3568 ‘We note that, since 6, # 0, the deflection at Dis not the maxi mum defleetion of the beam.410 MECHANICS OF MATERIALS Fig. 8.18 (repeated) "8.4, USE OF SINGULARITY FUNCTIONS Reviewing the work done in the first three sections ofthis chapter, we rote that the integration method provides a convenient and effective way for determining the slope and deflection at any point of a prismatic bean. long as the bending moment may be represented by a single analytical function M(x), However, when the loading of the beam is such that two Afferent functions are needed to represent the bending moment over the éntire length ofthe beam, as in Example 6,03, four constants of integration ate required, and an equal number of equations, expressing continuity com ditions at point D, as well as boundary conditions at the supports A and B, must be used to determine these constants. If three or more functions were needed to represent the bending moment, additional constants and a corre. sponding number of additional equations would be, required, resulting in rather lengthy computations. We shall see in this section how the ue of singularity functions may simplify the computations, Considering again the beam and loading of Example 8.03 (Fig. 8.14), we recall from Eqs, (8.17) and (8.21) that the bending moment over the Portions AD and DB of the beam may be expressed by the functions Mya) = (o L/4, and by 2010 when x 0, we [(2— 0)" when x >a d GrOr= To whens a O when x 1 (647) rs = Singularity functions may be used to express the bending moments corresponding to various basic loadings. For example, in the case of 2 couple Mo applied at point A, we note that the bending moment M de tg his loading is equal to —M, when the point D where M ix computed ie jocated to the right of A (Fig. 24a), and to zero when Dis located to the {eh ofA (Fig. 8240). Recalling Eq, (8.45), we conclude that M may be expressed for any location of D as M = —M¢x — ay? Jn the ease of concentrated load P applied at A, we find that the bending moment ta point D isequal to —Plx ~ a) when isto thevight OF 4 (Fig, 8254), and to zero when D i to the let of A (Pig, 32S Therefore, it may be expressed for any location of D se M= =P ~ay! Considering now a uniformly distibuted load wy, we fst pass a see: tion trough a point D located to the right of A (Fig. 8260, Replacing tao ® Me) ~foges= a>? ——— Mo h —— Ma)= Seeia , whee Fig. 8.27 Basic loadings and corresponding bending moments or, recalling Ba, (8.46), Ve- nei Using now Eq. (7.6), we write eave DEFLECTION OF BEAMS BY INTEGRATION 443TAT MEURANICS OF MATERIALS, t-Th M [va Tal oo omtae OE k x i eer (8.48) m+ iin +2 = 1 +043) The expression for the bending moment corresponding to the loading of Fig, 827d may be obtained by making n = 1 in Eq. (8.48). We may also ‘verify the expression obtained for M(x) in the case of the loading of Fig. 8.27e by making n = 0 in Eq, (8.48), For most of the loadings encountered in engineering practice, the bending moment may be obtained by combining some of the expressions ‘given for M(x) in Fig, 8.27. The moments caused by the reactions at the Supports should also be included. Since all the distributed loadings shown in Fig. 8.27 are open-ended to the right, a distributed loading which does not extend to the right end of the beam (Fig, 8.28), or which is discontiny- wl) eg <8 29 = wy ous, should be replaced as shown by an equivalent combination of basic Fig. 0.28, open-ended loadings (see also Sample Prob. 8.3), Example 8.05 PR Lahn ug= LS kN Using singularity functions, express the bending moment corresponding to the beam and loading shown (Fig. 8.254), We fist determine the reaction at A by drawing the free- body diagram of the Beam (Fig. §.29b) and summing moments about w a oem 08m LOw +1EM, ~AJ3.6 m) + (1.2 RNY m) + (L8KN)24 m) 4 144N-im = 0 Pa12kn A, = 2.60 kN ke My =LALLN-m Neat, wo replace the given distributed load by two equiv- tee alent open-ended loadings (Fig. §.25). The given loads have thus been reduced to basic loadings ofthe type considered in Figs. 824 through 826. Following the procedute indicated at that time, we determine the bending moments corresponding to each ofthe loadings. Adding the expressions obtained, and including the moment due to the reaction at A, we write M = Ax — P(x — 06)! — jug ¢x — 062 + ht — 18)? — Myce — 26)" Substituting the numerical values of the reaction and of the loads, and being careful not to expand any of the products or Squares involving brackets, we obtain the following expression for the bending moment M = 261 ~ 19¢r — 06)! — +07 3 18)! ~ L4dcs ~ 26" Fig, 6.29DEFLECTION OF BEAMS BY INTEGRATION 4 *8.5. DIRECT DETERMINATION OF THE ELASTIC CURVE FROM THE LOAD DISTRIBUTION We saw in Sec. 82 that the equation of the elastic curve may be obtained by integrating twice the differential equation d’y _ M(x) aE eo where f(a is the bending moment in the beam. We now recall from Sec. 7.Sthat when abeam supports distributed load w(x) wehave dM dx = V and dV/de = ~w at any point of the beam. Differentiating both members of Eq, (84) with respect to x and assuming EI to be constant, we have therefore dy _ 1 aM _ Via) ane (49) and, difeectiating si, ay _ _ wl) re =e We conclude that, when a prismatic beam supports distributed load w(x), its elastic curve is governed by the fourth-order linear differential equation, (850) Multiplying both members of Bq. (8.50) by the constant EI and inte- grating four times, we write ap B14 = -wl2) #Y _ va) = — f wide + aa dy BIS = Mia) = = fd f w(x)de + x + Cy (851) 1M = B10(2) =~ f def ae f vlads + Leet + Oe +0, Ex yx) = — foe f de fae f w(x) de + LO ya? + 5 Ops? + Cox + Cy ‘The four constants of integration .aay be determined from the boundary ‘conditions. ‘These conditions inchude (a) the conditions imposed on the deflection or slope of the beam by its supports (cf. Sec. 6.2), and (b) the ‘condition that Vand M be zero at the free end of a cantilever beam, or that ‘Mbe zero at both ends ofa simply supported beam (ef. Sec. 7.3). This has been illustrated in Fig, 8.30. A [y,=0] [vn =0] [¢,=0] [uty = 0} (e) Cantilever beam [u,=0] [v5 =0] [m,=0] [M,=0] (@) Sip suppoeted bear Fig. 8.30 Boundary conditions for beams ‘ying a distrbuted load‘The method presented here m HARP supported beams carrying a distributed load. Ia the can over- - By expressing the distributed load t0(x) itself in terms of singularity functions, the method may be extended to the solution of problems involving beams subjected to discontinuous dis- tributed loads (see Sample Prob. 5.7) Example 8,06 ‘The simply supported prismatic beam AB cartes « une formly dtributed load w per unit length (Fig. 831)" De, Imine the equation of the clastic curve and the maahoure oe a IP aston ofthe beam. (This isthe same beam and loadings Example 8:02) te ca L Since w = constant, the frst three of Eqs. (8.51) yield Fig. 631 wr +6, Ft Mey = —dort 4 cn 4c (652) | © a Noting that the boundary conditions require thet af - Doth ends of the beam (Fig. 8.9), we fist let x = Q and af 4 aie in Eq. (52) and obtain C, = 0. We then make ro ang ‘M = 0m the same equation and obtain C, = fe, tm Carrying the values of C, and ©, back into Eq, (8.52), and [x=0, =o] [e=L.m=9] integrating twice, we wrt [r= y=0) [e=ny=0) Fig, 8.32 eel ext + Dole Se CEE a B= jet totic Canying the values of C, and C, back into Ei ‘lividing both members by El, we obtain the equa clastic curve: (6.39), ana et GE tet ac Bly = ~Jrort + Lule +0046, (659) But the boundary conditions a ‘ends of the beam. Letting x obtain C, = 0; letting x = we write require that y = 0 at both and y= 0in By, 659) ap age ean) sy y= Oi he min equation ‘The value ofthe maximum deflection is cbtained by mal a ae ing x = 1/2 in Eq. (854). We have HL! + ult + cy Eas C= — gui? hoes = SgaRr‘SAMPLE PROBLEM 8.3 For the prismatic beam and loading shown, determine (a) the equation of the clastic curve, (b) the slape at A, (c) the maximum deflection Modified Loading Diagram. We note that the given distributed loading is equivalent to the two open-ended loadings shown (Fig. a). Each loading and each reaction (Fig, b) corresponds to one of the basic loadings of Fig: 8:27 ‘Bending Moment. ‘The bending moment M(x) is obtained by adding, for cach reaction and loading in Fig. b, the bending moment given in Fig, 8:27, ) = Rx — Leys — Lbs - 4)? M(x) = yx — Liye? — Lis ~ Ly ak 1 (2160) 31 (400) og — gio (ay = +hugls— 4 (222) x2 — § (-492) ce -a0 (e) Equation of the Elastic Curve. Using Eq, 8.4, we have Wo 53 4 ocr — Lys soph — 23 4 BOs — Ly Bro = Mab gt _ eat + EOC yt + EI Bees GEG der + Substituting x = 0 into Eq, (3), we note that the quantity Boundary [x=0, y= between brackets is negative; ths the bracket is equal to zero. Therefore, Ce = fe=Ly=0} Again using Eq, (8), we write apl* — wool4 | Om tray 0s OE: ¢ 3 ie - eee re B19 = “oh Bote — aL) — spp tol @ we) Megs 4 0 (e —pty — sagt By = Sth gs — Beat 4 Bee HL) — aggitol ©) Slope at A, Substituting x = 0 into Bq. (2), we find B16, = =o (c) Maximum Deflection. At point C, where fea 5 wot ¥en = 00l4[ gigs ~ ona *°~ Toa) ~~ TETkN —36mn festa ya d E- MN Tig ie 292 S0kNim TA, SAMPLE PROBLEM 5.4 bar DEF is welded at point D to the uniform steel beam AB. For the shown, determine (a) the equatiosraf the elastic curve ofthe bean, (0) the deflection at the midpoint C of the beam. Use E = 200 GPa, TResctions, Considering the entire beam as a free body, we find Ry =S2kNT Ry = S2KNT Modified Loading Diagram. We replace the two concentrated loads by & force and couple at point D. Each element of the loading diagram is now equal to ‘one of the basic loadings given in Fig. 827. Bending Moment, ‘The bending moment M(x) is obtained by adding, for .ch reaction and loading, the bending moment given in Fig, 8.27. M(x) = Ryx — dope? — MgCx — 08)" — P(x ~ 08)! M(x) = [92x — }(00)x? — 5:6¢x — 0.8)° — 28¢x ~ 0.8)"105N +m (@) Equation of the Elastic Curve, Using Bq. 84 dy? Soames naan a P= 28KN Lae id [x-0,y=0) [x=12i,y=0] 1 SE = (ade — tse? — 56¢x — 08° — 28¢x — 0.8)*)0° Integrating twice in x 5c? — 56¢x — 08)" — 14x —08)7N07 +0, @) [5.3304 — 125e4 — 28¢x — 0.8)? — 467¢x — 08)3}10" +0240, @) [x=0,y =O} » Using Bq, (3), and noting that each bracket ¢ is equal to zero, we find Cy = Ir ©]; Again using Eq. (3), we write 0 = [5.93(1.2) — 1.95(1.2}* — 2810.4)" — 4.67(0.4" 10" + C12) C, = —489 x 10° Substituting for C, and G, into Eq, (3), we have Ely = [5.339 — 1.9544 — 2.8(e — 08)" — 467(x — 0.8)" — 489x]10" (B) Deflection at Midpoint C, For x = 0.6m, we write El yp = (53406) — 1.25(06)' — 0 — 0 — 4.8(0.6)]10" = —1.945 x 10° We recall that E = 200GPa and note that, for the given cross section, ‘bh _ (0.036 mY0.100 m)® we “ = 3x 10m We then write (200 x 10°X8 x 10-ye = = 1.945 x 108 Yo= Simm)[z-L.0=0] [r=Ly=0] 1, (©) determine the reaction at A and the deflection tinder the’ load, when’ La ‘Rections, For the given vertical load P the reactions are as shown. We note that they are statically indeterminate. Bending Moment, The bending moment M(x) is obtained by adding the moments of the reaction R, and of the load P. Recalling Fig, 8.27, we write Mix) = Ryx— Pex — a)" ‘Equation of the Elastic Curve. Using Fa. (8), ay me Ryz — P(x ay Integrating twice in x, y prem 2 ns Lee ay? ett = e10 = Last hPa +, riya dng Arent + G04 G IRL! ~ APL ~ 08 + Gy [x= Ly =0} IRL — APL ~ oP + CL + =0 (@) Reaction at A. Multiplying Eq, (2) by L, subtracting Eq. (3) member by rmamber from the equation obtained, and noting that C, = 0, we have ay, 42 r(-sQrgt < ‘We note thatthe reaction is independent of E and. (b) Renction at A and Defection at B when @ =}. Making expression obtained for Ri, we have Ry = PHA +2) Substituting @ = 1/2 and Ry C, = =PL#/32. Making x ‘obtained for y, we have oR 1p —1 pu — ae a) FR? — FPL — aIBL — — al sP/16 Ya =~ Teer [Note that the deflection obtained is not the maximum deflection. a9