K 1 Majid Theory of Structures with matrix notation Newnes-Butterworths= a aap gene pepsi she ig age fro Theory of Structures with matrix notation K. 1. MAJID, B.Sc., Ph.D., D.Sc., C.Eng,., F.LC.E., F.l.Struct.£. Prafessor of Civil Engineering, University of Aston in Birmingham 23097 Newnes — Butterworths LONDON = BOSTON Sydney — Wellington — Durban — Toronto 1 A‘The Butterworth Group United Kingdom — Butterworth and:Co (Publishers! Ltd ‘London: 88 Kinoway, WC22 BAB Austratia Butterworths Pry Ltd Sytiney: $86 Pacific Highway, Chatrwood, NSW 2067 Alo at Malhourna, Britbane, Adetmie ang Berth Ceneds Butterworth and Co (Canada) Led ‘Toronto: 2766 Midland Avenue, Scarborough, Ontaria MIP-451 New Zealand Butterworths of New Zealand Ltd ‘Wailington: 77-85 Customhouse Guay. 1 CPO Box 472 South Africa Burterworth and Co {South Atrical (Pty! Ltd Durban: 152-154 Gale Straot USA Butterworth (Publishers) Ine Bortorr, 19 Cummings Park, Woburn, Mass. 01801 First published 1978 ©Butteworth & Co. (Publishers) Ltd, 1978 ALL rights reserved. No part of this publication may be reproduced or tranamitted in any form or by any means, including photocopying and recording. without the written parmision of the copyright holder, application tor which should Ibe addressed to the Publishers. Such written pemmnission must also be obtoined before any part of this publi- cation Isstored ina retrieval system of any neture This book is sold subject to the Standard Conditions of Sake ol Net Books and may note re-sold in the UK balow the net price given by the Publishers in their current price list. ISBN 0 408 .00303 0 Cased 0408003235 Limp British Library Cataloguing in Publication Data Majic, Kamat Izekseen Theary of Structures. 1. Structures, Theory of 1. Tite 62417 TABS 77-90307 ISBN @-408-00303-0 0-408-00323-5 Pbk Typeset By Reproduction Drawings Lid Sutton Surrey Printed in Scotiend by Thomson Litho Lid., Cast Kitbrige Preface The use of computers in structural engineering is now well established. The methods at present used in structural analyses are therefore based on matrix algebra which is more suitable to computers. However, these matrix methods are often taught too late during degree courses io the extent that the student, well versed in matrix algebra fram schol, does not appreciate the advantages of them, On the other hand, present day courses in the theory of structures suffer from duplication which ‘causes congestion ax well a3 confusion as to which of the many methods prescribed is the most useful. ‘This book is therefore written to fulfill two main objectives. First, to introduce the modern matrix methods gradually from an carly stage and in a simple manner Second, to remove duplication PEON aae topics that are out of use so that those currently practised are covered thoroughly. Prominence is given to the fundamental aspects of structural analysis. Thus whenever possible the equations of equilibrium and the steess-strain relationships of the material are emphasised. Such structural properties as stiffness, flexibility, energy, elasticity, plasticity etc. are not only described in detail but-also used extensively. To be consistent, a single set of sign conventions is adopted throughout. [1 is considered that the right hand screw rule ix store advantageous than the left particularly because students become familiar with this mule at school, This volume treats simple structures and covers the course given in the first two years of undergraduate study, Chapter 1 covers the analysis of pin jointed isostatic plune frames. Once the member forces are calculated, these are expressed in terms of the extemal forces in the form of a matnx equation. This expression later forms the basis of the matrix force method. Chapter 2 introduces the structural properties such as stress, strain, stiffness, flexibility and strain energy of axially loaded members. The relationship between the stiffness, the stress and the strain which forms the basis of the finite element method is introduced im this chapter with the aid of a single member. Chapter 3 is dewoted to laterally loaded beams while Chapter 4 covers bent and curved members in which bending moments and shearing forces as well as axial forces are present. The local axes, as opposed to the global axes, of a member are defined in Chupter 4. Identification of members and their ends is introduced here ‘with the aim of training the student for the future use of member numbering. This chapter is conchided with the analysis of three pinned archs as isostatic frames can- sisting of bent or curved members. Hoth the matrix force and matrix displacement methods are based on the slope deflection equations. These eee ave derived in Chapter 5 and used extensively to calculate the deflections of isostatic and simplo hyperstatic members and frames. ‘The slope deflection equations are derived from a general displacement function. In this manner the procedure used to present the finite element method is illustrated, ‘The method of dividing uniformly distributed loads into two loading systems, as adopted by matrix methods for discrete problems, is also described.vi Preface Chapter 6 introduces the influence lines. The virtual work method it adapted to draw the influence lines for shearing forces, bending moments, etc. for isostatic fembers and frames, The dangerous section far design is defined in this chapter, ‘Chapter 7 defines the flexural properties of structures and members, and the strain energies of a member in pure bending and pure shear are introduced, The virtual work equation in its general form has been derived in this one and it is demonstrated that in fact this equation is another form of defining the stale of equilibrium in a member. Chapter 8 is devoted to the plastic theory. It has been emphasised that frames at incipient collapse are isostatic and the ‘equations of static equilibrium are sufficient to calculate the failure load of such frames. Chapter 9 introduces the foroe method for calculating deflections of pin jointed and rigidly jointed frames. ‘It is considered that calculating deflections by the fore method is simpler than the use of the virtual work equations, for, once the relation- ships between momber forces and extemal loads ure established, they can be used directly to calculate joint defections. ‘This chapter is concluded by extending the force method to cover the analysis of hyperstatic structures. Chapter 10 is concemed with the analysis of columns. The modem method of using the stability functions in the analysis of columns is adopted throughout. In this chapter the significance of the clastic instability on the one hand and of the non- linear analysis of columns on the other hand are both explained in detail. ‘Chapter 11 is concemed with the analysis of pin jointed space frames. The method of tension coefficients is brought up to date and presented as the basis of the matrix force method. Tike chapter then deals with the matrix displacement method for the analysis of hyperstatic pin jointed space frames. Chapter 12 covers the displacement method for the analysis of rigidly jointed plane frames, This covers the linear and the non-linear elastic analysis of frames including the state of elastic instability. ‘The volume contains more than 120 worked examples and each chapter is con- cluded with exercises with answers. ‘The suthor wishes to thank Mr David J, Newman, B.A. {Cantab.) of Taylor ‘Woodrow International and Dr David Just of Aston University for reading the manu- script and resolving and correcting the examples aswell as for their valuable suggestions and contributions: Dr ‘David Anderson of the University of Warwick for reading the text and making final suggestions; Mrs Janet Allen for typing the script and Mrs Patricia Taber for producing the drawings. Contents we Pin jointed isostatic plane structures 1 1L-Introductiog, 1.2—Representation of supports, }.3~Equilibrium of coplanas forces, 1.4—Steuctural equisteium, 1-$ Calculation of reactions, 1.6—Member forces, 1.7—The method of sections, 1.4=Comibination of methods of joint and wection, 1.9-Unstuble, isostatic and hyperstatic frames, 1.10-Consideration of merabers and supports, 1.11—Exaruples, Exereites on chapter 1 Structural properties 3 ‘LA- Direct stress ins member, 2.2—Direct strains, 2.3—The stress-strain evrve 2.4-Hooke’s law, 2.5 —Stiffness, 2.6~- Flexibility, 2.7— Examples, 2.8- enperatre “The ptinciple of wuperpaiition, 2.11-Strain energy, Relationships among stress, strain and stiffivess, Principles of structural analyds, 2.15-Examples, 2.16—Propertics of symmetrical structures 2.17~The equivalent half frame, 2.18—The equivaieet lous, 2.19—An fr matrix methods, 2.20—The critical design toad, 2.21—-Example, Exercises on chapte: 2, Isostatic beams 3.1 -Introduction, 3.2-Shearing foree and bending moments im beam Definitions, 3.2.2-A cantilever with an end ford, 22 3-Sign See cantikever carry ing an inclined force, 3.2.5—A eniformly loaded cantilever, 3.2.6- A cantilever with several loads, 3.2.7—Simply supported beams, .2.8—Examples 3.3— Relationships for the load, the shearing force and the bending moment, 3.4—~ Esomples, 3.5 — Bending stresies in beams, 3.6—- Example, 3,7 ~Generalived Macaulay's method for stieaiing forces and bending moments, 3.7.1—A beam with paint loads, 3.7.2-A beam with uniform loads, 3.8- Example, Exercises om chapter 3. Bent members and structures 90 itreduction, 4.2-Plane beat members, 4.3=Plane curved members, 4.4 — Esai oe at eee ‘out of its plans, 4.4—Curved member loaded out ne, 4.7-Three pioned arches, 4.8~The local aves of a member, 4.9— rs eee. is mber, 4.9—-Exampies, Slopes and deflections Md : oductiog, 5.2>Elatic bending of beams, 5.3—The deflection function, 54 Ez slope = WB (tensile) Because i =O, the tat A acts as a roller. Thus the frame is Bess ee otra lly. For this reason the forces in the members are also symmetrical, and py = P5, P= Po and py = 27. Suny the forces in members 5, 6 and 7 are calculated by cutting these members” Pin Jointed Isostatic Plane Strecteres i To calculate the force in member 4, the frame is mow cut by section ¢-c a5 shown in Figure J. 4a. The part of the frame to the left of this section is shown in Figure 1.4 1d, Resolving the forces acting on this part vertically Vy + pa + Pe cosa = 0 Pa = -Fq — py cosa = —0.5W —0.6W x (LS/1.8) Le. Pa = —W (compressive) 1.8. Combination of methods of joint and section Sometimes, it may be easier to use the method of section in conjunction with the method of joints. For instance, in the frame of Figure f.41, while the forces p, .p3. Ps, Ps) Pe and p, are calculated by the method of section, the force py can be calculated by considering the vertical equilibrium. ‘of joint F. This directly gives py “Ag another example consider the pin jointed frame shown in Figure 1.12 where it is required to calculate the forces in members 1, 2, 3 and 4. By resolving the external rie the reactions honzontally, it is found that Aly = 0. Taking moments about point yee 4 - Wee We ww a= 0 y= 302 =F, To calculate the forces in members 1, 2, 3 and 4, two sections and one joint are considered. The part of the structure to the right of section A-A is shown in Figure 142b. For this: taking moments about paint G aly +p, x the O. Thus Py = -al_/2h = —3aW/4h (compressive) Figure 1.1212 Pin Jointed Isostatic Plane Strectures Similarly by taking moments about point E, it is found that ps = 3ab//4h which isa tele force, tos Sete Next consider the part of the structure to the right of section B-B, Fiuere 1.420, and resolve the forces vertically Ve + py sind — pysin@- WO (7) ‘Now using the method of joints at F, Figure 1.124, and resolving the forces horizontally =p; conh — Py cor 6 = 0 ie. Pa. Pt (8) Substituting for ps from equation (1.8) into equation (1.7), it is found that py © —0.25 Wisin @ (compressive) Thus. py = 0.25 W/sin @ (tensile) 1.9. Unstable, isostatic and hyperstatic frames ‘So far it has been stated that isostatic plane frames can be analysed using the equal cof static equilibrium. Before doing this il is necessary to find out whether the frame is isostatic. A plane frame is isostatic if it has a sufficient number of members and joints, arranged in a suitable manner so that it is just stable. For instanee, the tri- angular frame shown in Figure 1. 13a is just stable because its shape cannot be ‘changed by a force without altering the length of its members. On the other hand the frame shown in Figure i. [3b is unstable because its shape can be changed fram the position ACDB to AC'D'B with a very small force acting at C or D. This frame can be rendered stable and also isostatic if a member such as CB (Figure J, 3c) it added to it, Finally the frame shown in Figure 113d is “over stiff’ or “over stable” because any one of its members can be etnoved without it becoming unstable, This frame is sald to be hyperstatic. The equations of static equilibrium, on their own, are not sufficient to analyse such a frame. ‘A stable frame can be developed by starting with a triangle, with three members and three joints, and then successively connecting each additional joint by two new members. Thus starting with the stable triangular frame ABC shown in Figure 1t3e, the stable frame shown in Figure 1.13c is developed by connecting the new joint Db by two members CD and BD. This process may be continued as shown in Figure 7.13f, Joint E is connected lo ABCD by another two members DE and EB and then: joint F is added by means of members EF and BE. Thus for a stable isostatic frame © © pte 2 © mi ras oe A DX] i eel 5 ” a a) ee oo, t eg ‘ “ o o Figure 1.13, Pin Jointed Jeastatic Plane Strectures 13 k nivers and j joints the first three member j selalcaoce fet pall one eal sevanten "secant ~ 3 new joints. Now since the number of W the number of new joints, it follows that Ties an m —3= 4) - 3) maya 19) __In developing an isostatic frame, care must be taken to joints so that each part of the frame is stable. For inseance, tha | uit Fae laa has nine members and six joints and satisfies equation (1.9), It is stable and internal isostatic (reactions are excluded here). The frame shown in Figure /.14b also has i with a total of mm which gives a eo t r (o) © a — Figure (14 oe. le to nine m joi rs Renee six joints but itis not internally isostatic. Here joint Dis connected ingle ABC by three new members instead therefore hyperstatic, On the other hand the Ma TORNE Lau Fite oneceat i ABCD by three members instead of four. Thus part DEFB is unstable. In Figure J./4c, Joints CDE are on a straight line and do not fi joint Dio Ta do not form a triangle, A small force can move ibe ccna Ae ed asl the frame is therefore unstable. The rest of the frame on 1.10. Consideration of members and supports It i eee out tt when using the equations of equilibrium to calculate the emnerak fone tine he ee eae it is possible to derive two ‘a Pus for a frame with / joints the j lane supported with F reaction components and Manoa ee Funknown componentsandm member forces, m +r must be equal to ay be. mere y (1.10) is oO rain a frame asa whole th cl a aecesary- thus y= 3at eat For lata he tame show aha iinke: this fame equiten I-10) tele oo aca corponants = 3} For - 3 je structure is i 1 tion components can thus be calculated, Removing the sepparte teawie crake It was stated earhicr th:14 Pin Jointed Isostatic Plane Structures ‘ * f etme ie tue Rigi Fo op ry o Figure 1.15 frame shown in Figure [142 and equation (1.9) will be satisfied, thus the member forces can be calculated and the frame is internally isostatic, The frame in Figure 1156 has m= 9, = 4 and /= 6. This frame does not satisfy equation (1.10), The reason is that the number of reactions is one too many and the frame is extemally hyperstatic. Removing the supports once again results in the frame shown in Figure J.J 4a which satisfies equation. t19) Thus the frame is internally isostatic. Once the extra reaction (say 1") is calculated, by methods given later, the member forces can mined by equations of static equilibrium. The unknown force Vp is called a redundant force and for the time being the other reactions and the member forces can only be calculated in terms of Mg. 1.11. Examples Example 1. Calculate the member forces in the frame shown in Figure 1.76. Hence write down equations P= BW. Answer: In this frame mt =9,r= 3 and/= 6, pea is satisfied. Re- moving the reaction components, it is found equation (1.9) is alse satisfied. ‘The frame is thus isostatic both externally and internally. Pin Jointed Isostaric Plane Structures 15 From the horizontal equilibrium of the frame as a whole, we obtain Ha — Wy coe =O E Hy = Wy cos = 0.60, Taking moments of the external loads and the reactions about B —Fa x 12 + W, x 6+ Wysind x 6 = 0 ‘ Va = GW,/12 + GW; x 08/12 = 056, + O4W, Resolving the extemal loads and the reactions vertically Va + Fg — W, — Wy sind = 0 Vy = W, + Wy x O8 — Vy = W, + 08K, — 0.5m, — O4W, Va = OSW, + Oa, : Haves ‘ nes Sa the reactions, the values of the member forces can now be AUB, resolving vertically Va +p, sin? =0 P= =Vpisin ® = (0.50, + 0.4, \sind mp, = -0.51,/0.8 — 0.4, /0.8 Py = 0.6251, — 0.51; (compression) At B, resolving horizontally Pr — pod = 0 Pr = Py % 0.6 = 0.625W, x 06+ OSH, x 0.6 Pi = O375W, + 0.31 (tension) ALE, resolving vertically —p, sin® — pysind = 0 Pa = =P) = 0.625W, + 0.5W's (tension) ALE, resolving horizontally “Pa = Py ore + p, cosd = 0 Pa = —Pi kK 06 + py x 0.6 = ~0.75W, — ‘O.6W, (compression) AUD, resolving vertically Wy, —ps = 0 Ps = —W, (compression) rizen tally Pa — Pp = 0 s Po = Pa = —O75W, — O60, i ALA, resolving ver ically : err ia Va + pa sind = 0 Pa = —Vaisind = -0625W', — 0.5, (compression) ACD, resolving16 Pin Jointed Hsostattc Plane Structures Pin Jointed Ieostatic Plane Structures 7 At A, resolving horizontally Hy + py cusd + py = 0 a Py = Hy — py coad = —0,6W, + 0.625W, x 0.6 405K, x 0.6 Po = O.37SW, = 0.3K; ‘Thus if 0.3751; 20.36; Le: ff W) >0.3W/2/0.375 the foree in member 9 will be tensile, Otherwise it will be a compressive force. ALC, resolving the forces vertically —ps sind — p,din@ = 0 a Py = —Ps = 0.625, + OSH, (tension) Equations P= BW are Pi 0.625 O57, Pa 0.375 03 Fel Ps 0.625 05 Pa =0.75 0.6 Ps} = | -1 9 Pe 0,75 =0.6 Pa 0,625 Os Ps -0.625 -05 Po 0.375 -03 Example 2. ls the frame shown in Figure LJ? isostatic? Calculate the forces in members 1, 2, 3 and 4. Answer: m= 19,7 = 3andj = UL merm 1943 = 22 = of Pigare 1.27, (a) Therefore equation (1.10) is satisfied and all the member forces and reaction com=— ponents can be calculated. The frame is isostatic. Considering the part of the frame: to the right of section A-A, as shown in Figure 1.176, and taking moments about ; point H Similarly, taking moments about paint F px 6+Wx 4nd = 0 SF 6 Wie BAT wenoeeew eG ‘The lever arm of the force W with respect to H is HM = 4 sin 0. i Pa = 4W x 0.6/6 + GW x 0.8/6 = 1.2 : py = —4Wain 0/6 = —4W x 0.6/6 = -0.4W Meh components of W have moments about point F, ‘The force pry can also be obtained as follows. sidering next the part of the frame te thi ctl 7 ‘The hovizontal component af the force W passes through point H and thus has and resolving the forces vertically i Fight of section B-B, Figure 7.76, no moment about this point. The vertical component of | is force is W sin @ and Pi sin @ + Wsin @ — py si the lever arm of this component with. respect 10 Af is equal to.HL = 4m. The ie, Ps sin 8 = 0 moment of the vertical component of W about # is thus Wsin @ x 4. The moment ei Pat+W—py=0 (any of py about H is 6p, , thus nally, front the horizontal ia on ante one 6, + Wind x 4=0 Pz 608 6 + py cosh = 0 py = Wx 06 x 4/6 = -0AW. zs Ps =18 Pin Jointed Isostatic Plane Structures Substituting for ps in equation (1.11) Pot Wtp, = 0 es P= -0.5W and Ps = OSW Exampk 3, Are the frames in Figure 1.18 isostatic? Answer: The frame shown in Figure 1.48ahasm =5,/= 4andr= 3. These satisfy ‘equation (1,10), When the reaction components are removed, the frame with m = 5 and j= 4 will satisfy equation (1.9). The frame is therefore stable and internally isostatic, With 2 = 8 there are sufficient equations to calculate the reaction com- ponents and the member forces. ‘The frame shown in Figure J./8bhasm=4,r= 3 and f= 4, thuson += 7< 2. ‘This frame is unstable because it has not a sufficient number ‘of members. ‘The stability of the frame shown in Figure ‘L18b can be restored either by an extra reaction component as in Figure 1. /8e or by an extra member as in Figure 1.18 a and d. The frame in Fi 1.1 Se satisfies equation (1.10) but not equation (1.9), Nevertheless with 2/= 8, there are a sufficient number of equillbrium uations to solve for four unknown reaction components and four member forces. ‘This frame is therefore isostatic. Frame d satisfies equation (1-9) and (1.10) and is therefore isostatic externally and internally. = Sand /=4 and m +r=9 which The frame shown in Figure 1. 18¢ has m= 4,7 is more than 2, There is an insufficient number of equations to solve for the nine unknowns. This frame is hyperstatic. ' 1 4 aS is 3 ia) to) ; : Pe 2 aa © @ Figure 1.18 @ 1 Pin Jointed isostatic Plane Streetures 19 Example 4. In the hyperstatic frame shown in Figure #18 |. £8e (a) soles i reaction Rp as the redundant force and eran TaneE fora eat fc toe member 4 as the redundant force and calculate the Note: Often when a structure is hyperstatic, the redundant 5 ss fe et nee a re ns two such ways are Sigealed) Gan eat often forees in terms of this fortes pee terrae: erp see menter Answer: (a) Rp is redundant, Resolve forces at D vertically Rp + pa x Uy? =o ‘ Pa = —V2 x Rp: Resolve forces at D horizontally “Ps — Pa % 2 = 0 e Ps = —pal V2 = Rp Resolving forces at C vertically (ra x T2) = ey x 2) = 0 : Pa = Ps ™W2 x Ry Resolving forces at C horizontally “Pi ~ Wy X Uv2) + es « 1/2) = 0 an Py = —2Rp ‘The answers can be expressed in matrix form as rm : 3 [ eo] 1 Py Ps Ps af (b) py is redundant Resolving forces at D horizontally “Ps — Py X U2 20 Ps = pal 2 Resolving forces at C vertically Gar * 1fs/2) - @y x 12) = 0 P2 = Pa Resolving farces at C horizontally “Pi — Pa ® U2) + (py * 12) = 0 Py = v2 % ByPin Jointed fsastatie Plane Structures 1 8 Pin Jointed! Isostatic Plane Sanectures mf ele] Ps -1 Ps v2 oe Exercises on Chapter 1 1. Calculate the forces in members AB, CB and AD in the pin jointed frame shown in Figure 1.29. ARS. Pg = —446W, Pop™ 287M, Pap = 245%. Figure 1.22 4. The horizontal thrust at A and B of the bridge shown in Figure #.22 is H. Cal- culate the member forces in terms of A’ and tabulate the results in the matrix from P= BW. For/f= 10 kN what is the largest compressive force in the brigge? Make use of the symmetry of the structure, Ans, 24.7 KN in members CD and DE. I Foo td +++ Figure ted 9 Figure 1.20 2, Calculate the forces in members 2, 3, 6, 7,9 and 10 in the pin jointed frame shown in Figure 1.20 (a) by the method of joint resolution, (b) by the method of sections when W'= 10 KN. ’ Ans. Py = ~30,p5 ~ —20,pg * 10,7) = —10.p, = 17-32.pjq = 17.32 NE 3. With W'= 10 KN calculate the force in member AB of the pin jointed frame shown in Figure 1.21. Ans. 31.38 KN. Figure 1.23 5. The forces in the members of the pin j z jointed frame shown in Figure J, 23 can be eeeieted in tens of the forcoe in the redundant members p, tt |. Tabulate noe fom, | 1fpe = 10 kN and p,, = 20 kN what ix the Ans, p, = —10 KN.22 Pia Jointed Jsostatic Plane Structures 2 | . Structural properties 2.1. Direct stress in a member | Bes Consider a pin ended member AB of length L. and cross sectional area A as shown in . Foqure 2.1. When a tensile force p acts on the member, it is assumed that this force isevenly distributed over the cross sectional area of the member. If the member is cut somewhere such as D, then equal forces p are required to bi fed at D, as shawn in Figure 2.16, so that the two parts AD and DB are kept in equilibrium. The member is thus under a tensile force p throughout. The intensity. of the force acting. on the unit cross sectional area of the member is known as the stress in the member. by 4 The tensile stress is thus given 0 pid kN/mm? (21) © When the forces p, acting on the member, are reversed in direction at each end they Figure 1.24 tend to compress the member and this produces a compressive stress therein. Both tensile and compressive stresses are referred to a5 direct stresces, 6. Indicate, giving reasons, whether the pin jointed frames in Figure 1.24 are stable, ’ ecatail the. ie e isostatic or hypersta: z Y oe L - * * . 7 Greens o ‘p (oy (oy «&) fa) Figure 2.1 Consider a pin jointed structure with W members 1, Bes these members. Pi. Px... S p members be Ay.As,..edy.. Ay he stresses 94,03,.4.0)... Oy in the members are given by equation (2.1) as 1 = PylAy 92 = pilAs & = Pda, on™ pwld (2) 23a (Structural Properties Strictueral Properties: 25 ‘These equations can be written in matrix form 3s 2.2. Direct strains ay WA 0 7 [= When a tensile load p acts on a member, it elongates the member by a small, but [ 2 4 measurable, amount, In Figure 2.2 the member AB is shown to be pinned to its 72 ia VAs 4 \} g (23a) support at A and subject to a tensile axial force p at B. The original length of the a WAj=.. Pr member is Z, The force p causes the member to elongate by an amount 62 and for [3 ° == 1A bw | this reason point B moves to B’ and the final length of the member becomes L + 51. i ! or simply o-aP (2.36) a ae where 0= (0, 02 --.0/-.-9N) is acolumn vector of member stresses, P= (piyPa.---Pr-- Py} the column vector of member forces and the diagonal matrix 1A, 0 os WA, Qa) “WAL 0 “=HAy ins th J of the member cross sectional areas as elements on the feoiling dlagonal. THE matrix is square, i.e. it has the same number of rows and. columns. Once the forces in the members are calculated, by the methods given in ‘the last cane the stresses in the members are calculated using equation (2.3). Alt ti if the limiting stresses in the members are O44, Gwas + + Put ~~ «Ores then neuer’ area of cach member is obtained from equation (2.1) as ‘The sirain ¢ in the member is the ratio of the elongation of the member and its original length, ic. €= SL/L (27) ah ‘The strain is thus the elongation of a unit length of the original member. Since both Ai See SL and 4 ave measured in units of length, e.g. in mm, the strain has no units. [tis Az ™ Palys (sy? just a ratio, aeaaieo Equations similar to (2.7) can be written foreach member of a structure and these equations can be put in matrix form as Ay = Pylowy e=6 (2.8) Equations (2.5) can also be written in matrix form as whereelsl{es,€2,....€¢. ..€y} is the column vector of the strains in each member, Ay [ioe 0 PL 6" [6L,,0L3,...5L;.. 5éy) is the vector of the deformation of the members. |} i Px The matria # is square and diagonal with the reciprocal 1/Z for each member Ay} 2 | Qe i (2.5.8) appearing on the leading diagonal. Ar Mee (Vi [Aw 0 Wey] [Pv 2.3. The stress-strain curve A For many. engineering materials, such as steel, the graph of the stress in a member a ligt (2.5.b) against the resulting: imilar to that shown in #igure 2.3. Between the origin or simply ‘O and polat A, the graph is nearly a straight line and the material is said to behave ‘The column vector A = (A, Az...4;.--Aiy} contains the unknown member elustically. Point A is called the limit of proportionality. After A, the graph tends areas and the matrix towards the horizontal axis and the increase in strain becomes large for small in- Sreases in the stress. Between points B and C, it is said that the material behaves a plastically. The portion BC of the graph is nearly horizontal, After point C, the Mews slope of the graph increases Uy and itis said that the material is strain hardened. Hawa = _ Ae material fails suddenly. ~, the member is loaded to a stress level below A and then unloaded, the str il Te train curve, for unloading, coincides with that for loading the member. However 0 oN | after A, if the member is unloaded, ata point much as F, 1 the unloading stress- Strain curve follows the PG. When the load is tely removed, a residual Strain is left in the member as a permanent set. A material with a stress-strain curve is square and diagonal and called the limiting stress matrix.26 Sorucrural Properties Figure 2.3 such as that shown in Figure 2,3 is called a ductile material. A brittle material, such as concrete, has a ree of stress-strain curve. The linear part of the curve is much shorter and due to local failures, such as the cracking of the concrete at some parts, the ordinate of the curve may siddenly fal! before total failure. 2.4, Hooke’s law It was stated in the last section that the part between paints O and A of Figur 2.3 iis nearly linear. Engineers assume that for small loads applied to a member, the stresatrain curve is linear. That is to say the deformation of the member is directly ‘proportional to the applied loa. This assumption is known as Hooke's law. Fora © material that obeys Hooke's law the ratio E = ole (29) between stress and strain is constant. This ratio is called Young's modulus of elasticity. The units of stress are in KN/mmn? while strain has no unit. Thus the units of Young's modulus of elasticity are also in kN/mm7*. For structural steel the value. of Elis about 207 KN/mm?, It is evident that & is the slope of the stress-strain curve for the linear portion OA. Now @ = pi and € = 6L/L, it follows, from equation (2.9) that E = ofe = (piAN(BLIL) = pL|AbL (2.10) © 2.5, Stiffness Consider a force p applied to a member that obeys Hooke’s law. The member has area A and length E, Let the deformation (clongation or contraction) of the member be 6L. The equation (2.10) holds true for the member. Rearranging this equation p= (EAILL eal or p= RL (2:12), where the property k= Ea/Z (kN/mm) is known as the stiffness of the member and is defined as the load applied to the member which causes unit deformation of the member. This is because equation (2.12) can be rearranged to became k= pibL (2.13) rs Structural Properties 27 ‘Thus the stiffness & is the force per unit deformation. Equations similar to (2.11) and (2.12) can be written foreach member of a pin, jointed structure. For a structure wi otal of V members, Py = Rybhy = (EAL DL, pa = Kablis = (EyAgi/Ly) SE, ; (2.04) my = bbb) = (B AJL BL, pu 7 AnSl= (EwAyltoe) Bly ‘These equations can be written in matrix form thus Pr BiAriby oO na Fs AziL ean Ell bie by ° “ERAnity of simply P= ke (2.15b) The diagonal matrix kis known as the. stiffness matrix. of the members. The feading diagonal of the matrix consists of the stiffnesses k = EA/E of the members. 2.6. Flexibility Equation (2.11) can be rearranged in a different manner, thus SL = (LIEA\p 2.16) oF SL = fp (17) The property f= L/E4 (mm/kN} is known as the flexibility of the member. It is defined as the deformation caused in a member by a unit load. This is because equation (2.17) can be rearranged to read f= BLip. _ Equations dmilar to (2.16) and (2.17) can be written for every member of a pin jointed structure and in matrix form these become neo —— [eer] fees o | fa §L: LilEsAs L = nt fa) (ga Ly “tule a Y tty] Lo LylBAw | | pw or simply e=1P (2.18b) ‘The diagonal square matrix fis called the flexibility matrix of the members. Cx paring equations (2. and (2.17) it is noticed thet f= 1k. This means that the28 Structural Properties flexibility of a member is the reciprocal of its stiffness. Similarly comparing equations (2.15) and (2.18), we deduce that f=h-! Thus the flexibility matrix is the inverse of the stiffness matrix. Since k is a dit matrix, its inverse is given by taking the reciprocal of the ele- ments £A/L of its leading diagonal. In equations (2.11) and (2. 0). the original length £ of a member and its area d are constants, The value of £, for the | part of the stress-strain curve, Le. for OA in Figure 2.3, is also constant. Thus the stiffness and the flexibility of a linearly elastic member are constant properties of the member. However, the member his atiffness and flexibility even afler the limit of preporticnality, A in Figure 2.3, is execeded, Here, because £ is no longer constant, the stiffness and the flexibility of the member changes from paint te point in the region AD of the stress-strain curve. ‘The stiffness of the member is reduced while its flexibility is increased. Once member fails, at D in Figure 2.3, its stiffness becomes zero and its flexibility be- comes infinite, Thus failure of a member can be defined as the total loss of its stiff ness, (2.19) 2.7. Examples Example 1. A concrete block is cylindrical in shape with a diameter of 100 mm and a height of 300 mm. A compression load of 70 KN causes the cylinder to con+ tract by 0.2 mm. Calculate the stress, the strain, the modulus of elasticity, the stiff neas and the flexibility of the block. Answer: ‘The area of the cross section = = x 100 x 100/4 A = 7860 mm? ‘Compressive stress @. = pid = 70/7860 = 8.9 x 107° kN/mm? ‘Compressive strain e, = SL/L = 0.2/300 = 0.67 x 10-2 Modulus of elasticity F = o,/e. = (89 x 107>)/(0.67 x 1073) & = 13.3 kN/may* Stiffness k = EA/L = 133 x 7860/300 = 348 KN/mm Flexibility f= 1/k = 1/348 = 2.88 x 107 mm/kN Example 2. The structure shown in Figure 1/0), section 16, is to be made from mild steel members with modulus of elasticity 200 kN/mm?. Calculate the area of each member so that the stress developed is limited to 0.15 KN/mm? in terion and 0.1 kN/mm? in compression.|What is the axial deformation in each member? Answer: From equations (1.5), P= B/W, with W'= 10 KN, is found to be P=(p Py Pa Ps Pe Py Pap = {-2 W2 ? -t -l v2 0 0} [io] = {20 M14 10-10-10 1414 0 0} kN ‘The areas of the members are given by equations (2.5), i. A= Sw P. There are no forces in members 7 and 8 and they are disregarded, thus a aan ag a Sa a a eo ey ay Structunal Properties 2B ‘Ay 1-01 [=207 Ay 10.15 0 eats sett (ATi 170.15 10 Ay 1-0. =10 As 0 1-01 |-10 Ay 10.15 jist Say = (1/01) x (20) = 200mm, 4. = 14,14/0.15 = 94.3 mm? and Ay = 10/0, is = era: Similarly Ag = 100mm?, A, = 100 mm? Ag = 94,3 mm? ‘The axial deformation in cach member is calculated using the flexibility equations (2.18), i2.6 =P. The member lengths are shown in Figure #70, thus & Ly 10 000/200 = 20 | St, 14 140/94.3 0 14.14 Bs | 10 000/66.7 10 tha |” 200 10 000/100 -10 tls 10 000/100 -10 [an | 0 14 1agj94.3 | (14.14 | 6L, = (1/200) x (10000/200) x —20 = —S mmi.e, contraction SL, = (1/200) x (14140/94.3) « 14.14 = 10.6 mmix. extension Similarly 54; = 7.5mm, 6L_ = —3mm, 6L, = —3mmand6l, = 10.6mm Sar 3. Calculate the areas of the members of the structure ADB shown in ae 2.4 when the load Wis 10 KN. What are the horizontal and the vertical deflections at joint D? The stress in a tensile member is 0.15 kN/mum? and in a compressive member is —0.1 KN/mm?, Take £ as 200 KN/mm*, Answer, Resolving the forces at D horizontally and then vertically we obtain Pan = Py = -v3 % W = =104/3KN (compression) Pun =P; ™ 2W = 20KN (tension) arcacimember | = A, = (—10./3W-0.1 = 173mm? area of member 2 = Az = 20/015 = 133mm? Contraction ofimembert = 62; ¢ Sop, = Uvex 10x IP EA, 200 x ri Bstension of mémbet'2'= "nyo! SUS NOX 20s ym, "300 « 133 x 330 B. The detections of join D Figure 24 Had member AD been unrestrained by member BD at D. the compressive force would have shortened it by —5 mm vertically. On the other hand had member BD been fee, the tensile force would have elongated it by $8.66 mm. The end of AD would have moved to point ain Figure 2.4 and the end of BD would have moved to point b. In the figure Db and Da are drawn to an exaggerated scale for clarity. As the members are connected together at D, AD tends to rotate BD and b moves to d on ‘the circumference of a circle with centre at B. Similarly BD tends to rotate AD and moves to don the circumference of a circle with centre at A. In this manner the ‘ends of the members remain attached to ane another. Point D moves to d and the structure takes the deformed shape BdA. This means the compatibility of joint D is preserved and the structure remains in one piece. Because the lengths of the curves bal and ad are very small compared to the lengths of the members, these are shown a straight lines perpendicular to BD and AD respectively. The deflection diagram Dbda is reproduced in Figure 2.4. From this diagram it is clear that the vertical deflection » of joint Dis equal to Da = 3: mm downwards. ‘The horizontal deflection x of joint Dis equal to ad. Now ae = 8.66 cos 60 bg = 8.66 sin 60 fe be = bg + Da = 8.66310 60 + 5 ed = be cot 30 ee ag iF aed po Structural Properties 31 ‘ x = ed + ae = 8.66.cos 60 + be cot 30 x = &,66cos 60 # (5 + $,66 sin 60) cot 30 = 25.98 mm Alternatively, using matrix algebra, consider a horizontal load Wy and a vertical joad W, are acting md, remembering that in the\actual example W, = 0. Resolving the forees at 1D horizontally and vectically, we obtain py = —VEe Wt We py Ww, +O +. The equations P= BW become [im fava i || =V3 x W, + W Ls as Lee olliw.| > L2W, # 0% MW 2:20) va] and B= [- 4 th [-v3 The transpose of Bis BT = ' ie.the rows of Bibecome the columns of BT With A, = 173 mm* and A, = 133 mm? equations (2.18), using the results obtained in equation (2,20), give Bi] fee 10” ° tesco ae 3 3 bts 200 0 Ose 1S: WW, + 0% Wy fax (133 Since W, = Qand W, = 10kN = BL, = (1/200) x (10000/173) x =10.73 = —3 mm and By = (1/200) x (20000/133 V3) x 2» 10 = 8.66mm Bhi | ei me bb, |] ~ (8.66 Now premultiply this vector by the matrix Br Feat Layfix <5 42% 866] jae 1 gel line thee LS 259d] fe So Neale ty It is noticed that the first element of vector is 25,98 which is in fact the horizontal deflection of joint D. The second element of Xis —5 which is the vertical deflection32 Structural Properties of D. In will be proved, in Chapter 9, dhat generally the deflection X of all the joints im a structure is in fact given by X= BTS (2.21) Example 4. With A, = 173 mm™, A = 133 mm? and £= 200 kN/mm? what external load applied at D to the structure shown in Figure 2.4 would cause joint D to deflect horizontally by 259.8 mm ond vertically by —50 mm? Answer: Intuition suggests that the load is 100 kN acting horizontally Deflections, like forces, are vector, They have resultants and components. In Figure 2.5 the vectors x and y are the horizontal and the vertical deflection com« ponents of joint B. Members AD and BD remain connected to joint D as it defects, Sdehenoe cogem force egram Figure 25 The deformation of member AD is equal to the vertical deflection of joint D and therefore 8L, = y = <50mm G22) The component of x in the direction BD is equal tox cos 60, where x = 259.8 mm, while the component of y in the same direction is y sin 60, where » = —50 mm. The net deflection of joint D in the direction BD is thus equal tox cos 60 +/y sin 60, The deformation 5£; of member BD has to be equal to the deflection of Din the direction BD, so that member BD remains connected to joint D, thus BL, = x cox 60 + y sin 60 (2.23) Le. SL, = 259.8 x 0,880 x OSv3 = 866mm Equations (2.22) and (2.23) can be written in matrix form as BL, 0 |) fx Bt; | * [corso sin 60, is G24) Structural Properties 33 or simply b= DX (2.240) Matrix D contains sines and cosines of the angles of incli ons for members AD and BD and it is known as the displacement transformation matrix. [1 relates the deformations 6= (6L, 613) to the deflections X = {x » } of joi ‘Once the member deformations are known, the member forces can be calculated from the stiffness equations (2.14) or (2.15), thus PB, 173/10 000 Oo.) fase a,| 7 ° 133-/3/20000 | | 86.6. Py = 200 x (173/10000) x —30 = —173.KN and Ps = 200 x (1334/3/20000) x 36.6 = 200kN So far we do not know the magnitude or the direction of the extemal force at D. Let the horizontal and vertical component of this force be IW’, and I, as shown in Figure 2.5b. Resolving the forces at D horizontally Wy = pz.cos60 = 0 W, = p; cos 60 * 200 x 1/2 = 1OOKN Resolving the forces at D vertically Wy — BP, — pr sin 60 = 0 Wy = py + py sin60 = 173 + 200 x V9 = 0 Thus the extemal force acting at Dis W= W, = 100 KN. - Altematively, notice that the transpose of matrix D, equation (2.24), is ; Q 0s 60 Dit sin 60 and post multiplying this matrix by the vector of member forces P= tp, pp} gives 0 cos 60] [=173 | [300 cos 60 fia07) [i 2.28) 1 sin éo || 200° =| 173 + 200sin 6a |"| 0 |") We ate ee Iiwill be proved, in Chapter 12, that the results obtained in equations (2.24) and (2.25) are generally true, that is 6 = Dx and W= DIP (2.26) 2.8, Temperature stresses When the temperature of a pin ended member ts increased or decreased, the member extends ar contracts. If this extension or contraction is prevented, temperature stresses develop in the member. Consider a pin ended member of initial length £ and of linear coefficient of thermal expansion @ per degree centigrade. Let the temperature of the member be increased by /° centigrade, If the member is free to34 Structural Properties expand, its length increases by @£1 to become £ +022. If the expansion of the member is prevented, it & as if a member of length 2+ Ohi were compressed to a length £. Thus the compressive temperature strain ¢° is given by = —OLrL( + bt) * Or (2.27) Here it is assumed that 82 is small compared to £ and Or is small compared to unity. The compressive stress ° is obtained from E = ofe be. io * Ba; thus oO = -£or (228) Similarly the tensile temperature stress in amember prevented from contracting; after a fall of ” in its temperature, is £67. The linear coefficient of thermal expan- sion for steel is 4x 10-* ey degree centigrade and the product £8, with F = 200 KNimm? is 0.8 x 10-? KN/mm*. As an example if the temperature of a vended member is changed by 50°C while the member is restrained, the strain €” of the member will be 4x 10°" x 50=0.2 x 10-7 and its stress 0” will be 0.8 x 10-* x 50= 40% L0-? kN/mm®. If the area of the member is 1000 mm’, then the force developed in it isp =a" = 40 kN. ‘Although the linear coefficient of thermal expansion is a property of the structural material, isostatic structures are at an advantage because thermal stresses do not develop in them. These stresses only become important in hyperstatic simactures. 2.9, Example Example 1. The pin ended member shown in Figure 26 is 10 m long and has an area of 173 mm. The member is 2.89 mm too short. What force should be applied at B before it can be pinned into its support? Once it is pinned, what is the safe tempera- ture 10 which it can be heated before the stress in the member reaches 0.1 KN/mm?? Take £= 200 kN/mm* and #=4 x 10-*mm/*C. Answer: p= BASIL = 200 x 173 x 2.89/10000 = 10kN tom Figure 26 Structural Properties 3s The tensile stressin the member = a, = plA = 10/173 = 0.0878 KNimm? Ifthe menther is now heated by # a= Far Le 0.0878 = 200 x 4x 10-%, Thus fy = 0.0578/(200 x 4% 10°") = 72.25°C then all the tensile stress in the member will be relieved. The member can then be heated further by 13 ty = 0 JEO = 0.1 x 10°/(200 x 4) = 125°C before the compressive stress of 0,1 kN/mmm is developed. The total safe temperature is therefore 15 72.25 + 125 = 197.25°C 2.10. The principle of superposition In examples 3 and 4 of section 2.7, the structure shown in Figure 2.4 was subjected we aa different horizontal loads of 1OKN and 100 KN. Under the first load, the resulting deflections were 25.98 mum horizontally and —3 mm vertically. Under the second load these defections increased by ten times to 239.8 mm and —50 mm. If the structure ts analysed with an external horizontal load of W = 50 KN, the resulting Py twa) 2, (am) Figure 2.76 Structural Properties horizontal and vertical deflection will be 129.9 mm and —25 mm respectively. The resulis of these three analyses are plotted in Figure 2. 7 and it is noticed that the graph for cach deflection against the applied loads is a straight line. This indicates that the deflection of a structure is linearly related to the external applied loads. This is known as the principle of superposition. If the applied load is doubled, the defections are doubled. The load of 100 kN applied to the structure can be treated as two loads. The first is say 10 KN which gives the first set of deflections of 25.98 mm and —5 mm. The second is 90 KN which produces deflections of 233.82 mm and —45 mm, and the two loads acting together produced deflections ‘of 259.8 mm and -50 mm, ‘Similarly the deflections at any point such us A, B, C of the structure shown in Figure 2.8 are obtained as the sum of the deflections ul that point due to the sepa- rate effects of the loads IV, and W. Figure 2.8 ¥ * Me ‘There are two exceptions to the principle of superposition. It dees not apply to ‘Siructures manufactured from material whose stress-strain relationship is not near elastic or if the limit of proportionality of the stress-strain curve is exceeded. In example 4 of section 2.7, the structure js subject to a horizontal load of 100 KN at D. This produces a tensile force of 200 KN in member BD. With an area of 133 mm? the stress im this member is 1.50 KN/mm? which is five times the stress at the limit ‘of proportionality for steel. Thus the resulting stresses are by far beyond the linear pe of the stress-strain curve for this material. In fact when the acting horizontal load W' exceeds 20 KN, the material of the structure begins to yield and for higher loads pene develops in the material. For such high loads the stress-strain curve becomes non-linear and the principle of superposition becomes inapplicable, ‘The second exception is when the geometry of the structure is basically altered by the effect of the external loads. Here again the principle of superposition does Figure 2.3, a Pee werticar » enbens Structural Properties 7 not apply, Consider, for instance, a pin jointed structure with twa vertical members: as shown in Figure 2.94, As soon as the force W ix applied to B, the members extend and the pin at B moves to the right to B by an amount x, The steucture takes the new shape shown in Figure 2.95, in which x= OL Resolving the forces at B horizontally 2:29) W— 2psina= 0 ies p= O5W/sin 8 Since 0 is small, sin @ = 8 and P= OSW/O (2,30) The strain in each member is given by © * change in lengthy original length wiE? + € (2.31) ‘On the other hand, the material is assumed to obey Hooke's law which gives the strain as @ = off = pica (2.32) where A is the cross sectional area and £ is the modulus of elasticity, Equating this strain to that given by equation (2.41), we obtain, using equations (2.29) and (21.30), O.5x7/17 = piAE = W/(2PAE) = WL/(2xAE) Hence x = L(WAE? 2.32) ‘Thus the deflection x is not lineasly related to the load W. This is in spite of the fact that the material obeys Hooke's law and the deflection x is small. A graph of x against has the form shown in Figure 2.10. With members subject to axial forces and bending moments (see next chapter) ‘the principle of superposition becomes also inapplicable. either when the axial forces are too large or when the deflections are too lange or both. Figure 21038 Strectural Properties Structural Properties 39 In spite of these exceptions, the principle of superposition remains important Figure 2, / Jc shows the force-deformation graph of one of the members of the because engineers often design their structures so that the limit St peanoct nally is structure, The strain energy stored in the member is, by a similar argument to the not exceeded, the deflections are small and shapes are not grossly changed by the Shove, given by applied loads. ae 2.11, Strain energy es -[. er oo Under the action of gradually increasing external loads, the joints of a structure icare 0 SRE TRIAL deaoanmtitn (exes nk or Commaction] of the member The deflect and the members detorm. The applied loads produce work at the joints to Tot strain eneray stored in all the members of the sructire isi, whers which they are applied and this work is stored in the structure in the form of energy which is known as strain cmergy. If the material of the structure is elastic, whether x No pa" linear or non-linear, then gradual unloading of the structure relieves all the stresses ou =Suey [ pda (2.35) and the strain energy is recovered, 7 T da —4h. and Nis the total number of members. As the member force @ occurs under the Be ¥ action of the external load HW and the member deformation A corresponds to joint x. then by the principle of the conservation energy, U; = w and thus he wath S [Ppa EN mim 6) ong Oa io aca Consider now a single member of the structure, [f the limit of proportionality of ens the stress-strain diagram is not exceeded for this member then the force deformation diagram will be linear and the area under the graph (Figure 2.'e) is that of a tangle, a“ thus ate & ‘* i us | pds = 0.5pa' 2.97) Ja where A’ is the total deformation of the member and p is the total force in it, But for linear elastic material which obeys Hooke’s law, we have al» phjAk It follows from equation (2.37) that = pLiQAk) (2.38) = Now p/ is the stress @ in the member, thus nm |S U=O507ALE kN mm (2.39) Furthermore, since AL is the volume of the member, it follows that the strain © work doar by Werk gone by forne energy per unit volume of the member u, is given by Bilis w= O507/ — kN/mm? (40) Consiler an elastic structure subject to # single gradually increasing extemal load. For instance, for a steel member with £ = 200 KN/mm. and with the tensile stress at W as shown in Figure 2.17. The graph of HW against the deflection x at D is of the the limit of proportionality @, = 0.3 kN/mm, itis found from equation (2.40) that shape shown in Figure 2.1 1b. If at a given instant the value of the load is W, and he amount of strain energy per mm? which can be stored nt the elastic range incteases by an amount 8W/,, then the work done is W,8x as shown hy the shaded isu = 0.225 Nim’. area under the graph. The total work done by the load, during its gradual increase from zero to its final value W" is the sum of all such clemental areas. Hence the total work w is given by 2.12. The work equation es " -[ Wie = kN mm (2.33) Consider a Bis Jolnied linear elastic structure with V members and subject to a r. a , tots] of M external loads W= {W, Wy... Wy... Wy, Let the deform stion in the wheve'x’ ts the total deflection a D when W= W's members be & = (8) 2 5;...5y} and let the deflection under the external40 Srrecterral Properties loads be X= fary ary). -.xy .- .2¢qy ]. The work done by the extemal loads is thus w= (Wha, + Whey t+ 00. + Woy to. + Wy 2 and the work done by the member forces is U = DB, + pib, +... + pb) +... + pydyii2 Here E/is also the total strain energy stored in the members and since L = wit follows that Wy, + Weep +... + Way te. + Wye = pid, + pbs +... + pd ply (2.41) This is known as the work equation. In matrix form this equation is written as (Hy, Wy We Ma) (noe ee ey = Wri Pa ees os Pwd {81 Ba Be ‘or simply WX = PTE (2.42h) where WT = [¥#, We Wj... Way] and PT = [py ps... pj-s.Pyr) are OW vectors. They are the transpose of the column vectors W and P respectively. The vectors X and 6 are column vectors, as usual. Equations (2.41) or (2.42) are extremely useful for calculating the joint deflections in a structure. (2.420) 2.13. Relationships among stress, strain and stiffness Consider a gradvally increasing load W’ applied to a pin ended linear elastic member as shown in Figure 2.12, Let the value of W increase to i" when the total deflection is A= 1 unit, ‘The total work done by the applied load is w -f O.SW5A = OSW'Al ° but a’ = Dunit : w= OSW" Structural Properties al When W= HW", the force in the member isp’ and from equililyrium pW =o . pew Thus the total work done by the external load is w= 0.5p" ‘The total extension of the member is 6" which is equal to A’ = p= (FAILS = ks! = pekBbiekx 1l]k The total work done & thus w= O.5p' = 0.56 (2.43) Now the strain energy per unit volume, «, is from equation (2.40) equal to 0.50 /E and the total strain energy is u- Woe a where v is the total volume of the member. Furthermore, from Hooke’s law o = ef, thus us {"(oizmretae = 05 f edu (244) Je 5 ‘Equating the total strain energy in the member to the total work done by the extemal load, we obtain using equations (2.43) and (2.44) (2.45) integral of the product of stress and strain over the volume for a unit deformation. This important formula is extensively used by engineers in the powerful ‘finite element’ method to derive the stiffness matrices of members or other elements, such as plates or solids. In matrix form equation (2.45) is written as k= ih atedy (2.46) where a! is the transpose of the stress vector fa} . 2.14. Principles of structural analysis MWith the aid of single members and pin jointed structures, the last two chapters have summarised the entire theory of structures using both classical and matrix ‘methods, It should have been noticed that there are three basic principles involved. ese are: |. Make use of the stress-strain relationships of the material of the structure. 2. Make use of the equilibrium of the forces applied to the structure, those developed in the members and those developed at the supports. 3. Make use of the compatibility conditions in the structure, i.e. ensure that after applying the loads, the members remain connected to their supports and to the joints to which they were connected before loading.a2 Structural Properties ‘Compatibility conditions are obtained by equating the deformation of the members to the deflections of the joints to which they are connected. It was stated that deflections and deformations aré alse vectors. Thus by equating these defor mations and deflections we are in fact deriving equations of equilibrium for deflece tion vectors. Thus compatibility equations are merely atalements of equilibrium of deflection. Looked at in this way, there are only two factors involved in the analysis of structures, These are stresestrain relationships and equilibrium of forces and deflections. Using these two factors every structure, linear or non-linear, isostatic or hyperstatic cum be analysed. Nothing more is required. 2.15, Examples Example 1, me the iat igs frame shown in Figure 2.13, the area of each member is 100 mm? KN/mm?, Calculate the strain enerey per unit volume and the paraears each member, Using the work equation, what is the vertical deflection ¥ at point D? Fieure 213 Answer: Resolving the forces at D vertically W = Ipiv2=0 * p = O.SW/2 = 10//2kN The stress = o = p/A = 10/(100/2)= 0.071 KNimm? The strain energy per unit volume =a = 0.507) w= 05 x 0.071 x 0.071 x 1/200 = 12.5 x 10-* kNimm? Stiffness = RAjE = 220% 100 20 erijram 1000 The strain energy in each member = U = uu * Us 125 x 10-8 x 100% 1000 = 1.25 KN mm Total strain energy in both members = 2% 1.25 = 2S kN mm. The work done by the external load = 0.5 Wy O5hy = 2: y= 2« 25/W = 5/10 = 0.5 mm Altematively the extension in cach member is 6, given by pm ks * 1/2 = 208 and 8 = 1(2V2) Structural Propersies 43 From the work equation (2.41) or (242) pd) + pid, = Wr 2p5 = Wy 2% (LO/V2) ™ IM2V2) = Ly giving ¥ = 05mm Example 2. Calculate the vertical deflection of joint D for the structure shown in Figure 2/4. All other data.are given in the last example. What is the deflection at 0 when W = 20:kN? Answer: This structure is hyperstatic because there is not a sufficient number of equilibrium equations to find the forces in each member. Resolving forces at D horizontally P= Pa (247) Resolving forces at D vertically, using (2.47), We py - miiv2=0 y= WWD DB ee ‘There are two unknowns p, and py in this equation. To solve the problem a further equation is required. From the triangle DED! (Figure 2.14) it is noticed that the com- patibility condition is satisfied by & = yeosds = p/y/2 (2.49) Now for member AD, p, = EAS/L, thus 6 = p,LiEA Similarly for member ED, py = GAy/(E) 2) - Y= pali(w2 * EA) (2.50) ‘Using these values for & andy, equation (2.49) becomes Po = Uy substituting for ja in equation (2.48) 2p = W—V2x p, By = WH + 4/2) = 10/3.141 and Py = 20/3141aS Structural Properties Hence equation (2.50) gives y= 20L/(G.1415/2 x FA) Y= 20x 1000(3.141.V2 x 200 x 100) = 0.225 mm. From the principle of superposition when W’= 20 KN then y = 0.45 mm. Check that the stresses are below the limit of proportionality. ample 3, Calculate the forces in the members of the pin jointed frame, shown in Figure 2.15 (a) when a horizontal load W, = 5 KN is acting at E, (b) when a hori- zontal load W; =.10 kN is acting at D and (c) when both forces are acting together, wpe 10 hae — ae Fguere 2.15 a . Answer: (a) With W, = 5 KN acting alone, resolving forces horizontally at E W, — pen/v2 = 0 is ep = V2 x Wy = SV2kN Resolving forces vertically atE = pp = ~Pep/ V2 = —5 kN At Dresolving horizontally pox V2 * Poni v2 = 0 “ Ppp © —Pog = =5 V2 kN Resolving vertically at 2 pp, — Pop! V2 + ape! v2 * Poa = ~Powl V2 + Poe/ v2 = LOKN (b) With W; = 10 KN acting alone, resolving the forces at E and D Pen = Pen = 9, Pog © —10V2andppa = LOKN By the principle of superposition, the member forces, due to both Wy and Hy acting together are (Peo Pew Pon Pow} = (3-2 S$ 20 —ISV/2}kN 2.16. Properties of symmetrical structures. In asymmetrical structure which ix loaded symmetrically, such as that shown in Figure 2, 16a, the forces p, and p, in two members f and f, which are arranged symmetrically on either side of the axis of symmetry, are equal. Thus to analyse Structural Properties 45 such a structure it is sufficient to find the member forces in one half of the struc- ture only. For each member, such as &, in that half there is a corresponding member iin the other half whose force p,, is equal to the known value of py. On the other hand, when a symmetrical structure is loaded antisymmetrically (Figure 2.165) the forces p, and p; in the members é andj, which are arrenged symmetrically on either side of the axis of symmetry, are equal in magnitude and opposite in sign. Thus, again it is sufficient to analyse only one half of the structure. *) * ‘ f NSM 4 1A symmetnca) frome leaded ymmetrically ‘, Wye BA Apmeetrica! frome leofed estiarmee incally Petre 2.16 When a symmetrical structure is loaded in a general unsymmetrical manner, it is Possible to separate the actual loads into two systems, a symmetrical system and an Anthymmetrical one. Half of the structure is then analysed once for each load on The forces in the members due to the actual loads are then obtained from Cofulis of the two analyses using the prineiple of superposition, ‘onsidder the jointed frame shown in Figure 2 17 which is symmetrical about He vertical line FG and subject to forees W, at D, Wy at E and Wy at C. The “ertical reactions R, and Ry are found by takir arescans Ry and Ay ry taking moments about B and A W, x ISL + Wy x OSL +W, x L —R, x 2b = 0 Ry = 0.25(3W, + W, + 2W5) (2.52) Rp * 2 —W, x OSL -— Wy x 1.5L — Wy * LO Ry = 0.25(W, + 3W, + 2s) (2.53)46 Structural Propertics Figuee 217 Notice that the coefficients of W', and W/; in equations (2.52) and (2.53) have interchanged, lhe member Forces Pa Pass and > are now found by the method of joints, thus At A, resolving vertically Ry — p: c0s30 = 0 Using equation (2.52) P2 = OSH, + Wy + IW ALA, resolving horizentally Pi + Py cos 60 = 0 ‘ mi = —OSpy = = At D, resolving vertically Pz cos 30 + py cos 30 — Wy = 0 4 Py = O.5(W, — Wy — ZW Ni v3 At D, resolving horizontally GW, + W, + Wy V3 Py + Py con 60 —p, cos 60 = 0 i Py = O5(K, + Wy + DW fd ‘The method of joints may be continued to calculate pa.pe and py. Howeve: evident that ps. pe and p, are associated to lv’, and W as py, py and py are astociated to WW; and M,, thus interchanging the coefficients of W, and H’ Ps = O.5(-W, + ty — Wy) 3 Ps = O.5(M, + 3 + JW) 3 Ps = -O.25(H, + 3, + Thy fs Structural Properties 47 ‘These member forces are written in matrix form P= BW as Py LS O5 —1 || Hy Ps 3 1 2 || Ws Ps 1 -I 2 [Ws | ps | = 05003] -1 1 =a | ps 1 a 2 (2.54) Pe 05 =1S -1 ie 1 1 2 Consider now the following three loading cases. 1. The unsymmetrical loading case where W, = —W, and Wy #0 That is, W, and W, are acting downwards while b*, = —W, acting upwards as shown in Figere 2 18a, Equations (2,54) give py = OS(-LSW, +058, — WyV3 = 0.50, + Wy V3 Similarly Pa = (Wy + WV, py = Wy — WiV3, og = —OW, + WV ps = (Wy + WS V3. pe = O50", — WS undp, = Wa V3 (2.55). 2. The antisymmetrical loading case where W, = -W, and Wy = 0 ey Annnmeetse ipsa Symmes saeaieg Reps mo er Figure 2.0848 Structural Properties That is W, is acting downwards while W, = —W, is acting upwards and WW, is not acting a8 shown in Figure 2. 1b, Equations ) give By = —O.5Wy/ V3 Pe = OSH) V3 Py = Wil V3 py = Wy I /3 2.56) Ps = WilW3 Pa = WI 3 Go P= 0 It is noticed that when this symmetrical frame is subjected to an antisymmetrical loading system then pg = —P 1, Ps = Py and py = —py. This indicates that once the force in a member to the left of the axis FG of symmetry is determined, the force in a corresponding member to the right of this axis can be found by chunging the sign of the force im the former member. Notice that members 4, 5 and 6 are the mirror images of the members 3, 2 and | respectively. Member 7 crosses the axis. symmetry and it is symmetrical about this axis. The force in this member is zero, 4. The symmetrical loading case where W, = W, = OandW, #0 That is only Wy is acting as shown in Figure 2. 18c. Equations (2.54) give Py = 0.58 / 3 Pe = O50, / 9 Pa © Wal v3 Ps = Ws V3 onl Ps ® —Wy/ 5/3 Pa = Wy) /3 j Py = Wil V3 The load Wy is symmetrically placed on the frame and it is noticed that p, = p;, Ps = Ps and py = py. This indicates that the values of the member forces are symmetrically distributed about the centre line of the frame and once these are calculated for ane half of the frame the forces in the second half become known. ‘The forces in members crossing the axis of symmetry, p> in this case, have also to ‘be calculated. Using the principle of superposition and adding the loading cases 2 and 3 (Figure 2.18b and c) the general unsymmetzical loading case [, shown in Figure 2. 18a, is obtained, Similarly, adding the values of the member forces obtained for cases 2 and 3, given by equations (2.56) and (2.57), results in the forces given by equations (2.55) which is for the general case 1, It is therefore concluded that the: analysis of a symmetrical frame subject to a general state of loading is obtained by adding the results of two analyses, one with the symmetrical part of the loads and. the second with the antisymmetrical. 7 2.17. The equivalent half frame Cates 2 and 3 show that, for the symmetrical and the antisymmetrical load syst only half the frame needs to be analysed. Thus the original symmetrical frame can replaced by two equivalent half frames. Under the symmetrical loads, points on axis of symmetry move only along this axis, Thusin the equivalent frame these points are supported by rollers free to move along this axis only. Any extemal low ‘that may be acting along the axis of symmetry is shared by the two halves of the frame, thus in the equivalent half frame such loads are halved. For the symmetrical loading case shown in Figure 2.18c, the equivalent half frame is shown in Figure 2.194 where it is noticed that joint C and it H are supported by rollers free to move vertically. A load W’,/2 is acting at C. Structural Properties 49 Figure 219 : Lu i Equceaion! Bait toome (Gevinymmerricat tooding) 4. Equirolret molt tome . (remtiest cesemg) Under the antisymmetsical loads, points on the axis of symmetry move perpend- icular to this axis. Thus in the equivalent frame these points are supported by rollers free to move perpendicular to the axis of symmetry. The force in a member which intersects the axis of aymimeitry perpendicularly is zero, thus in the equivalent frame such members are removed. For the antisymmetrical loading case shown in Figure 2.184, the equivalent half frame is shown in Figure 2 196, where joint C is supported by a roller free to move horizontally. Member 7 is removed. b Sguiens ait Gearon ees